03_Model-Training-DataAugmentation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 12:59:45.123588: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n",
      "To enable the following instructions: AVX2 AVX512F FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2024-02-10 12:59:46.220277: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'2.12.1'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os \n",
    "import tensorflow as tf\n",
    "\n",
    "# Importing matplotlib to plot images.\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "\n",
    "# Importing SK-learn to calculate precision and recall\n",
    "import sklearn\n",
    "import sklearn.metrics\n",
    "tf.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# !pip3 install git+https://github.com/pvigier/perlin-numpy\n",
    "import perlin_numpy #generate_perlin_noise_2d\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "N = 28"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7f2b16da0750>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "noiseMap  = (np.random.random((N,N))-.5) * .2\n",
    "plt.imshow(noiseMap)\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7f2b14c98490>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "noiseMap  = perlin_numpy.generate_perlin_noise_2d((N, N), (4, 4)) *.2\n",
    "plt.imshow(noiseMap)\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6.22 µs ± 34.3 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "np.random.random((N,N))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "220 µs ± 6.6 µs per loop (mean ± std. dev. of 7 runs, 1,000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "perlin_numpy.generate_perlin_noise_2d((N, N), (4, 4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# the data, split between train and test sets\n",
    "(x_train, y_train), (x_val, y_val) = tf.keras.datasets.mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000 train samples\n",
      "10000 test samples\n"
     ]
    }
   ],
   "source": [
    "shape = (28, 28, 1)\n",
    "# reshape image data\n",
    "x_train = x_train.reshape(-1, shape[0] *  shape[1] * shape[2])\n",
    "x_val = x_val.reshape(-1, shape[0] *  shape[1] * shape[2])\n",
    "\n",
    "# convert to float\n",
    "x_train = x_train.astype(np.float32)\n",
    "x_val = x_val.astype(np.float32)\n",
    "\n",
    "# normalize to a scale between -1 and 1\n",
    "x_train = x_train / 127.5 -1\n",
    "x_val = x_val / 127.5 -1\n",
    "\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_val.shape[0], 'test samples')\n",
    "\n",
    "# convert class vectors to binary class matrices (one-hot notation)\n",
    "num_classes = 10\n",
    "y_train_one_hot = tf.keras.utils.to_categorical(y_train, num_classes)\n",
    "y_val_one_hot = tf.keras.utils.to_categorical(y_val, num_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7f2b14cd9fd0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.random.seed(42)\n",
    "i = np.random.randint(0, len(x_train))\n",
    "plt.imshow(x_train[i].reshape(shape))\n",
    "plt.title(\"Label for image %i is: %s\" % (i, y_train[i]))\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## GPU and CPU settings\n",
    "If GPU is not available, comment out the bottom block."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "if True:\n",
    "    # Enter either 0 or 1 below to select the GPU to use\n",
    "    os.environ[\"CUDA_VISIBLE_DEVICES\"]=\"0\"\n",
    "\n",
    "    physical_devices = tf.config.experimental.list_physical_devices('GPU')\n",
    "    tf.config.set_visible_devices(physical_devices[0], 'GPU')\n",
    "    tf.config.experimental.set_memory_growth(physical_devices[0], True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Data Augmentation using a Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "class DataGenerator:\n",
    "    \n",
    "\n",
    "    def __init__(self, dataX, dataY, batch_size = 64, shuffle = True, isTrain=True):\n",
    "        self.batch_size = batch_size\n",
    "        self.shuffle = shuffle\n",
    "        self.isTrain = isTrain\n",
    "        self.dataX = dataX\n",
    "        self.dataY = dataY\n",
    "        self.n_batches = int(np.round(len(self.dataX)/batch_size))\n",
    "        self.indices = list(range(len(self.dataX)))\n",
    "        \n",
    "    def __call__(self, shuffle=True):\n",
    "    \n",
    "        if self.shuffle:\n",
    "            np.random.shuffle(self.indices)\n",
    "\n",
    "        for j in range(self.n_batches-1):\n",
    "            returnX = []\n",
    "            returnY = []\n",
    "            \n",
    "            # It would be much better to do this within the next loop, but its to slow.\n",
    "            noiseMap = perlin_numpy.generate_perlin_noise_2d((28, 28), (4, 4)) *.4\n",
    "            noiseMap = noiseMap.reshape(-1)\n",
    "            \n",
    "            for i in self.indices[j*batch_size:(j+1)*batch_size]:\n",
    "                \n",
    "                x = self.dataX[i]\n",
    "                y = self.dataY[i]\n",
    "                \n",
    "                \n",
    "                if self.isTrain:\n",
    "                    # Data Augmentation is typically only applyed to the train set to reduce overfitting.\n",
    "                    # If a test set is available it might be an idea to also use Data Augmentation for the validation set.\n",
    "                    \n",
    "                    ## A) One option is to do this per semple using 2D noise but as we seen this is slow.\n",
    "                    #noiseMap = perlin_numpy.generate_perlin_noise_2d\n",
    "                    ## B) using the same for each sampel is a cheapter option\n",
    "                    x += noiseMap\n",
    "                    None\n",
    "                    \n",
    "                #img = cv2.imread(f)\n",
    "                #img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
    "                \n",
    "                #if not((img.shape[0] == self.res[0]) & (img.shape[1] == self.res[1])):\n",
    "                #    img = cv2.resize(img, (self.res[1], self.res[0]))\n",
    "\n",
    "                # Shift Blob (img_A) and Template (img_B) using normal distribution\n",
    "                # stdev = 1 -> 99.73% samples within range +- 3 around mean = 0\n",
    "                #for shift_axis in [0, 1]:\n",
    "                #    shift_amount = np.random.randint(-1,2)\n",
    "                #    img_A = np.roll(img_A, shift_amount, axis = shift_axis)\n",
    "                #    img_B = np.roll(img_B, shift_amount*4, axis = shift_axis)\n",
    "\n",
    "                returnX.append(x)\n",
    "                returnY.append(y)\n",
    "\n",
    "            # We did this in the pre processing, could also be done here. \n",
    "            #returnX = np.array(returnX)/127.5 - 1. \n",
    "            \n",
    "            if self.isTrain:\n",
    "                if False:\n",
    "                    noiseMap = perlin_numpy.generate_perlin_noise_3d((len(returnX), 28, 28), (4, 4, 4))\n",
    "                    noiseMap = noiseMap.reshape(len(returnX), -1)\n",
    "                    returnX = np.array(returnX)\n",
    "                    returnX += noiseMap\n",
    "                    #returnX[returnX<-1] = -1\n",
    "                    #returnX[returnX>1] = 1\n",
    "\n",
    "            yield returnX, returnY"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 12:59:56.227279: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1635] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 29458 MB memory:  -> device: 0, name: Tesla V100-SXM2-32GB, pci bus id: 0000:62:00.0, compute capability: 7.0\n"
     ]
    }
   ],
   "source": [
    "batch_size = 32\n",
    "\n",
    "gen_train = DataGenerator(x_train, y_train_one_hot, batch_size=batch_size)\n",
    "gen_val = DataGenerator(x_val, y_val_one_hot, batch_size=batch_size, isTrain=False)\n",
    "\n",
    "train_dataset = tf.data.Dataset.from_generator(gen_train, \n",
    "                                         output_signature=(\n",
    "                                             tf.TensorSpec(shape=(batch_size, shape[0] *  shape[1] * shape[2]), dtype=tf.float32),\n",
    "         tf.TensorSpec(shape=(batch_size, 10), dtype=tf.float32)\n",
    "         )).prefetch(tf.data.AUTOTUNE).cache().repeat()\n",
    "\n",
    "val_dataset = tf.data.Dataset.from_generator(gen_val, \n",
    "                                         output_signature=(\n",
    "                                             tf.TensorSpec(shape=(batch_size, shape[0] *  shape[1] * shape[2]), dtype=tf.float32),\n",
    "         tf.TensorSpec(shape=(batch_size, 10), dtype=tf.float32))).prefetch(tf.data.AUTOTUNE).cache().repeat()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 12:59:56.904485: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n",
      "2024-02-10 12:59:56.904685: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n",
      "2024-02-10 12:59:56.952300: W tensorflow/core/kernels/data/cache_dataset_ops.cc:856] The calling iterator did not fully read the dataset being cached. In order to avoid unexpected truncation of the dataset, the partially cached contents of the dataset  will be discarded. This can happen if you have an input pipeline similar to `dataset.cache().take(k).repeat()`. You should use `dataset.take(k).cache().repeat()` instead.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLkAAAKfCAYAAACCHhrcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOydd5wV1fn/n5m7dyvLwtI7KM2uQbF3xBaV2JMYa6JJwPo1fiXfWGJi0MT+00QTe9TEiN1Ejb0FrNEEKyooiHRpu+zde++c3x+wyz6FvbOXuW338369eOkznJk5c+Yzzzn3MOcznnPOEQAAAAAAAAAAAAAAJYxf6AoAAAAAAAAAAAAAALCpYJILAAAAAAAAAAAAAJQ8mOQCAAAAAAAAAAAAACUPJrkAAAAAAAAAAAAAQMmDSS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDyY5AIAAAAAAAAAAAAAJQ8muQAAAAAAAAAAAABAyYNJLgAAAAAAAAAAAABQ8mCSKw/MnTuXPM+jq666KmPZSy+9lDzPy0OtQKkCPYGogaZAlEBPIEqgJxA10BSIEugJRA00telgkmsj3HnnneR5Hr311luFrkqkXH755eR5Hm299daFrkqXAnoCUdNZNPX+++/TMcccQ5ttthlVV1dT7969aa+99qLHH3+80FXrUnQWPRERvf3223TQQQdR9+7dqba2liZOnEjvvvtuoavVpegsekJ+Kh46i6YkGEcVhs6ipzfffJOmTJlCW221FdXU1NDQoUPp2GOPpU8++aTQVetydBZNvfjii+R5nvln5syZha5eaMoKXQGQP+bPn0+/+c1vqKamptBVAZ0A6AlEwRdffEGrV6+mk046iQYOHEiNjY304IMP0uGHH0633HILnX766YWuIigh3nnnHdpjjz1oyJAhdMkll1AQBPT73/+e9t57b3rjjTdozJgxha4iKCGQn0AuwTgKbCpXXnklvfbaa3TMMcfQtttuSwsXLqQbb7yRvvWtb9HMmTMxeQqy5qyzzqKddtqJbRs5cmSBatNxMMnVhTj//PNpl112oXQ6TUuXLi10dUCJAz2BKDjkkEPokEMOYdumTJlC48aNo2uuuQY/IkGHuOiii6iqqopmzJhBvXr1IiKiE044gUaPHk0///nP6cEHHyxwDUEpgfwEcgnGUWBTOe+88+i+++6j8vLy1m3HHXccbbPNNnTFFVfQPffcU8DagVJmzz33pKOPPrrQ1cgaLFfcBJqbm+niiy+mcePGUV1dHdXU1NCee+5JL7zwwkb3ufbaa2nYsGFUVVVFe++9N82aNSvUue655x4aN24cVVVVUX19PR1//PE0b9680HV9+eWXafr06XTdddeF3gfkF+gJRE0paaotsViMhgwZQitWrMhqf5AbSkFPr7zyCk2YMKF1gouIaMCAAbT33nvTE088QWvWrAl1fpB7SkFPFshPxUspaQrjqOKnFPS02267sQkuIqJRo0bRVlttRR9++GGoc4P8UQqaasvq1asplUp1aJ9iAZNcm8CqVavo1ltvpX322YeuvPJKuvTSS2nJkiV04IEHmv4fd999N91www00efJkmjp1Ks2aNYv2228/WrRoUbvnufzyy+nEE0+kUaNG0TXXXEPnnHMOPffcc7TXXnuFGmSl02k688wz6Yc//CFts802WV4tyDXQE4iaUtEUEVFDQwMtXbqUPvvsM7r22mvpySefpP333z+Lqwa5ohT0lEgkqKqqSm2vrq6m5ubm0INDkHtKQU8tID+VBqWiKYyjSoNS0ZPEOUeLFi2i3r17d3hfkFtKSVOnnHIKde/enSorK2nfffctPa8xB0zuuOMOR0TuzTff3GiZVCrlEokE2/bNN9+4fv36uVNPPbV125w5cxwRuaqqKjd//vzW7a+//rojInfuuee2brvkkktc29syd+5cF4vF3OWXX87O89///teVlZWp7RY33nijq6urc4sXL3bOObf33nu7rbbaKuN+IDqgJxA1nUlTzjl3xhlnOCJyROR833dHH320W758eah9wabTWfS0zTbbuNGjR7tUKtW6LZFIuKFDhzoictOnT293fxANnUVPLSA/FZ7OpCmMowpPZ9KT5M9//rMjInfbbbd1eF+QPZ1FU6+99po76qij3G233eYeffRRN23aNNerVy9XWVnp3nnnnfYboYjAm1ybQCwWa31FNAgCWr58OaVSKdpxxx3pnXfeUeUnTZpEgwYNao3Hjx9PO++8M/3jH//Y6DkeeughCoKAjj32WFq6dGnrn/79+9OoUaPafb2RiGjZsmV08cUX00UXXUR9+vTJ8kpBPoCeQNSUgqZaOOecc+iZZ56hu+66iw4++GBKp9PU3NzcwSsGuaQU9PTTn/6UPvnkEzrttNPogw8+oFmzZtGJJ55IX3/9NRERrV27NptLBzmgFPTUAvJTaVAKmsI4qnQoBT1JPvroI5o8eTLtuuuudNJJJ3VoX5B7SkFTu+22G02fPp1OPfVUOvzww+nCCy+kmTNnkud5NHXq1CyvPP/AeH4Tueuuu+jqq6+mjz76iJLJZOv2ESNGqLKjRo1S20aPHk1/+9vfNnr82bNnk3PO3JeIKB6Pt1u/X/ziF1RfX09nnnlmu+VAcQA9gagpdk21MHbsWBo7diwREZ144ok0ceJEOuyww+j1118nz/NCHQPknmLX049//GOaN28e/e53v6O77rqLiIh23HFHuuCCC+jyyy+nbt26tbs/yC/FrqcWkJ9Kh2LXFMZRpUWx66ktCxcupEMPPZTq6upo+vTpFIvFQu8L8kcpaaqFkSNH0hFHHEEPPfQQpdPpktAWJrk2gXvuuYdOPvlkmjRpEv3sZz+jvn37UiwWo2nTptFnn30WyTmCICDP8+jJJ580BdXegH327Nn0xz/+ka677jpasGBB6/ampiZKJpM0d+5c6t69O9XX10dSV7BpQE8gaopdU+1x9NFH0xlnnEGffPIJjRkzZlOrCSKgVPR0+eWX0/nnn0/vv/8+1dXV0TbbbEM///nPiWjd4BAUB6WiJwvkp+Kk2DWFcVRpUex6asvKlSvp4IMPphUrVtArr7xCAwcOjKR+IFpKSVOSIUOGUHNzMzU0NFD37t03tZo5B5Ncm8D06dNps802o4ceeoj9S94ll1xilp89e7ba9sknn9Dw4cM3eo7NN9+cnHM0YsSIDg/Ov/rqKwqCgM466yw666yz1N+PGDGCzj77bHzZpUiAnkDUFLum2qNlWdnKlSsjOybYNEpJTz179qQ99tijNX722Wdp8ODBrW/jgMJTSnqSID8VJ8WuKYyjSoti11MLTU1NdNhhh9Enn3xCzz77LG255ZZZHQfknlLRlMXnn39OlZWVJfNGPDy5NoGW2VHnXOu2119/nWbMmGGWf+SRR+irr75qjd944w16/fXX6eCDD97oOY488kiKxWL0y1/+kp2n5bzLli3b6L5bb701Pfzww+rPVlttRUOHDqWHH36YTjvttFDXCnIP9ASiptg1RUS0ePFitS2ZTNLdd99NVVVVGKwVEaWgJ4v777+f3nzzTTrnnHPI9zHsKRZKQU/IT6VFsWsK46jSotj1RLTuS53HHXcczZgxgx544AHaddddM14XKByloKklS5aobe+99x499thjNHHixJIZR+FNrgzcfvvt9NRTT6ntZ599Nn3729+mhx56iL7zne/QoYceSnPmzKGbb76ZttxyS1qzZo3aZ+TIkbTHHnvQT37yE0okEnTddddRr1696IILLtjo+TfffHP69a9/TVOnTqW5c+fSpEmTqLa2lubMmUMPP/wwnX766XT++eeb+/bu3ZsmTZqktrf8C5H1dyC3QE8gakpZU0REZ5xxBq1atYr22msvGjRoEC1cuJDuvfde+uijj+jqq68umX8x6iyUup5efvlluuyyy2jixInUq1cvmjlzJt1xxx100EEH0dlnn51do4CsKXU9IT8VH6WsKYyjio9S1hMR0f/8z//QY489RocddhgtX76c7rnnHvb3J5xwQgdaA0RBqWvquOOOo6qqKtptt92ob9++9MEHH9Af//hHqq6upiuuuCK7RikEuftwY2nT8hnQjf2ZN2+eC4LA/eY3v3HDhg1zFRUVbocddnBPPPGEO+mkk9ywYcNaj9XyGdDf/e537uqrr3ZDhgxxFRUVbs8993TvvfceO6/8DGgLDz74oNtjjz1cTU2Nq6mpcWPHjnWTJ092H3/8cYevDZ8qzj/QE4iazqKpv/zlL27ChAmuX79+rqyszPXs2dNNmDDBPfroo5G0EwhHZ9HTp59+6iZOnOh69+7tKioq3NixY920adPUJ7tBbuksekJ+Kh46i6YsMI7KP51FT3vvvXe71wHyR2fR1PXXX+/Gjx/v6uvrXVlZmRswYIA74YQT3OzZsyNpp3zhOSfeYwMAAAAAAAAAAAAAoMQojUWVAAAAAAAAAAAAAAC0Aya5AAAAAAAAAAAAAEDJg0kuAAAAAAAAAAAAAFDyYJILAAAAAAAAAAAAAJQ8mOQCAAAAAAAAAAAAACVPzia5brrpJho+fDhVVlbSzjvvTG+88UauTgUAAAAAAAAAAAAAujiec85FfdD777+fTjzxRLr55ptp5513puuuu44eeOAB+vjjj6lv377t7hsEAS1YsIBqa2vJ87yoqwaKDOccrV69mgYOHEi+n5s5V2iq6wA9gaiBpkCUQE8ganKtKeipa4EcBaIEegJRE1pTLgeMHz/eTZ48uTVOp9Nu4MCBbtq0aRn3nTdvniMi/Olif+bNm5cLKUJTXfQP9IQ/0BT+FPMf6Al/SkVT0FPX/IMchT/QE/4U859MmiqjiGlubqa3336bpk6d2rrN932aMGECzZgxQ5VPJBKUSCRaY7f+xbK9h59BZX7Fum3VFWo/F4ux2AsCVcZrTok4qSss9pPHpbjRRGm+j5c0jpvg25xVvzJ+LlfJr9PVVOpTV8V5mTJjBlO8nBdLpFQRf63YlmjmdUvpfSiV5nHgdBlfzKCLa1xXoQ3bUkEzvfjFLVRbW6vLZUkoTVUamornSVMyXldJftxcacp4lgKhqSCu6+eJe+03G5pqFHUWbRNKU2l9TRQTGm9HU9BTayX5caGnDXRAT0QF1FQh+70i01S6upyXybbfi0JTm9jvIUe1VpIfFzlqAwXOUdBTmzLQUyRsTFP7DDuDyvx1+T2o0r95SOZ6656JXG/ee7mf0JDS2LpKZj6u1JTR/lJTVMU1ZF13UCM0ZfR5nuiKfLPPE7/t5POWzKwp8zmRb89Ymipr0+d9+cf89Hmb/WRDjqoqV/u5uKh3WhUhP9H+c0ikc5srE3MFVntIPYnf3ea5rOdZ5iihH3sMJfRUbuhJ5ig5T2Bsk3oyn5GIcpRjOSpBL83NnKMin+RaunQppdNp6tevH9ver18/+uijj1T5adOm0S9/+UtdMb+CymLrhRoLMdj3jIdQljFfacvUmVpNJCa50sZxxaDXkZUk+LHldVrX7ZVlMdg3OkY/JoQY4/X1AuMB9WU2MIQq29i3jmMMACJ8vbT4NRUi+eVRU4HQVGAkF5X80pamRJ1FHEpTzupxYu3HxjboCXpaV8Fo9EQETRVSU16ZKJN1vxeBpiLq96An5Kh1FSy+HAU9tT039BQFG9dUeeukRGC0rfrxG0ZT1r2X91pcb6hJrjCacpamZPuLSa5QmgoxyZXSEwy+/G0Xqq/i2rSfE3kc47dywfs8Y5JL3Wf9j1aZnkMiIk/lKHH9YfQk/3GMSI8lzOd50+cOTD3JHFWWee7AkznKfEaiyVHWM5pJU5FPcnWUqVOn0nnnndcar1q1ioYMGUKec60NbuQMInGtzuoo5b+mWI0hRKca0RpMG7PaEk++7WX9C4xKEjKZWw8A3xbIWeks8UQipITxgKa5MOUDQUTk5ENrvQnX5jotrW8qYTQl7zuRTnWhNBUCPUAz7mtanN2oX1aakhoyNCXvmSszypDcZtxXlcBlRxFCU8YMv1Mdzsb/xQh6Wg/0tG6fTdQTUQE1Je8hEbli6vfyqCkpF0tTslAQ4kX1rDQVpt9DjuIbcpmjMukHOarDdGk9oc/La44i51qv23rzz8nvoxmacvJ3hjV5IPO2fBvG6vOy0RQZmpJvjan+wtCLJzVlXLfcLUSfp/yLrAnZFD+OZ1l3q/oVR59HQbBhIjSM5bjV9uXi+q1nXo6h4vIeW5NIQt/GWMJ6BhRq4lfcixA5KrByqNoWQk9yH0sH8m1B676E0GXbZyCspiKf5OrduzfFYjFatGgR275o0SLq37+/Kl9RUUEVFcbsPQBZAk2BKIGeQNRAUyBKoCcQJdATiBpoCkQJ9ATCEPlnDsrLy2ncuHH03HPPtW4LgoCee+452nXXXaM+HQAAAAAAAAAAAAAAuVmueN5559FJJ51EO+64I40fP56uu+46amhooFNOOSUXpwMAAAAAAAAAAAAAXZycTHIdd9xxtGTJErr44otp4cKFtP3229NTTz2lzOgBAAAAAAAAAAAAAIiCnBnPT5kyhaZMmZL9AdLpVmcxz/rEqTDhM035YvILBIbJmzLgzWwy6RnVyYhltCYM5tR1JvXt8ZLCZNL4DKg0lAvKMxt9+sKUTpkTGue2PisrMb9Y0qZ+zot8xezGSaY2fLHBMMT3wmhKfQEic/2VCaBh+OfJ79iGMIs2NSWMVL2UqG9Su/V5KXEfDb3I58L6/CwR/0St0pB13UnR5tanZaVJp2U8uv7Y2tw1h0BPnV5PRHnWVJt+T312mbLTlPlFmmz6vaTLWEZXJkS/J69T6ofCaSoQZs+uIkS/J6/B+uJQFv0eclQec5Q03pZ6yjJHST0RclTX0BP6vKLKUeq3nvXhAi/Eh1Nkn6c+/GUcVrpbh+jzpCE5kf7AQRhN+UJTgdGfKTNxq8/L8OEE+cya9bE0JTG+QNjSxs7l73eel05vuG/GWELNA1g5SprIWybtEqEn9VEAIiLx8ctQT5k13kiLPWUfGCJHmXMH4kN2trd7+x95sPQkP6BhmutL83yrzdseJ6Sm8jjDAAAAAAAAAAAAAABAbsAkFwAAAAAAAAAAAAAoeTDJBQAAAAAAAAAAAABKnpx5cm0yiWTrWk/P8BRxcu22sU5brem0rjadYY2+tb5a+mJZnmFJvvjWpYwyYp2zXIHqGeu/fXFN1vp0VynWE1vL08XaWxe4dmMi0l4q1qJj0V7m2ts2+3lh1npHRTLVuqbXM/zOtKb09UmPCdPnTbad1JDUHBnrpS1NCQ2ZmhIeZ14z16G1BtwP4aORrhT+dtLjgfR6bnndfmAIUbSF9AoyMT0P1m2DntbvAz2tq98m6okoz5rqcL+XWVNk5vIM7ZKtpqLo94zqqH4vjAenIY/MmjLudSA8J8LIwdJUuuW/ITQZFZ0tRwl9rdtR3B/513qPUDnK+bxtpNcpERF1tRzV2fSEPm/DfgXq81wy1erdZGlK+nS5uC4if8+Q1QQZvBStvBxGUyonNesc5YRflZdoP2cRaR8jX14jhdOUE96CQcDHFb7pcxhCU5meUdrw+8/8HZgr2oyhqMzwIxW/mQPDF1o9v8ZgItMzonIYrfMLY7GVo4R+zD5P+E6G8faKhfIxz+yLFci2EFqJWeObmGirMHMHhiEY80MNmaPwJhcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkgeTXAAAAAAAAAAAAACg5Cla43mXDsi58KZ1po9ZrH0TRyJtNqiNBrX7mbc2wTc0JVQZZRaXNlzUhOFdy/W2/rVx3V5MmhEa5sTCfFCZWZJl7Cn+PmWcW26zTPNkGxvm+axMuln/fY7IpClpruiMuitjR6OMFwhTwDCaahLtkE9NCXNGU1MxaVpoXLf66IA0EsxsxCgNFdedrANGzWnDpDFHQE9dQE9ERa2pwDKalf2eZxnCZtBUs87tpaYps9/LRlOy39vkHFW8fV7O9BQiRzUPrVdl5u9TxeJtDvhYldmlxxwWV/hcgzW+1mmfslUsThvGwme/+l0W93irQpXpPWstizt7jkKfhz4vctJpIre+zuaHqkTd/cxG4dbl+knRlvIDYiH6vHSvWlXm86MGsXjI+K9UmcMG/IfF178ykcVbXvqFPncYTYlnSRnwExGJJpWG6eZvPakp47eeZay+Mbwgj31eEJBbf9GWObyTOUr2b9a2dj4k0xonw+QokZMSul1cs9hmtbMvrkEcx/xQgHhuYpae5EcKrI+tyJ/4sn5WfgwzhpL7Wf1C2/8PmaPwJhcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkqdoPbm8mE9eyxpSy9dJLpm11tWK9crWenbK4EWi/LeIiNY28fMY62rJyfWlun7zfrQFiyef/CiLr334cLXPyNu+5odN6LWtflxsC+FF5jeJ607o9a5qm+VnoHYy7l3bvw5CHCMiMmnKifXHQVkITVnrpaXPm9SU9I4gCqWp1NihLF6+ZbUq0+/pL/lxmsRxrfXSSXnvM2vq6526qTIHfH8mizerWsLiR86YoPYpX9nAN4Tw32vP580LDJ+4HFHMelo5rj+Ll22t19+ffNQzLPY93fZxYTwQEwvy455u76uf4Hlr5N3fqDJSTy6wvCSyyFGyLTZRT0TFralQ/Z7lz5BJU9n2e7JPiGndkeH7xI4bUY4ynyXpcSOOkw9N5VNP5Hsb2jtbPWW2UIqkz0v8XB/45IGvstjKUXJbIPy1GgLtpVWZ5l5fQ+LLVJmn97uBxecMP0aVWZwYxuI+b61kcWR6ykC+NFXMfV5W+cnKRXnKT6sG6/Fb+Rp+7Op5PA93xj6PYjEif30/YXq4Cb8twycozG+9TL6TVp/3zc4DWTz+grdUmf0r+PhGjpGsbeft+TSLr/nNAWqfMTdxvz85/iHSHlxWVy99qfxm0ec1Wb/1hKYMbyuF9dy0PNv5HEP5meYORE4yfSiFnlKWD6UYn2YzhpL+W0Q6RxlzB/K63MDeLP5qv55qlz+ddT2Lj3vxx6pM2ZJyFgdl+rr7vMPj5VvyfYY8q6+pYuVyFnuWJ5ck49xBOE3hTS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDyY5AIAAAAAAAAAAAAAJQ8muQAAAAAAAAAAAABAyVO0xvMULyPy40RE5OK6mi7ODW1NM0JpHmd4nXnSpFEYSFKzNuVzyRAG7MJw95uDx6giPzv1b3q/NvzfUQ+obX/bfScWf/HcQFVm6JOrWOwbxp5eUpgRCvNK0zRPto1lcCmNhssM4+FCEYGmSG6zDC5lu0iTvTCaGtBHlRlx/ScsPqxmgSpz07BDWLz5LV+oMgph8GcZVQcVvG2+/YNXVZmjenBTzmXpGn6MuDGnLtsiW015GzdXzhlFoqfZp+scUDGWGyCfPPJNVUaaylf4WpeVnt7WlphhBH3FpHtZfNtdB6syygQ2zEdBhOluTnNUWx2VmqbkY1bAfo/K4/rcVZUsXjRhEIunnP+g2qfW54atv/n4IFWm/C/cyLn7pw2qTEH7vQLkKK88Tp6/zhi22Pq85Aj+cYzB3b5UZaSpfLWvTW2zyVEy96WNf+9tcrxtrh4xXZWZ/b/c8PeKX5zI4p6v6z46J+OofGmqSPq8nOansix+FoUYQyXq+QcQLvnVHarMT549icVbXN/5+zwvXkZehzRljCNDaEq1i2i3D6b2U7scueMbLO5fsVKVkaby1b6+R5XG2Kotv9ztEbXtve34h6aenad/Q1Y+yD8A1fOjNaqMlxRG82tFn9dkaEo+X5ajvdSU395HZvI4hipvk6Mq9POt9FQW4kMG1nmkxqIaQwmjea+iXBUJenNj+dVX8nt45cjb9HEFz+53vdpWL0z544bp/V0HjWLxATUf8XPvd6Da519Pbsvize6cpysk5ykimjvAm1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASp6i9eRy5XFysXjr/0uCcl51Z/j9OLmuPMw6c2vtcSaMdeTpUYNZPPF/X1Fl+pat5vuEWLd8/pCnWfzOccNVmX8+uRuLPcuTS64FluuJpQcCEbkUL+P5xhypWEdr3bu298FZa+dzRCZN6bXaxvVlISmlKcMzwavrzne5Sa+t37bbfBb3iDWqMol+/B65qgpVRiH0a133nDP4Nfyuh/Z4WuX4uRpF7Cf1dYfSlPBgsHwbaP1+XUFPjTfxdvzfIQ+rMqsD7nsUOH1u6R1x9eOHqzKurP32vObwu9W2GnHcDyfXqTJjbhfatZ6JZIYcJWOKXk9ERa4py+dOEMoNI5t+T3pyEJHXg+exVD9975t/xX1NLh5+F4sD49/epO/STVv9RZX501l7s3jeuZurMn6j8MQogKaKWk957PO+Gcs91A6q0/6RMkddN2s/VSb4jPvSjHhY952Spr5VLF5+mt7n3h1uZ3HS0GWvGN/v2IufYvEzE7dU+yg9WQ1apDmqmPWkMPKT9ODyKvX4yGzrTIQYQ80/geee/rFVqoyXFo3RBfo8Ko+3eihZHmlO+MGa3q4hNJWq4898rz8sZvElvV5T+6xO832s32jSJ/Dt1cNUmWc+3ILFF47neWJofJna55Ae77H4e/UzVZmPthjA4tt/OkmVqVjIc5QXxjtK/P7zLB80v1j7vHJysfU+lOW6TkG50JPxrDp5m3PlUWeNoeL8GUgP6K3KlF/H9XLN0Mf5YUm3d63whasIcUlNTs8DHFLzIYtXBry+x/TWvw0PO+FdFp/b73uqzJB/8rjbh8t1hfyOzx3gTS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDyY5AIAAAAAAAAAAAAAJU/RenJReZyoZe1/hV636mJ8Qanz9QJTua5WrbMlIifWFav17KbvFG82y1Phs2O438QF3T5SZa6dN5HF7384hMXTD7pR7RP3uBfByIqFqszToi28tOF3I9azqjLSs8vCaBu5LttVGmvs26xDDkKcJjLaaiquNSXXZkuNWVi6U+u3ZTsZ67DTdTUsPqzf26qM9OBakqpVZfq+Jtq/OrMnlxP1aRxWo8rcvssfWWx5EyxPc2+UC//9HRaPWM39VYiISPrFWevjhU+DqyrXxynA2v986WnxLj1YfNlm3INrWYq3O5H2MLrl0z1VmYFnNbB4VOUSXZ8KrqclO/dkcd/vcF/BdecWng+Gr5e8Ti9p+AYK35VQOSoiPZVMjrLyjyCrHGV5UIh+LxjaTxWZdyG/R1dvd78qU+uvZbH0j7Byi9RUpadvynn9n2Hxt384VpXZ4mp+bqWpMF4/Vr/XAU0VtZ7y2Of1feErFt885kBVpv9M3lgjPtc+RkTfmPVsj6qv+H0e8Fs9Tjnj4u+z+A9j71Nlko4/E8onM0yOMnxjijZHFbGewozLqYK3o7M8uSo6/rMoqjFUv9cyjN1z2OcVgw+l9FomInLSTyxEn9fUR9/X7037O4uHly9l8bK0HkfFRb9T4el2mfbyofy42iKVxszn46Q/jz6MxSN+xn2OiIiO7P0Oi4eU6dy3U+WXLH7/Ku2H9M7p27E41sD7QGf1eU5s84zf4BVCU8az1KJFw9opd8TLiGLrdGTlKFfG9ROUZdaT+TqQ3BbCA9wTOcocZ4kctXraWlXkLyOmGxVqn8sXHsDiF/+5vSpz4EFvsTgR6OdxZZL71K1q5v6/y28bqvap+5T3iyPLmlWZ2Fr+m8Xs89rktgCeXAAAAAAAAAAAAACgq4BJLgAAAAAAAAAAAABQ8mCSCwAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlT/Eaz3veBlM2wxfPk2aLhgeZsnSzTDClUaYwaCRnGCILI8RFBwxSZW6ddAuLpYEhEdEXfx/BN2zGy/SLaXO2SmFUtzCVVGWCcm62FzP8vp04jjLpNMx1VevJtiIiJ0zzgmptHtfWnDSdyqMjYVtNWX8tNKU0RrrdrOMps0Np+mm02/yLeNwrtkaVqY1xA8Lf3HmcKjP4I25wma4JYzzPr2H4/+qPJAwv4/VZHWhDx6s+4x9S2PxCYUqe0HqmMnGcuGE82gFNpVIhjKOjIk966vMOb/tLpp3C4hVb6uP24f6RNHDGAn1uYRRqGT3KXLLNabNY3C+mTTHj4hJG/lnnPvmMGH6uRM0Z2sYwH1Z5bBP1RNQ1c5RlwO7q+Ycu+t7whSpzZp83WFxv5THxUYT/Ng9g8X8atXHpvrUfsHjL+EpVRvaEW2xmaN7vzuKs+r1N1BT0tP4wSZ4XRt38lSoj2zVMjpJG1Rayz2sYVKnK3Dj2TyyWH/0hIvLFNl8OVjtbjiqUnsQHdOYeoz96MfRpbswdW6qNuvOlpxWb6Xvax+fmy4FhPN/crTj6vEKNozzj95Y01jc1JczokzX6nu1X8wmLl6S5gbb1IYBbv9yDxaseGKjKbPHacl4Xw/g/qOb5sHoBHzctPE/8FiSiC87i2/61x+9VmaRor5PrZ6gyz4/ZjcW9xHOhPrZGpHVn/GaRBunyGok2PMfpPOrJxbz2P3ihPrpmaE7oyfywnchR6oMVTv/28uTvHSP3Ld6bj4cuHXmHKtPb520/fQ3X5a8eO0btM/zvTSwe4uuJgQ+f2ZrFfkL3K34j/x3nNfLj9Gqap/ZR8ygVRv6pFc9jTfu5OWyOwptcAAAAAAAAAAAAAKDkwSQXAAAAAAAAAAAAACh5OjzJ9fLLL9Nhhx1GAwcOJM/z6JFHHmF/75yjiy++mAYMGEBVVVU0YcIEmj17dlT1BQAAAAAAAAAAAABA0WFProaGBtpuu+3o1FNPpSOPPFL9/W9/+1u64YYb6K677qIRI0bQRRddRAceeCB98MEHVFmpfQ82SipN5NatBw01E2csq7XWRkuCOC/jO9EkxiGahvVk8bnn/02VqRbrXY975qeqzKh/8TX6s0fzNc09fH17Kjxe5sf/OFWVGVHBvS4sv5tYkq+1dWIdv2d4aMh1tXIfIiJXybelK3WZoGzDOuR0SnsM5Iy2mrK81lzmtf9BGE2JMr5og8+O5d42REQ3bXsri2OG/8fCVA8WD3lGe05k8pNwZYYPRHdev/16ak+uOp8fN+30Wu2v5vRmcfeyZSz2AiPdSE2ZPm/83EGF1kw63uJ30/n11Ps94XXmaz0t2oWfq/sPtZ4uHsHz1v/7eoIq89shj7FYegLGjQRZ4YkcUGX4iKSkp6LRftKXJSl8vPKgJ6ICaipE8VCaMnwfMuWoVA/dVy+7gPuInNtH+39U+twZq9LTeeKw1yazuPeT3Lsi3qCvad+ruCdXfUz7XTQ53u9Vl2kPwAbpwVkWkaZE+3X1HCX1ZIo5mXn46YQHkTWmU94o0k/F6PMC4dmybGtdZkyc58zlgdbyMpFWr7zrWBYPLzP8SUjUt5RyVIH09Pmx3Vh811E3qn1+N/FgFjf/uLsqE0pPMkdkMYY67/TpqsxAsd9La3upMj1n898NkeWnYtUTEdOU5aQka2L5dkm9dPtKj3cO/cv5/LS9eV9V9572ABrwygoWV61dosrI30HKT45CaMrwkBp4Nz/OCf2OVWUeGv0wi9OB7vNkbqt/XWjK8tsKMmsqUH2e8VuvPP99npcKyHPr73+g/WBjMkdZNnDxjucoquRt5FteZ6LPW7JzvSry919exeJqT7ddtc/HZ5e/dwiLhz6tdZCqEjo1HjZf+pNZfujST1fOA6RDzB1UGHqqCDF3UN4mRyXDaarDk1wHH3wwHXzwwebfOefouuuuo1/84hd0xBFHEBHR3XffTf369aNHHnmEjj/++I6eDgAAAAAAAAAAAACAjET6dcU5c+bQwoULacKEDW8E1NXV0c4770wzZswwJ7kSiQQlEhv+9WLVKv12CgAdAZoCUQI9gaiBpkCUQE8gSqAnEDXQFIgS6AmEIVLj+YULFxIRUb9+/PO+/fr1a/07ybRp06iurq71z5AhQ6KsEuiCQFMgSqAnEDXQFIgS6AlECfQEogaaAlECPYEwFPzrilOnTqWVK1e2/pk3z/IwACA80BSIEugJRA00BaIEegJRAj2BqIGmQJRATyAMkS5X7N+/PxERLVq0iAYMGNC6fdGiRbT99tub+1RUVFBFhTaR9VIp8oL1xmLNSfX3fkIYnSYMI7Mqvi0oNy5XmK8pI0pPH3fbK99j8ebxxarMqf8+mcVj/rhWlVHnMvzaJHFhQhdr1POUykw/pQ8szRE9YeLmQhhcUplh/GaYHOebcJoy3AaFzmLN+t57wjBPGnoSETnRBtIMfqfdtbF7d7+JV0XZbRJdfs9xLB6WWqHrJ0x35f1whrXn/P35tpO7az0nHTflfHjtAFVm2BOiTaUhp/ExA2VMGzPm3UNoqsU80jKR3FSKTU8yb9z1f9eofRrEBzSSLrNJ488GPqW2NQl3Sl984aPCMNeUxvMLx2tD135vcmNMP53ZAFgaXOZDT/L/oyKqfo+Mfi+UppQ5Nz/usq2r1D6/3/o2Fls5SnLEs1PUtlG38+tyIvd9erLue/aq5HqJG/2yL/7N7r8vj1JlRpTxjzYoTRmmxqXQ7xU8R2XQkznUlLnDGKeouhhlPJ8bwi/ZiX+Io6mPPs5W+3/C4ruHPaLKVHnVLK42PqLQFPC2GfYY/9iK9XGeqHJULim0nr4ZU8Pi679zB4trfW2sPL7nXBa/EttB10/WxdRTx8dQS3/APyIVZgx1zgxt3zJ6tbgu9HnkNYlxgNXnVfK2dYZx+IhH+T3yhEG2l9a/0cLkfmWEb2jKl78JpKYM43l5Xz9dqBNZ9Vh+3QmnjdZ9uSmMUbi8zDB9Xp5fmdmYnqg52foM+MaHL1xC6knndfnhFPkbjihzn9fUU9et7IJFLL5wyJOqTLPQ05uJHqrMz2/kH5wb+CHPG854gJ24P5bmApn6jOfIlx9NkB9aSIeYO7BM+WXejYhIZTlixAjq378/Pffcc63bVq1aRa+//jrtuuuuUZ4KAAAAAAAAAAAAAIBWOvwm15o1a+jTTz9tjefMmUPvvvsu1dfX09ChQ+mcc86hX//61zRq1CgaMWIEXXTRRTRw4ECaNGlSlPUGAAAAAAAAAAAAAKCVDk9yvfXWW7Tvvvu2xueddx4REZ100kl055130gUXXEANDQ10+umn04oVK2iPPfagp556iiorK6OrNQAAAAAAAAAAAAAAbejwJNc+++xj+1asx/M8uuyyy+iyyy7bpIpROk3k1q+VtdZpq7XS+lL8JF8T61XrNbJO+HQJCxpavmU3tc/J9a+x+MXGMarMwKv5unF/9Rp97gp+bq+Z+6CsDvT66grhCxFmFav09Fm3TcRifa5nreGXbW6s5faSvH7KQ4aIvNSGY7uUXg+dM6LQVBXXkAsMTYk1yovGd2fxVYP0OuxG4e0hvT6IiAb8K8Fib632pfBTQs+V8lnV1/TGd6SnU40qI3liybZqW+Vi7q0jdecZ67CdNKIz1tCH09S6/bqCnuqv/6qDFSVqclpPSeHb1WCYAjY5rrkefkKU0O0dFyYQH/z096rMLxZvw+L7/rOTKjP6/wnPgwLoiahraEr6a6QOWpGxqlaOenolv6+jbzVy1Bq+bdmOPVn8h31uV/skHG8L6U1JRPRJkh+3+6eqiPbXkZrKtt9r5pr3Db8LShcgRyVTRP76tkoaepIbcqQnE9GuXtpoF9HWqe56PLb0Qt7v/HpLrp/typeqfep97mVT7evjpl1mj7Afv/YDFo/xhO9PDnNUi56I8qipPOWnpTvxNhpe9g2vhjHy/fND+7N4RHqJKqOe3aQ+TjZjqC372V+Pb4vMTwMe1/mTPOFXKJ6jL47TXqhrh/DcM+wxraeqr/jvj6Lq81JpohZfPSNHyd+bXsL6rcfbwIX4rRcKqWfDE41EW3lJ/btNej+S8jzWWpC/ya7c8aGNVrMF6f1FRFQ7V1ZG+oEZWpDPSZg+z/BU8tb3C/nUk5dOk9eSo6x7Ie6h16THKC4pXsqx9CS8BJt78DJb/eo/ap8L+z3H4tWBbvuVAT/uma9+X5UZ8wLPh4GYSwiqtNalnqSnGBHpAYHVjWfwLDT7/hDPkfRGixn1Y969KX1vLQr+dUUAAAAAAAAAAAAAADYVTHIBAAAAAAAAAAAAgJIHk1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHmycOIrAIbRvZMGhYYJmTLBNPBS3AAtqORNsu85M9Q+DcLE+YY391dltvhyUeZzJ/m5/EQPFq+W7vBEVBkI83HDe01ek5c2jE2FkaDcxzSGk+aBhhmhvFcxeVwiorI2c6tpbTSZD5xpIJlqPyZDU4bGvApuartiR25suDrgf09E1CDMV3/z+aGqTM3nwkDXMHOUBpdOnGvlKG2w2yiuoTHQRoyBMMsNDG2qjxeUiTn0pmg05Rt6btGU6wJ6OmvAsyxeYRj3SqP5FWn9MYEGQ4eSGp9r4QdPfZfFPT7U/06yy6n/ZvFVA15WZX7e+20WT95H59mzNzuCxWtO78VipR0inbc2UU9EnUBTBl4514d8dnvVrFX7rAq4GeuKoFqVeejf32Lxll8ZhswiR8WP531l/9gqtYv8CItPOkfd+83OLO4xW1+D1Ic0oY2q37M05bdoKp968r1Wo1jzg0E5ylFSTxbSXN0yAJbGwV/v1k8VuXTswyyu8fhxliuTZ6JyYfBtfHNDV8W47uO3fYvF7wZjWZzLHOUXeBwVVX5K9alVZX6w56sslmOmZtIfnhj8rDD9t/QUwgA70xhq3kT+ISEioruGPcLiRqNvlWb5X++lz714Jz4+673NYhZfOOJvap+Yx9tz/u71qsxLR/EPBRVtn2c9C0JDztKUpUVZRvwWccZHIdQ+Iv94xscWLHNzhdCUH2LstWYw72+HxJepMgknY6P/kptkH2i1ufgIiNm+8reeoSnXOi43nsV8YHzMROnH+h0l8681B5EWH5f79QoWXyBM5om00fxKY+x+0ZxJLB7yoM51XgP/2Iqf1seRKCP3EAbx9tyB/BBPCFP5EGMo+TRamnNZ5Ci8yQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkgeTXAAAAAAAAAAAAACg5ClaTy5XVkYutq569pph6aVhrL0V66c9aw2qOM7sC7nPyJm1H6ld3ly7GYvHTlupz92UUNskat1vitdvhbFuu9Ljx/WT+ppiyUCUMdYcJ8Q22VYJa+252Ga0p1x7a3lSuNiGuVUvnbmdoiKTplwg1xZHoylK1LFwSVp7OjQFfH33qLolqszMo7Zjce9Z+h4FcV6fhv58PffwE2frcwt/rTVOHzcttHr50EdVmSX38WdneZr7S1x884lqnyH3fqa2SbxUXMQb11RX0NPZl0xh8f9dfJfaRXpyrU5XqTKNwg/gxocOUWVG3vwFi8dWcY8QV629AD7792gWH1GxhSoz6qoPWXxOX+1fcPbAZ1j8f9cdyeKa0wwvDJmjDDqiJ6JOqil5zTH+711rEvq+yry1Ol2pypy4E/dW+8tvd1Rl9t38Exaf0fM9Fq9y+tzVjt/rmJGjHv6U58ehhh9kJP2eQcnmKOlZYpXJps+LhfC7kW2U0L4tDdsPYfEh3/+XKrNaeMXFhVGpjImIajzu3+T7Rl8qz2P4UO7f/X0W//2APVg8+EGeP4nI9KmSFGuOykV+6vHb+arMVlV828I0H0O9uIp7nxERfXqq+DmT7KvKUJDZK47KhC9NBb+G/9vpIbWL1IaVn/r4/Lh/P+xaVWZJ0P4YakVa+yDOaerN4jvf2F2V2XLt1ywuFj0R0Tq/Kn+9piwfHqkhy2dJ5GhvrZF/5G+TEJ5c6lxGjlL+0FZ+rOL6TQzkfWmsSV/TipG8fmPiOm80iuetQT5/RCRTm+rjDJ8xdU0GMkeZfneF6PPicXKx+PrzZjl3oPSk7+ny8X1YfOog7j27JG14Lzu+TfZdRETx47nf1pdnaU+ubgP7s7jfC3xc7ks/ZCJycX4cZ/1+l3MSzcbcQXP7PnWmb521TR5XPmtJPT3lxTZcgx9SU3iTCwAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlDya5AAAAAAAAAAAAAEDJU7SeXBQvI1q/9t/yifCEr4GzfA7SfP20WttNRMv2HcriM7/1JIsXJnuofWIk1j3f3KjKfPQFP27Z13p9bvkKviY2Wc/XrUpfHSKihoCXMawkyGsWnlxNum08sbbcaxJtZaw9p6SxTVLGJaXWbRNfV+sFeV7736IpY/24l+JzvtlqSq4l9oU/wMIkX59PRJR0fJ+R1YtVmREnct+imKefi7oY12IPEdfH1qh9GgN+7hgZnlzEtboy0L450i9Cxr4hH9fUpDdKxH1oT1NdQU+9n/mcxbc8s6faxcmcaT27wr9h85j2H6QKnreUl4fleyS2WR4+/7r7WyxecLR+Jn406CUWnzr0VRb/acej1D61LxjXIOmAnogKqKkc9ntSU/Ierfo3950kIloyopbFiUC3W9zj5zpta+2h1C3Gn3nZz0l/QiKiRo/X9ztvnarKDL6ee074TVrzOev3SjVHif7apTNffzZ6sgiTozyRowLDF0vqpcnn8SMrxql9HnppZxbHV+l/7/3t9+5kcf8y7b3aIPrB5p78783+LYQnVyg9tfEUypumctDnrUnWqCJLUsKzSDikja3iHlNERJvvzH1MGw1P27Tj9Su1MdRlMw9T+/R7mmtjiyc/VGVcEY/Luc+bkVtE3S1PLun5o3KLdRyf3w+vXOtl5c6DWLx6kPZHahZDl+SYtarMwN4rWHz2iAdY/MJK7Vt6UvdPeV2MvJsWP0XvXLGzKtPzw9V8g+zj4noaIDmC+9nF39PeufLZLpk+T46hDN8prSddZm0fnkuGxpexeEmaj5eIwo11vv/qOyyO0VuqTG2Ma+z507dk8T8e20XtM/RJrgPf6KI9oSc/YcwdrBVjqLX83pp+5KH6POGZV67bhrLo8/AmFwAAAAAAAAAAAAAoeTDJBQAAAAAAAAAAAABKHkxyAQAAAAAAAAAAAICSB5NcAAAAAAAAAAAAAKDkKV7j+cBtcEFzrv2yRKYZoWsWhqnGcYRfJK1JV7J45gptwLt17QIW71w/V5XZoxc36quNaQPS+jJuYFnrczO5wOk5yKSYlzz2uBdVmb+N24HFjUu6qTK93uRuiWUJ3jYVK7RxY9Uz77HYNJeNG4aPkjYGcy4wnPNzRVtNGSh9ZKkp2S4DXuNl/jFuG7XPtnVfbbReLUhT50rjqwOyjG8Yq0qahaYanE4L3337NBanZ2mjcCnxbvP5dQ94iOuHiCiQ5pVhNGXlA3+dwWVX0JPzQ/zbRBDiOYxxE1XPMCCVZUjeH6u+wnjequ2AV1awuPlpnaOWPMbNh6UJ77IttQlszT+4CDdVT0QF1JRRr3xpavP7lqp9pu/EPxawV79PVZnqGD+3zEdERJVest3YMoOeuZb3w3UPaFPXsjWrWCwNUtdtEyapa9eKWPfT0nQWOSp/OaryVW6i/dwftaHuE/vzcVT1szyX9PxYG9SOXsG1snaANj///GhuvlxuaPnTRH8WD/0HN/d1jdqE2iWFCa/hAKxMsDPmqBDGvlGQAz01XTlalbl2/2Es7r0VN5Uv87V2QvxKoJioe1WZfhZry3kOOLbfmxmPK8dQ/zNXfxTl/Rk8h8US+r5vdj/Pu95y/rGDMav1h1WkAbgrtT7PufZ/44mPT1im8lJTch8i0s/ZliNZuHqaflYnDnhl4/Vaj+zjqmM638iPF8g+7ther6t9unv8OA2BzqmL0jzXPXHTXqpM/6+/YHFiNM9Zn51o6CXg28b8WPeLXloYg1ttHltXZxeE+HhLVLTVk6WrUDlK6N/oqwY9zj9+cfYux7F43OB5ap/aOG/HbbvNV2VqfH7frfGQ/PjdMT15jvrhKfwjTUREh444k8Wb36GP66/h98lvjGgMFeJ3nuzznDPGB7GO5yi8yQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkgeTXAAAAAAAAAAAAACg5ClaTy4vmSQvvX4OLmWsmRXrsi2fCFUmpddw1t8xg8XPz9udxeUv/Vft81q8nsWrDtUeS+kKvuY0VaHXoDYM4du2O4Cvtz9nwDNqH8nRdW+rbRN2mMXiJenuqszCvbin0krhd7MyVaX2eb7Hbizucf87qoxfbqzvVoXalglRPiIyaUquzZZeLOu2ZdaU1GLNg3y9fepV7vVBRDRzy51YHMT1/HO6nG9LVem2W7A/P/cfJtzFYstXRHq/HTXjR6rMyCv4Omt/lV5L7tY08FiszQ6Mtdok112XxVURpSnLc6IAFEpP4SrH28ivqNBFxBp4r7JSlaEKsU5e+HY56dllVSWl19Z7zWKNvvRAIKLVaZ6DVqZ4jmre0vC7kd4cJaQnoiLS1Iez1aaag3k7Pf/93VSZxTvz4560l/Y0GVbOfWeUJ5fhrnP9/UewePhHK1QZpamE9pNwTU0i5v4SQZP2U+nyOUr4afg9e+rj9OHbgk8+b/cY6yrH20j5UBGRL7YNePwLVYb+KXOU8AgxvAaDan4PvzhSH3b7yi9ZvDClfShnrhjB6zuLe7GG0ZNn6EnRmfQk8lP502+pMps/laM+r4r3KV61Hucu3Y7f0+438BxmjaHmJXuxeMmNw1WZUe8v4+duNLxrxBhKjpk62xiKKISmArHN0pT0uZP7EJG/7Vhe5Crud7ZTj4VqnzVpriHpObluG3/G62NrVBnpySV9l2QfSESUFr+NGg2v3I8SA1jc59/63K6O+3bNOY0/W8dupX/H/e0N/ntEaczC8BYsBF6imbyWuoSZO7A83oSerDEUfTqHhcO5JRfxp30dK0ZvzuKXj99BlUlvye/hDTver8pIjdV4vL7llq/p/jew+OBep+gKPs7zWP9nFqgicswkf+e5hO7zVD9g5CilHi+ad7DwJhcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkgeTXAAAAAAAAAAAAACg5Cla43lqThL56+fgDKNBZd5pGSnGxBye0ybJ0hAt/iw3cjftL8W5657+UBXxarnZn6vRBpeunJuvvVk3mtdl0lNqn6QwCb9iwYGqTGOKG7F2L9cGzSuauYnz6mZusLh0TY3aZ/B/VvAN5YZhqjD79Cq1+Sf7+yALg9FsaWreMK0bwszbs0y1Y0KLITQlz5VetFjtU7ZsOYv96mpVJoymluzATe1rfX3vM1FRqU0WV43hprs9ZmqDS4W4bs80puTt54XRVIU2KG4xUjX8FnNHkegpzLksU2dpuhsmR6l/FrHyrqxf2qivNPs0zD+l8arC08fV171peiKCpjZ2rvrHPlBler3Ac9Tz/9xDldnhmntYLE14k4bBblmD2BAYN0VqKMwHG7pQjnJWmwksPfn9eJ/y2bW9VJmDNudjoMdmjmPxgBf1uWof5mMt0yg8ghyV6KvHMnOO5YXu2f8WVWZ1wM/1YdNAVWb2fWNY3J/eZXG2epIfATH11Pbv8zWOKqH8FKbPC3rpjzJt/6t/szjMGOp/H/sei0d9uEIXCtHnKQqQn4gKmKPC5GxDU574uESsdz9VJnUtH7NuYRjNSx79eFsW93hKj8vF93Ao0UPfo5O//zSL967hHxmzPrYiSRrPyWbl/LfEyfc8ocosSXGN7yY+6LM6rT84FK/jfbJfp58T9eEQI39v6PPyaEqfaN5ggh+qzzPe9ZEfKzGM97PJUcHn/MMpI65bogsN4tq9cugP9HHEh8m8s7gOzhnxrNpnt8pFLJ6+7e2qzKGJH7PYvRTioyghCDUurxJ9npG/237cIGyOwptcAAAAAAAAAAAAAKDkwSQXAAAAAAAAAAAAACh5MMkFAAAAAAAAAAAAAEqeovXkcqk0OT9FRLbfllfGq+6sddpi3adnrIF3qRTfINfZGmt65VpRuc6fiMhV8/WlrlKvQVV1FpdpedIEwpPr/Qe2UGUGPcnX5y5L6vpRopmFVU2rWTy4aanexxfrgLtprwvloWGtvW2zdtkZdmu5wgUBOWpnIa+8H1UhNCX1Q0ROeS9EpKkqvubd0tRmu3zJYrnW39JUs/B0+MGoN1SZFz4fzzeUWb4Ioj5VYo265d8kNVWp/QG8mhCaavm7LqinL8/ZXu3T+z/8ON3e+kKVCaMnmaM86TtgeR6ITWofg2V7DVbbav23+GGFd0L5B4ZvXRXXyqbqiahraiqqHLVqmB5i9Ig1sjju8XO/s3aE2mfwU9yzMIymZG4hiihHGdftCT8JZ3nerK9zqelpyR95P3/asH+pMomAt2vtp/w4dS9/qs8tfCez7fOa+/D6rR7M2/6Yn/1T7bNb9WwWWz5wzcID5+G79lZlhjz8Oa+f9KXJQ59HROTS+fG86Wz5acn4nqrML+tn8FOJMbc1hhr0ovAIDpOfinQMRZTnHJUOyBmeR634YgxSmVlT6b76vu7dh48nLI8rSdkHPLf0fkrnsYadhvNzH7halTmghvtXZvQbNepn7RMTxkSVvvbTrY1xT7kw5z55y9dZ/NLm41WZskUrWFwsfR6lg3ZPqDy4Yvp5objIJZbfmMhJYTy65LyFV6PHsDL3VS5q1GXKxDVM5Tr91Tjt4/XoL37HYuuJe2DHP7H4rEFTVJnKNaI+gWgba9wl81aVlaN4W7gKI0cFHZ87wJtcAAAAAAAAAAAAAKDkwSQXAAAAAAAAAAAAACh5OjTJNW3aNNppp52otraW+vbtS5MmTaKPP/6YlWlqaqLJkydTr169qFu3bnTUUUfRokWLNnJEAAAAAAAAAAAAAAA2nQ5Ncr300ks0efJkmjlzJj3zzDOUTCZp4sSJ1NDQ0Frm3HPPpccff5weeOABeumll2jBggV05JFHRl5xAAAAAAAAAAAAAABa6JDx/FNPPcXiO++8k/r27Utvv/027bXXXrRy5Uq67bbb6L777qP99tuPiIjuuOMO2mKLLWjmzJm0yy67hD6XVxYjz19fPcPIjIR5nGVOr7DMKjMYXDprH2EeR5aJozSVtMzzhTGd8N+lRmnoZuBZ7nGibZxv1M/n9VHtJ6/RKlOp6+cquWGfKzcklm5rHteO4WTEME1ZepGGhJbuJGE0JYxVnWVIKM9ltC3FeVsu3lWba1497G8slqbyTYHWQiDmuv924wRVpt/aZSy2jEulPjxpQusbhqGyzS1NSfPhuHGcFk35+XO4LJSeVu7CTdrPPOFRtc81jxzO4lHvW+aamdO/J+uTljnLaO9A6NvXbbNk9z4svvznt6oyjY7f97jHn6NeH+hz+9LIM0s9tXXlLFiOsvQi2zKqHBVVvyc01TBY57py4udqcvw4jYE2sDV1JpH9XjY5Ktt+rwOaKrU+b1UDzx1D4stVGXkPUzLd9NZ9lYvx+gTGWCFdybet2kybiZ/wv/9g8XZV/CMblsF0gxhbyT6QiOjsl77H4i3+vliVkQa6ofQUUZ9XiBxV6mOoVJ/uLB576oe6OsJoXo6hfvi3n6h9Rn5mfKhJEkV+yscYiqgA46j19TH7PGkUnllTsW+0+fud/9iPxecfocdNkn0Of4fFz9bsoMqcffgTLN6p6nNVRpq9N4kPXbzeOFLt88DlE1mcqtLP29LduNH80d96W5VJBPxcwyv5WP6Thr5qn/f+ujWLBy/5SpVxlVJTRq5reU7z2OdRWYyoJUdZWpF6sn6byw3WB5bENckPaDRsP0TtUnH+1yye/+xQVWbYY8vUNn1uMQ4X9fP1tzwoEeJbGJViEiIo1+0nP6qn8m4Z/6idWSbrPm/DdSqj/42wSV9XXLlyJRER1dfXExHR22+/TclkkiZM2PAjeezYsTR06FCaMWOGOcmVSCQokUi0xqtWrdqUKgEATYFIgZ5A1EBTIEqgJxAl0BOIGmgKRAn0BMKQtfF8EAR0zjnn0O67705bb71u1nfhwoVUXl5OPXr0YGX79etHCxcuNI8zbdo0qqura/0zZIie/QSgI0BTIEqgJxA10BSIEugJRAn0BKIGmgJRAj2BMGQ9yTV58mSaNWsW/fWvf92kCkydOpVWrlzZ+mfevHmbdDwAoCkQJdATiBpoCkQJ9ASiBHoCUQNNgSiBnkAYslquOGXKFHriiSfo5ZdfpsGDN/jD9O/fn5qbm2nFihXsba5FixZR//79zWNVVFRQRYXl7VRBLrZ+ewhPrlBr/411/GqtsFjn6Ul/GeuwIdaIe0lj/aioz+DneJkVhwt/GSLyidd31Wh93IHP8RXFrkzXT66RlVep1vkbWGuwnfQQsHwb/DZnC7FOuKMUXFMpcU/EeumsNSX2SxyoX8+V3ijSg8vy5Lpj/u4s7vXfRn1ucR+te6/bT/pLRKQp4760eAi5MN58HaTY9LRwV36Nfcq0/0T/cfzN2aUfDVRler2+hMVmjpJaFfnxm/H91C5LduD1G7aD9nO4erM/sDhG2vOg3PFznf/YCSwe87Y+LlUIT8As9WR6MERIwTWVo34vMZB73uy5z39VmRUB79eahWfS8lSN2sdLcO8RJz0viZQfmKsorhyVS03lUk/B591Y3Gec7neSwmNm2ql38r8/JfNQs9LXXh5x4d8WM0xI08JDaUWa60v2ievKcI1Ne/lQVWaLG/h1FlufV6p60gfN3xjq8yO5Ns7v+5oqI/PTZ83cs2jws4bnTBPfZmpF5qci01NbH8Z8jqOoqoLIX7fdHPdm4cllaWrk7dxT7/4dd2TxmcOeV/sc1OM/LD786H+rMtIr1PJS/jJZz+JfvfNtFo/+ufY5rKcFLLbua5+ZfNssf4wqI/0sP0sM4gWS2sBpMH3Jz10Rwnc6ZmgmlYMfeOvZaI6qqtyQoywdi3pLb0gLL21ch/Kr5XGsWefn7wzg+tnl9AdVmXPem8LiLybpUw97mMerhnEdnHIm96kkIlomdFlujLl/8O4pLB64IqHKODHGlm0cKkcZfoRynsK6L55rsy1kjurQm1zOOZoyZQo9/PDD9Pzzz9OIESPY348bN47i8Tg999xzrds+/vhj+vLLL2nXXXftyKkAAAAAAAAAAAAAAAhNh97kmjx5Mt1333306KOPUm1tbavPVl1dHVVVVVFdXR2ddtppdN5551F9fT11796dzjzzTNp111079GVFAAAAAAAAAAAAAAA6Qocmuf7wh3VLTPbZZx+2/Y477qCTTz6ZiIiuvfZa8n2fjjrqKEokEnTggQfS73//+0gqCwAAAAAAAAAAAACARYcmuZy1dl5QWVlJN910E910001ZV4qIyNVsWFfrzLX/Yj1mmPWZhoeBWmubEmUs3wPZDpY/gPQ8sc4tfAeqP1nK4seXba/22bvHRyw+aOf3VJk5Nw9nsYtr/xJpU+HJ9pNr+In0dVu+KB1c32yudc4Rrrp9Pwm1BjiMn4R5X0W7yGvMUlPpuioW/2Tsi6rMsjT3T5EeXCvThs/b2Xwfv3GlKhNUV/LqWpoivk36voVqz2w1tb79vBA5KioKpadRd6xgca9Ja9Q+v9j8CX7Yi/W508TrJ311iLR+GsS6/qTT9yvu8bzWq0zXT1LjJfVGsa3Hh8IXTnoDWGyinuT/55q2/Z7Vp0WnKeknIe59ljlq4Xiuj5/W/0eVyZSjnvxiC7XPkATPSR7pex8InwczR8WQo9jfh9DTZg/y5/fi7Y9QZa4dcz+LpQ+WlSekl1ZguGckRZ9i+d1I/SxJcV+4a16dqPbZ/C88R2355RJVRvV5FYaegsLrSf5/Lin1MVT3z3m8OqhSZaR2X/tmJIsrFjboczfzvkqNpylkfiqCMZT8/1wT1FRS0F6fJz2Fw/zWM35XSG8qd0UPFj9/he53juj5jjiskaPEuOlPi3ZXZT67lXtljX5qDq+L9KAjIk/cR0tT0tvI0hTJ9svG48zQVBgvq9ZnskB9nllH6UedrSdXSvoE8rh8kc4Tv33rQBY/spd+AeiG3/8/FjcZfWfj/hWiTLzdmEh7VS5Jd1dl4k/XsdhL67G7qxRjRdl+YTy5jPkFc55HFXL2/7dD1l9XBAAAAAAAAAAAAACgWMAkFwAAAAAAAAAAAAAoeTDJBQAAAAAAAAAAAABKHkxyAQAAAAAAAAAAAICSp0PG8/kkXRUnr2yduaxpDBfG9E6Yxfkpax9u+Oc5YW5szANKA0PLAE2aylMypU+dFGbLiWYWvrVwiNpl+9p5LB5YsUKVmSNi24CXX5cnynhJw8BQen2GmCJVBsdE3DTUMhDNEenqcvLKWkxTDRNHY5vEE2amXjrEPrLhnLGP1JChqQV717K4xk+oMtJc8MtELxa/csUuap8eX3/CN3Sr0fUTuLhhOlsmNdW+xoi0wakLYSoqDR6JiKjleZPPZg4pFj2d/5szVJle3+N54qxhz6oyTQE3724ItJm3NHqWsTR9JiKq9JPtxkREfWKrWPydZ6eoMqPu4PmwV4qbYAY13BiaKAd6Isqvptr0e6E0ZeXgUP0eJ6oc5YumWh3oeyQ/XpAQGvJe6KlP3bSIl7HMleU+Vr8n2k/lKMNcXPV7ISjZHBVCT4kH+6kix2/Pn9/fHfAXFvcp48+7hWVO/8iycSx++dltVRlP6Hv4E6tZPPYj/rEeIiKvTAx9u3dTZSRZ6SkffR5R3jRVLH1etvmpZiE/zvVz9ldl1iZ5Pur1S56v/C+/0KeOcz2Fy09FOoYiymuOCqriFLT81gujKeP6lKaSmetfsZgbg79y506qTPwUfpwD6t5XZa7+gn/YIj5F93l9ln7KYpdsFrHx+7Ccj8dUzjKwNBUITfnyt125Pm5WHx4omj6vzdxBWUQ5KqmvTba0E7fQc/qaR1/Pf7Md0XCWKnP/QfyjfdYYSn4wQ46prHH5jCb+AY0Xr91Vlen33goWp2v0b4JMcwe+oaesPjxgfTyizXyCJz/utxHwJhcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkqdoPblcud+6Pj2IWWvXxbraQK/fjMl1/MayZ+kZpTykpLcWkVp7bK4NleuTjfXKTu4n1y2/qL1JVm/G1+c+8PkOqsyggB/HWqMfyLX+sj3LDC8y2cSWJ4tcu2ysS2b+ZGnjpuSIjJoSm9T1EpEv9ZJsf93wujKiDcJoytDzkCeWsPjZ72ypyixay327Gm8ZxOK6x95V+7iY8GOL6/XcXg1fA276cQhNKX+AcqOt5CbjumWbU8Jov3T+1/4Xi576vrJYH/glfg9/tcvJ+rjf53o6Y8QrqkzS8S5CrvW/9y/a00Ra66Sr9DVtfu8yFo/urv3lAuGPVBA9yf/PMR3VlIXq95qzyFEh+jSr39v80M/4YYwKS++lP324O4tH3Pux2sc1cX2EyVFmv1cuc52on+UDkU2/15lylNBTr/fWqDJ93uD9+G03HcQLZNnnybYamTL8kBIJEXO/G+V9SkROnlvuQ0afZxBJn7epepL/n0OKpc/LVk/d31nAN5yqj9Mtwf3jpJ6coSfZ/iU9hpL/n2OCeIyC9R5mlodSIPyXTU01i3tvnEd6KUtNDfgn930kIvrwqWEs/iitfZLLpRZTq1UZJ3+TCQ8u12xoSuaFcq0pCjquqbT0VLLyrvK300V80X5ekx4PtHhTe0Ee9VTut/qQyd+5RIYnl5l/+baY4clFGXKU8u4mIq+J3+ctrm5UZU6dezY/TY2u32XH38diOS6X3rlERK//nHvO9X5XuncTufo6UeGOzx24wBikyuaz2ly0p58y5gay8A3Em1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkqdojecpoA1mZTH91074oXmG4Z40O1PmlUTkNXNzMy8hDACzNWCUhm1l+iI8YZTpUrzMwD++q/Z55WZuRj+oRpslUk9uHucny1URJ4zlXUzW1zAwlG1uGBbKTTHtJ8328wwDupyRdhtcKw1NSdNGZxkySk0Z5stZaUqey9Czt5af65vjuqky5UlukhpPcBNns7XT0hTTMMEUxphe0jDBlB8rEJqSmrPq4xnGz05oxDQVXd9+0FPLcfm5er3ylS7zKq/fA+nxuozQgjThHbr2Hb2PwKup1htFjipWPREVt6Ysk+FC5qheFQ0sjnvaPLTSE7nkXf6xDLeGH8MiTI4K0+/Jf+YLpSlDD/LZRo6KRk9qHOUbH8QpE8NYYVjrDKmE6fO8Jp7rvCojR8VFoxZATxs7Rk7oinoS9QtjPF/KYyii/OYoL3Ct53ZmzUIgr88wCvcSQlPS7N36mIFsB8OIW+ksZjwYQlNeTOQoX/eTThrlG+b08rnwjD4vo6Yss3pPmItnqanW9stnn+doIz901v+1/B1r1NxzwgTd0JMvcpR+vg3j9BB6GvY3MVY32u6uP+zKizTJj6/oH95V9D4vU218WEVdg9aT/FiP1I+aSyBSM03W3EEoibSdMzE+5meBN7kAAAAAAAAAAAAAQMmDSS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDxF68kVS6Qo1uKt4KzF/9Y2jlz36aWMNZxy3ayKjbX1ch1tuV5b76QHl7H21hN+AKqE0/V1cl22se7XW8vX4/pG/SSB8JYw19Va29TJRWx4HrRdw+vkAukc0lZTnmUV4TLP+Ur/iIJqKq4fX0+UyZmmjHPLa5CasnzznPR+M26Bk21jPEstmoKeWg4EPa07sD5sR/REVDhNWbgwmipgv/f1CX1Y/Fevrz6M8D4Z0fQ5i11lhdonMk0JstKUgfI16eo5SvndGJrOV44yCKMnEv6DfqMxjooiR1n164CeiPKnKfR5hD4vYvzmFPnplvYy2kR4AFkPuOrzDM825ZEkYtOXT7ZT3MgBwoPLhdCUfHSsp8Y18/zjjBzqrW3ixwnR5zn1W884e4jXX5SmrH1aylheZjki1pSiWFlLjjLmCaSeDLyUuEMhcpTyTDM81NTvYeN+qRxl+QZmM3cg9W/qieeoWIi5A6Uny89b+nZZ+ceTsTFn0nZbSE3hTS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDyY5AIAAAAAAAAAAAAAJQ8muQAAAAAAAAAAAABAyVO0xvN+Y5L8FkM8p83PpOmlaebpRCHLqCyTeZn199IILmaZ2/GmdVaZMm6O6Mn6BobZnTDtdIbBIjVxM0JpekhE5ItzedI8Lq73CcrlNYUwUDXKeG3bxjfqnyO4pgzXVOI6U+avRNoxMiqkpgyzwUg0lY5IU0b95JbINCUNCS0jxta2yPxBiqiAnqhz6KmsPT0R5VVTa1PkxwzD0vUETlyfpakghKii6Pdypamo+r18akrQvqa6QI4KYwpb7DmqcS0/btHqiShfmkKfR52jz2t3DEVUVDkqTJ8XQlNyv1C21TKPGSbtymje+L3lYu1rytSL3GaUcWtFjjL0LGss62tpylWEMKeX+xjnbjFRd+k86qmhjZ4Cwzg9ENcfZu4gKoSelMk8kfpghvk8y49jhBhDyY8zKCN6IiKpJ6N+MaldmX/MHJXZnF5h9HmUxdwB3uQCAAAAAAAAAAAAACUPJrkAAAAAAAAAAAAAQMlTdMsV3fpX4VLpROu2wHhNM0iJV+SMl0+9lHyFWL+e57U5DxGRFzSLE1mvxIlzGctDnHx92Xj91xPX5QW8Lk7WhYicE9uc9VqieC030LfZpcVrk36s3ZiIKEjx9nPOaHPRFr5x7/z0hmtouc8uV6+H0sY0pdstSPG6Wq9Fa03ppUVaU6JMYLwm6on5ZuN1U5cWr4kWVFPGq8lpfg351dS664SeWgpBT+vqsml6IiqkpkL0e5FpKpt+D5racC7kKBZ3hhwl3S7McVTh9USUe01BT2326eT5iajYcpS4PktT8r6m9T3K1Oc5Q1OefA/E/K0nE4WxBC0STelzy2Vq8rjr6ic1JZ9RQ4cpsY8z3ofpSJ8XFEhPKWMMFYtm7sDPOIbKnKOUdoiIVI7KrPcweiKhJ+d02+gcZelJzB2kpZ60VoKUyGPWYmH1DEczd+C5XKouC+bPn09DhgwpdDVAnpk3bx4NHjw4J8eGproe0BOIGmgKRAn0BKImV5qCnromyFEgSqAnEDWZNFV0k1xBENCCBQuotraWVq9eTUOGDKF58+ZR9+7dC121TseqVasK3r7OOVq9ejUNHDiQfMsoNAJaNOWco6FDh0JPOQJ6AlEDTYEogZ5A1HQFTWFcnj+6gp6IkKPyBfQEoqaUNFV0yxV932+dlfPWf4mge/fuEGsOKXT71tXV5fT4LZpatWoVERX+ejs7hW5f6KnzUeg2hqY6F4VuX+ip81HoNs6lpjAuzz+Fbl/kqM5FodsXeup8FLqNw2gKxvMAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHmKepKroqKCLrnkEqqoqCh0VTolXa19u9r15puu1r5d7XoLQVdr4652vfmmq7VvV7veQtDV2rirXW++6Wrt29WuN990tfbtatdbCEqpjYvOeB4AAAAAAAAAAAAAgI5S1G9yAQAAAAAAAAAAAAAQBkxyAQAAAAAAAAAAAICSB5NcAAAAAAAAAAAAAKDkwSQXAAAAAAAAAAAAACh5MMkFAAAAAAAAAAAAAEoeTHIBAAAAAAAAAAAAgJIHk1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McuWBuXPnkud5dNVVV2Use+mll5LneXmoFShVoCcQNdAUiBLoCUQNNAWiBHoCUQI9gaiBpjYdTHJthDvvvJM8z6O33nqr0FXZJF588UXyPM/8M3PmzEJXr8vQWfTUwjvvvEOHH3441dfXU3V1NW299dZ0ww03FLpaXYrOoqn333+fjjnmGNpss82ourqaevfuTXvttRc9/vjjha5alwJ6AlHTWTTVAvq9wgI9gSiBnkDUdBZNdZZxVFmhKwDyw1lnnUU77bQT2zZy5MgC1QaUMv/85z/psMMOox122IEuuugi6tatG3322Wc0f/78QlcNlCBffPEFrV69mk466SQaOHAgNTY20oMPPkiHH3443XLLLXT66acXuoqghICeQC5AvweiBHoCUQI9gSjpLOMoTHJ1Efbcc086+uijC10NUOKsWrWKTjzxRDr00ENp+vTp5Pt4GRRsGocccggdcsghbNuUKVNo3LhxdM0115RMZwqKA+gJRA36PRAl0BOIEugJRE1nGUfhSdgEmpub6eKLL6Zx48ZRXV0d1dTU0J577kkvvPDCRve59tpradiwYVRVVUV77703zZo1K9S57rnnHho3bhxVVVVRfX09HX/88TRv3rwO1Xf16tWUSqU6tA/IH6Wgp/vuu48WLVpEl19+Ofm+Tw0NDRQEQehrBPmlFDRlEYvFaMiQIbRixYqs9ge5AXoCUVMKmkK/VzpATyBKoCcQNaWgKYtSHEdhkmsTWLVqFd166620zz770JVXXkmXXnopLVmyhA488EB69913Vfm7776bbrjhBpo8eTJNnTqVZs2aRfvttx8tWrSo3fNcfvnldOKJJ9KoUaPommuuoXPOOYeee+452muvvUKL7ZRTTqHu3btTZWUl7bvvviW/XrgzUgp6evbZZ6l79+701Vdf0ZgxY6hbt27UvXt3+slPfkJNTU2bcPUgF5SCplpoaGigpUuX0meffUbXXnstPfnkk7T//vtncdUgV0BPIGpKQVPo90oH6AlECfQEoqYUNNVCyY+jHDC54447HBG5N998c6NlUqmUSyQSbNs333zj+vXr50499dTWbXPmzHFE5Kqqqtz8+fNbt7/++uuOiNy5557buu2SSy5xbW/L3LlzXSwWc5dffjk7z3//+19XVlamtktee+01d9RRR7nbbrvNPfroo27atGmuV69errKy0r3zzjvtNwKIjM6ip2233dZVV1e76upqd+aZZ7oHH3zQnXnmmY6I3PHHH99+I4BI6SyaauGMM85wROSIyPm+744++mi3fPnyUPuCTQd6AlHTWTSFfq84gJ5AlEBPIGo6i6ZaKPVxFN7k2gRisRiVl5cTEVEQBLR8+XJKpVK044470jvvvKPKT5o0iQYNGtQajx8/nnbeeWf6xz/+sdFzPPTQQxQEAR177LG0dOnS1j/9+/enUaNGtft6IxHRbrvtRtOnT6dTTz2VDj/8cLrwwgtp5syZ5HkeTZ06NcsrB7mgFPS0Zs0aamxspBNPPJFuuOEGOvLII+mGG26gM844g/7617/S7Nmzs7x6kAtKQVMtnHPOOfTMM8/QXXfdRQcffDCl02lqbm7u4BWDXAI9gagpBU2h3ysdoCcQJdATiJpS0FQLpT6OgvH8JnLXXXfR1VdfTR999BElk8nW7SNGjFBlR40apbaNHj2a/va3v230+LNnzybnnLkvEVE8Hu9wnUeOHElHHHEEPfTQQ5ROpykWi3X4GCA3FLueqqqqiIjou9/9Ltv+ve99j2655RaaMWPGRo8NCkOxa6qFsWPH0tixY4mI6MQTT6SJEyfSYYcdRq+//jp5nhfqGCD3QE8gaopdU+j3SgvoCUQJ9ASiptg11UKpj6MwybUJ3HPPPXTyySfTpEmT6Gc/+xn17duXYrEYTZs2jT777LNIzhEEAXmeR08++aQ5GdWtW7esjjtkyBBqbm6mhoYG6t69+6ZWE0RAKehp4MCB9P7771O/fv3Y9r59+xIR0TfffBNJPUE0lIKmNsbRRx9NZ5xxBn3yySc0ZsyYTa0miADoCURNKWgK/V7pAD2BKIGeQNSUgqY2RqmNozDJtQlMnz6dNttsM3rooYfYjOYll1xilrdeGf3kk09o+PDhGz3H5ptvTs45GjFiBI0ePXqT69zC559/TpWVlVkLHURPKehp3Lhx9Mwzz7QaXLawYMECIiLq06dPh48JckcpaGpjrF27loiIVq5cGdkxwaYBPYGoKQVNod8rHaAnECXQE4iaUtDUxii1cRQ8uTaBltlR51zrttdff51mzJhhln/kkUfoq6++ao3feOMNev311+nggw/e6DmOPPJIisVi9Mtf/pKdp+W8y5Yta7eOS5YsUdvee+89euyxx2jixInk+5BAsVAKejr22GOJiOi2225j22+99VYqKyujffbZp939QX4pBU0tXrxYbUsmk3T33XdTVVUVbbnllu3uD/IH9ASiphQ0hX6vdICeQJRATyBqSkFTnWUchTe5MnD77bfTU089pbafffbZ9O1vf5seeugh+s53vkOHHnoozZkzh26++Wbacsstac2aNWqfkSNH0h577EE/+clPKJFI0HXXXUe9evWiCy64YKPn33zzzenXv/41TZ06lebOnUuTJk2i2tpamjNnDj388MN0+umn0/nnn7/R/Y877jiqqqqi3Xbbjfr27UsffPAB/fGPf6Tq6mq64oorsmsUkDWlrqcddtiBTj31VLr99tsplUrR3nvvTS+++CI98MADNHXqVBo4cGB2DQOyptQ1dcYZZ9CqVator732okGDBtHChQvp3nvvpY8++oiuvvpqvG2aZ6AnEDWlrin0e8UF9ASiBHoCUVPqmuo046jcfbixtGn5DOjG/sybN88FQeB+85vfuGHDhrmKigq3ww47uCeeeMKddNJJbtiwYa3HavkM6O9+9zt39dVXuyFDhriKigq35557uvfee4+dV34GtIUHH3zQ7bHHHq6mpsbV1NS4sWPHusmTJ7uPP/643eu4/vrr3fjx4119fb0rKytzAwYMcCeccIKbPXt2JO0EwtFZ9OScc83Nze7SSy91w4YNc/F43I0cOdJde+21m9pEoIN0Fk395S9/cRMmTHD9+vVzZWVlrmfPnm7ChAnu0UcfjaSdQDigJxA1nUVTzqHfKwagJxAl0BOIms6iqc4yjvKcE++xAQAAAAAAAAAAAABQYsCQCQAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlDya5AAAAAAAAAAAAAEDJg0kuAAAAAAAAAAAAAFDyYJILAAAAAAAAAAAAAJQ8Zbk68E033US/+93vaOHChbTddtvR//t//4/Gjx+fcb8gCGjBggVUW1tLnuflqnqgSHDO0erVq2ngwIHk+7mZc4Wmug7QE4gaaApECfQEoibXmoKeuhbIUSBKoCcQNaE15XLAX//6V1deXu5uv/129/7777sf/ehHrkePHm7RokUZ9503b54jIvzpYn/mzZuXCylCU130D/SEP9AU/hTzH+gJf0pFU9BT1/yDHIU/0BP+FPOfTJrKyZtc11xzDf3oRz+iU045hYiIbr75Zvr73/9Ot99+O1144YWsbCKRoEQi0Ro754iIaO/hZ1CZX7FuW3WFOoeLxVjsBYEq4zWnRJzUlRX7yeNS3GiiNN/HSxrHTfBtzqpfGT+Xq+TX6Woq9amr4rxMmTGDub4NW4glUqqIv1ZsSzTzuqX0PpRK8zhwuowvZtDFNa6r0IZtqaCZXvziFqqtrdXlsmRjmtpn2BlU5pcTEVFQpduWZFta90y0pXnv5X5CU0pj6yqZ+bhSU+nMmqIqrqnA0FQgNBUYmvLErfYtTTUKDcnnLZlZU+ZzImfpLU2t35YKmunFL/+YFz3tPeLH7eco0Y5eWj8vMkeRkaNkblP6sdojCKEneS5DT/LY4XIUz5lBua6fbAsrR3kiR3lhcpTUmJWjYpn15FiOStBLc/OTo9Dvdf5+L599Xi5ylHwO15+Qh1HlqCZxLme0fSY9GdedrhZ9XpY5ym8UdRbPWqnmqFB6qipvt05ERJ5xvzyRI0LlpzKRjyw95So/ibGiqSeZn8qN/CT11JxWRTLqybqmEh+Xsz6vUretGpdbmmoK0U6ZclRU4/Js+jxDU2pcHjdylLjXvhxPUpY5Sh7HyrshclRB+jyMoTr9GIoovKYin+Rqbm6mt99+m6ZOndq6zfd9mjBhAs2YMUOVnzZtGv3yl7/UFfMrqCy2XqixEEL1DCHIMuYrbRmEGrOaSAg1bRxX3DBH1o93fmx5ndZ1e2V8YBFKqIbo/Jh4uGK8vl5gCMyXnbLxo1i2sW8dx0jWEb5eunFNlbcmv8BoW5W0w2jKuveyXcT1hprkCqMpZ2lKHNsXk1zGdQdCU6EmuVI6OftSQ6G0wLVpPyfyOMYz6ct8kA89ZcpR4gckGYMzlW+Mts+Yo4x29cLoSQ4e9aBbtnW4HCUnTa3nPXOO8kSO8kLlKLnNylHyGWn/B2Tr+YpCU+j3iDpHv1cceup4jvLkYJQoux+QoXJU++dZV2bT9ZRtjvJlvpbtWaI5Kpyespzkkm0UKj+JfGTpqdjzk9RT2pjkyqQn85pKfVzefo5SY6IwmrLaKYpJrjxqSo/LQ0xypaPKUZnbPEyOKs4+D2Moos4xhiLKrKnIF8cuXbqU0uk09evXj23v168fLVy4UJWfOnUqrVy5svXPvHnzoq4S6GJAUyBKoCcQNdAUiBLoCUQJ9ASiBpoCUQI9gTDkzHg+LBUVFVRRYcw6Otc6S228sEIkJvScNdMqX++0ZvwyvmZvHFe+Pmj9y4J8LdF6hU/WWdbPqq/YFMSjmaeUb0lQwnodnM/Gyn9FICJy8l95rX8dafMvEtbLI5vKxjRFzrXeK+tfGNUWQ1NOvoJq/au2bBf5aqmlKblMJIymyNCUXBop62doyoltVv2c2i2uykjUv4ZYS/FSvD2t+yLr7NpZrlgoPZn/0iWLy2eMSOUo618lnMxR4v6Y/6IfJkcZr0HrQu3rR2qHSGvOum69TXdF6qrkPs2GDuTzGUJPmZYC5VNTJdXvGWTV72XqB4lUni36fs9aqrD+OguWo4yliPKxC5OjyNvIedqGucpRctkDUeYcZfXjEeUoVRW5j9VXCW3IZ2bduTu2XDFqTW00PwVB65vFofKTpady0Y7Fnp/CjMtlfjLeyNYdmoHj46pQehLLX7PNT23bOEhHb9wdps+zcoDcYvZ5FUU0Lrc0lUlDYcZRZUYZ+YPQylHybTTZflZultus+6LeECuOcXmXHENlMXdg6snr+LiqEGMoovCaivxNrt69e1MsFqNFixax7YsWLaL+/ftHfToAAAAAAAAAAAAAAKKf5CovL6dx48bRc88917otCAJ67rnnaNddd436dAAAAAAAAAAAAAAA5Ga54nnnnUcnnXQS7bjjjjR+/Hi67rrrqKGhofVriwAAAAAAAAAAAAAARElOJrmOO+44WrJkCV188cW0cOFC2n777empp55SZvQAAAAAAAAAAAAAAERBzoznp0yZQlOmTMn+AOl0q7OYl7QMtoUpnzSKI/3ZYRczTAOVeVxmAzdPVifMZ1EtA+RAGswJJ7WkdlbzUpk/0RyIawjKMzte+sL0zTfM+DxRHxfCRM8ynG1777IxusuaZGrDJ0iz1BR5IcwFpaaU0bJxWOmiF0JT0pCciMiTBn8hNOWn+H0MKoxnSRqrllv3jZum+vIaLBNaWR/DhFdhmvCuO3Y+9eSlUhs+l5ts38iViEytONGOrsy4frkpRI5yKVkkhBFrGFNMoSeZE4iIvLR0szZOHZf1se4bP44vP1UcIkeFMdc3DWfbHjufOSpTvycNvS1N5avfC0OYfk9ep/Esqftq5J9S6fcKlqMsI1fxCXNzHCWeVfN5ySJHUdJlLiNzVDqz22yYHOWLcZSlFWnMa/d5GXJUwsiPsj5h9GSZI7NzRW8UbtLR/GRoXWrMuraSy09i3LJiVJUqc8fUa1m8Vbxcldnhav6bacCrq1ls5iexzcxPIT7GUBA9EfFxuZWjZL6xPhLhi+fXur5sxuUysYUwIDc1lcW43BPjcrJylOrzQozLlfG8cU3SVN4al0ujcOsDHy19Xj711BXHUFnMHZD5O0+cpkjHUEThx1F5HL0DAAAAAAAAAAAAAJAbMMkFAAAAAAAAAAAAAEoeTHIBAAAAAAAAAAAAgJInZ55cm0wi2brm1zPWZjq5drvcWKct18j6xtpWuUmuf00ba3GlP4i17jeZ5IdNGWXEOmd5BZaPjvR8MD00KqXngT61i8t1/K7dmIj0OmCXeX266YnTZj/pZZBLXDLVug7fM3wVpB+AszQlvYSMZsq0Rt+zNJXK4ANBREF9LYu/OKJelWnuwY8z5qq5/LhqD+3p4Me1YNKVom3kenwyPEwCrkPf9CoQbWOuP8/sHdWixXzqiZpTG3wIDC8JTzybgdGuTnofGP/uIK/Jy5SzKGSOEjnJiZy1bkdenzBOWsrbLoSHRlBm5RJxbqEDryxzjqIwOcp4HplnQD411dF+z/SiFNds9fLymkP4iqDfW1+/Tez3iilHSf046TlJRE54IIbJ0SpHWe0Ros9TOSqEnlaPH8Li+YfrfU781kwWb1U1X5W54fP9Wbzkbf3hpIGv8GPHV2eRo0LIwWpzZuNp5LCckCk/VQiNGXoKhAeO2UYllp/SfWtYfMn5d6ky0oMrZnjKNPUV+Ug+n51wXL7Ok2vj43Ld51njCTnm0KeR4welIcvKLJscZY6jhFef+Ovluw1Qu1z3q5tYfMb1Z6oyvf+b4Oc2xuVq3CnawQ+MjjKbcbnRgAUZl+diDJXN3EGR5ajIxlDS06zI5w7wJhcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkgeTXAAAAAAAAAAAAACg5Cla43mXDsi5dcZilgGZNMILLB8zaepsmMd5SWGIJo0Gm7Xpm9fUzDc0JVQZZT6YTqsy0tS55Xpb/9q67pg0CTfM44T5oGV4rU0Yxd+njHPLbZZpnmVQ2B7p5sxloiKd3uDWaprayUawzNXlPdOH8aW5YBaaWrzvIFVm55++w+IDqt5QZd5aOYzFK5qF4aVlhiiMhcNoSho4E5HynVSG6Zam5HNhaEpptR28IH96ckFAbv1FWyaITm6yzPpj8v4YJ5KpQ7Tj3El1apfY1itZ3L9O1+/UIW+y+N2GoarMg//5FovrX+Hmuf2e+kLtI/UUs/QU4jlSWpVxDnNU27vi5TFHZer3VPkw/Z71zEtjWWWwq/srb63o59DvtTlO5r6jtUw+9ZSDHGWJzgvEfRZt5ht9HiVEO6xt0uduFmUMPa3deRSLd7v4dRaPq5mj9hlU9g2Lh5Q1qjJHbHM/iyu2jasy++w0iW+4VOTiMHpKGc9IkeaojPkpxHc/pMbM3F9i+WnOaby+Y8uXqDJrhRv60w19VZmhT4k6y3wV4qNFWecndh7DPD1HME0Z90OOA1yIcbn1QSiZo7LSlJWjZB5TSZWUpubd2pPFZ495RO2yOM0/NLVqC31f+77DzyU//kRkPEthPoQl70OYHGUU2fB3BdJTVGOobOYOiixHRTaGUuNy8ff5GEMRhR5H4U0uAAAAAAAAAAAAAFDyYJILAAAAAAAAAAAAAJQ8mOQCAAAAAAAAAAAAACVP0XpyeTGfPH/9HJy1Blv4RKg1tKTXl1prj6VngvRLUmtoidS6bLUmm0ivozW8LsxtbY9rrasVa1m9hF6v68fFthCeLH6TPK5eQ628yEKsd860rtYLjLW5uSIWI/LXr0O2NOWLdc7Si4TCaSqTB5elqSX7cA+uA89+VZWpE74ht328qypT/jz3BBlAH7HYGb4sUlOWf4r04LJ9NKQPi/BAaDI0JZ8do34K67lp8a7Io548zyOvRUeWzmWOsuotNZbSDTvvAO7NcM4Jj7B4eFz7f7y8ZiyLn56/hSpz9ccTWOwZhmDn7vQsi+t2aWDxH4/ZS+1T9xOhd0NPfoL7AShPDTL0lBA6zTZHddCfJJ85KlO/J/9ZKp/93lff5jmqqY8+7mlH/pPFlV7mtvM9Xpe4p/u0a/8yicVDntMeStILMbJ+T26zPDLUThvXlPKvyiH5ylH6WeVtpp5LIj2Okv5bRKqtvaoqVWSLX8/icdUCFt/y5d5qnwWvDmbxwJe13j8/lueon+7xnCrz8Bb3sXj3805n8fCLQ+SoTdQTUf5yVIfzk5HXQ+WnoHjH5R9cor0rb9zlbhavDrR/2wqP12fq9O+rMsMTPK/J/OQbz1HU+Ykoz+Ny39vQ5r6hF/VbL7txufQzU5oKk6Oy1NTn5/Dx1+RRf2dx0ml/pCs/O4jFY29cpcoE1VxnLtDHUblZ9nlrDU01C01ZnlySdp6bUhtDpap4O64ZVK7KlCW4nmLNIm7SY0+5Tf5GItK+oVafXD5nMS+T4Pmw6OYOcpKjwo2j8CYXAAAAAAAAAAAAACh5MMkFAAAAAAAAAAAAAEoeTHIBAAAAAAAAAAAAgJIHk1wAAAAAAAAAAAAAoOQpWuN5ipcR+etM9VxcV9PFhXGxZUYYk2aE+jSeNGgT5mwkDfiIyCVDmKjFhAFguTai9MqyaH5hxuZZxnDCVNgPjDJJaUbIzQdNY0/ZNpaps7zuMm2EyK4hg7lclHjxMvI6pCljDljqLIRpqmy3OT/gprdERCccw01tZ34zQpX54K3hLB7zuzn63N4yHpdrw0S9j7gm02xQGMFaxvPC+FkaWnpNhqbk82U52ktN+YamWs0Z86cnqogT+evaN5Se4iFMeElff6IXb/trZ+3P4pqnuql9+r60kMV9li5SZcgT9TFy1CNbcXP6b85Zw+LTRv5L7fN4D25GL3VBROSLHOUszYn9/LXCzNrKUUa+1icX122YI/MT5VFTeer3gip+7CU7dWfxVqe8r/bZvYZ/xCLmGcaqQr8Vvr4f5SHM6CXTfsCNnadu8x1VZtiVwhC12Pq9QtDRHBXmQwZGjlJGz9Ko2BpHSaN5Yxzl1fKPbsz+bS9V5tQ6noM+b+7D4sbbB6p9NpvJzemD6kpVZsxtPB/+/fH9VJl5v6xn8SljZ7D4uYpd1D6xlWvUNoXMUcZzXpBxVIfzU2YjcXNcLvVUwHH5Bxfx8dqV+92v9qn0+LkbnB53HXnx2Swe8YHWQcYPOHS2/EREXnmcvPZyVCyacXnOfuuJ/Lh6v7GqyM+OfUjv1warL+3xE34uV53duFx9ZEZqyjLTt7ZJpKassXd7Hz3JFRGMoVYN58/v/Rf9TpVZkuYfQVmY5h/9WpziYyoiosaAH7cxXaHKSC1U+/pe3HzvoSyuXMZvfPVirafu7xm/AQRFP3eQBXiTCwAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlDya5AAAAAAAAAAAAAEDJU7SeXK48Ti4Wb/1/SVDOq2763ch1wGHWBVueQJlQa5NJrfX3KvXaW2u9cEbi8rozr1v1rHW1cm25XDMrYyJyKb7Nk74RRERivbx5jW32c5anVa4oj7eu1ba8GFwFb8vA0JRcdh5GUp+d3J/FU77z94z7fH2X9uQa/finfIOhKamPMMh75DJ5FBGpdf7WNi+M54HwbvEsvwU/vKbyqScXb5OjKnSdgnKhJ+vapB2aoafRd6/iuzQ08bhR+LARkUuIdfFGjvLi4hkw9FS+uIHFW/f5msXDy5eofdYOqmFx1VcNqoyqi+WhITUWIkcpzwzrAVU51NBTrDA5Khf93tJtq1WZzb4/m8VT+nNfo+Vp7fO2Is2PEzh97mqf665HTN976V8Thu4+1/z/bvO0KvP7rY9icf2s1apMxn5PeklRyH5P+EdY965Fi0Wdo0L0eVaO0id27cdE+tm0xlHC72bLgQtVkV4x7m30p8V7srjnm4t19SqFZ1KIPq9y0Vq17YPzt2Hxe7W8jWsal+sDZZOjLD0VIEe5inJysfX+SZUh8lOZ4RkYZlyezThcEmJcTnW1qshXB/Zj8fl7Pc7iHn6j2ictcuFP7j9dldnMyEcS5V9ZgPxEVFx9Xq7G5QrLJ0gSQlOTp/1NFZH9ouTyy05S23qJsVWocblxDRk1Zf3WE8fxrAaVv/Xa7fNCtG1ERDKGOoKPqa2WT0fg/xv3ddtLD65aX/c7l518Dz+O8DmdvnRHtc+rb2zJ4gGv6frUfs770qzmDoosR+FNLgAAAAAAAAAAAABQ8mCSCwAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlT9F6clF5nKjVS0Kvg3Yxvh7W+dba//bjdfvxeT61VtRai1xWJooYZSq454OrrtTnln5aIRaSy/paXhKJeu6tk+ym2y8QXgnpCh43dzfqIpZVO0M9PT7lfitV87UPgWuzrj3Qy3dzRtu12nJdNhGRk2vMDU2pZdjGPZs/sZ7FFx9zf8a6XTRjEou3eP4rXUh4Jrkqw+dNrmMOM40tPa+stf/yOq21/3Kttigj1/mv3ygOYjzrFfyanOVFVtbiyaX/Kme0zVFh9GT4kygMPclcp3KUsb7dKwuR2kPoacU2PVj8+4G3svj1piFqn6qvuQ+Tlb8lyn+LiEisufekr4PUGxE54eVitYN8RlxVuS7T5hkI8u0bGHG/F/u29mz7xWDuC7gw3Z3Fvkz2pL20PmoaoMr88+HxLE5X6LZzIkUdMXEmi4/p+YZxbn6v+8XmqjLX1MsO3/KT6LimFMbzpnwNDb+iln4vn31e3nKU1/EcJf1trHHUnB/w/PLnodepMrU+1+XHsweyeGyN9oVTz5HpBSmuKaX1VL6EH7viS+G9s5Z7yRGVeI6KlxHF1tXX8oOVY4cgFmJMG2ZcLrVherp0fFy+aK++qsjd/3ONXdF2OPq+c1k84pE1qozM1Sr3kJGfpL9NHvITUZ5zVCZNyRwVQlPmb6lMXnCW31aIcdTs8zZncX3sBVVGekp+78mfsnjMe9+ofYJqMR7L0ivXk/5aYfo86cVk9R1SUxU6RxV8XJ7lGGr5wjoWx7bTp4mLMYkcM5V72m8rEJqTxyDSHlx9y/RvaFlGHueMfi+qfaYc/hyLT1l0tirTbW4WOaoAYyii8DkKb3IBAAAAAAAAAAAAgJIHk1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASp7iNZ73vA3GgIYnsSeN8QzfTWUnZxkWCjM/ZdxtGWyXZTaMl8cJKg3jUGGKJ81bl21brfZZuTc3nLtqp+mqzKj4EhZL014ioiXpKn6uoIbHqW5qnyDEnOg/l23Fj/M/2pg6aNPmaWmsmUvaaMozTIml2aLSGGmTQss09X9+yO/J5vHFLP7eE9x0koho7A28jGWaLI3BLTPaoEI80qq+RoVlGcOIUbaFb7Vfho82KDNiowzJ549ImcUG1YYh4XrD0rRlYJ4jnOdtaE+rPaTBsGEOrD6OYeQoZbwq73GgdeBJI2UrR4mcFHTTH8dYfQw3vewT4+f+9fuHqH0GVIuPbhjXJNvCN541T1yCMu61DPflhrhhmCoMLdM1Rm6Ot81R+dNULvq9M0c+r8pUiD5BGpneuWx3tc+Ks7mhd2yZNkQdXs7zmDLPJd3vzdm9F4v79xbm3URUKfT7fONgXWaZaAzL3FTmpBCG1kpTRo6SprtBtZGb1/d7+ezzMuUoqaesc1Q8NznqJ9/lH0jo4TerMnIsM/QJ8VEd6/mWfV6WOUr1p7nKUYaeXHkBclQH85NnVUs2QDbjckvLIfT0zXj+sYyxp36oytT73DBaWlcf8u8fqn2G/YPnz6DK+Gkl9aSlnLf8lK4x8nJZEfR5FnIcFeI5ND/IIjTlyY87hNDU8t0HqTI3HcU/xtMrpj90ce3XE1k86s9cL9b9yGZcPvuHegznNfNrGDya99EL395M7dP3HX7/6976WpWRH4AKum28z0sV8RjK0lP9W/z5jR+ojyN/V8uPC1i/l5s84/eNoDbGtVEf0x+xqBUfAaoUidZq7biQz2Un36PKXDX3eyyum218tKUIxlBE4cdReJMLAAAAAAAAAAAAAJQ8mOQCAAAAAAAAAAAAACVPhye5Xn75ZTrssMNo4MCB5HkePfLII+zvnXN08cUX04ABA6iqqoomTJhAs2fPjqq+AAAAAAAAAAAAAAAoOuzJ1dDQQNtttx2deuqpdOSRR6q//+1vf0s33HAD3XXXXTRixAi66KKL6MADD6QPPviAKiv1euGNkkoTuXVrLkPNxFmWOGWZ9wyEl4TvpKeR3mf5Nj1YvPRbuszm289n8fDar1SZi/o/w+Jan6/+r/T07akQa3rnp/R63dNmH8/iT2YPVGWq5vFjd/+Cr+KNGX4B3T9eyWJvrS4UdOf+GOlu+hqCsg0rdNMp6XiQQ9poynIAkDWxfLsCoalPT9Jri/er/pzFx846mcVjbtdeNtKXw8V0uzhZJm6UkV4WwlNg3gH6Gey5E1+jf+WYB1WZzcq4zpYH+r4uSXNft4eX78jip17UD8roP2X2Igsq+bksv4sNfjf505OXTpNH69eFaxshrSfLSyKeOUfJ+6yayPKzSGX2MJHHnfvtWlXmvZ2vZ/HH3AqAetyrvftS1WKtvFE/T3o0OH3fvCQv48lnwvCykSgvFyJyFcIvsUKfO93G7yZVoBwVVb935V3HqjJjDub/8PT+yyNZPOQ5LejyRp4DpKcbkZGTjD740+O41+QTQx5jcZ2v27va4+d6edUYVaZ2Pu+PgnJ9nFiSa1PmVM/w4JTPjtyHiMiJHJWu3Hi/V6gc5TUZYpGYeSI/OWr5Dj1VkQk1f2FxhXGYE18/lcVDV3NPpbTxfAfSoCSqHJXMUY6qtK6hjT+JMV7ICUHQarTlN6cyFCblc0oU0bjc8vFKZm7rr/fm9+tvg59QZWqEGeTVy3Zmcfdbu6t90tX8uJZXq58SXkDW41gWQX6y9CTHUNYzwTze8tjnpQNqcRHyE0n113Icbo2j5LjcIpB9kzxuCE39+ld/UkUqlT+S9gpafP4wfmrpYWh6sfJtC8dr366Djp7J4lv7vKzKLEzz/Zan+ZhtxXDt/UxH8fCa3xyvitTPWsViU1Pr70uQr/xEFMkYqv4D7q914vfPVGW8NH8W/WZ+3z0jP+o+JfO4fPb/Vqkyj+9+E4t7CB8s3+jP5HzC4TXfqDLTavh+xTqGIgqfozo8yXXwwQfTwQcfbP6dc46uu+46+sUvfkFHHHEEERHdfffd1K9fP3rkkUfo+OP1gwIAAAAAAAAAAAAAwKYS6dcV58yZQwsXLqQJEya0bqurq6Odd96ZZsyYYU5yJRIJSiQ2/KvxqlWrVBkAOgI0BaIEegJRA02BKIGeQJRATyBqoCkQJdATCEOkxvMLFy4kIqJ+/fqx7f369Wv9O8m0adOorq6u9c+QIUOirBLogkBTIEqgJxA10BSIEugJRAn0BKIGmgJRAj2BMBT864pTp06llStXtv6ZN29eoasEShxoCkQJ9ASiBpoCUQI9gSiBnkDUQFMgSqAnEIZIlyv279+fiIgWLVpEAwYMaN2+aNEi2n777c19KioqqKJCG+p5qRR5wXpjsWZtRugnhNFaQpstBlV8m2nU7XOjNWmK+fWe9Wqf3595I4tXBNq47+kV27C4wtcmdL9fthuLX128OYuXrNKmzjX/5NsqVmrjuvJV3Bhu5FrdfrHGtSz2G4WJvNHmyiTPMLg0DWbzTLaa8pr4vfcMTXnCbPme/W9XZeTMcfLRPix2MeO12nJh3icNCskwwjfKNIzg5u87XPAuix8d8Irap9rn17QyWKvKfNjMzQ+bnG6buDDc/G6vGSw++DvvqX1uGb8Pi5Nn6+et0JramJ4omSJaryfLMNXJHGU8LzJHWUaP8vqlKa9nfKDC84Uhclqboc79Djd6/ttJ16gynya55r7/+/9hcd8V2qC8rWk7kW2q6glzXz+d2fBamVWGMLikMiPnFz5F5bXfG/Sibqc1L/GPkQwPGng9rPshcpSVfySpGl2/nx/0CIt7CK3GjGFJ3OP3sco3PnoizaqN+skxgDQKd8YHNcJoqhRylDTKJSLym/m1RJajyjqeo6pO+lofV/BWor/a1v15Pv7y0tw02DR1luM+5CjGRvNTc5K8FpNjwzQ5m/wUyLxCpL4MJE3DLWNlqcH5E3qoIrdM5MbhjUZDL0jz+j19wx4s7t5g5B7V5+nqBaJ+Vo71o8hPMePdhWIelyeT5KU3rik1Lm/WHzzxxIcanGGC7kQbqN+DITQlTeYtvv3k2Wrb6JQYU8u6lOlzB2Lbsce/qMr8ovcsFn8T6ONcs+BAFr/xzigWX3fQn/W5xa8YZ70OU6R9XrHMHZBn/D72xTjcGkOJTSOv0vnmwW34R7xO6/EWiyuN95d80Z/JMRURkUwvli5LbQwV6ZtcI0aMoP79+9Nzzz3Xum3VqlX0+uuv06677hrlqQAAAAAAAAAAAAAAaKXDb3KtWbOGPv3009Z4zpw59O6771J9fT0NHTqUzjnnHPr1r39No0aNohEjRtBFF11EAwcOpEmTJkVZbwAAAAAAAAAAAAAAWunwJNdbb71F++67b2t83nnnERHRSSedRHfeeSddcMEF1NDQQKeffjqtWLGC9thjD3rqqaeosrIyuloDAAAAAAAAAAAAANCGDk9y7bPPPuSMddMteJ5Hl112GV122WWbVDFKp4nc+vWrln+SWuOpL8VP8vW6rlqv35U+I2qJvnGpNWJd9lkfHq7K9LtA7BhYfg783NWVPB5cqdetpiv5+lzpBWAh16Kv2yZi4UnhWetjZZsbHh9ekq85jjUZXhdt1i67lPYKyhmp9IY10UnDQ0lcn5ewNMW91So97bW2OhAeApmta3TbWl4eoq3SPapUkR/88gkW7109m8VrjGd362d/zOIe/9LPSd0c3l7WWu1Ed67XlSN5O+x3+NtqnzMGvcjiCy85UpUZfAlvY983NLV+nXg+9ZTR4022dVyv0Zc5yjNylPLAkV4Ahl/Al4fUsrjfbgtUmbtGXc/PY/iTxDx+DXX78a/kfj6it9qn2xz+3PR63/CxkPnHyjfyPkvfHN/IffK5MZ4jL8m3+QmtGS/V5rrzmaNy0O9Zmsrc7228j289rpWjmnlbzdtfn3tsBdfi8kDqV+fUCuH3d2TPt1SZ187ajMWNz/VVZQa8sprFuer3pO8HEZGXWnfsLpGj1HmsbXzj+N5fqDIrA36u8/5xgioz5vVvWOwq5DXoZ0Te9yBb4w6Zo+RxssxRvniOnOGz1DZH5U1TUeSnFP8Hb69KeyxlzE8GzfV8PHTY919VZcpFHlmS1mOo0//9AxYP+e8aFn+zhfbKXTmSx/1n6hymrsFKNXkal9v5acN+LqXrnzM6Oi63fJaqxG+9dGZNhfEA+ugMPo6yvGiTjh93i+uW63OL33quQt4jfdzH/nwzi6V3LhGRtHUb/8KZqszo/8fbq347fu9jB2u9XPP5RBb3eW2xKuPEc2tral0F8/o7r2jmDrIcQ4m2WnCA9inuXcbHMXIMVSu9v4go7nhbWJ5cks4wd1DwrysCAAAAAAAAAAAAALCpYJILAAAAAAAAAAAAAJQ8mOQCAAAAAAAAAAAAACUPJrkAAAAAAAAAAAAAQMnTYeP5QuBMc7ZU+zEZBnMGbc0WibSJ2sAXV6h9Jh/0XRZfMvYJVeaSK7gZfc+btFll1adL+bkDbSyokObLIfCkOyERecL4TbaDadgsjd4M8zhpMOfL4xKR38Y8jtKGKXUecFbdhYYs801pFLgsXaPKvNQ0hMX1H6zlx7CMroWhvWWuSaLM2n7aDHFgnJvwPt2wJYunX3yg2mfzxfJjBs2qjMRv0G1TuaCRxT3f5ceZ9fa2ap+tr5rP4l9t/Zgq88e13AQzZujZrdeUS2eue2R43gbzUiPXOGmimmWO8lPclLS5L9fczle/qfa5qucbLF4R6K/b/lfo9KUVo1WZqhi/hgMGfMTin2ylz93L52a+v18xQpW55vUDWDz4cd0VxVclWBxVjvLFR0C8tH4eXWyDKacrUI7KpaYy9XvmPsIY1EtkzlE0ukEVWZbmfWFa/FubTzyPEBFV+/y5Hl+h+8pXtrufxd9s06TK7Dbgf1i82YM8N5c1cs0RZdfvxYx+j1r6vXzqKRc5yjiOnxL3I4SeGoZzU+fNK7W58YqgmsX9ZurjeOKeWeMdidS78wzTZHEYz/h4kHyO1LmN3KL0ZBnoinPFjFxX6BwV2bjcyleiHT2pJ2Of+UfwvPKz2o9VGZl7/t04TJ96Vh2LD7rzHyzepnKe2udb5dwIevYPtJH4IyvGsfhfP99ZlSmTfZ7UU0Tj8nbzE1F+c5TvtX7AwfyoWZhxeQhNybaUOWDJ+J5qn1sP/COLVxjj/akPfZ/Foxq+1OeWxvPigxq/ffA2tc8ax+/rJQu1Xp68f1cWb/6a7jsXj+Oav+Znt7BY5lgiosQd/Vlc06w/XCTxrby7vo3zOi5vSycYQzXupO9pfRn/GMYSocuk4+MaIqI4yXPp8dEup/6bxbPP20KVKbW5A7zJBQAAAAAAAAAAAABKHkxyAQAAAAAAAAAAAICSB5NcAAAAAAAAAAAAAKDkKVpPLldWRi62rnpyDSgRkZP+CIHhqyJ8jTzPWDMr14rGMs/71U3h8fmnnaTKfPeQl1lccZVe93vrK/uweMwfV7HYazbWCsdjLPaNa5Lrif2k4QshfaFkW5lrhTOvgfWEf5Bai0tErk0be2ntgZIz4mVE/npNWR4XUkOGB4dcz/3vtcNVmTGVX7N4zWDuh9Tz6xW6btK7I6HXsMtzzz9A129Fmq+vv3n6ISwe8fFytU/QTfipWI+AaC6/Wd9XX+hVaqjmP3pd/9WPce+6C494WJXx1gqNWB4m6zWVTz1lylFq3bmVo8Q99dYaz7O43q93rWfx4XV8HT0R0btNg1n8yzcPU2VG3sSPG5uzUJ87zn1z5tUNZfHT2+yl9lm4J9fl44dep8qcedAXLH5u75gqc/Yfz2Dx0Id4/bwmw+fB8FeQeEnR7aW1f4rnb3gI/FLTVJh+Tx7HD+EnIZ87I0d99uPNWHzuNo+qMjJHScpJX1NNwD0mAk/fkzRx3a028vc9R97E4p9+yjvzAX9fpisk+z2rzxVt016/V3I5StxnS0/y2C7EOKq5lnsf9Yhp/7bFKZ5/Ys2G147IA07Ur2F0D7XPgj15meP2/ZcqM/2jHVg8/AZVRI/RwoyjwuQo5WFSHDnKxePkYuvqkstxeSY9NY7upfa55vg7WGzlmd99wr0gY9P1cS77v3tZ3CvG/W/mNvdR+yQd71P2rVqjymzd5y0W7/7jLVWZqv8TG7IZl29ifiIqYI6y6iXHxpbusunzhKZWbKFzi/RwW52uUmWEXaTKl0T6N1njLbx+qw0/5hqPezE9fe+uqsyAN3i/+OlJOk/csz/v8+Q1/WnfvdU+9ak5aptEaUr6cRKRK1s3riu5Pq+AYygnyqRTui+VOqz0eH19T193teP3J+b0dV/U71kW3/17PR66+Q2uly2mcZ0WW47Cm1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASp6i9eSieBnR+nW11hpsT/gaOMvnIC28Goy1tyT8Wcy1twIn6jPyyg9UmTfv3ILFH57fU5U5b5+nWPzgKO4BUfEL7kdBZPhrWdUVS8v9JsPbS/psSF8LY60wJY1tEnEfpEcXEZEX2+C/4wWFWqttaKFMPA4Jo27Ci+quvx2gipz1Xe5Ds7YPn0vu0cDXMBOR9uRq1uua1+w7lsU/3fM5VeadhuEsHvEg9+CyrtsTmjI9MqTPW8J43qRmROyamtQuZY38XCsNHw23lvsOtKepfOqJ5Shj/bgnPB/M/CPus/KfICJq5n5VI/7Cvc1+9uFP1S5Vi3g7jPlgrj6uPFdM+2LJtfNeI7+HPf77jdql+2fcX+KnT52tyizegT9rfzjpZlXmqh/dxuLfvXUCiyv+86Wub5gcJZ5z2ZcQEWuLgmkqp/1eBs8JIwdIrwirrX2RthqDCl3G47lkeZo/J6+sHqP26VvO/SqHlS9VZXaq+IrFKwyfkxUBzy8HnfEai2fM3VntUzXzE7VNoTTViXKUuM/OGT4nIke19YvaGGVNQgepbqpMWvw7rJ/WvjlyzLF0d+4beMkvuFcTEdF2Qj+1vs595/SeweLv9vmuKlP2U+FBKsdRTca9LuUcVRbbcN64/vmQr/w0/3vcz42IaGGyB4ut3JN4tTeLh7yr+6+mgN/Tq786kMWBMZZPV/Nc89z176oy5/V+hcX3bKd1+YM9z2PxgMd5H5f1uLwD+YmouHKU8rCzfJZkjrI0lRA5SuS+Cw55TO2yJNWdxY1GnzLid7P4BsOP8LCX+G/E+jLu2dbk9P045oqfsbj/u9qzsP813Dvr8v76N8H/fno0ixO3DmBxj6bZap8wvoFaU0Y+KHSfV4JjKOkFHVuo89jDi/hcwemDXhKH0BpMim1NhieX7F4Prv2vKrP13vNYvGx33m//vyuPUfv0eVjPkShylKPwJhcAAAAAAAAAAAAAKHkwyQUAAAAAAAAAAAAASh5McgEAAAAAAAAAAACAkgeTXAAAAAAAAAAAAACg5Cle4/nAEbUY1DptNurkNsOw2TVLw1R9HGkW50IYplJgGB9K5n/Nwi0uXK6K3H7coSz+5bl3svjs07+n9hlzEzfh9pNGXVJ8m7dWm9t5a7lpmzT3dmu1SbhLCoM+3zAoLxfGjEabUxuTVxdog/Wc4ZxdnxaE4Z/8wACR1tSIu+epMv86ZHMW9/0ONw9NfDBE7RN/bZbaJmmu5dqs9LRhYnPAH2l/tbivldo40xNug+rjBkTkZaOpRmEYX6NN5cfs/xmLX1w2WpVxTav5hsC4h+vNPl0QwoQ1KjLpSSKfH9J6CnXa1dy0tNtTizPvY22MtW/ESkT6YwwhPswh82PlorWqyLC/8zLnrPixKnP5WbezeOF4bsA5dKb+gIM0ETU/olAunnNVgtg1uCCECWtUZOj3FPns9yRGPzjsem5Ueufag1SZtEhBMeEfOvgpbSr/3wpuQtowXJuU73kxNwrfp/ZDVWZZmu8nDe1TVbodAtEXmh+niYcYE6w3nS2qPk8SKkdlNnwNo6eyRt5mq4NKVUb2cc5o+7VbD2bx4ee9wOL+MX6PiYgWpHkuueAjbZZ725h7WFxVpu+ba+TPnxpHGcbzkeWoNibGedNUgcblwRZ8THXGdi+rfRYluRn97f/cV5UZfetHLJ7747GqTLnH6/zxe/xDBmPW6rE8icf/H9N3VUVOPZ1/5GJZUKXKrO0j8nJXGJe31VQYItLUF+d9i8VfN+uxfCC+7PXy/+6mylQG3FT78/O30WV83hfFxBP9oyd+pPYZM2Mli+tuWqjKnN3/WRZf8NnRqkzzn4TR/DMfs9gZH9iKTFPrf0MV0xiqoHMHEmsuIc7HOqOu/FgVcZW8r/zZNfy+l5fr9q6M820/2vxVVWb7yi9YvML4EJj8wMfydA2LV/FUTUREfSp4f+tW6j65I2MoovA5Cm9yAQAAAAAAAAAAAICSB5NcAAAAAAAAAAAAAKDkwSQXAAAAAAAAAAAAACh5itaTy0smyUuvn4NL6TWzch2ts9bVCu8Dq8yCc/na+cG3v89Ps4Kvi15XOb4WV61NJiJfbPOq9Pr7Af/kvl1PnbIti+/Z749qn1/feALfYPknNQvPh4T2/XFNfG2/XOsfGF4S5IQvU1lcFVH+EmE8fPJERk0FYpulKbFWPfWFXsf/xW/Gs3jI1E94gV9zTyUiojnX7cDiHq99qcrUfdLA4rjhyTWxB/f2mrYnP26vtw0/CbHe3ItIU1411/yHP9NeZEfVvs7ix/6+iyqzGf2HbzD8AQqB15wkz9+4nqSnm7WuX5VJGevMO+Kp01o53ka+WBNPRORV8XX9XqX2xFEebnHeZbiYcS/E0nrP8B3whZ4GPf6VKjP9ezuxONld+NRYXhLymbVylNpQPP/WkylHRaUpa7/MlQuhKeHhNuT+L1SZzJriXnFERJ6ob7c5q1WZfy3ZjMXjauaqMg0Br/OaNNd8rDmzD6OlKb88RE5qzVv5y1/FnKN80daB089htTBsS/5omSqTEn4+E2u5L1zc0/f05N+fx+K+E+erMqsD4QOX1OO8qgT3BZQeXDnNUawfzI+m8jUulzpcvBv321qarFX7xETHM/zvht+l8PNsGqDPXR/j47M+b4m6Gd476W5cG+MP/68qs9rx+2z53fSeJT3eMC7PSlMpw1tQxL6Qx7JkDUniwp+t4oX/qDINB23H4guPm67K9IjxPPE/f+e/4+r/q+/HLx++S22TSB/DxtsHqjI9n5/NYpmTstWUokg0lfl3nhhHFvkYypo7kGOozS8V3n3GGEr6695w4CRVpHEgv+4rDv6LKiM1J8dQ3z6Y/6YjIvrgr9xr2S1focp0bAxFFLbPK57RPQAAAAAAAAAAAAAAWYJJLgAAAAAAAAAAAABQ8mCSCwAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlT9Eaz1NT84YpuBAGb14IozWL5h782Inp3OAydf0YtU/Vk++w2DSPE6bbrlqbOn89sR+L/6f+MRY/tvJbusKyLay2EaZ52ZjCeqa5N29jr9wwIxRt4VVos9a25nuGJ2zuaKOpUKaBlgGyNEk2PHgrH3+DxW/uzD9ucPDBb6p9as5/j8XzftJTlVl2bzcWSzNLIqL+Mf6hhGUHcuPS+neN+yrbwronUlMyJqLGXTZn8ZJTef2O2Wym2ue1RdwsevMbP9PnlgbpxvPWoikvyKP5ZaJ5gxGiYYaqzDstY86Y+HcGpzWnjFazyIfWxzGkoaWrMQwuK0QODWFgK03CzfoKM/rE8F6qyNiaD1j8Gm2tjyPPra7byFFST0bbsL8PsjAYzZbmJFGLsXEpaipEv+fkPZH/1GZdk6jfkvE9VJFzhz3N4qRx3XLbR2v6s7h8uTarDqUp2e9VGjmq5e/yqacizlHlL/I+75YX9lNl/t8h3Hz5urH3qzK1wkHaFxbTR99yvtqn7394x/2DU3Xf1CyEOe8/A1SZ0W622paJks5RmfKTINtxudKTkKU0mSciivt8n3S5/jf8FRP5eP7Ph/xelanxuDaWjON/v3iiNoz/+143sjhpvD+wOuD38KL3j1BlBr38KYvlXY1sXN5OfiIq4Ljc+EiN+jBSVJoKgS8+amH91lu4C69Pn7JVqkytz8fh9x5+Ey9weOa6WP3Zr+Z8m8X1r+gPaGRjkB4mR8kPFZm/9Vr+DmOodecqojHUkCf1R8i8BP9gwyXLvq/K/OjYp/SxM/Dx+fyjDmMmR5GjwmkKb3IBAAAAAAAAAAAAgJIHk1wAAAAAAAAAAAAAoOTBJBcAAAAAAAAAAAAAKHmK1pPLBQE50xxoPXJddpXhnyTWrXqplCqz2TXc92XBkq1Y/O1fvaL2ue+YnVhcN1OvmZW+NN/spM2b7tqH+wEEjs85vnAd93IiIuqV/qbd85j4ei7Ti4s1sVXiONaaXnEcuXbY2qbWDhOxtcGu48vks8alA3KuHU35Yr10ZWZNUVJrSq7V3mzaf1j8WE9h8kBEZ+z1PIvH185RZfpfyMv0ia1WZSo9fu7pu93M4hMOPlftE1/D47q5WqvfjObXndpjpSrz3VH8WekW4z4Ejy7YTu1Td3IDiy3fBK+Ge2A4a+3/ek3lU0+UDto9oSd8Ipzl8SZzlOF15mTeEm205sBt1D5lP13I4uoTE/q4VXzNu6vUz6orE7lDVM8L4aHRXhpvYc5pOo9tUfkVP4y0B7O8EFW/YORmmaMqDT0FbXNU/nzeXCpNzl93vy2viHxpSnqmrTsubycr/0v/CFNT0pdC6SWzYBr76rbpE+NeKI1O66PG8efgtf+MZvHY/7yvTya86za13+uKOSrMuUf/WXtM9vg23yb9t4h0n5cWBk4xnfpoyen8uN+q/FKVWZiqZXHfN1QRnYOy8Y8Nk6MsPbX9+zzlqLb5ySIrPRnjcic0lhbNXFe2Vu1T4fNxy6op2htp2z5fs9jSU1x0WI9951peF2kQRkQJ4dkj/dyIiM6c9V0WD/o/I8dm0lNnHJdn+q0n/bb83GiqZ1znnwqPH2fS64tUmf5lb7G4h6+PUymOIzUkf/sREV32BTfqWvjn4apM31eX8A1leuyufJ+El5EfRlOVRo6qKc4+r5jHUG4r7ltcds0yVaZXBf9d9+7dw1WZfjP4769sxlBeOnNftf9hb6tt8nedlQ8lA/7O29zrVqPK5CpH4U0uAAAAAAAAAAAAAFDyYJILAAAAAAAAAAAAAJQ8HZrkmjZtGu20005UW1tLffv2pUmTJtHHH3/MyjQ1NdHkyZOpV69e1K1bNzrqqKNo0SL9iicAAAAAAAAAAAAAAFHRoUmul156iSZPnkwzZ86kZ555hpLJJE2cOJEaGjb46px77rn0+OOP0wMPPEAvvfQSLViwgI488sjIKw4AAAAAAAAAAAAAQAsdMp5/6qmnWHznnXdS37596e2336a99tqLVq5cSbfddhvdd999tN9++xER0R133EFbbLEFzZw5k3bZZZfQ5/LKYuT566tnGKeTL8zOrDISy4xNGMoNmv45i5/9Zne1y3fO5Q6ke+/6kSrTV5iCS+NBIqKbFu/H4ndv2ZbFfd5YqusrsUwDY7wtLAM3acinzFCl2aOxD1VqY19p4uzKDYml25rHhXCljoh1mlp/naamxLZYCE0Z9Zdm3NKgcItff6H2ueu0A1jcvIU2Vj1i7HssPq7+dVUmKQwsk8IQ9fdn8I8dEBE1BVwfSdJGjLU+NxuUxyUiWpLqzuL/e3MSi0fcprXq3HwWS5N5IiIndOYqDG22aCqPeqKyGFFLjgrxHHpZ5igvJRwWhcHl0u9po9PN49xQ1xnmxhSXTu7GNUijeXFuLxWivY3jfnheDxbfv7vW5WMrv8Xi4f/gDtKmeaU8l3HdynDfylFt7oP1MYRckbHfk5qy7pnE0pR8TsQ1Omsf2SdYxqBlIi9Yxq/SJFVoSGqMiOjDn9Wx+JLdpqsy5Z7cTzuO+zF+rpF/4Ua9oTS1if1ePvu8DueoMHoyzNUz5SjTkF2MOWLLrA+p8L7TJ30caXzbJPqmq376J7VPrxj/4InsN4mIznj+ZBZv8R9tEixzicrxYcZRYXJUXPe33IQ3PzmqUPkpJbzUB8T5B5iI9PN/2RaPqTK9fH7fY4aeggx6anL6nl49/0AWf/WnkapMn9n83BToMZ7Sk2y/bMflYfRU0HF5+BxllpGEyFH1H/N4QHyF2icuNNUjpsdaNT7vZyxN6RzF+87L53xb7RP7Kc8LfX3j96D4HRzVbz15XC9LTRViXF7MY6jv3vs0i2uMD1/0EH3TT372vCojeXIV/6jXs5fuqcp8NYmPdf60x12qjNSu7+lrWJ7uxuJKj1/DB2sGqn1qvuK/Hz2rz8vR3MEmfV1x5cp1Dv/19fVERPT2229TMpmkCRMmtJYZO3YsDR06lGbMmGFOciUSCUokNiSJVav0F1EA6AjQFIgS6AlEDTQFogR6AlECPYGogaZAlEBPIAxZG88HQUDnnHMO7b777rT11lsTEdHChQupvLycevTowcr269ePFi5caBxlnc9XXV1d658hQ4ZkWyUAiAiaAtECPYGogaZAlEBPIEqgJxA10BSIEugJhCHrSa7JkyfTrFmz6K9//esmVWDq1Km0cuXK1j/z5s3bpOMBAE2BKIGeQNRAUyBKoCcQJdATiBpoCkQJ9ATCkNVyxSlTptATTzxBL7/8Mg0ePLh1e//+/am5uZlWrFjB3uZatGgR9e/f3zxWRUUFVVRYHhcV5GLrt4dYVxvKk8vyhciwrrbXv/QbaB/8dwyL39h8J1Xmm9F8fXLFcn3ufq9yj4c+zfxcas0vkVqP7qQHCpHy2rE8jJR3ilyDHcKPyklPHyJycr23ud7ZuA8RsjFNUVUFkb9uu7OuLxtPrhCaUn4lxj4j7v2KFzHO/X6PLVl86gS9/LdqD75u//sjuH9cpcfXZRMR9Y+vzFhmRsMoFj/45faqzNoX+rB47EMLeIGk9qWjcrEO29BUGJ+fXGoqqhzlQuQo5VdEpP0AhJ5qnuZr5ImIzp96L4svvFF//CP4Wy8Wd5+rPYzKFwmfHHHudC997gV78m3d99c59NrN/8LiFWntxfbY3dxXYPB8kR8tnzGBmaPKxH2xfKMyHnnTKHS/l+7Gz/3FobUsHrAbz0dERAtfHcQPG9PH7f0frtXuH2jvnC+O6M3ipr58n6u/fY/aJy68meKkPYhkmWpf6/nOpVxTfkJ4zBneI5Ks+z1/fXvlIFUVPEfJPi6QPmt6LONC+KrOXLs5i3ep+kyVSbv2/W4sVgc8d3zSrMen/V7i99TyCJEeXDKXePIaifQYLkyOMtrG8q6Lio3qqapyg57C+CcZ9VZtFMIrd+jf+RjlqrXHql32+e6bLD6i5zv6sCKzNxj+Wv9q5GOd61+ZyOLhj2j9V369hsW91i5WZdS43PIwyqSnMONyw5dJ6ae9/ERUuBxl1SvM+E9uC6Gp7u/ye3Tn5YepXcpPXMTi/938SX1Y4efXQOW6TMDLyBy16rbBJKlv/prF5thYaioW5pkUmjK1kNnrK0yf15KarTHWplLoMVSmuQMv0Pvc+Nm+LP7F6L+rMkmRk6yxcSDeT9qhei6Lx1zBtUOk/bUaAt120l+u0hhn1fraS7At7925tdo24CvuvZytnrLJUR16k8s5R1OmTKGHH36Ynn/+eRoxYgT7+3HjxlE8HqfnnnuuddvHH39MX375Je26664dORUAAAAAAAAAAAAAAKHp0JtckydPpvvuu48effRRqq2tbfXZqquro6qqKqqrq6PTTjuNzjvvPKqvr6fu3bvTmWeeSbvuumuHvqwIAAAAAAAAAAAAAEBH6NAk1x/+8AciItpnn33Y9jvuuINOPvlkIiK69tpryfd9OuqooyiRSNCBBx5Iv/+9/jw8AAAAAAAAAAAAAABR0aFJLmetSxVUVlbSTTfdRDfddFPWlSIicjUb1v6bPhFy7XGYNb/GOm0vLa4pJcpY+zRz74/a95epMrWz+HGVLxORWsMrfS1cSnsYeWX8lnnG+u9ArHe11v5L2wq1LjvMOmXLD8xaEy5o6+lh+nvkiKCmkoJ21v5LH4xQmpL6IcMrIy0eM8svQK7fNtrFX9nI4mH3r1Zl6F5+7qf67s7iVK1eh722L9dQ2rA0qX+Xe+vUG/4kRNw3w1WJc4XxODM0ZfoMqEKO/zcPsBxl1VH5a2Q+prWOX2pM6qv32/rTyac880MW33rAbapM7SVNLJ5t+NLcPZ8vM08ILU8a9LLaZ7Ny7nXRK7ZGlZny3++xuPut3VWZgV/z63JVItdlm6PC7Bds5P9zTL76va9/zvuW27a9kcXSF4KIKL0ZP5f0IiEimnUk/8LR3KZeqsxRNVwzMeEVsTiltVDpNbO4e6xJlVmdqmLxz+//vioz6CV+nIo1DSx2lieXzCnZ9nvrn2M15sghTE+m3000OcpLiTaRmrOuWbarkbsf/ukBLL720G+rMr88/G8slp6S9y7cWe2z4rKhLPZT+tw9l6/g1TP6PDm28qQOOlmOctUVHcpPyneNiDx12y09CU+3JO/zBj/JvUeJiD79x3AWXx0M0/WT57LG92IcvmWS+8mocTsRkfCTUWMf0s9fZONyeU2Wd1OYMVSh+rzq9j2U1NgqTJ8XRlMirv/3cn2ct/g9uskdmvFcn5zRVxXps9USFi/7Ny/TN2E0eBP3R7K8s6RPl6kp4tt8qcOsc1Tm3Vp+3+Xzd14xzx1U3NqTV+V3ukyz422ddDqXSE+3pNjH8tuKe8Jv3BiXx4iPq6p97c/88DfjWPzAq7x/HfKVzo+uQozdN3EMJf+/PbL+uiIAAAAAAAAAAAAAAMUCJrkAAAAAAAAAAAAAQMmDSS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDwdMp7PJ+mqOHll68zKTFPnMEaKwpjMT1n7cJM0TzivOmMeUBmLmyaHokxSm8i7Zm6CS83c5M0ynidhKu9ZBm7yPJYZoTAB96SBatI4rvTIC2M8KM34iLjBvmXimSOCqjgFLZoyDFFlm1hI011piGojNGY1XBCRphJcU/5n3DQ1buxTXs5NAb1uNfq43fk2FzcMQqVxv4wN494whpTSsFU9f0QbNJVHPaWr2+QoU08dM1Iksp8Xn6Q2xLNrXPMWN6xg8bT7T1JlvvopzzfplL6nZ2/3PItr/bUsfmTxDmqf6z/bn8W9/qU/jtHvLW4qH1QmVJl0pfjIhshRvvXxA6mnEKaiVpu3bVNTbzkiX/2eJ5K5NCW1D8v7nsDIY5tV8I8ODIx/o8qsCrhB/MpUNYuliSoRUbXPzz1z9UhV5vknuRaH/6NBlZE5Pugmcl8u+70WI1rrwyM5Il1VRl7Zuraz+rfocpTYR3QzzjLuVSbgui+o+JKbQY+5Qfdf9129Iz9XUpjlJnVuqYzP4XUx+rygjm8LzHGU6JuyyVEWss8rkhyVqc/LLj8Zz4Noo1Djcqkxl9lU3hpDkdCPHFOZH4QKMS6XH7UotnF5wfq86nLyylqM541xlKUzQahxubzkmPygj3FgeW6r3YSmxlw3T5WROalnM89rlqac/MhYmZFLqqSmMo/LZRkvmXlcbn60RO5TLOPyIp47qP6Kfzzs/244VZWROqw7YoEqc8zgt1mcCHj+aQz0mLtCmMjLj/kQEfUS6eXdpsGqzDN/5B+jGvVvPs6ynr2gppJvyOPcAd7kAgAAAAAAAAAAAAAlDya5AAAAAAAAAAAAAEDJg0kuAAAAAAAAAAAAAFDyFK0nlyv3W9cSB5aXRJlYI2v4OcTkImxj+b1c96nWgco1/ETKU8NLt+O/sbGYSK0pVeuypbcEaW8LLx5XZbwa7nliracO5Lps2Z7ST4mIPNnEhreEn5S+CEb7pTu+rjYKgniMgvVeCUo/RBRIfw3DOsNvFu1vnEeuTVdrlEPoJZSmrLYT3l5OeE64Zq0pdR/LtaZI+OiY3glCU2npL2E8o+rcVpvL9jMuocUXwZPeZjkkiPsUtOQoQ0/Kr8W8Npl/dBmVo2R7mLmFbyufp72RNvuZOI6R6/6eGMs3JLi/jWterfYZG/uIxX73WlXG1dexOKjSXZHUkxNFlAecgaU5+WyZ/h0F8g3MV7/X91rukfCDiWexuHzLlWqfK7Z5iMVJeUOIqFn4acmYSPtHNIn4T//aW+1T9wE/18CnF6sym8WWsThVV6XKFLTfa8nFaeOG5IigfEOfV9AcFWIcZXpVhSjjZF+p+jztPSKfaa9ce5hILzKrs88qRym/KV1Etp/pXVmAHNVWT5bfjdpm5CdfXrDpD5XBBy6H43I5Di/kuJxkvjJ8vJSerD4v0/NJxJ+bPOYoV9amz7M8pcQmM0ep69NlpGdbzjRlPasZfuuZOUqeO6lzlDp3iHG59n3Lclwu28/SVMt1d8IxVDZzB7Fv+H0e9IQeP8t7mn6jThV5pHICi+VvJPP59nlbpLrpHNVcxzuwVIVuv/pF/DdAUCFyUpHNHeBNLgAAAAAAAAAAAABQ8mCSCwAAAAAAAAAAAACUPJjkAgAAAAAAAAAAAAAlDya5AAAAAAAAAAAAAEDJU7TG8xSs/0NEZHktCu84zzDck95xlhmb18wd5byEcCw0zb2FaZp1bmkqWaYvwhOGu9Is2hkGl8rA0CojzMV9w7BQGqIqw1DLpFa2uWHYJzfFEqoI2880VM0VzrUxvLMs40Meow2eNMsjIi8hNCXN3i2DS9kOhimp0lnMeDDK+CPtxYTBpa8dFJ004TXM6eVzYZpgSsNBoSlnmWJ6wgRTmv0SkZNtrs+8of3yqCfPuTb6zZwDnOXeKVOJZcIrcpR8vr1s9SS3GSadnshbLiU052c2Vs46R0nDVKEfy6RWYuUoP4xG2pbJZ47KU78XX7mWxSNvEx8mMPq939MB+lwSqSmr7cS9d028kxibmKX3kQa7NdW6TE9u0Fqs/V4++7yc5KgQ46hOkaMS3CQ4mxxlfeBG9XlhcpSlpzbjRWmqnSu8tCNvvYuwaTwfIj95LoNhMxnjcPExAS+ZpTF6iY/LLT2p59r8wI+xmzxKEYzLrfM68fw66/pkn2cYxOfst57EyDeeGJeT/JiBlftCaKrYx+Ut9yFf+YmIMo6hQh+jDZHNHYRBaCG2eIUqEsswhnJr+fjOotwYQ5WLMVTQo0aVSVfx/Fjscwd4kwsAAAAAAAAAAAAAlDyY5AIAAAAAAAAAAAAAJQ8muQAAAAAAAAAAAABAyVO0nlyxRIpiLeuWnbWwNvNiW7nu01r7L9f669hYWy/XT5fHVREn1/oba649sXZbrTF1ur5O+iUZ3gTeWr6Y1TfqJwniwtfC8FuQa7lNZBHjutuuAXdysW4OaaupwBnSlx4/lkWI1JSx7lr5RYjYXFsv2ylu3DPhweXi+hqUP4n4e2tW2zVz7xEn/QKIyFvbxI9jnFsdV2nKOHuIaXYn28bap6WM5W2QI/y1KfLL1rdVoNvDso6QyJxkrf1XfiRy7b9xv5QvhHG/dI4y9CQ0p1rXylFS/5bHm8xRVv2kvZPyK9FCUJ4lplGE9KQwPDTalsmjprpkv6cO0gn6PUNTLf1ePvu8nOSoEHrqFDmqSXhyNRgmIQLpE2j2efIRDpGjzHFUm2O7PP17dcb8ZOYsjpfqeH7SHl2dND+J+ig9Wf5Jsl+0rqkD+YmoyMblIZAeXKH6POWVa+Qo5eFm/W7g21x5Zt9A1bpGIlbjcMuHTvoGNmbWVJhxlEonlq2h8kvceI7KV34iypyjnAvxvGAMRUTFO4YiCp+j8CYXAAAAAAAAAAAAACh5MMkFAAAAAAAAAAAAAEoeTHIBAAAAAAAAAAAAgJIHk1wAAAAAAAAAAAAAoOQpWuN5vzFJfoshntPmZ570aDcMGUkauVtkMhW2/l4aohkGacqMMGYYcpZxk2llPJ82zO7ENmcYn1MTNwmXpodERL44lyfN4+J6n6BcXlNm4zdpckgkjWIzG5VGBdeUoQ1hemlrKvN55H6h7PGUiaNhuieNeY376mLta8rUi9xmlHFN3JBQmv0S6RlzWV9LU64ihDm93Mc4d4tZo0vnUU9rU+TH1pk5yny0rjLi2qxnIYSelDYsXWbcJ0SOMupHKWGYKp+bQOcoadppam7tWr6Pcd/lnVQfMjBzVGY9Kb9KK4+1fba8QuWoEP2eNNon0rnN6sOi6PesPi2Kfs/QVMn1e0YZr6VtfKPuOSJvOUqduGvkKLklnJ6y6POs+1KAHJUpPymsZyVf43Ir9wgjZeujI0WVn2TuyVZP0ny8vfxElN8c1dFxuXV92eSoMB+UkfqwxtwV4jmwDLOl8XyY33phxu6NIkcZ5/ZlrqsIMS4P89Eo+WEg6wM+rc9g8Yyh5McNCjp3gDHURokqR+FNLgAAAAAAAAAAAABQ8mCSCwAAAAAAAAAAAACUPEW3XNGtfxUuld6wPCowXqsLUplfOfTU63lJo4xYhhU0ixNZr8SJcxmvBrq0eEXUeF3TE9flBbwuTtaFiJwT25z1Kr5YLhfo2+zS4vVlP9ZuTEQUpFKiLkabi1djfePe+W3avOU+uzCvh2aJrSndbur6TE3JV4j1PcqkKRfw8xAReXK+OdDtITVlLlGIRFP63J6TmjJecU7za3DidVJLUy4l9nHGvHsHNJUKCqMn69XftFhG44x/U/BTUeQorSfypJ5C5Ci5D+kcGkZPJPTknG6bcDlK6knmKF3fQLa5pSe1DEC3n9/muS5cjgrR7xmLoT2lKSPfRNLvWTlKnBv9Xptt666hUHrKb44SZZCjWgmTo7SeCpujwuantMhPem14ofOT0JPV9sWUn9JyDNU59NT22B0el1vXV0Q5iswxbMdzVChNyZ8EWWnKGJd7IcblrgPj8mIeQxV07gBjqA3nCj+GIgqvKc/lUnVZMH/+fBoyZEihqwHyzLx582jw4ME5OTY01fWAnkDUQFMgSqAnEDW50hT01DVBjgJRAj2BqMmkqaKb5AqCgBYsWEC1tbW0evVqGjJkCM2bN4+6d+9e6Kp1OlatWlXw9nXO0erVq2ngwIHkWyagEdCiKeccDR06FHrKEdATiBpoCkQJ9ASipitoCuPy/NEV9ESEHJUvoCcQNaWkqaJbruj7fuusnLf+6wTdu3eHWHNIodu3rq4up8dv0dSqVauIqPDX29kpdPtCT52PQrcxNNW5KHT7Qk+dj0K3cS41hXF5/il0+yJHdS4K3b7QU+ej0G0cRlMwngcAAAAAAAAAAAAAJQ8muQAAAAAAAAAAAABAyVPUk1wVFRV0ySWXUEVFRaGr0inpau3b1a4333S19u1q11sIulobd7XrzTddrX272vUWgq7Wxl3tevNNV2vfrna9+aartW9Xu95CUEptXHTG8wAAAAAAAAAAAAAAdJSifpMLAAAAAAAAAAAAAIAwYJILAAAAAAAAAAAAAJQ8mOQCAAAAAAAAAAAAACUPJrkAAAAAAAAAAAAAQMmDSS4AAAAAAAAAAAAAUPJgkgsAAAAAAAAAAAAAlDyY5AIAAAAAAAAAAAAAJQ8muQAAAAAAAAAAAABAyYNJrjwwd+5c8jyPrrrqqoxlL730UvI8Lw+1AqUK9ASiBpoCUQI9gSiBnkDUQFMgSqAnEDXQ1KaDSa6NcOedd5LnefTWW28VuiqbxMknn0ye5230z1dffVXoKnYJOoueiIhmz55Nxx9/PA0ePJiqq6tp7NixdNlll1FjY2Ohq9al6CyaevPNN2nKlCm01VZbUU1NDQ0dOpSOPfZY+uSTTwpdtS5FZ9FTC++88w4dfvjhVF9fT9XV1bT11lvTDTfcUOhqdRk6i57WrFlDl1xyCR100EFUX19PnufRnXfeWehqdUk6i6aIMI4qBjqLnvA7r3iApoqLskJXAOSWM844gyZMmMC2Oefoxz/+MQ0fPpwGDRpUoJqBUmTevHk0fvx4qquroylTplB9fT3NmDGDLrnkEnr77bfp0UcfLXQVQYlx5ZVX0muvvUbHHHMMbbvttrRw4UK68cYb6Vvf+hbNnDmTtt5660JXEZQY//znP+mwww6jHXbYgS666CLq1q0bffbZZzR//vxCVw2UGEuXLqXLLruMhg4dSttttx29+OKLha4SKHEwjgJRgt95IGo6i6YwydXJ2XXXXWnXXXdl21599VVqbGyk73//+wWqFShV/vznP9OKFSvo1Vdfpa222oqIiE4//XQKgoDuvvtu+uabb6hnz54FriUoJc477zy67777qLy8vHXbcccdR9tssw1dccUVdM899xSwdqDUWLVqFZ144ol06KGH0vTp08n38cI6yJ4BAwbQ119/Tf3796e33nqLdtppp0JXCZQ4GEeBKMHvPBA1nUVTGP1tAs3NzXTxxRfTuHHjqK6ujmpqamjPPfekF154YaP7XHvttTRs2DCqqqqivffem2bNmhXqXPfccw+NGzeOqqqqqL6+no4//niaN29eVvW+7777yPM8+t73vpfV/iA3lIKeVq1aRURE/fr1Y9sHDBhAvu+ziQpQeEpBU7vttpvSzahR/7+9846zoyr//zNz7927JZsNSchuQjpJCEUEA8RIFSMYpERAwAYISjFBARF/UalfMZavgnylKV2kV0FAITSREIq0UEILEFKWhJRN291775zfH8lu9inJnb2ZW2bzeb9eeemZPTNz5sxnnufM4Z7PjKYdd9yR3nzzzVDnBqUhDnq6+eabqbm5mS666CLyfZ9Wr15NQRCEvkZQOuKgp3Q6TU1NTaGvCZSXOGgK46j4EAc9WeA9r3KBpkoHJrk2g5aWFrr66qtpv/32o9/85jd0/vnn0+LFi+nAAw+kl19+WdW/8cYb6dJLL6UpU6bQtGnTaPbs2bT//vtTc3PzJs9z0UUX0bHHHkujR4+mP/zhD3T66afTjBkzaJ999qHly5d3q82ZTIZuv/12+sIXvkDDhw/v1r6guMRBT/vttx8REZ144on08ssv07x58+i2226jK664gn74wx9SXV1dgVcPikEcNGXhnKPm5mbq379/t/cFxSMOenr00Uepd+/eNH/+fNpuu+2oV69e1Lt3bzr11FOptbV1M64eRE0c9ATiRRw0hXFUfIiDniR4z6tsoKkS4oDJdddd54jIPf/88xutk81mXVtbG9u2bNky19jY6E444YTObXPnznVE5GpqatzHH3/cuX3WrFmOiNwZZ5zRue28885zXW/LBx984BKJhLvooovYeV577TWXTCbV9nzcf//9jojc5Zdf3q39wObRk/T0P//zP66mpsYRUee/n//853n3A9HSkzQl+etf/+qIyF1zzTXd3hcURk/R08477+xqa2tdbW2tO+2009xdd93lTjvtNEdE7phjjtl0J4DI6Cl66srzzz/viMhdd911ofcB0dGTNIVxVPnpSXrqCt7zygc0VVngl1ybQSKR6PxZcRAEtHTpUspms7TbbrvRf//7X1V/8uTJzKxtjz32oPHjx9ODDz640XPcfffdFAQBHXXUUbRkyZLOf01NTTR69OhN/rzR4uabb6ZUKkVHHXVUt/YDxScueho+fDjts88+9Oc//5nuuusuOuGEE+hXv/oV/elPfyrgqkExiYumuvLWW2/RlClTaMKECXTcccd1a19QXOKgp1WrVtGaNWvo2GOPpUsvvZQOP/xwuvTSS+nkk0+mW2+9ld55550Crx5ETRz0BOJFXDSFcVQ8iIueuoL3vMoGmiodMJ7fTG644Qb6/e9/T2+99RZlMpnO7SNGjFB1R48erbaNGTOGbr/99o0e/5133iHnnLkvEVEqlQrd1lWrVtF9991HBx54IPXr1y/0fqB0VLqebr31VjrppJPo7bffpsGDBxMR0eGHH05BENBPf/pT+sY3vgFtVRiVrqmuLFq0iL761a9SQ0MD3XnnnZRIJELvC0pDpeuppqaGiIi+8Y1vsO3f/OY36aqrrqKZM2du9Nig9FS6nkD8qHRNYRwVLypdT13Be148gKZKAya5NoObbrqJjj/+eJo8eTL95Cc/oQEDBlAikaDp06fTe++9F8k5giAgz/PooYceMl/4evXqFfpY9957b+y+jLAlEQc9XX755bTrrrt2Dsw6OPTQQ+n666+nl156SX12FpSPOGiqgxUrVtCkSZNo+fLl9O9//5sGDRoUSftAdMRBT4MGDaLXX39dmToPGDCAiIiWLVsWSTvB5hMHPYF4EQdNYRwVH+Kgp67gPa/ygaZKBya5NoM777yTRo4cSXfffTd5nte5/bzzzjPrW8sk3n777U2auG277bbknKMRI0bQmDFjNqu9f/vb36hXr1506KGHbtZxQHGIg56am5vNT1t3/JeIbDbb7WOC4hEHTRERtba20iGHHEJvv/02Pfroo7TDDjsUdBxQXOKgp3HjxtEjjzzSaTzfwYIFC4iIaOutt+72MUFxiIOeQLyIg6YwjooPcdBTV/CeV/lAU6UDnlybQcfsqHOuc9usWbNo5syZZv17772X5s+f31l+7rnnaNasWTRp0qSNnuPwww+nRCJBF1xwATtPx3k//fTTUG1dvHgxPfroo/S1r32NamtrQ+0DSksc9DRmzBh66aWX6O2332bbb7nlFvJ9n3beeedN7g9KSxw0lcvl6Oijj6aZM2fSHXfcQRMmTMh7XaA8xEFPHZ4R11xzDdt+9dVXUzKZ7PyyGSg/cdATiBdx0BTGUfEhDnrqAO958QCaKh34JVcerr32Wnr44YfV9h/96Ed08MEH0913301f+9rX6Ktf/SrNnTuXrrzyStphhx1o1apVap9Ro0bRXnvtRaeeeiq1tbXRJZdcQv369aOzzz57o+ffdttt6Ze//CVNmzaNPvjgA5o8eTLV19fT3Llz6Z577qGTTjqJzjrrrLzXcdttt1E2m43lzw17EnHX009+8hN66KGHaO+996apU6dSv3796IEHHqCHHnqIvve972GJWRmIu6Z+/OMf09///nc65JBDaOnSpXTTTTexv3/729/uRm+AzSXuetp1113phBNOoGuvvZay2Sztu+++9MQTT9Add9xB06ZNQ4wqMXHXExHRn/70J1q+fHnnrwHvv/9++vjjj4mI6LTTTqOGhobudAnYTOKuKYyjKou466kDvOdVDtBUhVC8DzfGm47PgG7s37x581wQBO5Xv/qVGzZsmEun027XXXd1DzzwgDvuuOPcsGHDOo/V8RnQ3/3ud+73v/+9GzJkiEun027vvfd2r7zyCjuv/AxoB3fddZfba6+9XF1dnaurq3Njx451U6ZMcXPmzAl1PZ///OfdgAEDXDab3ax+AYXRk/Q0a9YsN2nSJNfU1ORSqZQbM2aMu+iii1wmk9nsfgLh6Sma2nfffTd5HaA09BQ9Oedce3u7O//8892wYcNcKpVyo0aNchdffPHmdhHoBj1JT8OGDdvodcydO3dzuwqEpCdpCuOo8tOT9OQc3vMqAWiqsvCcE79jAwAAAAAAAAAAAAAgZsCTCwAAAAAAAAAAAADEHkxyAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7MMkFAAAAAAAAAAAAAGIPJrkAAAAAAAAAAAAAQOwp2iTXZZddRsOHD6fq6moaP348Pffcc8U6FQAAAAAAAAAAAADYwvGccy7qg95222107LHH0pVXXknjx4+nSy65hO644w6aM2cODRgwYJP7BkFACxYsoPr6evI8L+qmgQrDOUcrV66kQYMGke8XZ84VmtpygJ5A1EBTIEqgJxA1xdYU9LRlgRgFogR6AlETWlOuCOyxxx5uypQpneVcLucGDRrkpk+fnnffefPmOSLCvy3s37x584ohRWhqC/0HPeEfNIV/lfwPesK/uGgKetoy/yFG4R/0hH+V/C+fppIUMe3t7fTiiy/StGnTOrf5vk8TJ06kmTNnqvptbW3U1tbWWXbrf1i27/CTKemn122rTav9XCLByl4QqDpee1aUM7rBYj95XJLldY3kx80Yx23j25zVviQ/tqvm12ldd662itdJGjOYon2Jtqyq4q8RbRZ942X1PpTN8XLgdB1fzKAnjf7r0qfZoJ2e+PAqqq+v1/UKZKOaGnHKpjUl+tLL6euTmpL9RqS1qDRl9UkQQlPyXDmtKXlspam6arVLroaHgaBKt0/2RWSayohtlqYSQuNG/3X0cTZooyc/KJGeusaoaq0nks+m8aNZr1X0kXXfxX6RxajWdrGL0b4CYlRQk+LllKEncZ99+VxRgXqSx7F+qBxCT+WKUfsNO5mS/roYH9ToZ1Vpysor4tk0773cT2hIaWxdI/MfV+Y9I0ZJTVEN15B13UGd0JSR9zxxq30zRnHNqzGBjEdEKu+ZuVz+l0RLU8mOGNVOT3z059LnvJoqtZ8Tz2aYnOe1tas6eWNUoTlPxCjzeS5oHMX15FLGOEo+IhWa84iiz3tFzXlR6KnCxuXIefkpVFMqRm0h73pl1ZR817PeNbqhqbLpqZxzByljekX0Y9H0ZL7niZxX6NzBWrFNxO9KmzuIfJJryZIllMvlqLGxkW1vbGykt956S9WfPn06XXDBBbphfpqSifVCTYQQqmcIQdYxf9IWQeDLGccVN8zJERMReT7vfnmd1nV7SVEnjFAN0fkyQImyFxjX7QuhGtdEso996zhGsI7w56WFa0r0ARkDNKUP42WrEE15YTQlB5DyfhBRQZqSL5DW/SmWpuQ2S1OJTZfJigfl15PShjXgl31k3fcoBvymnuQuRvsK0FOQrBLlEIOzXFR6yt/nYfRUvhhV1TlAC8JoKkzes+69fM7E9Yaa5AqT95yV92T/i0muUJoKMcmV1QNIP8Hbp8YEpha4Nu1cLo9jDK38cscoY5JLaiVEzvPkYJSosBgVKudt+jzr6hSS80L8x0I5yVXhOY8oOk0VN+cVSU9lHJcj5+UH73pdz13pmhLvFua7Rvc1tWXpyZpeEZNcJZ07KOAHMmbOE+MqOaaqsLmDsn9dcdq0abRixYrOf/PmzSt3k0DMgaZAlEBPIGqgKRAl0BOIEugJRA00BaIEegJhiPyXXP3796dEIkHNzc1se3NzMzU1Nan66XSa0mlj1tG5DbPUxk/onZjQc9ZMq/x5pzXjl/dn9tZ/4ROzkNZ/rZI/S7R+wifbLNtntVdscklrFpNvC0KsSlUz1dZ/1cjx2Vj5XxGIiJz8r3LWcbr8FwnrPxBsLhvTFDm34V6F+N6CSxh9K386bdwj+YsYOWNu/hfYMJoyfraqK21aQ85qr/wvB8Z1620hNCWfnXZDC+InvfK5WXfu8Es3SqknFqOM+yW3mDEqLfrR+q/a8jmTP1W2YpSMmYXGqHwxKYyejBjlZCCz9CT/K2qIGKV+aWndF/VfNiszRlnPu5P/XcrQlJM/k49IU15W/JfLMJqi/HlP5QzrPzDKOGa0z8nLDJH3lGGp9V/LsyJGWblDtW/jmipbzguRPsLkPCL9izB1nDA5Lyd/LWDoSS1zMC6igBgltwXGL7JVSJKDTiIix38F7cn+24QOOvexxlHdXK4YtaYKzXlBmL4PoyfkvHWn2gJyXiTjKGMpXz70L28MrUalqQLe9cqpKfWuF2ZcvglNlUtPxg/JiTB3sO7Ulp488Su//K/JOue1VdbcQeS/5KqqqqJx48bRjBkzOrcFQUAzZsygCRMmRH06AAAAAAAAAAAAAACi/yUXEdGZZ55Jxx13HO222260xx570CWXXEKrV6+m7373u8U4HQAAAAAAAAAAAADYwinKJNfRRx9NixcvpnPPPZcWLVpEu+yyCz388MPKjB4AAAAAAAAAAAAAgCgoyiQXEdHUqVNp6tSpxTo8AAAAAAAAAAAAAACdFG2Sa7PJ5TY4i0nzUdKm1qYRbYjPoksTWWW8ahi4eRmXt45ujOHgFkiDOXGdWePToXJblb6mQBjKuXR+U0ZfXoNlPJzh7XNhjNAtY+QufayNE4uHl81u+LxpxvrEu9CUYWbpqqTujD6Qm0JoisJoSpr1SeNeA6kpeQ+JiPwsv87A0JQ0KQyqLDs/3qe+/LSsoQXVnhCaMk1nO47tRW4zuHEy2Q2ftJVm30T6GbIM/dVnmPN/1lcZZRq7eFKEIYwzzRglDSNljDL0FCZGyThr64mbOiuTcMsoVpqhGoapyoi9gmJUdzVl5TRlmhriU9FhNKUMbMOYsZrGqnk0ZeQ9X5reG3qR5uFBiLwnnwv1sQwirXFLUxLDyLyjj50rXYxiOc/SUy7EOCoVwvS7kJwntRFmHFVAzjNjlLqHIWKUpSfRZl/EFmUKTjqOmeMo+SEDKy+w9pUoRuWJT4WMy5HzuoKcJ4lMU3IfqQ+jbz0S97pQTQnzcE+MuStNU15G9Plmaqqkeuoyd+BlDNN2cb3me17c5w5MPYlrMPKZHJrkrDrqscmvg0LmDsyxbdcYFfJdr4RvhAAAAAAAAAAAAAAAFAdMcgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQeyrXk6st07nm1/I1cHLtdsrwu5Frt6Wn0ca2dSVneIqItdLmut9Mhrcla9QR65zlFVirdaXPkbmeWPSXZQESpMRGse7Xt9bMBsJvQi7ONY5jrSf2cl3/f57+j5L27Ib1wpY/ifQe8ay+lZ1prD+W91VeotW3YTQlNFSIpsxZbaEp3/A8yKk16vreS0154tnykvnXlitvFwNLM51rvMP45URFZoOePMPjLVSMEr4GzrJ9kTFKPlOWpUsIPb115hBWnnLAv1SdW/5wID+VuKSgSl9Tn3faWTm1MqPq5Kr5gWQ/EBkxKpAxylCz0Ib0TbCOY3Vg19im/XuKh8tkO/1nLE3JuOVSuoqKUVaIzeOJYD1jYTQlY9LbZ22r6pzx1QdY+bafTWLl+pcX6XMLnwdPaoNIedNYmnJiPykF38pHvtBUGD1YeW99n3thvCyjops5j4ycF4h8YD1TefN4GXOeOY4S11RozpNelS7g+6jYbRzHHEcJzDiW3bCtZDEqipynfAWtZ6U4OU/pKaNzk/Lpk3/Weyj9WL5RyHkboVI0Vei7XpgYJfyDvHauu5JqSr3rFagpiZUX1seokuqpu3MHxhg21NyB3BSVnqKYOzC8vkLNHVTnnzuQfRPIGBXGQyzE3IE5TiogRuGXXAAAAAAAAAAAAAAg9mCSCwAAAAAAAAAAAADEHkxyAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7KtZ43uUCcm7jRq3SCC+wTGalAa9vGMG153hZGsO1a9M3r5UbK1Nrmz63NLTM5VQdaUbYcb2df7auO8mN4XzLPE70hTIeJN1/yiQ8hGkeWaZ5YQwKu5Jrz18nIlwQkFtvuGmZ1jm5ydSUNI01TiRvteg339AUtYl+WNuqqsz7xkhW/v6J/1B16hNrWbk14M7UGacf+fG177Lyt2Z+X9UZerXYYJlM5vvogNQPGZrKGs9JCE113BWvlHrqGqOM51v2hzPMIJXBpXGpXpAnRmWMc68VMcnQ0ykHPMLKa4IqVefIMx5lZd/j56739XHvWbgrK7/38mBVZ8S9vH3SwJlIxyAVs6wYJe9DGD0ZVfjfDXPiYpHLEbn1DTKNN6VTuo7tyoTX8lJvFxqKKO99dPwoVj550j9VHamz4T99i5U//NVYtU/dm4tZOUzekybzRKT8ltVHQqycKzVlPev5PmDD6pYp5xnPgsx5prm63GaYxso+kjGp0Bjl2kVfWeMo+aEdsYt1TxddXM3KJ4x6XNX5+w/2Z2VpwL/uZKIsdRBmHFVgzuPnKU2M6u643Mx5clxl1Cko5xVrXE75TcI98a5hxif5YQzkPCLaQjUV1bteGE0V8q5XiKY2RQXrKbA80MPMHWSKM4Yqmp7CxKgo5g7CvOeVcO4Av+QCAAAAAAAAAAAAALEHk1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxJ6K9eTyEj55HX4j1hps4ROh1tCSXl9qrT2WnglyHa3yjSBS3hFO+ikR6XW0Cb3+1fJ8Yse11tWKtaxem16v66fE2nPLL0SuoxXH8dr0Gmq1ntjyjFE7WQueN5zbC4y1uUXC8zzyOtpjtUtqyro/0p8ka61nFx5coi9VPxIpTS0+cISq8q3vcg+lP/znAFVn2L28XLWCnzuxQvue/N+3D2Llvx/ze1Xn6BfPYuXG59aqOvq6s6IcQlPW+nO108afm1LqiXxvwzNseSN5Mkbl95IoKEaF0JMVoy5/9ot6P0HfxhZW/taIF3jTjP9Ocuzgmax8X3oXVWfRv7dl5erFun1KT60yNhv7tAuNWV4SkjxxuKSa8vwNfgum94jwYihQU8r7pVXEKEMvTmhqyaRtVZ3vfJvHqJSXv//HN8xl5Te33lHV6fWqiCWG34X04LIsHpSmpCdnqxGjZF8Yfo4KS1MdObdcOc+IUVHlPDkWKChGSf8tIp0PPOMaNpEPiOxx1HGjnmXlL9S+o+rcMJrnxa3m6NyZL0b5xnWrsdVm5jyiEsaorjkv1Li8snKe6mtL74WMy7Ny/GyMyxP8Glyg3wm2xJzX3Xc9yxtPacp654mbpgp413OB5UPJr1tpqgTvepWsp4LnDvJ4cBU8dyCNMq2cVyI9md7L8jkKM3cQVc4rYO4Av+QCAAAAAAAAAAAAALEHk1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxB5McgEAAAAAAAAAAACA2FOxxvOUShL5KSIicindTJfipo2WwaWawjO8zjxp0CbM2UiaOhKRy4QwUZNG81Upfe5kAd0vzNg8yxhOGOJJ40EiIi8jDS65AZ5pmif7xjIjlNedNAz3u15DHkPVSEmniPwqIgqpqVQIQ0LKb8ynjEAtTQnT3ezhS1Wdl1qGsPIOFy1SdfKajhr9Pepa3p6DGk/T++3I2zfwP/o8XoZv89cKM2tLU0ZfKKRhsmE8uuEkpdOTl0qS160Yld/4WZqjEhUvRm1/ljBbTlWpOsHwJla++bMHsvJJP75P7SM5eeATatuPxm7HyoPn6w8ZKGNP+QEHy7TT2iZRHwExNFOmGOWlU+R1J0ZZmlLtzW8UrvRh9ONHx49i5dQXdIySVHtam77H21PlCUNmI2WoazINUYVeLON5EaOkkbPXGiJGWY72UlP+pj40U8Kc1+1xVEQ5L0yMkkbz1jhKmu4WMI5qG92ktjUln2flM94+WtXZ6i0ek3w5riIj54X5yIzsC2scJa/JGt+WIUZ5VXnik3gOLJPwcuY8+Zy6IY36OKLO2sF1/BBr9XGTq3n7grR+/pMreWyx4lO+nNc+sLfapfaCBazcO6U/kLDkh4NZObF4hT53ucblUbzrFaKpMOPymL3rhTJIlzGqFO96cdMT5g6IiMjvAXMH+CUXAAAAAAAAAAAAAIg9mOQCAAAAAAAAAAAAALEHk1wAAAAAAAAAAAAAIPZUrCeXq0qRS6Q6/7/6ewj/JEmoFZyW30Y+lMcLqXW0wVC99n/OqWlW/sEej7PyDtXz1T7S46SPr71s/t6yKysf1fCCqvPPVTuy8sMn7c3KqZVr1D4uy9fVetIriYhIeHpYa6K7eiw5aw15kXCpLppK63YFVcJPwvS7EccMIyqpKUtjWzWw4imj/62qXH7FZFbexvtYn6q2OkSDBGLt88i/6irDf/UWK39MI1Ud5QUn12HLMpFek26tsxYaMjW1/l6VVE/pKnKJ9f4kVowSvhyBFaPE5ZrLzPPFJGt9uyREjPKq06qKv2wVKw94gpcvr5ms9vn1j65h5Xpfe4R87rDZrPzJI0NUHel3E0ZPTvSFZ3WojFHGvet6I1wuRP9GRVWq00/C8mKQeS6UpozHTu0ivBcWHTJC1fnesQ+ycsLwZkqJk22dXJm3zv975XBWHjFjodrHpblfnNuUL18HITyUvDA+GtK7xdeakrlwU3mvbDHKynmp7seoUDlPNcQySBMHMmKUlxLPgBGjzL7uwrLt9D4jqz5h5Xlzt1Z1tmsTYyvLw6SAGBUq54WJUYnSj6PyjstFzjOf0zA5Lx8hcp7Xv6/a9uaZfBx+7sR7VJ1a3/CUyYPlPSj5+V+OZ+VtZmhfLKmnNSP7sPI3//cfap81AY+NSzL1qs7C3tuysr/C0FO5xuXdfdeLSlPyeS7iOCpfjDKR494wOc+4hoLGURG/61WSnoIq+T5h+FAKASWyIdpfwNxB+7hRapv3i8WsvKhFP89DfmJ42OajAD1Jn9N12/LoyfCFDqWnpJzTiWbuAL/kAgAAAAAAAAAAAACxB5NcAAAAAAAAAAAAACD2YJILAAAAAAAAAAAAAMSeivXkoqoUUce62pReBx2I9aTO8MmQmHXk4m25VtRa3J1Miiq6zoIj+Rr4SSc+reqc3/CS3dBNkPL4etdaw3DlmIbnWbnN6f77XM0HrPyPmi/y88h1txbGulq5BtrVVOk6Xda1ByFOExldNVWlpe/kGvNkiIX9xr13Cb5NrT82+k2ufX5jzSBVZ9UwXsfVGGv/5XWFMCeQ7U2uald1Dur7KitflTDWkgsPHLWe29CUE+vYvaRxX8Joan08CEq49p9SSaLEuvZaMUrpKVGYnvLGJMsnwuhHdSrhHeEsLwnDx6crOcMCrl74BFZ7+r6/+PedWHlIYPiTSH+tEHoi4S1F1nMuvQnSWk+U7Lr2X/+5WDA/CUtT0p8kRN6zPKSkhqRW64/UvljSgyvtax+aOuFnI7Vg1el1P/ecCHoZPkbS88r0ZRHXaXmsyG2iLD3d1m/kp/EtLyIRo4xnqUNTpdRTvnGU7Edn/WdPqR8r54ltakxkerrw9pj+eTJGFZDzstX6uP2EBvvMtsYD4poMXxZP5JvIcp6MUZvIeUQlzHulynlyW4ict3Y37hNadfYiVefcbbgH19bJFlWnztPjn3xIn8E6w6Prdydxr8rzFp6g6qwczq/zh9+4j5Uzxlg+EA/t32ZOUHW2/4j7/GwqPhGV1kOp2+96ITRV0LteiHGUFaNW7DmclZd9c5Wqk05xfTgn46Xu74TPt/W7wIgT8jHJGDEqinGUkW/DjMvL4cnF9JS2YpQY+4SZO7CqFDB3sGL/0aw88IfvqTpHNfL39wuv/Zaqs2gi93DO1oZ5JkTZkLv0tmvrq+/b8H9wj92qlWKMV+jcQZicV8DcAX7JBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9mOQCAAAAAAAAAAAAALGnco3nPW+TptmeMMaTZSJthmoapkqjQ2GmZ5nXekm+z/xDB6s6F552PStvk1yu6kjO+/AwVn5j9lBVxwkzwgN2e03VOafpEVYOAn3dD7XszMqp5dxMzjRHlxtShhGiMHEOarV5XFcjyVy2dC68zvM2aMIZRojSENYwS5QGhJYJpjJflebdge4TL8ONKd9a0ajqTP3Kw6z84N37qTrKMFI2z3oGhPGidU0/v+XbrDw0WK2PI/cTx1UG/EbzTE1V82cyZ2jKVa07di5rGEcXiy4xygujpxAxyjLBlObQnjT8NM7tJfKbYMpn1TJ6DKpEfBTts4znhyTXsHK1cW71CFhG4u154rdxXHXdpp646W7Qa9MxKlsmTZmEyXvyeTYOJ/Peov/lfTB9JDdoJiJqCfjNrjI+KFDrcUPvd9qbVJ1rLzmYlfvP4bEkV2eYIoe4Jund61vPZB4NWTFK7SPHCEREKu8Z5vTr+7yUMYrlPAM9jjKOQbJjjTopkfNEzLbybagYVS36tdp4Vqul8Twv9j14vtpnYIIfp6pFty8Qz4gfZowZVc4TsdjMeWwcVSJN+b79EYEOypjz1jTyOt/d5llVp3eCj3P7+GtUnXqfG8/LD278+P0j1T7LW2v4MdJtqs6Vo25h5SN/+i9VZ1jVElbOOK6NlUbCvfyVfVl57GXaTD/OOa+c73oLDx3Gyn0P/1jV+cXw61nZ0lSD+NBFQiQrqTEiopS4hNRdqgrt+58pvM5r9aqOkDMlxavegOdWqn2S8/iHCjb3Xa9sejJOG0pPYT62kkdPi/caoPYZ9t13WPkbjc+pOn0SfDx09Un/p+psneBm71IrVoSuEtdg1ZHdtdzov2PeP4uVB87L/5EZ1XvGGKpYcwf4JRcAAAAAAAAAAAAAiD2Y5AIAAAAAAAAAAAAAsafbk1xPPfUUHXLIITRo0CDyPI/uvfde9nfnHJ177rk0cOBAqqmpoYkTJ9I777xjHwwAAAAAAAAAAAAAgAjotifX6tWr6bOf/SydcMIJdPjhh6u///a3v6VLL72UbrjhBhoxYgSdc845dOCBB9Ibb7xB1dWGgcvGyOaI3Lo1l2Fm4qx1tYH0ebG8YUQdX/o7WFYxfWtZ+cST/6Hq9EusYuWUYXZx5M1nsPKwh/hi6ZHJjG6v8D16unGkqtNnEL+GFUbfPHjLF1h5yMpmXsFYgy39Fpy1rlb0X5BOqDq5VNd1tfrvxcLL5cij9et4tWUCyZaYa7VT+dUo12or+wjLfyDL+3LpDX1Vlc+cyz257jhHey/U/rI3P7fQt+V/EcgF3Ub7mp7lWvzgh6oK1ffiGl86b2tW3v5Py9U+3hqueVNTab4tqDY0VdWx9r90eqJcQB2r2L32rPpzGD2pGGUgvWGU74nhoUaZ/KHdJYROjbYo/QgfnVxaX1O9x+ukPd2WIOVE2VjHL9sjvBClT4uFrSd+HCtGdb0vQaKEmuqS9yyfN6Upo47yBDTk8cnu3Lvj5NF/Z2XLV8QXjg0JI6elhE/XTR/uoer0m82PLfVt+xyG8OSSzbF8oDKiBzMhNBXkz3vSFyqQPoxEFHT6BpYn53mthkdZiD4rWs4LE6NUDDBilNSGqHPv9reqfdrkWMZoXi4tzuX0ffMygSiLOtY4SmDGqKr846gOPRGVUFNdx+XW8xXCk6tYOS8rXi/6JVepOtUeH8dcOPdQVaf1kkG8LUl+rqoWnevX7sQ9ry47/RJVp0GMvQ6rf1XVWZSrY+WluV6sfMlvjlL7jHn6E7VNEtTzzqnUnFdUTeV513v32K3UPpcd/hdWzqgMTFQr/LZkDiQiqvZk7hTHMOJjQjgZWeOoZ/a6nJVXfEH3zXyhoeU5/v76h7kHqH2S3xceeAWOozre9UqZ87o7d2DYoZljYcl7R4hn88jrWHnrhPY6a3W8H+9b9jlV5+6X+LbUJ0bfCylIr24jVZFL8Dqf2/U9VefGEQ+y8pnztTYGvMCvy4kc5xnedmruwPTKLc7cQbcnuSZNmkSTJk0y/+aco0suuYR+8Ytf0GGHrTNRv/HGG6mxsZHuvfdeOuaYY9Q+bW1t1Na2IUi0tOgXdwC6AzQFogR6AlEDTYEogZ5AlEBPIGqgKRAl0BMIQ6SeXHPnzqVFixbRxIkTO7c1NDTQ+PHjaebMmeY+06dPp4aGhs5/Q4YMibJJYAsEmgJRAj2BqIGmQJRATyBKoCcQNdAUiBLoCYQh0kmuRYsWERFRY2Mj297Y2Nj5N8m0adNoxYoVnf/mzZsXZZPAFgg0BaIEegJRA02BKIGeQJRATyBqoCkQJdATCEO3lytGTTqdpnQ6nb8iACGBpkCUQE8gaqApECXQE4gS6AlEDTQFogR6AmGIdJKrqamJiIiam5tp4MCBndubm5tpl1126daxvGyWvGC9sVi7NmD328SP0Nq0OZsnDKulUR6RNuLWhnO6i5aO5Q/WDtUfqzorgxpW/vYTJ6k6Y+7lxphBlbIVVvss3oWf+9Hd/6jq1HjcWPD1dm2ouM1jYv2yMIJzhgGkcpNNGsZvlsFspZDJEq3XlN+mNeWEpnzLYLhGmKCre0aqD6ThpWcYSHo+N+vr99IyVef7//geK//j0ItVncOmnMrKwy6XhoSW8Txv3/LvacPEm3e5lpVzhjYPe2wqK4+5kZvKU0YbtjqpIUNTlimwxHP8f0uBl8mQl1vfd5Zhaqu47+1Vuk6YGCX0JE2ezWdObsvl7xgvqw0jPfmhghDPd63Prynt6ecovdNyVnaP6I+SyOv0siJGGQaXypQ2YfxYuYJjFM97hqZE3PKMvKc+1GDEqLUDeB98uW4OK7c73W/SUDdjuJvKuLD4ja1VnYbMalZeuQ2/90t20ecORMj84cEPqjqnbfUhP05utaqzOCfaF/Bc+f9+doraZ6tZC/mGMHkv0t/IF47XniHP33iM8tsTohzCVL/KGDb6m855vjGOcr6Ijzlt2Kwbo595PyvOLe5Fg8/HYkRE/2nlx0kapvwyL/pZI4bKD3NIQ90QJrxWjFIfiKmQkJUv55HQU8LQU7FynvyAjjSZt3h7ziC1bUzzWlZu68/j0/vH6Pv17wN/y8qDk71UnTVCCq1eu6pz+cL9WXn56duw8tbL9IoYmfNk/F9XqUIEZJAv58n3v0I1tXhXburfsje/z3dOuESfWgTyNUH+SRUrL7aKfFonjeiNBzzMOCohBscBaU31DrgxfiDi7mnDH1P7XJvgOix0HFWWcXk35w58Ywwl3/OWfqa3qnPXMfz9q97j7zfynhMRfeM/fB5g5JW6Y8Yu5/MCKvaRYfqv8pA+d07s885QPTar3Za/o8x8b4SqM5pEn4rxkJODNaKC5g5kDiiUSIdiI0aMoKamJpoxY0bntpaWFpo1axZNmDAhylMBAAAAAAAAAAAAANBJt3/JtWrVKnr33Xc7y3PnzqWXX36Z+vbtS0OHDqXTTz+dfvnLX9Lo0aNpxIgRdM4559CgQYNo8uTJUbYbAAAAAAAAAAAAAIBOuj3J9cILL9AXv/jFzvKZZ55JRETHHXccXX/99XT22WfT6tWr6aSTTqLly5fTXnvtRQ8//DBVV+vlKAAAAAAAAAAAAAAAREG3J7n2228/269pPZ7n0YUXXkgXXnjhZjWMcjkit96jwVhX66k1nvpS/Bq+ftoZ66md5S+Rh8aZy1n51Oe/pepUvcTX5G9/7yf63MK/Qa6tD5xu2yHfepqVBxpr/19t515Iv774m6rOAE+s+xXrXz1rPazs85zhidPO1yX7hgdUV38glw3hwxER+dZqK02l9FptP8M15NVqTSkPHOGvEWZ9umd4KI297FNWPqj+NFXntv2uZOXvvPdDVm6apft74P97l5WfHKHX6O/6/HdZudf1DarOmIXc48BfI/wBfGN1tPQssXyWMnyb36avwVvvl1JKPVE2R+SvP1/G8HgTevIsfwAZo3Lat0vFqALWqpt+N7KvMvr++NIHKy3uRYZ7GhERrQn4dSZ9Y/19GPKs0TdjFIWIURl+3crfkTboiajCNdVm5C+hKd+IUb44dKvhIyKRXiOtTsfHloD/x6wxVy3WBxJ5b4ez3mflcwb+U+3SmODXYPmThCGRJ/h+edq/1bbb3/4cK4/4+RpVR+U96alERN56b42S6qm746gCc570DQnnoyi0bHgCyudXPrvWuRd+uS8rrwm0T82Tqz7DylUrQ9wTa0gkrbNkKAmT80yvNN4eZ3jilCVG5YlP8lo8S09FynlbvcPv8+Mrd1B1dq3lvn0/2GeGqnPrKwew8smn38fKX6p9W+1TJ258xun7sULocO9HT1d1RtzM+y+9VvgKGnpyJPQUt5xXhHe9RRdo3Z046nFW3rv2HVa2cqDMeaud1mou4H0ZGC5ArY7njFrHfbJyPi+vPzsrWeOo3CbexTcH5QkovU6pe5qKnZ5Ezqv5tE7V+e2Cr7Dys3O5f9U2t2oNjv6YP8+WN7Qcl3uWl3FGzB0I38ycMbWzchS/plt3uUbV+US8dzb+w8j1Pm+P9HkueO5A6CnRaoyhuniPhdVUhdijAgAAAAAAAAAAAABQOJjkAgAAAAAAAAAAAACxB5NcAAAAAAAAAAAAACD2YJILAAAAAAAAAAAAAMSe7ruulwFnmFFTNrvpMhkGcwbS7FSaqJn7CIO00T9bqytlluY/jjDgdQE3Nfzw4N5qnz/1ncnKawJtbnf4M6ew8ojXjPaJPvXCGIBLozfDPE4azPmGgbrfxTyOcobxXrHwvA0GpoY2nDRSLVBTflbckwI05bVqs1wSBoQ7/I9u39GZH7DyLd/8EyvnvqnbslMVN73c7pozVZ1t/s37JtGm2yefJWWeb5ifh9GUL0wvLRN1l1hv6lxKPfle50cFzI9xCP24MHoyjiP7UcUo4zGUfRRGTxY6RnEjyrZt+UcuiIhWO96gtNPnaX2jD9+QM2KU1JOMWVb8kddt5Q6BFaO6PrMuZ/RdCXDW9YXRVJhrFpe0UuQe39PHkKa7/2r5jKrzzIXjWXn+VH3uP0z6Gyv3TfCPoKwMtMFurSc/YqGP65M0ANZknKgjyuNq56p9dvzsx6z8x88eo+o0vMwN9hOGptz6vFdSPRUj51mxLs214RUp55nHETFq1XB+3KWG8fxNd3yJlZuy+e+J/DCQta2gnGcZ6IrjJoxnuiPnEZUw73U351nXH0XOM6iet4KVb3t4L1Vn6GH84z19E6tVnTPPvJ2V+4n49KnxAauUx8dQq418tu9dZ7HyqLt07kys5fdR5TxLK7KPQ5j0V2zOi+hd7/ztH1Z16oS5u3x3aqcQxvPGvV9jbJPkRJ655OMvs/Ki1fVqn/o0b29dUt+TQHwN4/IRd+pzizqybBHVux515rwSjsu7EJWe6t5aouos/X4jK49OyGdexwCV8yzj+YLG5fqDCJKW/fnHD0altG6//PoRrFw3X38QQY671Qdjopo7MD5EU8jcAX7JBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9FevJ5ZJJcol1zbO8V5z0RwgMn552sb7dWqsu14om8s/7qfWlhj+Rs9bOy+NILwKft+9P37lK7ZMgvs+Vy8eqOk1387W2ftZYG9wm2if7ylwrnH8NrCf8qCwPAdelj72cXvNbLPJpSq0TtjQl7rWlKXnNrhBNtep+UZpq17obOEP4CkzkxaYEX5dNRDT5zW+xsvTfIiLy28U67HZj3XW7WEseRlPGenh13IwIUzntQ+f56/rYL5eeLJ1Lrwwrjsk+sk4k94tMT6LvPX1cFaPEua/f61q1zwrhqfTHJeNVna1fFuvv2w3vlsymY5QqE4XTk+gbz/BAcMkN11DKGEWpJJG/XlOWD4+MSZY/nYjT3lp9X5PCDmZRroGVfcPRamVQw8rLMzWqzvzDeF/+eI9/qTqfZnvlPZek1uO+OAHp2Cev8tOcfpqWC/+U5UEtK1ueKytz1ay86FB97j4zhUas/L/+2SlbzjP9fArIecZ5ZD4tWs6zjiNi1KDR3B9tjdMtHvAif0ZkfjPPU8qcp2KUHqp3jVGlynt5x+WF5DxrXB5BztvmSd3P7x04gJUbUy2qTsbxvkx52U2WiYiOeYCbD/b7r27vyHd50DVzXp5xeWQ5L0bj8kLe9RZntZ/xSp/XqfZ42fKqyjj+3C0VuYuI6Mb/PUhtk0jr5L5zuBa2Wqbf0YJakZvS2jNMsuIGXefTXB0ryzz+aU5fkxWLJd1514u7nuI2d0B1ehxzymf+zcqrAn1Pljw2iJW3yaxSdSKZOwjx3hxVjMIvuQAAAAAAAAAAAABA7MEkFwAAAAAAAAAAAACIPZjkAgAAAAAAAAAAAACxp2I9uSiVJFq/rtZa1++JdejKX4aIKMfXuyo/EyIi4XVgrr0VONmejF5XS3Ldb0KvlQ4G9Wflt6YYa6MFK8Xi7mv/+hVVZ9D81azst+q+8aTPRqvoK2OtsHmdEnEf5LptIiKvS194xrrgotFVU8Z6X0+sqbb1IvrJGV4e7fxed/hFbQq1DtvUs6gzcICq8rmzXmLlDzJcY997WfsHPLIb937b6zunqTrbXsGv09RUq1ibLTVlrfMPo6mkeEatvlmvqUrSk7qHlg+c1JOlObnmPcTa/08OGsnKvmXbIUKUMw4rvSQajpnPyq2yAhFlPH4Pnl8yTNXp9RH3hvPbCohRrcJYiiiUP4nWk06D5YpR3E/C0EJS+tPl14vKV0SUaOU3/xPhYZIwfLJaHb/X29U2qzojxy1h5TZDH6kEb7P0PQkMIbaKbQkr7grWOH1u6UfSIvy25N+JiK5+Z09W3m7Ke6qOS0lNbTzvVVKMUuOoMDnPrCNiVLFynjGOah87mJXPGHkXK8sxExGRnxWegBlDT8L3pLw5r0LGUfn0JNrtcvnHkVGNy1smDGflz/zsFVWnWvgyWfnL94JN1vl/009S+4yZLTwDk/n9Lcub84xzV+q4PCvG5SHe9VbkalWVNUJn0lstZ/zmY9aKEaw87/wxqs7WM1/nG4wY5aWr+IZq7pnkarSHkvQkNf1/haZWBlWqzhqXFnV4zrvq3b3UPo2tOrcruvGuVzY9lXPuwLhfKucVOHcg/b+8LK8zd7KeS5jUazYrnzZPvws2PSt8A0s5d1CknIdfcgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQezDJBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDsqVzj+cAReesN2IS53rpNYpthwOvapUm4Po40i3MhDFMpyG96Sylumub52oRu6Wf7sPL/7ve3vIc9+ubTWXnEjBZVx8vy9nlrtembt5abtrm1a0VZG1y6XH4jREqJvjH6nPwN5nEuyOi/Fwvn7PZsjIw23ZOaIspvfhdKU2EQBoRv/aivqvLFGm48f8n/HM3KQ19apvbZ++dTWflf+/yfqnP4i2ez8jb/XKzqeGu4ZpSmDBNeafpommtW8Xtm3sH1z6QLQpiwRkXXGBWGiGJU2xd2ZOX3j9LGlN8Y/x9+auO/Z9T6/NwpT7evf2olK9f5/B4GxnFXC8Pvrw9+UdW5fvuDWXnr/3yi6ig9rRF6ajP0JJ9ZI+56VcKc1YoJXQxLS6qpfDFKmJJapvIqRkkjUyJKrebbPmrrx/9uaCHjuM4Snj63L57OdDJ/fJfHaXdaz21iWybQugscv9eLcr1VnU+z3JC1OdvAyje8/Xm1z9Cz+UcSAiM3ejlhkmr0eYdZrAtCmLBGRVFyXojTRpTz/Cb+cZVlezSpOiN/+BYr1/trVR1J8zgeA4Y8auRxMY7y1+g6Zc15XUyMSzaOypPzSjUu97bROqidyj+K0ie5RtWRSCN6IqJqj2+rErEw3WKYWQuTcN+6pjDjcqmnVqkvY1we95xXBE01Z3Tsl8gcJ/MHEdGCn41i5eqZs1UdJ8blXsp4rZbm4SE+pODl5Mcx9HW/ewzPZ58GdarOogzPcdJ4vupm/R4RrP2Qt2Uz3/VK+p5XyXMHEmsuIcTcgdST/HBBr50/VbtkxFj92bkjVJ1RK3i+6glzB/glFwAAAAAAAAAAAACIPZjkAgAAAAAAAAAAAACxB5NcAAAAAAAAAAAAACD2VKwnl5fJkJdbPweX1Wtm5TpaueZz3Ta+xtNl9RpOa61t/sbx9aR+Oq2rJHnXrt1tpKpz+E8eZWXpJfHUqrFqn6EP8TpynT8RkdcuPB/a9LpatdZfeEdYPhyqj5MpVcevEmttQ6w9LxVee4Y8f+OaUnqx1nOH0FS3PFA6G5dfU8sP25mVp3/pNlXn3HuOYeVRLy7hTUtpv5uRl/HyecMPVnWye6/gzb3P0IfQmdKU5aEUQlNKQV5lzM177e0b1ssXGqOEP4ssE2k/lh1/+yorb+90f7QFPP6kfX3c2gS/Hw0J7WXTL7GKlauFj5f0LyEiSogW717zvqrzyPF8W+ZxrXepF1kODL8bcsL3xNCTopJiVL68F4htlqakR4vch4gSGX6PWrLcp8M3PFISxPu2V1L3f32C55W+Qj9ERH0S3CtH+rxZmpJkDM2vdtx3ZnWgNbUyqGHlNTleZ/Ui7WmSe/d1Vpa53cTy0SgDLOdZflsVnvPcCu45utXT+tzPfXZ7Vp569AxWPva/31X7DHmUew1K/y0iIl+Oo9qN8WM5cx7TWGn0Fsm4XOY8q04ePTVP0p5ce/Xm3o/SQ5CIqJfH78f8tj6qzu695rJyrYhPZleL9oYal5t6kn43PJ5ukTlP6OMLT2v/zhVZHtfl+IeIKBtwPTw4c5e8bRv9xHN8g/WuJ/zOvOpqVYfSvI4Tvl3S18vC0lTtKD4uX5mrUXXWiDy4KidyfdaI+fL9L0bvevnHUMLvsMCcV6q5A69G31NXzfX07nf6s/KNO/5J7fPvNaNZecDf9bn9ttWsvHS3/qpOTtz3rWdx/y/3th7vl3PuoDLeFgEAAAAAAAAAAAAA2AwwyQUAAAAAAAAAAAAAYg8muQAAAAAAAAAAAABA7MEkFwAAAAAAAAAAAACIPRVrPE+t7Rum4EIYvHmWcV9CmJ0ZRpTKEK2Ac0njQSIir5abxX1ykjZ1nlD3DivPy/Rj5X/+dm+1z1YrudGg2V5hmhfKFLag6zYMLoWRnpfWfdPVUM7TXorFo619g1mrYXaqDDwt47uEmBculqbq61Wd8We9wMrPrtpW1dn25mX81FXiETeuKbGam0z+5+Uxqs70L93Byjcmv6jqyP4LQyhN1VSLOoamOv4WFGAGWShtmU49ucC4dmlGa8WoECbWUk+BMN22TLj/9c5OrNz3/lpVJ1vLtZCt1tpYsTM33Lx6/2tZOeVpM+tWx+/hn+bvr+qsOn8bVk7ntJlsIWbWYfQkjWHNGNX176XUVJe8J81PTQxNedLU1vAJb3h+ASs/9PaOrDxpDDdbJyJKCHnU+vrjEw3CVH5AcqWqI43lZTlnODu3i/8eJzVGRNQapDZZJtLPzs1/35eVR/1T52nyQ+R7GaMMQ9mO2OsFJTTojVPOM/tVGEqPblR1/vfIG1i5XujSvdygj5tZzjdY8VtuM2N8+XMeUQljVNf4ZPWHIKqcJ/XU8L7OO74YTPZKaJP2t1Zz/cz93faqzoALeMwaX/seK6/cRl9T/Ttig2ESrsblYeK7wDM/aNFzcl5BH7Ag/aGUh97ZQdXZ+nYeS0bfNSvvcZUJeIh3PVdnGIXLeyKHbFbcFX3x6S59VJWDh/2HlduN2Gx9gIGd2ujyqN/1Svqe154h6vjYShxzntRTrf6QgdRTtje/ppSnr/t/Zx7IymPfXK4bKGJSv6fn63Ov5mM8t8YYMwnKOXeAX3IBAAAAAAAAAAAAgNiDSS4AAAAAAAAAAAAAEHswyQUAAAAAAAAAAAAAYk/FenK5ICBHm1h0Kdf61xjeJGLdp5fV6/iV74lcZ2v4Dsh1tHINLRHRnDOHsvIfdr5R1anzuHfEb688mpUHvbZU7SO7xAuzht3Xc5leSqyJrRHHsdYpi+NY1y29SZy1rrZLm52xZLpo5IJNnlCuv3eW343UlOGr4KTOrHXhec79yWGjVJ2fb3UZK//w11NUna1TLXyD1EsY7yjDM2Z0VTPfkNKhQ/nQFOKlV6PXn6s16tZ67o6/5Urnd5M3RknvET+aGPXixdw/6Ys/eUbtU1PNfY6yNbpfpAdXTnc93TbxclaWnkb/XqP92+787QGs3O9FHceqskKnSatvROwQPiF+mBhVbeipLoSeyhSjXC4gtymfH+kPVZ1fU5TRmgqW8HtS+/wQVu6342q1T7XPNdXf8NtqSi1n5b6JVapO1aaeGSJqdTq2SN1ZJIRJQ53hGfbrvx7Fyttexr3HXLthYCb9kawYVcc97zYdozb6p+iJUc6TfUhE5Gp4Tlk2WnudfaaK+/nJ1vV/Vetf+21tvJ2dWH1TQM4LNY7qRs4jKl3e6/a4PG2MPSMYl6c/1X5bW6W4V0zfpI5h99y8HysPfFvnpr++swcrT9iFG26176XjHs0QY6gwOrBynhyXp+UArgfmvG5q6tqXv6CPIfS//RnvGnXEOWpFvCnwXU/GKFet+1bGVaUP0xOQF1v763u/Z/3brLw60PFR+mS+0jKYldPL9PMnvRAL1lTH30qpp2yOnL/umiy/rZLlvEL1JDy4LD21Dahj5W/uxd8BVjv93t1vprymEDGqkJxX6NxBmJxXQIzCL7kAAAAAAAAAAAAAQOzBJBcAAAAAAAAAAAAAiD3dmuSaPn067b777lRfX08DBgygyZMn05w5c1id1tZWmjJlCvXr14969epFRxxxBDU3N2/kiAAAAAAAAAAAAAAAbD7dmuR68sknacqUKfTss8/SI488QplMhg444ABavXrD2vczzjiD7r//frrjjjvoySefpAULFtDhhx8eecMBAAAAAAAAAAAAAOigW8bzDz/8MCtff/31NGDAAHrxxRdpn332oRUrVtA111xDN998M+2///5ERHTdddfR9ttvT88++yx9/vOfV8dsa2ujtrYNJpItLesMib1kgjx/ffMM43TyhbmZVUdiGcHlMbh01j7CuC67TT9V5yv7vMTKAxLarPLPi/dl5UGPLePnyRTo1pfgfWEZuElDPmUALk2zjX2oWpscKhPGlDau62qoqMwgI2BjmqJkgqhDU5Y5nug3z9KUNNkzDEa9rLhv0pDQMiUV/b/qAG3YXOdxU+S+c1r1uaWZoNCvl9X9vXLsVqz88y/dp+r8fO7X+HGsvhH3XtUJoynD1FlpqsoIW+uv04UwPO4uoWJUCD2ZdSQh9NTwNtfGowu2U/ucOuYpVq49Sxv1DhDG4dWeNt3OOK5LaWx6w9s6pg+dtZhvkLHa2BZVjJLH9QqMUew5KmGMWqep9e0x857YJjVmYbRffoBim5u5UW/j91foU3tcm32EoS0RUYryP3/SWD5H/J5Z5rnSVD5nGNFXefzcZz12jKqzw83z+YY6buBKSf2cyNhsaqpaasqIdR3PdiXnvIhiVCE5j4yP1OS24vdn2Le1oXSr0MIrbduwcq93lutzF2CW69KGEa6MN/I4xRxHMRPeaPNepeU8qSd/3hK1z98e2ZuVzz34TlWn32Ef8w1PGh87EM1pDfh9/8zABWqflnQTK3tr9EcvSpbzwoyhKi7nhdfU9ud9mv9kDb3VJvWuJ4zDnaVDmW+NZ9X6CJM6jnyPFON0z3p+xYd2Rh76Xt7z1Pk6fwWi/577YDgrj3n1I7WP+gjI5o6j/DKNy0PMHZjvMpIizR2QZa4uP0hhPBMfHchz5a96v8zKZ739dbXPgP+ImBniWXPWFFGxcl41vybzPS/XNeeFi1Gb5cm1YsW6gXDfvn2JiOjFF1+kTCZDEydO7KwzduxYGjp0KM2cOdM8xvTp06mhoaHz35AhQ8x6AIQFmgJRAj2BqIGmQJRATyBKoCcQNdAUiBLoCYSh4EmuIAjo9NNPpz333JN22mknIiJatGgRVVVVUZ8+fVjdxsZGWrRokXmcadOm0YoVKzr/zZs3r9AmAUBE0BSIFugJRA00BaIEegJRAj2BqIGmQJRATyAM3Vqu2JUpU6bQ7Nmz6emnn96sBqTTaUqnjZ+AAlAg0BSIEugJRA00BaIEegJRAj2BqIGmQJRATyAMBU1yTZ06lR544AF66qmnaPDgwZ3bm5qaqL29nZYvX85+zdXc3ExNTU3GkTaOq06TS6wXsLVmVq79D7Ou1vKFyLOu1guMfSTTl6lNh271X1b++4pdVZ2n7+Xbhq1eyM9trTkVa1udXL9LpNaIu7Rxm1X/SS+A/P1p+Y446SFg3RdjrXIp6K6mXAhNeZampJ+E9MUy7qsT/T9wqxZVR3rXeBljzXeb8FUSbWn57NZqn29c+CArP71itKqz+k+DWblXQrdPeqo45S9RoKaS4r4Ya8lDOH9EDtOTtb49xNp6tc16NoRe/OWrWXnrqXqfq756GCuf96Mb9XEFrU77A8htF3/4ZVYedqb2GlQxyvKskDEqEcLfJZ8XgFHH8j1RMco4jvO6/v8SqqsmTeSv05SznpdCPLlCxCgZk/56/iFql0+P5Lq75HO3qzrSO2uN4a/VLnzeMsTLrYH2ZpLHvXr+3qpO8IN6Vh7r6xglfR+UB06ImG/qWT7rlp6zIcYSEROnnGe1b8kuvVj5V0P+oeqsEBo75x7uxTZmrfZQChOj1DbLc0b6doXJefLcVoySOc/oG9PHp8iUK+eF0VPf2fy4vQ/TnqVnDv8XK5924vGqTq3Hc9rKoIaVX3h/mNpn+6XcJ6qsOS/MuLyCcl7J3vXyxagQ73pmThb7mV7K8tjyPTOEv9D3t3lSbZM+qilPn1v6ZCbe555tXq32cJMU+q7X4UvVI/UUwdyBqacQ7PL5d1j504B7Vy56rVHt09DG5xcKnTvwcnw/eZWRzR2Ymun+GKpbPeyco6lTp9I999xDjz32GI0YMYL9fdy4cZRKpWjGjBmd2+bMmUMfffQRTZgwoduNAwAAAAAAAAAAAAAgDN36JdeUKVPo5ptvpvvuu4/q6+s7fbYaGhqopqaGGhoa6MQTT6QzzzyT+vbtS71796bTTjuNJkyYYH5ZEQAAAAAAAAAAAACAKOjWJNcVV1xBRET77bcf237dddfR8ccfT0REF198Mfm+T0cccQS1tbXRgQceSJdffnkkjQUAAAAAAAAAAAAAwKJbk1zOWpcqqK6upssuu4wuu+yyghtFROTqqje59l+tZw+zrtZY++9l5bpa0SXGPgv368/KFwy9XtWRPiO3Pq2Xa27/N/E1CLFm3GWzah8vydvnkfYvCYTHgzO8JOT6V+VFEqY/rTW9ls+AoKunh+nvUSS6asr0Q1D+GvmPaa279rKiX6SGcsY1i35ouaOfqrL4bO458+GP9HGyC/h+QQ0/93n73aP2ufL9fVi5+rKtVJ1e81bw5lZZHiZCU4U8o5amQj3b4n9LgKvd9Np/pbEwvgSGNjzpHSFiljNiVN832lj5/P87VtU5/zTu09UaaG+YKz/cl5Wrf8416NY26/aKvrB8ROSafDNGiRjqSy+bgvWUf7dyxaigrpqCTeU95clVoKakn4+o0/Aa95ghImp4hdf56VdO1McVUmwZqz1C/mfinax8xVyusbbbtZ+EsPGird5cq+qkPL7N0hQl8+S5MB4ZhqbMfKIqOf6/JYDlPCv+JOQzFSJ/R5XzQnhzruZWkLRc+CMREb3VNoiVR/91Ka+wVnszkfKTMa5b+JGYOU/cy7LkPPn/i0i3c551HVL/xvOgx+WijqGd3h/wnDdz1ShVZ4+691n515NuUXWkj+CibANvbptxv9q5N5KV80KNy7fAnBeJpiTmu57olxCaUtq0fJbk+MzwypV+TYW86602/C1TPj9OH1/nxSpRZ+Rt3EPapfU7pGIz3/XK9Z5njqHkGCDMuDyiuYNQehJeXwv30e9jP2vkPsqfZrl35dBHhDczEVFbOysWPHcgXhNUrJO5lUhfdwnnDgpzPQMAAAAAAAAAAAAAoILAJBcAAAAAAAAAAAAAiD2Y5AIAAAAAAAAAAAAAsQeTXAAAAAAAAAAAAAAg9nTLeL6U5GpS5CXXG6MZpqDKINWarhPGgn42hLGZcPDMbFWn6nzzB/9k5YR02yWiC+Ycwspjr1yu6rhWbpQpzSuVWSFpw2bPMnBT+xjmcaL/vJQwi04bxy3A3FSaZhMRN100rrFY5GqS5CXXueY5w2A4jHmwJzSlzAdJS9ETnpKWUbg00Wt8crGqc0YTN3o+4mtPqzq77PohK7+6Zigr//r2I9Q+217F96Fkm6oTNPDnIDA/ZiA1JUxUDeNeZUgYwgTS6vMOI21lqF1EcrVV5CU7DFNDxCgDaeLsZYz2Sw/JhNjHCH7phS2sPGiePu5f7v+SOI82P63NrOEb2sUHCAzDVJIfNLD/cQAAZxFJREFUx0ga972Gm6i6lPE8CoNQWcfL6OPK58g025b7WJopU4wKalIUrM97ln7UNuP6QmlKwet45lc3+HG3eXCRPrcw1N3mHq2Pv53/OVbuneEm9y5jfMxA5r36elXH9eYxytJUIOK+L+OYEaMKMs3dVN4rV86z4pE0nrf0VKycJ8+V1f0y6ClulvvWkYNUnUte5nFs9KKP+LmNuEYhxlGOeIyycp6M+zLneWFyXgjkPVi3LWf+/2JSspwn5OI5UceIT1XNK1n5gdu/oOqM+A4fV6U8fe6VQTUrr8lxHQx6xNDKGm74rT7kZICct/5UxdBUrvvveuaXpkJ9JGHTpvJERK6dx7FC3vWu+HA/Vee8kfezcq0MvERU5fNzB7XcOVzGLKICNVUxOa8y5g6ccWD13IXQ0+oJa1Sd5Tk+1lmV4zEruVrrwLXyD7AUPHcgxlCeLOcKnDuQ3wMwxhmFxCj8kgsAAAAAAAAAAAAAxB5McgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQeyrWk8tV+Z3r06WPBhGRCzE9l5ALQdut9a+bXiz60VS97nO79EJWbg2qVJ3sA/35eZZ/qOo4saZU+dtkpBkPqXWoXiqlqnh1Nfy4xnrqoEp4R0h/AMsDQm4y1hP7GdGfbca62VyZ/G6qEhSsX4ccJI212tJTyrDOkNdnLIFXmlKeE8aafeXhYvTtyBs/ZuVXrh2i6rzc1sg3tHF/rRHtL+lTJ4R3Vm/td6M8QSz7AqEhJ6KL9JtYt1F4KRh9LvvP9Mgpw9p/l+wSoyx/DbnG3NKT0oquI70OVMwKoScvZ8S5EJqT/SljlPKasM6d0fFRndvyTpB6kmv/U0Z7lYeGriL7nCxPmHLFqNSGGOWMGBWEiVHt/Pp865mSeSREjFIa2tRz2IHVd4HQVEZqyngI5LlM3fG8Z2qqimsoJ2KS9HYxz23mBdGfrYYH1fq+8YLy5DzLc1LmwVB6yuQfR0WV82re5R5KDx6ws6ozum0+36D0ZGhF6r9KxygVZ42cFwg9yaBveWkVlPMs78MyeCjFKecNv22BqnLNnnuy8okj/6PqZBwfD7WJgUyvD1arfaTGrJyn/Hh8XaegnCcxYpjKeZZeypTzuvuuF05TxYlRZn6QcczyporgXW/l7dupKqt/aoytBGkRt4IaMTCvCvGuF0ZTm3rXqyQ9yXGVcW1q7qCQ97wQY25rXJ7tz9+/Dt/+ZX0Y4R93zd++wspDX/yv2scl5Dt/gXMH0idQ9mdUcwfSh5GooBiFX3IBAAAAAAAAAAAAgNiDSS4AAAAAAAAAAAAAEHswyQUAAAAAAAAAAAAAYg8muQAAAAAAAAAAAABA7KlY43nKuQ1uZQnj79JU1jQE5EXL5M1r545yC7+0NStfsMstap+VwuD2F48foers8JAwQ03qi/ACYfwm2udCmBGadYRxr2+YYEoTcGVKaxmzi7JlAO7EfTC8Wpl5o2kiXiQ857qcz2iZMNlzlsOw2CQNh4m0puT98CyDS9kPhuGf2maYKnpCZy4rdOfnN38vWFPSNFU8o5aRtvPyGz/7sm/aVJUN11BCPZFznecznwXR1/LZIKKCYpTXJu6Pae4thWo9iQJDG15SpAhpmGrpNESMktdgmtPLDxWIGCX1RWToyTDBdNL4WZ+Z3QdpglxMvMB1PgPObFkIpBaNj6t4bUJT0uy90Bgl70nCSN5CU15CGIX72uVVfqRFtZeKpymlEEMOTjy35p0rQ4zyco689eMoy3g+1DHk82LpqVg5TxJVzitoHJVWVYqW86xcIShLjIp5ztv6bF6+49Jxqs7kga+w8rVvf46Vh7/5gT53WXOe+HhEzHJed9/1Kk1TKm6FGUcV8K434O45qsplj09i5aChVh+mRrxnij62Pgil3vUK1VSw8VhRNPLpKQxSTyHe86LS0zvf4ffwu7X6o3V/eHciKw+/eR4/jT5zxc8dyK5JGO95hcwd4JdcAAAAAAAAAAAAACD2YJILAAAAAAAAAAAAAMQeTHIBAAAAAAAAAAAAgNhTsZ5cifYcJaw1retxLv/8nPQ6sLwkglq+5rT+kIWs3CL8tyxG3dSu2yc9uAy/CU+s3VY1nG6vk14kGe1f4q3li1n9VP7bHKSkj4WxrtZYa6vqyP2M6+5ax7kCfWcKwF+bJT+5vr8C3SchbDCUhixNyXui1mpLXyMivY7fuGdaU7qOJzxwQmlKasjyuylIU3LttvHMysOYC/vza6pzWxhfl4hItGUpsf5eBq6wUCq9I8LoSd0fS0+yH6QnBJHSmKvK73ejetd4aJxsjxGjqI3HTH9NSteRHnnKC8DQk9xkWdApbzsjRnU5tivhfwvy27Pk5zr6y7pnBeQ9I4968p6IsukpI/stZdwzEX+cESeUh5L4u3WFrp3rRWmMiLy1rfw4IWKUE3nP9LyxdJavjrVLmWOUqSc5jrKalss/jioo50mqtJ6c3GaNo0qV84z2SVTOszx8DA8cRZhxVBliVKxynnW/xHHSp+preDi7CyuPXPsJK1vxwOXEtkrPeRWiJyKuKc+y2wrzrleIplTZ8CgKoSk1Lg+R80r6rieuQb7reWHe9QrV1Ppjl/I9r7tzB+o6KNzcQbH0VLWU35/lOe2zlvkH9w6n9vf5aaoN/8hC9BQq54kxlOX9GcYPVFaJaO4Av+QCAAAAAAAAAAAAALEHk1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxB5McgEAAAAAAAAAAACA2FOxxvP+mgz5HUaITpufSdNL0+DT5XcS91vWsnL997g5233ZnY2d+NxgVdUyfepqbmjvhDkqEREluTmeJ9sbGGZ3wmDRWQZ7rdyAV5oeEunZTU+ax0kjeiIKqkSfhzCTk8aZRNIo1uiXIuGvzZKfWHd/LYNLcqIPrLaHMKdXhnmWNvPuY8w/S6Nwy8A2K0wlQ2hKmiyamlrLnxPPMEAtTFPStNAwdRVdY96XDo17JdQTi1GGMGSMskxjw+hJEsa4WurHiAEuLeKqoTlpqqr0lDNilAuhpzVCT8a5fanLtOhPQ0/KSNzqc6kn49xli1Hd1ZSZ9/KfR+4XysJTGfYbepHGt5buEpvOe6Ze5DajjlMxKn/ek+01NZUOoSm5j3HujmfJ5cqkJwuR8yyD2FA5rxDkc2eNkcqZ88KMo8S5PDlGKjDnSTaZ84hKlvdilfNCjKEiy3lhxuXIeSZ5NUV8nFJoziuIAjRV0LteoZqSMcrSlChH9q7XLU2VS0/FmzvIG5Osv4fQ08hbF7PyfXfsreoMzPE6lObzDUWdO5A5r8LnDvBLLgAAAAAAAAAAAAAQezDJBQAAAAAAAAAAAABiT8UtV3TrfwqXzbV1bguMn9UF2fw/OfTUz/MyRp02Xg7axYmMn/TJ32kaPw10OfGzY+OndZ64Li/gbXGyLUTknNjmrJ/ii6Uogb7NLsfrOD+xyTIRUZDNirYYfS5+bu0b987v0ucd99mF+XlogViasn6qmRPLHpwxB+xno9AU78d1lcS5wmhK7kNa82E0RUJTzum+Cacp3h6tKd3eQPa5s5Y3yKUluv/83LprKJeeAuPn5vp5Me5XQXoSdSLSE1l9H3RfT6FilDh1qBglnlkrRslnIoyeKjVGhdKUmffk86LvUb4Y5QxNeTIeBro/lKasZVSR5D19brkERB53XftkjAqhqWwITXUj72WDytGTzHlqbTgReVmpp2KNo/LrqfJyXr4YVVjO0+Oojec8ouLHKOS8Lrsg50VCeE3JPrBiVOVoqrzveoY+8o7Li/+uVz49Ye6gc9sWOnfguWKqrgA+/vhjGjJkSLmbAUrMvHnzaPDgwUU5NjS15QE9gaiBpkCUQE8gaoqlKehpywQxCkQJ9ASiJp+mKm6SKwgCWrBgAdXX19PKlStpyJAhNG/ePOrdu3e5m9bjaGlpKXv/Oudo5cqVNGjQIPItU8cI6NCUc46GDh0KPRUJ6AlEDTQFogR6AlGzJWgK4/LSsSXoiQgxqlRATyBq4qSpiluu6Pt+56yct/7rBL1794ZYi0i5+7ehoaGox+/QVEtLCxGV/3p7OuXuX+ip51HuPoamehbl7l/oqedR7j4upqYwLi895e5fxKieRbn7F3rqeZS7j8NoCsbzAAAAAAAAAAAAACD2YJILAAAAAAAAAAAAAMSeip7kSqfTdN5551E6nS53U3okW1r/bmnXW2q2tP7d0q63HGxpfbylXW+p2dL6d0u73nKwpfXxlna9pWZL698t7XpLzZbWv1va9ZaDOPVxxRnPAwAAAAAAAAAAAADQXSr6l1wAAAAAAAAAAAAAAIQBk1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxB5McgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQezDJBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9mOQqAR988AF5nkf/+7//m7fu+eefT57nlaBVIK5ATyBqoCkQJdATiBLoCUQNNAWiBHoCUQNNbT6Y5NoI119/PXmeRy+88EK5mxIJ//3vf+nQQw+lvn37Um1tLe2000506aWXlrtZWww9SU9tbW3005/+lAYNGkQ1NTU0fvx4euSRR8rdrC2OnqQpIsSoctNT9PT666/T17/+dRo5ciTV1tZS//79aZ999qH777+/3E3booCeQNT0FE0RYRxVCUBPIGp6iqZWrVpF5513Hn3lK1+hvn37kud5dP3115e7Wd0Gk1xbAP/6179owoQJ9Mknn9A555xDf/zjH+nggw+mjz/+uNxNAzHk+OOPpz/84Q/0rW99i/74xz9SIpGggw46iJ5++ulyNw3EFMQoEBUffvghrVy5ko477jj64x//SOeccw4RER166KH05z//ucytA3EDegLFAOMoECXQE4iSJUuW0IUXXkhvvvkmffazny13cwomWe4GgOLS0tJCxx57LH31q1+lO++8k3wf85qgcJ577jm69dZb6Xe/+x2dddZZRER07LHH0k477URnn302PfPMM2VuIYgbiFEgSg466CA66KCD2LapU6fSuHHj6A9/+AOddNJJZWoZiCPQE4gajKNAlEBPIGoGDhxICxcupKamJnrhhRdo9913L3eTCgJvE5tBe3s7nXvuuTRu3DhqaGiguro62nvvvenxxx/f6D4XX3wxDRs2jGpqamjfffel2bNnhzrXTTfdROPGjaOamhrq27cvHXPMMTRv3ry8+918883U3NxMF110Efm+T6tXr6YgCEJfIygdcdDTnXfeSYlEgg3sq6ur6cQTT6SZM2eGOgYoHXHQFGJUfIiDniwSiQQNGTKEli9fXtD+oDhATyBq4qApjKPiA/QEoiYOmkqn09TU1BT6mioVTHJtBi0tLXT11VfTfvvtR7/5zW/o/PPPp8WLF9OBBx5IL7/8sqp/44030qWXXkpTpkyhadOm0ezZs2n//fen5ubmTZ7noosuomOPPZZGjx5Nf/jDH+j000+nGTNm0D777JN3kPXoo49S7969af78+bTddttRr169qHfv3nTqqadSa2vrZlw9iJo46Omll16iMWPGUO/evdn2PfbYg4jIbCcoH3HQFGJUfIiDnjpYvXo1LVmyhN577z26+OKL6aGHHqIvfelLBVw1KBbQE4iaOGgK46j4AD2BqImDpnoMDphcd911jojc888/v9E62WzWtbW1sW3Lli1zjY2N7oQTTujcNnfuXEdErqamxn388ced22fNmuWIyJ1xxhmd28477zzX9bZ88MEHLpFIuIsuuoid57XXXnPJZFJtl+y8886utrbW1dbWutNOO83ddddd7rTTTnNE5I455phNdwKIjJ6ipx133NHtv//+avvrr7/uiMhdeeWVm9wfREdP0RRiVGXQU/TUwcknn+yIyBGR833fHXnkkW7p0qWh9gWbD/QEoqanaArjqMoAegJR01M01ZXnn3/eEZG77rrrQu9TKeCXXJtBIpGgqqoqIiIKgoCWLl1K2WyWdtttN/rvf/+r6k+ePJm22WabzvIee+xB48ePpwcffHCj57j77rspCAI66qijaMmSJZ3/mpqaaPTo0Zv8eSPRui8krFmzho499li69NJL6fDDD6dLL72UTj75ZLr11lvpnXfeKfDqQdTEQU9r166ldDqttldXV3f+HVQOcdAUYlR8iIOeOjj99NPpkUceoRtuuIEmTZpEuVyO2tvbu3nFoJhATyBq4qApjKPiA/QEoiYOmuopwHh+M7nhhhvo97//Pb311luUyWQ6t48YMULVHT16tNo2ZswYuv322zd6/HfeeYecc+a+RESpVGqT7aupqSEiom984xts+ze/+U266qqraObMmRs9Nig9cdBTW1ub2t6xrKxDb6ByiIOmiBCj4kKl66mDsWPH0tixY4lonQnvAQccQIcccgjNmjWLPM8LdQxQfKAnEDWVrimMo+IF9ASiptI11VPAJNdmcNNNN9Hxxx9PkydPpp/85Cc0YMAASiQSNH36dHrvvfciOUcQBOR5Hj300EOUSCTU33v16rXJ/QcNGkSvv/46NTY2su0DBgwgIqJly5ZF0k6w+cRBTwMHDqT58+er7QsXLiSidXoDlUMcNIUYFR/ioKeNceSRR9LJJ59Mb7/9Nm233Xab20wQAdATiJo4aArjqPgAPYGoiYOmegqY5NoM7rzzTho5ciTdfffd7L/knXfeeWZ9a9nN22+/TcOHD9/oObbddltyztGIESNozJgx3W7juHHj6JFHHuk0de5gwYIFRES09dZbd/uYoDjEQU+77LILPf7449TS0sJMLmfNmtX5d1A5xEFTiFHxIQ562hgdSzZWrFgR2THB5gE9gaiJg6YwjooP0BOImjhoqqcAT67NoGN21DnXuW3WrFk0c+ZMs/69997LZtufe+45mjVrFk2aNGmj5zj88MMpkUjQBRdcwM7Tcd5PP/10k2086qijiIjommuuYduvvvpqSiaTtN9++21yf1A64qCnI488knK5HP35z3/u3NbW1kbXXXcdjR8/noYMGbLJ/UFpiYOmEKPiQxz09Mknn6htmUyGbrzxRqqpqaEddthhk/uD0gE9gaiJg6YwjooP0BOImjhoqqeAX3Ll4dprr6WHH35Ybf/Rj35EBx98MN199930ta99jb761a/S3Llz6corr6QddtiBVq1apfYZNWoU7bXXXnTqqadSW1sbXXLJJdSvXz86++yzN3r+bbfdln75y1/StGnT6IMPPqDJkydTfX09zZ07l+655x466aST6Kyzztro/rvuuiudcMIJdO2111I2m6V9992XnnjiCbrjjjto2rRp+BlriYm7nsaPH09f//rXadq0afTJJ5/QqFGj6IYbbqAPPvhATVKA0hB3TSFGVRZx19PJJ59MLS0ttM8++9A222xDixYtor/97W/01ltv0e9///st5mf6lQL0BKIm7prCOKqygJ5A1MRdU0REf/rTn2j58uWdqyruv/9++vjjj4mI6LTTTqOGhobudEl5KN6HG+NNx2dAN/Zv3rx5LggC96tf/coNGzbMpdNpt+uuu7oHHnjAHXfccW7YsGGdx+r4DOjvfvc79/vf/94NGTLEpdNpt/fee7tXXnmFnVd+BrSDu+66y+21116urq7O1dXVubFjx7opU6a4OXPm5L2W9vZ2d/7557thw4a5VCrlRo0a5S6++OLN7SLQDXqSntauXevOOuss19TU5NLptNt9993dww8/vNl9BLpHT9IUYlT56Sl6uuWWW9zEiRNdY2OjSyaTbquttnITJ0509913XyT9BMIBPYGo6Smacg7jqEoAegJR05M0NWzYsI1ex9y5cze3q0qC55z4HRsAAAAAAAAAAAAAADEDnlwAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxB5McgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQezDJBQAAAAAAAAAAAABiT9EmuS677DIaPnw4VVdX0/jx4+m5554r1qkAAAAAAAAAAAAAwBaO55xzUR/0tttuo2OPPZauvPJKGj9+PF1yySV0xx130Jw5c2jAgAGb3DcIAlqwYAHV19eT53lRNw1UGM45WrlyJQ0aNIh8vzhzrtDUlgP0BKIGmgJRAj2BqCm2pqCnLQvEKBAl0BOImtCackVgjz32cFOmTOks53I5N2jQIDd9+vS8+86bN88REf5tYf/mzZtXDClCU1voP+gJ/6Ap/Kvkf9AT/sVFU9DTlvkPMQr/oCf8q+R/+TSVpIhpb2+nF198kaZNm9a5zfd9mjhxIs2cOVPVb2tro7a2ts6yW//Dsn1HnkpJP71uW02V2s+lxMxdTrfFb8uIxmVUHS8I+HGTokuSCX3gHN/Hy+jjqnOJfaxju5pqXq5N61PXplg5qNIzmF7gWNlfm1V15DZPtNe8pqzoZOuaEqI9Rv+5xIZt2aCNnvzgKqqvr9fHKpCNamrEKXk0xdvq5Zyq47WLfmtr1w0QP47ser1EZGtK3DOz/1vFuawfYUpNVXMNhdGUeraIiMStTrQZmlqz6efNy+p9KCO2BcY1dUNTZdOT0a8uydsdRk+hYlRUeiokRkk91fGYRUSUq+ExNKjS7ZN9Ual6IoKmOoGm1p88pjGqB+Q8F2g9eSmujVA5r0bkPGMcVdac54tfIqT0UL2YMaqU8Sl2eurh8YmosnOefC6JCnzXq6RxeUSa8i1NyXc98bz1uJw3/ORN60ncd6kLIiNGGXqifHoyYnaouQOhZTNGFaQnkfPkc0WktGvGKDmfEEZPcu4gTM6znkcWo9rpiQ/zayrySa4lS5ZQLpejxsZGtr2xsZHeeustVX/69Ol0wQUX6Ib5aUom1gs1YQzOpKBId5ovH0JZJiKPpFBFl6jzEMko6+UMscifzzlrFk4k00R6k2UiIi/J+yIwhKomuZKGUBMieSbk4MS6JnEN5jUlNl0m695RpD8vjUpTnqEpT9aRDyZR/gGapSlPJFOz/zd9nnV1Nl9TZvCTA34jkOV73rzAuG6lD2sE031NlV5PVjIV1x9CT+FiVFR6KlaMEhPxRrLyRPsqXU9E0BQ01Xn2/HUqMkbFP+c5o+89Xwzee0TOE8f1Nz3J1Xm+iDRVyvjUM/TU8+ITUWXmPDKaVNi7XqWPy7uvKT8X5l2Pd2DPznkhJrk8IwaoGGUth8unJ2t6JczcAe8jO0YV6T1PTnKZMUpMzIXSkxwHhsl5YXSZX1ORT3J1l2nTptGZZ57ZWW5paaEhQ4asmyXtEF8Y27CEvlBXxS/P6gwnk6n4L5vOELc181tIHRWIRfuc1V7xAATGdeu+yH+bfbmPNYsqZ/fNYB5idr/Lw2W9A20uG9WUcxvaHOL2OKtvU/J69IuDOo4MJlbAlHox+tZTs+HGRUjN5Csb2wJjgKAk5Ax9OJ6UZTI1NSW2yQnadacK/1+MyqanEDEqjJ5CxSihJ3NwGkZPYWJUHv2EiVHWdett+WOUJ5+jdkNP4r+eecYvibqjJyJoqvM40NS6fTZTUz0y5+UZR0WV88z/SlxIzlPjKKN96r+jVmbOI4peU6H0ZPxKSz5yofTk6RcypacixSc1piLa4uMTUWXnPPNdr5JyXoVpypNNlvv0sJznOdcZU52V80QzrXd8HaMKmIg3/8NJiLkD+WsvK+epCaFNzyWsqyNynrVipwBUzmsz9JQTedzKeSF+vdz1OsNqKvJJrv79+1MikaDm5ma2vbm5mZqamlT9dDpN6bSR5AAoEGgKRAn0BKIGmgJRAj2BKIGeQNRAUyBKoCcQhsg/c1BVVUXjxo2jGTNmdG4LgoBmzJhBEyZMiPp0AAAAAAAAAAAAAAAUZ7nimWeeSccddxzttttutMcee9All1xCq1evpu9+97vFOB0AAAAAAAAAAAAA2MIpyiTX0UcfTYsXL6Zzzz2XFi1aRLvssgs9/PDDyoweAAAAAAAAAAAAAIAoKJrx/NSpU2nq1KkF7+/lcuR1OItlLdM7YVxmmLwpE/kQXxOQBm7OMu4XHxewqrz7/cGs/PMj71B1/nryIaycmr+cHzejndU80Ree8elraShn+7OJWy9M35ThJekvTJgmw9JQ0TIe7XocF/mK2Y3iZbMbvvxgft5V3PswmjK/uiGQfWCYAnryS7JhvkKSy++8p4xVLU1Jo0PjnsiveAZp66ujHF9oyrc+4iCftzCasvq8s4+j+3pLPpieMps2SSQyzJiJ1KfrXdK4frkphJ5cVlYx+kWaP4YxxZRG0Iae/Cy/TuvT10FKtsd6jnifyo9jWF+8Ue0JcU1mXuh6bK9MMQqa6qRHaapceiplzpNVzH6VG0LEbsuYOp+ZeKE5T4yjAq9Scx5RqfJet/VUwvhEGZe/joxPBYyhekR8svSUrMycpw29839krKw5LyJNeTn5FQfj1EJTzrxvW1bOo1yu05Xckx9LI1JzBWbOS8ivFxofHFDG8yHe84zm5CVMzpPXaTxH8p6acwfiGqw4JgmV88S5zZwnsD4E0fWZtbVutC9ULQAAAAAAAAAAAAAAKhhMcgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQe4rmybXZtGU2+EQl9dpMT6z1D4z1mWqtrbG0VXozqL/L9da0zi+MlY11v36Wr1PtnWhVdTK1vPurWts32RYiokSo9cRe3jrqsnJ8QyJndFZC9JVlWCbWD3vG8nS2HjdP/0dKe3bDGnHLT0L6RxgeHNKbykta+jD6jh3EuGbptWatJc9wExOXNepIrxHZNqM5vvTbstbWS00ZXmtOeJjIpeTOeJbU2nLTBI9EHeM42XXb8j3PkdJNPTlDTy6hzGtUHe0fIw9iaFCuv7f0JPTjMtIkh5SXQhhXEakVqS8iolWDa1h56U76MKce9E9WvuuCL7Ny/Qdr9E7y2QoVo3T/xSVGQVNdzuWLMYHh3SI9IGV+t+J55JqqYD1FlvOkfio85+lnRvuTlDTnST1Z/ddlv5LlvUqJT2XUU8XHpxBSsMZQLluZOU+966WMvlXvemXMeUXSlBl/ktI7StdR/eVCvOvFOed1mTvwDF8nJzVm+QbKfrRmSvLlvEL1FEXOM/zApBeb6ZdYLf039alVzhMxqtJyHn7JBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9mOQCAAAAAAAAAAAAALGnYo3nXRCQW++gaBmMOWFSZhmHqm2WEZwwRvcy0rxSO6d7rW18Q5thGC9O1e60yZsvjOucOI5psO3za0pY5nHC0FIaqFrtUwb7humbJ4w9KWu4ysv9DAO8rlu8nGFMXCSYpoy2O+m1aBk9ym2GgZ7+MMGmy0RE3lqhqbX6QwXKxDln9L8wdZYKssz8vER+TUkTTNOcXpnDig2GMWUoTVnPwcYol56sGCU3+fn1pMxQiYhkl4g+89sNY0oZkyw9tYs6lp6EcacTu1imzp74UMh73+ut6nxv38dY+Wf95xhH4vxlB37c+ndDxCjL2DOEnniMyv9BkKiApsJpys57mzYBJzJyqixL/VD0miqbnqLKeYbfqzQdln1WzJzX9rlRrFxz7gJWrvJXq31efqWRn6dOH7fvrBQrJ9eqKlT/kbiGcuQ8opLlve7qqeLik9DT+z/aTtXxtl/FyiN/soI3raFO7bNorwZW7j9bn9vPivF+seJTgXqq2JwnN1ia8is85+Ubl+s9Csp51scMpKbUu15Py3m5gNz6QGSal8v61utxqLmDCHKenEugAt/zROAN857nG3p65wQ+JXT8bs+oOk//cDzfIOcSotKTMXfA6oTUFH7JBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9FevJ5XkeeR1rMq21mcKbKtTa/6yxrlasT5brstUaWiK1LlutoSWipr3ms3LGGV0t10ZLjLW4nljL6hnryP023jdeSs9lyr7w2/hx/LWGx0M732Z5MugTWavNu/w5MNbmFgmmKakfIuU7Jdf5W3XI0JT0JZP3yGs11hJLTcl1/kRaD551DZuet3bWWm2hqbe/00vVefbw37Ny2jj3OYv2YeUnbtmdlQc/uFifuy3E+nO108Y1VTY9WW2KSE/Sp6KgGFWonvI8vy6r+3vBV7jfzZf2eEXV+WFfvq3N8Cz82tuHsXLdfBmzdIxSz5blyRBm7X/XP0NT6yijpsLlPeH1ZfgGquuWea8Emop9zrOGLflynqUn4UfiDH8SafI0/7s7qirf/d6DrNwnsYaVfcNE7MRJ/2bllbkaVWf3L81j5dfbB6g6F114HCtvNbuFn9vI9VHnPKLSaapUYyj9nIqxZ4gxFKWrVJU5Zw9n5VO+/C9VZ1BqOSuf/4OjWPnAL/5X7fOTvi+w8s2LJ6g67/3P9qycajHiXBTxKUZ6IgqR88Q2y39Z6c7wwvOjeNcLEaPCjMsXfGM0K+/6zdfULt8dwOPakMQqVefo2d9l5YZf67G7esctg6ZKqqeET15HbDLnDnjR9POWHm+GnqSXWaicJ/Vk+nl3X0/6EPnf83K9dHz8xRceYOU31gzSx2kXMapVXLelp6LkvBDHIPySCwAAAAAAAAAAAAD0ADDJBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9FWs8T+kUkb/OGM2lU+rPLiVNZg3jeWkeZ51HGsoJczZptk6kjeaDYU2qzqnDuIHbMytHqzo1by3iGwyjTHVuaTgnDW+JyBOGeJZRrJ/hpm3KPM4yzZN9Y5xbmZEmtaF02Uglifx1WnIpLX2lKcuQ0JOaskxTC9CUNHG2jPmkAWGVfi5W7jmClVcP4PtYRrArtuN6ufbgP6s6q4V5Zbun2/ejrR9n5QNP5WaaP02eoPYZdtMHfINl6pwU90oa1xJtMCnMY1YYKV1jVBg9GR+A0DGqsvTkyb4XvP/94WrbFd+5kpX7CiNoIqKVwoh03/+crOqM/F9eZ8DaT3nbrBhl9IVCxag8/60HmiIiosH3rWDl+Wv6qDrvPsnjj0vqc1cv5u0bfMcHonH5zYcNL3HyRd5zVm6UeU98YKUkmiqlniol58kyGTFKGu4SqRgV7LNcVRmUWsbK18zbm5U/emqo2mfgTH5u66NE077O++amA69UdfY541lWfv3bfJxnGqRLPW1uzpP/v5h0V09RxSf5kaMQOU+azBMRnTPpblau9fXzviZI8+P6vC3vr+qn9kn14+376cB/qjo//HEDP+7P9HEiiU9x0hNR93NeiHc93/qYgXwvimocleDtW7vHtvo4P17Cyr8eeQ0/rP12ysgYdR7Y+XpW/tUl+6o6T9ywBysPepS3BTkvxEfrrOGzfM4KmDsIo6dCxuUm4n59dEC1qvKFmvdZ+bmVI1QdOVcgP64Sau7AilHyuiOaO8AvuQAAAAAAAAAAAABA7MEkFwAAAAAAAAAAAACIPZjkAgAAAAAAAAAAAACxp2I9uVxVFbnE+nXaVbqZQRVfrxkY64GdXGpbyLpgy3dKMO+ABrWtt9/Kyv+Ys5OqM7qmhW8I076kXE+cf92q9Coh0mv/w3hoUFZss84t22esJ6Yu6+ed9F4oIi7dRVNpQ1NirXZg+EmoZfEhlq7rhhjXLO+90bdeivdl8yEjVZ3vnsG94MamF/DDGv4X1R6/iHpf+4jcs3JnVn537QBVZ0wt95j7TPU8Vv7esQ+qfW7/4CusvNVTH6g6JHxiTE2tr1NSPaVS5BLr1/5bepIxyvC7kXpSMcs8sdt0maggPVF1WtWRngZtQ7Zi5c9MnKP2kR5cdZ6OJXv+/cesPOYG7duV10PDilHS48CKqeKaLN+GrpqDptafy/FzHdX0vKpT943/sHKfxGpVR8ab1tO4Dk+e9R197nZ+3b/+wp2qzqtruPfSAb1fU3XOPucUVu773Ce8Qgk0FbucF0lDQlxziBi19RW1qs5FO36Llbd5eDErj8gK71Mifb+MMeZ21/L2nNyodfnNUS/wDVHFqG7kPKLSaSqvnsoYn5ZOGsPKU778L1VHenD9v2eOVHW2vYbnnTGLeIxYOl57vK05n+fO6uQqVeeybW9j5a9M/omqM+pm7i/X0/VEJHJetW5XmHc9dcxijaOMuC5j1NqtdZ1LRt3FDyNMnjJOx75an+fJas+6J7x9vxjwpKox9pSFrPz3h8bzCj0t51V10ZOh86BKtju/D2Wod/MwOU5i9Znw2/JCjMsVVnvFu/mOe7+bt3n/vWIXta1/Lk+Mkv6JROTE3IEn/dyM9pkxyuv+3AF+yQUAAAAAAAAAAAAAYg8muQAAAAAAAAAAAABA7MEkFwAAAAAAAAAAAACIPRXryUWpJFFiXfNcSq9Xdkm+5jRIhlgza03pyW1yLauxdtRLi/X3X1ii6vTxucdM4t0aVcfVae+jfDixHl2WifR6YuVtQ0Qk1rN6ch2tXJNNRE6sOfaMdb9yrbCrqdJ1urQ5KOFabapKEXWs1TY1Jfo2xBSw840+kP0fQlPKj6R3L1XlzZ9zP6RzP3+XqvO56o/4YQ0PLknK4/q4asneqs5zv9ydlevmaa+d2YO5bxddxD0xtk6uVPusHsj7YivDl6U7miqbngxPFyd9MMLEKOuZSgg9Sf2E0JP1rEoPLldjrP0X1zXq12+w8ukDZqh96nyuJ2NlPY2+aa04kXHfZIzKiTgWJkYljfsi1vpXbIyqME2lfe6rUGd490kPrn6+9lqrF14j8kyP7PkntU9KXEK1cU1fruHeI61O571Tzrmble846Au8Qgk0VVI95R1HlTHnSe+REDGqet4KVWVQcx5/Les5ktdteEetHcTHbFftcqWq8+hK7rUaWYwKk/P8MsSoCo5PA77/ASuPTH+i6vz81cNYeexvWlQdeV2urpqVt5qt97nkaO7tdfFdf1F1Bom+eOqbv1N1jnqOe1X2frmZVyimnioh51kxKiWf1RCaMuoUaxzVNnYbVt7/x8+oOo0JPt6R+ezAK89W+/R9k9/rsf9vtqrz20GPsHLOGO9/vuZ9Vr74uENZedRlc9U+sc55XfWUNvQkdGDns02X1+3H9RJKT2FyXpr3o6utVlXUcxLCM+z9I/uw8m1DblJ1HlrF81m/V3Ws8woYlyuMvlExyvDn6+pBHoQ4DRF+yQUAAAAAAAAAAAAAegCY5AIAAAAAAAAAAAAAsQeTXAAAAAAAAAAAAAAg9mCSCwAAAAAAAAAAAADEnoo1nncJb9MGg8r8zDC280MYzAkDN5cWXeK0GfOy3bZm5T/vdKmqszLgZnEDXtQuablaYawmzOMssztpqGhdkxfwvpAmgkREnpjelMfxDANwZaxXZRjDCfNBdY1EFFRtOHbWMsUvEs7zlEFuV2S/eUbTlDGvMU2sjDKlgZ51P4Rha+sVWVXnryO5mWlfv1XVOXb28az8uQHzWPnXA7VR+NUruGH8s7/eXdWpXcJNpi3d1b3LTQr/+OxEVp6+lzbK3/tbL7Ly3PuaVB2tqY0bXObKpSfLOF2adxoxShlcmoapQk8yRgW6P6SeTHPfar5fYPRr8/gGVr5r0A18H+MBeKGNfzThR5eeouoMSHEjVj8oIEZZHwWRG1KGYap4HnN1hp66PMPQ1Doe+s+urLx8D/0xlWfnjmDlP46/RdW56qP9WHne8j6s7Hn6mhI+3zawXhui3jXqH6ycC3R87JPgRvjl0FRJ9eT7yiC3K2FynuqAMDlPjg1C5LxQMara6FdhLqxyk3VcUWf1ID3O877LjcuHJfVHFK5/dk9W3oEW8wqF6qkbOY+odJqqlPg09+ihap8rBusPA0iqq/hHLyw9BdXifsjmhdDTQY/9UNV5+8CrWDnj6fjUesIyVq4/PaL4JPUUp5wX4l1PmYmbHzMozjjq/a/zWHd9P208X+/zGLXbzWeyctOb+v2weikfc7/x+8+oOpf+nOvl9H4vqjq14oMxe018jZUX3DFE7eN/wo8bp5xHnrfhPhmnlTnP+jaXusvWXIT8eElUOU8cJ1TOk3MHVTpOTDroeVYeltSau/yVfVl5lK/fQwv5yIy6SmvuIJ3/fSRgOS+c8zx+yQUAAAAAAAAAAAAAYg8muQAAAAAAAAAAAABA7On2JNdTTz1FhxxyCA0aNIg8z6N7772X/d05R+eeey4NHDiQampqaOLEifTOO+9E1V4AAAAAAAAAAAAAABTd9uRavXo1ffazn6UTTjiBDj/8cPX33/72t3TppZfSDTfcQCNGjKBzzjmHDjzwQHrjjTeourraOKKNlw3Ic+sX1AZ6XWhCrv031tUGqfxzeIFYV0tinbFvrC/99DN8hWlfv13V+f5r32HlrZv1+vtA+oEl8/ttyWuyfLt8uWbdWpcszu3JNdc5y5BKeH2lDb8tsc49J/0NiCjosl44l9HeX8XCy+XIo3XreL1WSzCybKz9l3qx/EnkWm3RT7l6/RzU/mohK1849H5VJyUMU7765FRVZ9u/8DrP/Jh75Gw1uFbtc+dvDuBtWamft1yaX5MX6PvmUvzc9a/zNdX1+3IfJiKi3km+Ta1rJyJXxTUUpPW5OzSVy5ZHT9Sm/y5borwASHvZWCjfQHkYy2cum98fQB73k917qzq/OZP7wKU9fi8S0jiLiE67kntw9Z+TUXWknsgZHoAZriflE2j4REhMPYnn0dJTrkuMykJTRES03V+4T8fiW7Uvzui1PM/9btB3VJ1EG7+v/aSHpOF/EYjc2JJsUHVOOYf7SVw+5HFVp9YTnVoGTZVLTyXNeTWijyxPk0yIvpbnlmUyxnDiXIH0QSGieQfw3HTF1/+s6nyphnt+bPvYaarOdn8R+Uvoxwvye9Fsbs4jKl3ey6snWb9I8WnaN29X+2ydWM3Kr7UN0ge+v584zyp9bunvJLUdIj4NuU9f46/GcU+l/9f/FVXnrp2vZeWDjjiblYfd9rHaR4KcZxPFOOqDyf1UlScO+S0rS/8tIqJxt3APruEP8gu1NJWt5sdJrdQeRE/+5AusfN71b6g6awKu8R81PcrKZ271A7VPegV/luKU8yibI3Lr+irUr3gsa0Ejz0jku7jvRD6zDiHHFyHG5VbOU/5yIv7M31v7pT7YNIuVT5p3oKoz+GaRd6oMf9QM12GonCfnDkyPtxBzB12uM2zO6/Yk16RJk2jSpEnm35xzdMkll9AvfvELOuyww4iI6MYbb6TGxka699576Zhjjunu6QAAAAAAAAAAAAAAyEukX1ecO3cuLVq0iCZO3PBVtYaGBho/fjzNnDnTnORqa2ujtrYNs9otLfqrSQB0B2gKRAn0BKIGmgJRAj2BKIGeQNRAUyBKoCcQhkiN5xctWkRERI2NjWx7Y2Nj598k06dPp4aGhs5/Q4boz5kC0B2gKRAl0BOIGmgKRAn0BKIEegJRA02BKIGeQBjK/nXFadOm0YoVKzr/zZs3r9xNAjEHmgJRAj2BqIGmQJRATyBKoCcQNdAUiBLoCYQh0uWKTU1NRETU3NxMAwcO7Nze3NxMu+yyi7lPOp2mdDqt/9CeIVpvruZbRoNtwgi7TZvy+dXSRM0wy/Y3bTJp2YYmRnEjv2rDr3BZMzdxHpDVptueNJYXJveBYUbolF+zYYIpNnmGKaMvjeuk2V3OkIY0yTNM+ckwyy81G9OU154hr6PNhuGfLzTlt1epOoHSlNFPog+kMW7zT7S75q3D7mHljDH/fNjDP2TlMdfrjxmsbeKm9rfvejkr7/Xqt9U+NUu50bzhAU5OGIwHWaP/xHVWL+V1PmjfWu0zsGoFKz89tq9u30Lx7JRYYhuNUZks0XoDfr9Nm6vLGOUb5p2BMGi2YpQ0RJV68jytQc8Xpu05HR/lM/D3n/1OVamT9514P2x/46lqn0GzeV9IU14iIk8cV30sg7SZrDKrDGFwSUkj5pc/RPUITflr9AdXJNWfrNHtEzEzjLHz0h14LE4YptezFgxj5fRQ3Te3LPk8P1cP0VSl5zzfGmpKo2ej71eP5OOoeQcZLsHigyd+kpcnbPue2uXhYTP4qY2R3vZ//hErD/+34YotrqEgPRnG+KYJdgkJE5884+NEfjt/NqKKT/O+XM/K+9d+oI8ryjOW76DqDHiWfzxDjXuJyMuJ+yG+amXFJ/ke4Rnp9pb7+IcxzvueNgmv9nhMPeZbj7HyM3fra4pDfCKqwJyXyp/zFu3PdXfJd/+i6mRE9//h091UnUFPc0F40ojbGNSqDyBYHyITm3Z4Ro/vnxp/FSunxJNijc+c1FAFamqjOS+b3fBhrHatJ5Xz2sLoycpfvKjM6j19XPJFYMjm/zCJVccT2mjry9/7/u+7/J4TEaU8fg///d4oVWeIaJ40tCfSMdMTH49zQYi5A0NPxcp5kf6Sa8SIEdTU1EQzZmwYQLS0tNCsWbNowoQJUZ4KAAAAAAAAAAAAAIBOuv1LrlWrVtG7777bWZ47dy69/PLL1LdvXxo6dCidfvrp9Mtf/pJGjx5NI0aMoHPOOYcGDRpEkydPjrLdAAAAAAAAAAAAAAB00u1JrhdeeIG++MUvdpbPPPNMIiI67rjj6Prrr6ezzz6bVq9eTSeddBItX76c9tprL3r44Yepurp6Y4cEAAAAAAAAAAAAAGCz6PYk13777UfO8HbowPM8uvDCC+nCCy/crIZ5uRx5bv0C0UxW/114HXit2h/EZcTEWq1ev+vSxtrQPHxhyFxWNuxBaMC/edf6q7Wfg5cRXhdpeTv07Vk2hl9D8qAlqk5OLNRubddrg/v9rY6Ve70vPr9qrY+V993wm5DeaAljHbnf1fMgq+9t0cjliDo0ZazVluvkyfAH8DO8/z1LU2LN8rLte7Hy73a6We0TiAXe/14zWtUZ/Veu8cQK7fPmb8XbU+2J9ffXD9D7iOfLWqOvsOQhtmVreLlfgnvZERGlhHnFstFa83Uf8jrO8DDx1nuEuaxhhlEk8q39V3pK6ecwjJ6Uv4T0/9h4ON5Qx/C8ItFXzTntx1Mt7s9fPhnPyv1fMnx+xLkCa1G89DOw4o3yLJR/Nw4sY5IVozLCw8fwc/S6es5BU3ZbQ2jKyxh9K7a5NL+GZdvxeElE9MiZ3C/u05zWy1+Xcb+tL71xqKoT/I7Hv2pfxKRSaKqEeipKzqvRccJJn7UC7DXamvR9n/rb21h5aHKpqlPr8zb38Xk+a/D1GO8fa7g/5M+uP1bVafwvP65necNK/1MpnzDjKMPbSj4j0keGiD9/pcp75YpPh339aVauNvo1I8796Kvav2r7Nu7JVUh8ssbl0j/JynlNz3Jd/vSQXVSd/9f/GVb+VsMLrHzz0furfYbfJr5cj5xXUM6zaDr4I1beJtmi6pzxwZGs/O6/Rqo6wz4QXnDS46laa0qNw0PE1Nb5OoYmxI5V4p3A7IcQfomhNBXEIOcldd+Hes/Ll/M2MU/SeVzLs7FdjKE8/Y7shGf2R4fw9u2Q4l7HRES/XMK94gb83fDHIxH7jDgrr1PGPq+YOS+74dhhc17Zv64IAAAAAAAAAAAAAMDmgkkuAAAAAAAAAAAAABB7MMkFAAAAAAAAAAAAAGIPJrkAAAAAAAAAAAAAQOzptvF8WchpgzEnDcsNEzJpMOcbRnAuJ0wMhfFstkGbs53c+DgrN+dqVJ3qZcI8bk2rqkPCeN7P8XN9slu92uWEKf9g5cPqX1d1+vvcGLbW10axu1cfxcp1FwijvTWGUb7sY8M8zhfmrJaxnktu6GOX08aARcPzNhgqWlrIiLYYpvjKtNA4zqd7NrLy4BPeZeWm5Eq1z8ttg1n5tp9NUnXqP1qgtklSK/kHBb705A9Zecx7q9U+uVoZBvJ/jIGZlHZsE/c+04sbEPY1jOdv/GRPVh5613xVx1XxZzRhaSqxrs2x15OBnxUxKpHfgVSaOFof5nj/Nzy+fJqrU3WqhanzA49z88pR72s9Zet4e51yY9bG5r5lRJkVHxeRRuemGaro4zAxysoviQ3PADS1/rhCU2tGNKg6da8u5HV2aFJ1Ui38GgLx8Zfk4YvVPq3img6++8eqzsBneJ2axVrzqdU8D5dDU7HXk6Uv0Y9eAXqyzOqrxIcvFuW05t5YtQ0r3/LeOFa+8rM3qX32qm5m5Uzv/B/QCJPzQukpxDhKYuU8NlbNaa0XhRLpyc/yMevybC0rrzA+AvBiG9fBqJuMc1vjcIkwmjc/sCGQ5sthct4/bv2CqnPiD/7DyjnHj7PDgW+rfdbexPvGehdCzsuf894+S3yojIge2PZ2Vl4c1Ko6bzw/nJW3u/EjVUd+vMDL6ncytY9oXy7Eb1LSg/R4rNUFoszzraXvWOc81oDi6UmOT+X9MveR43LjQy/Wh/Yka4fzcdVNE69i5eXGly9uvoN/tKLpU/2OLzG1IbQg+6HgnCf6PCGPS0SU7JrzwmkKv+QCAAAAAAAAAAAAALEHk1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxJ6K9eRyqRS5xLp1zHINKBHpNZ2BsR5YrL311hprZuX6UeHJlRmofWokf/lkX7Wt7tm5vC2+Prf0c3j3RL7O9p/f/K3a54hXTmTlO649UNXJVfFr+HQnfe4Lj7mZl/f/FisPu3W52ofa86+B9aTHmbFO2+vSx34u/7rgqHDJJLnEOsmrNcJE4TTVxr0vrFXYKybxdfHTt3mMlRcb3kf/c+vRrLztrPf1ua02C6o+4H42Y/+He625Wu0x56W4XnwvxNpyY720187Xkgs7J1oZaO+6N6/fnpWbWueqOvJeeRkdtlxy3dr/sump0Bgl9WT0vbx+l8j/3ybkPi27am+kLw5/hZUX53qrOnWO9+eYqz7hbTH05As9mQ+JWOrvtefXk4w/ls+Y5a+gjiv1I2MWVUiMKqOmFh6o9bLdN95i5d17aT/If/52b968b36q6ixe0IeVj9rteVYeU71I7XPg1Wez8oh/G946MdFUT8x5UquhYpTwHql9tUXV+c2532bl1q30cfu/uoaVhzTz45zfeILaZ/tLuHa33V376GQf4t6afsYYy+TTk+W5EkZP3ch5REReiTRVLj0tbedeSIsNH9y3WweyctXbC1WdMGMo6cfjRPz0klqDoXKeYNBT2j/pme+OYOUd0tyjdI3h5eS15fdj0544yHmrdhjAyt/emfuhEWkPrqdXbafqDH2YP+Ou1fAzLpGmXpxwjdr2cY7vePzs41h564XaI5jCaCoOOa9QPck4br0TyeMY7/gS9Rwa/SxjlBvUX9XZ49cvsPLqgI/Dz377SLXP4Ed5vJFeqESGj7mV89rEtjA5T/qghXg+rfzSdVwRNufhl1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxB5McgEAAAAAAAAAAACA2FOxnlyUShKtX1dLxtpMT/gauHZjrb1YK+osf6g8a0WX7Mx9GYiIFufqWfnx53dUdca2vck3JPT6V+n/tce+fJ/n27ZRuzRcxs+dXK3X9HrCm2TN1rWqzo5V3PfEielOt8bwPAnhJSHvlVel12l3vW4vKN1a7W5rylirTRne32t3HaaqfGf7p1l5cZZ7Hf1s1tfUPmP/9A4/t9XXUr+WpoQXipcVj7ixxtrPiLXl1vpzuVa7VbevZWwfVv7u9x9k5SdbtJ/BVm/z++/WrtXnTvJr8LLG2v/1fVFRehL3IoyenDM8BNr5ffb8/P9tQq7rX7invqffqpvHyjmn67Q7fu7ln+M+FtlqvU/fN1bxDQXqyWvdtF+S5X0h+9NE6cl41ro8W1uipnLaao0m93+JlVudfg73PetZVu6b1L4z243hXjl9fO6p9F471xgRUZ93+DX4ht9WXDRVUXqSOc/MO6KPTM2Jfi0gRlnji63+8QbfYOQ8Ly18imqqWTG1lOuLiOjV83Zh5esuv1jVOXK7n7DygFl6TFQ6PW085xGVUFMl0tOanQex8tH9+JhqUa5B7bMk04sf1+r7KMZQhh9pVGOopVl+DUuTvDy56WW1z207TGLl9Ivv6nP3pJxXaIwSfZCpE+9fde+pXf62ZAIrzz19jKpTPZuP3UupqfbzlrPyKqf9kO5p2YWVB5zK46H1nLiekvMMT66i5Tx5f4z7pXKe1c/Cq3vuEX1UlSOFb6n003XX6TGUn5Hjcn1q6WtqjqGk/53MeZafWxlzHn7JBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9mOQCAAAAAAAAAAAAALGnco3ng4DIW28aJwwbzW2GqbxpPCnryA3CMDWlfXNpUaYPP0adcW7RPmmeSESUG9SPlSf1e5iVX10zVO2TXsKNuZ1xXE+Y7S0fp03fPg1qWHnAS8Kkf402a5UGfcpoj4i8Km4Ca5oedzGPc4E2Siwaztla2hgZbbrn2nlffrpjlarjC1V92N6flQf8S+/jVhtCkySkWbThHChNLy0DVIkwvTRnvkWdlaN6qSpHnPsvVq72+L198Mlxap9RM7mZtTM1xfvTvIPrDR3jrieiEDErhKmzZOpXHlbbpHF4Nem+S4jePvjnj7NyfUKbMf/j2L1Y2TNuiYxRfqvxAQ3x8Qv5UQLTMDVUjAqhp2BD+1wQ4oMbUVEhmhp8f7OqM+OoHVh5x17zVZ0R6cWsXOvrc//itcP4hlncRDqrv5NC7XvyHJts0874Kwfz2DfokcWqTiVoqqL1ZI2j2uWHDPTxZJ8UEqNMipTzqj/hOvjVogNVnfbe/Dje2gqLUV1MjEuW90qkp9Rq3kcft/OxckNSj09XZvgHB6SRMZHR94Xoybr+gOezQnPeFa/uw8rnj7uflTPO+PCCMKo233tSxjhcUg49EZUtRn16ONfQ+236I2P/evEzrDz2v6/qc6f4OMrUlNRimHG50NSi8fWqyi9H3M7r5LQ+Zi4dycry3c6t1WM4+ZE2K0Z1R1Ml1VPgNnx9LYyuypnzAt2H7XvyD9kdO/kxVSfl8Tj2p/O/zspbvbZUnyrNdao+bECk3vPMnLd20x8LK4WeiMJrCr/kAgAAAAAAAAAAAACxB5NcAAAAAAAAAAAAACD2YJILAAAAAAAAAAAAAMSeivXk8toz5HWscc0aa2bF+nZZNutkjTWcedbsDnhhldq2IsdNQ47c5UVV5/XG4azstelztzby43y++kNWfrdVrxGXyHX+RES5Gr72du/t31Z1vv/8say87TNzWDloM7wk5NrlZErVUSttvcqZR2WasrxspF4Mbcg6TQd/pOqsyHG/s0fmbcf3+ed7ap+cWNfsp7XnjFdTLco1qo6rFn5fKf6Iu2SI+5HVmppzch0rHz/h36pO3yR/Vi7819dYeewvucaIiHLSP6hQTXX6IITwOoiIbseoEHoqJEbZjcvfD30S3JuhKblC1ZGeStvWc5+jtKeve86VTaz82gWfVXXq5vJzeYY/iWsTvgjCa8T19BhVTk29rWPU+z/jnnpz6ndUdbJp3pfZaq3DwW/wOOGv4prK9OOxhojo44k8Vy6coNuf68f1ss390JTX1r7BIyainGfpydovf+N4r5k5T2wrKOcZXimByIPDaj5VdeY/yLd57cZ1l1NPzPunNHkvkjGU8MVS/UFE/pPcq/Mvz3KvqmPHP6P22baWx5EXvvUlVWebG95k5cjGUGLI5BleO3678AMzcl6wiHuPLRfvGnPW8NxKRJRaJnzhjHcCv0roo0L0RBRRzguhKbnftgOWsPKSrPaZ9TL5+8ETflsFaSphnEfcxrNOul1VGZ3i1/Bbw1uw9awBvH2t7/PTWB5u0kvZiFHd01QJ9ZTJkJcr/hgqipyXHLyNqjLve1zLg6t0brq7mY/FtnqZ13HSV5CIPNnejO4bT2zz2owxVKuINzLnKU/YcDlP6cmigBhVGSMxAAAAAAAAAAAAAAA2A0xyAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7MMkFAAAAAAAAAAAAAGJPxRrPU1v7BpMxw0RQ4iWM+Tq5zWkzNmWIJs3ZZr2m9rnyfm7ud97h2hDw7qnjWXm7K5aoOqmV3GDutXZuKlntG4bBwg9w5Shtljjmx2+w8skDnlB13vzLVFa2jBolnjCz86q0eRxJc/SqKl2ni3mcpz0yi0c+TQmzRc8y7xaamtD/fVWlLeCP1eq13Dw3t5gbpBIZ5pVGv0lDS1dbreq4anFP5DUY1xTU8HO/830dFk7Z7XFW3q56oarz0LLPsPLY/+PXGaxcqfaJTFMdfwsKMIMslCLoqaAYZSD7tdrTsaTe5waSfRP6IxvVwlhelnNOX9MZA2aw8q/P0Xqaf/wgvsEw6lXmpyGIWk9E0FQHqSdfYeV0ba2q4/Wu54etM0x408J0t4qXk6u0cWm/2TyGnvvra1WdV1qHsvITwQ6qTiVoqqR6as8QdZg6V5iedL8aOa+Oa6ygnGeQEznv0N4vqzrPZHfiGxCjoolPyfyvHVJPgx4VmuPDayIiSojBZM0Bn6g63p1CTyHiU5gxlDJ1tvQv9WPo6av76I9YdeWJj0epbYNm8w9LyY81EBGR/IBDtVGn69/jlvNCmL+/N40/z0f2+U/eptXNKyBGFaCphftspfY5fcqdrDyuep6q0yZi8cIpQ1Ud/8MFvC1Cm55vxUvkPL5PcXLeh9/S9+sXu/L5hDWBflYXXDeSlfullokTGdcUJkZJo/5CzPUNQuW8IsUo/JILAAAAAAAAAAAAAMQeTHIBAAAAAAAAAAAAgNiDSS4AAAAAAAAAAAAAEHsq15MrFxC5jftESQ8u5+v1m16KrxH2sllVx8ltIbyphv6zjZWXHqp9se6ZfAkr377f7qrOowu2Y+VBSb6udreGRWqfh3/LfUamb/s3VWcb4a0z+bKzVZ2h//qQlZ1c/+rr+U+1drlG+2Oo9elpY+1tl7W0m7jF0ZNPU2Jdv0voddhyLXFjSq+Tl2uoU//l+vAtLxvpKVCn67iatCjrNfAuKe6bsH3wDB+IxT/ler5z57+oOq2Ot+8Hr35L1Rk8ZYXYaTk/t+UVIXTm1Wo/A7nNWeu5168dj7uevJy+P4XEKHnuptQKVWfrRAsr9/PbVJ2UeOQPv/An/O9HaN+TG3a4kZVPb3xU1TlhtzNZuf9/tMeb0ksBngdmjJJ6qja8JFiMyu/xExkx0lSoGCX9kqiwGJVo5ff+nys+o+o0pYXGrb6pAE1BT/a5C9ZTSrQ5hJ6WfHbTfh9ExLxDiSgyPUWd84hKmPfKpKe6+dw/sjXQ/bFNFR8//2DkE6rOeed8jZVH36i9Kv1WsS2EntR9D2HVtmb7JrVtZA33AK7317JyOqn7XvpPFaynrn+PeYwKhum+PXPy31k5I8a0OdLXPOSq2fw8EeW8xbv1ZuVzTrtJ7bNLmntptTv9TnbUrO+z8uhlLaqO9DpSWrX8mwqJURWS81w2R85fFz8sv62S5TwjTriduKfe17/xhKozIMm9i69euIuq0/cVcZ/l+7oVowReLkSu6gE5D7/kAgAAAAAAAAAAAACxB5NcAAAAAAAAAAAAACD2dGuSa/r06bT77rtTfX09DRgwgCZPnkxz5sxhdVpbW2nKlCnUr18/6tWrFx1xxBHU3NwcaaMBAAAAAAAAAAAAAOhKtya5nnzySZoyZQo9++yz9Mgjj1Amk6EDDjiAVq9e3VnnjDPOoPvvv5/uuOMOevLJJ2nBggV0+OGHR95wAAAAAAAAAAAAAAA66Jbx/MMPP8zK119/PQ0YMIBefPFF2meffWjFihV0zTXX0M0330z7778/ERFdd911tP3229Ozzz5Ln//857vRsgSRv755CWMuThjKeYZRusIwY1OGctJMzjBVS7/DDeFvO+8rqs7dJ+/CytO3vVvVOb3fs6ycEedqNfzcrh7Djeatq57y/lGsnFyt60izRGXQl2xX+6g+lmb1xnGVKSwRuw8uhEFtZHTVlGW2KHRmmRZKhlctUdtWC+P5vnO4pvze9ca5RT+lDRPHVP7HVZkJCs3/8j5uCk5ElHH83JYB58kvf4eVh/5c6yO/pnT7VZ1CNZXrMLgM4fgaFd3VU6ExKiueEfnMWMaPQk99/DWqSrWnP8QhaXX8urZ+jpv7es/oe7HmH/kNXY+d9gAr3zj9YFWn30weZ1X/hdGTYRKu9FRlPFdxiVFl1FSoGGVdgzRyFudu71+ndvnocL7Tx0/toeq4Kt7mseml+tx+nnFDCTRV0XoKkfNy/XT+SixaJioVoCfLaDaEnvLlvI++2lftc+w3H2Hl4189TtVp8vhxzY/oSD2VIecRlTDvFUFPljZkfPLf57ng9uf1h5yu2J+PbWqND6k8cOClrHz01t/T5362gZUHvMiPk17IjaEt1g7prbYt2Ifr5+zD71F1dkzPZ+WHVu7Myn1+o2OjL83QkfPM4zy1bAwrf6txJitLI3oiIq+ef+ir0Jwnjebv+fnvWHl1oK+pVYzLp5z6Q1Vn1EfcgDxUjOrhOc9LJsjr0JOllUJiVJi5A3GNzthnznf4R8hOqf1I1ZnRwjX36S+HqzrVTozns/xcckxlEuJZc34Jc574cIGppwJy3mZ9XXHFinVfM+rbd90g4sUXX6RMJkMTJ07srDN27FgaOnQozZw505zkamtro7a2DUmkpcX4OgQA3QCaAlECPYGogaZAlEBPIEqgJxA10BSIEugJhKFg4/kgCOj000+nPffck3baaSciIlq0aBFVVVVRnz59WN3GxkZatGiRcZR1Pl8NDQ2d/4YMGVJokwAgImgKRAv0BKIGmgJRAj2BKIGeQNRAUyBKoCcQhoInuaZMmUKzZ8+mW2+9dbMaMG3aNFqxYkXnv3nz5m3W8QCApkCUQE8gaqApECXQE4gS6AlEDTQFogR6AmEoaLni1KlT6YEHHqCnnnqKBg8e3Lm9qamJ2tvbafny5ezXXM3NzdTU1GQeK51OUzpteRJUk0us326tHU3y9coukX9drfJuINJrbeU6T2NdraT3a5/qjT/i13RB6tu6jvAiaN5zK1b2DtLHvWanv7Ly/FyDqvPOrGGsvO0zK/Sp5dpy6XFm+aDJYxgeGi4p1vQa98Xr6vMTZj10N9mopqrT3dNUiLX/Wye0P4P0ghjw4/dZecUv9H9xSM1fzjcY524dzNf1L/6s9gdo+BL/xeRXBr3B98lpP5WnV3Gvgr89O0HV2e6ataxseYTItf4ujKZEnVCaMvrGc+vXoBuP+OYSSk8h1v6H0ZNnedfIGCT8SpQ3ABE5sW7+z4v2VXXOHsQ9FlcbnhQZJ/wL1nIvNuvcLY73VbWXUXX2r+Vf5b1im0NUnX4iRoXSk8AZPnZKT5bPT94jbx49QVNm++S5M4Y3hIgduV68H1aM0HGtpjePs42316g61R/zOpanQ0ExSlCJmipmzluxUx9WPv68v6s6H7T2Z+VnFo9k5bpT9KlD6Ulot23rWlVl0Xh+3ftN/i8rf6/hQbXP++1bs3LNX/sY7RPjJit9VULOI4o875U9PknPQBGvdvj1YrXLBUO4r+Mfx+r/AC/z2c2fu0af+3O8+N/Woaz8+zcnksQT/m03fvYyVSft8WtKePq6ZftueXAfVh790ce6vdXSv63y4hNRaTXlr9Vjjmdf5uPcH056lJXbSY9ps0N4XFszUOedgT9+l5/buK9H9+PnWpzjOe7KT76o9nn1T9yPre9C4z1O3kfLx0h6Hyo/a+POyxhVqKY6tpXrPS+EnsKMY0xPyTyeXO2D+fs8EdHX9+E+3NbY+M6Xx7Hy9vOMZZiyPRnur2uN39Q9TRo+j+I+y3tMZOQ8+fdCx1B5dEpERH6Xs4XMed36JZdzjqZOnUr33HMPPfbYYzRixAj293HjxlEqlaIZM2Z0bpszZw599NFHNGGCfnEGAAAAAAAAAAAAACAKuvVLrilTptDNN99M9913H9XX13f6bDU0NFBNTQ01NDTQiSeeSGeeeSb17duXevfuTaeddhpNmDChe19WBAAAAAAAAAAAAACgG3RrkuuKK64gIqL99tuPbb/uuuvo+OOPJyKiiy++mHzfpyOOOILa2trowAMPpMsvvzySxgIAAAAAAAAAAAAAYNGtSS5nrUsVVFdX02WXXUaXXabXpHfrXLUb1tWafltyXWihnlxZvnbVUx5deh9VJ8R6XU+smbW2Dbxf+Dvdo/f5efJwfuoavSZ5VO1yXsdYp636S5VDrKuVa2gp3Hp51l8hNBUVTFPmWm2xZtlaqy54Wfg1EBF9vmYuK/9i8D9Y+Q+//rLaZ87lO7LyiFPmqDo71nBvry/3nq3qtDru77Em4Pr41XsHqX28P3F/ku3f1H4Xrlb4PliaEvfSk5oKow1rnXiY/QLxvyXA1VV3L0aFsCXwAiuWiH4Va/+tGKX8/s4dqar86gKuhamDZqg6K4NqVn7vl71YOcjpe7M0x+v08deoOvUJ7kXgjFsc1HAtl0VP8v8XmUrWVLYP9yOxzi3tSN47olrVaRi1jJUPG/YyKz/2873UPo0/WsrPbeQ9FaMM30C5rcfHqAhyXltvvu3LtW+rOq013JfmmK2eY+WLb9Y5LxPwfgwMQQXCKWhy3/+oOtukuDZmr+Welz/55zfUPqNuaWXlhk8+UXXC6YmXi6Un21dvI/+/iJQqPnlZ0SchxuXVf+SepUcfNkXVGb8z1+lpAx9VdeQYqm9iFSufv+MDap+Ux8fqrYa/Za3P69Qb3k0nf/A1Vh78uMiT1TruKXp6zrPqCA1ZmnJV/KL6JngMkH5oREQ7XcbH2Nukl6s6+9a9xcrWvZea+t2Cr7Dysh8PJkn/eR+ycqE5j6Tln7zMYua8jufUGqMWCaanEDkvlF+Y4c2t5hPEXELy3Ga1z6Ter7LyL96ZrOps/3sxD9Cufbs86Vko5xfkeI5Ie7P5hp6UF1uInCf7z5gXUO/5lp66OYdjzucYFPx1RQAAAAAAAAAAAAAAKgVMcgEAAAAAAAAAAACA2INJLgAAAAAAAAAAAAAQezDJBQAAAAAAAAAAAABiT7eM50tJrjZFXrKKiIhcUhuShTEFl+aDXkabx8lZPif82zxnGLhJN82c4dIojN8s43lqa+eHzQiDOWufFL9lnmHg5ogbygUpPZfphLG8JwzmfMNYvCCTeMu4v4tBn2f1XZFgmrL0I014DUNCaXZ39f9MVnXW/PwhVt5fGFP+qEmbnWYueJyV20nfV2kivzKoUXVeWjOMlW95Yk9W3u7cN9Q+lFrBil6vOlVFa8oyouT9JTXlWZqS+ggx7W4ZDnYYZysD7SKST0+FfAzDy1oxSlyTMEh1hkusl+XHTX+8QtVZfWw9K/8mc7BuoIhJ22Y+Fn/XMerPqT14Www9BQ182zZ1q1WdXLWIdYXEqBCmolafd9XRlqCpT3bfipWX78NNeYmILv38Lay8dWKlqpPvwxdERHcs2Z2Xb9mPlWsG6Oe7rpW3J1zeM+okNh2jSqGpStJTmJxXP58/44e99H1V58tD+IdSvtfvaVb+UaPOee0i2EvtEBG1BnzbA8t3UXX+7/bJrDz0xvdYeeyaN9U+ahwVJudZ2siX86w8KaWxmTlP/v9iUrr4JPYRacYZRtA1H/Ect/0lus6K9ABWPu64H6g6QYq3b49x77DyflvpjwLV+m2sXCfKRETnvDuZld9/ZRtVZ+R9fL+kMJ0O6vWHPLa4nGd98ES+67Xr6xt5K9/233242fvnqsXYhnQcW2OYyq92Vaz8WusQVefqyw5h5YGPcFPyxKIP1D6uoHc95LxcTRc9WfEoRIyS7yV+1tpHvOOLeYF3XtY68IfzPsr8rVHV8dYu5BuseQAxLndiLsE0ng+jp+oQH1vJM3fgZSLKeYae2AcAjBxggV9yAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7MMkFAAAAAAAAAAAAAGJPxXpyBVU+Bcl1c3Cmp5Rcu234RfkZvi1heHKRWPfpZcQ6W2ttq/SRMtaGKq8pw3vKyW1i7a1rF+tsibTXV1WVquJlhVeTsZ5a9qknfM9cYMx/ykuw+lz0p5+11hN32ZYz/l4kgqoEBevXIVtrtQPRB55hQeaLtf6931+j6tx23ldYec15/B5NqOMeD0REGcfXMeec7v92Uecn//qGqjPqZu5ds91s7sFVuKbEzTeWqAdVos3iGixfEakhq8/VM2lpKru+TgnX/gepLjHK8g2UXgiWnkRMkt4j67ZtOkaZa9Oz0sfLOLncz6hTtBgltWDoyYkYJe0wXDL/f6ORXh1Ehp6svunaf1uAplYO53+/fc+r1D6t4gZc1fxFVWdozVJWfm9Nf1Xn369tx8rbXz6blS1NuYT094uxpkqppzw5L4yHUnI1b+/AX2vvrDdax7LyjzOjeQUZj4hU/DH9OWWMMvxJhrTzfOrauK9RZDnPQI1N5bjK8uSKOucRlUxTFZPzQujJHJ+2cC1s94cWVccJH6yWNt7A+9q11w6J+OT3rldVUn1rWXnbhrWqTpDmx3Gy3BNzXtd3PStGyWs22p4Q43IrqqWW8rHxzx45ipXv/uqlap+cONLza0eoOpdddxgr1zbr9g18SIz5o4pRheQ8Eb6dHLcTqXc9U1O5PM8oUVlynqvyOzUTJIy5Axm3jGtLyA4IE6NEeburlqh9fn3F4azcP7dQ1VFxzBjfq3G56N+C9STOZXl05ps7ICNGqTxgztfI+G1opoC5A/ySCwAAAAAAAAAAAADEHkxyAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7MMkFAAAAAAAAAAAAAGJPxRrPk1v/b2N/Fl5nnuG45zlhgm4Yz/vtwrxMmE56htGpMk0zzNnUtqQ2IPVywlQyK+r4xhykOLfLZHQddQ2GwZw0IxRG/qYhrVCLZUZo+VluslKoHaLByzny1jvghTHcNY8hTWMNc9rebyxj5Se/thMrP5XdXh84jKYE27sFeqO499Kw2dSUNC0MoSm/XWtKGVxKTRlGtdIp05KDb+hMHWW9EaM0TiwmnnNd9JA/BjgroIlNvmHe6eWLUZYJbyExyjDp9JKliVG+EaMK0ZM0yrRilC/7pk1VKV+MKpOm6nbkMWt5ID5eQkSnX3ky32ehftYW/UcYqRrP4w7t83hzw8SonqSpCsp5ahxlmaCHyHlFi1ESaxwlzXLDxKhCcp6lJ2HaLLVSipxHVLq81+NyXiF6sgijpzZuBl1IfLI+wiUpND6VQ0/rzuU2tLnAcbnSlBGj/DZ+T8Ze8gkr//x/v24cN7+mhnrz+Qar7/KNoywiilFBlTiX9F03xn0kd7GMwttC5LEy5DwKaINxvvXdD5nz5EfsOo7RtU6IGOUJfRVsth9FjAqT8yxz+jA5TxjLq3GFOYbiZStGyU0JK0Z1qWR+PMMAv+QCAAAAAAAAAAAAALEHk1wAAAAAAAAAAAAAIPZgkgsAAAAAAAAAAAAAxJ6K9eRKtGYpkVy35tVzxsLaMGvTs2LNprFOm4TnlifWpMo1qkSk17umdDc6uY7WWCPriW1qJavT7XXSIyyrPcO8tXwxa6Iqpeqo46bEmnFjXa1cuy39J9Ztk2XDK63rtjA+HBGRaMtSorO/DOk7cY+spuXy+5MoTcm12sY9Uxj3zMltVt8Kf5uCNGVoXmrKtzTlbdo/QvpLEOn13SZyP0t367XpSjhv76/Nkr8+RlGg9RTCVkXpJzI9FRKjPF2nZHqy2idjifICMO61PIxpiSN8Y6zYXKYYVS5NDZrK/Rn+kP2y2mWo/wHfYN0zuQ2a6lLFM+sWE5bzrHGUtU1Szpwn72GxxlFxzHld6jgp7CJR0TlPEkZPUcWnEO8NYfRUSHxSY/UQ8WlTYyiiHjCOMjyUYjeOKkRTRvukZZJ61zNjVIh4EtecZxl1CaRnVJgYpcvG3IHsBysGKD0Z73mF5DypH8NvPFTOEwRST5avXhivPVnF0FMhOQ+/5AIAAAAAAAAAAAAAsQeTXAAAAAAAAAAAAAAg9mCSCwAAAAAAAAAAAADEHkxyAQAAAAAAAAAAAIDYU7HG8/7qDPkdRoiBYX4mDApNozwXwrGwEKThnjSKI1KGcqYZm9jPk+0NtHmcNMRTBqpERGvX8n2M9iXEuVyV6M+U3serym9Or7BMVruaI/qGQWSR8Nd00ZSlDSceB+OeeUWSlDLZS1gfW5CaN/o2G4GmcsY9aW3l+xia8sW5vBCaCqSmLONngXXdne3xQhgpR4S/Nkt+Yp2Zo6kLJ58Xo91h9CSNJw2j0Pz75HkOqcR6kjHKuO/yTirD1AL1pPwq88VmaGoj+0BTndu6o6lS6onlvPwmsmbOC+P8XAhhcp4cR1mak+Mo+feekPOs+1KGcVTJ4lMhlHEMRTm+rVA9lSo+bXIMRVRZOU+8pprPgtgv7weviMLlPEkJc14oTRUr56Xzv1/HNecprYSZO7CM8/OZ6Vt/LyRGWXWSXAuljFEq54WKUfmfYUlUOQ+/5AIAAAAAAAAAAAAAsQeTXAAAAAAAAAAAAAAg9lTcckW3/qdw2Vxb57Ygq3+WFiTEz9/0j9TJy8qf5+mlfX6X8xAReUG7OJGxHNDjc4MuZ/zGVmwz2yfqeAFvi5NtISJyfJtzum+8wBPlNlXH5cSSS/HTRWtZQCB+cmtdk7pu4yeRfm7DNXTcZ1espaW0EU0Z7cplxeOgfo9L5GXl9WV0nbyasn5mKc5lLBGROnOe8ZNz8ZPU4mlKh45INOWMaxJ94ZvP8bprKJeeLJ3n1POir81XMaoQPeWPUdbPlytfTyLO+qI/C9ST/Cm6nRfKH6Ogqa4be46m4pfz8o+jSpbzzPywZec8ouJrqvLiU2WNoZzQEzlriWw84hNRpeU88a5nxCipKft5KSDnbbGayv9+rd/14pHzwlxbeXOeOLdacErkiesqrZ5EzlN6MpYrZnn/Wc+wznnRzB14rpiqK4CPP/6YhgwZUu5mgBIzb948Gjx4cFGODU1teUBPIGqgKRAl0BOImmJpCnraMkGMAlECPYGoyaepipvkCoKAFixYQPX19bRy5UoaMmQIzZs3j3r37l3upvU4Wlpayt6/zjlauXIlDRo0iHzL1DECOjTlnKOhQ4dCT0UCegJRA02BKIGeQNRsCZrCuLx0bAl6IkKMKhXQE4iaOGmq4pYr+r7fOSvX8TWM3r17Q6xFpNz929DQUNTjd2iqpaWFiMp/vT2dcvcv9NTzKHcfQ1M9i3L3L/TU8yh3HxdTUxiXl55y9y9iVM+i3P0LPfU8yt3HYTQF43kAAAAAAAAAAAAAEHswyQUAAAAAAAAAAAAAYk9FT3Kl02k677zzKJ1Ol7spPZItrX+3tOstNVta/25p11sOtrQ+3tKut9Rsaf27pV1vOdjS+nhLu95Ss6X175Z2vaVmS+vfLe16y0Gc+rjijOcBAAAAAAAAAAAAAOguFf1LLgAAAAAAAAAAAAAAwoBJLgAAAAAAAAAAAAAQezDJBQAAAAAAAAAAAABiDya5AAAAAAAAAAAAAEDswSQXAAAAAAAAAAAAAIg9FTvJddlll9Hw4cOpurqaxo8fT88991y5mxRLpk+fTrvvvjvV19fTgAEDaPLkyTRnzhxWp7W1laZMmUL9+vWjXr160RFHHEHNzc1lanFxgJ6iA5paBzQVDdDTOqCnaICeNgBNRQM0tQ7oKRqgp3VAT9EBTa0DmoqGHqMnV4Hceuutrqqqyl177bXu9ddfd9///vddnz59XHNzc7mbFjsOPPBAd91117nZs2e7l19+2R100EFu6NChbtWqVZ11TjnlFDdkyBA3Y8YM98ILL7jPf/7z7gtf+EIZWx0t0FO0QFPQVJRAT9BTlEBP64CmogOagp6iBHqCnqIGmoKmoqSn6KkiJ7n22GMPN2XKlM5yLpdzgwYNctOnTy9jq3oGn3zyiSMi9+STTzrnnFu+fLlLpVLujjvu6Kzz5ptvOiJyM2fOLFczIwV6Ki7QFDQVJdAT9BQlW6KenIOmismWqCnoqXhAT9BT1EBT0FSUxFVPFbdcsb29nV588UWaOHFi5zbf92nixIk0c+bMMrasZ7BixQoiIurbty8REb344ouUyWRYf48dO5aGDh3aI/obeio+0BQ0FSXQE/QUJVuanoigqWKzpWkKeiou0BP0FDXQFDQVJXHVU8VNci1ZsoRyuRw1Njay7Y2NjbRo0aIytapnEAQBnX766bTnnnvSTjvtREREixYtoqqqKurTpw+r21P6G3oqLtDUBnrK9ZUT6GkDPeX6ysmWqCciaKqYbImagp6KB/S0gZ5yfeUGmtpAT7m+chJnPSXL3QBQOqZMmUKzZ8+mp59+utxNAT0EaApECfQEogR6AlEDTYEogZ5A1EBTIErirKeK+yVX//79KZFIKIf+5uZmampqKlOr4s/UqVPpgQceoMcff5wGDx7cub2pqYna29tp+fLlrH5P6W/oqXhAU9BUlEBP0FOUbKl6IoKmisWWqinoqThAT9BT1EBT0FSUxF1PFTfJVVVVRePGjaMZM2Z0bguCgGbMmEETJkwoY8viiXOOpk6dSvfccw899thjNGLECPb3cePGUSqVYv09Z84c+uijj3pEf0NP0QNNQVNRAj1BT1GypeuJCJqKmi1dU9BTtEBP0FPUQFPQVJT0GD2Vz/N+49x6660unU6766+/3r3xxhvupJNOcn369HGLFi0qd9Nix6mnnuoaGhrcE0884RYuXNj5b82aNZ11TjnlFDd06FD32GOPuRdeeMFNmDDBTZgwoYytjhboKVqgKWgqSqAn6ClKoKd1QFPRAU1BT1ECPUFPUQNNQVNR0lP0VJGTXM4593//939u6NChrqqqyu2xxx7u2WefLXeTYgkRmf+uu+66zjpr1651P/jBD9xWW23lamtr3de+9jW3cOHC8jW6CEBP0QFNrQOaigboaR3QUzRATxuApqIBmloH9BQN0NM6oKfogKbWAU1FQ0/Rk+ecc9H/PgwAAAAAAAAAAAAAgNJRcZ5cAAAAAAAAAAAAAAB0F0xyAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7MMkFAAAAAAAAAAAAAGIPJrkAAAAAAAAAAAAAQOzBJBcAAAAAAAAAAAAAiD2Y5AIAAAAAAAAAAAAAsQeTXAAAAAAAAAAAAAAg9mCSCwAAAAAAAAAAAADEHkxyAQAAAAAAAAAAAIDYg0kuAAAAAAAAAAAAABB7/j/znZ0QZbv3SwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1500x800 with 32 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imagesPerRow = 8\n",
    "columns = batch_size // imagesPerRow\n",
    "\n",
    "fig, ax = plt.subplots(columns, imagesPerRow, figsize=(15,8), sharex=True, sharey=True)\n",
    "ax = ax.flatten()\n",
    "for elem in train_dataset.take(1):\n",
    "    \n",
    "    dataXFromGenerator = elem[0].numpy()\n",
    "    dataYFromGenerator = elem[1].numpy()\n",
    "    for i in range (dataXFromGenerator.shape[0]):\n",
    "        ax[i].imshow(dataXFromGenerator[i].reshape(shape))\n",
    "        ax[i].set_title(f\"Lable {np.argmax(dataYFromGenerator[i])}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " reshape (Reshape)           (None, 28, 28, 1)         0         \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 26, 26, 5)         50        \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 3380)              0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 10)                33810     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 33,860\n",
      "Trainable params: 33,860\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 13:00:00.280414: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n",
      "2024-02-10 13:00:00.280588: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n",
      "2024-02-10 13:00:00.899864: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:424] Loaded cuDNN version 8600\n",
      "2024-02-10 13:00:01.145774: I tensorflow/compiler/xla/service/service.cc:169] XLA service 0x7f2719506280 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2024-02-10 13:00:01.145807: I tensorflow/compiler/xla/service/service.cc:177]   StreamExecutor device (0): Tesla V100-SXM2-32GB, Compute Capability 7.0\n",
      "2024-02-10 13:00:01.151761: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:269] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n",
      "2024-02-10 13:00:01.320053: I ./tensorflow/compiler/jit/device_compiler.h:180] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "59994/60000 [============================>.] - ETA: 0s - loss: 0.0364 - accuracy: 0.9894"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 13:04:52.536409: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n",
      "2024-02-10 13:04:52.536925: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 322s 5ms/step - loss: 0.0364 - accuracy: 0.9894 - val_loss: 0.2079 - val_accuracy: 0.9703\n",
      "Epoch 2/2\n",
      "60000/60000 [==============================] - 320s 5ms/step - loss: 0.0037 - accuracy: 0.9988 - val_loss: 0.2696 - val_accuracy: 0.9707\n"
     ]
    }
   ],
   "source": [
    "########## HYPER PARAMETERS\n",
    "#batch_size = 128 is specified when setting up the generator\n",
    "epochs = 2\n",
    "optimizer = tf.keras.optimizers.Adam()\n",
    "########## HYPER PARAMETERS\n",
    "\n",
    "########## MODEL ARCHITECTURE\n",
    "model = tf.keras.Sequential()\n",
    "model.add(tf.keras.layers.Input(shape=(None, shape[0] *  shape[1] * shape[2]), name=\"input\"))\n",
    "model.add(tf.keras.layers.Reshape((shape[0], shape[1], shape[2])))\n",
    "model.add(tf.keras.layers.Conv2D(5, kernel_size=(3, 3), activation='relu'))\n",
    "model.add(tf.keras.layers.Flatten())\n",
    "#model.add(tf.keras.layers.Dense(5, activation='relu'))\n",
    "model.add(tf.keras.layers.Dense(num_classes, activation='softmax'))\n",
    "########## MODEL ARCHITECTURE\n",
    "\n",
    "# Print summary\n",
    "model.summary()\n",
    "\n",
    "# compile model for training\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer=optimizer,\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "history = model.fit(train_dataset,\n",
    "                    steps_per_epoch=len(x_train),\n",
    "                    batch_size=batch_size,\n",
    "                    epochs=epochs,\n",
    "                    verbose=1,\n",
    "                    validation_data=val_dataset,\n",
    "                    validation_steps=len(x_val))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot training and test accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 2ms/step\n",
      "\n",
      " Summary of the precision, recall, F1 score for each class:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.98      0.98       980\n",
      "           1       0.99      0.99      0.99      1135\n",
      "           2       0.98      0.96      0.97      1032\n",
      "           3       0.95      0.99      0.97      1010\n",
      "           4       0.98      0.97      0.98       982\n",
      "           5       0.99      0.94      0.97       892\n",
      "           6       0.96      0.98      0.97       958\n",
      "           7       0.96      0.97      0.96      1028\n",
      "           8       0.95      0.97      0.96       974\n",
      "           9       0.98      0.95      0.96      1009\n",
      "\n",
      "    accuracy                           0.97     10000\n",
      "   macro avg       0.97      0.97      0.97     10000\n",
      "weighted avg       0.97      0.97      0.97     10000\n",
      "\n",
      "\n",
      " Confusion matrix: \n",
      "[[ 962    0    4    0    1    2    5    2    4    0]\n",
      " [   0 1124    2    1    1    0    3    2    2    0]\n",
      " [   2    7  989   13    2    0    1    5   12    1]\n",
      " [   2    0    3  995    0    4    1    3    2    0]\n",
      " [   0    0    2    1  956    0    6    4    4    9]\n",
      " [   2    1    2   15    0  842   14    2   11    3]\n",
      " [   7    2    1    1    4    1  939    1    2    0]\n",
      " [   0    4    7    6    1    0    0 1002    4    4]\n",
      " [   6    0    3    6    0    1    6    6  943    3]\n",
      " [   4    3    1    6    8    2    1   22    7  955]]\n"
     ]
    }
   ],
   "source": [
    "# use model for inference to get test accuracy\n",
    "y_val_pred = model.predict(x_val)\n",
    "y_val_pred = np.argmax(y_val_pred, axis=1)\n",
    "\n",
    "print ('\\n Summary of the precision, recall, F1 score for each class:')\n",
    "print (sklearn.metrics.classification_report(y_val, y_val_pred))\n",
    "\n",
    "print ('\\n Confusion matrix: ')\n",
    "print (sklearn.metrics.confusion_matrix(y_val, y_val_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f2b74123490>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.array(history.history['val_accuracy'])*100, label=\"Validation Accuracy\")\n",
    "plt.plot(np.array(history.history['accuracy'])*100, label=\"Training Accuracy\")\n",
    "plt.ylabel(\"Accuracy [%]\")\n",
    "plt.xlabel(\"Epochs\")\n",
    "plt.ylim(0,100)\n",
    "plt.xlim(0,)\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 13:10:43.765308: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node input}}]]\n",
      "2024-02-10 13:10:43.813687: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node input}}]]\n",
      "2024-02-10 13:10:43.831245: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:43.849931: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node input}}]]\n",
      "2024-02-10 13:10:43.859475: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:43.878618: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node input}}]]\n",
      "2024-02-10 13:10:43.885407: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:43.915151: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:43.924729: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:43.978288: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node input}}]]\n",
      "2024-02-10 13:10:43.997900: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:44.007519: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "2024-02-10 13:10:44.092426: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'inputs' with dtype float and shape [?,?,784]\n",
      "\t [[{{node inputs}}]]\n",
      "WARNING:absl:Found untraced functions such as _jit_compiled_convolution_op while saving (showing 1 of 1). These functions will not be directly callable after loading.\n",
      "2024-02-10 13:10:44.184460: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'serving_default_input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node serving_default_input}}]]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /tmp/tmp2kqd4i_i/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /tmp/tmp2kqd4i_i/assets\n",
      "2024-02-10 13:10:45.294084: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'serving_default_input' with dtype float and shape [?,?,784]\n",
      "\t [[{{node serving_default_input}}]]\n",
      "2024-02-10 13:10:46.055189: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:364] Ignored output_format.\n",
      "2024-02-10 13:10:46.055223: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:367] Ignored drop_control_dependency.\n",
      "2024-02-10 13:10:46.055682: I tensorflow/cc/saved_model/reader.cc:45] Reading SavedModel from: /tmp/tmp2kqd4i_i\n",
      "2024-02-10 13:10:46.056971: I tensorflow/cc/saved_model/reader.cc:89] Reading meta graph with tags { serve }\n",
      "2024-02-10 13:10:46.056988: I tensorflow/cc/saved_model/reader.cc:130] Reading SavedModel debug info (if present) from: /tmp/tmp2kqd4i_i\n",
      "2024-02-10 13:10:46.062507: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:353] MLIR V1 optimization pass is not enabled\n",
      "2024-02-10 13:10:46.063390: I tensorflow/cc/saved_model/loader.cc:231] Restoring SavedModel bundle.\n",
      "2024-02-10 13:10:46.091740: I tensorflow/cc/saved_model/loader.cc:215] Running initialization op on SavedModel bundle at path: /tmp/tmp2kqd4i_i\n",
      "2024-02-10 13:10:46.100123: I tensorflow/cc/saved_model/loader.cc:314] SavedModel load for tags { serve }; Status: success: OK. Took 44441 microseconds.\n"
     ]
    }
   ],
   "source": [
    "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
    "tflite_model = converter.convert()\n",
    "\n",
    "\n",
    "with open('model.tflite', 'wb') as f:\n",
    "    f.write(tflite_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}