LeNet5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e09477ad-c2dc-4cf4-8901-366282f218d2",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-09-12 00:10:42.799183: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-09-12 00:10:43.100452: E tensorflow/stream_executor/cuda/cuda_blas.cc:2981] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n",
      "2022-09-12 00:10:43.554889: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /opt/intel/oneapi/mpi/latest/lib:/opt/intel/oneapi/compiler/latest/linux/comcompiler/lib/intel64:/opt/intel/oneapi/mkl/latest/lib/intel64::/home/hagi/.local/lib\n",
      "2022-09-12 00:10:43.554948: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /opt/intel/oneapi/mpi/latest/lib:/opt/intel/oneapi/compiler/latest/linux/comcompiler/lib/intel64:/opt/intel/oneapi/mkl/latest/lib/intel64::/home/hagi/.local/lib\n",
      "2022-09-12 00:10:43.554954: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.\n"
     ]
    }
   ],
   "source": [
    "# Load the TensorBoard notebook extension\n",
    "%load_ext tensorboard\n",
    "\n",
    "# Import the necessary packages and modules\n",
    "import datetime, os\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('./sci.mplstyle')\n",
    "from tensorflow.keras import datasets, layers, models, losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "953270d2-da71-4468-990e-4f51cfb14beb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train shape: (60000, 28, 28)\n",
      "Train samples: 60000\n",
      "Test samples: 10000\n",
      "Image shape: (28, 28)\n"
     ]
    }
   ],
   "source": [
    "# Read MNIST dataset\n",
    "(x_train, y_train), (x_test, y_test)=tf.keras.datasets.mnist.load_data()\n",
    "\n",
    "print('Train shape:', x_train.shape)\n",
    "print('Train samples:', x_train.shape[0])\n",
    "print('Test samples:', x_test.shape[0], )\n",
    "print('Image shape:', x_train[0].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a3ea61d3-0988-4540-8e4e-ee495b36c7f3",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-09-12 00:11:10.291123: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-09-12 00:11:10.438551: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudnn.so.8'; dlerror: libcudnn.so.8: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /opt/intel/oneapi/mpi/latest/lib:/opt/intel/oneapi/compiler/latest/linux/comcompiler/lib/intel64:/opt/intel/oneapi/mkl/latest/lib/intel64::/home/hagi/.local/lib\n",
      "2022-09-12 00:11:10.438610: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1934] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.\n",
      "Skipping registering GPU devices...\n",
      "2022-09-12 00:11:10.440432: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "TensorShape([60000, 32, 32])"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add pad to extend size for input matrix\n",
    "x_train = tf.pad(x_train, [[0, 0], [2,2], [2,2]])/255\n",
    "x_test = tf.pad(x_test, [[0, 0], [2,2], [2,2]])/255\n",
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a47b7359-0b1e-4dbf-a313-b112f78b3c8b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorShape([60000, 32, 32, 1])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add dimension for input matrix\n",
    "x_train = tf.expand_dims(x_train, axis=3, name=None)\n",
    "x_test = tf.expand_dims(x_test, axis=3, name=None)\n",
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "11c719bb-f90d-4f11-9db4-839e5e227800",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorShape([58000, 32, 32, 1])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Validation data extraction \n",
    "x_val = x_train[-2000:,:,:,:]\n",
    "y_val = y_train[-2000:]\n",
    "x_train = x_train[:-2000,:,:,:]\n",
    "y_train = y_train[:-2000]\n",
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "bbac0440-24b4-4889-8691-31838487cabc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 28, 28, 6)         156       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 14, 14, 6)        0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " activation (Activation)     (None, 14, 14, 6)         0         \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 10, 10, 16)        2416      \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 5, 5, 16)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " activation_1 (Activation)   (None, 5, 5, 16)          0         \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 1, 1, 120)         48120     \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 120)               0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 84)                10164     \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 10)                850       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 61,706\n",
      "Trainable params: 61,706\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Build the CNN model\n",
    "model = models.Sequential()\n",
    "model.add(layers.Conv2D(6, 5, activation='tanh', input_shape=x_train.shape[1:]))\n",
    "model.add(layers.MaxPool2D(2))\n",
    "model.add(layers.Activation('sigmoid'))\n",
    "model.add(layers.Conv2D(16, 5, activation='tanh'))\n",
    "model.add(layers.MaxPool2D(2))\n",
    "model.add(layers.Activation('sigmoid'))\n",
    "model.add(layers.Conv2D(120, 5, activation='tanh'))\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(84, activation='tanh'))\n",
    "model.add(layers.Dense(10, activation='softmax'))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "cb502732-97cc-435f-a968-3f6d9cb09d95",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "907/907 [==============================] - 16s 17ms/step - loss: 1.6318 - accuracy: 0.3951 - val_loss: 0.3101 - val_accuracy: 0.9125\n",
      "Epoch 2/20\n",
      "907/907 [==============================] - 21s 23ms/step - loss: 0.2744 - accuracy: 0.9147 - val_loss: 0.1495 - val_accuracy: 0.9595\n",
      "Epoch 3/20\n",
      "907/907 [==============================] - 16s 18ms/step - loss: 0.1897 - accuracy: 0.9393 - val_loss: 0.1488 - val_accuracy: 0.9565\n",
      "Epoch 4/20\n",
      "907/907 [==============================] - 17s 18ms/step - loss: 0.1474 - accuracy: 0.9540 - val_loss: 0.1011 - val_accuracy: 0.9735\n",
      "Epoch 5/20\n",
      "907/907 [==============================] - 17s 19ms/step - loss: 0.1303 - accuracy: 0.9592 - val_loss: 0.0879 - val_accuracy: 0.9755\n",
      "Epoch 6/20\n",
      "907/907 [==============================] - 17s 19ms/step - loss: 0.1177 - accuracy: 0.9630 - val_loss: 0.0817 - val_accuracy: 0.9765\n",
      "Epoch 7/20\n",
      "907/907 [==============================] - 17s 18ms/step - loss: 0.1078 - accuracy: 0.9650 - val_loss: 0.0988 - val_accuracy: 0.9775\n",
      "Epoch 8/20\n",
      "907/907 [==============================] - 17s 19ms/step - loss: 0.1016 - accuracy: 0.9676 - val_loss: 0.0902 - val_accuracy: 0.9695\n",
      "Epoch 9/20\n",
      "907/907 [==============================] - 18s 20ms/step - loss: 0.0924 - accuracy: 0.9709 - val_loss: 0.0864 - val_accuracy: 0.9780\n",
      "Epoch 10/20\n",
      "907/907 [==============================] - 16s 18ms/step - loss: 0.0895 - accuracy: 0.9717 - val_loss: 0.0842 - val_accuracy: 0.9795\n",
      "Epoch 11/20\n",
      "907/907 [==============================] - 17s 18ms/step - loss: 0.0845 - accuracy: 0.9729 - val_loss: 0.0722 - val_accuracy: 0.9775\n",
      "Epoch 12/20\n",
      "907/907 [==============================] - 15s 17ms/step - loss: 0.0810 - accuracy: 0.9745 - val_loss: 0.0724 - val_accuracy: 0.9800\n",
      "Epoch 13/20\n",
      "907/907 [==============================] - 17s 19ms/step - loss: 0.0778 - accuracy: 0.9750 - val_loss: 0.0713 - val_accuracy: 0.9835\n",
      "Epoch 14/20\n",
      "907/907 [==============================] - 19s 21ms/step - loss: 0.0748 - accuracy: 0.9760 - val_loss: 0.0621 - val_accuracy: 0.9835\n",
      "Epoch 15/20\n",
      "907/907 [==============================] - 16s 18ms/step - loss: 0.0712 - accuracy: 0.9769 - val_loss: 0.0562 - val_accuracy: 0.9830\n",
      "Epoch 16/20\n",
      "907/907 [==============================] - 16s 17ms/step - loss: 0.0691 - accuracy: 0.9778 - val_loss: 0.0817 - val_accuracy: 0.9820\n",
      "Epoch 17/20\n",
      "907/907 [==============================] - 20s 21ms/step - loss: 0.0633 - accuracy: 0.9795 - val_loss: 0.0717 - val_accuracy: 0.9830\n",
      "Epoch 18/20\n",
      "907/907 [==============================] - 20s 22ms/step - loss: 0.0622 - accuracy: 0.9804 - val_loss: 0.0597 - val_accuracy: 0.9865\n",
      "Epoch 19/20\n",
      "907/907 [==============================] - 19s 21ms/step - loss: 0.0632 - accuracy: 0.9792 - val_loss: 0.0514 - val_accuracy: 0.9890\n",
      "Epoch 20/20\n",
      "907/907 [==============================] - 19s 21ms/step - loss: 0.0595 - accuracy: 0.9809 - val_loss: 0.0610 - val_accuracy: 0.9820\n"
     ]
    }
   ],
   "source": [
    "# Configures the CNN model for training\n",
    "model.compile(optimizer='adam', loss=losses.sparse_categorical_crossentropy, metrics=['accuracy'])\n",
    "# Store log for TensorBoard\n",
    "logdir = os.path.join(\"logs\", datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\"))\n",
    "tensorboard_callback = tf.keras.callbacks.TensorBoard(logdir, histogram_freq=1)\n",
    "# Training the CNN model with the given inputs\n",
    "history = model.fit(x=x_train, y=y_train, batch_size=64, epochs=20, validation_data=(x_val, y_val), callbacks=[tensorboard_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "089221b9-d339-4618-b853-63dd5d128240",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhkAAAMHCAYAAACUsUXUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOydB3wU1fbHT3qlJITeUURAVLChgr2h2LA+e3vqs+uzl/fs9Vmxd1H/drE8FX2KvVdEUURpIjWBkJCQnv/vzN5ZJstusrvZMpP9fT+fH3d25s7M3dkN9+y5556b1tzcLIQQQgghsSY91hckhBBCCKGRQQghhJC4QU8GIYQQQmhkEEIIIcQ70JNBCCGEkLiQGZerpihpaWmcqkMIISTlaG5uTgu2n54MQgghhMQFejISZ+VF4xmJ+txEX9cr14zXdVO9rXz//Pxj/Z2K1/fKK9eM13Xbc0373EigJ4MQQgghcYGejDgQa2uWEEIIcWPf1pZ3g54MQgghhMQFGhmEEEIIiQtpdO3HfgprrJ5pvIKJSOzhZ+Ud+Fl5B35WnvqMOIWVEEIIIYmDwyWEEEIIiQucXeJiOEziHfhZeQd+Vt6Bn5X3oZERB4JN6eEfCyGEEC/DZFyEEEIIcQ30ZMQBei0IIYR0NJiMq4Px7YKV8uxXC2VhWXWym0IIIYREDGeXuJgnPlsgF788U75buCrZTSGEEEIihkaGiykuyLbK0jW1SW4JIYQQEjk0MlxMSaHPyCirqktySwghhJDIoZHhYroV5lhlGT0ZhBDSgiuvvNKaUhmOPvjgg3Y/Pb2GXuvGG2/kJxEBnF3iYrqZ4ZKyNfRkEEKIk0mTJsmGG27YYt/RRx8tG2+8sVx22WUt9g8fPrzdD0+v8eSTT8ro0aP5QUQAF0hz8QJpOrvkoPs+l837d5VXTt8+JtckhJCOinoadtxxx5h4Lkj4z7y1BdLoyXBxxs9uBWa4pIqBn4QQQpILM352MLrZgZ8cLiGEkKiZP3++1UHecMMNsv/++0tOTo7069dP/vrrL+v4s88+KzvvvLMUFRVJVlaW9OrVS4444giZN29eqzEZ+vr888+3hlE23XRTyc3Nlf79+8tFF10ktbX8cajQkxEHYjVcUpiTKdmZ6VJd1wg1SH42Py5CCImW66+/XrbaaiuZPHmyLFq0SPr27Su33HKLXHjhhbLPPvvIddddZxkOH3/8sWV4/PzzzzJjxoxWr/nSSy/JI488IqeffroMGDBAnnnmGbn55pslPT3dMmpSPeMney0Xox9eCYI/F6+usbwZ+cX8uAgh4TPo4jdc/bjm37hPQu+XnZ0tL7/8snTt2tV63djYKDfddJOMHz9eXn/9dX+H+Y9//EMaGhrkhRdekFmzZsmIESNCXnPhwoWWIbLJJptYr4877jjLmzFlypQOZ2REA6ewupxiM2SykrkyCCGkXWy//fZ+A0PJyMiwhkxee+21Fr/Iy8vLpaCgwNpes2ZNq9dU48I2MGxDRo2SZcuW8dMC/Gnschj8SQjxiqfA7fTs2XO9fRqf8eGHH1pei9mzZ1txGDqUYtPU1BTVNdVLQujJ8EzwZymDPwkhpF2o5yKQE088Ufbcc0/5+uuvrVwYGsipMRmXXHJJWNfU2AsSGnoyXE6JP+snE3IRQkgs+eSTT+TRRx+V448/3iqdaEwFaT80wTyT9ZPToQghJJaUlpZa5ahRo1rs1yETnTWiaAAoiR56MlycjMu5EisDPwkhJPaBoMXFxdY6KKtWrZKBAwfKL7/8Yk1JrampseqsXr2aj93AZFwdeLiklLNLCCEkpnTv3l3efvttf+6Mc889V6ZNmyannXaaFZehvPPOO3zqqbR2CSyp41A8Bo1H2z8J8xz12Og8JF+PvT5/4Vr9As7JQ3EOdBQ0GCqHdNL5v1B3SSLWLlF+XFQu+939qYzs01neOGt8zK5LCCGEtJcOtXYJ3sy2KCZHceoIY2D8AX0R5PjKgPtkoXgV2h36E3oT2hg6CZqI49vggS6Moh3tWO6dgZ+EEEK8hWeMDHTsk1A8DhVGcbq9Nu9jMA6uC6P+mcbAUM/FQTin1rRBz70UugfaN4p2RB/4WVVreUiiGRMjhBBCkoHrZ5egU+0H6VwiDfXVSc7L2mFkfBvG/bQXPw/SMY8zbAPDcAU023gzdAgl7uRmZUhBdobUNzZLZS2jnAkhhHgH1xsZ4FroaOgbaCz0azuMjO/CqKtzmfpCM2BgzHcewGtN/faaeTkhinZEBYdMCCGEeBEvGBlqVBwLaRzEzEhPNp6JzaGl0H54+RVUCa2AnoGGBZwy0pQ/tdIepeXE6oQs+c5cGYQQQryD640MGBY3QlOMFyEahkCdoV7QA5BOfn7flIdDX8PQ2N5Rv7cpg84gcexfP2F9nNcvYWpxQgghXsL1RkYMsIdK/oK2gLGyA7QftjWm4laoE/QcDI1cU8+39J5IdYjrrTVlyABUdZ60pkgpKVwX/EkIIYQkivb2Z6lgZGjA6ABoaxgXP9g7sa1RlBeaYFCNwTjAHLI9JqGSXdhPNWHTPPxZPzmNlRBCiIfo8EYGjAnlT2hxkGNNJgeGsoUpNWmXosm4gmF7PKpauWerijrwk1k/CSGEJJD29mcd3sgIAw0IVfJNaRsjGsMRjLZiNmKOPVxSysBPQgghHqLDGxkYMzod0piL3UJUsfNdLAqYVaJZQoMx3JQRz3Rpb+Ans34SQgjxEh3eyDCzSw4102BbYII9DzEvrVVw4P75BcUCaAyO9w+or89Lg0bVR/R2HNscfAorAz8JIYR4iA5lZMAIGABtDJU4dj8CNUJHYv9BAeuT6DooA6G3YFw4s4HeZ7KLPoJ69mwT5RpoI2gq6us6KJLI1OJc7p0QQoiX6FBGBtD04+qJOMPeAWNglkkTrrxoknG9iO25ZsEzTa6lK7s6uR36xKxfMkfrQzPNuiULnddPBEUOI6OxyVur5hJCSKzZbbfdrOmT337b+koRV111lVXv+eefD/vaRx11lHVOaWmp9frhhx+2Xr/4onYbrdOvXz/ZZJNNwr5XIL///nurbfEiHc3ICAoMjbuMwaBDHEOhiSYPhi54thWOLw+or0ue7mk8F9WmvubFeBAaG2qp93iRlZEuXfOzRO2L8mquxkoISW2OP/54q3zmmWdarff0009LUVGR7L///lHfa+edd5Ynn3xStt5666ivEQ4333yzbLrppi32nXbaada9O3XSdE7exDOrsNqgg98pymPvoXgvgvuocfEvo4gIlqAkmqmrgUMm5dX11jRWe0orIYSkIpMmTZLOnTvLc889J7fcckvQ/3O//PJLmTNnjpx++umSkxP9/5kbbLCBpXjzzjvvSF1dyx+R2223nSW3EE0yyZTwZHQEbMOC01gJIalOXl6eHHbYYbJo0SL5+OOPg9Z56qmnrPKEE05IZNNIADQy4kAsEnAFwuBPQghZx3HH+ULpnn322fUeS0NDg+Xl0OGHMWPGWPtWrlwpF110kYwYMULy8/MtQ2XkyJFy4403SlNT6KWxgsVk1NfXW/EeQ4YMsa6z1VZbyXvvBXeUf/HFF3LwwQdL7969JTs72xq+2X333eWjjz5qEcuh5zc2Nlr30piTUDEZlZWV1vvQe+v19LpqSP35558t7qvXPPzww+WNN96whnq0nT179pR//OMfsnr16oT1bZ4bLklV1q3EypgMQgjRYYRhw4bJCy+8IHfddZdkZq7rzqZNmyYrVqyQyy67zHqtwxA77rijzJ8/34pz2GijjaSsrEwef/xxueSSSyyj46yzzgr7oR500EHy+uuvyyGHHCI77bSTFYC69957WwZB165d/fU+giGhBsPw4cPlggsusIZ4fv75Z3nwwQdlzz33tNqjHf/kyZPliiuukF9++UWeeOIJy3AIRkVFhYwfP15mzpwpRx99tIwdO1Zmz54t999/v2VMfPbZZy2GdvT1f//7Xzn55JPl73//u1VH665Zs8aK9XBFylAq/Gegj9P3SGPPbe/Mbh540X+bb33717hcnxDSMZk1bOOQWvnsc/56ut1aXSdzD5wUst7iy6/w16ue+VNY14uW66+/3vo/9623NAvBOjCU0oxf+c0wNKzXMESsejAqWtSDodEM46R522239e878sgjrbr2uQ899JD1Wq+h6L309fnnn9/iWvfcc4+1H94R/7599tmnuUuXLs3l5eUt6t55551W3QceeMC/b9ddd23OyMhoUS+wLTCIrNd33313i3rTp0+39k+YMMG/r2/fvta+t99+u0XdTTbZxHo2NTU1LfZHi6PfC9ovcrjEI/hTi3P9EkIIsTjmmGMEHXOLWSY6nPDaa6/JvvvuKyUlvpRJOlyhQw7otFs8OfVmqHdBf9mHy6uvvmqVZ5zRMpOBegtgUKxX9/fff2+xv7a2VtLTfV1vJPdVXnrpJenRo4eceuqp682AUWnwqHo7bIqLi2WPPfZoUXfzzTe3PDs6fJQIOFziEfyLpHH9EkJIBAz/VVMHtU3RYYdaCofBL+vi1m2Tt8nIsO8fDfi1bsU3vPLKK4Jf5pKbm2vFTqxdu9Y/zdUmKytL4G2wAkX/+OMPS2qQKKGGJ4Ixd+5c61oDB2oex3XocM1GGIaprtaJiT4yYAAtXbpUbrjhBmuIY968edYQicaMKK3FggRDz9f4Cr1uIBpf8v7778vChQv9uTp0KCYQe6aNxn8kAnoyPIJ/uXd6MgghpEUAqP56f/PNN/2zStRo2Guvvfx1tOPVTlgDJtUA0V/3anCowRCJgaFo3IV20Br8GUig0fDoo4/KqFGjrGRgvXr1sgI01buhHolYTyO1jQbndF3bY5JM6Mnw2HAJAz8JIWQdBxxwgDVjQzvy7bffXj744AMryNL5a/9f//qXNd31008/bZF3Qg0FHTbQYYVw0cBKxDnIb7/9ZhkuTgNDPQ220aIejTMwpLLZZptZAZgaXGozZYomp46cQYMGWfdVgyLQmzFr1izLm9KnT5+orh0vkm/mdEDU0gxUrFZiZZ4MQghZh/5y16ma6snQ6aza2QcOlWg8hv6q1+mrTnRWh8ZI2MMX4aAzSpTrrtOE0et47LHHWsQ5qJGxFl4TNUqcBobuu/fee61t533VaGhr+OTAAw+U5cuXWzNEnHz44YeW1ENTUOBcbiv5fRs9GR6hS16WZKSnSUVNg9Q1NEl2Ju1DQghR1Ki47777rNwV9tRWJxoEqtM3tRM+9thjrc5Rp7nqNFSN44gkb4ROWdVr6FTT8vJymThxouVF0Hwa6lGx0aDTsWPHytSpU61ATc2lsWTJEus89aoozvtq/ITOxrj88sut8/S6gVx66aVWUOuZZ54pX3/9tVVPPRv63rt16yZ33HGH674Q7KniQLBpPO0lHQYG4zIIIWR9tAPXoYtVq1at58VQTjnlFLntttusTv3888+Xa665xgr61PgIzZuhwZkamBkuGmtx6623WsGj5513njVEo14UDfx08hJiL3RGi95Hh07U2zFu3DjLKOnevbs1G8Tm4osvltGjR1trmOh2MHSWig69nHPOOdY9NbeHDhPpLJsffvhBhg7Vpbnc1belxaIDJOZhpqX5kmXE6ZnudcdH8uvSSnnjrHEysk/LqVKEEEJIorGHTNDvBR07oSfDQzDrJyGEEC9BI8ND2MGfZVW1SW4JIYQQ0jY0MjwEPRmEEEK8BI0MD1HiX+6di6QRQghxPzQyPMS62SUcLiGEEOJ+aGR4iG7GyGDWT0IIIV6AybjiQLAsaLGY1movksaVWAkhhCSaaLJX05PhyfVLOFxCCCHE/dCTEQfilYxr3XLvDPwkhBCS/L6tLe8GPRkeoiA7w1qzZG19o1TXhb+gDyGEEJIMaGR4CLUYSxj8SQghxCPQyPAY/iGTKg6ZEEIIcTc0Mjyb9ZPBn4QQQtwNjQyvrl/C4E9CCCEuh0aGVz0ZHC4hhBDicjiF1UPJuFpm/eRwCSGEkMTBZFwpAAM/CSGEeIV0D1pSx0HN0LgIzxsLTYWWQXXQEuhpaHiI+u+Z+4TShqHupV6LQMV6uKSUngxCCCEJJJq+zVPDJejYt0UxOYrzjkTxBJQBfQ99Bo2AjoAOwPEJeFgfBZy2ObQGejXEZSsjbUcsKGHgJyGEEI/gGSMDhsAkFI9DhRGeV4LiPt2EDoQx8YrZr68vhq6HnlTPBI7Vm2MDURRD72LfUbF7F+2n2HgyVjLwkxBCiMtx/XAJOvx+0BRsvmQ8EcsivMRBUCfoKdvAULCt3GA8GwOg7RznjDblt1E3PE74Az+rauO2RgohhBCSEkYGuBY6GvoGGgv9GsV7VENieojjc0zZxwtGRm5WhhTmZEp9Y7NU1HD9EkIIIe7FC0aGGhXHQtvgl/vMSE/GOfdBYyCNyWgBPCT6/rcwLxcFMTKKUWcatAKqhN6H9oy0DbGGWT8JIYR4AdcbGTAOboSmQE1xuPxx0AbQYujzIEbG/VBfSINC50M7QWp0nB2HtkQxZML1SwghhLgX1xsZ8QKGgnow7jQvL4YR0+AIFO0H6esjsH8UdJCWeH2Y2X8r6m3eyrVbVXsp5gwTQgghCaC9/VlKGhl4MFuheMfMVLkfBsST9jFsl6LoDo3A9jPO8/D6eRR3mwDUUxPX4paU+FOLM+snIYQQ95KeggbG3iYIVKeoPgadHlhHDQ3IDggN5HVT2rEcYSUsiWVirnUxGRwuIYQQEj/a25+lp5iBcQqK14wH4zboxChiPZaaMj+WbYtuJVZ6MgghhLiX9BQyMP5lAjl1EOlcGBf/1EQZQertBmlyrnNCXGpwkNkoCcWfWpyBn4QQQlxMShgZMBjORHEVVG+COe9opXoepFk+z8Z5wTKiHmPKt2Pbysg9GSs5XEIIIcTFdCgjA0bBAGhjM0PE3qezQm41L4+BgfFcG5dR42EBNAi6GednOK51PIpDoSXQIzFtfDQxGQz8JIQQ4mI8s3ZJmGj68R2N1+JKs+8yKAsqhybCUJgY4twHYIB8DOkKrUcYY+NcaD+8noFyiGPRNJ3SujqO76NVGPhJCCHEC3Q0IyMYmkBL6QrpaqyheBf6WDdgQHwGw2IMNi+Hdof2hVaYBdquxvF58Wtu2xTnm0XSquuksalZMtLbn3uDEEIIkVQ3MtDB20ZDWMewr1eU95lj0plHTLAEJbFczCwzI1265mdJeXW9rIKhUVLoi9EghBBC4kU0ySQ7VExGKmGnFueS74QQQtwKjYw4EI8EXIF0M96LUubKIIQQkgCYjCuF8KcW5zRWQgghLoWeDI/CrJ+EEELcDo0Mj1LM5d4JIYS4HBoZHmXdSqxcJI0QQog7oZHhUezATy6SRgghxK3QyPD4FFYGfhJCCHErNDK87sngcAkhhBCX4rmMn14g3hk/nZ4M5skghBCSCJjxM4XokpdlrVlSWdMgdQ1NyW4OIYQQsh4cLvFoxs90GBj2NFamFieEEBJvmPEzxeCQCSGEEDdDT4aHsVdfZfAnIYQQN0IjoyNk/eQiaYQQQlwIjQwP081k/WRMBiGEEDdCI6MDDJeUciVWQgghLoRGRofI+lmb5JYQQggh68NkXB5NxqUw6ychhJBEwWRcKQYDPwkhhLgZejLiQDy8FsHgcu+EEEKS2be15d1gTEaHWO69LsktIYQQQtaHRoaHKcjOkJzMdFlb3yjVdQ3Jbg4hhBDSAhoZHkbdVP6sn/RmEEIIcRk0MjpI8CeXfCeEEOI2aGR4HGb9JIQQ4lZoZHicbgUcLiGEEOJOOIXVw8m4nNNYS6uY9ZMQQkj8YDKuFB4uYeAnIYQQt0FPhoeTcSnF/uESejIIIYTEj5RIxoU3dBzUDI2L8Lw+0APQXGgtNBu6AsoJUT8PugT6GaqGFkMPQb1j805i7MmoYkIuQggh7sJTRgY6+G1RTI7ivH4ovoROhsqhN6DO0NXQNBzPCqivr1+Froc6QW9CK6GToO9wfEA73kZMKWHgJyGEEJfiGSMDHfskFG9DhVGcfi+khsYVcPeMgQ7G9obQu9BO0FkB9c+EdjfGyFCtD21ijI5e0D3RvYt4ejI4XEIIIcRduN7IUC8ENAWbL0EZ0LIIzx+GYiL0hzESLGA0VKE4EWo0RoVdXweYztMq0Bmo5+y9r4Bm6/VQbXBUbyhuK7HWJTQWhBBCCPG8kQGuhY6GvoHGQr9GeP6ekBoOr6MTbnIewOuFKL6DBsJoGGF2j4L6QjNwfH5AfT3/NfNyQoTtiAu5WRlSmJMpDU3NUrGW65cQQghxD14wMtSoOBbaBp38zCjOH2nKn1q5vm1cRFM/6XDIhBBCiBtx/RRWGBY3tvMS9myQJSGO2/t7Rlk/4ik9sR7W6IYhkwVl1dYMkyHdY3ppQgghKUxaG/1ZR/BktJcCU1aHOL7WlHZAaaT1k043/0qsDP4khBDiHlzvyYgBdhxGKPeBbabZZaT11yPRAZj+1OJc7p0QQkgMaas/a8vTkQqejDWmzAtxPNeUOtskmvqummFCCCGEuIVUMDIWm1LzWwQjMAYj0vquWYl1JXNlEEIIcRGpYGTYs0TsKaqBDDelPXMl0vqumV1SytTihBBCXEQqGBnTTLkfxo5avF+THnw0tADjTrN0H8pf9DU0Bsf7B9TX8/fTaib7qCsoYeAnIYQQF9KhjAw1GqCNoRJ7H4yGecbQGGbWKrHr6iySh00W0VsDLnWf2f+IqWdzDbQRNBXX1QyiroDLvRNCCOkQRgY6XTdnYtD04+qJOCNg/+nQUugytH8m9CK255j1Sd4yRoWT26FPzPE5Wl/Pw/al0MIg13dH4CeHSwghhHjck/EhOlztgD0DvA5zUWwNPQ6pkbQPtAq6BJqE4y3yceN1nUlHfo3JlzHR5MV4EBqL464J+lSK831GxqrqOmls4volhBBC3EFapDkdYGBoMqp3cN7+8WmSd8GzaU5W7ozRV78DI6Nevrl8N3+MBiGEEBIrWsuJgT4uLVaejEVQi4BI4qasn8yVQQghxB1EY2ScAm0Ii+Z5aDzUOdaN8jrqtQhUvNH1SxSmFieEEOKWvi2atOK6YFkFdJCRulA0FXewu6ENzfTdJwDmyiCEEOI2ojEytgyyT6d7Ejdk/eQiaYQQQrxqZMAz0aFya3QU/LkyOI2VEEKIS6DB0MECP7kSKyGEEM8v9W6Scp0M7WgWE6uFlkEfQE/B46HJr0iCKGHgJyGEkI5gZMDA0ERVz0Bd9GXA4QnQJahzFAwNzaZJEgCzfhJCCOkIacV1DZCXjIHxBLQXNMysWqqZMZ+CukLPoe6GsWuqtxKWBCpRwyUrGZNBCCHEJX1bNJ4MXb8jDzoRngpN0+3kV+hN3FiHTB6BzodOjeIeJEJK7OXeObuEEEKIhwM/d4V+DGJg+MGxx1DMgPaItmFeJhnJuDrnZklmeppU1jRIbUNj3O9HCCEktWiOom+LxsjQgM/ZYdT7DeodxfVJFKTDwCgywZ8cMiGEEOIGojEySqGNwqg31Kx0ShLEutTiXL+EEEKIN42M6dBmiLv4W6gKOHYEis1NXZIg7NVXmZCLEEKIG4gm8PMG6GBoii6QhvIFaD6kgzNDoEOgk0zeDF3nhCQ66yeDPwkhhHg0rfgs48V42swc0VVZneicliroGNT9qf1NJJGuX8LhEkIIIZ5NxgXj4RUYGhsYA0O9GX2McbEY+gh6CHV0myRlJVZ1IhFCCCEeMzJgXByD4gcYET+ivCr2TfI+wRKUhDPVp70w8JMQQki8iCaxZDSejJuhNVBKZvN0M8z6SQghxE1EY2R0hj6JdUM6EonwWgSDgZ+EEEIS2be15d2IZgrru9COuLDGYRAXUWICP7ncOyGEEK96Mm6BHoJ+gqEx1aQPXwk1hbB8/i/65pFIKLansCLwUy3ORCzMRgghhMTSyPgQUp+J9mDHm+3WoJGRIAqyMyQnM11q6pukuq5RCnKimjxECCGExIRoeqEpUHKCDkirqOdCs37+Vb7WWr+ERgYhhBCvGRkXwBW/IuYtITEL/lQjQ5d871+cz6dKCCEkaUQT+PkBfjFzdolLYa4MQgghXjYydH2Sslg3pKMNWwQqURTbqcWZ9ZMQQkiS+7ZojIxFUP8oziMJoMROLc7l3gkhhCSZaIwMXa9kQ1gwz+sqrJAm5yIOdPpooBKdkEsDPwkhhJBk9m3RGBm6fHsFdBD0AbQKhkY9VBdEMVmpC9fZDZoOlUIV0PvQnmGe+zjUHIYeDzjvvTbqb+julVi5SBohhBDvzS7ZMsi+jPY2JBTozI9D8RikveZ0c6+doWk4dgosqQfbuMRnrbxPvdahxtj6PuDY5maNlldDnFvZduuTmFqcngxCCCFeMzLQqUfj/YgKGBG9UdwPrYbG4d4/mf1bmfTmd2L7Dez/K9Q1jBES1BDBuVej0PfzMurd6dg/EEWx3gP7j4rV+0mkJ4MxGYQQQpJNmwYDOtwBkHa4EYFzDoCuiK5Zfs6EtNe83TYwFGx/bVaDzYVOjubCaNs4FJdBy6C/Bxwebcpvo7l2MuEiaYQQQtxCOF6JedDtITrq26AjQpynMRtXRtsww16mfCXIMV03RZkQ6UXRZn3f90BangejRdde6RBGRnHBusDPZK0GSwghhIRrZOhE2FCTYc9xGAIxBYaA3nOEWXjtlyBVfjPHRpq6kaCei02hL0Ms4GYbGcW4tMZ+rIAqIwk4TRa5WRnSKSdTGpqapWJtQ7KbQwghJIVJWHxFFBSZoZIyGALrzcfEPu1BSyHNnd0p3IvCSNA4lMvNy6tCVLONDI0H6Qt9BM2HdoLU6Dg73Pslc8iklAm5CCGEJBE3GxkFpqxupc5aUxZGcN3DoH7QDBgqbwUehAFRYo6rEXME6oyCDtLSnKv7b0U9nX0Sdla0RGb/tIdMypiQixBCSDtob3/mZiNDh0KU1gIL7HcYSc9teyFuCXYQxoR6R7pDI7D9TMCx51Hcbaa+nhrBPRNKt0LmyiCEEJJ83GxkaI4KJa+VOjq7RKkK54Kwugaj2MrUtwNHgxoa0JwQh1835RatnN+qEpVanLkyCCGEtIf29mduNjIqjKFRYuIoWmD26dBGDd5oeZjXnGTK13FOa5u2exoAACAASURBVMMwrbHUlK5dR31d1k+mFieEEJI8XGtkNPtMpFlmaGKjIFWGmfbPjOCye5vy5TZSmD8J6cyZYKg3xF4oTtyd9ZOpxQkhhCQP1xoZhmmmPCDIMXvfm+FcyExztVOif9pKVR2e0SyfZwfzoIBjTPl2OPdNBgz8JIQQ4qW04pq9c26Q/eptCHVMhzLai65ZciF0Ee7xNpwbVnIsbG9p9uvsknud2UnNMEapCeAM9HzoirF/4djiVu6pxsMCaBB0M655Aeo3musfb9Y6WQI9EoP3FxdKTOBnKRdJI4QQ4gEjo7CVaaKtHWtXlCM69/no2P9psnN+riujolSPxC6m7cegznLHKVOgHU3+iytDDHPMbeOedSaLqRob50L74fUMlEMci6bplFZdT8WVcLl3QgghXjEydMXTpIHO/F508guN52I8pIEGn0DX4ZgaHeGi01LDiqXAdT/DPceYpF27Q/tCKyBdDv5qHNdU6+L6wE+uxEoIISSJpHF9ixg+zLS0kJ6bRD7nhsYmGXq5L8/YnGsnSGaG20NvCCGEuJ3Wkm+hjwt6kL1PB0SNiq55WfjQRVZV1ye7OYQQQlIUGhlxIBkJuEJm/eQ0VkIIITGgoyXjIu2gm73kOxNyEUIISRI0Mjoo/mmsDP4khBCSJGhkdFD8WT+ZK4MQQkiSoJHRQWHWT0IIIcmGRkYHhYGfhBBCkg2NjA5KiQn85EqshBBC3J5WnLQzYUmip7Gu82RwuXdCCCHxTcYVCnoyOigM/CSEEJJs6MmIA25I1W7nyeBwCSGEkHj1bW15N+jJ6KB0zs2SzPQ0qaxtkNoGa6V6QgghJKHQyOigpMPAsKexrmRcBiGEkGT0RUm4J0l08CdTixNCCEkCNDI6MCUm62cps34SQghJAjQyOjDM+kkIISSZ0MjowHQr4HLvhBBCkgensHbQZFwtcmUw8JMQQkg7YTIuEjQmg4GfhBBCkgE9GR00GVeL4RIGfhJCCGknTMZFWlDM4RJCCCFJhIGfHZgSvyeDi6QRQghJPDQyOjDrAj9rXTOEQwghJHWgkdGByc/OkNysdKmpb5LqOq5fQgghJLHQyOjg043WBX9yyIQQQkhioZGRIkMmpRgyIYQQQhIJjYwOTjezEis9GYQQQhIN82R04IyfzpVYV9KTQQghpB0w4ycJPVzCmAxCCCEJJt0j1tNu0HSoFKqA3of2jOD8TKgGag6hRUHOyYMugX6GqqHF0ENQ77bup16LQCULDpcQQgiJBdH0ba4fLkGnfhyKxyCNXJwOZUA7Q9Nw7BS8yQfDuMwISMcN/oC+CHJ8ZcA9s1C8Cu0O/Qm9CW0MnQRNxPFtcN+F0b2jxMKVWAkhhCQLVxsZxmtwP7QaGoeO/SezfysU70J3YvsN7P+rjUuNNuVjqHtdGLc+0xgYb0AH4RxragbupedeCt0D7Rvp+0lqQi4OlxBCCEkwbh8uOdN4IG63DQwF21+juBnKhU4O4zq2kfFtWxXTfFGb5+ltoDNsA8NwBTTbeDMGh/UOkkyJCfzkcu+EEEISjduNjL1M+UqQY1NNOSECI+O7MOqOgvpCM2BgzHcewOsmFK9FcF8XeTKYJ4MQQkhica2RYTwKGkuhHfsvQar8Zo6NNHVbu87m0FJoP7z8CqqEVkDPQMMCThlpSr/nJIBfHcaI6yk2eTJWVtVJUxPXLyGEEJI4XGtkgCJIff1l8CCslxMb+xpQlEL5UKdWrjME6gz1gh6AaqD3TXk49DUMje0d9e3ZI0tCXM/e3zO8t5FccjIzpFNOpjTAwKioqU92cwghhKQQbjYyCkxZ3UqdtaYsDGOoRINDt4BxsgO0H7Y1puJWY6A8B0MjN8z7tnlPdZ60pkTDXBmEEEKiob39mZuNDB0KUVrz8dvvsLV3+hI0ANoaxsUPAZ6QC00wqMZgHBDmfcO5p6tYl/WTi6QRQghJHG42MtaYMq+VOrb3oSpUBRgTyp/Q4iDHmkwODGWLMO8bzj1bVfIScjH4kxBCSPi0tz9zs5FRYTr8Es3YGXjQ7CuBavBGy9txHw0IFRPbodjGiMZwBKOtmA0Xr8RKTwYhhJDE4VojQ90PKGaZDJ8bBakyzLR/ZmvXgTFyOqQxF7uFqGLnu1gUMKtEZ7YEY7gpW72vK7N+0pNBCCEkgbjWyDBMM6UdL+HE3mcPd7Q2u+RQ6NjAAybY8xDz8h39B7aNTpddAI3B8f4B9fV5adCoGkBvh9F+V8Csn4QQQpKB242Mx8xU04vQwW/h6Oy3NEGbOtPjXsf+AdDGkA6j2DwCNUJHYv9BAeuTTIYGQm/BuHBmA73PeFAeQT17tolyjfGqTEV9XQfFEzDwkxBCSDJwtZFhMm7+0+S5+Bwd/luQejc+M1NPT0ad5Y5TpkDqiTjDcY1ZJk248qJJxvUitueaBc80uZYuwubkdugTs37JHK0PzTTrlix0Xt8LlJjAz1IOlxBCCEkgrjYyjJFwr1mMTFdPHQ9tZRsAOPZUmNe4S+ubIY6h0ESTB0MXPNsqwFDR+hohuafxXFSb+poXQ1d8HYvjngn6VIrt1OIM/CSEEJJA0pIxpbKjAm9HyIeZzOe8orJWtrruXSnKz5Lv/7VH0tpBCCHEu7SWfAt9XJonPRmk/ahxod+NVdX10tBo5xojhBBC4guNjDjghgRcTjIz0mFo+IZM1NAghBBCIqWjJeMi8cj6WcWsn4QQQhIDjYwUwV7yvWwNs34SQghJDDQyUoQSs0gap7ESQghJFDQyUgRm/SSEEJJoaGSkCPb6JVzunRBCSKKgkZFqngwGfhJCCEkQ6y2hTuKTsCTZ01jt2SWlDPwkhBAS42RcoaAnI0WwF0njcu+EEEISBT0ZcSDZXovWh0s4hZUQQkhs+ra2vBv0ZKQIJXbgJ4dLCCGEJAgaGSlC57xMyUxPk8raBqmpb0x2cwghhKQANDJSBHVp2Vk/OY2VEEJIIqCRkZLBn4zLIIQQEn9oZKQQJSb4s5S5MgghhCQAGhkphJ0rg8GfhBBCEgGnsKZIMq4WwyX0ZBBCCIkQJuMircLl3gkhhCQSejLigBu8Fq3GZDDwkxBCSIQwGRcJayVWDpcQQghJBAz8TCHs1OLMk0EIISQR0MhIIUqYJ4MQQkgCoZGRgoGfpWtqXRs3QgghpONAIyOFyM/OkNysdKltaJKqOq5fQgghJL7QyHA5VV98IZXvvhuzOc7+4E94MwghhJB4QiPDxaz98UdZePwJsvjiS6Ru0V8xncZaVsX1SwghhMQXGhlxQD0GgYqG3FGjpNNuu0rTmjWy+IILpLmhod1t4yJphBBCEtW30chwMfoB9rr6asns2VPWfv+9lN57XwyzfnK4hBBCSHxJ90hnuxs0HSqFKqD3oT0jvMZYaCq0DKqDlkBPQ8ND1H8Pam5FG4a6l87cCFS0ZBYVSZ+bbtIGSen990v1N99EfS1nrgwOlxBCCImEaPo21xsZ6MyPQ/E/aDvoK+hzaHtoGo6dHOY1jkTxCXQApMENb0AV0BHQNzi+Q5DTNofWQE+HUGX07yoyCsZuI91OxlttapK/LrhQGlevjvpaJSbwU6exEkIIISm7dgk6/94o7oe0Vx0Hq+kns38rFDrl4k5sv4H9IaMicbwEhY4z6ODRgaj7itmvry+GroeeVM8EjtWbYwNRFOs9sO+ouL3BCOh+xulS9fnnUr94sdT9uUjyunRplyeD65cQQgiJN273ZJwJ6U/v220DQ8H21yhuhnKhtrwZB0GdoKdsA8NcQ7kBm99DA4ynxGa0Kb9t9zuIEWlZWdLv9ttkyKuvSN4mI6O+zoDifKt848fFMvm9OdLYxKRchBBCUtPI2MuUfuPAwVRTTgjjPaohMT3E8Tmm7ONmI0PJ6ttXMrt1879urrccLxGxxcAiOW2nDURti1v/95sc+fAXsqyiJpbNJIQQQtxtZJjhjBFQE/RLkCq/mWMjTd2gwFtxHzQGeiLIPfT9b2FeLgpiZBSjjsZ+rIAqowk4jQc6lXX5nXfK/L8dIU11keW70Ed14V4by5QTtrZyZnwxd6VMuPNjmf7rsji1lhBCSKriWiMDFJmhkjIYCOv1pNinSSNKoXwzHBINGlS6AbTYBJQGGhkaD9IX+giaD+0EqdFxdpT3iwlNNTVS8fp/peann2TFHXdGdY0dNuoub529g4wfWmKtynrC49/INf+dJbUNTDdOCCEkNqS5daEsdOT9USyEFqCNg0LU0Y5fgzT7os7iCK+vHowPoELoGJz/pCNQdAXUYPY/4zjnUDOzRD0nW+LYDwHXDOthxuKZV3//vSw46miRxkbp//DDUjhOJ9xEThPGTR78eK785+3Z0oDtTfp2lsl/GyODSwra3UZCCCHeJtxkkujX0rzmydChEKW1Htl+UxGl1DSzU94xBsb9toFh3ay5Wb0j3aERTgPDHHsexd1QBnRqJPeMNfmjR1szTpTFF18sDWVlUV0nPT1NTt1xA3nh1G2lX1Ge/PRXhUy862OZ+r1z9IgQQgiJoo+J/JSEoTkqlLxW6ujsEqUqAgNjbxMEqlNUH4N8PbUDNTQgOyA0kNdNacdyhJWwJBaJuQLR3Bn5W24pjaWlsuTSy9p17dEDiuTNs8fLxE17Wyu0nvvcDPnn8zOkqrb9qcwJIYR4k/b2Z242MiqMoVECw2C9fB5mnw5t1OCNlodzQZxzCorXjAfjNuhEnGt7TMJlqSl9c0GTSFpGhvS55WZJ79xZ1nz4oZQ/+2y7rtc5NwtDJaPlpoNGWUvCv/TdItl38ifwbkSf/IsQQkjq4lojQ5NYoJhlhiY2ClJlmGn/zDANjH+ZQE4dWjkXl/+nuUewFOaanOucEJcabEpXjCdk9e4tva+5Rgq2204Kd901JuNvh201QF4/Y5wM69lJ5pZWyaR7P5PHPp0XUy8MIYSQjo9rjQzDNFNqOvBA7H1vhtFxalKvqyBNLHEEOss7WqmuwzOa5fPsYB4UcIwp327rvomi8557SP9HHpasHj1ids2hMDBePWN7OWrsAKlrbJKrXp8lf5/yjaziEvGEEEK8PrtEQSc/yOTI0Cmsu6Ct35r9W6J4D8qCBmH/crN/gBnGKDUBnLpvFIpvTd2/YX+rYwqon21ycOislduhC3BOozl2PIpHoSXQcOxfHWx2STKfaXNjo1R9+qkU7hBsOZboeGvmErnopR+loqZBenXOlTsO31zGDlmXFIwQQkhqzz5pDjG7xNVGhnkDp6G4x3gh1LDQN7ILlGmmmD7lqKtTUneErsL+K80+NSoOg8rNwmiheADnfGzO2c54KjR24w9oBjTEsWjaHqjrzKvhCiND7/vniSdJ1WefSb9775FOu+hjig2LVlXL2c/+IN8uWCXp+ATO3GUotKFkZrjdGUYIISReeN7IMG9iIooLoTFQren0r0Pb3wuoF8zI0EDNnmHc5nic87jjWkNRXA7tDtm5M3Ta69WoNy9EO5PuySh7/HFZfuNNktG1qwx+9VXJ6hm7IZQGDJvc8e4cueeD3/EeRbYeXCx3wqvRu0trE4AIIYR0VDqEkeEVWkvGlajn3NzUJH+efIpUffKJ5G87VgY88oikpcfW2/DZ76VyznM/yPLKWumanyW3HLyZ7D4iHDuOEEJIR0zMRSMjRYwMpaG0VObuf4A0lpVJj/P/Kd1OOinm9yhbUyv/fGGGfDB7hS8/+3aD5OIJG0tulk4GIoQQ0tGgkZH8DyDpwyU2mjfjz1NOReRKpgx65v8kb5TGv8YWTUn+6Kfz5KZpv0p9Y7MM6V4gh2/VX/bfvK/0RIAoIYSQjg2HS1LUyFCWXn+9rJrypHSZNEn6XH9d3O7z46JyOeuZ72V+WbX1WgNDxw3tLpNG95U9RvaU/OxgM4EJIYR4HRoZKWxk6DLwq1+eKl0POdjKDhpP6hqa5P3Zy+Xl7xbJ9F+XW54NpSA7QyaM6m0ZHDrtVddKIYQQ0jGgkZHCRkay0IRd/525xDI4vl+4LuN7ny65sj+MjYPG9JUNe3RKYgsJIYTEAhoZCcTNRkb9smWy7NprpcdFF0t2v74Ju++80iqZCmPj5e//kkWr1vr3b9qvixwIg2O/zfpIt8KchLWHEEJI7KCRkUDcbGQsvuRSWT11quSNHi0Dn5wiaQgITSQaJPrNglWWd+ONH5dIpVndNRPDJztuhPiNMf1k1+E9ODuFEEI8BI2MxD5s1xoZDatWybwDDpQGeDRKTjtNup+ly7kkh5r6Rnn3l2UwOP6SD39bIY0wQJROuZnWUvNqcGw5sKjV6VKEEEKSD42MFMyTEYqqL76Uhccfrw2Vbn//u6Tn50vhTjtJ7jDfIrc1s2ZJ1Vdf+b40aemCKE0ribuVzAuBo0WHHuq/VuW770pjRaU1lcQ6jnOy+vSRvDFjIjIOVlTWyuszFmM4ZZH89FeFf3//4jwMp/STAzbvI4NLCmhwEEJIkmGejOR/AK42MpTlt90uZQ8+6H/d+8YbpOsBvgVtV06ZIsuuvyHoeWnZ2bLxj5rN3ccfEydK3e+6rEtLsgcPlpLTT5cuE/eJuG2/Lau0vBuvIH5jaUWNf7/OUBncvQDGRqFlcAyBBkG63SVP170jhBDiRiODCQzigFsMimDoMEn2gP5St/BPbajkDNUlWnzkbLyxFB97jDTr8IW+h6YmFE3Wdlp6yymwnXbZVRpGbmKOY0djo1R/843UzZsnjeXrZpQ0NzSEHf+xUc9OVtbQC/YcJl/MLZOXzHTY8up6y8vh9HTYdCvItowNS2qIdPOVg1Ay+yghhMS3b2vLc821S1IkJiMRNNfXS+X770vBtttKRiffFNXlt95qDcEU/e1v0nnCBEnPyYlqSuzc0iprpsq80jWmrLa2a+ph5ARBv/d9uuStM0AcRki/ojyuHksIITGAMRkJJNWNjED0Oczda4LULVhgvdaVYTX7aNHhh8GbMiAmM1aWVdbIvBVVfiNkvikXrqyWBhNQGkhWRpr06pIrvTvn+UpoXZlnlSWFOZLBxGGEENIqNDISCI2M9Wlau1Yq3nxTVv3fM1Lz88/+/QXjx2Po5izJG7VJXD6L+sYmKy+HejvmwgiZX2Y8IdhevHpdvEco1MDo2SlHetrGBwySlsZIrvTolCvZmbFd4ZYQQrwEjYzEPmx6Mlph7cyZlrGhRkdzba0MfOpJyd9yS/8S9bFekj5kO+oaZcnqtVZw6VIYHEsgf1mB/ShL19S1eR0dklGPR6/OPqOjZ+cc67UmFytBrIiW3QqzsZ0jnfMyOUOGENLhoJGR2IdNIyMMNDC04n//k64HH+z/gv552umSnpeH2I3DJW+LLZLeIdc2NMryilrL8LAMEssAaWmULMdQTYgRmfXQpGNqcHQrMIaHGiBOQ8RxTMu87PiuNUMIIbGARkYCoZERHQ2lpTJnp52x4csCqjNeusLY6LLffpJRWBi7DyjGNGBIZsWaWr/RsQxGyEoEqaoXpAz7y7BtlXhtZzgNF522qwZIMQwRTVKmKsjOlELdzvGVhTlZUpCTYR3T7ULd7z+WyaEcQkjcoZGRQGhkRE/94sWy6oUXpPyFF6URRoeiycI677OP9LjoIhgbBf4Yj7Tc3KR7OqLJcqoGiBocpVU+w8M2REqNIVLm318ndTBgrGnESpTvVeNF/IaHw/hQdc3PkqL8bMuICZTuZ6wJISQcaGQkEC8k43I7zXV1VjbRVc88K9Vff23tG/bjDEnPzra25x95lBVAmtWrl2T27oWyt2ShzMTr3JEjJQ/yGpryvW7+fGiB1C3Qcr7UzJ0v9QsXyNq7H5WK7n1lDTwh+c8/KZ2+/Ehq8wulOrdQqnIKpBKqyMqXZflF8nW/Ta16qqbKSqlIz4naQFFvSTGGboIaImZfkWNfZxgwXjP8CCGRwWRcxPNoZtHOe+9tqXbOHFnz6ad+A0NpqqiQ5poa0ynPb3Fu0ZFH+o0MTZG++KKL1xkifXpbhohtlGT16ydpGYmLe2iqqrKm8qrSCwqkcIcdfO2c/ZvM23//kOdt1FghhcNGW9uLn1ktq/+cKzqA1C2gXu6IEXLNf87wB9H+uskoy8BI69xZpHMXaerUWRoKOktdQScpHb+HLBkwzMo/0vjH75L/+y9S3pgOiaysT5dVGNlZm5YldRlZ8kPXdSv25tfXSEN6htSlZ65nvGjMSee8LMnOSJeszDRfCalHxN7OMtvZOG4da7FP66Std45u52VlWDEq+X5lWmWevY3j6ZxuTIgrYcbPOECvRWzQ2AxnRlJlyOuvSeOaNdKwZInUL12KYRYtl+D1Usnfcgt/vbo/F1lGiioYG374oWT17GFtL770Mln7/fdW55+OGJB0DM1kFGhZaHlHuk460KrXBOOmSo0ec0yHcOxzAodwKqZNk6rPPvcbQw3Ll/uP5W871m9kaPZVPV/zhmQPGuTTYFMOHCgZaiQYepx3rmVIaeCsJXhA7O3MHr73Yhs0GkSrZTPqCKQt0wTsqqETdpYu2w+26pY9/rksf+OR4B9Ado5Uvz7dGr5ZVV0nG1xwkhQs/dM61AADpD4zS2rTVZny1oCt5blhu/o+o/K/5OSZr0ojDJKGtHRTZqDENspHRu4jK/O6WHV3WfitDKlYLPWot9ZRpwHlsvxi+aTvZv7mlFSXSzk8OA1q5ASQm5VuGRxqkNjGiBohGseyzkDxbWu8i2aDzVFjBsrJDNjO8hk5Wupr3377uO8YvTYkFWmOIuMnjQziOTQYNCOIAeKkcNz2MuilF6VBDREYIA0wRLRUw0QNFDvGQ6lftMhKhx6MTnvt5TcyGlaskEWnnxGiURky5NVXJGfDDa2XamCUP/+8/3BaVpZkGUMib7N1HacaAxt9bRala4PM7t0ttYVmWx327TfW0FPj6tV+Q6TBlHmbb+6vm7PhUOly8EHSXFtneYia6mr925oOfvjAYn/duV3ypW5ltnXdzMZ6S3nm2Nlje8v5p+1m5Sep+epLqf1gbsj2bXjBOVLdo7fUNzRL38mvSfHvHwWtt2zISOl3wERZi3iW2jXVcvZ/zrf2V+Z1klX5XaUst7OsyOksy7I7yUd9N5dFnXyGVkZTIwwV36J98cI2OGwDxTZCnAaJ7aHJVkPFGC1WGXDc99pnzNj7nUaN37vTwuvj9PikdZihqsY1VVLz00zrb0T/NpTSBx+SShjtGd26SWZxsa/shrK4m7UoY8E2Wye51aQ1mFY8hjDw05s0lJVZnXETPCQq9ZQ0VVVb21n9+kqnnXf2B6cuveZaX50qrVvlO0c9BrW10u/uydJpt92sulVffAEvyu9+z0RW794JHZ6JJzoco4aGvuemGhglMEzUo6MdgKLPsuaXX6W5scGaMaTr1zQ3NEL11uvCXXfzG3mahr5u7jxTp94qxarbIFn9+0nxEUdY9dQ4nH/IodZnpWvqBNL9rski2423cqBUPfygND31mDQVl0hDUTep7dpNaroUS1XnYqnsUiILRm5jpaPXacqFGH6qb2iSOmzrbCENuNXX9Y3Nsjyvs6zMKkS9Jsles1qKVpdaxxs1KBc4u/XZRf2lWQ0b0K9yueQ01kkTXjejVhMMgGZIX1dl5sIb08lvDHWpXWOOpbWor9s6XKVlODiHnAKNkMDhJ6cx5DeSsloaTDnG0+Ovo8cd3p0W+822XjOSLLn6PVIv39ofZkA/yNoZM3yeR+wf8PjjUjB2G6ve4ksuldVTpwa9Rs7w4TJk6sv+X9lzth9nGdl+Q8QyTLqhLJaCbbeTnCE+D571PcPfY0cxzpIJAz8T+7CZJyNF0U7XioGAx4LE8Tmjc1BDo2HZMmsIql7LZcul66GHSDbibJRlN9wgK5+YEvT8rIEDZMO33/a//nXUptaaO8HoedllUnz0UdZ2+ctTZcmll4ZsV+a7nyJWBfEqMEgyzz5ZMn/5KWi95dvuKrOOPdeqlwMDZ8ebzgl5zccOu1T+6LWBVXf8N2/KZr9/YwX6rs4uhPKlPLtAViHod0luV/mm53D/eWnNMALDNE5ijfbZWek+YyPTxNjodhaUI03W34fu71S/Vi59+jLJr61ucb4Ora3oOUC+3PlQWTx0M6tuUU2llKwtly4oO62tlIK1FZJXVSE5MPzSeyK+6pTTrRlTebhW2S6+Ychg9Ln5JmtavLLy6adl+S3/sYYZM3t0l6wePSWzZ0/rdVavntY6SyTczzzNNvK4Cish8QxYJfFHh3Cy0BmoQtHzkkuslPX1MELUAGlYoeUyGCTL/Qv3OVcetn7VWhe3C99GRnGRv15GUVfJHTXK0RB/ZYtBPQr934HFwzaUGmn0eVzslYwbfdvDRw6SHXfawKpX+0e6LHy0u/WLfl3dZmtFYy2vmrSpf2ht8eXTZPWXvniYQHKGDZO+D/kMl1oMLa3YHh4AHWro0lWaoSYE/jZ28qlqmx2kevBQyzvTVFoKLZcaeFfWQtWZ2VKtMTYYxlIvj9axVG9vm9J4gdY/ru1HPBTeaxq2+8OjM3zlAtl4FYRSH9Wpu17gb/eadNwvN1N+LR4ovxYNlF9Q/t61n+XBEbU9Zix2vEs1mjSOp8u6gf6u+hChuz7x7cPNC/e+WrrVr5FejWulR2O1lDRUSXFdlRTVrZGnZtfL2ud+kAIYJKO/+FWGY0iwfuFCS2sdd6or7Cyf5g2VdHzGGdDm152N99Mg9UUlUDdpgGekAduN6iUbuIE0du9pGVL6lbAWr8Y/mqTPKpvWbetnqtuN1r7mNus2mmOWgRbMsxTgSdJSY5NaHDfnOD02llENQzBRnlUOl8TyYdKTQQiJAxoPVL90GYJ9V1oBvw0rV/kCf/E6s3sPGFVn+mMafttqq3U5VgLofeMN0vWAA6ztlVOmyLLrb2hZAZ2P5qfJ6NJFNnzvXf/updddb8XzpBfk+4KdoQxT5mw8XPI28c3qqvz0Myl9+GGpnTlTmjGUCjAAVAAAIABJREFU2AIYYZmv/U8aESTdgOGoBgyrNeQVCGwa63U9jKxG3d/kG67SzlhjfDQmR6dlV1nybVfX6VRtDI2Z/c7jVo6ZMMirr5FuNRXQaukG74hV4rXG8zw0yufx0Of46uuXSHZT8GR6j42YIM9v5At43mrpLDn2l2lSCe9SJTxNFdm+co15Pb3/GH/QckHdWsuYqs+IXVikBlz3riqTLjCoutRWoayyhuO0/KO4vzw/ej/L8OhfXSq5sC+uOmMf2ahnS6M7GujJIIQQjxN20C9iXTae+aM0VqCzXLm+QeLMI5PeubMV06AxRbas4F81DgJ+5a6ZPl3q//or6D2LTzjBb2QI4nNqPv/c12bEIeVtvpnljclHsHHOiBEtpqNLX+OViDHq0fEZIeuMEqcxUo24HU2OF+hBaDQehHS8ONU+BoNl2mYPSM7qMsldvVJyy1FWrJI8vM6vWCl9Rm8iB27U16o7/OtZssFqp/elJT0PPtDyxKmHZLe7L5FuC+dIQ3au1OUXSr2qoJOl0hFbyJLtdod3BF6M6irp/uUHkl5RLpkqDBFlV66WHAwX5UKPHHetrESMj3qRjvx+ioxc8GPI+1fWNEil4P3XZ0m/NctD2aExh56MWD5MJuMihHgYHTpqqkbQ89oa/xRvZc2HH1pJ43zGSHULw6Rwxx2k8157WfXUuKn6/AvLuGhtSKsjogHPOnW+cXW5NOmsLluryq1Mxb2vvspfd/5hh8taXZXaHqpzUHTUUdLr8sus7ervv5cFf/MFPwdj8CtTJRdDfkrpQw9ZwbOZRTrzphgBr0VWmd61SJoRc9I8YJBvmMsa7sJwVnGe5dmIdzIuGhkxhEYGIYSQcGhWT4kabKt908ttwySrX3/JG7WJVadm9mxZ9cwzvqm7lvFQ5Nu2XmO7W7eEzlrrsEYG3pjOC9TQ7k0h9bd9C92Itr8dwTX6oPg3tDvUG1oIPQXdjOvUBqmvk7Q19FvDy3XeUzn0BvQv1F8S4h6cXUIIISRlSGtjdonrjQy8geNQPAapITAdUrNNExfoXMFT0P4Hw7iGzm3TgUItv4c0U9D2UC/oA2gPXKfeUT/LGBRqkGhI91eQ+qR04HEptA3qLwxyHxoZhBBCUoa0NoyM5EymDr/x6nG4H1oNbYk3sTe0pzEQKqA7UWfd4gqhudcYGFfg/DHQwdjW1IwaPr0TdFZA/TONgaGGxlCtD6n/6npjmNzT/ndHCCGEdGxcbWSYzj4Huh2dvD+7DbZ1ec6boVzo5NYuACNkGIqJ0B/GSLCvUYXiRKjR3Meur9bYeVoFOiNgKOUKaLZeD9V8qeMIIYQQ4kkjwxeyLPJKkGN2ntm2UrOp50MNh9dhMLSYQG2GPL6DBsJoGGF2a8Yd9Y7MwPEWy3ya818L877tRu0d2xVF3A0/K+/Az8o78LPyPq41MoxHQTt+7dh/CVLlN3NspKkbCntiePA8vyK/mtJO5xdpfUIIIYR4ycgARWaopAwehLrAg9inE4xLoXyotbRlGtehBJ0R4thvT+qOtD4hhBBCPLbUu70Wd8sVdFpip5wvNIGg0VzHeY1o6q9HW0McyZ7R44gGTrlrxvO68cArzzXVPysvvX+vXDNeeOn9p7mgrXb9jujJsOMnWnsS9rtPa8d1Aq8RaX1CCCGEeMyTYa+uo0mxQqGzS5Sqdlwn8BqR1l8PL1jzhBBCSFu01Z+15elwsyejwnT4JXgT6xlDZl8JVIOHoNk4Q2GvWKP5LYIRGIMRaX1CCCGEeMnIaPaZT7NMhs+NglQZZto/s41L2bNE7CmqgQw3pX2dSOsTQgghxGPDJco0aGvoAGNwONF9ypthXEPZD96Pi525MvB6AIrR0ALst66P8hfsX4DNMSj74/Wfjvpq1Oyn1aC34xUoE+/rxfO6XrlmvK6b6m3l++fnHw+88l1N9e+/pzwZBl2zpAa6CA9kC3sntrdEcaGZ6XGvY/8AaGNIh1EsYCTMM4aGej6udtTVWSQPG0/JrQH3vc/sf8TUs7nGeFWm4rqaQZQQQgghIfDCAmmnmbVCdAGz93QXtIvxwhyD9j/lqKuLne0IXYX9Vzr2D0HxqYmz+MmkBt/OxFe8Be1n8m7Y9bPNvcaZ2IvPjJGi65doltCxoVZiJYQQQog3PBnqiVBPxb7QF9B4aCvoE2h3p4HRxjXmmmGXx6Hu0D7QKugSaJLTwDD160w68mtMvoyJJi+GrvhKA4MQQgjpCJ4MQgghhHgT13syUhUM2ewGTYdKoQrofUi9K8RF4DM5GmpuRdcmu42pDJ7/ceZzGBfi+EbQM9CfUDX0I3SGCfImLvmsNAi/jb8z9W4TF+L22SUp+8dmgl51mfnpJgh1Z2gajp0C75MO2xB3oLOTlP9By4Mc/yGBbSEO8LeyLYrJrRzfDMVHUGcTs/W1+TvTc8ZCR/GBuuOzcvyd/RgifYDG2REXQiPDfX9sGox6P7QaGgeD4iezX2NR3oXuxPYb2P9XEptJ1v/P73h+Ju4BfyOTTAxWYYjjGkA+xRgYR9vxXdjd3fydHYltnUX2UoKanLK09VkF/J3djM/k6fi3isQKugTdx5lm9dnbbQNDwbb+yrrZpDU/OUltI+uzObSMBoZrOqx+kBoPLxkP4LIQVXeHNoU+cAaQY3sFCp3RppwVz7amOhF8Vk4j49u4N4zEFBoZ7mMvU74S5NhUU05IUFtIK+A/yMEoukL8j889aAzM0dA3Zsjj10j/zmBofGqGvsbhM+4Uj0aSiD4r28jQZSZ+47PzFjQyXIRx4Wo6c81K+kuQKr+ZYyNNXZJc7F9Xy/BxTIZ+h2qg2dAVkL2YHkkc2lEdC20DY6G11P8jTen3FgYZ409vZXkBkqDPCn9HxSgGmP//zsPrGSZIdzH0INSHH4Z7YUyGuygyQyUrTK6OFmg+D51tgs0eUCeziBxxQTyGybvyMaSxMlua7LJ76SwhfG6amZYkADzrG8Os2tZCh/b+nu1rEYnBZ2X/nY2BRkEfQosgjVP7O7Qv/s52wvUY/OlC6MlwF3YKc00AFgq7w2otSIokBvs/v+chXedmf2hH8yt5hskqyyms3vxb49+Z+/7OfoaG4W9MEzFqQkUdrnzGZHJmMKhLoZHhLuzF21rLkGYPk3C4JPkcbAwKnZ1QZe/E9nwUx5nP8WT8yspKUvtI9H9r/DtzD7dDujTETmYtKgvzN3eS8R5ugb8zjesgLoNGhrvQwCYlr5U69ji/v1MjyQH/ydXo6r0hhrZ+MC7dQrOoHvHW3xr/zlwC/pYa1biASoMcqza5hBT/IprEPdDIcBcV5j+/Eljl68XLmH26wqx2buWJbhyJmKWmzOezcx2LTamu9mhiNoh74N+Zi6GR4SJgOKjrdpaZMx7s1+8w85m1FjVPEoBObTSR7S8GMwgNOmasMHGa+7Bnlaw3e8TM3NoYajR/jySJ4OP4t/k706DP1v7O1HNIXAaNDPcxzZQHBDlm73szQW0hoVGP04HQQZAGe7YA/yHuZbxOM2E72r+aiTf+zjRgVzN/foLPrjJxTSIh2NT8nR0a5O9MZ9rtAdVD7/MJug8aGe5D1yypgS7CH5B/jBHbOi3yQhP1fm+S2kZaep0eMi8nO+fqY3sDx2fE2SXu5EMzW2F3fF46DVIcacXtz+7WZDSMrMcDpvwnPp/t7Z3Y1ninR01q+IfxJ2kPmxAXwTwZLkNnJuCP55/YvAf6HNvvoVT37S7m8zoGdYItxEUSzzXQeEhXjZztWAlyZ5Pv5DZ8Vjq9lbgMfC5N+LxOwKb+femw14ko1eO0k8lX8xDqvJ7MNhIf+BzewedzGzbPgz7CtmZkLTV/eyUmP835fF7uhJ4Md/5R6S+pfaEvzB+SJp3RDmx35zoLJLmYJFu7QhdD841xsZ353A7CcTUWiUvB5/MVim3M2hlDjdt9AXQq9I8kNo0EYP6WdLjkU5M3Yy8TlKve3V3NLBPiQtJ8Xl9CCCGEkNhCTwYhhBBC4gKNDEIIIYTEBRoZhBBCCKGRQQghhBDvQE8GIYQQQmhkEEIIIcQ70JNBCCGEEBoZhBBCCPEO9GQQQgghhEYGIYQQQrwDPRmEEEIIoZFBCCGEEO9ATwYhhBBCaGQQQrxHGkh2GzoSbnyebmwTcQf0ZBBPgP/DjoOaoUpoUBt1G7RuotrWGtpW0+7fk92WWHQk4HJoHlQLLYf2a6V+BnQ6Nm+Lc7v0+Ta08xofmOuMi1W7Yv2dweuh0DRsDkxS04KCNu2O4q2AfTuZ9/BukppFXAKNDOI1CqGHkt2IFOUY6BqoN/Q/6BNoTiv1D4Puhori37SU4A1oz2Q3wgmMiD4o3oE2TnZbiDvJTHYDCImC3fCf20nNzc0P8+kllK1NeR2evRobbvkRMxxqjoEBlQ8taH9z2s1f5j3VeeBHYag2fWXeQ1UC20JcCI0M4jUWQf2gW9V1jM5OX5PEkGPKP930wPEd+DUG11gYi7bEArSlHkW731OS30O1198DiQ1utIwJaY23oVegztAD4T4qGCTzzRhxvyDHHjbHjnPsu9Ls2wuaBH0FVUNLoLuhfBNzcDE0xxz7CTq5lTZsAE2FKqDV0H+hbULULYJuMteugVZAL0KbhXhvpdDm0A8mXmIutGkYz2U70ya9fp251r3O52SPr2PzRLPrMfNsHm/luh+geNK8PNbUv9Ice9y83gF6DVoLLQt4/oepEWniPrRdq6APob+FE5Nh9n1jnuM90F/mOf4CXQhlthWT4djXBbrAnFtjrqXXLAnSlkzobOhH853QurdDnbSN5rlEFJPheP4bmCoaE9PCc6OfF3QftMB8/ovNcx4c4nnps9nV8f3S99bTHC+GroG+N99Vvd4i6ElomOM6VzoMzoHmuh842xwsJiOc71w7Psd06Bzz91oOrTF/E5dB6qkiSbA4KT4D138HgHZAuqFDJL2gleb1MUHqaofTHLBvvqnfL0h9vaZuHOfYp/+BNptxcC2/hl6DVpvXL0DPQWuh90yMQqM5dprjOhqkqhtl0DJoFfQy9KnZr79a9wtozwDoD3Nc2z0V+gxqgmqgiUHe2xrjZv/NGGFaZrXxTE9ztPlz857mmNcroDGmnrq9n4K002s2bdHXJ7dy7csc7/EPU3+SOabGiW7MNs9Fn8dcaHNz/F5zvBLSQMeXoFlmn+qsgHvpPw1B9s0251Ua4/Qd+7sB/SegvnaOujEuyL6p5jl9Yrbt78A3AddIN8f1RbnZ/sCcq98fLT8I47tuf2d+D3j+lY72POWoP9p8Xs3Ge6DP6zvzWr9vWwV5NovN9X6AXtfP3xzraT6LZvNdeNl8BmWO99Xf1J1kjjeb75+28TJzbCez/91ovnPt+Bxvd1xL39erjra/D6Ul+/+y5hRT0htA8RlEamSY18ea12ps9IqjkaE6wbF/M9PZ2/ce4Tj2d7P/hyAdRrP5j7/Ecexwcy39D7GTY//Hpv6NUIZj/+7GqNH/6HsEeW9f2IYFSG/jeY42/9nrmPluAR3lv8319Lo5rT2nNu5xlKn/eMD+x4N0WFZ7wVbm2BznszLHzrWPhWlkqL6F+gQ8Q92ogArCNDK0nds49vcxBqO+GO/Yr54e3fge6u7Yv5N5zs3RGBmO/baRN8ixL9thkJ4RUP8Y8/2ap/WCPJuX7E7X8fzvMcduc3bI+v00RlazbUiY/f3s70rAvdczMqL8zoX9ORrj3Da0Ch11i4yhoi92Yp/TnNB+l8MlxJPgy/sEijfNfyD3xfFW+mv1Ucd9ZzjGmu/Ea/2FZfOiKTcMcS3tBEod13rW/EevbvdDdB9cumNRjDMGySWo0+io/z8zW6OLY+jCyX2oU2/qaufSGmdB+vd/Ler6Xdp6HnSV6WB1quShbVynPUzFvf4MaG8X8+v4UuezMjxoSu1MwuUiXEd/tTuf4WzTaYY7I+JunPel4xqLjTfBGQxrP1PleNRZ4aivz/KGCNocCepNGKLtwX30u+EHr6eY75caLQcFOfeuZrsnX/f8S43n4kr7mDmuXoRnonj+sfrOhfM5qodTWYnjaxx11Zujw5gnGC8NSSA0MoiXOcX8kjlAx/DjdA/1DgRidyDqanaiv3iV3CDnLMR/djrMEMh/TTnelDub8gPnf/IO1F2s7BjkmBpA4bKDKZ8PcfzZVu4TK9Zrr3Y+0EGQutEt8NnmaryJ+WXebH69h4vfOHCw1JQFsboG2tcDxabmF33g90Lxv58YY39fdChA2vt9Qdv/DU2A9O9KzHvrrvEbju9oJM8/Vt+5cD7Hn4x3cVu09yPoNDsmBe/nQ+gxNwX4pgqcXUI8C/7D0GC0C0wA6GRsT3f+gowRK4Pd2pRlAe3RILVQ11E3cDDs2THqglf6m/I8XOu8Vtpl12urraHo00a71MXu/HUYD4K2F+87D8Xx0P7QCKiv7jbPPZLMkk3mF3ggDRH+yCoP4xr25xGqEwv1nNuLfd+78NzuCqOe89mUhwpQRnEGtD00zARZO7/30Wb3jPY7F9bniDoaaHuo8biMt40i7JttvGP36v8ZUbadRAmNDOJp8J/Gg+Y/Fv2lNdnEOURKRivHrOGHGKCxFMFIC7hPhsODomPtoQhmTLU1RBLsvqGw21EbwTUjZb324rPUjugjSDu6cpNv4QXzq3u66Yha+7ycBPMERUM418lqw3CJV9pt+1m8a+JEQuEc1gv5nvD8j0AxxVx3jsnkOcsErvaPZEZXsMtH+Z1rjuD/g/d0dg4294X2Mf8vqKF0CXSmemRQR79TJEHQyCAdAQ24nAnptEfb5RqqQwv2ne8al1YF/xUXiJ0i2v6FtcSUb+A/w2vj2B4d31ZXsipYynMd55c2Oq54cJ0xMDQO5lQ7xkTRaaARGBiJZlEb8QrBPE+xwP6+PIFnpbM7ogbPV7Pp3g9pHNC+uF5gqvAz23P9RH3n1KNhZn49Z9q9mfleqdFxNbRXe65PIoMxGcTz4D+VeeaXipgg0GC/mNYEc8XiP6AMM6Mh3owwv9IDsQPy7PwJ+itemYD6670P7DrF5OO4op3tse9jBZwGwQ6+azOvQ4w9CXbekJudBoZhD2c+hKhbFQeMG147zgEh8pPsG4vbtPI57h3sBJPvQvNEqCHeFjo0pYbcd4EGRsDzT4/yM47rdw7v8RDNLQJd6txvgrUvamfQKokSV/2hEtIO7jZT7HqF+F6rp0OMyzTN0VFdb8b8440aM084EwJpYBqKCSaV9VRHAJ/+p7gddJ0z2RC2R5v2joR+bGd7JptfrLrg2a6Oeyj/MkF6CxyBqdGgOT3sGSPhYid3arHwmpl145w9ESy4NtncaUpN7lZs78T2GBRXxGAIJ9jzfNYEQP7NLEbnB691nRONWdrUDHeE++w3wblDAha607wnE4M8e7tNmmwsLcnfuV+MF0yToQXO8LKTuIXzHEgMoZFBOgRmJsYJrcQ+aFCc/jLWMeef8Z/Qi+aX57mtRLvHkl+Nx+QPvbdmMTQ5CTSg7XA0v9bxPg43bvBLTHbHV00mRT1HO6/JqKZJhqIG539r3rsGWb6L638G6XPQILmrTMzHocb1HC324mn7mcyeIbOhOrjDlDebrI0vmGf1uUlrvjQBAanRcp9JFGV/zprV8n9mZkRZDGJ87Oep3x9VZ/P5qAdAZ4PcbTJ46n0/N1NR9ZmdH2LGSwtQZ4kZYtBhk5m4xhvQK8b4uNYR19ErIHh3pflefoL6dyTrO4fzfjLJuHqYv/H3zHPSdl9mhmE0HwdJIDQySIcB/8nof8L/CnHsSzPd739mfFxdv3NMBH2o6X+xZIGJdv/JeC+GmIBGzcbYYposXqtBMtr8h6m/FPc0WR81SdfB0NmxaBDuM9lMF9TMiBsZ70Gaue/m7Q2Qw/k/GkNpuXne24dxzlvG9a9eqQ3NeTmmA9/MYRDGYvghppicJvoMLzfveW8zBKHP82hTTbOFRssF5rloAqxdoEHmvh+b78vD5lntbYYF3jJJrzSxVricYDr8P809NA/IPDPbRO+xykwRLXEYxUcbQ2HLtj6XeH/nwPmQegh/MG3f13he7jbXj9csHxICO9sbIYSQdmACDFc4k0Y5jh1oplHehOMX80GTVIGeDEIIiQ06vVMX8FJPlR/zq9920+vwAyEpAz0ZhBASA3R2gz1tEnxlEnN1NcNEGvBLLwZJOWhkEEJIrP5DTUvbzsTMbGMCJCvMgmn3tjdYlxAvQiODEEIIIXGBMRmEEEIIoZFBCCGEEO9ATwYhhBBCaGQQQgghxDvQk0EIIYQQGhmEEEII8Q70ZBBCCCEkLviXkSYxScTDhWAIIYSkHM3NzbrQ3XrQk0EIIYSQuEBPRnwsulh5RmJ6vXhe1yvXjNd1U72tfP/8/GP9nYrX98or14zXdWN9Tft6oejwngw8gON0GAMaF+F5faAHoLnQWmg2dAWUE6+2EkIIIR2JDm1kwCDYFsXkKM7rh+JL6GSoHHoD6gxdDU3D8axYtpMQQgjpiHRYIwOGwCQUb0OFUZx+L6SGxhVwKY2BDsb2htC70E7QWTFrKCGEENJB6XBGhnohoCnYfAnKgJZFeP4wFBOhP6Dr7f0wNKpQnAg1QmfGrMGEEEJIB6XDGRngWuho6BtoLPRrhOfvCWkky+swLJqcB/B6IYrvoIEwRkbEoK1JQQN+Yh2g5JVrxvO68cArzzXVPysvvX+vXDNeeOn9N3uoralkZKhRcSy0DR7kzCjOH2nKn1q5vjIqimsTQgghKUOHm8IKw+LGdl6itymXhDhu7+/ZzvsQQgghHZoOZ2TEgAJTVoc4vtaUhdHOGw7XVZVIlxZpH/ysvAM/K+/Azyr5tNWfpeJwSXux4zBC9fD2E2/fkyeEEEI6OPRkrM8aU+aFeGa5ptTZJkGh9U0IIaQj0FZ/lvIZP6NgsSl7RRmzQQghhBDA4ZL1sWeVhJqiOtyU0cxcIYQQQlIGDpeszzRT7gc30MXOXBl4PQDFaGgB9s9KxAdECCHEuzTV1EhTZaU0rlkjTSp7u3KNZA8eJPljxlj1an//XVZMvrtlXSijqEhyNhgi3c86S7IHDfIPYbQ3IDNRpLSRYYyGfKgUH1qp7kM5D/vV0NjLrFVyuamrs04eNllEb01OiwkhhCQlLkE79nSf879uwQKp/u57aVy1ClopDVZZbhkIzbW1Mui5Z/3nzjv4YKn7XRNIr0/REX/zGxlNVVVS+bauhNGShuXLpXb2bOl+zjn+fUuuuELWfv2NZG+wgWWA+MoNYLQMkYxCe4KkO0hpIwNo+vEdoaugKx37T4c+hS6DcbE/ytnQdiYe4y3ovgS3kxCSQjStXevrwMrLpblRVzLwkbvxxpKW5VufsXbePOuXbjAyOnVa96u3rk5q0EkFkpaB30tQVt9+/o6pcfVqacQv7LQMdKbpGb4SdazONSOzRQfW3AQnL35Ne+UXtRN9po0VFdJcXY3339e3D0ZE2f33+wyGlWo0rML2Sv92z0svlaLDDrXqVn3xpSz9979DXx/PPC0729rOLCqWxqKVko7PJKOw0CrTUep27qab+s/Rz6vvbbf6jhfgeCfULSiQhtJSqf1jrr+dSt2c3y1DR7Vm+vQW9+68377S9+ab/V6Ump9+soyQTHhEkkGqGxlBwZdNl3ff2ngyJpjF0eZCd0F34HhDMttHCPEW2nk3lKHDKtcOzPzyXenbLhi3vRSOH2/Vq3zvPfnr/AukGUZGMIZ+8rFklpRY28tuuEGqPvo4aL3CXXaR/vfeY2034B7zD/F1jsHod/990mknXfdRZOUTT0jpvcF/Q2X26CFDP/rQ/3rO+B2ksawMxogxRFQwgFTdTjpJup1wvFWv6quvZPmNN/mPpWXDSLK3od5XXikZXbpYdVc9/7zUzZu/rm5WpmXMNNfXS86QIdJl33197wn3XX7Lf6z9wdTzkoslb5QvKXPZww/Lyqf/z9ov5rgaceqZyB48WDZ4602rnhpLZQ89LE0wPEJ9hjY5Q4danbl23BkwInRII6NrV8norAZEJ+t52Ax8Un/Lto0+g857773e/qw+fSTPYYwoA6Y8IXXzYWTM/QPDLBDKOhgidTA8M4u7+evV/vabLDjqaN/10cZseD1yhmwgxccdh3JwWO1qLx3eyIBB4PvrifzYnyh8fyWEkLgROL5cv2QJfk373M6+X4Q5+EWXL+n5+dYvdP11l2is9R60Y8rMlHTzC1V/Rdb88iu8CZXWr/+mygqrQ1fjQel3153+8/+YOFEaV1gjsuuRlpPjNzL+v73zgJOiyrr4mzzDkBmQrGJAglkUFcwIZsW4rgEThlV3jZh11zXuZ1oVc9YVA2JAQTFgREWMBAEFJEmYGYZhmBy+c6tvNTVNd09Pd1fH8/9xeJW66nV1T79b9913X2ZBgXUdaWCzOnsarwxc04uj8crdcivTMNhzLV9y+/bddH68Pn/w4OYHSMMtngg80cs9tZEGMrsnHLYNsr/BKuUYOVaevn1uivdc1vmkAceTs7ULn5uNeGOq5wYOYesO17/Nhg+mmY1ffOH3uLaHHOw1MuS7sf7NNwOeUzwPVtnYZKrXbzD1+E750oj3WpFTYKbNXW3q8T5roQ5Hn2oaYDTVFLaH2pnqNh1MJcqqgnamJjvX1E351Tq2riHH1O19urVc39BkvbaxuMlkl2aa3OwGk/vbXJMLL1BetqxnWstW6bO+aX9WwH11OHd1XYOpqm00lbX1pgrLsl5Z28ZUdR5oqnbpb6oGYr8cU1Vr6qqqzYYJP2B/gylaPM8ctMXWpqh0pcnHPan6bpalMaV9zEM3nmy222LTZ+8WGczpEMWbmZERcEAx7zPxfhfkB7m+Hj/E8lRV633Cyu7a1etirV261NSvXWv9eIsbVfYlIlL3mt9/N9Vz5praZUvRyNSYxppqq+yAJ73Cvff2PqEXP/IoGgcEwdXgPUswnBgD8Q6TAAAgAElEQVQRKKVB2mHObG9/9+LjT8D55vi9XodjjjY977rLWq5ZtNgsPfccy/jILPAYIU51Oe9cb2Nb+f33MAqWWo24GCziLpfAOzEQMtu3Nx2OOMI6Tp5wV1x+hSM4b4PVJSHL0tj2/L//Mx2O9Bxb8vQzZo26pX2Rz7H/Tz96jaclp/4VT9/FJrtjJ6/xkN3Z8xRcsOuu6Jff1ds4N+J7IXVMhm4I63dNjRCD73QD6l6Hz7UhJ8805rcx9dheu77c833G596A92eppg7Ce8X3v3qv4aYeXTHSWGd/Od1krFrpMVisv486U28yTG1Gpikr6mWW7DjUVKPxrN1YaXr99JWpbsowVU2ZXlVivbIh0/xW2NWUZORbjX/b2krTpr7a1GXiGugCqs/IMrVZOTAmNhlsKU9TkymqXm/6blht+mxYY6ZsNdRM/PuBZsfeHg9SqAT7TuK7kJGWngySvlg/gPiRshtu+SGs/vlnyx3aiB86p/tUGn35sReXrFD1y2xTMX36JhesGAW2QVCPxuauTVPk/HnTzXBbLvI51rPcHo1Xt8s8AVtVuPaSv5xq/Sj7Y+s3J1l97kLxY4+Z9RPf8O7L6tLF5Pfvb/Kwv2CXnU37Qw915Z4FQ4wCeSK0+/plfcFeQ71Pr77kD9jBa2Q0lG+w+oYDIU+mGTAAhJy+fSxDJTMv33qitxpe+cwgaZS99YGBUL8ycLqajief7F1eP+lNU/baa/7rCbe6bWTI9So++SSgx8H5hJ637Tam3YgR2sdeCI9Ae4fx0Lz/e6v/vRSwns2uge9qFtSIJ3BpoOVJvF6kT9p1aIjr6qV0rEtZ71mXhnqzfbJe33zd37LnCd3/MVIHOYeU9vXqdNl+kpfXY1MrkNgSaJ5MbG0jXqrtNq067YBS6NNFjpuF4wK1edaYwEYJGbE8Fg05HU1hbpYpyIFQipcgG0ZtDsqczAyTnZVhcuBBEGVb6+JR8JSyT7wLcrx3WffJa3N0PQvL1r3AvanRe15rC3/zm5ax37HPu+x8TcOmfTk4d4Fdd62/XbaB8nW7vdwmNxv7MvWY7Oav09fepfcgFtDIcAF6LaJDQ8VGU/X9LI/rHE+YdkPTiKeYTqf91eRs4ZmjruTZZ82GDz+09jVVOo6D8gcNMlu/9qr9wZglp/wl4PW6o2/YNjLEvVv8sKdP2x897rjd++QtT92BnrwbStFn7XR1q4Gxqc95U/+0E2nICyTqHMeLp0D6vjd+9ZWlNnvt5TUyxFhadTPq3X97GCgDoP5W33CkyL2r/nW+dR9syRC7rE4dzfafe+IAMtHo5vRGMFpdPe7zQCuyXTwIGfno3sjPNwU77+w9X9v99zNbvfqK1VDL6zKw31rW0r6XQu/77gupjmKQbfuR53P3Sr4nGz3LOb16eo8t2GUXr7HSIMdhWxOC65raFJq6nr3NnJXr4YLGjzpczk3X3W5qcvNNdU6+qYQ7fSPKjdl5eGrONNX40a9+a7blrq6ugzt9yBmWm7pGXouGoWEdGlp85I1NeFKf8ZlpwHdODAZP6XHfO7fJerP9ui2ZkUbRbpDtBtvTkDu2SQOvDbU0zvZrpMzyeU1+jjaYUL42ks7G1W44PQ1s83VpSJPBI5TMbVtL95fdJS50l9DIiBxxby899zwr+tsfW778P9MGngdh1a3/Nute8v+kmIen/35vbeq7XfLX06w+6oy8XDTsuc0a+w5HH20Kh+5lHVc1e471ROvcL0Fo9nJ79A3bDWPVTz8hirtG9+dagW32cdKPbUd1233g0q/fmh8++T7VrVhpaub/asUA5PToYToeP9raJ43/4tHHNzs+u3t3j9cDnoROeJqX44MhRpzVN6/GSRk8KH/ecMOmPncbvN9cBIttNWGCFRnvjKKXOspTl/1kW2c/hdtPyfrUK9vkCdnqy3Y8pVvbGpvvr7Of4P083TV76rO2NTTb5nuM82kxWRpxaWzxz/uEnaNP19bTt2PdapD1SVyesq191jGep277Sdu7r9m6/dQur7Eb+k3nc75u01P8pv3NjkVlpc5s1NOLDP0tC9RdQiMjujebRkaYSLR9zfwF3kZenjgX7rc/GrV+1hOpb7+79PdL1LVQg64KGeZl7dcAQc/xBZt5CFINed/lH3xgasTrACOkZsHCZiMT+r33njeKvOSpp6y4hDx4O2RMvuWhQCxFHfrLiy652HT92988owG++dYsPQexDltvY+q22c5s7LutKe25tfmzqLcpbcgypZW1Zt1GyCrrrPUySIyDZEGejuVJNz87y7ucZz0Zb77ds8/zNO3Zt2m7HJOnx8sxWfjB9RgHntIaCbrZNscySmtwxmbb+PRNkgMaGbG92TQyWoE82Vd+8w36yl83G6ZNs1zoMkROjANBIvWzESRHWnFPpe8XhoR4PSpheOScc76pgCt/QzXiRC4622TP3zzKvx5BcN/ufKB5ZchoGA51pnwDuhzEk4BgvNbgfDL2uMmbu8U32+brUrdc5Zv6yO3XbIq2d0Tgy3pWkGh9e9kZta/HyXXsH0ZCSGTQyIghNDJCo271GgTiTYJbfqKpW7bMvnmmcNgw0+OWm5slnUlH7K6HjTX1UIMpr66zjIQNviX2l1d5ljc/xjPUzcmgksVmu3XLzNblf5rqrFyzsGNv81vHXmZZuy02i7SXILlOhbmmUxvIKnM8y1DnwpxN+6z9nn3yZE8ISS8y2F0S05tNT0YL1MKo+H3UYd4AyGyJLxg92ooxsLs/kg0J3FuPxr4MEsNAxqdvrEUJA8FTwljANhnjLkaDVcq6Y/umYz1lNMIGxOPeviDHtMvPNu3ytMxHPAnKDjAaOsMw6AhjobPDUOiM9Y7YJ14DQgiJ1Mjg6BLiKjI+fuOXX5pOf/GM6sjt08fkDx5kcrptYTqedKIp3GcfT3rjBENGD5Qg7qCkogaqNWu1lPViWca+YqzLcimWox1MKK7+wjzP8DMxDtrDOLBK22hQg8FpODhL2S6R9vYPACGExAMGfqZpMq5yxEBI5kJJZZvZvoPJ6gi1b988u2CYSP6EDdM+NGWvv24qv/7a2tbv3cnWBD6C5JGIxnVai4xUWL2hxqwsqzJrysVQqPEaCrYx4VmutboiWoM07tKFUJibDePAM1bdW8p4dhgMVul3PwwCLNtlGxgWEo9ACCGJRLCHFnoySDP+vO56K5uhLzLksuvf/246n36aN4FU2esTYYy09xgkkJRZaphIDn/bE1G9YIFlWJS/9bY3z78Ec7YfObKZt8ItA6MKXQ8ryiqharNiXZW1vNK7XGVWlVeH7HGQgMMubXNNl8I8U9QOggEh60Vt81Bi3bvsOUaCCgkhhPj8ljZfJdEg0bwW/urXdv/9PTMuWiozjWUy+6InhbI17k6pWbjQlL2qyaz80P+H761MjTLkdMnJp3iHT+YNGGA6nniC6XDkkZaHJBp1lpEPtvGwmSEBI0K6LVqiKwyGnh0LTPf2aizAeBAjQgwF24gQA6IDuiXY1UAIIc1/h31p6XeS3SVpEvi5bsIEa4a+buPGNcuu6DukVIwM8TRIngl7fojKb78xDWVqkJR7Sssoqdxotp440fslW3X77VZypo4nnmgKBg2KqL7SpfHeL3+aL34rNstKPYaE72gJX2R4Yo+O+aYXjAgxJKTs1UlLSPYxoJEQQqIHR5fEkEQ1Msrfe8+suOJKK4Nj32efMYVDh8a7Sn5Zje4MMSwm//ynmfXH5rNLSjBjLx/joacu90YpXggmMSKEkNjB0SVpTsXnX5gV466xDIyu//hHwhkYazZUm6mzV5nJP/1pZv5R6s1kLZkUD9qhmxk5qLvp372dZUzICAtCCCHJA2MyUpjKH34wyy+91JpptPOYMabL+WPjXSULGcFhGRY/rzTfLN5kWEjw5IH9u5ojduppDoaBIUM4CSGEJC/8FU9RZKTHsgsutAIxOxx3HGIxro5rIKMEZb4/x2NYzPi9xJtsSuIo9tu+qzlypx7m4AHdrBwPhBBCUgMaGSnKmjvvMo0I0Gx78MGmx63/iouBIZNmfTBntZmMOIsvEcBpDx+VuSMO2A4eix17mEMGbmGN5CCEEJJ6cHRJigZ+yqymxY88YrpdcYXJzMuL2XUlvfa0uavNu/BYfL6w2JrKW5DZJffdtsjyWIwc2N1Ka00IISS54eiSNMr42YiuEUl+FWuvhby3D+etMa/MXGo+W1BsahsavXNn7LNNEWIselgBnDIvBiGEkOSEGT/TGDEwlp57nsnbdlvT/aYbYzYfyOwV682tk+daAZyC2Dd79+tiGRajBne3hpUSQghJTxiT4QKx7i5pqqszK/5xmamaNcvUrVhhGkr/ZrK7dnU9p8V/3p9vJn6/3BodIlOBX3TAtuaYXXuabu3yXb02IYSQ5Mj4SSMjyZEsnSuvu95UfPqpyerY0fR96klXDQyZnfSJzxaZRz793ZrSXII4z9x7K3PJQdsxzoIQQkgzaGQkuVW5+vY7TPk775iMNm1Mnyce98506sa13v5ppblryq9m5fpqa9uhA7cw1x4+wGxdVOjKNQkhhCQ3NDKSmOLx4826F180GTk5ps/DD5mCHXd05TrfL11n/vXOXPPjsjJrfUCP9ubGIwdYQZ2EEEJIIGhkJCmNNTVmw0cfGZOZaXre83+mcO+9o34NmdlUPBfiwRAkiPOqkdubE3bvYw1JJYQQQoLBPBlJnCdDpmavnDnTtDvooKied2NNvXlk+u/mic8XmZr6Rivd93nDtzYXIrCzLVN9E0IIUZgnI8WMjJqFC00u4i4CTdceCY2NTeb175dbo0bWbqixth21c08zblR/07uTZ+p3QgghxIazsKYQlbNmmaXnnGvajxplevz7VpORHb3erq8XlVj5LuasLLfWd+7T0dyEuIvdt+wctWsQQghJLxiT4QL+xg1H6t2onj/fM+FZdTUCPfGxRSnZ1h8lG83t780z789Zba336JAPz8UO5mh4MDIZd0EIIUQJJ5s0YzKSIK147dKlZsmpfzUNxcWm3YgRptf990Wc0VPmGHno44Xm2a+WmLqGJlOQk4WYi20Qe9HPFOTGJlsoIYSQ1E4rTiMjwWMy6tasMX/AwKhbvty02Xuo6fPYYyYzN/w5QKRuL3+7zPzfB/Ot6deF43frba4a2d90hxeDEEIIiVZMRvSjBxPjTR8CfQwVQ+XQJ9DIVp5jP+hdqBSqgGZBF0Ix62JqWL/eLEMMhhgY+TvuaHo/+FBEBobw3R/rzHWTfrEMjD236mzeuXiYueeknWlgEEIIiTopZ2TACBiDYhq0D/QtNAPaF5qKfWNDPMcFKD6BDofWQh9BnaDxkBgehbFKGZ7RpsDk9utn+jz+mMlqG/llV6yrsspDBnQzr5w/1OzYu0PE5ySEEEJSPvATjX8PFI9C66FhcN/M1u1DUHwIPSDeCWxfEeQc/VE8qAbYJTj2Id0ugQp3QldC/9TSVbI7dTJbPv20aajYaC1Hg+IKz9BUGZIa6ynhCSGEpBep5sm4BJK5xe+zDQwByzNR3A1J0EFL3oyz1Pj6n21g6DkaUFwDzYEuRQPt7jSnSmZhocnZolvUzmfHYXQpjKzbhRBCCEk3I2OUlm/62TdJy8NaOIc9Acg7vjvU0PgMyoEODqeC8aakwmNkdG5LI4MQQoi7pIyRkeHx/Q+EGqF5fg5ZoPsG6bEt3ZMNAfbXazkgnHrGmxKvJ0McPoQQQoh7pIyRASRoQVrOEngcPC2pA2wT46AYkvzY7YKcZ76Ww313qHEiAaVCTLpLok3pRk9MRhd6MgghhLhMKhkZ9tCLyiDHeIZWGNM2yDHPQ5Lo4jLYFEf4GBg3QrvrpoCuADk0mBLBk9GZMRmEEEJaINL2LJWMDOkKEYJlwrLvSMA7A4/H92pMSNDCZNzE76FJ2t1yA/S4HloXWXXjQ6nGZBSxu4QQQojLpJKRUaFlQZBj7JSWG4OdCIbGbShGQ19BO0AHQHOhodCnelhZkNcHVbyoqW8wG2rqTXZmhmlfkFKjlwkhhLhApO1ZKrU05WpoFElWTo3B8KKZOougauwLaCDY4JhJjhEpzvMcq4tLI69ybCl1dJXEu9uGEEJI6pMynowmj0kl3gZJmrW9n0P66/v9Jdh5JP8FdBC0XYBDDtLyu3DrGvfhq4zHIIQQEgNSxshQpmppexuc2Nvea+Ecu2ka8Zt9d8Dw2FFHlyxOSiPDHr7KkSWEEEJiQKoZGc9A1dA4GAS7O4yDPVBcraNLxju294V2gKQbxUaSbUl3yknYPtRxbHcUL8ki9E84TuxA0+QbvsqgT0IIITEgpYwMNPxLUFwBtYdmwDCYAk3VAE7JjTEWx6zxGa4qibsudpxDDJHzNavn53j9B9BkLP8GiSdjPI55LiZvKMqwu4QQQkgsSaXATwsYAONhFCxVz4Uk1JLH9y+g27DvoxDP8SrOUaVzlQyDZHkWJHOZvO5KxWPYXVLE7hJCCCExIOWMDDUSxPMwOYTjZGhqoH3v+Ju/JJmxc2R0ZncJIYSQGJCSRka88Tc8NJ75MWxKNCaDo0sIIYS0lnBSH6RUTAYJDrtLCCGExBJ6MlwgEbwW/mDgJyGEkGi2bS15N+jJSCPsjJ9d2nKad0IIIe5DIyNNqK5rMBU19SYnK8O0z6cDixBCiPvQyEgzL0anNpy3hBBCSGygkZEmsKuEEEJIrKGRkSYUV9gpxXPjXBNCCCHpAo2MtPNk0MgghBASG2hkpJmRwURchBBCYgWHGaRJxs9iTSlexOGrhBBCwoAZP0mL07zTk0EIISRW0JPhAvH2WviD2T4JIYREAjN+koBw3hJCCCGxhoGfaRf4yZTihBBCYgONjDShxM6TwSGshBBCYgSNjDSZt2RjbYM1b0m7PIbhEEIIiQ00MtIoHkNGloQzBIkQQggJBxoZaUCp5sjowngMQgghMYS+8zRIxlWiOTIYj0EIISRcmIyLBM2RwcnRCCGExBJ6MtIgGReHrxJCCIkUJuMifilmdwkhhJA4wMDPtAr85DTvhBBCYgeNjDSA07wTQgiJBzQy0oBizZPRhdO8E0IIiSE0MtJomnd2lxBCCIklNDLSAO8075y3hBBCSAyhkZEG85ZU1jaY3KxMzltCCCEkpjBPRopn/OS8JYQQQqIBM36SzeAU74QQQuIFPRkpnvHT6ckghBBCwoUZPze5dA6BPoaKoXLoE2hka24mjh8ATYDWQHXQcugJqE+4H1A84LwlhBBC4kXKBX7CCBiDYhq0D/QtNAPaF5qKfWNDPMeeKGZCJ0OroHegauhc6Hvs396Fqrs7fJU5MgghhMSYlDIy0Pj3QPEotB7aA66dw6GRamSUQw/gmF4hnGo8VAhdhtfvBI3Gcn89dxF0jytvwAXYXUIIISRepJSRAS6B8qD7YBjMtjdiWbwSd0P5UFBvBoyQzih2h0rxuvsd52hAcZOu7hflerveXVLEHBmEEEJiTKoZGaO0fNPPvklaHtbCOcSYENrB4Ojgs0+8GEJpGHWLC5zmnRBCSLxIGSMjw5OcYiDUCM3zc8gC3TdIj/ULPBbS1fI1lANNxKGDoQKN0/ifHvZ/Ua18DIawcnQJIYSQWJNKQ1g7aVfJWhgKnj4CB9hWL6NNsNgNaqcxGoE4DZoMHQz94ti+AToV53o5arWOUUwGu0sIIYTEmpTxZGigplAZ5JgqLdu2cK6V0LOQtNA/Q29Di9U4uRLGylbBXiyOkmCKJZzmnRBCSLhE2p6lkidDukKEYJmw7DsS8M7gpuXpENih6rV4VbeLQXazBn9Ow/ogfx6TRKKqVuctyc40bfNS6aMmhBCSDKSSJ6NCy4Igx8joEmFjkGPO0iGvj9kGhoDlRkiMjI+hbTWHRsCsaMEUK0ocU7zH2oNCCCEk+Ym0PUslI6NcDY0iNKibPbbrNhkdUo0bUxbkPAdoKd4Mf0zRctdwKxrzKd6ZUpwQQkgcSBkjowmgmAtlQf4ycvbX9+sM5PRHRy3rA+y3tyf8ZCB2PAazfRJCCIkHKWNkKFO1PNbPPnvbey2c41ctDw+wf4SWP7WiXnGhuGJTdwkhhBASa1LNyHhG5xgZh+4RydppgeU9UFyto0vGO7b3hXaA7CRbwlMaRHoeth/nPDnWL1PjYy00wbV3EW1PBo0MQgghcSCljAz0mCxBcQXUHpoBo2AKJN6Nr3T46Vgcs8bxkuc1cdfFjnP8ounJ5d68gdf/AEk5H+v3atzHiThOcmYkNN7hq0wpTgghJA6klJEhoPEXT8VRmrVzODQE+gIagX0vtuIcMj/JW1BvPZ/k1ngO2g37P3Wh6lGnWAM/6ckghBASD1xLniCpuNEY28mvYgquK9k6J4dwnD2SxN++L1GIkhbvNO+FkvqDEEIISR1PxioYGk9C4k1IK+Kd5XOzad7ZXUIIISQObZubRoY8Rp8NTUdFfoNukEBLF69HAk3zTk8GIYSQOOCmkdFDh43KFOu9oH9Bi2BofAidJt0pLl47rkjKDl/FAwZ+EkIIiWfb5pqRgYs3QG9DJ6jB8TfoO+ggDaBM2+6UWFBZW2+q6jzzlhTmSn4yQgghJAVHl8DQKIMegWTSse2ga6HZ0BjtTlkAXQPJNOwkql0lnLeEEEJI+gxhlREt8mido9eXyJF+0O3QHzA0btMZT0kEMOiTEEJIvIlJYw6joSt0KTRT5xf5NzQYelnTdEvyrDOhP6FroLtiUa9UhsNXCSGEpHKejHwN/DxdDYks9VrM1PTfL6P7ZL3jJS/gNZKXYqFOt36VW3VLp+4SJuIihBCSckYGWK1ZMsWwkFTekm3zGRgWcwK9APtk9Im0jp4sUiTy7hLOW0IIISQFjYw2mnXzaehdGBCBpk73AgNDUlP+HQpoiJDQ4DTvhBBCUtnI6A3DQrwZgYyJehnm6tyOdfFgPO5inWKCvyxosc6VwWneCSGERJNwsle7mSdjNSrUGXoYeslnt0yXvgbbx0OFbtUhndnkyciNc00IIYSkK24GfhahmAFtA/3ms1taPgkMPR/aD8cOk1wabtUl1sQrw6ffbJ+MySCEEOJS29aSd8PNIazXq4HxFLSnT0VfQdFF9w2EbnSxHmk+uoQzsBJCCIkPbhoZ0iWyABrrz0uBbdUoLoQWQaNdrEdaWpsl9jTv7C4hhBCSgkZGH+jnpiB9Bxr4+QPU08V6pB2VtQ2muq7R5GVnmjact4QQQkgKGhlrdJ6SlpDp30tdrEfaYcdjFLXNCysamBBCCEl0I2MatBMaOUkX7hfskxlah0CfuFiPtIOJuAghhKR6noz/QCdDT8OYOFwTcy3Tfb2hw6CToCroNhfrkXaUVHjiMTiyhBBCSEoaGYi3kOnbJaDzOehESLwWTsSPXwKdFizVOAnfk8GgT0IIIanqyRBD40MYGttj8RjoAKiHXnMVJJOhvYJjNrhZh3TM+MnJ0QghhESbcGL8XDUytHHdiOJ/KhLLad4R+EkIIYSkYuBnqJZRDnRUvOsRTcRr4atYwsBPQgghidC2uerJUOPhYh2mKqnEM3wMHEkt3knrkeVmXdIJdpcQQghJ9blLDkXxpiy2cKjEZHAIaxThNO+EEEJSvbvkMjUw7ocGQ7dAjdDu0E7QVVAFtA4a42I90nYIaxdOjkYIISRFjYw9oEXos7kcmovlKXq9bbE+G7oHy3/RrpSrXaxHGs5bwiGshBBCUtvI6AD97Fi3c2Hs5mgQ30UhBsgoF+uRdvOW1NQ3mvwcmbfE9cFDhBBCSFyMjHIox2FQVKJYrVO7O5mn3gwSBTjFOyGEkETBzUfdX6A9EQCaBwOjxmFQyFwlTrq5XI+0SsbFKd4JIYQkSjIuNz0ZE6Cu0DRUbF/d9j60BdZv0vwYkmp8GLTAxXqkpSeD85YQQgiJN24aGU9C76kRcaVue1TnK7kZqoZe0e0yAiVliGcyLu/w1UJm+ySEEBLfts01IwMXb4CO1JlWJ+i29SgOhD5TI2MhdBG2RzXlODwkh0AfQ8VQOfQJNDLE1z4LNYWgZ6NZ52jBkSWEEELSIRnXCBTfw4B43bldZ1w90MXrSs6NZyCJA/lYM4nK9aZi3/m4/uMtnOKrIPclS40mMc5+iE6NowuneSeEEJIouBlw+TRUD23t4jWaASOih3bJiMdkmOTj0O0SbPoh9ACW38X2FYHOoUaIX0MEr/2XGhhv4LgHol3/6HaXSBZ3QgghJDVjMiToc5aL5/fHJZAEI9xnGxgClmeiuFvnShkbzolhYEhsyfU6DPe8yKvqDsVMxEUIISQNjIxvdQhrWxev4Yud1EvmTPFlkpaHtfakeA9ynx6GpJQMpqXhVS+G07wz8JMQQkgKd5dcAL0B/YhG+jmUP0GlOn/JZqDhlliIsMnwDOAdqOeXfBy+LNB9g+TYptYN+ThP51v5JtpBqtGmlENYCSGEpIGRId0V0pBn6ORowWiKQl06aVfJWhgCtZtdoKmpXkabaPKvdpqRtEXwGqnXDbr6zwjr6CpiN7G7hBBCSDoYGTJMNXYJIowp1FLSlweiSsu2oRoZ4GSoN/QTGnGZ5C3irGhu5c3YWNtgausbTUFOFuctIYQQEpcsnzExMtCQHuDWuQNgd8MEa8Htu9Wau/Z3Lf/T6hrFGA5fJYQQkkik0pwhFVoWBDlGRpcIG0O04GT47RA93g4cbZFYZvj0l4irqC2HrxJCCImcltqzljwdbibj2q+Vb0S6VyKhXA2NIomjkBgMn/rIey2CqrGvLMRzjtbyHZ1FNqFh0CchhJB08WRMh1rzSC/ZNMNGRovAkJiLxT2h7SFZdtJfh6DK7LChcriWMkom4bFnYO3M4auEEEJS3Mj4OICRIcZERx1uKn79twMMOQ2HqWpkHOvHyJBtRidtC3VI7B66+mVUaucy7C4hhBCSSLgZ+HlIsP1ow9tpCnBJjnVFlC4rc5ZcDY3D+ZwAEiMAACAASURBVN9HHayMo1jeQ7fL6JLxjjr0RdEGKsaxMrzV1/PRHlqBfSujVD9X4TTvhBBC0iXjZ1DQcG9AMUYb/juidM4larCIcTADRsQUSLwbkuhLjJqxOGaN4yXPqxflYj+ns+dcWRSNusV03pK2nOadEEJIGhsZAhr8OhQzoIOieE7xVBwFfQ0N19EhX0AjsO/FVs69IiyPVt1iNs07J0cjhBCSACTCENYtHUNLowKMickoJkeSywP7xMshShqYJ4MQQkgiETcjA90Ykt77Umj3ZAmsDBV/44ZjkTtjU3cJ82QQQgiJf/ZPN/NkbDZ/SJDhqne5VY90QYwYO/CTM7ASQghJdU9GsHM3aeIsyVlxr3ZvpAzxyPhZUVNvahsaTZvcLFMAEUIIIW63bXHL+InKxDWoNN2wu0o6M+iTEEJIghATQwCWjj1Sw17vAUksBokSxd6uEsZjEEIISQMjA4bEXtAcLE702XUg9K3sgwa7WYd0gTkyCCGEpI2RocaDzF8ywM+sp8ugz3TfVzhWShIBHL5KCCEknTwZN0OSevJMxGdI6nAvWP8cOlAzfraFbnKxHmmBNxEXh68SQghJAyNjL2gGjIkXWkh49R0UdJ4T0oruEsZkEEIISQMjowu0IoTjlui8IiQK3SXMkUEIISQd8mSI8TAU8RZZ8Fg0+DsA+8TI2T2Z5gdJ1IyfdndJZ3aXEEIISZCMn256MmRESW/oYVQsO4CBcQ+0FfSmi/VICzZl++QQVkIIIanvyRAD4hToPOhIGBXv66gSo8aHxGH0gZZCt7tYj7TI+MkhrIQQQtIp4+d6XFymcJep14+EzvJz2DToXBxb6lY90uWDZ+AnIYSQtJqFFY2fxFocDWOjJ0qZVr2HXnMV9CX2/+bm9dOFDTpvSWFulsnP4bwlhBBC0muq9zoYFP+zV9ToEIODRIFSjcdg0CchhJBEIl5pxcWrwbTiUaJko2f4audCyX1GCCGEJAZMK55CI0uKOLKEEEJIAsG04ikAp3knhBCSiDCteErNW8LuEkIIIYkD04q7gIwb9pWbMBEXIYSQRGzbMmORVjzQAamaVjx+gZ/M9kkIISRxYFpxl5Jj+So22T5pZBBCCEmcto1pxVOATd0ljMkghBCSODCteAp1l9CTQQghJJGIe1px2QfdhOV/uVmXdJi3hDEZhBBC0i6tOBrClSi8acUFGBajoP9g8QhIgkNpZIRBeXW9qWto4rwlhBBC0nbuEgsYFd1RnK3Tv/eVTbprYSzrkUpwindCCCFpbWTAuBiB4nzoKL2mGBcyvfsr0AvwdHwdi3qkIqUcvkoIISTdjAwYFt1QnKVei61lk+6SMS/HQ5NhXNS5df10odiet4TDVwkhhKS6kQHj4iD1WhwD5cgm6EfoGegMaDcYF5Oifd1Ewl8WNLdyZTDokxBCSCwIJ3t1dhQvfiWKsdA2sqrdIS9DT6OB/UGPGR2t6xEPJRWc5p0QQkjqezLuhqqgl6DXoCkwLuqjeP6kwe0Mn/4mR2N3CSGEkFi3bS15N6KdVjwf2gs6BDpI5yaJObjuIdDHUDFUDn0CjWzlOXKgy6EfoUo9z0caxJowsLuEEEJIohJNI2BH6H6oA3QJNAVagUb5bmhgFK8TFFxrDIpp0D7Qt9AMaF9oKvaNDfEceVr/e6De0AfQXEjiTd7Hfok3SayU4pzmnRBCSKoaGXCjzIGuwGIv6DjoHagzJLEav6BhlgZ/22hdzx+4hmQUfRRaD+2B+hwOjVQjoxx6AMdI/VriOuhgaDrUD+c4FhqK5ZN0/1M4T0xzjLTUXdKFM7ASQghJMKLenYHGuAF6SxpmrEqDLobHbGgPXRdjQLwKYyDxekQT8aCIF+I+XH+2o04zNWZEunOCejNQpwIUl0Fl0El4bbnjPBJr8hYk2wZFue4RBn5yBlZCCCGJhasxE2iUiyFp8HdWI+NhaB10KPQUtBqN+pvQKVG65Cgt3/Szzx42e1gL5xDPRztNErbWdye2HQeJd+On8KsZvSCcdZWct4QQQkhikhnDBvF7SDwN0qVxMvS+zllytI5IiYgMT4irxH40QvP8HLJA9w3SYwOxm5bf4rBc6C/QQ9Aj0GkSEBppXaM9b0nbvGyTnyO3khBCCEkcYh5XoFk+pdvhNZ3L5ExVpHTSrpK1uEatn+vWy2gTLHZTT4W3G8SHbR0G2Hca0GpzAXQlznMYzvdnoIq0NKQnWkNc7a4STvFOCCEkURJwOYnLEFNHY7sKuguKxuiTQi0rgxwjeTyEtkGOseNEHoDaaPeJbNtVA0Gl6+f1FrwhMYHDVwkhhCQyCTFCIkpIV4gQzE1gGwbBDAQJDrXLQ2EALdJ1yZdxOMpfdXjsITpUNm7JuOx5SziyhBBCiBu01J7FOhlXPKnQUkaHtGRAbAxyjL1vqsPAsMC6eEJe0NUDWl1Dt6Z5L5ReIkIIISSxSCUjo1wNjSJ/OSx0WxFUDWNBhqcGwh5RsiTA/j+0lHMlxjTvnIGVEEJIApIyRkaTx6cjWTllmMX2fg7pr+/3lxZOZe/vGWC/BKsKmw1vjTXsLiGEEJLIpIyRoUzVUhKB+WJve6+Fc0g6ceHQAMnC7DlQPm9l3dzrLqEngxBCSAKSakbGM1A1NA4Gwu72RixLIrCrIYmpGO/Y3hfaAfJ2fcAhMh/Fu1BH6FnNAGoff5WmKJc8HB+6/F5apMTuLmFMBiGEkAQkpYwMGAhLNI15e2gGjIIpksIcy19pboyxOGaN4yXPq8Fwsc+pzoPmq/djMc4xCZqtqcklnuN0SZ/u7rtpxeRoTClOCCEkAUkpI0NA4y+eiqOgr6Hh0BDoC2gE9r0Y4jn+1NfdppOtHaaTvcnIkiHYP8uFqrcadpcQQghJZDJildMhHYC3I+DNjPZ9lvNtd/0UU9/YZOb/e5TJy2ZacUIIIe4RLCcG2qSMtPBkpAvlVfWWgdEuL5sGBiGEkIQklTJ+Jgyx8A4VM0cGIYSQOLdt6ZTxM63YlO0zN841IYQQQvxDIyNJsUeWcPgqIYSQRIVGRpJi58goYiIuQgghCQqNjCSl1OvJYHcJIYSQxIRGRpJSojEZNDIIIYQkKjQyktzIKGrLad4JIYQkJjQykhTvNO/sLiGEEJKgME+GC/gbNxzt3BneeUsY+EkIISQGtJQTwx/0ZCR5d0kXzsBKCCEkQaEnIwkzfjY2NnmTcXUqzHH1WoQQQojAjJ9pQnl1nWmQeUvyOW8JIYSQxIXdJUndVcIcGYQQQhIXGhlJyKagTw5fJYQQkrjQyEhCOHyVEEJIMkAjIwkptj0Z7C4hhBCSwNDISOZp3pkjgxBCSAJDIyOJjQxO804IISSRYZ6MJMz4WVzBad4JIYTEFmb8TDtPBoewEkIISVzoyUjCjJ/2EFYaGYQQQmIFM36mCZzmnRBCSDLAwM8kQ+YtWVep85a0YXcJIYSQxIVGRpKxvsozb0n7/GyTm82PjxBCSOLCVipZ5y1hSnFCCCEJDo2MJKNEh68y6JMQQkiiQyMjWbN9cvgqIYSQBIdGRtJ2lzDokxBCSGLDPBlJlvHTO817Iad5J4QQEjuY8TMN4DTvhBBCkgV6MpIs42cxu0sIIYTEAWb83OTSOQT6GCqGyqFPoJGh3kgcmw1VQ00BtNydj7BlStldQgghJElIOU8GDIAxKJ6BZKznx1AWdCA0FfvOhyX2eAinGQhJ0MPv0Nd+9pdGqbqthpOjEUIISRZSysiAEdEDxaPQemgYDIrZun0Iig+hB7D8LravaOFUu2r5DI69zbUKh0HJRk7zTgghJDlItSGsl6gH4j7bwBCwPBPF3VA+NDaE89hGxqyo1zDCeUtsT0Yn5skghBCS4KSakTFKyzf97Juk5WGtMDK+j7hGUaSsqs7AzrDmLcnJSrWPjhBCSKqRMi1Vhic5hcRSNELz/ByyQPcN0mODnWcXaBV0NFa/hTZAa6GXof4uVL9Vw1eLOG8JIYSQJCBljAzQSbtKStA94ulTcIBt9SiKoTZQuyDn6Qe1h7pDj0HV0CdangLNhKGxb3Sr3rpEXJy3hBBCSDKQSkZGoZaVQY6p0rJtCF0lEhy6O4yT/aCjsbw1dI8aKK/A0JD4Dr+IMySYwoUpxQkhhMSSSNuzVDIypCtECJYJy74jwe7MRKgvtCeMix99PCFXazBoL+jY8KsamZHRmSnFCSGEJAGpZGRUaFkQ5Bjb+7Ax0AEwJoRl0Eo/+8SQeU9Xdw9yjqCKdJp3zsBKCCEkFkTanqWSkVGuhkaRZOz03anbiqBq3JiyCK4jAaFGYzviM807Z2AlhBCSBKSMkSHuBxRzNcPn9n4O6a/v95dg54Ex8jdIYi4OCXCIxGYIy+PXXcJp3gkh6c0tt9zSYryArenTp0d8PTmHnOvOO++MQu3Th5TK+AmmQntqvIQYHE7sGAq7uyPY6JKToFrNEupFgz1P1NUPIqppBN0lHMJKCEl3Ro8ebbbddttm204//XSzww47mOuvv77Z9gEDBkR8PTnHCy+8YHbd1R4bQEIhw80ZQ2MNjICtNEeGGAgH4b3N0u17oPgIyoG2wvY1ur2vdnsUY1uxbpNcGz+r1+NEbJ+o2+W146FzoSnYfrif61s30617euh9n5oFqyvMlL8PNwN6yChbQgghjt9gs//++0fFc0FCwx5hgnYvI6W7S/RNLkFxBSQt8Ay8+SmQeDe+0qGnY20DQ3lejZKLHecQD8jluvq6JuN6HcuL1MD4FZJJ2OKWJ4OBn4QQQpKBlDIy1EgQb8NROnvqcEgmR/sCGoF9L4Z4jv/K8dD70HbQkZp/QyZLG+JjqMRs3pJ1lZy3hBBCWsuSJUusJ+477rjDHHPMMSYvL8/07t3brFjhmStzwoQJ5sADDzSdOnUyOTk5pnv37ubUU081ixcvDhqTIetXXnml1Y2y0047mfz8fNOnTx8zbtw4U1Pj6d5Od1ItJsMCRsBkFJNDOO6AIPuke0XUavwlKIm0C8Wet6RDQQ7nLSGEkDC4/fbbzZAhQ8yDDz5oli9fbnr16mX+85//mKuvvtocccQR5rbbbrN+vz///HPL8JgzZ4756aefgp5z4sSJ5qmnnjJ/+9vfTN++fc3LL79s7r77bpOZmWkZNalEOMkkU9LISEW8OTI4fJUQEiJbXfNuQt+rJXceEdPr5ebmmjfeeMN07NjRWm9oaDB33XWXGT58uHnnnXe8jeiFF15o6uvrzWuvvWbmzp1rBg6UUD3/LF261DJEBg8ebK2PGTPG8mY8//zzKWdkhAONDBdwI/DTm1Kcw1cJISQs9t13X6+BIWRlZVldJlVVVc2e0svKykxhoWemiooKO8+jf8S4sA0M25ARo0S8IenQtrXk3aCRkSRwcjRCSKJ7ChKdLbbYYrNtEp/x6aefWl6L+fPnW3EY0pVi09jYGNY5xUtCUjDwM1Wxp3nvwmneCSEkLMRz4cs555xjRo4caWbOnGnlwpBATvFCXHvttSGdU2IvSGDoyUgS2F1CCCHR5YsvvjBPP/20Oeuss6zSicRUkMihCZYkMEcGIYREl+JiKwej2XHHHZttly4TGTUiSAAoCR96MpIEe3K0zuwuIYSQqAWCdu7c2ZoHZd26dWbLLbc08+bNs4akVldXW8esX7+edzsC6MlIEoo5zTshhESVrl27mvfff9+bO+Oyyy4zU6dONRdddJF3dMgHH8R8mqqUIqXmLok3bs5dMuLeT83CNRVm6j+Gmx26c94SQgghiT93CbtLXLzpTiI1PLzdJcyTQQghJEkyfrK7JAloaGwypTpvSec2uXGuDSGEEBIa9GS4QLS7S8pgYMgpO7bJMdlZtAsJIYQkR8ZPtlhJAHNkEEIISUZoZCRVjoy8ONeEEEIICR0aGUkAgz4JIYQkIzQykoAS77wlDPokhBCSPNDISAKYUpwQQkgyQiMjqTwZjMkghBCSPNDISAIYk0EIISQZYZ6MJMj4ye4SQggh8YYZP1M9Twa7SwghhCQR9GQkQcZPdpcQQgiJN8z4maLzlqyrrIWbyphObXLiXR1CCIk7hxxyiOW6nzVrVtDj/vnPf1rHvfrqqyGf+7TTTrNeU1xcbK0/+eST1vrrr7/e4mt79+5tBg8eHPK1fPntt9+C1iUZYeBngrPOnrekgPOWEEKIcNZZZ1nlyy+/HPSGvPTSS6ZTp07mmGOOCfvGHXjggeaFF14we+65p6s3/+677zY77bRTs20XXXSRde127dq5em03oZGR4LCrhBBCmjN69GjTvn1788orrwTsnv7mm2/MwoULzamnnmry8sIf/r/NNttYHoW+ffu6+jF88MEHprbWE39ns88++1jXjqT+8YZGRoJTXMEcGYQQ4qSgoMCcfPLJZvny5ebzzz/3e3NefPFFqzz77LN58+IIjYwk8WR0KWRKcUIIsRkzZoxVTpgwYbObUl9fb3k5pPtht9128/yWlpaacePGmYEDB5o2bdpYhsqgQYPMnXfeaRobGwPeWH8xGXV1dVa8R79+/azzDBkyxHz00Ud+X//111+bE044wfTo0cPk5uZa3TcjRowwn332WbNYDnl9Q0ODdS2JOQkUk7Fhwwbrfci15XxyXjGkli1b1uy6cs5TTjnFvPvuu1ZXj9Rziy22MBdeeKFZv359zL5IHF2S4HhzZHDeEkIIadaV0L9/f/Paa6+Z//73vyY7e1NzNnXqVLN27Vpz/fXXW+vSDbH//vubJUuWWHEO22+/vSkpKTHPPvusufbaay2j49JLLw357h5//PHmnXfeMSeeeKI54IADrADUww8/3DIIOnbs6D3uMxgSYjAMGDDAXHXVVVYXz5w5c8zjjz9uRo4cadVHGv4HH3zQ3HjjjWbevHnmueeeswwHf5SXl5vhw4ebX375xZx++ulm6NChZv78+ebRRx+1jImvvvrK6t6xkfXJkyebsWPHmvPOO886Ro6tqKiwYj1igvRnUdG5B3I7Pbc0etzzwfymLcdNtkpCCGktc/vvEFClE17xHifLwY51sui40QGPW3nDjd7jKn+ZHdL5wuX222+3fnOnTJnSbDu6UprwlN8EQ8NahyFiHQejotlxMDSaYJw07b333t5tf/3rX61j7dc+8cQT1rqcQ5BryfqVV17Z7FwPP/ywtR3eEe+2I444oqlDhw5NZWVlzY594IEHrGMfe+wx77aDDz64KSsrq9lxvnWBQWStP/TQQ82O+/jjj63thx12mHdbr169rG3vv/9+s2MHDx5s3Zvq6upm28PF0e75bRfZXeICYs36KlxK7XlL2F1CCCHNOOOMMwwa5majTKQ74e233zZHHXWUKSoqsrZJd4V0OaDRbvZ68WaId0Ge7EPlrbfessqLL7642XbxFsCg2OzY3377rdn2mpoak5npaXpbc11h4sSJplu3buaCCy7YbASMSIJHxdth07lzZ3PooYc2O3aXXXaxPDvSfRSLto3dJQkOu0sIIZEw4Nd5IR3X6eSTLIXC1m9MDOm4gsGDQr5+OOBp3YpvePPNNw2ezE1+fr4VO1FVVeUd5mqTk5Nj4G2wAkV///13S2KQCIG6J/yxaNEi61xbbrlls+3SXSPdMJWVld5tWTCAVq1aZe644w6ri2Px4sVWF4nEjAjBYkH8Ia+X+Ao5ry8SX/LJJ5+YpUuXenN1SFeML/ZIFYn/iAX0ZLiAP5dRpCnFO9OTQQghfgNA5en9vffe844qEaNh1KhR3mOk4ZVGWAImxQCRp3sxOMRgaI2BIcjTuzTQEvzpi6/R8PTTT5sdd9zRSgbWvXt3K0BTvBvikQiXQN4D22hwDne1PSbxbNtS0pOBD0FCc6+DJLOJDMuQtHB34oa8H8E5H0MxFjoL53k2KhUNgRIdwlrEeUsIIWQzjj32WGvEhjTk++67r5k+fboVZOl82r/pppus4a5ffvmlFTBqI4aCdBtIt0KoSGAl4hzMggULLMPFaWCIp8E2WsSjcTG6VHbeeWcrAFOCS22ef/75sD7JrbbayrquGBS+3oy5c+da3pSePXuGdW63SDlPBowBGdc0DZJv0rfQDGhfaCr2jQ3znGISh/XaSGEyLkIICYw8uctQTfFkyHBWaex9u0okHkOe6mX4qhMZ1SExEnb3RSjIiBLhtttua7b9mWeeaRbnIEZGFbwmYpQ4DQzZNn78eGvZeV0xGlrqPjnuuOPMmjVrrBEiTj799FNL4qEpLCwM+b3EgpTyZMAYEBNS7r4MAh4Gj8Ns3T4ExYfQA1h+F9tXtOKcYuI+7UZ9W6K+odGUVdXpvCXMk0EIIf4Qo+KRRx6xclfYQ1udSBCoDN+URvjMM8+0uhxkmKsMQ5U4jtbkjZAhq3IOGWpaVlZmjjzySMuLIPk0xKNiI0GnQ4cONZMmTbICNSWXxp9//mm9TrwqgvO6Ej8h3Q833HCD9To5ry/XXXedFdR6ySWXmJkzZ1rHiWdD3nuXLl3M/fffn3BfkFTzZFwCSYfUfbaBIWB5Joq7ofwwPBJicnaBvolWJUNlXWWdNW+JGBhZmeGPUCGEkFRGGnDpuli3bt1mXgzh/PPPN/fee6/VqF955ZXm1ltvtYI+JT5C8mZIcKYEZoaKxFrcc889VvDo5ZdfbnXRiBdFAj+dTETshYxoketI14l4O4YNG2YZJV27drVGg9hcc801Ztddd7XmMJFlf8goFel6+cc//mFdU3J7SDeRjLL58ccfzXbbbRfye4gVGZEEJSYasE6/R7ErtAve108++8RPNgeaiX0hzXSD1/wFxf8gyejSDzonWEwGjvcky4jSPZ2/aoMZef9nZttubc2Hl+8flXMSQggh0cIOREW7l5HSngy8UXmDYkhIp5a/MVMLdN8gPbal80n0zMPqwbgrilVtddAnc2QQQghJRlLGyACdtKukBBZV86nsALZJhI0kgJcInFDmzX1Ku1fOxGtjM6A4wPBVphQnhBCSjKRS4KcdUrspE8rmVGnZFtqUFm1zL8aFKGREyWUwMOa3tiItOUpC7U7hyBJCCCHxJJKM1anmybDH/gRrwe27lRHkhsrsMv+BZIq8B6JTtUi7SzYlVyGEEEKShVTyZNhJ4AuCHCPdH8LGAAaGZDd5Xg2VMTr5S6uJVuAnu0sIIYTEk5bas5Y8HalkZJSroVGEN52tMRheZJvsg2TqubIA5zhOk3gtgm71uXlDtRyrGUXfwHneiOYbCDhvCT0ZhBBCkpCUMTLE64DGfy4WZXiqDFaWZSf9tXso2GBoidUwOlxV5I+9Vb9BrhoZ95+yiylGl0lHJuIihBCShKRSTIYwVctj/eyzt3lm0fGD5L8QW8WfdLSJ0TwZsu2WKNbbL/k5WaZ3pzambV7K2IKEEELSiFQzMp6BqqFx8Grsbm/E8h4ortbRJeMd2/tCO0DSjUIIIYSQKJJSRga8C0tQXAG1h2bAeJgCiXfjK82NMRbHrHG85HlN3HVxzCtLCCGEpDgpZWQIMCLEU3EU9DU0HJLJ0b6ARmDfi/GsGyGEEJJOpNTcJfEm2nOXOHLCR+V8xD34WSUP/KySB35WyT93CSMKXbzpTmgoEEIISbfsnynXXUIIIYSQxICeDBdIdK+FGy7IZDmnm+d1g2S5r+n+WSXT+0+Wc7pFMr3/jASrq7/X2OcLBD0ZhBBCCHEFGhmEEEIIoZFBCCGEkOSBngxCCCGEuALzZLiQJ4MQQghJJwLlyaAngxBCCCGuQE8GIYQQQlyBngxCCCGEuAKNDEIIIYTQyEizINJDoI+hYqgc+gQaGe96kebgMzldAn6D6N+8Z/ED93+Mfg7DAuzfHnoZWgZVQj9DF0N8AEugzwrb+rTwdyYzbZMEhGnFE/SPDcUzUA30MZQFHQhNxb7zEcX7eDzrR5qxq5bToDV+7s2PvF/xAX8re6N4MMj+nVF8BrWHvoRm6t+ZvGYodFoMqklC+Kwcf2c/Q7/42T+fNzIxoZGReH9sPVA8Cq2HhsGgmK3bh6D4EHoAy+9i+4o4VpNs/uN3Fj+TxAF/I6NRPAu1DbBfhts9rwbG6fjsXtTtXfXv7K9YnoTtE2NU5bSlpc/K5+/sbnwmL7lfKxIt6BJMPC6B8qD7bANDwLI8Zd0N5UNj41Q3sjm7QKtpYCRMg9UbEuNhonoAVwc4dAS0EzTdNjAELK9FcZGuXupmXdOdVnxWTiNjlusVI1GFRkbiMUrLN/3sm6TlYTGqCwkCfiC3RtER4g9f4iAxMKdD32mXx6+t/TuDofGldn0Nw2fczo1KklZ9VraRUQEt4L1LLmhkJBDqwh0INULz/ByyQPcN0mNJfLGfrlbj43gQ+g2qhuZDN0LidSKxRRqqM6G9YCz467u3GaSl11vop48/U/8eSRw/K/wddUbRV3//Lsf6TxqkuxJ6HOrJDyhxYUxGYtFJu0rW4o+u1ncnttXLaBMsdoPkCas8xvUjAeIxoHXQ55DEyuwB/QsaJaOE8LlV8cbFBtzrO0M8VGKfhD8D7Le3bxFZjUgUPiv772w3aEfoU2g5JHFq50FH4e/sAJyPwZ8JCD0ZiUWhlpVBjrEbrGBBUiQ22D9+r0J98CN3DLS/PiX/BO0DcQhrcv6t8e8s8f7O5kD98Tc2AjoCy9Jd+TLUHWIwaIJCIyOxkK4QIdhEa3Y3CbtL4s8JalDI6ISN9kYsL0ExRj/HsXjKyolT/Uj4f2v8O0sc7oP6QeKtWGxv1L+5c9V7uDv+ziSugyQYNDISCwlsEgqCHGP383sbNRIf8CNXDc0N0LX1o7p0xeO0fcwrRyL9W+PfWYKAv6UGMS6gYj/7KjWXkLB7bGtGQoFGRmJRrj9+RbDKN4uX0W1FkDRuZbGuHGk1q7Rsw3uXcKzUUlzt4cRskMSBf2cJDI2MBAKGg7hu5+qYcX9Pv/31MwsWNU9igAxt1Mj21/0ZhIr0GQtMnJZ42KNKNhs9oiO3doAa9O+R7qiW/AAADgFJREFUxBF8HDfr35kEfQb7OxPPIUkwaGQkHlO1PNbPPnvbezGqCwmMeJyOg46HJNizGfhBHKVep19gO9pPzSQ5/s4kYFcyf36Bz25D7KpEArCT/p2d5OfvTEbaHQrVQZ/wDiYeNDISD5mzpBoahz8gbx8jlmVY5NUa9T4+TnUjzb1OT+jqg86x+ljexvEZcXRJYvKpjlYYgc9LhkEaR1px+7O7Jx4VI5vxmJZX4PPZ196IZYl3elpTwz+JP0m724QkEMyTkWDIyAT88VyBxYehGVj+CKW4bw/Sz+sMHONvIi4Se26FhkMya+R8x0yQB2q+k3vxWcnwVpJg4HNpxOd1Nhbl70u6vc5BKR6nAzRfzRM45p141pF4wOfwAT6fe7F4OfQZliUja7H+7RVpfporeb8SE3oyEvOPSp6kjoK+1j8kSTojDdgI5zwLJL5okq2DoWugJWpc7KOf2/HYL8YiSVDw+XyLYi+dO2M7dbv/AV0AXRjHqhEf9G9Juku+1LwZozQoV7y7B+soE5KAZHi8voQQQggh0YWeDEIIIYS4Ao0MQgghhNDIIIQQQkjyQE8GIYQQQmhkEEIIISR5oCeDEEIIITQyCCGEEJI80JNBCCGEEBoZhBBCCEke6MkghBBCCI0MQgghhCQP9GQQQgghhEYGIST5yADxrkMqkYj3MxHrRBIDejJIUoDfsDFQE7QB2qqFY+vl2FjVLRhSV633b/GuSzQaEnADtBiqgdZARwc5Pgv6Gxbvdblecn/rIzzHdD3PsGjVK9rfGaxvB03F4pZxqppfUKcRKKb4bDtA38OHcaoWSRBoZJBkoy30RLwrkaacAd0K9YCmQV9AC4McfzL0ENTJ/aqlBe9CI+NdCScwInqi+ADaId51IYlJdrwrQEgYHIIft3Obmpqe5N2LKXtqeRvuvRgbifIQMwBqioIB1Qb6I/LqRMwKfU+1SfBQGKhO3+p72BjDupAEhEYGSTaWQ72he8R1jMZO1klsyNNyWSLdcHwHfo3COZZGoy7RAHWpQxHxe4rze6hM9vdAokMiWsaEBON96E2oPfRYqLcKBskS7SPu7Wffk7pvjGPbLbptFDQa+haqhP6EHoLaaMzBNdBC3TcbGhukDttAk6ByaD00GdorwLGdoLv03NXQWuh1aOcA760Y2gX6UeMlFkE7hXBf9tE6yflr9VzjnffJ7l/H4jm66Rm9N88GOe90FC/o6pl6/C2671ld3w96G6qCVvvc/5PFiNS4D6nXOuhT6C+hxGTotu/0Pj4MrdD7OA+6GspuKSbDsa0DdJW+tlrPJecs8lOXbOjv0M/6nZBj74PaSR31vrQqJsNx/7fRQyQmppnnRj4v6BHoD/38V+p93jrA/ZJ7c7Dj+yXvbQvd3xm6FfpBv6tyvuXQC1B/x3lucRicW+p5pzvr7C8mI5TvXASfYyb0D/17LYMq9G/iekg8VSQOFifFe5Dw3wEgDZAsSBdJd6hU18/wc6w0OE0+25bo8b39HC/nlIUxjm3yA9qk/eBSzoTehtbr+mvQK1AV9JHGKDTovosc55EgVVkogVZD66A3oC91uzy1Hu1Tn77Q77pf6j0J+gpqhKqhI/28twp1sy9QI0zKnBbu6UWOOs/Q97RQ19dCu+lx4vZ+EZJGr0nrIutjg5z7esd7/F2PH637xDiRhfl6X+R+LIJ20f3jdf8GSAIdJ0JzdZvoUp9ryX/1frbN19dtUOP0A/u7Af2fz/HSOMrCMD/bJul9+kKX7e/Adz7nyNT9slKmy9P1tfL9kXJ6CN91+zvzm8/93+Coz4uO43fVz6tJvQdyv77Xdfm+DfFzb1bq+X6E3pHPX/dtoZ9Fk34X3tDPoMTxvvrosaN1f5N+/6SO1+u+A3T7h+F85yL4HO9znEve11uOun8CZcT7t6wpzRT3ClC8B601MnT9TF0XY6O7i0aG6GzH9p21sbevPdCx7zzd/qOfBqNJf/iLHPtO0XPJD2I7x/bP9fg7oSzH9hFq1MgPfTc/7+1r27AAmS3cz131x176zA/xaShv1vPJefOC3acWrnGaHv+sz/Zn/TRYVn3BEN230HmvdN9l9r4QjQzRLKinzz2UhXKoMEQjQ+q5l2N7TzUYZWW4Y7t4emThB6irY/sBep+bwjEyHNttI28rx7Zch0F6sc/xZ+j3a7Ec5+feTLQbXcf9f1j33etskOX7qUZWk21I6Pbe9nfF59qbGRlhfudC/hzVOLcNrbaOYzupoSIrB7DNaYppu8vuEpKU4Mv7HIr39AfkERcvJU+rTzuu+5Ojr/kBrMsTls3rWm4b4FzSCBQ7zjVBf+jF7X6ibINLdyiKYWqQXItjGhzHT9PRGh0cXRdOHsExdXqsNC7BuBSSv/9/41ivS1teB/1TG1gZKnlSC+eJhEm41jKf+nbQp+PrnPdKeVxLaUxCZRzOI0/tzns4XxvNUEdEPITXfeM4x0r1JjiDYe17KpyFY9Y6jpd7eUcr6twaxJvQT+qD68h3wwvWn9fvlxgtx/t57X+b7JZ80/0vVs/FLfY+3S9ehJfDuP/R+s6F8jmKh1Moxf4Kx7HizZFuzLPVS0NiCI0Mksycr08yx0ofvkvXEO+AL3YDIq5mJ/LEK+T7ec1S/NhJN4Mvk7UcruWBWk53/sg7EHexsL+ffWIAhcp+Wr4aYP+EINeJFpvVVxof6HhI3OgW+GzzJd5En8yb9Ok9VLzGgYNVWhZG6xyoXzcUO+kTve/3QvC+nyhjf1+kK8BE+n1B3W+GDoPk78roe+sq8RuO72hr7n+0vnOhfI6z1bu4N+r7GXSRHZOC9/Mp9EwiBfimCxxdQpIW/GBIMNpVGgD6IJY/dj5BRolSf5fWssSnPhKkFug84gb2hz06RlzwQh8tL8e5Lg9SL/u4luoaiJ4t1Etc7M6nQzfwW1+87wIUZ0HHQAOhXrJZ73trMks26hO4L/WtfMgqC+Ec9ucRqBELdJ8jxb7uf3Hf/hvCcc57UxYoQBnFxdC+UH8NsnZ+78PN7hnudy6kzxHHSKDtSepxGW4bRdg2X71j4+U3I8y6kzChkUGSGvxoPK4/LPKk9aDGObSWrCD7rO6HKCCxFP7I8LlOlsODIn3tgfBnTLXUReLvuoGw61HTinO2ls3qi89SGqLPIGnoyjTfwmv61P2xNkTBPi8n/jxB4RDKeXJaMFzcSrtt34sPNU4kEM5uvYDvCff/VBTP63kXaibPuRq42qc1I7r8nT7M71xTK34PPpLROVg8CjpCfxfEULoWukQ8MjhGvlMkRtDIIKmABFz+AsmwR9vlGqhB8/ed7+hKrfw/xflip4i2n7D+1PJd/Bj+28X6SP+2uJJF/lKeSz+/aaHhcoPb1MCQOJgL7BgTQYaBtsLAiDXLW4hX8Od5igb29+U53CsZ3RE2uL+STfdRSOKAjsL5fFOFXxLJ+WP1nROPho78ekXrvbN+r8To+Bc0KpLzk9bBmAyS9OBHZbE+qRgNAvX3xFThzxWLH6AsHdHgNgP1Kd0XOyDPzp8gT/HCYTh+s/eBTedrPo4bI6yPfR0r4NQPdvBdi3kdouxJsPOG3O00MJRDnfkQwq6VC6gbXhrOvgHykxwVjcsE+RwP9/cCzXcheSLEEG8J6ZoSQ+57XwPD5/5nhvkZu/qdw3s8UXKLQNc5t2uw9rgIg1ZJmCTUHyohEfCQDrHrHuB7LZ4Ooy7TDEdDdbv2+buNGDPPORMCSWAaisM0lfUkRwCf/CjuA93mTDaE5V21voOgnyOsz4P6xCoTnh3suIZwkwbp/eEITA0HyelhjxgJFTu5U7OJ13TUjXP0hL/g2njzgJaS3K2zvRHLu6G4MQpdOP7u5wQNgPyLTkbnBesyz4nELO2k3R2h3vvBeG0/n4nuJO/JkX7uvV0nSTaWEefv3Dz1gkkyNN8RXnYSt1DuA4kiNDJIStAEdIhaoNgHCYqTJ2Ppc56DH6HX9cnzsiDR7tHkV/WY/C7XliyGmpNAAtpOQfVrHO/jFHWDX6vZHd/STIryGmm8HsRhkmQobPD6WfreJcjyQ5z/K0jugwTJ/VNjPk5S13O42JOnHa2ZPQNmQ3Vwv5Z3a9bG1/RezdC05qtiEJAaLo9ooij7c5asltN0ZERJFGJ87Psp3x9Re/18xAMgo0Ee0gyect0ZOhRV7tmVAUa8NAPH/KldDNJt8gvO8S70phof/3bEdXT3Cd4t1e/lFzj+/nh95/C62ZqMq5v+jX+k90nqfb12w0g+DhJDaGSQlAE/MvIjfFOAfd/ocL9p2j8urt+FGkEfaPhfNPlDo91nq/einwY0SjbGZsNksS4Gya76gylPiiM166Mk6ToB+ns0KoTrPKjDBSUz4vbqPcjQ6+4SaYAcXv+zGkpr9H7vG8JrpqjrX7xS2+rr8rQB39lhEEaj+yGqoO4Neg9v0Pd8uHZByP08XQ+TbKHhcpXeF0mAdRC0lV73c/2+PKn36nDtFpiiSa8ksVaonK0N/jK9huQBWayjTeQa63SIaJHDKD5dDYU9Wvpc3P7OgSsh8RD+qHU/Sj0vD+n53RrlQwJgZ3sjhBASARpguNaZNMqx7zgdRnkX9l/DG03SBXoyCCEkOsjwTpnASzxVXvSp33bTS/cDIWkDPRmEEBKNH9OMjBPtYZPgW03M1VG7iSTgl14MknbQyCCEkGj9oGZk7KMxM3tpgGS5Tpg2PtJgXUKSERoZhBBCCHEFxmQQQgghhEYGIYQQQpIHejIIIYQQQiODEEIIIckDPRmEEEIIoZFBCCGEkOTh/wF3PZr15fhSFQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 576x936 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot results with Matplotlib\n",
    "fig, axs = plt.subplots(2, 1, figsize=(8,13))\n",
    "\n",
    "axs[0].plot(history.history['loss'])\n",
    "axs[0].plot(history.history['val_loss'],ls='--')\n",
    "axs[0].set_xlabel('Number of training iterations')\n",
    "axs[0].set_ylabel('Error')\n",
    "axs[0].legend(['Train', 'Validation'])\n",
    "\n",
    "axs[1].plot(history.history['accuracy'])\n",
    "axs[1].plot(history.history['val_accuracy'],ls='--')\n",
    "axs[1].set_xlabel('Number of training iterations')\n",
    "axs[1].set_ylabel('Accuracy')\n",
    "axs[1].legend(['Train', 'Validation'])\n",
    "\n",
    "# Show figure\n",
    "plt.savefig('LeNet5.png', format=\"png\", dpi=600)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ab9799ae-5a4f-4a12-a73d-5c700d047ac9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 3s 10ms/step - loss: 0.0607 - accuracy: 0.9817\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.060676831752061844, 0.9817000031471252]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Evaluating the already trained CNN model using the test data\n",
    "model.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "49cedf49-8631-4e18-a532-c2868ca9d528",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Reusing TensorBoard on port 6006 (pid 31954), started 1:58:11 ago. (Use '!kill 31954' to kill it.)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "\n",
       "      <iframe id=\"tensorboard-frame-1bf63efc8d4861cc\" width=\"100%\" height=\"800\" frameborder=\"0\">\n",
       "      </iframe>\n",
       "      <script>\n",
       "        (function() {\n",
       "          const frame = document.getElementById(\"tensorboard-frame-1bf63efc8d4861cc\");\n",
       "          const url = new URL(\"/\", window.location);\n",
       "          const port = 6006;\n",
       "          if (port) {\n",
       "            url.port = port;\n",
       "          }\n",
       "          frame.src = url;\n",
       "        })();\n",
       "      </script>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Show data with TensorBoard\n",
    "%tensorboard --logdir logs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "506e56cc-2050-4efc-9593-7400ff3df694",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}