notebooks/4 - tf2_cnn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.7.3 (default, Apr 24 2019, 15:29:51) [MSC v.1915 64 bit (AMD64)]\n",
      "2.0.0-rc1\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "print(sys.version)\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from tensorflow.keras.datasets import mnist\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout, Flatten\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D\n",
    "from tensorflow.keras import backend as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MNIST data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28)\n",
      "(10000, 28, 28)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape)\n",
    "print(x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000,)\n",
      "(10000,)\n"
     ]
    }
   ],
   "source": [
    "print(y_train.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 128\n",
    "num_classes = 10\n",
    "epochs = 12\n",
    "\n",
    "# input image dimensions\n",
    "img_rows, img_cols = 28, 28"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "if K.image_data_format() == 'channels_first':\n",
    "    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)\n",
    "    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)\n",
    "    input_shape = (1, img_rows, img_cols)\n",
    "else:\n",
    "    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)\n",
    "    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)\n",
    "    input_shape = (img_rows, img_cols, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1)\n",
      "(10000, 28, 28, 1)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape)\n",
    "print(x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: (60000, 28, 28, 1)\n",
      "60000 train samples\n",
      "10000 test samples\n"
     ]
    }
   ],
   "source": [
    "x_train = x_train.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "x_train /= 255\n",
    "x_test /= 255\n",
    "print('x_train shape:', x_train.shape)\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_test.shape[0], 'test samples')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 10)\n",
      "(10000, 10)\n"
     ]
    }
   ],
   "source": [
    "y_train = tf.keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = tf.keras.utils.to_categorical(y_test, num_classes)\n",
    "print(y_train.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Conv2D(32, kernel_size=(3, 3),\n",
    "                 activation='relu',\n",
    "                 input_shape=input_shape))\n",
    "model.add(Conv2D(64, (3, 3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(num_classes, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss=tf.keras.losses.categorical_crossentropy,\n",
    "              optimizer=tf.keras.optimizers.Adam(),\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d (Conv2D)              (None, 26, 26, 32)        320       \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 24, 24, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 12, 12, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 12, 12, 64)        0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 9216)              0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 128)               1179776   \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 1,199,882\n",
      "Trainable params: 1,199,882\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/12\n",
      "60000/60000 [==============================] - 17s 276us/sample - loss: 0.2478 - accuracy: 0.9252 - val_loss: 0.0538 - val_accuracy: 0.9821\n",
      "Epoch 2/12\n",
      "60000/60000 [==============================] - 11s 191us/sample - loss: 0.0865 - accuracy: 0.9741 - val_loss: 0.0404 - val_accuracy: 0.9856\n",
      "Epoch 3/12\n",
      "60000/60000 [==============================] - 9s 148us/sample - loss: 0.0670 - accuracy: 0.9800 - val_loss: 0.0332 - val_accuracy: 0.9887\n",
      "Epoch 4/12\n",
      "60000/60000 [==============================] - 12s 195us/sample - loss: 0.0543 - accuracy: 0.9835 - val_loss: 0.0306 - val_accuracy: 0.9905\n",
      "Epoch 5/12\n",
      "60000/60000 [==============================] - 9s 155us/sample - loss: 0.0441 - accuracy: 0.9860 - val_loss: 0.0313 - val_accuracy: 0.9891\n",
      "Epoch 6/12\n",
      "60000/60000 [==============================] - 8s 138us/sample - loss: 0.0409 - accuracy: 0.9870 - val_loss: 0.0265 - val_accuracy: 0.9918\n",
      "Epoch 7/12\n",
      "60000/60000 [==============================] - 10s 164us/sample - loss: 0.0355 - accuracy: 0.9886 - val_loss: 0.0310 - val_accuracy: 0.9892\n",
      "Epoch 8/12\n",
      "60000/60000 [==============================] - 10s 159us/sample - loss: 0.0351 - accuracy: 0.9889 - val_loss: 0.0320 - val_accuracy: 0.99040352 - accuracy: 0.\n",
      "Epoch 9/12\n",
      "60000/60000 [==============================] - 10s 173us/sample - loss: 0.0281 - accuracy: 0.9912 - val_loss: 0.0286 - val_accuracy: 0.9915\n",
      "Epoch 10/12\n",
      "60000/60000 [==============================] - 11s 182us/sample - loss: 0.0269 - accuracy: 0.9911 - val_loss: 0.0272 - val_accuracy: 0.9919\n",
      "Epoch 11/12\n",
      "60000/60000 [==============================] - 10s 171us/sample - loss: 0.0252 - accuracy: 0.9921 - val_loss: 0.0274 - val_accuracy: 0.9920\n",
      "Epoch 12/12\n",
      "60000/60000 [==============================] - 10s 158us/sample - loss: 0.0244 - accuracy: 0.9919 - val_loss: 0.0283 - val_accuracy: 0.9913\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x1b45770a048>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(x_train, y_train,\n",
    "          batch_size=batch_size,\n",
    "          epochs=epochs,\n",
    "          verbose=1,\n",
    "          validation_data=(x_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test loss: 0.028271958214271534\n",
      "Test accuracy: 0.9913\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(x_test, y_test, verbose=0)\n",
    "print('Test loss:', score[0])\n",
    "print('Test accuracy:', score[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## CNN autoencoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras import backend as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_img = Input(shape=(28, 28, 1))  # adapt this if using `channels_first` image data format\n",
    "\n",
    "x = Conv2D(16, (3, 3), activation='relu', padding='same')(input_img)\n",
    "x = MaxPooling2D((2, 2), padding='same')(x)\n",
    "x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)\n",
    "x = MaxPooling2D((2, 2), padding='same')(x)\n",
    "x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)\n",
    "encoded = MaxPooling2D((2, 2), padding='same')(x)\n",
    "\n",
    "# at this point the representation is (4, 4, 8) i.e. 128-dimensional\n",
    "\n",
    "x = Conv2D(8, (3, 3), activation='relu', padding='same')(encoded)\n",
    "x = UpSampling2D((2, 2))(x)\n",
    "x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)\n",
    "x = UpSampling2D((2, 2))(x)\n",
    "x = Conv2D(16, (3, 3), activation='relu')(x)\n",
    "x = UpSampling2D((2, 2))(x)\n",
    "decoded = Conv2D(1, (3, 3), activation='sigmoid', padding='same')(x)\n",
    "\n",
    "autoencoder = Model(input_img, decoded)\n",
    "autoencoder.compile(optimizer='adam', loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_3 (InputLayer)         [(None, 28, 28, 1)]       0         \n",
      "_________________________________________________________________\n",
      "conv2d_16 (Conv2D)           (None, 28, 28, 16)        160       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_7 (MaxPooling2 (None, 14, 14, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_17 (Conv2D)           (None, 14, 14, 8)         1160      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_8 (MaxPooling2 (None, 7, 7, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_18 (Conv2D)           (None, 7, 7, 8)           584       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_9 (MaxPooling2 (None, 4, 4, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_19 (Conv2D)           (None, 4, 4, 8)           584       \n",
      "_________________________________________________________________\n",
      "up_sampling2d_6 (UpSampling2 (None, 8, 8, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_20 (Conv2D)           (None, 8, 8, 8)           584       \n",
      "_________________________________________________________________\n",
      "up_sampling2d_7 (UpSampling2 (None, 16, 16, 8)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_21 (Conv2D)           (None, 14, 14, 16)        1168      \n",
      "_________________________________________________________________\n",
      "up_sampling2d_8 (UpSampling2 (None, 28, 28, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_22 (Conv2D)           (None, 28, 28, 1)         145       \n",
      "=================================================================\n",
      "Total params: 4,385\n",
      "Trainable params: 4,385\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext tensorboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ERROR: Timed out waiting for TensorBoard to start. It may still be running as pid 6788."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%tensorboard --logdir ."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Open the http://localhost:6006/ in browser to use Tensorboard.\n",
    "\n",
    "More on Tensorboard: https://www.tensorflow.org/tensorboard/r2/get_started#using_tensorboard_with_keras_modelfit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/50\n",
      "60000/60000 [==============================] - 9s 152us/sample - loss: 0.2009 - val_loss: 0.1429\n",
      "Epoch 2/50\n",
      "60000/60000 [==============================] - 7s 121us/sample - loss: 0.1339 - val_loss: 0.1254\n",
      "Epoch 3/50\n",
      "60000/60000 [==============================] - 8s 130us/sample - loss: 0.1220 - val_loss: 0.1172\n",
      "Epoch 4/50\n",
      "60000/60000 [==============================] - 8s 141us/sample - loss: 0.1157 - val_loss: 0.1122\n",
      "Epoch 5/50\n",
      "60000/60000 [==============================] - 10s 161us/sample - loss: 0.1115 - val_loss: 0.1089\n",
      "Epoch 6/50\n",
      "60000/60000 [==============================] - 10s 158us/sample - loss: 0.1086 - val_loss: 0.1065\n",
      "Epoch 7/50\n",
      "60000/60000 [==============================] - 9s 153us/sample - loss: 0.1066 - val_loss: 0.1047\n",
      "Epoch 8/50\n",
      "60000/60000 [==============================] - 9s 156us/sample - loss: 0.1048 - val_loss: 0.1030\n",
      "Epoch 9/50\n",
      "60000/60000 [==============================] - 12s 207us/sample - loss: 0.1033 - val_loss: 0.1013\n",
      "Epoch 10/50\n",
      "60000/60000 [==============================] - 8s 140us/sample - loss: 0.1020 - val_loss: 0.1001\n",
      "Epoch 11/50\n",
      "60000/60000 [==============================] - 9s 157us/sample - loss: 0.1008 - val_loss: 0.0989\n",
      "Epoch 12/50\n",
      "60000/60000 [==============================] - 13s 215us/sample - loss: 0.0998 - val_loss: 0.0985\n",
      "Epoch 13/50\n",
      "60000/60000 [==============================] - 12s 195us/sample - loss: 0.0990 - val_loss: 0.0974\n",
      "Epoch 14/50\n",
      "60000/60000 [==============================] - 10s 162us/sample - loss: 0.0983 - val_loss: 0.0969\n",
      "Epoch 15/50\n",
      "60000/60000 [==============================] - 7s 121us/sample - loss: 0.0977 - val_loss: 0.0966\n",
      "Epoch 16/50\n",
      "60000/60000 [==============================] - 7s 114us/sample - loss: 0.0970 - val_loss: 0.0957\n",
      "Epoch 17/50\n",
      "60000/60000 [==============================] - 8s 127us/sample - loss: 0.0966 - val_loss: 0.0954\n",
      "Epoch 18/50\n",
      "60000/60000 [==============================] - 10s 160us/sample - loss: 0.0961 - val_loss: 0.0946\n",
      "Epoch 19/50\n",
      "60000/60000 [==============================] - 10s 171us/sample - loss: 0.0957 - val_loss: 0.0944\n",
      "Epoch 20/50\n",
      "60000/60000 [==============================] - 10s 159us/sample - loss: 0.0952 - val_loss: 0.0938\n",
      "Epoch 21/50\n",
      "60000/60000 [==============================] - 9s 149us/sample - loss: 0.0948 - val_loss: 0.0937\n",
      "Epoch 22/50\n",
      "60000/60000 [==============================] - 10s 167us/sample - loss: 0.0944 - val_loss: 0.0931\n",
      "Epoch 23/50\n",
      "60000/60000 [==============================] - 10s 165us/sample - loss: 0.0941 - val_loss: 0.0928\n",
      "Epoch 24/50\n",
      "60000/60000 [==============================] - 13s 212us/sample - loss: 0.0937 - val_loss: 0.0923ETA: 1 - ETA\n",
      "Epoch 25/50\n",
      "60000/60000 [==============================] - 17s 278us/sample - loss: 0.0934 - val_loss: 0.0922\n",
      "Epoch 26/50\n",
      "60000/60000 [==============================] - 16s 264us/sample - loss: 0.0932 - val_loss: 0.0921\n",
      "Epoch 27/50\n",
      "60000/60000 [==============================] - 16s 263us/sample - loss: 0.0929 - val_loss: 0.0916\n",
      "Epoch 28/50\n",
      "60000/60000 [==============================] - 14s 227us/sample - loss: 0.0926 - val_loss: 0.0917\n",
      "Epoch 29/50\n",
      "60000/60000 [==============================] - 12s 196us/sample - loss: 0.0924 - val_loss: 0.0916\n",
      "Epoch 30/50\n",
      "60000/60000 [==============================] - 14s 233us/sample - loss: 0.0922 - val_loss: 0.0910\n",
      "Epoch 31/50\n",
      "60000/60000 [==============================] - 13s 218us/sample - loss: 0.0920 - val_loss: 0.0911\n",
      "Epoch 32/50\n",
      "60000/60000 [==============================] - 12s 208us/sample - loss: 0.0918 - val_loss: 0.0906\n",
      "Epoch 33/50\n",
      "60000/60000 [==============================] - 12s 203us/sample - loss: 0.0916 - val_loss: 0.0905\n",
      "Epoch 34/50\n",
      "60000/60000 [==============================] - 13s 213us/sample - loss: 0.0914 - val_loss: 0.0902loss: 0.\n",
      "Epoch 35/50\n",
      "60000/60000 [==============================] - 13s 216us/sample - loss: 0.0913 - val_loss: 0.0905\n",
      "Epoch 36/50\n",
      "60000/60000 [==============================] - 11s 179us/sample - loss: 0.0911 - val_loss: 0.0900\n",
      "Epoch 37/50\n",
      "60000/60000 [==============================] - 12s 197us/sample - loss: 0.0910 - val_loss: 0.0899\n",
      "Epoch 38/50\n",
      "60000/60000 [==============================] - 10s 172us/sample - loss: 0.0908 - val_loss: 0.0897\n",
      "Epoch 39/50\n",
      "60000/60000 [==============================] - 10s 162us/sample - loss: 0.0907 - val_loss: 0.0897\n",
      "Epoch 40/50\n",
      "60000/60000 [==============================] - 9s 146us/sample - loss: 0.0906 - val_loss: 0.0894\n",
      "Epoch 41/50\n",
      "60000/60000 [==============================] - 9s 146us/sample - loss: 0.0904 - val_loss: 0.0893\n",
      "Epoch 42/50\n",
      "60000/60000 [==============================] - 8s 141us/sample - loss: 0.0903 - val_loss: 0.0892\n",
      "Epoch 43/50\n",
      "60000/60000 [==============================] - 8s 137us/sample - loss: 0.0902 - val_loss: 0.0893\n",
      "Epoch 44/50\n",
      "60000/60000 [==============================] - 8s 138us/sample - loss: 0.0901 - val_loss: 0.0889\n",
      "Epoch 45/50\n",
      "60000/60000 [==============================] - 10s 166us/sample - loss: 0.0900 - val_loss: 0.0891\n",
      "Epoch 46/50\n",
      "60000/60000 [==============================] - 8s 133us/sample - loss: 0.0899 - val_loss: 0.0888\n",
      "Epoch 47/50\n",
      "60000/60000 [==============================] - 8s 135us/sample - loss: 0.0898 - val_loss: 0.0886oss\n",
      "Epoch 48/50\n",
      "60000/60000 [==============================] - 9s 151us/sample - loss: 0.0897 - val_loss: 0.0888\n",
      "Epoch 49/50\n",
      "60000/60000 [==============================] - 9s 153us/sample - loss: 0.0896 - val_loss: 0.0885 - loss: 0.0\n",
      "Epoch 50/50\n",
      "60000/60000 [==============================] - 8s 138us/sample - loss: 0.0895 - val_loss: 0.0887\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x1b5c0c272e8>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from tensorflow.keras.callbacks import TensorBoard\n",
    "\n",
    "autoencoder.fit(x_train, x_train,\n",
    "                epochs=50,\n",
    "                batch_size=128,\n",
    "                shuffle=True,\n",
    "                validation_data=(x_test, x_test),\n",
    "                callbacks=[TensorBoard(log_dir='.')])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoded_imgs = autoencoder.predict(x_test)\n",
    "\n",
    "n = 10\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(n):\n",
    "    # display original\n",
    "    ax = plt.subplot(2, n, i+1)\n",
    "    plt.imshow(x_test[i].reshape(28, 28))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display reconstruction\n",
    "    ax = plt.subplot(2, n, i + n + 1)\n",
    "    plt.imshow(decoded_imgs[i].reshape(28, 28))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "score = autoencoder.evaluate(x_test, x_test, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.088706576192379"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "cnn_encoder = Model(input_img, encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 28, 28, 1)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoded_imgs = cnn_encoder.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10\n",
    "plt.figure(figsize=(20, 8))\n",
    "for i in range(n):\n",
    "    ax = plt.subplot(1, n, i+1)\n",
    "    plt.imshow(encoded_imgs[i].reshape(4, 4 * 8).T)\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model saving"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder.save(\"cnn_autoencoder.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import load_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "del autoencoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\grigo\\Anaconda3\\envs\\tf2\\lib\\site-packages\\tensorflow_core\\python\\ops\\math_grad.py:1424: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
     ]
    }
   ],
   "source": [
    "restored_model = load_model(\"cnn_autoencoder.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_3 (InputLayer)         [(None, 28, 28, 1)]       0         \n",
      "_________________________________________________________________\n",
      "conv2d_16 (Conv2D)           (None, 28, 28, 16)        160       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_7 (MaxPooling2 (None, 14, 14, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_17 (Conv2D)           (None, 14, 14, 8)         1160      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_8 (MaxPooling2 (None, 7, 7, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_18 (Conv2D)           (None, 7, 7, 8)           584       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_9 (MaxPooling2 (None, 4, 4, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_19 (Conv2D)           (None, 4, 4, 8)           584       \n",
      "_________________________________________________________________\n",
      "up_sampling2d_6 (UpSampling2 (None, 8, 8, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_20 (Conv2D)           (None, 8, 8, 8)           584       \n",
      "_________________________________________________________________\n",
      "up_sampling2d_7 (UpSampling2 (None, 16, 16, 8)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_21 (Conv2D)           (None, 14, 14, 16)        1168      \n",
      "_________________________________________________________________\n",
      "up_sampling2d_8 (UpSampling2 (None, 28, 28, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_22 (Conv2D)           (None, 28, 28, 1)         145       \n",
      "=================================================================\n",
      "Total params: 4,385\n",
      "Trainable params: 4,385\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "restored_model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using trained models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://www.tensorflow.org/api_docs/python/tf/keras/applications"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg19 = tf.keras.applications.vgg19.VGG19(  # 570 Mb\n",
    "    include_top=True,\n",
    "    weights='imagenet', \n",
    "    input_tensor=None,\n",
    "    input_shape=None,\n",
    "    pooling=None,\n",
    "    classes=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "img_path = 'dog.jpg'\n",
    "\n",
    "from tensorflow.keras.preprocessing import image\n",
    "import numpy as np\n",
    "\n",
    "img = image.load_img(img_path, target_size=(224, 224))\n",
    "plt.imshow(img)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'<PIL.Image.Image image mode=RGB size=224x224 at 0x289F505B7F0>'"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(224, 224, 3)\n",
      "(1, 224, 224, 3)\n",
      "(1, 224, 224, 3)\n"
     ]
    }
   ],
   "source": [
    "x = image.img_to_array(img)\n",
    "print(x.shape)\n",
    "x = np.expand_dims(x, axis=0)\n",
    "print(x.shape)\n",
    "x = tf.keras.applications.vgg19.preprocess_input(x)\n",
    "print(x.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 1000)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res=vgg19.predict(x)\n",
    "res.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.10930387e-05, 2.56522253e-06, 1.26345071e-06, 8.49679395e-07,\n",
       "        5.63886033e-06, 3.41120312e-06, 3.44479901e-07, 4.43124190e-06,\n",
       "        4.76909372e-05, 5.05355501e-07, 6.16952889e-07, 8.27683778e-07,\n",
       "        3.51625204e-06, 1.83759653e-06, 1.51337736e-06, 2.18312425e-06,\n",
       "        7.05130844e-07, 2.92731102e-06, 6.97464372e-07, 1.54960310e-06,\n",
       "        7.18754450e-07, 2.18273067e-06, 5.36023344e-06, 1.41509699e-05,\n",
       "        2.98485361e-06, 6.66250025e-06, 9.03315777e-06, 6.51917117e-06,\n",
       "        1.93163578e-06, 1.02317358e-06, 1.90026697e-06, 2.25005806e-06,\n",
       "        4.63641845e-06, 2.71096815e-06, 3.27298130e-06, 9.78269327e-06,\n",
       "        7.49322362e-06, 9.39795245e-06, 1.31981069e-05, 1.70080984e-05,\n",
       "        3.14756435e-05, 7.54196371e-05, 3.69662193e-05, 4.26325423e-05,\n",
       "        9.59702993e-06, 3.17903937e-06, 3.85292151e-05, 7.99869940e-06,\n",
       "        5.87192881e-05, 4.40627809e-06, 4.42231294e-06, 5.36274410e-06,\n",
       "        8.77452931e-06, 2.05343463e-06, 3.23875202e-06, 2.74551530e-06,\n",
       "        1.26230179e-06, 2.44622152e-06, 4.44865282e-06, 1.41511864e-05,\n",
       "        1.39089043e-05, 5.47111767e-06, 2.38445245e-05, 1.05393710e-05,\n",
       "        1.17179125e-05, 2.38143548e-06, 5.81023824e-06, 6.39014706e-06,\n",
       "        2.13181215e-06, 2.95719275e-07, 3.35019735e-07, 1.58665944e-06,\n",
       "        3.56349176e-07, 1.66777409e-06, 2.42114282e-07, 3.60977737e-07,\n",
       "        1.93366327e-06, 2.56571661e-06, 4.02816113e-07, 3.25988231e-06,\n",
       "        1.17450713e-06, 8.67110714e-07, 1.51707027e-05, 2.51596407e-06,\n",
       "        2.56724138e-06, 3.44565137e-06, 9.14016073e-06, 2.04352045e-06,\n",
       "        3.31429510e-06, 3.68101769e-06, 1.38768212e-06, 5.03814454e-06,\n",
       "        2.04663115e-06, 2.22483254e-06, 6.45397085e-06, 1.77425392e-07,\n",
       "        5.26571921e-06, 7.60845160e-06, 1.36018662e-06, 2.65423569e-05,\n",
       "        1.60100205e-06, 8.56652696e-06, 1.58359016e-05, 7.72579187e-06,\n",
       "        1.07298929e-05, 6.46909598e-07, 1.43028944e-06, 1.85817282e-06,\n",
       "        1.64158098e-06, 4.62635455e-07, 1.56249564e-06, 6.02869068e-06,\n",
       "        6.48139985e-06, 1.46051059e-06, 2.39372184e-06, 3.76369189e-06,\n",
       "        1.05703782e-06, 2.09500669e-07, 1.05352956e-06, 2.68239614e-06,\n",
       "        1.15474541e-06, 2.83984332e-06, 3.86626027e-07, 8.01211627e-07,\n",
       "        2.73543492e-06, 1.96307451e-06, 2.14713100e-06, 9.79528704e-06,\n",
       "        1.92968969e-06, 8.34943057e-06, 1.23430891e-06, 1.94577410e-06,\n",
       "        3.89642219e-06, 2.47907628e-05, 6.79900722e-06, 1.82317797e-06,\n",
       "        5.11016287e-06, 2.11903193e-06, 1.23369932e-06, 5.42026669e-07,\n",
       "        1.47535945e-06, 6.14328155e-06, 6.95579843e-07, 5.46342199e-06,\n",
       "        4.67743303e-06, 8.02566774e-07, 1.16721285e-05, 3.91954586e-07,\n",
       "        2.09438539e-07, 1.41982468e-06, 2.73452879e-05, 2.33816914e-04,\n",
       "        2.77290674e-05, 3.01919590e-05, 3.38707068e-05, 2.66022416e-05,\n",
       "        1.97014728e-04, 1.05728932e-05, 7.12966648e-05, 3.21906758e-04,\n",
       "        3.40088550e-03, 9.07387643e-04, 1.52647018e-03, 2.56117713e-03,\n",
       "        7.62504642e-05, 1.51772940e-04, 4.83940064e-04, 6.71924266e-04,\n",
       "        2.94499286e-03, 6.85187915e-05, 1.58279057e-04, 4.60941301e-05,\n",
       "        9.77690652e-05, 7.27547857e-04, 1.64590529e-05, 1.41055568e-03,\n",
       "        3.42827998e-02, 2.99946514e-05, 1.15542125e-03, 1.17792115e-05,\n",
       "        3.94811068e-05, 2.44117167e-04, 1.82570977e-04, 3.86488773e-06,\n",
       "        3.20549501e-04, 2.72899546e-04, 2.12164887e-04, 1.66295467e-05,\n",
       "        4.49769213e-05, 1.84349279e-04, 3.11879267e-05, 1.17871285e-04,\n",
       "        9.14171687e-05, 6.34460739e-05, 7.47776125e-04, 4.26064662e-06,\n",
       "        3.76485332e-05, 6.01387765e-05, 2.83947102e-05, 5.15422980e-05,\n",
       "        2.32758335e-04, 2.42282676e-05, 1.48143343e-04, 1.88770227e-05,\n",
       "        3.96919968e-05, 2.19411985e-03, 8.55102786e-04, 7.84063756e-01,\n",
       "        2.83820834e-02, 7.92817737e-04, 3.06559195e-05, 3.29017127e-03,\n",
       "        1.68112654e-03, 4.81322082e-03, 1.05045806e-03, 1.02854811e-03,\n",
       "        6.24197442e-03, 1.07977205e-04, 7.49712577e-04, 3.83924246e-02,\n",
       "        1.05969859e-02, 1.56471302e-04, 1.35854690e-03, 9.21584615e-06,\n",
       "        2.30446603e-05, 8.12654034e-05, 5.48899407e-05, 1.20857754e-03,\n",
       "        1.20218778e-04, 1.30217841e-05, 4.38699644e-05, 1.60557087e-04,\n",
       "        1.50014166e-04, 2.23031566e-06, 5.16215005e-05, 4.86160206e-05,\n",
       "        8.16554602e-05, 2.27053497e-05, 4.12813708e-04, 5.10831851e-05,\n",
       "        2.78610212e-04, 1.99176691e-04, 1.56128699e-05, 3.42502026e-05,\n",
       "        3.67390196e-04, 5.00358738e-06, 5.57183484e-05, 4.63402539e-05,\n",
       "        2.23132854e-04, 3.18201637e-05, 5.91326061e-05, 2.99340409e-05,\n",
       "        4.17695537e-06, 1.51083223e-04, 6.70794498e-06, 7.53675777e-05,\n",
       "        1.36776245e-04, 2.42571798e-04, 1.12833304e-05, 2.16734279e-05,\n",
       "        7.29835883e-05, 7.10877430e-07, 4.23579368e-05, 3.38348880e-04,\n",
       "        1.85359779e-04, 8.47325500e-05, 3.09807569e-04, 3.59870633e-03,\n",
       "        7.37732580e-06, 1.17786672e-04, 2.90562224e-04, 1.39228068e-04,\n",
       "        2.31788785e-04, 3.34151685e-02, 5.63599519e-04, 5.85487078e-06,\n",
       "        1.54816953e-05, 4.70280502e-05, 4.52934219e-05, 3.74492520e-05,\n",
       "        2.26173033e-05, 1.39051781e-05, 5.34081555e-05, 2.12553914e-06,\n",
       "        5.06125525e-06, 8.98408416e-06, 7.75185530e-04, 3.56497767e-05,\n",
       "        6.16405077e-06, 4.16867152e-07, 7.91380171e-06, 5.13633189e-04,\n",
       "        2.19179728e-05, 1.08519316e-05, 1.76969712e-04, 1.28724823e-05,\n",
       "        1.36626986e-04, 2.46616855e-05, 3.01182095e-04, 2.97430292e-04,\n",
       "        1.21324592e-06, 2.12889449e-06, 1.30760861e-06, 2.62675741e-07,\n",
       "        7.38428298e-07, 1.00115278e-06, 9.52226799e-07, 6.00767578e-07,\n",
       "        1.80284758e-07, 3.74076677e-07, 3.07088385e-06, 7.46504156e-06,\n",
       "        9.45311785e-06, 1.14520799e-05, 5.42681209e-06, 1.16122737e-05,\n",
       "        1.48725206e-07, 2.01179250e-06, 5.02204216e-07, 7.54681309e-07,\n",
       "        4.99762336e-07, 1.74614698e-07, 1.20714049e-06, 7.20594528e-07,\n",
       "        8.41473422e-07, 2.96156031e-06, 3.84093539e-07, 1.81108462e-05,\n",
       "        1.01716785e-06, 2.32370297e-07, 9.97739062e-06, 1.71119827e-05,\n",
       "        1.90441267e-06, 1.33249625e-06, 1.28054944e-05, 4.73683167e-06,\n",
       "        7.40304313e-06, 7.04289732e-06, 3.19996025e-06, 1.72917462e-05,\n",
       "        2.46621630e-06, 2.92921723e-05, 1.42853833e-05, 1.03031725e-05,\n",
       "        1.14401871e-06, 4.88234800e-05, 1.00963198e-05, 1.40091925e-06,\n",
       "        2.29784928e-05, 3.92513466e-05, 9.02534375e-05, 7.80803839e-06,\n",
       "        9.11616517e-05, 1.20764264e-04, 1.05288818e-05, 1.64401499e-05,\n",
       "        1.61421922e-04, 2.13389612e-05, 1.65856436e-05, 2.30579153e-05,\n",
       "        4.12012923e-05, 1.43871785e-06, 8.42730788e-06, 3.57148565e-06,\n",
       "        1.91216095e-05, 1.62595770e-05, 3.92243828e-06, 1.63173536e-06,\n",
       "        2.55561667e-04, 2.58583641e-05, 9.00934356e-06, 1.58759343e-04,\n",
       "        3.40340011e-05, 5.31665610e-05, 9.22819891e-05, 6.59194257e-06,\n",
       "        2.56823969e-05, 3.62780065e-06, 1.99663064e-05, 2.06261320e-05,\n",
       "        2.36805663e-05, 2.53011349e-05, 4.23066194e-05, 3.08339668e-06,\n",
       "        2.51711367e-06, 3.28501164e-06, 3.41852901e-05, 2.70474611e-05,\n",
       "        1.06347015e-05, 4.73915816e-06, 1.91702770e-06, 9.33710112e-07,\n",
       "        5.55009592e-07, 2.59479833e-07, 1.33053763e-05, 1.40233160e-05,\n",
       "        3.08269847e-07, 6.05094954e-07, 1.31370700e-06, 4.90454022e-06,\n",
       "        8.15864155e-07, 6.10316465e-06, 1.26020841e-05, 2.35000243e-07,\n",
       "        4.09120929e-07, 1.81800056e-06, 1.72086425e-06, 4.19973503e-06,\n",
       "        3.50985306e-06, 1.63102050e-05, 7.57306725e-06, 5.16347463e-05,\n",
       "        4.91989449e-05, 2.33258311e-06, 8.19097568e-06, 3.77883725e-06,\n",
       "        2.64693299e-05, 1.79824133e-06, 7.17547721e-07, 1.77009133e-05,\n",
       "        7.76440720e-06, 5.58182501e-06, 2.73850969e-06, 1.58923922e-05,\n",
       "        3.54073291e-06, 1.27175299e-05, 2.09964628e-06, 1.52058701e-05,\n",
       "        3.64513427e-04, 6.80140056e-06, 1.95380699e-06, 2.00898849e-05,\n",
       "        2.18681862e-06, 5.88011835e-06, 8.91575619e-05, 2.07508474e-05,\n",
       "        6.87840802e-05, 3.62688047e-06, 1.15476962e-06, 1.27163685e-06,\n",
       "        1.05713407e-05, 1.39981216e-06, 4.58586328e-05, 7.10018730e-06,\n",
       "        9.99200620e-06, 2.41746766e-05, 5.55132374e-07, 5.59188948e-06,\n",
       "        2.05845899e-05, 2.35033422e-05, 9.31420004e-07, 2.81050552e-06,\n",
       "        1.64506728e-05, 1.94732047e-06, 2.16719081e-06, 9.50629726e-07,\n",
       "        1.48428526e-05, 1.67650251e-05, 8.49892615e-07, 2.20390075e-06,\n",
       "        2.11002794e-06, 2.54948463e-06, 1.58130744e-04, 1.68772298e-04,\n",
       "        5.27898828e-06, 2.00297291e-06, 1.19750595e-07, 3.58970647e-06,\n",
       "        1.41983237e-05, 1.02972590e-05, 1.59025740e-06, 9.80959499e-07,\n",
       "        3.14883597e-04, 2.91731703e-06, 2.86183808e-06, 2.49397658e-06,\n",
       "        1.61879791e-06, 1.50546600e-06, 2.21753398e-05, 6.19263665e-05,\n",
       "        1.33832782e-06, 1.37473648e-06, 2.06610071e-06, 1.35106063e-06,\n",
       "        5.69079532e-07, 1.14987608e-06, 1.84544513e-06, 4.35199581e-06,\n",
       "        1.03283182e-05, 1.17920135e-05, 8.38584697e-07, 2.62331050e-05,\n",
       "        2.45940410e-05, 4.05599894e-06, 9.45765350e-06, 1.04462526e-06,\n",
       "        2.88703291e-06, 6.57539158e-06, 9.89341743e-07, 1.44396188e-06,\n",
       "        3.42170970e-05, 1.63986697e-05, 4.53279063e-05, 1.25479312e-06,\n",
       "        1.21958794e-06, 1.91272352e-06, 6.59259558e-06, 8.99719794e-07,\n",
       "        3.27334055e-06, 1.15846376e-06, 1.30167663e-07, 3.45648477e-05,\n",
       "        1.13930867e-06, 1.26643033e-06, 1.79460447e-04, 2.64734954e-05,\n",
       "        4.27935629e-05, 2.72496027e-06, 5.25981659e-06, 2.66944917e-05,\n",
       "        3.42070916e-06, 3.78438722e-06, 4.97068504e-06, 2.87170387e-05,\n",
       "        1.09270934e-06, 1.75778155e-06, 2.28210229e-06, 5.41862278e-07,\n",
       "        2.38491748e-06, 1.28893371e-05, 3.08588255e-06, 3.17985473e-06,\n",
       "        8.66198661e-06, 2.31777926e-06, 3.26289482e-05, 1.37334371e-06,\n",
       "        2.30720480e-06, 1.97424833e-05, 6.44309739e-06, 3.19523067e-04,\n",
       "        1.74333752e-06, 8.29675719e-06, 3.16156979e-06, 1.01365704e-05,\n",
       "        5.37307460e-06, 6.31607918e-06, 1.36406811e-06, 6.18839692e-07,\n",
       "        1.51718370e-06, 7.34042487e-06, 1.98658040e-06, 6.81647180e-06,\n",
       "        1.67552753e-05, 5.81138511e-06, 4.76321588e-07, 9.13771964e-06,\n",
       "        1.57098077e-06, 1.49304969e-06, 5.88213743e-06, 1.20519435e-04,\n",
       "        2.10408621e-06, 7.35600622e-07, 1.34797529e-05, 1.38092673e-06,\n",
       "        9.01085116e-07, 5.95986592e-07, 1.84213070e-06, 1.90352384e-05,\n",
       "        2.62096983e-05, 9.05965578e-07, 6.79653795e-06, 1.00682346e-05,\n",
       "        7.43735995e-07, 9.23985738e-07, 1.01534633e-05, 1.14430413e-05,\n",
       "        3.58310018e-07, 2.18129298e-06, 3.61431084e-05, 4.28776957e-06,\n",
       "        3.47773675e-06, 2.04088428e-05, 7.33341449e-06, 1.90823243e-06,\n",
       "        1.28320528e-06, 1.47864716e-06, 3.75771697e-06, 1.19483102e-05,\n",
       "        3.37895654e-05, 4.25647813e-05, 3.15136361e-07, 4.35650963e-06,\n",
       "        1.61738029e-07, 1.56670121e-05, 2.08930533e-06, 6.58012777e-06,\n",
       "        3.79245685e-05, 2.11337510e-06, 2.91671086e-06, 3.57826138e-05,\n",
       "        2.79924939e-06, 1.80434236e-05, 1.06633306e-05, 6.62009006e-06,\n",
       "        8.63473190e-07, 1.32367597e-06, 7.07963227e-06, 5.10358313e-06,\n",
       "        1.25785018e-05, 4.48234969e-05, 3.08200506e-05, 1.64343408e-06,\n",
       "        6.27696409e-06, 1.31257502e-06, 2.50959442e-06, 1.41745852e-06,\n",
       "        5.75956910e-05, 4.97459496e-07, 1.42677945e-05, 3.55329939e-06,\n",
       "        6.19093908e-06, 3.94887138e-05, 1.70670728e-06, 2.70174382e-06,\n",
       "        4.77607102e-07, 1.21356993e-06, 8.99354291e-07, 7.27122824e-07,\n",
       "        1.22079868e-07, 2.95018867e-06, 4.35037282e-06, 4.66973097e-06,\n",
       "        8.07672052e-07, 8.04403953e-06, 4.77483663e-05, 9.37000834e-07,\n",
       "        1.13172673e-05, 1.11062382e-05, 3.01626897e-05, 3.40054467e-05,\n",
       "        2.00711270e-06, 4.17656520e-06, 8.36399465e-07, 3.25740257e-06,\n",
       "        1.00927434e-06, 6.10558288e-07, 1.64397734e-05, 2.21892924e-06,\n",
       "        6.14933413e-07, 5.36958123e-06, 1.08738061e-06, 9.61806745e-06,\n",
       "        1.99952615e-06, 8.23614118e-06, 7.44421823e-06, 9.75564253e-05,\n",
       "        5.16410228e-06, 5.72179829e-07, 2.15165687e-06, 4.23569463e-06,\n",
       "        6.90846082e-06, 1.61637035e-05, 3.31347508e-07, 6.99239581e-06,\n",
       "        1.12695159e-06, 2.25454642e-05, 8.59885040e-06, 4.92776644e-06,\n",
       "        5.75232889e-07, 2.39920018e-06, 1.19143679e-05, 4.53597568e-06,\n",
       "        5.82928405e-05, 1.27527414e-06, 5.24379948e-06, 1.80589416e-06,\n",
       "        8.26101095e-05, 1.80554798e-06, 9.37625191e-06, 4.71089834e-07,\n",
       "        7.80889513e-06, 3.79206449e-06, 2.19752906e-06, 4.03390686e-06,\n",
       "        4.03710510e-06, 7.84385179e-07, 3.08374069e-06, 5.87025397e-07,\n",
       "        3.50553933e-07, 8.92556000e-06, 1.90041501e-05, 1.57188470e-06,\n",
       "        1.21036655e-05, 9.68961613e-05, 8.77748676e-07, 1.90604419e-06,\n",
       "        1.28536167e-05, 1.41731343e-05, 2.39914084e-06, 5.93525192e-06,\n",
       "        4.17158335e-05, 4.07779794e-07, 1.44574296e-05, 5.69416770e-05,\n",
       "        6.59783836e-07, 5.49861579e-06, 7.43924375e-06, 1.17654645e-06,\n",
       "        2.12998020e-05, 1.21504763e-06, 7.93862546e-06, 2.63603647e-06,\n",
       "        1.65564154e-06, 3.28896590e-07, 1.48420793e-06, 7.06584569e-06,\n",
       "        7.60900321e-06, 8.54216414e-05, 4.05690071e-06, 4.42693135e-07,\n",
       "        3.12881843e-06, 1.26832874e-05, 2.49991190e-05, 8.42988413e-07,\n",
       "        3.95171764e-06, 1.02153956e-06, 1.94167001e-06, 3.76104168e-07,\n",
       "        4.50428124e-05, 1.10654560e-06, 1.97885147e-05, 7.82904408e-06,\n",
       "        3.74348451e-06, 1.72019486e-06, 6.47128616e-07, 1.61863102e-06,\n",
       "        7.04608237e-06, 6.49306185e-06, 1.25388106e-05, 4.90371021e-05,\n",
       "        1.84625651e-06, 5.26180338e-06, 2.60440902e-06, 2.60916227e-06,\n",
       "        1.15615251e-06, 1.30859655e-06, 3.06945651e-07, 6.33272293e-07,\n",
       "        7.01066392e-06, 5.31698333e-06, 3.69820118e-05, 1.00166608e-05,\n",
       "        1.87644819e-05, 2.63534503e-06, 4.21007690e-06, 7.31043713e-07,\n",
       "        9.96864037e-06, 1.78102309e-05, 7.26532699e-06, 2.35049086e-07,\n",
       "        5.95551091e-06, 1.67552673e-06, 4.89626336e-06, 1.73426190e-06,\n",
       "        8.03783587e-06, 1.94070555e-04, 1.78330833e-06, 1.35403388e-06,\n",
       "        3.72761338e-06, 4.47187813e-06, 1.72801811e-06, 2.57724037e-06,\n",
       "        4.24777511e-07, 3.11788972e-06, 4.22592984e-06, 1.29273267e-05,\n",
       "        3.11951067e-06, 3.25228416e-06, 1.97327699e-05, 1.65276488e-05,\n",
       "        4.80700419e-06, 6.13768691e-07, 1.12141588e-05, 5.24533220e-07,\n",
       "        1.42032047e-06, 7.73000647e-06, 9.40601331e-06, 1.76098308e-06,\n",
       "        6.26976941e-07, 4.25680491e-06, 1.15244784e-05, 9.40813970e-06,\n",
       "        3.76301177e-05, 6.30714703e-06, 9.43167208e-07, 6.33794912e-07,\n",
       "        2.43017985e-06, 2.04150328e-05, 2.35782579e-07, 4.01591205e-06,\n",
       "        2.46869195e-06, 4.42962147e-07, 3.52878260e-06, 4.20043170e-06,\n",
       "        5.98190582e-06, 1.38785297e-04, 6.21784920e-06, 7.99142128e-07,\n",
       "        8.58538333e-05, 2.66589109e-06, 7.83712210e-07, 2.21712799e-05,\n",
       "        2.07972633e-07, 1.15776411e-05, 4.32185828e-07, 7.81108611e-06,\n",
       "        4.47729946e-07, 7.95384221e-06, 5.86129545e-06, 1.70691226e-06,\n",
       "        7.91600201e-07, 2.87961416e-06, 4.13074360e-07, 3.16367800e-06,\n",
       "        3.82599592e-06, 5.07475743e-05, 9.10724907e-07, 5.44074974e-05,\n",
       "        1.71506872e-05, 1.62133313e-06, 5.39086432e-06, 7.47563899e-05,\n",
       "        2.47886260e-06, 2.50803822e-07, 3.11994245e-06, 6.53083907e-06,\n",
       "        1.16409246e-05, 1.24626440e-05, 4.36144364e-06, 1.65458539e-06,\n",
       "        8.75801416e-05, 1.82154890e-05, 1.30309536e-05, 3.90772184e-05,\n",
       "        1.32527759e-06, 7.33167963e-06, 6.99232851e-06, 8.88908801e-07,\n",
       "        3.90659852e-06, 1.53728297e-06, 3.48045251e-05, 6.77341131e-06,\n",
       "        7.19879579e-04, 3.04918813e-05, 1.37583959e-06, 5.97773169e-06,\n",
       "        7.89133537e-06, 1.58088751e-05, 2.96839626e-05, 1.66467385e-06,\n",
       "        6.30939076e-06, 2.00520044e-05, 1.36396932e-06, 3.57801423e-06,\n",
       "        8.73356203e-06, 1.22310739e-05, 9.53552171e-06, 3.63591585e-06,\n",
       "        4.52680724e-05, 1.06417328e-05, 2.91631324e-04, 1.11389772e-06,\n",
       "        4.06165782e-06, 1.45612262e-06, 5.69610620e-07, 1.19728281e-06,\n",
       "        3.11086274e-04, 1.36494793e-06, 1.31687545e-06, 1.99143160e-05,\n",
       "        1.92653315e-05, 1.25308961e-05, 1.33726571e-05, 1.06212667e-06,\n",
       "        1.28466309e-05, 5.12222250e-05, 1.10862743e-06, 1.78805772e-06,\n",
       "        2.62334902e-06, 3.42490694e-06, 2.67260634e-06, 1.26229213e-06,\n",
       "        5.69757049e-06, 3.86865258e-06, 1.05514027e-06, 9.28451811e-07,\n",
       "        8.61346871e-06, 1.10435540e-05, 2.16444696e-05, 4.48295441e-06,\n",
       "        1.19124525e-06, 7.41541157e-07, 6.30457271e-06, 4.01168036e-05,\n",
       "        1.23438958e-05, 3.66564223e-06, 4.35453416e-07, 5.42772295e-06,\n",
       "        1.54809629e-06, 3.49218726e-06, 2.53336584e-05, 5.09489291e-06,\n",
       "        9.71289774e-06, 1.46700438e-06, 5.60565695e-07, 4.07693515e-05,\n",
       "        8.78393985e-06, 1.53247152e-06, 1.51930192e-06, 1.22984159e-06,\n",
       "        2.65496482e-07, 5.72527824e-06, 7.71620023e-07, 1.17575780e-06,\n",
       "        5.04890340e-06, 1.41397652e-06, 4.96138318e-06, 2.79341549e-07,\n",
       "        2.67882626e-07, 4.60224874e-05, 1.26431933e-05, 2.00334471e-05,\n",
       "        1.16472329e-05, 3.35005143e-06, 2.03800664e-06, 4.83102383e-07,\n",
       "        1.44942817e-06, 2.82905130e-05, 3.76255412e-06, 4.89722561e-06,\n",
       "        3.94224708e-06, 5.33307400e-07, 1.27650865e-05, 1.17935087e-05,\n",
       "        5.23259189e-07, 9.02078170e-07, 4.17165808e-07, 2.05728458e-04,\n",
       "        1.99915621e-06, 5.06420236e-07, 1.30109777e-06, 3.72998534e-06,\n",
       "        9.71776046e-08, 1.29920090e-06, 4.45056685e-06, 2.63412039e-05,\n",
       "        1.05597883e-06, 1.05627841e-07, 2.04276530e-05, 9.74569275e-07,\n",
       "        3.32665621e-07, 1.66591035e-05, 6.77322760e-07, 7.57825717e-07,\n",
       "        5.93531186e-05, 2.84172324e-06, 1.04403864e-06, 1.04681033e-06,\n",
       "        4.72753482e-06, 5.44919521e-06, 1.34520701e-06, 1.01055575e-05,\n",
       "        1.66677473e-05, 1.97940358e-06, 1.71717886e-06, 6.76930140e-05,\n",
       "        2.61354808e-06, 3.78394179e-05, 1.41365781e-05, 5.51479570e-06,\n",
       "        1.86759212e-06, 1.07814708e-06, 8.87328770e-06, 2.46865994e-07,\n",
       "        1.23773625e-05, 1.44746105e-06, 1.90245942e-06, 2.77419196e-04,\n",
       "        4.22726453e-06, 7.22088714e-07, 4.29058809e-06, 2.92609584e-05,\n",
       "        7.99162735e-05, 9.03017735e-06, 1.23807367e-05, 6.96553689e-05,\n",
       "        8.76179765e-05, 5.03619158e-05, 2.24509393e-04, 1.85757835e-05]],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([207], dtype=int64)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(res, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[('n02099601', 'golden_retriever', 0.78406376),\n",
       "  ('n02102318', 'cocker_spaniel', 0.038392425),\n",
       "  ('n02091831', 'Saluki', 0.0342828),\n",
       "  ('n02115641', 'dingo', 0.03341517),\n",
       "  ('n02099712', 'Labrador_retriever', 0.028382083)]]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.applications.vgg19.decode_predictions(res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"vgg19\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 224, 224, 3)]     0         \n",
      "_________________________________________________________________\n",
      "block1_conv1 (Conv2D)        (None, 224, 224, 64)      1792      \n",
      "_________________________________________________________________\n",
      "block1_conv2 (Conv2D)        (None, 224, 224, 64)      36928     \n",
      "_________________________________________________________________\n",
      "block1_pool (MaxPooling2D)   (None, 112, 112, 64)      0         \n",
      "_________________________________________________________________\n",
      "block2_conv1 (Conv2D)        (None, 112, 112, 128)     73856     \n",
      "_________________________________________________________________\n",
      "block2_conv2 (Conv2D)        (None, 112, 112, 128)     147584    \n",
      "_________________________________________________________________\n",
      "block2_pool (MaxPooling2D)   (None, 56, 56, 128)       0         \n",
      "_________________________________________________________________\n",
      "block3_conv1 (Conv2D)        (None, 56, 56, 256)       295168    \n",
      "_________________________________________________________________\n",
      "block3_conv2 (Conv2D)        (None, 56, 56, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv3 (Conv2D)        (None, 56, 56, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv4 (Conv2D)        (None, 56, 56, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_pool (MaxPooling2D)   (None, 28, 28, 256)       0         \n",
      "_________________________________________________________________\n",
      "block4_conv1 (Conv2D)        (None, 28, 28, 512)       1180160   \n",
      "_________________________________________________________________\n",
      "block4_conv2 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv3 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv4 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_pool (MaxPooling2D)   (None, 14, 14, 512)       0         \n",
      "_________________________________________________________________\n",
      "block5_conv1 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv2 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv3 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv4 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block5_pool (MaxPooling2D)   (None, 7, 7, 512)         0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 25088)             0         \n",
      "_________________________________________________________________\n",
      "fc1 (Dense)                  (None, 4096)              102764544 \n",
      "_________________________________________________________________\n",
      "fc2 (Dense)                  (None, 4096)              16781312  \n",
      "_________________________________________________________________\n",
      "predictions (Dense)          (None, 1000)              4097000   \n",
      "=================================================================\n",
      "Total params: 143,667,240\n",
      "Trainable params: 143,667,240\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "vgg19.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "xcept = tf.keras.applications.xception.Xception(\n",
    "    include_top=True,\n",
    "    weights='imagenet',\n",
    "    input_tensor=None,\n",
    "    input_shape=None,\n",
    "    pooling=None,\n",
    "    classes=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"xception\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_2 (InputLayer)            [(None, 299, 299, 3) 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1 (Conv2D)           (None, 149, 149, 32) 864         input_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_bn (BatchNormaliza (None, 149, 149, 32) 128         block1_conv1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_act (Activation)   (None, 149, 149, 32) 0           block1_conv1_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2 (Conv2D)           (None, 147, 147, 64) 18432       block1_conv1_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_bn (BatchNormaliza (None, 147, 147, 64) 256         block1_conv2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_act (Activation)   (None, 147, 147, 64) 0           block1_conv2_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1 (SeparableConv2 (None, 147, 147, 128 8768        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_act (Activation (None, 147, 147, 128 0           block2_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2 (SeparableConv2 (None, 147, 147, 128 17536       block2_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d (Conv2D)                 (None, 74, 74, 128)  8192        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_pool (MaxPooling2D)      (None, 74, 74, 128)  0           block2_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization (BatchNorma (None, 74, 74, 128)  512         conv2d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "add (Add)                       (None, 74, 74, 128)  0           block2_pool[0][0]                \n",
      "                                                                 batch_normalization[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_act (Activation (None, 74, 74, 128)  0           add[0][0]                        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1 (SeparableConv2 (None, 74, 74, 256)  33920       block3_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_act (Activation (None, 74, 74, 256)  0           block3_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2 (SeparableConv2 (None, 74, 74, 256)  67840       block3_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 37, 37, 256)  32768       add[0][0]                        \n",
      "__________________________________________________________________________________________________\n",
      "block3_pool (MaxPooling2D)      (None, 37, 37, 256)  0           block3_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 37, 37, 256)  1024        conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_1 (Add)                     (None, 37, 37, 256)  0           block3_pool[0][0]                \n",
      "                                                                 batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_act (Activation (None, 37, 37, 256)  0           add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1 (SeparableConv2 (None, 37, 37, 728)  188672      block4_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_act (Activation (None, 37, 37, 728)  0           block4_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2 (SeparableConv2 (None, 37, 37, 728)  536536      block4_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 19, 19, 728)  186368      add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block4_pool (MaxPooling2D)      (None, 19, 19, 728)  0           block4_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 19, 19, 728)  2912        conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_2 (Add)                     (None, 19, 19, 728)  0           block4_pool[0][0]                \n",
      "                                                                 batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_act (Activation (None, 19, 19, 728)  0           add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_act (Activation (None, 19, 19, 728)  0           block5_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_act (Activation (None, 19, 19, 728)  0           block5_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_3 (Add)                     (None, 19, 19, 728)  0           block5_sepconv3_bn[0][0]         \n",
      "                                                                 add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_act (Activation (None, 19, 19, 728)  0           add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_act (Activation (None, 19, 19, 728)  0           block6_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_act (Activation (None, 19, 19, 728)  0           block6_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_4 (Add)                     (None, 19, 19, 728)  0           block6_sepconv3_bn[0][0]         \n",
      "                                                                 add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_act (Activation (None, 19, 19, 728)  0           add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_act (Activation (None, 19, 19, 728)  0           block7_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_act (Activation (None, 19, 19, 728)  0           block7_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_5 (Add)                     (None, 19, 19, 728)  0           block7_sepconv3_bn[0][0]         \n",
      "                                                                 add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_act (Activation (None, 19, 19, 728)  0           add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_act (Activation (None, 19, 19, 728)  0           block8_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_act (Activation (None, 19, 19, 728)  0           block8_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_6 (Add)                     (None, 19, 19, 728)  0           block8_sepconv3_bn[0][0]         \n",
      "                                                                 add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_act (Activation (None, 19, 19, 728)  0           add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_act (Activation (None, 19, 19, 728)  0           block9_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_act (Activation (None, 19, 19, 728)  0           block9_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_7 (Add)                     (None, 19, 19, 728)  0           block9_sepconv3_bn[0][0]         \n",
      "                                                                 add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_act (Activatio (None, 19, 19, 728)  0           block10_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_act (Activatio (None, 19, 19, 728)  0           block10_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_8 (Add)                     (None, 19, 19, 728)  0           block10_sepconv3_bn[0][0]        \n",
      "                                                                 add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_act (Activatio (None, 19, 19, 728)  0           block11_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_act (Activatio (None, 19, 19, 728)  0           block11_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_9 (Add)                     (None, 19, 19, 728)  0           block11_sepconv3_bn[0][0]        \n",
      "                                                                 add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_9[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_act (Activatio (None, 19, 19, 728)  0           block12_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_act (Activatio (None, 19, 19, 728)  0           block12_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_10 (Add)                    (None, 19, 19, 728)  0           block12_sepconv3_bn[0][0]        \n",
      "                                                                 add_9[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_10[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block13_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block13_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_act (Activatio (None, 19, 19, 728)  0           block13_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2 (SeparableConv (None, 19, 19, 1024) 752024      block13_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_bn (BatchNorma (None, 19, 19, 1024) 4096        block13_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 10, 10, 1024) 745472      add_10[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_pool (MaxPooling2D)     (None, 10, 10, 1024) 0           block13_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 10, 10, 1024) 4096        conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_11 (Add)                    (None, 10, 10, 1024) 0           block13_pool[0][0]               \n",
      "                                                                 batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1 (SeparableConv (None, 10, 10, 1536) 1582080     add_11[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_bn (BatchNorma (None, 10, 10, 1536) 6144        block14_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_act (Activatio (None, 10, 10, 1536) 0           block14_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2 (SeparableConv (None, 10, 10, 2048) 3159552     block14_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_bn (BatchNorma (None, 10, 10, 2048) 8192        block14_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_act (Activatio (None, 10, 10, 2048) 0           block14_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "avg_pool (GlobalAveragePooling2 (None, 2048)         0           block14_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "predictions (Dense)             (None, 1000)         2049000     avg_pool[0][0]                   \n",
      "==================================================================================================\n",
      "Total params: 22,910,480\n",
      "Trainable params: 22,855,952\n",
      "Non-trainable params: 54,528\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "xcept.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.utils import plot_model, model_to_dot\n",
    "from IPython.display import SVG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# YOU MAY NEED TO INSTALL THESE MODULES\n",
    "#!conda install graphviz #(I use Anaconda)\n",
    "#!pip install pydotplus\n",
    "#!pip install pydot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"12227pt\" viewBox=\"0.00 0.00 703.00 9170.00\" width=\"937pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1.33333 1.33333) rotate(0) translate(4 9166)\">\n",
       "<title>G</title>\n",
       "<polygon fill=\"white\" points=\"-4,4 -4,-9166 699,-9166 699,4 -4,4\" stroke=\"none\"/>\n",
       "<!-- 2791550437752 -->\n",
       "<g class=\"node\" id=\"node1\"><title>2791550437752</title>\n",
       "<polygon fill=\"none\" points=\"164,-9125.5 164,-9161.5 290,-9161.5 290,-9125.5 164,-9125.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-9139.8\">input_2: InputLayer</text>\n",
       "</g>\n",
       "<!-- 2791550438032 -->\n",
       "<g class=\"node\" id=\"node2\"><title>2791550438032</title>\n",
       "<polygon fill=\"none\" points=\"151.5,-9052.5 151.5,-9088.5 302.5,-9088.5 302.5,-9052.5 151.5,-9052.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-9066.8\">block1_conv1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791550437752&#45;&gt;2791550438032 -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>2791550437752-&gt;2791550438032</title>\n",
       "<path d=\"M227,-9125.31C227,-9117.29 227,-9107.55 227,-9098.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"230.5,-9098.53 227,-9088.53 223.5,-9098.53 230.5,-9098.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791550438984 -->\n",
       "<g class=\"node\" id=\"node3\"><title>2791550438984</title>\n",
       "<polygon fill=\"none\" points=\"109.5,-8979.5 109.5,-9015.5 344.5,-9015.5 344.5,-8979.5 109.5,-8979.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-8993.8\">block1_conv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791550438032&#45;&gt;2791550438984 -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>2791550438032-&gt;2791550438984</title>\n",
       "<path d=\"M227,-9052.31C227,-9044.29 227,-9034.55 227,-9025.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"230.5,-9025.53 227,-9015.53 223.5,-9025.53 230.5,-9025.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791550454080 -->\n",
       "<g class=\"node\" id=\"node4\"><title>2791550454080</title>\n",
       "<polygon fill=\"none\" points=\"134.5,-8906.5 134.5,-8942.5 319.5,-8942.5 319.5,-8906.5 134.5,-8906.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-8920.8\">block1_conv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791550438984&#45;&gt;2791550454080 -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>2791550438984-&gt;2791550454080</title>\n",
       "<path d=\"M227,-8979.31C227,-8971.29 227,-8961.55 227,-8952.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"230.5,-8952.53 227,-8942.53 223.5,-8952.53 230.5,-8952.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791550454024 -->\n",
       "<g class=\"node\" id=\"node5\"><title>2791550454024</title>\n",
       "<polygon fill=\"none\" points=\"151.5,-8833.5 151.5,-8869.5 302.5,-8869.5 302.5,-8833.5 151.5,-8833.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-8847.8\">block1_conv2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791550454080&#45;&gt;2791550454024 -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>2791550454080-&gt;2791550454024</title>\n",
       "<path d=\"M227,-8906.31C227,-8898.29 227,-8888.55 227,-8879.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"230.5,-8879.53 227,-8869.53 223.5,-8879.53 230.5,-8879.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791550439208 -->\n",
       "<g class=\"node\" id=\"node6\"><title>2791550439208</title>\n",
       "<polygon fill=\"none\" points=\"109.5,-8760.5 109.5,-8796.5 344.5,-8796.5 344.5,-8760.5 109.5,-8760.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-8774.8\">block1_conv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791550454024&#45;&gt;2791550439208 -->\n",
       "<g class=\"edge\" id=\"edge5\"><title>2791550454024-&gt;2791550439208</title>\n",
       "<path d=\"M227,-8833.31C227,-8825.29 227,-8815.55 227,-8806.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"230.5,-8806.53 227,-8796.53 223.5,-8806.53 230.5,-8806.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791552390424 -->\n",
       "<g class=\"node\" id=\"node7\"><title>2791552390424</title>\n",
       "<polygon fill=\"none\" points=\"134.5,-8687.5 134.5,-8723.5 319.5,-8723.5 319.5,-8687.5 134.5,-8687.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"227\" y=\"-8701.8\">block1_conv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791550439208&#45;&gt;2791552390424 -->\n",
       "<g class=\"edge\" id=\"edge6\"><title>2791550439208-&gt;2791552390424</title>\n",
       "<path d=\"M227,-8760.31C227,-8752.29 227,-8742.55 227,-8733.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"230.5,-8733.53 227,-8723.53 223.5,-8733.53 230.5,-8733.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791553221408 -->\n",
       "<g class=\"node\" id=\"node8\"><title>2791553221408</title>\n",
       "<polygon fill=\"none\" points=\"29.5,-8614.5 29.5,-8650.5 254.5,-8650.5 254.5,-8614.5 29.5,-8614.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"142\" y=\"-8628.8\">block2_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791552390424&#45;&gt;2791553221408 -->\n",
       "<g class=\"edge\" id=\"edge7\"><title>2791552390424-&gt;2791553221408</title>\n",
       "<path d=\"M206.424,-8687.31C195.566,-8678.24 182.078,-8666.98 170.254,-8657.1\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"172.305,-8654.25 162.387,-8650.53 167.818,-8659.63 172.305,-8654.25\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791552682304 -->\n",
       "<g class=\"node\" id=\"node13\"><title>2791552682304</title>\n",
       "<polygon fill=\"none\" points=\"292.5,-8541.5 292.5,-8577.5 405.5,-8577.5 405.5,-8541.5 292.5,-8541.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"349\" y=\"-8555.8\">conv2d: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791552390424&#45;&gt;2791552682304 -->\n",
       "<g class=\"edge\" id=\"edge12\"><title>2791552390424-&gt;2791552682304</title>\n",
       "<path d=\"M241.474,-8687.42C262.936,-8662.08 303.317,-8614.42 328.021,-8585.26\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"330.751,-8587.45 334.545,-8577.56 325.41,-8582.93 330.751,-8587.45\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791553654288 -->\n",
       "<g class=\"node\" id=\"node9\"><title>2791553654288</title>\n",
       "<polygon fill=\"none\" points=\"0,-8541.5 0,-8577.5 254,-8577.5 254,-8541.5 0,-8541.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"127\" y=\"-8555.8\">block2_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791553221408&#45;&gt;2791553654288 -->\n",
       "<g class=\"edge\" id=\"edge8\"><title>2791553221408-&gt;2791553654288</title>\n",
       "<path d=\"M138.369,-8614.31C136.674,-8606.29 134.616,-8596.55 132.719,-8587.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"136.089,-8586.59 130.598,-8577.53 129.24,-8588.04 136.089,-8586.59\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791553804048 -->\n",
       "<g class=\"node\" id=\"node10\"><title>2791553804048</title>\n",
       "<polygon fill=\"none\" points=\"15,-8468.5 15,-8504.5 219,-8504.5 219,-8468.5 15,-8468.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"117\" y=\"-8482.8\">block2_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791553654288&#45;&gt;2791553804048 -->\n",
       "<g class=\"edge\" id=\"edge9\"><title>2791553654288-&gt;2791553804048</title>\n",
       "<path d=\"M124.579,-8541.31C123.449,-8533.29 122.077,-8523.55 120.813,-8514.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"124.259,-8513.94 119.398,-8504.53 117.327,-8514.92 124.259,-8513.94\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791556066888 -->\n",
       "<g class=\"node\" id=\"node11\"><title>2791556066888</title>\n",
       "<polygon fill=\"none\" points=\"48.5,-8395.5 48.5,-8431.5 273.5,-8431.5 273.5,-8395.5 48.5,-8395.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"161\" y=\"-8409.8\">block2_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791553804048&#45;&gt;2791556066888 -->\n",
       "<g class=\"edge\" id=\"edge10\"><title>2791553804048-&gt;2791556066888</title>\n",
       "<path d=\"M127.651,-8468.31C132.894,-8459.85 139.32,-8449.48 145.127,-8440.11\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"148.154,-8441.87 150.447,-8431.53 142.204,-8438.19 148.154,-8441.87\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791556294696 -->\n",
       "<g class=\"node\" id=\"node12\"><title>2791556294696</title>\n",
       "<polygon fill=\"none\" points=\"56,-8322.5 56,-8358.5 310,-8358.5 310,-8322.5 56,-8322.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"183\" y=\"-8336.8\">block2_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791556066888&#45;&gt;2791556294696 -->\n",
       "<g class=\"edge\" id=\"edge11\"><title>2791556066888-&gt;2791556294696</title>\n",
       "<path d=\"M166.326,-8395.31C168.839,-8387.2 171.896,-8377.34 174.703,-8368.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"178.107,-8369.12 177.723,-8358.53 171.42,-8367.04 178.107,-8369.12\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791556328696 -->\n",
       "<g class=\"node\" id=\"node14\"><title>2791556328696</title>\n",
       "<polygon fill=\"none\" points=\"117.5,-8249.5 117.5,-8285.5 296.5,-8285.5 296.5,-8249.5 117.5,-8249.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"207\" y=\"-8263.8\">block2_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791556294696&#45;&gt;2791556328696 -->\n",
       "<g class=\"edge\" id=\"edge13\"><title>2791556294696-&gt;2791556328696</title>\n",
       "<path d=\"M188.81,-8322.31C191.551,-8314.2 194.887,-8304.34 197.949,-8295.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"201.357,-8296.12 201.244,-8285.53 194.726,-8293.88 201.357,-8296.12\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791553177808 -->\n",
       "<g class=\"node\" id=\"node15\"><title>2791553177808</title>\n",
       "<polygon fill=\"none\" points=\"237.5,-8468.5 237.5,-8504.5 482.5,-8504.5 482.5,-8468.5 237.5,-8468.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"360\" y=\"-8482.8\">batch_normalization: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791552682304&#45;&gt;2791553177808 -->\n",
       "<g class=\"edge\" id=\"edge14\"><title>2791552682304-&gt;2791553177808</title>\n",
       "<path d=\"M351.663,-8541.31C352.906,-8533.29 354.415,-8523.55 355.806,-8514.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"359.289,-8514.95 357.362,-8504.53 352.372,-8513.88 359.289,-8514.95\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791581831008 -->\n",
       "<g class=\"node\" id=\"node16\"><title>2791581831008</title>\n",
       "<polygon fill=\"none\" points=\"243.5,-8176.5 243.5,-8212.5 312.5,-8212.5 312.5,-8176.5 243.5,-8176.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"278\" y=\"-8190.8\">add: Add</text>\n",
       "</g>\n",
       "<!-- 2791556328696&#45;&gt;2791581831008 -->\n",
       "<g class=\"edge\" id=\"edge15\"><title>2791556328696-&gt;2791581831008</title>\n",
       "<path d=\"M224.187,-8249.31C233.083,-8240.42 244.091,-8229.41 253.828,-8219.67\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"256.375,-8222.07 260.971,-8212.53 251.425,-8217.13 256.375,-8222.07\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791553177808&#45;&gt;2791581831008 -->\n",
       "<g class=\"edge\" id=\"edge16\"><title>2791553177808-&gt;2791581831008</title>\n",
       "<path d=\"M357.475,-8468.47C351.319,-8429.02 333.89,-8328.65 305,-8249 301.701,-8239.9 297.236,-8230.36 292.887,-8221.89\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"295.876,-8220.05 288.095,-8212.86 289.693,-8223.33 295.876,-8220.05\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791582301600 -->\n",
       "<g class=\"node\" id=\"node17\"><title>2791582301600</title>\n",
       "<polygon fill=\"none\" points=\"111,-8103.5 111,-8139.5 315,-8139.5 315,-8103.5 111,-8103.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"213\" y=\"-8117.8\">block3_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791581831008&#45;&gt;2791582301600 -->\n",
       "<g class=\"edge\" id=\"edge17\"><title>2791581831008-&gt;2791582301600</title>\n",
       "<path d=\"M262.265,-8176.31C254.201,-8167.5 244.241,-8156.63 235.392,-8146.96\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"237.924,-8144.54 228.59,-8139.53 232.761,-8149.27 237.924,-8144.54\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791581957088 -->\n",
       "<g class=\"node\" id=\"node23\"><title>2791581957088</title>\n",
       "<polygon fill=\"none\" points=\"367.5,-8030.5 367.5,-8066.5 494.5,-8066.5 494.5,-8030.5 367.5,-8030.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"431\" y=\"-8044.8\">conv2d_1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791581831008&#45;&gt;2791581957088 -->\n",
       "<g class=\"edge\" id=\"edge23\"><title>2791581831008-&gt;2791581957088</title>\n",
       "<path d=\"M296.152,-8176.42C323.298,-8150.87 374.576,-8102.6 405.481,-8073.52\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"407.988,-8075.96 412.872,-8066.56 403.191,-8070.87 407.988,-8075.96\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791582034688 -->\n",
       "<g class=\"node\" id=\"node18\"><title>2791582034688</title>\n",
       "<polygon fill=\"none\" points=\"84.5,-8030.5 84.5,-8066.5 309.5,-8066.5 309.5,-8030.5 84.5,-8030.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"197\" y=\"-8044.8\">block3_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791582301600&#45;&gt;2791582034688 -->\n",
       "<g class=\"edge\" id=\"edge18\"><title>2791582301600-&gt;2791582034688</title>\n",
       "<path d=\"M209.127,-8103.31C207.319,-8095.29 205.123,-8085.55 203.1,-8076.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"206.45,-8075.51 200.837,-8066.53 199.622,-8077.05 206.45,-8075.51\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791675678672 -->\n",
       "<g class=\"node\" id=\"node19\"><title>2791675678672</title>\n",
       "<polygon fill=\"none\" points=\"50,-7957.5 50,-7993.5 304,-7993.5 304,-7957.5 50,-7957.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"177\" y=\"-7971.8\">block3_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791582034688&#45;&gt;2791675678672 -->\n",
       "<g class=\"edge\" id=\"edge19\"><title>2791582034688-&gt;2791675678672</title>\n",
       "<path d=\"M192.159,-8030.31C189.874,-8022.2 187.095,-8012.34 184.543,-8003.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"187.877,-8002.21 181.797,-7993.53 181.139,-8004.1 187.877,-8002.21\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791675823328 -->\n",
       "<g class=\"node\" id=\"node20\"><title>2791675823328</title>\n",
       "<polygon fill=\"none\" points=\"135,-7884.5 135,-7920.5 339,-7920.5 339,-7884.5 135,-7884.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"237\" y=\"-7898.8\">block3_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791675678672&#45;&gt;2791675823328 -->\n",
       "<g class=\"edge\" id=\"edge20\"><title>2791675678672-&gt;2791675823328</title>\n",
       "<path d=\"M191.524,-7957.31C198.894,-7948.59 207.98,-7937.84 216.088,-7928.25\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"218.828,-7930.43 222.609,-7920.53 213.482,-7925.91 218.828,-7930.43\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791734529104 -->\n",
       "<g class=\"node\" id=\"node21\"><title>2791734529104</title>\n",
       "<polygon fill=\"none\" points=\"154.5,-7811.5 154.5,-7847.5 379.5,-7847.5 379.5,-7811.5 154.5,-7811.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"267\" y=\"-7825.8\">block3_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791675823328&#45;&gt;2791734529104 -->\n",
       "<g class=\"edge\" id=\"edge21\"><title>2791675823328-&gt;2791734529104</title>\n",
       "<path d=\"M244.262,-7884.31C247.726,-7876.12 251.948,-7866.12 255.809,-7856.98\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"259.137,-7858.1 259.805,-7847.53 252.688,-7855.38 259.137,-7858.1\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791734736432 -->\n",
       "<g class=\"node\" id=\"node22\"><title>2791734736432</title>\n",
       "<polygon fill=\"none\" points=\"155,-7738.5 155,-7774.5 409,-7774.5 409,-7738.5 155,-7738.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"282\" y=\"-7752.8\">block3_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791734529104&#45;&gt;2791734736432 -->\n",
       "<g class=\"edge\" id=\"edge22\"><title>2791734529104-&gt;2791734736432</title>\n",
       "<path d=\"M270.631,-7811.31C272.326,-7803.29 274.384,-7793.55 276.281,-7784.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"279.76,-7785.04 278.402,-7774.53 272.911,-7783.59 279.76,-7785.04\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791734762240 -->\n",
       "<g class=\"node\" id=\"node24\"><title>2791734762240</title>\n",
       "<polygon fill=\"none\" points=\"214.5,-7665.5 214.5,-7701.5 393.5,-7701.5 393.5,-7665.5 214.5,-7665.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"304\" y=\"-7679.8\">block3_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791734736432&#45;&gt;2791734762240 -->\n",
       "<g class=\"edge\" id=\"edge24\"><title>2791734736432-&gt;2791734762240</title>\n",
       "<path d=\"M287.326,-7738.31C289.839,-7730.2 292.896,-7720.34 295.703,-7711.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"299.107,-7712.12 298.723,-7701.53 292.42,-7710.04 299.107,-7712.12\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791581957984 -->\n",
       "<g class=\"node\" id=\"node25\"><title>2791581957984</title>\n",
       "<polygon fill=\"none\" points=\"322,-7957.5 322,-7993.5 582,-7993.5 582,-7957.5 322,-7957.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"452\" y=\"-7971.8\">batch_normalization_1: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791581957088&#45;&gt;2791581957984 -->\n",
       "<g class=\"edge\" id=\"edge25\"><title>2791581957088-&gt;2791581957984</title>\n",
       "<path d=\"M436.084,-8030.31C438.482,-8022.2 441.401,-8012.34 444.08,-8003.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"447.483,-8004.11 446.963,-7993.53 440.771,-8002.13 447.483,-8004.11\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791734762184 -->\n",
       "<g class=\"node\" id=\"node26\"><title>2791734762184</title>\n",
       "<polygon fill=\"none\" points=\"332.5,-7592.5 332.5,-7628.5 415.5,-7628.5 415.5,-7592.5 332.5,-7592.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"374\" y=\"-7606.8\">add_1: Add</text>\n",
       "</g>\n",
       "<!-- 2791734762240&#45;&gt;2791734762184 -->\n",
       "<g class=\"edge\" id=\"edge26\"><title>2791734762240-&gt;2791734762184</title>\n",
       "<path d=\"M320.945,-7665.31C329.715,-7656.42 340.568,-7645.41 350.169,-7635.67\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"352.683,-7638.11 357.211,-7628.53 347.698,-7633.19 352.683,-7638.11\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791581957984&#45;&gt;2791734762184 -->\n",
       "<g class=\"edge\" id=\"edge27\"><title>2791581957984-&gt;2791734762184</title>\n",
       "<path d=\"M450.419,-7957.23C448.088,-7930.22 444,-7876.32 444,-7830.5 444,-7830.5 444,-7830.5 444,-7755.5 444,-7710.35 415.822,-7664.53 395.222,-7637.06\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"397.73,-7634.58 388.845,-7628.81 392.192,-7638.86 397.73,-7634.58\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791783619104 -->\n",
       "<g class=\"node\" id=\"node27\"><title>2791783619104</title>\n",
       "<polygon fill=\"none\" points=\"207,-7519.5 207,-7555.5 411,-7555.5 411,-7519.5 207,-7519.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"309\" y=\"-7533.8\">block4_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791734762184&#45;&gt;2791783619104 -->\n",
       "<g class=\"edge\" id=\"edge28\"><title>2791734762184-&gt;2791783619104</title>\n",
       "<path d=\"M358.265,-7592.31C350.201,-7583.5 340.241,-7572.63 331.392,-7562.96\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"333.924,-7560.54 324.59,-7555.53 328.761,-7565.27 333.924,-7560.54\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791783545432 -->\n",
       "<g class=\"node\" id=\"node33\"><title>2791783545432</title>\n",
       "<polygon fill=\"none\" points=\"463.5,-7446.5 463.5,-7482.5 590.5,-7482.5 590.5,-7446.5 463.5,-7446.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"527\" y=\"-7460.8\">conv2d_2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791734762184&#45;&gt;2791783545432 -->\n",
       "<g class=\"edge\" id=\"edge34\"><title>2791734762184-&gt;2791783545432</title>\n",
       "<path d=\"M392.152,-7592.42C419.298,-7566.87 470.576,-7518.6 501.481,-7489.52\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"503.988,-7491.96 508.872,-7482.56 499.191,-7486.87 503.988,-7491.96\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791783866320 -->\n",
       "<g class=\"node\" id=\"node28\"><title>2791783866320</title>\n",
       "<polygon fill=\"none\" points=\"180.5,-7446.5 180.5,-7482.5 405.5,-7482.5 405.5,-7446.5 180.5,-7446.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"293\" y=\"-7460.8\">block4_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791783619104&#45;&gt;2791783866320 -->\n",
       "<g class=\"edge\" id=\"edge29\"><title>2791783619104-&gt;2791783866320</title>\n",
       "<path d=\"M305.127,-7519.31C303.319,-7511.29 301.123,-7501.55 299.1,-7492.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"302.45,-7491.51 296.837,-7482.53 295.622,-7493.05 302.45,-7491.51\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791784119656 -->\n",
       "<g class=\"node\" id=\"node29\"><title>2791784119656</title>\n",
       "<polygon fill=\"none\" points=\"146,-7373.5 146,-7409.5 400,-7409.5 400,-7373.5 146,-7373.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"273\" y=\"-7387.8\">block4_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791783866320&#45;&gt;2791784119656 -->\n",
       "<g class=\"edge\" id=\"edge30\"><title>2791783866320-&gt;2791784119656</title>\n",
       "<path d=\"M288.159,-7446.31C285.874,-7438.2 283.095,-7428.34 280.543,-7419.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"283.877,-7418.21 277.797,-7409.53 277.139,-7420.1 283.877,-7418.21\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791784173520 -->\n",
       "<g class=\"node\" id=\"node30\"><title>2791784173520</title>\n",
       "<polygon fill=\"none\" points=\"231,-7300.5 231,-7336.5 435,-7336.5 435,-7300.5 231,-7300.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"333\" y=\"-7314.8\">block4_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791784119656&#45;&gt;2791784173520 -->\n",
       "<g class=\"edge\" id=\"edge31\"><title>2791784119656-&gt;2791784173520</title>\n",
       "<path d=\"M287.524,-7373.31C294.894,-7364.59 303.98,-7353.84 312.088,-7344.25\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"314.828,-7346.43 318.609,-7336.53 309.482,-7341.91 314.828,-7346.43\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791784251856 -->\n",
       "<g class=\"node\" id=\"node31\"><title>2791784251856</title>\n",
       "<polygon fill=\"none\" points=\"250.5,-7227.5 250.5,-7263.5 475.5,-7263.5 475.5,-7227.5 250.5,-7227.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"363\" y=\"-7241.8\">block4_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791784173520&#45;&gt;2791784251856 -->\n",
       "<g class=\"edge\" id=\"edge32\"><title>2791784173520-&gt;2791784251856</title>\n",
       "<path d=\"M340.262,-7300.31C343.726,-7292.12 347.948,-7282.12 351.809,-7272.98\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"355.137,-7274.1 355.805,-7263.53 348.688,-7271.38 355.137,-7274.1\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791927557984 -->\n",
       "<g class=\"node\" id=\"node32\"><title>2791927557984</title>\n",
       "<polygon fill=\"none\" points=\"251,-7154.5 251,-7190.5 505,-7190.5 505,-7154.5 251,-7154.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"378\" y=\"-7168.8\">block4_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791784251856&#45;&gt;2791927557984 -->\n",
       "<g class=\"edge\" id=\"edge33\"><title>2791784251856-&gt;2791927557984</title>\n",
       "<path d=\"M366.631,-7227.31C368.326,-7219.29 370.384,-7209.55 372.281,-7200.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"375.76,-7201.04 374.402,-7190.53 368.911,-7199.59 375.76,-7201.04\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791927711784 -->\n",
       "<g class=\"node\" id=\"node34\"><title>2791927711784</title>\n",
       "<polygon fill=\"none\" points=\"310.5,-7081.5 310.5,-7117.5 489.5,-7117.5 489.5,-7081.5 310.5,-7081.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"400\" y=\"-7095.8\">block4_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791927557984&#45;&gt;2791927711784 -->\n",
       "<g class=\"edge\" id=\"edge35\"><title>2791927557984-&gt;2791927711784</title>\n",
       "<path d=\"M383.326,-7154.31C385.839,-7146.2 388.896,-7136.34 391.703,-7127.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"395.107,-7128.12 394.723,-7117.53 388.42,-7126.04 395.107,-7128.12\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791783546216 -->\n",
       "<g class=\"node\" id=\"node35\"><title>2791783546216</title>\n",
       "<polygon fill=\"none\" points=\"418,-7373.5 418,-7409.5 678,-7409.5 678,-7373.5 418,-7373.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"548\" y=\"-7387.8\">batch_normalization_2: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791783545432&#45;&gt;2791783546216 -->\n",
       "<g class=\"edge\" id=\"edge36\"><title>2791783545432-&gt;2791783546216</title>\n",
       "<path d=\"M532.084,-7446.31C534.482,-7438.2 537.401,-7428.34 540.08,-7419.28\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"543.483,-7420.11 542.963,-7409.53 536.771,-7418.13 543.483,-7420.11\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791927711728 -->\n",
       "<g class=\"node\" id=\"node36\"><title>2791927711728</title>\n",
       "<polygon fill=\"none\" points=\"428.5,-7008.5 428.5,-7044.5 511.5,-7044.5 511.5,-7008.5 428.5,-7008.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"470\" y=\"-7022.8\">add_2: Add</text>\n",
       "</g>\n",
       "<!-- 2791927711784&#45;&gt;2791927711728 -->\n",
       "<g class=\"edge\" id=\"edge37\"><title>2791927711784-&gt;2791927711728</title>\n",
       "<path d=\"M416.945,-7081.31C425.715,-7072.42 436.568,-7061.41 446.169,-7051.67\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"448.683,-7054.11 453.211,-7044.53 443.698,-7049.19 448.683,-7054.11\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791783546216&#45;&gt;2791927711728 -->\n",
       "<g class=\"edge\" id=\"edge38\"><title>2791783546216-&gt;2791927711728</title>\n",
       "<path d=\"M546.419,-7373.23C544.088,-7346.22 540,-7292.32 540,-7246.5 540,-7246.5 540,-7246.5 540,-7171.5 540,-7126.35 511.822,-7080.53 491.222,-7053.06\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"493.73,-7050.58 484.845,-7044.81 488.192,-7054.86 493.73,-7050.58\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791928231864 -->\n",
       "<g class=\"node\" id=\"node37\"><title>2791928231864</title>\n",
       "<polygon fill=\"none\" points=\"298,-6935.5 298,-6971.5 502,-6971.5 502,-6935.5 298,-6935.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"400\" y=\"-6949.8\">block5_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791927711728&#45;&gt;2791928231864 -->\n",
       "<g class=\"edge\" id=\"edge39\"><title>2791927711728-&gt;2791928231864</title>\n",
       "<path d=\"M453.055,-7008.31C444.285,-6999.42 433.432,-6988.41 423.831,-6978.67\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"426.302,-6976.19 416.789,-6971.53 421.317,-6981.11 426.302,-6976.19\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792201315440 -->\n",
       "<g class=\"node\" id=\"node46\"><title>2792201315440</title>\n",
       "<polygon fill=\"none\" points=\"423.5,-6278.5 423.5,-6314.5 506.5,-6314.5 506.5,-6278.5 423.5,-6278.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"465\" y=\"-6292.8\">add_3: Add</text>\n",
       "</g>\n",
       "<!-- 2791927711728&#45;&gt;2792201315440 -->\n",
       "<g class=\"edge\" id=\"edge49\"><title>2791927711728-&gt;2792201315440</title>\n",
       "<path d=\"M485.481,-7008.3C506.91,-6982.65 543,-6931.95 543,-6881.5 543,-6881.5 543,-6881.5 543,-6441.5 543,-6400.4 544.442,-6386.65 524,-6351 517.488,-6339.64 507.721,-6329.37 497.999,-6320.93\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"500.219,-6318.22 490.274,-6314.57 495.77,-6323.63 500.219,-6318.22\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791928231808 -->\n",
       "<g class=\"node\" id=\"node38\"><title>2791928231808</title>\n",
       "<polygon fill=\"none\" points=\"282.5,-6862.5 282.5,-6898.5 507.5,-6898.5 507.5,-6862.5 282.5,-6862.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"395\" y=\"-6876.8\">block5_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791928231864&#45;&gt;2791928231808 -->\n",
       "<g class=\"edge\" id=\"edge40\"><title>2791928231864-&gt;2791928231808</title>\n",
       "<path d=\"M398.79,-6935.31C398.225,-6927.29 397.539,-6917.55 396.906,-6908.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"400.393,-6908.26 396.199,-6898.53 393.41,-6908.75 400.393,-6908.26\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791928233880 -->\n",
       "<g class=\"node\" id=\"node39\"><title>2791928233880</title>\n",
       "<polygon fill=\"none\" points=\"261,-6789.5 261,-6825.5 515,-6825.5 515,-6789.5 261,-6789.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6803.8\">block5_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791928231808&#45;&gt;2791928233880 -->\n",
       "<g class=\"edge\" id=\"edge41\"><title>2791928231808-&gt;2791928233880</title>\n",
       "<path d=\"M393.305,-6862.31C392.514,-6854.29 391.554,-6844.55 390.669,-6835.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"394.143,-6835.14 389.679,-6825.53 387.177,-6835.82 394.143,-6835.14\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791928096640 -->\n",
       "<g class=\"node\" id=\"node40\"><title>2791928096640</title>\n",
       "<polygon fill=\"none\" points=\"286,-6716.5 286,-6752.5 490,-6752.5 490,-6716.5 286,-6716.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6730.8\">block5_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791928233880&#45;&gt;2791928096640 -->\n",
       "<g class=\"edge\" id=\"edge42\"><title>2791928233880-&gt;2791928096640</title>\n",
       "<path d=\"M388,-6789.31C388,-6781.29 388,-6771.55 388,-6762.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"391.5,-6762.53 388,-6752.53 384.5,-6762.53 391.5,-6762.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791928744760 -->\n",
       "<g class=\"node\" id=\"node41\"><title>2791928744760</title>\n",
       "<polygon fill=\"none\" points=\"275.5,-6643.5 275.5,-6679.5 500.5,-6679.5 500.5,-6643.5 275.5,-6643.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6657.8\">block5_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791928096640&#45;&gt;2791928744760 -->\n",
       "<g class=\"edge\" id=\"edge43\"><title>2791928096640-&gt;2791928744760</title>\n",
       "<path d=\"M388,-6716.31C388,-6708.29 388,-6698.55 388,-6689.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"391.5,-6689.53 388,-6679.53 384.5,-6689.53 391.5,-6689.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791932021112 -->\n",
       "<g class=\"node\" id=\"node42\"><title>2791932021112</title>\n",
       "<polygon fill=\"none\" points=\"261,-6570.5 261,-6606.5 515,-6606.5 515,-6570.5 261,-6570.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6584.8\">block5_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791928744760&#45;&gt;2791932021112 -->\n",
       "<g class=\"edge\" id=\"edge44\"><title>2791928744760-&gt;2791932021112</title>\n",
       "<path d=\"M388,-6643.31C388,-6635.29 388,-6625.55 388,-6616.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"391.5,-6616.53 388,-6606.53 384.5,-6616.53 391.5,-6616.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791932067400 -->\n",
       "<g class=\"node\" id=\"node43\"><title>2791932067400</title>\n",
       "<polygon fill=\"none\" points=\"286,-6497.5 286,-6533.5 490,-6533.5 490,-6497.5 286,-6497.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6511.8\">block5_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791932021112&#45;&gt;2791932067400 -->\n",
       "<g class=\"edge\" id=\"edge45\"><title>2791932021112-&gt;2791932067400</title>\n",
       "<path d=\"M388,-6570.31C388,-6562.29 388,-6552.55 388,-6543.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"391.5,-6543.53 388,-6533.53 384.5,-6543.53 391.5,-6543.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791932067232 -->\n",
       "<g class=\"node\" id=\"node44\"><title>2791932067232</title>\n",
       "<polygon fill=\"none\" points=\"275.5,-6424.5 275.5,-6460.5 500.5,-6460.5 500.5,-6424.5 275.5,-6424.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6438.8\">block5_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791932067400&#45;&gt;2791932067232 -->\n",
       "<g class=\"edge\" id=\"edge46\"><title>2791932067400-&gt;2791932067232</title>\n",
       "<path d=\"M388,-6497.31C388,-6489.29 388,-6479.55 388,-6470.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"391.5,-6470.53 388,-6460.53 384.5,-6470.53 391.5,-6470.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791961436000 -->\n",
       "<g class=\"node\" id=\"node45\"><title>2791961436000</title>\n",
       "<polygon fill=\"none\" points=\"261,-6351.5 261,-6387.5 515,-6387.5 515,-6351.5 261,-6351.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"388\" y=\"-6365.8\">block5_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791932067232&#45;&gt;2791961436000 -->\n",
       "<g class=\"edge\" id=\"edge47\"><title>2791932067232-&gt;2791961436000</title>\n",
       "<path d=\"M388,-6424.31C388,-6416.29 388,-6406.55 388,-6397.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"391.5,-6397.53 388,-6387.53 384.5,-6397.53 391.5,-6397.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791961436000&#45;&gt;2792201315440 -->\n",
       "<g class=\"edge\" id=\"edge48\"><title>2791961436000-&gt;2792201315440</title>\n",
       "<path d=\"M406.64,-6351.31C416.381,-6342.33 428.459,-6331.19 439.096,-6321.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"441.553,-6323.88 446.532,-6314.53 436.808,-6318.73 441.553,-6323.88\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791928231416 -->\n",
       "<g class=\"node\" id=\"node47\"><title>2791928231416</title>\n",
       "<polygon fill=\"none\" points=\"293,-6205.5 293,-6241.5 497,-6241.5 497,-6205.5 293,-6205.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"395\" y=\"-6219.8\">block6_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792201315440&#45;&gt;2791928231416 -->\n",
       "<g class=\"edge\" id=\"edge50\"><title>2792201315440-&gt;2791928231416</title>\n",
       "<path d=\"M448.055,-6278.31C439.285,-6269.42 428.432,-6258.41 418.831,-6248.67\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"421.302,-6246.19 411.789,-6241.53 416.317,-6251.11 421.302,-6246.19\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792463261480 -->\n",
       "<g class=\"node\" id=\"node56\"><title>2792463261480</title>\n",
       "<polygon fill=\"none\" points=\"418.5,-5548.5 418.5,-5584.5 501.5,-5584.5 501.5,-5548.5 418.5,-5548.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"460\" y=\"-5562.8\">add_4: Add</text>\n",
       "</g>\n",
       "<!-- 2792201315440&#45;&gt;2792463261480 -->\n",
       "<g class=\"edge\" id=\"edge60\"><title>2792201315440-&gt;2792463261480</title>\n",
       "<path d=\"M480.481,-6278.3C501.91,-6252.65 538,-6201.95 538,-6151.5 538,-6151.5 538,-6151.5 538,-5711.5 538,-5670.4 539.442,-5656.65 519,-5621 512.488,-5609.64 502.721,-5599.37 492.999,-5590.93\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"495.219,-5588.22 485.274,-5584.57 490.77,-5593.63 495.219,-5588.22\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791548016792 -->\n",
       "<g class=\"node\" id=\"node48\"><title>2791548016792</title>\n",
       "<polygon fill=\"none\" points=\"277.5,-6132.5 277.5,-6168.5 502.5,-6168.5 502.5,-6132.5 277.5,-6132.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"390\" y=\"-6146.8\">block6_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791928231416&#45;&gt;2791548016792 -->\n",
       "<g class=\"edge\" id=\"edge51\"><title>2791928231416-&gt;2791548016792</title>\n",
       "<path d=\"M393.79,-6205.31C393.225,-6197.29 392.539,-6187.55 391.906,-6178.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"395.393,-6178.26 391.199,-6168.53 388.41,-6178.75 395.393,-6178.26\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791548080424 -->\n",
       "<g class=\"node\" id=\"node49\"><title>2791548080424</title>\n",
       "<polygon fill=\"none\" points=\"256,-6059.5 256,-6095.5 510,-6095.5 510,-6059.5 256,-6059.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-6073.8\">block6_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791548016792&#45;&gt;2791548080424 -->\n",
       "<g class=\"edge\" id=\"edge52\"><title>2791548016792-&gt;2791548080424</title>\n",
       "<path d=\"M388.305,-6132.31C387.514,-6124.29 386.554,-6114.55 385.669,-6105.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"389.143,-6105.14 384.679,-6095.53 382.177,-6105.82 389.143,-6105.14\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791548048832 -->\n",
       "<g class=\"node\" id=\"node50\"><title>2791548048832</title>\n",
       "<polygon fill=\"none\" points=\"281,-5986.5 281,-6022.5 485,-6022.5 485,-5986.5 281,-5986.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-6000.8\">block6_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2791548080424&#45;&gt;2791548048832 -->\n",
       "<g class=\"edge\" id=\"edge53\"><title>2791548080424-&gt;2791548048832</title>\n",
       "<path d=\"M383,-6059.31C383,-6051.29 383,-6041.55 383,-6032.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"386.5,-6032.53 383,-6022.53 379.5,-6032.53 386.5,-6032.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791548047824 -->\n",
       "<g class=\"node\" id=\"node51\"><title>2791548047824</title>\n",
       "<polygon fill=\"none\" points=\"270.5,-5913.5 270.5,-5949.5 495.5,-5949.5 495.5,-5913.5 270.5,-5913.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-5927.8\">block6_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2791548048832&#45;&gt;2791548047824 -->\n",
       "<g class=\"edge\" id=\"edge54\"><title>2791548048832-&gt;2791548047824</title>\n",
       "<path d=\"M383,-5986.31C383,-5978.29 383,-5968.55 383,-5959.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"386.5,-5959.53 383,-5949.53 379.5,-5959.53 386.5,-5959.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792420914288 -->\n",
       "<g class=\"node\" id=\"node52\"><title>2792420914288</title>\n",
       "<polygon fill=\"none\" points=\"256,-5840.5 256,-5876.5 510,-5876.5 510,-5840.5 256,-5840.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-5854.8\">block6_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2791548047824&#45;&gt;2792420914288 -->\n",
       "<g class=\"edge\" id=\"edge55\"><title>2791548047824-&gt;2792420914288</title>\n",
       "<path d=\"M383,-5913.31C383,-5905.29 383,-5895.55 383,-5886.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"386.5,-5886.53 383,-5876.53 379.5,-5886.53 386.5,-5886.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792420940096 -->\n",
       "<g class=\"node\" id=\"node53\"><title>2792420940096</title>\n",
       "<polygon fill=\"none\" points=\"281,-5767.5 281,-5803.5 485,-5803.5 485,-5767.5 281,-5767.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-5781.8\">block6_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792420914288&#45;&gt;2792420940096 -->\n",
       "<g class=\"edge\" id=\"edge56\"><title>2792420914288-&gt;2792420940096</title>\n",
       "<path d=\"M383,-5840.31C383,-5832.29 383,-5822.55 383,-5813.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"386.5,-5813.53 383,-5803.53 379.5,-5813.53 386.5,-5813.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792420939928 -->\n",
       "<g class=\"node\" id=\"node54\"><title>2792420939928</title>\n",
       "<polygon fill=\"none\" points=\"270.5,-5694.5 270.5,-5730.5 495.5,-5730.5 495.5,-5694.5 270.5,-5694.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-5708.8\">block6_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792420940096&#45;&gt;2792420939928 -->\n",
       "<g class=\"edge\" id=\"edge57\"><title>2792420940096-&gt;2792420939928</title>\n",
       "<path d=\"M383,-5767.31C383,-5759.29 383,-5749.55 383,-5740.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"386.5,-5740.53 383,-5730.53 379.5,-5740.53 386.5,-5740.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792421452376 -->\n",
       "<g class=\"node\" id=\"node55\"><title>2792421452376</title>\n",
       "<polygon fill=\"none\" points=\"256,-5621.5 256,-5657.5 510,-5657.5 510,-5621.5 256,-5621.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-5635.8\">block6_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792420939928&#45;&gt;2792421452376 -->\n",
       "<g class=\"edge\" id=\"edge58\"><title>2792420939928-&gt;2792421452376</title>\n",
       "<path d=\"M383,-5694.31C383,-5686.29 383,-5676.55 383,-5667.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"386.5,-5667.53 383,-5657.53 379.5,-5667.53 386.5,-5667.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792421452376&#45;&gt;2792463261480 -->\n",
       "<g class=\"edge\" id=\"edge59\"><title>2792421452376-&gt;2792463261480</title>\n",
       "<path d=\"M401.64,-5621.31C411.381,-5612.33 423.459,-5601.19 434.096,-5591.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"436.553,-5593.88 441.532,-5584.53 431.808,-5588.73 436.553,-5593.88\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792201314992 -->\n",
       "<g class=\"node\" id=\"node57\"><title>2792201314992</title>\n",
       "<polygon fill=\"none\" points=\"281,-5475.5 281,-5511.5 485,-5511.5 485,-5475.5 281,-5475.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"383\" y=\"-5489.8\">block7_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792463261480&#45;&gt;2792201314992 -->\n",
       "<g class=\"edge\" id=\"edge61\"><title>2792463261480-&gt;2792201314992</title>\n",
       "<path d=\"M441.36,-5548.31C431.619,-5539.33 419.541,-5528.19 408.904,-5518.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"411.192,-5515.73 401.468,-5511.53 406.447,-5520.88 411.192,-5515.73\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792518575496 -->\n",
       "<g class=\"node\" id=\"node66\"><title>2792518575496</title>\n",
       "<polygon fill=\"none\" points=\"400.5,-4818.5 400.5,-4854.5 483.5,-4854.5 483.5,-4818.5 400.5,-4818.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"442\" y=\"-4832.8\">add_5: Add</text>\n",
       "</g>\n",
       "<!-- 2792463261480&#45;&gt;2792518575496 -->\n",
       "<g class=\"edge\" id=\"edge71\"><title>2792463261480-&gt;2792518575496</title>\n",
       "<path d=\"M472.432,-5548.29C490.006,-5522.21 520,-5470.52 520,-5421.5 520,-5421.5 520,-5421.5 520,-4981.5 520,-4940.4 521.442,-4926.65 501,-4891 494.488,-4879.64 484.721,-4869.37 474.999,-4860.93\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"477.219,-4858.22 467.274,-4854.57 472.77,-4863.63 477.219,-4858.22\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792463261424 -->\n",
       "<g class=\"node\" id=\"node58\"><title>2792463261424</title>\n",
       "<polygon fill=\"none\" points=\"258.5,-5402.5 258.5,-5438.5 483.5,-5438.5 483.5,-5402.5 258.5,-5402.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"371\" y=\"-5416.8\">block7_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792201314992&#45;&gt;2792463261424 -->\n",
       "<g class=\"edge\" id=\"edge62\"><title>2792201314992-&gt;2792463261424</title>\n",
       "<path d=\"M380.095,-5475.31C378.739,-5467.29 377.093,-5457.55 375.575,-5448.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"378.996,-5447.81 373.878,-5438.53 372.094,-5448.97 378.996,-5447.81\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792517272800 -->\n",
       "<g class=\"node\" id=\"node59\"><title>2792517272800</title>\n",
       "<polygon fill=\"none\" points=\"238,-5329.5 238,-5365.5 492,-5365.5 492,-5329.5 238,-5329.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-5343.8\">block7_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792463261424&#45;&gt;2792517272800 -->\n",
       "<g class=\"edge\" id=\"edge63\"><title>2792463261424-&gt;2792517272800</title>\n",
       "<path d=\"M369.548,-5402.31C368.87,-5394.29 368.046,-5384.55 367.288,-5375.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"370.769,-5375.2 366.439,-5365.53 363.794,-5375.79 370.769,-5375.2\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792517327224 -->\n",
       "<g class=\"node\" id=\"node60\"><title>2792517327224</title>\n",
       "<polygon fill=\"none\" points=\"263,-5256.5 263,-5292.5 467,-5292.5 467,-5256.5 263,-5256.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-5270.8\">block7_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792517272800&#45;&gt;2792517327224 -->\n",
       "<g class=\"edge\" id=\"edge64\"><title>2792517272800-&gt;2792517327224</title>\n",
       "<path d=\"M365,-5329.31C365,-5321.29 365,-5311.55 365,-5302.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.5,-5302.53 365,-5292.53 361.5,-5302.53 368.5,-5302.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792517326832 -->\n",
       "<g class=\"node\" id=\"node61\"><title>2792517326832</title>\n",
       "<polygon fill=\"none\" points=\"252.5,-5183.5 252.5,-5219.5 477.5,-5219.5 477.5,-5183.5 252.5,-5183.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-5197.8\">block7_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792517327224&#45;&gt;2792517326832 -->\n",
       "<g class=\"edge\" id=\"edge65\"><title>2792517327224-&gt;2792517326832</title>\n",
       "<path d=\"M365,-5256.31C365,-5248.29 365,-5238.55 365,-5229.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.5,-5229.53 365,-5219.53 361.5,-5229.53 368.5,-5229.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792517831368 -->\n",
       "<g class=\"node\" id=\"node62\"><title>2792517831368</title>\n",
       "<polygon fill=\"none\" points=\"238,-5110.5 238,-5146.5 492,-5146.5 492,-5110.5 238,-5110.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-5124.8\">block7_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792517326832&#45;&gt;2792517831368 -->\n",
       "<g class=\"edge\" id=\"edge66\"><title>2792517326832-&gt;2792517831368</title>\n",
       "<path d=\"M365,-5183.31C365,-5175.29 365,-5165.55 365,-5156.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.5,-5156.53 365,-5146.53 361.5,-5156.53 368.5,-5156.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792517705624 -->\n",
       "<g class=\"node\" id=\"node63\"><title>2792517705624</title>\n",
       "<polygon fill=\"none\" points=\"263,-5037.5 263,-5073.5 467,-5073.5 467,-5037.5 263,-5037.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-5051.8\">block7_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792517831368&#45;&gt;2792517705624 -->\n",
       "<g class=\"edge\" id=\"edge67\"><title>2792517831368-&gt;2792517705624</title>\n",
       "<path d=\"M365,-5110.31C365,-5102.29 365,-5092.55 365,-5083.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.5,-5083.53 365,-5073.53 361.5,-5083.53 368.5,-5083.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792517988480 -->\n",
       "<g class=\"node\" id=\"node64\"><title>2792517988480</title>\n",
       "<polygon fill=\"none\" points=\"252.5,-4964.5 252.5,-5000.5 477.5,-5000.5 477.5,-4964.5 252.5,-4964.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-4978.8\">block7_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792517705624&#45;&gt;2792517988480 -->\n",
       "<g class=\"edge\" id=\"edge68\"><title>2792517705624-&gt;2792517988480</title>\n",
       "<path d=\"M365,-5037.31C365,-5029.29 365,-5019.55 365,-5010.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.5,-5010.53 365,-5000.53 361.5,-5010.53 368.5,-5010.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792518418272 -->\n",
       "<g class=\"node\" id=\"node65\"><title>2792518418272</title>\n",
       "<polygon fill=\"none\" points=\"238,-4891.5 238,-4927.5 492,-4927.5 492,-4891.5 238,-4891.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"365\" y=\"-4905.8\">block7_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792517988480&#45;&gt;2792518418272 -->\n",
       "<g class=\"edge\" id=\"edge69\"><title>2792517988480-&gt;2792518418272</title>\n",
       "<path d=\"M365,-4964.31C365,-4956.29 365,-4946.55 365,-4937.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.5,-4937.53 365,-4927.53 361.5,-4937.53 368.5,-4937.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792518418272&#45;&gt;2792518575496 -->\n",
       "<g class=\"edge\" id=\"edge70\"><title>2792518418272-&gt;2792518575496</title>\n",
       "<path d=\"M383.64,-4891.31C393.381,-4882.33 405.459,-4871.19 416.096,-4861.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"418.553,-4863.88 423.532,-4854.53 413.808,-4858.73 418.553,-4863.88\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792463260528 -->\n",
       "<g class=\"node\" id=\"node67\"><title>2792463260528</title>\n",
       "<polygon fill=\"none\" points=\"275,-4745.5 275,-4781.5 479,-4781.5 479,-4745.5 275,-4745.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"377\" y=\"-4759.8\">block8_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792518575496&#45;&gt;2792463260528 -->\n",
       "<g class=\"edge\" id=\"edge72\"><title>2792518575496-&gt;2792463260528</title>\n",
       "<path d=\"M426.265,-4818.31C418.201,-4809.5 408.241,-4798.63 399.392,-4788.96\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"401.924,-4786.54 392.59,-4781.53 396.761,-4791.27 401.924,-4786.54\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520362976 -->\n",
       "<g class=\"node\" id=\"node76\"><title>2792520362976</title>\n",
       "<polygon fill=\"none\" points=\"392.5,-4088.5 392.5,-4124.5 475.5,-4124.5 475.5,-4088.5 392.5,-4088.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"434\" y=\"-4102.8\">add_6: Add</text>\n",
       "</g>\n",
       "<!-- 2792518575496&#45;&gt;2792520362976 -->\n",
       "<g class=\"edge\" id=\"edge82\"><title>2792518575496-&gt;2792520362976</title>\n",
       "<path d=\"M460.433,-4818.22C469.909,-4808.41 480.923,-4795.44 488,-4782 507.389,-4745.18 512,-4733.11 512,-4691.5 512,-4691.5 512,-4691.5 512,-4251.5 512,-4210.4 513.442,-4196.65 493,-4161 486.488,-4149.64 476.721,-4139.37 466.999,-4130.93\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"469.219,-4128.22 459.274,-4124.57 464.77,-4133.63 469.219,-4128.22\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792518574880 -->\n",
       "<g class=\"node\" id=\"node68\"><title>2792518574880</title>\n",
       "<polygon fill=\"none\" points=\"251.5,-4672.5 251.5,-4708.5 476.5,-4708.5 476.5,-4672.5 251.5,-4672.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"364\" y=\"-4686.8\">block8_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792463260528&#45;&gt;2792518574880 -->\n",
       "<g class=\"edge\" id=\"edge73\"><title>2792463260528-&gt;2792518574880</title>\n",
       "<path d=\"M373.853,-4745.31C372.384,-4737.29 370.6,-4727.55 368.956,-4718.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"372.362,-4717.73 367.118,-4708.53 365.476,-4719 372.362,-4717.73\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792519093048 -->\n",
       "<g class=\"node\" id=\"node69\"><title>2792519093048</title>\n",
       "<polygon fill=\"none\" points=\"230,-4599.5 230,-4635.5 484,-4635.5 484,-4599.5 230,-4599.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4613.8\">block8_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792518574880&#45;&gt;2792519093048 -->\n",
       "<g class=\"edge\" id=\"edge74\"><title>2792518574880-&gt;2792519093048</title>\n",
       "<path d=\"M362.305,-4672.31C361.514,-4664.29 360.554,-4654.55 359.669,-4645.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"363.143,-4645.14 358.679,-4635.53 356.177,-4645.82 363.143,-4645.14\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792518970560 -->\n",
       "<g class=\"node\" id=\"node70\"><title>2792518970560</title>\n",
       "<polygon fill=\"none\" points=\"255,-4526.5 255,-4562.5 459,-4562.5 459,-4526.5 255,-4526.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4540.8\">block8_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792519093048&#45;&gt;2792518970560 -->\n",
       "<g class=\"edge\" id=\"edge75\"><title>2792519093048-&gt;2792518970560</title>\n",
       "<path d=\"M357,-4599.31C357,-4591.29 357,-4581.55 357,-4572.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"360.5,-4572.53 357,-4562.53 353.5,-4572.53 360.5,-4572.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792519134792 -->\n",
       "<g class=\"node\" id=\"node71\"><title>2792519134792</title>\n",
       "<polygon fill=\"none\" points=\"244.5,-4453.5 244.5,-4489.5 469.5,-4489.5 469.5,-4453.5 244.5,-4453.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4467.8\">block8_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792518970560&#45;&gt;2792519134792 -->\n",
       "<g class=\"edge\" id=\"edge76\"><title>2792518970560-&gt;2792519134792</title>\n",
       "<path d=\"M357,-4526.31C357,-4518.29 357,-4508.55 357,-4499.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"360.5,-4499.53 357,-4489.53 353.5,-4499.53 360.5,-4499.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792519659416 -->\n",
       "<g class=\"node\" id=\"node72\"><title>2792519659416</title>\n",
       "<polygon fill=\"none\" points=\"230,-4380.5 230,-4416.5 484,-4416.5 484,-4380.5 230,-4380.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4394.8\">block8_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792519134792&#45;&gt;2792519659416 -->\n",
       "<g class=\"edge\" id=\"edge77\"><title>2792519134792-&gt;2792519659416</title>\n",
       "<path d=\"M357,-4453.31C357,-4445.29 357,-4435.55 357,-4426.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"360.5,-4426.53 357,-4416.53 353.5,-4426.53 360.5,-4426.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792519804408 -->\n",
       "<g class=\"node\" id=\"node73\"><title>2792519804408</title>\n",
       "<polygon fill=\"none\" points=\"255,-4307.5 255,-4343.5 459,-4343.5 459,-4307.5 255,-4307.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4321.8\">block8_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792519659416&#45;&gt;2792519804408 -->\n",
       "<g class=\"edge\" id=\"edge78\"><title>2792519659416-&gt;2792519804408</title>\n",
       "<path d=\"M357,-4380.31C357,-4372.29 357,-4362.55 357,-4353.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"360.5,-4353.53 357,-4343.53 353.5,-4353.53 360.5,-4353.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520121648 -->\n",
       "<g class=\"node\" id=\"node74\"><title>2792520121648</title>\n",
       "<polygon fill=\"none\" points=\"244.5,-4234.5 244.5,-4270.5 469.5,-4270.5 469.5,-4234.5 244.5,-4234.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4248.8\">block8_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792519804408&#45;&gt;2792520121648 -->\n",
       "<g class=\"edge\" id=\"edge79\"><title>2792519804408-&gt;2792520121648</title>\n",
       "<path d=\"M357,-4307.31C357,-4299.29 357,-4289.55 357,-4280.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"360.5,-4280.53 357,-4270.53 353.5,-4280.53 360.5,-4280.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520333072 -->\n",
       "<g class=\"node\" id=\"node75\"><title>2792520333072</title>\n",
       "<polygon fill=\"none\" points=\"230,-4161.5 230,-4197.5 484,-4197.5 484,-4161.5 230,-4161.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4175.8\">block8_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792520121648&#45;&gt;2792520333072 -->\n",
       "<g class=\"edge\" id=\"edge80\"><title>2792520121648-&gt;2792520333072</title>\n",
       "<path d=\"M357,-4234.31C357,-4226.29 357,-4216.55 357,-4207.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"360.5,-4207.53 357,-4197.53 353.5,-4207.53 360.5,-4207.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520333072&#45;&gt;2792520362976 -->\n",
       "<g class=\"edge\" id=\"edge81\"><title>2792520333072-&gt;2792520362976</title>\n",
       "<path d=\"M375.64,-4161.31C385.381,-4152.33 397.459,-4141.19 408.096,-4131.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"410.553,-4133.88 415.532,-4124.53 405.808,-4128.73 410.553,-4133.88\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792518575048 -->\n",
       "<g class=\"node\" id=\"node77\"><title>2792518575048</title>\n",
       "<polygon fill=\"none\" points=\"255,-4015.5 255,-4051.5 459,-4051.5 459,-4015.5 255,-4015.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"357\" y=\"-4029.8\">block9_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792520362976&#45;&gt;2792518575048 -->\n",
       "<g class=\"edge\" id=\"edge83\"><title>2792520362976-&gt;2792518575048</title>\n",
       "<path d=\"M415.36,-4088.31C405.619,-4079.33 393.541,-4068.19 382.904,-4058.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"385.192,-4055.73 375.468,-4051.53 380.447,-4060.88 385.192,-4055.73\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522203088 -->\n",
       "<g class=\"node\" id=\"node86\"><title>2792522203088</title>\n",
       "<polygon fill=\"none\" points=\"374.5,-3358.5 374.5,-3394.5 457.5,-3394.5 457.5,-3358.5 374.5,-3358.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"416\" y=\"-3372.8\">add_7: Add</text>\n",
       "</g>\n",
       "<!-- 2792520362976&#45;&gt;2792522203088 -->\n",
       "<g class=\"edge\" id=\"edge93\"><title>2792520362976-&gt;2792522203088</title>\n",
       "<path d=\"M446.432,-4088.29C464.006,-4062.21 494,-4010.52 494,-3961.5 494,-3961.5 494,-3961.5 494,-3521.5 494,-3480.4 495.442,-3466.65 475,-3431 468.488,-3419.64 458.721,-3409.37 448.999,-3400.93\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"451.219,-3398.22 441.274,-3394.57 446.77,-3403.63 451.219,-3398.22\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520650592 -->\n",
       "<g class=\"node\" id=\"node78\"><title>2792520650592</title>\n",
       "<polygon fill=\"none\" points=\"232.5,-3942.5 232.5,-3978.5 457.5,-3978.5 457.5,-3942.5 232.5,-3942.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"345\" y=\"-3956.8\">block9_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792518575048&#45;&gt;2792520650592 -->\n",
       "<g class=\"edge\" id=\"edge84\"><title>2792518575048-&gt;2792520650592</title>\n",
       "<path d=\"M354.095,-4015.31C352.739,-4007.29 351.093,-3997.55 349.575,-3988.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"352.996,-3987.81 347.878,-3978.53 346.094,-3988.97 352.996,-3987.81\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520900560 -->\n",
       "<g class=\"node\" id=\"node79\"><title>2792520900560</title>\n",
       "<polygon fill=\"none\" points=\"212,-3869.5 212,-3905.5 466,-3905.5 466,-3869.5 212,-3869.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3883.8\">block9_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792520650592&#45;&gt;2792520900560 -->\n",
       "<g class=\"edge\" id=\"edge85\"><title>2792520650592-&gt;2792520900560</title>\n",
       "<path d=\"M343.548,-3942.31C342.87,-3934.29 342.046,-3924.55 341.288,-3915.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"344.769,-3915.2 340.439,-3905.53 337.794,-3915.79 344.769,-3915.2\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792521045664 -->\n",
       "<g class=\"node\" id=\"node80\"><title>2792521045664</title>\n",
       "<polygon fill=\"none\" points=\"237,-3796.5 237,-3832.5 441,-3832.5 441,-3796.5 237,-3796.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3810.8\">block9_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792520900560&#45;&gt;2792521045664 -->\n",
       "<g class=\"edge\" id=\"edge86\"><title>2792520900560-&gt;2792521045664</title>\n",
       "<path d=\"M339,-3869.31C339,-3861.29 339,-3851.55 339,-3842.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"342.5,-3842.53 339,-3832.53 335.5,-3842.53 342.5,-3842.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792521354712 -->\n",
       "<g class=\"node\" id=\"node81\"><title>2792521354712</title>\n",
       "<polygon fill=\"none\" points=\"226.5,-3723.5 226.5,-3759.5 451.5,-3759.5 451.5,-3723.5 226.5,-3723.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3737.8\">block9_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792521045664&#45;&gt;2792521354712 -->\n",
       "<g class=\"edge\" id=\"edge87\"><title>2792521045664-&gt;2792521354712</title>\n",
       "<path d=\"M339,-3796.31C339,-3788.29 339,-3778.55 339,-3769.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"342.5,-3769.53 339,-3759.53 335.5,-3769.53 342.5,-3769.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792521562040 -->\n",
       "<g class=\"node\" id=\"node82\"><title>2792521562040</title>\n",
       "<polygon fill=\"none\" points=\"212,-3650.5 212,-3686.5 466,-3686.5 466,-3650.5 212,-3650.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3664.8\">block9_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792521354712&#45;&gt;2792521562040 -->\n",
       "<g class=\"edge\" id=\"edge88\"><title>2792521354712-&gt;2792521562040</title>\n",
       "<path d=\"M339,-3723.31C339,-3715.29 339,-3705.55 339,-3696.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"342.5,-3696.53 339,-3686.53 335.5,-3696.53 342.5,-3696.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792521596040 -->\n",
       "<g class=\"node\" id=\"node83\"><title>2792521596040</title>\n",
       "<polygon fill=\"none\" points=\"237,-3577.5 237,-3613.5 441,-3613.5 441,-3577.5 237,-3577.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3591.8\">block9_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792521562040&#45;&gt;2792521596040 -->\n",
       "<g class=\"edge\" id=\"edge89\"><title>2792521562040-&gt;2792521596040</title>\n",
       "<path d=\"M339,-3650.31C339,-3642.29 339,-3632.55 339,-3623.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"342.5,-3623.53 339,-3613.53 335.5,-3623.53 342.5,-3623.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792521887528 -->\n",
       "<g class=\"node\" id=\"node84\"><title>2792521887528</title>\n",
       "<polygon fill=\"none\" points=\"226.5,-3504.5 226.5,-3540.5 451.5,-3540.5 451.5,-3504.5 226.5,-3504.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3518.8\">block9_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792521596040&#45;&gt;2792521887528 -->\n",
       "<g class=\"edge\" id=\"edge90\"><title>2792521596040-&gt;2792521887528</title>\n",
       "<path d=\"M339,-3577.31C339,-3569.29 339,-3559.55 339,-3550.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"342.5,-3550.53 339,-3540.53 335.5,-3550.53 342.5,-3550.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522132896 -->\n",
       "<g class=\"node\" id=\"node85\"><title>2792522132896</title>\n",
       "<polygon fill=\"none\" points=\"212,-3431.5 212,-3467.5 466,-3467.5 466,-3431.5 212,-3431.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"339\" y=\"-3445.8\">block9_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792521887528&#45;&gt;2792522132896 -->\n",
       "<g class=\"edge\" id=\"edge91\"><title>2792521887528-&gt;2792522132896</title>\n",
       "<path d=\"M339,-3504.31C339,-3496.29 339,-3486.55 339,-3477.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"342.5,-3477.53 339,-3467.53 335.5,-3477.53 342.5,-3477.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522132896&#45;&gt;2792522203088 -->\n",
       "<g class=\"edge\" id=\"edge92\"><title>2792522132896-&gt;2792522203088</title>\n",
       "<path d=\"M357.64,-3431.31C367.381,-3422.33 379.459,-3411.19 390.096,-3401.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"392.553,-3403.88 397.532,-3394.53 387.808,-3398.73 392.553,-3403.88\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792520362528 -->\n",
       "<g class=\"node\" id=\"node87\"><title>2792520362528</title>\n",
       "<polygon fill=\"none\" points=\"243.5,-3285.5 243.5,-3321.5 454.5,-3321.5 454.5,-3285.5 243.5,-3285.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"349\" y=\"-3299.8\">block10_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792522203088&#45;&gt;2792520362528 -->\n",
       "<g class=\"edge\" id=\"edge94\"><title>2792522203088-&gt;2792520362528</title>\n",
       "<path d=\"M399.781,-3358.31C391.469,-3349.5 381.202,-3338.63 372.08,-3328.96\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"374.478,-3326.4 365.069,-3321.53 369.387,-3331.2 374.478,-3326.4\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524089608 -->\n",
       "<g class=\"node\" id=\"node96\"><title>2792524089608</title>\n",
       "<polygon fill=\"none\" points=\"366.5,-2628.5 366.5,-2664.5 449.5,-2664.5 449.5,-2628.5 366.5,-2628.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"408\" y=\"-2642.8\">add_8: Add</text>\n",
       "</g>\n",
       "<!-- 2792522203088&#45;&gt;2792524089608 -->\n",
       "<g class=\"edge\" id=\"edge104\"><title>2792522203088-&gt;2792524089608</title>\n",
       "<path d=\"M435.259,-3358.36C445.155,-3348.59 456.646,-3335.61 464,-3322 483.776,-3285.39 488,-3273.11 488,-3231.5 488,-3231.5 488,-3231.5 488,-2791.5 488,-2750.4 489.737,-2736.48 469,-2701 462.326,-2689.58 452.342,-2679.34 442.373,-2670.95\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"444.446,-2668.13 434.445,-2664.63 440.083,-2673.6 444.446,-2668.13\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522244336 -->\n",
       "<g class=\"node\" id=\"node88\"><title>2792522244336</title>\n",
       "<polygon fill=\"none\" points=\"221,-3212.5 221,-3248.5 453,-3248.5 453,-3212.5 221,-3212.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"337\" y=\"-3226.8\">block10_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792520362528&#45;&gt;2792522244336 -->\n",
       "<g class=\"edge\" id=\"edge95\"><title>2792520362528-&gt;2792522244336</title>\n",
       "<path d=\"M346.095,-3285.31C344.739,-3277.29 343.093,-3267.55 341.575,-3258.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"344.996,-3257.81 339.878,-3248.53 338.094,-3258.97 344.996,-3257.81\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522807336 -->\n",
       "<g class=\"node\" id=\"node89\"><title>2792522807336</title>\n",
       "<polygon fill=\"none\" points=\"198.5,-3139.5 198.5,-3175.5 459.5,-3175.5 459.5,-3139.5 198.5,-3139.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-3153.8\">block10_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792522244336&#45;&gt;2792522807336 -->\n",
       "<g class=\"edge\" id=\"edge96\"><title>2792522244336-&gt;2792522807336</title>\n",
       "<path d=\"M335.063,-3212.31C334.159,-3204.29 333.062,-3194.55 332.05,-3185.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"335.516,-3185.07 330.919,-3175.53 328.561,-3185.86 335.516,-3185.07\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522845432 -->\n",
       "<g class=\"node\" id=\"node90\"><title>2792522845432</title>\n",
       "<polygon fill=\"none\" points=\"223.5,-3066.5 223.5,-3102.5 434.5,-3102.5 434.5,-3066.5 223.5,-3066.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-3080.8\">block10_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792522807336&#45;&gt;2792522845432 -->\n",
       "<g class=\"edge\" id=\"edge97\"><title>2792522807336-&gt;2792522845432</title>\n",
       "<path d=\"M329,-3139.31C329,-3131.29 329,-3121.55 329,-3112.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"332.5,-3112.53 329,-3102.53 325.5,-3112.53 332.5,-3112.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792523145112 -->\n",
       "<g class=\"node\" id=\"node91\"><title>2792523145112</title>\n",
       "<polygon fill=\"none\" points=\"213,-2993.5 213,-3029.5 445,-3029.5 445,-2993.5 213,-2993.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-3007.8\">block10_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792522845432&#45;&gt;2792523145112 -->\n",
       "<g class=\"edge\" id=\"edge98\"><title>2792522845432-&gt;2792523145112</title>\n",
       "<path d=\"M329,-3066.31C329,-3058.29 329,-3048.55 329,-3039.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"332.5,-3039.53 329,-3029.53 325.5,-3039.53 332.5,-3039.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792523374040 -->\n",
       "<g class=\"node\" id=\"node92\"><title>2792523374040</title>\n",
       "<polygon fill=\"none\" points=\"198.5,-2920.5 198.5,-2956.5 459.5,-2956.5 459.5,-2920.5 198.5,-2920.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-2934.8\">block10_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792523145112&#45;&gt;2792523374040 -->\n",
       "<g class=\"edge\" id=\"edge99\"><title>2792523145112-&gt;2792523374040</title>\n",
       "<path d=\"M329,-2993.31C329,-2985.29 329,-2975.55 329,-2966.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"332.5,-2966.53 329,-2956.53 325.5,-2966.53 332.5,-2966.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792523439968 -->\n",
       "<g class=\"node\" id=\"node93\"><title>2792523439968</title>\n",
       "<polygon fill=\"none\" points=\"223.5,-2847.5 223.5,-2883.5 434.5,-2883.5 434.5,-2847.5 223.5,-2847.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-2861.8\">block10_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792523374040&#45;&gt;2792523439968 -->\n",
       "<g class=\"edge\" id=\"edge100\"><title>2792523374040-&gt;2792523439968</title>\n",
       "<path d=\"M329,-2920.31C329,-2912.29 329,-2902.55 329,-2893.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"332.5,-2893.53 329,-2883.53 325.5,-2893.53 332.5,-2893.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792523856472 -->\n",
       "<g class=\"node\" id=\"node94\"><title>2792523856472</title>\n",
       "<polygon fill=\"none\" points=\"213,-2774.5 213,-2810.5 445,-2810.5 445,-2774.5 213,-2774.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-2788.8\">block10_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792523439968&#45;&gt;2792523856472 -->\n",
       "<g class=\"edge\" id=\"edge101\"><title>2792523439968-&gt;2792523856472</title>\n",
       "<path d=\"M329,-2847.31C329,-2839.29 329,-2829.55 329,-2820.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"332.5,-2820.53 329,-2810.53 325.5,-2820.53 332.5,-2820.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524043320 -->\n",
       "<g class=\"node\" id=\"node95\"><title>2792524043320</title>\n",
       "<polygon fill=\"none\" points=\"198.5,-2701.5 198.5,-2737.5 459.5,-2737.5 459.5,-2701.5 198.5,-2701.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"329\" y=\"-2715.8\">block10_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792523856472&#45;&gt;2792524043320 -->\n",
       "<g class=\"edge\" id=\"edge102\"><title>2792523856472-&gt;2792524043320</title>\n",
       "<path d=\"M329,-2774.31C329,-2766.29 329,-2756.55 329,-2747.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"332.5,-2747.53 329,-2737.53 325.5,-2747.53 332.5,-2747.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524043320&#45;&gt;2792524089608 -->\n",
       "<g class=\"edge\" id=\"edge103\"><title>2792524043320-&gt;2792524089608</title>\n",
       "<path d=\"M348.124,-2701.31C358.119,-2692.33 370.51,-2681.19 381.423,-2671.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"383.954,-2673.82 389.052,-2664.53 379.275,-2668.61 383.954,-2673.82\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792522244224 -->\n",
       "<g class=\"node\" id=\"node97\"><title>2792522244224</title>\n",
       "<polygon fill=\"none\" points=\"235.5,-2555.5 235.5,-2591.5 446.5,-2591.5 446.5,-2555.5 235.5,-2555.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"341\" y=\"-2569.8\">block11_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792524089608&#45;&gt;2792522244224 -->\n",
       "<g class=\"edge\" id=\"edge105\"><title>2792524089608-&gt;2792522244224</title>\n",
       "<path d=\"M391.781,-2628.31C383.469,-2619.5 373.202,-2608.63 364.08,-2598.96\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"366.478,-2596.4 357.069,-2591.53 361.387,-2601.2 366.478,-2596.4\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792526004296 -->\n",
       "<g class=\"node\" id=\"node106\"><title>2792526004296</title>\n",
       "<polygon fill=\"none\" points=\"365.5,-1898.5 365.5,-1934.5 448.5,-1934.5 448.5,-1898.5 365.5,-1898.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"407\" y=\"-1912.8\">add_9: Add</text>\n",
       "</g>\n",
       "<!-- 2792524089608&#45;&gt;2792526004296 -->\n",
       "<g class=\"edge\" id=\"edge115\"><title>2792524089608-&gt;2792526004296</title>\n",
       "<path d=\"M426.879,-2628.14C436.663,-2618.31 448.169,-2605.35 456,-2592 477.517,-2555.33 487,-2544.02 487,-2501.5 487,-2501.5 487,-2501.5 487,-2061.5 487,-2020.4 488.737,-2006.48 468,-1971 461.326,-1959.58 451.342,-1949.34 441.373,-1940.95\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"443.446,-1938.13 433.445,-1934.63 439.083,-1943.6 443.446,-1938.13\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524482656 -->\n",
       "<g class=\"node\" id=\"node98\"><title>2792524482656</title>\n",
       "<polygon fill=\"none\" points=\"218,-2482.5 218,-2518.5 450,-2518.5 450,-2482.5 218,-2482.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"334\" y=\"-2496.8\">block11_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792522244224&#45;&gt;2792524482656 -->\n",
       "<g class=\"edge\" id=\"edge106\"><title>2792522244224-&gt;2792524482656</title>\n",
       "<path d=\"M339.305,-2555.31C338.514,-2547.29 337.554,-2537.55 336.669,-2528.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"340.143,-2528.14 335.679,-2518.53 333.177,-2528.82 340.143,-2528.14\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524611200 -->\n",
       "<g class=\"node\" id=\"node99\"><title>2792524611200</title>\n",
       "<polygon fill=\"none\" points=\"197.5,-2409.5 197.5,-2445.5 458.5,-2445.5 458.5,-2409.5 197.5,-2409.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-2423.8\">block11_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792524482656&#45;&gt;2792524611200 -->\n",
       "<g class=\"edge\" id=\"edge107\"><title>2792524482656-&gt;2792524611200</title>\n",
       "<path d=\"M332.548,-2482.31C331.87,-2474.29 331.046,-2464.55 330.288,-2455.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"333.769,-2455.2 329.439,-2445.53 326.794,-2455.79 333.769,-2455.2\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524764048 -->\n",
       "<g class=\"node\" id=\"node100\"><title>2792524764048</title>\n",
       "<polygon fill=\"none\" points=\"222.5,-2336.5 222.5,-2372.5 433.5,-2372.5 433.5,-2336.5 222.5,-2336.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-2350.8\">block11_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792524611200&#45;&gt;2792524764048 -->\n",
       "<g class=\"edge\" id=\"edge108\"><title>2792524611200-&gt;2792524764048</title>\n",
       "<path d=\"M328,-2409.31C328,-2401.29 328,-2391.55 328,-2382.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"331.5,-2382.53 328,-2372.53 324.5,-2382.53 331.5,-2382.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525101768 -->\n",
       "<g class=\"node\" id=\"node101\"><title>2792525101768</title>\n",
       "<polygon fill=\"none\" points=\"212,-2263.5 212,-2299.5 444,-2299.5 444,-2263.5 212,-2263.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-2277.8\">block11_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792524764048&#45;&gt;2792525101768 -->\n",
       "<g class=\"edge\" id=\"edge109\"><title>2792524764048-&gt;2792525101768</title>\n",
       "<path d=\"M328,-2336.31C328,-2328.29 328,-2318.55 328,-2309.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"331.5,-2309.53 328,-2299.53 324.5,-2309.53 331.5,-2309.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525284520 -->\n",
       "<g class=\"node\" id=\"node102\"><title>2792525284520</title>\n",
       "<polygon fill=\"none\" points=\"197.5,-2190.5 197.5,-2226.5 458.5,-2226.5 458.5,-2190.5 197.5,-2190.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-2204.8\">block11_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792525101768&#45;&gt;2792525284520 -->\n",
       "<g class=\"edge\" id=\"edge110\"><title>2792525101768-&gt;2792525284520</title>\n",
       "<path d=\"M328,-2263.31C328,-2255.29 328,-2245.55 328,-2236.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"331.5,-2236.53 328,-2226.53 324.5,-2236.53 331.5,-2236.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525343096 -->\n",
       "<g class=\"node\" id=\"node103\"><title>2792525343096</title>\n",
       "<polygon fill=\"none\" points=\"222.5,-2117.5 222.5,-2153.5 433.5,-2153.5 433.5,-2117.5 222.5,-2117.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-2131.8\">block11_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792525284520&#45;&gt;2792525343096 -->\n",
       "<g class=\"edge\" id=\"edge111\"><title>2792525284520-&gt;2792525343096</title>\n",
       "<path d=\"M328,-2190.31C328,-2182.29 328,-2172.55 328,-2163.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"331.5,-2163.53 328,-2153.53 324.5,-2163.53 331.5,-2163.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525342928 -->\n",
       "<g class=\"node\" id=\"node104\"><title>2792525342928</title>\n",
       "<polygon fill=\"none\" points=\"212,-2044.5 212,-2080.5 444,-2080.5 444,-2044.5 212,-2044.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-2058.8\">block11_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792525343096&#45;&gt;2792525342928 -->\n",
       "<g class=\"edge\" id=\"edge112\"><title>2792525343096-&gt;2792525342928</title>\n",
       "<path d=\"M328,-2117.31C328,-2109.29 328,-2099.55 328,-2090.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"331.5,-2090.53 328,-2080.53 324.5,-2090.53 331.5,-2090.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525855376 -->\n",
       "<g class=\"node\" id=\"node105\"><title>2792525855376</title>\n",
       "<polygon fill=\"none\" points=\"197.5,-1971.5 197.5,-2007.5 458.5,-2007.5 458.5,-1971.5 197.5,-1971.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-1985.8\">block11_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792525342928&#45;&gt;2792525855376 -->\n",
       "<g class=\"edge\" id=\"edge113\"><title>2792525342928-&gt;2792525855376</title>\n",
       "<path d=\"M328,-2044.31C328,-2036.29 328,-2026.55 328,-2017.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"331.5,-2017.53 328,-2007.53 324.5,-2017.53 331.5,-2017.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525855376&#45;&gt;2792526004296 -->\n",
       "<g class=\"edge\" id=\"edge114\"><title>2792525855376-&gt;2792526004296</title>\n",
       "<path d=\"M347.124,-1971.31C357.119,-1962.33 369.51,-1951.19 380.423,-1941.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"382.954,-1943.82 388.052,-1934.53 378.275,-1938.61 382.954,-1943.82\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792524089048 -->\n",
       "<g class=\"node\" id=\"node107\"><title>2792524089048</title>\n",
       "<polygon fill=\"none\" points=\"222.5,-1825.5 222.5,-1861.5 433.5,-1861.5 433.5,-1825.5 222.5,-1825.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"328\" y=\"-1839.8\">block12_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792526004296&#45;&gt;2792524089048 -->\n",
       "<g class=\"edge\" id=\"edge116\"><title>2792526004296-&gt;2792524089048</title>\n",
       "<path d=\"M387.876,-1898.31C377.881,-1889.33 365.49,-1878.19 354.577,-1868.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"356.725,-1865.61 346.948,-1861.53 352.046,-1870.82 356.725,-1865.61\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792527824152 -->\n",
       "<g class=\"node\" id=\"node116\"><title>2792527824152</title>\n",
       "<polygon fill=\"none\" points=\"344,-1168.5 344,-1204.5 434,-1204.5 434,-1168.5 344,-1168.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"389\" y=\"-1182.8\">add_10: Add</text>\n",
       "</g>\n",
       "<!-- 2792526004296&#45;&gt;2792527824152 -->\n",
       "<g class=\"edge\" id=\"edge126\"><title>2792526004296-&gt;2792527824152</title>\n",
       "<path d=\"M420.461,-1898.49C427.948,-1888.36 436.948,-1874.99 443,-1862 460.675,-1824.07 469,-1813.35 469,-1771.5 469,-1771.5 469,-1771.5 469,-1331.5 469,-1290.4 470.737,-1276.48 450,-1241 443.326,-1229.58 433.342,-1219.34 423.373,-1210.95\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"425.446,-1208.13 415.445,-1204.63 421.083,-1213.6 425.446,-1208.13\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792525913896 -->\n",
       "<g class=\"node\" id=\"node108\"><title>2792525913896</title>\n",
       "<polygon fill=\"none\" points=\"200,-1752.5 200,-1788.5 432,-1788.5 432,-1752.5 200,-1752.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"316\" y=\"-1766.8\">block12_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792524089048&#45;&gt;2792525913896 -->\n",
       "<g class=\"edge\" id=\"edge117\"><title>2792524089048-&gt;2792525913896</title>\n",
       "<path d=\"M325.095,-1825.31C323.739,-1817.29 322.093,-1807.55 320.575,-1798.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"323.996,-1797.81 318.878,-1788.53 317.094,-1798.97 323.996,-1797.81\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792526538008 -->\n",
       "<g class=\"node\" id=\"node109\"><title>2792526538008</title>\n",
       "<polygon fill=\"none\" points=\"179.5,-1679.5 179.5,-1715.5 440.5,-1715.5 440.5,-1679.5 179.5,-1679.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1693.8\">block12_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792525913896&#45;&gt;2792526538008 -->\n",
       "<g class=\"edge\" id=\"edge118\"><title>2792525913896-&gt;2792526538008</title>\n",
       "<path d=\"M314.548,-1752.31C313.87,-1744.29 313.046,-1734.55 312.288,-1725.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"315.769,-1725.2 311.439,-1715.53 308.794,-1725.79 315.769,-1725.2\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792526571952 -->\n",
       "<g class=\"node\" id=\"node110\"><title>2792526571952</title>\n",
       "<polygon fill=\"none\" points=\"204.5,-1606.5 204.5,-1642.5 415.5,-1642.5 415.5,-1606.5 204.5,-1606.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1620.8\">block12_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792526538008&#45;&gt;2792526571952 -->\n",
       "<g class=\"edge\" id=\"edge119\"><title>2792526538008-&gt;2792526571952</title>\n",
       "<path d=\"M310,-1679.31C310,-1671.29 310,-1661.55 310,-1652.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"313.5,-1652.53 310,-1642.53 306.5,-1652.53 313.5,-1652.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792526571560 -->\n",
       "<g class=\"node\" id=\"node111\"><title>2792526571560</title>\n",
       "<polygon fill=\"none\" points=\"194,-1533.5 194,-1569.5 426,-1569.5 426,-1533.5 194,-1533.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1547.8\">block12_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792526571952&#45;&gt;2792526571560 -->\n",
       "<g class=\"edge\" id=\"edge120\"><title>2792526571952-&gt;2792526571560</title>\n",
       "<path d=\"M310,-1606.31C310,-1598.29 310,-1588.55 310,-1579.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"313.5,-1579.53 310,-1569.53 306.5,-1579.53 313.5,-1579.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792527096576 -->\n",
       "<g class=\"node\" id=\"node112\"><title>2792527096576</title>\n",
       "<polygon fill=\"none\" points=\"179.5,-1460.5 179.5,-1496.5 440.5,-1496.5 440.5,-1460.5 179.5,-1460.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1474.8\">block12_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792526571560&#45;&gt;2792527096576 -->\n",
       "<g class=\"edge\" id=\"edge121\"><title>2792526571560-&gt;2792527096576</title>\n",
       "<path d=\"M310,-1533.31C310,-1525.29 310,-1515.55 310,-1506.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"313.5,-1506.53 310,-1496.53 306.5,-1506.53 313.5,-1506.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792526979024 -->\n",
       "<g class=\"node\" id=\"node113\"><title>2792526979024</title>\n",
       "<polygon fill=\"none\" points=\"204.5,-1387.5 204.5,-1423.5 415.5,-1423.5 415.5,-1387.5 204.5,-1387.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1401.8\">block12_sepconv3_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792527096576&#45;&gt;2792526979024 -->\n",
       "<g class=\"edge\" id=\"edge122\"><title>2792527096576-&gt;2792526979024</title>\n",
       "<path d=\"M310,-1460.31C310,-1452.29 310,-1442.55 310,-1433.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"313.5,-1433.53 310,-1423.53 306.5,-1433.53 313.5,-1433.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792527154200 -->\n",
       "<g class=\"node\" id=\"node114\"><title>2792527154200</title>\n",
       "<polygon fill=\"none\" points=\"194,-1314.5 194,-1350.5 426,-1350.5 426,-1314.5 194,-1314.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1328.8\">block12_sepconv3: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792526979024&#45;&gt;2792527154200 -->\n",
       "<g class=\"edge\" id=\"edge123\"><title>2792526979024-&gt;2792527154200</title>\n",
       "<path d=\"M310,-1387.31C310,-1379.29 310,-1369.55 310,-1360.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"313.5,-1360.53 310,-1350.53 306.5,-1360.53 313.5,-1360.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792527679328 -->\n",
       "<g class=\"node\" id=\"node115\"><title>2792527679328</title>\n",
       "<polygon fill=\"none\" points=\"179.5,-1241.5 179.5,-1277.5 440.5,-1277.5 440.5,-1241.5 179.5,-1241.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"310\" y=\"-1255.8\">block12_sepconv3_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792527154200&#45;&gt;2792527679328 -->\n",
       "<g class=\"edge\" id=\"edge124\"><title>2792527154200-&gt;2792527679328</title>\n",
       "<path d=\"M310,-1314.31C310,-1306.29 310,-1296.55 310,-1287.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"313.5,-1287.53 310,-1277.53 306.5,-1287.53 313.5,-1287.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792527679328&#45;&gt;2792527824152 -->\n",
       "<g class=\"edge\" id=\"edge125\"><title>2792527679328-&gt;2792527824152</title>\n",
       "<path d=\"M329.124,-1241.31C339.119,-1232.33 351.51,-1221.19 362.423,-1211.39\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"364.954,-1213.82 370.052,-1204.53 360.275,-1208.61 364.954,-1213.82\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792528363648 -->\n",
       "<g class=\"node\" id=\"node117\"><title>2792528363648</title>\n",
       "<polygon fill=\"none\" points=\"216.5,-1095.5 216.5,-1131.5 427.5,-1131.5 427.5,-1095.5 216.5,-1095.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"322\" y=\"-1109.8\">block13_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792527824152&#45;&gt;2792528363648 -->\n",
       "<g class=\"edge\" id=\"edge127\"><title>2792527824152-&gt;2792528363648</title>\n",
       "<path d=\"M372.781,-1168.31C364.469,-1159.5 354.202,-1148.63 345.08,-1138.96\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"347.478,-1136.4 338.069,-1131.53 342.387,-1141.2 347.478,-1136.4\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792527823928 -->\n",
       "<g class=\"node\" id=\"node123\"><title>2792527823928</title>\n",
       "<polygon fill=\"none\" points=\"480.5,-1022.5 480.5,-1058.5 607.5,-1058.5 607.5,-1022.5 480.5,-1022.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"544\" y=\"-1036.8\">conv2d_3: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2792527824152&#45;&gt;2792527823928 -->\n",
       "<g class=\"edge\" id=\"edge133\"><title>2792527824152-&gt;2792527823928</title>\n",
       "<path d=\"M407.389,-1168.42C434.89,-1142.87 486.839,-1094.6 518.147,-1065.52\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"520.691,-1067.93 525.635,-1058.56 515.926,-1062.8 520.691,-1067.93\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792528334464 -->\n",
       "<g class=\"node\" id=\"node118\"><title>2792528334464</title>\n",
       "<polygon fill=\"none\" points=\"190,-1022.5 190,-1058.5 422,-1058.5 422,-1022.5 190,-1022.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"306\" y=\"-1036.8\">block13_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792528363648&#45;&gt;2792528334464 -->\n",
       "<g class=\"edge\" id=\"edge128\"><title>2792528363648-&gt;2792528334464</title>\n",
       "<path d=\"M318.127,-1095.31C316.319,-1087.29 314.123,-1077.55 312.1,-1068.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"315.45,-1067.51 309.837,-1058.53 308.622,-1069.05 315.45,-1067.51\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792528830192 -->\n",
       "<g class=\"node\" id=\"node119\"><title>2792528830192</title>\n",
       "<polygon fill=\"none\" points=\"155.5,-949.5 155.5,-985.5 416.5,-985.5 416.5,-949.5 155.5,-949.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"286\" y=\"-963.8\">block13_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792528334464&#45;&gt;2792528830192 -->\n",
       "<g class=\"edge\" id=\"edge129\"><title>2792528334464-&gt;2792528830192</title>\n",
       "<path d=\"M301.159,-1022.31C298.874,-1014.2 296.095,-1004.34 293.543,-995.277\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"296.877,-994.205 290.797,-985.529 290.139,-996.103 296.877,-994.205\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792528983992 -->\n",
       "<g class=\"node\" id=\"node120\"><title>2792528983992</title>\n",
       "<polygon fill=\"none\" points=\"240.5,-876.5 240.5,-912.5 451.5,-912.5 451.5,-876.5 240.5,-876.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"346\" y=\"-890.8\">block13_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792528830192&#45;&gt;2792528983992 -->\n",
       "<g class=\"edge\" id=\"edge130\"><title>2792528830192-&gt;2792528983992</title>\n",
       "<path d=\"M300.524,-949.313C307.894,-940.592 316.98,-929.84 325.088,-920.246\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"327.828,-922.426 331.609,-912.529 322.482,-917.908 327.828,-922.426\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792529254608 -->\n",
       "<g class=\"node\" id=\"node121\"><title>2792529254608</title>\n",
       "<polygon fill=\"none\" points=\"260,-803.5 260,-839.5 492,-839.5 492,-803.5 260,-803.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"376\" y=\"-817.8\">block13_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792528983992&#45;&gt;2792529254608 -->\n",
       "<g class=\"edge\" id=\"edge131\"><title>2792528983992-&gt;2792529254608</title>\n",
       "<path d=\"M353.262,-876.313C356.726,-868.115 360.948,-858.123 364.809,-848.985\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"368.137,-850.103 368.805,-839.529 361.688,-847.378 368.137,-850.103\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792529508616 -->\n",
       "<g class=\"node\" id=\"node122\"><title>2792529508616</title>\n",
       "<polygon fill=\"none\" points=\"260.5,-730.5 260.5,-766.5 521.5,-766.5 521.5,-730.5 260.5,-730.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"391\" y=\"-744.8\">block13_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792529254608&#45;&gt;2792529508616 -->\n",
       "<g class=\"edge\" id=\"edge132\"><title>2792529254608-&gt;2792529508616</title>\n",
       "<path d=\"M379.631,-803.313C381.326,-795.289 383.384,-785.547 385.281,-776.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"388.76,-777.036 387.402,-766.529 381.911,-775.589 388.76,-777.036\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792529550752 -->\n",
       "<g class=\"node\" id=\"node124\"><title>2792529550752</title>\n",
       "<polygon fill=\"none\" points=\"320.5,-657.5 320.5,-693.5 505.5,-693.5 505.5,-657.5 320.5,-657.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"413\" y=\"-671.8\">block13_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2792529508616&#45;&gt;2792529550752 -->\n",
       "<g class=\"edge\" id=\"edge134\"><title>2792529508616-&gt;2792529550752</title>\n",
       "<path d=\"M396.326,-730.313C398.839,-722.202 401.896,-712.336 404.703,-703.277\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"408.107,-704.117 407.723,-693.529 401.42,-702.045 408.107,-704.117\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792528332168 -->\n",
       "<g class=\"node\" id=\"node125\"><title>2792528332168</title>\n",
       "<polygon fill=\"none\" points=\"435,-949.5 435,-985.5 695,-985.5 695,-949.5 435,-949.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"565\" y=\"-963.8\">batch_normalization_3: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792527823928&#45;&gt;2792528332168 -->\n",
       "<g class=\"edge\" id=\"edge135\"><title>2792527823928-&gt;2792528332168</title>\n",
       "<path d=\"M549.084,-1022.31C551.482,-1014.2 554.401,-1004.34 557.08,-995.277\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"560.483,-996.111 559.963,-985.529 553.771,-994.125 560.483,-996.111\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792529550696 -->\n",
       "<g class=\"node\" id=\"node126\"><title>2792529550696</title>\n",
       "<polygon fill=\"none\" points=\"440,-584.5 440,-620.5 530,-620.5 530,-584.5 440,-584.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-598.8\">add_11: Add</text>\n",
       "</g>\n",
       "<!-- 2792529550752&#45;&gt;2792529550696 -->\n",
       "<g class=\"edge\" id=\"edge136\"><title>2792529550752-&gt;2792529550696</title>\n",
       "<path d=\"M430.429,-657.313C439.45,-648.417 450.613,-637.409 460.488,-627.672\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"463.068,-630.042 467.731,-620.529 458.153,-625.058 463.068,-630.042\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792528332168&#45;&gt;2792529550696 -->\n",
       "<g class=\"edge\" id=\"edge137\"><title>2792528332168-&gt;2792529550696</title>\n",
       "<path d=\"M563.419,-949.228C561.088,-922.216 557,-868.321 557,-822.5 557,-822.5 557,-822.5 557,-747.5 557,-701.905 527.766,-655.948 506.554,-628.604\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"509.245,-626.365 500.269,-620.736 503.776,-630.734 509.245,-626.365\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792529960520 -->\n",
       "<g class=\"node\" id=\"node127\"><title>2792529960520</title>\n",
       "<polygon fill=\"none\" points=\"369,-511.5 369,-547.5 601,-547.5 601,-511.5 369,-511.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-525.8\">block14_sepconv1: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792529550696&#45;&gt;2792529960520 -->\n",
       "<g class=\"edge\" id=\"edge138\"><title>2792529550696-&gt;2792529960520</title>\n",
       "<path d=\"M485,-584.313C485,-576.289 485,-566.547 485,-557.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-557.529 485,-547.529 481.5,-557.529 488.5,-557.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792530162856 -->\n",
       "<g class=\"node\" id=\"node128\"><title>2792530162856</title>\n",
       "<polygon fill=\"none\" points=\"354.5,-438.5 354.5,-474.5 615.5,-474.5 615.5,-438.5 354.5,-438.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-452.8\">block14_sepconv1_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792529960520&#45;&gt;2792530162856 -->\n",
       "<g class=\"edge\" id=\"edge139\"><title>2792529960520-&gt;2792530162856</title>\n",
       "<path d=\"M485,-511.313C485,-503.289 485,-493.547 485,-484.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-484.529 485,-474.529 481.5,-484.529 488.5,-484.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792530201008 -->\n",
       "<g class=\"node\" id=\"node129\"><title>2792530201008</title>\n",
       "<polygon fill=\"none\" points=\"379.5,-365.5 379.5,-401.5 590.5,-401.5 590.5,-365.5 379.5,-365.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-379.8\">block14_sepconv1_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792530162856&#45;&gt;2792530201008 -->\n",
       "<g class=\"edge\" id=\"edge140\"><title>2792530162856-&gt;2792530201008</title>\n",
       "<path d=\"M485,-438.313C485,-430.289 485,-420.547 485,-411.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-411.529 485,-401.529 481.5,-411.529 488.5,-411.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792530200896 -->\n",
       "<g class=\"node\" id=\"node130\"><title>2792530200896</title>\n",
       "<polygon fill=\"none\" points=\"369,-292.5 369,-328.5 601,-328.5 601,-292.5 369,-292.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-306.8\">block14_sepconv2: SeparableConv2D</text>\n",
       "</g>\n",
       "<!-- 2792530201008&#45;&gt;2792530200896 -->\n",
       "<g class=\"edge\" id=\"edge141\"><title>2792530201008-&gt;2792530200896</title>\n",
       "<path d=\"M485,-365.313C485,-357.289 485,-347.547 485,-338.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-338.529 485,-328.529 481.5,-338.529 488.5,-338.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792547314376 -->\n",
       "<g class=\"node\" id=\"node131\"><title>2792547314376</title>\n",
       "<polygon fill=\"none\" points=\"354.5,-219.5 354.5,-255.5 615.5,-255.5 615.5,-219.5 354.5,-219.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-233.8\">block14_sepconv2_bn: BatchNormalization</text>\n",
       "</g>\n",
       "<!-- 2792530200896&#45;&gt;2792547314376 -->\n",
       "<g class=\"edge\" id=\"edge142\"><title>2792530200896-&gt;2792547314376</title>\n",
       "<path d=\"M485,-292.313C485,-284.289 485,-274.547 485,-265.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-265.529 485,-255.529 481.5,-265.529 488.5,-265.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792547176344 -->\n",
       "<g class=\"node\" id=\"node132\"><title>2792547176344</title>\n",
       "<polygon fill=\"none\" points=\"379.5,-146.5 379.5,-182.5 590.5,-182.5 590.5,-146.5 379.5,-146.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-160.8\">block14_sepconv2_act: Activation</text>\n",
       "</g>\n",
       "<!-- 2792547314376&#45;&gt;2792547176344 -->\n",
       "<g class=\"edge\" id=\"edge143\"><title>2792547314376-&gt;2792547176344</title>\n",
       "<path d=\"M485,-219.313C485,-211.289 485,-201.547 485,-192.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-192.529 485,-182.529 481.5,-192.529 488.5,-192.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792547471488 -->\n",
       "<g class=\"node\" id=\"node133\"><title>2792547471488</title>\n",
       "<polygon fill=\"none\" points=\"376,-73.5 376,-109.5 594,-109.5 594,-73.5 376,-73.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-87.8\">avg_pool: GlobalAveragePooling2D</text>\n",
       "</g>\n",
       "<!-- 2792547176344&#45;&gt;2792547471488 -->\n",
       "<g class=\"edge\" id=\"edge144\"><title>2792547176344-&gt;2792547471488</title>\n",
       "<path d=\"M485,-146.313C485,-138.289 485,-128.547 485,-119.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-119.529 485,-109.529 481.5,-119.529 488.5,-119.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2792547782384 -->\n",
       "<g class=\"node\" id=\"node134\"><title>2792547782384</title>\n",
       "<polygon fill=\"none\" points=\"425,-0.5 425,-36.5 545,-36.5 545,-0.5 425,-0.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"485\" y=\"-14.8\">predictions: Dense</text>\n",
       "</g>\n",
       "<!-- 2792547471488&#45;&gt;2792547782384 -->\n",
       "<g class=\"edge\" id=\"edge145\"><title>2792547471488-&gt;2792547782384</title>\n",
       "<path d=\"M485,-73.3129C485,-65.2895 485,-55.5475 485,-46.5691\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"488.5,-46.5288 485,-36.5288 481.5,-46.5289 488.5,-46.5288\" stroke=\"black\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SVG(model_to_dot(xcept).create(prog='dot', format='svg'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"2493pt\" viewBox=\"0.00 0.00 187.00 1870.00\" width=\"249pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1.33333 1.33333) rotate(0) translate(4 1866)\">\n",
       "<title>G</title>\n",
       "<polygon fill=\"white\" points=\"-4,4 -4,-1866 183,-1866 183,4 -4,4\" stroke=\"none\"/>\n",
       "<!-- 2787143678104 -->\n",
       "<g class=\"node\" id=\"node1\"><title>2787143678104</title>\n",
       "<polygon fill=\"none\" points=\"26.5,-1825.5 26.5,-1861.5 152.5,-1861.5 152.5,-1825.5 26.5,-1825.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1839.8\">input_1: InputLayer</text>\n",
       "</g>\n",
       "<!-- 2787143678552 -->\n",
       "<g class=\"node\" id=\"node2\"><title>2787143678552</title>\n",
       "<polygon fill=\"none\" points=\"14,-1752.5 14,-1788.5 165,-1788.5 165,-1752.5 14,-1752.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1766.8\">block1_conv1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2787143678104&#45;&gt;2787143678552 -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>2787143678104-&gt;2787143678552</title>\n",
       "<path d=\"M89.5,-1825.31C89.5,-1817.29 89.5,-1807.55 89.5,-1798.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1798.53 89.5,-1788.53 86.0001,-1798.53 93.0001,-1798.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791526224112 -->\n",
       "<g class=\"node\" id=\"node3\"><title>2791526224112</title>\n",
       "<polygon fill=\"none\" points=\"14,-1679.5 14,-1715.5 165,-1715.5 165,-1679.5 14,-1679.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1693.8\">block1_conv2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2787143678552&#45;&gt;2791526224112 -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>2787143678552-&gt;2791526224112</title>\n",
       "<path d=\"M89.5,-1752.31C89.5,-1744.29 89.5,-1734.55 89.5,-1725.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1725.53 89.5,-1715.53 86.0001,-1725.53 93.0001,-1725.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2787143530480 -->\n",
       "<g class=\"node\" id=\"node4\"><title>2787143530480</title>\n",
       "<polygon fill=\"none\" points=\"0,-1606.5 0,-1642.5 179,-1642.5 179,-1606.5 0,-1606.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1620.8\">block1_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791526224112&#45;&gt;2787143530480 -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>2791526224112-&gt;2787143530480</title>\n",
       "<path d=\"M89.5,-1679.31C89.5,-1671.29 89.5,-1661.55 89.5,-1652.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1652.53 89.5,-1642.53 86.0001,-1652.53 93.0001,-1652.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791525834032 -->\n",
       "<g class=\"node\" id=\"node5\"><title>2791525834032</title>\n",
       "<polygon fill=\"none\" points=\"14,-1533.5 14,-1569.5 165,-1569.5 165,-1533.5 14,-1533.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1547.8\">block2_conv1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2787143530480&#45;&gt;2791525834032 -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>2787143530480-&gt;2791525834032</title>\n",
       "<path d=\"M89.5,-1606.31C89.5,-1598.29 89.5,-1588.55 89.5,-1579.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1579.53 89.5,-1569.53 86.0001,-1579.53 93.0001,-1579.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791522785264 -->\n",
       "<g class=\"node\" id=\"node6\"><title>2791522785264</title>\n",
       "<polygon fill=\"none\" points=\"14,-1460.5 14,-1496.5 165,-1496.5 165,-1460.5 14,-1460.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1474.8\">block2_conv2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791525834032&#45;&gt;2791522785264 -->\n",
       "<g class=\"edge\" id=\"edge5\"><title>2791525834032-&gt;2791522785264</title>\n",
       "<path d=\"M89.5,-1533.31C89.5,-1525.29 89.5,-1515.55 89.5,-1506.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1506.53 89.5,-1496.53 86.0001,-1506.53 93.0001,-1506.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791525179744 -->\n",
       "<g class=\"node\" id=\"node7\"><title>2791525179744</title>\n",
       "<polygon fill=\"none\" points=\"0,-1387.5 0,-1423.5 179,-1423.5 179,-1387.5 0,-1387.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1401.8\">block2_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791522785264&#45;&gt;2791525179744 -->\n",
       "<g class=\"edge\" id=\"edge6\"><title>2791522785264-&gt;2791525179744</title>\n",
       "<path d=\"M89.5,-1460.31C89.5,-1452.29 89.5,-1442.55 89.5,-1433.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1433.53 89.5,-1423.53 86.0001,-1433.53 93.0001,-1433.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791525108032 -->\n",
       "<g class=\"node\" id=\"node8\"><title>2791525108032</title>\n",
       "<polygon fill=\"none\" points=\"14,-1314.5 14,-1350.5 165,-1350.5 165,-1314.5 14,-1314.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1328.8\">block3_conv1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791525179744&#45;&gt;2791525108032 -->\n",
       "<g class=\"edge\" id=\"edge7\"><title>2791525179744-&gt;2791525108032</title>\n",
       "<path d=\"M89.5,-1387.31C89.5,-1379.29 89.5,-1369.55 89.5,-1360.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1360.53 89.5,-1350.53 86.0001,-1360.53 93.0001,-1360.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791524714984 -->\n",
       "<g class=\"node\" id=\"node9\"><title>2791524714984</title>\n",
       "<polygon fill=\"none\" points=\"14,-1241.5 14,-1277.5 165,-1277.5 165,-1241.5 14,-1241.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1255.8\">block3_conv2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791525108032&#45;&gt;2791524714984 -->\n",
       "<g class=\"edge\" id=\"edge8\"><title>2791525108032-&gt;2791524714984</title>\n",
       "<path d=\"M89.5,-1314.31C89.5,-1306.29 89.5,-1296.55 89.5,-1287.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1287.53 89.5,-1277.53 86.0001,-1287.53 93.0001,-1287.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791524394824 -->\n",
       "<g class=\"node\" id=\"node10\"><title>2791524394824</title>\n",
       "<polygon fill=\"none\" points=\"14,-1168.5 14,-1204.5 165,-1204.5 165,-1168.5 14,-1168.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1182.8\">block3_conv3: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791524714984&#45;&gt;2791524394824 -->\n",
       "<g class=\"edge\" id=\"edge9\"><title>2791524714984-&gt;2791524394824</title>\n",
       "<path d=\"M89.5,-1241.31C89.5,-1233.29 89.5,-1223.55 89.5,-1214.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1214.53 89.5,-1204.53 86.0001,-1214.53 93.0001,-1214.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791524074720 -->\n",
       "<g class=\"node\" id=\"node11\"><title>2791524074720</title>\n",
       "<polygon fill=\"none\" points=\"14,-1095.5 14,-1131.5 165,-1131.5 165,-1095.5 14,-1095.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1109.8\">block3_conv4: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791524394824&#45;&gt;2791524074720 -->\n",
       "<g class=\"edge\" id=\"edge10\"><title>2791524394824-&gt;2791524074720</title>\n",
       "<path d=\"M89.5,-1168.31C89.5,-1160.29 89.5,-1150.55 89.5,-1141.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1141.53 89.5,-1131.53 86.0001,-1141.53 93.0001,-1141.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791523771616 -->\n",
       "<g class=\"node\" id=\"node12\"><title>2791523771616</title>\n",
       "<polygon fill=\"none\" points=\"0,-1022.5 0,-1058.5 179,-1058.5 179,-1022.5 0,-1022.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-1036.8\">block3_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791524074720&#45;&gt;2791523771616 -->\n",
       "<g class=\"edge\" id=\"edge11\"><title>2791524074720-&gt;2791523771616</title>\n",
       "<path d=\"M89.5,-1095.31C89.5,-1087.29 89.5,-1077.55 89.5,-1068.57\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-1068.53 89.5,-1058.53 86.0001,-1068.53 93.0001,-1068.53\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791523771336 -->\n",
       "<g class=\"node\" id=\"node13\"><title>2791523771336</title>\n",
       "<polygon fill=\"none\" points=\"14,-949.5 14,-985.5 165,-985.5 165,-949.5 14,-949.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-963.8\">block4_conv1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791523771616&#45;&gt;2791523771336 -->\n",
       "<g class=\"edge\" id=\"edge12\"><title>2791523771616-&gt;2791523771336</title>\n",
       "<path d=\"M89.5,-1022.31C89.5,-1014.29 89.5,-1004.55 89.5,-995.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-995.529 89.5,-985.529 86.0001,-995.529 93.0001,-995.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791526418024 -->\n",
       "<g class=\"node\" id=\"node14\"><title>2791526418024</title>\n",
       "<polygon fill=\"none\" points=\"14,-876.5 14,-912.5 165,-912.5 165,-876.5 14,-876.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-890.8\">block4_conv2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791523771336&#45;&gt;2791526418024 -->\n",
       "<g class=\"edge\" id=\"edge13\"><title>2791523771336-&gt;2791526418024</title>\n",
       "<path d=\"M89.5,-949.313C89.5,-941.289 89.5,-931.547 89.5,-922.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-922.529 89.5,-912.529 86.0001,-922.529 93.0001,-922.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791543634184 -->\n",
       "<g class=\"node\" id=\"node15\"><title>2791543634184</title>\n",
       "<polygon fill=\"none\" points=\"14,-803.5 14,-839.5 165,-839.5 165,-803.5 14,-803.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-817.8\">block4_conv3: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791526418024&#45;&gt;2791543634184 -->\n",
       "<g class=\"edge\" id=\"edge14\"><title>2791526418024-&gt;2791543634184</title>\n",
       "<path d=\"M89.5,-876.313C89.5,-868.289 89.5,-858.547 89.5,-849.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-849.529 89.5,-839.529 86.0001,-849.529 93.0001,-849.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791543724968 -->\n",
       "<g class=\"node\" id=\"node16\"><title>2791543724968</title>\n",
       "<polygon fill=\"none\" points=\"14,-730.5 14,-766.5 165,-766.5 165,-730.5 14,-730.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-744.8\">block4_conv4: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791543634184&#45;&gt;2791543724968 -->\n",
       "<g class=\"edge\" id=\"edge15\"><title>2791543634184-&gt;2791543724968</title>\n",
       "<path d=\"M89.5,-803.313C89.5,-795.289 89.5,-785.547 89.5,-776.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-776.529 89.5,-766.529 86.0001,-776.529 93.0001,-776.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791543795272 -->\n",
       "<g class=\"node\" id=\"node17\"><title>2791543795272</title>\n",
       "<polygon fill=\"none\" points=\"0,-657.5 0,-693.5 179,-693.5 179,-657.5 0,-657.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-671.8\">block4_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791543724968&#45;&gt;2791543795272 -->\n",
       "<g class=\"edge\" id=\"edge16\"><title>2791543724968-&gt;2791543795272</title>\n",
       "<path d=\"M89.5,-730.313C89.5,-722.289 89.5,-712.547 89.5,-703.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-703.529 89.5,-693.529 86.0001,-703.529 93.0001,-703.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791543795384 -->\n",
       "<g class=\"node\" id=\"node18\"><title>2791543795384</title>\n",
       "<polygon fill=\"none\" points=\"14,-584.5 14,-620.5 165,-620.5 165,-584.5 14,-584.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-598.8\">block5_conv1: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791543795272&#45;&gt;2791543795384 -->\n",
       "<g class=\"edge\" id=\"edge17\"><title>2791543795272-&gt;2791543795384</title>\n",
       "<path d=\"M89.5,-657.313C89.5,-649.289 89.5,-639.547 89.5,-630.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-630.529 89.5,-620.529 86.0001,-630.529 93.0001,-630.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791543953000 -->\n",
       "<g class=\"node\" id=\"node19\"><title>2791543953000</title>\n",
       "<polygon fill=\"none\" points=\"14,-511.5 14,-547.5 165,-547.5 165,-511.5 14,-511.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-525.8\">block5_conv2: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791543795384&#45;&gt;2791543953000 -->\n",
       "<g class=\"edge\" id=\"edge18\"><title>2791543795384-&gt;2791543953000</title>\n",
       "<path d=\"M89.5,-584.313C89.5,-576.289 89.5,-566.547 89.5,-557.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-557.529 89.5,-547.529 86.0001,-557.529 93.0001,-557.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791543990536 -->\n",
       "<g class=\"node\" id=\"node20\"><title>2791543990536</title>\n",
       "<polygon fill=\"none\" points=\"14,-438.5 14,-474.5 165,-474.5 165,-438.5 14,-438.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-452.8\">block5_conv3: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791543953000&#45;&gt;2791543990536 -->\n",
       "<g class=\"edge\" id=\"edge19\"><title>2791543953000-&gt;2791543990536</title>\n",
       "<path d=\"M89.5,-511.313C89.5,-503.289 89.5,-493.547 89.5,-484.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-484.529 89.5,-474.529 86.0001,-484.529 93.0001,-484.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791544073128 -->\n",
       "<g class=\"node\" id=\"node21\"><title>2791544073128</title>\n",
       "<polygon fill=\"none\" points=\"14,-365.5 14,-401.5 165,-401.5 165,-365.5 14,-365.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-379.8\">block5_conv4: Conv2D</text>\n",
       "</g>\n",
       "<!-- 2791543990536&#45;&gt;2791544073128 -->\n",
       "<g class=\"edge\" id=\"edge20\"><title>2791543990536-&gt;2791544073128</title>\n",
       "<path d=\"M89.5,-438.313C89.5,-430.289 89.5,-420.547 89.5,-411.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-411.529 89.5,-401.529 86.0001,-411.529 93.0001,-411.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791544155720 -->\n",
       "<g class=\"node\" id=\"node22\"><title>2791544155720</title>\n",
       "<polygon fill=\"none\" points=\"0,-292.5 0,-328.5 179,-328.5 179,-292.5 0,-292.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-306.8\">block5_pool: MaxPooling2D</text>\n",
       "</g>\n",
       "<!-- 2791544073128&#45;&gt;2791544155720 -->\n",
       "<g class=\"edge\" id=\"edge21\"><title>2791544073128-&gt;2791544155720</title>\n",
       "<path d=\"M89.5,-365.313C89.5,-357.289 89.5,-347.547 89.5,-338.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-338.529 89.5,-328.529 86.0001,-338.529 93.0001,-338.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791544155832 -->\n",
       "<g class=\"node\" id=\"node23\"><title>2791544155832</title>\n",
       "<polygon fill=\"none\" points=\"42,-219.5 42,-255.5 137,-255.5 137,-219.5 42,-219.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-233.8\">flatten: Flatten</text>\n",
       "</g>\n",
       "<!-- 2791544155720&#45;&gt;2791544155832 -->\n",
       "<g class=\"edge\" id=\"edge22\"><title>2791544155720-&gt;2791544155832</title>\n",
       "<path d=\"M89.5,-292.313C89.5,-284.289 89.5,-274.547 89.5,-265.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-265.529 89.5,-255.529 86.0001,-265.529 93.0001,-265.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791544233544 -->\n",
       "<g class=\"node\" id=\"node24\"><title>2791544233544</title>\n",
       "<polygon fill=\"none\" points=\"51.5,-146.5 51.5,-182.5 127.5,-182.5 127.5,-146.5 51.5,-146.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-160.8\">fc1: Dense</text>\n",
       "</g>\n",
       "<!-- 2791544155832&#45;&gt;2791544233544 -->\n",
       "<g class=\"edge\" id=\"edge23\"><title>2791544155832-&gt;2791544233544</title>\n",
       "<path d=\"M89.5,-219.313C89.5,-211.289 89.5,-201.547 89.5,-192.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-192.529 89.5,-182.529 86.0001,-192.529 93.0001,-192.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791544314456 -->\n",
       "<g class=\"node\" id=\"node25\"><title>2791544314456</title>\n",
       "<polygon fill=\"none\" points=\"51.5,-73.5 51.5,-109.5 127.5,-109.5 127.5,-73.5 51.5,-73.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-87.8\">fc2: Dense</text>\n",
       "</g>\n",
       "<!-- 2791544233544&#45;&gt;2791544314456 -->\n",
       "<g class=\"edge\" id=\"edge24\"><title>2791544233544-&gt;2791544314456</title>\n",
       "<path d=\"M89.5,-146.313C89.5,-138.289 89.5,-128.547 89.5,-119.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-119.529 89.5,-109.529 86.0001,-119.529 93.0001,-119.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 2791544314736 -->\n",
       "<g class=\"node\" id=\"node26\"><title>2791544314736</title>\n",
       "<polygon fill=\"none\" points=\"29.5,-0.5 29.5,-36.5 149.5,-36.5 149.5,-0.5 29.5,-0.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times New Roman,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"89.5\" y=\"-14.8\">predictions: Dense</text>\n",
       "</g>\n",
       "<!-- 2791544314456&#45;&gt;2791544314736 -->\n",
       "<g class=\"edge\" id=\"edge25\"><title>2791544314456-&gt;2791544314736</title>\n",
       "<path d=\"M89.5,-73.3129C89.5,-65.2895 89.5,-55.5475 89.5,-46.5691\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"93.0001,-46.5288 89.5,-36.5288 86.0001,-46.5289 93.0001,-46.5288\" stroke=\"black\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SVG(model_to_dot(vgg19).create(prog='dot', format='svg'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOy8Wax1W3bf9Zvt6nZ7zvn69t5btymXqyqxy0lsJ8R24iQKtvwEBDAKEsISChI80r3BAxI8wCN5QEJ5SIRARALZKSudY2MTl8uVal11b93u60+729XNlod9DQbVRSBzlSvV9384R2tp7rXmXnON/x5zjP8YIufMS7zES/zwQv7znsBLvMRL/PPFSxJ4iZf4IcdLEniJl/ghx0sSeImX+CHHSxJ4iZf4IcdLEniJl/ghxydGAkKIvyKE+J4Q4vtCiP/gk7rPS7zES/zxID4JnYAQQgFvAz8PPAG+AvyrOefv/P9+s5d4iZf4Y+GT8gT+FPD9nPN7OWcH/B3glz6he73ES7zEHwP6E7ruHeDxHzl+AvzpjxtcVWWeTBqkFGRACEmKgXw4QClJ9J6UMikHnPPknBEIEof/Vkuk1iilkMpgrUarRIqZGDM5Q4oDCIlAMrqA94FhPFwXONw7QxYgcv5oLoc5cZgKWkqqQh3O58N1BaCkxFoDZHJKxJQRwB96WilDJiGRIA6f+cO5H04IjFIIJQBQBKSCmCEliBFyFvDRvLz3xJRBKWZ1SV0ZrnYjUihCTJQmE1OkqUrqouL55SVWK0LKOOcZx4TUmsIaCgtSCkYXAIkgE0Ii5YyUhxlmIMXDVJVUaH2YCwiE1Fij2HfD4UHlRAYkmZjBuUBOmbJQ6I+ud3LtGMhobUgJumFgt91ytJxhjCYkaLctPmWszXgHgoxWClMUmKIihpFJM2Ho9piiZLXeMPQjSkq8G0kJshAoo9FaobXCKIFRkpAFWiaECDiXUTITk0SKw/ONMRNCpqoqfMwfrTeIHMk5kVKClJFSkojEnIkJpJRo9Ye/rQKBOKyyEIiPXpaYPno+IpNzQuvio+coyTGQ+D+9c5EzWUgymeADUh3W5/B5ifro/dTqsLZaS4QQkAU5p4/e60yMiWfPry5yztf+7/b3SZGA+AHn/i/7DiHErwC/AjCbTfmlX/p5ZpVkt10RIohigjGK5fFNJJ7T8zPSrufD50+pjURUDaVQjO6MRsDs9hvYekq9WGBE4MY1OKng3Wcel2eUkxmnj75KITO7IfEHX/8+739wymoLwghsyDgtAIn4aNFDStRlSQqBIUZSgkLCFx4eUzaSIhtc7hjGhJaGN1+9hdaaSmY653E+slnvkSYx9p5WZESGqS6YLOcYAQNQIw4sASQJQkveeOCJpWT9/Z4n24h3cLXq6aKgvnGNoVvzxr0JD+7fY97UPLx5zForus7hR0nFI5IKPDnT/PyX/izGZn7za3+PQhp83GPL65R1hZDX2G8fM7pAaAOb9cCuF5xtISRBXWWOjhusiKw2A7aoaWxJNdcUjIjcg74LWbHbB1zs2HQj+4s1bujp3MjZ43MukucnXj3mxnwKwnPr7kM+/9Y9Hr75Fo+ebbh94zr/xX/93/Dv/PIv0KkVtXmdX//1X+f28oh1eEK/L1jYkihLQgj84r/0L/P49DkzI3n6wXf5/Bd+jL//u1/hu19/zFGZ+O63vknnEt5FZm89YGpqPv/6TaQQTIQn1Ymj5oxCtQzryCpYLvdLpOvRxZzRNUgV+LM/+SW+9o136LNExsisgLYd6Lod3nsEMA6Oy3ZP33YU0zleZG6eXAcNc+OIYkIIgd55ZoVh7APtbs8gJds8cqeeYU2FNyWSRJaZcjahyImUHIVQuJjY7/dcP1kgzQTnNjy5Oke5RIqBo2snBJe4cbxEWQNZIIOj7R0ptLig+I/+s7/94Q8y1k+KBJ4A9/7I8V3g2R8dkHP+m8DfBLhz91Zu5lMmU0u7X2G8Jw4jtmg4mkW6TY/uA6ebK/oAN2/epKoqTC2xY8ZttzSFwUvD2MPtm5FlEbFW01QnLBd3mE5q+vN38N5z9fwRL043tGNgUlbYpjiw8zgQfaKwhmwUE2DWWPa7ERsEo4/IUvOiG5l6KHVgqhPCRyDx/PKKmanpp5ZGWUodcHXBzmek0tDvGVNGmYHCWYTQNMayGfY4H3Bdh0CDEYxtYMyCi/OBy52iIzGbHXF0MuMLb9yhjzPuHVuUiJSmxEfQPGVYd5x38MYNTaHBVprT9TM+d/8Nri+vowswdg7aIgSkuKOSjt3WsdqCipGmKThRDZ3PlIVgXgekVFSzBdk3ICWSgqQDMTQ0VYnMloe3Le98+D4TnfjO+uBFAeyExCbwLpF9pqgqzp694PFEcfvhfT58/Jz7N29z7/ZDjNG4MXNcC56d7ZgVUx5vBcW+Q08CairJEb72e7/J62++xfL6A56+9x1yHrh78wbff+eco2ODUZapGmmPZqSosKVkvdsxm0152q25I3tGOlZ9JHtHGxWVtjhf4tyAwvDm658h9XsuVyPNrMTIzD5qQoxIqZhVgtEPrMcBnEPFjI8RYQNWBcbB4dLAZH7CKDqCA60tXiYwBd4HcpdwKmKKSPaB2bJAypJCW1LyjHFASsmsqZE5gTbU2rC5dEyVxZtIOyZ8O7B1gcWkoYqgbU1IoFhj9A5J+bHG+kmRwFeA14UQrwBPgb8G/GsfN1hJiZKGMKSPXCFHFwqaNFLmnpGe7bilHXrqomRiGkwlaUzCJ01WkiQkwXmG0FFohVIjMgqwlklTUVqJLQwh9my3W7reEWJmVkt0qbAYYoyE4LF1QSk1SguMEpQ2sREZJRWlUvg+sAuZqDNlY3FCYvDQO1ZBclxpUh4ZM0gUOva46Bj7yBATXgmU2dEUFUPXsR07sgffe8bsSVqyHwNDSPQbTy8mqMJy//YNjk8WNLOSsl2jZCJ1mdgctiZ9u6PzgUZYkgaPoy5OGIMnSomuDt8JRmQcgURKIAhkL+jbHQmN8SPGNiyExZYCPQ7oWiCNJeiAUooxSYywxOHg0mcUiczNqeXscjh4HCqjlackEoLER08UEZJBCU3fBZLreXq2ZvSeazeP0Cozjj1CCGLMvP/sBUkExs5zqQXXy0yfAhfnL7j16gPuVyXKKlKEpqopi4KiKlCFQeVE1VhySEgpabuAKNbMSoNPe7QfCS7hoqT3mqoIKCqsUSgRWE4WXGxOGVOicBGhEj4O5JAgGZzwuBgIwePjYVukUyILRVIZmTMuwjiOaCnRCIQVSG/QQlDsW3JVYLWmMQanLJUt8EERlSAOAwUCFxzCWJJIDK4Dn1kPI9NK4V1inyNV8HgfGPuBzjgWQhKJgEeLxC64jzXWT4QEcs5BCPHvAl8GFPDf5py//XHjpVLsdpesYyQNiXXrmdU91l8S2h39tqPfJaSE5bVb2GmJZ8Vuc8lurEixwPeeGDz1JMNgWIUe71p6cRO3OeNid85qs+Hi+QXfeP85LgvKaYU9WnL/5jWq2rB89ITVfgStKcsJAKerS5bHc+4Wmv2mx4XIZr9HBNiQWLsWrSVKKVzoEamj3WwJIjFEgRCZdevpfWLXDxTZMJKwZxrEBbJQTKuSnDPTqkCVkpmBhzcabp7UfOtRAWbKrRvHPLx3h27o2fU71Nhz8aJl8AteO5IQE1I1VHVDFJLj5pKQ7iOaL5By5NFFy+byAlJgUhlspdAKQujZt5oX60SSU8Z0hbGK60VHdc0ShcQGga4yR0cRWyZSGplMvsg7z3YwiWy7jiwcr1WS2595yMk1z9Pt9zn3K1IwjEnipCNTMisrotGsOod7tuWLmw2ff+0Gz89P+bmf/jOo7oJHT8557b7hs6/d4ve//ZT5HI6vTUgq4xWkpLlsJf/LP/6HhOy4Ol3x/Pic2yf3WMxLPvPqCV/7vQZ75wbGCDYXAWJi3I8sj0oeXn+Ty83vMgbD1o/MbaTtA09XK+rasFABbCTZgj/4YAVCsO1G5ouK9cUaW9QI3bPbdNRSYo0gexA202vw/cDl6ZajWcNCH7PePkaXlqtuStVkNJlCSKLUVE1i0VRMq4JYLunHAcGI9ZK12zFtJuQoef7sHOccxUSi9R7nB1ZBE0MLY+TZfoX3idW+Y95UdNcERira3RYVW25cf/Nj7fWT8gTIOf8q8Kv/r8amyJ2TmkJPuLoquNp2DEPHhejhVNC2BqkTdXXMslHMFwXrXcnGGUq5Z8wNSkzIemS/G/mDK0WhFBNTU1aJb739Hca25eLiKVcvNsyaCfXxgfFPTk64MZ0ghEDevEbpAv2+JwqNGPeYpDmqSsx0QiME0SdSHIlCouIh8HV87YRSQm4Htl3L04uWMQtm1hBloB8SYQRhamKM5JRREtAlr96bsJiWKFVw3JSURWZRCu7dziybhqFcAEeUhUFVFWVZwGDZ+BE/dqg6klJAGIX1gklZIVRCTeeofJOrrUN6yYthS9q0OCEoXUSNkt5qVFLILDmZGraqZzapWRYKLyIy9+z6kpu3BESLEJLQrjGl5OTmAy78h1RD5jxKGu2o1IrdrkWLOaooKZRmzI5SBoKUqJyIKFTShNGxiSPtruPzb/xJvvv+Byg8A5HzlaMwBa+/do0nmzVNyhSlRdkaIQRJCIYo8Enxa7/19/hXfuovczX07PozklphyzuIes7xtOS99QodB1QE5iO7dUW4HSAnHImYNBZHowf2tsFny2qE124d8/zZhxR6QypLIgUER1MVSJOJXqByYvQRQYkwQBaIGKiKEpJgVhUE4Kp1zKMnjJFu7ZlOp0SRmZYamRNDcOAtPm7wbkBHyLVDi4rgPD5lEhnvPToUSJnQQnJ+eYmRAWtLxiSY2wJZT7GFodSCq6fPWF28YHFyDFl9rP19YiTw/wU5J+bTkqJZMIxbpM4ED7nMDG3PxT4xKsm0MFiZUaZEZ42SBbaSaDPB1BUxK9brLclBMS3QtqLvW9bbHdl1rLYdMUZmlWU6rTFKM6srZKVIUbBcTJhFz2lIeGFw2VOXCWkkc1PQlhVKRrZ1yYDApoQUE24cnaCVoOOKNjpkn9A5ooqS7AashlJkyrJkcJHN0FE1BcczwWduNdR1TZsMTWEoZWBSS4ROhCSZlBahJbUV1FVDzIlGa+gDzm6YNXusPGxbzlYe6owxCisiXlSAxI09ISVUVvRjj4yW0gvUxDKEiJ02eBcROVELSUJgtaAfE4PzCKdwOeJ2gUimlomYE1IVoDzTWqDR4ALb7Q7b1KTgESJDlESlKAKMUjBKgREZiUKQ6AdH8B6pJTEcovI5JnLONJMZaiJQLXRRME0Q1CHqvA8BnwIjkcXkGl2+wu32FDlglGF60qCEJgwDMUTMEDi5l9idtqic2A0Jaw45cjcElJQsa8c+JISGyWRCt3rGUdHTlhC9ZgwZrTI5ZXxOSC2ILqK0IniBCYIkIkoeEideRIj9IZMlI1ZmBBHUYSslrcB1I35I9PRYK3DjQJAgnCCJhM+SPo6U2pAbg4uBHC2TRtL3ljhEhIKFLqjLhnU7kkVit8usdns+ytcwDulj7e9TQQLRB0QIbNdXtJstOUKbe2oH3z/f0HvDm299gXJRMDORCk80GTmZMDlZcny04GojWW82TKcNi/mczEDoM9/+5lfZbK/Yp5H9LvLGvRNSkujaMqss2RYsG8XaCRa6IYTMrTs3kbrmbL3mybvn1I1hTAWf+7FjSrHl+EXNMApW+8DRcsq1o2OEgK4oSdLQVBf0rqQuNXtfMZ3MuDafopVgte/JruUX/8IDlM7YtOHFZeS8nTGbJIyAuR1Zry29gT/xuYBWKyTXuMhLghtITGmOT1hyxX58hzQafJZUpUOoHSdTQ/ANuzHQbtbsh5HRRx4cf4YX3/k+x3g2ssamgT4LJktHXdSM/SUXQ6R1MybTgHGCISd2312zmJbMliXtqNExofBUcs1lu0FGwRBGRgouO8Vmc8pf/Pwd/vvfuCKEgEiHNN35LjC/2jOrLVol1r3g3WeX3Lz1IY3QrNodYXvFzByhdcPdB2/xlwr4337rHVb7DRfrnqY0RC1IosZkyYjj3edPeOP+a3z3e19nlY7RpuLP/MgrNOWSdVB84/e/gT5q8EKwWnmauuDr30u8ekNTzAOjLLHBcGS3XDMjpSm5MXuLD7ZvIyvFTM4Y44pFMlzmu4x+oFIBtEFazz4EprZBdJFuvWKMntnxjP2wZeh23FkeYURmIi3ZZ5z3jEKggmCIkmfnF6yGQNUIFs2U+WRKKhIZxThEfEjUc4PNDauLK0a/Z3Gy5MHdO3TDyNV2x8X5hvO0JYSEqUpWzyJ1mbl14w5JaNarq4+1v08FCfSj459+7XvErBndnma6pFYzcuxJpcdFQUeiHB3KOpAVIu0JXcfx/B7HSyi0QYqKs4vA08dP6EbHfrPm0dtv41QCp5gvKkZZMDkpMWOLsAV1VVBWhiOr6YYRZOb+/YaYphRGcFxrhtbT68jnX5Os1yUUN3CD4jh4ZEgIU2BwqJPr3LMTRneDwYMuG2bTihvLKVbBarOi3O2YFDdYHlVI9gzrRGzusaznGHvJQjnmNvP4meX2PFCYANR0+QiRHVo7QlSUUiG1I44eIRLBRUIUNLoFb7nY14xZUhUFWhrGfuDB3Vf45gdbPBG/G9lQoBBsTwOZPWVOGGWoZUbpEu0duJYgNUkqhj7SdY4oFDHtOW4CYf2Up8MrhxfbCFQ9obsSrCP8zBfv8T//Vo/IiTYpCufo+p5pqQkRssn0XWCz7pjPJnzw4glTJdgmiGFg13Y8fOVLXDzztF/7Z4zZMwjDrbLk0ZiQwKyY83vfeBvZ7/Buh0SjE6QgEEXBL/zcl/hnv/91bt7QPP5gh8gGhKLr4HItuGEFrrR02XMsJKQRUy9oux7ICKMI0oD3CNNyNWzBG7SMZFOQYmazO6cynqqoKCrD2gVEHEijBTljFAafRlLKuNzh9o4YLSIksvect3uEMohUM/aOVnu0Vqz7PUImTuoGiWG93bDajuSc6VYteymZaEsXIiFmhp2jqAu6rcPIQFHXvLhoyT4Rf1DS/iN8KkhgHHr+4O13kFkwWyyw1RHHc0OJ4+I8QIy8//QdnmXJyVQzO2q5utrgO8ez7VdR2oOZMPaCD5+csTt7QedGSIdItq0MtrLUkwYbBxpbM58siEpw81pDYEklS7ZPn6BtZNJ4Ti/2VHpgZMDZgofLiqKExXKCnS7A3EEqyzhuSPs9KQ8YfUwQmUlVYquaHDxCK1L2iAS3hlvs2paxd1xevsft6Z5Vb0FUTBYn/Pire4SWbNuSf+EVS6EG9kPmbLOky1NuTj8gBo8SJY3Yo9kwn3jkMCJEw9gGhm3g0dUWqUuqWYuuZ6isqWpDXU157dWHpEFS+Q68IcnMbDKjjD3ePWfbDQy+pvUZoS64Oas5Hy05jXzt3RZZVvyJSc3q6jH3br7O4xdX+KgZ+kBne64ZwbmxvPt4Rxe2vPG5hvefaWgTQkkWiwVlXeOjoAqB8/WeF7uewmoeffiUGydz8IIQAtgJx9fuE/htohJUTU3ImlWQvDKzXPWaL91/jS9/7V3uXyxQy8T1+YLHT55hmjmX23Oenr3g/mev0/qRR9+7YLGYEceBQihWnaTKx7xSXxG9pii39EPDbP5ZHr14H1vUkGG76ii0oTAFJQppDdllUm6pippr1ZIkAsEF9s6z2+zo9j0ByXw5Zbdf0dgaz4pnp1uEgOVkyYv1ltoWXC8rZFVwdLQkJs9uv+fpfsvT0zOsUVzOlyA1/W6PlCClZRg6fN9xhUTJw9bDjwm13SGUxJrMB91A3/d0g6MpP14c/KkgAXKmzZFJFIzjyPlmhaRGy4528GgEdQiMEdp9IlWOdnSM/cCT00ualIhFiRsz/WaHa3tkEiQDKEFZaYpCMasURanJ0pPrApsFXnII1IkAKWMEpJiRIbEbA4PLpAxRlozBkGzCykC2Gq0sKkhCfUiQjRmKbJBGU9mCIAWeAEEgtaEsMilGZIJtC0eD43InGG3ETiK6EhhriSKh0wyRIs5nMuaQZdi3VEqBcSjdkT+ar1KZJBRSGLb9nvO956hR6CIgix5hDMSE9yPzyZTeeOgNoZFIoThaTJHOEUaLlJL1MKXzLUqW7ENL6wWbzhGF5Kg0dGPCDyukqFDmGCMSndDIMGKqmlJ4LoWgHTLSBl574zofvLuiRFErRV0VdD4xRM8YElJEYswsZw1953BjS/QjfRhJEcbRo6VBJEmQB9Xm0HZkqZlbkEHR+8gSzau3F4jtOaqZcTWs2e9a/tzn3uTvfuV3UB5CCOQkkNagTYU1c6y+IiaHRlE2NTlnUmjJxiBlwAVJoSXeO6w6qA/HMZFCJuARMqGQ+JzZxUSfMip5eudJFqb2kDrsRaLrRyQRJaYkF4jKQKUQSh6CxlngvGfXepRSJKUZx5HCCBSCEANGCrTQRFsQhpGUIzJEFIoQEiElfEzI5OmSx8dIivZjze9TQQI+ZcY24jJYt6J9cs4za0kFhD6xvF4QN4mOxMWlo1rtDqqrbuTsfIvSgiBBZrAcUjbCgJWaoi6Z1Zaiqpgsp0xLya3KIktBTIJKDggZiIOl1AajBJcXjhfbhA5zlGqYVhWFbMh2jsznJGmYN1CYKy7dC6a6JaUtBYaQF6CKQ3CoKDEp4qXEx0AfRgY/MIbMe086tkUAM2dWTjmuCzInyDxgpGEIc/pRkOUEWyzJEfzaoKtMITQu9vhRUzeaJCJkwx7JNs2Y1kfMl/cwZcE+7JDCoSoISXLr+JgXVwFTZXI4CK+bcgtyyxjWyEVDMSS6wbG5NBS24fr1O6yuLrk/X4EUaClw7gwXPfP5Ay5Wj5jT4OIlJ3pkYVo23sAYuXwWeO1zUz776pTf/O3nVIVhXpeUzuFG6HrPRCXabuDVm1O8mfG9D7/CftgwPb6LG3veePU+HzzfUGpH3u9xfU85FSxky6TRCOkJEm7NF9SVwesFwjacn3ZcXYz8+Gs1f/rhW/wP3/xNlEuo5KnqgiwUqld4V9CUmcErbh3fZtc6hNuwUROW0lIVAhcSw1bgho5RBqKS+FGydh0TpSFHfIowRipbYRTILEmDp5wuSUIxkYJrsxntkEhuZG4thS4otSb4gO9Gdt7hhsy0bpB8pMnwEed37MeMwkOXcSJSWE1WgewiPmZ8ikQXyDHQJk9EIkWCmNgO8WPt71NBAikmdrs9Kgt2QpJToh/CQVWvYP/BgIiBqDUmC7wdkC4SQ0YVAik1tQGjNEII6kaBVJRGsJgWLGZT9KThwd0buLhlPhW8f7qnKw0/8wpUnLPdGfpxQZaK8/M1RTVDFIKmKGjqBZNZQ10vcMMSVSiKwiHd29TFKamPuDhQFCtW6xElG4J3SDOhtJLt7hDsVFozmR0zaQZ+cvlTuHaFEJH5yQl1NeG8uwN0ZO+RqiRzgzFFPAkpNdNFR1loOtfQXkwIsaGZ7JEl9LvIw5MjxqMjkr6LLSfgI6ZvWPcbfBw4H1bcX8wIbseu3VNONJOy5+mjd2BoAY3LIyJvqcoZ6nikrgoe3J9ycW3K2aMNDCvKqkRmw9nFh8wWb+ACBDKmhtFF7h2PuLymKC1tX/G1//Upso78zF9+QFXOKINl955jbiylEly/PuOD7z/h+s0f4frNt/j9987RumY+OyKRSQJ++V/885x2Pb/6q/+Ajd9ju8DD10+wEvq8R+WMKY6597mf5N1vfY2vfuXX+OD9HS4m/qevvofYbPkbf+0v8t/93d/Bpcz33/2QWdWwXl3ixhnTCgZheXj/Fuebb+NLgRGRqzZSBNjLgqvNlNokUt7gowcjacYNygjOR0PsoTIHpZ83Ge0FUkq2lyv22TPJEvdRlqhsplRC4mNiOwSSTOyGDXF0uLFjHTVlJRmHxGoYSW2HlQKkJpFxgNES5yM5RDofEYNn5w9ZAOEjWYEXAh3z/yFL/0H4VJCAEAItDT6ng+JKSnRKoBVSCWJIZGOoEnglWMiMqw0iQ1VrTJHQQdMLmJaaUguMFJTWEJuSoqpoJlMmtiPlSMiCutE0KfF0k7h7EqgnJfVOE4KG2TG2KpHGMLUGVdcYfUjPKdsglKTbO7Z7QyUCRVFg1KEoychA3XSkwTEMO6Kdk2OiricU04ZSAr6n211hZEnICVNMKAsIeEYvkWICVEiTkaml1JqiNIi9w40BoQumi88ipGW9+xaFlkhgEyRSNlgEafQkIj4GtKqQyXJ6ccbdZo+tRuL5GapSqNTRblZMyBSNQUXB1idOyi1jVWPLiItXoGouzrcUJtGYkTg6zrfnfGb5BsW0ovBwrb5gDNcQqcAUO0qjmRWCXe8YT3u+qp7y6uue6XHB5N6C/jmkUaFEwgXBZt+ycANffP0WrrskuQdkKzh/+pTnz97jzc98jq3zzE3B/RuKeyfHtMOeeZ2YqMzV1VM27/wD/ItztJxSlT2uHfhP/q2/wb/xH/6n/JWypo0HXcVu2+LHwIwpHz52FHXFretz1qsN5+srlPCMMdLvBTrVSGtIUXKVRtzYMboIOYKAonf4EMnKUFhBGiNxG7jc7dFKYUNgVIJ9ltQcagCafmCwBSEE0uCIMrMNDpPVoVIt9/i9IpqD1DqJRA4Q08HrUckTosbLTBzhauvY5UgKEZUFUimyy4iUGEio+KkngcMfKwRIQyIihUFqiVQCnSNIiVAglUQVionRVClRN5LaGrYBTBA0pQERyVpjS0tWxSFQUoPICWOBnDiuNSJnOifZuMyxTiAUUkqMMSilKJWkKBuEskjJYS+pJSklQhIMoaEoj6iaPSJK1psJE2uZNi1XbSIlTUFCVSVlabFSIyUEIvhIFBllKwptMHrA9Wek0JBlidUaKSW4QG3BGMcIyJQRKVMUSyBzvoJaOJbFBFJJTI6sJS5HBBqZPIWCJGA9CmJOFNozDGtsK1F5QIZErzTWO5AlUhUo3SFjh4iGkHuc09h82Iv6j6rp/NCTCSghMdqgTWYcIpAxozrIsHPACHAJutVAOu0ISqGPRpqbM8LZHokiJujHjmHoubac47fnBD+gc6ZoKt57+wNm8xsgFTIFqqamsRXb4Yq6UVAPgzAAACAASURBVCThcf2AcBuEyEynSxwXOBe42l+xXB6zaluW08mhGlMKGHvuHj/g9PwRKWWaB5LT1YrOObTyiNHS7hTJKmqjiUrStp7gPDkrTFZ4mRiAKghESjiR8VKyDYHsD8FCLxIyaQKBFoFE0G53JDMwkkmjR8hDxam2ApIiAd6P4Dw6CUSCLkLM6eB15YPXK4OiyyN7Hw5iJTJJGRCH9ySQUEHg+ZRvB6RSTKbVR2W8CVFJpmiUTsSscWPCagkk6sZgTEE5qbg30bjCcr2RbNqO1oEWBYPzGDthMjPUpuZouWdZRnQDVbI4M6EoAt3O4XrN9lJzZjP3TwY2W4HkmJQhYSgmU4wxiKxZbS4OcYeqpK6OqeopWr6CyO8yhjWmeRM/rpHxEW6UhCwp5rdIRQlSkASIFOhTySgyWUQm8yNEkdi7gSfPVhSVZTaBsizJSWDlI8r8DBn2XG229GFBqkro9uxGTxiWeJtRSrHbOXQxQYqW9c5RyJKkBMN+RyIyV5rTyyuOblSYZUtj9mR/IKJZMePx+RW3jzO1NdSyZ7OFUxSv1zBRiZ+6p+jGxIej4+n2JkEc8eJ8TcyWaaVoO8EQNtQ0JA8qOY6mjpgF0UTGTeDsvOWOqLjabXHTPfqkIEpJIQTbZyvelh/y42/9KE8vntDtLhC64Uc++0Xu33uVX/vtf8K//Ys/y14anjz/DWQh+e73nvJKsWReTbn+4CGlSrxTP+UnXvsRHrz2Kv/V3/rbfPnrX+av/+yXmDc7/vW/UHDvM58lp7/PHs+DGzO+/X4i9y2/8ubr/Od/63+kajJHlcSpmhAl16uD6i/nns45lCwpJyXt5ZYhBjTgVQAXGPcDMUt2fSBmRYgZFTNZ+EMAXBtqn3Bkdn04pB6FIglQZPqNI2a4HBIaCIPASU8MkpjjwRvNAmSmUJrCarSWVJWidwmFRuVMTJkYIR6q1DHCMPKD6wc+HSQg4WhaMHpQOjGtLMXUYFRk7DLbPrAoLUopmtJiZcXRrZrZJFPUxxSm46i1bKKm2wt0Uixqi1KCk0VkMZOgNIVxlBL6nSNpz3QSebFtOa5n5K5jWxqKyZLYSWolGRWQBEM34sWWOFzwYr1ncXwTKTKjd5gMiCkpzbBqjZHnbHcNvW5oymOCkqh8UIrprHEZlPYY6QiiBGnZD4Gs5xwfzagmJ9hqSpKZFAZW6zP24weIief0KlMUkZNFZLM/xw0eW2iq0oIaKOx1CglKGZJr2fqB2UxTyhWDW9O3nnUy3L/1E3grEdMjjBsxKXDVb/HdllNG7FQjS4XLDbP5Pfqw5WK9Qw97vM7cuKZI0wcoueD5ZocWjuh6yirSR0PIGcLIrBR0vsBIRSkMO++42G4P0trpgsnW0Nkd4c0HzI6ntF1kd7Hm4vQxWmb63Qu6fct26Ll16xX+6k//Ob78j3+DP/+n/iQxDMQYeXYpubMs6IgoIRnHnmZMvLN6xmev3eeXf+Fn+Ucf/lNevfkGs6rm3/87X+a/vH3E6D05enQeSEHy2sObJAPbvcMPAtcUzBsBKvHh6TPawRNS5GQyh9LQhcTQbXFtiywOXtu47zkbAirDuO3I0hByIrmepDU5SIwc2GXIUSBTwkmFiMOh4jIl+hiJURCDPCguUyTnTJQSQyIpixAZHxMhBlxIKJ0py5qJgNO+RShF8gkrQGRFJvw/hQQ+HSRgteaVB0uGACppyjJQVxZbSFxfEJM46KaV4GhpyZXg/q07HE0bVm1kWsxhtuEoJ3ZdxaSISJVBKq4f7xm6inUHcipRyZHSNXR0+H5PIwL7zUifa3o9pVV7HjQVhZxTqpr9/gLvR7TWFLbm5FhjtGS3O0MkqM05enpEWVlU6ilV4FLc587RXaydMfQbkjQYY2lHR0oBF3sKcwutLLiR2j9FSMPTneCdp4+4ce0GF+dbRPBcnH7IYuK4ew2OloZJkxBccm3WEOZThFxgVMSIETlmRg7ZAysipbyi0S1Bn5PdQHl8jawNbdjxyqTlIk4IqWB7tUPXUJWBi8uEWz1lMZ9g5xMqNeWqNfh0wTC1VI1kcVLgqor9YGmdY2q3NP6csZ0gQmIYLNtQkUOilJmsNEoJfDdytYkEF3lgS6RQiFGwPbtiUhi8Mxxfr/jm+494eFQyLRy9Oyfud5yOntnxbb7z9mN+93vv8R//e/8mSz3hp3/U4MSKu3dfR2TN1WrHj7714/zDb/wOKTgK7en3I8+uXvDKjWtonwj9jtXeMaTAyfECYSR//ed+jF/7R7/HsO4YraIaIquVwxiNIBEyRB95b3OKyJAARSLGTFKKEBWxHwlWcW2uae4sKGVil0rOT6+YTGvuHUfON4FSKlZbQeczS60ORb5SkDUUQhBUpmkqtqstQ/DIlHlx5eiDQ8UEKdFkQVCCQIJ8UCHqytAki4mRUSnGGFEp4YRCHVql/ED7+1SQgFKCotSoUSIlKKVAaGQ8dG4p68zQ18ToD3XiNdTlFHRzkAfHntlEISPURqK1JKaAMBorBWMSh+48AQavD7++MZP8gqwHfB/xSXG9mBE8XI1XnEwmlCnTx8ikLDFWoikoRIIs8SGD84hqQm0zkhaJByIqR5AJISOjdyhjCCEwjh6Ep48OKw1RCHLck8KawWm69oTLZ0/YxZZ8kbEqInPAj5LTVeSV+wXSRMawp841iYhVFkQi5wGZJUJGckhYM6AYSWKDEB4XDqWnQmlC8DTK0RM5Q+C0RhEp8oQhOQoBykzIGKxWyCgobcRmhS0LhILG9MQ4J3jD7VlN3q3ZrAIRUChiCuwjGHHooCOEQmTJkDO1z4zOkbXCScEwDDRFSQ4ZoxVH16Z4P0ASZAGD80S/59n2fT77o2/y2c+8wZ3mOlEo7GLJMZkKASJyNKlJUdBgOV+f0QjBGCLoQBag6hKFQEhFkaEpNNdvHGMmBW8/OsOHRBCHhioTFejjIcIfgqRPgeg9Mh20Hg6ISeLw5CQPXaRcpC5KfuwzR8QEXTHlbKE4mU+5fhQZhoSLjtOzwNl5xBzaUqEQyGKKIZCUp7Fwfbag0JF922MqwePnmXFIJAk+HzoKJQQqfdRjK2QqLYhJEQgIcdgOkCP5B/b5OeBTQQJWS64vKnatQEh7YN40UlQGIywpJSYzhU4Vcj7j7nFL3RjaqEFrrp9ELIeWYz5rkAolJTEmulEw6gUxRzb9FpeOmFX3kFKSdaQeO4bYUemSqr7B7TDh65tvEc0FyyIxq2vqSUudp4xyiRWeFDJX+w4Kw6RQZDqIK/b7nqoKbLeRGKY0pcJ3gUHs6Yh439PHDcP/ztyb/Fq7puddv6d729Xv7uu/09apKpfLcco2OE5i7MIh2EaOAlgIJgxQZkBmySRjEH8BREICCYEiBpEQAkRikdhRKMvEkcuuKp9Tzem+djerfdunZfB+RgxcCQMsnXey9traa23trfe51/Pc93X9LmHJsxnaFHTiQHe4JYXId779Mfv9kSvRcW92j3ndMR4y9kNNPB5pYiQPmiK13O5vQV2RVR0iJXQVWM8HQmpox4zZ4kCVtaRxROeKcr5ClG+TdEXbnoj6MWemIYmAnc9h3DMUFdLM2DrPxeItLs/vMYyJ5O4ozRYpJNFAEhqlblAxY0gFDy83XA9HuqOFckkxkxw+t4QI5Twj04lWpAmXFmAbA/W2w7uR0hh2d5ZV3XLqMrJdx+XmHl0aiVETx5GX2yMfXOTsBslf+5V/g48++YT//v/4L3mvfAdTb3iZ9lx1OStTInTik2ff5/75mj/47iuOdcevfe1nKFcLQuz46nv3kFXOB2/XCGWoFmv+7Z//CofDwD/93qfgEtqAEY6dlCgREWH6oOp8nHYSUiONhKTRBs4XC+Z5znw5o4/wjbcNX393zT5JsvqMZqmgKrj3pMJai5aJ9njit36nJSERRIQQbFYPaX2LRqAzy0Wl+fK9hhR7rrcj3/uk53f+8Mh2sDBMbAqVND5FlJoeF3nBMYyEACZGvJDINPWjftz1xSgCueFyNScvZrgIy8qTC0GHQsYAQlPpnjFJsionKUlzPOERPJ7fIMOWg52jgqC1PWezM6gybNdy7M9BzgnR07tEWSbq/ITI3kMZjbWWamEJRJxNUCz5uc2X+Naz7/B58YJfePLnySiRRYUcI51zDMNAdB6TdXTNxwy+pzmNfPbKc3mmGC1wd4vSLUkHxrZndbbmh9d/RJs1rLNL/KtrkA3LVcHrU0+hFU/fOuMDeY7Re+zljg2CQa84NxsG/4jnL76Nc4Faa/a7gbPLirR1JA3zK8UxbInJ0nYFWdaR3Innz0BnCq0E1ftvE4PE5VdssgrXfx9lO77XvqZKBfeLJe88WROzMzbrhPR3NM3APO4ZXCDoGmEFg9ZUs09QvsUcM6z/Bvn8J1ktv0XMcrztebCWqFxgUmQ9Kzm2fpq62EQr4EXnMKOnUB1PbhqqzJCS47f+z4/5N//1c9776l+BlJit7vOVfyWB8FxsG3xQ/N4ffIfZeeAf/NM/4M995V1++i/8EqI78tn3v8vZ/Yf88+/+Pr/0F77JT3zwFqf9AeEsJgWSyvh3/q2nFLnhV/7iuzy5r9jc33Bx/yH/w//829w/uyCJCBKWxYLN2ZJMZ7x91nG5Njxras6Xl8SkMMIj6wotDTFGtMnRRjI2DWfqluZ0w/FgyTeWVy9uWeSGsHoMck4XB6TOeOvtjNO4pLbXKLfDzy1zMcPrFcfbZ/RkzGcdUmas7p/x6G3F+mLL89cNH/3wwPPbBh8dPk1TrSxX1Dqn0YnMeqIxqJiIWpLCF3w6IITAlDmZEAhnmOX55BtHYNIJTIYMI7VQZHmk9xJcpI97NvWRtg+4IEhREIeeKBtGWeBtxAUQo+PUW0SVSAKcC5hSgJhGKTrjzTnPATAvJGuT83rYcWtPVGaJCiWSERG3aBJJOTJlaZsGHxKjBZ9ymm4kBcUwDkQ5TiInmaj6il40eAU2BYb2RK57spmk7zw7PI/zhBGJGMHkCRsTmUrEPFEEsG2g99AKS58KLlTxps+QEKGgPwVOIWEHT6YPyAJ2p4xReOZVQeo9Jq+xJLQUtCFCUASluDt6ntQCjEAVmmQPEBx5mqzbvQ9Ib9FSAwkZEyn1BCHpLZhsRV5GWgfIkeVCQooEL0hKY4QkCUUjJnluSgnrEyLAZ89vKWXk6uKCD772M1zee4fl+hIhM0xWghLEJLlXbhDB85u/8e/yv/yT/4zHD6/YnRp2xx3ZeGDXHZnFM7Yhsd99TjMeiR5Cagg242L9mKr7CKkS+o1JSipD23UMJJ6884Rca+a5JDMV98/OiAKuqmvO6oitF5T5BoIllwI9WyC1Ync6orRGSOglbEeJGiVDH9Any+AKhHPs9h3ZwtC1jsoE5gWk5JBJk8uSIUhMmbGsDXbIuJwp8hy8zBG5oTKai/OCjMDd9sRtI0m9RKfp/s2FAa0oZGIUkxUZBPoNEPbHXV+IIqAEqKSojKJcLIkmkAtFlSyl1DQuQTVnXVqE6Dm1F3yyf0mlM1Tc0h0gRUWKin1refGyR69rsjdjOSUnX/xaeZz07JTkqk4IKYkqkXRJJSSCASkSRXXiy1fvsDo1/OD2B4jwhMtVyYoTJnzGrJgaRXZseXHsWWiF9ol1NafrLc3piLMNUkp2+yMBg38QCDM5NWh6B95DCgx3J17tO1w7snwgmS8V3oEaJafcUYsDqkvcDhJr9QRf62Fz74qyLBAyMowOmyyjKxntRBmaVRmygrPVOYdY4YPADg6tHW3TUy0yeg1WB8q2oVM5G2XxssXELdf758zzEhtyTCYRUbNZg85yMnli71bs+zXB53zy4sBPPH7AYHOMKDj2FlOuiP0Wj0TERGYSTknyJLEp4lKElDEkz+tTz/3ulnv6gl//q3+dxfIcIx1/YsxXyoCeekTeex7cu+LX//LfoutP3H36If/oW/8rv/YTX8WddvjwgEZF6FravuHL736d3/7ot/mSWuJUhjOaUkAzLPnsWnL/DG6Olq+++2W+8eWMICM6qxjHcaJBO8GsGBj8nkUJp37EJsesnJELiQxMZIToKYuCmc452TXHUTDILa+vO8a0pJWwbiXz1ZFDSjRNx+VKcn/p+OHnjvOl4bh1qDrwznnLvcpztjLoVY5kjhcFC9Xw5EFGO/Mc2yXbneMIuCERJaAFWkRSnqO9RzuBzxW5SJxG92PX3xeiCIQIm80KnTkWtSakOUP/kkNviLJAR0+uPUWm6XzFcdTkokZxRITAvXXBbWtwnSaaGR/+6BkXLnF+lvPwXk0ICjcGvv/ZLep8xpONoelachOIWpIphRcJWWbo1FPIa9brr5PNH9P98Lt8+8Udi+GWv7wM2GGLKSJpBJ8yoje8OhzYN4bX22f0fQtRklVzBIroPdvdM/bXLyh/fkXd5zS7LeL1iFEddpVhfUFoJO3tjpM2mKJEHzWb2cB6ldHtelpZ0lVLqjpyJs+ZnT9FGsmcBQpBaY6s14EiBEJoKWeGe8uMQXwZuz8xOgvBEjo4bHcs310R1oLCWH7u6RXf/nzH7x13PERxHo/I055n+wy1yfHlErXMuSuXpOTRJif6OcO8xFWKrtvxyG/Ypq8xVwVdBq59SXKaqppR6Z6qWtBuW6IB4UHpbEK664L/9D/+m3z1/fdZbC7RqiCFEZEkyTtMVhFCRMSRJARaCbRQvPXkXVKIxPd+kre+9rNsP/4Dzp9krNdX/HvfuMAO11zmNWQL6vg2l7Mzrl98yuvrnvN31uSmo1IGkubFfkCagjbB4XDC64aZMJyOR5CJv/CNwOE0h11iuzuiVE1ykf31lojlfHOJdYnOKqQumS1rUqZZB7h9eUaeEtpIhuMt7ZVAhEhlHVrlSNHx3soyhkgYIiY4ZNvx4NwCGX3ckGWS2PeI5Zz5IlLUhj+XNG3v+fb3Gk56OhKIEDnGgMMCkaLUeARaQiUkJ/o/df19IbIIhRDILCOFAmE0SjlK7cljQovJAacJyDeNmV0/EqViXhkQkSgm3jtSkcmMsqhZzAXr0pOpCWuldEY5OyPPDMRJdBFSRMgJ8TxGjyZg7YC1kRQVQkYerXOWKdDsbzmNLXFU2DERBJQqEFVicILbQ8PnL1+zOzSc+gE3HGjaI3bsES5MKCsvGKPHj5bOWw77gd7DosjZLJa0XtPHjOQdKXoqJRhFJOUSbQRGTaQgoRU+Tc2kMSSkhEBilnnO6o65GhA+olROVs2JJHwICHfAhJHoA95FhBNIAfOl5GI94y54mqGlSJZAYgwDpy6QB4fvRz5pO573jusxcIgjDQ0jHUM+sPVHQpkhUKRMQ0zUeYEUJaZWKDVNJoIASaQuq2ls9uAhX/3SV6hmc4gO+aZJ9if3RRKClMKUyZCmmXkKkeAsQiSUgPXqjHtPfwonFLv2hCrnaC0nK/DoCB5E0gzdkTjmeBfog0dqgIiKA9ZaZEooJoTcyTpCUGgRwSiGIBnHEZ8iLo40XUvne6z1CJWoZ3oyf2USIRO1AoTAFQnMdI+e/Eg2Snxrsc6RbKIfQCgJKDobkFoxOk2wgqQM3hpEEISk8U6SG40ScDYreXRZoesMLeWEr4+RJOIb6bEiao1MQBTTxO3HXF+InYAQksyU+ORwLuBGRzreEIYSl2ZIFWHYYYPg1W7G8Sby5Ueax5uIS4mkJL6LjM6zPRxYzCoeryNV2ZGSQIhAUef8+atLxrQj0yPDuMUOax6eqWl8JyPBdmhxx66/ROU1Wo3cW9UQOj7c3/Fxu2YdNbILSDSj95xuT9wdA8+ebdm+6vAE3nnvPsej5ea050IaZGnYGDXZYAfL/jCQdYmtE5wZRVnmnOUSL2fMyxHXg9s55rXgO68+46xacD97gliseGkHUgSRLC5AjIFmlCjtmVe3eDpslpgVDhlazpbn1EXJJx9/yL3c0dtPWApLko/xvqLvX1IsVzyxjt++abnRiTMruGstuXFkFriZ3Jz5XDMvNKtFZF5uEdLR+hVIzacvEkNSLEvF66FntZyDnKMyxb15yaGNaBXIksBLQ7CBr3/5S/yHv/nXmc1mqJQgRYSY+uUyNyQ7YcaklBPKO6Vp9JoCkskUk1KilBJzcZ9v/Owv8/GHv0trO+bFA5SIfPjDH+BHT4iRQ9txdvkUrSq8dywWF4g4YptrBrNBiBU6K1kYhbCegx+JaH708Zbru4Znz+Hxek7ygflMsMkyopHMdQvFGU2nKApPpKcZLId2ZKnm9K57gx/PeXV7x8sXPVrBB3NBkSuC9JjS8PV3PEcKfnAYeZJgdVlyc2wAyUwKfNEjCcgUyXXi6VsrvvS64cOXGXd3A1kAlRUoEQh5DikQlEBrg8L/2PX3hSgCSkLfHLFDRy4ip9cvJhlw77h8q+SsaLE7y+EIZhx4f514tBJkxtE2EikFh77n9d2AiCWjz/j05kCRdVysPE4tGFzBcq0RMdEMiRCOeCy1buhdQsmMbN5AFLzufo4qzxGxw/uO3MCXZzW/9WxHtqn4elnSvez44cfX/PEnOx6cz3AmoLSnnGcsvnSP/R+/IGsC5xcXLNb3yAt4Lvf4LGe26Hn/6TlORD4MjrfvWx7MDad9T1Vqbu40z/YFi2PPOMDH8Y6byvK1xRX5TmKqnLKWDM4yq2e0oyUh+fRZxosB6rTj/FyxO2bc/8qcrJjx3gc1lQn46+8wL0Zev/wu7cHz2TFQlVvak+dxUXPdd/zjUfLNR+esF4HtXcsfv9iTZzVffzfjcpOoc8vxNGIwbGZ7jDH0TjLE90mnI1V3RswDzw8dq0VJ9vgx95cDx2NL6TzWev6Lv/N3ePre24gU0VqBFIj4JtlJG4SPCD2l8ighcCFNen+ZQBtSnLIQRBgRTClV2fkl9fqvcnvziuRH0uh49MCyWDeMQvP4/AJdn6G14qe//oQHs5qPvv9dTumKVbXifH1OMKB8pOtaVB4Zxh1Fu+SdSlO/03DTPWMYDLvrknFzTqoSr3aCi/pAAm63noDHHltkKDAzz+AK8sJSiDU/fBXR2QX77sRXpGZWCj7reobesDmric0cn8747NnHPL85IqTHO0GeayqdMVeWWe4YUKTFBT/73oazeeAffvuG9tSTx0BZ5Cgs7RA4yzR1XfGHN7sfu/6+EEUgJkFvPTFFou0Yu4B3CakS0ac39PSMJCNSQ1mAzhqUVgxO01vLdpc4NhaZVfTOM173VJlnvfQUKtFhiHYgRU8fBUpkSG9x/kgIBUl6RtuQy3uYbDY5xOIR7y3BGdrBMxwsok7cuZyYEqdupG8HwkVFYTRNKalqhRUKKSIbrdBFzWyek2UaQYsWkVhkiALyBLoXnM8NVRkZugVKGFShJlhonlHZHUNIjG5A+AbhG8gMSUh0nCTXMk3xTtbnuN5ySIJzFzlGwZVLk+EJiR093iUIju72lv24Q0VJ0ygGH1lVmiZoTiGiaoMKnlmuqHPLphjJTT5t61Mkk4myAuEEmfIE17FYbfj0cACdoYOh9GCiAi/IteHs/jmnQ8sDU/LWe+9OcW3JoUU1RWvJQAoKqaYoOCUyXLKAmY4JUiFiIEoNAlIKk6xWTbqQlALKRRbVnN1tg3Mti+XlJHDKSrrdjuQCNo1U2ST1TbLi6mKFqmpklkMY8X6kcS9JbLlXKIqQ8TJJ7vYB7ysWMSOb1wQEmRcIFaiqPbiCsZNoregKTeEt8wxaI9EuIwjLYraaRDxeMIQOGyzzUtI5aLuCKB06K7EhI7UHMIIqB+cSHYJCCNoxUJpEkgKVK8pqul8AAgIvFJnRnLxDm5KgBbgv+IgwonBD5Ngrml3g1K25uoxcFJHdseMzZXh6/5KSimXxmkzsSbomM4piFxg7gY0FQ1xxUZ3xeDajE2+RUs9sceLhZcF6kAwpZ1lUyNaCz0gcsY3ltrkg5Z7gL9iU9yjzawqekZmX3ImM0ULbgPy8o30VeP0zOabQyKsV6kcnBgQX6zVP1ivEWrHvPSoVXN2/ICtyYpDIN5oHqwRZXaGzAYRnmQrcOOPGLpDynFYkgnEsFw0+Od5fnfGtz29pB8gujzw6l1y7PYWvJkpRMlT5iPKKRZ0hZw949/wr+PQ7XImW4/4jbvYRj+IvffVd/M7y8nRgmTKkCthO4lMiiQ0P64FlvuKHbuQ7r+64KHPKJvHVh4oyc1RpZNiOlPM3W0yXYWpBkol51rFaVjzPBSsG9keDlwVZUaOkRKeOIKFel/zN/+g/gWARMaB0TooOKTMQeoJ3oglxwKURFSRBORTqjXPxRFbUSD1hy6VWKDX5SrwPFEWBlBqVVzjbTZOW45axu+Wdr/0i3//R76LVmnk60h7uqNdzPrn7IcVuJEZDXVekJCilYNct+EHnyaqSdX7OT33wDbbbW5xzSK0YhoEQJWdzjTZHzkTHmN2nD44qL9g2J1prqcsronaUpSIeR4TPePIo4+U2IaNjc2V4Z6b46PmcLz0+0gyarXif5y8Va3PiwVnGpzctu9MRqQaEyrlvJFI3tEEyKwIP1guGvsc6R5EK6vWcfdtPUNYg0D8eNvzFKAIhTp/0UZQcY8H5w5zziwpBx8iOWV0zX99D6Uhd3OE6Q/CCAIxC8s6TEpXPeeJnBM6plzPQA9J1LEuPShbFjmVp8KLAtidOTUedCV4FhZdrlEtkeskQM3L/OVo/IwTLsdGcRkWQGbt9w9GPxGcljz+4YPNgjnt0zlLUqJlirhOh0BxOAZ1VrJZLEIIYAjYovAxkOoMMuhGGkNN08JFLlBlYsWVZlFR1jdSWve1Y+5y5WPIqnmiHgSw37G9fsTYXxOhJcUTKjCBG8lyxuXjEg8s1Hz1T1Cbw+pM/oo1XfPXLX+PQXPPRzXMO9iWXywf4vkSJiuXV+8gkudl+n0wM1FbyapgCRi+puEwdWQ13R8Es08weSu5uJdsYmYfEcm5ISiC04tG9cz7/4RBS5wAAIABJREFU/IdcM0daj4iO6BqMHfj5L3/Ar/2VX0ULjbeOLMsQOiExIAVSaWRyON9P0wMXcCqioyQlj3ORFCcEnc41SkhStCQhscFN4aAhYLREmAKtNTFGsiLHqLd49uIHPLr/NZ49+zZ/9Pz/4nXbEIPkG4tzzPl9nD7jFHOsd/huYNu+wLc9yXluC4/1I8v5imq14bS7ZbmcT4ix6NhdR26OPeVs4Gw2Z9tEZrMZTXfCMZJioN96Li8sMVouFoKXryV9mqFFS9IZbz0MCJ3wbkTnsDi7YiE1mTrwwds1rhesBoGfLXm2VVSnBm2gqCUXmyXX2y1tP3IcR9ZpjtCTSSnPC9arkrvt6U9df1+IIhATeFESjWH0BXk5JzMF1ufUdaTQGVFkFBqkiiSRGHvwInKyGnSgykHmghFJXQd8UmgVyUuQJlDJHuciLli6MDArErmB0ygpS4NIk06d4LGxY1SRGCXBRpI1hNHjrcMPAzfblrNmQzkzrO5vMH2BEg2VNpyiIgSY6wqjc5KGGCMuOkJShDde9lOnacZJLNJ7SxBHyqhxSiJCTlSCtgtkSlEoQ4lkO4ysk2BwLSJN+cECh7WKFMR0hCgrkpaMfgoMkTHw7pMHXC4M3/v0e2yHHZkOODJGCo6M1DIgdElEM/SCKDVOjmTSMARIImCMoR0LVjM3mbOCJxeScfCQTaCLEBzL+YoX4sQw5FRM8WBdd6LOC37j1/99VOxJIaLV9P8Wgens/4blkEIixsl/kVJCCkMKHkFApInOFEiIFEAmUkgoKREpghCEGFB6YjKaPEOENyGzCurFnDgMuPqcu6ZFRMeFyDH1Y7pYMo5g5XQ8BUhMfgfrpgLThESet+RVSVaVIDNibBk7i3SC3mnMEAl5Yj6rSCGiCknXB4xKDHJkXSlOQ6TKArmJ+OQJIpKkwGQdbYAuWZw8UZpsCjaJJ5ZLRaMELkiUFOQqA3tCFposBuoiIUyGiQPeO3yKSJmhM0NQ/wILIV+QIpBCpLOTSnBebaj1HXbwNL5kdyopTOQ8j/Sh5VxE7l4FPn4h8XpNXt3n1H9ONA1GjCwKSxKK/d3I1dpRlQWl0Uhb8vsvT5xlnrIQPFzMybTg5rQiiCVCBNzwjCE2bO8aPop7HtQrztae/XHk5esDXk3Mt+puJH1ywj1ecPbukvQs4bs13fwOKQou9JzN4wVReHRIjMHhnUOb6UznY2TfCdroQCSa8RUbk2PWD+idR7YduUnQF8gLy6NSY8YZN6eOOZJkDoik8TFxuztR1EsWZY3t75gtSsaxZbedcwo9v/jzv4FNLdft5yxUw69+fU2ma+72CTWPrCPsji+JcsXty4EYNMlC+bCgPTZUVaReG6oy8d7TBfO6wcVIFhMhBaITfHTTYHzAhuc8fe8d5ssa/4llMIKYNL/887/I+uw+9ayEkEOyEATB9zjbUS5LgrfgLUomVHqTxR4DhCki3dsRqQVCFATXoFU+LfxgiXZSJcLU4E0EtJrSoaIMGA0xKVbzczq15e2zh9ysv8aXVmv++as9u67i0LcopSbYa1K0IfHo/JLnakvTn5hpzWazIpiMVy/vaFPDg80GpQSH7Q6jCxZmxfZoaYdbHj5c8fTp2/g4MnQ34A8knXP7+hrlYCtbrjaS5EGQTYAZCVVKLC4ih+NrXtk5bW/J65yzzDAzASsS2XhiSc/ng2OuA6uLisWx5OH5gtftQJ0S0TuePr5PgeKmP7E/jT92/X0xigCRwTWUeoJgJHeHtQeybImLmkoa5HCLNjsyPN5JiuIJsryinlfI7AbfO1xylKaZEnO6yFgyWXfjgiE9ZVH+IdYFVFaCTkidkYklxzQyzzIK1XC8bSBkjPuSZ8c9/eh5/gKub3egBNUiZ7E4R4uC5SHilw6zKTDesneJTBvqomRZlkDk1LXMiorMNHwqA3kwWClxYpwcfwK6EHCxwx86ysJgYyTLC87PDKtaIOKSwuU04oYQJTMVuD7ekHAoWZLPVwgtJutulnMYD3zaWL75jW/SsWcMPdFnPFi8QOqBISSIJ1LX48OAODhsCGy0ptOJ9WyGSo7XOucQHbM8wwWH0S2eQLv3dMeRdpjm4U2syGzL/viKq/4Rjx78LPPvf4eQItW8op6dc352RXSeGCxlWWFTTwwOaeYIJRFhMoBprUlM1GPnHDIpREzozPw/40KtZ0Rvp+cYICKFIYZJYq4wxOARQiKEQJkKETwahcgKsvU5Tx/9FNJfc9r3uFoRYsIgGboe6x2nU8v6bM46Nyz0ktVqRV6VKC/ZVpZNmNMNkctZzr3zM3ZdR0qRqpRkwvH5Z3vyL80xOufhVQUisTvO+DQG6jxSmIHwhhiUsoJ+Z1msNbcHy7xKFFpwth44qIg0Gfseci2oSoHtJdt2nCLUyoyyUrSimIAy9ZHUDrTjyGx5ideJstWTWOzHXF+IIiCFIFcCLbckcZ84JoT2lEIyq9fkZUKkW3LVYIMnjBGZG/JyRp47qkoSveTQaEQKU0AoZsqCEo4kV0g5bZfbJFEBQog4BixT99ylnjCO2DiJgfrjSCcGRht4eX3icGhRRlPmBVJP2nc/JnwfIPfktcE5yInkSpK0nLavSlFmJRGHGKDWkhQDXsEQ4kSazaDwGX0X0Dj0LMOIRK4lvfPkwpBCQS8ko1e40ZDyEakdKitwwTJ6jRfT7w7JcvXwA3IyutBAUHTNCNXA4CwnW9KNA8fG4fuAu23ZqsgsW7M2EiFGZgLuSoPbj9gg0DGSZZZxjAw2sjsmnAvoLCFygSwyhOgZ7ZZ5fcV6vSYKS1nMOV9vEHGK3tZK4nyHlIosLxHCEJwHBLzxTSTBZD0WAiUN6U2YlhR6csMpOZ0hRSAhUVIT44CUmoR+E/s1WW2VkPwJXm/6/hTTtqhKkp3TjpN5SypDLwXRe8beowQM3YAPnqv1GebNzwijOM82lNpzHOw0/dGaoDKOXYOeeCoMwxFhA0klZosRKQInm5gtL5gzUNfX7I+BGGEpJaPSyAAparwDnQXyGDBG4lNk7B0mUyiVk3KDlx2bImCwpKTJlEZgMHmGdREXPN71lClDioT8s2oMCiE+AU5MtAKfUvoZIcQG+HvAW8AnwG+mlH78kBKQUvDkLOJcw017oBUKQkIuahaFodSwLl7hfGS7U+hcUwqNyi3VXFHVirqSzJuW6DOcEDy4l/AxEVNGTNcYaTmrK06j4+AdG28YrOfFqz/m832BcJaXH3/C9hjw1tE3B05jT1kpooa1rLl3fsnyYslSCYbo2e86sqSJjwbEUjG8hhAHzpVgDB4pYT6f0wBD01OYBH2HLkqCkUQLPtPoqiRTBUs/RymFkgO3bccqG3m9E9SznCxpmmGkrSZj1Ti85NSeuDj/afZuZCdveP/emlID4yuuVis+b18ynI40R8ncPODFXYXJHTIVjNsjrz9vePZyQM7uI03B2XuPsOMOv9tTycTDHF4slnzvGu7XkU1o6btISiWby5JwHHFtT7lKFDIjkyPb7Q8oy7f5uV/4JlVmUDJHKQHJYUwBDJBylJIIAWMApSTt9oZisYSkpmZiHNF6cujJNC3ggEPHBFFCaIg+kcoZqb9FmClUVqipB0OMSG2IKUCS085CaITMSGGkLDRy8Zhf/IWMH/3oEw5jw/XtDSYlqiJD6RzXNShlyOo5MUTWqxnJeYRzFPlAEh29U/ghkpcF59nAqjzxvc9glq15fXtHWeYcqi3rc8mqgrPzJxTBoUzC9SN9FxndDUOEw+sjbYK6gkUFLw+e9pRogaMSLGtQT3+W5fqnef9r8PwP/z55+owgJBeLE7tFwRDO2XNivN3y+bMX5JVBhACFgeOfLhv+/2Mn8Esppdv/1/O/DfxWSuk/F0L87TfP/9a/sAiIiIwN41BQSo+aX9AMAhlmZOORsYzoEqyNjH1B21kOwwuKvGLhzxndhqp4hskisggQc0rtiUGiM08aPbK8w7bnyA6OwbMbM06953f/2WcM0bM9OexdixMCfIfSguVMUxY1y3pBNpuzKntmwpJUTTbuOasMPnq404TzQOElRgW8t/gYkMrgVeDYn2jCgFeRbeOhljgtEASiFUQc1piJb59a2nGkiCXHEFChJIuCyMhCawZzoG80z65f0l13kG7Jl4FkG/J37hHiQDO2xDAD6dk+e01YXLCZacZtRhg0iyKn33n2W4upr3j7/a/i/EA39Iyt5dhKvvmrf40Hl+dUpuLZx7/PyZ749Ee/h7OGMofzmSc7z4nac9u2xGJOfvbTbO59FT2/oNB6Uv4hpjGeyBDCAdnEfXAOkSJKZQQ/YvIpHCMmR0xT6q6zLVm9Zjgd0FqjVU1KjhAtiAykRdgOIQ1+bJFao3QOISK0mhSFwU5HJTQYQfA5CcVydU57aihV4BBGCpF4ul7y+m6LHzuCdFysz8lnc3SM2OAI7sDdyxe83jbcu9Ach54iveZwGikWM2zXkhY191bnpGiwscFz4FvfjszmJxQNpcppsorF7An1bKDIPc+3A2IciLbH+QJ9GRD5yIqOm7jmdpfohp75qubee09ZbJ5wXtXY0y1m+z8Sgeujo8oUyERVZlRG0409x2OP9YEiBg4/Zv39WRwHfgP41958/d8C/4h/SREgWY5t4MVBM4aRr20ucDGgzALbb9k3UKXIqa948SoRwoLV1YqZvublySDEhkX1A/RMQconXXySWCkZ9yeGUcNO87/93g/wLhIuBb///WfcPN9jpOL8Uc7Tx0vkeoOWjqbvqJcLxNxQHBN1NUcYRetHKuPp/DQCHMKk0z+c1lRihV15msHx4vqONDqEyCF5TK4ZcseRge4mstA1ywhdBBcCXkauj3fUC0/wFmEWzOuOECNZtcHKRBpPhH7GSY9o5VB1xv0HMxaLBYUe8aLhyaMvEyj44Cf/A4ZXP+S/+9b/xN3+ALtb7totv7xpudmO3F+s+e6P/ojZxWPqck1dGYIq+JVf+IsMPrC/u+W//nv/DYSOYqbY1QUPZkveS5c0hx3enPjEWmQQHBrHYAWLZca7/+q7VEVNhkSlMG3dYyLFiSCd8JP3PgkEGudHhB3ROsPZ/s0NaQh9QyqWkAR+bIgh4JFEOuToscFjokcVhmQ9qSiQCLSUJD8ihCE6T2LCi6cgQXmIkZQiQiiyLEMuZwiu2F3/Y2ISZLqknM9YFAvWmzNyk3EMI8+ef8a+O/Lh5zvORk9dKm5ulpAZyvlIsuCwbLuO48kRQk9enbGclzh6vM05HGsuH94nas3r5x9yc+NZ1Gs285zzMxgHzThc0OwP9AdHXUqyeUZ+usfV6ik/+vCP6G3JfPUuVgjmpuLp+3+Jb/3Oc4bbj7DDyGgj67zg1lmmGpjIpGQUDvdnyBhMwP8uhEjAf5VS+rvAVUrpJUBK6aUQ4vJPe6EQ4m8AfwNgvSzpQo5D0HYn7u5uSdkGGRwh1jT9C15uBTpJwjCiNnM2VY2NIzYFPBc0/QVKT1vEu7sb7u4Ex31g3/Yc+jnWtXz84TWBxMasMDKwqQouz68QC4WZRTIiRZ54ulF8omHMgT7DtY48G1ktrkipZ7XM8V1DsqCyxEU+4rWlFoKjGIgM+FEzDh1BGQpdkNKAcIIyWLLjidNa0XUjulJErbAhcewDfucRs0CkJyXH/cWGEDxNM6KsxmhFqAaWM8OsylibHVYLkrRvutvn5Pmc33v2A/yhR2aB1kd2u9e8yA3SVRgMnSgphaHO57z/6IohRv7h7/wOu+uXbK9fgd2hoqY5aNTc8sPTS4QtSIcBlRouhWPEcDt4Yl5R37tPWSwodY5UAXSNiBbUtJVHMDUG42R0ESqSZQuCbYkRdJnjh5aRGns6UWU1MUlwU2zY1AcIE3//DSDE2hGlBCkEUpji3ogWXS+IKUKatv/Ejhg1QmhkssQwgj2Sh4DMDO89fo+2b8iqmrPliqRzSmP59oefMJwG9qdnlLmioaVazal9omkONEPBwWXUvqNzxYT4tv0kG95FjJthRSTLamQ1J6/mLOczjqc7xv33+PTzhvhow7zfQzIUWWB/2mMHx6ZcE6scmc95cr7Gnq5YPThnc3FJqTMKk3FKR3ZdRtvO0IxcjxOP0IRElSlucXghIUaE/LPLHfiFlNKLNwv9Hwgh/vj/6wvfFIy/C/D4wToN3pALwWASvT1gyWj6OSIecaOlyQS1TCRliN5x8gJ8jqkkPhqsW6JC4HC44/PnDZ+9EOxah+stjRuIw0A7RGzsmHNGscjIPSxnc8ZsJJOSPjZkEspMIH0gBE1mPT4IRi+BjCy3SOGxXoEDsZpmyYUJPLMOx4DSFcGOUz6gmgwfLnq0CrRe0QFZVMjgEVZjVGJMATw4BKkd0YUmBU9Ydgjfo1RCmBxCojSaYxpwqiSmKXV2uVlRmAxMSUTw+vUriB5DhhCTVXVvHat8QUITiznS1CBGnO/45OWONDp0DMgisVQZY5cIXuDJyYLFITj1I8l66kzShAlxtZgbkuvIywIhJVrlJCQ+CiJpGtcFR7IOVdSkcUCZafsvJ2QAMEXUq0yhtJkafzIgxCSHTdG+4QswUXhDgwgQlEH5REwexoiqiikXQGUIlXC2QaCww3SkSEKiUoZTGSEMIGG1WWPGEiNAmxxrPYPYczxtOe16jPZEqfHWc2g8Y/Q0nSXkhsiUjkyyJKmwKZIXgjH17HtIUTJfZpiUaNueYrYm1wZRCcaXjpvdnkYnfBp4vIBlkWOkJAVL3+dIXeKykryecT67ePM3QJCRwXpieyAlwWnwhL4nMxlSZ4g8RwuJ9WHyXIg/I6hISunFm8drIcTfB34OeC2EuP9mF3AfuP6XvY+NkugiRe0ock8YLLF/xUe3O5wbqE3k6ZXDeihMxvXuOV1nuTzbsFrBzfV3+Wc/es6xO/DxH39C73dcXVZARttkUyCnKrh8EFmYBXurWNw/Z1FWuGBhGHjZ7lkNBaGsIQ+40eN0Trk5Rw4O4ScIxqHTXObXmP+buTeJtS09z/Oev1vd7k5/m+ouq4pNkUWxFSlTkSzFImM4cmwlQUIDChzERhAEyChDAwGSTJxM0oyCIKMISOTIdixHDS0nNqVIoiWxZ7GKrO7W7e897W5W9/cZrEtnYspIY6DWZJ9zBhs4e6//W///fe/7vEkwoDlcwOATcejo724ormdG2RGrBVVS7B8fIU3gyo2MAYTtUaNCXRekdiIH/bWPv8Lf+d73UTTM9IJoPZU1DKbinbt3WS722WsKCtlgXeL67GX+47/6y3z/3g/42lf/Fh967gNTLn2zQgp47/7bHB5UnHaao/lNFuNjTseez7/0l8h+x7h5wkc+9AlEjjSlZm91zLF1oK5z132LZXWAGrb4arq5rFdspeZSdoyVZPNk5LG31Id7vLRn+Nitl/ncl75MkTWEiMsjqEBGI6Ui+57oe0JMMHToqiF2AymuqeuakCxGacysRquM3Dsk5SleK+rJ5YkdMGVBdoFgLdQGJRNhTCAHwhiQSjE6T7OaI3Im2ozIGdtegKmmwJX6gCQi0XmydSgpWdU1WxcZ7TnnF+9w/8kFIlnCKLBdZHE9sx7WxDZwL29YX/XooqBcFcxrRRsslXCUEsSswROpF3Dx6ILzywBPLkjpHWam4uWPfQw39FxeRrJ3XJ0n8vE+fRtJrmW+zCxMzdit8XtHfOYz/wrNfI9rx7d4561v8X/+0de4f+c+166vmJmS7C+YLeac3/cIH+nsiJSCSisWlWHrJrdlCP8CvANCiBkgc867pz9/CfjPgL8P/FXgbz59/fV/7pvlKUyx9oGddBRCMTMGKQ1idczRynG4b7lsA5UpoYQRiciO8/OW9cOWuw9O2Q4jsao5aQr29hNHy5r7Tyqy0Awh8pHnjsHtSKEmisTOOfrLLXZYM9sbKJoFlTI0zZzu4hxVSsqiwKhEFxVV0eDdlooNrlwiKOgGh6Zg0/d4DPgRsV/T+GIyxkiNj54hgQgCgUKqGmsMaMe/8/nP8d+99nUYem7OV4hSklziyo7M9JIkj9htYegCN05K5ktJyo5eNnz8xU9iH/2Qfjylc2uqasa63XH7yRlP7IBQjmeOr/FSTizLFW9fPORE7RB+YL48xPuCwgjOru7Qbzu++Iu/hBdn2It3ON0ltEkcFYL1MCILwcOo2FtJbr78POHqCXvzzMnBjE//ub9CWVQk4YhJorIgR4kyEzs/pUASkmF9TrN3jTh0iJhRSpKcR+Q4JeVQ4UNAxEAS4IaRai6QBHyO5LbDDWtk8hT18wTrEWkgkMgUpEIxn8+JOSOQhGBJaRppxvYKJTNlmQGPMgXRjfS7c16+fsTXv/d9gjzneBZ44UBzb2tpt45hcFwNBrKgaRpyCuSwJOSC0Dl6sWGv0gQ/EmRJ17aw6WnrxHYoUSJSioAwNc51vPfWG0ipqYpMEoaYgFRTmUDfZba+Q+1H9vcbXnq2YXWwIOuKw8MjXvvuwJPbb9HbLWFXsD/bx+ydUEvB+egoI4yDR1SKgCAag0weExSD+PFW4v8vUJFrwO8LIb4D/DHwmznnrzxd/F8UQrwFfPHp7//cK6RIHwKd1URRUJQaoQU3n32GerFHyBOYomoyhdZUcmRjt9y/2PHocodqNIezGYu6QauCrCuqsmG1rFgsl+wv5tRNwawuUWSsn1R8oxvZLwNLOWGi6gyIgMhpqujSUcpMIXkqZ030QQIl9eKENtT0MaJEoNAFbQIhPbmUNFIS/ISpitFPtBylsMEjhCC6yNXVBTJFchKEFAkJut0GksNFy14zZ1kVkDXKGJqyIObAt+59m85uSHGk254i1BKyZHSBbhjwLlNkiZaeQmn26yWvn95GMDEYVE4Y4zHqim775KlLN7FcHjFGj5bT562IGDxVllid0FWJmks+dFCyWlTEENDKEL0nhQjREUKYZL1pkvqmNHEBlC4Yxw1ZKIwpiTEyeocu5k9HgI7sHcmPQCKG6cZ1Q4tSmugDIvO02RjJfpzksUzBHTpnXN8R+5ac7FNnnaRdX1EUBTF5UugQWaPziNSKsiwRGV66fkQKjk0vcP0E5sg4koCYp++HnBl8QcSgxRQs0reeXevxY6QdA4wjGx+xW0cYJgFUFmJySEpBDIHM1KhzMSBTxtsBmwLOWgbbc7lxtDYiVY1CTVi1PH2GVVkwqytE1hNP0GgCUCrNmBXSTI5N+dRxmAXTceBPWer/r3cCOed3gU/8M/5+Afy5/2fvlQi54Gwn6KInLRcsFs+hU+b5Z29xvl3y3p3X2faW68eSIfScnfU82XgGH7h6vOW5Wy8wOyoo+i3CBuxQU1wrufX8gjE0iCw5nDkuzhwhOwZr8Wnk5FrDXo5kU9CGQDnviTkyZo+Sma0LHNeWm3Xm4VajhWTwS3zzHPP5PqcXCtVcUNUj2TjSaFEHIETCjpFybkA4ooqoDbgxQjPwn3zyp/jD5SN+84e3KWeCQUH2Hje2XOw2vHB8ghGC5XKJMJpVshytNIcLw5haXnSeuz/4GrPZCX/797/DX/vlfw8XHT+4/SZ9d8X2dKAxS95++BYfWGjMdotRPUt9xJ/YFpFqllVgNR94fL9nKwK/+bv/BweF493HglVuqOKO0WUKJA8fd8yuVZyqgcNSIGwgR81grmE3j9H1FNQahCYFRzHbR+sChEd4iGNGVzVKGVy/AQGyqIndmlDUSCJZKFx3Rc4KHSYaVLQOu9uQhh0p56lgSMvu9AFSBbJrycvr6GoOORNsItgtddWglUDKxPLoAN93tJdnMFwyv/kyoetRxmCqBe7sbT7/yY/y9fuP+OY3v4eXEd96jmYGHyAlRb8TrFtHCiW6Lim0xHcZH0eGtqBIkRhGylJhUQgJSvYo0TB2O5IqWVRzpNYsCklWPQwl0iQuzi5ZNiWXuxYhLHfDyJNHku/ee8gv/Kt32VsdsW03mKZinvah7ZBJI2Jis92SnWPeFKic8Lmkbg5w4gKnoELgVJr8Hj/mel/gxWLMKJcpBVTRQFMwqxa8cP0mXYZSKZR6js1FxfffWPPGmzveee+Ss6sdZTOppdrtBbW+ZF47UnNCU5SM0ZKyRWjJvNKctZ6tDdR1iReJZlkyeri/TpzuLHUe8SnTEbA2oy3oMNIPjl2cI8bIegjTsGvsudpafJBstiPtOtIGy4hgtJKiSeztTaOrXTsSXGDtBnb0DMnye69/j+89eY0kAhGPVQkZM+VCMF8YXPIkKWjtjmWhubZquHnScXx4yY2jlrff+g4Ow6y5xs999uewux390HHnzm3uP37AwiwIo6ZvT5GmYu0iz5YFhMhpuyOZjGROUCdcxYIwKg5nFUW9x8vPvkwoDuhiRRsM2yFgxwm6WQUYleDJZmraPffiszR1CcLgugG/vUQbQ44jcdzihhY/OHKcFknOGVWUUM4Yu55UzOh359i2x27Ose2OHAO62Sf7EZ88qqywdkSmhMoBU62mncPokOUMrQuUSISnpGPdzElhJMeA95728gJ0RVlWSLMkjgPCRdzVE0BSHdxC9iMfO3kWJSRVBytdYa2k1CV1eUxkRTdExjDgu5aULdFbSiWphWCwluwTWcCymqO15pl55GSVmS+PKYVil0b2io4qPmLhL5G1QFDhUmR0llmdEUojRMnQJd69/ZBv//Hv0W4f0m+vqKKA5CnTtEszs5psI9kOODtJApOzXOyuIETqlKZdwFNi04+73hey4RAj79y+TVCJulkgB3jc76h2LfrygsJo7jzYcLX1nD06JWvNXjlHFwJdSxCeaCu6XlPmGTMn2YVACp55GRjClA5zdbUhJgNGQ3BYGVnlkXJ/RtsG1tTUQjOEET96XMq0lFxdJWqr2V08oSpL2rmnMpEYWz77zB5fe3ePMdUQ7lCgSd2I27O0ybM53XA6XpLnDhGhMoJWBHbSo8yA5ACRJAhLGCPN3oxrN5Y8vPc3KxMvAAAgAElEQVSYbf8EO+wzL2uemR3xqE9sGSELlqlDFpJf/+YP+NKnP8XRTHN29oSmabhmSo7qBU/ufpdT4Si85NQNHJs9+sGS8chFyVnnOVR73LrxKturJxwuFlyNgqNrhiw6erOP7j0PvvF19ppIIxouhGPnPNVzJ8ig+eBLn2TcbpGyRVQzUDVjtwUkY3dJGseJBCTAZEWSER8sdjgHm6hXK/rtmrqaUnar+Yx+u8EnMOUMQ8aOW4Ibcd0ls+NbiKtHmKJAmIgSUFQlWZYUpUDPriNVSbA9WSQapWhnh4ToELokOo/Wmq0fKVRBd3EX0yyRzYKf/cxH+Ppbb9BfdiSVmRUV6Mys2md2lLgIke3YU+oakzUqC+zgUMmjlCRLiXeQYktykfs2UZrA/rJCH87onSV294muZXAe2yyY1UsOmwU2OWoUuQiE7JHCwPqCb/3h73L77h1uHB1gpJqQ8EESUsdcHKBN5OHjDT4MkDLGaHa7HUozSbWEQKqCmh+vG35fFAElBDs/YoQgK8mIZHt1QTAVhaxwfUfbtrhhRImINgVFUaBrRUg9Ao0I0LUSM9dYmelHBxaiyhjv0EJgihl59EQPmUhOkf2ZQWiDFAahRrJMdGPCP62cRS5oxw7FQFMWNPOGGCXeRYL2KKOY1RojNG+7TAKKGAjJs8sdPoyE5KcpjVEQBUZl+hJcJ8ki4yWYLNBaIkVBNdMopZBpiqIenGU3WAbnCcFS4FmZQN9d0lQNty8uub56hrBdowvDTBUUuiRITSMm5V4AGmUYvZtm0iozRDdBLmVGlxqkJ7mMlIIsDRnJfL4gmpJoHXvWgzRoGel0ZlFUKFlOoNNujaxWKJERWTJcPmEcdqAVwjlsiCxnDSJppDKUxYxhPKXvQEvw3qGEwBST/NcUGmU00Xmi93g7YoxAxABqgq1qURJDxsfJqiuYdAHSVCihiGNHjiOX773GbO86sb1gdfIswSeqesXFvdeZr5YQQaiJuCSlZNYUxATzuiHgSUz3ZxZQqOKpdcFh/fQ9a6mJObNMBh+ZILA5Y4NBSehsoCgjSkS6URBdwnqJNIFegjETDnw3OsahIxSGSiTmRUEKnvbJXTZAU0vK5QxEQZYZZKZQFWiFlgpvByIZnRMiSkIUkCZ4avhT9vzviyKgjeTZTxzRNA2zes6mv2CzvmDrFPOdIfmB+6fvkceI1JIQIjORUAF6WaFFZG91wmgdHFW47RUya7ZjZhkyMxXIbY9WMzyQkificFlD2VDUgpOZIEuFlJ71/UuQk95cZokREh8tz9y4TkyWx5cC5y0KxbhoeeH6CdcOKv7wd18nikRdGEK0aHHJ+rLDzjN1EoiocMJTFBWPdppFeUBfTrsUJRWLYg6pQAsoTEVVSGJ0XF1tsF6yG3cosaOuDPMbDXm+5Wh+nYv1BQ+2Bzw6v4tKBc47dt6TU8WN1RLnAjZ5VFGy2zgqu2Il4XV/xat5xaYNhG6LcHDVt1SqoSr3ycwYVM/ihWd48OYd0qCRGAo94JaST33s59BSQbCMIWK6K9L8kEplssyo+R5NvcLZHSZHlFQEb5E5Yfsdpl5g20vq1RHZtQgmbUR98gJ+HNAkbLSkOBF5BRmBR0iJKTXJdchijqobpKrIShF8gkIgqhmmqmHoELMVNoyoeo4Ml6h8HUGgbPZJfsDllloahNAUQsGsoZGCVNTTYu4HdPTsFKjCTI1lGyYoqsiIUqM99DlAsChZocSUaNXlgMtnNDFjdGQ9CGqxoMuZOiqC3eBcDcIzDpa+3VCaCrk6QFbTw8GPlmF3Cnk2TYdUhUpz8APGFBhVk7VD07MdHDkFnNKUQhCRBCIqvs/DR8qF4aOfvIbvJFdXFpkf4XJHMjNifx0RBMeL62zFjhRH2r7jNAdqU1Jf2+OVz/4UtTH0Vxckcc5Hb3hSUbGYO4z23D7dYK8igUwMgWbP4FtQdMwWFf3QgUgsG0M/JK4GRVYK71qehC2HiwWqrGgWe3jrWHcVpTFUM0OgpVEG33m0zCSbYVHQ28hSJaJSLEsYZKRPHT4LDuuKMgbudxUHlaXrp8nEk3VLmXvKmaYuDAnB8eEMKUrc0GPPNoxypJkZvnX6Dj9/6yU+89Kr/MNv/D6XFz9kISv+5K0fkqynMJpZKXl2teSNh/fJKjKbGzb9EaWx6GLBteWWMmW6q46d6xizZdj2WOnZ29+nSI7Q7qgWNR/61EeR9yJnw4blmPmLr/5rVFJjypI4WiQTBjxtzyiaAooD8tBytXlCP1i2jx/w4qufxOiSFCyLvWPq1QKRnyFFzzjOKZUGIVC1wO86EnOk67FjSx636NmC6HbkFCjTM+QkKRtDVhrne1QyQIbOoqsVFAWmWvHcM89wcXrOsDnFyT3S5j6i2McYxWgle7Ml7dVdzOomv/iTH+bvfe0NWhJVYRijpdQVhWzYnxfUZs6m37Hb7JjNK3bbnnmzgDgV2hAyMU7HBuEdWYIyBb0PaBeRqiIoRVWqCZaioFSSzieEziwWe0QSo+2JIbOsS5q9OUUzp55FCuGRMmPqEmcjKUXC2FGJyFbKp6nGEpLGEwjKkJNA/AsaEf7/dkkBtktsLyJ2k9j2nljsGBjI2Gk0WBuM0mgpqTGUGXKMeKWgKKCZMgvmlaCZBZpyajiOQyLLxJgCPgWUktSFIMaIy5kQR2S2eDtiO09IfnK8hYxOU86BVxIXPIN1hOgmSKkUZJGwMU4Nw80lK2UmWaucRkw2Z0zS8FT1JmSB1BNtRxpFTA6jE0YIRJ6Q6SSBkAnK6UgQRERKiVGCciaYNSWzQrJYGWbVjIO9PT5883n6oWU2q3lydka72WDtwK6zkCMuT6MqRYETgbkqcE5yUs7JOSFzQjgxNfZsh1ICmSIiBpTQ5KRwIpCUppGSFJksrUoQNqfY3SVFvUQqcHFE1yvGzSNyGKjKhjBsufGBD1HWC1AaXe8xPzpB6gpRlgglma+WxOin/x9NzoIsEyCJNhAFxKHFOwdISCOqWBKSJPlJW2B3a+Iw0j55QBrXpLGb8gnUj4hLBcYYdL1PCBZTNEg5MSei8yTg2rVrWDs1k93T0WCpDLNSURQVupoxm82pmoq6bKiKGmQmCk1RlBTaoJSiQD7FpykKrRFFRRblFFcmEmWp0HWJ0RVogVIGlSQ+OVTMEOOk+FNQzGoOFiuaao5IChJoVeBkxtqBrJ8+5XMmItB5cm2GLKf7UUCW7/OdgLOBqzct46hpigybGWt9j0r3tHrGnj6mTBEvL5GqYHY4Z6+qaP2IE4IUt+RwQNnMCL7DMqIinO8Mq2XNcenQ+5JoM9FotrGHMdLGHq7PKXctZxcBvzefwj5yBgJRSGLKpOhIwMXVBa4fkVJS1gUb0eFTyejOeNyteWV/xVnfUyfJkDI14IpA0whUlGjh8UYidMQYz6IuqNXIagbry4gLG7Q4IktNVTbY0NK3kt60XG9q9Gw24agQXDtY0F7d4e6Dd1hqS0BwbbXPMkVuP3hIM9ujKSHfOmGzvWK+KEmpZyNGnCx4dP+cn/jQTcS4oWNL3/fcf/Au223AScPlcE5TRMYUKKKh0yAOJHEYKI1Huh26WeLGDtt23H3thxy+/Dz7Nz7E5uwuOUmqg+uYokbWazbbltXxIVWzRzE/nLQYGrCOFDwpJcr5irHboKKmqhrs5oKkQJaZsBO4HKn2JSkqVHOCKA0iOpIQjG3L9t5r6OYANT9ESI0sFeV8D2lq3PgQU8zQs8OnkuNENBotAjlnZscfeoqXa/mln/44f/Dme3TBEQks9uckdiSKKUbOC0xdY13Eoxg2W6p6cie65BBKUDUFTVHjkMRoacqa/b09unFEq4DWFeVsTk6SsetpisRQNoSxZecDjAPIwEEjuXk8o14dISV42+FCJOPQ1rLdXhDGiBaBIQQ0GYWkCx5BwuSEe9o/+HHX+6IIjH1ic5FAD6hxx7y4hoobunCG5wFrMeegqyhlSbNQ6HKPg9WSecrcdXc5j4+5VVxn9C1PthEnIotG4qzizXuOowM/MeF3HUWV2ShJzBLvA5dbz4FXbLxg3cGzzfSkmy1LjmZz/C4yWstMaGwe2NkRKWE9XjCyIc9uEoGrzvL5F064v1vTJoXQ4GOifrak9xZfTIyBslIMKdC1gflMc++e59XP3KCYdQy2hLkiFpDPHdQV3eUFKksuwg5xWFLKRGEKZi88z5959Qv89le/wisf/xk+9srnOH18ny985CZv334bFktEWXP77oa373U8d0OybTc8PH+AzxDPMx9//hArAvc4R6eRO3fXIDUhXfJoATf8iiIYTC84zI539RXzQ0k5+wBS12TX40dPf9lx4+XnSarBi8Rs7wZSaMx8n8f336Euau68+Q1eevVTVMtDEBJZzEjWEiQk5xBFg0s9uqiIZPrNOTGNCK8oZ0t86wl2Q4wH5OzISiKNIUrN9vF9ukfvMmxbDm98lCznqNmc7BNXD+6xuvEcpqxpH73DVRhYXjsi+R0yTg3lcXtKcXBrQtLnwMvPXOObr99h/7jh07de4evvXBJTRRoE52FAxMhi/4CiWTHymLfvvQ1ZYuYNM6MoTUExX5BjRgNCl8yqkrqZkbVmVY34nFiall2oiZVBCbg2P2YYBtpuYHCWZlZytDejnmsebU+5MT9GF3NcGOi3A4PO7LYjm/WWqtQYB3rWQBRcPXrIbrcjC0mhoLXv8+OAyJBVJslEHzwNepL2UjKKQIhXOBJKKfZEhSeRY6JQmiAyUYKXEVFIbBRkDDJJKuUR2eGCQOuIYCSFkfLpljsgiDFNNlIDqi4wKpCkogyaQipkUSKyYhRh+jKXC3IW9HSMwhHcZJntCBTGcLPeQ8RAgSBJQTMvAEGdJVJAVApNnuLDBDitqIWmrA1FpdFS4DMMwVHkTIqR3lkGH0gGYspYEfm5j/4U/ehYzeZ0LqGUoR06jJaMKVDXNXpWsrGBHOHSWbK36OAQxqJEwMnMrFqQkWxwdONAH6aINCUTzvfYceRiu+ZiO+IDDAgimbjrCDaQvKNa1JTVj4AbETt2FOU+u/OHjGNLMBo938c0B2TEZIIJT7kAyT0d/20QUiOkRimDlOJpPN0EME0pIcqanEDKhuB6EIqcC0BDUeJ9oNtanN3RbjcTIDQNxDiNBW23xlQa/EAKkex6Ip7kJ/x5TJYUpki7GycrKiOI0aOERsViymOIeTqWFlP2RFVOR4y6VsykoJ6XzBYN148bxuAZh4GkEi67CVHuBoQPHDQZLRzEHi0FVVEhpUSliUZVloZCaaTUhJjR0mBtILjE6DOjG7Ftj7M9IkmyTyQzGbJs9Oy6ntZZButoB0cc3+cJRDdPjvjZL7zKctEwdBtO5gvGy2dIfsfNWyf8V7/xaySzokyGDseLJw3ed/R5j2vzPVobCbZHioCsKlaLE5Q/ZaU6SjPjapxCHF/YT2RleHuUYAR5kKx7MHpFpRyj1jweNJuU2bc/SsDVDNsebyVFaUBoXHasL7fszSqygOAirZdgDB99dp/vfq/HHCuUC8yXJUMS7C6vCBIKEZFJsl17innNzeOKP/zOIz790ZcZuCIFjcwJN3ZcjjsGO2Iqg50bcAnZ1PwHX/olfut/+p8Js4IPf/Al6thx++ED2rYlCMl+odkNVyzUIffGLWUsGHvF4yctbnOOG2r0qqC/eMjzH/5J3LYjrJY8fPsdqsWMONsn3h85vbhNJRrysqGoVxRVzblt+fKLP0WyPW7b49sNy+Ob+OgmKXS/xvYtqbukWl7nuC757rdfY3FQTaM90xCyZ7x8QHf1gNBe4XdnVMcv4QUs968RnSUmgVYFImdSN4mMXNdTHEus33D1JHGgSjwFY3vF+b17XDzo+Htf+W/4t/7SFyk//EnOH/wAU84o9noUmv2bH2B28hwmRvp2TexahFAEJ1if3sUoiS5Kxn7g5z79ab751tf49mtvsOMQS6IIkSASwSU2V5k7dx9QlZqf+NRPAjCOjlvXKublmhuLzO1lSegl3rWoBO9eOXweCd2as4uS/WUGNediC1r1hBC4OjulKApc1sQwslqtONw3yGLJm5eXrDfnpN5iQ0dnwWRLlAXrNrLb7ShKzRAy7dUOkTJZePKfEksO75Mi0NqOS3vBWXdJe/WE37lzn2ZRcbjSPHnrdX7hhY/wO3eupjgmveRkplgPGW1KSl+zs5cUMeOIpAjkGX6AmCp2OTNkxULu0TQJLRzCa0pZUGsYoiDpJZepQ3k/jfGSQJQWVGIueyIjXkScLZB6QVloooRa1gRRkuTImEfWseJmfQjDW8g4B2PoLMyqmq3ZkQuHQeCyxPRycgseSOR54vV373Ky2COOU2NRyEjfjewdzQg1hFngwwfP89e/+Od56+EbEBMya/q2ZXmjYNhueOXDP8l3v/lVdFEz9C2X3jNnauIFBw9OB958KJDykoMXblDIzFl3yfbxhudfPuR2VRPrir29BY/efI0CTStHrr38HFHC5eWWT9x4nhwjYdcidCangmym0FiVHcPuktJUaKEpVCKtDrj/7nf55MHnpibcsGXY3KfrOpwT9BYKC1f336Tce45yvo9KUw6J3zyhmB+QvUOWAt8mguspF3MSM7aP3iNJhet2bO4/5Fe/+nVe20p++7/8X/iV/3zJW995nRdurJjvH3Px6BHJtoTdnGyKyc5czomjI0rHuF2Tm5ocM0UzIwrHbm35xg8fIPR62oXOZiSjUVGw2Z6x3Vyg5gfohWCxv4eQmaoW6LTjbH1Ks5phmjl3H4yMznM5bjm/uGIlA7Al2hKvExdXjtH32N7yeN1NKPUgCSFgCsXycE6VMufbEd95bLboJHEpo/U0wj7vO4TjaRJTJuaJXgyC9DSp6sdd74siEKLn8cM7PL+Y887pPbT21GVBnROzbHnYdlwvDUaXVLkkZ1CyoZKaNimynrDNKg2oIBHNPsPwiJlzjDlTzhY4CgzTdjMLg5EKkyVFUuhqTuEFWVqaBswmoaiRWeKCpYg9Fo1ICtkP+DoRQqAMCh8zWgRsTJy5wMlckEJGBEGuM6631KrBmJpaWUSSJCGARB81BylOevPdDv3ciuw0QgS8nI4/FImsEnKU/Js/8wtsNk/4h99+jetljcrTbsXoJVJYmnqOkAYhgTEiBXgp0EZRFwU2ZsZuhBCZPwuazJP1Ja63PLM44Y3qHotmTikVeXCIVUFZ1HgfcUNGrDe8/MpPUmnDLjhImtnxESqM7AZHIWtyHohaUzUNZr5Hu90x7kaUUWTvGTfn7K4u/mkar8wSR8JUK0a3IVtPkkC0CKmnII6qxrYbyJLBDcycQpQLYkrEZIk50bYt9y4tUhQkBX/y7XdZ9QPd2aQz6K62qDggVU1RmMmpaAqkeqqsy5CsJSoFIiJSpJgdcn5+SVYts9kMkaagVIFi27VUEkQaOd/t0IuGveUCkx1SBWzMVD5x1ScenbU4F9hcBHabgbWMHBpFSiNOZi77njgGxhAY+2FKWgpqcouOBh8SMQtsP072avLEXRTgLYxCTNmNQk0kJQEmS1KKT9uE8KcIBt8fRUBLybLJ/MTHfpo7/SWXT3qSX7IWiZ3N3OMBxklSKLDjmkN7jJ5BnTOn/oy4SAyuJcSRXY7sHR6g4gvcPz1Fq4F5panFwNaVlLEEMUIR0VpxNlqa1jPaHaIpkGSamULEipgNd84vGbcCrRR7M0+ygbXoMLGhni9wEqL1OGv56nff5fAnPoJI04ir1JonriPPFqxWK7LtCSLje09tEtoJHl0mbpxoHsqaJjmqZzSnm4bVzWOi7dHPzNivK55Z3uCf3Pkm33v4HqWo8Mt9uLog2Mje3nXuP7zLD975IT/x6qfougFk5mR/RhIV6+CIfUezOuSVz36Bb333Dda7J4R+5Pe+/084XO1z7fIBq+vXWBpNXLcsn30eUxuiTbz3je/htyP/9pf+LLcaQ3SJlCKEDBFyXRAe3KM8ujGp1DZb6g9/mu5qTZaSw/2Dpw2822wev4tLkaxnMGw5f+89Hl9suDYvOHn1Y2wvHkMISLcjjCMyeygLTJ4TvUVmCINlfryPCZ4Q4P53/hFvnFpctcCkSSd//uiM5Z7m5OOfJbodr33nO3z4xgpVTmO6setRUSDN5KgUMk6W9mAxo6aoVvzUpz/N//Crv8GT01MMmrp5jG5maJEI1mEOjyii5d7b3+f89D4fvvUBymsVSUBvG7799pbbj3Y8unuGIExJWzFjc+IqTFiukNdIQEeJA0gClyZXakAAnp31T9dwRghBSokoprhMAf80yj3lAFpBTAQySkwCq5wziff5dGA+O+CmmfPH3/kuX/7Zf52v/9EfIcuacbji7r2SnHo2acTFjkII+piw5xsa0aKKhFzB+uIKRw8qcRkd4xgZvWBRJGJoQSWCNwgjsDlOH6ZMJBe5vNyyTluuzY657CxGBUSUdC7RiIKqUghR0YeAzInNbkuSiSIphpwJNtCJTLltaduWel7RO4sRktpIfDeglnOEzGid8IPDDZpwHLD9yLg342iRObsIPPdMy2xVsGkT+nDO0fERn3vhVf78c6/ya29+hQ9Uz+C7zPAWJDvgreHy7C63H73OJ6+9wOluxue/8PO8+/a38FJiPdzc28NqwQsffAVdzvn4Sy/yrTe+wehHaj3jg7dO2Jyesrh5jBkcth+5df0mb9w/Je82uI0lKs3nP/JBYtGgbCS0I/XxIdl1uFERosbHgBGSXNYE21LuH7J++/vca3teSpJud4W3FrXcY/fGa+wd7/PD137Ab337HX7xsx8mJw8vOUylKY0gO4tKEmEKqnKJEy1x8BNhyPbockYIjjd+8B7vrHuE9ywWmiqXHBYFtz5wk9VzHyLHnm59RnrhmAS06w1CC3Rdkn1PSh4ZEsG3FHJF2z6ibBYk1/If/rtf5m/8F/8tSSd8H9BuZKlLdFMh+55OOVLMXDx5wre7njv3l8giMOx2nD22bLoeP0asiqSYyXGiH1sEJouJKRATVoinC/bpXD/JCa6kIJIhZ0SGKDICgUni/+YG5jQRk0QmB6aGqYwTnk2kieaOgB9TCN4XRSClTK0WDNtLlotjRicI0bJuR4YuoaRGLXq0gyAsNo4wBsakqYQiyExvewQjuRKst5eEzpHI+BjobKbSk81SPhWkJJ2JAkxK2GCxKmCVBi9xKGYqo4gk7QidQquSYC0yJZKM6CgZomeeDdFnZMzIFLEx0RwVbHtHTAoBE9Mu1aTIJG1FkMdM9hljCnyCEhh6z+mF5+hkjpprytIgZcHPv/Qq26s1JiuasqJTgbEZaa9AZYUII8v5iqreZ76Y0F7KDwQ38vrdlno257mlobx6jC72WGrFreMZYws/Pf8QWgkejJbPP3+LKhqc6+kfn/HWbsB2kaZssMnRtwP1ssBLiBKiT4gc0NaQhSD6RLUoGUfHsDmjKVcUKdCeXtB3A7bbEMJIOn1E3cy5/d5DfuN793jkMn/05m325yXXXvYECuocsc5TNtMZXiuDqSvUpCFG+YFMACeICKp6wfIoMWxPURFO9guWNw6QIhMwyHLBbL6iqCpccmhpkBmSMKTUI6UkCEWOkbKekcaWrCWvvPgiN69d5+GTh9OUQGiCzBghKHMi+YRUgjEEhqs1676fItNcZOw8PnpSmp7EU/7BlKgcn3osRI7T0syJnKYEbvX0VTw9008CoIwXAvV0NzCpAIH8NJ04T0/8jJioShIQoDITC+H9rhMIY8eNT/ws//hX/nv+5ld+h5P7P+QHjwfWhaNJ0J0K5vM9VN0z2B1DusKYBbILbAeLsIqm2Wde1Tx+8jZuPOXs9Jx2dKTBsX+0Ym0EB6s9dv1AKCwilYCit45Sj+QcCbGnCy3WRRaLDbrc43y3Ybf2lH7LXnOCaEpGH9hHkyNYO9I7TxQK7zq++dZdvvSFV/ntO9/DOUdR1lydbSH3VKZkpMcYQ7lx2I3l8KUD7NojapClYfAFj21gf75isCP/6V/49/m7v/2/8av/4G/xM5/+GHl/hV1pypcVppkzbAeoGn7h57/M63/w9/nbf/Jd+s0Z/9GXf5lvf/13ubWCg8MSieK//rX/kbyRfOYzf4ZlecaiD/yv33+Pv/Hlv8xvDZazr/4+KQbK5Yrn8py/8i99Hmstv/KVP2BVwWw+Z3d2iSgk28tzVFGy3V2xf3hIt1tTr1ZoU+POztCLV8nW0u6uGMY173znGyzVR3DnF5xvBn73vTXfevNdTDIUleD1y8Dxe0+4+Yrj8HgPb3ua5T7Fah9VLBjblv0bz5P9AMEz+J6ZKimrBV/4/Oe4dtbx3nt3ufHilyjywCc++wKLm7cI40hZLamaOaY0KJmJY4tvz4kSyroiu0wQgbxtIQWGlBl2LfOTE+Zlx71790i1waQJJee8xYpEVRoKBaUwhNYSksdtBFUGKwXOwZDF9IDwYlqYJFQG/TSjMmWBEBGdJZZJ8ZcApQU5CnKKaBRIMS3WDIlEfJquxFOjW35aDEiQREKlH3UCJVL8iIT4z24MvC90Av3o+MO3t/z1L/9lHvzwT/jFv/AXqeuaWhf0CMIYkRtNHiRKKZLqMFqQNHjrEEJx6+UP8sKLH0AXPwq2TKigyASG3UA3ZqrZgpSgVhJDxES4ftCwX0lUjMQUSM6SXUahyCnRAtshoFNJVYIqImPwxCwReZzOZ8mjhCRYz+l6TWMzMmUqLRGFpJCBcRwp5zXJSQoRsdYSBot3oLVBSAPLSNqTk5S2zciu4Fd+4+/w6//g7xKLwNXbFzx584KX1HWeXz7H8fExzz/7IjJK6mpOe7FlvHqEbUd8yLz8wVf5379/j3dP1+BHpEjENGCCRXUJWZdcf2FGUVfc3JPcv/Mm926/zf03f8CDsyf8zmtv83tvP0Di+Te++C8z7LbE9oqh21Bpxfbq/2LuzWNky9zv4mkAACAASURBVO77vs9Z7rlLrb13v/3NvNmH5HCGpIakJJOSYsuCZScSBEt2YEZxIsuIpGww4ARILAdy7MBIDDiBHMi2FtqAYgFaKFkKJUqWRNLcyRly9pk3b5m39N5d693Olj9uS1EMjSUoDsILNKr6dnVVAVXn3N/5ne/38z0miZFqXpIIRb62zv1XnkdrQ6hbfDNjuL7KufVN5osF0UUaBK6a8txzL+CJBOMYxIjBcus0UB0e0tSW/niFweo2RbGJFBl5f410vEU2WOuit3qrtEEifcl4c521OOfDT53nmasZzzy6wWDrQWJQpOMNGu+x3mHGY6JMWJzuc3LnDuXubaa3rtMc7XUXSiVZzJaApjCKJM9omxP+9PufxQQFMhCjRWhBoSQnbct+7UFZvIs4BMoJFkjqxtE4C60D0ZGsoTMERSkItiNJRecJUhGFPItZ62T0Kgrk2bI1xogmkMqAEJ2OQCmBVl2vSitJkhsGmSHPElKtWSkM41GfUT8lG/Y65sPbHF8XkwBKc//+PXa2rtG2LRcuPkCa9UiKDCksTdvSLB2hTdASvLII00kkkxaCiyRZhiLFiIyoJA6JNpCoTqwTiARx5vc2LQrOWPOdvLRbh7U4EWhDxJcRHxUyxG7N5hpslEQZMU5Q4zCie+Y2dAm6VgjKM35+4jSKiBIQpEaErvR30eGEx56VkstFB800KsfGgAttF8wxb/mLz34Tv/3JT9KKliAF3sGKyrgoe2zKHO0Ey9pRNgtaZ7GhIrSWxXLC6eSIIh8xP5jy2q0JdbOkDBYroXENygastCACl688zDmTEJwn1FAtLC7W3N2/x2uvvE5ZtTz58KO4aknQkmY+w7WW0FokAuuWIDX4CiEjqijwoSRIEFIzm80IMkH3xxTb51jf3uGB8ZDFdIaIcFwvIOtYetSW8ugOSTFCZwUYjSkytEkwxpCkfbQZopIMqRKiTFBZwsXHr7F66QK91SG98RBkRBQpRIWXSQcgkRopuytoVBq/qLGzBeQGV83PPASebHXceUjaEt/UvPudj/D4gw93Hn8vkS6QGIFoI2Vt6Xt5lnYlaEQk2EDjAx5BojrR0+815rzsGnlRQECgEKjfw4ApSaY1SdItr6QCoQRKC2SiSbRmaAyJ0phEdirBRJMYwUimpL2CUa+gPygY9nuMxj1WxkM2BgOGw8HbDr+vi+XA6WTCN3/Dgxwc7fK3//Jf4yd+51XOjXIOJhOaXkRkGqyh3XNQ5bA2xdldGpWxMlgl2MD1218jbReMxjDOLdmVIe1ScjKpIE07zl1zn+31httO0kgIRMqyRKsEGT1GK6KS0CYIxtBatNBsiQFJukqxOuag3KVd1sihpmcygrccIMBGmqUlSsFJ47jYX+dL129y6bFt1ncGTHcbksyQypTGtlgC09MZtXG07ZiLF1YRfsH8vmQeBD/+t/8u/+gX/3d6D2Y8OH6SRGnGto+1FT/7pU+hE0XmKzYaQ2wazl14GOkltl5yOitZ7r/AL/3KG/ylb3s/r9+7xUAJpFX8R9/1nbz1xj0SmfAtH/rzfN/GOX7tp/8+j+6s4ZaRwkeElhzWS1aLDCNSfvQ//uu0p0dYK1geHlE7jwgwXCk43dujN94gmMjy6BgbJISGxaKmJ3OUMFy6dInjZUOWGQa9VdTFh/juWckTt/bwsmA/JKz7UwqZsrEyZu3aNUxvBdPvo/M+wlqc8AjrMVmGG12AtkFlOVIIlK6QO5cwWUqrwNmIGYwJIaCkZTlRrI00aZp35iAzZHDhItXRISiNP56Qr21gNkb0sz5ZMUKurHSQ0sN7iGbJt3zwMb76yus0jaUWlkkTSFJDIsBkGSZTKHU2kWrJdj7GKE2RdSxCF0SH+Go8RZoRUkNoKrQDYRQD6Vk0YIzBpoK2tCwWJdo7Uuu7ODatEL0Cpbt+ghpmpNmARChWBkOiyTAEQgBsRMmAyDTKwdJafuqtvT90/H1dTAJCCX7r5c+x/Z73cxQiszuv8vTVc3z2S7eQWuB01y3VPuJQiNZQaYdULQoFVpDFY3rKs1AeGRWpb/BRIlVKkeUEEehnDSopCJVHtZGgIsErqkC35pKdYWgQFBmKVgYyYRDjHJOlBNXhuVQjELGTKhMklvD7xiOHZVI6BolgUVpiHchST1U4goZM5dShRicVIXYTUagqmrIiUwlBt0TO8bVbL/Hy/qtcu3wNGSXlrEILuH+6ZP9gRk8ItrZ6uLpiWc+JISCM5PLWKqenFcIJLKD6ORfW1vHecvXiY6wXAz61rPhz732SzdVtbt+/xX7dcD46dFSsrY0pBmNut0dYleBkRLiaen6M0BrblISgkIlCeNfRkyUUWYGzLVInYAP1Ykp/OEJmmjzPmN+4iUo0SmcEItuXLiBUyunRIatBsbFyDldX9K9cRWUGnRbdOjbaM4CIQAj5fy+9okebtINm6A6rHYiYtAexwbkapRQ+CIrMYIareAdN02CkIB2MO+jnoiYmGpUakiQFb7HNAu0lIh/ga8dgrcenP/0y/TxDRYuTXQmdKkFhNDJKsjQhk4qYSDAJ434BxpCbrhqREaxoCJauulGGoDQxWJIkJStSZOhyGVOhqNOawnlCULSFgzKSJ5K832HZowCT9VGDIUposnEPGQTKhQ5iKh1JYhBKEBUk8u2H+tfHJKAFs+lNPv16inct3/+Nj7PzxLv5+ef/FXWqWd+JrMlIVSWIZEApJfthhhk0zE4PaCcFvXHK5KTEDzSHZctJWTGWKTvnrjBZnFBXNZM6YeDBxbbbdvGCU9eibeTIey6TkYiCVZNS+pLgAlHD6qYhFYZ7010O5hOqtuVqsUaiNMFbljFiWih9JPGSL1+/zXc9+yg75i4neyeIQcpwo8d8UTFY7xNFh1YPs0jrIj46dk+nXL6S8r07z/DOZ/4MP/faJ3jqHe/i+vVbXL95n/Zuw3c+/U5ev/46cx0RDLmSrbF3csqyLQmu5GuzI8xqzoc/8ChlbJnPG372M1/ge9//NJqSH/jQB/jol36HvXtv8thHvpff/NV/xr/43KeQCFYMPH3tIZ669gD9Ucb/8rFf4bAf+d4PfAC7mHYBn74iy0eIeoELjqouyVbG6P4AaQzTk0OkMsznc3SALLlLMR5z4co2r776GrK3iqAinByz+chTDC6eEssZQWuwEZ32KUYDskRjit4ZYSjF10tEdATrUEISQo3p9QjWdTbrNKOxAd94lDSkhe4MTjFAVBze/QrrKz1Gm5sc7d2gZyyjjSsMx+sc3LxNeu4cvd4aIVrs7LArx4NDq5yAZ7Y/4fFLK7z78rP8yhdfgtoiRMTkY3TPQOJ4RGzj8pSwmBPRpP0hxhTotDMpOS3I6MRGKoCtLQs5w/lIkQ/prY46q7pz3e7AoqXRhiJbYZhKlDJo3xKzHkop5k3D6s4WOuloySezGfXxDJ1rLmxuM8wiUieUtaPoa1L7/0HuwL/Tw3vuvnKDP3P1Wd7ilNdqzUbUPLq+wedmJVt5irGw0IL+UJJUGVVbIUTJoZ2xliSclBLlJYmP2LBEuyUij6xtbHLjzk1cXYEomMgAA0EMEhEEsbE41+P8+W0eeOIpzj1Ss2L67O7ucjC5S9tYjsolA9MwLWuqqgYPWVYgtSKJER9g1loSOsT1vcMDdPI4D29t8+WTm6ybiJh68v4aAUiTjCrLuhgtEZFCU1jYzi7w5FPfyZ2DG+ysrfHm7h4qTknDknmInDRLRKxwSpOmFqoWlxmkapAS3vPgg8yXBwxUwXSyZNLU9PM1ZpMT+ptrvPbGV7j55nV+8K9+hM/9nz/LWBjGOuWl+ye8lM8598AajYZNmSCdpB8U3/TU0/i9O7Sn8y4lSWhMv2B5fIzJRp0/XymaxYLER+q6QqYJRkPdzCkYsnHhCuP+F7FoYszJ16+BDOT9VVi7iJ8cwKohIkFnXUqRkCRGE63AW4s6g5e2IaC1PpsANDFarGtQIsHhkNoQSRCJBN/inaecH2PxmMEYfftlQlToVCBVTm9nC20V6WDAfHafgCTJUqrDCaY3JBms8uCG58Wbh7y0d0LfGMgLvAtkw35X3YQa28/JMskyKLSrqK2jVSV9V+BjQHtFlB0urY2CxrU0rnNQHtTHFHXNYtGFkrbaYxqLi3CyLEkaRb8XMLM5ZV+QKE1PJbigmO4egvUs5lMCS/rpDqmStFlGZhL2Tu5iRM5gvPa2w+/rozEoIFjPOE9YLBt+4ROf5HRZ8aGnnyGVjjZYGqGQOkFESS4DSmegUtrgaV2NUDAqEhIfSaJGR9eRhILruP4+YK2lqT0Bj+z2UkhCQA36rK1tMB6MubixzfbWBufP77C2s4PznmW9pG4sWXDIKhKVQitBCA6lVJd7H+LZradqapRMGRUZdeuJiaBdtqxkEqkjbRsQShBEQGiBMB1k82rvCoP1Lc6fu8JsvuTu3gEqBnRu6G+NWJY1ZdWJRFzrCXVJUBEZUprG0h+PWZrAfmjAtjRtwAhFjmUymbGYl7zzycd5+uJVmkXD6bJiPu+4enemc9aGBSHXCAUqSB68cI6mqjFpijIKj8C1y+5zUAmSTpO+mByhBVjhQUSi8wgSRND4ICkKg8kzCKprcOKJQiOyMa5ZdA0xNcIva2JVninmup2fEOuzbrhCaUEiJF7oLpcwBKJUaFXgWo/JC2SkIyqKLuRF6oyTuwfoKHFtS9Yf4FBE3+B9RSY1oTxCSo1yHuUCoVlibdWBT1c2EVGxvjkmQ6OSDHNW5isliIkiRtE1HU2K1pJGKETwHZpOKoxWHR5NaSoibduiEfRUAjrBtiWLkxnT6SltUyNKhxYREz3KV4TG0ZaeJnbJWNFaZm1NOauZLxYsqwWurVFtgogWLyDT3TLN1o7ZsqRt67cdfl8XlUAUgW9/9t9jmMBLn7uOcXN+7Kf+Of/zD/01vuXmdX7jues0KiUhItsZSy9gEZkBjYwczhsu9lJGSnFUd6Kddu44un8P615mWVfgIkYoWttio4S6Cx+Rw3Xe8eTjTOuWw7YknFZ46djZ3uDa1iMsqgXyzX1W+0PC8Qk9IVnqLvCxDYKhEbRNQNhIOWlxbUSOHHu24dzaiPa5krDTRW3tnRyxtb1OvtPntRfu0WjBWr9AJgoVJR964ptJ8ow393b53Bde4aSecsMu+eC7LqD0kN3fvUUcpFzdFHzxN/ZZvzTn/X/qHfiq5c4rv8uVp76Jzx9+DdKanUyz/cQ68zuHeJvzma98kf/mR/8uyjf8zL/8xwysZVFOmMwq3rm+wb2TOU9fPsdnb+wS0oxxkfGei5fx8xnLuqS3vY1pW2ZHRyxnS1ZWOiR44wO+aalFRAhFZiK2XtAu+wQvKWdT1s49wJWHrhKjR4QMmY9pmlNEvUCJHDlYYdnMQOZII1FS4mOAGGlbi/IV0TfEKNBSIZoKsryLHUcjYsD0C7z1uMUBanQeQuz2hHxDsAuuPf0OdGxxIiHRHbdPy0g+3sDXpywne4TFFOctWq1A1kemPYrBkOf2vsRLtw6pWs3muTFVDMynC5RRbBQS1d+iti1t7fBJTj/pLlZRgJGCkOZdEIpIGC7nVElClmXY+RIdPSsEFmFBT1ic10gZMNLgRILAonxkXs3piUhfeFxTUZFwuHeboe7s7jpLKISknVfs3rtPNloh14L2cI8qWib7J287/r4uJgEJvP+ha9y5dYPje3usrWbcfP55KiFpshW+54lH+Nev3ePG/hEhLJEWylpQNZZiVTEaJBzXXcrKwbTB9S27pzV4zeSNV0E4BmpAs5QIZbp0IStpW7hwdYfKexbR8uaNVzFHDSfzijtXLrBxeZULl85zctQy1imTXkt/LVCXSxIhzhDPCcokRBqq0mJx5LZg73TCzuqYunU0LeS9hOOTE3ISVtbH1E5h+ookSajaQC5a1tbPcTS/wyu7z2NVjVYpD25soQYKWTWYQrCaCVzRwzZ7fKnJeF/VseXbxjHor7PWW+V0eYRJC66cW+HV2xWait39CbGd8buf/Q0+8eJL/IeXHmTZJGw9eJnZ7hGv315gFy0bOme5aLDzmvF4FecjmZS4psbIgHCQDwZIESiXc9pFRb7Sx1mYTU8pTIpIDD40yFhQ1y2J1jz0rvfhZWS4vtYhu8wa6ISmnpylFidASaLXCTH+PmhU4hBSIqVGqIRmOUPGFNs4kB4VIyRJJzE2GdW0JMunyLRPazsnnYmO5ckh48uPY/IZs/0baDxqMGZ2eAeVFtQn+4wvP0LbLGl9wPuuypC24vzOOm8d1yS+YXhlk0GzpJwtyXTCuvfUwx5tpWnakqIQSAzadXwMUaRE53BekqaKkBeY4DFpghKBYZTcmCdoIbE6IG3EywSnNFprdFDQdheZmElkZhAiZVg1pJnE6ByLQOZZh1qrGuqyZDnvSM9GBURMifXy3zr+/n8/QgTlA1969ToRj1TQhCV1XfO5k3tcvHCJ2DbMD6bMT1rqhWc6KVk2ljJIpIy0AtJ0QE6KaQO+jQgvkY3DO0HwkbZy4D0hdMjoGLtS/LQtGSjJ8mhKOV0Qm4rj3XtMDw/wUbE6GtPWneegMJq80EilsbFTgUmliUKdqbM1SDipK6KSNDHQBAcioa48J8spZdvS6+ckSYL3EVxDIz1RBPZP73F/ekxULVJLxptDouucbSENBK1RSqISTbFmuD6bUbYOJ7vX7qkRMYguHrttyTMBUWN7Y55/6TloBNopZEwJhWFlPeH+pMLaltO6YTVLaXyFHg87mm6iO8998ESdEELA5BnqLLOxLGvaRdux7rzHWovRkqgMQQmEUAgh6A36tK1jMZ0RBeSDVYSMSJ2hkhQIEAXWVp2hxzZ474kq7dxxApxvQUVsU9MuTs80HAIpElRS4L1DJCmxKgnOI0SkKSdo4Zkd7yFVl3mZ9g1mtNZpSbRGtB7Xeu5/7auopkWIhHaxBB9o2hk7g4LVsUaZtPvsEwlaMCxyoMXkhiAcOoCJCiUkpl+Q6LQLZBWSxofu+8bZ9yR68sQzzC0qTxGJJiadxsVIhYkdY1IYDQn0jCbLMvLBkKLok66MGfYLVJ6gdSdnVqlBGI3UGilBLUukSkikRBVvLxb6uqgELm5e4j//0X/YCWkiXHhok9mrLXY2J6HkJz/7Sf6HH/mv+MJv/0v+u3/ym4i0xsQUqbvElenUM8z6PHr+IuNsxEF1l72DY7TQKAvLkabC8cbxWzAc8MiFdZqg2NjcYWYrxr2EJtacTA85Pq55Zn2HaayY3rvO3bfusaK2aQnce3WfBsH4Sg9joJw0mGFOniUMR30Owz5KCDAJrxzc531XznM+HbOYW/J+wtFd25lXFhXvfv+jHJ7W7L6yT2MEUUuODm7xd37yn3A0mXDtyXV89GRjwY3rE3w9IyOQJArZwBPfvMX2hS2+8uIdhuev4NuSxWLB45uP8dLNl9lKFdePjnj2gVUe/MYP81Pf/X18x0f+e370B/40o2yEUQlHiyX3fcXaOzdo11I+9vnX+C++6wME58jHglZE5tNTrIokyiBzQ7+Xo/oFwXnu7d3FNTVtyFkuK7SQ6KgQOsPXC4arV2mi53RynzzfoN/roaWirhb4eoYyg04l5y0q7TF98xVGj70LISEmXbS2EJG2LEmLHjFR+LnAS4dvl5QL2YnHmgXS9BB51jkfrcMQumSn2nL79k0u7Lwbge5ArzKnXU66/kE+oKwg27xKUh2yWFa03pKON2nqBWU15+e/9DJRJGxfHHBSL3H7FUOTsa0c+5MZ5ryipwtaUWHynCxPKNKuotJKIIWm9jVVPaGuPN46tDZcXBP4EDh39SK2jiwXNd5ZIrbLczAaSSTRkry3Sm9oqVtonODiAKKwnE4XLGaCdhHoq8Ayz4miU83SOKLuoxNYy7K3HX9/ZCUghPhJIcSBEOLFP3BuVQjxCSHEG2e3K2fnhRDiHwohrgshviaEePqPMwksyq6T7YKnrjzFeg/nIkEE8rZroP3TT/wW7/3Qn+XRKwXSnuG+Iig0ic6ZT454+XgOBFY3hmxd2GBR1uzOFoSFZ7osOTmeQ1UiHAx7q2SjEQNdMJ0vOD44JgC+apBOkEuFloHTe3f52vMvsDvdx59WtMenACihWVpLU7d4nTLKUxwRGyNJ8JRlxf1Zy5XNTUTTYbd0A9Vxy9E8YKNkY2PAJDSICOurl/jdr3yc2zcPmNyeYMjYuHCek70ZlHNOZw1Zv6ARjhgCKzsbHByXtG3JLNZoDM+9dZ2VdMRVn5OpyAOjnA9+y7dzYXOb/+2n/jGBOa70bK0UOJNQLlvu3jhmvD7k8mNb7E6n3Lx7wupmj5XxGsv5lOHKCG0FOsup20CyMupot70+2WiNbDDABU+aJvT6OXKQd42vxGCbGl2XtNUM304plzMQnl6akRRDktSQFyOEEvjpIbrIO1+/DRiTgQg00wU6SRBaUC9LhArYkwMIgjTrI/Ic4GxiqLowFSVxrsOWRVcRYuezb5dHXYXgYwcU0Rk4z/TO65xcf5HJ/rTjBXgHoaE+3kNged9jjzIa5MzzSDytCTbQVh7l5qwVCbQNMjOYvGBcFNgYOVksqVyNS8AmhizLUGkkNwmZ7rBlx5Vhry6Q6QjyDNnLWN0aokcrFIOU0TBj2B/QX9tkuLFGko3woSKLHcSmDpqqTQhSI7MAQpKnBl0YxsMeetRjsDVg+8IOp+LtgQJ/nOXATwPf/m+c+5vAb8UYHwJ+6+x3gD8LPHT28wPAP/rjTAKHx8fsPDYiGydMW0t/dR0bHTtrY+rFEh0Vr3zlU3zHD/4N/tsf+S8plzXltGJxtKA8LSmdQzVLbt76Ii+99QZOSD70oUf53r/yQS4/NEKI0GG0U0EsuvVlr1hFCo3fPeXN596gWc4gRPQCbuwekrhIFjOMUNR3jrj74g0G1wqKnSGqUBQyoOpAWbbMJpa0SEiVREXf7Rg0NZ955QbvvnKZ49ZRVZF8oMnHOVIl3Lm1z2S65MGHLrAyXudvfd+P8Ol7L7K+Aq4OrG1f4srFS0yO7nUJRW3gHU89jumvIkTg6MYJu6/vY4Vlt1ownc144/q/5t7u67y0N2EeBevjIUko+egvfJQ7WnDuvec4ySLvvfQQNA13TpfUk5bZCzMWeyXnnzzPL735Flme8pUv3uCw8lR1c5acFNi9fovlsmP3hRDor44pts4z7I9YXV9DCkOxuspo5yK9fEiiOupPtJrgG+bzKdOTPeazA9rFhGpxyPJ0l9hUlLMD0tEOAY9KDNWiIjZ0WPOmZHl6ArbGO0e2tkW+vg1JRJocih5pOiLtFb9vlk2SFCVTXFljoidGQX18n2AX+FDjyxmhqnAS1q5dYuvao6xd2CGcGW5EcJyezpD1khfv3OC0X7NYLKgOppwfK77tnTuc2xiz/sBVfHRoBCpPOW1a3MIS67YLTykrmnJJL81ZH62Q9IdEleGFpF5YXBmJ6I6alWmCb8mERQqNRZHlBiMX+GgJjWCcC9SwpGoaqgZkSDASjND4RJCPM2SaINdzBus9slxhNlZZiuRPPgnEGD8J/Jutxb8A/MzZ/Z8B/v0/cP6jsTs+B4yFEDt/1GsIISiynFoKiKLT2iuJjhZtAwJoQyBLAv/glz7LR77tCbCRGBzl0mIbh7MS0oQmNCwrR+MUSap44PIWPu1EP1EKXAKeBIRCKM3xyQxlLV4JlI9UeBbLFttYgnCMTU4iBc4qTB/kSoZWiixKAuBsoK5bRFB44hkvP+J9ZD6fU6QpPa9orCcWmiRLyHXGbF5Rn9ZsDfqsre6wNVrtOPWpQWuJznKCkEgJxoPUAhU0G4M+PjomiyWqacmiRDRdH0UTkMIzaSt6eZ+t1R2O7lzHqsCw32Pn3CoyVehehouAD9iyC2Sppk1nZMkiM1dhS8trN+9wMDk5A31GTIwYnaMTgbMNSih0CJjegOA8uldgkh7aGBACoRQRQesdioQ7X32O2d3blCeHeFvirEVnI2y7QAqFzrrVaVMuqWZHRN1BP51zpEUfhyDgcbHpxF4qErTqegbR4lwgSfIu5yHKLseAgDQJbbmgnk9QOkOGCD5gihwD+Mp3zEkAoVGZQQhJ0u9jfWBtdZU8NzRzR91aRIgURqLyFKcNzkaC810vICrq6BHC4pTABo8MHncWs64TQAq8DARvCVpik0iepyQ+wQvZpTrphJgoaiwhOkKzwEaH0ZpcGPpGdr2tJCJ1ZzWOUqCUJzqLJe3yIgjgS/qD/E8+CbzNsRVj3AU4u908O38euPMHHnf37NwfNvB/QAjxJSHEl4KPVN7STEskiiw4EjHCtQ3L2iNc5NRJdGb41C9/jD/3nX+Bb3xiu/NUN47ZdIloLFIXGG2Yn8z4/Nfu8taR5cGHtnnfux5iuN4nzzNMz5DHAq0UcVly5859UqnJvMALh7aR5bzktGkpraOfGfpbPaKzrKxlnFuTEDWV7Tq+betoTic00ZPqhERopIhI5zg6OWbYS1Exoa5bVCYww4KLWzuc3J9x+9YexWrBB9/xHprlnJAJesOEopcyHKzRLOeIILBJwGjPwemcc5tbBGkIyxqdSjhLURbFgBgCN09vM8w1FzYu863v+kamkwlv7s05mZ5Q9HvUoaV2DWmasHZxnYuP7ZArSTi17O2ekijNaydHIDwvv3qDz7z8JkmW0iznyPQM8T1b4JwlH/YZjIZkRY/xxnlCCPRGqygtEGeNzuzcJZrDA6pyRjU74eDeHeb7R9TLU0Jb4pqyCwgxiuVyhvO2g3xIWLY1dV1D3qdxHucFi0WgWi5ZljVNbQne0c5PaMslCkEILSoxBD9HIZif7BGMIuv1EPkG0QucPxswRtE0ltHWBRSCaB2J6rIFYpKQiMDrd2Y8eeUC3i2ZvXWMb0oSa6mrOT5I9meOum5ZLE/xTUltS1TbrfvTaDFpSprm1JXjZFahXR+kuwAAIABJREFUZCQtUjZHI4YbG/SyEX2h0BISAxbIsozB6pi+yVDSc7qoKJeHVPM96rYCEToDFJ4oBXm/R0MkCEHW61ELGBY9RH8V7RPaUpIXvbcdzP+uG4N/GNb0D6UZxBh/AvgJgNFaL6q0o9Fm4x7tYsH7PvAUt9/4Mt7PkXbIovEMC4VIAt/5w3+HH/sf/yY/WO3xn/z9f46btUzamsFiiXMpuYS7r9xi/61d/Ief5OrDGzzwyAafKhT9/pCeKTg4OOatN6+TyM7u22pP23TvzQfBvWnFxtAgVcZDl67i1ydk/RSbtDR7nlBIhqOc49MjZOs5nFWc2+xxsDdHGEWwEdE4rNA8fnGLkxj4kW/6IB7P3mHLX//wt3D/ZMLv3L/N6+IFfv1n/xUjJcn6PbIdgTae3TfvsmgkmbBEEZjcv4fefIRnH34Pbz33C4hegqgci6Zl4nK+9clnuHv/89h2wZXNNT72iV8kq5Yc7R+QZ5L7uzXl6pR3XDmHF5rs8gbvuHYZIxoeip7f/OXP8+pbhxzsbvKf/uVn+aXP3uD2dILK+iSNZYlCSoP3DVIOWU4WSJVQVZYgG7JegYgNtpqTSE072ac3WEVuXKC/foG1tTc4OTqiXHyVtfUhgYAFvHco1Uf3x/joSPMCHwW2nJPpnFt33iJNU/ZuvEl1WpOvjJnaQD4YcunKefoG+rrAxgzzex1/Z5HC0+uPKIY5wgFSYLKcUBnylVWETDE6UlYTfLVkfnjEuccvM10coXWB7ktUv+bjd55nw47Zb47Is4zUSNqq5cZ0yo2JZ3vzPJUVmNBijCSoHOsT+ivrFPmQ6COniynN3hKnNdGkIBP6ecIyNoS2RSaK2kNf9VGpxBOpXcsoyTHDSJr1ka1lYUuaCvqyhxYCrSowfYq1QGwDTexjadAqQ8bAcH2FdjJntn/vbQftn7QS2P+9Mv/s9uDs/F3g4h943AXg/h/5bAIKBErSYbiM5tFrVziY7hOdIHhP9OC1QQhFgeTHf+ZXKZM+Nnici91+fLSAwCiNSiLaBu69dcxx2YKC0cVV0l6BVY6qnSGtRUeJ1w4fFKlQiBhxiUAQyLxAKJjhsSKexYGBp8EkgiRThMYSo8A3lmGedFdBA8kZ7KGxLY8+cJXvfuYdtCFw57jk4197AScDW6sjnri4QTU7YlgooEHpQLExoq0qJstTlIiEGNEoVHAsqxJZpGz0R9goEVKDi4ToKBLNovTMXcMLb7zK3rTE+W6rTIaIaLtOfJSBqBRRdK5JM+zRH49Yv3yuC1OdtFy7uE7/wpjBMKdJFKQa6Tr7caJyivGQJNVMFhXe1UQ8SZYyny4QTUNVNSRpv3Py0bAolxjh8bHjLwQpiM6TZj20NrjogUBvuE5e9NE4QuOYllOqqmJ374DGCd46nHD//h4Hp4KvPvcad+7s4mWGNClBdrJhkRiQHRF6uHkOGwKeBmsbXF12oahJH2ebjvCjJC4GghIkvRVQBhcsgch4PKRcVOwfn5AkKTo1xNhd65wV6BBxMtBLC7xOwAl0anDEziCkJFZ00NiQJUQ67HwuXLfc8xEtuyzKDiPWoIVDUpEagRcCiUCkPUJSdJmYxRCZSTgLxY14TJpjpKDynugdy8UxtV3grUMkCm/btx1+f9JJ4JeBj5zd/wjwsT9w/q+c7RI8C0x/b9nwbztccMzDkv7IsHl+nfGgz3/wzBN84vNfIHjVadZtJNcRLSDKyOGbL/LDP/5z/K8/9BexLtA0FmEjpAkmEXiRULaeF7/4Op/+7C1ev9ew2htTjFep6gWFLVnvD1jd7iPRJMJDdIjQpRF70X3h8yCZLo+Z1Q1L70Em4DLKXBL6Atsr2H5ik4cf2mDrfY9w6ZmrGJOx9BJd5Dz40AN8+KEHGBvDP/3sF/joJz/L8y++wf/08U/y3OGCDz3wLr5hfYdcO6IvUEmfhx97gN37R9x54wDlI9YGam0QynB0ckwpLM9ePs+qMYgomCSQ9zI+/6lP8qmXXuO//r6/xGeu3yTzdfc+hhlJMWbZLIi6m0ict8QIUUgOj0puHs154IktPvDUwzz1+CUubl7g6Ue22Vjfog6aJklJR318VVFWFTrpkQ5W0W7G+vYGShtC42nmc5KVNfJej6ZqqWdzXOlIBmOKfsq0KvGiYDZbIBE0syOE0hihESKhbkpq6zi8f4uDO69Te8VH/9lv83/8/Bf5xd99k9vLjC/cXtJmKVoYXv7ic2jd7+yzSY6QKWgDQNFfId+5ylG0VPMlwoFMe5AP8UJSTSckaYpJByit2bj0IK68z3jnYUw+5u7eLk554kyzPLLIzDA2msp7TG4oE0l/UJDYQGgbhEqoo6UNDf3xiFh5mqo5IwVpsjQhRkgzic8kieoG78JWCBy2baljhbRzCjtloCQ2Snzoo1VB6QSjfA0nc1JTkPRSvJPQQFABpzvJ9Pqwj20caVth24bDZYMW5m3H3x+5HBBC/CzwIWBdCHEX+FvA3wN+TgjxV4G3gO85e/ivAd8BXAdK4Pv/qOeHDoK4qE6JK5GdB8aMR0P80W2e35+wrRPEagq3K1qdEAQQIg8+sMPadsF1OeJvfM8385MvvoLc6PTtCItOIlIF1qSjevM+L9SO937zkzw0HmEqxbxOuLk4prjcY80U3Kl3udu0JIOEONbkiUB7zagY01svOtPQsmRpAo9dXKfNYOXiOd7zznfSihZrBCeLOVfetc17x5cpyorawa984Xl20jmfPriHthmLwyOiFCyPb/H67gaqPqEpDzjZL/nWxx4gNz2y9W0mr97gT33wm/jV15+nDYI0atomML13wnxrndfnS9492OZzs+vQODbWLjBvFvzYd3w/n37ps5ye7HGQSC7nQ0b9HhkpzUbDuf4AbwST+8fcrSs2r56jcA4fGpK1Ie6iZigNbdvy5x95F29un/K5N77MxY1zXGgDoq8ZjIZYCdEIhptbzCYTorMM+itMwhHV/X2SQQ+TaZTIiTS46Sk7ly7gvvp6t+OTZKAzkmKEBJyW1It9dBziGo3auMrixmv8+N/7B4h0jQhIGZkuF5xOJ+z/+m9Rl8c8eWUbL1qSfEwiFK2tUDJHa4WTBtsE3nzhTd73/qfZeOAqPqgzgo+kv7qG1AlJcJg8w/QHlCe7mEFAUtIf9PjCvdew90uE8MS8R2+gGYjI/HSCkD2yfs6ibSBGesUQbGCQKfJxwWTa0Jwcka2uInTAGklST+jJHvvLFpUqVIBMBrz3jAZjMnogK5zztM6hlMEJuPvWGxQmYZGfwy2PaLKMPGisqDByTtv26A8GrKytMV1YZod7JP2sIzTXJf3i/4WVOMb4fW/zp2/9Qx4bgf/sjzPw/x//h0A3AqtBpgIXGl6+/jo+qo6wQvfBKX9mh5SCfp6wnQ948c4NfujZb+Cnv/ZVbJQo51DCkPYzRFkTx2PkWDJaG7KZFuQxMOqNmZgFc18zSsZsDPscqYJ0OsdJ6AlJ0AlCK3oJeJkidcrCtdjoEGmkt7nJ1fEO04mj1gHpPW3leWRtm1VtOHBzbp/Oef3WLfKrfYQItInvzCXCkQwHrPcKnr91i8rPOZ2copKH2XOe3De8cPuAtZUh09AgpAAfWISGedMQGs+JW7IvE1bHK9RNQ2IUzzx5jaWb8OmXbuIbi80SStdSL2oaGRFpgpLgassgTembmmhrijzBOc3e6ZQNk5OmgqW1DHLNIO+R9yTXVlepZ0eExhNdS2wdghatFSbrMz89oAoVIgp8IjsDDRqKBLO6ibclRif00py6rZA6I8nyjkmgDFElpNqA7HYVNDn5cIUkL5gul7hQ0laRdPc+i2ZOtSzZNAbReuz8hDbp0SpBf9gJkJAZ3keUUti2Rqc5oa7wMkGmKfiW1np0Dq4pUTJDa0MUAW8dvlpwfXrACRKTROog0SKwOhzQ9yV3TxdolRC1oBfTTtilIlFLwFM2HXItSonhLLXYaSBhWTlUq5FJpBWRVBmW0aO9RhvF/AxPl2iNc4okOGy5JOltQ+tQiUFKSdO2GOG6pOM2dEVq0yClQmloAqgkp3EThubttwi/LhSDAsFobZXUGB67eJ6if56P3/0y5y6u0k6PCCIyHOQYBXmW4RDsLyzV0Zx1N+Env2z44W/9Bj5x9xZeBNZEzuMPbWMGPUbjghbQQnIxGxFjpCk8s1GDHStuL+9zfjTgcrrOC7PbVJOWDZMj2hYRIgduykgUtGO4tXfM1vqAh959jYe2nyC0DZ95/Xe4Ot7iwa11jOlz781TfuTj/wJnFBujFbKR5+bEEYRCxJZkAO008vCD16jkkufevM7Fcz0MknvLGcXaORZNzee//BWKlT6sSzZXCxLRMplXiGnL4VvHbJ7f4KWXrvP9H/wQSinOr+9w9/av8bHP32YtGG67BWpacD8J1MuGVkZWR6sUTvHVvX2S1U1yl+KnFb/96VcYn99E9xNCEXh0c8RbreTdsUIsF9yZzhgPDni0ECynJ/RXt5EmpZlXXWJQMocwZH58ynBzk7ZtiQJkloN3hOWEeduysb2BlhEjOzm1Hp+jn2XYckm7PGbt/KPY8oQQYDBcx6YpjzzyCL/+67/OwkXqqsWkGhUCJyczqkGPc+k7UGaEFyVZMsbFgPYC8n6XZRjhgYuXKbVmuncX3euh+yv4uktyblwDSYpOJcuje/xfzL1ZrGbpdZ73fNOe/vmMNVdXVVePZLO7JVIkRYqUSEWWyWiIlCgyYAUCEgG5iBIgF0FgIHAunAsjQYaLDIpvYkCCI0SWIjuQY8lSZBGUSIri3Ozu6uru6hrP/I97+MZc7ENdpXNjJOgNHBygUAc4qPq/b6/1rne9z3D7KrZds1ks+e7qlLNjR4oSa1uUgKGMHB/XNBuHrTp0DsSAEqJnN7rEpumI68jubIqq+ki5tGnQmaOOhkpVINaIDDISmzaRnMXiOVo4RIjY5MgrT1ZYBqVmMJlS5jmt70jBI41laVcoo2gbwWRcEELHycMnrCNcmCqO14p2c8rO7pR1U7/v+ftA7A6IlDg5aTmYR+ou8okr1zgMklE+Y+M1WT7m2csXmRQ5L378Qzx16xIxdSwfn/Dadw/4s7/4Cj/z87+G6+DW9g6di1CVdFHxzsGSu++e8c7Rgj++833+7N4bfPPxQ05OGka5oCok7y5OOG5WNK1jq9Bc3BojRE7XJh6fnLCUNSo42sZzYfs6l2eX2bRLvv3gdX7imau8vDukXS/5jT/4U373y1/GJ4uSsHELQmOxQqCrirptGJkefrqlQG8sn7gw49Jkn5cmu9zdNLgkaOsVf+sLH2eWF30F5AJSDhBCMso1i5Mlu4MBZBm/883Xea9ukKxYvteh1oE/fecdrBO4FYgqZ7RbIKpEVpQsdGKalby1PEFWieuVItclphpSVRrrPF1ecG04Zd3AVAV+8WM/z4+//DOUO5coxzO0Vr2gV+ToakCQGXowIB9URCMpqhJdDMirAUYL8mzA3s2XSNFgNy3oyOLoGCkKTDlChoasnFLbrs8T0Bmr+QFpecjNa5cJSbJZrsD24NAuem59+MPs7+9zcTqiXRywWZ7h6gXReQIeHxMpSdq24clizuLgEFmM8a4DYZBZQTGeoJLGNTXtekWM4Op+DTzoxLqJ2JMa5zoUCZ0ZXFMzX88xWUbwLU3tCBbwkaYOKFmhTcFsOkblBWWVM5wOyMcjRDSUqaQFsvEEG3yfZ7leUJYlKBgIqExgezjpvQpKs/SBshoR44bgGqTtWB4vcS4gVYFUWe+TCDmb1RnatixDzslyTsSSl4aYvT+P8ANxCSgEmyPH+mTB2dyTNw3r+Ypl19HF3h482xpQqIIXbt3k9gu3GI5yooxoErr1lMMJ0UXGZsDSNmwSLOY1i6NTlvMFab7i6P57HDx4xPLJKYtVwzECnyKtq7GupiwKJqMSkQTOJnyK/cTBejrVMh7tcvXGJQyGh/MDZGYgBu4drvjD1+7RrRYkKTAZVIUnRs982SvlImmMzrkwGvDM5apfM6XF6AFJRFoCZZEhhKBuArvbu1y6coGB1uhyxK3ZPllWMBqMUU6idI7KBG7+kHvLmnJzCFFxf74k+gA2komcZAUyVxgdIdOshKD1CT3MGBdQ5RXXn77K5YtDdCYImUHphHOBs82GYrTN5d0bmDxHqEReDJCDAU10mKLE+R6KQfSYKu+XZrIcJT3K5KxOzihnOz1deX3KYFIhbeLkyRE+xX69uKz6BZgUSSnSOktsPcFbsriCfEC1s8t4bxufQAvN0cEBUUUKLXolvnM450gp0AMpIYQ+BKVuGmzbQmix5/N8rQq0yIlujdTQNktEntG2pxAsyTmstYDEpQ4fBLlIBG/7kBOlsM5hIiilcdL0qUA4VIoUQmGSojAZWiiyfEAyBiEdWVZgtCQ7Xx5rU+yJ0UKjc0nSkmxUITNDCl2/YecCYWP7YFnvaayidZ5oIzr2YbzCgyqGxExg64YYJBmWhKP8VxEG//94ZlVGih0m5Nz5xtf4e9/9EpvDU+rpFH/iibMB1X7FyG3z8sdf5NC+y+ODt+nWNeGcUJNS4KMffg5JwkrFRA949/CI0+MVl/d3kFXFav6IMqtRM8koZSg8Z6uaehhphOKZnS3KS0O+decBugl00pMKwZEPTDvJ3/7bP01Vzkjtmme0oo6Jf/C7/4K3v/cYOSh55fY+9zeOrYtbLE5q6ihRy5oDm3Gxqnl+d0KynjBJPHp4wPH8jE2QDLOC+mzDpycv8Madu1RZzp2ZZv/KHl/cv8lxK2ibyMm05pYa45vA4yJnf7bNr/2bv8T1wTaivEDTfJV6s2GSCa6Mp+T5hK9+/032n9pma3tCYs7ZsiGsHZkueCAFbxy+yzO3n4NRwaf3XmFndJnL0z0m4x2IAqLDWUvwLdXkIg1HSCGwmxVNpF/gSaCyiriqyfemrJ48QCbLevUe1c4eiyf3cW5Ftlpxsl4TmwbOHrFaLSgHxTnx2BC1Idk1UrXUcYkUkdms5POf/VG+9KUvc//4gCJpYkyMhvDqU9e4vFsiw7kwKBXeNb0RKAwIGFarDYNCc9I0yCxhuMT64C6j7cskoensGtdEgutYvHeIGlQ0TcODKqdeJDpRMwuaRjuGOnB2tuG0S+SZIdQlTdORj88dkkYSvKRVlsViTrUNIlUs5qdsVwXToWApRwgTMSLnzJ4h84JhkRFjpBqURC84Pl1wY6/g8vaExw/fA2eRRcZ6KRFsiDFiTCTVkhuXFwzGkcOF4s2TAZdmM7qmpVktuDgxdKqgdZD+X2zDH4hLwGQZ0QhSrqlmF/jOo9dRMVKv1gyGhnE5wIeWdWbY1AtOzw45dWs6JfFWsb1b4oLmpQtX+c2vf5X9GzOqkKNItIuGeK3nvAcSPh/QmZyuDaxqj5JQ6oLBZMTq8Rl5CuxsS05aj7UCWSokkuvPPE3SBc5u2Go6vvreO3z/5JiTeQ1SEokMTM649ZQonnQJFTwhCIZbklxmfPv4CbI2NCJjpGp89BTVlCIrSMpyPF/wzr23qXTB83KXO68/5tnpq1wqK4rxkK/f/Q6PWkkMiXhRMRiNeeHKFf7iL7/Hj33mVYbTGR+9/TTfPHjIc9mEOkV865k/XrKzs8fF4Zj93RFPX7jBg+aUveE2W5/c5tmnXkELCdIQokMmSXAtLvaUZuc3fZhmNaN+/BbZ3lOsHj0gtS3F3i5RSkRVIZo1ZjxBnz0iNYpqPMQ1NWZvytat56kfv03x598laIOJkc3JfVaDIbOdi8jkiClgQ2R1eIht5tSrlqaJ3L68y/rF24xeVzw+OqS1kZev7COVx3ZLWlszqi3rGJmMxihZ4KyncS1P7t8lRcVwOgHv0MIi926wWT0h0xXL41NMkVOOdojeYes1YTRBVYbaWmLwJDNkaCTBdsxToBOSQhsmQ0nbJIwIFFFgpUbOMnRXAIHgAp1dkVLECUXsakRQyDpRywbpIyF6xtMLNOtT1o0nKUmhClJyFKWiLAqWqxOGMsdUCh0GBLsilxUX9mqu3JB0MTGVnu2lZLk6o8wGlJkk6YrYWhKRYP/VFoj+P3+EELQqoX3iM69+HKF7jzgx0eWGvfGITWepSqh9y9lmA7E39IgooBK0PpBJwdn8iFY6kIlKJFITyVKf6y6N7v35MeFjIHTteehoTnARG6FtO0plmI4NKUsYAdO9C1ze3iIGR9uteXz6hG8fPSalQAigTL/TXSjD0PQpxSqEHhbpI1lRETJB3Xha1ZLnOdYlqqAwVYYKmp7juaHb1HRnZ8TQsWyW/OXhMet6jZKOput4dHTE0XpFDJq81Ny5+waLhaBpLaIo2d8aUo3GbIhIG5DSMJlsc2E04/aFG1y7eIPd/af51Muf56Mf+nGefepVlOx3/310CCHwyeFcR/KBFOM57toRk0OaPphFhQZZDHrqjUkImUAoXNshs4wYHTY5FvMzsukWy/e+z/HhGVEKZFaQlwWHh0/oNnNijyFBCImMMBrtUk52qQrD7OIVjJJsj3JGu0NeuHmJvFCUwwHTsiS2LW69ZtGc9EtgQEge6za0zZr54SFbs93eZmsFUFBISbSOrj4jeIspcoTKQRtUPuIsdkSjsMkDPRAwSYENlq7tPStBRMpSkFea5CUuRULyDKRCa0mMCVe3NEmSkqTzvaakCVgstQ29kSt4BlVBYUZI21DoHDOaIIMik4o8z8mzAhf6NGlxTqvOZEKXEjMICD2kGg7IZwUIi9CGpA2td3jraJoN683qfc/fB6ISWLSWcZdjrtzm5M0/QQRHk1L/ASzG7OyO+fbrd/j0C9dp2kO+/dV3KLIcN2op9gxXr+/w21/5Gr/80sdY/94foHF4FPtXp7zxl30cU1SR7YmmDZbYRfLKgoic1XB2/yFPUmLbjJFScvCoQWyXvHh1yoUPPc/FnW2W8xXHb3yNJwdnbIYtQUpC7ZhlElsqSinY35mRCsFDOkazitXSoy9kVCJw9GSFOoMwCkx8x6PYocuC3V3NyetHjLZ3iCcbtjDcfPY27x3fB2F58M6Kt2zBweGGf+PVl9FrAcny6k/+JKFd8/u//X/wuX/tC/zzP/ot5mvL1dmYC4s5L330Y2gHP/sr/yGj0QSZ5ZR5QYweIXNS7BApkUQiut5FJ6TGentebua9sy9ZQucweYnVkcmNH8HaNcV0m6Qk68eP6LqOyYXr5LMt8qKPcB9fv4FbL9m59Qynb71OdfUGJ69/hWyQo6Kk2ax5/K1v0D26x+6v/ifEZPDthqbd9PzGboMZ79LWxzw5fEyWC25f6G24zz5zHeECg1FFpxQ2RMoINknqzQqh+/ASg+fWjX2On7qKqxf8k3/6j/nUi0+z/+pPMNi5AkkyGV9mVT9iniyPN2u+dfgW12/ukeIO2abDS08eO5qNY3er4JHR7O1nlLkjG9dUueTkODLMBap0bBUnNI3gtbsziuk2eE8XAg2GrCyxQePcmlGhSVqx6jYcPXmMqQz5dk7yERUcq3ZJd5Cxe+ka5WgP1xyzrjc0zSmFzLGdo151rF1JKwdkZeTGpZw3bE7yMN7fw803ZFJRL1ZsFR9wApELkXER+Lmf+DzfvHcPbCL3gkhEyUTyDVmRQQ62aTHnNuNMa3ResZUV3H9yzHi8BQqkB6wn6YytvVH/hosNS9+LR5kwJJkotEIJgSSQXEPdBM6iY2taslgt2bSJ7eGAzWZDc3jM6dmGebL9myH2yTKl0SgEpshx3pN8RxMc+jznfZyX1A5S3VGYDCVy5l3NOBMIHRExsjscMZsUSBSzcsKF/a1edJKJTnvWDoQs+PrRCavgKcsBA1PynXffwAhNEIKOwCsf+RhP33qRL/zoj/GRFz7BR175FLu7uwyHFbnJIKlzrFVPTEo/QFrHvlR0roPzVegYI8FbSAlTVgihkXicc5AkUmiUSBTFAEUCBd46Uuswwy2EHpBN9pAqQ2iB36ywPiIsjK73pp0QHVpKbNcQZE/k6VqH7WpUMWWzPGK9sAyUoJIZhRTkw4x61SCHY/LBiImUSJ33qO4YCVH2Fcy5eKdVRpZrykGJVBlt3eGcx7U1Wmc9EShJVFQc5lC3Hq0UQQh0jOACVy/skhmFyE3vA9AZyN5tKTKIMoFJZDqHPJGX/TandQnr+zVjH8GlfqvQaIE0Ci80ySmsa8HZfqRpW3COFALBtYS2b1lDlEQfkN4jYqBQihaJclBIMFlAqgitIMVelM6SwCYDSIT+gHMH9icz/oNf+imkfcgfqH5urPtFXSZZycODQ6azIb5u+NLXvsNSWGLjkGieyScsDzsOn3ydfPRLlARW1uLnS1anDT/0sVt8+927+CajqqGrA/lupMlLUtmgTiBpmBbw6PGGyaU9iqbgI7MpW1dmrB4c8ODuY5aLFaEc0AzgufEWHx7v8PbhGWw1bA0HXJwOubdc8t7iBFcptosxJ6VjfyvnrbcPyaRg50PXmHpNqQIdHTYmmi7ywtU9vrtZc0nC7jPP8MyN5/jSN79MKkt2Lu+yGUneM5Jn9VX+6C9f5xe/+NN8/3DOrSuf4FOv/DJew79180MoafA+cl1KREygFDF1xBCRKfY+eZEAfx7NZfHOYbIMKXQfH+Y9JPrcfVMipMbbmnA+PhNa4VKiGG3TdivMlmRSZUgBaTikWa4wpSEWQ1y7YH34DnpyEboVksjWtSscvvEGIlOkCEdnp6yOHpFXE0ymkQpklKzOnmCjJB8UbF++yvHBA3aLHf78q9+jXh1yo92QX5+gix3e+d73KEcjRqMRp0eHaJ+4cusaSo5Q+QA9LKmP5gyHY/50PGT0f/0+n7n5NIvxht2h4ve++TWGt29xcjDHbVqmSnE/RLSTtIuaT//K5xh86c+IqSM7OCVFwUBnOBPIoqAoJOtlxMYMpQMDmRiWlsPFnGQEozKnQBM9mOQYzkpMNKyhu2zxAAAgAElEQVS1xvuI7Dyth301odOCfGwQItD5jscn9ylFxXL+CF8vGKoMiFBpCjvl7btr8uGKyWTDk6WgSzNmZcFqvsEMZpTNEj8YsXh/SeCDcQlEIfnYRz7Bv/df/6f42EMlkpYImVGVhocHB+xMd3lwcMT9+ydM9zPW1qFsx5ubh/i8oNwa8f2TUz554yKPO8vRZsMwWURmGO/OSGeeOS2DAnzQ6BBIpsCaDiEkk2xAmWuG4xHXb2wzTR1rKXjw2ve5t14R9ZhnLl9hnHl+6sbTfO3B2/zFwT1uxIzSSE6WK9Yx0FkQqcWMhkx1RhCSTGn2L++wNxyiApgscv/Y4qXAGM1rJyesFzWPYuTmCy+gMo2NkZ1qi49cucnuaMz+aMyLT73Kz31eIZQBRB9zjidLEFPPUVAiQLSECDIKpOxz+BCC1MyROiMI3bvqSOg8o2tqtClJdoXKxoTYtwZdWyOEwC6WhODQozG23pCwdLZGmIzoLTIf4oJFrNZEITl58C7VuCbhSEkidMCfrOiSo15tEJlCiUQ1nrJz6QKPv/tldp9+BfKcPCtJUuIT5Cajs56Dd9+gWy85Wsx598EDnn/qAoMykJaO1dk9pApsDoDLl4m2RWcKJZ8B2XP8TDakKKYcyBpzuKKer/lvfut3GI4VLzz/YU6EZCQSYbGkzIaM1DZ2vaJrWoaq6N/Qg5z5oyPGe6DFnPl6l+EMYnScLCWuS+zmgflCUAuHixmTQY4QgiYaQhag8zglGOVDQoRRsHTFAD3q4+tlLpBeI6RnXXdoLfoIdhqU90ifUVQVbbcCn0jCcHhmKK3h8bxgOtllXCQaDE1oics1QUC1NSTW/n3P3weiHfDeYsOaW5M9EIFAQGnQsWfAr2qL0gFf92YRIySVEcQuUK+WpOiZasNX7j3gx599uefWBU9mJKdRkhcDfCHRQoOSSJ9wNvVjHaHRSaFNLyjV3nJxWmHyErepOQmW3GQUWeLGJGO8tY2LmnVbgwtsgmfRdDxc9Y6sjXPgAlYIyrIkGxbsb1eUuSC1ER8SlojsImGzZhg8x8slLvXbYlWRQYjkIeepvX2MUEzKba5feoEyL0BKUoLoG2KwPetAnEMvU0v0geT7nIG+pP8BujqANufBJ33JL3Wfxyek7lkM+ZDgPYSIXa8hSZA9x3Ew26WrN4gYSFHiVIZtOxAOVZbIGJBFSVwvsOsNMdnz8j9RZDmjp56nizm+zxOFIBAejh49IrRruvUTUmgJrqNr1tjNCt82LI8eU6/mtPWS0ig+cmOKzgyqGBB94IVP/RjD8S5buztkRY4uqnOgJ+D68JDhdBvvGlwZSZsVRhqkFuRB87V37qEHGcHL3kJtNKUa4+uE7wJ5lVO3G3TssL5hUCRaFQiux87ZKLHxPKMgJQKJjc96SrHoqcKGQIqa4DzGRayDZFuCSNhNR5ZlTKoc31laEXCdxYuAkZpKGYwXCJn34iOWDEUIPQU5lxk6GZIv0bpAaciUZDIYAh5vA4VQZFn1vufvA1EJhJh47c5X+fkv/Ov8zaMT/qP/5TdRRUZQCttEmm7OMLuKMZ7YSuZNYDqucE2NnwNjhcgS/8N//1/xz/7Lv4/7ra+iUs2mdRx8Z8nTt69ycSejOtMMZeLxYkVr4fJ4xLRoOFtbll3ghQ89x62bN3lrveHD0qB9wF++xIOmo5psMx5sIfLAP37323zm2lP88NWb/PO/eoO5ENzIDTsXp1xZr6hdjcygTYnVcsOpb1ALRyBxtppDTMiFZXJ5l3ceHNItPFk55ekXbvPCtWeZDab8w7/zG0gJqIKYAhJJcBukEKQoiBESNTJpiD28xLauD9VQBqkMMbjzS6LXAgQGoWSf4hsiKTjwFplpEILoA8E3dPMV+XRCu1n3byCtOD17go6JaEpSjCSfcLJAiIKwOEBnBdl4hxAdWyngVR9Nbvb2OPved3jt4REpSHKj8cogQ4c1CrqO777xFrfUAPnW62ztXsLhWR+dsNmcEpyl0pJQjeiipdiaIo1m99oF5octB3fvsnv7BdKm5ejR2xTlCC8l67MnDCf7xOB4+pmPcOqWzL79Pd46POPpC9tURjIsCr62POLl3U9xeHbIqGsQkzFZscOjg69gx5LnnrnGd77/TXbihijgHet5dL/ms89Kmi7HycA6CCZqwOnaMtsuILRc3ut4cJSovSOVFamzyODQgyG+6Xh87yGoiPWO0d4uNibqbkWZVbTWE1OkcR3yB4E36zXz+YLpeAQmY7FYM94v8QmcDxg9YLnQ/Z4Nto9Uz3KGqmPTNgyHO+97/j4Ql0AicWl7xJ3Hf0w1/SL/zsde5ffeuIOWgqPFAVnsKLOEyzRawsxKRAXjUrERfS69MJq9YkPtBa4LlCkx944qGI4O52xfHlPmkS4mFu2cIEtWPqOsMlyruHD5Ateu7BLChtPjQ7rtyzyOhk8+e4M6eS4+9TJP7r3JQzXn9niP0dYNXrn0HNnORe6sHmCbwOPFgo1StM4QbSKJyPHdExa+Yf/SRULK0AJscFy5dIFVEdmzE575yHO89PRTfO4TX0BIhRSCkPpYqhQdMgVkD7EmxkQi9N9jIslEkr1DzhhFcJJgW4QJ/RhVyX4iIBRAn4asBLbe9NWDlsik8bbDr9ZAT2RybUdRDFl3K4TW5DIndBtUiriuJggBKRE8tH6AXa64MPQU5QAuVdTLFTRzJIqHDx5Qzx35eIA9R5rnwxG23ZBriTEZr33zr6h8h35FknAgJIPhkBRj79tP4LucTmdIqcmQTCaCxnpGZ2ds6lOqakhZloSUaA4fYJLoQaPDkp1ql/ttyzBTSCkodYbKBGMkWki+9/Z9PpsZxtUY5xzz0xNmkwEXt7ZoFytGOyX+SHByWOObjhzNm4cLRFYxMDnJeo5PLNXYUChDkbVc3l/z6HSE1xopFevkkSSqAJpAkImtImNRz8l1ToqC5WaNNgqRSnIitj1lSSSExKAo6KzAdh22FaAyYufJjIGQ8HYOVUmeAo2K5FLjNoqu65DyA04gKo2iDjl74xG/+9Wv8Kuf+hxfefwQrftE3JgrIhKbeTyKosw5E45KB3ymiBqQimFW8GAxxzcBYz0qSopBTuM6NIkQFC6u8d6jQ4I6oocVV6/scvXKJTbMOZ2fYbRkNBpS+I4wGFFlhlkx4slQMFBDtPVsDbbQwxE+SVwL7y42HB/OadqI0VBoCEazqAN5VlAIQVi2DJNmkVq2tkdcrwZ8+KMf4dMvf5KyGiKk7Ed2BFKCEHq7sRDy/OD7cyus6vHUscTHXgUXQhC6lhT6EhQgiogSPcIrISEKYkp/PQ2QukBKgbNt753wFpnnGJXhk6RzjhTA2w6d+lShpl1j26ZvLVLE2bYXF/OypwKXE9rVKSJuqA/P8Ntw6hXImtQGfJ7wUiJ035ZZICsy7OEhw2GFNBXR1cjkwFQ0viUvhoQQKEpFngJd15FXU4qtXfzyjJRpVKc4j51g5/JN3OoxzfoIM96mmuywvXuRkBS723v41JGpiBI9vruxlvnRnOLKNnmm2MQW6oaiNBiT43LHsNynCkd03QlRKpyHkzOHGSX2S9nvAND//1hrcTKhZKBUjnVMdKmjQNKtHdJEJuOSRdNgqgKIFKGjKiccbxaYCFpFovcIlVMhWFUdpRbny0aeVms6r4giYlRO3baIvKC0kpSX5GFNYy2t7WhFxHfd+56/D8QloLKS/+6f/jH/8U+/zFXzx/zb/+PX+dmPf5SHh4ccuw5RVMTSMPA5JMdmbVHrgJ9KZvuSjv6DXkzG/P3/9X/nlSLjK82SunHs39jn6rbhdNkg9hVP3tsgtGFkwMaCz73wSXRe4GTk0Te+gwsNn/zQJ3j26RcxywOOVcMP7T7Pe8tHZEXFBQ/f+OY3KF6eIn1GCDnf/Jdv8Gj1gMFsAkXJcnXCwktKo5mVBReuXuWZ/SG/8Km/ybWrtzD5ACnN+WENgEQEi4iORD++00KQJP0egJCAJNIbcqRIpKhQWqHjuZ899JWD1P1cRSSJSj30RMpzBHoKiHMZKKWE7zZ9H+vaPjWpKnA24XyH3TQkASk4dK6RImNzckiSgrwYsKnXEBMpJCSC1jY8fOMNwjmApB9BZixfe5dMKmJe9aPalaUcjdhsaozJGZiMaAqq0jBfrrj35h02i1NaH1BaEJPi4tWrhBjZuXwB7yLGjOhqi3CAqDCmYHzpKZKF4FZsFqfs3HiZ8XiK35xi25qt2Q5SwXqzYXtWMTIZb1iLT44v/dVd2qOHVDf3KKqSJ81jDg9OuXDpMlsq4+ZTNzlaRVYpcLAIPDcdkzDofIKZDqg3NXO7ohhG1vMNo1xwECW3Jycch5oUMqSHZEAk2GyWXL52jfjkPs57hqrAhpbOCGazGcE6lssNHsH+ZJu2sUzyASLB8ZNHWHrQyL0npxTWwbYiqxQmFyyOH8GoohxMsKsHGJNBzEjuA34JHB4/5h/88t/hT770x9y+OiX7q/v84V99hUu7F/udaJEo0ZRZgSSgk2JNIm9BbUd8E4lSsGk77t/9Fr/2t36MP/z2P8F5TVg6qv2SMLC0rWdnq+BJTESdcSHPeOvJA2KRszWyLDeRwXjMphBYJckzwXa+y0lzSn18n7vvPGY3H3K8rFFv32FxPOfC1haXtw0Pa0XRJg6On3BlXJKqjBsXbvJzv/RZblx9mqKaIqXslWZinyAbExLQMhJTIEkNKFIIJOiVddH/3eATxvRrqTGGntAb+kiuFM79D0VFiL2jzNoaLRXRt3SdhRRACpC9ftCnKEGMtj+0QmKyCqUdzfIUYQzERFEUtK1lUx9gdAZC07Rrku9IWIbDHU4ODshyg02JFPrLJ6Wcpl1hdIbNHDpFlvNTtoqcGD1ps0JtF3TOgusvJlMY5sePcc4RUXgbKcuS1dkZWmTM9SnleIs2WXKpiSfHjHe20ALq40O0KTBEgvfopPGdRQ5GAMTk2b5yi0d3v4tKFTEzrOyK1EjuHd1hK+/bm73xjK8/ecBQw/PbW+RlQe0ET+YnbLxlWipm4wKbHDEDnUkWZw1G1IyMRsRELgeMM08NGNHSNpZoO4rZlCzTCD3DtWu6usG1jlYXSCnJh1CUOakY4JslZWEYjqdEOcetOpoY2XSOUiq8F6TVmmW7IWUrZlygjJHGWeK8bwUGo21qK2lOzzCD958BfCCmA2UWuf3Dn0EYTZnlmCSIMSJTjpGBIiiUUWjVu56cSKRo0NJQKkFpMgSRxkLWtWxP9wjn7PrS5Fgn+s3A5CmKjMJIkoKm9Zw0Z5i0BLdh6+KUCzvbeFcDkkJV7JoRD48fsD5eYFc12nmscywWa85WaxbBsZVlGKlQQnNxuMsXP/t5/t3P/yy/9NnP8eHnfoRycBEtzxX66CFJREr9l+g5fFKZ3uQRHCl6kv9BDychyfOSPoCISAlSm17hlxKdFVTDIa5rIERECmSm7HWD4BCyD9eIPvQXkexbjEAiOIvWsncIBodtaqK1iNSnKocg6OoFyhg854nMgp7jIHKccyjT72VoU/YW8K7DCHCrmqzIGRYlCoEOEFVOsJ4gIXkPEaSKCFRvn06ScZYRgkMIRQqx31LUEt92GK0IEUw5IMSIJVIMKrJqiEog0MSm6wXQ6EhBEkMfE379qVsAtN4jQm8d90mhbI2WgmhBFxWrZUumBaOyQtK3T65pWXpPpg1aCFyKoA0xgYkwNJJh0U+uQkgok5OSoMygtQ2bziF8JImI0KmnGCcJQpBJQV5WDMoK13m0EhRaMR11ENp+cqEiBI+PASEDGYlsOu1BOxPJqOrQRqLLHOsCHb2JzAWP9/3v9X7PB6ISEHLIndde4yd+/tfp3vsjfuy2R6YSV414chTRraXMRrTRkOcGkcBoSSsFWim2RgEXE7mrubhVcXHrEoPJhK0qZ2YU774zJ7tYMKwmFDuOv3HzWf7P199kkXleurnHxdkEGyK7lyTB5Lzz6C5n6zUTVfHmV77Kwcl7XBwMuXd0xjO7Oxysjvju0WP2BsfMFoe8sLPFL199hp/5G7+A0RGZNOK8TycFRFwRk0T+oFdPlugTUkJKAs8aRQHJQ+wlwHD+xg62xkdQyqBljkiBlBTQr+EmH0hKklIgkxIfPZECQoAQCSmgTUUUoAX9z3tHiKBkjij7gx6aFevVCXlRIbOMkDzJeZCCvCxJPtK6Fl1OsH5JCImExK0W5MUYpaFuW7TJmW7tUc9PGO9s4duGLkbyQUXpmt5s5CwyZSgE1keyPFGZkqg9ZxuLGo0pNhuUSMgsx+QZAUEEdDkgbztUXjG5dA1Cr4mMZzv0O8EteqDxdkFWjhEhIvKC4BtuPf00X/4XmsW84Z3NhiAiLzz/Kn/+lb9gv9rmLHqOQsc3vn+PH33+RZa1IPoVg5FiMz/GRselaY6UkdW6Rg6HLM+WDOdrBiMHGnLT0lrL47Mp1/Y1k6lieJIR8hy0Zn9rzLKZI+WUTEZirplNxpSjCbkx3Hn4LoOzxIevbZC5xLnXCEtPGn+EtVxiBoIyQTkeMB3MuDYuMeUZy6Xl1J6yffFpzOmCtKlZCUOpAtnIsAn/j6HfwAfkEhjkA447wUuVonnhC3zh/hm/+e27fOLGbWbB88/u3SEThmgco+kEuQzEor9xVw1UY08pNflwiK07Lt54iYs7U3a3thkMC4RUpEySnGWVJN8Inleff4Z15cnHE7qk6ASImGiXa8zK8ebhCZ+6dIU7b7/FD13c5s7ijHebh/zZXfjJa9epiwHXrz9FUpofev6HqcoxRgBBIrUk0aOvlOirGqVEL9opDVEgdX8IE32SsqdFYhAqkWIghAhtQxQJbQq0kaQYkSpHiEQ4B55GGXqAZ0x9eq0Goid6i5ACLQtCcKQYgdjPz5MkRY+PHdF6OmcxxjCa7GCtR8SW2EZ8alkfnlEOx9TNhnI07rcChzPSag5J4lVGVInYdeTGEJOgSy1CKELXIjLdW4OD6/8tnEPjEYOCqBWqswxMhi81qXHkJkMoQ1SRUZnTxQbh+2RjmStUmZPmER1bzHBApjVJpl43sRafQHSBKHRf+RiJlpKEIfiOvCxwds1xu+b6rWuMx2OQiVmeMRsNeOP+Ha5UFYox69Wcs5i4RsYqrPGiQ1BgpGHuAmGzpjtZcmGWse46bKu49VS/NGZ91zs4g2VYDGhFpFA90LVezIkFWJHhZWCzamhWNeV0l2Fe4UOD1h21mKHlnKqUNEWGsyUiZFgRmFVD1l3Nk7ohnQVOF5YkAsMdi/WBTdNQiBo/HDEohv1L4X2eD8QlEFPixQtbzB++xW984z3+s1/8Ff7e7/8qW2XBuydD/osv/hTvashlyWx/m8XqhCgtqol0wdNUkqKStOvEcbPCBc3tl66SDRRCAG6CjYnYWnaLAVFrIpoX8y3eC2sOF0vscsXx0Yqjszn//qc/w1+c3ENdvsrHbz/Lv3zv+7y7fsJnnv0hXrn5Cq88/SLD6QwZ+7chgJB9tLdIkma96IW2rCBlJUb1Dj0hJYkIEpSQxND1f05EKolI/fxfSUkmBFEKnOtNKN7FfsyXPN52pOjQqiflgEAaSfKehOyThEPA6D46WwpBFKkfz5H6gxYT1npQEp00MUba9RkqrzDZgMXxHRSKcrJPUy/IBkNk7Gk+ybXovOLswSPy6YjYdpSTQR9kIaHMKxgF2m7F+mxBlmcMNNimICaLykZ0doVJOV0ClAJk/9b3AWlyJtNdrAelDT4lqrJgMpoyLCvSaEJeDim2tsEFih+EsdgzykFGNtonWIcwBdbWSJ1jTE5KnuFkyKqR7Ny4wbMvvIo6mzPMDJVWTLYm3H9wynMXt/jT732Lly5c5+7ZBtG9w53DI5bSMdxVVFFhTMb88ZKjg1Oeu3wbZQKn65qV3SaTCYHFhsTmpOXe0SGZUcQwZ+ZKYr1g4xOHR6dUkwHGNGQK6sURymiMGvOX3z/l1Q+veXQUWJ8NOfH38H7J1taEWTUlBstTN2/QNZbaJWT9PYwsSH5DkWv8MpHnitWqptGaKD/gyUIxRWKUKDPh4Z//ARvn+YWPvYQLifj4DbYv30T3i8NoI4nSE7wijwJjoXOJGaC1pgOa5NBlPxbbOI8QmjIqyipHKUWnoc0li9jweLHk/tEp7z06Zr6cI33Lxkee2ZphbY2NPdZLJYgqYzYeIlUfiRak73tt2QOyerON7UWeokLl1V+7+Xqbbq/PidRjtn2wxOCQ56W/lLJfFgmBEBLB9Su9P3hS9H1SrtE96guIGHyU2M7jm64v4X0gy7I+mfkHP+sDuICWCu/7UVaWZYjUO9+UNpiyIkqNc448L0mZQUtBOz8h+kQSkhQESkiaZoMoMkye9avRXSIGgUgQlMCUBcpkKAkYhTxfsqp0hpb92zIK+kx93bdOuVHooux/P5VB8Ki8+Gvtw+T9YQ8pIsqMTIqeOViMiKqgnG7hW0sMG0w+6CuxH7RlUuOl4nTVUGnB9sX9vu1qGjQKKTXWd5S+H8NBg3cBlWd03rK0ntY7nIeYS7QxNM5ThMjD4zXLdUe9cZyeWU7ngScHNd0qMN842uWSpmmorWNUjUjZGJEN+3697ehaT5cSrU9gHTE4bNCsFpHlacd61WA3S6osZ1iNIcvQOifXFbYoGJQVg0GFLjKE7iPpcqGw1hOCZbPZINz724Y/EJVACpFv/Nkfsnt5m//27/5dfue3/yd++Vd+nW//yT/if/7Wu/zn/9uX+ckv3sCYXiCzmUT52NtsdYZsBLu5Juk1169tsVmsiTIgu/4D5p0jAEiDLAyf2X2KZr3g9bv3EKcL7t9/Ex0j84VhNhL8o9ff5Nd/5FOMhhWP1yec1i0hws9+9HNsTbcZj7aIUpCREZM//0Cf0218AAxJ9WIlPuBjBHrIhRSBGM4P9HmkdBIRt14jtSYJhUiB4PsKoRpMCKmnGkfZawTE1Cvfqu+TlcpAJmSuSaEX/zqfUMoQfQOpDwmSWmHrNURB3a6QQpCVFUkoFscPST5QDGcEAjIfUCIQMpJt7SG8p7YrhpML+EZBWDCaDAEBRY5qakTWtyopeJrOE5uWshpBTLTeYZQgigzjfG9jjZaUCXCJIldIkWOCw3drREqY8YDRcIqSknw0AKlJOsfkFcPB/83cmwbbmp31fb81vOud9nSme+7cw+1Bak2tlloiQqUglQEjQEAc5kDiSgUXgZC48sEZqmKqEueDA3GVHcdlnBAXocDGGFM2BgsLJCHAojUgtVpqdaun23c+095nD++0pnxYh44qhVAK+0O/n+557777nNrnrvU+63n+/99/m+AS30AhKUczun6TfBM2NTjNaJQ2ZTFOTWIzwkWNtx2XLj/My198hXx+m1FWUGnDvc0pXpR85sUXMUVkDYxnFd3xBiUtaoj0fYvQOeiAWi+pxorb1+9xaAReKlxzggoK5R2nqxV9F7h7dwU649KVK/Q2EmMFg2C6e55Nv0a7BvQO23mJtRbXB5qV4rkbimE9SZ8LgaqeUO2dR0RB03SsiYTW0oj0AOyGDB08PkT66CmrjNXRhk2z4HTzOu8JrNqe5a3rHJzc5f43vJ0f/O4f5nc+9ms8+si7ePBNL8B2wbvOXePzyyOqeko2OYXlGqsCJsvwUZHtjnlop+aV4wWvnt4mHFqoFX1wWC3QWvN9+0/w20//If/o+lNI71k2ge/Z2mPUHfOpo0OO5iuWaC70DXnoePXwVd76yKNceeDNzKb7XLvyCFIFovcMmy5p+MqMTJk0h7fDGWaqJA4Ng/cURQUyQS/TXhAIMaDzDBEjtlniVzGBOGyP0DLxB6sRMUba5RGmKAhK4/sGgUaXBt9vcINAl2OESPZS1ydSjXMDQmb0NOR5jo8BkPTrOUKnjrFSGVpLZAx0XY9UGcqM6LsleTkjqgQ8lUNSuzmtKeWYYTMn6IyirgBNtzrGFBcQ9ZgiBobBE3yHiB5hNJkwhMGS1TXeWnrXo9CIkJNJgW/TNKSskiBIhYgWBTbCZDqjbTZM986TqRxZlUidJWGTbamn54lCMb/1HPV4giwKApK8zBlPtnCxo7eSOhNY69Fa8vjjT3DnzrN8/o9fYvnM59m/MKHqIuO8xNuSjz37PJlYIcYlGx/YUoIvny6oyxITI/OTNeRjOjlgXYM2FTvDQAiGYqdg/0rFdu65fRhYtlDkhp2dhsPFglu3bvLcS3fo7cDe3h6zOkMGRz7NGPnAqvcEYHADulT0G4GpK3Ktia5HK4NtB4pyzHTLEJ0kWsvdmzeoZgWT6RZeFnT3jjCFpVclzmSIBsLp+muuv9fFccA6SzWZsrx1m2W7oZOCq/szuvkJf/XHfpIHLl1CyylGKvYmI2ZKM0RLFOCcY/viPlmZo8qaiczAB1pj6YVHKYH2grDpEWLDaXOCcAPCBDIcT50seGi6y25uEDoSXOTJ/Uu0Q8OrR8dMts4RhOOh+x86OwZIunaDjKRutCxTeSpSYy5oA65LwNHc4H1SuCWtvksjOanwa4ft+zQ1kBKIBD8wNBtsP4CzSB/J6ym6HKWEn7PSObgzC3CWIaVMR40gzlKFIiozZ6SeFPWlgiQ6i1LirETOMEWJNvnZMWJFsA3BrslMgRQBnCNGTzusk78gRIZ+dWaSSWz+6P/EDQTSx9Sf8GmioJRKvRBvk6ox0yij0DrRnTKVpSRjrZBKoUSGjgrlU98hzww6N1RVhc41UZ0FvfoENg3OJ/UkksnWLoGIkJoMjSKjHxqKfERmJMFFlNIgAtPplHxWs3r1NqNcw5CmKLlLFWnfdgwuTS2kC+RSYHHkhWBUZ0iXGmxKB2KmKGRAkiN0xOSSyVbFaJZT1DlZMcKMKs5dvMK5i5egbxHDGkRLszlhdbKk95Gu6Tg6OmJ+eEhmFNIPlHmODZG+7+nsgKpyVGHYbDY0y1PoZer3uIgILWOdY7qRQMEAACAASURBVIoxSoLJBSKmo5NyjmoypX+9jwjXbccXDk7YnxX8wYd/i+/43u/n2hufZHnac7QZ+MFv+SDPvvoSb5/ez8Glm/jnBaXPIQr2H73I/VdmdE3P8WbDlf1LtETGVckSQbPw3H7pAF8KeKRg8+I9DscVYjIlyyS3+pZ74z2+69F38sevfJxr2/uQGX7ni59jzYQPPPFNfNN9b0PFLCXo9h2mqNFZjpAZHkfUSZevs5xou/Qf1jtsn/oDhUmx0M62KUY7BGw3R5Uj2s0JMSp0ZiimE3IgCpXm0GEgMyNsPxDtgJaSMCT9QDHdSrN5o8+kvQNSn6XraI0qpkgC3eoYZ3syU4HOyUSBDxE79HSbE5RQFKNdNs0aU2ZsTpdYY+mbFcrUeNugsxrvB4zMOD485vyVh1nHVaLgoomuJQLWK5SyDH3KfkRrmvUaYSQGTVFXydk42LT4tcGYAikDUQhUCQVnjkSTgRHMql1QiryukNKgtKCxLaWPYHtEXmGm53Gn94g24lWgb5dUswmr5Ql1PUZlmsEOaOXZ3t7lnjvP4vSYIVpuDz1VlCz8krgWbFzD4AM7doYtGkyjsC4wHlWo6Ra3T57DdJ5hq+Lc5RnbWlBsKXoZsVowqg2TIvLwxYwv35kxnC7Zv3qe8+fP87zdcPdgTh0UwXcEerql5eZBh46C6bTE64hbrdnd20aRAlzKsmRUFczqGTe7A46XLYv1ht466Ncslwv09g7jPqHr6npCtzkl9CsG1zGtcqbydQ4VubI75eqVi0y3JrQvP8fvfvQjPPiWd/OGy+cpF3f5/Cf+IY+883vp+g2tUpjWMd7d4v6LFyj2FUebhjDfYEYFm8xxY3Wbrb7guVdeZb0MXJ1toTL4n3/xn/MzP/4T/ODf/Fmqbs3WVkEz5Hz0jmP9yCV+5sf+Mrv1hOvhJu8T72W89yh1uYMKkhAseVkhznTszjkQqXOv8jxBOZwluEg40/OX5Zi+3+BcsvyGGBDW46InH4+xvaOa7SDIiNFjN6tkN80LokpSaNuuEVqB0oShQZ5ZhK1zKKUJ0Z75CSDTJoVaBo8KPYN3mGqC8gM4jzQm2VWbI0xeU9U7DC5xBbPK0C5XoEATOV0u0VmPdZayEkhT0tx5me2L92FDQ15OWM8P0WdCIJEZcJ7BCYKL5HmB6zumsx2iFDQnx2SmoNCBqD1ZYcEHNq2lGo+IUqBNhhRrQKLLDCNrZHVmDY6Ccjyi6SxVOUP4gNIZOoDKM0bTczi7wa0bimJEu15RlGP6wZF7C1Hgg0bGQNhk1H4g+J4bG8+Tu2NqI7m5bIhOIbanbBW7DIfXWZoFmalpzRiWntpLvIw0TUDszrAE8iayv21SXP3QsJIeGVsmRQGhp6pnHBwd8Mi1B9nat8m6nY84vvcSqt3g7q4JJmN+uOTmzU8jIhRVzva5GRe2ziNVxrppOT5ac/HBh8iyjNObt+nlgJUdrqo4eP4F9H2RkyhpV6d0zZLV6hiGjqOj1I/6WtfrYhM4WPdMZzs42WB3dtgb7/PRj/0m2x/6fnbH24y3rnBvfZu9ch8Gx3RUUhUjZAmHyw7rA8oYZmVO5y3LrkUNHcNqw2hrwmyUs2g94k/SaAeffPe2p8omfOdbHufCuSmP7j7Asl1RuV2y/CLL3lFXjsGlMRuyQkZouxVSFEiVtPxCiKTmwxOiw2QjQnQMQ0fw6UhiTEHbdMgsQwpFCB4QDEOPUh4JxCAxOsMNLTKv8Wfd7eggyqT6j2cqQ3nW+ldaEHWG8mcGI5fm8X3fY4NDMyT0lnX4rmNoVsQ4EKhQMpXn1lq69SFG58SYsV6u6XuL1IoqK+iallFVIsoCKTQ+dPSbJSIzSFMQvUUWmhATq6Ac1wz9BvyAykqCCIikjII8uRYLkrstU45Ma9L5SiClBp2RSYXQAmMMEoGXMVVfg8MYg1IqTTiUwrmAQKRRYF6gZJrIOOcoCs0wDAQUJmRYnxq6PheEPmUBCyIjqSiEQmrFVl6gowIiWRQMUlISWbUthAytBLa12JGmV57BN4hGYzKFzATD4FG9w8vAuLYMzrJZrRnVgqKuqOsxWzu7ZO2cJhdM1hu0MWhp6BaR0AysuwGOlygUmIJRqWm7FVfLEpXV2PKE0SDZaMsoDLjMUNc1DIHGW7rG0m4anD+L7yv+LZDjQoifB74DOIgxvvns3k8D/xlwePay/y7G+Jtnf/ffAv8pyRnzUzHGD3+979GuV0xyxYIZ7/3G9/DsM58izA/4x7/4P/Gd3/aTPPzOb+WTn/4odnyTH3/Xt/Pzm5bPvvwK12/f5OjOMTsPvZF3PLbN7YNjvviJL/K+9+YU7cC52TbrCl44WVAqx1al+Vu/9s/4jg88yQ888W6KeoSVFc/dvMkrRyt2ucPh4gQvSjZbpynXXpdkWiFkTrc6QUrSuVnVOOHRCGIMZ4ASgc5KnO0JISAYkBKsDfh2Q0CBa1MHPyuQWkHnaNcLMqURQmEJ+BjJlcS5M30/nhAkRqVy3w0OpKIoxjhrkXhQqTegYsC71CHOlMINlkwbNr5BRJDaJOHQ4Oj9Ch9Tc9CUE8J6g7UbfL9h7+JF+q5jaFsmO/u0pwsEOsFHXSQrcnS1zbBekJVVqnS6gaye0B4fIzQoXYGRaDJGu/v06wXGaOSZ3gFpyHSJLNKYER9ovAMC5WgbaUqkCuiqJNM5upww6noGrcirLYQySZEZBTIMxOCQwhDPJjEmLxmCw7cdWhuW/ZqTxSmEjp3piGdvH3IuV4zG21Qmp8fx9vuvcmV3j4PDY/JiTCMiUWYsF5b5rUO2LmiyKLh9fEz0NWp3ygXVYNdLWmlYdDVZDjU51aQlzzXd6dMcH55g+n0isH/5Mrtbu9wpKmwRueYlweS8dWebkNdsrCBTgqc/8QecHB+xPDimGm3h6Cm3LzLb2qNUOUvX4lcLpHRcfvA+Yl5gbYsLgs36hMXpirUTtE3Hw++6+uffBIB/CPxvwC/8f+7/rRjjz3z1DSHEY8APAG8CLgIfEUI8EmP82nIlQCvBpWtv5QGt+ae/9X/xwEPvIJ+O+Nytm9z+lb/LT/zH/yVvf8O7+cinfp2++SQ/9K5v483Fh/mXn/oSr84dO0h8FNw7XjDYhsPjlkc1LIsMtAM9sFiu+dbtc/zdP36K3/xf/ld+5zNP8dy9I54/fh6O4bGLewyTJ9jb2qKLU5CautqiKCpicAjhyc6eQNYHXGhfM/QYbYh+gOBfm/cH3xLRuG6DbzvQMunvfUQpie/WeGOwqyVFWdPbDkkgLyuy3BCcI89z2mWDzhWZkgxNS5YbpMmRUWG7DWiF1DnReezQIhEgNFpJogz03ZLOeaTJse0KU03A17jQInVGHCw+r3AnR/Rdm2zGWcHx0SFSFUx2z9OtTsm3dwlDx9A0aK1B5vSbJSpPisigBD4OiKYlm9TYxZJyMkqAzDCQGYEcT5ExYK3FDgll7nNJWVZkxjDYnrKPWF0jjSDLNSKmnopUBpkX5MqQV1NMUdJuGoSSZErSbiwmy8hVxIk+pQK5HqKi6Ru09ljXcHpym2U7UIzGeHdACAJJRGUZL16/xQfe/QReGa5ff5lZIVgAtvdsDha4Vcv5+/ZZNI7Y9JyGQFlG4jTHHnsmOcyHNaIX9LlBrAc2K8nVLXjLAyU3DpYIX3Aul6xOD/BCMYSk1nTOYbRGTrbZKkuULOiuPktvt1mHjE995QVi7/j47/0e5WybBy6e59L2lNZaVKHJg+D24RG3n3+OTTOnXbQcDw1xExDTnN0rW19z/X3d6UCM8feAk6/3urPru4B/FGPsY4wvkyLK3/X1/lGhFSo3rBe3eO7unOvLU3bqbUbZmFXo+NQf/i4qH/GW+97Kcn2bKAX3X3qUH/6Gd0IQbKmSjXfo2KGlp/M9ZaYwImNsxuznkkoZnr5xA0SNkAW/8JGP8usf/wSbxQGXL57n2vaMG0f3UFjW3QJRTjB5dRaVFZPZ5UxoE7rujMjrUULi/YD1HqEkQmQQLYSA1Gfd+Twjq6egdRIEIcnqEb7t0VomJZ9tyOsaO7gkM84yXAgoIwkhJLEPATdY/ODS8UBKVFakDaDbpPwAKYlSpGmFc+AcKitxfkAojUAhhEdE6PsOo8B3K3yzoihLuvWKTIDwA0qQjiOhJ3QNMdNIUqy3ipZoeyRnQqio0HlBxKS47WKURqM+osoafBrtSSlRWuJ9QGQZuTQJoCI1OgpEVlAUJZlMAiFhNCiDyM5iytBpI9Saui7xdsBHzo5kAoInBkWIKrkRRcSFiLUDp+tjTk6OWS8XqYEmPX0X0qZNok6XowqlBG3bI4FeakLX0XUD2Mgoelrvaa1jbQf6pWXQGZsusvEDhUwI9EIkqXZOpA+K2cxQZJ5MCDrbsDedYrF0XlGfpTE3Tct6vcb7iAgKkylmW2N2ZoodLRAx0i7mrO5c58WvvMTJYslm0+JWHVZq2PSE5Qmb1YpF0zK0AYtlNDX/dseBP+P6SSHEjwKfBv7rGOMcuAR88qtec/Ps3p957e7u8sLtO/z9n/lp2u0LOPUMk/NjvvHJ93H8wtNcffARfunn/zbnL1ziAx/8Uf7Fv/47vOv84/z6Z5+nvnIfDz9wkXvtAasFOCTDZs6Fyw9zdPuIV456/t6P/4/8yod/kUcfehP/VSF48vt+nJ/7sfdzZAP74xE//9RneN+3/Q/82v/+N9n6Dz7E/fvnKHcuIlSGItIPPcZo+vUSKcGUs/TRiZDcjjiESDoAiU/oL6GSFXe6lRRby1NMnuGCZ3ARMbRI4ek2a6JQqLLEtn2y2bqKod0QfESXJUIKgghokROVIgyWodsgdYYe1thgE6gSBVIgQ6BdHiAwKFOzXh1T1Ts4aen7HpFlBGfJpGHdbHBSsNqsqeqa2cULRJlRqhK72uA2p5jJiPntO4y3J1gHSkcCEm0yhAsIk6fkXRQxrJDljKjS+MqMCprTQ4pygs4Kgoe6HiW2oRBYayknk2QKm4zQgyQGQVWM8NGCqcjLCmPGRCwmr3FDqppQMJluMbQdFvBioB1WmLzm5OgOQgW0HrFZHKGKCf/0V/8J5y+PuTufMy63uf/KI5zPcqoiNev+0oc+yGe/9BVmmeNNb3qAuFzzpVVDWCx45Jzh8kOXedVbTo8W+NPIth241xyzc2mXdnfCC0/d4cIlj9x0jC9OeOuDW9gYWawj6yJDZh337+7xzBdfwk/m3HnpZcIo42i2g3GC5arHHTzNHQr2rlwhltvMLl+F04aHH474l15ksWxoB0tz4wb/6vYtKiIUmqsPnNIu5qyPl/Qu5Tl6JXjkHVe5/9ELhNXXFgv9eXUCfw+4BjwO3AF+9uz+nyZQ/lO/uxDix4QQnxZCfHpxumJ+3HF1J5UsUkpy7ziNkCvJfH7CiRa8dP0VDg6Pubq7z+TqW/j2J9/M7v4W83WPCxEfA0po+qGlnkwpJyPe98gbGVzP/uwqD++d5/rhXYo48NmTDpUXbO2d53ve+VaOXn6GJ9/4IFt7F6hG2yhUgmrEiDE5wQ1oBFroxAKIjujTCNBG+RrNx7mBECOZKfAxJOGOTWV8360JzuKDg+jpmw3KZOgyR6uckCXllxACIRWqMq+RgLQyyWPf9oSuR6scITXIJNXNlEGeWW+jD2d+gw1Du8RkAiE91rVE6fCuQZgCZxukCmwO7zGZbaOl5PToBO89QkakcFgXCDaCsyhdQJtIxiDTk9Yni7IWCiMlkaR7cDEgdIZ1PTIKkAqpsqSdiIkRoUQk1xlKF0TvMFmBkikQNcTEz1dnUxIpZZIXKxBnuQjS5HiXPp8QHVqkvoj3EeccwzDg+g47DCznC7quS5OLKFHDKdVkxrxr2Z1OyMoKU5a0qyWZkpwfVXSZwg2paauqjHGhkbGnNhmlNkxVTiElbezICkU+NfRW4KJm3Q5s+p6m8bS9wnWCLkRsbzlcd7SbPiUJRYUvNL1QdMIRvEf0Le1qQ8xL5m3H2qcFLTKD1kk1641IQq8YGXrP0cExw6bHEVJAaxTkmWE8KQiFYbGxX3Mx/7kqgRjjva9azP8A+I2zL28CV77qpZeB21/jPX4O+DmAa1fPx5k+4Z3vfJyTgwPyUYaLkpevv8QF53n6xad5fO8ij75hi1//5V/gP/+pv8YLL3ySx97zLfxQNuIXPvZH7FydsnKWwSt+6b/5WTZtz/TeLU6Hni9fv01ZX2K8dx9P/cb/QX015zee+TxBCX7qPX+B8ewc14qW1ZveyrMHLzEy72SUOZQw4CI2dLjVEpUbRJals6xzCBHpuiYRghH0zYayqlAqY3CpVO7bNSorCNaCzJEmQ0uJ3TRk5ShJe4cWazuMMEkY0vdk4wnCerwAk0ns0OK6pDMQeQa5Qnjw1oIUr/nvh35DdD4p9rRAeoWLgmEY0GfleBygXy4QMqLrLab7BT502MYyns1YvHoTMy5SE1FGlnfuYkYTrIsMcqAQI6JWyHbD6aql2h5T6DHWNuiqSv0MrVitVogQ0eMR7eqUemuK8w4jFLmpcDEhyaXwSJUhSwO9QESFyAIqVigpQWfITJKJHKkzGi8helzXIMoy+QJCQHhJUe9yOj+ib1pQ0A1HHB4c8+LxMYeLQ6YryWQ8Zc/UvLBeczAELlzYRynF7dM5pXdsho7qVBCDwq7XxCKwk5fYzFFkhrwI+EqhTGAcNYe3Vlx9cIf9a3tsvnADLzNOrjc8a9aoQnNxNqPvW0ZiixcO1jTO0x7fJQwWAuTjHKQmbwfu3fN0Q4evN+zXu9y9dRdvA10zcH5Uw+AwsmeDxxOxPhCCoDleYIoKERSoQKU103NTts5tY5Th8KUbX3M9/7kqASHEha/68nuAZ87+/M+BHxBC5EKIB4CHgae+3vtthp7xZMbbv+GDDC7y0gv3uP3yhuzUcevemgvG4YPl+VeOeezNj/Or/+BnKGePUuvIX3zy3djVGhMcE68xzvGX/vqPcO/Fz/MbH/9n/P1f/0Ue3Z9ybafjxZd/n+98/3/Ef/IXrvKWK5pLU8Hdu9d59+NPsrAtd597mktmymS6RZZnROuwwWJXK7SQ6KzEW0ezWkK0tH1HYQx907I+OcLkOc45unaDsBZJQBuDDA6ERxiFyRW+bwl+Q7QDQ3OK1CaFbuiMYjQm5jkyglQaosN2aUqAjUilES7iug7fd1jb07cNXddBiOhqhihHdEHSNz15OUWr1F3vhw3DpqXbHGHGI7yQtMsT2tNjVNQI17GZnzDarrHWUk9n0A4473FuwLZzFDKJjpShGSwqkynIRDiEVun+Zk4zPwEf0Fogup6irCnKCVifpL+lQhcVJs9RMseSKhpTVGRlhapqzGhMkWmMMoQAQmX0fUcxLnG+BaERDkxZU0/2EMDp/AjXN2zWC/rFKc1yzsGdV+ju3oM8yZHPVzlRRqYRtuqS0dY57NDzwif/mBfuHbCVj5l3nkmtWR/1XCtKprnhRutoVo6TY0u76nj1zhx7Eli8eML15+5QbUsmQlEvN8jTllefus6Lf3CTG7dOaecND57fZeUUY1PTb5YsNqfMT44YtKfeLXnw0TezNSmZFIGjuy/w/EsvoW3DzshgmwW5kuzkBbMqZ3s0YVIXbI1r9rYKxlsVJrNUE8lsmlPOCi4/uofTgpN7K65/6U99FgP//0aEvwx8E7ArhLgJ/HXgm4QQj5NK/VeAvwIQY/yiEOJXgC8BDviJrzcZAPDRs1kcc/7cZRa3TtlsNuxd3KIJkelsylaR0RSS3briw595hg+95QrzO8+TF29A4vmOxy7xTIzY4BFScn77MlvTbeo+8sarl/FmyvYW/NJv/xyP7j/GlatblLXkcNmhyVDZFqNJzUoEbi6WPGwdWhd43+OHASli8vefUX+cGxCdQhmNtT22a1+T7/qzLD+ls0QECiS12miEApx1aGnwJBCI1Kly8N6ji4ph6MiETLgxBCLwmhMuKIF0nqDlWYpwoGvXCAkSjTLmLJPAo4VGTXcSqVcpkIpc56yXK8yoSFJmOyAzg9YaZy3WDxRViQ8txWhM21uCd4xGI7oQiE4itUhUohjou456NMETMdIQsQhSjoELA1JLfEgoM1MU+Cggy1L2g66J3qXmKQGpSP0MrfEh4G1A54ookx5AKPVaXkIMDpOVaKOwLv1+pM6wQZ5FqaVMAxcDXZ+alkophLVMxzNcEDjfoU3GrEgUI+Uic9en34XJqbQgi44HZiOMbOkGT9MPrJzH9iBlMkM1XUBKxbD2DN7jJgV2viRaT2MDXm+4e+eY8ZUdDuYtolniiirJm4WkCSl415M+17KeELylMgqVK6LtGIYCbRTRW6KJ5FHTRwtRo6PECJ9i1pVCIShkpC0FqtLE3rJcOcSfsQq/7iYQY/zBP+X2//lnvP5vAH/j673vV19DO/DpZ54l332I77y2z795+Zjoe/pecSsErpkJbam4IzdsTXt+qz/lP8Rzeus6o9kW3/Wh7+fZf/J/U872ee9br7Bfl8RqxI986Ed49eg684MjfC35xB++zL/ov8hf/dZvpN2M+IZLI566dYfTZsHueJ9/XbzK+6fJ8OPcQKYE1ve061PK6RZuvcQhKIoExPBDEuQU+YiiKgl2g/AWQYJcxGCJBFRdw1mnPvQbeh8oiil9e4KOSQMf4kC7XqFk0v7joet78qokBI9tO7LcEF1EkRpqSgp0XiCkQUnNydFNap0TbMdodo6274hhjpAGvCArakamQohAf7pKSUbBE5UmE5LSTBj8BolCK8lqfczQdBACo3N7QMJ3NcsTlIS6PLMKk3ohcejxSmFyjXUande0yxOq3d1U/tuearydFp0UCbFa1vihxQiPzid4NkSb/PohxrNjjkCclb5oje8s3lp86EAnMVMIgSwvCLHH2YF133Lz9g2mecFoXLJZNSw2S+7LC+4MljJK8jznka0Zn/3Si0yVYHn3EDLNyg+87ZEHufHcDd7y8EVu3LvDc3dO8SZyeMMSgmWS51zdntAZj3v5hGHZ8nweedvDl1gHy+HzR6x8AA/3D44r9Xl+62N/wF5RcVJkrF2gjDGJpE4bTnYE4t4t9GybMAxMZ1vs1/vc+PLzNOsNZV7QrtaUCBoDZczO+BMBrzS50sgA+IBTinOXavJRwcGtFUevHGBKYP6nr7/XhYGoyBT3Tlb8wee/yIe+74eZVprT5oB2WHFjcZP18Sl3b9zhC8/f4l6XutAfvvciXz64i+s2rG99mf/+g9/CN7/nUR546FHed/VtCAxffvbzrL70ebK84/dffZVu7ri/nvHvvec7aZXgk4uB9zz6BGWRUciSzeKQfHqRLEssQ28D7XJNPMtxi0qj8wKvMoauRaDIyxm6KuiGFmsHgtRkeUWMoGSGLAoyIUBK+nZDUIZ6vI0QPZkZIU1G224weZ2y+GQCfPTdhsxorLes50eYIkMgkWUyJQ3daWp0hUQLskOLMSUOga63GLoWJSSmqAmoZG4iJQ11XQc6RWh7n6y/i8U8GaFQ9Fa8xmg0xRSiRGudMgbWK4qiSJFnRXWWaSDQOjkptc6JQSGzCqkF0hiENGey4AIZz8aORiMziQSKskbIHESSBxflCJRMP7MwZy5IEDGd02OMBCXQ1Ripy7SJhoAbetr1is3JCc++9BKcrs50GfpsrGdZ9wPrkxMuXrr0Wn7icy/c46UbB+A9daFQwxqTjVh2LVWhOeg8s7Hh2u6It+2PePuFKeemGZ2AtpcMncU0lv7lJcjI6P4tjIyUEkZZztsee4QbXc/+1oh6ZLiQZ9QZydloHQcnC3QILLtTDk83LNyKu4tTbt2ao4uaSmuM0mSlQRhJXZTM6oJKBAyBMgpqEclJxOmsVKhZQWMjw71T2ASqTH3N9fe62ASq3HA0X3Pr1i28GNFuGpbOElqHzCLzvse3LfODIxZNy82jOXdXHZ+d36bpA1LXqHqP997/EB+49hYYOlZtR7884OTgFOkNo9PrfNcHvoG/9qM/hG0cH7j2BtQ4o+8SVnsQjod2z1GaMglmYmQYOqJrUxfa9oBEygS+1ErhfJfKwgCCkErxkGCQyLP5vvdEHciMIqtr6q3tVB2YGpVJVGFQVY4XijAMSKWx/YDOSwQSu14ynW2DNsg8I/Q2CYI8ROEQOuXdBWJK4B1NUHmG1gm6QTQokyfKkQhkuiYzNQoBusANDc1yg4iW3g9keYHvVgzOp9yCXJBXJT6C8z1NsyEQGbxLbsiQNrsgJKg02UhRYwFPikj7k+OMlBJESJWLlmht0oIO8YzKlLiJSiYMF0KiMn0WnJKmLz7Y1P3vO2JQ5IVJORJaY13L0A703Rrb9aBgOio5WvRUmUQGg+9autYigLKoyEvDqJQUKsmRM0DFiJAKvKVznlprXCaxSqFNRlHVTPd22J7UXDQZou9pQ0RuAsumx5QmNZFHhmK7wJiCk+MVQZd4o+mNxpgMFCgvsJ0lWkdUkr47YcBjlyuO5gd42+OjQ2eCkSrQRUlejxB1QWlylNAMBITICDqhyLNSUtQFsom0LSmIlq+9CbwuvAPjnXPIaInHN/jcH32Ut1/epls4dre3uXtjySvCMZoKxitJ0y2o24bbt2+wqHZ45vIe97clw/ozrE57Htq9hpvu8PztF3nPB3+Yw5svIAvN5Ue/gcvXPIU9ZG43PPCGx3nzZ+9y9cG38NQzn+fKVs6//8RfJK+nAPh+A9Fj+45MaYagGI0rokiLW5sqjcSkxLdrhPAU5ZgQIyrTSbyjNZlShEGeBXOkQE+PR0gBaHCCTBgUgZBlWGvRozHKW2KeUVfnCNYjvWB1codysst6taCoJghJkriEIY3SMk2/mqPzArIagcMonSzXQ48xxRn7MNC6HikUzbKjKCv6ZkWMLcfLU4rMYIUniyqd1c/wXa6XjGcz1qdzRrMpSppEEA6B0hRgC4QyXZc7RQAAIABJREFUFIXHOZ1ER1WeFIZ/Ui1EgSCg1AzXL5AiVT5CZwTfJ8KwCsQOiskEpSI+eIRO4aRts8L2G/KyxrYNCEc9GrNplnRDTyUGnru7RAvFfY9eo8gUw427lEXBG3dqlsenOL/BdRv2L55jPN3mbW97I69+5RV0sebyTg0xI3hHHBwvHZ7wyKUp8xB5cX5MDBrnYWtScPG+HTKnGQfL08/fpO86Xn7ukOaBGeff/zBbJyBVycc++yxh3mAKz3UtMEXOLC/ZInBiNM3tBYdaMrq6z6A6JhK2dcHpesnhJlCf8Q+diGyPdxAyQxiBn2zjMsnp6Sm3b92jj5GLV/dgrNG14eDZ27BuiSKSmdd5JdB2Hd/95Jt48MI2Vy5e4Hv/8l/hTY89wMX7dqikxIXIvpIU5Qglc/ZkzeXxNm+7eI5CZHzhxiv0sWYkIr/5q7+MH1acDoe4YHnpmWc4fPmP+PAf/2N+5d98gnpymUvlmFu37/LEG7+Jj3zyt2ldx8H8JRbNANoQg2Poe1ywlGWdZtIKgnD4vkvhnQqUMYjoyIuCvBgxeJdAot6hzRkevf+T0juJi2QmyXSBH1LIiHUDSqcEG0yF0Aq/WSPyGr9pCc0AzjH4HlNssTo+QecRMoWQBbZPISBSaoa2ASXJijo9iYUiKoWP8YwlEPCuJVMSoQzNvRvs7Ozgh5bpzgyZGbb3ziGrMVWW4Cau6dBa0x7eQcSOiCRah8zqM2FTj7eWKATIpEqTWqOyDIsjy0cJK36GOh/smrSzBLTW/68XIytQWialo4/ITBF8i4ukFCBrEw5NgcmqBHaxm8RDcA7btYi+Y7S7x/H8iK29HWbjkme+8hKr1YqLVca73vEmGiL0Em8ds+k2XW8ZQuRkMUdrTZ5rTF5iosTLwGLtaJY9V6uaaT5C1uDjksMXbrN4dY2lpxeSx998jXFmqLxjvdgQNByfNFzLDet7pzh6Nj6yWAwcnPTccR23fMRaz7BuaI83xK5htVkynDoqlXPx2mUmWlH4SNsMuKGnG1pQERkiJivYHc84t7vHrCrZyhW2hGykCWvH/GBB63ukH8j111YMvi42gegteV3hhCcWI3o1YW+nZlwW5CojSEUmJdIoiqLkdhy4UFZs7cwoYsaujhzfuo0SGWJ1xDAMiONXGNYNfnGXzkXePN0jqoaAx5c1mRuYzw/p7AEffeY5Xjg8SUiw4AkhWYWFEAgF2uREXYCX6LwgL8cIZVKEeoxJcqoVeV4mbUDTEnwa+WmdIaXG+jOXoReoTCFzAVIQA2flbE6wA/qsBJauS/ZXQupuC0G7OUJrjQo50gui6M74eSo1Iv0AIk0CVPRnacMBfKTvOoTSSDKiCLTrY+rpNjFT5HWB9QP1eJZYBn1DDBkiWPro8e2AbTuiPOPlS9Jmd9atz7IMby1CpCrIu/gaA1HpPNlYhSIqjSAifCRg08+pFDGmI4MimbASqDCJnkRIWrPUK3FgI0GAKWoIoCLJ0u0Hom+SsMcFtgpF2/bMj4+4OM6oTeAzL94+i1xr6KKgyATODbSLNV0fiULhRMTIHGJHrQuKqmS+bmn7hsIkm/Vi4/HzOfO7t+k2EbG1TYaiqHJUlRF6T9+0GNsxNQITfSJRExiCp+17bOPpyyIdH7NEf+6tww2Rdtkmx6mHvK6IpU7cyRjOyNSSDsGqa1K2pA9MRhWzUYaSgqAMvkmbY4wRnMKp1zlo9LQLvPXJx1HGo0dX+J2nP8q3P/F+vvz8XZarOdlIMslL7jeevekWb37XVc49cY1HLzzIRGR85MWX+Nsf+13+5Z05j77zCW5df543Xh7xlRtfYJhpXr11l7KveOd957l568u8+sJn+exTT+HjmPc/9h6+9YGK+cE99rd2Cd4Rhha8w64WRJEh8pIqyzH1hIBNrH/nCMKnp2+IhAB2aOibBWY0Ajsk8Kf3uKFFa03wnrzMcV2PiBIzGiHwyZDkemyzxgZH0Ia2PWP0dwPDZs7QLqnGM4rJCFXmWNfRLlaIfAxnG1aWz8iLcdLTC4VrN8izPDuZVyhB8iosl4y395GjEt83DMPAsO5w7SlKRMbjMcN6yTA4RpMxtl+y7DdURRIV6axA2BYhNfXWBVQ9wtuBzFTJDUhKNZIqwwmZNgRhsN0Sk1dEJRBAP3SEAH1/ivWWznb4oaHdJLfm0B3jbIMDXDyrzoYlgUizPEHGmFxzbkC6Nbu7uzz73HNYEemE4c5zr/COS1u82lmeeuGI516+xfFijfCC/auXcQE++blnuHv3gLbtuXd4xDx6JuUGIQK+rnjs8iU+8pW7fPKVAyZWc+N0DUNHHBybL73CFz7+WcYPjRk9dh/lzjbDgSPMHUIUfPOTj/Dq8QZdCzpSYMuoysmUYnPcowy43YrgBVFH5usT3LFjvWk5XsxZzNfU4wlb413qyZhKCjbtkrtH9zi6eZ2mWXP75JTFfJOA9XvblLOaiZLI+YaiLBmVGlVGVsO/Y8Xgv+tLhoG7q4yHtmccns4p9JhOFLxtCp93jsxb2j6gROBuDDx26QG++f734tsNv/3pf8XpYs1P/+R/wdXdKzz/qU9StUtuHzmml17h4YefYL7ZcOXqNdbLU6QcePbpT/KGt7yDz33645zcuc63ftd38+DFC7ihRWU5IojUwEK/1r0eNksYOop664wG7NHSUE32cHYDNoV1BOcZVqcEo6jyMbZbJSlwDMgY8P2AUgKhSw5uvkiZzxAZ4CN5XROISGtBKYKzIBIw1OGTQYhEDtJ2QE2m5EVBrxXgESrBV6PICX0HWU4IPUM/R8kK6xzt6h55UeG8hRAxdZloRyYn9hvWy1NijAzOMp3scHD7HpmAyex8IiZ1HcIodD5hszlmOLmNnm6dhXwka7PSORCxVqUzfhQI0SO8QxUjYvR4F/AhoKNPX/fJDq2VgigYbIfJRkQUrh8AUEoi8hK7mRNxRAF5TJLrLK/QeGyENzx8lWo4Jt/WxKj43PPP8c5rl7gwKnDeM3iYjWcczRtW9+6yVZbYfoOwPa0f2AhPVJ57bcuOMlycFKyHwPG6ZSwNRgRGxnEsBGNnaTtPiWUdPVsmY/cNl9i77zInJw0HzQF5UGS2p60ypJcoASiHjgKTSbqRJpSC0AiyLKOyFlMXlHnBdDql87Cttzm59QpbOo1K+1xy8YEH2d45TzM4vvSp36cTElkJwmlgIgq2CoWPGTsSVk3Pq19j/b0uNoH/h7n3ivU8vc/7Pm/7tX89dfrM7mzjklwuuUtS0kqkKNGWrKgkcdxjx3AuEidBkAhQEAW2YSdBYOTKgIHkIohhODKQQLQARlSxmhmSIsW6JLfv7M62KeecmdP+5Vfflov37MoXWTuXOzczc6acg8G87/stz/N5XHQ8e/s+/+Ff+Fv81u99GW/gC1/6Tf7GX/lP+Kc//G9Z1T1vD5Hthy7yzLlL/Pz1z/AHz36Vvl3wsz/xizw62uY3//mX+IPXb/KFf/j3ee7ZIy5snEM7ixwteP07N9idbLL/2nf55M/9RYpizmAFn/v5R1jceZnv/ckf8PBHPkGMljhoQmyTGafSqGqCkBlmNkebkohMcIvQJrluGFDSMLQN2ihUYVBmnKjCIXH+Qoj4fo10DllUCQVdHzOqZjTrhlKMCBqCsyxOTpltzPGDQxiD6zqy+RyRAgnIMkOzXpNVFVrl2GAxWUUMAzoo2n5g6BaUozGua5IGIB+fYck6xhvnGNo12g2sHYAiLyr6MOCjZbVYs7O7QRSwampGGxsIqSiMxPaOYjQn+Jbm9B4iSwKh2A/kVZUgK5xZpoeWcjxPNOMwIFXOMDTJkRlc+rq0RGqFGATB1WgxRpZTcIFueR+hC7TKcK7BB0n0jqHtQEhMNkbnBm1GCFdT5hXawE98+pPcf/NVvv2d14gx46BbMwwDR71lZ3PE6d0jOim48dKLfPOFG4xjS2vPeIVScHiwZH+n4TGjmTxiCMscdE4QjjcPD/mxCzuc5jmHQ8CXGo9E3bzH95qa8Ybg4SefIEzHPKwqfu+1H5KPArKUWCmp15Y6U8hCo4gMfZvCVC9PqKKi9CPEeYXyA4Y11kQoKlTXU4tILHNEZyFGLmye49y5ixx3qa2U5+fEecAsA4d3a7JFQyYyonZoVTAtMmD///P8fSDagRgjz928iTRTjtaHGO8oo+eVm/fIMk3mW9ie08xzPvPYM7z2+kuEIfDRh54AqXmrsWSxR/UDXo34rZfvcLK/hxk/QN2/w9aW4rm3b1IUBXf2VrRN4O7em7zz8vc5Ojpi9/pD3Lu7h7OWKM4MKCG9wjorEFpD1GkzIFNQZ1QKbEAhcLYnK/IzK61Kw60QzmhD4NsE9KAYJ7NLN2AyhQoDeZkAokNT4/qG3GhclCAFtu8oplOiC+C6lAvgHJmWyXAkIGUVJlSXCxE/WIrxmKGrGbqUWWD77mwV5xm6nizLGKzHKEEmJV4rtM7wTjAfj0AoghOYokDlJXmRpQBSMSANiLxAExjW6wRAtTUxCEASVMS5QNQQUmzw2RwhtQfybFWloyB4gVTpJU85DgFQhOjT+lAIpIgIFyBElPAQLUpGxNATLMQYkL5NoxAEIDi6fYuvvbFHVIYr5y4AkWh7tDAMKHQl6ZZr9LrGoBliSioSEYQD26RMBFdIoggomSLa19bigmRrOqUVEQ3kmWLROHTbYMYZepwnf0fvEKdNgpcSUHmRtCM+YEkZDlYIVl2HlkkfobUmaMMgIjEapIy0g6OLA1nMMdIgfI8Y/BlOPXL/8A6r0wVeW4Rx9J2la2pOrSeIiBQaqZOt/v2+fSAugcFa/uSr/4obB8fEZo/Du7fZ7zp+7f/+F/w7D2ww2Jby3CYf2n0EY9f877/9ZX7qY89wYeMCf/Ts1/n+Ky+y8pqnr27wey+9zl99+lFaEbl581U24xaT6ZrF899mKLc5ufEn7L39IqK7x/3FAaPZlPHGg/zJvecTK8D3aYcOxGjSYEsmsYzSFVoJorRIERAa+q5BZyohx0PAZBVC5SidIYVG5SUizxEqQwZLcCnQoqs7nA+EoaVvlilPoW0Q2Zg4NPgYMXmeWg8CwhTYbkUUgqgM7oxviPCJUBTTz7PJhMwUhMGh86R58DGmNkKU5JlkGIZkXjIVWZaD87TLIwbboMoRXdeR5QVmNqUQAhk1Axmj8S5Sa6rJBl6XOBsSo1Al16DSGk0JBEwsUarC2uE9V2AaBIoUwCIzlElRYSJGotEonWNdQ1+f4nUGtsVj6LpkHe6HFqkVUgKZIS9HaJ3Q6Zkfko3beU4asE2CnT586Qq5ECytpShGZFJx+dIDaAQdHlkUjCYjxue2cCiGvqduGzItCVGiiTx6+Rof2t1muypZ+kCnW6p5wUTmbJsCf9ozpSDf2KDNc3YrTVN3XCjmyM6QXb1IO5sRFKADYvDQe4Y+YIxBiBSbtlitsK5jWTe4HoZeYFcrjk9qhtURmgytM0Z5jouGvYO72MPbnN66QVFCgWFYddhlR+Y7VAzIKOmcxf0bdMMfiEtAG81Pfe4jfPFffoG/8R/9Mq2E2KwYc8ozT38GmV3g6GDJmy8+x//wP/93vPbSTf7Tv/c/8jd/9R8wXQ2IKmdpjymK85x883e5f1zTV2PGOudkadgaX+f6j/1Zeuv45nPPcWG7wi56PvnJn+HL915F2Tm/+NN/KWG72zWDd0RZoKsx6Jwo06pNyFTiS2Ew2QghS8rJDEi+AaEMWTFCSJXy9PAJOiokUUkG59FlDjpD+IEYzpRuvktT63KKVp7BWYL1BDcQvSefTLG9xbmzzyNAeIcbOkSIxL4jxjQJHpYn9OsleTVOQA6SBDcvR2kWoSqETklMSmf0w4CRkmw0pcpnqOmcaHLy0Rhz1oI431NVFVEEZDFiaBvaxRFZVqR2xzm6+hStDc3yLuPRNkVRoGSgKnOsNQiVkxVzogAhJUoLlC4I0SFVhTI5PnpsPxCiRwmDGZ1HRU+wa9ywSERlQJsqCXukIgSLUTk6H6NE0iQMozlGGjanU77z8mtcnmpy79mYbGHmYz720HWi9ZR5RuEgzyrmZcXWpTmucajOEb1HeUWhKx6+doH7DmblhPNbFzFqyhMPX+Li9i6FGXP+0hYfeuQqF85P2Aye8YnlleUKqzQqlgwmsnP9Eur8ObY3MibzSOMccwmzWYZrLSKPKCPpujWybVjUA/v7+9y9e0C7XrC32GNRHzBIg5mOGA7vsXznJsH3sCHoCmhtZLi/Yp6DLFMQTJGFtJnowvuevw/EJSCUpC8FpQxENaYZBM4LCimJYaD2sBw67hzvUa811vcc2Zoh9KxPF2gsjTBc3Z1wRwsaM+bu6Qmv3HkLPd7FxzHGj/H33mF7ktG4hmpqOD46oJpOkFnGTG4TYsSRlG3GGGIAIWPy7MssYfGUIXC2svE9IXiCVCiTI88yB9/1tIPG+5C4+zHx9Lu6SwMuIZBS0dc18qxy0DrRcACGvktBI1mOG3w6+AS8d4AA5zBIpDYg0osSbY/MchLPxGOMAaHIsgxrHUEmuGemDVFJmnqBVB5xlkIkxyPkYFGCZJoaBoJrMUYhhUBlBuEcrNbIEEA4tIG+tSgiMQa8c0QRCDL9GUQiE4no8cEipcQHhSAjRoFUBjSpLTjTD6isRJ4ZhnwMZLrAB0FZVkg0QqRgVgDCgIoeFAgRcUQ2p2PGs4Ld8YTBWZwX5HnO5QfOcfX8LhsbM6SJZEVaYw7DQO9dYihK8BKEkDgbabuBI9ty3KzofeDi9pgLF3dZVwK0YpCC+WSMnOSovGRiBc3Q0zmHGywFSYKdVYrNrTkaQaENkwx0Bcpo8pARh+QZaWxPdJLaeewQWNkWbwPBA6YkGIOTik4EnEwzp0FaxsEg2vQ4xJDKf00yJkkEg/yAXwKZEFwoCwbX8cqtu7x194BFex+Mpc1G/Gef/1FUc0q9us1GbuhdDziyUcltBuygWdlACBG5u83pXDC/+hDPfObz2KHhuDYct3e5+MBVHn3kcS5ff4abqxX/7JtfoFwotHd42yfqbdS44Om9o5xsImKGEme9vjRIYVAm5fwV4w18TJWBOANa9D4Qg0UJn1oIIdGqJGYZKi9RIuDbllV7SIwRnWcURYXvO7rlkuBb8mLOZGMjrQujItiByeb5M2y0wHdDIgyZRBnCQ3N6Sh+SfFbqHJWVDF2fEoy9R+UFRil0ntHbgdVimfIGvcK7VLLbeo21NnEVGQg2oc5ijLT1MkFC2oY4HqFHJaG3dOsaGSXeC4L15FlFVuQpu/C9/14BHwU6K1O8mnRY1xCjRcs8fZ2Alpo8LyknF9H5FEQg2oHGduASNkxhUUpjdJl8EX2LMHlyImY5Mjr27x3y2cceoTy3w4VzF7hXW/7aX//L7DzyGM/8yMexosTpDCNHhBgRwRGFJhfJz9MPIQW7Nh17t48ohkhTr7m7d0gnFfk4o8l65O4UsTuhuDCnLjQhyxgNAZ955NAjjSeLlmbd0vctl6/skI+mFMqwvaUpZukiy62jPm1Y+1Ok9RActu6JuJRLODTkQlFNZ5STeYK7OI/toCklamawWtIfrzFCUqlIJgVVVVEYTW4EhXn/HcAH4hKwjeX1O8e8tX/Ab3zrCzxYrsnGYBnx+ve+z+XLD9N3K0SRUZRT4hBxXvC3P/UUq9yjZpGtCyVmYvjc5g5PyBHhrT2+99U/4Vs39/md7zzLD9/Z53TRI3Y+zpe+/hu8tTrhabXBlaYmqwpWR3u4GAi+Yahb8tEEJTVSCXxweN8SCYSQHHNSK1xXk5kchAeR0Nd+sAhp0t8VHFKnyoEocOsVMqsgyxiNtxm6FYNtUnS4UQmsKTVdfYptW7KqSnZjqekXRyhpMHmZ2hNyoov0Q4OPElEWjGe75OM5Is9xtgdVIkyG8xEfHV2zIuJo+oZyNELpjHw8S9yE9ZpyPKMYZSkJ2KaXNSsmqLJkfv4CznYI7/G2oz05QRQKITTOdUit6foTRhszhnaNKSRK57gY0SYFivT1CmJMoat2jYwk23JukhpQZpSzC+SjEYXWaFUgszwJqs7aJl3NycwUGwInB6+hvCe6Gi1SFHyIkdl8wu+9dJuu62lF4O/8N/8FizDl5W9/l9Glh8lGOcEGXAgcrBcgJd26pe47qmpMdAY/KA5fX7MIPTrT9EtY2Z5f+6M/5v/47rd54/V7FNJTCMVzi0MOFz3nQsZKOE4XjlV/n0wqhtpSdXD39gl764HN89uUcozwBaOtEjsE1l3H0PXUzhEVLLxgNor0ncM6RxwCuqoYLBzfO+Do3gF9vWIoLMwVoQZ/t6E9XFOcnyAun6M0GSLPMVlFlY3RVfW+5+8DcQnIGDncX9A3loBlY1ZSjCc0Y0kuHOV4lKK2MfRAea7gkUvbaOmJAnoNG+MMIwKyLDk4OOJkaHjjzj2OuiU+Nhy09+iN4ltf+ypPX/844sZ91nXP7uWLSX5b1xAU0iSXmlbFexFi3rqzsnZASZAiWXCFVrgI0Ue6NoWMSJmiwZVSyQkXBeJMWTf4AWVSkSaFQZ4lyCbNQSLOSgTaZIg8w7YdeZZhhybFimmVVIp+IPiGvmtQOifgUCL59LuuI/qAPINveudSWzA4pE4aCCMieTEFo2hXp3jvsMOQtg1REGNPsA6hwIzTRsMFD86elbk9k9GY2Hmk0MlMpHXChEWBjGBkkdKW/7VoZJOJtHUJSfMA/1qJGkWaOeg8mYeUoijHdE2LCBYRLMF3KQsSwHWJp+8bRIwp3VgIEsbV0bdL/BA5Oa4ZT2b81rdeIM9L7u8fsO4sIles+gajBbZ3rDqHlTm5yZBKEYSjWaUsx+gjayWJUVHKgAg12ZBRWsidpFsPhFySO8lKepraIXsHuYTgKKRGi0DjavTgKY1m7Swmz9MWRQqCiqgIIi8ZjbME340e6QI68+QyEJzDWsvgHa0diEahzxSnje3JM4mcZZjSYJXCi+RZqaVHhA/4YFBKyemNdzi5c5c77xzx05/9BVQ1w0wyHvnEkwijqUxkvQ6c5ivOX6r4cz/ycb5+922MqxmoOeki96uMo7bnD26+xqFtuLl3G2sCdRmQY8Xt1+9zSR4zuBE/9ePP0O7fZ+94QJrASAWk8OAEnvheXHiSDieqL8LQ2xatZFoDKYkIHu8dWabIygohYrLdBonApP+YRYZ1A+PJZtpH+4BUHm3ASIMfapp6gckrAioZgyJkWtG2a0QQDB66Zk3brNJ6qe/PaMMlOhsRlURESSTtvEVRIAkIP+BiSBeMJOG/dUXf19jVKWTQtyuq6QxhW7QRhKHHaDDzXbyPFFlBWK+JUZCPMnQxxhcVuijPSMyejd2LVKN5es0nc1wEpUC9a1yRkWEQCBzeR8rRRgKzShAyUIzmVGWJNjmZyohGgY6Mxhvk1QyVT8lUhhY5KIVd30WKQJHlSMJ7vEeRad7Y32MsFJV2rNdrnn31Fm9/7UtcvjznB8+9yBe++Fs8/tiH+ewnn2Y02kCbyOF6zeAc5fYUypRLqYRmVxnuHy3IQ6SIiouzLbJyylY2Q8opV3dnFC4QVi3dqqaNhmKWYdQcpyNyNsErSZCSYDyrzrIQDZO8IMsVo1FF0BqTa+wQmM8KLl85h8sMRaaZ5gqZK05ZY12PVhGVKVwIyLFmfdLSNA30DrNtyMeTdOlIUpgJBhEMiA94O+B8YESgr9eoN9/gt7/7Attv7vGV//NZdLHBr/5vX+Sv/8TnsGvNo/PLrO7WuP2ar/7gNW5+54DTVx2X8/M8uHGBJx78MMNpR5EbHvzoNcyG4oFHr3Mp3+DNxV2+O2yyvPEsy9MD1MVd6BqMyrn5ykv4qBiGNX6o0wvd9+/ZdF3XJfinyZKBSBZgEzRU6fw9d57QKZeAKLC2P8sgsMSQ3r0QUn8eQ/LKS2GJdgASr7BvG9q2xtvuPSuydWuyLEv4MH22h9Yj2rZB4pL+fOgY6iUSgSTg+gEf0rrTdT3C9/RDjRQ5NoLru4Qed0BW4qJCV1NWJ8focoye7aRXkIFh6DHjOaYa4wdPUeWMJvPk5ptMOXftUYRMll6jNCf3bqOUYt02aJFcg5kZYTQoWSK1oO1qRuMpRmVU5QyVlahigvCOul6B9axO1rRtTYxJSSezghBTYpKRGcV4hB/WRCnOch4FKsAv/Nk/w+7FbZxSDDbwm//Xr/Oh6+c5v3uOdv8WMXq++81vcTQM/Owv/vtcu/4QeVlweLSmqirUDmSFYzorqQtJmGZ89mOP8vFHriI3R8xUjrA5r925y6u3D5HHS55SgdMyozGB4ypiJpuI2Yz5zphweIJpHeVsxmQrZysvmZ0bs1hFhLTszFO1NcGzsTkm3y6ZXp7BVk65OWKalXS6Qc4FuhihdEF59RzrfmB1VFNKqLYL8vkM+07P8vYpQ58CcYwxTKoS8UH3DtgQkASk1pTS0C57nvzEw6yXNZ2HN2+8zIWdK7i+Z2oqsjDiu7duoVTOrKj48HyLQpXY3nLS9zRdi9ea2c5GIq/WHW/c2kd6wXr/DS4//jGm56+Rd0eoaGlDlwQyIYkrUNmZGjAlH0sEUihcPEvjFTKV5xKk0YgzUIY+w45rnb1nrIlSnGG0DEKCUpooBSqvkFKmMlZl2L7GDg3WB4rxFCUzmmZNjJ6smNG2ixRS4VOQSN+ckGlwLmURBucJOELXpEqAQBSGvu2QpMtMCk3vLEKmLIR2sDidsFVCKdZNndyOUaEwqQ1yAqU0fdcwhJhMSFmeLjghENGBKgne8i5mKMQOEMgoiD5ivTtDhCUc27urRXG2PUAIgoDgk1kozwx5njO0S6Kr0xo0DAhvE6vgXXNXkGT5CKkM0Se9ASHy+suv8gt/7lO0dcejjz2CoeXD16+zanq8c2QSjNLcf+cuz73wOpcuPsDWtELnAqEiRWGI3mKlJ3jFfDZmNh8v3AOZAAAgAElEQVQxaE3jLEZIGAZCN3DSLBnnigA4GdBEkBavklCryyNrmy7lXDqMlmiVIaRJFnEhGGcKFaA0GYjAOFeYqWF7XKGrCllWKVchj2xszBLZKlcMtUPlILTEGYmoBe1iTTe0ONtjo2OQ4KTEufdvBz4QsuHcSGYX5nTOM8gpt954i1/4ub/FF/7+mDcODnliZ5vJuUf527/web7/0h4/8/STWBE5f/Eq0XVcmI1YqJwX7t9hrznhz/zkJ7hBoJxv8bFFQ7eqeerBBwjScG+o+fLLL5GvVlx/9McoplMyAn3rCWFN9Gn4tz7cZ7p7kWg9XqeVXpGlqaxSAtu2mCxD9B06LxL7P0aUMGevc0RKhUBSTDYJbkgUYtUhg8OqDmNyOpkz1C3ldIuurTFFiY6Crl2jizGZFNTLQyabF7HRJdKxbTG6wjMkSpBUqCzH45DB4boaIQwm94gqCY6kCogQaNuaPB+x7mqE1IgYcc3AaLpFhsArT1Q51XjE4b1bTOdTmrpHZTnVaErISoYIUjpUVqFMwWxnG4Ug+oCjZ7p5hRh6qqokG01BWBCSenGMztNLNvRrlu2a+cYOppoytKfobMLh/h6jyqD1hK45SalPUqGzCoRiWN1ie/sqTpaUShERyOBBKYTv0bni7sEpylge2J7y1Oef4qPbGdvnNviVf/hPeOKxczz58GVWdkAMLS9+7yuM7MM8tTPhmSce42Bxws50Rt04ciX5/suv8iMP7nK4t+IHe/d48kO7bFMy28hpGo9oa7YunOdffPVFHp2N2Li2SWk0p8USvZDsr05wfQe3LK5S3Dl/DjebYqTgQW+JoWeB5vK5czw6HfPyO8cswzHbI8nW9QuMqTA+4/ZQIZVgd+MSV69cQpwc8OtH/4rd3QtoC7GpGGqLL8GtFtS+5+rFKYaM03VH177/ivADcQkEAZicHMdUGzwaZxUbV6/Rv3rExnyMzku2Ny/z0v4rbI4yRibjzdV9snYg01cwWxWjcsy+DZzrHdVmTqYUv//cDRSJzrrqT7i5GriqQA4dT33yR7nzzk3mO0ukFukLER7rOnxMiTbee6RIVlZrU9RV8C3GGIIPSKNx1hGEQIkzboB3SCUIPqDPGPmQqEMxnsUwhEiEM/FRGgD1TYsqU5/tECgRiNkIKTNiGLAuomVARIWSHcF6RJ5T5jk+CoamJxIxVYXrB5qmIS9HSa6qDTEOGJMnFLlzBKUSIm2SE/yQ+AEogl3R9AVKw+AcQ7RkzYDPp4n2Q8RrkxgLrufdaIkQPZqELbNeosucAKh8RL8+TdJiEQj+TwEXPgZUCGhp8H44sxZHXEwiFxEGTDlBa01nBdPRHKVEwqETUaI8M3oNxCiIShKExagpfVSIbsW5C+d5c38vcQfaHj94dHRIGYluYLxxgVjcxwLTPG0yrFOMRiWZB+kFp6s10vbv5qaS66S/8PWa1bIBZ4negvU4GajRqOUavEdrhRkiFoWXikJINjLDaBkIWnPXtQwucNqWOBU5Ol1SiYrZIHBmoHMR4QXCOa49tMlYG75894RLOxucn23Qrzu8KrnbWmqbHjGfjZhMC4YmMhMVdb543/P3gWgHCFAGzaZI8Ihbhwc8/9ZdDlaWu8crusUBqz6yfeVBnn/zJk3dsVyveOPWHkOI+LHkvl6yXjUcv/wmmzjUItK8c5dXfvgqt15/hVdv/oD1/huMspbN3YoHHzzPN77/AndffZX65ISNyzsMQ0d7fERcW1Q+olscIzONkjqRfkMaukVHEhAJBcHjfJsoPkVJP9Q4JCKKlClwxhsQunwPsSVkMs5IDEInA8jgVuSZQYuMrm+J3qKzDN83RKVYr9cI3+NdxHY1jkh21h+2A/TOY6RC52MgI0hNUU1AZ5hqgosaCIkuZMG6niqf4JuaopqdeRVadF4y2bpEVx9RTLZQeoQhoHTG0NRk1QgbHO3pMTGkVidYh7UpKMRH6G1EGo1WORKwbY3zkXxcUVbTJFByFqMEhJjchQZce5qSjlUOwaLLEaH3SD1C6jmr+wcUoyneuTSnsS02klozCmL0mKB54MJlPvYjP05x9UP88Ld/C6cEr7zyFr/yVz7D6wcL3jo6xjlHLyJBKCbbMzavXUd0a06NYD3OudkPOKkZTStGk4IhWsrKsJY9bx+e8trxmhgjT1y/xtHBAfPC0K57mpjcpLksuN+dsGobphdn5Fd3EZMp+bKnGBwbJpBlGUfrgdCu2egCPzw4oHYrjNIEW6EXPXfu3Gd1ukenB575yMfZO6354pf/kMv5mgfmO1ye7vCRS5d47OoWu9MRq9UpUUk+duUSF8sRJ/sDvbVsFe//3n8gKoHBBT57YZdbh7e5cf+Ev/CpT7E936V/6zmWb79MXkhWx8ec257xd/7SL/Evv/IN3jo8onCRaz/+MENn4U7LRyaa6fkdbs0nHL/0XbwdmG8IeuM4DQOf/8RP0svAi/dO+dqrr2CGF/ipBy7xyld+jz56Ll9/hnJjk/1791menLCxc452VTPa2MQIiS4KhBsQOvWgIQ5oWaBVROqMoVkQhUOJlEeXZQZJnnToYSCSEUVM7YJIe/i8mhGGHoHCCktwPdV4G9Vr6sO9pKQrKtrmGCFmKAUizxnaNVYk335ZztKUPisY+pRrKLSgbzuGrk19qCmpmwW2t+AblK5ouhOi0XgRMTLHjQzednTL0zRAdI7QLRFeQp6TT85TH91C6Ug0hji0KDGj7VZJ8OQdQhiEjJTVNM1CBEQB4/EYnY/x3Zq+WyRdQFamVGEB9aoBqSjzAhcHbNsmTkJZMnSW48M3qDLP8vQuk3JEvTplOt/CNaeEfAujc1yISK2YbVR848Ydmh98jU996tN85Q9/lx9/6lPs3T/m33v6I/isZFZ0nLQD+syqfXCyYu066t0JR9ZyWQYuXt3ggcd2yJ3gmemYtw6PuHt4i1s3lth4yKcfP0/PlMeuX+GNgyMW95eMHtkkGkOYaW53adZVnZ+z0Y5oTzqO3AmzzR3i+RGv3D2kOTnh8SvXcMbhwmniRMYRsQ4crhec25px/cFHyItNfv3/+WN+9a/9ND/+oc/xxpuv883vvc7pyGKznJ94+kkme7f4d5+6wtbFizRWcH/vGBM7+sZS+PZ9z98HohIY5YZ764bzH32C+/eOuXhxxg9ffAmVCT7xkauMR4aT03sgFOcvXWZZr2lsZGdnh5h5RsGnIEdRYjanhNYx1J4i11gFbSEZVRN0iNxcrnn51Vdx3QqhA5944sPsPPY4Dz39NCeLJUfH99FaE+nx3uIHS3SWGD3xTD0HqYzPzBikTP1qTFkBWkiELDBSne3NI94OiVmHgrPWQEiZ5LJSE5XG+0iWFVhb062XSX1o+7P5Q42z4Uy+rJFRo7QmoijHW9RNYha0do21K6ztic7jfAci4CIMvsV1Pd6tkx06L1FCwtDjBosLHt+siD7ghUbpSHeyT5QasoxQZCxP3iDaAVPMiMM6CYW8TWGiKklwi7JCGZNKej+ktN086f3jmX5C6yQJJlpkhCgF1qYtQIwxAVZjxEcJKJanJwS7JjcZnGkWbLP60zVuFEQh4d0INxRf/vq3eP32AYujY/7w1QOE6xhv7WKHmsW9uzhVoVTOaDLBdj2r5TGMJlgXyIKkyA0mKpatpVeSamPCw5enzCYVDz+0w7SUbGYpZi1owWha0b/bximLjzENA5Vk2TmGoUN7T1FqiqpMBGfnmRow0wAzyRhNocbIQrLsGmQm+PBjj7K7tc0f33iZpx6ao8wMR+D5W/foLHipGLqeWydr3mxA5BVaQNP2LOxAKWKC5Bbl+56/D0QlEEPkn3/zLfj6Kzz1uZ+jlxtcqiybj3yEBzdL/Pr3uXXzTc5fvoaXGVnv2SlLZo9eYPHOgg1Rcnp/wf1zFximU/aefYnTU0dR5kyvbnBxe5e8LHj57h3WIiKDoF8MLPPIlSd+gn/6v/4jKq24duUKD128ShcNqlrTVR3zc7t4ZwnRI5VARkmMlqyo0iwgxjPElyDPCmzwWNdjzCQpBiXAu7TXiJSBGDUyK4hSoRxoobGxZeh7TFawXtxjPDuPEoaoSlx3gggJWBI09EONsGfKvxgpJ1Nc3+C8Q/hAVVXp9wiBBEwxxg0NeSYRQ0Z0nqZeIgiMtjfpVjX96QE2Qjae0a/vk5cFUmZnswuHjpDnJbqc4K1HyYKmadi6/AAiFtgQyLTBux4lNSjQ5Hg3kOkNbBgSsSlaPDlFaej7HoTFDTmZMkgt6UIk2IB1kJVj6pMjhuUxk0myIRtjsP1Alpd0Q4OXGSYqYohokwRS83HJt7/9Xf7rp7b57gsv4LuBf/DPfp+/+5c/ww/uHlIvayYjjQgd1y5d4Ouvv0W5m9MXmktS0aiS4XCNbtd8Z++YzYsVHYFH8pyHzm+zc32Dm+8ccTEbcWIk0aU+/+jeivNDjQ4DwWVsNIJyZGh9z4nOmOrAxYvnODefsFd3XJ1uECYFS2EZx4K9+45rVyZEQOoFn/vRn+R4yLnx0k0+/VjFJx/9ND948x7f/t6zNPffRshtvMwojef5597iwZ0rLLs9Dm8d8vLbt3l0a4djO9B1a2bjc+97/j4Ql8DxOoVHlOMMue75n/7xr/Er//nP8J2XbpM3e3zk8Uf53ktv0qxarl6c8ZOff4a3spY399asvvka2dPX+NiHr7E8rdk9v8lLwwnF1Q26UvJwMaZcRTbajD3bcXV7l4eelNw/WKPG29y4dY/bukUNgklbc6ldcrxsQFiy0YRhmFJNpyANwTmEERiVk9xFERECQsT0spH4eEaVKczTthSmIGAQ0TIMS5QsEcLh2oDWBq9TToH3FlkPWFuji4x+WDHYAXf0Fq6zFFsXsa6DDob1EdV4OzECxFluQLNmNp7ShwYbFK5psUOgmG9i2zUhOtadQ3qL0AWT6YShaalP2jScK+cY7+gXa0yuWJ+cMNm5TH+6QG1cor73FqMLD9Mtj1FaMHQ18/OPsP/qC1x44klKOYII3kf0meEqtKcImdMNPX6oMZkgKzcYugYGT3QtXufk3uPOCM6lMAzSsXjzRXrrUKqiKrOzYV1LKUcYrZBFgY+evh/QckWOZF2voF8gi4JrdPzGS/c5vzXl4XnJEBX//a9/GbGyTDLBSho++eknkBtjFqf3CDG9xP7giJPFPdrhFKMkeQ/b68Dp0Zpnu1Py0ZjT7pRHMsNzdaDNMy5szhi2Kh4Vm2wMki01I6tyzEevIc+Q5MumpY+JKH3oHEe3TtnZPYcoBbdeO+Dc1SlVnDBRMx66fhl3/Umm45L6cI/PPn2dczvnee71Pe4+/yz9yTF0GbUYWC0aTJ7x9hvPc37zE2TTHX7jS19jcyRZ5IaT5SFPPHiJdxYfcAMREQKevrOcni4hOszWA8mEUi+oZiNunfQIk5HnU9TOjON1w9vPv0Ocpolqpg3HwjJ2PeONCdJ4/CCxyxU70zkmRha64IGtTeza8crRgun2RYQPzMcznHNsGk3CgDuKbMRgW6y1RCWJMYly1Fk4SGoLkosr/TgZiIwugJCQWuHs184COt7955biT92GMuF2UxpQJnHW4kl6eyE1AtL60Q30TQ3OJeCkFBR5RRh6UEmXQBCYvKK1DUErzChhs/qupus6SiXJp/OEd3cJXNKtF6A0Q9vRDR3WWwIGnVWsj05Zr45ZH76TkONdjVIaKQt0XjEQKbamhD5tPezQEaKkdwPODaA1q+U+7foEwYBWFc4PaDR+SG1EnucEKVAmodxdPEsZ8oIQHO3yBB9Sa+Z82r2fZZYl/YZKlVjwEd+tcX2P1pqHz024e7LmaNmyDiCDRS17okzBJbIsWeeCgxiIIU8A1s4T+55zWlJkhlwrhHfcXfT0QWMqQ90tCU3NJoGD5Qn98pTQDhwf1fRdYL0aOK17FnXHqfX0TtLV4GzSj2gEth8gNyxEpLWRYB1SZ4zyjEeuX6HIJ7x+821GZc6ly9eYT7Z47fCYOzffYh0FQ++TPDpElusVx+slUQpOFimT4oFLG+xONcd1h5GKc5vbZP+G4/eBqAQee+gyqjni7SPLjb1bTHYM/+RL3+Hv/s1fpL7vCLNd3nr5Bv7nBU4Y3j5s2FzD409eofCOOnQ8v7iFmY+47nu+tL/HtK64cG6Tn3/mIzRBcvPmTSZR8Mc/vMGFi7v80kd3KLZ2mciGxzZ3ea5e8PrNPTpyCtcwhAHZrJnM5vi2JcsKsnKKUiJFaHmfqDdK4X0qQ6PwKQQiOkTsUKbEYlFS4pB/2iKojKIsz0hGABpTjhikxC56FBJv18ShhawgU46mXjAZz2hO9uijwPdvJw99WXJ05zW2Lz6WglDdACHQtymBKAhPUc6RwtP1FoZAiJFuvWAYHAD+dB8vJEYadJ7AHsv9PWY71yinu5jJlG5dMylnLPZeIUhJMdmiOb5HuXGOYjpn3a4otMK5gWw8ozs9YvADtlmhTEbXCbSu6dtIJRXSCIwaJem1AFXN0cWIdrFmuVjT20jTC7p3XqPcmuEyjW2bs8oKJrMZUmgybeiWR0h1RHCeen2CM1MmeUHoBm7tnfB2CFzZnvLRK5t8/84SHxVHY084XnKu9AjVs77fMj1pOHJLdvKKhy9fYlxEbr+2z0tv3CZTitG84roq0KMMpwrq9ghtIq8u9imXLfbektWdSLYF5y9O+ezDH0GFjJfXt7HSUuQZInfIheP8RkYrT7jv12xNRzz+6ON8/FObfPEr36A+vsuf/4lPUMwvEup9/ugHb/G9b/wOWo7ZO1mxU1TcPWnZmE0os5zDw2NOFh2rk31C7PjZH/sYs7zg7f0jHry0QdtDla3e9/x9IC6Bt/eO+PYX/xHPff0P+S//l9/l/JNT/t4v/zK/89u/ySNzj7izzwPXtvHdgqPlwOM7ORc+9Em+9cbLfON7b7M9EmTVFtPxLsd1z1/9+KfZmu4gygl6NOIbf/RH3Lh3iJKRGzfvsbN/RKkD6qW3+I//g1/idDXwqfOX+Mo3b9ABj13YAWmYnb+CtT3VePKeQcUHl5DaWiNFgXc9MtMEFxFRQogYkxOCQgmFkjolFQmF9wNa50nlFxVSeqRUoBVSZhhnKUYF3kFVFgy+RqkMGxoQgj5GbO8Y2hWxKEEV2OUSU4xp6gXt8gBVzSm0QZsqGZZ0Qd0sqNcLNnYvUh+vyExG8BYpIsVsjK0lyoekehsGehcpptusnccv9hAnOaPtDZYnd+nJyKSmjxVxOkLnGcPQUZgMEQVCpCFg0y7p1isUnmG9StmJtqMYz5PBymjApPyE6JDKoKWgaVcsju9zevcOq9V9tkuTWATWkkuZkGN9T+havABVbRLjmrauyYxO2oW258tv3wcv8coxACs30JiMaiS4cuUcNIJq5mlzSXOrZXnrHo8/colX99c0q4GNxy9RFmuubNXM46MQLS/cP8JO5zSrhtdPD5iIEUEbXL8kjiDuSjItqUrNsQi81B6zqTLauce24FXPYBVGwGbp0aeBvpX80s//efY6y/Pff5b5+RGPX/8w+0jCnRf4zW+8CKsjlMlY9DVlVnLcDYhC0gVFKQ25UZS5ZDTKqdcLjo4My+kG1WRMUAVfufkWi4N773v+/q3tgBDiihDiy0KIl4UQLwoh/quzj28KIf5ACPHa2fcbZx8XQoh/LIR4XQjxnBDiqX/b5+iblq+/fIcnP/M5/uKPPpBK3JMD3n7jbdxqzSiruLhZ0K5PWNUdtxct664lqy1zM8KSEfIxF7IRVmmm5ZyX9g5Z955Xnn+Jo+M1BRLhB5rWcm9/RV4JingKvsN7z4O72zy0OWerLKFvcQiMTFPu4CNSZTjfI4Uh+rQTdz6x82XUaf9/1hokwKjCx+THFyQ9AQTCWTyXTBPD97YNQiikyNLAUEAgkYo874I5AypYhmixtkcJjfM90UMYPMI7hMkQrkvBH9YivKNvB2zdUJmcde8xOjIERzc4ZJaBzPCRZDYyBhsjcugJZMT1IdV4lxgC5egC69UhWZFjvac5usdwco/lqoZhjR0GvJBEpRiGDu9AhJTmK7Um1wGdKdzQglBJam1yRJQopd/LVIwxcnp4jza6NAfwLnkNQ0xhrC6mzMSYsiGia4jRY4dUQaV4Ns267bDSU4UMoySFVNQ9fPiBC1zc3WWzKHClpJeRo/1jVNugywnOw/7xEZkUaJmjS4nSGhc0gwvIUUU1qVIAi/ZUKqfKA9pI8mlGMZV0WmKt4s3jA75/dMRBvWAVW9a+Y9WtGURHWc2RsuLKbIf91vL1V2/Qi5ZPPfQQo8mIVxYrvn5wn/r0kFZ6vM4QWcbV7S0eOH+eBy+co3MN/syfMhuVGBmYlhneDezdO2B/1bJ32rJ/74S9Rf2+5+//TyXg+H+Ze/Mgy677vu9zzrn729/rfZl9BoMZEMRKggBBiKQoUaIsSpZlK4rsWInslJ24osSOrUpVEsWVKrsqpSR2lKjKiZTEkhItRUqUKS7iBkk0QIDAAJgZYDD7TE+v7/Xb776d/HGbsioh5SR2UrxVr1737XdvVy/nvLP8vp8P/G2t9SUhRAN4XQjxJeCvAl/RWv9DIcTPAT8H/D3gB4CzR4/3A7909Pwdj9OnjvPco6eh0Py5T3wfP7N6jN3dV3n2tMNwMibXkuefusiLA4ON0bt86PgG0/mMjm1zoqu43J9y/KzNCQq+9s51dCg4/fBDqOmAd+/cwXVdGqVmod5i7+YBhQmPXzjPbjBhYWmZ0Y3rfHWnyYeee4ygKLFjaLW75Cgsx0UZkjxNELIkJ8VyPJIkOnLkUf0DmpXuSxomeRpgHll4hTySjBY5SEGeBuRpVunJqECZluNUAhJMbMNGRjHTOEaaJrYyKl25UyPWLsqpU+Yls3BOOcmwLbDbq4yGu+TKptnrMd66iVGv49h1iijCW1rHn4wJ9veQNYss17i2RRjG6GyMUJJUWCg0hjSJdMy8f5dWd5358IDWyXPsPrhNISCY7eC6NQ72t1g6doH5g7sknR5tL8OqbWDWamTJiGBwh9SvQCRebwEtrCovYdbReUimBXkeooscp+4hyNFScHDvLsPZiOHOHZY9C6us0pCGKbHrHvPhDkiBIkPXGshgSDCbQxoyyWO0MDmMIj58dpE8zBAOPPK+5/nG669QLyQvPP84v393n0ceWSELc3Zu7fPkQpveqRO8tXXA/q0dfD/CcGo0ai6PL59iYud84e1rpNEMXxpY3UVUtEXjeI1CaLrOGqI00KUizTO8VBInPhUJNWLJU5R+SWpqDn2NrxSPtZfZaLfYn8749Jc/hfBa/PgzT9Ns9Bgc+iT9KUbN5cy500xCn73hgFO1FrXCYBqnxH5GGIYMC81ar81IWwSJoNM2ieKMeX9K4Mw4/cQFPnqqy3UZcPnqvf93nYDWeg/YO/p4LoS4BqwDnwS+5+hl/yvw4lEn8Engn+rqLe4bQoi2EGL16D7f9rh7b5uf/vlf4JGTDVI/5+MfbHDh/JMsnpRc+sJnyPwJ7sYjGPu3uLa7jwo8ZoaLXW/z/KNrvM8f07daGL6PIxQf/97HcFpdfuWffoqzF84xDafo1GZ5eZlur0H95ArHlnpYh3C3P2LeUpw9cZKb20POnTmOcE0QObbModRkSVot7JXgmJIkjcnTBK/WpEhjlGFTFBmW6RKFAcooKbK0UnblBcIUlDpHSUVZyqqyjgyJotDVEDeMQ8QRCmw+n5LHEUqUTIMppuWSZhFRfxu3u47MC9Jckwc+0miQDvqgM+xOk/69uyhdUHc8DE9RWi7RbEgehtiOx2Q2wVYmYZaglEmmXDTVkDKNI/JkXi3WNRcJ532EUAy37pCVgng0wO54ZKMQWWqmg11Mt45VpkTaw0jnTKM5rtMkng8pC0Gz3cawXEzXq+Kv0RRheeRxhFNTGK6HQFVyVgRJkhEOZ/QMRaPTxhUay7IwlEO1y6qqdRgEZiIJ5mOm4yGO16DwA7zVDTyR8+SFc9zZ2aNhFHzgfY/j5TP645A/2jvk+JpL5EfoQcp5z6NTb3Njp08+GtAfTdCxRNQ8SpUzjw6xeyt45g0O5wmn8wTdtFk5uYLhuFAaBL4mTwv8JKXw54RxTpmnPLGxAsdLan6d8fCQpmvw6InjnFjcJIjgt7/xBZyax4lTD3NhocWpk+e48+4NynmfbJwR96fQ9hjHIZmfcVhGvLYfMNzu46kCXZSsthyaDYsHD+Z4Kwvs9n26LYeoSDje6dKotZg1cx47D//bly592/b3/2h3QAhxAngceAVY/lbDPnpeOnrZOvDgT122fXTu/3yvvy6EeE0I8VpRFPjBkP3xLq9dvc/2bM4sLdGFYm25S6PucO3yJU4vN4jDGcV0ThYnzPpDLl2/zTAX1OMEaUj+wvc/TyE0+7sHWEaBXXeJDcVUzinThOOLC8iaRX8WsT9PmI8O2Wh0MdwW4dhnb3eEW/MwjRqe1/uTZJsyzArWIQ0kGtt2KcqcOAn/hDScZVVRUVmWFbY8/9Y5/SfnpSiPYKHi6HxVY4AUR78XTZ7nFHHVAcW+T1pWvDnTsYnCCYZTmYhzw6h2IbIMXSQoCXkaIYoUVRQUSYihE7IsJclT4mCEaVpoJYnjlFJXnMFwPibN/KOSZ4GWAvKsQll5TQzTpQinOK0G/qjCaJs1F5DsPbhHISQyS8jynCI8ROuI8mjdxPKa2LUmaabJQ//IK5DhuRZ5XpIkUZVGNKopwnx8yM72PUrDQIcJpWmilImQOcqUlZosL9F5QZrMjqZblXKLIkY5FrWmy9uHAfdmITf9kijV+HlOocEwJceObaBnKZtujRMrywxnIdFwSIFAW4JCKhQuZZkzOTwkiHysvMBKCtqejWdIak6HcaCYjnOCrYCD+4eMbg84uD/BfzAiPPAp4gyZW2hXMc8Frl3j5NIxBAbBdM5pR+DmBc7if34AACAASURBVA3XoNXrIoRkbzLALwys1CceHTDZP8CfDEjDgIP5jGbdxDFLjKNp0Xqvi2dZGMbRm0uREmmFcDwcy6QowbUESfGvIUUohKgDnwJ+Vms9q7a8vv1Lv805/X85ofU/Af4JgGFIvbrqYScW51Y9zh87wa/92qd5/sIaZs1Dagj2r3FrlPL+Z16gnR/y6PGLpMmYcDyntbzKaDznnbev4EwS7G6XfLLHuROblEHIuZpLItpM53POX9jgD+7tMPVCxMGQIXWeefT9WI0a7mqNwXBMo7fMytomQiRIWUMrhS5KLNclCudVRFjnpFGEQlAWOVmSkssEZVhVLJmELI2PTMAVSKQsUoRyEDKtAKUoNJCECQITLaEoBSKtag8K08Vrm0RRQNzfw2j10MmcpIA0HFNfO0sRDIl0ho3NZHhYcf0Ng9lohBP6GI5HoUwqt4Z5ZL8xSaMJZehjOA5G3a0Aq8oGy0ZLiW5I5pMRUkqSvMDPSpQ/wqw3SLSmDEqiYBu33eXwYIua69FcWodCMhr0sR0Lp7WILmA+n6LSGUZtgWSyh2XYZLgIQ2ALo1ojEAot4PalV2iRkMwLQqNNGUvajTa6ZpHMA6xGA51kFGGItA3mkxlKuGRxjG3VmI5n7N7fIbu7xcnlHpZn8eDedUZ5jS0j4md/5C+y2Gjzdvp1dicR9/fGvHb1He7uDDl/vIt0BfE0ZzaboRsCy/bY3XpAu15j9XiPJ0+f50Z/n1m9w/Vf/S2aqzXM4ZwskaSiJIw1UhSYluLrL91h5aDF937yOS4+d5zBXPPS62/iIHj6kYdZVo9zYzRkzTTZfnDAzVlKkRRcvXyZ/f1DGp5LODrA7BkIWWNhGMKZFc584iR33rxNMR6ytrnGN+9sc36zQ9Fy6XU3aQpBaZSkhuT1nbu0dMrJlvOv1gkIIcyjDuDXtdafPjp98K1hvhBiFfjW8uM2sPmnLt8Adv+s+0spWGn3yPoxe+MJP/03f45f/8d/l9/9/FeYOx7/9sYCJ554gWtffJm+k9FZbXHl3h4ty+XhJz/O7q1XycMxLVVwvz/iRL3F9SRnHM+I8MlnBvMwYanMkXHCc+0WdVMxXVrE8jIGQcS6U2csV3j42WcxhUEQRzSabcoiR8mSMA6xrHol38xLtE7I0oLS0JBFZElYjRIMhS4lUpqVkUeoI/Z/WcEwNRRpclSaXI0W+v19mq0OWZHiug38MMVxbRSC6XRCqQT142eZ7++SlxLSmMJuko760O5h2w10lhAMtrCtGlkRYzouYZnj2R7xaIjtuWR5SJhoUBme55IKhzgDIy8gS5GWiwKiKCQJfKxGmyxJCAf3cGodtFkFk3SWocvKhRDN59QWmxheDZIIu7tEur+F5bUxDJM4zzDKjDIFaVblzWEcUHdcKCDVMaZo8q03lYZnYpQmjhSUcUyoc7LeEk5c4ciFlSOKgpICnYHpWMRBSBjMkd0ew90B/uEhP/Uzn2R5cw3HNpmFMSdOO3xl6002Fo7z9luX+fpr17h1+z6GYXBrd5egyDFZpF5zUDJiMpmzvLSOn5dMdUSj6fL442c58EOywKe70aOx3IRuDXcSYTsaOxfs5RkNBIUUbC44PLS4zmKxyJXxPrde+Sbrq2t0FutMo4D1Y6t85PwF9g5GXHvtdW4Fd/jAC0+yFqYk4ZTl3gKjPCRz68RRzmKvxa2tPqlW1FY66Dhibzhlya0RNx3iLGN/d84Zr8Z0ltIwbB4cHHBDJXRqm9+x/f1LOwFR/XV+Gbimtf6v/9SXfg/4t4B/ePT8mT91/t8XQvwG1YLg9M9aDwCwXZO93TkH+yOkH/CJH36BsH2G8e1f4naiST7+DJYuEfke797xcd33cOOrnyHo9didbnFvdshTvQ1aJ89ysrXCL/6jf4x+9ASpcOi1BVu7GcM8AdPCOuwTNzxOdjp4Xp1vXLnHoxd79G9fYzqdMr9zhTNPPEV74RhxFFC3HPxZQK3WQKpKoFEWMdF4RKOzSI4mCWbUW0tVrX5WYbLTMqk4AnFMkkbYrlMBSbSmyGPKwkaXBbN5zGw8I5z7NHurhFGIH87RZpskCBlNMwxLMpsfYrtN6vUGiT+jiOYIQzHbP8RstbFEyeKJMwy2tqGMUQXY9SbpeEhpWYRhiHIaaBGglIcfz3BqNkqnZKnG9VrEIicajSruXlmg/TlJliJri/hJQpqEtBoNkkJjGBKtJYnWHPbnWGYdwysqV4FlkQLZaB/DMimzEsO1EVFOIgSW26JM44oSnMuKhmzaCGmwfvEiydZdgiDAKAqMQhGNDwm0puZ5xONDlG1h2QZZHOM6NXSWE841s9mMpc01nnz2Asr0yLIx/uGU37l8lSceeYJ/5xN/hX/wj/57dt6+hid9Docz4lhSr0uk5+HHGY1GDXNRUuqEQiiyJGKh4TIwJNrroDITs26AFDzxg89y9ZXLNG2bRJQUeUEjKFlaavEjH3mB3sY6r9/d45+99hLtpuDxR47xgWefRMc2neVV7PYye4cjfvt/+S3SYMJ8GnKj22bzzCaj6UMcHuyw+NA5RLfFyy++STyZ4pkwvX4T0XOpt2skcYBXr/OFV69RzCtZbt5u8IFHTnFnd5tMacIioziqCfl2x/+dNYHngL8MfEQI8ebR4wePGv/HhBA3gY8dfQ7wOeAOcAv4H4G/+S/7BkpKskyz1LY4VbP4wMXz/MbvfJFOy8asuRxOIwql6NkWH3z0EbIgYqJzZmFGtr1FoCz2b9zgYBiRC5N1S9HWPj1lMJzm9OcjsllAmISkOiXWGrveYmlhkZtbO1jSQtVqjKOKx/ctVHhZVBBHyzDI9NGcqtQoIbFch0KWKENAngBlNQqQuqK9S6DM0bKEPOdbl4u8rBwGZUYpDfw4IEs1ZVEwGhwwHvSZz2KCMGEynrK3+wDLsJjOR8znc4KsZD6dYRRQ2Db1llWx/1ODxLCrkYo0KfIUMKoahKLiIeb+BNtzEVGIgUY6NYR0wFQVMTktMcgxnRpKWszzjDLXJGGAsAwkJUUOaRwQZQVJoSGKUUoxDn2chksyPsQqNTLPQRtY0kLW6pQFVblwVpBGU3IEqT8ly0PKMkfoSjluIEjSDMO0wDBQSiELjSg14gjZVmYFZZT/C5hqkVBo0ElWZfxLGOzvsr21w51b95gOMgQG4BANDmnVIAlzjLzAyQtMZYCSyEJXaDZboGwHlUTYRo1ISBacOnYq6daapEU1GpTSxPBzSpnjCI2rJI+dXOBHP/JBNjfW+NpbN3n5jTc52ejw/s2TvPCBD1DzeritDnGhuf9gl6s3b6KTKijmmgb93R1G+1PaC4sMRhGq20ILSRFlDKdT8qO1pnSSYpngODUKrUlmEVkcsWwZHARzDFMR+RHDWUoyL7DUv0KUWGv9db79PB/go9/m9Rr49/5l9/3Th5KSxbqJjBRxTXF3e8iD179K2nDYONbgrdcuc/bP/xjrJ0+xevI9vP75z9JtdcnXe4TFHHsWcf7J91Uls8P73LE1P3nyOM7xlP/is3+ITCwuPnQBe3CAUUhCX6MSyUjEdAxBV00hTVBWgV2vcN1KFVXctxREaYyIZliLawTTffI8w2v1KJMMy6vjtFbJS41hWKR5gmW75EVMUWiEhqLU5HmKP53hujWUUgThDGEUJDHc23nAPI5pWSZaKGbjOc0wpe1JdJRxcHBAEmUEQR8/Klle6+A4NtPhBFM1qdUcynRMFpXU146RPrhDLiRZFmHW2oTTMb2lTfJawWx/j0QXdDbPM9m/R7e3RB7GzKYByjGISovZ/gCEwrZM/MxHqxIZxuDUGE2HNJodZnGMIUrSLELnBqpoYDaWId3BqDWJZj61llNxI00T27DQpkNWFIgsIw3GKNvFMk0oqwpMBChVp7QdSn+OACI/wzAdGq0aYexXUwINiQTLsNFWTFaUSKNg7kuWFOzce8DL37iMUXfx8oTGwnEeP/9BfuU3f49RMOE9q+u8E0QEkwTpJhSmoKQgSkIMDKZK0/RMVKnw6jUGZoKfFOh5QHPNwb8eI9OEfJ7h6JRH3/swluFimoI4L0i8Or/y4iuooM/HLqzy/c8+jaksEiF59d09nj23yd0b1/mlz73CppEQJhmZFozTFLZ3mQQxH/7ER3nf4xtEhmJ7e0iSRcRRgdPIcYSiSFKEZRCaDrfvHeAAUtoI10BFmp3JDB8JfoiXlDQaS9+x/X1XVAwmecng/j47sxREyeAPPs/a8QU2F1oEmNy594BSx9jNE7xy6RIvvv4uP/GRJ7jW9njCtbl8aZf7zhbHzp/lnb0Rf+en/jrD4U3+8Ppb/ORTj3Hx1CZffu06l3bGrLTqbJ5YZxQUNIyYj73wHHt+yu7t+zz6xMMcP3aapeW1KuGnFEk4qfDilkc0OcSwa3imU6HHTYc4mGJaDpbVJIgmOG6NoijIo6Qy7irQwiTyA9BVbfxsMkWV1YJZWZYsrp5g8MbrnLl4kpff3mJ8/Q26mydJJgecf+z97F1/h6WNU9RWViln+1y/9IDeQgez1aRhucz290hVBbSst+oYnTWM0T4HgzHrKCxpEYVjgiAgTisL0Ohgl9DPCcbXyaSk1lxkun9ALkxkDkk5Z7AbYJmCLI/JhSYuNDXTZjKZkJYJWZzgWTbHTp4jnM+YHA6ouQqUjbSm6CytSEu2R1YWWNJAWTamBMN0KMojZYMQlWMx16ydPsPk9ptESpLOpqhajSyeM44j3LqHkJooqSCs1EqiYcZweAhS0ux1+Oofvkrcn9C0JQeTOVd2D/mNX/sv+f0vfJGLPc3ntg5535kT1HpLPLayhqEsru1v4ccT7uylXDy+wvTGgDSO8RlwenmFd4db3Lr6Lk+sr1JmBScaizSzBgMRsvnsMzz12HspiowoCLEMhZSKR08tY7oOo1nAL/7WHzDtjxnN5vyNv/ARhHGGL/zRJcqDO8hug9ZijVkW07Zs7m/7uOk+V9+6xkeefZIv/vElDq7fIokKHCRplGCbNqUC0485SAqUqZHCwrY186Rah3r99j0++sg5dnd3aa2uE2j7O7a/74oAkamgXsSEcYFQikCUKKNGICRhUEKUY5jVvnHSHyCTDLdlVjqtAlQOi6sLUKTs9O/wxq0tegsneHz9LF+/so+fRjy5YZMlIaYuGCq4v3WXnQeHSNOgrUzqNYcTmydYWt6oiDmi2hEo8wKtQehKASYVpEdwT6UUhll59tIsgbL4F+rto4BRnpekWUxZFBRlheeWZUGu9VGjzJmMx8wmQ8JMMHzwAFMp7KKgLHLiJGHj9CbT0GdycEB7/TTKKDCkIJ+GFVq63iWJ50RpRJFrkBahlBiioESSlppw5pNlObkuSbOCYDQi8Uc0WktY0sT3fUCQhhHD2Qjf90nLGGVAlFQ1DxbVXn4QzlBJiW26RFnGOEho9xaoNTxMq4shTSxlVRZir4a0TFy3gZAxhhLIehNUBVotURR5XoWsdEG91QTLQFFt1QVBSJQkaC3IihRZKIosJ0tyoqTSiSvTRFg2o0wwtTRLdY+9ecD27gAnjoiDkLt33qLbrLO82MR0TLanAaMoJ8pKTqyu0l1cQucCgVXh23JFHKcYrokIU8wU2u0WQRRhKsnI9zksSoZZjihyDKnw6g2ceg1p2USlYn8w4+13dwj398izhPMnlvC1pEhSdJrjKYPCMWidaLO4UWf1ZIu10026qx5xMcd0XZYbdaRx5FQwLIpcExYlKZqhHxHmKaJmguJPFGrKKHGVpOFW6Pam43H3wbfXksN3yUhgbWmFrj1nNJ/w0bPHuTyYUStTZlHGcORjUXA4GpCO5yzaOR86vc7aQoebgeT4sVX+g3NPomuLyLzkR97/NF/+2kv87q1LLHQW+Zkf+gHeuXmdNC/5xKPL+NR45e4Ox2ybVDucXOhxEMU8f+7DdJY2iIoMBxNURlbEIEWFydIS5agKCtpoV26BOEdaJspyiKYHFEpVsV0BcZaiygKQSF2S5jmGKMmzmKwoSErFzA+59darDMc+RhbQH025cG6d/UGNWRyglGL/wX1or7HYabN96xaTScTpxx5lvL9Ly3MqtfrSCqaycByLyWRCvdWjwMWqS8ZBRSmWxhEKLQhI8yof0F3epD/YAcOiEAnz/oh6p0Wt3mQy2MVWirkfoWyLPK2G5VEYoDUUOqd/OKVXswjGA7LldZJc0VzsURY5eQCWUyecjmh2FpBCUTrLIDUizTFci2ieYxsJZelWxVaGTbvXxXXq2IaJKRW2aZGWmlwWlFFJJlO8eo2yTCvvXpJSsxWzwmbsJ3RWFxj4e+zuHKAKwd/7+b/Pwd5ldnfGNL+3y0ceP0s6DZndfcCDpEDUm7z/0ZMcq3dZvuiQioQLDx3j5v4hJ1sZq2sLJAd9jnc7TCc+twbXOPTnxIai3rRhGPP7r7zNYruNYwhGwwmHRUSvZXBjL2Lr5n2aSrGx1ObRhzb46rX7rC4t0zq1RO8RGyzAlGQzRZaanGy2sNwmuF12win2Wo0T4iTa2Sbop+RFtUvilZLdrQnNJZfmQhezlDhWgaUEvqicjn6RojuLHFtb5vr9m9+x/X1XdAJxFPO3/+df5t+4+RJ//NWv09cOSa2kfzDk9MIqwrL51Ode4Wf/6seJG6dRxR/gCIvvOfUo9179bbqPPM8v/Fc/zw+99yT3tocMYoElBTcfbHFvFCPjMZ3VY/zoT/wQw/Eu7zvc5XNffZNbg5hnF7o83lvAUSbxdEiOjU+C0y5ASqSu43h1LEMiUORmQRzOqLea5OSI3CBPhmglcUyXcf8+ym1QppUxJ8kTbKtOUWQIIaHMyHLNgwf3uH3tHaThMNi6hGPa3HnjJdoL6xjNFjqeMUsi0nibRpqzdXuE0+0xHmxx6Qv3cR2HYb3F5uYa16+/w/rmKeJwRmtxib3dA5pOwTgu8SwDYTaxLIPBrWs0eh1KZWNYJYPtB5XpKJwRBCGmaTPaPwBDYQjFLAwqFHzhUhQ5NSdlSkkQ5ziOwd3tPXxH8n3HF5n277P2yNMIZSPKEm9hBaHBbXexXY9MCJw8AZGTxwnK0NiGIM9zHFNXHaYU2KagubJBsLeN0TZIopQ4npNHCSIH03PIspJWt02MYjabMMVkqaN45tEL/MY/+zzmsSbtlXX+7n/ynxLaLl/+zV/lk9//AUKrxbW7h2xt3cOW0K67iDzk0rvXMF2FoRt8z/FN7IbiyuVL2O9/lMahz2///jdZ8xwunD3JTMfsZTkXnjxHo9lEuAn7r11mP3e5M5gwLxO8jSWazSaNns2ZD56jo2CSwEt+iKqnXNq9wzgeMdmdsrU/wkoks3mMiFIWFx3kouahi8tcObjMubNnMJrrbFxcYTyIeOelu/g7EwoBnaaNP4zx80PaS03imU8/KECkGI7F1Xt9nnv6SZTbJgzj79j+viumA1mec+nyTTZPvZe/9tf+XWp1jzhNMcsc0zZomIp5lDKYpnzqK6+wZDkkRcnGQgetJZ5l8thGh2bLZRLM6ff72Fpw3Onw3Ic/xl/5i59gtWHRT2Bl9Tze8jqf+PAjfPTJcyx32jiOB0dBniScYTfqpHlJoamy66KkLCu3oGFb1R5/niGEqGg6SYyhHLIkQNkWZZZjyqqCy5Cy8hdoDUqSFRlJHBD6s0oAgsZzLLQEURQEYYgeT6qYrDQQGEyjOUazxfhwjzyeUxoGsT9mmmTMyJBZwuH+ARLB/v6YxaZHKUx0keDWGswDn0xLEgOEWSOeD0njHCy7AnlKiSrB92cUoiIM50VGRonQVFBTVVmTHQlJEB65/AqMsuTWwZBkcoghFaalsJVCZxFSgut6FTDGgGg2osgyMKBMS5LUx/GaKKMSl4CkLAuarQ4lBYaoIC7KNJDCpCwyMnLSNCHKCnSWkpQ53YZD3bFpNlpk4ykty+aZD32QDIN4soMsIjqdNvt7e0SDIZFfwJFSDEfSNSQNR1I3QJWS/tRnlEQUjkWQJSRhjjIkWkhE6KOqHwGRV7tEV+9NeO3uNvuHQ5LBnMm7u+y9c5ebV+4z3fUZbwfcfnuHW9e3GN4/xJv4sDMiOpgx252RzWJEUhLnGYmfEPQnuHFJNM2JoggRQhRIak2b7mIbt2EhC4EyK8OTAziugakUng0cTd38eYbKU5Io+ZYS4tse3xUjgSJLePDGF3n7FcnC+af4W//R3+HMsS579+9z5etf4/V3rvMjH3icZLjHv/m+OqZ3HnPjONevvkZWSl5/8wpnzz7Enbe+yVdevgOOyV5/iswLXrz8n/PwqTVKobFlQby4xIXHPsi+scKzayXxfM74YEiuNZYpcdsdkiJHJxHCWMA0q3SgaVcocIrqny9XCi0UYTQFKejv3atq3E2PNPFRpkGWppUaHHBsh/3JkLC/w3A8Zfv6HYYHhxQ6JTc9VJ7i2DZpMKJwa6SJj2FU8hKDlDCIsKSGMiMO9xlNY8b7B9zZuUOvu8zGmmAaDahJk3uHKW6nSxpGzJKENE4Y7G0T5zaDvQF5ZpASYBg5s0hSxlOKRNNdqJEkEYEfI2s1wtmYyHApZyE3+od86JET3N8/5NbuCM9cp1szkULwqa9e5uPPpHyo0aJII4QyqXdWUUqRxQFFNCdOC+rNqiDKNj3cRocsLymUIklSXLdGGvko2+DME0+yc/mbBGWJ67qURV7lGeoNZrMJbr1Ow5EkWc7KygL9BxG+73MzuMyZdgO969P9nmM0rQF5rpmGCW+/+hZvXrnKpa0Deq5gEGfMZwWO0HQXa3z47Dk2Fpf4zJW32N0+4L/7ub/FZ954hcCP+OALj7C20eP0+ipf+uKQY2fWMZKCWzv36WwsM45SGkpSWibh2CdNM55bW+fsxgq3+hNeuXwbYYGlHIZSU25m6DRna3uKSEs2Vk3Gs4x7Y42fCEyRsmjCrUCwtTfANhU/ceoh/sNf+E3+/AfPwfljfPnN29jKpt1tkBQlXqNFz2uhi5C8gERa1KVJq72A4zY4u7kG/GvIDvx/dQghkEmMPxnzzqv/nK2771Bgs7KyQG1lg/c/chav5fG5q7dYO3aC/b373N7fxk5jDMPmwdSnJhPevd+vmOtCUmYlwzDB1iWHh4dcH874yktvc+Xt69zY2kYVATIvyfKALAopkwirXq9y/mVJrnPKPMMwTEzToNQZWeqTpyG51pR5Ufn14rBiCZQplu1V9BspyNMqbixLjbJdsjJDZyVxXjAZjClltVMgrAoyKinQpqqEpnleRY2PPIiGVBhIlLQxDDCUINclthDs9cccDidcvb2N6XbZDzKSwmA890F5OMoDkZOkJXkcUKDBMEnigqw0CWbVnn0sSgxRoISBIEEphSug6Th4tsF6s1Eh06YpSZZi6gKExjFLkjRna29MEAwQZklRppRxSBJnmLaFaymUAcr0aDTbOM0OqZZ4Tg2JQmlNGkdVWlBZ2K5Doqu6C6lzMAWm4WLbZiVIERJtmrj1GqYwSHXGLA4Zzqco12X95HHIBtheiytX73NuuUujU6fjeng1l6SsQKqmKCml5rFzm1w8tsm9aUJeRJhasLF+isCPGKcxynEIIsUkkyANpK0oI82aZTPzE86e3KRTdzAMg9KQeLbB8WMbbM1Srm8NiPOscgcYGlNWEJokq/Imrq04trjAUtvDFCapKDBtA9/PMbMUy7LQQuAnGQsLTSSKZs3EQZNLWcWohSbLMtAVp8GrO/TaHl7NAlUwS2KkZX7H9vdd0QkYStLt1lluGHTrKe+++Sb/zS/+D+wEFg8/dIbHX/gka4vL/OTzH+bO7oT9O7f45ksvItEYlgO5wXh3wJt3RkQ5FeE1i7G1oFNz8TOFVbd4/eo9vvRHV1gSKeAQBQOyIGDUPyCPEgY7D5jvPGA+nZIHKU69BUpiCEmRQ5kXFEU1GsjymCINEAKyPEYqmyyNUCUoTITSFKUmzSKkKZnNQ/z5jOFoRqYrhoJpKrK0QGVpZSlGYHs2UqdYloOhJIaiElaaNoVIQFYLZV7NRBwNjR88eMDgYI/b9/YIxmM++89fZrS9zaW3X+fGrTsMD+doMvwcRmEA+ZTSNtgfTmjVHUxp0qo7+AkMo4R6q4ef5iSlQSklpmniKs3TTz/NwoJHSyj25xEiFmQamkbBnQd9vvb1VymSEtOAXJWU4Qh/NMJwm9i2g5AlkR+QhRNMAWERI2UVv6asrLuGKTGtBmVRVRLGWtFw6riOgdtw6GysYEmNlJXOPBIWiShIdc48innPUxf4wR/6Ud738EW2+od8+Ytfpn3sOL1Ol8cev0i7V2Ma5tRshdc2Wex0+NBTT7AdwWdfeY121+WJZy+SCAOvZjGI5sRJQv/2fW5v79I4tobdaRNP57gtBfM5T73wJK3VJZqW4vhKmx954b2srvR459Y9oumYqNQUaUZeaBxXQJlRBjFPPrTODzz7XrxGG2laLHcsVhs2p1YXeDA4JPFj2som0iUDCn7mRz/COwcjZGFQKhsLyUQb1C2LNE0rNgIGUa1Op+txZr1bIeh0gWPVv3P7+/+tpf8Zh7ItiiQGt82CJek24dTDT/C13/wVfuwv/wz7966wc+Mak2nEiZMr3Nqb8okf+BDBaISp4PkLZ/jlX3+NsxsrvHxjn7iUnFhpc3KhQe42yeMEK034yb/0YRzMI4W4gdleIutPMRybLI+oNdfAtTAMQWv9DIaoKtbCMCSO44rf73jYbg1/PMBp1kmSBK1j0kJQojBUVRugMklRptR76/T7+8ynY6688SZSFKRpSpylBEGAsCxKU5IVGgyJkgrbNVBKIrOCqMwxhSYuUoQ0oMwxLYfdfZ8sL4jjDEtClIfcufUmp089xPc9/R7eeON1mq5FGY4I0pIoylle2USWKf39Adp2OOgPKFybbsclCAoKrVhc7PD4uQV+97Ovoh2Tcj4n9meMhnPc5gL9IOHkIqYpewAAIABJREFUksJ2NKnt0VSatbqN7ZgMb3yTz81GPPuhZ3DJaPaW0f6U4XCE6zo02w3cRqsSuOiStlcjR5DHIY5Xp8wThNYIJbENxSRLWajZdNc3CIMxpuVRKxLUokDWa4SJpq4sLnYXCOczgiTg2Q//IF9+9Q0eOn2cq69d4XueukAURbQXl2g/fJEfL03+/i/+Ko+/9yTPPXSeRneB37l3DZnMOf9Ij9SFMYL/+Nf+J1pug5iIM6fWufTGPbJ79znx+HGiaEDTjLDMFse8GqppsXz+BMFwyF/62AfIlMntG7d5+uwapDntuoM0TDzLIslLVla7HE6mvOfcMUrL5osvfqOqLdGKIC/oWiZrp1bY/qNd5ntT1s53eftglxdWjhOMp6Try6yv9Li93ScNExrHl7GzkqVzK6iaR7g9ZsNpVYZqlRDOhrxx88Z3bH/fFSOBNIz48Me+l9noEDuKScwa/cE2x7qSYX8bQ2mG0xmvbx0gvDaPrNVJ0oLx+IBcp9iezY3dOe3FGr22y8ZCjTMrS7heC6/VYBCEzKYxQaBpbKyhS1mZbrIS07NwHAevvYTZalKr16k1WxRZNRylKCvjTZFVtuAiJ45D4iQhDmOybIYyHPIoRAtNGgeIoiA/UmWneXWd0jmBP0GVkKZxVW2oDCzLwBYCoarnEo0pBEJItCUrFReqUluXlWR0rd0hyDLIBRQQF4IiAxODBddmGIRkusQUJkhNFE9J8oK98Zwkk2SUDMYTcgVJHuJICZkmCEOMms1onmHKkoVmG8uEdDbH1Jooqzj6NjaRlhRFhm1aOLYNlsmZpR5LjQr4mUVziiTGdBqUOiEJJ+RpRjQfUwQBQkv8JCeezo9Eq/pPHAKFLkmyFBdFrGM0WaUrp8AQFqUhSTNJHKbsHUz4vVdv8Pk3b/PanV3u7m3zxtVr7A5HeJ0WG0tdRFbgegZpqektdDCE5MLaGm6jwd40IJyNIM3JghjTz9lQmnIwxByN2dm6Ty8Df2dAPgiwtUSnJUGZoAzIdEmWgtdpcHZzBathMwt84kKTlSalaVIIhWFZGJZJrV5iCUWYFWxPfCajKYtNi07N5fySx0LNItcl2rQpUOikwDMsIqUppeLE5jG0AZsrXaIkZhoF1eJhmtPudjCkyUa3huXV0VJjFwXzqOTPqBr+7ugEvHqN7XnI9/3oR+n0GhjTmC997uv03vNBPv97v82Vl19i8/gJPva+R9kPcj7wiU/wey+/yCiakUYlMpqwWVfgdHnPmWUM0+R6f8bnrm6ThAmN+RxPZ4xGfdqeR5EkUArCYIZsdlk4fpLFk2dZWD1Oe3ETy6mAoGmRkeQp49EuQTgii+eE0z7BdIzp2KRRROKPiOfDynMXVKz/MM4qerCwSKMUKSyGkzFLqyv40xnRZAKmQFpgliaGYeDYCmxV1cILgTIMpGFSSkUiQQmwBGwdjrgV5JzsOSSZRogKWxZFc84eW2dn0Gdr+4AiLQl1ycwPMWyPaDaG2YT9w12EqDEZ9amnBaIoCHXOOJyx2PFYsAzevr2PJ6EoUuZhAVJh1qpU32gaMIlS/EmEIU3cmsdaz2HRtnj0sYvIRpObe9so22a0c5skDqh3N2h2ljFtE6fWpZSS8fgQGY+RllnhxpUgK7PKIiQ0qSlwOm1caZPM50eayAwchyLX2G2PPM8J9rb4xtf/mFdfe5MyK/kH/+3/zsWOYqXW5OETPUKrwSjO0JjM5wGbax3Onu8Ryzqff+U2n/nay5j9iE5hsJC7fGTjHCftLuXBhAe7O2S7eyyZiiSrOn3T8pCpjecrRlnCKNojC2M6jsvZJx/lzXe2yScznFJTdww8pZDSIEVjt1sI7TEKfJJpyOUb27x554DM8jDqDRzP4z3HOzzS7bHQrHN6uUY+mZIcxOSl5P484Md/+EMkJVzYbHG2Z/P+MwssdW06621Uu05xf4BnKcZRQm+5SeEYNJomlvGdKwa/K6YDea6JDmdcffcWKx2XxvIyP/3MC6RxQauzhNFb5NbOLl3LJZhFXO3f5ZRWyBiEKhmOx/zY9z/Jr376j7hw/iy72zd5kCsee+osjMbUl5Y51+3RWz+GZVmgcyJ/jNfosri0hmE75Elc0XdV9c5rmgryDH9ySBnOUaZDaZvkFCTTAzQxorQopGbYv0Fv9TSTyT71epdSacKk8gwGfsztN68wn88Z7G9hYBCmOd1mg6LWROcatESlBRiCsgSpCvJSVyGWPEegKfOErXnAYZQweucWZx9/Lz/1uOJXP/0NckOw4Bi8c3uLIJiz1PBIHIsy9EnmgodPrCIXNrl07T5mdwmDgOVaRUpqtpbo9/t4hmD9xBqX372L78f4fooZ9mkvLDMLNXa9zs72IbNZyAhNTMQp1+FMrc7L92ec3+yxN5pQVwn5JCNcWMOwmgTD+2Ra0Gk0KYsMw7ax6nVk6BPO5hh2iVNrIO2yUrvnGQA/+GM/wYu/+7uUpmbsy8qwLAtUaQKSeBoxGw05mPv8Z3/jJ3jozHluD8ck4wnnH9rkay9+gyTwyVXG3oMJqdastDt86dpVfvrP/TC/9YUvIcucTkvhSQN/njMah6w+s8Rn37hCreVwe2eKLDzSKMW1HQwbRuOQg6u3GTRMHt7LUSE0CQhTQWthEbeIWFhYIitg0TURheDeJMbRBpNZRLDbx3RdfOAwjEknKc89cpx5nOGHEff2Biyt2zzTbjGKLCwjwhkMaV/c4P58xPP2Et/7/FO8+o3XeN+FM6xsLJPEIeNeg7sPBqwuWfjknFryWFpd5NZ4G9Gd0WvV4NPfvv19V4wEirJkEoa4uqDWW2Z45zZJlhIFMxa6bWzX5f9g7s2CJMvOw7zvnLvfm3vtS1dXV1Xv2yw9gxlgAAxWgoQoyCABkCK4mJZJmqRlm7TNkBWSLPmBphQKW1ZIECmGuUALIS4AIRIrgVmAWbunZ7pnel+qqmuvzMr97svxQ44j/MBh0G/zkJEZkXlyi/z/vPc///9909USSR7j2DX85j41S1IuuSiliKKIo6eP88ipOYZKR5EzMVdnvOygyYwzs5NM10pUSt5I75X4aLqB6TnopsuoED9qMFKZGimtVEqR5QiZkcsCLI08iVFxShJnCKwRjccfIhRIp4ouQBkaqhjBSJMwYtjt0uu2SdIIWUAiRtNvhWUiJdiahSV1MCWC0fsQQqApSU5BZkj0fOQocL1RwTEsUjbvrnKrlbM0WaJi6VRMjZbfxRSKKEmY8iyOznsjGpLh0h6EGIDyfaqWQc21STUdzwTXK5NqFr1Wl0GQkgdDaoZFkhR04gilQ5ak7HX7+HmBUjlZUrDdD0gTyZyjcfLYIirL0XKTaDigMjZJnCYo3SEdNtEMgTTL5EmKUAUlS8P2qkhdoDujjkQY7YwUWY7TmKJAYTkOqYI8SYlSRVLkJCqlOxgyTBV77T4PP/Y4/VzynWevEIQxm/sHbG4+wLQURlFQcQUVlSKEoiRzLFPDtyVNodPKM/pZTrMb04oCDFtSBDFVw3ybbagRJAmDJMcsewy7ATIY0ih5aHmBKw0mnTJWXiB1wUStQqQpNFMgFERFQTwc0u336LQHKCmJioJhFOFlCpOC2XIJXRgUeU5FNzFy0ESGngb0/JR+WjDmOMSaztAf4HomuSrootjwffbDmG4ckYgUfcymMldCHy8Y6AWRVryNb3vn//t3RxLIYkpTBo//0Ic48/gHmJsoc/H7/5nW1i0cQyMOct7Y2Ofm+ibb6ZCddsDS0SX2YjCKiFSENCPJj/ztTyPDNmfPnWBlqoLoDUZ8O13n7k6TRmWEDicGuzxBWgiK1CfNJEKziLOYOBqQqoJg0CKMBsR+gIaGyASGYRAHXSzPJEkDYj/AbzbRnAbN9ZsUpkdvd5t+Zxs/GJCnio1794j9PgyHCA1knGDVylimgyNcHDmi8ApdA9PEdq2RPlxTaFmBkWuYhj6aXejlzDsak7ZBngzYuHaVj3/kCZ44dwS95jIpBYUuOTRm8tjxBk59gnLJJU4Vtx/ssT/ok6RDSjWLN/cHZFGMNB322yHbmwcEQUTUa+E4Dj4Zc+NVsr7PQaaxP0i4vbHPmNQplaoc8gwqaUgnCBnaHhPLU8QKgmDICzfv4yuNWq1CrhRjiyfRjTKDdpNOZ49+ax80G3Sw7RIUkIvR2LYmQJEhTYfFEydod3y0IibMcrQ4IYlDVJLx1o01wjTh3Mll7uy3+fV/8dtEwx1UmrBzaxWZFVR1SaxyPvLoURqVMvPjFm8GCT4RaZpTqIiOn7MVhuyFI2LyMJcMCBCejVu1KHkGB72Ag96A448c5c6b2wxyODY1jl0qUapWyLUUt2YSyxx9vEbgp6QhrHf6NHt9krjH5naTIOgwiGMOgpyoKCgMhdRzbrd73Om0Wd3rgqfhOQJNJRxdmmDGM5FxCPlIs/bMzV3eXN9gU6XsyJA77X3upH06cUB9KiaftmDcJq07DM2A6lSZqYVJGpPvcg2ZYbq0dyOuvPoci0fXeeXSDRqOyZ+/+Aw/98s/zp9/+UswPo4h4OmFw3z5u89T+sQTZK++Snn5PN2m4lt/9l3s2QV+/ud/BEWF+9fe4PLrN2j6Pvc3dpiqVlCGg8hzZKmO0AWaMsiyAsNIUQjCfotCSLIihBxIBiglMDQTWSRkSEzDJvS7RIMYs1oHJYn7HTIFudqgSDUKzaLQQ27euAQHPVSW4OcZRZ6hl108TWJoGkoX6KZJEYSYmj7Sf2cxujBHA0oCYplRpDkRGrWpCcqWQIsTXrm9RWiaPPf918k9h7GKyw+fP0QqdcxKnatX3yIKc7rdkF4QMzc1RkVUqNVdtno+cyULdJ3OQZuqiDlyaoF0GOIakmzQZXl2liJPODqhc3K+hm3qCEtn9twScRSgywoly0XIiF/80Y9y5fptnjx/lNc3mszUHG5efJ6jZx9idm4RXaX4nT5KhBDFxFlAlseYQiezE0SgcGrjSG00HKQS0MyY8x/8QTbXNxGGwkOCoaFpNpWqR3LlJp/96V8gM1y+9Fu/yQePz2G5Jt39XXY6ARs7uwSDNkeW5vCTAr9IqWk27A64PNym6BRsdSN67T6Nuosz5fH+E0v8u29fotEokVmCmZk6eaJI7RIPX1giEhZnSgFhouGVNPrtBD9SNLdajNUrmGUPv97g9uv3UPEA3dJJi5zTc9OcmResbbe41hzi2gmnVo7Q6fTp9Ye8dv0uKoHpiTrVWgnLc7D9hKYycJenWOt02b+7zqOPnmQ92eQbl69ycrKODFIGmk6W5hDm5KZLWETohmLWmObK7Tvosc5qq8+ZsXf5FmEchwRpzowrMc0KvdinEDo1V7LaGw3biGiIIkEzJGEU8GCtyaMPH0PHRRUdbuwdUNrp4n3y/XQwOfO+jzI9NUH3IODW2gMOOn2CICK1bcqlCgDybZOuGmmAkJrCMDzScIhQBVmcoFkauRQYho3SFGmooYoAp1zFj3yyokAJnSjsIylBkVDIHL+ZIKOYfp4g8wwlINMllpBIwwSVoksBhY2hSUZ9nQVSKYoiwdIgKQR6WtBOC8wiZb3TIRc6E47GmKtRaVQwlKDfP8AuTZPoFhvtiKqMSZRGFg2YGnNIkCRZRqVmcKsZ4g9jVBIiTYsUQRzDqZkJHtzfoGJ6bIYdykKQqBw9grnJBnkMb641ybIUqSsO1zxSw6TiaOgiYbPZRWgFhe6SiJgoF2hJimaapFGCkBlpIvCHfTzPwTJ1RC7JihS7VEYlOZnIELmOYUqQFoUocCsVBt2DUUehVEhGFuOf/OwP41ZqfOV7rzBR82hUSqSqoJVldOOcIExJPJ2iyOmnCb2o4LAwiIuM+KDFvd3+27s9BWEEtRzGxypc2rjP5HQNS0jCQlHxbApXMTlVo9f3MYRJYGUjUnPgE2UZRBVEDqqIMEyDwgARJjjKJCsyLNMgSQtaYUZSFMRhxEyUoQDXsfjw6WNcu7+B65aYalRIlWDQjzCrJQovwctdskEfspwEqJklqlLHKtUo6ZJ797cYDhI8xnH3BV3hs2yNc+fKJnVp0ml3iIt3OU/ATzLOHj/NjTu3Wd/bJJiyqCQKwQSp1+DX/puf5P/6V/8B1Y2JMsXEeJmXv/k1Hn/8AleuXWJpeZkn5xuUnCr/+De/yOefeogvXNzm0HyDea3gdnPA7OQYWxurTDtzFI1pCjUSYZS9yogInCYUuSDsblLoNoZp4tZLmLXD5Jqg5hmEgyFtGdDvHZD3t5FytM2VKsVwkNHdeoDjlVlf22Zl0qG1u03JKpEXgrzIcUwHx3JAU2gIDCS6TBGWRVbkoy7EQiByKOKcTEpiTWd9Yx0VDFHCoKbHWFInMgo6+z0efewEDx17nKYquPnWPY5M19nbaXOoWiGsevS6A8Ioxx/2ubgXchApPE2RBgkxGqJQPPae0zx0ZJqbr91mvz/Ea5SYa3gkhUs/iHj18k1MXaPqGLT9iJ1UwyXF6uwT1mtc3+2ztd5kb3/IzdurZEXKysmzCFOSxjGF36M/CClZGuXJBUzTIB12MWyPLA8IO20CTUOnwCpXSFMb1yuTJQXv/fAP8JXf+QJFGqHZHjutPWbkMgcq4/q1r5ClOUq3yXSDWAkCFbAy5nJ2+Qy50rF1jf1uhobG6t4uk/Oz/O5//Dp2FiOFialrkCVIZ5zTx5Z45qXrSGGhgozcj/jYw4/ychziYTO8ehPLNDi+tMDlW1sMd9tEuWR5SUPqFWaLnKjSYObYIvtv+ui6xFYmr9zaxDI01rs9LEzCOGJ9p4V0JCXL4/EPPMaFJx8l7B4gNJdur0WmCU4cm+V7N+5TiSRxO+CYNsZuPcX3Ien7ONUaJ2bHuXGtxf7aBnN2mQd3d5H9Ph9bWuLC+BjH5qfx6jZxmgLP/KXx965IAg3b4Kvf+B4fe/r93H3zIjPmDLEZcK/ZYuKNF7hnXuBX/9HfZ+vi8yRCEakMvd4gTbp88md+gv071xnrSe5tH0Cnw5tdxXRyn9vPvEGrVgY9YTdsUzn5EAKLQkmStxHlo/3pgizPIQenNk6uJGaphqYJ/FTx3J/8Pufe8zirb7xCmJpMzBwiixSbnX06zQ5jnoHSbDRN8pVvPsePPnWWIEzwjJGWOk1TdEPDcXWUrnBsFz2LkbqBJjVMIUilJExT8niIRKNX5BiaoNltsdXvY+c6RRDRNAVjZky1ZjNzuMGpw4f4you3MMgp5wlfW2/zkXPL3HuwyyDJONjt0o5TSpbFyrRHutfDLyQ9JcjiBFUIlhdnuXzvAStLEzxkzpFicP3BNq4jOdjtMF8rURgGW02fjXbKuK3QpMnkxARZkbO51eXckXlyleEIaA59Lt1apz7ZQDhVPJVhmzlJnlMMuiRSYUmTLEuwpY2fR4g8wR2fAiFAaCRJAoBVqYEGmZTsHfQ5c+4UJx96nJtvvk4xNY3c3eW1zV2u5gVRkPDpD51nEESYSQ6mZLOXcPHmbQ6GPp95zxmefuQ4f/BH30Aoi4pWECowBBCOGsEeObFCrV7mxbv3ODY2zczEDM7+HttbLQwTBqlgedzmhdduYboFNcultbvHzKEauVmjYth4jTJVzQC7jGYazHqKTIXIzQNyo8DVdPwwoGp5bPk+axvbTFTGkJqJdG1EWqNIM+pSoO320FTCyeMraIbO/ESZm0EPETr09vbpNyr86A9+gN/+7T+iYVhMnzhEf3+XXlbgNsa40Qvo73VZbb7LeQK257C316Rc9vBKDtPFOEHcxZ9IUKnCUBq/9Sff5dM/8BSl/gHd2Ob8/Didfp+dzSZKwkZ7n77fx9F0nnvpdX7pQ6e5v/4cQTBEOgKzsNDzEABd08izUetpjiKPgtH5dxZTrtQwhCA3ykRZRNLe4ljDYuuN1xhu7ZKrgh1h8OAgQCU9Fj2TgTTZ7vSpEDFjpQjdRFMJwrHRwwRDN7AsA0c3UQVoeYYw9BFPv8hReoH+9qSijkZapFhpDFJnc+Cji1HVPFM55Dl+bLFUabB0aJ6Ldx7Q3N6gapmUqh6DXpeLDzZZ8jyEECS6wpUGUioKBGNjHvWwoCIhq3nkcUTDEQzaCbm0eDAMCfo9bKBhlWG8YKcfkSTgmTBbNtBkxmLVochyFuarPHN1Ew5PE8aKXGYkGexvbrPxYImJxgFapY5hWCSDYMRhyEOkMToNCooMTWVoTgmpQIkC8baOjCKjKAxMXRLmgkQoFlZOchDkHHRDXEuSSYltCEhgslSQ6B7S0DAJ2I4yVrfb7O/sEyQxURIzVq3g6CZZFpMJDQGUpEboD0jSFLdWY7JqUREWsxN1pKOPahRxQjQscMdLI7hKnKCVc2zHI+wX5PnIKxGpAkN30CwX3dCZME22BglBCIUuqVgCJXX8XoKW5JRUTqc1gExS0SJkmmBrktfW2nywVB4xEHVBxbJY3wnxjkhct0SpZhM2dyisgrmxGoYEYZhMVMtovs/qdsBWs0mn06U58CmZ7zxK/K5IAgU6P/v5H+Y3f+8r2FmbQBk8tDJFeWICEaTcvr+Ke/82X/jHz/DZz3+epz/5MR4az3j5G99j+0GLiZqiYtsYYzrbW03m3D7/5pn7/P1f+xW++h9/H7NSYdDr8dKlq8wsnaCuG0iVkKUxaBKVhaRJjO14JGlEnmtka8/hTdZIgwOu3l1DmB69SOEnIWZ0m/b+DrWyw5/dSGn7ORVTY2G2yuOnz0KhgWmhxcNR15aek6YxSpmYmokwdTzDQqmc3LZQuUI3dOw4JcwgjhTDt208WdAjl6BywcyEy3itRNlyqFo6d1fXWXvQpTZZZsJOEIbOfNVC7h+wOm3y8GyF9S2TnlJ4pkQbpKSy4NBEg1ONCrozwp7dunmfXqDY3t7GqpbQ8py5Q/NkccKVBz4VGTK5eJgbtzvULYsjMxOEecpDpxd59sYO5xc8qgy4cS9mpjaGrjSEqbj65iWOHZlkbmGcQXN3xALMc1SmELpBkqZQgF4ujVDpYQq6ohAxhhRomk6a5rS1MnkS89EPPUW73Wb14CqD9ibrvRjL83jqiXMEfkyucqJBhGMK/smfPscx1+HM8Tn2HY2+a9Nth9RqDd771MO8+v3L5AjyOOfEyjQzs+PoFYfydA2FZKXu8ch73w+GxoM3vkdvv83CqQlOLFYZxAmHZ0vsdHy8+TrVakbe9knaMWW7T6zDqw+2aLcLrLLGo+fn8PU9zr9vHHeqgWFLSrZFpzPyFbTyXXbbW/SaIdEB/I2nH+HTF2b5ixcuYaqEB/shQbRGydtHbYxz5iPH2TGavPzNPcypcZbmDoNpUq54fOrp95FkF/gf/9nvME7KhdNHWJ4/yqkzJ3n6+//gL42/d0USSNKUi2++Sdy7zs//1Gf45p9/lzu7bTbXtzl8fIFDnQPe9/gp5vsrpL5P961X+Y3VNo88cYEFM2TjQcCUGfD1tYDpsqB3MCQJ+/zxX7zC4vQYm80BueNy0O1y8/YDTh0/gxzZ70YiT6mTFwFhmkEScOzkWaK5KoNOm9f++OuE/QDLg8aYx0x5mluvvU6ew537HVq5xMgUiSw4OrkCMqJQFnkWU+QpOQV5blJyq9i2hcoDSkYJzdLfPjXJUFqOhkkhfITuYBDT7mdk5FRKFtkgJNbhyMw4mmXz0LFD/P7XX2XgZ8y4irvbBxw0qtjpPqGmGHccav0ed4VivmHi9WM0oTh1eIqEjEJqvLy+zfrOkDiOWZmv0RpqPHJolkbNIo0jdGETFTmTpmJudoGxxcURYefQNHd2d3n0xCGu7Q0J9pokcYE+btEeDBmfGiPKchxnHE9XrG1tMz0zRxyGuLoi8ttYpkeW5IT9Fp5XxqROFveRQqdIM6IgwDMdhBJkWcTDp5c5evwMUZzSHvRI93WGgxAVJswfGqOXKBJpk4chL7xwmS1/yCM1h5s7LaJrKeeWZtlt9kmzmKEfcGJljldfvI6p55w+e4gnTh0ls6BhVtDMA3rDgOmpBgdJSkOmRM0medAnuBuzXjEJ4j7jC5O4doRpe6jOkGuDDh8+PI5iAltKtP6QlXmLxiGTiSMmc8Zh7LIHeUpuWQx7OnlHcHdrnSc/cJbFaY9BGrGz2uFaf4NK+TgLjQbf2WoRpRkbrR7zQhGvBezcHSOqaLzn6BR31loYlbt8/ic+xPsf+gD9eMBOR/H0yjif/OgFsG1SoWjMHn/H+HtX9AkIJJoWkJfKaJMrXO+nxP6Qsq5wlUFuSopqmR/6qV9As3U2+wX0e2zfuEt9fppBoDj/+Dn8fhOVF9Qcm+WZKer4PHTuPL0wZ3FqHgtJmiQIodBMDU3XIS9QOSOIiRRU62M044zE9Mj7myAljmMR54IX7myws9mkkBqWytAMmHJ1jLLGbNlC13VEro8+TwF5JnA1C9OzsD0bx5K4loshdUQx6oATQlAogaYr8nwEMJFCJ09CwjCkatZBkxhSYHoWj5xapFbyaPeGaFlMkAlKCgadAfOeQdCJ6PdCdEdjwtCZrLgcGq8yP1HFF4JemNPq+VhxRE0UaKpgtx2QRD7VkkeQFewOAzSpaPb75CgiISl5gpprMghStGz0Ha+v7zE1UaFWHglD8XNkWiCLnN3mNiIcsr62TRwcIHRBFMSjWkCREwcdbNslERlpOkATGnGWIqVOXoRQCJJsxNE/eeYxEmyub2xCIcjiiEGU4ZQdxupV+oOAwWDA2naXvf4BSbeHY1rMz03T7AZ0/FFdpiQlg4GPoyyEVtBwLR5eXiaXBkUuyEWOoWnoukZW8QhERl4UDOOQVAlCU1BXKUkUolzJbKOMJkEVGmmkeFsvgykFwnGYWC5RHS8TDDLCwEb4FoNAsrcXU3RMOvsB/a2A5lZMsx9haSZmWcPTYCPoojSB61iUTRNTgwM/QUUR/l5nNOI9XiYKQ9JhmyfOPUYsQMsVphnx8Q+fxak0EF4Zxylhus5+DkJAAAAgAElEQVQ7xt+74khAioIFp8bViTH+5e/+J37kMz/O7uvf4XlNx9Ftrq6n3Hj1G/zT9/8XYDk0jJyzH/0g1+5v8X///ld5/w99hO9cXOM3/odP8btfvszigsft5pDXLt7k2Y2Yf/i5T/Gfv/Esf/OTT7F4ZBolDJSWY+Y5iQDHcnDHVlhdu05/7YBGpYS6e5vvffNFyp5DK1T0o4Clikc/zXAsg9QwmPYSqhML3Nu8ix8rDEvHtG3SKCBO+xhSwyiXQSU4holpKizLI89TpCwQ0kDoBlkBQtPJadHq+ZBlbPR9et2AhXEXMzMZX3KYn5rgi995DT3OqZkGUZ4wzFMcQ6ciNFb9gkePjKGh8dbdJrBH2bEJTI1x28X0u7SHfYJA8b6zR3HtHRalRqPqEuUG3bRgplJBFIqtdshup8eYmXHqUIMSNiqO2euGzEy5rLczZCH52hu7fPZDR7h4u0UJWDs4wDUNgmjIIAI76mKWJ6hU6oTi1sjjIAWpcLBsi0EUUAhJrLuoZEAQ9jDsOmF3l7RI8bwxvn/pEkIljFUmubR2nyRLWZ6dYLMXs7o14He++hzHp8tMupJS7JMZgufv7vPJDz7Mydk5nr96FTuVDIuc8mtvULdMlsYqDJRgr6jB6iqlqslOs01hKCZck62KSZFFWN2CclWnqOjYkyWGJY2KUSft+vTqDR70OtQdk6r0Rnr5JAFLcOJCmV5P0trwiVSIoZrkmkatVGGq4aHslDzoMunZ3H/hFreeD6gaBrE0MHRoHCo4NmkzdniC3rjPpOaS+wmbmzuEr99iWqX86Affy2c+d4JIUyRJwsKch2FPs5jnDAc9Bs1tCAu0UpmDdvKO8feuSAJCM+jkCrfi8fM/9ln8oEv9I3+bD8kW//u/+wY/9smPcf1r2whT5+b126wsLXNnr8VkxcAIK3RW93ns9ALfuvSAiggZ9BTLk5Mcn6qhy5z17TbTM3XqkxWu3N1jellhYqDbNpKQ9dV79PfXWDx+HqWnNHdv4DRbpOhMzs9x5fYLWF6ZAovUz5iaOYTT63EQ9bh9bx2EYHlqHNPUMU1JEqdkfoph2lAInKIgVj5lcxyJQGo6wtQRuUYmBTqCnALdspFJix4GehIjREwiyxxfmGJhrsGfvX4fpGBYMYgR6JmNK6DlF5S0jMJXrLk2Zw+NE+90MQud/TgjCXP0TLGgWex2M6xCozvsYXglpiarvHVzg6qIWY1MrDOnGGtMsXblCjMlHV+MMTE7ze3tA7ICLCNhYmqaF968S78z5ONHa5AV7Pd61GfGKeIAX+jUKza6O4ZmGnR2diBTkKZv10RMtDCgN8woDpokho1KfBzTRZkjgpNlWBBDmvo0GhWC7pA7u02cksZsyeMP/vQSD5pdut0uVhahT9jUx+fxB13qmGRZn437WzzyyFEKYaPLmIOujy0KwrCDsCUVlXP78vPU63W2mzEP2hln33sUp1QiVglpEHLQ6iHGdGy94PzsFK9e3cEqbAwlGQ56YAm2RYjnwGq7x9OLK/R9nwd3Q/xOBkIjDhNKtkGpVGG2bCEHOb39Dc6uzMBQcn93j2EfhK7hRzmmbrC32eWxiQVu+j3CIKUVNxmfbnD48BJOYfL5D72PO1FGmYwJXTK3fJRcScI4G01hRoq91oCrb73J5z77k7xx850lYO+KJLC936HT6vHP/sG/xO+2QJjcePlPef3WfT56ep5w9wYf/pmfYxAlfPDx4/zel76Op+u80exyZnacN6+8hanmEHaJhck6q+2cfNimXKox3hin07xPczdgY/2AbpTx4Q8FaNImywpsAfOHT5DNzdO8+Qa6ZjLQFSVnhmNndL71zMvMzMxyaa1N3YkRheS1O2us1D1aoUPDyegOYdor4XoWWjFi7Pk6aCUPW2YYjoMlcizLAnNUFVeGThEmGIZJFsVgGgzvDwilQ1Ur8IRG4Xkcm6wTK5217X12VlsIEkwlqY6VafV8EiWI0wzb0LBcxf5Wn2/tdfnp9x7j1laT9WaALQsOQtiOBA45tbpNFGV0e0POHJ6kO4y4nytmjJC8v8tBaDM3N0dn2MfJCp574QrN7V1mGg4Pn3uIqxsBJ8ZMnvj4k9zY7vDvn72GSiXn5nJ01yOMc4Tr0RuE9LpdHDeh01lnolIjkzlZ+wDHcciTIdb0FE7JonfQH5mK0xTHMSDPMD2XIktZ3Wxh6wlTlRqvbG5gDkJu3bpFGCRoAiYOjXP+1BKXbm8zW52iHfR4cmWOVBXsr23wQxeO4VqSJx8/zZdevsulV24wMzvOlCsht0AFbA977LfbDJ9Y4FIqcExB1jwgDGO8pMxjh+dIA8n9N/coigJT16hUSiAyqjMTOHWb23stHhokPNhr8tTJ4+xttXG1DFFIEqnjOS5VW2eY5ax3cnbe3KVsSm5strDTFOm5XDi9wsR0ldCHc+cPMbZfJ/ATDh9ZYH5xhfX9LTrNHSynwtGS5OSRFTIp6MaKZHDA9voGt+7c5TsvXuH662/xhX/6KzjVMf7ou3/8jvH3rqgJFHnKz/y3v8a//eLvkoQJvUGHtb0uC8uH8QTopRLDfpO8u0+iJDIaMkwySobBYBCS5SENz2b//i5apcTCeINjhyY5vDCHXa/QOghwJxsM2x1SS6PI4tH2U+KjmTaoHJEVWJ6LkilFp8X4ymkGyqA1HLDgmPjtbXqdDiIf0FAw7PVptzoMspyqZ5BrAqOQFPJtTRkadpGSxj7KsEBzQMmRTksINGmgTI0g9knymCLMydKcQmaoIifVFTWzwLEtelGPm2sdtCJBFmKkBIsT9CQBlSNRxGIENvVMqOs694cpJ2YnKLkOu2lOmCX0+wN0ZzQhmecp0tKwXIvtnk9F5SQZ9KOM7V4f17GYK9tU9BHfL1WSMVej5FVwioCl5Rk0WxKGimGQQ5aD0rGMURem74eUtILF8QqJkJi2Q6EJVBAiNBNpWmheGVFAFhdYmqTIcqTKkKaLkgaaYSGlxMpj4liw2+8TRymdpCBWgkJAKgRVz8GSJvvBAKnp1EpV7vSG+HGCUilTYy5PPHYGd6zB9Wv3MBQEScpWJyHMChanp6mXXaZdiaNbhP0IXWVsrx8QBAmO0GnUKgz8BNuAKFakUU6apmipQe9gCIOEkumy1j6gFwxRJY3UH6ApyZ6fIoWinyQMw4Re6LPVbNJqtYj9AUaW0isKbM3g+PIU01OzTMzUyKWO8EosnDjG7NJx/DynUq5x9ugyE5MNJhoTDKOQIIyJBm167R6X37rJt1+5xvr6OnGSUp6YphAG3a3Nd4y/d8WRQNnS+OJv/BO6loXz2Z/iD3/j10h0nahtM1l3aW3sIqICrz6LVXH4xf/p7/Hvf++LeHqAzGN0BNPLR3HubnHhY5/j/lsvkwmTesXl+WcvstHJ+MjRGqsipmG6gEAVGnHep6TVMQwX09TJgwZ79++yeOJRvLFJtq6/zrDTp/hgjR/TH+F7V26zP4jQbYsl20CPmiyvHCGLBJamk+YFprAohKKQgjgKcWWBqZmYtjkqCJo6Qgh0y6TQBDJIUCqjvbeJ60qioUXP73N6pkrbjxnEIakfMxwOMQwDT1MYpqQb5uSahpFBlOdYgCMzUsPEtnSu39xiuDjLfrtPEA0xK3VkFtKwqgyHMaeXZlmYrKASjbMzZcqeg6UZZHmMjeCte+vodonzCwuYUlFTDeamJnlrbZvNjs9Tjxxic7OPVy4xaUmOjXm4poEwBQmKIuxQnlngE5/+MFdubTHemMKZsdENC5H2kO4YWpjilct02/uYjkUQdqlU60RBjFvy0DSJYVk0hz00Q6dRrmO6NhWlj5KOMHA0wenZSfpFQTZIOXA7lBvTdLbW6Dg6um3xd3/5Z/DTlH/7lRcIez0mKx6KhFwW7KfwvtoCjI1xcvkEW90e0bBHK4B09QHbtTofODlPveJxJB9JX+I4xKu6tHoJjxx1ubfZxnM1aiWLW7tbpGnKasnDL7mkWYGUIyvQwTDG10J6ux2OVS2u78cMVcrS3ATdKGHx8CFOnj6FLgW6ljJen2Y8B023CPOIyUoJ6U6S5gXCdDGBb7/+LKenV7j8xg2+/eo1Du7eZnenw0RVp+7qNKbmSBPFhWnJN94h/v46VuJDwO8D00AB/JZS6l8IIf5X4L8Gmm8/9H9RSn3t7TV/D/ivGI3h/F2l1Df/qteojU2ytnqPnVTy0sWrDA/6XNvZplyucHZ5BmFo3N0dsrvxJ5SNjH0/5/xjj+M07/DVSxscPn2Um509Zi7M8u3rv8dc9Qz9tTZR74BJV+fIoRm2V1eZXVxidm6CJIkpkFS8SXIEhqYhzRJRGrBw9hh56LN14xJnz5/m0p0HfOtbb7F47ASf/oHHuLsXUURdclIeK5c46PXRTIc0U2RFgWtb2M44YZqhqQK73sA05EiUKTIMZYDK0ZIUVwhSx0BLJXMLc2xsbDBQQ9wiQXdsFqemeebFN/AzRZhJTjQs9v2IIBdMe5JBopHFElvklGwdzbAYBgW9wOfs3Az3Vzd57KFjOImPOznBiy+/QVDkfPKDj3Jna4Nbr+0gZIptO0iRUSiLKE0o4owsTlgeryNlwTD0UbrOnz5/hVutAX/rvStsHkAoDWTc4/Bknd3BkNOeR5CkHFuY48n3fgzl1rh0axM17JNrEUyXkVJiOFVEoaHpgijsUa5V0DSTHIEix5IK0x1HKFCa4j0PH8OyKkjTZuvqWzx88hBb62vcWdsmTDOOnlrimVdukcYhXmOBc0fHObX0NE88eoaxmTn+4W99mSsXL/Pw8WXOHTlEu9tl72BIQsZcVdAetzi6eBzNqjP4/mXSoM8vfvpv8H++cIcF0yQfX+Drz77EzOwCn/nQI2zv9fnW9TUuTLpEccaplXG6tsEgDXlCVHlubY+zj83BE1O0unvU1wVxEuCaEaZUzJ0do1ap8pmzS/iDAo0uJddj4cR5CmWSMnJWuGWbcddEGDZRPs4wCvFbPfIsJRg84A+fv0R3/y7Pz1ymfXGL9c0D/tEv/CBfu3wPV0XUyyV0s8Ld+w/4xpV77xh/f53TgQz4VaXUSeAJ4JeEEKfevu//UEo99Pbl/00Ap4AfA04DnwD+tRBC+6teoN3p0O33CP2EG9evIXTBmKHRTwqEpjPo9fn4UxcoV2yiLOPUQ+cImzsor8rRlRmqRzz6WciYV0HXyzxoKZp795CGTea47DaH3NkfUHerjJVqI7qxlCBHH18IQYHAc+sMehlF/ThRa5f68eOoFBq1Eq6IGDtxihOHZ9EMkzHLIVaCOEroRxFZWiBUTkSO0nVs0yBVYEoDpEAVBSQRRZaM3HtSUAiFJgTStjAMnVrJpexJDHQMzyURbxuS31aHKU0fbaXlKbapUzcNLFvD0kaqsXZaQBZiSkEzCuhnitdW93BLJTTDoDAlF07OMTvp0No5QIsyVKBxkCq6ic5uNJJeqjilZOsYhkY0HOBHMakhmbLA8wzOnTpCmIzce7f3e2x1Q+ICGtOTeFUPt+4hPJc3bqzS3j4gigsMt0GejSYyo7fBIXmeIy0HwzJBSfI0QeQWXqU+QqkphcrBVwYpBYiUiXKJ+flZTi7NsnxogomSg6HpOIZG1TZG48iOx3vec4763AKdJGP79luIrOCgE6A0wDZolGwWXYvZhomyNURhkKiEuNcl388wzBrH5ud58qGTXLy+jmvqRP6Qqflpzh5boOZZVKtlHC2nmQzohD6Hax5BFNLZa+L3YqRMKFc9Ys/A8TyiMMU0daShkwsNy61zeHmWE2cusHD0JLpRAkJso2C6VqJc8yhMj1QIhoOQbrtDu7nBxvptvn3xMldevMh0lhPGKVMT00jLYDdKOLw4SzfMWVw+wjBOWd3bofRXhPpfx0q8A+y8fXsghLgBzP0VSz4F/IFSKgZWhRB3gceBl95pQR6FVJwa3WHMxRef5bjrM7UwyzGvjJZrLC2f4PLNmzz+5OMc7B0gcsXYkz9M1LnPeyq3uHLjPvOHV5hyy+i55GC4y3DhKM2te3iVCfqtkCdPLvDKgwHffeUa//N/N0+pPIFhmQgpSRPI6ZOIBNPIeXDxz3jsEz/Kc3/8ZTQEpmtAHvCd52+xMOfywpv3+cC5FSYmp7n7YJ/FeoWu32MibmAGQ+Kyh7RtVBRTSDE6ZI+HWNUqpAmWU0KpAlNqlFyPPClQJiMwJxLP0Ljd6ZP2hwjLYlwqojBgP8xxpEJL4XYnwhIaQkCoJHGSM6n7lL0Kk5Mltre7NMyC/nDIjf0B+f6Qv/ODT9FOUv7guauMeyVW2z0ajsmknrHbSzBFRjOSlCScX5yim8WUihQPiW15dA+t8NMXSjSbA9IooSQyOvst+n5IjuLlS29QG5/i1OmjPPPsNfZ2tqhVGxyamCHQHEplGyU0TNvC73fQTQNHKxMnCXkyxLRN3OqIo6+KmHK5TJIlqFRnt9cjiHL0sobUNT74/ieYLJuUShVyy+Dk8gTvOTZBtV5hfH6OwnD4T89epr2zy7HJMSYrMTvtDl1hQZYyOT2GbZrkhkvQC6kfnub6jXXGJRR1nWs3rvPJT36EHX/IjHyAtHQe9IYk0qByaIxHzszTaq/TN1IOAnhqZYYfOH+cf/WlZ3H9AbeuvUEjXmKQHTBbO0RJExySdZZXpqk3GrimQa1WIlYaue6h6Ro1z8Bq1Mg1nRSTVj8kGjTpJzF333qTrf02r93aoT/w6WxuUauPsTMQmHnM7ENH+MEljRdvb/Bj732Iyy/d4OTJk1y5epnvfPO7zM1MAX+5iuz/V01ACLEIPAy8ArwP+GUhxE8BlxgdLXTeThAv/3+WbfJXJw28ksPykSXSVshP/Jd/i87qTe5tHDAoFMtlg/Mf+AS//qu/wvVLb3HmxByyVGXpUMCX/vCrNPfa/Ozf+Rxff+4un/v4DNe/ewnddrGNgsnaLOsba9y510LYdXbvXBthwvIUw7BBjvwBGjFxOCALFf2uj2maWGPLOGWTKAkYrzq88OpNHr3g8eqzV/jEkw+zOFnGrM3Q8yO+/fINlqqSyVqNcKAxxgyG0HBsG1SGH45mFvLcRSYpiR3jYIMnSMMUx3UoUFARTJcWaaRziHvXaT7Ywp8ex0xjNNVmo+NzQEqt4pH3fEJRUHFt6pqOrRX0Eo1jh+tEcYpdMgjDjEdOLnFqtsHMdJXf+tJz5FbO0swks40qJxbHEaKgFyQ0yj6R0im7FtvNLoWhwXDIt282eXR5kuDeAz7xQ4/wF6/fo7XX57EzR/nu1VvYpkOmRWjofP7HP0FUGDx36Tp3b9yhbFs40iHKU7x0SLPTR8syNBSmY2NLjSjqoNIMW9ewDIvUH2JXqiP2YJYhpcby0gqXXnqGNMlZmZog7m1jioQTpw5zoDyOHF7BsCSO5XJz/T7feukady99mSjLKTSHhq2oT05x8pjN4sJh/CBhQwuIpSKNQux6hde+dY3a4D718TpFZnGnv87tt3ZQrkkkIpJwSFr4vHozZpgpdrZb1IwqKhxQyxUy1bh8vwlRiDFfpxsEvKfkcndryPJ0yJ1OxA88/QiTc4u0goI/+fb32P/WSxyvWvjdIU9/4n1MjU2xt9/k3uo9hoOUZrKHGjiEqaDZaWGQMWh12R0kfOo9C3zl8i6aJsiaCufoOEfGF2i9+jw7/YTHn3oEd2yBf/7rv85UVWeuVnnH+PtrJwEhRAn4Y+C/V0r1hRBfAP43QL19/c+BnwXEX7Jc/SXP93PAzwGMVUuk1TIfOLLCS1/9Mok0mPNGP5KX37zH1PLLpI7kbhpw3PT4m48vcXdvm+HWAzzT4j/869/lp37pF9FNh7I/YGxxEa8IePPuFjfubOPqJvVKma6fMFvVoFAokRMHIeXxaQ7W36LklElMiS4SlKuzsdHELhnoBahk1FnX2t9H6AYf+dSn+Ddf/HMunAn5/ovXmDBhKExa/S5KaXj/D3PvGSzJYV/3/jr39OR0587NOW1e7C6ABbBIBEkQFAlmiRJFihItlfxoSXbpyZat5DIl2XzPlkpPVqAo0hQFmmAmCIIIIoDFYhE3x7u7N6e5d3Ls6fw+XPiV6tn086dX6E8zPTPVVV19/vNP55xUBCMSI51N0WzWsbsOhqYTBAJiJIpgWSipJE7goao6geciAEoAgSJjaCrZ/n4y6V7UxRU6lk9HEOhxurz78BwP3j5HMzGCromEWxU6nSqNeouFW5so4RAJQ4euRbXr8cD73sPawhKltUWisoeoijQqTS5eWyMSj4ImMpCIYUphRhSXYrlENixQrLfYLjW5azLD5ZUie/vjPPvSeYZzKQ7sncAKBxw/PENhfY1b203u2tvLWtHk5OnTxHFISZDIZhkbztIxXeZyg2wtXiMci+P7Pn3pFJ4q4LRsMvlBHMfBsrqEE7HdMgENBA9BkMnkhzh+9z1c295iZCCPJ6oIdoDpd2iv3eJP/+qrDI0Mk85luWdvPz//rtt5VrG5vrhJo+1yeK6fZDRKtQsvr9botmtEBjRsH7y2x9bVy+ScBpW2gCdBf0LB2gkoyjaDkSiJkSxdO4K8uslwpJdkOsU1o8VOYRNB0VivVNk/mWS74VAPhQkHPoqq8aPzN5jqyRBoaVbXLvKt9mVS8ZtcvLxKrVnjA3fMsH98kr//0Yu89qM3abg1FsouZqvJscke0sMxFrZc1kslIk6bwUyYWjqJYVhksyki/hoRRebo/gmO7T/CmeY6v/CRD5DORMh22qyurvPInhRXS0200E+G+v9SEBAEQXkrAPx9EATfBgiCYPsfff4F4AdvvV0HBv/RzweAzf8uKgTBXwN/DTDUlw3qbcgkgW6HesumpkaJRhU+/ouf4tzJUyTjaZK+wK2SzWJbI9av8r/93EO8cGGdyaE+wsNznH/yCxx/9/08dmoRvbpGtdLEdzwGp8foLJ1BEXzMrkXXAavdQDNCeKVNsgMT2G0LOg3CiRRmu0H95jMcePiXGH3lKmcv3iAUjjKUUfnQZ/89jdJ1rr/6Ig/d/at8/j/87/zV5/6Qi7UOgt0lGw6zsVkhm1epVmtE8qNM9MHi0hqRIMDuWoSTcQLHQ3TbSGochADb7oIWJ6Qp+L5PMpUFP8BIpnBdn9s8lwvXb3Lx0g1euvQEET2E7llUzS6Br5COKWhGkm5zFSOsU2jafPpTH2LhwmsUTPjRy9cYySWwkbmyVEKQFMRAIC4JvH5xhcNjCS7UHTzRRSg6dPUuj8yOUOrYfOzjR9hYKTHkWAghCV2S+PYzb5CL73bf/+A3P4krSJjzF7h/7wACCssbW4yO9RKPJqgYfbzw5nlGkxJBfYdIKEI38PFrJZRI5i3bbQFdS6NpBr66a0gTsNu2UcQAOTdIe32dU2cuU67WaJdLuxqNiTRzgxnEwMTZXub3fvg0o71p9t33AD9z13Gi4TDnX32DRqXM9OQ0l2uvIRgB/UIYPRej60mkozHabptII0CzbATBRYnHeOS+A5gu3Lq1zuKSjesYbF65SliP8JlPvR89fAdPPneaT+y7n4rl87UXT9JjZNjcKRLWBQ4P5Nh75+28/PI5xI5AoFbRewa4/eAIyZDB+GA/526u06x1OL1Zx7U7fPidt+HLMbaabfoGUrhbZVJBlftPvJelhUVuXLxF1xIod0Wmp6dJJZKYskFXyTKseojhCOevXOXYsYN02yXyk2n2H53hyZMrPxHf/5+NQUEQBOCLwLUgCP7jPzqf/0df+wBw+a3X3wd+WhAETRCEUWASeP1/dg0vgKjoo9gtlIjBVrmOqkkEQcCrL7xAqVJnYavG0naBlYVFVssB3//uj+k9epxj953gbNWnWq6QSeYwrTKN+XNEEmFWiiXQQkRTObx4mqneLJN7Z7HNBn7gouoxhFgcwVcRVQVVlJFlGUXV8B0XKwihyxKdwCWWStMqd1CcFqWGx6/+2i/x2GOP8cwrF/jsL3+MoN3BMT0sz6bWbeN2TVxXxu40Kas9jA31I2khXEHBDjwQTIxYElnxQPJRZQ1F8HGdDvgeiqwha1F0fXe5KJnvYe/4IO8+vpejMxP4SOiKgYOCKQi7QqQyCGEFSVS4c9QgisvXTi3z3Klr9GgQj0WpNtpIoYC24+1atvkKshjQMzwEEgS+gOeLHBkZZGTPJO941wkuX7xFc2eDICKSjISwPRtLcnF8ha7gUq5UuHZlGSucRUmkSeayHD86y9DEIFI8xur8dWTX2jUsFTS0UIhupYKg6ASOA5KHLOtv2bJpeJb7/zyaghAAIpKs0hNPs7Cyg+77BEC95bC5UWK90ablekiazqHJPN1ul+888QJPnL7GleUtRobyHD1yCCmqMpWLYbsuqZ4EvqSxvdwkMzTA1Ox+5vZPkxzqpX8oy77b9iGjsVHeplmsYJc3yAgK+WSOsaEU41NzICvk00nksMb15TXMZomQJBHRZDzXYbvlsV12yCYTHJ3Is29okOZOFc+xSSaTrBYqBL6DhUAsqvH+u/eRiSosWRU8x0Iy27TbbTxZJZ7N0lZ21amT4RCSrNOyXRzPplgsYjWb4AiYWFSbJornIAoKeiRLuWux2aj/RPz9r2QCdwGfAC4JgnD+rXO/DfyMIAgH2U31l4FfBgiC4IogCI8DV9mdLPzTIAi8/9kFZMEnrQdUuha+IKNhc6VQYXbPHl65uMBIb5iI20aSRHpVm+uvnUIsrfCNr36Tj7//LpY1n5uv/JByYYM7j00wddcR4sU1Ateh1FV57dU3SWZ7yPb1IWX7kYIAIxShUbxJtn8Wy64jCjqSIuN6Iq1Og0g4yfzL32Ztq8BQb4YT9x/n+NE9tNtNDhy5kyf+5A/4zc/9CY986Ofx7r+DL/7Zb/GlP/9LXl9roghQNQxkWUXSJZpry2w1a0zPTSBZDr2jR7ly6gfEe30ikR7ig73QKmA7Pr4v4Nm7TsEiLg6iNGgAACAASURBVJ4gMTA5RbFUJTk4TLQnT//Mfo7sFOiYLXzfI2h3sM068UyOaE+SWkugJ5/h4tkLPDgV4eqtBo3mrtPTRErHsk2QPFqOSV6CvqTB9k6Vaq3F3bM5pvdMMH7oCI22zbnlHc69eZljc0MkYnHKhU1eW6+Sjerksgk+8L57+fL33mAml+XlaxdoOi6+EOLuw+MEKztUGx36c324boAva3hBwM7WDrlkCDFIokXDuFaA6zQIJ9K45q4ZSRB4yIoMgYDwlnb/odn9CI0NOhZ87fsvIIsKPg5d22csn2FiJMfM7ChzcxqXrlxjJuSitcucKThUrDYvXLjErz94jCtXW0SyKfo9neP3DRMbnSEWT1MoVnj2+mlS0SiLW22mRQm5XGI0m6ZfgEwuxdLaFj/34Y9S9gzOXrxE2TV5+dIip8/OkwwbGGGNLEl828FQfGhsofsmY9MTnF0u4XQgkzdQ1Cinz19kbjTH7/3LXyERS7NW3OTU6VeZzsdpd7ponQiBXGZkaJxCocCNrQrZfApZjVI1fQ7MjLKyWcJutWhU1phfrRMZ1RgdTLOx3eD1Sxd54PhtBG5AXr72E/EnBMF/V67//34M5uLB7/7ie7m4usGJ26b55jd/THR8nKlsGIkQpe1l2r6C4bq0HIs7f/Zj9AgSzYVlXr7wCscefIRXvvZlthouk/unubpYxO92SQ5nufzjN0gkElQlgwemY5RMh4cf/WkO7t1H0+wSTubwuxVkWaVV3QTXp7C1QDY3vrvRV9jg7OIGM5MTPPv8SyTiBgfG+vj85/6M6Q//MsVXn2GoJ4KvhxgZH6MvpXD6xdOstSTyUYl0Tz+irKDKAtF4Ak1PoYoBjfI8swduQ5Z09HB0tw5WBbBFXBEkzyMILFxCaHoYy3d3fQk9Ac9zwHVxAUQJz26iCgo1yaAvEeFHT3yH9fVNBuIKpVodRU0xcfvdtK89T+/MAcLdBvV6HUmVSKYijO3dQ70T8Pwrl3Asm4QY8OaVW0zPTPLmzQIfuHOSSH6QL331cYrFJhHR4d/8q1+hq2jcnF/l5KlncU2BoUyanaqJHA4zmh9C0HUyCQ1Z0LHeEk0pLi4wMZgllkoT1kUGR+cQtBAeFlEjiZHuwdBDeJ6J5wfIsoqq6Lji7uj0mWcf54cvnOHB/cOkDQVUlZPnrpAydPryQxghBVWNIScSWKbFyvUFIpqFILrooRDDMwcpmQ0W1qrstC1M26RSKpFyttmwA6byvcQGsmRiMlfnV0kZKWKxfrL5QdIhjWqtSFTVobqAaQcsbO2wulyk67pkk0kc2yekK0yNDTO2Zwaz3UCTFLR4nK7lcHOrjI5Pt9tlaKAflAiVrSUWFhYpVz3kboF8rpds/xCmLzG59xCurPH9L3+BkXSSyzslmvUuPb1JIppALJC5uLRBOGqwsV3iZx+9jaW1FoJiMDs3TuCpXFy8Siro8M8/99iZIAiO/L/x97bYGBSC3S2kiK5RbHZRRIdH3/suvv53X+OuuTHU/gzLax1EQ6W+VkUKJNY3Npm/dJFyzaJ47hQf/+l38A/PnePFxQLvmpvm5TcvsnZlhXv3DPLq8g51C9qmiN/1KZV2aNseoOP7Lnooiu856EqIhl3Dc81d2W9RZvy9n8J98Qc4lUWq1Q5Gt0nvQ8f4mUcf5i8e/zL/4uGDfHulRa/qUbgxT2Sil7vvPsA3nr5Aw5GROy5xJaAtiYR0H9Mr7z4sQpRGu4OhC1hyCENVEQMBURPxuy0k5F2FIimEHziIvo8XSAiSgOALBOIuQARBQvBUIskEpYUVvMQoR4/dxiPvvhOfgPLWBp4goqo+kT3vQ0ICUWJEUwmCALPj8Pr1BS5fXiIm+OybG6avN8l6ocTaZpn19U36R9/J956/QrncJqX6OJZEoOpUdyqcu7xMLJwCxSUcjiI3u3iuiBA4FKpdMuksZ5YLVJs2w1EJTRbRjRCVnS169u2h7VhIvosqK0jKrtKS47mIgCrvSq8j7mr4e4LH6Ngs+5fXeOlGidmpLIcHhrhrv4QRjxJJ5CmVC2xWWxR3Oki4zC+s8Y5j48R64uTyg9QcgUK5yenXr9BpN9EkkQNDEeRQDqdYRZWjdB2RzS2HXKAQJ6BVvE65vUF6co7CtQusiAo9UZWOD74oMdcbxfcdtLCBHwR4gkqyJ0Pb8VBDUVxRQnRhuVBkc7OBoYfw3TZGpYwotwnpLrMj/Zx31hkdGySfyvPC1TUiqV7EzSKeK+MRkB5II2+VkAOPnZqDr7vEU2GmB9I0uy16RrJcW9mhVDDpyckochg5kSe5uEjX7fxE/L0tgkDXdujUS5SKTcLhME4owtqVs+yZmOW5hSKP3jFEySzRFzM4ceIwbtdEdVt86Bc/wB/9/p/TMzLMrS2LOx99hO2/+BswAhy3g2KoXC52uf3ABMWWSblc4/i73sH1KzfYv38/uhZDJkzArvxWt9VEEyXKlR1GJlVc4G///G+4Z49GOxHn0q1bpCfzhNN7+dPXn+Tzv/FR/vUXT5L1iizWc8yOJKnHBjj50mk++KH7ibW2+Pt/WEKOR5BCGoVKhZimUhdFgm4Nu5QgOp5Bdm3cwEKO9mB1dtD0OIIg0ChWCKUjSIoKoofVqQO7HAE5lMFXVBQJEpk0mxs3MUIxHvvyV5maGOW1jQKrpR3evLFJRghYNh2me2IEQUAiN0w4KpOOKvRnEjz5zGuMzU3z3g/czUun53nh1FVSmoQYdNgTD3j8OydxmhtMJg0QAx5++B5evbTE5q1VsLskkn30JTQuzi8hhhNM96TJZFLo9QqnL22ydPESw309yPE8yUyeQtNmJJun64Bgdkj39JNMp0EUUXUNQRR2eRaivKs56AeIgQ+ixODQNHedcFn+0lfQaiHml6psd0SUnQZtp0h3Z5nX5xfoTWXYM9rLnXsHGJyeAiXMxRurnDp1lqTURatUEQULIRziB2famK7DPcMRspkIJddlcXuD/mySgmxR6zSZ9AOuXT7LzO33gF1hc/kW33r6AklDpDfeA1YdLyiSTqbQNJ3qletEdI2a5VIsNYhHI2RSKaxWA7vqs91qsLq8xt37E7x+qYotKcz0xVlsC5y6ucSAZBGJu9w4e5qpnhAHx+PYnsw7bx+jsFVmq9BgeCyD50PHUJhOpollEnz+K6fxVZW1doF6cJZyW0JsFZgbTv9E/L0tgkA8HiUxMMRRbZt/mF/jg+95iHgyhojN3L4hVE0nGsqxXavz2HdP8+F33kZqeJrCrS18dL7y3Ze47cAUxVqTmaMHOLhniBurVU7sG+VHL12gLBqMZlXk0V6efeoFrEBGUQ1sUUDUI4h2m2bbQosmCRyHSC6P77RBMnjH3X24mTvY/ts/YHxskvF9ozz79S+SqF/mt//zRX7zQ8c5deYy13fqnHj3x/nCX36Bf/6bv85jX/p7JtJh7rrvEJOKw/NvXKMe2HhdCbNrEdM8Gp0yvVIO0zMQugFIdSQljuO6qCEdI5tHEtl1KA5cBF3HFxUCJPy3SEitRpVAzBJJTdBqXqU3niAZNbjebjI71MvSjWVUWWJfWkUVA3ZaJnJplZ7kGLnxWVJCGSGkEo+H+cFXv0tH70FRZZqNCrIm8IGPPMLffOn7KIrCZ37pUeLDQ6wsrfOdv3+CfFoliswrC5vEMilu2zOG4Tps1tvsNLtkohJ3z2YItzP4roNXa1JotDDCEhOTI7i+h+e5GJEoXhAgChKBJCNKISQ8fN9GFGQQFAIRhGC32TU0uZ/f+8M/xlOj+EjcPP9jzp18ioFsFkYOkslECIKA4cE+nj59javzy9i+wLXlDWb3HMSRwtTEKqIcQbVFhmMdbEfAy8fRsjHccpX7Ds5y+cw6ayubHJzN4RV9in6H0GvP4ppwwUnwaz//CDtbBV65uMFOJ0RYdInJBmUrQK5VmZrNo3oCH/z0ryP5Xa6feYmJiV4cRSUsBfzDi6f5yrde50MPHefycolOYHF8eprrwjJOtUJ/uEs2Hd11burG+dvHnyKZytKb1jhy5134rooRiRMxW1SLV1k5s0LCq/CJRx7BVRWGJw/xra9+gWPHJih1nZ+Iv7cFi9BGYnpmhLljtzGZCrG4USHbP4oYyXJxdYdOs4LTKfLUyTPUa0X65w6w3igjRzNQLyN5AbUu1GoNyt0OZVfmaFLi5eUio2mdpdUtLCGMHO+lYfvYvo0Q7K7gyqKIY5mEQiFkWUZWRNKJPJKsIQiQzWX4t3/yBd71O3/E3sl+/EaFerVGoMmkcz3E3vkJZF0jECQGx2ZIRKI8d/oyh0diCFGJU69epJOK8a579mG1ugS+jN9pI4Qz+FJAdOAgzVoT0XURAhHH2W0KIsggBgiBi+/7IKm7/4yBgBD4BJ6NFPhIoTie4yL4MuVWHbvdRgx8kobMhdUCE/0ZwrJENG6QTYaYyWfYPzfFRrVJ4cZ10v2DxKQui1cuMj01zERe5sjkKAPZKGNjw7x8s4QjSKybbTJDg1gdjydfvkSv4mIgYBLQtD0ajSanzq1xvdzE6jTZLNdYa1hUTRFklUwsimlb2E6HVgfMjoPviYiihGW7BJJMgAiBSBAIBIIM4u6GIYDE7gREFEXEABzRwPd9/ECiZ3gvo3OHGJ07yKuvXWBnY4veXJRGq8WejEgyGiIeVYkZIUrNJnLgkJB9Qnh0rBYNV8ONZtnyBYKOgG2KzA4dIxw00XHZannMl1oEKJS7gK5jbSyw0PRJ5gbo0CWjdYlEVRYLW+wf7QFdYbg3yp7BNDtbm1RqJXL9A+yUG3QaTQRP4Mx8EdF1CRs6+/sizEwOIYdUHnjXw/QPDBJLx0jEwvT3pgmHBe7ek2cyBlLgEzQr1EyTcqPGcmUbLT3MqudxaM8kXc0gmh6mjUoslmS92aVh/uTe/NuiMbh/Ziz4j79wnGK5zuWmTFRTGOiN4nS6zC/vMN4fZ9+Dj+I1lmk0XP7uq98kN9CH3WzzsU9+hPWzb/DM+VssLFXZ2N7kkWPjWJ7P4dlx7K5JtCfPTrVKKqpx9eoiod4hPvzIh1ASPUQiEQTfxPV93MYmoqDTqW8hG1GSQ31UVit87tc+y0udHNMRiwfuHOPVN24hhBUefedxXj17lV/4pV/gd/71HzO7b5b7bhvlT/76W9z7sU9yQN9mfq1CKqWjNcqEegfpjwusLhU4s+ahY6EpOncen6RVl2mi0pvL0LGaRKJpFCOOFpj4gkwg/jf6hYwgSXidGpYroGsSbteh0g344eNf5v0PPYBpOWzUSpx/4wKhRC9GzGB0KM/llU18xySGSG44j+T4tNtt5lfrbK0v8p775ggkiZatYrW6nLp6C6VeQYmp/Lvf+DTfOHWDysoKjUaDluAxmuthZadJt1hCi8rkwhFsx6fouLzznccRips8c/o6B6eHiIfhI5/5J1y/dJbVYhXdVUll0oRkkezQBHo0RuD7JDO9yKqGpGlIdrC7uQiIgYjv28iyiudaCKK0a9gSyAiKgu90ee4HXycT81kr1ZlMhBGjYRzTpdNu0KzVSQ2M0BFldpausVpxsQSR1tY2PYkQO56IFlEoeFG2zl9F9wR64j6xsMexgwe4sVGnL5MgFdcYzvfScXYp4X0jY7R9gevnL2GoJgPjR9kyXbZW56lXymTUbYxQHs2IE+uf5IfPn6a8cIODOQVfVOh0fJp+By9QCHsyiakBbC3GIyce4tSLP+LAgWmyg3PcuvAKgW/hOCqSqFGpN1htC6wtLdPu+NAs44k+77r/ABkjSlfyIRxneqCXhWvXaLeb/Nyv//n/sDH49ggCc5PBr/7KT9GX6KF/dIRYJoLqdhFtm0Z5i7bpcn3bRqsV2HECIp0yL792GSGkk+5PM9abJZnKU13b5PSly8QiEiceejdbNy6zXCjtzksljUR/jvLKMttunN/47KcR5RR6aJdM5HQbuwYkto0QgJHPUCysEIvm8c0O//5f/DOGBkbpagYLt67zjncf59aOyfkzl/jgQ0f5+jee4cSxaQQtTj4scvCBY5TkET7/uT/iWFbntY02H75/D7YEiUScAT1ge7vBasnkjg9+kue//3ccPXYnju1h15qkRqcQBRlVA10NI0giji8gCv6u6cn2MrKm44kyl64usn/PGK8+/RRDI0NYLZNitwPotDsN6s0WxY7PbE+U+Z0GpZ1tQoLIRqnMsZlhzE6d2akRGn4ER1R58+wF0lGNWq2Bogj8u9/7DUxf4PO/+5/wdZB9kd54nLZj47otAjWBHYgUyyXMtomgqJgWhFSRlGjyK5/5ILmZvXzp8edYW9tgX/8A2VyE3lgUQ5Dp338EVdWJpzO7fAI9hmm20XQRAgnb6qBIAr7gI6rpXQKWvyuXJcoStieiK7vB4fEffJ2pXIS66eOYJhg6tGq7gi5Wh063TKMFxfVNNrZ2yKbTLBTKTByeZmRiH7mNW6T6ezByffzFF7+Fr4XJZKJMDOSod2xOzhfZPzPL8WOz2O021y5fxZVF+hIhRgYGUCQPz+xSqpYo1hvkUj3sFMsMDAzQ3FrjxoWLBAmDphfj3gNDZBIxGqJMTIGooRB4EfygS2JkD1tLqyQzSRqOzc2LZwirIaqlLfqHJlFUWL+5wqGjR3B8l8e+8g3UaJg9Bw/S0z+OISs0auskUgMEYQNX0Lj3roffvtMB0/EYTKeoF1c4u3INNTPM+N5ZMvEU8cEkEbuD171OVwlTuLJMRxToyyUQxd0UcjgfRUtkSfVkiWUjGEpAYaeAFwQcOrKHV545SckTmZuaIL5/H9d+fHFXbVh8y4pM2JUbbzs2gSCghUM0rTaSKyOKMr6a5Fd//Z8gRHr40fe/ixCO43VMSqvLTPan+OaPL7NnvIepmWnOXLzCkbvu57UbJb575hqLm2UemNnDhCvQv+cgreI6IVlHTIToi6cJpbq8/PI5lNY2rbZDu22R0DW6po2ueQS+gouAIirIgo/2FuuuU4DAdejaHV47fwUlopBNpTAtFz/wEEWZFipdBNK5LNJOibW6TadaxzSbOK5ALhlmcv8sYcWmG8j4LYO10jZts0NCaxBWNSS6dHyPqzfXsZWAvKERDRmUUAicgC1Hot6oIAQB3apJs2XSosV4LEReCyEHOmt1m1tvLvKDF84ymYsw1htCTkbxTJtO0AVRIhB8JFkjEES8QMAx22h6dPf+BzaBqIOvYrVbKNEYkiAgC7tcdUkA2+qiqBpKt4EvxVi8cRXf6jCQidN2HGKJLDvVOvNrDa4tbjGZiXCrBZutCv/0o8fpzQ9yZdumksmzVOrgbK8TkwJeXVxiOsgxENdJKCL93W2K10y2eiQ8B/KqiaYpJIwosgCuY9MNwAoMompAVFdpxnV0ISDeP8RITw5LkajUO4zO7kMKICN5hGIR5JCBLMvgCHRcj7agIda6nL8xT4+RYH57h+lsmrYURVdc0v39lFwBIYjT15vnwN5Jzq5VCCkblMUIIwN5nj9zBtsV2S6UfiL+3haZwN7pkeD3/vifsbpa4NRrl0jVN2mUi3QDlUMH5rAFiXqrzvvvOcStUotsOoEU+MiaTGftBsvrFk6rQdUIE1U8/GIDK6QjiXBlfhVZ9AkbOq1iDT+qk5+8jV/59KeR9RDJTA5FkAm8Lq1WCUW2qW1cIZY7jLjr6IDoO5hWgCC7PPf1v6bre5S3TXRDo1LY4ti+fmwlgqZpiE6d+eUOAxOTXH/hBZqtFl+9sMlgPsX+tE6spxfX67KxXGJPXy8TExlW5AyTg3me+f732P/g3fSHoliiArbLyPgEWjxD4EsEgYciR7DtGk55Acd0aJgKid4wN0+9SMN16M+muHR9BSHRSzyRQLGq2LaN1awj6joXLt/EVeDdD72bZO8Ezco27XYDL4CXXznPRF+KnfIWJy8v84G7j/HJTz3MZ//Vn3HHzDAT0yOcubzJ4uIt0qGAjCZwaatBOplCklWu3NpAUyWa7Rb7Jnvw/YB0Ks9YbwQpEJi7515UI47Q2mFldRPLNjFUmdtPvB8jlkSSBGKxCI5r4/s+irBrqy7Iu45Eum7gdlu4joAei+1mA4KCoGgIgU+72aCwscyT33uM3kySiKrR25fDtFwm9h1m4+Z5XnjuBTRR5rb3P8KN5U3Wri9TLWzScGz27D2IWdzgwFQ/V4tNZpIahgQ/Pnmeht2hioEgxdCSMe47MkPKrbO1XaTZ6jI6nGa10KUnFef8wgYP3T7DamGHl968SblmMjWcxVF8RvpHePn8Cn/8h7/Nf/r8X2FWSwwP5wjHk0zMHmCq36NabfPm+ZuYFQ/La9HtOhixMIPpGIGoEk/HcVot/CBg9vg7KFw7h2W3aTsWXlfA9Xdt3rVYhkhCRQ9EzI7FL/zGH7x9y4FsOhaEY1Ek32I2qbG41aFlhBjvj/HT9x4gGjJoOxaHjt/L2TfPMDE2BqEwxe0NXnv+FbJJnff8/CfYfPUVrHicb3/lcYamp2iUipw5cx1T0clENLqutyv0ETf4l7/1b4hmhkile5AkAVkIKBRukjAMRM1AlnW8AFzTQRQcUBMEYsBTX/4/EKwqRiRCpdzmalOiuLTM+GgOUxXwi2XC6SxzvRmEiI6ayHP5me8TTqQ5c32ZoYE8d5+4l43rV6hXCqQNmaeubqNIMDE5glgqofT08uUXb/Gr77uPselZhJBOLJrGcbtYHQdF8ZDNCo7nEglrLJy9RMet8OPzt+jpzdET1TBLRaLpNOWqheU2mRwcYX55gw+/511IiRhLpYBKpYIXOPQnImytrfLA8UlkLYSuyMyfO8dWsUOovwc5CJEeGuC/PPY9hlMRxK5J1XLpyjrJoMnF5W1WdqqERYUwPk3HoXd8kLgu8+AdB6kWd1ACeKNgEY0ZDGdixEUXXZVxXJnjD76HkB4mFI0hSRKBaxIEAeFoBEHR0ENRqqUt9GgKPAuzvE0kP4oQ7Eq2e4GI4As4vofTblBYu8VLT32LcNjAcn0G5w7TCjRO3D7L5ZvzVNa2+f43v83h4SRuPE15aZlPfvpjnDz1BnfecyflZovRmQOceeJx+kf7+e6pJeKpEBnVIxWL7qoLOTaSIFKqWDRCcbRQlIxic/jgHFulMjU7S6O1TeXWdaKag7ld4c3lLWy1h6RikUqmSYZ8QKQqKJTqbabHe9mY36Dj2hyYG6NoSvSEFQKnC55NfnCQoamDrCzOo8biSL6NVa/QtkR8AVRZJCQJmK5AOhnG7Noomk691kSVfT7xa597+5YDdtckKYuYok8yYaCXPFKqTcY3aXZ9JFVCE1UcJFLJMN12lTdfv4bktthc2aAnu5fF8+fITgzSnl/EMU3mt2vEg4BozGCnAtmIR0o1qHcqhEJp8AUIBDzRx3ctQkaMWChMoXCL/PDtCHII2TV36caCgu+5BLaFEli0XAnPlegqOoYQcGA4xdFDw/z4xYskkgbtZpNXPYnbh7I0O2v09/dRaFrsnRqg7emcvHCLwaBFvdEhEk1Sq9XoTcRZuLnIiaP7iceSPJK/itmtUi0XEUJhBEmh0Whw6Y3XGB7IMjWUpFE2kYUwO7Xd/Xmz4VJzd6joGsMDYURNoG22KNdbpFJt7jh6gPjIXsxujWrxBr5ngijgajn6R3OUvRCy4+MJLjulOuWdGhdXmnzqfe+kXu3y6oV5rJF+BlMhBmIKnW4HWTHoj8JMKk1YlUhEIliOS08uy9DoEIlMlh+tF+jr6+G+qMWNkk11Y5PkcA9DE9NcuTqPbVuoqorj20iysjsF+G9ybbKI53koioLregiOjxKKgBcgCCK+Dz67YHJNE3wfRw7RbpukMz3Y9SqlwhYDCY1qdZjX/+FVRlIKRtKgocaIySoHDk6THDhIdnSTG6UWEWSeOnmRarmDnhOYy4dY2m5Sl0APRbi82cRpt5gZTmJ7XV584SxOO+Dheyb5vq2SCSlMZSV8fFZMhxtLqwxm4pyYzuErOuFIgq4SpVnrEIgh2vU6gtWhXmvSIeD2o4fJ5vvYfuV1TClGNhNBsLuYqBQKBQRkgm4NBI1YTw67XN2dLMgaa8s7aIGJZESQvS6Nrk0sbGC/ZcH3PzreFkEgGo3y2d/6NJWaT4g6J+4LWNrY5OD0CGa0j8mwhxiLsLOxwqXra/zpV55h/2iCIzMD/OYvf4Qnrs7zzOs3KTxxipvzN/nL//KfKT/7FW50VBbX6sTDFiFNZKg/y5V5E63j0PV9jMAncCwkNUSzs0Fp4xJ9o8epmB003ycWTeG3N7CtOnooiStI4FpMHX8vUWubWwvXGJkZ4cb1BS4vlqkpCSQHEDpEPJ/N4jaJkMboob0Yyzs8//TTKLpGTypG8sFHOah0ePPNN+nJ99EpVbj7o5/B3zzLEydf5cDEDIbhsG2aZL0KLU1ks1pAV1q0Si0WO9vkh4Z47oUXCWyXWtMHWcCXfIrVBsWdMg8cmaXb6LBdDZhzfV65vM3Cxre4ut5GEtvUTRM1pKNeXePg1AAnr+1gdgMapW0iho5EBOwuX/7e01Q2FsgP9dORBa6ub3Pd8alYHkene1je7pJNGhwaGMFqm5S7JptLO5xbaZLJ9HL54g1uLBeIp+JEA5eBsRGmjx5hZ3WFUq1Jo7xD13cRKhVyw1METoDkQ0iWwHZxBQ83kOm22+CJOKZH0nDxAwHbMnFME1GSUUSwfRdfgKs7TQb6Gxw+MkPQ7rDVbTK/skpY6bBScHnP3gk820OPJZGjGVbOvUSqaVFvb+CpOnKpwL0n3svm0jlCfXPsS1ShUWNtp8jPfPwTvPTU07y6WOYzn/o4mf5TpHqHaJYqaIaG07VwnTaeZXP7PfegujaW5xJNZ2iXFhCUGFubt9ifzyNrFu+Z+0X85nVOnl/lwXfPsHXjHK7T5ugdRwgZcQqlMh2rQlzSKNfKRJI5PFWhuLPD0akRPNHgxo5Fs3gDoeMyMjWFtzOegwAAIABJREFUb1uU2j6ZvhS6nMCtrP5E/L0tyoGhgVTw6KEZblRbRL0mQyOztEIpbjs4zO2TOc7euEXfQI5qwyMndZgcH6Zhqzz94imee/okSQmWbQmhXabZ9mhkJvi3Hz/KwEgWQxT46t/+kFgiiSZ7mF4AgsM73vuziMkRxvsNWt0W86/+mGMn3o2vJLh1bZ7c4DBGOIJfXUXQNHw3IAgCXvnOF7AVgY7lc+zBR/j6F7/EcF+MJ54/Q38qjmg7rDa7ZLNZstksH/qpB7h8dQXX8RjMxbhw4zLlqkMmsIkOjCNJFsl0ioWCSdfz2K4JfORIktevbaFFwyjVbearbUJqiIF8gnRvksmhPN/+r0+QDIdRB+Yw2w3WF+bxCWh3yvhKjOnBEa4VthnGJRwW0GSNH1zZZDgRRdddapt1hsb7MDyTG+UO+47ejrN6E8SA2w9Ncr0m8uYrr9FBJKa45IbGmb+ygqS4HDlyiNn+JCvVBqs3d+jqOmo4RiiWIhGNkQ77uOV1KvUOHmGsTpGoobLV9Zga6qdWaaCFNGTbolJrM7HvEKnMLnW6d2wvnUadvsEBVF1BlIxdmTE8qvUKgWMTcVtI2UECx8H1bEKqgmU1kWUDT45gFlfwFZdmscaT3/4qdxzsp9wNsV3ZoGd4gGRkhHOvX6fUaXPPbYeJahKWmkEKeXjlZVKGgda/h+LqCuF0L1Zth1QqzXKlRFSJIoQNvved5zl8eAbHqXDPHfdTqrZoVLdwWhVGJ8ZxLOgZ7CUk+1xaK1PaWCWR6KNUq9EXkbDcgJXlG7zyyjmOHBoh1z9Db28PdTMgqkfQ9BAELpmeLIXV64xOHuTJ732b0elZWrUyYmOHoX1znL90hampKXwM3MYOiy3YMzZMLqxw8+YCid48ni/z2pnL/MHv/+7btxywTJtyp06PrpNPD1K3W9yzd5ipgQjPnj5Ly68zMbuHwYiP57W5VXeo7RRwStu8//79tGw44nuIkkKtYdJst4jrUbpmQLla4cMfuI/V5UUMw2CtssVSEax2nb68AHSwnC5j+T6kaD++1UWWVWRJIfB8PGFX7MMPLARBQ9ZDGJksnbbP+StFxnp01ms2B0dyhCJxWh2bgUGRiclh+sfGmC/bVBoNRrMaP3rpdWqdOiPJAY49dC/VSgNHFXn1wjInLy9zW38Ox/fwJx7Ge/0LlO0as3mDu3virBZqZMNh8rk+zt0oEQnHCcVDXD7/MkLHww8pJOI6oh/lYskmYexQbTbJ5dO4gcrYUC+/dWA/ybEhzK1N/u7xHxDP95PUPFbsJbqlAvffuxfdiNDyJMydVWZmJ1grVMikE0RCOuMjSTwvoFJtcUF2aVWbDA3EubGwweLSCnsmJimWtrnU6jLSnyKRTSG2HSQlSrXeolZs0Emld5180TDlMIVum3S7ja7KmLaAnqpgmw1sJ4+qheh0TXRVw3Y9TNOj2+6ihEGxBVzHwrbaCLZLuePhtArE0oOYYgi7apIY2k8okiKc24MhRhnoHeXGSpGkEWZ0KM9sYBNLJwnrMVpLt0iOTFGN9VINJMRKF1HTwGszMDbOhZUSlbJDsh8KK6scHI4yFmsTjcRIhHzSso6Xy9GwMzQ6Fqbd4sknz1Fv+zQLq4hOk/z4AWbHBhkYHMb2RZLxBIpTZerg3VxbKbFx6Sb752aJxlRq9TqaEcIKogRGD5vVLocPjHB5/jqTE6OkpoYp1h2G+4cRlB5CsoAfjpC6dQ3BD6iaLnXLJSPIILq89NyPfyL+3haZwOBgb/B//v4vktUVVmoB3dIqt99xGz3ZJMvrG5x88xJJSaYja/ToEldrHim6XH/jPMPTQ2QGxjHaNWKBDUaIbVFhYrCPnc0tItE41fUV1FiKte0Ge+68C8kX2XtoH1fOX6Rv7ACa7OHaDvFYBKdr0amX0WIZvOYWgRzCUAK6rTZeZJBbr/+AC1dvkjJAC6lMjfZz5Y03ILurrSf6ZSoFi8zEHO76OsmBQa5dPMOVy5s0UkM0ttdxzA57BrLYkgxmh2gkzMD0LI8/9nWaHYuZsT7e9fAJYvEkF9eqGJVVetIpnnhzlevnzzKSz5CePcSxY8d55cUnaS1eIjJ+hOjOOpcaJtmEjhIy0AFX7yHXn+G2uXGe/tEzTM1NYwQO15suLVsiiKQYat8kIdg8e/4m9YrDvpkRuu0q0XCMzUaXhGGAIXH42DGcjSXmVwoUHZAlgZ+9/yC/9fm/5R13Hebi0hZp2UX0VHK5HB3PR9E1AjFLvXgTz9AxFI3R4WFeeu0KJ+69kxtLa8iBhxD4xFSfagsW1tZ430c/yvjYHPNrq5Q2N/n4w3dS2Fqg68usbrUZiAkY8TRGKkS5UMO3bZK9gzQ8CdH1OP3MU+w5dohkTEcxkgSOTL26iUiAq0QwFAARVZVRFAUnEPEth7hcQ5YE/utXv8Geu+4mHI6wcu0SPZk4bdtHtm2aapy4JhPP9BAKJ3AdWLh8hv0n3oPvthG6DnZnB1NJQuCC28XsBmTTKVqNAqF4P41GnVgIOp5KOtNLq7qN4LmYYohEzKBeb5PLJPHNGi1X49rrTyFLFsfu/Si1Rh0fiUajgSAr6LKO7XSIJpIUttbRzE3OXFrh0Y98hJbl4VkmyUSGu+594O07HZiZHA7+r//wO3imBWxjihJKYLFx7RamYlCuW3RW19B68kTtdaJ6jAvza0SjUfbPDROK59iqtak0q6xu1zk0lmer3OSdxw/z4tmrqIkU5VKLllNn9c1zPPpT76Rghpid28f41CiKZCOQIJZQWF3aIhGL4JlttEgIIfBQJBXwqTsaV57/GrYsUyrWSE3OsWd4nGgqzNLaEsvzqxiyjddu0Km3iKXC+MlR+noNdCNGJpfEc038Tot2uUtt7Qr/d3vvGWtZdt13/vbJ4eb0cq73KnZkh+pms5lJiUFNSpZEeyx7AHs0HsjG2MbMQAN9MTwB1oxtGM4YWxpThmjStiiTskiDolqMTXZ3dXV35fjqvXr55nRymg+viCEItmQRYlcBfD/g4Jy77wHOH+vcs+7a++y91uT8MfZ2O7x55QbDjSs0Vh9iZBT4+Ac+SK5c5Iv/8h9w6pn3YNoGUapw0MvY2bjK6uoSF195mef/3M8zGncomxat0ZDxUFCrTzBle/Tb+2zc2Eaz4PVrTR5emebijVvMFXNsjiPKssv+Xo+PfuoTvPR7X8ZqTFIuFwkGQ7qBIPW6HJ+d4cbGPqqlMDNbIzOmOPX4Q1w/f4GNVo92t8Nf/Cu/Qqc7RB1tce3GBjNTVWZWT9Dvdej3fR47e5aiEvD737qOJQJmpuq0d7YxhSC2SyxPKdzZ6jIzM4dWWSTzRuwOYjQ5xlRS8MeEskX/YBcjZ4FZIi+53Lq9i6FrqGYZ1dAoTy9h9K7z0ivneeRdH2TYOSAMQ0w9IicdFlzZ7wypL63i9ALUbIRemWXv1gUko8HM0gIvff63eOIjfwHDlBn5Afv7B5w+Vqe1uU5leg0Zhe2dTaJYot0Z8/g7n+DujesszE/Ta3dQJ1bJaTH76+tMz03gjCO+/cp53v/Rn0GJBvTu3sCeXeON8zdZnM4hyyqt1gGVchHPy5ibKeFECTvbB8RRC9MoUCgWad29Sbm2gJ0rUrFswpzFuN0jcdsYtWmkRGDogkTNkUYRnucTCpVczsLb30LRdD708Z97cJ3A7EQp+6Mv/UtutENuvH6O4ZULFGyJnufz7NmzJEKiPXBZXpnmjW+8zIkzc7huyub1AyamZOYXZ7DKU8SKwJXzhOOYW6+cp1pO6A8iNtZ36Acp11p99HGPu27CmekKJ9ce4omP/ywThQZCSyGLuHtzneWV49y98QaTK6cYd/aYmF8DyeT3fvczvP+dj3Dzwjle7Urk3D3kcZd3vPt5JpdO8J3/9FlIAyTJABEzPX+cIHN44vkPMxj22F6/Qq02wxd+70vYssTtA5/nf/oDfPn3XyRoHfDC82cwF0/R6rs0Gg1Wl+pk/RbDVGWrG5DTBUbcpZov0hk7WLkCvYOb5OuTvPSlr5GbXmUYZhD0ePLd7yHqd7En5ti8+B0krUin5yCXZ6lUSlRtCLyA2O1DOEQxbD7zn1/mz3/8XQRxwMrCAs2hw8HuJrak8tLLb/C//B9/j7t3t1FlhRe/9oe857EK+8EcB1cucPy5dxE5Ablijc3r51hcPYUzaLKx3eWDn/hZrr38TZR8DU2VyVKBlY5YP+hw7dyrfPTnfpaN/Rb1gkaSahxsNzl2cgVVknHikChKyJs6nb09drouey2fdz21TDgeoRfKBCMHN0txez0efXSNu9tbBFKN7d09qsUCUgq+24Y4pLm7Q6lSwyoVUCybhakpfN9DUvO4rksh7TIcj1A0g+07t5hcOU5xco44TkkSnZuv/GcmZo9Tn1+lv32FVrPPmYdOstF0CUOf9fUuE3WDhQmDWC2CUUFOffLFBoN+G7x9NLtM/6DP7PEliCRiVSfyR6RCY7C/RWrXsLWQfnvA/FyFOJVp7fURImNybhpNK3Ds+DFe/C9fxVJcKtUc165sks+rzCyuYRBRqZTI1Wb48uf/HXOzCygi5qkP//kHd0ygUCzyO9+9yc7123zsZI34PU+wcWmdcX8fRzX5g2+dYzJfJosjjEKN4tQqmuNSqC9QW5pm6/YdtNIcxAE5q0iStjn98AyRpFIaDpHVmO4Qzm3tMTdVQPVi6kUZN/M56LvYyoDQS6kWSvQHLWQxz9RkDSn1CIMRbpShKCFFPDIhOPXeT/Lir/0NzvzUR6iXTuEOHG7f2mLlmeeQ4zF7XVAkl61uwjvf9SxycY6ymuPff+nrvL6xQanT4d1PrtHrbrH16ussqB7bUkYgK6S7d9CJuHXxDlbuPSjDHS6+dJlWmHH1wOGdTz3OsWWTb3/nGu97dhWtWEBSFU48usYrr63z3Lue5g//4DYvv3qD6bLOxsYOS2pMYarOdK3EtZ020qDPMC5StiTCXBFFqaLZRR4uv4JSnSenpmiNYySDV8mZFuXJCR6OEzb3R7j9DpJus3n+FX73YIYnnzR4+gPP4oc5EhkUVWfh5BkKhRquF/PEYw0217cp1iYJA5ckNnAcD3OqjHPpCiINGIUWxXyDdugzVcozUerj+WMiRSFOD9cOuJGEVSlSdbpMz4woVOr4skaiSBw/tUK/t0WPHGZpGrvjIdKU3uVXUOYnmKk2kLQEQ0upP/YIiTuiOL1A0D1AtfIMYxl8j87eDn0l5viJFXZ2+9ROPEa1XMEhhyYn+M4Wg16T6dV3MPYiivNrFCdD8pPzFP1rNJsDzsznqcw0UHJl4gS2tu5SLddxnA4iiVGq86iyxvxyifNv3uGh0yvsrl+hP05ozE4jVI2NS2+wtnaMhh6RxSoCj6LsYE0skAUBXhKw33IZuwMqkxaypDM9UUC2J6nWagizwJA8V27s8vDZ9yIUgWWX3vL5eyAigdXluew3/rdfZiwkonjI1e9cYaxJTCshk0tP0N66xcnnnuHgxjpr7/sFwt46569sk5N9di++zpSlcGsU8PBDp/F7TVafeYbNTpvXv/MtvvbKNs9PGXzhlRv88l/771GG22w2B+zv73P2zDIf+On3k597gn/z9/9v+uUZqtGQK72Qn/3AQwybbR558h1s7ockZHjjFpe+9HlqC/P0FJvCqM/xp59EFYLYLGIQo5SWSZKEW1/9HZ56/6N86cXvsvjcX2VhwuKbn/vHnDxzmpml43z9i/8BmQRPsblx8QY/9Ut/ifb6FUYHd6jqCjc2WjgYBJqEUV9hwQpRbZOiKnPz0gWee+Hn6Q63qRgpfiLzrXNbPP7oMyi5SXLFIgdvfIGCrVCaPUPX12gUdGyREHoHjIYOUm4G5IQsTTDMCpIUEyMIu01CtU6pnGecSBiZT8nKESVj/ARcJ6JSqaAKic997rf5xf/mF3j9zVsszM2ia6DbFqPuHo3JNW5v3IYg4viZNe5uNdF0BVXTsOWInYMm9ZJBJJcIDrZ59c0rpIng6WffwSgyKBQKrCzU+OY3XqY+NUOmCsqmxpuXt5gq9MliifzMGqvL84z2t8lNH2PsOlgqZEnKXrNPFI6o1ma4fvkqQecWxaWnMQsGShYxGiWU8xrNrddR9DyKbGKUFzEVl85ek+rcPF67T7M/oFKt44zGyPKIzN1DLzRQcwvUSzHrly7TDfI0ZmoEwWEhU9s22dzcQgubzJ96nK0bt7nVHJE3FJaXj1Fp5Bm1Rqxv71Ku1SmXi1QLFqaq0B8HzC3OIAmdrds3SYVGvpJHs4uINGDz9l0kWYZ4yMkz7+DO+hVee+0OC2urDDsH6EaeiZkK1778Gd7787/E+k6f+elJbl6/ygu/8Bcf3O7A0vJc9nf/xicJRiNkWaAVG8SSxd1WH7e9w8mnzpDdvoHjQrmoIusGqpGj76hYk1V0/wBtco7OwR5bV9ZZOn6czO0yOz3DwvIs3TRllORYmlvFiTyuX1sn8H2iKKIwukxYe4TW/h63zn2TM8dWKFVz/OtPf5nnP/I8nZtXWTv7Acr1CT739/93wsosn3zhp8jQGLkd7u50eOrJs3RimdHN1/DDlJXjVdxOyNTCEpdfe4W1R86wefUu8w+dwG+1wcqRjvaoTM6yefUGmZKSFZYRg7vMr51m482X+K2vnOPdP/cJTmguueo8pmXhDVqs33qdVjfCzOt89etv8I7FEr/01/8WnQOP46eWkYXL3sYuQ6lImKSYEsixQyibeFGIG0Hgp5xYnibq7lGYmKXT6VDMVei5Q+r1CdIoplwqsNvuc/Mbf4T/5rdYPfsBJt75LIYUYdUqyEKQZSlhFtHsxITjNu1mRq5ioGQBll0gSg9n9IWBQIgUTYTkajP0WpuHqxmtAqE+wfbmTQq5IpkzoFCvEMcxip0nZxxOvMrCMbEksPI1Dg42MLUEM1dn/fpNTiwvY5UtdMXGKNbZunKebn9MkEQgWyiaxvzMBGEi6Hd6uHFK0U6Q1ALBsE2nPWTrzhbzJ1dYv3yJp599msgLKVZLdEcR+YLFP/9Xv8Go0+PMgsnzL/y3xKM+pAZjv0ej0aDdblFUImR0ZGlIu+sTxgGpMUV39yZnn3kWkZ/BH7dIMxlZN0lQIAmIvT52YZI07UJsI+kGaTwkkmqHi4Suvoaia1SrFdp7TSaXTnD+q/+JU88+h5Grkjp9WqMMVU3J5/PYSoDQ8yhKju7eJophH97fapX3vfdDD64TmKzY2d/9W59io+2TlwyK9cN5+FNTDQaDAbk0ZiDZTBYTvvvSRba6Lt3tXZ558mGmyxrv/ZmPkZtaZmfgkDeL3N68yWj9IikSnc4AP8vTHXQ4szLN1s1beGaZybkZAmdM0twiySTs6gJJ/zZPve/j/PvP/UfOvueD7PR6JNoE49Zd6tVJ7LzBIydWOXf+m6zOVtneHZPJGV8/t4EI+jxychVD9Zk//jDt/Q6FfI6dgwFmwaZhqzT7fWqFAr0Qbl08x/KJk+xcu8Py6UWQFVq7+9RnZpCEydzKAnFng06gE8QJWdQnyiwk1eD613+H7/7+71GZKvOe970buTyDWa7hJAazM4tMzy9y+5VvY9UaOOMDTL1MmrqMxy7VyVlkIRH4HczaCfqtPUxvk6mZRdRCEaSMME1wAoWta5cJFZv6xCqK5KJGMY3FBXKWza2rl/AdF6tQxA8HGNYE25vXmZydxXNCxnvrFItFwshFkUsIW9DvZeRKNuPWNoXqBIokkIVElsREqkkqJDRFIwoCxq5LpVDgyje+zNw7nkfXDfxhG9nQuPTd73Dm8cdBUZlfXsPt7RFrRcoTc+Acpo67cn2DZ59+lBsXL6JaBa5cvsDExBJyNqQ+u0jzzk3K08eJ/T32b51j6aH340UZmqYiSzFut4NZbtBu7tK79RrLT72b3lCwffMKpXqJWnWKNHUxFJvtlkPOkDGLZVZPniJKwAt8YqfN/vplaotnMBWDg/0tqvU6vu9TzmmMAoEaezhJQugGJLJARmXQ7zM1UaHfPsC0i+TyFut3W6w+fJJhq8UoENTMjO3dfSbn51Dw6WzvsXLqUW688h36ssXc4gx5kRGELoNui1Fs8ou/+ACPCZSqZc7vRDw0myf1FfKlBhPzS5SKdaqNLqPLLzJ3bJ5B4PLkQws8a+Q4vbZKYhWwlk5x+c4B+rbL5l4XJeuwd+0lhDekNLnMxPIZXn71AvPTNazJWZ5YWmb7zg7tfpuHHn2aUaWArJrIuSpFfQnJLPLeDz/PrVbAoOdz4cLvk8tc1NUpnPkn2Wz3MfQiWv0ERv8y3z5/GUtkaKaKmg2witMMOmN220M2Q41J4WCgIysK9akpPM/D8SMefexhjGIVebBDoTyNP95nZnYBoUhkIqG5NyRnlvD9FnF/l6GfIhkF4iTj/R/5GGdOLuMQszS3wvZQxzRs+ht3CasRL379ZVZmygyGLnGcYRdUorhIpV48XKEXhIzGMvVKB7usYS0dx43h5u1NSvkSvu+TJoclzuL2Nl3JolbSyPQK+z2BdS9LjZErE2UpUmaws91CZDKD3hC/cxNbM5HMHIam4Q77ZJi43TaGNYN8r05jliREQsUwdIbjgEF/xOrsBGPHxZIFSexTqtrkCnm272wzOV1EAWZXFtGLJfa3DphfGBPLNm+ef42ZhRG1epFBq8NnP/1p4thCSsfMSCpyOGDn5hssnzqO1xsxszTHxvoN6lOTzB97AkPL0evtEedL5KTDytFSGpLTZXJPfIxeu8/djYtUKjVOrCzSHUZIUoFE0xg332Tp6Xfi9btcunAZmQxvPGJqYQmjPIsSjfDCEN3W2e46FBWJxMihipjAPaxhaVoRreYQJQ2xjAKFvIHb15mslYmFwGBEf699mJRlZ5fG6RMIfOIgRsgG+XKVgRsyf2IeZbtDPpVw/T0yNQeqTUHX3vL5eyAigcceOZV99b/8WzIS3FjHGfXByHPh1YucnDFxYhm328TQMvauvImtqsw9+V6EFKOaBoEfYxgGuqEQOC6ZWUKSIkxZQZMS9m7dYLc3RsvlkNVpYsUkpykcdNrUajXsYokk9knTmMiP2Ok6aIlLY3oep39AImkosiCnxgihoBensZWI61ev0Wx3WF0+jqJDFgsSWSYMMorVAomfIOsq3X6fUrnM1l4Ty9RxgpDJ3OGae1kWlG0VL5GIghaBn6FbEjmzgh9n2DKQCbYPDpicnQPJpNvZx7JsXv/mN9B0wfLyMSTTIkk8/P3XkAqraPl5qrUC7sEORmOBMHLoN9vYxRL5fIEwGEMc4jseo1igKyqmJWHIJpKmI9IQ32lRLBfYHeuEXpecplHI2WSpxPrlN5icmyF2RxRnTiLJKUno0DsYUF9aImrdwff6yFYNv3UNP8lYe/rnkKSQnY114vQwk5MTu5iqgpzENHfaLK6t4Xge23euMT05AXGEbOrcvXieIElZe+ossijgBC6DgwOkxGNy4SRCSxkfbFKcOoEXBWhpj93Xv8btN2+z/L4P0RqqTM0skMYuUhbijzokwmJqtsG+Z7P+yld4+l3vYf/mSxQbJxgM9gjHDrW50zTvvI5m5kgThYWHH0eRVCTNZOvKNfp7bzD/6AcY9prEmYYldylOrXL+D7/Jo0/MImtFzPwkchYiqzae0wNZkGYSWZah20VUkdEauUxUCgiRkYY+47FHmEFr54Dawgq3z32F48vTTK09xs72PlImMbHyMLduXyPauQHlZdaOT3JnZ0RDlVHKRbZ3muRwmZoqEptTnFlbe3C7A1NTjezX/uoHieUCJ97xNN2tO3zz9ducPfsk337xJeZrKpPz88yeeZpOe4wbdFmolBn6AaPmNqWpJUgjKrU6w3EfRVjoeRsl8bh78zr7SYmTS7Ps371CbXqNSAgm8gqRpJC3TGqlIru7XaxSCVNPGRzsE6cJlUqNKIH+yCeXN0nCEWkC23e2WTuzzN1bt5hdOk57GGPoEqoUkzo9MnMRUw/ot3coTq6wee0yEwtLSChYBZ1xr81ux6FhBxRyRVpth0xWULUKadZElQxCKY+qKwS+xOOPzbF97Saf/cxvU24UmJyYYfH4Y7T2btG6eo1hu8lHf+Vvo8ka3W6XMIjxBl3kqEuam8WUQnb6LjnDwjBskEPy+Tr7u3ssrK4w7jYJQgXZu4NdmeVga4uJqQqmXQAk9gc+uhSjSjbCyOi0R9RKZRLFQImHxKKE6zbxQhlLTtDiPhcvXuL6xduoRoaCxulTk8w9/gK9nRskeExMzuJFCUkUUy7VGYc+IklJI5dCuUGr76KqMjklIJF1xv0B48E+Rq5+6LT9MbJkgKKREuM4PqptUi4WGbW2kI0KWZqSChURdFFEgiQ0hFVic/0qty5doqdUWFhYYELaY2puDaHYDHyFMAtQ4gAtX6ezt83a8SWaG3cIQpfgYIPaQ8+gCoNK1aZ/sMne1haNuTVkWSAkDUXE5OqTjAYOhUKBMAzJsgzTUBFJSJCArOgEnkeGevjaNOyDXsAPEiBGzVuMrl9Cq9RQMhW1WGDs+EShT6k0BUrM6GCPouGj6mXu3N1nanoZkfXJtPJhQho1olBZIA6HvHHhGr/w8w/wPIG142vZ5//Dv6G9u4ewpui0tlB1malalW+/cp7VY0touoWpZRQLVc5984ucfc8LfOG3f5N3vu/93No4YHp+gULRIHRS8pU8ruchC/VwhZnXRzLLuKMOKDqxyDFRKnB74waEAbVGFb+/xZWLu5x84gSVagOkHIqISe0yd77yD5l97GcIY9ALVdJgRDTyyDQD27ZxvBhFTahUSuzujygY4LouuWKVm7euIwmVSl7GdSK80QGBVsW0y+xt7PLYYytocoofSeR0Qac7oDQxd7h6MQVFhvEoQE56ZBhIBmy//i0e//BfoL13l1RSSWOfhdUncEctUHI0N64ShX2UeIDFecgwAAAS6ElEQVRaPQaqjWwoKOM+kVApSoL2qI8qK+ilPLGfkJEQhiFxmJBKAktT0HJ5Qg8sM0UQEyUmod9FkYs4ww5GqUQUuqSRS5iaxN6QJAXiCNMUYOSQRZ7RoElOV0mTAMnMkcUSmpEx6I/Qy3OMm3fIkAh8h4WFRYgyRt19Usums7NDY/EYveYddE0GaYJC0WTsuUxMToPfJQgEty6+yszqSSStSkKIrMqoiobwPQ7cGD31KBoyMSmSkSMY7DMeD9GkhCyzMcoNmhvXqcyuoBsKIglJUYi9MUvTJTJ3gD27ysWrW0hRn26oITKVQERMVSbRFJNERBTsHGgSg1YHwjFpKjGzvIbrjYjHfXRDoTwxRxClbNy8imXlsGRIDYNKscTrF25Sn5mi22zx6OnjIKfIqoXXG3Pr7g3KjWV2717hzInj3LrTZGm6hGoXAZAVjUxKIPDwExuzYOIOB1iGidAk1pZWf6gTeCASjcpCECQKVqWKnvQZt3ZQVAMn8ZFa1/G6HXTNpD1MWN9rkreKeJmJLo144+pVHnn8NNV6jSSVUG2DIMjIohhn0EOEY1BURu1drFIeOYuJ3R53br9J1ZKoVnJoRpE4v8SpFZP65Cq6ZqOIEFkTXPjai9hnXqDlSuw1h2RJzHA4Qssf1qePREbqdVEFtJoDEr+JlTOw9JQwDGmUNWbqGvWZBWRVp7F0nFw8RunfZTrfIVfMIZs2tqWRaFXKUzO0Ok2c8ZgoGJJGYzK3Q6E+g54vkLopjZXH2djYJ1NM+oMxYxfCKEZSTfZ2biKpKUEso1eX0eKQ8f4ug902vmwfJt2cn0XTNBRJ4PWHpEmEFKtkqUwsFAbNJrJiIwmFcfuADAM3KZJEAb4vEcdjCiUFKQzxBw6qkkPEHrqhoJoKmV1AmHWCcYDrD/HDjPzkNFJpEoGKJKfEboxpWwy2NnBdnyzLUElI4whV9tFyCl57h3JBR04cDKEhFBO/v0W/18Lb2SIJXEIpj25ILK4sYubKRMmYOMiQwohgPCRWNZpXX0L1e1iFHOhFFCWHnmtQri/geQGaniETsHLqDMXaNIqkkSY6EyWDlYU5KseOUz39JHJ+AtnrU18+iQhdot2LKK3bDIY9wiTGc8c4vkM4HJHGHlP1PFbeIowTdNNAyZXRrSKd7oAoyvDGA3bu3kSxNfL5OpJ2WPYujTM8b4RqFUBopKmEVs5j52s47pDRwMELEoqqh1mbwSgUUG2bjBhFkjFyVeJgH6fVxDZ0hGaSpupbPn8PRCRw8uSJ7J/8vf+ZmRMP840Xv8aTz30AI/YxLY1zF64wNT3N5tU3mFl5GMuQ2Vy/yf4owBjcYfnsJwlHLSzVpDJZZNBqki/YJNYkUhoRhjFZKqNqAdvr+xQnq8xPTbC3vQVyjsgfMOofYOUblKolvPEIRdcY9DqUGkvEgztoUkAaq8R6gcgfkMUShVqN7l6HXKWCbak43hgPAymMcV2PRq1ItVbg1//FZ/jYhz7EYH+L+ZkS7c1NpMoi+bJBTgisSpV+u4VAhXSApNWQFIUvffYfgaYh2bMoisJ8I8+xk8+QRT6ZpuG6LoahMPZilCxk4yBjpmawd+cKC7N1YqVMoXyYq89STaJwjOuOEeIwS4+XaSzMz5BmAW5/hKIKXDcmb6RIVpX2/gGSHDC1cJpRv4fsdxnEJfIllWDgouU1+nttSlPTh92eyiTdjkMwbpMv6WSBg5+qTEzP07t7ifzMGts3rpGvTFOfmsIfN+He0uwo7FOrHaO7fYPy7CIxgiQICYRFrTFNEnaIwpTB0ENOuuiqTqwYBK6HpJWQhUSxWGQcDpG8iE7rgMbiAs7AIcoi8ppMtVYky0yuv/ktipMPoZo6cdRFEzLtfp9GqYAi5ZFiF03PaMxWSc06vuMQOx7Xr7xJIhmYlQnUJCRLYzRdRlF1bl25QnXmBPm8DVlMr9dDKAqzyycJI59Bc59yIUcah6hGHl3XcYcHDIdDbFPBKpRxE42CLYFqEzshirdLbnaNYbvPbnPEzEyRNIl48+XvYuaLLC5NgTKNpifoZhGQaG7dZGr5OMFgiG4JkC3a+wNkEWLYKg8/9PiD2x04cXwl+2v/w3/H+z72UXp3dylUqwyHTTQ1jz9oMhgHLB5bZdjaoVyZpDdsIasacaRSrhZJ44SEkCRNkaIEydTpdloYioqiWbidXQwpxpg+QRrGyGpM6ktYZYtxt8/AG+H2x8zOzyA4TM0UhTDoNSmVc6DIOC7oIiJfncIfDECViSMJPZej39pE0QsoXo928yYrj3+Q8dhHNzU6nQ5mlmFXG3h+SBaO8R0PP5KoTFTI0phub0BRz5BzZUzZxg0dvvyvfx1v6JJqARO1aR5678fYunqO0+9+gSQ10fBRNeswqac/QgiTQb9D5PZx3AGTk9M4fsbs/CrBeBev3cWsVfBCQRj69NYvMXX6aUJvSMeBnK5QLhfot/oYhkS3N2BudQ019mjutTFyeTxnTKqoyFmGpFpksYMfJgSxgyVlBLFKvlwiEymxN8Afe1SmFhn0+uiWyXjYwbJyjAd9SrU6ZBlOe4vpxRPEcYyqSMjhmICIOJbIF2uEUcbd9YtUFx4iHrUx8lXU1KO9s0FpaoFCoXD4b17Ko2R5dravYRWrjDpd7JxKuVwGJFTNQFEUAt+nVK4y3HwNu9QgjYbEsoaIY3rtAZJiMBo7jCIJzbIplOrYdp7hoE0aesRJiGKUMaSQvf0NcnaFbq9NdxgwXzIQtoVplinnNQb9hEEQMz1RYGuvyclTqxRMlSxTUQwTZzwklU00SUDiMug7bO3s0agWKU3NIpKU7Z0dnN6AlVNrJJlJ6HZIUw1nsE+5ZCMZZTRVIooiZEVFSKBJCn4coioG45GPrikYusTi0g8fGHwgXhGmKJxaW8YfxsRxSKfZ5PWr25xeXaBSm8AdXODu5g4TU5MMPB/DqpOkfSaKNjuDENvU8TstNM0ATcHveTj7A5icInQCJicmUDQTPzFJJJ8wsFBUFy+x6XWu0W5usnr8UWQ0DCtPLBKGgz1UTYCwSVFIxgfIE5NEkYqWKxM7HTJhEEYBxbyJH6Qo5QmmKg0kSSIOYlwXtDRByhdAycN4k8yyySQPvWARRRGSyJBEgmZVSRIZlIg4CHnuE58iDYZs9AMW55eQMoOF04/gOCnD7haLK/OYZp7BYEiMQuS5yEEXYVgUCkWEpGEXFMajHkKz0KsSKDZhfw9dSplZexhd1UiiHFqwg26XiKKILHORtCqGbpNGKalmoudk0jQkSRJkVcaQYsbxANM0YTxAixWELaET4jkuzmDI1MI8WdIkiiIip0ccx+imga6reGlMGnlosomQFRxfIKkCVTbxww4xGe7gAMnKk2CimSWC0RCRHvZfVdsmV58mQsVPNYSUEAYGGT4iDfAGDvFgF4xZhGSiioRUSCiKjhAeUgKliUUSEZMJjcT10CQdPxMItYSwDKwsJEGwt7tJrVIHkaJZeeQ0I0kPxxwkKUeqyCj+mMx3UIonUBIZTZVR83XmyoJ8p49RqtAYObiOj12ookkSqdAI0whN0Rh6DtVCAVkZIaUJaRTiDRwyzUBVZCr1GppqE7gxirDwksN8kZmsIkhJM4Uoihm7DsVCiUSSSJPD8Q8R9JBVG6HX3vL5eyAigVOnTmb/9B/9n7T7I5TQxZb66FlGLMek7pjR0CNJDRpnztKYO0ZzdxsvGjNsHdDevArbVykvnYLlZ2mU65AKcsUSWSY42L5Dr3vA9NwEaTRGMUukkUs+VyYLM3RdRdYEjheTyx/WJAwzlU5z695EjTxIMoP2NoXyNImkYcoJvYF7mOtP1dm9fBHV0rCri6zMl9jfblKs5nF8j0wpMW7tYxTzqHqBNItAkunvbaIbeRRdo98+wLBrWDkTVVYYOi6FXJ6DzWtImoaqWTgJzE5PEIxaDL2ISmWawBsSJxl+HJEEIbac4TgOhWqZTDIJQ4+d3Rb9oY+QZJZX5pAin+mlNWShkGYZ3f0NhCojkowg8gGII0HOVEiSiDSVkOQYbzDEqk0j0hQhS3jj7mF9lNREUiWSyCX2YoRp43sO5XKZMIqIHI9C2UJSTNxeB18xqJUtvL09pHwZSbeJ/QGyLJHFMjESIeD2+5SrZYSQkBTwh0MiFMqlIpJQyUKX4XiEZhrIkYsqIJc3sYo6WZYhhHxPv4WhqyRJhCzpBKGDoqk4nowzHpJlCflaDd2oIKSMy6+fJ/U7FKvzkEWEkYNqFJFlGUnWGPYcDFtBVaDfbSMLhWpJY//uXaTyEitzdWTdwPVipheOEccxgT8mV7DZu30b1xkxs3Sc3tCns3OTXKmCQoxhFdCtHKQeo/4IWbMPHY9qEKQyge9TKdikSYDQTJxhBy/MaFSqhP1tBm4KpDQWF4mihJsXXqNaazC9dBzSFCfweeTk6Qe3O7C6upz95v/7T1HjDDOfZ+D4CDSiYIgwC+hpRq+9R21ilkjIqLJPtz8gZ+VRpAhFL+GMxuiGiqxKSBKMumMkRUc3dVRZIQl8fHdIYlk427eoTi+iF8p093bQCyWSJEMXAaHnIswa/riDbQjubu6ytHocJA3Pcxgd7JFoBfJliyQCXYlIgpjp2QWmFlZwvRGaYYNQ6bf2ae3eIUwhjUO0fBU5dojcAK1QIQxDIneAn8mUyhMoYgyZTOQHKLqGIhQO1q+RCIVMMdDtHKqqYhdtklAwGB32KTVdx3EjnH4T3cwjKxlf/cq32e4PKBeLSJKErmgUKyXsnM7MRINqqYiUREhphmrnCNOQsNdBtTTQcrhjB0OWkBSBFyQk8eFDFiEIPI/u5jpWqUihWCbTCkT++F7JsAQ5OyyDLokMZzhAzlI0TUMt1PGHbWJVp1owSZ0hTmpj2zZZlhH4LpKqINIUWVVJM4EzHqClMYmi4Acx040aZs5GBF3SNCWTJPrNA8ZOHxKZzCgiSSqWZRC4AappIKsWYXzYjw8GXTSzQBCFKGmMkBWG/QHFag1ZZMiaRCopxEJBuB0yoeL5DlqmoJYqjMYD3IO7NMomWW6BwXjAXKOBZpuopoWcJShqgSQ7LOzSHQQYijjs72cZUSrjOy5p6qPbOdIwoFAsE2Uc1lyUIYoEu90ReVujXqvgDkakaUqpXCBLI2RJoz8aksQy7riDpWRY8YCkNEUq5RFCMB4FjAd3mGjUDsvaKQZPnH3/g+sEhBAtwAHeOjn620+NIz1/Eg+apiM9fzwLWZbVf7DxgXACAEKIcz/MS90vjvT8yTxomo70/Gg8EPMEjjjiiPvHkRM44oifcB4kJ/D/3G8BP8CRnj+ZB03TkZ4fgQdmTOCII464PzxIkcARRxxxH7jvTkAI8VNCiOtCiFtCiF+9Txo2hBAXhRBvCCHO3WurCCH+QAhx896+/GPW8JtCiKYQ4tL3tf1QDeKQf3zPZheEEI+/TXr+jhBi556d3hBCfOT7vvtf7+m5LoT48I9Bz5wQ4o+EEFeFEJeFEP/jvfb7aaO30nTf7PQjkWXZfdsAGbgNLAMa8CZw6j7o2ABqP9D2fwG/eu/4V4Ff/zFreB54HLj0J2kAPgJ8GRDAWeDlt0nP3wH+px9y7ql7904Hlu7dU/nPWM8U8Pi94zxw495176eN3krTfbPTj7Ld70jgKeBWlmXrWZaFwGeBF+6zpu/xAvDpe8efBj7x47xYlmXfALr/lRpeAH4rO+S7QEkIMfU26HkrXgA+m2VZkGXZHeAWh/f2z1LPXpZl5+8dj4CrwAz310Zvpemt+LHb6UfhfjuBGWDr+z5v88cb8cdFBnxFCPGaEOKX77VNZFm2B4c3G2jcB11vpeF+2u2v3wuvf/P7ukhvqx4hxCLwGPAyD4iNfkATPAB2+q/lfjsB8UPa7sfrindmWfY48NPArwghnr8PGv403C+7/QtgBXgU2AP+wdutRwiRA34H+JtZlg3/uFPvo6b7bqc/DffbCWwDc9/3eRbYfbtFZFm2e2/fBH6XwxDt4Hvh47198+3W9cdouC92y7LsIMuyJMuyFPhX/P+h7NuiRwihcviw/XaWZZ+/13xfbfTDNN1vO/1pud9O4FVgVQixJITQgE8BX3w7BQghbCFE/nvHwIeAS/d0/OV7p/1l4Atvp657vJWGLwJ/6d4I+Flg8L2Q+MfJD/SpP8mhnb6n51NCCF0IsQSsAq/8GV9bAL8BXM2y7B9+31f3zUZvpel+2ulH4n6PTHI4inuDw5HSX7sP11/mcMT2TeDy9zQAVeAPgZv39pUfs45/x2HoGHH4j/FX3koDh2HlP7tns4vAE2+Tnn9773oXOPxBT33f+b92T8914Kd/DHqe4zB0vgC8cW/7yH220Vtpum92+lG2oxmDRxzxE8797g4cccQR95kjJ3DEET/hHDmBI474CefICRxxxE84R07giCN+wjlyAkcc8RPOkRM44oifcI6cwBFH/ITz/wH1u8Q+KinYaAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "img = image.load_img(img_path, target_size=(299, 299))\n",
    "plt.imshow(img)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(299, 299, 3)\n",
      "(1, 299, 299, 3)\n",
      "(1, 299, 299, 3)\n",
      "(1, 1000)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[[('n02099601', 'golden_retriever', 0.47158465),\n",
       "  ('n02102318', 'cocker_spaniel', 0.12101043),\n",
       "  ('n02099712', 'Labrador_retriever', 0.08217219),\n",
       "  ('n02101556', 'clumber', 0.021299073),\n",
       "  ('n02102480', 'Sussex_spaniel', 0.01892862)]]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = image.img_to_array(img)\n",
    "print(x.shape)\n",
    "x = np.expand_dims(x, axis=0)\n",
    "print(x.shape)\n",
    "x = tf.keras.applications.xception.preprocess_input(x)\n",
    "print(x.shape)\n",
    "res=xcept.predict(x)\n",
    "print(res.shape)\n",
    "tf.keras.applications.xception.decode_predictions(res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "del vgg19, xcept"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:tf2] *",
   "language": "python",
   "name": "conda-env-tf2-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}