Kannada_MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "5884ef47",
   "metadata": {},
   "source": [
    "# Classification of Kannada MNIST\n",
    "<!-- ##  ![Preamble of kannada digits](img/kannada_numbers.jpg = 100x200) -->\n",
    "<div>\n",
    "<img src=\"attachment:Screenshot.png\" width=\"500\"/>\n",
    "</div>\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "9f09057d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Author: Vikas Chidananda \n",
    "\n",
    "# Dataset credits: Vinay Uday Prabhu\n",
    "# Prabhu, V. (2019, August 03). Kannada-MNIST: A new handwritten DIGITS dataset \n",
    "#     for the Kannada language. Retrieved April 19, 2021, \n",
    "#     from https://arxiv.org/abs/1908.01242v1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5e946c2",
   "metadata": {},
   "source": [
    "### Numbers in Kannada"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cb3c3ed4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 2,
     "metadata": {
      "image/jpeg": {
       "width": 400
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image('img/kannada_numbers.jpg', width=400)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "105ef983",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "import pathlib\n",
    "import matplotlib.pyplot as plt \n",
    "import matplotlib as mpl\n",
    "\n",
    "from sklearn import svm\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis as LDA\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
    "\n",
    "font = {'size' : 10}\n",
    "mpl.rc('font', **font)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "908e3150",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "51c99430",
   "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>label</th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   label  pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  \\\n",
       "0      0       0       0       0       0       0       0       0       0   \n",
       "1      1       0       0       0       0       0       0       0       0   \n",
       "2      2       0       0       0       0       0       0       0       0   \n",
       "3      3       0       0       0       0       0       0       0       0   \n",
       "4      4       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   pixel8  ...  pixel774  pixel775  pixel776  pixel777  pixel778  pixel779  \\\n",
       "0       0  ...         0         0         0         0         0         0   \n",
       "1       0  ...         0         0         0         0         0         0   \n",
       "2       0  ...         0         0         0         0         0         0   \n",
       "3       0  ...         0         0         0         0         0         0   \n",
       "4       0  ...         0         0         0         0         0         0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0  \n",
       "1         0         0         0         0  \n",
       "2         0         0         0         0  \n",
       "3         0         0         0         0  \n",
       "4         0         0         0         0  \n",
       "\n",
       "[5 rows x 785 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path_train_data = 'Kannada-MNIST_data/train.csv'\n",
    "df_train = pd.read_csv(path_train_data)\n",
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d8daf5f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert data to numpy arrays and seggregate train and test sets\n",
    "labels, data = df_train['label'].to_numpy(), df_train.iloc[:, 1:].to_numpy()\n",
    "train_data, test_data, train_labels, test_labels = train_test_split(data, labels, \n",
    "                                                                    test_size = 0.05, \n",
    "                                                                    random_state = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "197b61e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10 784\n",
      "57000 3000\n"
     ]
    }
   ],
   "source": [
    "# Useful statistics \n",
    "num_labels, dim_data = len(np.unique(train_labels)), train_data.shape[-1], \n",
    "N_train_data, N_test_data = train_data.shape[0], test_data.shape[0]\n",
    "\n",
    "# Normalize the data\n",
    "train_data = train_data/255.0\n",
    "test_data = test_data/255.0\n",
    "\n",
    "print(num_labels, dim_data)\n",
    "print(N_train_data, N_test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "099b0e9b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "28\n"
     ]
    }
   ],
   "source": [
    "#size of the image = m x m\n",
    "m = int(np.sqrt(dim_data))\n",
    "print(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "69b1c94d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# store data as images\n",
    "train_images = train_data.reshape(N_train_data, m, m, 1)\n",
    "test_images = test_data.reshape(N_test_data, m, m, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "7f6d0053",
   "metadata": {},
   "outputs": [],
   "source": [
    "#helper function\n",
    "def print_misclssified(predictions, true_labels):\n",
    "    idxs = np.where(predictions != true_labels)\n",
    "    idxs = idxs[0]\n",
    "    num = len(idxs)\n",
    "#     print(num)\n",
    "    n = int(np.sqrt(num)) + 1\n",
    "#     print(n)\n",
    "    plt.figure(figsize=(20,20))\n",
    "    \n",
    "    for i in range(num):\n",
    "        plt.subplot(n, n, i+1)\n",
    "\n",
    "        image_ind = idxs[i]\n",
    "        digit = test_data[image_ind]\n",
    "        plt.imshow(np.resize(digit, (28, 28)), cmap = plt.cm.binary)\n",
    "        plt.xlabel('true:'+str(true_labels[image_ind])+' pred:'+str(predictions[image_ind] ))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29542897",
   "metadata": {},
   "source": [
    "## Visualize some of the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ed59f51b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2cAAARPCAYAAACbE5mwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOzdeZwV1Zn/8e8jEVABZVNxbXHHNQkoLjEYYyAYA5o4g4kOECOJETOOBgMug5hEiVuYuOOo7Y6JqOEXE8FhUJMoKhp3RFEwEpFFNHELiJ7fH33N9Km63fd21b23Tt37eb9evLqf6lqetr9Wc6g6VeacEwAAAAAgWxtk3QAAAAAAgMEZAAAAAASBwRkAAAAABIDBGQAAAAAEgMEZAAAAAASAwRkAAAAABCDV4MzMhpnZIjNbbGYTK9UUGgP5QVJkB2mQHyRFdpAG+UE5LOl7zsysk6SXJB0uaZmkxyUd65x7oa1t+vTp45qamhIdD+F74oknVjvn+pazbkfzQ3bqWzWzI5GferZ06VKtXr3ayl2fcw9a49yDpDj3II32zj2fSbHf/SQtds69KklmNkPSCEltnqSampq0YMGCFIdEyMzstQ6s3qH8kJ36Vs3sSOSnng0cOLCjm3DuwT9x7kFSnHuQRnvnnjS3NW4t6fVW9bLCsujBx5nZAjNbsGrVqhSHQ50pmR+ygzZw7kEanHuQFOcepMG5B2VJMzgrdik3do+kc266c26gc25g375l3TmAxlAyP2QHbeDcgzQ49yApzj1Ig3MPypJmcLZM0rat6m0kvZGuHTQQ8oOkyA7SID9IiuwgDfKDsqQZnD0uaWcz28HMOksaJWlWZdpCAyA/SIrsIA3yg6TIDtIgPyhL4geCOOfWm9l4SbMldZJ0vXPu+Yp1hrpGfpAU2UEa5AdJkR2kQX5QrjRPa5Rz7neSflehXtBgyA+SIjtIg/wgKbKDNMgPypHqJdQAAAAAgMpgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAUj1tMZ65pz30nZ99NFHXj137tzYNsOHD/fqk046yaunTZsW26ZTp07t1gAAAAAaA1fOAAAAACAADM4AAAAAIAAMzgAAAAAgAMw5K4jOMXv22We9+vOf/7xX9+jRI7aPXr16efUdd9zh1c3NzbFtLrzwQq8eP358yV4BAAAA1B+unAEAAABAABicAQAAAEAAGJwBAAAAQACYc1awaNEir47OMRs6dKhX//a3v+3wMaZMmRJbtmbNGq9+//33vXqTTTbp8HEQFjNLvY/onEgAACpp3bp1Xt29e/d2vy7V5ndTsd+h/E7Ml+OOOy627NZbb213m/333z+2bP78+al7if5d/Nxzz+3wPqqdP66cAQAAAEAAGJwBAAAAQAAYnAEAAABAABicAQAAAEAAeCBIwYABA7z6y1/+slcneQBI1Oabbx5bdvbZZ3v1Lrvs4tWjRo1KfVyEpZyJpNEJ0EyIRns++eQTr77yyiu9+pRTToltc9lll3n1D37wA6/eYAP+7S50K1as8Oott9yy3fWPOOKI2LJK/G5L4r333vPqbt26ZdIH/k+XLl28Ovo7phIPtwLK9fzzz8eWHXDAAan3+9e//tWrow/8mzx5cupjpMVvXwAAAAAIAIMzAAAAAAgAgzMAAAAACABzzgq6du3q1XPmzKn4MU466aTYsg8//NCr+/TpU/Hjorai9+UnmRvGvf5oy/r162PLbrvtNq+eMGGCV/fq1Su2TXSdrbfe2quPOuqopC2iRvr16+fV0TnL//jHP7w6OjexVpYsWRJbFp3nHf1dCCC/fvKTn3j13nvvHVsn+myHrOy4445eXYl5bWlx5QwAAAAAAsDgDAAAAAACwOAMAAAAAALAnLOMnXbaaV79zDPPePVbb73l1b179656TwDC9dhjj8WWff/73/fq008/3at/+tOfxrYpNgcW+bLxxht79dNPP+3Vzc3NXp3VO82+9KUvxZYNGjQog07QEc8++2wQx91rr70y6QPJ3XvvvV49fvz42DrHHXdcrdrJHa6cAQAAAEAAGJwBAAAAQAAYnAEAAABAABicAQAAAEAAeCBIQfTlnDNnzvTqzTbbzKsPO+ywqvRxwQUXePV3vvMdrz788MOrclyEpRIvskZ9WLdunVe/+uqrsXWGDRvm1cUeABIVfen0O++849UffPCBV0cfPoHsHX300V49ffp0r54yZYpXT5w4seo9leu+++7LugWUMG7cOK8ePHhwTY4bfWExv//y59BDD/Xq7bffPqNO8okrZwAAAAAQAAZnAAAAABCAkoMzM7vezFaa2XOtlvUys/vN7OXCx57VbRN5RX6QFNlBGuQHSZEdpEF+kFY5c86aJV0u6aZWyyZKmuucm2pmEwv1jyvfXu0cc8wxXj169Givjr7UtVpzzqKic49yqFkNkJ/WovfHlzN/rA5+ztXQrAbLTjGvv/66V//gBz+IrXPyySd3eL9nn322V0fnL+26665efeCBB3b4GBlrVp3n56qrrvLqbt26eXX0Rc8TJkyoek+SdMstt3j13/72t9g6gc8jaladZ6ccjzzySNYt5FWzGiw/0ReHn3POOV7NnOWOKXnlzDn3kKQ1kcUjJN1Y+PxGSSMr3BfqBPlBUmQHaZAfJEV2kAb5QVpJ55xt4ZxbLkmFj5u3taKZjTOzBWa2YNWqVQkPhzpTVn7IDorg3IM0OPcgKc49SINzD8pW9QeCOOemO+cGOucG9u3bt9qHQx0hO0iD/CApsoM0yA+SIjuQkr/nbIWZ9XPOLTezfpJWVrKpLFx33XVevcMOO3h19H0xSKXu8pNW4PMvQtLw2dlzzz1jy6LvR0Sb6io/nTt39uqzzjrLq+fMmVPLdv7pl7/8pVcXmycZ7T0H6io7xUTnnWYllD4qrK7zM3DgQK+eN2+eVxd7R94GG/DA+LYk/S8zS9KnT8wYLek3lWkHDYL8ICmygzTID5IiO0iD/KBs5TxK/3ZJj0ja1cyWmdkJkqZKOtzMXpZ0eKEGYsgPkiI7SIP8ICmygzTID9IqeVujc+7YNr5Um2fJI9fID5IiO0iD/CApsoM0yA/SSjrnrO5E738/77zzMurEx1yk+lPsnWaV+DmX8z41hG/9+vVe/cc//tGrhwwZUsNuELINN9zQq7/73e96da3mnF100UVe/corr3h1tC8p3juyN3PmzEyO29zcHEQfSG7RokVeHZ1nOm7cuNg2I0fyNoG2MBsPAAAAAALA4AwAAAAAAsDgDAAAAAACwOAMAAAAAALAA0ECM2rUKK/eaaedMuoElRJ9MEexB4IkeZgHDwCpT2vXrvXq6MOJog9bAD615ZZbevWpp57q1ZMmTYptU4kXmD/44INe/bOf/azdvhCmyy+/3KvHjx9fk+NEHyYxZsyYqhwX1dPU1OTVa9as8er33nuvht3kH1fOAAAAACAADM4AAAAAIAAMzgAAAAAgAMw5C8yIESOybgFVVmxuWLF5aKW+zhwzAK117drVq48++uiS20TnoZUzBy26zbe+9a12jxvtC2E65ZRTvLpac86ix+F3Wf27+eabY8seffTRdreJzmOTpNNPP71SLQWNK2cAAAAAEAAGZwAAAAAQAAZnAAAAABAA5pwF5uqrr/bqgw8+2Kv33HPPWraDGinnXWhoDKHMv4jOEdh9991j6/Ts2bNW7SCB6FyvzTbbLLbOm2++2e4+zjnnnNiySy65xKvvueeedo+LfDrzzDOrst/zzz+/KvtFuObMmVPWstb233//2DLmnAEAAAAAaobBGQAAAAAEgMEZAAAAAASAwRkAAAAABIAHggTmwQcf9Oodd9zRq3kgCFDfpkyZknULqFPDhw+PLVu5cqVXn3rqqSX3M3369JL7Rf5cdtllFd9n9IXTUjgPPULt7LXXXrFl/fr1a3eb3XbbrVrtBI8rZwAAAAAQAAZnAAAAABAABmcAAAAAEADmnCX05JNPxpZNmjSpw/s5+eST2/06LyMGGkt03kd0fk+trF692qs//vjjTPpAdW266aZefdttt5Xc5otf/GK12kGGxo8fn3ofY8eO9eobbrgh0TrIl6VLl3p1r169vHrixImxbUaOHFnNlnKNK2cAAAAAEAAGZwAAAAAQAAZnAAAAABAA5pyV6aWXXvLq4447LrbOwoULU+/3vffe8+oTTjihw/tE/Sn2XpjofETeHVMfNtxwQ6/+t3/7t5ocN/o+onHjxnn13/72t9g2ffr0qWpPqL6jjjrKqxcvXuzVr776amybAw44oKo9Ib+am5u9mjlnjWHXXXf16nnz5nn14MGDa9lO7nHlDAAAAAACwOAMAAAAAALA4AwAAAAAAsCcsza8+eabXn3EEUd4dbG5Fs8880yHj3PBBRd49e233+7VJ554old37949to85c+Z49ZZbbtnhPgA0tkMPPdSr+/btW3Kb6NzbW265paI9IXtXX311bFl0/sjo0aNr1Q4CM3ToUK+ePXt2Rp0gS+vWrfPqTp06efUGG3AtqCP4rwUAAAAAAWBwBgAAAAABYHAGAAAAAAEoOTgzs23NbJ6ZLTSz583s3wvLe5nZ/Wb2cuFjz+q3izwhO0iD/CApsoM0yA+SIjuohHIeCLJe0unOuSfNrLukJ8zsfkljJM11zk01s4mSJkr6cfVara2PPvrIq7t06eLV99xzT2ybLbbYot19Tps2Lbbs3nvv9epf/vKXXn3kkUe2u09JGjFihFevXLnSq5csWVJyH1XSkNmplehLp+vwpdTkJyDRF8dK0hlnnJFBJ2UhO2V67rnnvPrDDz/06tNOOy22zeuvv+7V0d8xO+ywQ4W6ywz5KVP0gWQ8EITsIL2SV86cc8udc08WPn9X0kJJW0saIenGwmo3ShpZrSaRT2QHaZAfJEV2kAb5QVJkB5XQoTlnZtYk6bOSHpW0hXNuudQSRkmbt7HNODNbYGYLVq1ala5b5BbZQRrkB0mRHaRBfpAU2UFSZQ/OzKybpJmSTnXO/b3c7Zxz051zA51zA8t5bw7qD9lBGuQHSZEdpEF+kBTZQRplvYTazDZUS8hudc7dVVi8wsz6OeeWm1k/SSvb3kP+LVq0yKsHDBjQ4X1E7+WXpMmTJ3v1uHHjvDo6162Y++67z6s/+eSTDvdWLWQHaZCfcPzwhz+MLfvqV7+aQSflITvJDBkyxKsPOeSQ2DrR+YcLFy706jqYc0Z+ith7771jy5555pkMOgkb2ZGuuuoqry72d9l99923Vu3kTjlPazRJ10la6Jy7tNWXZkkaXfh8tKTfVL495BnZQRrkB0mRHaRBfpAU2UEllHPl7CBJx0t61syeKiw7U9JUSb8ysxMk/UXSMdVpETlGdpAG+UFSZAdpkB8kRXaQWsnBmXPuj5KsjS8fVtl2UE/IDtIgP0iK7CAN8oOkyA4qoaw5Z41om2228eo//elPXr3//vt3eJ9nn312bNmPfvQjr+7UqVOH99uzJ+8yBFBZt9xyi1dHz4lSsvMVwrLnnntm3QJy4tlnn40t22uvvdrdph7mH6K0MWPGeHX0Hb5du3aNbTN48OAOHyc6d+3oo49u9+t51aFH6QMAAAAAqoPBGQAAAAAEgMEZAAAAAASAwRkAAAAABIAHgrSh5VUV/2e//fbzaudcLdtBg4nmK5pHoNr69++fdQsAcm7p0qWxZatXr659I6iqG264watPPPFEr77rrrsUdc0117S7z27dusWWHXTQQV59xBFHeDUPBAEAAAAAVAyDMwAAAAAIAIMzAAAAAAgAc84AAIAk6brrrvPqCy+8MLbOm2++6dXHHHNMVXtCuErN8Vm7dm1sWefOnavVDgJx7bXXevWHH34YW+fWW29tdx977LFHbNl9992XrrGc4MoZAAAAAASAwRkAAAAABIDBGQAAAAAEgDlnQA7wXj0AtbBmzRqvfumllzLqBKHh9xCSuuWWW8pahhZcOQMAAACAADA4AwAAAIAAMDgDAAAAgAAw5wwAAvLuu+9m3QIa2IQJE9qtAQDVxZUzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACIA552p3MLNVkl6T1EfS6podOJ289BpCn9s75/pWY8etsiOF8b2WIy99Stn3WrXsSJx7qizrPmuVHSn777VceelTyr5Xzj1xeek16z4598TlpU8p+17bzE9NB2f/PKjZAufcwJofOIG89JqXPishL99rXvqU8tVrGnn6PvPSa176rIS8fK956VPKV69p5On7zEuveemzEvLyvealTynsXrmtEQAAAAACwOAMAAAAAAKQ1eBsekbHTSIvvealz0rIy/ealz6lfPWaRp6+z7z0mpc+KyEv32te+pTy1Wsaefo+89JrXvqshLx8r3npUwq410zmnAEAAAAAfNzWCAAAAAABYHAGAAAAAAGo+eDMzIaZ2SIzW2xmE2t9/PaY2fVmttLMnmu1rJeZ3W9mLxc+9syyx0JP25rZPDNbaGbPm9m/h9prJZGd9MhOeNmRyE/oQs4P2Qkb2akM8kN+kspjdmo6ODOzTpKukPRVSQMkHWtmA2rZQwnNkoZFlk2UNNc5t7OkuYU6a+slne6c213SYEknF/47hthrRZCdiiE74WVHIj/BykF+mkV2gkR2Kor8kJ+k8pcd51zN/kg6QNLsVvUkSZNq2UMZPTZJeq5VvUhSv8Ln/SQtyrrHIj3/RtLheeiV7GTfJ9kJLzvkJ9w/ecgP2QnzD9khP+QnvJ9HHrJT69sat5b0eqt6WWFZyLZwzi2XpMLHzTPux2NmTZI+K+lRBd5rSmSnwshO8IL+mZCfoAX98yA7QQv+50F+ghb0zyMv2an14MyKLONZ/gmZWTdJMyWd6pz7e9b9VBnZqSCyQ3bSID/kJymyQ3bSID/kJ6k8ZafWg7NlkrZtVW8j6Y0a99BRK8ysnyQVPq7MuB9JkpltqJaQ3eqcu6uwOMheK4TsVAjZyUV2pEB/JuQnF/kJ8udBdshOGuSH/CSVt+zUenD2uKSdzWwHM+ssaZSkWTXuoaNmSRpd+Hy0Wu5VzZSZmaTrJC10zl3a6kvB9VpBZKcCyE5usiMF+DMhP7nJT3A/D7JDdtIgP+QnqVxmJ4OJeMMlvSTpFUlnZT3pLtLb7ZKWS/pILf9icYKk3mp5isvLhY+9AujzYLVc2n5G0lOFP8ND7JXshPXzIDvhZYf8hP8n5PyQnbD/kB3yQ34y7zN32bFC4wAAAACADNX8JdQAAAAAgDgGZwAAAAAQAAZnAAAAABAABmcAAAAAEAAGZwAAAAAQAAZnFWRm3zSzh83sLTP7h5ktMrOzC++nAMpmZlub2Xtm5gpvtQfaZWY7mdk1Zva0mX1sZg9k3RPyh3MP0iA/6AgzG2lmz5jZWjNbYmanZd1TCBicVVZvSfMkfVfSVyVdL+ksSZe2txFQxEWS3su6CeTKHvq/d+K8lHEvyC/OPUiD/KAsZnaQpLskPSbpSLX8nfnnZnZqpo0FgPecVZmZ/UzSyZJ6Ov5jowxm9gW1vKn+fLX8ouvunOOXHdplZhs45z4pfH6npD7OuSHZdoU84dyDNMgPOsLMZkvayDl3SKtll0oaI2lL59y6rHrLGlfOqu8tSdzWiLKYWSdJl0k6T9LqjNtBjnw6MAOS4NyDNMgPEthX0v9Els2R1FPSAbVvJxwMzqrAzDqZ2cZmdrCkH0q6iqtmKNP3JXWVdEXWjQBoKJx7kAb5QUd1lRS9Ora28HH3GvcSlM9k3UCdel9Sl8LnN0makGEvyAkz6y3pJ5KOc859ZGZZtwSgAXDuQRrkBwktljQosmy/wsdeNe4lKFw5q44DJX1B0umSRki6PNt2kBM/k/Soc+53WTcCoKFw7kEa5AdJXC1phJmdaGY9zWyoWv7eLEkfZ9hX5rhyVgXOuScLn/7RzFZLutHMLnHOvZJlXwiXme0h6TuSDjGzzQqLNy583NTMPnbOfZhNdwDqFecepEF+kML1kvaRdJWk6ZI+kPRjtcxdXJFhX5ljcFZ9nw7UdpDE4Axt2VnShpIeKfK1ZZKuU8srGgCgkjj3IA3yg0Sccx9LGm9m50jaRtISSbsVvjw/s8YCwOCs+g4qfFySaRcI3R8lHRpZNkwt/4o0XNKrNe8IQCPg3IM0yA9Scc69LeltSTKzH0h62Dn3YrZdZYvBWQWZ2X1qeSzo82q5X/Ygtdw/ewe3NKI9zrnVkh5ovczMmgqf/oF3xaAUM9tYLX8ZkqStJfUws28W6t855z7IpjOEjHMP0iA/SMrMBks6WNJTknpIOlbS0MKyhsbgrLIeV8vL85okrVfLvxhNUsukRwCops0l/Tqy7NN6B0lLa9oNAABt+0jSv0o6V9Inkv4g6SDn3LNZNhUC4/VbAAAAAJA9HqUPAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFgcAYAAAAAAWBwBgAAAAABYHAGAAAAAAFINTgzs2FmtsjMFpvZxEo1hcZAfpAU2UEa5AdJkR2kQX5QDnPOJdvQrJOklyQdLmmZpMclHeuce6Gtbfr06eOampoSHQ/he+KJJ1Y75/qWs25H80N26ls1syORn3q2dOlSrV692spdn3MPWuPcg6Q49yCN9s49n0mx3/0kLXbOvSpJZjZD0ghJbZ6kmpqatGDBghSHRMjM7LUOrN6h/JCd+lbN7Ejkp54NHDiwo5tw7sE/ce5BUpx7kEZ75540tzVuLen1VvWywrLowceZ2QIzW7Bq1aoUh0OdKZkfsoM2cO5BGpx7kBTnHqTBuQdlSTM4K3YpN3aPpHNuunNuoHNuYN++Zd05gMZQMj9kB23g3IM0OPcgKc49SINzD8qSZnC2TNK2reptJL2Rrh00EPKDpMgO0iA/SIrsIA3yg7KkGZw9LmlnM9vBzDpLGiVpVmXaQgMgP0iK7CAN8oOkyA7SID8oS+IHgjjn1pvZeEmzJXWSdL1z7vmKdYa6Rn6QFNlBGuQHSZEdpEF+UK40T2uUc+53kn5XoV7QYMgPkiI7SIP8ICmygzTID8qRanAGAKitdevWxZa9++67Xn3aaad59YgRI2LbRJd16tSpAt0BABrdlClTvPrcc88tuU3Pnj29evHixbF1evXqlaqvvEgz5wwAAAAAUCEMzgAAAAAgAAzOAAAAACAAzDkDgIC9//77Xv2LX/wits4555zT7j5uuumm2LJXXnnFq/v375+gOwBAo4vOMbvuuusy6qQ+cOUMAAAAAALA4AwAAAAAAsDgDAAAAAACwJyzhKLzQCTp8MMP9+qhQ4d69eTJk6vaE4D68+KLL3p1sXv5x4wZ0+4+Zs+eXcmWADSADTbw//1+1KhRsXVuu+02r46+c3GTTTaJbbPNNtt49RtvvJG0RQTi5Zdf9urXX3/dqwcPHhzbJvp35ssvv7zyjeUUV84AAAAAIAAMzgAAAAAgAAzOAAAAACAAzDkr0xNPPOHVxd41FH1v0D333OPVS5cuLXmc8ePHe/XnP//5MjtEPTn22GO9esaMGbF1unbt6tV33323Vw8bNqzyjaHqPvzwQ6++4oorvHrQoEGxbW644YZ293n77bfHlvXq1StBdwAaxZZbbunV0fllxXz729/26muuuaaiPSEMN954o1fvuOOOXj1p0iSv3nfffWP7aGpq8uqLL77Yq4v9Pfukk07y6q222qpkr3nElTMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACAAPBGnD/PnzvTo6EfaOO+6IbRN9od4WW2zR4ePOmzfPqz/66COvLvYiP4Rt5syZXv3www+X3Oapp54quc769eu9+tprr/Xq+++/36t79OgR20f05cXbb799yeOiuv7xj3949YMPPujV5UzKj4o+YAaopOi56JNPPim5zcEHH+zVTz/9tFevW7cuto1zLkF3SCr6cKK5c+fG1undu7dX77zzzl690UYbVb4xZO7FF1/06uHDh3v1F77whZL7iL64+stf/rJXX3jhhbFtvva1r3k1DwQBAAAAAFQNgzMAAAAACACDMwAAAAAIAHPO2hB9wV60jt5XLUk/+clPvDrJC6SPPvpor47OPYq+FFKKv8gPYfmf//kfry42X3Gvvfby6ujPudjPPTqvIzpf8e233/bqnj17xvbRt29fr46+uLp///6xbVBbm222mVfvv//+GXWCRhSdAylJ77//vldHf/fdfPPNJff7wAMPeHX0HLjxxhuX2SGqJfrC3+9///uxdfr16+fVZ5xxhldHz1+oDxdccEHqfUT/Hh09b0Tnl0mNc17gyhkAAAAABIDBGQAAAAAEgMEZAAAAAASAOWdlOuSQQ7x6+vTpsXWSvNeslFmzZnn1n/70p9g6S5YsqfhxUT3F5iLefvvtXt2lS5eS+1m7dq1XR99lFZ3rFp2DJkk/+tGPvHrFihVePWXKlJJ9AMiv6LusoueAu+++O7bNL3/5S6+OniemTZvW4T6WLVvm1dttt12H94HK+s///M+S67zyyitenWSuPRpT9L2F0Xfi/eEPf6hlO0HhyhkAAAAABIDBGQAAAAAEgMEZAAAAAASAwRkAAAAABIAHgrTBzLw6+uK7ak1WPuWUU7y6e/fuXn3TTTdV5bgoz4wZM2LLog9tiXr88ce9eqeddoqt061bN6/u2rVryV6imTzrrLO8+jvf+Y5Xv/POO7F9XHrppV69bt26kscFkA8fffSRV7/44ouxdebPn+/V0ZcIR2upMg+hWrx4sVefeOKJXv3UU0+lPgbSif4emjRpUmyd6ANlNt9886r2hPrx9MvkAoYAACAASURBVNNPe/XYsWO9et68ebFtGuWl5lw5AwAAAIAAMDgDAAAAgACUHJyZ2fVmttLMnmu1rJeZ3W9mLxc+9qxum8gr8oOkyA7SID9IiuwgDfKDtMqZc9Ys6XJJrSc7TZQ01zk31cwmFuofV769xnPooYd69aJFi7w6h3POmlVH+YneIy3FXyAdNWjQIK8eMmRIbJ1OnTp1uJfoNsX229qbb74ZWxbN0wsvvODVxb636Muuq6hZdZQd1FyzGjw/0f/n995779g60TkcP/zhD7262DyjSvjpT3/q1dH5vOXMu62iZjV4doqJzoFva1lrTU1NVeomaM0iPyWtXbvWq1977TWv/uSTT2rZTlBKXjlzzj0kaU1k8QhJNxY+v1HSyAr3hTpBfpAU2UEa5AdJkR2kQX6QVtI5Z1s455ZLUuEjj+dBR5AfJEV2kAb5QVJkB2mQH5St6g8EMbNxZrbAzBasWrWq2odDHSE7SIP8ICmygzTID5IiO5CSv+dshZn1c84tN7N+kla2taJzbrqk6ZI0cOBAl/B4NedcblrNo7LyE0J23n33Xa+Ovv9OKn3P/fe//32vjr5/rFaK9b7RRht59ezZs7262DvcajjnrJi6P/egqnJz7kni448/9uoHH3zQq3v06BHbJnp+mjJlSuUbK6K5ubkmx6kgzj1l+PrXv+7VxeYNvfHGG7VqJyR1fe4pZcGCBbFll1xySQad5EPSK2ezJI0ufD5a0m8q0w4aBPlBUmQHaZAfJEV2kAb5QdnKeZT+7ZIekbSrmS0zsxMkTZV0uJm9LOnwQg3EkB8kRXaQBvlBUmQHaZAfpFXytkbnXFv3MB1W4V5Qh8gPkiI7SIP8ICmygzTID9JKOucMaAjR++cPOyx+bi02L6u1XXbZpaI9JdWzZ/ydlxdffLFXX3PNNV49ffr0qvaE0pYtW+bV119/fWydrOYxIizr16/36ttuu82re/XqFdtm+PDhVe0J9e2VV17x6uhDLObPn1/LdhCIxx57zKuvuOKK2DoPPfSQV48bN86rM37XYaaq/rRGAAAAAEBpDM4AAAAAIAAMzgAAAAAgAAzOAAAAACAAPBCkYOzYsV79yCOPePWAAQNq2Q4C8cADD3j1sGHDYusMGTKkNs2k1Llz59iyz33uc17dv3//WrWDNnTr1s2rzznnHK8+66yzYtvccccd7e5z6tT4U5s/+9nPJugOIevSpYtX33LLLV49ZsyY2DYnnXSSV//oRz/y6t133z22zf7775+wQ9SbaMb69Onj1Ztsskkt20Egfv/733v1nDlzYuv867/+q1cX+z3VqLhyBgAAAAABYHAGAAAAAAFgcAYAAAAAAWDOWUFzc3O7X2fOGYBa+OCDD7z6/PPP9+o333wztk2xZa1NmDAhfWMVEn2x+7PPPuvVS5YsqWU7dS360ukrr7wyts7Pf/5zr54xY4ZXf/jhh7FtLrroIq/eb7/9kraIHHnuuedKrnPZZZd59aabblqtdhCQ1atXe/U777zj1Ycddlhsm//6r/+qak95xpUzAAAAAAgAgzMAAAAACACDMwAAAAAIAHPOACAgm222mVdH309X7H11UdG5RV/+8pfTN1bExx9/3G597LHHxra57777vNrMKt8Yitpmm21iy0477TSvPvPMM7161apVsW3efffdyjaGXFi6dGnJdZqamqreB8ITfT/izTff7NXFfhegbVw5AwAAAIAAMDgDAAAAgAAwOAMAAACAADA4AwAAAIAA8EAQoEEUm8R/ww03eHX0hbMXX3xxVXtCXPSBIG+//XYmfaxbt86rFy1aFFvnzjvv9OrzzjvPq7fbbrvYNk899ZRXDxgwIGmL6KA33ngjtuyss87y6hdeeMGroy+cloq/UBb1Z+XKlV592223xdYp9mJz1L9vfetbXv373//eq7/3ve959aRJk6reUz3hyhkAAAAABIDBGQAAAAAEgMEZAAAAAASAOWdAg3j//fdjy26//XavHjJkiFefe+65VewIIfvv//5vrz755JNj60RfOHv00Ud7dfTFpBJzzKopOk/w9NNP9+quXbvGtjnwwAO9evTo0V79la98pULdIW+iL5V/7733YutE58iiMbz66qteHT1vROecbbvttlXvqZ5w5QwAAAAAAsDgDAAAAAACwOAMAAAAAALAnDMAgC644AKvPvPMM706Or9Mkn7xi1949ciRIyveF8q3fv16r7788su9et99941t8+c//7mqPSG/PvMZ/6+I22+/fUadIEvR3wWS9Je//MWrt9hiC6+++eabq9rTp/7lX/7Fq9966y2vnjt3bof3ef7556fqqRK4cgYAAAAAAWBwBgAAAAABYHAGAAAAAAFgzllBdD5F9L5VIG+i91rfeuutsXWOP/54rz7ooIOq2hNK+/DDD716ypQpXj158uQO7/Oiiy6KLXvhhRe8urm52atvuOEGr95qq61i++AdWGH59re/7dV9+vTx6osvvriW7SDn+vbt69WXXXZZRp0gS9H5yMXMmjWr3bpa1qxZ49UrVqzw6nvuuafD+2TOGQAAAABAEoMzAAAAAAgCgzMAAAAACEDJwZmZbWtm88xsoZk9b2b/Xljey8zuN7OXCx97Vr9d5AnZQRrkB0mRHaRBfpAU2UEllPNAkPWSTnfOPWlm3SU9YWb3Sxojaa5zbqqZTZQ0UdKPq9dqdc2bN8+rzz77bK/+4IMPatlOvch9do499livXrduXWydOXPmePUhhxzi1V27dk3dx2uvvRZb9vDDD3dom2J9RB/osNNOOyXormpyn58kohOpv/GNb3j1fffd1+F9Lly4MLbsb3/7m1dH83TAAQd0+DgBacjs/P73v/fqUucmtKkh8xMVfQDIKaeckkkfv/71r2PLFi9e7NWTJk2qVTulNER29tprL6/eZJNNUu/z448/9uo///nPsXXWr1/v1ddcc03q44ao5JUz59xy59yThc/flbRQ0taSRki6sbDajZJGVqtJ5BPZQRrkB0mRHaRBfpAU2UEldGjOmZk1SfqspEclbeGcWy61hFHS5m1sM87MFpjZglWrVqXrFrlFdpAG+UFSZAdpkB8kRXaQVNmDMzPrJmmmpFOdc38vdzvn3HTn3EDn3MDoOzPQGMgO0iA/SIrsIA3yg6TIDtIo6yXUZrahWkJ2q3PursLiFWbWzzm33Mz6SVpZrSZrIfoS6h49enh1JeacRe+RluIvBV27dq1X9+vXz6ujL4YNXd6zc9ttt3l1sfvao/MTb7/9dq/efPOi/0DWIdE5kZI0duxYr47OKfve977n1VdeeWXqPmot7/lJYtCgQV793nvvefX8+fM7vM+ZM2fGln3xi1/06t69e3d4vyFrxOxssIH/763MMUuuEfMT1aVLF6+eNm1abJ1TTz214seNvjj49NNPj63z+uuve3VAc84aIjvXXnutV++///6p9xl9oXSxOfBvv/126uPkQTlPazRJ10la6Jy7tNWXZkkaXfh8tKTfVL495BnZQRrkB0mRHaRBfpAU2UEllHPl7CBJx0t61syeKiw7U9JUSb8ysxMk/UXSMdVpETlGdpAG+UFSZAdpkB8kRXaQWsnBmXPuj5KsjS8fVtl2UE/IDtIgP0iK7CAN8oOkyA4qoaw5Z4i/M6q5ubnD+yg25yy6n+ics5NOOsmrx4wZ0+HjoroWLVrk1ccff7xXR+/bT2LFihWxZZ07d/bqKVOmePXIkTypN4969vTfTeqcy6gTAI3syCOP9OoBAwbE1unevbtXn3DCCSX3W+odin/961+9+oc//GFsnYMOOqjkcVAZtfod1KtXL6+OzkFrJB16lD4AAAAAoDoYnAEAAABAABicAQAAAEAAGJwBAAAAQAB4IEgbohMgn3zySa+OvgC4Vn0gW1/4whdiy6IPi5k1a5ZXv//++x0+zte+9jWv/tznPhdbZ8MNN/TqESNGePUuu+zS4eMCyK+FCxdm3QLqyBZbbOHVV1xxRWyd7373u159xhlnlNxv9EEPQ4cO9erf/va3Xh196IgUfyAWUE+4cgYAAAAAAWBwBgAAAAABYHAGAAAAAAFgzlkbrrrqqnZrNKbhw4fHlkVfnDhv3jyvTjLn7Nvf/rZXjxo1qsP7ANBYtt9+e6+Ovsx37733jm0zZMgQr545c2bF+0I+bbCB/+/33/rWt2LrFFsGIB2unAEAAABAABicAQAAAEAAGJwBAAAAQACYcwakNHjwYK9evnx5Rp0AaGQffPCBV/fv39+rBwwYENtmxowZVe0JANAxXDkDAAAAgAAwOAMAAACAADA4AwAAAIAAMOcMAIA6sPHGG3v12rVrM+oEAJAUV84AAAAAIAAMzgAAAAAgAAzOAAAAACAADM4AAAAAIAAMzgAAAAAgAAzOAAAAACAADM4AAAAAIAAMzgAAAAAgAOacq93BzFZJek1SH0mra3bgdPLSawh9bu+c61uNHbfKjhTG91qOvPQpZd9r1bIjce6psqz7rFV2pOy/13LlpU8p+14598Tlpdes++TcE5eXPqXse20zPzUdnP3zoGYLnHMDa37gBPLSa176rIS8fK956VPKV69p5On7zEuveemzEvLyvealTylfvaaRp+8zL73mpc9KyMv3mpc+pbB75bZGAAAAAAgAgzMAAAAACEBWg7PpGR03ibz0mpc+KyEv32te+pTy1Wsaefo+89JrXvqshLx8r3npU8pXr2nk6fvMS6956bMS8vK95qVPKeBeM5lzBgAAAADwcVsjAAAAAASAwRkAAAAABKDmgzMzG2Zmi8xssZlNrPXx22Nm15vZSjN7rtWyXmZ2v5m9XPjYM8seCz1ta2bzzGyhmT1vZv8eaq+VRHbSIzvhZUciP6ELOT9kJ2xkpzLID/lJKo/ZqengzMw6SbpC0lclDZB0rJkNqGUPJTRLGhZZNlHSXOfczpLmFuqsrZd0unNud0mDJZ1c+O8YYq8VQXYqhuyElx2J/AQrB/lpFtkJEtmpKPJDfpLKX3acczX7I+kASbNb1ZMkTaplD2X02CTpuVb1Ikn9Cp/3k7Qo6x6L9PwbSYfnoVeyk32fZCe87JCfcP/kIT9kJ8w/ZIf8kJ/wfh55yE6tb2vcWtLrreplhWUh28I5t1ySCh83z7gfj5k1SfqspEcVeK8pkZ0KIzvBC/pnQn6CFvTPg+wELfifB/kJWtA/j7xkp9aDMyuyjGf5J2Rm3STNlHSqc+7vWfdTZWSngsgO2UmD/JCfpMgO2UmD/JCfpPKUnVoPzpZJ2rZVvY2kN2rcQ0etMLN+klT4uDLjfiRJZrahWkJ2q3PursLiIHutELJTIWQnF9mRAv2ZkJ9c5CfInwfZITtpkB/yk1TeslPrwdnjknY2sx3MrLOkUZJm1biHjpolaXTh89FquVc1U2Zmkq6TtNA5d2mrLwXXawWRnQogO7nJjhTgz4T85CY/wf08yA7ZSYP8kJ+kcpmdDCbiDZf0kqRXJJ2V9aS7SG+3S1ou6SO1/IvFCZJ6q+UpLi8XPvYKoM+D1XJp+xlJTxX+DA+xV7IT1s+D7ISXHfIT/p+Q80N2wv5DdsgP+cm8z9xlxwqNAwAAAAAyVPOXUAMAAAAA4hicAQAAAEAAGJwBAAAAQAAYnAEAAABAABicAQAAAEAAGJxVmJkNMLO5ZvaBmb1hZueZWaes+0L4yA7SMLPPmNlEM3vZzNaa2TIz+0XWfSF8ZjbSzJ4p5GaJmZ2WdU/IBzN7wMxcG38OyLo/hI38FPeZrBuoJ2bWU9L/SHpB0ghJO0q6RC2D4LMzbA2BIzuogBskHSZpiqQXJW0raUCmHSF4ZnaQpLskXS/pR5L2l/RzM/vEOTct0+aQBz+Q1COy7DxJn1XLS5SB9pCfInjPWQWZ2SRJZ0ja3jn398KyMySdK2nLT5cBUWQHaZjZMEn/T9I+zrkXsu4H+WFmsyVt5Jw7pNWySyWNUcu5Z11WvSF/zKyzpDcl3eGcOynrfpAv5KcFtzVW1lclzY78RXqGpI0kfTGblpATZAdpfEfS/zIwQwL7quWqfWtzJPWU1LC3FSGxYWrJzu1ZN4JcIj9icFZpu6nldqJ/cs79RdIHha8BbSE7SGN/SS+Z2eVm9vfCvMW7zGyrrBtD8LpKil4dW1v4uHuNe0H+jZL0V0l/yLoR5BL5EYOzSusp6Z0iy98ufA1oC9lBGluq5Ta0fdXyy22spM9LutvMLMO+EL7FkgZFlu1X+Nirxr0gx8xsY0lHquWWNObMoEPIz//hgSCVVyxQ1sZyoDWyg6Ss8GeEc+4tSTKz5ZIelPQlSXMz7A1hu1rSVWZ2oqQ71TIwO73wtY8z6wp5dKSkbmrwW9KQGPkp4MpZZb0tabMiyzdV8asiwKfIDtJ4W9Kznw7MCv6oltvVeGIj2nO9CgM0SWvU8uTG8wpfW5FVU8ilUZIWO+cWZN0Icon8FDA4q6wXFZkfZGbbStpEkflEQATZQRoL21hukj6pZSPIF+fcx8658ZL6Stpb0haS5he+PL/NDYFWzGxTtTzYquGveqDjyI+PwVll/V7SUDPr3mrZv0r6UC23FwFtITtI47eS9jazPq2WHSJpQ0lPZ9MS8sQ597Zz7lnn3HtqeffQw845/mEI5TpKUhfxl2skQ35a4T1nFVR4kfALkp6T9HNJ/SVdKmmac44XCaNNZAdpmFkPtWTnr5LOl9RdLTl60Tl3eJa9IWxmNljSwZKeUsvLYI+VNFTSwc65Z7LsDflhZvep5b14+2bdC/KH/Pi4clZBzrm3JR0mqZNaXgg7RdIvJE3Osi+Ej+wgjcL78b6klrlnMyRdoZaHgPxLln0hFz5Sy1X6eyQ1S9pY0kEMzFCuwhX7w9Ry7gE6hPzEceUMAAAAAALAlTMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACACDMwAAAAAIAIMzAAAAAAgAgzMAAAAACECqwZmZDTOzRWa22MwmVqopNAbyg6TIDtIgP0iK7CAN8oNymHMu2YZmnSS9JOlwScskPS7pWOfcC21t06dPH9fU1JToeAjfE088sdo517ecdTuaH7JT36qZHal2+XniiSe8eo899vDqrl27pj7GO++8E1u2fPlyr959991THycvli5dqtWrV1u563PuQWv1cu5B7XHuQRrtnXs+k2K/+0la7Jx7VZLMbIakEZLaPEk1NTVpwYIFKQ6JkJnZax1YvUP5ITv1rZrZkWqXHzP/9/Sdd97p1bvttlvqY9x9992xZVOnTvXqRx99NPVx8mLgwIEd3YRzD/6pXs49qD3OPUijvXNPmtsat5b0eqt6WWFZ9ODjzGyBmS1YtWpVisOhzpTMD9lBGzj3IA3OPUiKcw/S4NyDsqQZnBW7lBu7R9I5N905N9A5N7Bv37LuHEBjKJkfsoM2cO5BGpx7kBTnHqTBuQdlSXNb4zJJ27aqt5H0Rrp20EDID5Jq6OxEb51EhzV0fpAK2UEa5AdlSXPl7HFJO5vZDmbWWdIoSbMq0xYaAPlBUmQHaZAfJEV2kAb5QVkSXzlzzq03s/GSZkvqJOl659zzFesMdY38ICmygzTID5IiO0iD/KBcaW5rlHPud5J+V6Fe0GDID5IiO0iD/CApsoM0yA/Kkeol1AAAAACAykh15ayR/Md//IdXT5s2LbbOyJEjvbrY+4gA1K958+Z59XbbbZdRJwAAII+4cgYAAAAAAWBwBgAAAAABYHAGAAAAAAFgzlkbJk6c6NUzZszIqBMAeTFkyJCsWwAAADnGlTMAAAAACACDMwAAAAAIAIMzAAAAAAgAc84KJk+e7NV9+/b16gsuuKDkPninEYBqc85l3QIAAKmsW7fOq7t06eLVnTt3LrlNvf4+5MoZAAAAAASAwRkAAAAABIDBGQAAAAAEgDlnBbNmzfLq6By0kSNH1rId1JHoPdLdu3fv8Db1el81ACAfzKwi++H3WeN56623Yst22WUXr+7Vq1fJbXr37t3uOtGv5xVXzgAAAAAgAAzOAAAAACAADM4AAAAAIAAMzgAAAAAgADwQpOBLX/qSV7/22mtevXjxYq/eaaedqt4T6kP0xYrlTIau1MRr1B+ygU8NGzbMq2fPnu3V5557rldHH3SVpSlTpnh1tNckeNBEPkTPYfzc6s/SpUu9eocddoitk+TnHn0ASL1miStnAAAAABAABmcAAAAAEAAGZwAAAAAQAOacFVxyySVefdRRR3n1G2+84dXHHXdcyX1utNFGXs08tcbU1NSUyXGj93xn1QfS+etf/+rVzz33XEadoBFE51cfccQRVTnO6tWrq7JfhKXYHKDoPKF6nTfUyKJzzPbaa6+MOsknrpwBAAAAQAAYnAEAAABAABicAQAAAEAAmHPWhs98xv9PM23aNK++9NJLY9usX7/eq/fZZx+vfuyxx2LbdO7cOWmLyIklS5ZkctzoPd/cx59Pd999t1f//Oc/j60Tfd8V6s/YsWNjy4455hivnjhxolc/+OCDXn3RRRfF9jFhwgSvXrdunVe/9NJLHeqzmOgxJGn48OGp94vqKueditHfK9XaBvn2zDPPZN1CrnDlDAAAAAACwOAMAAAAAALA4AwAAAAAAsDgDAAAAAACwANB2nDFFVd4dXQi9Q033BDb5rzzzvPqhQsXevWgQYNi2/zhD3/w6h49enSoT4TvkUce8eoDDjggo06QB2vXrvXq6AMaDj/88Ng2v/71r6vaE7L3+OOPx5aNGjXKq4cMGeLV0QeCvPjiiyWPs/XWW3v1ddddF1vn5JNP9uro775tttnGq/fbb7/YPnbccceSvSAs5TxUKsmDp0o9IISHWeFTc+bMybqFmuDKGQAAAAAEgMEZAAAAAASg5ODMzK43s5Vm9lyrZb3M7H4ze7nwsWd120RekR8kRXaQBvlBUmQHaZAfpFXOnLNmSZdLuqnVsomS5jrnpprZxEL948q3l53NN9/cq6+++mqv/tWvfhXbJjqX6Nxzz/Xq6IutJWnjjTdO2GFuNKsB89PagQce6NXVun++Du/FblYDZueCCy7w6qlTp3o1L+8tW7MaMD9pbbrppl593HHHxdb54IMPvHrZsmVe/fWvf92rczi/rFkNmB1eBl0xzWrA/KQ1c+bMkut885vf9Oqjjz66Wu1kquSVM+fcQ5LWRBaPkHRj4fMbJY2scF+oE+QHSZEdpEF+kBTZQRrkB2klnXO2hXNuuSQVPm7e1opmNs7MFpjZglWrViU8HOpMWfkhOyiCcw/S4NyDpDj3IA3OPShb1R8I4pyb7pwb6Jwb2Ldv32ofDnWE7CAN8oOkyA7SID9IiuxASv6esxVm1s85t9zM+klaWcmmQnDttdd6dfRfMKLveZGk/v37e/VXvvKVyjdWH+o+Px1Vzr3WpQwdOtSr6/Re7LrPTvS9ZtF77C+88MJatlNvcpuf0aNHx5Y1NTVV/bidO3eOLRs/frxX77nnnl596KGHevWuu+5a+cZqL7fZSSqU94sVmwsXSm8d0FD5GTNmTIe3+e1vf+vVzc3NJfdb7J3D9SDplbNZkj79TTFa0m8q0w4aBPlBUmQHaZAfJEV2kAb5QdnKeZT+7ZIekbSrmS0zsxMkTZV0uJm9LOnwQg3EkB8kRXaQBvlBUmQHaZAfpFXytkbn3LFtfOmwCveCOkR+kBTZQRrkB0mRHaRBfpBW0jlndWfGjBle/eqrr3p1dD7PkCFDqt0S6lSx+6jHjh3r1Unu146qxDw2ZK9nT/9dpVtttVVGnSBLEyZMyLoFoCai88l4/1r+ROeCXX755bF1onNXo9t069Ytts1ll11Wge7CV/WnNQIAAAAASmNwBgAAAAABYHAGAAAAAAFgcAYAAAAAAeCBIAUPPfSQVx933HFefeCBB9ayHdSx6MM/pPjE2HImvRabYIv8ueqqq7x6o4028uo6fZk4SnjggQe8etCgQbF1Ntlkk3b3sf3223v1xhtvnLovAOioU045JbYs+veeqEZ5+EcxXDkDAAAAgAAwOAMAAACAADA4AwAAAIAANOScs6VLl8aWTZs2zas7d+5co25Q7yZNmlRyne7du3d4v9F7uMs5DsIzf/58r95tt928euDAgbVsB4E49NBDvfree++NrfPFL37Rq9977z2vPvLII726d+/eFeoOANI588wzvfr888/PqJPwcOUMAAAAAALA4AwAAAAAAsDgDAAAAAAC0JBzznbdddfYsscee8yr99lnn1q1gzpXzn3UZubV7777blWOAyCfjjjiiNiyu+++u926f//+Xj158uTKNwYAJRR7Z1l03nz07z285wwAAAAAkCkGZwAAAAAQAAZnAAAAABAABmcAAAAAEICGfCDIunXrYsuccxl0ArSITnyNTpQtZxuE73//939jy7bcckuv3mOPPWrVDnLu2GOP9eqPP/7Yq88666xatgMART3xxBMl17n88su9upH/jsOVMwAAAAAIAIMzAAAAAAgAgzMAAAAACEBDzjkDQjN+/HivLmfOWTn3cCMsP/vZz2LLRo8e7dVf//rXUx9n7dq1Xv3II4+k3mcSPXv2jC3bZ599Mugk/+69997Ysp122qndbXr37l2tdgCgTd/4xje8+q677oqtc+edd9aqndzhyhkAAAAABIDBGQAAAAAEgMEZAAAAAASgIeecnX/++bFlM2fO9Opf/epXfjjWdQAAIABJREFUqY/Tr18/ry5nHhEaU/T+7Khi92Z/85vf9Oq///3vXh3NNML05ptvevVLL72Uep8rVqzw6kMPPTT1PpMYPHhwbNmNN97Y7ja77LJLtdrJlXnz5nn1lVdeGVtnwoQJXj1o0KCq9gQA5YjOMZs9e3Zsna985Su1aid3uHIGAAAAAAFgcAYAAAAAAWBwBgAAAAABaMg5Z5MmTYotmzx5slffdNNNXv2Xv/ylw8fZbrvtvLp79+4d3kcS0ePuu+++sXVmzZrl1WPGjKlmSyih1P3Zxe7Njq4zdOjQyjeGqvvxj3/cbp1n8+fPjy3bdddd293GOVetdnJlyJAhXh19F6IkrVmzpkbdAEByzC/rGK6cAQAAAEAAGJwBAAAAQABKDs7MbFszm2dmC83seTP798LyXmZ2v5m9XPjYs/rtIk/IDtIgP0iK7CAN8oOkyA4qoZwrZ+slne6c213SYEknm9kASRMlzXXO7SxpbqEGWiM7SIP8ICmygzTID5IiO0it5ANBnHPLJS0vfP6umS2UtLWkEZKGFFa7UdIDknI7k33KlClevXbtWq9+8MEHUx/jmmuu6fA2ixYt8uq333675DaHHHKIV59wwgmxdcaOHevV1XggSKNkpxrKmTwbXefhhx+uVjuZqMf8zJ07N+sWGkI9Zicr//jHP2LLfvrTn3r18ccf79WlHvoSukbNT/RhPGbW7tcR16jZQWV1aM6ZmTVJ+qykRyVtUQjhp2HcvNLNoX6QHaRBfpAU2UEa5AdJkR0kVfbgzMy6SZop6VTn3N87sN04M1tgZgtWrVqVpEfkHNlBGuQHSZEdpEF+kBTZQRplDc7MbEO1hOxW59ynL2RaYWb9Cl/vJ2llsW2dc9OdcwOdcwP79u1biZ6RI2QHaZAfJEV2kAb5QVJkB2mVnHNmLTcdXydpoXPu0lZfmiVptKSphY+/qUqHGZk6dWrWLUiSjjrqKK++5557Sm7z0EMPtVvXSqNmJysHHHBA1i1UFPlBUmSnctatWxdbdtlll3n1I4884tVNTU3VbKnqyE+28jzXjewUt8MOO8SWLVmyJINO8qHk4EzSQZKOl/SsmT1VWHamWgL2KzM7QdJfJB1TnRaRY2QHaZAfJEV2kAb5QVJkB6mV87TGP0qyNr58WGXbQT0hO0iD/CApsoM0yA+SIjuohA49rREAAAAAUB3l3NaIDN19991ZtwAAyNjw4cNjy6Lvwdxpp528escddyy53+XLl3v1Aw884NXr16+PbTNy5Eiv7tGjR8njIH9Kvfes2DqAJK1evdqr+/TpE1und+/etWqnXW+99VbWLcRw5QwAAAAAAsDgDAAAAAACwOAMAAAAAALA4AwAAAAAAsADQYAARCfPRifKljNhNck2APLhwgsvjC0bNmyYVy9btsyriz1EJOrxxx/36jPOOMOriz3s4/XXXy+5DhpDsYeEtFbOA0NK7QP5E/37yNq1azu8jy5dulSqndzhyhkAAAAABIDBGQAAAAAEgMEZAAAAAASAOWdAAKL3Z69Zs8ary7n3et26dRXtCUDYttpqK6+ePn26V1900UUl97HJJpt49S677NLu1yVpgw34d91GVGz+WKn5Yknmk/Fi6/rTuXPnDm/z/9m79ziryvLv498rFE+IgoCRmuNZKUhs8pDnA0r+VHqiEjyB2UERT1gJ5AktT6Xpz0zDQjIVNfNA5jHSkif1CdQUBcQUdIQElFTUSPR+/thbm3utPTN71lp7r3vt+bxfr3kN15p9uLbzdc1cs/a9VlfOAXtYAAAAAAgAwxkAAAAABIDhDAAAAAACwJozIEBd+b3WAKozZcoUr46u74l+vZI99tjDq++77770jaHL4GcVkD2OnAEAAABAABjOAAAAACAADGcAAAAAEADWnAEA0AB+9atftVsDAMLHkTMAAAAACADDGQAAAAAEgOEMAAAAAALAcAYAAAAAAWA4AwAAAIAAMJwBAAAAQAAYzgAAAAAgAAxnAAAAABAAc87V78nMlklaJKmPpOV1e+J0itJrCH1u7pzrW4sHbpUdKYzXWo2i9Cnl32vNsiOx76mxvPusV3ak/F9rtYrSp5R/r+x74orSa959su+JK0qfUv69tpmfug5nHz+p2SznXHPdnziBovRalD6zUJTXWpQ+pWL1mkaRXmdRei1Kn1koymstSp9SsXpNo0ivsyi9FqXPLBTltRalTynsXnlbIwAAAAAEgOEMAAAAAAKQ13A2OafnTaIovRalzywU5bUWpU+pWL2mUaTXWZRei9JnForyWovSp1SsXtMo0ussSq9F6TMLRXmtRelTCrjXXNacAQAAAAB8vK0RAAAAAAJQ9+HMzIaa2Xwze8HMxtf7+dtjZlPMbKmZzWm1rbeZPWhmC8qfe+XZY7mnzczsITOba2bPmtkpofaaJbKTHtkJLzsS+QldyPkhO2EjO9kgP+QnqSJmp67DmZl1k3SVpC9JGiBppJkNqGcPHZgqaWhk23hJM5xz20iaUa7ztlrS6c65HSTtKunE8n/HEHvNBNnJDNkJLzsS+QlWAfIzVWQnSGQnU+SH/CRVvOw45+r2IWk3Sfe3qidImlDPHqrosUnSnFb1fEn9y//uL2l+3j1W6PkuSUOK0CvZyb9PshNedshPuB9FyA/ZCfOD7JAf8hPe96MI2an32xo3kfRKq7qlvC1kGzvnlkhS+XO/nPvxmFmTpMGSHlfgvaZEdjJGdoIX9PeE/AQt6O8H2Qla8N8P8hO0oL8fRclOvYczq7CN00UmZGY9JP1O0qnOubfy7qfGyE6GyA7ZSYP8kJ+kyA7ZSYP8kJ+kipSdeg9nLZI2a1VvKmlxnXvorNfMrL8klT8vzbkfSZKZralSyG50zt1e3hxkrxkhOxkhO4XIjhTo94T8FCI/QX4/yA7ZSYP8kJ+kipadeg9nf5O0jZltYWbdJY2QNL3OPXTWdEmjyv8epdJ7VXNlZibpV5LmOucua/Wl4HrNENnJANkpTHakAL8n5Kcw+Qnu+0F2yE4a5If8JFXI7OSwEO9gSc9L+oekH+S96C7S2zRJSyS9r9JfLI6TtJFKZ3FZUP7cO4A+91Dp0PbTkp4qfxwcYq9kJ6zvB9kJLzvkJ/yPkPNDdsL+IDvkh/zk3mfhsmPlxgEAAAAAOar7RagBAAAAAHEMZwAAAAAQAIYzAAAAAAgAwxkAAAAABIDhDAAAAAACwHCWMTNbw8zGm9kCM1tlZi1m9tO8+0KxmNkmZrbSzFz5qvZAm8zs4XJWKn3slnd/CJuZjTCzJ8r7nFfN7Hoz+1TefaEYzGyAmc0ws3fNbLGZnWdm3fLuC+Hjd+bK1si7gQZ0naT9JU2SNE+lq7sPyLUjFNGPJa2UtF7ejaAQxkjqGdl2nqTBKl3IFKjIzA5T6XpFV0n6nqT+kn4o6W4za3bOfZhnfwibmfWS9EdJz0kaJmkrSZeq9Mf/M3NsDcXA78wVMJxlyMyGqnQF9885557Lux8Uk5ntKWmopAtUGtKAdkX3N2bWXVKzpFucc6vz6QoFcYSkJ5xzYz/aYGZvSbpL0naS5ubVGArheEnrSPqKc+4tSQ+aWU9J55rZJeVtQAy/M7eNtzVm6xuS/kTIkFT5rSBXqnTUY3nO7aC4hkrqpdIREaA9a0p6M7LtX+XPVudeUDxfknR/ZAi7WaWBbe98WkJB8DtzGxjOsrWLpOfN7Gdm9lb5/de38959dMLxktZW6S1GQFIjJL0q6ZG8G0Hwpkja08yOMbOeZratSm9rfIhfmlCF7VV6O9rHnHMvS3q3/DWgLfzO3AaGs2x9UtJoSTuq9MvRsZI+L+kOM+MvkGiXmW0k6XxJ45xz7+fdD4rJzNaVdKhKb2l0efeDsDnn/qDSz63JKh1Bmy+pm6Sv5NgWiqOX/nuktbUV5a8BbeF35jaw5ixbVv4Y5px7XZLMbImkP0vaT9KMHHtD+H4k6XHn3D15N4JCO1RSD/GWRlTBzPaVdI2kKyTdK2ljSeeq9AvSAc65D3JsD8VQ6Y9A1sZ24CP8ztwGhrNsrZD04kchK5sp6T8qnX2mywYN7TOzz6j0/uu9zGzD8uZ1y583MLMPnHPv5dMdCmaEpBecc7PybgSFcKmk6c65Mz7aYGZPqfRWtWGSbs+rMRTCCkkbVti+gSofUQM+wu/MbeBtjdlq66xWJonTEaM926i0MP9RlXZYK/TfdWctKp0kBGiXmW2g0gJ9jpqhWttLeqr1BufcfEnvqXRadKA98xRZW2Zmm6l0GZh5Fe8BlPA7cxsYzrJ1t6RBZtan1ba9VPql++/5tISCmClp38jHxeWvHSxOqY/q/B9Ja4nhDNVbJGmn1hvMbAeVzra3MI+GUCj3SjrIzNZvte1wlYb7P+fTEgqC35nbYKwXz0752h5zVDpL2gWS1lfpF+x5zrkhefaG4jGz0SpdoHF959zKnNtBAZjZfZI+6ZzbMe9eUAxmdoqkn5Y/Plpzdrak7pI+65x7J8f2ELjyRaifU+l3n4slbSnpMkmXO+e4CDXaxO/MbePIWYbK1/nYT6W3pN2s0tvSZkj6ep59AWh85b8+7q/Svgeo1v9KOlHSEJUuPH2JSm9z3J/BDB1xzq1Qab/TTdLvJU1SadA/J8++ED5+Z24bR84AAAAAIAAcOQMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAgAwxkAAAAABIDhDAAAAAACwHAGAAAAAAFgOAMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAgAwxkAAAAABIDhDAAAAAACwHAGAAAAAAFgOAMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAgAwxkAAAAABIDhDAAAAAACwHAGAAAAAAFgOAMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAgAwxkAAAAABIDhDAAAAAACkGo4M7OhZjbfzF4ws/FZNYWugfwgKbKDNMgPkiI7SIP8oBrmnEt2R7Nukp6XNERSi6S/SRrpnHuurfv06dPHNTU1JXo+hG/27NnLnXN9q7ltZ/NDdhpbLbMjkZ9GtnDhQi1fvtyqvT37HrTGvgdJse9BGu3te9ZI8bg7S3rBOfeiJJnZzZKGSWpzJ9XU1KRZs2aleEqEzMwWdeLmncoP2WlstcyORH4aWXNzc2fvwr4HH2Pfg6TY9yCN9vY9ad7WuImkV1rVLeVtQDXID5IiO0iD/CApsoM0yA+qkmY4q3QoN/YeSTP7tpnNMrNZy5YtS/F0aDAd5ofsoA3se5AG+x4kxb4HabDvQVXSvK2xRdJmrepNJS2O3sg5N1nSZElqbm5OtsANjajD/JAdtIF9D9Jg39PKnnvu6dUzZ87s8D5XXHGFV5988smZ9hQw9j1Ig30PqpLmyNnfJG1jZluYWXdJIyRNz6YtdAHkB0mRHaRBfpAU2UEa5AdVSXzkzDm32szGSrpfUjdJU5xzz2bWGRoa+UFSZAdpkB8kRXaQBvlBtdK8rVHOuXsk3ZNRL+hiyA+SIjtIg/wgKbKDNMgPqpHqItQAAAAAgGykOnIGAACK4Ytf/GJsW/Q6SjfeeKNXH3TQQbH7rLfeetk2BgD4GEfOAAAAACAADGcAAAAAEACGMwAAAAAIAGvOAABoQLvssotXX3zxxbHbNDU1eXW/fv28et111828L4Rh88039+qXX365049x5ZVXevXYsWNT9QSAI2cAAAAAEASGMwAAAAAIAMMZAAAAAASANWcAADSAQYMGefVVV13l1TvvvHPsPmuttVZNe0IYtttuu9i2xYsXe/Udd9zh1QMGDPDqcePGxR7j/PPPb/d5WYMGdB5HzgAAAAAgAAxnAAAAABAAhjMAAAAACABrzgAgIFtssYVXL1y4sMP7TJ061atHjRqVYUcoimeeecart956a69mfVnX9fzzz8e23XLLLV59wAEHeHWPHj28+pJLLok9xplnnunVl156qVd/+OGHsfucfPLJ7TcLdHEcOQMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAE4IAhTAmDFjYtuuvvrq1I972GGHefVdd92V+jGRrQcffNCrBw8eHLvNhAkTvPrtt9/2ai4E2zU55/JuAYG46aabYtsOOeQQr1533XXbfYzoRaklaeLEiV594YUXevWCBQuqbRGBip6kSoqfqOqvf/2rV++22261bKnhceQMAAAAAALAcAYAAAAAAWA4AwAAAIAAdMk1Z5MmTYptmz17tldPnz69Xu0AOuqoo7z63nvv9ep33323Js973333efVGG22U+XO8/vrrmT9mI4m+n3/atGleHV1jVulCwu+//75Xv/feexl1918tLS2xbXvuuadXv/TSS5k/L5Izs7xbQCBGjhxZk8dtbm726kGDBnn18uXLa/K8QCPjyBkAAAAABIDhDAAAAAACwHAGAAAAAAHoEmvOzj33XK++8cYbY7d56623vPqII47w6p/97Gde3bt372ya68ANN9zg1ZV6z8Iuu+zi1dH/ZshOpff+R68vlmTN0AknnODV3//+9zu8zx//+Eev/ta3vtXp543+vxBdv4n2Ra8X88lPftKrK60xy8Pq1atj26K9I1/XXXedV0f342effXbsPptsskktWwJQMNHfB1999dWcOsnH888/H9sWvSZgksfoDI6cAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAhAlzghyBNPPOHVTU1NsdtEF0U/8MADXr1q1arUfURPOiLFTzTy9ttve/XMmTPbrbPywgsveHX09V544YU1ed6u4Otf/7pX33PPPbHbRE8AcsEFF3h1NYtR+/bt69XRE0tU01t0IXA11ljD341U+v8LxTN37lyvPvLII3PqBNUaPXq0V0+cONGrTznllNh9OCEIgNaivw/edtttsducdtpp9Wonc4899phXn3zyyV5d6YRsCxYs8OoNNtjAqx988MGMuivhyBkAAAAABIDhDAAAAAAC0OFwZmZTzGypmc1pta23mT1oZgvKn3vVtk0UFflBUmQHaZAfJEV2kAb5QVrVrDmbKulnkq5vtW28pBnOuYvMbHy5PiP79pK5+eabvbqlpcWrt9pqq9h9+vXrV9OeJOmdd96JbYuuOVuyZEnmz7vnnnvGtnW0ziDDi8tOVcHyk7Xo+7cvuuii2G0+//nPe/XWW2/t1dH1ZFnp2bOnVw8cOLAmz5PQVHXx7FRy1llnxbbdf//9Xj1gwIDUz7Ny5UqvXrx4cYfPG5ipIj+eyy+/PLatT58+Xn3MMcd49fbbb1/TngI1VWQnkc0339yre/XqkjPIVBUoP/vvv79XR899MGjQoNh91ltvvZr2VK0777zTq6+88soO7/PGG2949VNPPdXp5+3WrZtXf+ELX+j0Y7SnwyNnzrm/SHojsnmYpF+X//1rSV/OtCs0DPKDpMgO0iA/SIrsIA3yg7SSrjnb2Dm3RJLKn9s87GRm3zazWWY2a9myZQmfDg2mqvyQHVTAvgdpsO9BUux7kAb7HlSt5icEcc5Nds41O+eaa/XWLDQmsoM0yA+SIjtIg/wgKbIDKfl1zl4zs/7OuSVm1l/S0iybSmvkyJHtfv3JJ5+MbevRo4dXDx482Kuj7y+tl+h6gGhdjUmTJsW27bvvvol7ykDQ+Unr2GOP9epx48Z59cEHHxy7T+/evWvaUwNp6OxI0pgxY7x6nXXW8eq//vWvsft89atf9ephw4Z1+nmj14M84wx/OUS0D0k68MADO/08OWv4/LTn2muv7fA2X/ziF726i645q6RLZ6ctS5f6/xkOPfRQr95oo43q2U7Igs3Pn/70p3a/fsIJJ8S2vfzyy7Vqp1233367V59//vlenWT92HbbbefVP/rRjzq8z1prrdXp5+mMpEfOpksaVf73KEl3ZdMOugjyg6TIDtIgP0iK7CAN8oOqVXMq/WmSHpW0nZm1mNlxki6SNMTMFkgaUq6BGPKDpMgO0iA/SIrsIA3yg7Q6fFujc66t9wju38Z24GPkB0mRHaRBfpAU2UEa5AdpJV1z1nCi1wK74447vLp79+71bOdjxx9/vFf/4Ac/6PRj5NV7VzVz5kyvPvnkk72a9WVoT/SaM9FrIVZaUxtdpxa9Tl41/vnPf3r1Sy+95NU//vGPO/2YqK9oVg4//HCvXr16dYeP0dTUlGVLaHDRNT/RdfHnnHNOPdtBDdx3332xbeedd55XV7p+cC3Mnz/fq6NrzHbffXevPu644zp8zOiJVw455JCE3WWn5mdrBAAAAAB0jOEMAAAAAALAcAYAAAAAAWA4AwAAAIAAdIkTgvzkJz/x6uiCQUnq1auXV9f6AnPVWmMN/1u09tpr59QJqvXhhx96dXRB9CWXXBK7Dxd6xUeGDx/u1ZtttplX77zzzrH7JDmJw+OPP+7VjzzyiFdffPHFXh290DXCc9JJJ3n14sWLvbp///71bAdAAVx22WVefdZZZ3n1uHHjYveJXpg6eiKYvGyzzTZefeyxx+bUSTocOQMAAACAADCcAQAAAEAAGM4AAAAAIABdYs3Zdttt59W77rprLn1ssMEGsW3RdR1c6LX4zj33XK+OXjh84sSJsftMmjTJqwcOHJh5XwjTBRdc4NW//OUvvTp6IfpNN920088RXU8mxdc+fvDBB1594YUXdvp5ADSuSmuPtthiC68eNWpUvdpBRk477TSv/uEPf+jV3/jGN2L3CWWNWaPiyBkAAAAABIDhDAAAAAACwHAGAAAAAAHoEmvOgHo6+uijvfqnP/2pV99xxx2x+3Tr1s2ro9fqOOyww2L3yWvtJLI1YcIEr46u4XjjjTe8+vTTT489xlZbbdXuczz22GOxbS+//LJXV1pPgmJzzuXdAhrIkiVLYtu23HJLr95oo43q1Q6gHXfc0asPPPDAnDrJFkfOAAAAACAADGcAAAAAEACGMwAAAAAIAMMZAAAAAASgS5wQZP78+V4dXUAoJbuwa2e9+eabsW1nnHGGV0cX3A4fPrymPSEMt912W7tfr5RPTgjSmKIXcb3mmmu8utJJHj7zmc+0+5jvvfdebNv3vvc9rz7qqKOqbREFYWZ5t4AGctJJJ8W2cTFi5GmPPfbw6pEjR+bUSbY4cgYAAAAAAWA4AwAAAIAAMJwBAAAAQAAacs3Z2muv7dXRtRUrV66M3eecc86paU/V6t69u1dHL06M4hkxYoRXL126NHabV199tV7tIHDnnnuuV3/iE/7f0F566aXYfa644gqvfuGFF7w6ug+UWGPWFcydO9ere/fuHbvNWmutVa92UDCTJ0/26i984Qux22y77bb1agc5ue+++2LbjjjiCK/u2bNnvdrxNOq6Wo6cAQAAAEAAGM4AAAAAIAAMZwAAAAAQgIZcc/bkk0969ZgxY3LqpPOia0e++tWv5tQJkrr99tu9ul+/fl69/vrrx+6z3377efWnP/1pr95+++0z6g5Fc/bZZ3d4m2OPPdaro2vOnn/++dh9Zs2a5dXNzc0JukNImpqavHrIkCFeXWntyDbbbOPV0f3Vuuuum7qvSuu8ly9fnvpxo6KvH+lEr79ZaV3R4MGD69UO6mTvvff26tNOOy12m913392rBw4cWNOe2lLpup+NgCNnAAAAABAAhjMAAAAACADDGQAAAAAEoCHXnEXX5/To0SOXPp544gmvPu+882K3WbFihVdvuummXt2nT5/sG0Om7r77bq8+8cQTvXrAgAHt1pI0fvx4r650PRmgLZ/5zGe8+sADD/Tq6PWuJOmqq67y6uuuuy77xlBX0Wvgffazn/XqoUOHxu7z4YcfevWNN97o1QcddFDqvm699dbYtlqsBW/U9SdAPUXXzW+xxRax2zz77LPt3iaL37tbWlpi21555RWvXmONhhxjOHIGAAAAACFgOAMAAACAADCcAQAAAEAAOhzOzGwzM3vIzOaa2bNmdkp5e28ze9DMFpQ/96p9uygSsoM0yA+SIjtIg/wgKbKDLFSzkm61pNOdc0+Y2fqSZpvZg5JGS5rhnLvIzMZLGi/pjNq1mtzw4cO9euHChbHbXHTRRZk/77x587x6xowZsdt885vf9Oqtttoq8z5yVPjsRD3wwAOxbUcffbRXRy/G+Itf/MKrt9566+wba0wNl59a+e53v9tu/eMf/zh2n+eee66mPeWM7EiaM2eOV0dPECLFF/YfeeSRNe2pIMhPBZX2GQ8//HD9G6lgn332ybuFjzRcdiqdEGT06NFeff3113t1Fhcnv+yyy2LbbrnlFq+udIHsRtDhkTPn3BLn3BPlf78taa6kTSQNk/Tr8s1+LenLtWoSxUR2kAb5QVJkB2mQHyRFdpCFTq05M7MmSYMlPS5pY+fcEqkURkn92rjPt81slpnNWrZsWbpuUVhkB2mQHyRFdpAG+UFSZAdJVT2cmVkPSb+TdKpz7q1q7+ecm+yca3bONfft2zdJjyg4soM0yA+SIjtIg/wgKbKDNKq6epuZralSyG50zn10dbrXzKy/c26JmfWXtLRWTaY1atQor77hhhtit4m+t/XJJ59M/bz9+vl/GDn00ENjt5k0aZJX9+7dO/XzhqTo2YmqdEHW6AWjp02b5tWbbLJJTXtqZI2WnyxUWjO7wQYbeHWvXqw1Jztx0TVoaBv5kTbddFOvvvzyy2O3Of/88+vVTrtCugB5o2XnT3/6U2xbdI3fUUcd5dXvv/9+6uetdCHrCRMmePXEiRNTP0+Iqjlbo0n6laS5zrnWE8x0SR9NPaMk3ZV9eygysoM0yA+SIjtIg/wgKbKDLFRz5Gx3SUdLesbMnipvmyjpIkm3mtlxkl6W9LXatIgCIztIg/wgKbKDNMgPkiI7SK3D4cw5N1OStfHl/bNtB42E7CAN8oOkyA7SID9IiuxPDVIRAAAgAElEQVQgC1WtOWs00ffGStKGG27o1ZXWh3XWLrvs4tU33XRT6sdEeP7whz94dZJFvIsWLfLq6BqiaD7RdY0ZMya27fDDD/fq6DpbAOisKVOmePV7770Xu83NN99cr3YQkOj17XbbbTevfuyxx1I/R6X1ZNE1Z42qU6fSBwAAAADUBsMZAAAAAASA4QwAAAAAAsBwBgAAAAAB6JInBKnkkEMO8eqQLmiIsL3yyite/c4773T6Mc4880yv3nvvvb16xIgRsfusv/76nX4eNKa3337bq19//XWvfvfdd+vZDoAGNG3atKq2oet59NFH826hoXDkDAAAAAACwHAGAAAAAAFgOAMAAACAALDmDOiE3r17x7Y1Nzd7dZL1ir/5zW+8evr06V79/vvvx+5T6WLEaHyf+ET8b2onnXRSu3Ulo0ePzqolAACQEY6cAQAAAEAAGM4AAAAAIAAMZwAAAAAQANacAZ0QvX6UJPXr18+rly1b1unHPfroo9v9+j777NPpx0Rjuvvuu/NuAQAA1AhHzgAAAAAgAAxnAAAAABAAhjMAAAAACABrzoCUli5dmncLAAAAaAAcOQMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAgAwxkAAAAABIDhDAAAAAACwHAGAAAAAAEw51z9nsxsmaRFkvpIWl63J06nKL2G0Ofmzrm+tXjgVtmRwnit1ShKn1L+vdYsOxL7nhrLu896ZUfK/7VWqyh9Svn3yr4nrii95t0n+564ovQp5d9rm/mp63D28ZOazXLONdf9iRMoSq9F6TMLRXmtRelTKlavaRTpdRal16L0mYWivNai9CkVq9c0ivQ6i9JrUfrMQlFea1H6lMLulbc1AgAAAEAAGM4AAAAAIAB5DWeTc3reJIrSa1H6zEJRXmtR+pSK1WsaRXqdRem1KH1moSivtSh9SsXqNY0ivc6i9FqUPrNQlNdalD6lgHvNZc0ZAAAAAMDH2xoBAAAAIAAMZwAAAAAQgLoPZ2Y21Mzmm9kLZja+3s/fHjObYmZLzWxOq229zexBM1tQ/twrzx7LPW1mZg+Z2Vwze9bMTgm11yyRnfTITnjZkchP6ELOD9kJG9nJBvkhP0kVMTt1Hc7MrJukqyR9SdIASSPNbEA9e+jAVElDI9vGS5rhnNtG0oxynbfVkk53zu0gaVdJJ5b/O4bYaybITmbITnjZkchPsAqQn6kiO0EiO5kiP+QnqeJlxzlXtw9Ju0m6v1U9QdKEevZQRY9Nkua0qudL6l/+d39J8/PusULPd0kaUoReyU7+fZKd8LJDfsL9KEJ+yE6YH2SH/JCf8L4fRchOvd/WuImkV1rVLeVtIdvYObdEksqf++Xcj8fMmiQNlvS4Au81JbKTMbITvKC/J+QnaEF/P8hO0IL/fpCfoAX9/ShKduo9nFmFbZzLPyEz6yHpd5JOdc69lXc/NUZ2MkR2yE4a5If8JEV2yE4a5If8JFWk7NR7OGuRtFmrelNJi+vcQ2e9Zmb9Jan8eWnO/UiSzGxNlUJ2o3Pu9vLmIHvNCNnJCNkpRHakQL8n5KcQ+Qny+0F2yE4a5If8JFW07NR7OPubpG3MbAsz6y5phKTpde6hs6ZLGlX+9yiV3quaKzMzSb+SNNc5d1mrLwXXa4bITgbITmGyIwX4PSE/hclPcN8PskN20iA/5CepQmYnh4V4B0t6XtI/JP0g70V3kd6mSVoi6X2V/mJxnKSNVDqLy4Ly594B9LmHSoe2n5b0VPnj4BB7JTthfT/ITnjZIT/hf4ScH7IT9gfZIT/kJ/c+C5cdKzcOAAAAAMhR3S9CDQAAAACIYzgDAAAAgAAwnAEAAABAABjOAAAAACAADGcAAAAAEACGs4yZ2Qgze8LMVprZq2Z2vZl9Ku++ED4z+7KZPW1mq8zsJTMbl3dPKCYz26S8D3Jm1iPvfhA2M1vDzMab2YLy/qfFzH6ad18In5k9XN7PVPrYLe/+EDZ+Z66MU+lnyMwOU+kidldJukNSf0k/lPSGpGbn3Ic5toeAmdnukh6RNEXSrZJ2kXS2pO855y7PszcUj5ndJGk/SRtLWt85tzLnlhAwM/uNpP0lTZI0T9JmkgY45ybm2hiCZ2YDJPWMbD5P0mBJ/Z1zq+vfFYqA35nbxnCWITO7WdI2zrnPt9r2UfgGOOfm5tYcgmZm90taxzm3V6ttl0kaLemTzrn/5NUbisXM9lRpn3OBpB+L4QztMLOhkn4v6XPOuefy7gfFZmbdJf1T0i3OuRPy7gfh4nfmtvG2xmytKenNyLZ/lT9bnXtBsewo6Y+RbQ9I6iWJt4agKmbWTdKVKv3lennO7aAYviHpTwxmyMhQlX5uTcu7EQSP35nbwHCWrSmS9jSzY8ysp5ltq9Ih2of4wYcOrC0penRsVfnzDnXuBcV1vEpZuirvRlAYu0h63sx+ZmZvmdm7ZnY76z6Q0AhJr6r0Nn2gPfzO3AaGsww55/6g0tvQJqv014D5krpJ+kqObaEYXpD0hci2ncufe9e5FxSQmW0k6XxJ45xz7+fdDwrjkyr93NpRpV+sj5X0eUl3mFmX/us1OsfM1pV0qEpvaWTNDNrF78xtYzjLkJntK+kaSVdI2lelH3S9Vfoh1y3P3hC8ayQNM7NvmVkvMztI0unlr32QY18ojh9Jetw5d0/ejaBQrPwxzDl3j3PuFklHq/THof1y7QxFc6ikHuItjagCvzO3bY28G2gwl0qa7pw746MNZvaUSme/Gibp9rwaQ/CmSPqcpKtV+ivSu5LOUGn90Gs59oUCMLPPqLR2aC8z27C8ed3y5w3M7APn3Hv5dIfArZD0onPu9VbbZqr0NusBkmbk0hWKaISkF5xzs/JuBIXA78xt4MhZtraX9FTrDc65+ZLek7RVLh2hEJxzHzjnxkrqK2mQSqdAf6z85cfavCNQso1Ki6sfVemX7RX677qzFpWGfKCSts6IZpK67Kms0TlmtoGkL4mjZqgevzO3gSNn2VokaafWG8xsB0nrSFqYR0MoFufcR79Yy8zGSPqrc25evl2hAGaq9LaQ1oaqdPT1YEkv1r0jFMXdkiaZWR/n3Edn+NxLpWH/7/m1hYL5P5LWEsMZqsfvzG3gOmcZMrNTJP20/HGvSkc/zpbUXdJnnXPv5NgeAmZmu0raQ6W/IvWUNFLSQZL2cM49nWdvKCYzGy3pOnGdM7TDzHpKmqPSGfYukLS+pIslzXPODcmzNxSHmd2n0jU5d8y7FxQDvzO3jSNn2fpfld6nf4JKp7T+l0p/0Z7QlUOGqrwv6XBJ56r0VqJHJO3unHsmz6YANDbn3Ftmtp9KP79uVuln2F2STsu1MRSGmfWRtL+ks/LuBYXC78xt4MgZAAAAAASAE4IAAAAAQAAYzgAAAAAgAAxnAAAAABAAhjMAAAAACADDGQAAAAAEgOEMAAAAAALAcAYAAAAAAWA4AwAAAIAAMJwBAAAAQAAYzgAAAAAgAAxnAAAAABAAhjMAAAAACADDGQAAAAAEgOEMAAAAAALAcAYAAAAAAWA4AwAAAIAAMJwBAAAAQAAYzgAAAAAgAAxnAAAAABAAhjMAAAAACADDGQAAAAAEgOEMAAAAAALAcAYAAAAAAWA4AwAAAIAAMJwBAAAAQAAYzgAAAAAgAAxnAAAAABAAhjMAAAAACADDGQAAAAAEgOEMAAAAAALAcAYAAAAAAUg1nJnZUDObb2YvmNn4rJpC10B+kBTZQRrkB0mRHaRBflANc84lu6NZN0nPSxoiqUXS3ySNdM4919Z9+vTp45qamhI9H8I3e/bs5c65vtXctrP5ITuNrZbZkchPI1u4cKGWL19u1d6efQ9aY9+DpNj3II329j1rpHjcnSW94Jx7UZLM7GZJwyS1uZNqamrSrFmzUjwlQmZmizpx807lh+w0tlpmRyI/jay5ubmzd2Hfg4+x70FS7HuQRnv7njRva9xE0iut6pbytuiTf9vMZpnZrGXLlqV4OjSYDvNDdtAG9j1Ig30PkmLfgzTY96AqaYazSodyY++RdM5Nds41O+ea+/at6p0D6Bo6zA/ZQRvY9yAN9j1Iin0P0mDfg6qkGc5aJG3Wqt5U0uJ07aALIT9IiuwgDfKDpMgO0iA/qEqa4exvkrYxsy3MrLukEZKmZ9MWugDyg6TIDtIgP0iK7CAN8oOqJD4hiHNutZmNlXS/pG6Spjjnns2sMzQ08oOkyA7SID9IiuwgDfKDaqU5W6Occ/dIuiejXtDFkB8kRXaQBvlBUmQHaZAfVCPVcAYAAAAAHzHzz31S6ZrK0dskkfRazaFLs+YMAAAAAJARhjMAAAAACADDGQAAAAAEgDVnAAAAADJRq7VgjbrGLIojZwAAAAAQAIYzAAAAAAgAwxkAAAAABIA1ZwA+9tRTT3n14MGDY7fpKu/5BgAA6VVzTTN+t/gvjpwBAAAAQAAYzgAAAAAgAAxnAAAAABAA1pxl6NVXX/XqN954w6sHDRrU6cfcc889vfqXv/xl5xurYOjQoV794osvZvK4KJY//vGPXv3Nb34zp04AAHm6+uqrvXrMmDEd3uf3v/+9Vx9yyCGZ9oTGxRqztnHkDAAAAAACwHAGAAAAAAFgOAMAAACAADCcAQAAAEAAOCFIQn//+99j2w4//HCvnj9/furneeSRR7x6u+22S/2YyNZZZ53l1TNnzvTq4cOHe/XYsWNr3lMlt912W2zbpEmTvHrRokX1ageBef3117361ltv9epKJwfo2bOnV6+xhv8jJfqYAMLxi1/8wquvv/76nDqpj8cee8yro6//uuuuq2c7XUqlk390dGHqrnzCEI6cAQAAAEAAGM4AAAAAIAAMZwAAAAAQANacJRR977IUX2P20EMPdfg45513Xrv3qeYxUD8TJkyIbbvhhhu8uqWlxat33HHHmvb0kcsuu8yrZ82a5dXz5s2L3WfOnDlevfnmm3v1hRdemFF3qKcFCxZ49fLlyzu8z5AhQ7x64MCBXr3rrrvG7nPFFVd49c4771xtiwjUqlWrvDq6llqS7rrrLq9ee+21vTq69vCdd96JPcbChQu9+tOf/nRn2kQG/vznP3v1ihUrvHrYsGFefe+999a8p6xU+h1t6tSp7dasOQtLpTVpXWUdGkfOAAAAACAADGcAAAAAEACGMwAAAAAIAGvOqvSPf/zDq0866aTYbR588EGv3meffTp83Gpug3Dcd999sW0HHHCAV0ezUi/Ra+LdeeedHd7ns5/9rFd///vf9+qRI0embwyZWrx4cWzbAw884NWTJ0/26kcffTR2n9GjR3v1uHHjvDq6HhZdw/HHH+/VldbubLnlll7929/+1qt32mknr957771jj7Hvvvt69T333OPVXNOz/vbbbz+vPuGEE7z64Ycfrksf0bVvb775Zqcf46qrropti64PR76i68c6uu5ZV8KRMwAAAAAIAMMZAAAAAASA4QwAAAAAAsBwBgAAAAAB4IQgVYqeOOGaa66J3SZ6Ygh0DdELdUYv7Bmyrbfe2quPPvronDpBWyZOnOjVL730Uuw2r776qlcfeuih7dZS5QuqA1GVTqwwfPjwTj1GNfvE6IWs//3vf3fqOdA4rr32Wq+++uqrO/0Yr7/+elbtIAOc7KNzOHIGAAAAAAFgOAMAAACAAHQ4nJnZFDNbamZzWm3rbWYPmtmC8udetW0TRUV+kBTZQRrkB0mRHaRBfpBWNWvOpkr6maTrW20bL2mGc+4iMxtfrs/Ivr1wbLjhhl79jW98I6dOCmeqCpyfadOmefWuu+4au010DdD8+fO9ul4XU91zzz29eu7cuV4d7asApqrA2clKdL3FL3/5y9httt12W68eOHBgTXsqiKkiP0hmqhosO8cee2xs24477ujVe+21V6cf94orrvDq3/3ud51+jKjZs2d79cKFCzv9GMccc0xsW/TC5zU0VQ2Wn85KssYselHqrqzDI2fOub9IeiOyeZikX5f//WtJX864LzQI8oOkyA7SID9IiuwgDfKDtJKuOdvYObdEksqf+2XXEroA8oOkyA7SID9IiuwgDfKDqtX8hCBm9m0zm2Vms5YtW1brp0MDITtIg/wgKbKDNMgPkiI7kJJf5+w1M+vvnFtiZv0lLW3rhs65yZImS1JzczNvKIVUZX5CyM4RRxzh1U8++WTsNpMmTfLqe++916vrteZs3LhxXt29e3evjq4NKKgut+9Za621vLqz15iCpzD7HgSn0PueqVOnxrbdeOONXh1dU/3MM890+nmefvppr37iiSc6/Rg77bSTV48ePbrTj1Hpmp377bdfpx8nQ+x7Ilhj1rakR86mSxpV/vcoSXdl0w66CPKDpMgO0iA/SIrsIA3yg6pVcyr9aZIelbSdmbWY2XGSLpI0xMwWSBpSroEY8oOkyA7SID9IiuwgDfKDtDp8W6NzbmQbX9o/417QgMgPkiI7SIP8ICmygzTID9JKuuasy1m1apVXP/zww7HbDBo0yKt79+5dy5aQg0WLFsW2rVy50qv79fNPwtS3b9+a9tSWsWPHevUHH3wQu02lHCMss2bN8uqDDz44dpt77rmnXu2gi7n11ltj26LXx9pqq61SPw/rT4rplFNO8eo5c+Z4dZI1Z1/72te8evz48Z1vDHXV0XXN+P+7c2p+tkYAAAAAQMcYzgAAAAAgAAxnAAAAABAAhjMAAAAACAAnBKnSySef7NX77rtv7DZf+cpXvLqpqcmrBw8eHLvPUUcdlb451M2Xv/zl2La1117bqy+99FKvHjNmTE17qhYLcospmq9KC+y/853veHU0gz169Mi+MTSkY445xqujF7eXpFNPPdWroxe433LLLTv9vB2dUABAODgBSG1x5AwAAAAAAsBwBgAAAAABYDgDAAAAgACw5qxKxx9/vFdPmjQpdpvbb7+93ccYMWJEbBtrzorvuuuu8+ro2kMgjT59+nh1pQtOjxw50qvPP/98rz7zzDNj91l//fUz6A6NJrqe+pprrondZvHixV694YYbdvp5Ro0a5dU33nhjpx8D4fmf//kfr37rrbdit7nwwgvbfYzf/va3Xr3WWmvFbnPaaacl6A61whqzbHHkDAAAAAACwHAGAAAAAAFgOAMAAACAALDmrA0vv/yyV++yyy5ePW3atNh9zjnnHK+eP39+9o0hV3feeWds2/777+/V3bt3r1c7ncJ1hBrDTjvtFNv285//3KvHjh3r1UceeWTsPoMGDcq2MbRpyZIlXj18+PBMHvfggw/26kprC9OK/uxLotLrnT59uldfe+21qZ8H+Rs4cKBXV1ovFv3/IWr27NlePWPGjNhtWHOGRsaRMwAAAAAIAMMZAAAAAASA4QwAAAAAAsBwBgAAAAAB6JInBHnjjTdi26IXEr7yyiu9OnoR6koXlK60DY1l8803j23r0aNHDp2gq5o3b15sW6UT1aB+3n//fa/efvvtvXr16tVeHT3hVFJz5szx6ujPsVBUer2PPfaYV6+55pr1agdtePvtt7162bJlXr3pppvG7rPuuuu2+5jbbrttbFtHOZ0wYYJXP/PMM+3eHrXFycTqjyNnAAAAABAAhjMAAAAACADDGQAAAAAEoEuuOVu6dGls23e/+12vvuqqq7x6zJgxNe0JxTBx4sTYtvPOO8+rm5ub69VOpzjn8m4BCaxYscKrK11Qesstt/TqK664wqu32GKL7BvDx3r16uXV0QvRZ7XGrLMuvfRSrz733HNz6ePZZ5+Nbdthhx28mnUt+bvrrru8esqUKV598803x+6z1VZb1bQnhKmr/z7R0f4q7X8fjpwBAAAAQAAYzgAAAAAgAAxnAAAAABCALrnmbMCAAbFtZ599tlezxgyS9JOf/MSro2s4JKmlpcWrQ1lzFl0/EF27JEmjR4+uUzdI6nOf+5xXNzU1xW4zefJkr46ugUJtrVy5Mu8WKjrnnHPardG1/eIXv/Dq//znP179zjvvePVmm20We4x11lkn+8aAjBR1LStHzgAAAAAgAAxnAAAAABAAhjMAAAAACECXXHO23XbbxbZNmjQph04QutNPP92rb7jhhthtomu7Pv3pT3v1TjvtlH1jVXj44Yc7vM2wYcNq3whSia4D+ctf/pJTJwAaSUf7koEDB9apEyAupPViHV23LOteOXIGAAAAAAFgOAMAAACAADCcAQAAAEAAOhzOzGwzM3vIzOaa2bNmdkp5e28ze9DMFpQ/c2EdeMgO0iA/SIrsIA3yg6TIDrJQzQlBVks63Tn3hJmtL2m2mT0oabSkGc65i8xsvKTxks6oXasooC6RnalTp3r1Bx984NWnnnqqV9fqBCGPPPKIV8+fP9+rK50IJ3BdIj+oCbKDNBouP5UuXr9s2TKv/sQn/L/Xf+pTn2r369X497//Hdv2z3/+s937vPnmm51+noA0XHYqCelkHe3p6EQeoT5Ph/+nOeeWOOeeKP/7bUlzJW0iaZikX5dv9mtJX860MxQe2UEa5AdJkR2kQX6QFNlBFjr1ZxAza5I0WNLjkjZ2zi2RSmGU1K+N+3zbzGaZ2azoX2nQdZAdpEF+kBTZQRrkB0mRHSRV9XBmZj0k/U7Sqc65t6q9n3NusnOu2TnX3Ldv3yQ9ouDIDtIgP0iK7CAN8oOkyA7SqOoi1Ga2pkohu9E5d3t582tm1t85t8TM+ktaWqsmUVyNlp3m5ubYtpaWFq/+zW9+49XvvfeeV5944onZNybprLPO8uro+/oPPvjgmjxvLTVafpJYtWqVV8+cOTN2mz322KNe7RQG2UEajZafhx56KLYtuh56/fXX9+qLL77Yq9dbb71OP+8zzzwT2/b1r3+93fusWLHCq4u2f2u07FRaT1WPNWf1Wi8WomrO1miSfiVprnPuslZfmi5pVPnfoyTdlX17KDKygzTID5IiO0iD/CApsoMsVHPkbHdJR0t6xsyeKm+bKOkiSbea2XGSXpb0tdq0iAIjO0iD/CApsoM0yA+SIjtIrcPhzDk3U1Jbxy/3z7YdNBKygzTID5IiO0iD/CApsoMsVLXmrNHMmzcvtu2II47w6muuucare/bsWdOePrLOOut4dXS9EvJ17bXXxratueaaXj1jxgyv/stf/uLVt912W/aNSerfv79XX3jhhV49atQooXiOOeYYrz7//PNjt7n//vvr1Q6AAqp0nbN1113Xq6Nrzj788EOvHjt2bOwxOvod5bXXXottW7hwYbv3+epXv+rV3/nOd9q9PeqvK68Hq4fOX1EQAAAAAJA5hjMAAAAACADDGQAAAAAEgOEMAAAAAALQJU8I8v3vfz+27ZJLLmn3PtGTLdTK6tWr6/I8yM7Pf/7zdr9++eWXe/Vpp51Wkz6uvvpqrx42bFhNngf1NXHiRK8ePHhw7Da33HKLVx9++OE17QlA8e29995evWjRIq++8847vXrVqlWdfo6NN944tm306NHt3mf48OFefcABB3T6eYEi48gZAAAAAASA4QwAAAAAAsBwBgAAAAAB6JJrzipdxDW61uuyyy6rVzuep59+OpfnRe2ceuqp7dZAe3r37u3VZ5xxRuw2s2bN8ur58+d79RFHHBG7z9Zbb51BdwCKKnpx5+i65Wj9yCOPxB5jgw02yL4xoIvjyBkAAAAABIDhDAAAAAACwHAGAAAAAAHokmvOunfvHtt26aWXtlsDQB7WXHNNrx4wYEDsNi0tLV49btw4r15nnXWybwxAQznhhBParQHUB0fOAAAAACAADGcAAAAAEACGMwAAAAAIQJdccwYARTVkyJCqtgEAgOLhyBkAAAAABIDhDAAAAAACwHAGAAAAAAFgOAMAAACAADCcAQAAAEAAGM4AAAAAIAAMZwAAAAAQAIYzAAAAAAiAOefq92RmyyQtktRH0vK6PXE6Rek1hD43d871rcUDt8qOFMZrrUZR+pTy77Vm2ZHY99RY3n3WKztS/q+1WkXpU8q/V/Y9cUXpNe8+2ffEFaVPKf9e28xPXYezj5/UbJZzrrnuT5xAUXotSp9ZKMprLUqfUrF6TaNIr7MovRalzywU5bUWpU+pWL2mUaTXWZRei9JnForyWovSpxR2r7ytEQAAAAACwHAGAAAAAAHIazibnNPzJlGUXovSZxaK8lqL0qdUrF7TKNLrLEqvRekzC0V5rUXpUypWr2kU6XUWpdei9JmForzWovQpBdxrLmvOAAAAAAA+3tYIAAAAAAFgOAMAAACAANR9ODOzoWY238xeMLPx9X7+9pjZFDNbamZzWm3rbWYPmtmC8udeefZY7mkzM3vIzOaa2bNmdkqovWaJ7KRHdsLLjkR+QhdyfshO2MhONsgP+UmqiNmp63BmZt0kXSXpS5IGSBppZgPq2UMHpkoaGtk2XtIM59w2kmaU67ytlnS6c24HSbtKOrH83zHEXjNBdjJDdsLLjkR+glWA/EwV2QkS2ckU+SE/SRUvO865un1I2k3S/a3qCZIm1LOHKnpskjSnVT1fUv/yv/tLmp93jxV6vkvSkCL0Snby75PshJcd8hPuRxHyQ3bC/CA75If8hPf9KEJ26v22xk0kvdKqbilvC9nGzrklklT+3C/nfjxm1iRpsKTHFXivKZGdjJGd4AX9PSE/QQv6+0F2ghb894P8BC3o70dRslPv4cwqbONc/gmZWQ9Jv5N0qnPurbz7qTGykyGyQ3bSID/kJymyQ3bSID/kJ6kiZafew1mLpM1a1ZtKWlznHjrrNTPrL0nlz0tz7keSZGZrqhSyG51zt5c3B9lrRshORshOIbIjBfo9IT+FyE+Q3w+yQ3bSID/kJ6miZafew9nfJG1jZluYWXdJIyRNr3MPnTVd0qjyv0ep9F7VXJmZSfqVpLnOuctafSm4XjNEdjJAdgqTHSnA7wn5KUx+gvt+kB2ykwb5IT9JFTI7OSzEO1jS85L+IaKOOBcAACAASURBVOkHeS+6i/Q2TdISSe+r9BeL4yRtpNJZXBaUP/cOoM89VDq0/bSkp8ofB4fYK9kJ6/tBdsLLDvkJ/yPk/JCdsD/IDvkhP7n3WbjsWLlxAAAAAECO6n4RagAAAABAHMMZAAAAAASA4QwAAAAAAsBwBgAAAAABYDgDAAAAgAAwnGXMzL5sZk+b2Soze8nMxuXdE4rBzEaY2RNmttLMXjWz683sU3n3hWJg34OkzOxrZja9vN9ZaWazzWxk3n2hGPjZhaTMbA0zG29mC8o/u1rM7Kd595U3hrMMmdnukm6X9P8kHSppiqSLzezUXBtD8MzsMJWuGfJXScMknSFpL0l3mxn/n6Jd7HuQ0jhJKyWdJukwSQ9JusnMTsq1KwSPn11I6TpJJ0v6iaQDJY2X9F6uHQWA65xlyMzul7SOc26vVtsukzRa0iedc//JqzeEzcxulrSNc+7zrbYdptIV6wc45+bm1hyCx74HaZhZH+fc8si2myTt5pzbIqe2UAD87EJSZjZU0u8lfc4591ze/YSEv2pka0dJf4xse0BSL0m71b8dFMiakt6MbPtX+bPVuRcUD/seJBYdzMqelNSv3r2gcPjZhaS+IelPDGZxDGfZWltS9C/Uq8qfd6hzLyiWKZL2NLNjzKynmW0r6YeSHmLHhSqw70HWviiJfQ86ws8uJLWLpOfN7Gdm9paZvWtmt7NekeEsay9I+kJk287lz73r3AsKxDn3B5XegjZZpb9CzpfUTdJXcmwLxcG+B5kxs/1VWj90Vd69IGz87EIKn1QpOztKGiHpWEmfl3SHmXXpo64MZ9m6RtIwM/uWmfUys4MknV7+2gc59oXAmdm+KuXnCkn7qrSj6q3STqpbnr2hENj3IBNm1iTpJkl3Oeem5toMgsfPLqRg5Y9hzrl7nHO3SDpapT8s7pdrZznjhCAZKu+IrpB0vEp/OXpXpTMXXSnpWH7QoS1m9oSkuc65I1tt207SPEnDnXO359Ycgse+B1kws96S/q9KZ27cxzn3Ts4tIXD87EJSZvaapBedc7u12vYJlc7W+F3n3JW5NZczjpxlyDn3gXNurKS+kgZJ2ljSY+UvP9bmHQFpe0lPtd7gnJuv0k5qq1w6QmGw70FaZraupLsldZf0PwxmqBI/u5BUW2fyNEkf1rOR0DCc1YBzboVz7hnn3EpJYyT91Tk3L+++ELRFknZqvcHMdpC0jqSFeTSE4mHfgyTMbA1Jv5W0jaQvOeeW5twSioOfXUjqbkmDzKxPq217qXQG0L/n01IYeFtjhsxsV0l7qPRXpJ6SRko6SNIezrmn8+wNYTOzUyT9tPxxr0pHPs5W6a/Yn+Wv2GgP+x6kYWaTJX1L0ikqXci8tSedc6vi9wL42YXkzKynpDmSXpV0gaT1JV0saZ5zbkieveWN4SxDZvZ5lRbG7qDSIdlHJI13zj2Ta2MIXvnMRMdLOkGlt4L8S9JMSROccy/m2RvCx74HaZjZQkmbt/HlLZxzC+vXDYqEn11Iw8y2lvS/kvZW6XIwd0k6zTm3ItfGcsZwBgAAAAABYM0ZAAAAAASA4QwAAAAAAsBwBgAAAAABYDgDAAAAgAAwnAEAAABAABjOAAAAACAADGcAAAAAEACGMwAAAAAIAMMZAAAAAASA4QwAAAAAAsBwBgAAAAABYDgDAAAAgAAwnAEAAABAABjOAAAAACAADGcAAAAAEACGMwAAAAAIAMMZAAAAAASA4QwAAAAAAsBwBgAAAAABYDgDAAAAgAAwnAEAAABAABjOAAAAACAADGcAAAAAEACGMwAAAAAIAMMZAAAAAASA4QwAAAAAAsBwBgAAAAABYDgDAAAAgAAwnAEAAABAABjOAAAAACAADGcAAAAAEIBUw5mZDTWz+Wb2gpmNz6opdA3kB0mRHaRBfpAU2UEa5AfVMOdcsjuadZP0vKQhklok/U3SSOfcc23dp0+fPq6pqSnR8yF8s2fPXu6c61vNbTubH7LT2GqZHYn8NLKFCxdq+fLlVu3t2fegNfY9SIp9D9Job9+zRorH3VnSC865FyXJzG6WNExSmzuppqYmzZo1K8VTImRmtqgTN+9UfshOY6tldiTy08iam5s7exf2PfgY+x4kxb4HabS370nztsZNJL3Sqm4pb4s++bfNbJaZzVq2bFmKp0OD6TA/ZAdtYN+DNNj3ICn2PUiDfQ+qkmY4q3QoN/YeSefcZOdcs3OuuW/fqt45gK6hw/yQHbSBfQ/SYN+DpNj3IA32PahKmuGsRdJmrepNJS1O1w66EPKDpMgO0iA/SIrsIA3yg6qkGc7+JmkbM9vCzLpLGiFpejZtoQsgP0iK7CAN8oOkyA7SID+oSuITgjjnVpvZWEn3S+omaYpz7tnMOkNDIz9IiuwgDfKDpMgO0iA/qFaaszXKOXePpHsy6gVdDPlBUmQHaZAfJEV2kAb5QTVSXYQaAAAAAJANhjMAAAAACADDGQAAAAAEgOEMAAAAAALAcAYAAAAAAWA4AwAAAIAAMJwBAAAAQAAYzgAAAAAgAAxnAAAAABCANfJuIA933313bNuhhx6a+nGjjzF9+v9v796jpKjOvY//Ho0xiKAQQVCJsOQmKzeBgCZwAq8LxRPUgOCFeDAXNV5e8QieKIS4EpdCMIHgqycmGhWNchGi0WgUEdFEiRpzQLzgiHoUuYkuJIqggOzzx/TxnV3VzPRUdXft6vl+1po1/Gqqqp9mHnpmU71r35f6nAAAAEAWDjnkkNi2DRs2NHrMn/70Jy+PGDGirDXVOq6cAQAAAEAAGJwBAAAAQAAYnAEAAABAAGpyztnUqVO9/Oyzz3q52Htl27Vr5+Vf/vKXjX59/vz5sXM8/vjjXv7+97/v5RkzZjT5uAAAAEConnzySS9PmDAho0p8Tz/9tJcvueSS2D7Lli2rVjmJceUMAAAAAALA4AwAAAAAAsDgDAAAAAACwOAMAAAAAAKQ+xuCTJs2Lbbt97//vZc7dOjg5aFDh8aOOemkk7w8ZswYL7dp08bLbdu2jZ1j9+7dXl64cKGXO3XqFDsmetOQ7t27x/YBgFqxY8eORr8efa39+OOPK1kOgBItX748tq1v374ZVBK3atUqLxera9u2bdUqJ9e6devm5eiC0pL05S9/2ct//OMfvRy9QUix1/1Ro0YlLfFTS5Ys8fK1117r5egNAiVp4MCBXo7eRCQEXDkDAAAAgAAwOAMAAACAADA4AwAAAIAA5H7O2eTJk2PbvvKVr3j57LPP9vK4ceNSP+6xxx4b29a6dWsvb9y40cszZ86MHfP1r3/dy8w5C0uvXr28/MorrzR5THQB84kTJ5a1pj1pqtY5c+bEjjnjjDMqWhNati1btsS2RV+f16xZU61yUMTcuXO9PHbs2IwqSa9nz55erqury6iS2rTvvvvGth122GFeXrt2rZcPPPDA2DEdO3Ysb2GSTj75ZC936dKl7I/RUrzxxhtePuSQQ2L77LPPPl6O3lNh69atXv7www/LU1zE9u3bvfzee+95uVivvfXWWxWppZy4cgYAAAAAAWBwBgAAAAABYHAGAAAAAAHI/ZyzYq666iovjxgxoiqPe/TRR3t5/vz5Xu7Xr19V6kByRx55pJej772+7bbbYsfcddddlSypbIrNfbvyyisr/rjR9Wdakujah6XMr4rO6+jcuXOzH/f999/38ubNm5s8pn379l6OruW4adOm2DFNrRs0fPjw2Lbo38EXvvAFL++1F/9nWC7Tp0/38i233BLbJ9orwJ5E/61K0s9+9jMvR9eZHTlyZOyY8ePHl7cwBOeaa67x8q9//evYPtGfdaeeempFa8oTfgoCAAAAQAAYnAEAAABAABicAQAAAEAAGJwBAAAAQABq8oYgQFIvv/xyo18vNiE6euOErMyePdvL559/vpefe+652DEbNmyoZEktjnPOy8uWLfPy4MGDY8d87nOf8/Lo0aO9HJ1gX4rbb7/dyxdddFGTx0yaNMnL0QXKi91QZvHixY2es0ePHrFtX/rSl7z8l7/8xcvFFq1FcZdccomXV6xY4eXXXnvNy0kWXy22mG+0v7ISfb6TJ0/28tChQ2PHLF26tKI11ZLoDX/uuOOO2D7RRX/5+6090Z9rpejdu7eX//nPf8b22bhxY+Kaah1XzgAAAAAgAAzOAAAAACAATQ7OzOwWM9tkZi802NbezBab2erC53aVLRN5Rf8gKXoHadA/SIreQRr0D9IqZc7ZbEnXS2r4JvPLJS1xzv3czC4v5MvKX16+RedOXHvttbF9Hn30US/v2rXLy9/+9rfLX1h1zRb9UxXHHHOMl9u181/7o3PQpPjC6YGZrZz1zs6dO70cnfNS7O+7e/fuXo4u1vnSSy95uU+fPk3W0alTpyYfNzr/aNq0aY3mXr16xc7RVP8Um6NyxBFHNHpMGc1WzvqnMRMmTIhtmzt3rpfffvttL3/nO9/x8re+9a1mP26bNm1i24YMGdLs81RC9N/CwQcfXK5Tz1YN9U5SO3bs8PL27dtj+0TnqkJSzvtn4cKFXr7ssniZU6dO9XKxualIrskrZ865v0jaHNl8sqTbCn++TVLuRxCoDPoHSdE7SIP+QVL0DtKgf5BW0jlnBzvnNkhS4XPHPe1oZuea2bNm9uw777yT8OFQY0rqH3oHRfDagzR47UFSvPYgDV57ULKK3xDEOXejc66/c65/hw4dKv1wqCH0DtKgf5AUvYM06B8kRe9ASr7O2dtm1tk5t8HMOkvaVM6imuPCCy+MbXvxxRe93LGj/x8UAwYMqGhN/6tVq1ZeHjNmTGyf6BpGBxxwgJdrYM5ZMcH0T3P97ne/i21buXKll/v161etcpql2Bpb0bWsciDo3lmyZImXo/9+FyxY0OQ5/va3v3n5pz/9qZcXLVrU5Dmia6VFsyTdeuutjT5OdP2+G264IXaOQYMGNVlLYILun8Y88MADsW2jRo3ycnQ+3ze/+U0v9+/fv/yFZSj6s73Cr2e57Z1SRefMXn/99V6eMmVKNcupNbnpn1NOOcXLxdbJHD9+vJevueYaLxdb47IaXnnlFS9He1iSZs2aVa1yEkt65ew+SWcV/nyWpHvLUw5aCPoHSdE7SIP+QVL0DtKgf1CyUm6lP1fS3yT1MrO1ZvYDST+XNMzMVksaVshADP2DpOgdpEH/ICl6B2nQP0irybc1Ouf29D6BY8tcC2oQ/YOk6B2kQf8gKXoHadA/SCvpnLNgFHs/aXQ9mOeee87L1ZpzhvyJzmG88847vbxt27bYMdE1f4466qiy14V8iM4x+/jjjzOqpGnf+973vLx8+XIv9+3b18s5nF+WazNmzPDyyJEjY/ucc845Xq7iGnKoQdE5ZzfddJOXmXPWMl199dWxbR9++KGX999//2qV06hNm/ypfI899lhsn+haoiGq+N0aAQAAAABNY3AGAAAAAAFgcAYAAAAAAWBwBgAAAAAByP0NQYqZOXNm1iWUzDnXaEZ1RW8ws3jxYi9HF16U4jcEQctlZs0+5r333vPy3//+dy8PGzYsVU2l4rUnLJdeeqmX6+rqYvtwAxAAlRa9eVSe7LVXPq9B5bNqAAAAAKgxDM4AAAAAIAAMzgAAAAAgADU55ywU7777rpfHjRsX22f06NFePv744ytaE4DKSTJva926dV6eN2+el5ctW5aqJgBIYv369V4uthD6PffcU61ygBaDK2cAAAAAEAAGZwAAAAAQAAZnAAAAABAA5pxV0M6dO728YsWK2D4XXHCBlw899NCK1gSgerZv3+7lWbNmxfb55JNPvFxsbmo1HHfccV7mtSgsb7zxRupztGvXLratQ4cOqc+L2rDffvt5+fnnn/dynz59Yse0bdu20XOec845sW0zZsxIUB1CMmfOHC9v3rzZyyNGjKhmOZ/q1KmTl8eOHZtJHWlx5QwAAAAAAsDgDAAAAAACwOAMAAAAAALA4AwAAAAAAsANQcpo7dq1Xn7wwQe9XGyC5GGHHVbRmgCELTqB+eyzz86kjhNPPDGTx0Vxbdq08fLw4cNj+zR30fMJEybEtnFzBuxJ7969vbx79+7YPo888oiXhw0b5uXrrrsudsxBBx3k5UmTJiUtERl54YUXvHzzzTd7uWvXrlWs5v/r3r27lydPnpxJHWlx5QwAAAAAAsDgDAAAAAACwOAMAAAAAALAnLOE3n333di2+++/38tXXnmll//xj3/EjuncuXN5CyvR1q1bvRydLxd9rznyZ8GCBV5u3bp1RpW0XK1atfIycytQqnXr1nm5X79+sX1Wr15drXLQAu3atcvLr7/+emyf6OLDPXv29PKZZ54ZO4bXwfybOnWql/fay7/WE+2VN998M3aOYq9pTVmzZo2Xly1b1uxz5AFXzgAAAAAgAAzOAAAAACAADM4AAAAAIADMOUvopptuim2LzjFr165dtcpptieeeMLLp556qpfff//9apaDMhgzZoyXo+/r79u3bzXLaZHMLOsSErvjjju8HF2nZtCgQVWsBtF1zh5++OHYPqeddpqXn3nmGS9fdtllXp44cWKZqkNLsGHDBi8XmyN07rnnermurq6iNSFMV111lZdPOukkLz/++OOxY0aNGtXsx4meZ9q0aV6ulZ9TXDkDAAAAgAAwOAMAAACAADA4AwAAAIAAMOesRFOmTPHynDlzYvt89atf9fIvfvELL7dv377Zj/vUU095+cc//nGzz1FMjx49vHzPPfeU5by17oorrohtu/rqq708ePDgapXjWblypZeZN1h9zrmsS0jsi1/8opdDnjPbEkXnAErSfvvt5+Xoz6nofKAOHTqUvS7Ujk8++cTLixcv9vJFF10UOya63hVQLtE5j5K0ZcsWL9fqmrxcOQMAAACAADA4AwAAAIAANDk4M7MuZrbUzFaZ2YtmdnFhe3szW2xmqwufeQ8MPPQO0qB/kBS9gzToHyRF76AcSrlytkvSROfckZKOlnShmfWRdLmkJc65HpKWFDLQEL2DNOgfJEXvIA36B0nRO0ityRuCOOc2SNpQ+PMHZrZK0qGSTpY0pLDbbZIek3RZkVPkwgcffODl66+/3svz58/3crdu3WLnGDJkiJf//Oc/N5pL8eqrr3r50UcfbfY5imndurWXjz322LKct6Fa6J2f/OQnXo4uNC5Ja9eurVY5LUot9E/Ibr31Vi9Hb1gUvcFRntRi70yYMCG27ZRTTvHyiSee6OUuXbpUtKZaVYv9U4pf/epXXv7oo4+8zM0/mtZSe+f111/3cik3JFu/fr2Xn3/+eS/feeedsWOiNwmJ3nhv+vTpTT5uHjRrzpmZdZV0lKSnJR1caML/bcaO5S4OtYPeQRr0D5Kid5AG/YOk6B0kVfLgzMz2l/QHSf/unCv5Ht1mdq6ZPWtmz77zzjtJakTO0TtIg/5BUvQO0qB/kBS9gzRKGpyZ2T6qb7I7nXN3Fza/bWadC1/vLGlTsWOdczc65/o75/qzxkrLQ+8gDfoHSdE7SIP+QVL0DtJqcs6ZmZmkmyWtcs7NbPCl+ySdJennhc/3VqTCKtm6dauXr7vuOi9H3+far1+/2Dm2b9/u5WnTpqWuq1OnTl4eOXJk6nNKUv/+/ctynsbUQu+ceeaZXr755ptj+9x7r1/+EUcc4eUBAwakrmPHjh2xbdEe3bx5c+rHCUke++eRRx7x8gknnODlBx98sJrlfOr222+PbVu3bp2Xo7XmWR57pykPPPBAbNtvf/tbLx9++OHVKqem1WL/lOKKK67w8rZt2zKqJL9qsXfeeuut2LboguWnnXaal9esWePlrl27xs5xww03eDn6enbggQfGjon+DtyqVSsvt2vn3wQz+nNOkg499NDYttA0OTiT9A1J/ybpeTNbUdg2WfUNdpeZ/UDSGkljKlMicozeQRr0D5Kid5AG/YOk6B2kVsrdGp+QZHv4cvlv8YeaQe8gDfoHSdE7SIP+QVL0DsqhWXdrBAAAAABURilva0QR0XXPKuVrX/ual+++++497IlqOO6442LbbrzxRi/PmjXLy9H1iUqZ7xddI2TevHmxfaZMmeLl4cOHezk6XxGVN3DgQC8/88wzmdQxd+5cL69evTq2z/jx473M5HMAQFSx+cjR+zBEzZkzx8vHH398bJ/Jkyd7OTrH7Ec/+lGpJX5q33339XKxNXyffPLJRs/x+c9/vtmPW25cOQMAAACAADA4AwAAAIAAMDgDAAAAgAAwOAMAAACAAHBDkILOnTt7ef369RlVgpBNmjQpts055+XoQoq7du3y8g9/+MMmH2ft2rXNPmb69Ole7tmzZ5PHoLK2bNni5cGDB8f2+etf/5r6cRYuXOjlCy+80MvnnXde7BhuAIJSfPDBB7FtDz30kJejC8N+5jP8aoE9mz17tpcPOOAAL0f7Cdl64YUXKnLeqVOnVuS8DdXV1cW2HXTQQY0eE/2dLgtcOQMAAACAADA4AwAAAIAAMDgDAAAAgADwxnAgpehCitH5FjNmzPDyggULYudo1aqVlw8//HAv9+7du8k6PvvZzza5Dyor+j14+eWXvdyrV6/YMdEFLwcMGODl6EKco0ePjp1jx44dXj799NO9HF2wHLVhxYoVFX+M6PxXSbrgggu8PHz4cC+3adOmojWhMj766CMvDxkyxMuPPfZYk+e4//77vTxx4sTYPmPHjvXyxRdfXFqBQBMGDRrk5RDmjyXBlTMAAAAACACDMwAAAAAIAIMzAAAAAAgAc86AMovOEdp77729fOmll8aOGThwoJeXLl1a/sJQdT169PBysTVXouvRLVq0yMtLlizx8ogRI2LnmDdvnpf32sv/fzfWncq/bt26xbZF57tu37697I9bbC5rnz59vBztN+TTxo0bvdypUycvt23btslz7Ny508tnnXVWbJ/oeqHMlwZ8vKICAAAAQAAYnAEAAABAABicAQAAAEAAmIgAVFh0nZdi676gZYjOQZPyuw4Lquuhhx6KbRs6dKiXS1mHqrm6du0a27Z8+fKyPw6y17FjRy/v3r07o0qAlo0rZwAAAAAQAAZnAAAAABAABmcAAAAAEAAGZwAAAAAQAG4IAgBADrFYPQDUHq6cAQAAAEAAGJwBAAAAQAAYnAEAAABAAKyaC6Ca2TuS3pR0kKR3q/bA6eSl1hDqPNw516ESJ27QO1IYz7UUealTyr7WivWOxGtPhWVdZ7V6R8r+uZYqL3VK2dfKa09cXmrNuk5ee+LyUqeUfa177J+qDs4+fVCzZ51z/av+wAnkpda81FkOeXmuealTyletaeTpeeal1rzUWQ55ea55qVPKV61p5Ol55qXWvNRZDnl5rnmpUwq7Vt7WCAAAAAABYHAGAAAAAAHIanB2Y0aPm0Reas1LneWQl+ealzqlfNWaRp6eZ15qzUud5ZCX55qXOqV81ZpGnp5nXmrNS53lkJfnmpc6pYBrzWTOGQAAAADAx9saAQAAACAADM4AAAAAIABVH5yZ2XAzqzOzV83s8mo/fmPM7BYz22RmLzTY1t7MFpvZ6sLndlnWWKipi5ktNbNVZvaimV0caq3lRO+kR++E1zsS/RO6kPuH3gkbvVMe9A/9k1Qee6eqgzMz21vSf0o6QVIfSWeYWZ9q1tCE2ZKGR7ZdLmmJc66HpCWFnLVdkiY6546UdLSkCwt/jyHWWhb0TtnQO+H1jkT/BCsH/TNb9E6Q6J2yon/on6Ty1zvOuap9SDpG0qIGeZKkSdWsoYQau0p6oUGuk9S58OfOkuqyrrFIzfdKGpaHWumd7Oukd8LrHfon3I889A+9E+YHvUP/0D/hfT/y0DvVflvjoZLeapDXFraF7GDn3AZJKnzumHE9HjPrKukoSU8r8FpTonfKjN4JXtDfE/onaEF/P+idoAX//aB/ghb09yMvvVPtwZkV2ca9/BMys/0l/UHSvzvn3s+6ngqjd8qI3qF30qB/6J+k6B16Jw36h/5JKk+9U+3B2VpJXRrkwyStr3INzfW2mXWWpMLnTRnXI0kys31U32R3OufuLmwOstYyoXfKhN7JRe9IgX5P6J9c9E+Q3w96h95Jg/6hf5LKW+9Ue3D2d0k9zKybmX1W0umS7qtyDc11n6SzCn8+S/XvVc2UmZmkmyWtcs7NbPCl4GotI3qnDOid3PSOFOD3hP7JTf8E9/2gd+idNOgf+iepXPZOBhPx/lXSK5Jek/TjrCfdRWqbK2mDpJ2q/x+LH0j6vOrv4rK68Ll9AHUOUv2l7ZWSVhQ+/jXEWumdsL4f9E54vUP/hP8Rcv/QO2F/0Dv0D/2TeZ256x0rFA4AAAAAyFDVF6EGAAAAAMQxOAMAAACAADA4AwAAAIAAMDgDAAAAgAAwOAMAAACAADA4KyMze8zM3B4+jsm6PoTNzE43s/8ys61mts7MbjezQ7KuC/lA/yApM/u2ma00s4/N7L/NbELWNSE/6B8kZWbf3cPvzOdlXVuWPpN1ATXmAkltI9uulHSU6hcTBIoys5NUv2bIf0r6D0mdJV0l6X4z6++c251lfQgb/YOkzOwbku6WdIukSyUNlDTdzHY752ZlWhyCR/+gTP6PpO0N8utZFRIC1jmroMKK7hslzXfOnZ91PQiXmc2T1MM516/BtpNUv2J9H+fcqsyKQ/DoHyRlZosktXLO/UuDbTMlfVdSJ+fcjqxqQ/joH6RhZt+VdKukNs65rRmXEwze1lhZwyW1U/3/aAON2UfSPyPbthQ+W5VrQf7QP0jqq5IeiWx7WPU/u3g7PppC/wBlxuCssk6XtE7SX7MuBMG7RdJgMxtnZm3NrKfq35a21Dn3Usa1IXz0D5L6nKTo1Y2PC5+PrHItyB/6B+XwmpntMrM6M/th1sVkjcFZhZjZfpJOVP1bGnnvKBrlnHtA9W8DuVH1V0DqJO0taVSGZSEn6B+k8Kqkr0W2DSh8bl/lWpA/uarFbwAAAkJJREFU9A/S2CDpJ5L+TfW/Mz8t6TdmdkmmVWWMwVnlnChpf/GWRpTAzIZK+o2kayUNVf1V1/aS7jGzvbOsDeGjf5DCbySdbGbnmFk7Mzte0sTC1z7JsC7kA/2DxJxzi5xzVznnHnbOPeicGyfpLklTzKzFjlG4IUiFmNk9kr7onOuRdS0In5n9l6RVzrnvNNjWS9LLkk5xzt2dWXEIHv2DpAqD92slnaf6q63bJF0m6TpJ33POzc6uOoSO/kG5mdkY1Q/QjnDOtci7NrbYUWklmdkBkk4QV81Qut6SVjTc4JyrU/2tZY/IpCLkCf2DRJxznzjn/q+kDpK+LOlgSU8VvvzUHg8ERP+golrs1SPWOauMkZL2FYMzlO5NSX0bbjCzIyW1kvRGFgUhV+gfpOKce0/Se5JkZhdIWuaceznbqpAX9A/K6BRJ76r+51qLxOCsMk6X9BxrC6EZfiPpV2a2XtKDqv/fxytU/4v1nzOsC/lA/yARMzta0iDVX3ltK+kMSccXtgGNon+Qhpn9QdIzklaq/m2xpxU+xjvndmdZW5YYnJWZmR0k6VjV330GKNX/U/3tiM9X/Xv3t0h6QtIk59yHWRaGXKB/kNRO1f8y9FNJu1W/9Ms3nHPPZ1kUcoP+QRp1kr4vqYvq1+R8SdI459zvM60qY9wQBAAAAAACwA1BAAAAACAADM4AAAAAIAAMzgAAAAAgAAzOAAAAACAADM4AAAAAIAAMzgAAAAAgAAzOAAAAACAADM4AAAAAIAD/A5ajnLAExhbyAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x1440 with 36 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,20))\n",
    "random_inds = np.random.choice(N_train_data,36)\n",
    "\n",
    "for i in range(36):\n",
    "    plt.subplot(6,6,i+1)\n",
    "\n",
    "    image_ind = random_inds[i]\n",
    "    digit = train_images[image_ind, :, :, 0]\n",
    "    plt.imshow(digit, cmap = plt.cm.binary)\n",
    "    plt.xlabel(train_labels[image_ind], fontsize = 15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "f78c21af",
   "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>Classifiers</th>\n",
       "      <th>Accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LDA</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Shallow_NN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Deep_NN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CNN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>LeNet</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Classifiers  Accuracy\n",
       "0         LDA       0.0\n",
       "1         SVM       0.0\n",
       "2  Shallow_NN       0.0\n",
       "3     Deep_NN       0.0\n",
       "4         CNN       0.0\n",
       "5       LeNet       0.0"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# store socres from each model\n",
    "score = pd.DataFrame({\n",
    "    \"Classifiers\":[\"LDA\", \"SVM\",\"Shallow_NN\", \"Deep_NN\", \"CNN\", \"LeNet\"],\n",
    "    \"Accuracy\":np.zeros(6)\n",
    "    })\n",
    "score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b647317e",
   "metadata": {},
   "source": [
    "### LDA and SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0f97e569",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21.4 s ± 5.06 s per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "clf_lda = LDA()\n",
    "%timeit clf_lda.fit(train_data, train_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "fb3e29f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score on trained data: 0.9351403508771929\n",
      "Score on test data: 0.9206666666666666\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.9206666666666666]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores = []\n",
    "print(f'Score on trained data: {clf_lda.score(train_data, train_labels)}')\n",
    "print(f'Score on test data: {clf_lda.score(test_data, test_labels)}')\n",
    "scores.append(clf_lda.score(test_data, test_labels))\n",
    "scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "53e457af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9206666666666666"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d19c77be",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since most of the data is empty (0's) let's use LDA to reduce the data space\n",
    "test_data_reduced = clf_lda.transform(test_data)\n",
    "train_data_reduced = clf_lda.transform(train_data)\n",
    "predict_lda = clf_lda.predict(test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "fb7e1346",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x2c95a774f70>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix for LDA classifier\n",
    "cm = confusion_matrix(test_labels, predict_lda, labels=np.unique(labels))\n",
    "disp = ConfusionMatrixDisplay(confusion_matrix=cm,\n",
    "                               display_labels=np.unique(labels))\n",
    "disp.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "10f21075",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.9206666666666666, 0.9553333333333334]"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svm = svm.LinearSVC(C = 0.01,\n",
    "                penalty='l2',\n",
    "                loss='squared_hinge',\n",
    "                tol=0.0001,\n",
    "                multi_class='ovr')\n",
    "\n",
    "clf_svm.fit(train_data, train_labels) \n",
    "      \n",
    "scores.append(clf_svm.score(test_data, test_labels))\n",
    "clf_svm.score(test_data, test_labels)\n",
    "scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "04b24cda",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9253333333333333"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svm_reduced = svm.LinearSVC(C = 0.01,\n",
    "                penalty='l2',\n",
    "                loss='squared_hinge',\n",
    "                tol=0.0001,\n",
    "                multi_class='ovr')\n",
    "clf_svm_reduced.fit(train_data_reduced, train_labels)\n",
    "clf_svm_reduced.score(test_data_reduced, test_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "15de88a4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x2c929dd8dc0>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix for linear svc\n",
    "predict_svm = clf_svm.predict(test_data)\n",
    "cm = confusion_matrix(test_labels, predict_svm, labels=np.unique(labels))\n",
    "disp = ConfusionMatrixDisplay(confusion_matrix=cm,\n",
    "                               display_labels=np.unique(labels))\n",
    "disp.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0f1d2fe",
   "metadata": {},
   "source": [
    "## Deeplearning (NN) with different architectures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e5997365",
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model():                                                 #layers as arguments\n",
    "    nn_model = tf.keras.Sequential([\n",
    "        tf.keras.layers.Dense(512, activation = 'relu'),\n",
    "        tf.keras.layers.Dense(128, activation = 'relu'),\n",
    "        \n",
    "        tf.keras.layers.Dropout(0.2),\n",
    "        tf.keras.layers.Dense(10, activation = 'sigmoid')\n",
    "    ])\n",
    "    return nn_model\n",
    "\n",
    "model = build_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ac13bbc6",
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext tensorboard\n",
    "import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "e7fbf010",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer = tf.keras.optimizers.SGD(learning_rate = 1e-1),\n",
    "    loss = 'sparse_categorical_crossentropy',\n",
    "    metrics = ['accuracy'])\n",
    "\n",
    "log_dir = \"logs/fit/\" + datetime.datetime.now().strftime(\"%Y%m%d-%H$M$S\")\n",
    "tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "e9ec006e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1782/1782 [==============================] - 46s 20ms/step - loss: 0.1864 - accuracy: 0.9468\n",
      "Epoch 2/5\n",
      "1782/1782 [==============================] - 36s 20ms/step - loss: 0.0796 - accuracy: 0.9770\n",
      "Epoch 3/5\n",
      "1782/1782 [==============================] - 31s 17ms/step - loss: 0.0536 - accuracy: 0.9838\n",
      "Epoch 4/5\n",
      "1782/1782 [==============================] - 35s 20ms/step - loss: 0.0399 - accuracy: 0.9886\n",
      "Epoch 5/5\n",
      "1782/1782 [==============================] - 30s 17ms/step - loss: 0.0295 - accuracy: 0.9916\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2c929e86580>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_size = 32\n",
    "epochs = 5\n",
    "\n",
    "model.fit(train_data, \n",
    "          train_labels, \n",
    "          batch_size = batch_size, \n",
    "          epochs = epochs,\n",
    "          callbacks = [tensorboard_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "1c195bb9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "      <iframe id=\"tensorboard-frame-d4a79b7ad5508f8c\" width=\"100%\" height=\"800\" frameborder=\"0\">\n",
       "      </iframe>\n",
       "      <script>\n",
       "        (function() {\n",
       "          const frame = document.getElementById(\"tensorboard-frame-d4a79b7ad5508f8c\");\n",
       "          const url = new URL(\"/\", window.location);\n",
       "          const port = 6006;\n",
       "          if (port) {\n",
       "            url.port = port;\n",
       "          }\n",
       "          frame.src = url;\n",
       "        })();\n",
       "      </script>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%tensorboard --logdir logs/fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "4a51eef1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 - 1s - loss: 0.0577 - accuracy: 0.9793\n",
      "Test accuracy: 0.9793333411216736 0.057670895010232925\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_acc = model.evaluate(test_data, test_labels, verbose=2)\n",
    "print('Test accuracy:', test_acc, test_loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "2bb49b27",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 62 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "prediction_nn = np.argmax(model.predict(test_data), axis = -1)\n",
    "print_misclssified(prediction_nn, test_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "fe6b8aa0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x2c9deb2d280>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm = confusion_matrix(test_labels, prediction_nn, labels=np.unique(labels))\n",
    "disp = ConfusionMatrixDisplay(confusion_matrix=cm,\n",
    "                               display_labels=np.unique(labels))\n",
    "disp.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "c86a3396",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 512)               401920    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 128)               65664     \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 468,874\n",
      "Trainable params: 468,874\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "c003dff2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "402192"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "784*512 + 784"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "794f5205",
   "metadata": {},
   "source": [
    "### With different architectures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "6973ffd8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model_arch(layers):\n",
    "    model = tf.keras.Sequential([])\n",
    "    for i in range(len(layers)):\n",
    "        model.add(tf.keras.layers.Dense(layers[i], activation = 'relu'))\n",
    "                  \n",
    "    model.add(tf.keras.layers.Dense(num_labels, activation = 'sigmoid'))\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8aa4d604",
   "metadata": {},
   "source": [
    "#### Deep v/s shallow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "0509c3f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "deep_layers = np.array([256, 128, 64], dtype = int)\n",
    "shallow_layers = np.array([128, 128])\n",
    "deep_nn = build_model_arch(deep_layers)\n",
    "shallow_nn = build_model_arch(shallow_layers)\n",
    "models = [deep_nn, shallow_nn]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "fba0cc81",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1782/1782 [==============================] - 35s 19ms/step - loss: 0.1818 - accuracy: 0.9454\n",
      "Epoch 2/5\n",
      "1782/1782 [==============================] - 28s 16ms/step - loss: 0.0642 - accuracy: 0.9809\n",
      "Epoch 3/5\n",
      "1782/1782 [==============================] - 31s 17ms/step - loss: 0.0420 - accuracy: 0.9874\n",
      "Epoch 4/5\n",
      "1782/1782 [==============================] - 33s 19ms/step - loss: 0.0277 - accuracy: 0.9917\n",
      "Epoch 5/5\n",
      "1782/1782 [==============================] - 35s 19ms/step - loss: 0.0191 - accuracy: 0.9943\n",
      "94/94 [==============================] - 2s 17ms/step - loss: 0.0579 - accuracy: 0.9797\n",
      "Epoch 1/5\n",
      "1782/1782 [==============================] - 30s 16ms/step - loss: 0.1858 - accuracy: 0.9462\n",
      "Epoch 2/5\n",
      "1782/1782 [==============================] - 32s 18ms/step - loss: 0.0785 - accuracy: 0.9769\n",
      "Epoch 3/5\n",
      "1782/1782 [==============================] - 33s 18ms/step - loss: 0.0543 - accuracy: 0.9840\n",
      "Epoch 4/5\n",
      "1782/1782 [==============================] - 26s 15ms/step - loss: 0.0397 - accuracy: 0.9881\n",
      "Epoch 5/5\n",
      "1782/1782 [==============================] - 33s 19ms/step - loss: 0.0296 - accuracy: 0.9915\n",
      "94/94 [==============================] - 1s 10ms/step - loss: 0.0600 - accuracy: 0.9783\n"
     ]
    }
   ],
   "source": [
    "batch_size = 32\n",
    "epochs = 5\n",
    "summary = []\n",
    "test_scores = []\n",
    "\n",
    "for model in models:\n",
    "    model.compile(optimizer = tf.keras.optimizers.SGD(learning_rate = 1e-1),\n",
    "    loss = 'sparse_categorical_crossentropy',\n",
    "    metrics = ['accuracy'])\n",
    "    \n",
    "    model.fit(train_data, \n",
    "          train_labels, \n",
    "          batch_size = batch_size, \n",
    "          epochs = epochs)\n",
    "    \n",
    "    _, test_score = model.evaluate(test_data, test_labels)\n",
    "    \n",
    "    test_scores.append(test_score)\n",
    "    summary.append(model.history)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "b7a63a2e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_7 (Dense)              (None, 128)               100480    \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 128)               16512     \n",
      "_________________________________________________________________\n",
      "dense_9 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 118,282\n",
      "Trainable params: 118,282\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.9783333539962769"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#histograms of wrongly predicted labels\n",
    "shallow_nn.summary()\n",
    "test_scores[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "879eb2aa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_3 (Dense)              (None, 256)               200960    \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 128)               32896     \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 64)                8256      \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 10)                650       \n",
      "=================================================================\n",
      "Total params: 242,762\n",
      "Trainable params: 242,762\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.9796666502952576"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deep_nn.summary()\n",
    "test_scores[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cea3d63d",
   "metadata": {},
   "source": [
    "### CNN with different architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "38c67f72",
   "metadata": {},
   "outputs": [],
   "source": [
    "def CNN_model():\n",
    "    cnn_model = tf.keras.Sequential([\n",
    "        tf.keras.layers.Conv2D(24, 3, activation = 'relu'),\n",
    "        tf.keras.layers.MaxPool2D(2),\n",
    "        tf.keras.layers.Conv2D(36, 2, activation = 'relu'),\n",
    "        tf.keras.layers.MaxPool2D(2),\n",
    "        tf.keras.layers.Flatten(),\n",
    "        tf.keras.layers.Dense(128, activation = 'relu'),\n",
    "        tf.keras.layers.Dense(10, activation = 'sigmoid')\n",
    "    ])\n",
    "    return cnn_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "05bd394a",
   "metadata": {},
   "outputs": [],
   "source": [
    "cnn_model = CNN_model()\n",
    "cnn_model.compile(optimizer = tf.keras.optimizers.SGD(learning_rate = 1e-1),\n",
    "                 loss = 'sparse_categorical_crossentropy',\n",
    "                 metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "02fe235e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "891/891 [==============================] - 70s 26ms/step - loss: 0.2098 - accuracy: 0.9363\n",
      "Epoch 2/5\n",
      "891/891 [==============================] - 19s 21ms/step - loss: 0.0570 - accuracy: 0.9829\n",
      "Epoch 3/5\n",
      "891/891 [==============================] - 24s 27ms/step - loss: 0.0376 - accuracy: 0.9882\n",
      "Epoch 4/5\n",
      "891/891 [==============================] - 24s 27ms/step - loss: 0.0281 - accuracy: 0.9911\n",
      "Epoch 5/5\n",
      "891/891 [==============================] - 23s 26ms/step - loss: 0.0224 - accuracy: 0.9928\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2c9e15cba00>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch = 64\n",
    "epochs = 5\n",
    "\n",
    "cnn_model.fit(train_images, train_labels, \n",
    "              batch_size = batch, \n",
    "              epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "4bdd233c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d (Conv2D)              (None, 26, 26, 24)        240       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 13, 13, 24)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 12, 12, 36)        3492      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 6, 6, 36)          0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 1296)              0         \n",
      "_________________________________________________________________\n",
      "dense_10 (Dense)             (None, 128)               166016    \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 171,038\n",
      "Trainable params: 171,038\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "cnn_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "b6e6d5c3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 [==============================] - 2s 18ms/step - loss: 0.0313 - accuracy: 0.9913\n",
      "test accuracy : 0.9913333058357239 \t test loss: 0.03129655122756958\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_acc = cnn_model.evaluate(test_images, test_labels)\n",
    "\n",
    "print(f'test accuracy : {test_acc} \\t test loss: {test_loss}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "8df9fbb2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#what did we misclassify?\n",
    "predictions = cnn_model.predict(test_images)\n",
    "predictions = np.argmax(predictions, axis = -1)\n",
    "idxs = np.where(predictions != test_labels)\n",
    "misclassified_labels = predictions[idxs]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "3f6710a5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABHoAAAOzCAYAAADHj2GHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOzdeZRdZZX//88mjBFcIaTAyBSIYRQITeTHqIjNIB2DMrVBMYzRVhRZKGNDoFkyyCCiQpsO6Qq2RKEBA4IIHaCZhwqEBDoK+UqAkJChab4kIENg//6om/4WOc9N3ap7zrn3PM/7tRarqnY9dc8+l0+d3Oyce465uwAAAAAAAFB9a7S6AQAAAAAAAOSDQQ8AAAAAAEAkGPQAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARIJBDwAAAAAAQCSaGvSY2cFm9mczm2tmZ+bVFFAF5B+pIvtIFdlHqsg+Ukb+UUXm7v37QbMBkp6XdICk+ZKelDTW3f+r3s8MGTLEhw0b1q/tAX0xb948LV261Ip6/L7mn+yjTDNmzFjq7h1FPDbHfrSzIo/9ZB/tjOwjZbzuQapWd+xfs4nH3V3SXHf/iySZ2W8kHSqpbuiHDRumrq6uJjYJNGbUqFFFb6JP+Sf7KJOZvVTgw3PsR9sq+NhP9tG2yD5SxusepGp1x/5m3rq1qaRXenw9v1YDUkD+kSqyj1SRfaSK7CNl5B+V1MygJ3SKUOZ9YGY23sy6zKxryZIlTWwOaCu95p/sI1Ic+5Eqso9UkX2kjPyjkpoZ9MyXtHmPrzeTtGDVRe4+0d1Hufuojo5C3joJtEKv+Sf7iBTHfqSK7CNVZB8pI/+opGau0fOkpBFmtpWkVyV9VdLRuXQFtD/yj1SRfaSK7CNVZH8VY8aMydR222234NoJEyYU3Q6KRf5RSf0e9Lj7CjM7WdIfJQ2QNNndn8utM6CNkX+kiuwjVWQfqSL7SBn5R1U1c0aP3P1OSXfm1AtQKeQfqSL7SBXZR6rIPlJG/lFFzVyjBwAAAAAAAG2EQQ8AAAAAAEAkmnrrFgAAaM5ee+0VrD/66KPB+rXXXpupfetb38q1JwCA9KUvfSlYP+iggzK1Aw88sOh2gLb1yiuvBOtbbrllplbv92ratGm59pQ6zugBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEd93K0f3335+pjRkzJrh22bJlmdrYsWODa2+44Yam+gIAVM+UKVOC9aOOOqrkTgAgTb///e+D9W9+85uZ2oIFC4JrZ8yYkanVe80PVIG7Z2ovvfRScO2hhx6aqd16662594QszugBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEhwMeZ+uP3224P1Rx99NFM777zzgmvNLFP761//Glx79tlnB+snnHBCpjZ8+PDgWgBA633uc5/L1Ood4/fff/9gfd111821JwBA8xYuXBisz507t+ROgGKF/s76ta99Lbi23kWaUTzO6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiERTd90ys3mSlkn6QNIKdx+VR1Ptbv78+cF66Ark3/72t4NrBw4cmKlNnz49uPbHP/5xsL7NNtvUazGDu3HlL9X8A2S//0J/fgwePDi4NvTnBFqL7CNl5H/1xo4d2+oWUBCyjyrK4/bqn3f3pTk8DlBF5B+pIvtIFdlHysg/UkX2USm8dQsAAAAAACASzQ56XNLdZjbDzMaHFpjZeDPrMrOuJUuWNLk5oK2sNv9kHxHj2I9UkX2kjNc9SBXHflROs4Oevd39byR9UdJ3zOyzqy5w94nuPsrdR3V0dDS5OaCtrDb/ZB8R49iPVJF9pIzXPUgVx35UTlPX6HH3BbWPi83sVkm7S3ogj8ba2fHHHx+sf/DBB5naeuut1/Djfu5znwvWd9xxx2B93Lhxmdrzzz8fXHvRRRc13Acak2r+y2RmDa919wI7QU9kv/9uuummTG3EiBEt6AT9QfarY9KkSZnaSSed1PTjXnrppcH66aef3vRjtzvyj1SR/d7xOrz99PuMHjP7mJltsPJzSQdKejavxoB2Rv6RKrKPVJF9pIz8I1VkH1XVzBk9m0i6tfYv7mtKusHd78qlK6D9kX+kiuwjVWQfKSP/SBXZRyX1e9Dj7n+RtEuOvQCVQf6RKrKPVJF9pIz8I1VkH1XF7dUBAAAAAAAiwaAHAAAAAAAgEk3ddStV66yzTiGPu+aa4f8dQ4YMCdZDd3gYPHhwrj0BZejL3bX6+hjcBQCtcNBBBwXrnZ2dmdoGG2xQcDdAtbz33nvB+mmnnZapzZs3L7h2++23z9Ruv/32pvqSpMcffzxYv/jiizO1s846q+ntoT1deOGFmdovf/nL4Nrdd989Uzv33HNz7wkAeuKMHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASHDXLQCl6ssdtvpyx6x6jxuqcycuFG38+PHBeuguPPXuzLPtttvm2hPQburdMWvSpEnB+k033ZSphe5oJEn77rtvpjZ69OjGm6vjwQcfDNafe+65ph8b7WfChAnB+tSpUzO1J554Irj2+eefz9QGDhwYXFvvzw7uzoh2l8cddJEvzugBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEhwMeYKeOedd4L1K664IlPbcccdg2s/9alPZWpc2A0xqXeBZS4Oh1Y4/PDDG157zTXXBOvf+MY3gvXddtutXz0B7abexZgvvvjiYD10geWTTz45uPbAAw/sd1+rs/322wfrgwYNKmR7aK3zzz8/WB8wYECmVu9izIsXL87UQq/hJenoo48O1nnNDjTm6quvztQ+/elPB9fuv//+RbfTUpzRAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNADAAAAAAAQiV7vumVmkyWNlrTY3T9dqw2W9FtJwyTNk3SUu/9PcW2mbb311gvWL7rookztD3/4Q3DtmDFjMrX77ruvucYSQP7jVO9OXPXu3JUisl+M0N246uXu5ptvDtY7OzsztUMOOSS49otf/GLjzUES2S/K8uXLG6pJ0pZbbhmsX3vttZlavbtgFeXYY48tdXtlI/+NOffcczO1ZcuWBdfeeeedmdoll1wSXLvGGvwbfKuQ/eYsWLAgWP/KV76Sqd16662F9fHCCy9kahtvvHFh22tnjRxNOiUdvErtTEnT3X2EpOm1r4EYdYr8I02dIvtIU6fIPtLVKfKPNHWK7CMivQ563P0BSa+vUj5U0pTa51MkfTnnvoC2QP6RKrKPVJF9pIz8I1VkH7Hp7/mBm7j7Qkmqfax7PpSZjTezLjPrWrJkST83B7SVhvJP9hEhjv1IFdlHynjdg1Rx7EdlFf5GUHef6O6j3H1UR0dH0ZsD2gbZR8rIP1JF9pEqso+UkX+0m14vxlzHIjMb6u4LzWyopMV5NoWPGjBgQLA+cuTITG369OnBtffff3+eLaWO/FdI6EK39S7GjF6R/QIcccQRwXq9nF522WWZ2pAhQ4JruRhzbsh+k0IXEb/77ruDax9++OFgvV7OUTjy34CJEycG608++WSmdtdddwXXkvG2Q/YDBg4cmKnNnj07uHaHHXbI1D7+8Y/n3tNK77zzTqa29957F7a9dtbfM3pukzSu9vk4SdPyaQeoBPKPVJF9pIrsI2XkH6ki+6isXgc9ZjZV0qOStjWz+WZ2gqRLJB1gZi9IOqD2NRAd8o9UkX2kiuwjZeQfqSL7iE2vb91y97F1vvWFnHsB2g75R6rIPlJF9pEy8o9UkX3EpvCLMQMAAAAAAKAcDHoAAAAAAAAi0d+7bgFAv3AXLKA5Bx10UKa2zz77tKAToHFvv/12prZs2bLg2qFDhxbdDpC75cuXB+tvvfVWprbJJpsU3Q5Qqu222y5Yf/rppzO1PfbYI7h2xYoVwfr3vve9TO2KK67oQ3dp4oweAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEF2MGUIh6F1gOXYwZSMX111+fqb355pvBtZMmTQrWv/Wtb2VqX/jCF5prDADQsKlTp2Zq9V7fHH300UW3A7St0EWap0yZElz72GOPBetceLl/OKMHAAAAAAAgEgx6AAAAAAAAIsGgBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACAS3HULQHK48xda5YILLsjUtt122+DaXXfdNVivtx5oB11dXcH6nDlzSu6kXAsXLszUZsyYEVw7evToottBwW644YZMbdSoUcG1Y8eOLbqdUjz//PN9qoeQ/Xi9++67wXrobqM33XRTcO3dd9+da0+p44weAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIhEr4MeM5tsZovN7NketfPN7FUzm1n775Bi2wRag/wjVWQfqSL7SBn5R6rIPmLTyF23OiX9XNKql8z+ibtfnntHQHvpFPlHmjpF9vvt5ptvDtb33XffTO2ss84KruXuWi3TKbLfbxdddFGwfs8992RqBx54YNHtlObpp5/O1MaPHx9cu2DBgqLbaUanyH+vzKzVLfTJBx98EKwvXbq04bW//OUvg/Urr7yy4T7a/K6nnSL7/bZ48eJg/fTTT2+ohvz1ekaPuz8g6fUSegHaDvlHqsg+UkX2kTLyj1SRfcSmmWv0nGxms2qnuW2YW0dANZB/pIrsI1VkHykj/0gV2Ucl9XfQc62k4ZJGSloo6Yp6C81svJl1mVnXkiVL+rk5oK00lH+yjwhx7EeqyD5SxusepIpjPyqrX4Med1/k7h+4+4eS/kXS7qtZO9HdR7n7qI6Ojv72CbSNRvNP9hEbjv1IFdlHynjdg1Rx7EeVNXIx5gwzG+ruC2tffkXSs6tbD8SE/FdfvYsotvlFAluO7DfuiCOOCNbnzp2bqQ0fPrzodtAkst+8cePGZWoXXHBBCzopxv7775+pdXV1taCT/JH/6gtddFmSDj744Ext0aJFwbXLly/PtacqIPthK1asyNSmTp0aXHveeedlaqeeemruPSGr10GPmU2VtJ+kIWY2X9IESfuZ2UhJLmmepG8W2CPQMuQfqSL7SBXZR8rIP1JF9hGbXgc97j42UL6ugF6AtkP+kSqyj1SRfaSM/CNVZB+xaeauWwAAAAAAAGgjDHoAAAAAAAAiwaAHAAAAAAAgEv266xYA9Fe9O14BMdljjz2C9Tlz5mRqG2+8cXDtBhtskGtP/fHOO+8E608//XSwvueeexbZDips/fXXz9Q22mijFnRSjHXXXTdT++QnP9mCTlCGsu/SeeGFF2ZqTzzxRMM//+677wbrf/rTnzK1esd9YKX33nsvU/v5z38eXPvyyy8X3U5LnXXWWZnaxRdf3IJOsjijBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACASDHoAAAAAAAAiwcWYK2DFihXB+n/+539man/961+Da8eOHZtrT0Ceyr6oIVC0Rx99NFgfPnx4pvarX/0quHavvfbKtaeV3nzzzWD9pZdeytQWL14cXHvqqacG67Nmzep/Y4gax3nE7v777w/W88j+1KlTM7Xnn3++4Z+vd3H/b3/72w2vBVY67rjjMrVYbrZy9tlnB+vz5s0L1g855JACu2kOZ/QAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARIJBDwAAAAAAQCQY9AAAAAAAAESCu271YtmyZZnaOuusE1y79tprN/y477zzTqb2/vvvB9e+9dZbwfqPf/zjTG233XYLrr3hhhsa7g0AUJ7XXnstWK93h4dmPfjgg8H6eeedl6ltscUWwbXcXQt9FcsdWYB66t11q169Wdtss03D9cGDBwfXnnPOOQ2vBVY6/vjjM7XHHnusBZ005sILLwzW77333kytq6sruPa6664L1o844oj+N1YwzugBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEj0ejFmM9tc0vWSPiHpQ0kT3f2nZjZY0m8lDZM0T9JR7v4/xbXaGmPGjMnU/uEf/iG49qijjmr4cS+//PJMbfr06cG19S7yPHbs2Ext3333bbgHrF7q2UfayH8xBg0alKmddNJJwbXLly8P1t97771Mbb311guuDdVHjx4dXPviiy8G66kh+0gV2W9cO1xc/Oijjw7WJ0yYUHIncSD/jQv9fXPBggXBtV/5ylcytVtvvbVP2/u3f/u3TK3eBZZDFi5cGKxfffXVmdovf/nL4Nqtt946WF9jjfY9b6aRzlZIOs3dt5e0h6TvmNkOks6UNN3dR0iaXvsaiAnZR8rIP1JF9pEqso+UkX9EpddBj7svdPenap8vkzRH0qaSDpU0pbZsiqQvF9Uk0ApkHykj/0gV2UeqyD5SRv4Rmz6da2RmwyTtKulxSZu4+0Kp+xdD0sZ5Nwe0C7KPlJF/pIrsI1VkHykj/4hBw4MeM1tf0s2Svu/ub/bh58abWZeZdS1ZsqQ/PQItRfaRMvKPVJF9pIrsI2XkH7FoaNBjZmupO/C/dvdbauVFZja09v2hkhaHftbdJ7r7KHcf1dHRkUfPQGnIPlJG/pEqso9UkX2kjPwjJo3cdcskXSdpjrtf2eNbt0kaJ+mS2sdphXTYYvfff3+mtsEGGwTXvvLKKw0/7rvvvpup/c3f/E1w7TrrrBOsh654Pnz48IZ7wOqlnn2kjfwXY8aMGQ2vPeyww4L10N0qzjjjjOBa7r7Sd2QfqSL7jbvtttta3QJyRv6bs2LFimB92rTs0/Xxj3+8T48dutto6O/SUvjO1uPHjw+uHThwYKY2YMCAPvXWznod9EjaW9Ixkmab2cxa7Wx1h/1GMztB0suSjiymRaBlyD5SRv6RKrKPVJF9pIz8Iyq9Dnrc/SFJVufbX8i3HaB9kH2kjPwjVWQfqSL7SBn5R2z6dNctAAAAAAAAtC8GPQAAAAAAAJFg0AMAAAAAABCJRi7GjFXcfvvtfaqH3HDDDZna2LFj+90TgMa5e6tbAHp1yy239L4IaKEvfelLmVpXV1dw7TbbbFN0OwCAEoTuVsVr6/bDGT0AAAAAAACRYNADAAAAAAAQCQY9AAAAAAAAkWDQAwAAAAAAEAkuxtwLLiwF9E+93x0zK7WP0Pb4vQaA5v3+979veC3HXQAAysMZPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNADAAAAAAAQCQY9AAAAAAAAkeCuWwCixp1eAKAYo0ePztS6urqCa8u+4yIAACnjjB4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiESvF2M2s80lXS/pE5I+lDTR3X9qZudLOknSktrSs939zqIaBcpG9pEy8o9Ukf3G3X777ZnaYYcd1oJOkAeyj5SRf8SmkbturZB0mrs/ZWYbSJphZvfUvvcTd7+8uPaAliL7SBn5R6rIPlJF9pEy8o+o9DrocfeFkhbWPl9mZnMkbVp0Y0CrkX2kjPwjVWQfqSL7SBn5R2z6dI0eMxsmaVdJj9dKJ5vZLDObbGYb1vmZ8WbWZWZdS5YsCS0B2h7ZR8rIP1JF9pEqso+UkX/EoOFBj5mtL+lmSd939zclXStpuKSR6p5+XhH6OXef6O6j3H1UR0dHDi0D5SL7SBn5R6rIPlJF9pEy8o9YNDToMbO11B34X7v7LZLk7ovc/QN3/1DSv0javbg2gdYg+0gZ+UeqyD5SRfaRMvKPmDRy1y2TdJ2kOe5+ZY/60Np7GSXpK5KeLaZFoDXIfjHcvdUtoAHkH6ki+8255ZZbWt0C+onsI2XkH7Fp5K5be0s6RtJsM5tZq50taayZjZTkkuZJ+mYhHQKtQ/aRMvKPVJF9pIrsI2XkH1Fp5K5bD0mywLfuzL8doH2QfaSM/CNVZB+pIvtIGflHbPp01y0AAAAAAAC0LwY9AAAAAAAAkWDQAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNADAAAAAAAQCQY9AAAAAAAAkTB3L29jZkskvVT7coikpaVtvFwx75tUjf3b0t07Wt3ESgllX4p7/6qyb+2a/6o8f/0V8/5VZd/aNftSdZ7D/oh536Rq7B/Zb42Y902qzv61a/6r8vz1V8z7V5V9q5v9Ugc9H9mwWZe7j2rJxgsW875J8e9f0WJ//mLev5j3rQyxP38x71/M+1aWmJ/DmPdNin//ihbz8xfzvknx71/RYn/+Yt6/GPaNt24BAAAAAABEgkEPAAAAAABAJFo56JnYwm0XLeZ9k+Lfv6LF/vzFvH8x71sZYn/+Yt6/mPetLDE/hzHvmxT//hUt5ucv5n2T4t+/osX+/MW8f5Xft5ZdowcAAAAAAAD54q1bAAAAAAAAkWDQAwAAAAAAEInSBz1mdrCZ/dnM5prZmWVvP29mNtnMFpvZsz1qg83sHjN7ofZxw1b22F9mtrmZ3Wdmc8zsOTM7pVaPYv/KRvarg+znj/xXB/nPF9mvDrKfL7JfHWQ/f+S/OmLNf6mDHjMbIOkXkr4oaQdJY81shzJ7KECnpINXqZ0pabq7j5A0vfZ1Fa2QdJq7by9pD0nfqf3/imX/SkP2K4fs54j8Vw75zwnZrxyynxOyXzlkP0fkv3KizH/ZZ/TsLmmuu//F3d+T9BtJh5bcQ67c/QFJr69SPlTSlNrnUyR9udSmcuLuC939qdrnyyTNkbSpItm/kpH9CiH7uSP/FUL+c0X2K4Ts54rsVwjZzx35r5BY81/2oGdTSa/0+Hp+rRabTdx9odQdHEkbt7ifppnZMEm7SnpcEe5fCch+RZH9XJD/iiL/TSP7FUX2m0b2K4rs54L8V1RM+S970GOBGvd3b3Nmtr6kmyV9393fbHU/FUX2K4js54b8VxD5zwXZryCynwuyX0FkPzfkv4Jiy3/Zg575kjbv8fVmkhaU3EMZFpnZUEmqfVzc4n76zczWUnfgf+3ut9TK0exfich+xZD9XJH/iiH/uSH7FUP2c0P2K4bs54r8V0yM+S970POkpBFmtpWZrS3pq5JuK7mHMtwmaVzt83GSprWwl34zM5N0naQ57n5lj29FsX8lI/sVQvZzR/4rhPzniuxXCNnPFdmvELKfO/JfIbHm39zLPYvMzA6RdJWkAZImu/uPSm0gZ2Y2VdJ+koZIWiRpgqTfSbpR0haSXpZ0pLuvevGqtmdm+0h6UNJsSR/Wymer+z2Lld+/spH96iD7+SP/1UH+80X2q4Ps54vsVwfZzx/5r45Y81/6oAcAAAAAAADFKPutWwAAAAAAACgIgx4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDnl6Y2SAz+3bB2zjLzOaa2Z/N7KAit9Vjm/eb2ahA/UIzm2VmM83sbjP7ZBn9oP0UnX0z28jM7jOz5Wb286K2E9huvewPNrN7zOyF2scNy+oJ7aWE7K9tZv9qZrPN7Bkz26+oba2y3WD2a9/7bu3PoOfM7Mdl9IP2lOCxfxcze7T2+3i7mX28rJ7QXkp6zb9zLW/P1TK3bpHbq22zXvZ/W3u9P9PM5pnZzKJ7QXvidU+cr3sY9PRukKRg8M1sQLMPbmY7SPqqpB0lHSzpmv4+rpmt2Ww/ki5z953dfaSk30s6L4fHRDUVmn1J70g6V9IPmn2gnLJ/pqTp7j5C0vTa10hT0dk/SZLcfSdJB0i6wsz69edxHtk3s89LOlTSzu6+o6TLm31MVFpqx/5Jks6s/T7eKumHOTwmqqno1/xrSvo3Sd+qHWv3k/R+E4/VFHf/e3cfWXvNf7OkW5p9TFQWr3sixKCnd5dIGl6bdl9mZvvV/iXqBkmzzWyYmT27crGZ/cDMzq99PtzM7jKzGWb2oJltF3j8QyX9xt3fdfcXJc2VtPuqi2qT9kvN7Inaf5+q1TvN7Eozu0/SpWb2MTObbGZPmtnTZnZobd16Zvab2tk6v5W0Xmhn3f3NHl9+TJL34zlDHArNvru/5e4PqftFf11lZV/dv4tTap9PkfTlvjxZiErRx/0d1D1MlLsvlvSGpNC/tpaV/X+QdIm7v9ujJ6QrtWP/tpIeqH1+j6TD+/JkISpFH/sPlDTL3Z+RJHf/b3f/YNVFJWZ/5fZM0lGSpvbjOUMceN0TIQY9vTtT0v+pTbxX/ivP7pLOcfcdevnZiZK+6+67qftfrq6RJDMbY2b/VFuzqaRXevzM/Fot5E13313SzyVd1aO+jaS/dffTJJ0j6V53/4ykz0u6zMw+pu5Av+3uO0v6kaTdVv6wmU2yHqe1mdmPzOwVSV8TZ/SkrOjs90UZ2d/E3RdKUu3jxv3oE3EoOvvPSDrUzNY0s63UncnN6zxeGdnfRtK+Zva4mf2nmX2ml31E3FI79j8raUzt8yNV/3cR8Ss6+9tIcjP7o5k9ZWanr+bxSnnNX7OvpEXu/kIv+4h48bonQnmc8pqiJ2pn39RlZutL2kvSTd2DcknSOpLk7rdJum3l0sCP1zuLZmqPjz/pUb+px78IHChpjJmtPCV6XUlbSPqspKtr259lZrP+d2PuJ35k4+7nSDrHzM6SdLKkCfX3FInJM/t9UUr2gdXIM/uTJW0vqUvSS5IekbSizsOWkf01JW0oaQ9Jn5F0o5lt7e6c0YmVYj72Hy/pajM7r9bje/3oE/HKM/trStpH3cfZtyVNN7MZ7j498LBlvu4ZK87mQRaveyqOQU//vNXj8xX66JlRKy+qtoakN2rve12d+froRHMzSQvqrPU6n/fsxyQd7u5/7vmDtV++vob3Bkl3iEEP/p88s98XZWR/kZkNdfeFZjZUUpSncaLfcsu+u6+QdOrKr83sEUn1/iW1jOzPl3RL7QXOE2b2oaQhkpY08LNIQ7THfnf/k7r/0iAz20bS3/W3WUQp79f8/+nuSyXJzO6U9DeqvaVlFaW85rfu650cph5nPgA1vO6pON661btlkjZYzfcXSdrYuu8isY6k0dL/XuvmRTM7UmGBlgEAACAASURBVOp+/6uZ7RL4+dskfdXM1qmdyjZC0hN1tvX3PT4+WmfNHyV912pJN7Nda/UH1P1WLJnZpyXtHPphMxvR48sxkv5UZzuIX9HZ74vCs6/u38Vxtc/HSZrWZM+orkKzb2YDa6cYy8wOkLTC3f+rzrbKyP7vJO1fW7eNpLUlLa2zFvFL6thvZhvXPq4h6R8l/XOTPaO6is7+HyXtXPszYE1Jn5PUymO/JP2tpD+5+/zVrEH8eN0T4eseBj29cPf/lvSwmT1rZpcFvv++pH+S9Li671LVczDyNUknmNkzkp5T98VeP/KeRXd/TtKN6j7Q3yXpO6ELs9WsY2aPSzpFPaaiq7hQ0lqSZln3RbMurNWvlbR+7RS209VjmGQffc/iJbV9naXuf+E6pc52ELmis1/7ep6kKyUda2bzrfsudCGlZF/SAWb2grrvCHBJne0gciVkf2NJT5nZHElnSDpmNe2Ukf3Jkrau/exvJI2L7fRlNC7BY/9YM3u+th8LJP1rne0gciW85v8fdef+SUkzJT3l7nfUaaeM7Evdd/7lbVuJ43VPnK97LMJ9ilLtRdGolad7Aqkg+0gV2UfKyD9SRfaRKrKfr6YGPWZ2sKSfShogaZK7r/ZfwIcMGeLDhg3r9/aARs2bN09Lly4NXeg6N33JP9lHmWbMmLHU3TuKenyO/WhXRR/7yT7aFdlHynjdg1St7tjf74sxm9kASb9Q91sc5kt60sxuW8377TRs2DB1dXX1d5NAw0aNWvXOkfnqa/7JPspkZi8V+Ngc+9G2ijz2k320M7KPlPG6B6la3bG/mWv07C5prrv/xd3fU/f72w5t4vGAKiH/SBXZR6rIPlJF9pEy8o9KambQs6mkV3p8Pb9W+wgzG29mXWbWtWRJVHcsQ9p6zT/ZR6Q49iNVZB+pIvtIGflHJTUz6Am9FyxzwR93n+juo9x9VEdHYW+dBMrWa/7JPiLFsR+pIvtIFdlHysg/KqmZQc98SZv3+Hozdd+WEkgB+UeqyD5SRfaRKrKPlJF/VFIzg54nJY0ws63MbG1JX5V0Wz5tAW2P/CNVZB+pIvtIFdlHysg/Kqnfd91y9xVmdrKkP6r7VnOT3f253DoD2hj5R6rIPlJF9pEqso+UkX9UVb8HPZLk7ndKujOnXoBKIf9IFdlHqsg+UkX2kTLyjypq5q1bAAAAAAAAaCMMegAAAAAAACLR1Fu3AKTluOOOC9Y7OzsztcMOOyy49uabb86zJQAAgFKNGDEiU5s7d27Tj/urX/0qWP/617/e9GMDSAtn9AAAAAAAAESCQQ8AAAAAAEAkGPQAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARIK7bgEoxP333x+sH3PMMZlavbtMAAAAtMp2220XrP/lL39p+DGmTZuWqV177bX97gkAGsEZPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNADAAAAAAAQCS7GDKBpo0ePztQOP/zw4Nrly5dnaieddFJw7cUXXxysDxkypA/dAQAArN5WW22Vqb388svBtffee2+mtuWWWwbXfuITn8jUrrvuuj52BwB9wxk9AAAAAAAAkWDQAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABCJpi7GbGbzJC2T9IGkFe4+Ko+mgCog/0gV2UeqyD5SRv6RKrKPKsrjrlufd/elOTwOUEXkX9JOO+2UqR177LHBtc8++2ymdtpppwXXnnPOOcE6d91qC2QfqSL7SFm0+Z83b17DazfbbLNMbdiwYfk1g3YUbfb7atddd83UZs6cGVx71VVXZWqnnHJK7j0hi7duAQAAAAAARKLZQY9LutvMZpjZ+DwaAiqE/CNVZB+pIvtIGflHqsg+KqfZt27t7e4LzGxjSfeY2Z/c/YGeC2q/DOMlaYsttmhyc0BbWW3+yT4ixrEfqSL7SBmve5Aqjv2onKbO6HH3BbWPiyXdKmn3wJqJ7j7K3Ud1dHQ0szmgrfSWf7KPWHHsR6rIPlLG6x6kimM/qqjfZ/SY2cckreHuy2qfHyjpn3LrLEHvvfdesH7FFVcE62effXbDj7148eJMjYNQ/5F/pIrsl+e4444L1js7OzO1ehcBve+++xpei9Uj+435r//6r2D9qKOOytSee+65ottpyPnnn5+pTZgwofxG2liq+b/jjjuC9U033bTkTtAqqWZfkvbcc89g/YUXXii5E/RHM2/d2kTSrWa28nFucPe7cukKaH/kH6ki+0gV2UfKyD9SRfZRSf0e9Lj7XyTtkmMvQGWQf6SK7CNVZB8pI/9IFdlHVXF7dQAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBINHMxZjQhdIetenfXmjhxYrA+evToTO2RRx5prjGgH5YsWZKpzZ49O7j2lVdeydR23HHH4Nq11167ucaANrN06dJM7dJLLw2ufeihh5p6XEn64Q9/mKmdcsopwbX77LNPw9sDJOnBBx/M1G655Zbg2tCfE0C7+8xnPhOsr7vuuiV3ApTvscceC9b/6Z+yNx279dZbg2tvuOGGTK3enZ+PPvroPnRXjFC/knTPPfc0/BhbbLFFsH7BBRf0q6f+4oweAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIcNetFvnHf/zHTO2mm24Krv3FL34RrB9yyCGZWmdnZ3DtNddck6lNmDBhNR0CWWuuGT5kTJ48OVObNGlScO1OO+2Uqc2aNau5xoA2M3fu3GB92rRpmVq943a9O2l96lOfytRGjhwZXHvXXXdlan/3d38XXMtdt9BX9957b6Z21VVXNfzzgwYN6lO9TPPmzQvWhw0bVmofANBuvvSlL2VqW221VXDt1Vdfnak9+eSTwbVF3XUr9PcUSVq0aFGmdscddwTXPvzwww1v7xOf+ESwHrpb31lnndXw4/YVZ/QAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARIJBDwAAAAAAQCS4GHOL7LzzzplavQtThS66XM+xxx4brG+88caZGhdjRl9dfPHFwfqGG26YqV122WVFtwO0rXoX7TvzzDMztXoX7at30dcvf/nLmdopp5wSXPvaa69lau+++25wbejiz0OGDAmuBfJw2GGHBesnnnhiqX3cfffdmdrYsWODax999NGi20Eb+d73vheshy6qev755wfXTp06NVP77Gc/G1y75557Nt4c0CLTp0/P1OodM2fMmFFID++//36w/oc//CFTu/DCC4NrQxfd32WXXYJr6/151RddXV1NP0ZfcEYPAAAAAABAJBj0AAAAAAAARIJBDwAAAAAAQCQY9AAAAAAAAESi10GPmU02s8Vm9myP2mAzu8fMXqh9zF6JFYgA+UeqyD5SRfaRMvKPVJF9xKaRu251Svq5pOt71M6UNN3dLzGzM2tfn5F/e/E66KCDMrXrrruuBZ2gF50i//+r3h14QnfdQuV1iuznLnSHrdAdWSRps802C9bXX3/9TG3QoEHBtaHH/tGPfhRcO3fu3Ewt0bvndSrh7L/++uuZ2l133dXwz5988skNr613V9Gy7zwUuuvWY489VmoPbaRTieb/Zz/7WVM/P3DgwGD9j3/8Y6Z26qmnBtcOHz68qR7QlE4lmv2++sEPfpCplZ3dt956K1gfN25cpvbGG28E12655ZaZWmjfJOnrX/96H7prD72e0ePuD0ha9U/9QyVNqX0+RVL2Xq9ABMg/UkX2kSqyj5SRf6SK7CM2/b1GzybuvlCSah83zq8loO2Rf6SK7CNVZB8pI/9IFdlHZRV+MWYzG29mXWbWtWTJkqI3B7QNso+UkX+kiuwjVWQfKSP/aDf9HfQsMrOhklT7uLjeQnef6O6j3H1UR0dHPzcHtJWG8k/2ESGO/UgV2UfKeN2DVHHsR2U1cjHmkNskjZN0Se3jtNw6SsTOO++cqW233XYt6AT9QP4bMHLkyGD99NNPL7kT5IjsB0yaNClTmzJlSmCltOaa2T926110ediwYU31Ve+x33nnneDaenVISij7r732WqZ20UUXBdceeeSRmVqzF7RFW0oi/325kDiSkUT282BmrW6hrrXWWitYv/rqqzO1MWPGFN1OaRq5vfpUSY9K2tbM5pvZCeoO+wFm9oKkA2pfA9Eh/0gV2UeqyD5SRv6RKrKP2PR6Ro+7j63zrS/k3AvQdsg/UkX2kSqyj5SRf6SK7CM2hV+MGQAAAAAAAOVg0AMAAAAAABAJBj0AAAAAAACR6O9dt9CkWbNmZWpHHXVUCzoBGvcf//Efwfr777+fqdW7e8XYsfXeAg1U08MPP5yphe5cJElHHHFEprb++uvn3hMAoFyTJ08O1vfZZ59MbcSIEUW3A6Dmd7/7XbC+3377ldtIyTijBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACASDHoAAAAAAAAiwcWYATTs17/+dbA+dOjQTO2EE04ouh2gbYUuvilJl112WcmdZA0ZMqTVLaANzJw5M1h/5JFHMrXPfvazwbW77LJLrj0BVfGrX/0qUzv//PODa6+//vpMbfvtt8+7JaCl3L3VLdT1yU9+MlgfOHBgyZ2UizN6AAAAAAAAIsGgBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACASDHoAAAAAAAAiwV23WuQ73/lOpnbRRRe1oBMgbOHChZnauuuuG1y74YYbFt0OgJycccYZrW4BbWDatGnB+k033ZSp3XjjjcG1O+ywQ649AVURusPWyJEjg2ufeeaZTG348OHBtZtvvnlTfQGtMmnSpGD91VdfzdT222+/gruBxBk9AAAAAAAA0WDQAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABCJXgc9ZjbZzBab2bM9aueb2atmNrP23yHFtgm0BvlHqsg+UkX2kTLyj1SRfcSmkbtudUr6uaTrV6n/xN0vz72jRNx///2ZWr27WqClOpVo/r/61a9mascee2xw7Te+8Y2Cu0ELdCrR7Ndz3HHHBesPPfRQprbPPvsU3U6/DRkypNUttLtOkX2kq1Pk/3/9/ve/D9bffvvtTO2CCy4Irv3d736Xqb322mvBtdx1q6U6RfYbEvo7wuOPPx5c++KLL2Zq3HWrHL2e0ePuD0h6vYRegLZD/pEqso9UkX2kjPwjVWQfsWnmGj0nm9ms2mluG9ZbZGbjzazLzLqWLFnSxOaAttJr/sk+IsWxH6ki+0gZr3uQKo79qKT+DnqulTRc0khJCyVdUW+hu09091HuPqqjo6OfmwPaSkP5J/uIEMd+pIrsI2W87kGqOPajsvo16HH3Re7+gbt/KOlfJO2eb1tA+yL/SBXZR6rIPlJG/pEqso8qa+RizBlmNtTdF9a+/IqkZ1e3HlmzZ89udQvop1Ty/95772Vqa6wRng0PGDCg6HbQBlLJfj2dnZ3BeujCy3vvvXfB3eRv7ty5mVroQtNS/Quzxyr17CNtqeT/7rvvztSOOeaY4NqddtopU9tggw2Ca+tdpBntL5Xs99UZZ5yRqU2bNi249pFHHsnUtt1229x7Qlavgx4zmyppP0lDzGy+pAmS9jOzkZJc0jxJ3yywR6BlyD9SRfaRKrKPlJF/pIrsIza9DnrcfWygfF0BvQBth/wjVWQfqSL7SBn5R6rIPmLTzF23AAAAAAAA0EYY9AAAAAAAAESCQQ8AAAAAAEAk+nXXLTQvdLX+xYsXt6ATpO7www8P1kN31dlvv/2KbQaooBNPPDFTGzduXAs6aczSpUuD9enTp2dql1xySXBtanfdAhC/gw46KFP7zGc+E1w7derUTG3TTTfNvafYvP/++5laV1dXcO2ee+5ZdDvoxbBhw4L10OuIH/7wh8G1AwcOzLMl9AFn9AAAAAAAAESCQQ8AAAAAAEAkGPQAAAAAAABEgkEPAAAAAABAJLgYM5C4W265JVhff/31M7WXXnqp6HYactxxx2VqI0aMaEEnQPVceumlwfry5csztfvuu6/odlCw0PFSkmbOnFlyJ0D13HHHHcF6R0dHyZ3E4Y033sjU9tprr+Bady+6HfTiz3/+c7C+9957Z2rnnntucO2YMWNy7QmN44weAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIcNctAEHXX399Uz+/5ZZbBusjR47M1KZNm9anx37//fcztR133LFPjxGDY489ttUtQNW7M8jSpUsbXjts2LDiGkEpOjs7g/Wtt946WP/CF76QqX3sYx/LsyWgMl555ZVg/a233iq5kzi8/vrrrW4BfbD22msH6ytWrMjUPvzww6LbQR9xRg8AAAAAAEAkGPQAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARKLXizGb2eaSrpf0CUkfSpro7j81s8GSfitpmKR5ko5y9/8prlWgXKlkf4899ijkcT//+c8H6yeccEKmtmjRoj499kMPPdRQrUjPPPNMsP7Xv/61tB6KvBhzKvkHVpVK9o844ohg/ayzzsrUNthgg6LbQRtIJfv1DB48OFM74IADgmvfeOONTI2L0f4/a6wRPpdg0KBBmVroeW+F1PPfrEsvvTRYX3PN7Lhh9OjRRbcDNXZGzwpJp7n79pL2kPQdM9tB0pmSprv7CEnTa18DMSH7SBn5R6rIPlJF9pEy8o+o9DrocfeF7v5U7fNlkuZI2lTSoZKm1JZNkfTlopoEWoHsI2XkH6ki+0gV2UfKyD9i06dr9JjZMEm7Snpc0ibuvlDq/sWQtHGdnxlvZl1m1rVkyZLmugVahOwjZeQfqSL7SBXZR8rIP2LQ8KDHzNaXdLOk77v7m43+nLtPdPdR7j6qo6OjPz0CLUX2kTLyj1SRfaSK7CNl5B+xaGjQY2ZrqTvwv3b3W2rlRWY2tPb9oZIWF9Mi0DpkHykj/0gV2UeqyD5SRv4Rk0buumWSrpM0x92v7PGt2ySNk3RJ7eO0QjpEn/z3f/93sP7aa6+V3En1pZL9Rx99tNUttEUPfbXzzjsH67Nnzy65k2Kkkv88dD9ViEUq2R84cGCwHrorTuy++MUvZmobbbRRCzpprVSyX0+919AhI0aMyNTmzp2bZzuVtvXWWwfrL7zwQsmdNC71/PfFd7/73Uzt8ssvD6796U9/2vDj1rsb18KFCzO1yy67LLj2nXfeaXh7set10CNpb0nHSJptZjNrtbPVHfYbzewESS9LOrKYFoGWIftIGflHqsg+UkX2kTLyj6j0Ouhx94ck1fvnyi/k2w7QPsg+Ukb+kSqyj1SRfaSM/CM2fbrrFgAAAAAAANoXgx4AAAAAAIBIMOgBAAAAAACIRCMXY0aFHHHEEcH6jTfemKl1dHQU3Q4QrVmzZrW6BQDo1R577NGn9TNnzszUttlmm+Daenfuagdvv/12pvb8888H1w4dOjRTO/nkk3PvCfFo57tHAUU7/vjjM7Xly5cH1/7zP/9zpvazn/0suHbp0qXB+iuvvJKp/eIXvwiuPfzwwzO1wYMHB9fGjjN6AAAAAAAAIsGgBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACASXIwZAIAmTJo0KVN7//33g2tPPPHEotsBPmLq1KnB+pQpU4L1008/PVOrd9HLESNG9L+xgr366quZWmjfJOmQQw7J1L7//e/n3hMAxOp73/tesO7umdpVV10VXHvcccc1vL1BgwYF69dcc03Da2PHGT0AAAAAAACRYNADAAAAAAAQCQY9AAAAAAAAkWDQAwAAAAAAEAkGPQAAAAAAAJHgrlsAADTg2GOPDdYfeuihTO3hhx8Orm2Hu27tvfferW4BJRo2bFif1i9YsCBTq3cXuXYW6jm0b5L0f//v/y26HQBI0imnnNLwWu52mC/O6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIRK+DHjPb3MzuM7M5ZvacmZ1Sq59vZq+a2czaf4cU3y5QHrKPlJF/pIrsI1VkHykj/4hNIxdjXiHpNHd/ysw2kDTDzO6pfe8n7n55ce2hr/793/89WN9www1L7iQKZB8pI/+r+Nd//ddg/bjjjiu5k+a0wwWh2xzZR6rIPlJG/ktS7wLNfblwM3rX66DH3RdKWlj7fJmZzZG0adGNAa1G9pEy8o9UkX2kiuwjZeQfsenTNXrMbJikXSU9XiudbGazzGyymXHKCKJF9pEy8o9UkX2kiuwjZeQfMWh40GNm60u6WdL33f1NSddKGi5ppLqnn1fU+bnxZtZlZl1LlizJoWWgXGQfKSP/SBXZR6rIPlJG/hGLhgY9ZraWugP/a3e/RZLcfZG7f+DuH0r6F0m7h37W3Se6+yh3H9XR0ZFX30ApyD5SRv6RKrKPVJF9pIz8IyaN3HXLJF0naY67X9mjPrTHsq9Iejb/9oDWIftIGflHqsg+UkX2kTLyj9g0ctetvSUdI2m2mc2s1c6WNNbMRkpySfMkfbOQDiO1ePHiQh53o402KuRxE0X2kTLy36B6d+NCZSWR/QkTJvSpXjU77LBDpvbss/z9rBdJZB+og/wjKo3cdeshSRb41p35twO0D7KPlJF/pIrsI1VkHykj/4hNn+66BQAAAAAAgPbFoAcAAAAAACASDHoAAAAAAAAiwaAHAAAAAAAgEgx6AAAAAAAAIsGgBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACASDHoAAAAAAAAiYe5e3sbMlkh6qfblEElLS9t4uWLeN6ka+7elu3e0uomVEsq+FPf+VWXf2jX/VXn++ivm/avKvrVr9qXqPIf9EfO+SdXYP7LfGjHvm1Sd/WvX/Ffl+euvmPevKvtWN/ulDno+smGzLncf1ZKNFyzmfZPi37+ixf78xbx/Me9bGWJ//mLev5j3rSwxP4cx75sU//4VLebnL+Z9k+Lfv6LF/vzFvH8x7Btv3QIAAAAAAIgEgx4AAAAAAIBItHLQM7GF2y5azPsmxb9/RYv9+Yt5/2LetzLE/vzFvH8x71tZYn4OY943Kf79K1rMz1/M+ybFv39Fi/35i3n/Kr9vLbtGDwAAAAAAAPLFW7cAAAAAAAAiUfqgx8wONrM/m9lcMzuz7O3nzcwmm9liM3u2R22wmd1jZi/UPm7Yyh77y8w2N7P7zGyOmT1nZqfU6lHsX9nIfnWQ/fyR/+og//ki+9VB9vNF9quD7OeP/FdHrPkvddBjZgMk/ULSFyXtIGmsme1QZg8F6JR08Cq1MyVNd/cRkqbXvq6iFZJOc/ftJe0h6Tu1/1+x7F9pyH7lkP0ckf/KIf85IfuVQ/ZzQvYrh+zniPxXTpT5L/uMnt0lzXX3v7j7e5J+I+nQknvIlbs/IOn1VcqHSppS+3yKpC+X2lRO3H2huz9V+3yZpDmSNlUk+1cysl8hZD935L9CyH+uyH6FkP1ckf0KIfu5I/8VEmv+yx70bCrplR5fz6/VYrOJuy+UuoMjaeMW99M0MxsmaVdJjyvC/SsB2a8osp8L8l9R5L9pZL+iyH7TyH5Fkf1ckP+Kiin/ZQ96LFDjtl9tzszWl3SzpO+7+5ut7qeiyH4Fkf3ckP8KIv+5IPsVRPZzQfYriOznhvxXUGz5L3vQM1/S5j2+3kzSgpJ7KMMiMxsqSbWPi1vcT7+Z2VrqDvyv3f2WWjma/SsR2a8Ysp8r8l8x5D83ZL9iyH5uyH7FkP1ckf+KiTH/ZQ96npQ0wsy2MrO1JX1V0m0l91CG2ySNq30+TtK0FvbSb2Zmkq6TNMfdr+zxrSj2r2Rkv0LIfu7If4WQ/1yR/Qoh+7ki+xVC9nNH/isk1vybe7lnkZnZIZKukjRA0mR3/1GpDeTMzKZK2k/SEEmLJE2Q9DtJN0raQtLLko5091UvXtX2zGwfSQ9Kmi3pw1r5bHW/Z7Hy+1c2sl8dZD9/5L86yH++yH51kP18kf3qIPv5I//VEWv+Sx/0AAAAAAAAoBhlv3ULAAAAAAAABWHQAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNDTCzMbZGbfLvDxNzKz+8xsuZn9vKjtBLZ7v5mNCtQHm9k9ZvZC7eOGZfWE9lJC9r9mZjN7/PehmY0sans9thvMfu173zWzP5vZc2b246J7QXviuM9xP2Xkn/ynitc9vO5JVQnZX8vMppjZbDObY2ZnFbWtVbZbN/u17//AzNzMhpTRT9kY9PRukKRg8M1sQA6P/46kcyX9oNkHMrM1m29HZ0qa7u4jJE2vfY00FZp9d/+1u49095GSjpE0z91n9uex8si+mX1e0qGSdnb3HSVd3uxjorI47iNl5B+p4nUPUlX0cf9ISeu4+06SdpP0TTMb1p8Hyum4LzPbXNIBkl7O4/HaEYOe3l0iaXht8n6Zme1X+5eoGyTNNrNhZvbsysW1yeD5tc+Hm9ldZjbDzB40s+1WfXB3f8vdH1L3C5+6zGyemV1qZk/U/vtUrd5pZlea2X2SLjWzj5nZZDN70syeNrNDa+vWM7PfmNksM/utpPXqbOpQSVNqn0+R9OW+PFmISqHZX8VYSVND3ygx+/8g6RJ3f1eS3H1xw88UYsNxHykj/0gVr3uQqqKz75I+Zt1DmvUkvSfpzVUXlZh9SfqJpNNrvUUpl4lY5M6U9Ona9F1mtp+k3Wu1F23108iJkr7l7i+Y2f8n6RpJ+5vZGEmj3P28PvbyprvvbmbfkHSVpNG1+jaS/tbdPzCziyTd6+7Hm9kgSU+Y2X9I+qakt919ZzPbWdJTKx/UzCZJ+md375K0ibsvlCR3X2hmG/exR8SjzOz/vbpfbNdTRva3kbSvmf1I3X8B+YG7P7n6pwiR4riPlJF/pIrXPbzuSVXR2f93ded9oaSBkk5199frPF7h2a/19qq7P2NmfXumKoRBT/884e4vrm6Bma0vaS9JN/UI0DqS5O63QtJXrAAAIABJREFUSbqtH9ud2uPjT3rUb3L3D2qfHyhpjJmtPCV6XUlbSPqspKtr259lZrNW/rC7n9iPXpCm3LNf+0PhbXd/VvWVkf01JW0oaQ9Jn5F0o5lt7e7RTvrRJxz3kTLyj1TxugepyjP7u0v6QNIn1Z25B83sP9z9L4GHLTT7ZjZQ0jm1x4gag57+eavH5yv00bfArVv7uIakN1ZORnPidT7v2Y9JOtzd/9zzB2u/fI0cuBeZ2dDav2oNlcRpnOipiOx/VXVOX+6hjOzPl3RL7QXOE2b2oaQhkpY08LOIH8d9pIz8I1W87kGq8sz+0ZLucvf3JS02s4cljZIUGvQUnf3hkraStPJsns0kPWVmu7v7a738bKVwjZ7eLZO0wWq+v0jSxtZ9F4l1VDu9zN3flPSimR0pSdZtlyZ7+fseHx+ts+aPkr5rteSa2a61+gOSvlarfVr/P3v3HiZHVed//PMVE8QE5JIEwiVMhHAJJNyGgKJsWAwgsoAil7iyXJQAAVd4QAz83MUbyhpBWRHXLISER0mQDWgWlBCQXUQQCLdcjEAWBojEMIAYLmKEfH9/TGd3SJ3O9HRXdXed8349D8/0fOdM1beaT2p6zlTXkcZW+f65kk6qPD5J0s8a7BnlVXj2zexd6rlB2+w+emlG9n8q6W8r43aSNFDSi330hThx3kfKyD9SxeseXvekqujsP6uet3OZmQ1Sz1Vkv6uyr0Kz7+6L3H2Yu3e4e4d6Jjz3jm2SR2Kip0/u/pKkX5vZYjObGvj6XyV9VdL9km7RO0P795I+Y2aPSVqiyntxzexIM/vq2kFm1iXpckknm9lyMxtdpZ0Nzex+SZ+XdG6VMV+TNEDSQuu5adbXKvUfSBpcuYTtAkkP9Nr/1fZ/S89dKmmCmT2pnjuRX1plP4hcM7Kvnkssl1e5dLO3ZmR/uqT3V753tqSTuHw5TZz3Oe+njPyT/1TxuofXPalqQva/L2mwpMWSHpR0rbsvVFgzsp8E499zOVReFHW6OzPtSArZR6rIPlJG/pEqso9Ukf18NXRFj5kdZmaPm9kyM5uSV1NAGZB/pIrsI1VkH6ki+0gZ+UcZ1X1Fj5ltIOkJ9Vzmulw9l2FNdPffVvueIUOGeEdHR137A/qjq6tLL774YmHr5fU3/2QfzfTQQw+96O5Di9g25360syLP/WQf7YzsI2W87kGq1nfub2TVrXGSlq19j6mZzVbPe/Kqhr6jo0MLFixoYJdAbTo7C38LZr/yT/bRTGb2TIGb59yPtlXwuZ/so22RfaSM1z1I1frO/Y28dWsbSc/1+nx5pQakgPwjVWQfqSL7SBXZR8rIP0qpkYme0CVCmfeBmdkkM1tgZgu6u7sb2B3QVvrMP9lHpDj3I1VkH6ki+0gZ+UcpNTLRs1zSdr0+31bS8+sOcvdp7t7p7p1Dhxby1kmgFfrMP9lHpDj3I1VkH6ki+0gZ+UcpNTLR86CkUWY20swGSjpB0tx82gLaHvlHqsg+UkX2kSqyj5SRf5RS3Tdjdve3zOxsSfMkbSBpursvya0zoI2Rf6SK7CNVZB+pIvtIGflHWTWy6pbc/eeSfp5TL0CpkH+kiuwjVWQfqSL7SBn5Rxk18tYtAAAAAAAAtBEmegAAAAAAACLR0Fu3UL9x48Zlag8++GDD27388suD9XPPPbfhbQMAAAAoxgUXXBCsT506NVM7+eSTg2OvvfbaPFsCGrL99ttnas8++2y/tjFr1qxM7YQTTqi7p1RwRQ8AAAAAAEAkmOgBAAAAAACIBBM9AAAAAAAAkWCiBwAAAAAAIBJM9AAAAAAAAESCVbcKtt9++wXrjzzySCH7M7NCtgsAaF+vvPJKsD5ixIhMbdWqVUW3AwBYjwsvvDBYv/rqq2vexk033RSsDxgwIFObNm1azdsF6rHFFlsE63/84x8ztbvuuis4dsyYMcH64MGD628sYVzRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkmOgBAAAAAACIBDdjLli1G2TOmTMnUxs7dmzD+/v+978frH/3u9/N1M4555yG9wcUqaibi7t7IdsFWuVDH/pQsL7rrrs2uRMAQF/+8Ic/BOvHH398zfV58+YFxy5atKj+xoA6vfzyy8H6vffem6nts88+wbEDBw7MtafUcUUPAAAAAABAJJjoAQAAAAAAiAQTPQAAAAAAAJFgogcAAAAAACASTPQAAAAAAABEoqFVt8ysS9Krkt6W9Ja7d+bRVFmNHz8+U3vmmWeCY7feeutMraOjo+Ee3n777WD9T3/6U8PbxjuR//rksZJWf1bNCu2vWg+sxlUbsl8e8+fPb3ULUSH7SBn5L9773//+YL0/v2M89thjebYEkf1aLFy4MFjfZZddMrUBAwYU3Q6Uz/LqB7n7izlsBygj8o9UkX2kiuwjZeQfqSL7KBXeugUAAAAAABCJRid6XNLtZvaQmU0KDTCzSWa2wMwWdHd3N7g7oK2sN/9kHxHj3I9UkX2kjNc9SBXnfpROoxM9B7j73pI+KuksMztw3QHuPs3dO929c+jQoQ3uDmgr680/2UfEOPcjVWQfKeN1D1LFuR+l09A9etz9+crHF8zsZknjJN2dR2Nl9Lvf/S5Tu+6664JjQzemQrmQ//wUeRPk0LbzuCF0ysh+a1177bWZ2sqVK1vQSXrIft+uuuqqYP2ss85qah9nnnlmplatN9SG/LcXFpBoHrLftzFjxrS6Bayj7it6zGyQmW289rGkQyQtzqsxoJ2Rf6SK7CNVZB8pI/9IFdlHWTVyRc+Wkm6u/GX83ZKud/fbcukKaH/kH6ki+0gV2UfKyD9SRfZRSnVP9Lj7U5L2yLEXoDTIP1JF9pEqso+UkX+kiuyjrFheHQAAAAAAIBJM9AAAAAAAAESioVW30Le99torWB88eHDD27788ssztWrL+R133HEN7w/oj3Ze2araShWhnlnVAu3mhhtuyNS+9KUvBcdutNFGRbeDhF122WWZ2ve///0WdJJ18803Z2rvelf475tXXnll0e0A73DooYcG61tttVXN29htt92C9WOOOaaunoBGVMv0j370o0yN5eebgyt6AAAAAAAAIsFEDwAAAAAAQCSY6AEAAAAAAIgEEz0AAAAAAACRYKIHAAAAAAAgEqy6VWKLFy/O1MaOHRscu8MOOxTdDlATVrECanPxxRcH6/fee2+mFlqFUZIGDBiQa09Ab//zP/+TqY0cOTI49vjjj8/Uqq3QdckllwTrG2+8cc29/fKXv8zUlixZUvP3A0U64YQTGt7Gvvvu2686UKTbb789WH/zzTeb3AnW4ooeAAAAAACASDDRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkuBlzCVx55ZXB+j333JOpVbsZM1AkM8vUuOkyYjN06NBgvbu7u6Htfv3rX+9Xfc2aNQ3tD+ivH/3oR8H6I488kqnttddewbGTJk3K1AYNGhQce+qppwbr1caH7LLLLpnarbfeGhz7zW9+M1O78MILa94XAKTu2muvDdZDN9f/y1/+0q9tn3HGGZnafvvt169tpIgregAAAAAAACLBRA8AAAAAAEAkmOgBAAAAAACIBBM9AAAAAAAAkehzosfMppvZC2a2uFdtczObb2ZPVj5uVmybQGuQf6SK7CNVZB8pI/9IFdlHbGpZdWuGpCslXderNkXSne5+qZlNqXz+xfzbgyTdddddwfq4ceMytfHjxxfcTXJmiPwjTTNE9t9h9erVhWz3S1/6UrD+2muvBetXXHFFIX3gf80Q2X+Ha665Jlj/zW9+k6lVW3Vr5MiRmVq17Odh//33z9TefPPN4NiDDjooU0t41a0ZIv+lsGrVqmD91VdfzdS22WabotuJwQyR/bqdfPLJNY/9p3/6p2B9+fLlwfqKFSsytVGjRgXHHn744ZnaRz/60Zp7i0mfV/S4+92SXl6nfJSkmZXHMyUdnXNfQFsg/0gV2UeqyD5SRv6RKrKP2NR7j54t3X2FJFU+DsuvJaDtkX+kiuwjVWQfKSP/SBXZR2kVfjNmM5tkZgvMbEF3d3fRuwPaBtlHysg/UkX2kSqyj5SRf7Sbeid6VprZcEmqfHyh2kB3n+bune7eOXTo0Dp3B7SVmvJP9hEhzv1IFdlHynjdg1Rx7kdp1XIz5pC5kk6SdGnl489y6yhh06dPD9aXLFkSrH/xi9l7ge2555659oSgZPNvZq1uoVDunqlVO+bQ2AQkm/1WuPTSS4P1Rx99tMmdQGQ/aPjw4Zna1ltv3YJOUDDy34bmzp0brN98882Z2pw5c4puJ1Zkv0GhmzSvWbMmOHblypXB+uzZszO1efPmBccOGTIkU+NmzFWY2SxJ90na2cyWm9ln1BP2CWb2pKQJlc+B6JB/pIrsI1VkHykj/0gV2Uds+ryix90nVvnSwTn3ArQd8o9UkX2kiuwjZeQfqSL7iE3hN2MGAAAAAABAczDRAwAAAAAAEAkmegAAAAAAACJR76pbKMCsWbOC9dGjRwfrY8eOLbIdoGaJrkCFxEyaNKnVLUiSjjzyyEztF7/4RXDsjjvumKkNHDgw954Qj9WrV2dq73pX+O+C//zP/5ypnXHGGbn3BCDr+OOPD9aPOeaYJncC9M+pp54arFf7Xfjoo4/O1D72sY8Fx37oQx+qv7HIcEUPAAAAAABAJJjoAQAAAAAAiAQTPQAAAAAAAJFgogcAAAAAACAS3Iw5R6EbS/3nf/5ncGzoZpj77LNPcOzEiROD9T322KMf3QEAGjF16tRWtyBJmjx5cqZW7ab9p512WqbGzZixPjvvvHOm9p3vfCc49uCDDy66HQBVDBgwoF91oF3ccMMNwfpjjz0WrB977LGZWrXfm/F/uKIHAAAAAAAgEkz0AAAAAAAARIKJHgAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiwapbOfrGN76RqYVWr5CkJ554IlO74447gmNZXQsAsD7vehd/t0E+urq6MrVNN900OHbjjTcuuBsAkjR79uxMLfRvVZKmTJlScDdA7W688cZM7Z577gmOPf3004P13XffPdeeUsErQwAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiwUQPAAAAAABAJPqc6DGz6Wb2gpkt7lX7spn93swerfx3eLFtAq1B/pEqso9UkX2kjPwjVWQfsall1a0Zkq6UdN069e+4+7dz76jEbrnllkxt1apVwbEHHXRQpjZ06NDce0LDZoj8I00zRPbb0kknnZSphVZyRN1miOwjXTNE/tvOc889l6k9/vjjLegkajNE9us2Z86cYH3y5MmZ2u233x4cy+pa+erzih53v1vSy03oBWg75B+pIvtIFdlHysg/UkX2EZtG7tFztpktrFzmtlluHQHlQP6RKrKPVJF9pIz8I1VkH6VU70TPDyTtIGlPSSskXVZtoJlNMrMFZragu7u7zt0BbaWm/JN9RIhzP1JF9pEyXvcgVZz7UVp1TfS4+0p3f9vd10j6d0nj1jN2mrt3unsn96BBDGrNP9lHbDj3I1VkHynjdQ9SxbkfZVbLzZgzzGy4u6+ofPpxSYvXNz4Vn/3sZzO1UaNGBcf+27/9W6a200475d5TO3njjTeC9QceeCBTGz9+fMHd1I/8I1Vkvz2sXLkyU3v00UeDYzfZZJOi20kC2UfKyH9trrjiimD9kEMOydR23XXX4NjQwi6S9Prrr2dqxx9/fD+6Qz3Ifu0++clPBusHHnhgprbxxhsX3Q5Uw0SPmc2SNF7SEDNbLuliSePNbE9JLqlL0ukF9gi0DPlHqsg+UkX2kTLyj1SRfcSmz4ked58YKF9TQC9A2yH/SBXZR6rIPlJG/pEqso/YNLLqFgAAAAAAANoIEz0AAAAAAACRYKIHAAAAAAAgEnWtupW6++67L1jfY489MrWrr746OHa77bbLtacy+P3vfx+sn3DCCZnaH/7wh6LbAQAAAPot9Pr+sssuC45duHBhpjZ69Ojg2FdffTVY7+zszNQOO+yw9bUItIUf/ehHmVqKvwe3Alf0AAAAAAAARIKJHgAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiwc2Y6/DBD34wWF+5cmWmNmzYsKLbaYpVq1YF688880zN26g29q9//WumtmjRouDYMWPG1Lw/ACizhx56KFgP/awBAORv1qxZwfpFF12UqXV3dwfHTp8+veb9TZw4MVjfeeedM7Xf/OY3wbEjRozI1P77v/87OHbQoEGZ2pFHHrm+FoEk/fnPfw7Wly1bVvM2mv17LFf0AAAAAAAARIKJHgAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiwUQPAAAAAABAJFh1K0fPPfdcpvbGG2+0oJP83XbbbcH6mWeeGaxvuOGGmdrw4cODYzfZZJNMrdod/59++ulqLQJAVK666qpgfZdddsnUhgwZUnQ7AJCcT33qUzWP3WuvvYL1973vfZna7373u+DYaqt8heqHHXZYcOz48eMztSlTpgTHbrXVVpla6HV5te0CfVm+fHmmtuWWWwbHDhw4sOH9vfLKKzXV+qva6lqnnXZazdto9u+xXNEDAAAAAAAQCSZ6AAAAAAAAIsFEDwAAAAAAQCSY6AEAAAAAAIhEnzdjNrPtJF0naStJayRNc/crzGxzSTdI6pDUJek4d/9jca22v87Ozla3UJhqN8fafPPNg/W99947U5s/f36uPRWN7NfOzDI1d29BJ8gL+W+en/70p8H6qFGjgvUTTzwxUxs2bFiuPaWM7CNVZL92HR0dmdrMmTODY8eMGZOpXXjhhcGxN910U809PPXUUzXXd9ppp5q3e/rppwfrjz/+eM3bKCPy35hqvxMecMABmdovf/nL4NjQv5X++ta3vlVTTQovHjRo0KDg2D322CNYb+eFgmq5ouctSee5+66S9pd0lpmNljRF0p3uPkrSnZXPgZiQfaSM/CNVZB+pIvtIGflHVPqc6HH3Fe7+cOXxq5KWStpG0lGS1k5dz5R0dFFNAq1A9pEy8o9UkX2kiuwjZeQfsenXPXrMrEPSXpLul7Slu6+Qev5hSApeN25mk8xsgZkt6O7ubqxboEXIPlJG/pEqso9UkX2kjPwjBjVP9JjZYElzJJ3j7qtq/T53n+bune7eOXTo0Hp6BFqK7CNl5B+pIvtIFdlHysg/YlHTRI+ZDVBP4H/s7mvvErbSzIZXvj5c0gvFtAi0DtlHysg/UkX2kSqyj5SRf8SkllW3TNI1kpa6++W9vjRX0kmSLq18/FkhHaItHHHEEcH6nDlzmtxJ85D9rGoraYVW3QrV8tpff7AiWH3If9YHPvCBYH3kyJGZ2uc///ng2HPOOSdT+/3vfx8ce/HFFwfr22yzTbUWkQOyj1SR/drdddddmVpoJa5qvvnNb/arjuKR/8a89NJLwfr222+fqR100EFFt1OTj3/845narFmzWtBJMfqc6JF0gKQTJS0ys0crtYvUE/afmNlnJD0r6dhiWgRahuwjZeQfqSL7SBXZR8rIP6LS50SPu98jqdqf5g/Otx2gfZB9pIz8I1VkH6ki+0gZ+Uds+rXqFgAAAAAAANoXEz0AAAAAAACRYKIHAAAAAAAgErXcjBnrYMUeIF9Fra7V322zQhfW5ze/+U2w/sADD2Rqt956a3DsqlWrMrULLrggOPbTn/50P7oDAOSN1wBAY5555plWt5AsrugBAAAAAACIBBM9AAAAAAAAkWCiBwAAAAAAIBJM9AAAAAAAAESCmzEDaFizb1bITZPRCmQMAAAAZcAVPQAAAAAAAJFgogcAAAAAACASTPQAAAAAAABEgokeAAAAAACASDDRAwAAAAAAEAlW3QJQOkWtfsSqSgAQhz/84Q+Z2vLly4Nj999//6LbAQCgqbiiBwAAAAAAIBJM9AAAAAAAAESCiR4AAAAAAIBIMNEDAAAAAAAQiT5vxmxm20m6TtJWktZImubuV5jZlyWdJqm7MvQid/95UY0CzUb2kTLyj1SR/TjcdNNNmdqNN94YHHvfffcV3U4pkH2kjPwjNrWsuvWWpPPc/WEz21jSQ2Y2v/K177j7t4trD2gpso+UkX+kiuwjVWQfKSP/iEqfEz3uvkLSisrjV81sqaRtim4MaDWyj5SRf6SK7CNVZB8pI/+ITb/u0WNmHZL2knR/pXS2mS00s+lmtlmV75lkZgvMbEF3d3doCND2yD5SRv6RKrKPVJF9pIz8IwY1T/SY2WBJcySd4+6rJP1A0g6S9lTP7Odloe9z92nu3ununUOHDs2hZaC5yD5SRv6RKrKPVJF9pIz8IxY1TfSY2QD1BP7H7n6TJLn7Snd/293XSPp3SeOKaxNoDbKPlJF/pIrsI1VkHykj/4hJLatumaRrJC1198t71YdX3ssoSR+XtLiYFoHWIPtIGflHqlLPvru3uoVcTJ48uaYa/k/q2UfayD9iU8uqWwdIOlHSIjN7tFK7SNJEM9tTkkvqknR6IR0CrUP2kTLyj1SRfaSK7CNl5B9RqWXVrXskWeBLP8+/HaB9kH2kjPwjVWQfqSL7SBn5R2z6teoWAAAAAAAA2hcTPQAAAAAAAJFgogcAAAAAACASTPQAAAAAAABEgokeAAAAAACASDDRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkzN2btzOzbknPVD4dIunFpu28uWI+Nqkcx7e9uw9tdRNrJZR9Ke7jK8uxtWv+y/L81Svm4yvLsbVr9qXyPIf1iPnYpHIcH9lvjZiPTSrP8bVr/svy/NUr5uMry7FVzX5TJ3resWOzBe7e2ZKdFyzmY5PiP76ixf78xXx8MR9bM8T+/MV8fDEfW7PE/BzGfGxS/MdXtJifv5iPTYr/+IoW+/MX8/HFcGy8dQsAAAAAACASTPQAAAAAAABEopUTPdNauO+ixXxsUvzHV7TYn7+Yjy/mY2uG2J+/mI8v5mNrlpifw5iPTYr/+IoW8/MX87FJ8R9f0WJ//mI+vtIfW8vu0QMAAAAAAIB88dYtAAAAAACASDR9osfMDjOzx81smZlNafb+82Zm083sBTNb3Ku2uZnNN7MnKx83a2WP9TKz7czsLjNbamZLzOzzlXoUx9dsZL88yH7+yH95kP98kf3yIPv5IvvlQfbzR/7LI9b8N3Wix8w2kPR9SR+VNFrSRDMb3cweCjBD0mHr1KZIutPdR0m6s/J5Gb0l6Tx331XS/pLOqvz/iuX4mobslw7ZzxH5Lx3ynxOyXzpkPydkv3TIfo7If+lEmf9mX9EzTtIyd3/K3VdLmi3pqCb3kCt3v1vSy+uUj5I0s/J4pqSjm9pUTtx9hbs/XHn8qqSlkrZRJMfXZGS/RMh+7sh/iZD/XJH9EiH7uSL7JUL2c0f+SyTW/Dd7omcbSc/1+nx5pRabLd19hdQTHEnDWtxPw8ysQ9Jeku5XhMfXBGS/pMh+Lsh/SZH/hpH9kiL7DSP7JUX2c0H+Syqm/Dd7oscCNZb9anNmNljSHEnnuPuqVvdTUmS/hMh+bsh/CZH/XJD9EiL7uSD7JUT2c0P+Syi2/Dd7ome5pO16fb6tpOeb3EMzrDSz4ZJU+fhCi/upm5kNUE/gf+zuN1XK0RxfE5H9kiH7uSL/JUP+c0P2S4bs54bslwzZzxX5L5kY89/siZ4HJY0ys5FmNlDSCZLmNrmHZpgr6aTK45Mk/ayFvdTNzEzSNZKWuvvlvb4UxfE1GdkvEbKfO/JfIuQ/V2S/RMh+rsh+iZD93JH/Eok1/+be3KvIzOxwSd+VtIGk6e5+SVMbyJmZzZI0XtIQSSslXSzpp5J+ImmEpGclHevu6968qu2Z2Yck/UrSIklrKuWL1POexdIfX7OR/fIg+/kj/+VB/vNF9suD7OeL7JcH2c8f+S+PWPPf9IkeAAAAAAAAFKPZb90CAAAAAABAQZjoAQAAAAAAiAQTPQAAAAAAAJFgogcAAAAAACASTPQAAAAAAABEgomePpjZpmY2ucDt/72ZPdrrvzVmtmdR++u13/8ys84qX/ucmT1uZkvM7FtF94L21ITsb2Fmd5nZa2Z2ZVH7Cew3mH0z29zM5pvZk5WPmzWrJ7SXJmR/gJnNNLNFZrbUzC4sal/r7Lda9m/o9TOoy8webUY/aE9F57+yjwvNbFnltcahRe6r1z6r5f/YyuudNdVeFyENTTj3j+t1rn3MzD5e1L7W2S/nfqxXE7I/0MyurbzueczMxhe1r3X2m/Tvu0z09G1TScHgm9kGjW7c3X/s7nu6+56STpTU5e51nWjN7N2N9mNmB0k6StJYd99N0rcb3SZKq9DsS3pT0j9JOr/RDeWRfUlTJN3p7qMk3Vn5HGkqOvvHStrQ3cdI2kfS6WbWUc+G8si+ux/f6+fQHEk3NbpNlFqh+Tez0ZJOkLSbpMMkXVXvdnM69y+W9AlJd+ewLZRb0ef+xZI6K+fawyT9sN4Mc+5HzorO/mmSVHndM0HSZWZW1zwEv+/Wjomevl0qaYfKbPdUMxtfuQrhekmLzKzDzBavHWxm55vZlyuPdzCz28zsITP7lZnt0se+JkqaFfpCZab9X8zsgcp/O1bqM8zscjO7S9K/mNkgM5tuZg+a2SNmdlRl3EZmNtvMFprZDZI2qtLDmZIudfe/SJK7v1DzM4XYFJp9d3/d3e9Rz4RPVU3M/lGSZlYez5R0dH+eLESl6PO+SxpUebGykaTVklatO6iJ2V+7P5N0nKr8HEIyis7/UZJmu/tf3P1pScskjVt3ULPy7+5L3f3xhp4xxKLo1z1vuPtblU/fo56fBRmc+9ECRZ/3R6vnj6hrf7d8RVLoKjN+381RHn8Jid0USbtXZrtlPZeajavUnrb1/xV2mqQz3P1JM9tP0lWS/tbMjlTPjP4/rzP+ePW8AKpmlbuPM7N/kPRdSUdU6jtJ+oi7v20FCYxiAAAgAElEQVRm35D0S3c/1cw2lfSAmd0h6XRJb7j7WDMbK+nhtRs1s6sl/Zu7L6hs68Nmdol6fgE/390fXP9ThEg1M/t9aUb2t3T3FZLk7ivMbFg/e0Q8is7+f6jnXL9C0nslnevuL1fZXjOyv9aHJa109ydrepYQq6Lzv42k3/T6nuWVWkgz8w8U/rqn8rXpkraXdGKviZ91ce5HMxWd/cckHWVmsyVtp56rmbeT9EBge/y+mxMmeurzQOWvUFWZ2WBJH5R0Y89EuSRpQ0ly97mS5q4zfj/1BHOxqpvV6+N3etVvdPe3K48PkXSkma19O8x7JI2QdKCkf63sf6GZLVz7ze7+2V7berekzSTtL2lfST8xs/e7e/CvDkhO7tmvUTOyD6xPntkfJ+ltSVur53z7KzO7w92fCmy2mdmvelUpkpdn/i373eErG8S5H62X6+sed79f0m5mtqukmWb2C3cPXdnMuR+tlmf2p0vaVdICSc9IuldStUlOft/NCRM99Xm91+O39M63wL2n8vFdkl5ZOzNagxPU90nWqzzu3Y9JOmbdy5Ar//hqCe9ySTdVgv6Ama2RNERSdw3fi/gVkf1aNCP7K81seOVqnuGSoryME3XLM/ufknSbu/9V0gtm9mv1XMIcmuhpRvbXvuf9E+r5Kxuwrjzzv1w9f8lda1tJz1cZ25T8A+tRyOsed19qZq9L2l09v/xmhlR5zLkfzZJb9itXrp279nMzu1dStSvI+H03J9yjp2+vStp4PV9fKWmY9awgtKEql5e5+ypJT5vZsVLP+1/NbI/QBqznZlTHSprdRy/H9/p4X5Ux8yR9zipJN7O9KvW7Jf19pba7pLFVvv+nkv62Mm4nSQMlvdhHX4hT4dnvh2Zkf66kkyqPT5L0swZ7RnkVnf1n1XNZs5nZIPX8Rel3VfbVjOxL0kck/c7dl69nDNJQdP7nSjrBzDY0s5GSRil8+b7UvPwDUsHZN7ORlYkVmdn2knaW1FVlX5z70UxFZ/+9ldc7MrMJkt5y999W2Re/7+aEiZ4+uPtLkn5tZovNbGrg63+V9FVJ90u6Re98sf73kj5jZo9JWqLK/XfM7Egz+2qvcQdKWl7lsv3eNjSz+yV9Xr1mRdfxNUkDJC20nptmfa1S/4GkwZVL2C5QrxdVZna1/d/Sc9Mlvb/yvbMlnRTbZWyoTTOyb2Zdki6XdLKZLbee1VhCmpH9SyVNMLMn1bMiwKVV9oPINSH735c0WD0rsDwo6Vp3X6iwZmRfqu2qUiSg6Py7+xJJP5H0W0m3STqr1+X46yo8/2b2cTNbLukDkm41s3nVnx3ErAnn/g9Jesx6ljG/WdJkd6/2yyXnfjRNE7I/TNLDZrZU0hfVs9J0Nfy+mxOL8JiiVPmFuHM9PxCAKJF9pIrsI2XkH6ki+0gV2c9XQ1f0mNlhZva4mS0zsyl5NQWUAflHqsg+UkX2kSqyj5SRf5RR3Vf0mNkGkp5Qz1sclqvn8vOJ63m/nYYMGeIdHR117Q/oj66uLr344ouhlT1y0d/8k30000MPPfSiuw8tYtuc+9HOijz3k320M7KPlPG6B6la37m/kVW3xklatva+MmY2Wz3vyasa+o6ODi1YELqxPJCvzs7Ovgc1pl/5J/toJjN7psDNc+5H2yr43E/20bbIPlLG6x6kan3n/kbeurWNpOd6fb68UgNSQP6RKrKPVJF9pIrsI2XkH6XUyERP6BKhzPvAzGySmS0wswXd3VEtTY+09Zl/so9Ice5Hqsg+UkX2kTLyj1JqZKJnuaTten2+raTn1x3k7tPcvdPdO4cOLeStk0Ar9Jl/so9Ice5Hqsg+UkX2kTLyj1JqZKLnQUmjzGykmQ2UdIKkufm0BbQ98o9UkX2kiuwjVWQfKSP/KKW6b8bs7m+Z2dmS5knaQNJ0d1+SW2dAGyP/SBXZR6rIPlJF9pEy8o+yamTVLbn7zyX9PKdegFIh/0gV2UeqyD5SRfaRMvKPMmrkrVsAAAAAAABoI0z0AAAAAAAARIKJHgAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiwUQPAAAAAABAJJjoAQAAAAAAiAQTPQAAAAAAAJF4d6sbAAAAAPoyderUTO2CCy5oag/u3tT9IS1mVvNYsghgfbiiBwAAAAAAIBJM9AAAAAAAAESCiR4AAAAAAIBIMNEDAAAAAAAQCW7GXLDdd989WF+yZEnN27juuuuC9RNPPLGungAAAFrtc5/7XLB+/fXXB+t//vOfa972hAkTMrVZs2bV/P1Au9tiiy2C9ZdeeqnJnQBoR1zRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkmOgBAAAAAACIBBM9AAAAAAAAkWho1S0z65L0qqS3Jb3l7p15NFVWoRW2li5dGhx76623Zmr/+q//Ghw7efLkYD20+sSkSZPW1yJyRP5bZ88998zUzjvvvOBYVqfLH9nP1x133BGsh1YN6q+999675v1tttlmDe8vdmS/fqeeemqmduONNwbHVluNqz+vcTbaaKNMrdoqRagN+c/PyJEj+zX+6aefbngbqB/Zb62pU6dmahdccEFw7Mc+9rFM7ZZbbsm9pzLIY3n1g9z9xRy2A5QR+UeqyD5SRfaRMvKPVJF9lApv3QIAAAAAAIhEoxM9Lul2M3vIzHjPEFJD/pEqso9UkX2kjPwjVWQfpdPoW7cOcPfnzWyYpPlm9jt3v7v3gMo/hkmSNGLEiAZ3B7SV9eaf7CNinPuRKrKPlPG6B6ni3I/SaWiix92fr3x8wcxuljRO0t3rjJkmaZokdXZ2eiP7axfVbn727LPPZmq33XZbcOyBBx6YqXV2hu/rdcYZZwTrb775ZrUW0QR95T/G7Och9G/iK1/5Sr+2sWzZskzta1/7WnDsn/70p0zt7LPP7tf+8E6pnvvzELr57NVXX92vbXzhC1/I1Lq6uoJj77zzzkztrLPOCo4999xzM7V99923X73Fjuz3beLEicH64YcfnqmFMidJw4cPD9aHDBlSf2NoGK978lPtnL1w4cJgvaOjo+axyB/n/nxVu+H+9ddfH6yHFiCqZv78+ZlatcUtQmNjUvdbt8xskJltvPaxpEMkLc6rMaCdkX+kiuwjVWQfKSP/SBXZR1k1ckXPlpJuNrO127ne3cOXrwDxIf9IFdlHqsg+Ukb+kSqyj1Kqe6LH3Z+StEeOvQClQf6RKrKPVJF9pIz8I1VkH2XF8uoAAAAAAACRYKIHAAAAAAAgEo0ur56k/qxucsABBwTHbrjhhpnasGHDgmM33njj2psD2sQNN9wQrD/yyCOZ2umnn97w/qZNmxas//CHP8zUqq1Yd/755zfcB9KzevXqTO32228Pjp05c2amtmjRouDYk08+ueZ6tZ9LgwYNCtZD/uu//itTe+utt4JjP/CBD9S8XcTrE5/4RKZ2wgknBMeGVj3ZbLPNcu8JaDeh8+W9994bHDtmzJiat9ufsUA7ee2114L1l19+OVifNGlSpnbhhRcGx951112Z2g9+8IN+dBcPrugBAAAAAACIBBM9AAAAAAAAkWCiBwAAAAAAIBJM9AAAAAAAAESCiR4AAAAAAIBIsOpWjnbddddM7ZRTTgmOnTJlSqY2duzY3HsCmmH27NmZ2lVXXRUc29HRkaldeumlDfdQbXWh0J3258+fHxzLqluox+uvv56pXXzxxcGxTz31VKb2sY99LDj22muvrbmH0aNHB+uHH354zds47rjjMrVly5YFx7LqFqTwKorVVkJhhS2k6itf+UqmxjkUqZg8eXKm9tOf/jQ49txzzw3WL7jggkxtq622Co795Cc/makdeOCB62sxWlzRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkmOgBAAAAAACIBDdjztHEiRMztdNPPz04dsSIETVv9wtf+EKwvskmm9S8DaBIoZvGvvLKK8Gx++67byE9HHvsscH6M888k6lVuxkzULTQz4RqNx8EyuiLX/xisP7tb387U9t7772LbgdouUMOOaTVLQCF+8xnPhOs33TTTZna2WefHRx75plnBuvVbrwcsvHGG9dUSwFX9AAAAAAAAESCiR4AAAAAAIBIMNEDAAAAAAAQCSZ6AAAAAAAAIsFEDwAAAAAAQCT6XHXLzKZLOkLSC+6+e6W2uaQbJHVI6pJ0nLv/sbg228t//Md/1Dx2woQJwXp/Vsy66667gvWRI0dmav1ZzQt9I//1+4d/+Idg/XOf+1xT+xgyZEim1tHR0dQeyojsZz344IPB+rRp0zK1U045JTj2b/7mbzK1LbfcsrHGcjJ58uRM7b777guOrXZ8oRX4yobs127q1KmZWrWVQkOrcW277bbBsccdd1yw/tGPfrQf3aEe5D9/ofNlDOfK2JD9xixatChYP++88zK1U089NTh26623zrWntR5//PFg/ZJLLsnUrrvuukJ6aIVaruiZIemwdWpTJN3p7qMk3Vn5HIjRDJF/pGmGyD7SNENkH+maIfKPNM0Q2UdE+pzocfe7Jb28TvkoSTMrj2dKOjrnvoC2QP6RKrKPVJF9pIz8I1VkH7Gp9x49W7r7CkmqfBxWbaCZTTKzBWa2oLu7u87dAW2lpvyTfUSIcz9SRfaRMl73IFWc+1Fahd+M2d2nuXunu3cOHTq06N0BbYPsI2XkH6ki+0gV2UfKyD/aTZ83Y65ipZkNd/cVZjZc0gt5NtXujjnmmKbub8GCBcG6uze1D/yvZPO/evXqYH2DDTaoqdYKJ598ck011CTZ7EtSV1dXsB66Qf8ZZ5wRHDtmzJg8W8rV+PHjM7UnnngiOHbGjBnBesQ3GE06+9V88pOfzNTWrFkTHPvVr341U7vjjjuCY5988slgPfQz6Kijjlpfi8gH+W9A6Hw5ePDghrf7ve99r+FtoE9kP+DMM8/M1Kqdt/fee+9MraibLlezatWqYP1Xv/pVU/totnqv6Jkr6aTK45Mk/SyfdoBSIP9IFdlHqsg+Ukb+kSqyj9Lqc6LHzGZJuk/Szma23Mw+I+lSSRPM7ElJEyqfA9Eh/0gV2UeqyD5SRv6RKrKP2PT51i13n1jlSwfn3AvQdsg/UkX2kSqyj5SRf6SK7CM2hd+MGQAAAAAAAM3BRA8AAAAAAEAk6l11C0CCdt5552B96tSpmdohhxxSdDtAYV566aVM7Y033giO3X777TO1gQMH5t4TUAbHHXdcsB5aeWjJkiXBsb/+9a+D9bfffjtT++tf/xocG1oRDGiFb3zjG5naRRdd1PA2gFa57bbbMrV//Md/DI4dO3Zs0e2gCq7oAQAAAAAAiAQTPQAAAAAAAJFgogcAAAAAACASTPQAAAAAAABEgpsxA6hZV1dXsP6+970vU9tkk00K7qZ+Dz/8cLB+4YUXZmrz5s0ruh20ocsuuyxTW7FiRXDs3LlzM7Wtttoq956AMgjddFmSnnrqqZq3cdBBBwXr+++/fz0tAaUReh2yvjrQLo444ohgfdttt21yJ1iLK3oAAAAAAAAiwUQPAAAAAABAJJjoAQAAAAAAiAQTPQAAAAAAAJFgogcAAAAAACASrLoFoGHu3uoW+uXFF18M1m+//fYmd4J29dJLL2Vqr7/+enDsiBEjim4HaEszZ87M1L785S8Hx44dOzZTmzBhQnDs0UcfHawffPDBtTcHtImLLrooUzv77LODY7/xjW8U3Q6ARHBFDwAAAAAAQCSY6AEAAAAAAIgEEz0AAAAAAACRYKIHAAAAAAAgEkz0AAAAAAAARKLPVbfMbLqkIyS94O67V2pflnSapO7KsIvc/edFNQm0CvlHqsg+UkX2axdaYWvMmDHBsV//+tcztdBKXGgt8t8c3/ve91rdAtZB9hGbWq7omSHpsED9O+6+Z+U/Ao9YzRD5R5pmiOwjTTNE9pGuGSL/SNMMkX1EpM+JHne/W9LLTegFaDvkH6ki+0gV2UfKyD9SRfYRm0bu0XO2mS00s+lmtlm1QWY2ycwWmNmC7u7uasOAsukz/2QfkeLcj1SRfaSM1z1IFed+lFK9Ez0/kLSDpD0lrZB0WbWB7j7N3TvdvXPo0KF17g5oKzXln+wjQpz7kSqyj5Txugep4tyP0urzZswh7r5y7WMz+3dJt+TWUcLuvPPOYH348OHB+q677lpkO6iC/GeZWatb6Jey9dsuYsx+V1dXsL7hhhtmakOGDCm4G7SrGLOfh8MOy97O4rzzzguO3XHHHYtuBwUh/435xCc+0eoWUCeyH3bSSSdlasOGDWtBJ1ifuq7oMbPeMw8fl7Q4n3aA9kf+kSqyj1SRfaSM/CNVZB9lVsvy6rMkjZc0xMyWS7pY0ngz21OSS+qSdHqBPQItQ/6RKrKPVJF9pIz8I1VkH7Hpc6LH3ScGytcU0AvQdsg/UkX2kSqyj5SRf6SK7CM2jay6BQAAAAAAgDbCRA8AAAAAAEAk6lp1C8W44oorgvVjjz02WJ8wYUKR7QBA9A466KBgPbRy0PHHH190O0Bbuv3224P1yy+/PFPbaKONim4HKJU5c+a0ugWgLjNmzAjWzz///Ext8ODBBXeD/uKKHgAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiwUQPAAAAAABAJLgZcwksWrQoWA/dHHGHHXYIjq1WB4CUdXV1BesDBw7M1IYOHVpwN0B7OvTQQ4P1p59+OlPr6OgouBugXO67775M7QMf+EALOgH655RTTgnWx48fn6lxM+b2wxU9AAAAAAAAkWCiBwAAAAAAIBJM9AAAAAAAAESCiR4AAAAAAIBIMNEDAAAAAAAQCVbdKtijjz4arP/5z3/O1F5++eXg2KlTp9ZcP/XUU4NjP/vZz2Zq1VaQ2XHHHYN1AACQnjFjxgTroVW3ttxyy+DYjTbaKNee8rR69epM7fHHHw+OrfZcANV88IMfzNTcvbD9VVutN4Q8Y30OOeSQYL2dz+f4P1zRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkmOgBAAAAAACIRJ83Yzaz7SRdJ2krSWskTXP3K8xsc0k3SOqQ1CXpOHf/Y3GtltOnP/3pYH3JkiWZ2vbbbx8c258bpd1yyy3B+vTp0zO1v/u7vwuOveSSSzK1QYMGBce+//3vr7m3siH7SBn5R6rIftbChQuD9ZEjR2ZqV1xxRXDs2LFjc+0pT8uXL8/UPvzhDwfHFnkT3VYj+83T1dVV2Lb7828t5jz3F/nPmjdvXrD+8MMPZ2rVflccPHhwrj2hdrVc0fOWpPPcfVdJ+0s6y8xGS5oi6U53HyXpzsrnQEzIPlJG/pEqso9UkX2kjPwjKn1O9Lj7Cnd/uPL4VUlLJW0j6ShJMyvDZko6uqgmgVYg+0gZ+UeqyD5SRfaRMvKP2PTrHj1m1iFpL0n3S9rS3VdIPf8wJA3LuzmgXZB9pIz8I1VkH6ki+0gZ+UcMap7oMbPBkuZIOsfdV/Xj+yaZ2QIzW9Dd3V1Pj0BLkX2kjPwjVWQfqSL7SBn5RyxqmugxswHqCfyP3f2mSnmlmQ2vfH24pBdC3+vu09y90907hw4dmkfPQNOQfaSM/CNVZB+pIvtIGflHTGpZdcskXSNpqbtf3utLcyWdJOnSysefFdJhpLbaaqtM7ZprrgmOPfjgg2ve7ne/+91gPbQKxqJFi4JjP/KRj2RqI0aMCI698cYbg/UNNtggU9tuu+2CY9sV2a9d2VZtKFu/rUD+kSqyX7tNN900UzvllFOCY1977bVMbfXq1bn3VI+e/+XvtPnmm7egk9Yi+8UIZWmfffYJjn355ZcL2R/6Rv5rd8wxx2RqP/nJT4Jj991336Lbqdtf/vKXTK3a78f9WQW7XfQ50SPpAEknSlpkZo9WahepJ+w/MbPPSHpW0rHFtAi0DNlHysg/UkX2kSqyj5SRf0Slz4ked79HUvZPHT1qv9QEKBmyj5SRf6SK7CNVZB8pI/+ITb9W3QIAAAAAAED7YqIHAAAAAAAgEkz0AAAAAAAARKKWmzGjAZtsskmwHlpha7/99mt4f+ecc06/6iHz5s3L1D71qU8Fx+61117B+nve855MbfHixcGxW2yxRc29AXnYeuutg/WTTz65uY2g5a699tpg/YknnsjUrrrqquDYyZMnN9TDb3/722B96tSpwfoXvvCFTG306NEN9QDU45FHHql57BFHHJGp3XrrrXm2U7ftt98+U3v66adb0Ali9NJLL9U8NrQCXJH7A/KyYsWKYL2rq6vmbWy22WbB+vve9756WvpfAwcODNY33HDDTO3II48Mji3jzwSu6AEAAAAAAIgEEz0AAAAAAACRYKIHAAAAAAAgEkz0AAAAAAAARIKbMRfs3nvvbXUL/XbooYdmatVu7FbtJqK77bZbpjZkyJDgWHfvR3doR3ncPLAozz//fKb2rneF57ir3ZgX8ap2A+7TTz89U7vjjjuCY9/73vc21MMrr7zS0PcXaaeddgrWuXE5+uuWW25pdQtA2+M1Mcpg0003zdROOeWU4NjXXnstU1u9enVw7Le+9a1gPbQIRX/ssccewXoZb7DcH1zRAwAAAAAAEAkmegAAAAAAACLBRA8AAAAAAEAkmOgBAAAAAACIBBM9AAAAAAAAkWDVLTRk9OjRwTqrBqBdXH/99Zna/Pnzg2PnzZtXdDsosQcffDBYr7bSRK0+8pGPBOvVctpM48eP71cdAADE7ZFHHql57BFHHJGp3XrrrXm2gyq4ogcAAAAAACASTPQAAAAAAABEgokeAAAAAACASDDRAwAAAAAAEIk+b8ZsZttJuk7SVpLWSJrm7leY2ZclnSapuzL0Inf/eVGNAs1G9rPKeJPt888/v6Ya3in1/P/whz+sqYb4pJ59pIvsI2Xkvxi33HJLq1tIVi2rbr0l6Tx3f9jMNpb0kJmtXQrkO+7+7eLaA1qK7CNl5B+pIvtIFdlHysg/otLnRI+7r5C0ovL4VTNbKmmbohsDWo3sI2XkH6ki+0gV2UfKyD9i06979JhZh6S9JN1fKZ1tZgvNbLqZbZZzb0DbIPtIGflHqsg+UkX2kTLyjxjUPNFjZoMlzZF0jruvkvQDSTtI2lM9s5+XVfm+SWa2wMwWdHd3h4YAbY3sI2XkH6ki+0gV2UfKyD9iUdNEj5kNUE/gf+zuN0mSu69097fdfY2kf5c0LvS97j7N3TvdvXPo0KF59Q00BdlHysg/UkX2kSqyj5SRf8Skz4keMzNJ10ha6u6X96oP7zXs45IW598e0DpkHykj/0gV2UeqyD5SRv4Rm1pW3TpA0omSFpnZo5XaRZImmtmeklxSl6TTC+kQaB2yj5SRf6SK7CNVZB8pI/+ISi2rbt0jyQJf+nn+7QDtg+wjZeQfqSL7SBXZR8rIP2LTr1W3AAAAAAAA0L6Y6AEAAAAAAIgEEz0AAAAAAACRYKIHAAAAAAAgEkz0AAAAAAAARIKJHgAAAAAAgEgw0QMAAAAAABAJJnoAAAAAAAAiYe7evJ2ZdUt6pvLpEEkvNm3nzRXzsUnlOL7t3X1oq5tYK6HsS3EfX1mOrV3zX5bnr14xH19Zjq1dsy+V5zmsR8zHJpXj+Mh+a8R8bFJ5jq9d81+W569eMR9fWY6tavabOtHzjh2bLXD3zpbsvGAxH5sU//EVLfbnL+bji/nYmiH25y/m44v52Jol5ucw5mOT4j++osX8/MV8bFL8x1e02J+/mI8vhmPjrVsAAAAAAACRYKIHAAAAAAAgEq2c6JnWwn0XLeZjk+I/vqLF/vzFfHwxH1szxP78xXx8MR9bs8T8HMZ8bFL8x1e0mJ+/mI9Niv/4ihb78xfz8ZX+2Fp2jx4AAAAAAADki7duAQAAAAAARIKJHgAAAAAAgEg0faLHzA4zs8fNbJmZTWn2/vNmZtPN7AUzW9yrtrmZzTezJysfN2tlj/Uys+3M7C4zW2pmS8zs85V6FMfXbGS/PMh+/sh/eZD/fJH98iD7+SL75UH280f+yyPW/Dd1osfMNpD0fUkflTRa0kQzG93MHgowQ9Jh69SmSLrT3UdJurPyeRm9Jek8d99V0v6Szqr8/4rl+JqG7JcO2c8R+S8d8p8Tsl86ZD8nZL90yH6OyH/pRJn/Zl/RM07SMnd/yt1XS5ot6agm95Ard79b0svrlI+SNLPyeKako5vaVE7cfYW7P1x5/KqkpZK2USTH12Rkv0TIfu7If4mQ/1yR/RIh+7ki+yVC9nNH/ksk1vw3e6JnG0nP9fp8eaUWmy3dfYXUExxJw1rcT8PMrEPSXpLuV4TH1wRkv6TIfi7If0mR/4aR/ZIi+w0j+yVF9nNB/ksqpvw3e6LHAjXWd29zZjZY0hxJ57j7qlb3U1Jkv4TIfm7IfwmR/1yQ/RIi+7kg+yVE9nND/ksotvw3e6JnuaTten2+raTnm9xDM6w0s+GSVPn4Qov7qZuZDVBP4H/s7jdVytEcXxOR/ZIh+7ki/yVD/nND9kuG7OeG7JcM2c8V+S+ZGPPf7ImeByWNMrORZjZQ0gmS5ja5h2aYK+mkyuOTJP2shb3UzcxM0jWSlrr75b2+FMXxNRnZLxGynzvyXyLkP1dkv0TIfq7IfomQ/dyR/xKJNf/m3tyryMzscEnflbSBpOnufklTG8iZmc2SNF7SEEkrJV0s6aeSfiJphKRnJR3r7uvevKrtmdmHJP1K0iJJayrli9TznsXSH1+zkf3yIPv5I//lQf7zRfbLg+zni+yXB9nPH/kvj1jz3/SJHgAAAAAAABSj2W/dAgAAAAAAQEGY6AEAAAAAAIgEEz0AAAAAAACRYKIHAAAAAAAgEkz0AAAAAAAARIKJnj6Y2aZmNrnA7Y8zs0cr/z1mZh8val/r7Pe/zKwzUL+hVz9dZvZoM/pB+yk6+5V9XGhmy8zscTM7tMh99dpntewfa2ZLzGxN6OtIRxPO+1uY2V1m9pqZXaLXbmEAACAASURBVFnUfgL7rZb9zc1svpk9Wfm4WbN6QvtpQv4Hmtm1Zrao8rpnfFH7Wme/wfxXvva5ys+hJWb2rWb0g/ZD9sl+qnjdE+frHiZ6+rappGDwzWyDHLa/WFKnu+8p6TBJPzSzd9ezoXq/rzd3P97d96z0M0fSTY1uE6VVaPbNbLSkEyTtpp7sX1XvdvPIvnr+LX5C0t05bAvlVvR5/01J/yTp/EY3lFP2p0i6091HSbqz8jnSVXT+T5Mkdx8jaYKky8ysrtejeeTfzA6SdJSkse6+m6RvN7pNlBbZR6p43RMhJnr6dqmkHSpXuEw1s/GVGcnrJS0ysw4zW7x2sJmdb2ZfrjzewcxuM7OHzOxXZrbLuht39zfc/a3Kp++R5KEmKlfX/IuZPVD5b8dKfYaZXW5md0n6FzMbZGbTzexBM3vEzI6qjNvIzGab2UIzu0HSRus7aDMzScdJmtXP5wvxKDT76nlxMdvd/+LuT0taJmncuoOalX13X+rujzf0jCEWRZ/3X3f3e9TzwqeqJp73j5I0s/J4pqSj+/NkITpFn/tHq+eFtdz9BUmvSAr9xbVZ+T9T0qXu/pdePSFNZB+p4nVPhJjo6dsUSf9TucrlC5XaOEn/z91H9/G90yR9zt33Uc8M5lWSZGZHmtlX1w4ys/3MbImkRZLO6DXxs65V7j5O0pWSvturvpOkj7j7eZL+n6Rfuvu+kg6SNNXMBqnnZP6Gu4+VdImkfXrt/2rLXtb2YUkr3f3JPo4R8So6+9tIeq7X9yyv1EKamX2g8PN+PzQj+1u6+wpJqnwcVkefiEfR+X9M0lFm9m4zG6meXG5XZXvNyP9Okj5sZveb2X+b2b59HCPiRfaRKl73RCiPS59S9EDlCoSqzGywpA9KutHM1pY3lCR3nytp7tqiu98vaTcz21XSTDP7hbuHZjxn9fr4nV71G9397crjQyQdaWZrL417j6QRkg6U9K+V/S00s4W99v/ZwL4miqt5kJVn9i373eEr2tTc7AMhuZ73+4Hsox3kmf/pknaVtEDSM5LulVTtD1zNyP+79f/bu/twq+o67+Ofr6g0iiXIAZF0SJRKEzXppDV3mQ/4kGRiKuYDXHVHZkxJjJNajhpmeuUDMzrl5aSCmQ91KUTYJXF7U06FGiiPotJtlIwEmJaoqQHf+4+zmDmyfpuzz9l7rbPX7/d+XRfXOed7fmev79p+zmL7Ze21pP6SDpf0AUk/NLN93b3W30dIC9lHqnjdU3EMenrm1U6fb9Jbz4x6W/ZxB0l/zq51Uxd3X2lmr0p6nzr+EsgtqfF5535M0qnbvgUl++Wr68BtHe99HKtOU1Ag08zsr9Fb/yXrnZKer7G2lOwD21HIcb8OZWR/nZkNcfe1ZjZEEqfvY1tNy3921vLkrV+b2a8l1Tp7uIz8r5F0f/Y/t4+Z2RZJAyVtqONnET+yj1TxuqfieOtW1zZK2m07318naZB1XE28r6STJMndX5b0OzM7Teq45o2ZHbztD5vZu7LBiszs7yW9W9LqGts6o9PHBTXWzJX0j5Yl3cwOzeoPSzorq71P0sjt7NMxkp5y9zXbWYP4FZp9dUz5x5lZ3+wU5v0lPVZjW2VlH5CKz353lJH92ZLGZ5+Pl/TjBntGtRX9umeX7BR7mdmxkja5+5M1tlVG/mdJOipbN0LSzpJeqLEWcSP7ZD9VvO6JEIOeLrj7nyT9ysyWm9m3A9//m6RvSHpU0hxJT3X69lmSPmtmSyStUMeFn7Z9z+I/SFpiHbcxnynpfHevdZDta2aPSvqyOv2LwDamStpJ0lLruGjW1Kz+XUn9slPY/lmd/ofa8tcpGSfetpW8orPv7isk/VDSk5IelPTFTqdkbqvw7JvZKWa2RtIRkh4ws7m1nx3ErITjvsxstaTrJU0wszXWcRe6kDKO+1dLOtbMVqnjTjBX19gOElBC/gdJetzMVkr6qqRzttNOGfm/TdK+2c/eI2k8b11JE9kn+6nidU+cr3uM3+dqyH45Rm1nCAREiewjVWQfKSP/SBXZR6rIfnM1NOgxs+Ml/aukPpK+5+7bnYYNHDjQhw0b1uPtAfVavXq1XnjhhdDFfpumO/kn+yjTokWLXnD3tqIen2M/WlXRx36yj1ZF9pGyol/3oBwMepqrxxdjNrM+kv5dHac7rZH0GzObvZ33mmrYsGFauDB0jWGguUaNKvaO2d3NP9lHmczs9wU+Nsd+tKwij/1kH62M7CNlRb7uQXncfVhv9xCTRq7R0y7pt+7+rLu/qY73dp7cnLaAlkf+kSqyj1SRfaSK7ANAxTQy6Bkq6blOX6/Jam9hZhPNbKGZLdywgbv1IRpd5p/sI1Ic+5Eqso9UkX0AqJhGBj2h9wHnLvjj7re4+yh3H9XWxlsnEY0u80/2ESmO/UgV2UeqyD4AVEwjg541kvbu9PU7JT3fWDtAZZB/pIrsI1VkH6ki+wBQMY0Men4jaX8ze5eZ7SxpnKTZzWkLaHnkH6ki+0gV2UeqyD4AVEyP77rl7pvMbJKkueq41eJt7r6iaZ0BLYz8I1VkH6ki+0gV2QeA6unxoEeS3P2nkn7apF6ASiH/SBXZR6rIPlJF9gGgWhp56xYAAAAAAABaCIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBI7NjID5vZakkbJW2WtMndRzWjKaAKyD9SRfaRKrKPlJF/AKiOhgY9mY+5+wtNeBygish/N91888252oUXXhhc+8orrzS8vX333TdXu/baa4NrTznllIa3lxCyj1SRfaSM/ANABfDWLQAAAAAAgEg0OuhxST8zs0VmNrEZDQEVQv6RKrKPVJF9pIz8A0BFNPrWrQ+7+/NmNkjSPDN7yt0f7rwg+4tgoiTts88+DW4OaCnbzT/ZR8Q49iNVZB8p43UPAFREQ2f0uPvz2cf1kmZKag+sucXdR7n7qLa2tkY2B7SUrvJP9hErjv1IFdlHynjdAwDV0eMzesxsV0k7uPvG7PPRkr7RtM5Kdu+99wbrF1xwQa72xz/+seh2SnH11VcH61/96ldL7qR6Yst/o55//vlcbenSpcG1TzzxRK62adOm4Nr3vve9wfqIESPq3h6ai+znzZw5M1cbO3Zs3T9/+eWXB+uXXXZZT1tCAcg+Ukb+u3b88ccH63Pnzq37MT7zmc8E67feemuPegKQrkbeujVY0kwz2/o4d7n7g03pCmh95B+pIvtIFdlHysg/AFRIjwc97v6spIOb2AtQGeQfqSL7SBXZR8rIPwBUC7dXBwAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIhEIxdjjoq7N/wYO+yQn5uNHDkyuHbfffdteHshzz77bLAeuiPR8uXLg2vnzJmTqx18cPht2XvvvXc3ukOsli1blqt961vfCq4N5e6NN94Irj3uuOOC9c997nO52lVXXRVcu2DBgmAd6K477rgjWJ8+fXpDjztr1qxgvU+fPsH6hRdemKv17du3oR6Aol1xxRW5Wq07zg0ZMiRYX7RoUd1rgSoKvQaXpMmTJ+dqN9xwQ9HtAKgwzugBAAAAAACIBIMeAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEd93KmFnDj7HTTjvlaqeddlpw7ac//emGtxdy1113BesrV67M1Wpd2f/pp5/O1S699NLgWu66lZabbropWF+yZEmu9uSTTwbXDh06NFc7/fTTg2tHjx4drA8fPjxXGzNmTHDtoEGDcrVQxqXw/k2aNCm4Ful55JFHgvXQ3Q4nTJhQ9+MuXrw4WJ8xY0aw3tbWlqudcMIJwbX77LNP3X0AzTB16tRgvdbrk5CXXnopWB8/fnyuduWVVwbXtre31709oLsuueSSXK3WnULHjRtX9+M+/PDDwXrodTwAbA9n9AAAAAAAAESCQQ8AAAAAAEAkGPQAAAAAAABEgkEPAAAAAABAJJK7GHOti8kuWrQoWH/ttddytfe///3BtR/96EdztY985CPBtcOGDavRYWNqbe/888/P1X7xi18E1z733HO5Wuh5QHqmTJkSrG/ZsiVX27RpU3Dt2LFjc7VvfvObwbXveMc7gvU+ffrkaqeeempw7SGHHJKrff3rXw+unT17dq7GxZixlbsH66G/E26//fa6H3fatGnB+jXXXBOsf+ELX8jV7rvvvuBaLsaMsj322GPB+jPPPFP3Y7z++uvB+rx583K1gQMHBtdOnDgxVzvyyCPr7gHYntDrheuuuy64ttZFmkM2b94crP/oRz+q+zEAQOKMHgAAAAAAgGgw6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIRJeDHjO7zczWm9nyTrUBZjbPzFZlH/sX2ybQO8g/UkX2kSqyj5SRfwCIQz133Zou6SZJd3SqXSTpIXe/2swuyr7+avPba77u3DVICt856Oijjw6uvfLKK3O10N2BinTEEUcE6+3t7blarTsPff/7329qTxU3XRHlv1FvvvlmsH7ggQfmaieccEJwbej3p1+/fsG13fn92XHH8OEsVK91V4ta+5eo6SL7pZgwYUKwXuuOWWeffXaB3UBkv25z5szJ1davX19qD3fffXfda3fZZZdgPfQaKWHTRf7/W607GoYyM3jw4ML6WLduXa4WugudJB177LGF9QGgOro8o8fdH5b04jblkyXNyD6fIemTTe4LaAnkH6ki+0gV2UfKyD8AxKGn1+gZ7O5rJSn7OKh5LQEtj/wjVWQfqSL7SBn5B4CKKfxizGY20cwWmtnCDRs2FL05oGWQfaSM/CNVZB+pIvsA0Dp6OuhZZ2ZDJCn7WPMN2e5+i7uPcvdRbW1tPdwc0FLqyj/ZR4Q49iNVZB8p43UPAFRMPRdjDpktabykq7OPP25aRwWrdbHVkSNHBusf//jHc7XRo0cH1+688849b6xJal28NlTvzoVu3b3HPUWosvkvyogRI3K18847L7h2+PDhRbeD4iSdfTMr5HF33333YH3PPfcM1kPH7htvvDG4duPGjbnaueee243ukEk6+7WMGTMmV+vfP3xDpiFDhhTSw0svvRSshy7S/MorrwTXzp49u6k9RSjZ/H/qU58K1lesWJGrHXDAAQ1vr9bvz5///OdcbfLkycG1y5cvD9YBpKWe26vfLWmBpHeb2Roz+6w6DvTHmtkqScdmXwPRIf9IFdlHqsg+Ukb+ASAOXZ7R4+5n1vhW+B7jQETIP1JF9pEqso+UkX8AiEPhF2MGAAAAAABAORj0AAAAAAAARIJBDwAAAAAAQCR6etet6NS6Uv6XvvSlXK3WnVCqZpdddgnWQ3eAqXW3sj/96U/B+h577NHzxgCgBR1++OHB+l/+8pdCtjdo0KBg/ayzzsrVHnzwweDaRx55JFfjrlso0rRp04L1Y445ppDtTZgwIVifN29eIdsDijR27Nhgfdddd83VpkyZUnQ7ACqMM3oAAAAAAAAiwaAHAAAAAAAgEgx6AAAAAAAAIsGgBwAAAAAAIBJcjDlhJ554YrAeutj0ihUrgmvnz58frN9+++09bwwAWlDZFzHeb7/9gvWbb745V6t1AU93b2pPQFcGDBgQrO+1116FbO/KK68M1gcOHJirbdy4sZAeAABoNZzRAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNADAAAAAAAQCe66lbDDDjssWB86dGiuNnny5ODae+65J1jnrlsAAKBo7e3twfqIESNytUWLFhXdDgAALYEzegAAAAAAACLBoAcAAAAAACASDHoAAAAAAAAiwaAHAAAAAAAgEl0OeszsNjNbb2bLO9UuN7P/MrPF2Z8Ti20T6B3kH6ki+0gV2UfKyD8AxKGeu25Nl3STpDu2qd/g7tc2vSOgtUwX+UeapovsV56Z9XYLVTRdZL/yQtlfv359cO2cOXNytZNOOqnpPVXEdJH/StiyZUtvtwCghXV5Ro+7PyzpxRJ6AVoO+UeqyD5SRfaRMvIPAHFo5Bo9k8xsaXaKZ/9ai8xsopktNLOFGzZsaGBzQEvpMv9kH5Hi2I9UkX2kjNc9AFAhPR30fFfScEmHSFor6bpaC939Fncf5e6j2traerg5oKXUlX+yjwhx7EeqyD5SxuseAKiYHg163H2du2929y2S/kNSe3PbAloX+UeqyD5SRfaRMvIPANVTz8WYc8xsiLuvzb48RdLy7a1vJaNHjw7WR44cGaz37du3yHZQQVXOP9AIso9Ukf3q6devX662atWq4NoxY8bkau7e9J6qKuX8T5o0KVgfMGBAyZ0AQPd0Oegxs7slHSlpoJmtkXSZpCPN7BBJLmm1pM8X2CPQa8g/UkX2kSqyj5SRfwCIQ5eDHnc/M1C+tYBegJZD/pEqso9UkX2kjPwDQBwauesWAAAAAAAAWgiDHgAAAAAAgEgw6AEAAAAAAIhEj+66VWXXXnttsL777rsH67vttluR7QAAUBjuHIRUTZw4MVdra2sLrh0/fnzR7aCibrzxxt5uoabXXnstWP/Zz36Wq9W66zCAeHFGDwAAAAAAQCQY9AAAAAAAAESCQQ8AAAAAAEAkGPQAAAAAAABEIrmLMR900EG93QIAAAAKFLqZxoABA3qhE6AYv//974P14447LlfjwvxAejijBwAAAAAAIBIMegAAAAAAACLBoAcAAAAAACASDHoAAAAAAAAiwaAHAAAAAAAgEsnddQtA8z3zzDO52s033xxce/TRR+dqRx11VHDtzjvv3Fhjkvr375+rjR07Nrj2Pe95T8PbA1qJmfV2CwAAACgZZ/QAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARIJBDwAAAAAAQCS6vBizme0t6Q5Je0raIukWd/9XMxsg6V5JwyStlnS6u79UXKtAuch+Xq2LIz/99NO52ooVK4JrX3755Vytvb09uLZfv37d6C4s9Binn356w48bO/KPVJF9pIrs540cOTJYnzlzZq42fPjwotsBgLrVc0bPJklT3P29kg6X9EUzO0DSRZIecvf9JT2UfQ3EhOwjZeQfqSL7SBXZB4BIdDnocfe17v549vlGSSslDZV0sqQZ2bIZkj5ZVJNAbyD7SBn5R6rIPlJF9gEgHt26Ro+ZDZN0qKRHJQ1297VSx18MkgbV+JmJZrbQzBZu2LChsW6BXkL2kTLyj1SRfaSK7ANAtdU96DGzfpLuk3SBu+cvslGDu9/i7qPcfVRbW1tPegR6FdlHysg/UkX2kSqyDwDVV9egx8x2UscB/wfufn9WXmdmQ7LvD5G0vpgWgd5D9pEy8o9UkX2kiuwDQBzqueuWSbpV0kp3v77Tt2ZLGi/p6uzjjwvpsMnefPPNYH2HHcIzrz59+uRqHU9J9W3evDlY/9vf/par1drnWndhikFs2W+G6667LlhfsmRJrjZr1qzg2gULFuRqX/va14Jrd9yxy0NU4W688cbebqFXpJ7/1atX52p77LFHcO1uu+1WcDc95+693ULlpJ797hgyZEiu9uqrrwbXbty4MVdr5d+dFJH9vGXLlgXrb7zxRsmdAED31PN/UR+WdI6kZWa2OKtdoo6D/Q/N7LOS/iDptGJaBHoN2UfKyD9SRfaRKrIPAJHoctDj7r+UVOsUlqOb2w7QOsg+Ukb+kSqyj1SRfQCIR7fuugUAAAAAAIDWxaAHAAAAAAAgEgx6AAAAAAAAItH7t7Qp2ZQpU4L1ww47LFgfO3Zsrvb2t7+9qT31lgceeCBYnz9/fq520EEHBdd+6EMfampPaG2TJk0K1ufOnZurPfPMM8G1f/jDH3K1Bx98sFt9hO4k9OKLLwbX/vWvf83VBgwYEFy7yy675Gqp3nUrdR/72MdytalTpwbXnn322UW3A7SkhQsX5mpf+cpXgmvXrFmTq9V6TQa0iquuuipYHzRoUMmd1N/DBRdcUHInAFoRZ/QAAAAAAABEgkEPAAAAAABAJBj0AAAAAAAARIJBDwAAAAAAQCSSuxjzTTfdFKyPGTMmWD/wwANztf322y+4tn///j1vrEk2btwYrIcuVFvrArgzZ87M1aZNmxZce8YZZ3SjO8QqdLHuiy++OLj21VdfbXh7oQss33nnncG1S5YsydXOOeec4NoPfvCDjTWGaKxevTpXe+WVV8pvpEFm1tstIGJ77bVXrlbrGN8Kvz8f+MAHgvWf/OQnJXeCqqj1WqYVtLW1Beut3DOA8nBGDwAAAAAAQCQY9AAAAAAAAESCQQ8AAAAAAEAkGPQAAAAAAABEgkEPAAAAAABAJJK761YtixcvDtavuOKKXG3cuHHBtWeffXZTe+qJBQsWBOu33357rvbEE0/U/bju3uOeEL/QnVdCtWYJ3dXl17/+dXDtqlWrcrUjjjgiuPaUU05prDFEY8KECbnau9/97vIbadBTTz2Vq02fPj24NrTPQLP8/Oc/z9VGjBgRXHvmmWcW0sPgwYOD9ZNOOqmQ7QEA0Fs4owcAAAAAACASDHoAAAAAAAAiwaAHAAAAAAAgEl0OesxsbzObb2YrzWyFmX05q19uZv9lZouzPycW3y5QHrKPlJF/pIrsI1VkHwDiUc/FmDdJmuLuj5vZbpIWmdm87Hs3uPu1xbVXnueee67u+kEHHVR0Oz327LPPBuuzZs3K1V5//fXg2j333DNXM7PGGqumJLJfRbvuumuu9p3vfKcXOola0vkPXcC+iubPn19XTeJizJ0knf1Gtbe3B+t33nlnrnbXXXcF1xZ1MWZ0iewDQCS6HPS4+1pJa7PPN5rZSklDi24M6G1kHykj/0gV2UeqyD4AxKNb1+gxs2GSDpX0aFaaZGZLzew2M+vf5N6AlkH2kTLyj1SRfaSK7ANAtdU96DGzfpLuk3SBu78s6buShks6RB3T/+tq/NxEM1toZgs3bNjQhJaBcpF9pIz8I1VkH6ki+wBQfXUNesxsJ3Uc8H/g7vdLkruvc/fN7r5F0n9ICr4p291vcfdR7j6qra2tWX0DpSD7SBn5R6rIPlJF9gEgDvXcdcsk3Spppbtf36k+pNOyUyQtb357QO8h+0gZ+UeqyD5SRfYBIB713HXrw5LOkbTMzBZntUsknWlmh0hySaslfb6QDpvM3Xu7hUKdd9553apju6LKPtBN5L9C7r///t5uISZkvwGXXnppsL5ly5ZcbdGiRUW3g+4h+y3quOOOq6sGAFvVc9etX0oK3Vv7p81vB2gdZB8pI/9IFdlHqsg+AMSjW3fdAgAAAAAAQOti0AMAAAAAABAJBj0AAAAAAACRqOdizAAAAECPXXbZZb3dAgAAyeCMHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIBIMeAAAAAACASJi7l7cxsw2Sfp99OVDSC6VtvFwx75tUjf37e3dv6+0mtkoo+1Lc+1eVfWvV/Ffl+eupmPevKvvWqtmXqvMc9kTM+yZVY//Ifu+Ied+k6uxfS+UfaAWlDnresmGzhe4+qlc2XrCY902Kf/+KFvvzF/P+xbxvZYj9+Yt5/2Let7LE/BzGvG9S/PtXtJifv5j3TYp//4CY8dYtAAAAAACASDDoAQAAAAAAiERvDnpu6cVtFy3mfZPi37+ixf78xbx/Me9bGWJ//mLev5j3rSwxP4cx75sU//4VLebnL+Z9k+LfPyBavXaNHgAAAAAAADQXb90CAAAAAACIROmDHjM73syeNrPfmtlFZW+/2czsNjNbb2bLO9UGmNk8M1uVfezfmz32lJntbWbzzWylma0wsy9n9Sj2r2xkvzrIfvOR/+og/81F9quD7DcX2a8Osg/Ep9RBj5n1kfTvkk6QdICkM83sgDJ7KMB0ScdvU7tI0kPuvr+kh7Kvq2iTpCnu/l5Jh0v6YvbfK5b9Kw3Zrxyy30Tkv3LIf5OQ/coh+01C9iuH7AORKfuMnnZJv3X3Z939TUn3SDq55B6ayt0flvTiNuWTJc3IPp8h6ZOlNtUk7r7W3R/PPt8oaaWkoYpk/0pG9iuE7Dcd+a8Q8t9UZL9CyH5Tkf0KIftAfMoe9AyV9Fynr9dktdgMdve1UseBU9KgXu6nYWY2TNKhkh5VhPtXArJfUWS/Kch/RZH/hpH9iiL7DSP7FUX2gTiUPeixQI3bfrU4M+sn6T5JF7j7y73dT0WR/Qoi+01D/iuI/DcF2a8gst8UZL+CyD4Qj7IHPWsk7d3p63dKer7kHsqwzsyGSFL2cX0v99NjZraTOg74P3D3+7NyNPtXIrJfMWS/qch/xZD/piH7FUP2m4bsVwzZB+JS9qDnN5L2N7N3mdnOksZJml1yD2WYLWl89vl4ST/uxV56zMxM0q2SVrr79Z2+FcX+lYzsVwjZbzryXyHkv6nIfoWQ/aYi+xVC9oH4mHu5Z1Ga2YmSpknqI+k2d/9mqQ00mZndLelISQMlrZN0maRZkn4oaR9Jf5B0mrtve/G2lmdm/yDpPyUtk7QlK1+ijvfsVn7/ykb2q4PsNx/5rw7y31xkvzrIfnOR/eog+0B8Sh/0AAAAAAAAoBhlv3ULAAAAAAAABWHQAwAAAAAAEAkGPQAAAAAAAJFg0AMAAAAAABAJBj0AAAAAAACRYNDTBTPb3czOL3gbF5vZb83saTM7rshtddrmz81sVKB+mpmtMLMtoe8jHUVn38yONbNFZrYs+3hUUdvaZru1sj/AzOaZ2arsY/8y+kHrKSH7e5jZfDN7xcxuKmo7ge3Wyv69ZrY4+7PazBaX1RNaT4L559gPSUlmf6qZM/ZiDAAABRdJREFULc2O/T8zs73K6glA8Rj0dG13ScGDvpn1afTBzewASeMkHSjpeEnf6enjmtmOjfYjabmksZIebsJjodoKzb6kFySNcfeDJI2X9P2ePlCTsn+RpIfcfX9JD2VfI01FZ/91SZdK+qdGH6gZ2Xf3M9z9EHc/RNJ9ku5v9DFRaUnlXxz78T9Sy/633X1kduyfI+lfmvCYAFoEg56uXS1peDbt/raZHZlN4++StMzMhpnZ8q2LzeyfzOzy7PPhZvZgdrbCf5rZewKPf7Kke9z9DXf/naTfSmrfdlH2r6zXmNlj2Z/9svp0M7vezOZLusbMdjWz28zsN2b2hJmdnK37OzO7J5vc3yvp70I76+4r3f3php4xxKLQ7Lv7E+7+fPblCklvM7O+264rK/vq+F2ckX0+Q9Ine/CcIQ5FZ/9Vd/+lOl7011Ri9rduzySdLunu7jxZiE5q+efYj62Syr67v9zpy10leXeeLACtrRnT4NhdJOl92bRbZnakOgYx73P335nZsO387C2SznP3VWb2QUnfkXSUmX1C0ih3/xdJQyU90uln1mS1kJfdvd3MzpU0TdJJWX2EpGPcfbOZXSXp/7r7Z8xsd0mPmdn/kfR5Sa+5+0gzGynp8a0Pambfk3Szuy+s90lBEorOfmenSnrC3d+o8XhlZH+wu6+VJHdfa2aD6niOEKcys9+VMo/7/0vSOndf1c0eEZfU8s+xH1ulln2Z2TclnSvpL5I+1s0eAbQwBj0981h29k1NZtZP0ock/ajjH0klSX0lyd1nS5q9dWngx2tN1O/u9PGGTvUfufvm7PPRkj5hZltPC32bpH0kfUTSv2XbX2pmS/97Y+7/e3v7AnTSzOxvXX+gpGvUkd1ayD56W9OzX6cys3+mOJsHYSnkHwiJOvvu/jVJXzOziyVNknRZD3oF0IIY9PTMq50+36S3vgXubdnHHST9eeu/CmzHGkl7d/r6nZKer7HWa3zeuR+TdOq2b7/K/uLhlEw0qpnZl5m9U9JMSee6+//bztIysr/OzIZk/6I7RNL6On4G6Whq9ruhlOO+dVzvYaykw3rWJiIXc/459mN7Ys5+Z3dJekAMeoBocI2erm2UtNt2vr9O0iDruJJ+X2WnVmbve/2dmZ0mdVz7wMwODvz8bEnjzKyvmb1L0v6SHquxrTM6fVxQY81cSf9o2VHezA7N6g9LOiurvU/SyO3sEyAVnP3sNOMHJF3s7r/qopcysj9bHReFVvbxx130hHgVfdzvjrKO+8dIesrd1zTYL6ovtfxz7MdWSWXfzPbv9OUnJD3VYM8AWgiDni64+58k/crMlpvZtwPf/5ukb0h6VB1XrO98kDxL0mfNbIk6Lja79SJpnzCzb2Q/v0LSDyU9KelBSV/sdErmtvqa2aOSvixpco01UyXtJGmpdVwwbmpW/66kftnpm/+sTsMkM/ueZbddNLNTzGyNpCMkPWBmc2s/O4hZ0dlXxynC+0m61P7n1s61ro1QePbVcRHGY81slaRjs6+RoBKyLzNbLel6SRPMbI113IExpIzsSx13f+RtW0gx/xz7ISnN7Gf7ulQdbwP7co3tAKggc+fdPFWQ/cUwyt1f6O1egDKRfaSK7CNl5B+pIvsAmoEzegAAAAAAACLBGT0AAAAAAACR4IweAAAAAACASDDoAQAAAAAAiASDHgAAAAAAgEgw6AEAAAAAAIgEgx4AAAAAAIBIMOgBAAAAAACIxP8HtWrd8Amu2iIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x1440 with 29 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_misclssified(predictions, test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d62db7b3",
   "metadata": {},
   "source": [
    "### LeNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "2a84b23f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def LeNet():\n",
    "    model = tf.keras.Sequential([\n",
    "        tf.keras.layers.Conv2D(32, 3, activation = 'relu'),\n",
    "        tf.keras.layers.MaxPool2D(2),\n",
    "        tf.keras.layers.Conv2D(64, 3, activation = 'relu'),\n",
    "        tf.keras.layers.MaxPool2D(2),\n",
    "        tf.keras.layers.Dropout(0.25),\n",
    "        tf.keras.layers.Flatten(),\n",
    "        tf.keras.layers.Dense(128, activation = 'relu'),\n",
    "        tf.keras.layers.Dense(84, activation = 'relu'),\n",
    "        tf.keras.layers.Dropout(0.2),\n",
    "        tf.keras.layers.Dense(10, activation = 'sigmoid')\n",
    "    ])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "572ba549",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "891/891 [==============================] - 12s 13ms/step - loss: 0.2382 - accuracy: 0.9249\n",
      "Epoch 2/5\n",
      "891/891 [==============================] - 11s 12ms/step - loss: 0.0594 - accuracy: 0.9821\n",
      "Epoch 3/5\n",
      "891/891 [==============================] - 11s 12ms/step - loss: 0.0431 - accuracy: 0.9865\n",
      "Epoch 4/5\n",
      "891/891 [==============================] - 11s 12ms/step - loss: 0.0351 - accuracy: 0.9892\n",
      "Epoch 5/5\n",
      "891/891 [==============================] - 11s 12ms/step - loss: 0.0289 - accuracy: 0.9910\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2c9ddfe25e0>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LeNet = LeNet()\n",
    "LeNet.compile(optimizer = tf.keras.optimizers.SGD(learning_rate = 1e-1),\n",
    "                 loss = 'sparse_categorical_crossentropy',\n",
    "                 metrics=['accuracy'])\n",
    "\n",
    "LeNet.fit(train_images, train_labels, \n",
    "              batch_size = batch, \n",
    "              epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "52611d14",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_5\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_4 (Conv2D)            (None, 26, 26, 32)        320       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 13, 13, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 11, 11, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 5, 5, 64)          0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 5, 5, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 1600)              0         \n",
      "_________________________________________________________________\n",
      "dense_15 (Dense)             (None, 128)               204928    \n",
      "_________________________________________________________________\n",
      "dense_16 (Dense)             (None, 84)                10836     \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 84)                0         \n",
      "_________________________________________________________________\n",
      "dense_17 (Dense)             (None, 10)                850       \n",
      "=================================================================\n",
      "Total params: 235,430\n",
      "Trainable params: 235,430\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "LeNet.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "60dbc2e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 [==============================] - 1s 6ms/step - loss: 0.0313 - accuracy: 0.9913\n",
      "test accuracy : 0.9913333058357239 \t test loss: 0.03129655122756958\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_acc = cnn_model.evaluate(test_images, test_labels)\n",
    "print(f'test accuracy : {test_acc} \\t test loss: {test_loss}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e553587",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "MIT6S191",
   "language": "python",
   "name": "mit6s191"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}