Heartbeat Classifier.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "import itertools\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "# Audio\n",
    "import librosa\n",
    "import librosa.display\n",
    "\n",
    "# Scikit learn\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report, accuracy_score, confusion_matrix\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.utils import shuffle\n",
    "from sklearn.utils import class_weight\n",
    "\n",
    "# Keras\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers import Convolution2D, Conv2D, MaxPooling2D, GlobalAveragePooling2D\n",
    "from keras.utils import to_categorical\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore', category=FutureWarning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading Dataset\n",
    "\n",
    "Considering only more than 3 sec files and\n",
    "labelling the files as Normal or Abnormal"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TODO: Add artifacts (i.e. noises without heartbeat data) and ask the user to try again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1281 entries, 1241 to 1126\n",
      "Data columns (total 3 columns):\n",
      "filename    1281 non-null object\n",
      "label       1281 non-null object\n",
      "offset      1281 non-null float64\n",
      "dtypes: float64(1), object(2)\n",
      "memory usage: 40.0+ KB\n"
     ]
    }
   ],
   "source": [
    "dataset = []\n",
    "for folder in [\"./set_a/**\",\"./set_b/**\"]:\n",
    "    for filename in glob.iglob(folder):\n",
    "        if os.path.exists(filename):\n",
    "            label = os.path.basename(filename).split(\"_\")[0]\n",
    "            duration = librosa.get_duration(filename=filename)\n",
    "            # skip audio smaller than 3 secs\n",
    "            if duration>=3:\n",
    "                slice_size = 3\n",
    "                iterations = int((duration-slice_size)/(slice_size-1))\n",
    "                iterations += 1\n",
    "#                 initial_offset = (duration % slice_size)/2\n",
    "                initial_offset = (duration - ((iterations*(slice_size-1))+1))/2\n",
    "                if label not in [\"Aunlabelledtest\", \"Bunlabelledtest\", \"artifact\"]:\n",
    "                    for i in range(iterations):\n",
    "                        offset = initial_offset + i*(slice_size-1)\n",
    "                        if (label == \"normal\"):\n",
    "                            dataset.append({\n",
    "                                \"filename\": filename,\n",
    "                                \"label\": \"normal\",\n",
    "                                \"offset\": offset\n",
    "                            })\n",
    "                        else:\n",
    "                            dataset.append({\n",
    "                                \"filename\": filename,\n",
    "                                \"label\": \"abnormal\",\n",
    "                                \"offset\": offset\n",
    "                            })\n",
    "                        \n",
    "dataset = pd.DataFrame(dataset)\n",
    "dataset = shuffle(dataset, random_state=42)\n",
    "dataset.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(4,6))\n",
    "dataset.label.value_counts().plot(kind='bar', title=\"Dataset distribution\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Split dataset in train and test\n",
    "80-20 Ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train: 1024\n",
      "Test: 257\n"
     ]
    }
   ],
   "source": [
    "train, test = train_test_split(dataset, test_size=0.2, random_state=42)\n",
    "\n",
    "print(\"Train: %i\" % len(train))\n",
    "print(\"Test: %i\" % len(test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualising the data and features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./set_a/murmur__201108222246.wav\n",
      "./set_b/normal__127_1306764300147_A.wav\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "idx = 0\n",
    "for label in dataset.label.unique():    \n",
    "    y, sr = librosa.load(dataset[dataset.label==label].filename.iloc[33], duration=3)\n",
    "    print(dataset[dataset.label==label].filename.iloc[33])\n",
    "    \n",
    "    # Wave plot\n",
    "    idx+=1\n",
    "    plt.subplot(2, 3, idx)\n",
    "    plt.title(\"%s waveplot\" % label)\n",
    "    librosa.display.waveplot(y, sr=sr)\n",
    "\n",
    "    # Mel Spectrogram\n",
    "    idx+=1\n",
    "    plt.subplot(2, 3, idx)\n",
    "    S = librosa.feature.melspectrogram(y, sr=sr, n_fft=2048, hop_length=512, n_mels=128)\n",
    "    S_DB = librosa.power_to_db(S, ref=np.max)\n",
    "    librosa.display.specshow(S_DB, sr=sr, hop_length=512, x_axis='time', y_axis='mel')\n",
    "    plt.title(\"%s mel spectogram\" % label)\n",
    "\n",
    "    # MFCC (Mel spectrogram)\n",
    "    idx+=1\n",
    "    mfccs = librosa.feature.mfcc(S=librosa.power_to_db(S), n_mfcc=40)\n",
    "    plt.subplot(2, 3, idx)\n",
    "    librosa.display.specshow(mfccs, x_axis='time')\n",
    "    plt.title(\"%s mfcc(Mel Spectrogram)\" % label)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extracting features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_features(audio_path,offset):\n",
    "#     y, sr = librosa.load(audio_path, duration=3)\n",
    "    y, sr = librosa.load(audio_path, offset=offset, duration=3)\n",
    "#     y = librosa.util.normalize(y)\n",
    "    \n",
    "    S = librosa.feature.melspectrogram(y, sr=sr, n_fft=2048, \n",
    "                                   hop_length=512, \n",
    "                                   n_mels=128)\n",
    "    mfccs = librosa.feature.mfcc(S=librosa.power_to_db(S), n_mfcc=40)\n",
    "\n",
    "#     mfccs = librosa.feature.mfcc(y=y, sr=sr, n_mfcc=40)\n",
    "    return mfccs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1024/1024 [01:14<00:00, 13.79it/s]\n",
      "100%|██████████| 257/257 [00:20<00:00, 12.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X train: (1024, 40, 130)\n",
      "X test: (257, 40, 130)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "x_train = []\n",
    "x_test = []\n",
    "\n",
    "for idx in tqdm(range(len(train))):\n",
    "    x_train.append(extract_features(train.filename.iloc[idx],train.offset.iloc[idx]))\n",
    "\n",
    "for idx in tqdm(range(len(test))):\n",
    "    x_test.append(extract_features(test.filename.iloc[idx],test.offset.iloc[idx]))\n",
    "    \n",
    "x_test = np.asarray(x_test)\n",
    "x_train = np.asarray(x_train)\n",
    "\n",
    "print(\"X train:\", x_train.shape)\n",
    "print(\"X test:\", x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Encode Labels\n",
    "encoder = LabelEncoder()\n",
    "encoder.fit(train.label)\n",
    "\n",
    "y_train = encoder.transform(train.label)\n",
    "y_test = encoder.transform(test.label)\n",
    "\n",
    "# Compute class weights\n",
    "class_weights = class_weight.compute_class_weight('balanced',np.unique(y_train),y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Input Shapes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X train: (1024, 40, 130, 1)\n",
      "Y train: (1024, 2)\n",
      "X test: (257, 40, 130, 1)\n",
      "Y test: (257, 2)\n"
     ]
    }
   ],
   "source": [
    "x_train = x_train.reshape(x_train.shape[0], x_train.shape[1], x_train.shape[2], 1)\n",
    "x_test = x_test.reshape(x_test.shape[0], x_test.shape[1], x_test.shape[2], 1)\n",
    "y_train = to_categorical(y_train)\n",
    "y_test = to_categorical(y_test)\n",
    "\n",
    "print(\"X train:\", x_train.shape)\n",
    "print(\"Y train:\", y_train.shape)\n",
    "print(\"X test:\", x_test.shape)\n",
    "print(\"Y test:\", y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Building Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /Users/manan/anaconda3/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:4070: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
      "\n",
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 39, 129, 16)       80        \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 19, 64, 16)        0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 19, 64, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 18, 63, 32)        2080      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 9, 31, 32)         0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 9, 31, 32)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 8, 30, 64)         8256      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 4, 15, 64)         0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 4, 15, 64)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 3, 14, 128)        32896     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 1, 7, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 1, 7, 128)         0         \n",
      "_________________________________________________________________\n",
      "global_average_pooling2d_1 ( (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 2)                 258       \n",
      "=================================================================\n",
      "Total params: 43,570\n",
      "Trainable params: 43,570\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Conv2D(filters=16, kernel_size=2, input_shape=(x_train.shape[1],x_train.shape[2],x_train.shape[3]), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "model.add(Conv2D(filters=32, kernel_size=2, activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "model.add(Conv2D(filters=64, kernel_size=2, activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Dropout(0.2))\n",
    "\n",
    "model.add(Conv2D(filters=128, kernel_size=2, activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(GlobalAveragePooling2D())\n",
    "\n",
    "model.add(Dense(len(encoder.classes_), activation='softmax'))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compiling Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "adam = keras.optimizers.Adam(learning_rate=0.001)\n",
    "model.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer='adam')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fitting Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /Users/manan/anaconda3/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n",
      "Train on 1024 samples, validate on 257 samples\n",
      "Epoch 1/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 8.1785 - accuracy: 0.5059 - val_loss: 0.6467 - val_accuracy: 0.6031\n",
      "Epoch 2/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 3.4987 - accuracy: 0.6016 - val_loss: 1.0331 - val_accuracy: 0.3891\n",
      "Epoch 3/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 2.6846 - accuracy: 0.5293 - val_loss: 1.1105 - val_accuracy: 0.6109\n",
      "Epoch 4/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 1.9415 - accuracy: 0.5635 - val_loss: 0.7511 - val_accuracy: 0.5175\n",
      "Epoch 5/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 1.4412 - accuracy: 0.5625 - val_loss: 0.6332 - val_accuracy: 0.6187\n",
      "Epoch 6/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 1.0230 - accuracy: 0.6006 - val_loss: 0.6269 - val_accuracy: 0.6654\n",
      "Epoch 7/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.9711 - accuracy: 0.5928 - val_loss: 0.6538 - val_accuracy: 0.6148\n",
      "Epoch 8/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.8383 - accuracy: 0.5928 - val_loss: 0.6330 - val_accuracy: 0.6537\n",
      "Epoch 9/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.7792 - accuracy: 0.6064 - val_loss: 0.6421 - val_accuracy: 0.6265\n",
      "Epoch 10/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.7617 - accuracy: 0.6045 - val_loss: 0.6444 - val_accuracy: 0.6498\n",
      "Epoch 11/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.7312 - accuracy: 0.6143 - val_loss: 0.6426 - val_accuracy: 0.6576\n",
      "Epoch 12/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.7079 - accuracy: 0.6289 - val_loss: 0.6372 - val_accuracy: 0.6576\n",
      "Epoch 13/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.6747 - accuracy: 0.6309 - val_loss: 0.6377 - val_accuracy: 0.6381\n",
      "Epoch 14/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6786 - accuracy: 0.6543 - val_loss: 0.6330 - val_accuracy: 0.6732\n",
      "Epoch 15/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6202 - accuracy: 0.6729 - val_loss: 0.6297 - val_accuracy: 0.6809\n",
      "Epoch 16/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.6241 - accuracy: 0.6484 - val_loss: 0.6304 - val_accuracy: 0.6576\n",
      "Epoch 17/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.6397 - accuracy: 0.6494 - val_loss: 0.6279 - val_accuracy: 0.6654\n",
      "Epoch 18/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6275 - accuracy: 0.6816 - val_loss: 0.6296 - val_accuracy: 0.6615\n",
      "Epoch 19/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6083 - accuracy: 0.6846 - val_loss: 0.6284 - val_accuracy: 0.6732\n",
      "Epoch 20/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6193 - accuracy: 0.6855 - val_loss: 0.6272 - val_accuracy: 0.6848\n",
      "Epoch 21/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6390 - accuracy: 0.6543 - val_loss: 0.6336 - val_accuracy: 0.6148\n",
      "Epoch 22/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.6223 - accuracy: 0.6777 - val_loss: 0.6321 - val_accuracy: 0.6693\n",
      "Epoch 23/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.6092 - accuracy: 0.6768 - val_loss: 0.6333 - val_accuracy: 0.6381\n",
      "Epoch 24/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5910 - accuracy: 0.6963 - val_loss: 0.6319 - val_accuracy: 0.6809\n",
      "Epoch 25/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5972 - accuracy: 0.6885 - val_loss: 0.6259 - val_accuracy: 0.6459\n",
      "Epoch 26/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5942 - accuracy: 0.6914 - val_loss: 0.6200 - val_accuracy: 0.6770\n",
      "Epoch 27/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.5828 - accuracy: 0.6924 - val_loss: 0.6128 - val_accuracy: 0.6926\n",
      "Epoch 28/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5812 - accuracy: 0.7285 - val_loss: 0.6110 - val_accuracy: 0.6848\n",
      "Epoch 29/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5773 - accuracy: 0.7129 - val_loss: 0.6123 - val_accuracy: 0.6770\n",
      "Epoch 30/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5809 - accuracy: 0.6982 - val_loss: 0.6132 - val_accuracy: 0.6926\n",
      "Epoch 31/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5827 - accuracy: 0.7002 - val_loss: 0.6186 - val_accuracy: 0.6654\n",
      "Epoch 32/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5643 - accuracy: 0.7246 - val_loss: 0.6173 - val_accuracy: 0.7043\n",
      "Epoch 33/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5722 - accuracy: 0.7051 - val_loss: 0.6129 - val_accuracy: 0.6926\n",
      "Epoch 34/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5826 - accuracy: 0.7021 - val_loss: 0.6054 - val_accuracy: 0.7004\n",
      "Epoch 35/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5708 - accuracy: 0.7197 - val_loss: 0.6031 - val_accuracy: 0.7004\n",
      "Epoch 36/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5773 - accuracy: 0.7100 - val_loss: 0.5987 - val_accuracy: 0.7004\n",
      "Epoch 37/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5436 - accuracy: 0.7383 - val_loss: 0.5906 - val_accuracy: 0.6965\n",
      "Epoch 38/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5584 - accuracy: 0.7236 - val_loss: 0.5897 - val_accuracy: 0.6965\n",
      "Epoch 39/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5547 - accuracy: 0.7246 - val_loss: 0.5934 - val_accuracy: 0.7004\n",
      "Epoch 40/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5512 - accuracy: 0.7246 - val_loss: 0.5892 - val_accuracy: 0.6926\n",
      "Epoch 41/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5350 - accuracy: 0.7471 - val_loss: 0.5860 - val_accuracy: 0.6926\n",
      "Epoch 42/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5573 - accuracy: 0.7119 - val_loss: 0.5837 - val_accuracy: 0.6965\n",
      "Epoch 43/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5451 - accuracy: 0.7148 - val_loss: 0.5844 - val_accuracy: 0.6926\n",
      "Epoch 44/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5442 - accuracy: 0.7373 - val_loss: 0.5805 - val_accuracy: 0.7082\n",
      "Epoch 45/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5558 - accuracy: 0.7305 - val_loss: 0.5810 - val_accuracy: 0.7043\n",
      "Epoch 46/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5417 - accuracy: 0.7383 - val_loss: 0.5774 - val_accuracy: 0.7004\n",
      "Epoch 47/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5460 - accuracy: 0.7314 - val_loss: 0.5752 - val_accuracy: 0.7121\n",
      "Epoch 48/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5488 - accuracy: 0.7314 - val_loss: 0.5723 - val_accuracy: 0.6965\n",
      "Epoch 49/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5452 - accuracy: 0.7334 - val_loss: 0.5730 - val_accuracy: 0.7160\n",
      "Epoch 50/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5505 - accuracy: 0.7256 - val_loss: 0.5762 - val_accuracy: 0.7121\n",
      "Epoch 51/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5371 - accuracy: 0.7461 - val_loss: 0.5708 - val_accuracy: 0.7082\n",
      "Epoch 52/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5436 - accuracy: 0.7451 - val_loss: 0.5676 - val_accuracy: 0.7121\n",
      "Epoch 53/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5523 - accuracy: 0.7207 - val_loss: 0.5707 - val_accuracy: 0.7121\n",
      "Epoch 54/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5340 - accuracy: 0.7217 - val_loss: 0.5686 - val_accuracy: 0.7043\n",
      "Epoch 55/300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5433 - accuracy: 0.7275 - val_loss: 0.5643 - val_accuracy: 0.7160\n",
      "Epoch 56/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5275 - accuracy: 0.7383 - val_loss: 0.5615 - val_accuracy: 0.7160\n",
      "Epoch 57/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5346 - accuracy: 0.7285 - val_loss: 0.5644 - val_accuracy: 0.7121\n",
      "Epoch 58/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5325 - accuracy: 0.7256 - val_loss: 0.5637 - val_accuracy: 0.7043\n",
      "Epoch 59/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5373 - accuracy: 0.7305 - val_loss: 0.5641 - val_accuracy: 0.7237\n",
      "Epoch 60/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5298 - accuracy: 0.7539 - val_loss: 0.5645 - val_accuracy: 0.7198\n",
      "Epoch 61/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5021 - accuracy: 0.7490 - val_loss: 0.5592 - val_accuracy: 0.7237\n",
      "Epoch 62/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5128 - accuracy: 0.7480 - val_loss: 0.5570 - val_accuracy: 0.7198\n",
      "Epoch 63/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5276 - accuracy: 0.7373 - val_loss: 0.5598 - val_accuracy: 0.7121\n",
      "Epoch 64/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5296 - accuracy: 0.7441 - val_loss: 0.5606 - val_accuracy: 0.7160\n",
      "Epoch 65/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5219 - accuracy: 0.7490 - val_loss: 0.5579 - val_accuracy: 0.7198\n",
      "Epoch 66/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5199 - accuracy: 0.7471 - val_loss: 0.5609 - val_accuracy: 0.7237\n",
      "Epoch 67/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5254 - accuracy: 0.7539 - val_loss: 0.5606 - val_accuracy: 0.7160\n",
      "Epoch 68/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5158 - accuracy: 0.7520 - val_loss: 0.5598 - val_accuracy: 0.7354\n",
      "Epoch 69/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5127 - accuracy: 0.7461 - val_loss: 0.5545 - val_accuracy: 0.7354\n",
      "Epoch 70/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5136 - accuracy: 0.7451 - val_loss: 0.5457 - val_accuracy: 0.7354\n",
      "Epoch 71/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5094 - accuracy: 0.7451 - val_loss: 0.5496 - val_accuracy: 0.7354\n",
      "Epoch 72/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5073 - accuracy: 0.7529 - val_loss: 0.5439 - val_accuracy: 0.7354\n",
      "Epoch 73/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5135 - accuracy: 0.7510 - val_loss: 0.5445 - val_accuracy: 0.7315\n",
      "Epoch 74/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5065 - accuracy: 0.7451 - val_loss: 0.5449 - val_accuracy: 0.7354\n",
      "Epoch 75/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5042 - accuracy: 0.7549 - val_loss: 0.5491 - val_accuracy: 0.7276\n",
      "Epoch 76/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5214 - accuracy: 0.7363 - val_loss: 0.5435 - val_accuracy: 0.7354\n",
      "Epoch 77/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4990 - accuracy: 0.7676 - val_loss: 0.5371 - val_accuracy: 0.7393\n",
      "Epoch 78/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5092 - accuracy: 0.7373 - val_loss: 0.5413 - val_accuracy: 0.7237\n",
      "Epoch 79/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5339 - accuracy: 0.7197 - val_loss: 0.5448 - val_accuracy: 0.7237\n",
      "Epoch 80/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4981 - accuracy: 0.7627 - val_loss: 0.5624 - val_accuracy: 0.7082\n",
      "Epoch 81/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4934 - accuracy: 0.7607 - val_loss: 0.5400 - val_accuracy: 0.7276\n",
      "Epoch 82/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4855 - accuracy: 0.7607 - val_loss: 0.5380 - val_accuracy: 0.7315\n",
      "Epoch 83/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5052 - accuracy: 0.7617 - val_loss: 0.5387 - val_accuracy: 0.7315\n",
      "Epoch 84/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4942 - accuracy: 0.7617 - val_loss: 0.5306 - val_accuracy: 0.7354\n",
      "Epoch 85/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.5041 - accuracy: 0.7568 - val_loss: 0.5421 - val_accuracy: 0.7237\n",
      "Epoch 86/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4889 - accuracy: 0.7754 - val_loss: 0.5369 - val_accuracy: 0.7315\n",
      "Epoch 87/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4852 - accuracy: 0.7705 - val_loss: 0.5417 - val_accuracy: 0.7315\n",
      "Epoch 88/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4999 - accuracy: 0.7695 - val_loss: 0.5344 - val_accuracy: 0.7276\n",
      "Epoch 89/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4898 - accuracy: 0.7656 - val_loss: 0.5382 - val_accuracy: 0.7432\n",
      "Epoch 90/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4799 - accuracy: 0.7666 - val_loss: 0.5307 - val_accuracy: 0.7393\n",
      "Epoch 91/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4758 - accuracy: 0.7725 - val_loss: 0.5334 - val_accuracy: 0.7315\n",
      "Epoch 92/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4930 - accuracy: 0.7656 - val_loss: 0.5451 - val_accuracy: 0.7198\n",
      "Epoch 93/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4941 - accuracy: 0.7490 - val_loss: 0.5520 - val_accuracy: 0.7237\n",
      "Epoch 94/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4912 - accuracy: 0.7637 - val_loss: 0.5383 - val_accuracy: 0.7354\n",
      "Epoch 95/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4867 - accuracy: 0.7637 - val_loss: 0.5371 - val_accuracy: 0.7315\n",
      "Epoch 96/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4787 - accuracy: 0.7725 - val_loss: 0.5383 - val_accuracy: 0.7276\n",
      "Epoch 97/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4722 - accuracy: 0.7832 - val_loss: 0.5312 - val_accuracy: 0.7354\n",
      "Epoch 98/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4748 - accuracy: 0.7754 - val_loss: 0.5298 - val_accuracy: 0.7354\n",
      "Epoch 99/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4800 - accuracy: 0.7686 - val_loss: 0.5435 - val_accuracy: 0.7354\n",
      "Epoch 100/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4704 - accuracy: 0.7793 - val_loss: 0.5343 - val_accuracy: 0.7393\n",
      "Epoch 101/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4698 - accuracy: 0.7852 - val_loss: 0.5335 - val_accuracy: 0.7393\n",
      "Epoch 102/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4843 - accuracy: 0.7656 - val_loss: 0.5180 - val_accuracy: 0.7432\n",
      "Epoch 103/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4781 - accuracy: 0.7715 - val_loss: 0.5370 - val_accuracy: 0.7393\n",
      "Epoch 104/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4679 - accuracy: 0.7744 - val_loss: 0.5371 - val_accuracy: 0.7393\n",
      "Epoch 105/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4612 - accuracy: 0.7852 - val_loss: 0.5241 - val_accuracy: 0.7471\n",
      "Epoch 106/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4783 - accuracy: 0.7754 - val_loss: 0.5292 - val_accuracy: 0.7393\n",
      "Epoch 107/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4731 - accuracy: 0.7773 - val_loss: 0.5338 - val_accuracy: 0.7315\n",
      "Epoch 108/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4682 - accuracy: 0.7764 - val_loss: 0.5314 - val_accuracy: 0.7432\n",
      "Epoch 109/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4745 - accuracy: 0.7832 - val_loss: 0.5249 - val_accuracy: 0.7432\n",
      "Epoch 110/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4771 - accuracy: 0.7695 - val_loss: 0.5589 - val_accuracy: 0.7237\n",
      "Epoch 111/300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4718 - accuracy: 0.7832 - val_loss: 0.5270 - val_accuracy: 0.7354\n",
      "Epoch 112/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4583 - accuracy: 0.7832 - val_loss: 0.5133 - val_accuracy: 0.7471\n",
      "Epoch 113/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4573 - accuracy: 0.7842 - val_loss: 0.5157 - val_accuracy: 0.7510\n",
      "Epoch 114/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4544 - accuracy: 0.7842 - val_loss: 0.5190 - val_accuracy: 0.7471\n",
      "Epoch 115/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4620 - accuracy: 0.7773 - val_loss: 0.5251 - val_accuracy: 0.7471\n",
      "Epoch 116/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4547 - accuracy: 0.7861 - val_loss: 0.5225 - val_accuracy: 0.7354\n",
      "Epoch 117/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4437 - accuracy: 0.7959 - val_loss: 0.5174 - val_accuracy: 0.7432\n",
      "Epoch 118/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4537 - accuracy: 0.7900 - val_loss: 0.5283 - val_accuracy: 0.7471\n",
      "Epoch 119/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4394 - accuracy: 0.7832 - val_loss: 0.5096 - val_accuracy: 0.7549\n",
      "Epoch 120/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4601 - accuracy: 0.7822 - val_loss: 0.5181 - val_accuracy: 0.7432\n",
      "Epoch 121/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4485 - accuracy: 0.7822 - val_loss: 0.5043 - val_accuracy: 0.7588\n",
      "Epoch 122/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4279 - accuracy: 0.7930 - val_loss: 0.5186 - val_accuracy: 0.7393\n",
      "Epoch 123/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4416 - accuracy: 0.7930 - val_loss: 0.5245 - val_accuracy: 0.7393\n",
      "Epoch 124/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4358 - accuracy: 0.7842 - val_loss: 0.5067 - val_accuracy: 0.7510\n",
      "Epoch 125/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4543 - accuracy: 0.7832 - val_loss: 0.5302 - val_accuracy: 0.7432\n",
      "Epoch 126/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4465 - accuracy: 0.7803 - val_loss: 0.5078 - val_accuracy: 0.7588\n",
      "Epoch 127/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4245 - accuracy: 0.8037 - val_loss: 0.5237 - val_accuracy: 0.7471\n",
      "Epoch 128/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4406 - accuracy: 0.7949 - val_loss: 0.5118 - val_accuracy: 0.7549\n",
      "Epoch 129/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4349 - accuracy: 0.7998 - val_loss: 0.4985 - val_accuracy: 0.7588\n",
      "Epoch 130/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4399 - accuracy: 0.7969 - val_loss: 0.5092 - val_accuracy: 0.7471\n",
      "Epoch 131/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4275 - accuracy: 0.7910 - val_loss: 0.5088 - val_accuracy: 0.7393\n",
      "Epoch 132/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4486 - accuracy: 0.7793 - val_loss: 0.5342 - val_accuracy: 0.7160\n",
      "Epoch 133/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4259 - accuracy: 0.7949 - val_loss: 0.4967 - val_accuracy: 0.7549\n",
      "Epoch 134/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4244 - accuracy: 0.7930 - val_loss: 0.5258 - val_accuracy: 0.7393\n",
      "Epoch 135/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4357 - accuracy: 0.7881 - val_loss: 0.5067 - val_accuracy: 0.7471\n",
      "Epoch 136/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4491 - accuracy: 0.7812 - val_loss: 0.5141 - val_accuracy: 0.7588\n",
      "Epoch 137/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4181 - accuracy: 0.8125 - val_loss: 0.4999 - val_accuracy: 0.7626\n",
      "Epoch 138/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4277 - accuracy: 0.7959 - val_loss: 0.5137 - val_accuracy: 0.7393\n",
      "Epoch 139/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4203 - accuracy: 0.7949 - val_loss: 0.5008 - val_accuracy: 0.7471\n",
      "Epoch 140/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4217 - accuracy: 0.7930 - val_loss: 0.5058 - val_accuracy: 0.7510\n",
      "Epoch 141/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4409 - accuracy: 0.7920 - val_loss: 0.4966 - val_accuracy: 0.7704\n",
      "Epoch 142/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4147 - accuracy: 0.8018 - val_loss: 0.5037 - val_accuracy: 0.7510\n",
      "Epoch 143/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4190 - accuracy: 0.7930 - val_loss: 0.5084 - val_accuracy: 0.7315\n",
      "Epoch 144/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4307 - accuracy: 0.7920 - val_loss: 0.4967 - val_accuracy: 0.7549\n",
      "Epoch 145/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4229 - accuracy: 0.7939 - val_loss: 0.5059 - val_accuracy: 0.7549\n",
      "Epoch 146/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4411 - accuracy: 0.7861 - val_loss: 0.4851 - val_accuracy: 0.7665\n",
      "Epoch 147/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.4323 - accuracy: 0.8018 - val_loss: 0.5090 - val_accuracy: 0.7432\n",
      "Epoch 148/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4431 - accuracy: 0.7822 - val_loss: 0.5092 - val_accuracy: 0.7393\n",
      "Epoch 149/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4161 - accuracy: 0.8066 - val_loss: 0.4758 - val_accuracy: 0.7782\n",
      "Epoch 150/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4074 - accuracy: 0.8135 - val_loss: 0.4935 - val_accuracy: 0.7782\n",
      "Epoch 151/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4011 - accuracy: 0.8203 - val_loss: 0.4842 - val_accuracy: 0.7665\n",
      "Epoch 152/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4042 - accuracy: 0.8066 - val_loss: 0.4946 - val_accuracy: 0.7704\n",
      "Epoch 153/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4022 - accuracy: 0.8164 - val_loss: 0.4919 - val_accuracy: 0.7471\n",
      "Epoch 154/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4303 - accuracy: 0.7910 - val_loss: 0.4853 - val_accuracy: 0.7549\n",
      "Epoch 155/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4184 - accuracy: 0.8037 - val_loss: 0.5075 - val_accuracy: 0.7704\n",
      "Epoch 156/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3947 - accuracy: 0.8164 - val_loss: 0.4628 - val_accuracy: 0.7704\n",
      "Epoch 157/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3979 - accuracy: 0.8193 - val_loss: 0.4933 - val_accuracy: 0.7782\n",
      "Epoch 158/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4110 - accuracy: 0.8018 - val_loss: 0.4862 - val_accuracy: 0.7510\n",
      "Epoch 159/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4079 - accuracy: 0.8193 - val_loss: 0.4845 - val_accuracy: 0.7393\n",
      "Epoch 160/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3946 - accuracy: 0.8223 - val_loss: 0.4922 - val_accuracy: 0.7471\n",
      "Epoch 161/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3973 - accuracy: 0.8037 - val_loss: 0.4796 - val_accuracy: 0.7510\n",
      "Epoch 162/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3894 - accuracy: 0.8213 - val_loss: 0.4783 - val_accuracy: 0.7588\n",
      "Epoch 163/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3984 - accuracy: 0.8223 - val_loss: 0.4770 - val_accuracy: 0.7782\n",
      "Epoch 164/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3998 - accuracy: 0.8125 - val_loss: 0.4852 - val_accuracy: 0.7704\n",
      "Epoch 165/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4029 - accuracy: 0.8037 - val_loss: 0.4782 - val_accuracy: 0.7588\n",
      "Epoch 166/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3947 - accuracy: 0.8105 - val_loss: 0.4699 - val_accuracy: 0.7704\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 167/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3844 - accuracy: 0.8174 - val_loss: 0.4729 - val_accuracy: 0.7704\n",
      "Epoch 168/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3916 - accuracy: 0.8203 - val_loss: 0.4739 - val_accuracy: 0.7704\n",
      "Epoch 169/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3909 - accuracy: 0.8271 - val_loss: 0.4737 - val_accuracy: 0.7471\n",
      "Epoch 170/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4104 - accuracy: 0.8066 - val_loss: 0.4689 - val_accuracy: 0.7782\n",
      "Epoch 171/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3900 - accuracy: 0.8193 - val_loss: 0.4717 - val_accuracy: 0.7626\n",
      "Epoch 172/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3998 - accuracy: 0.8105 - val_loss: 0.4720 - val_accuracy: 0.7704\n",
      "Epoch 173/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3861 - accuracy: 0.8203 - val_loss: 0.4720 - val_accuracy: 0.7704\n",
      "Epoch 174/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3971 - accuracy: 0.8184 - val_loss: 0.4594 - val_accuracy: 0.7626\n",
      "Epoch 175/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3943 - accuracy: 0.8086 - val_loss: 0.4645 - val_accuracy: 0.7704\n",
      "Epoch 176/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3824 - accuracy: 0.8193 - val_loss: 0.4521 - val_accuracy: 0.7704\n",
      "Epoch 177/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.4124 - accuracy: 0.8037 - val_loss: 0.4833 - val_accuracy: 0.7743\n",
      "Epoch 178/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3874 - accuracy: 0.8154 - val_loss: 0.4617 - val_accuracy: 0.7704\n",
      "Epoch 179/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3938 - accuracy: 0.8096 - val_loss: 0.4538 - val_accuracy: 0.7626\n",
      "Epoch 180/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3761 - accuracy: 0.8262 - val_loss: 0.4489 - val_accuracy: 0.7743\n",
      "Epoch 181/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3907 - accuracy: 0.8164 - val_loss: 0.4520 - val_accuracy: 0.7743\n",
      "Epoch 182/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3814 - accuracy: 0.8223 - val_loss: 0.4585 - val_accuracy: 0.7782\n",
      "Epoch 183/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3858 - accuracy: 0.8135 - val_loss: 0.4740 - val_accuracy: 0.7626\n",
      "Epoch 184/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3932 - accuracy: 0.8145 - val_loss: 0.4428 - val_accuracy: 0.7704\n",
      "Epoch 185/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.4038 - accuracy: 0.8057 - val_loss: 0.4601 - val_accuracy: 0.7588\n",
      "Epoch 186/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3824 - accuracy: 0.8232 - val_loss: 0.4643 - val_accuracy: 0.7510\n",
      "Epoch 187/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3670 - accuracy: 0.8379 - val_loss: 0.4431 - val_accuracy: 0.7977\n",
      "Epoch 188/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3738 - accuracy: 0.8213 - val_loss: 0.4511 - val_accuracy: 0.7626\n",
      "Epoch 189/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3656 - accuracy: 0.8232 - val_loss: 0.4569 - val_accuracy: 0.7588\n",
      "Epoch 190/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3676 - accuracy: 0.8359 - val_loss: 0.4398 - val_accuracy: 0.7899\n",
      "Epoch 191/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3837 - accuracy: 0.8291 - val_loss: 0.4569 - val_accuracy: 0.7704\n",
      "Epoch 192/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3588 - accuracy: 0.8330 - val_loss: 0.4483 - val_accuracy: 0.7704\n",
      "Epoch 193/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3623 - accuracy: 0.8320 - val_loss: 0.4533 - val_accuracy: 0.7665\n",
      "Epoch 194/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3745 - accuracy: 0.8125 - val_loss: 0.4320 - val_accuracy: 0.7821\n",
      "Epoch 195/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3748 - accuracy: 0.8252 - val_loss: 0.4468 - val_accuracy: 0.7743\n",
      "Epoch 196/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3807 - accuracy: 0.8213 - val_loss: 0.4475 - val_accuracy: 0.7743\n",
      "Epoch 197/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3548 - accuracy: 0.8350 - val_loss: 0.4606 - val_accuracy: 0.7393\n",
      "Epoch 198/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3847 - accuracy: 0.8184 - val_loss: 0.4460 - val_accuracy: 0.7549\n",
      "Epoch 199/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3826 - accuracy: 0.8223 - val_loss: 0.4453 - val_accuracy: 0.7626\n",
      "Epoch 200/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3502 - accuracy: 0.8350 - val_loss: 0.4301 - val_accuracy: 0.7743\n",
      "Epoch 201/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3548 - accuracy: 0.8398 - val_loss: 0.4403 - val_accuracy: 0.7938\n",
      "Epoch 202/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3830 - accuracy: 0.8174 - val_loss: 0.4284 - val_accuracy: 0.7743\n",
      "Epoch 203/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3499 - accuracy: 0.8359 - val_loss: 0.4314 - val_accuracy: 0.7860\n",
      "Epoch 204/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3500 - accuracy: 0.8389 - val_loss: 0.4318 - val_accuracy: 0.7860\n",
      "Epoch 205/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3516 - accuracy: 0.8340 - val_loss: 0.4278 - val_accuracy: 0.7899\n",
      "Epoch 206/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3560 - accuracy: 0.8291 - val_loss: 0.4296 - val_accuracy: 0.7860\n",
      "Epoch 207/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3480 - accuracy: 0.8418 - val_loss: 0.4300 - val_accuracy: 0.7860\n",
      "Epoch 208/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3535 - accuracy: 0.8438 - val_loss: 0.4432 - val_accuracy: 0.7704\n",
      "Epoch 209/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3874 - accuracy: 0.8203 - val_loss: 0.4577 - val_accuracy: 0.7510\n",
      "Epoch 210/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3631 - accuracy: 0.8271 - val_loss: 0.4408 - val_accuracy: 0.7665\n",
      "Epoch 211/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3511 - accuracy: 0.8486 - val_loss: 0.4314 - val_accuracy: 0.7977\n",
      "Epoch 212/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3593 - accuracy: 0.8340 - val_loss: 0.4376 - val_accuracy: 0.7704\n",
      "Epoch 213/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3526 - accuracy: 0.8330 - val_loss: 0.4311 - val_accuracy: 0.7743\n",
      "Epoch 214/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3337 - accuracy: 0.8486 - val_loss: 0.4291 - val_accuracy: 0.7782\n",
      "Epoch 215/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3478 - accuracy: 0.8398 - val_loss: 0.4475 - val_accuracy: 0.7510\n",
      "Epoch 216/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3569 - accuracy: 0.8369 - val_loss: 0.4236 - val_accuracy: 0.7782\n",
      "Epoch 217/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3665 - accuracy: 0.8350 - val_loss: 0.4352 - val_accuracy: 0.7899\n",
      "Epoch 218/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3553 - accuracy: 0.8340 - val_loss: 0.4416 - val_accuracy: 0.7743\n",
      "Epoch 219/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3536 - accuracy: 0.8350 - val_loss: 0.4327 - val_accuracy: 0.7704\n",
      "Epoch 220/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3289 - accuracy: 0.8457 - val_loss: 0.4211 - val_accuracy: 0.7977\n",
      "Epoch 221/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3426 - accuracy: 0.8477 - val_loss: 0.4255 - val_accuracy: 0.7626\n",
      "Epoch 222/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3593 - accuracy: 0.8389 - val_loss: 0.4326 - val_accuracy: 0.7626\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 223/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3387 - accuracy: 0.8438 - val_loss: 0.4194 - val_accuracy: 0.7938\n",
      "Epoch 224/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3275 - accuracy: 0.8633 - val_loss: 0.4209 - val_accuracy: 0.7821\n",
      "Epoch 225/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3132 - accuracy: 0.8623 - val_loss: 0.4273 - val_accuracy: 0.7782\n",
      "Epoch 226/300\n",
      "1024/1024 [==============================] - 2s 2ms/step - loss: 0.3339 - accuracy: 0.8525 - val_loss: 0.4149 - val_accuracy: 0.7938\n",
      "Epoch 227/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3336 - accuracy: 0.8340 - val_loss: 0.4147 - val_accuracy: 0.7860\n",
      "Epoch 228/300\n",
      "1024/1024 [==============================] - 2s 1ms/step - loss: 0.3373 - accuracy: 0.8516 - val_loss: 0.4117 - val_accuracy: 0.7860\n",
      "Epoch 229/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3158 - accuracy: 0.8555 - val_loss: 0.4150 - val_accuracy: 0.7860\n",
      "Epoch 230/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3425 - accuracy: 0.8398 - val_loss: 0.4309 - val_accuracy: 0.7704\n",
      "Epoch 231/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3368 - accuracy: 0.8408 - val_loss: 0.4088 - val_accuracy: 0.7860\n",
      "Epoch 232/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3415 - accuracy: 0.8350 - val_loss: 0.4028 - val_accuracy: 0.7821\n",
      "Epoch 233/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3290 - accuracy: 0.8418 - val_loss: 0.4075 - val_accuracy: 0.7860\n",
      "Epoch 234/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3248 - accuracy: 0.8545 - val_loss: 0.4059 - val_accuracy: 0.8093\n",
      "Epoch 235/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3315 - accuracy: 0.8516 - val_loss: 0.4142 - val_accuracy: 0.7821\n",
      "Epoch 236/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3057 - accuracy: 0.8604 - val_loss: 0.4072 - val_accuracy: 0.8016\n",
      "Epoch 237/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3183 - accuracy: 0.8535 - val_loss: 0.4155 - val_accuracy: 0.7977\n",
      "Epoch 238/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3156 - accuracy: 0.8574 - val_loss: 0.4015 - val_accuracy: 0.8054\n",
      "Epoch 239/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3244 - accuracy: 0.8438 - val_loss: 0.4096 - val_accuracy: 0.7704\n",
      "Epoch 240/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3274 - accuracy: 0.8516 - val_loss: 0.4277 - val_accuracy: 0.7743\n",
      "Epoch 241/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3420 - accuracy: 0.8418 - val_loss: 0.4135 - val_accuracy: 0.7743\n",
      "Epoch 242/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3259 - accuracy: 0.8535 - val_loss: 0.4018 - val_accuracy: 0.8016\n",
      "Epoch 243/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3086 - accuracy: 0.8584 - val_loss: 0.4054 - val_accuracy: 0.7860\n",
      "Epoch 244/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3156 - accuracy: 0.8496 - val_loss: 0.4022 - val_accuracy: 0.7821\n",
      "Epoch 245/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2938 - accuracy: 0.8604 - val_loss: 0.3991 - val_accuracy: 0.7899\n",
      "Epoch 246/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3143 - accuracy: 0.8604 - val_loss: 0.3924 - val_accuracy: 0.7938\n",
      "Epoch 247/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3026 - accuracy: 0.8662 - val_loss: 0.3910 - val_accuracy: 0.7821\n",
      "Epoch 248/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3158 - accuracy: 0.8633 - val_loss: 0.4009 - val_accuracy: 0.7938\n",
      "Epoch 249/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3072 - accuracy: 0.8574 - val_loss: 0.3928 - val_accuracy: 0.7860\n",
      "Epoch 250/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3100 - accuracy: 0.8672 - val_loss: 0.3938 - val_accuracy: 0.7938\n",
      "Epoch 251/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2973 - accuracy: 0.8691 - val_loss: 0.4048 - val_accuracy: 0.7899\n",
      "Epoch 252/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3016 - accuracy: 0.8574 - val_loss: 0.3885 - val_accuracy: 0.7821\n",
      "Epoch 253/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2965 - accuracy: 0.8672 - val_loss: 0.3894 - val_accuracy: 0.8093\n",
      "Epoch 254/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2904 - accuracy: 0.8662 - val_loss: 0.3863 - val_accuracy: 0.7821\n",
      "Epoch 255/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3030 - accuracy: 0.8613 - val_loss: 0.3859 - val_accuracy: 0.7977\n",
      "Epoch 256/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3135 - accuracy: 0.8545 - val_loss: 0.3724 - val_accuracy: 0.8171\n",
      "Epoch 257/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3126 - accuracy: 0.8477 - val_loss: 0.4043 - val_accuracy: 0.8210\n",
      "Epoch 258/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2905 - accuracy: 0.8701 - val_loss: 0.3924 - val_accuracy: 0.7977\n",
      "Epoch 259/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2783 - accuracy: 0.8789 - val_loss: 0.4149 - val_accuracy: 0.7860\n",
      "Epoch 260/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2929 - accuracy: 0.8721 - val_loss: 0.4053 - val_accuracy: 0.7938\n",
      "Epoch 261/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2781 - accuracy: 0.8838 - val_loss: 0.4122 - val_accuracy: 0.7977\n",
      "Epoch 262/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2984 - accuracy: 0.8574 - val_loss: 0.3845 - val_accuracy: 0.8093\n",
      "Epoch 263/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3138 - accuracy: 0.8584 - val_loss: 0.3882 - val_accuracy: 0.8054\n",
      "Epoch 264/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3022 - accuracy: 0.8613 - val_loss: 0.3863 - val_accuracy: 0.8210\n",
      "Epoch 265/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2844 - accuracy: 0.8711 - val_loss: 0.3857 - val_accuracy: 0.8093\n",
      "Epoch 266/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2820 - accuracy: 0.8857 - val_loss: 0.3827 - val_accuracy: 0.8054\n",
      "Epoch 267/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2736 - accuracy: 0.8789 - val_loss: 0.3874 - val_accuracy: 0.7977\n",
      "Epoch 268/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2781 - accuracy: 0.8682 - val_loss: 0.3826 - val_accuracy: 0.8054\n",
      "Epoch 269/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2940 - accuracy: 0.8594 - val_loss: 0.3909 - val_accuracy: 0.7977\n",
      "Epoch 270/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2720 - accuracy: 0.8809 - val_loss: 0.3844 - val_accuracy: 0.8132\n",
      "Epoch 271/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2955 - accuracy: 0.8682 - val_loss: 0.4023 - val_accuracy: 0.7821\n",
      "Epoch 272/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2827 - accuracy: 0.8701 - val_loss: 0.3799 - val_accuracy: 0.8132\n",
      "Epoch 273/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2892 - accuracy: 0.8672 - val_loss: 0.3717 - val_accuracy: 0.8249\n",
      "Epoch 274/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.3122 - accuracy: 0.8584 - val_loss: 0.3732 - val_accuracy: 0.8171\n",
      "Epoch 275/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2851 - accuracy: 0.8809 - val_loss: 0.3731 - val_accuracy: 0.8171\n",
      "Epoch 276/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2790 - accuracy: 0.8789 - val_loss: 0.3773 - val_accuracy: 0.8132\n",
      "Epoch 277/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2600 - accuracy: 0.8945 - val_loss: 0.3761 - val_accuracy: 0.8210\n",
      "Epoch 278/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2643 - accuracy: 0.8789 - val_loss: 0.3805 - val_accuracy: 0.8171\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 279/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2624 - accuracy: 0.8857 - val_loss: 0.3703 - val_accuracy: 0.8249\n",
      "Epoch 280/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2574 - accuracy: 0.8857 - val_loss: 0.3716 - val_accuracy: 0.8288\n",
      "Epoch 281/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2693 - accuracy: 0.8799 - val_loss: 0.3761 - val_accuracy: 0.8210\n",
      "Epoch 282/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2559 - accuracy: 0.8809 - val_loss: 0.3663 - val_accuracy: 0.8171\n",
      "Epoch 283/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2824 - accuracy: 0.8721 - val_loss: 0.3952 - val_accuracy: 0.7860\n",
      "Epoch 284/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2694 - accuracy: 0.8779 - val_loss: 0.3582 - val_accuracy: 0.8327\n",
      "Epoch 285/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2768 - accuracy: 0.8789 - val_loss: 0.3878 - val_accuracy: 0.8210\n",
      "Epoch 286/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2845 - accuracy: 0.8740 - val_loss: 0.3824 - val_accuracy: 0.8210\n",
      "Epoch 287/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2681 - accuracy: 0.8887 - val_loss: 0.3948 - val_accuracy: 0.8016\n",
      "Epoch 288/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2620 - accuracy: 0.8838 - val_loss: 0.3834 - val_accuracy: 0.8132\n",
      "Epoch 289/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2540 - accuracy: 0.8887 - val_loss: 0.3735 - val_accuracy: 0.8093\n",
      "Epoch 290/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2430 - accuracy: 0.8965 - val_loss: 0.3822 - val_accuracy: 0.8054\n",
      "Epoch 291/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2443 - accuracy: 0.8906 - val_loss: 0.3749 - val_accuracy: 0.8171\n",
      "Epoch 292/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2681 - accuracy: 0.8906 - val_loss: 0.3638 - val_accuracy: 0.8171\n",
      "Epoch 293/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2865 - accuracy: 0.8818 - val_loss: 0.3729 - val_accuracy: 0.8054\n",
      "Epoch 294/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2610 - accuracy: 0.8857 - val_loss: 0.3717 - val_accuracy: 0.8171\n",
      "Epoch 295/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2418 - accuracy: 0.8936 - val_loss: 0.3536 - val_accuracy: 0.8171\n",
      "Epoch 296/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2700 - accuracy: 0.8887 - val_loss: 0.3700 - val_accuracy: 0.8016\n",
      "Epoch 297/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2589 - accuracy: 0.8848 - val_loss: 0.3447 - val_accuracy: 0.8210\n",
      "Epoch 298/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2729 - accuracy: 0.8711 - val_loss: 0.3624 - val_accuracy: 0.8210\n",
      "Epoch 299/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2577 - accuracy: 0.8867 - val_loss: 0.3768 - val_accuracy: 0.8288\n",
      "Epoch 300/300\n",
      "1024/1024 [==============================] - 1s 1ms/step - loss: 0.2436 - accuracy: 0.8916 - val_loss: 0.3547 - val_accuracy: 0.8405\n",
      "CPU times: user 31min 12s, sys: 19min 30s, total: 50min 43s\n",
      "Wall time: 7min 11s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "history = model.fit(x_train, y_train,\n",
    "              batch_size=128,\n",
    "              epochs=300,\n",
    "              validation_data=(x_test, y_test),\n",
    "              class_weight=class_weights,\n",
    "              shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluating model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Accuracy Curves')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzsAAAE7CAYAAAD6sJ8GAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeXyU5bn/8c81+2QhCSQEZEdEVES0saKItqK21upBa3tsxQ160P5UtK1ttdpjF7qcLrS1apXW4nIUj1atrfaoKPSoVVRQiwsqKPsOAUKSyaz3748JYQaCTEKYmeD3/XrlNTPP88zzXDOI8OW+7+sx5xwiIiIiIiIHGk+hCxAREREREdkfFHZEREREROSApLAjIiIiIiIHJIUdERERERE5ICnsiIiIiIjIAUlhR0REREREDkgKOyIiH0NmdomZOTMbVuhaPoqZDTSzW8xssZm1mFmjmb1qZjeYWUWh6xMRkeLmK3QBIiIi7TGzk4C/AhuAm4G3AD8wBrgCqAa+XrACRUSk6CnsiIhI0TGzKuDPwCLgVOdcU8bup83sV8AJXXAdA/zOudi+nktERIqPprGJiMgemdlEM/tX6xSyTWZ2r5n13eWYr5jZ661TzLaZ2ZtmdlnG/mPNbLaZbTazZjP70Mxu28ul/wOoAa7aJegA4Jxrcs7Nbj3/p1qn5H1ql7p2TNUbnLFtmZn9t5lNMrN3gRhwjpnVtwaoXT//v7eeY3TGtpPN7Fkz225mTWb2lJmN3OV9nzGzF1u/j0Yze8/M/nMvn1lERLqYwo6IiLTLzKYA95IeXTkXuA74DPB/ZlbWesyJwH8D/wdMAL4I/AGobN1fBjwFJIFLgM8BP2TvMwtOBdY55+Z36YdK+zTwDeAHwGeB+cCDwFfMzLvLsROBt5xzbwCY2ZnAs0Bj676vAOXA82Y2oPWYoaSn3y0F/h04G5gOlO6HzyIiIh9B09hERGQ3rX/p/xHwD+fc+Rnb3wWeByaRXkczBtjqnLsm4+1PZzwfAVQB33bOLczYftdeShgALOts/XtRBXzCObduxwYzuxe4jHTIeqp1Ww3pMHRDxnt/C/yfc+7fMt47F/gQ+CZwDXAMEAC+5pxraD1szn76LCIi8hE0siMiIu05FOgN3Je50Tn3ArAcOLl106tAVevUsM+bWeUu51kMbAXuaJ0SN2A/152LeZlBB8A590/gA+DCjM3nk/5z8j4AMzsEOBi4z8x8O36AZuAl4KTW970BxIEHzOw8M+u9Xz+NiIjskcKOiIi0p2fr49p29q3bsd8593+kp64NAB4FNprZM2Y2qnX/NtLTxtYAtwErzOwtM/vCXq6/Ehi8rx9iD9r7TJCejnfOjil6pIPPHOfc6tbXO0LLnaTDTObP54FeAM65JaSn+3lITwNcZ2Yvm9mOgCgiInmisCMiIu2pb33s086+PsDmHS+cc392zp1MenrYOUBf4Ekz87Tuf8M59wXSAel40iMoD+66qH8XzwB9zOwTOdTa0voY2GV7rz0c7/aw/V6ghHTgGQ4c27pthx2f+frWfbv+nNV2AefmOuc+S3rt0qmkA9ETZladw+cREZEuorAjIiLteQ9YT3oqVxszOwEYRLohQRbnXKNz7nHgDtKBp9cu+xPOuXnA90j/+XPYR1z/j8Am4BYz221hv5mVmNmprS+Xtz7uGp4+9xHn341z7gPS09EubP1pAh7JOOQ90uuIjnDOzW/nZ2E754w65+YAPyfdoGBIR2oSEZF9owYFIiIfb581s3W7bNvmnJvd2ir5DjP7b9JTvPoBPya9DmcmgJn9EKgF5pKeqtYfmAq84ZzbaGafB6YAfyHdnay0df920sGiXc65+tapbn8FXjOz37HzpqKfBC4nfR+eZ5xza83s/4DrzWwT6ZuQTiS9vqaj7gFuBY4EHnXONWbU5MzsCuAxMwuQ7uC2qfXznwCscM5NN7PLSa/f+Tvp6XjVpEeD1rR+BhERyRNzbk+j+SIicqAys0toDSzteNs5N7L1uInAt0h3VWsk/Rf4bzvn1rbuP5N0eBlFepraBtLd2L7nnFtjZoeSbjX9SdKjPdtJNzX4gXPu5RzqHNR6/TNIB6k46VbYDwO37eh2Zmb9gd8D40hPa/sT6Q5pfwCGOOeWtR63DHjBOTdxD9erIr0mKQB8xjn3dDvHHE+6Q9tYINx6/Dzgt865l1r3X0e6K1tv0lMCXwBudM69t7fPLCIiXUdhR0REREREDkhasyMiIiIiIgckhR0RERERETkgKeyIiIiIiMgBSWFHREREREQOSAo7IiIiIiJyQCr6++xUV1e7wYMHF7oMEREREREpUgsWLNjknKvZdXvRh53Bgwczf/78QpchIiIiIiJFysyWt7dd09hEREREROSApLAjIiIiIiIHJIUdERERERE5ICnsiIiIiIjIAUlhR0REREREDkgKOyIiIiIickBS2BERERERkQNS0d9npxg0z5/P9meexSUSlNTV0eOznyl0SSIiInIAaWhoYMOGDcTj8UKXIlIUfD4foVCImpoaQqFQ58/ThTUdsFoWvUv9XXelX6RSCjsiIiLSZRoaGli/fj39+vUjHA5jZoUuSaSgnHMkEgkaGxtZsWIFtbW1VFRUdOpcCjs5MP/Or8klkwWsRERERA40GzZsoF+/fpSUlBS6FJGiYGb4/X6qqqoIBoOsW7eu02En72t2zOzrZva2mb1lZrPMrPPjUvni9bY9dQkNL4uIiEjXicfjhMPhQpchUpTC4TDRaLTT789r2DGzfsBUoM45NxLwAufns4bOMJ9/54tEonCFiIiIyAFJU9dE2revvzcK0Y3NB4TNzAeUAGsKUEOHmC9zZEfT2EREREREuoO8hh3n3Grgl8AKYC2wzTn3dD5r6AzzZazZ0ciOiIiIiEi3kO9pbFXAvwFDgIOAUjOb2M5xU8xsvpnN37hxYz5LbF/mmp2kwo6IiIhId3HddddhZqxbt65T729pacHMuPzyy7u4so65/fbbMTPmzZtX0Dq6m3xPYzsVWOqc2+iciwOPACfsepBzboZzrs45V1dTU5PnEneXtWYnrrAjIiIi0hFmlvPPsmXLCl2uHEDy3Xp6BTDGzEqACDAemJ/nGjpMradFREREOu/ee+/Nev38888zY8YMpkyZwrhx47L2dfU/dE+bNo3vf//7nb4xZSgUIhKJ4PPpji3dUV5/1ZxzL5vZn4HXgATwOjAjnzV0hmW1ntbIjoiIiEhHTJyYvWohkUgwY8YMjj/++N327YlzjubmZkpLSzt0bZ/Pt89BpbNBSQov793YnHM3OedGOOdGOucudM51vnF2vmQ1KNB9dkRERET2pyeffBIzY9asWfz2t79lxIgRBINBfve73wHw4osvctFFF3HIIYdQUlJCjx49OOmkk3j88cd3O1d7a3Z2bFu6dCnf+ta36NevH6FQiGOOOYbZs2dnvb+9NTuZ25577jlOPPFESkpKqKmp4fLLL6e5uXm3Op555hmOO+44QqEQffv25dprr+X111/HzPjZz37W6e9q/fr1XH755fTv359AIMCgQYO4+uqr2bJlS9ZxTU1N3HjjjQwfPpxwOExVVRWjRo3ihhtuyDruscce48QTT6RXr16Ew2EGDRrEeeedx4cfftjpGgtJ43E5yOzGhlpPi4iIiOTFf/3Xf7Ft2zYmTZpE7969GTp0KAAPPfQQH3zwAeeffz4DBw5k48aN3HXXXZx11lk8/PDDnHvuuTmd/8tf/jLhcJhvf/vbRCIRfv3rX3P22WezZMkS+vXrt9f3v/LKKzz00EN89atfZeLEiTz77LPccccdBAIBbr755rbjnn32Wc444wx69+7Nd7/7XcrLy3nggQf4xz/+0anvZYf6+nqOP/54li9fzn/8x39w1FFH8corr/C73/2OuXPnMm/ePEpKSgCYMmUKs2bN4pJLLmHMmDHEYjEWL17MnDlz2s739NNPc84553D00Udzww03UFFRwapVq5g9ezbLli1r+/67E4WdHKj1tIiIiEj+rVmzhnfffZeePXtmbZ82bdpu09mmTp3KqFGjmDZtWs5hp1+/fvz5z39uu3Hl2LFjOemkk/jjH//ITTfdtNf3L1y4kFdffZWjjz4agMsvv5zx48czY8YMfvGLXxAMBgH4xje+QSAQYN68eQwYMACAK664ghNO2K1PV4f8+Mc/ZunSpdx5551MmjQJgK997WuMHDmSa6+9ll//+tfccMMNOOd47LHHmDBhAn/605/2eL6//OUvAMyZM4eKioq27bl8F8VKYScHWWFHDQpEREQkDxaNOKzQJezRYe8uyst1Jk2atFvQAbKCTnNzM5FIBOccJ598MnfffTfRaLQtaHyUa665pi3oAJx44okEAgEWL16cU30nn3xyW9DZ4ZRTTmHOnDmsXLmSYcOGsXz5chYuXMhFF13UFnQAAoEAU6dO5eKLL87pWu159NFH6devH5dccknW9iuvvJIf/ehHPProo9xwww2YGeXl5SxcuJBFixZx2GHt/7dVUVGBc46HH36Yiy++GG/GuvXuKu9rdrolb+Y0Nq3ZEREREcmH4cOHt7t97dq1TJo0iZqaGkpLS6murqampoa77roL5xzbtm3L6fy7TssyM6qqqti8eXOn3g/Qq1cvgLZzLF26FIBDDz10t2Pb25Yr5xzLly/nsMMOw+PJ/it9MBhk2LBhWetsbr75ZtatW8fhhx/OIYccwpQpU3j88cdxzrUdc80113DkkUcyefJkevXqxVlnncWtt96a8/dRjBR2cpDVelr32RERERHJix3rTTIlk0nGjx/PrFmzmDx5Mg8++CBPPfUUs2fP5rzzzgMglUrldP49jVxkBoDOvD/zHLmeq6M6et4vfvGLLFu2jLvvvpuTTjqJp556irPOOovTTjuNROsyjdraWl577TWeeeYZvva1r1FfX8/UqVMZPnw4CxYs2B8fY7/TNLYcZLWe1jQ2ERERyYN8TRXrbubPn8+iRYv4yU9+wvXXX5+175ZbbilQVXs2ZMgQAN57773d9rW3LVcej4fBgwfz7rvvkkqlskZ3YrEYS5YsYdiwYVnvqa6u5qKLLuKiiy7COcfXv/51fvvb3/K///u/nHXWWUC6Vff48eMZP348AAsWLODYY4/lJz/5CQ8//HCn6y0UjezkQA0KRERERIrDjtGUXUc2XnvtNZ544olClPSRBg8ezMiRI/nzn//MypUr27bHYrGsjm2dMWHCBFatWsU999yTtf3WW29l27ZtnHPOOQDE43EaGhqyjjEzRo8eDaS7ugFs2rRpt2scfvjhBIPBtmO6G43s5MLn3/lcYUdERESkYEaNGsXw4cOZNm0aW7du5ZBDDmHRokX84Q9/YNSoUbz22muFLnE306dP54wzzmDMmDFcfvnllJeXM2vWrLbmCJlNEjrihhtu4JFHHuGrX/0qL7/8MqNGjWL+/PnMnDmTkSNH8vWvfx1Irx8aOnQoEyZM4KijjqKmpoYPPviA3//+91RXV/O5z30OgAsvvJAtW7Zw6qmnMmjQIJqamrj//vtpaWnhoosu6povI88UdnJgvoxpbAo7IiIiIgUTCAT4+9//zre+9S3+9Kc/EYlEOPLII5k1axYvvPBCUYad0047jSeeeIIbb7yRH//4x1RVVfGVr3yFCRMmcNJJJxEOhzt13p49e/LSSy9x00038dhjj3HnnXdSW1vLlVdeyQ9+8IO2NU8VFRVcddVVPPvsszz55JM0NzfTt29fvvCFL3D99ddTU1MDwKWXXso999zDzJkz2bRpExUVFYwcOZLHHnuMs88+u8u+j3yy/bVoqqvU1dW5+fPnF7SG5LZtvH/cGAA8PXpw6CsvF7QeEREROXB8VCtgObDdd999TJw4kUcffZQJEyYUupyilcvvETNb4Jyr23W71uzkwqs1OyIiIiLSOalUilgslrUtGo3ym9/8hmAwyLhx4wpU2YFP09hykNl6mrjusyMiIiIiuWtoaOCwww7jggsuYPjw4WzcuJFZs2bx9ttvc9NNN7Xdm0e6nsJODtR6WkREREQ6KxwOc/rpp/PII4+wbt06AEaMGMEdd9zBlClTClzdgU1hJxeZN4xKpXCpFObRDEARERER2btgMMjdd99d6DI+lvQ39hyYGfjVflpEREREpDtR2MlR1lQ2hR0RERERkaKnsJMj86kjm4iIiIhId6KwkyM1KRARERER6V4UdnKVsWbHqf20iIiIiEjRU9jJUeY0NjSyIyIiIiJS9PIadszsUDN7I+OnwcyuyWcNnaUGBSIiIiIi3Ute77PjnHsPGA1gZl5gNfBoPmvorKwGBXGFHRERERGRYlfIaWzjgQ+cc8sLWEPusqaxKeyIiIiIFJMTTzyRYcOGZW2bOHEiPl9u/7a/ZMkSzIxp06Z1eW2JRAIz46tf/WqXn1s+WiHDzvnArPZ2mNkUM5tvZvM3btyY57Lap9bTIiIiIp3zxS9+ETPjjTfe2OMxzjmGDBlCZWUlkUgkj9V1jfr6er7//e/z3HPPFbqUPTrxxBOprKwsdBl5VZCwY2YB4Gzgofb2O+dmOOfqnHN1NTU1+S1uD7LDjhoUiIiIiORq8uTJAMycOXOPx8ydO5dly5Zx/vnnEw6Hu+S6M2fOpKmpqUvOtTf19fX84Ac/aDfs+Hw+IpEIt99+e15qkZ0KNbJzBvCac259ga7fcb7MBgVqPS0iIiKSq9NPP50BAwZw3333EYvF2j1mRxDaEYy6gt/vJxgMdtn59kUoFMp5Sp10nUKFnS+zhylsxcp8O++zo9bTIiIiIrnzeDxccsklbN68mb/+9a+77W9oaOCRRx5h5MiRHHvssW3b77//fs466ywGDhxIMBikpqaGc889l7feeiun6+5pzc5zzz3HCSecQDgcpk+fPkydOrXdEaBEIsG0adMYN24ctbW1BAIBBg0axBVXXEF9fX3bcc888wyHHHIIAN/73vcwM8ysbQ3RR63ZueOOOzj66KMJh8NUVlbymc98hhdffHG3Ona8/4UXXmDcuHGUlJRQXV3NlClT9svo1cMPP8zxxx9PaWkpZWVljBs3jscff3y341544QU++9nPUltbSzAYpF+/fpx55pm88sorbcds3ryZq6++mqFDhxIKhejVqxd1dXVMnz69y+veVd7jpZmVAKcBl+X72vtCradFREREOu/SSy9l2rRpzJw5k/POOy9r3wMPPEBzc/Nuozq33HILtbW1XHbZZdTW1rJkyRJmzJjBCSecwOuvv87BBx/c4TpefPFFTjvtNCorK7nuuuvo0aMHs2bN4oUXXtjt2JaWFn71q1/xhS98gQkTJlBaWsorr7zCjBkz+Oc//8mrr76K3+9n5MiR/PKXv+Taa6/lvPPO49/+7d8AKC8v/8havvnNbzJ9+nTGjBnDT3/6U7Zt28Ydd9zBpz71KR5//HFOP/30rOMXLFjAo48+yuTJk5k4cSJz5szhD3/4Az6fj9tuu63D38We/O53v2Pq1Kkcdthh3HTTTaRSKWbOnMlZZ53FnXfeyaRJkwBYtGgRp59+OgcddBDXXHMNtbW1rFu3jueff54333yTT37ykwCce+65vPTSS1x++eWMGjWKpqYmFi1axD/+8Q++8Y1vdFnd7XLOFfXPJz7xCVcMll86yb1z6Aj3zqEj3Pbnni90OSIiInKAeOeddwpdQt6ccsopzuv1utWrV2dtHzNmjAsEAm7jxo1Z2xsbG3c7x5tvvun8fr+76qqrsraPHTvWHXzwwVnbLrjgAuf1erO2HXvssS4QCLjFixe3bWtpaXHHHHOMA9yPfvSjtu3JZNI1NzfvVsPtt9/uAPfwww+3bVu8ePFu798hHo87wE2ePLlt29tvv+0Ad9JJJ7lYLNa2feXKla68vNwNHTrUJZPJrPd7PB736quvZp379NNPd4FAoN06dzV27FhXUVHxkcds2rTJhcNhN3z4cNfQ0NC2fevWrW7QoEGuR48ebtu2bc455371q185wC1YsGCP59u8ebMDdvv16ohcfo8A8107WUITB3Plz2hQoNbTIiIisp8Nvu6JQpewR8t+dman3jd58mTmzJnDvffey3e+8x0A3n33XebNm8d5551HdXV11vGlpaVA+h/nt2/fTiwWo0+fPgwbNoyXX365w9dfs2YNr776Kueff35Wm+pgMMg111zDRRddlHW8x+Npa5aQTCbZvn07iUSCU045BYCXX36Zc889t8N1APzlL38B4Dvf+Q5+/87lEv379+eiiy7i1ltvZeHChYwePbpt34knnkhdXV3WeU455RSefvppli9fzogRIzpVS6annnqKSCTC1VdfnTUyVVFRwZVXXsm3vvUt5syZw4QJE6ioqGj7LIcffjihUGi385WUlOD3+5k3bx7Lly9n0KBB+1xjRxSy9XS3Yt6MXKhpbCIiIiIddu6551JZWZnVle1Pf/oTQNvUqEwLFizgc5/7HOXl5VRUVFBTU0NNTQ2LFi1iy5YtHb7+hx9+CNBuKDj88MPbfc8DDzzAscceSzgcpqqqipqaGoYPHw7QqRp2WLp0KQBHHHHEbvtGjhyZVe8OQ4cO3e3YXr16Ael1MV2hI3VdcMEFfPrTn+ZHP/oRPXv2ZPz48fz85z9n5cqVbe8JhUJMnz6dN954g8GDBzNy5EimTp3K3Llzu6TevVHYyZFaT4uIiIjsm1AoxFe+8hXee+89XnzxRZLJJPfeey/9+/ffbX3KsmXLOOmkk3jzzTf5z//8Tx599FGefvppZs+ezYgRI0ilUh2+fnq2E5jZHvdlevDBB/nyl7+Mz+fj5ptv5m9/+xuzZ8/miSfSo26dqeGjrrc33ow15F1xvn09TygUYs6cOcybN4/rrrsOM+PGG2/k0EMPzWpEceWVV7J06VLuuOMORo8ezYMPPsgpp5zCxIkTu6Tmj6JpbDkynxoUiIiISP50dqpYsZs8eTK33XYbM2fOpL6+nnXr1nHDDTfs9hf5hx9+mObmZp588knGjRvXtt05x6ZNm9qmUHXEjoYGixYt2m1fe9vuvfdeSkpKmDt3btYUrfa6wbUXoHKp5e23395tatc777wDtD+Ss79l1nXyySfnVNdxxx3HcccdB8Dy5csZPXo03/ve9zj77LPbjunXrx9TpkxhypQpJBIJLrjgAu677z6++c1vcvTRR++3z6ORnVxljezoPjsiIiIinXHMMccwevRo/ud//odbbrkFM+PSSy/d7bgd4WfXkYbbb7+dTZs2deraBx10EHV1dTzyyCMsWbKkbXs0GuU3v/lNuzV4PJ6sERznHNOmTdvt2LKyMoCsltQfZUfHtl/84hckMv4hffXq1dx9990MHTqUUaNG5fbButDpp59OOBzm5ptvzmpp3dDQwC233EKPHj0YP348QLu/DgMHDqS6urrte2hubiYSiWQd4/P5OPLII4Hcv6/O0shOjrLW7Og+OyIiIiKdNnnyZK666iqeeuopPvWpT7XbQvrMM8/ku9/9LhdccAFXXHEFFRUV/POf/+TJJ59kyJAhnb729OnTGT9+PGPHjuX//b//R0VFBffff3+707fOO+88HnvsMU455RQuvPBCotEojz76KC0tLbsdW1tby+DBg7nvvvsYPHgwvXv3pry8nDPPbH+E7vDDD+cb3/gG06dP5+STT+ZLX/oSDQ0N3H777UQiEW677TY8nq4fl4hGo+2GNUh/3hEjRvCzn/2Mq6++muOOO46LL76YVCrFXXfdxdKlS7nzzjvbGhd8//vfZ+7cuXz+859nyJAhpFIpHnvsMZYsWcJ3v/tdID0adOqpp3LOOedwxBFHUFVVxTvvvMPvf/97Dj74YMaOHdvlnzFLey3aiumnWFpPr7nxe22tp+tnPVDockREROQA8XFqPb1DfX29C4VCDnD33HPPHo+bO3euO+GEE1xZWZmrrKx0Z555pnv77bfbbTOda+vpHecdM2aMCwaDrnfv3u7KK690b7zxRruto3//+9+7ESNGuGAw6Pr27esuu+wyt2HDht1aSTvn3IsvvuiOP/54V1JS4oC2etprPb3D7bff7o466igXDAZdeXm5O+2009wLL7yQdcxHvf8Pf/iDA9zzz+/91ihjx451wB5/HnroobZjH3roITdmzBgXDoddSUmJGzt2rPvrX/+adb5nnnnGffGLX3QDBw50oVDIVVVVueOOO87deeedLpVKOeec27Bhg5s6daobNWqUq6iocKFQyA0bNsxdc801bu3atXut2bl9az1trosWM+0vdXV1bv78+YUug3U//CFb7p8FQO33bqTnBRcUuCIRERE5ECxatIjDDjus0GWIFK1cfo+Y2QLnXN2u27VmJ1dqPS0iIiIi0q0o7ORIradFRERERLoXhZ0cqfW0iIiIiEj3orCTK7WeFhERERHpVhR2cpQ5jU2tp0VEREREip/CTo4y77Pj4prGJiIiIiJS7BR2cmT+jLCjkR0RERERkaKnsJMrb2aDAq3ZERERka5T7Pc9FCmUff29obCTI/P5d75QNzYRERHpIn6/n0gkUugyRIpSJBIhGAx2+v0KOznSfXZERERkf+jduzerV6+mublZIzwipEdz4vE49fX1rFq1il69enX6XL69HyKg++yIiIjI/tGjRw8A1qxZQzyuqfIiAD6fj1AoxMCBAwmFQp0/TxfWlBMzqwT+CIwEHDDJOfdSvuvosKzW0wo7IiIi0nV69OjRFnpEpOsUYmTnt8CTzrnzzCwAlBSghg5T62kRERERke4lr2HHzHoAJwGXADjnYkAsnzV0VlbraU1jExEREREpevluUDAU2AjMNLPXzeyPZlaa5xo6J7P1tO6zIyIiIiJS9PIddnzAMcDvnXNHA03AdbseZGZTzGy+mc3fuHFjnktsX3Y3Ni0eFBEREREpdvkOO6uAVc65l1tf/5l0+MninJvhnKtzztXV1NTktcA9yb7PjkZ2RERERESKXV7DjnNuHbDSzA5t3TQeeCefNXSWWk+LiIiIiHQvhejGdhVwX2sntg+BSwtQQ4dlTWNT62kRERERkaKX97DjnHsDqMv3dfdZRutp1HpaRERERKTo5XvNTrel1imI9OYAACAASURBVNMiIiIiIt2Lwk6OsqexqUGBiIiIiEixU9jJVeZ9dtR6WkRERESk6Cns5Eitp0VEREREuheFnRyp9bSIiIiISPeisJOjrDU7CjsiIiIiIkVPYSdXus+OiIiIiEi3orCTo8yRHd1nR0RERESk+Cns5Eitp0VEREREuheFnRyZVw0KRERERES6E4WdXPkzW08r7IiIiIiIFDuFnRxpZEdEREREpHtR2MmReb1gln7hnNbtiIiIiIgUOYWdDlCTAhERERGR7kNhpyOy2k/HC1eHiIiIiIjslcJOB2hkR0RERESk+1DY6QA1KRARERER6T4UdjrCnzGyE1fYEREREREpZgo7HWC+jHvtJBV2RERERESKmW/vh3QtM1sGbAeSQMI5V5fvGjpL09hERERERLqPvIedVp92zm0q0LU7LatBQUINCkREREREipmmsXWEL3NkR62nRURERESKWSHCjgOeNrMFZjalANfvtKw1O5rGJiIiIiJS1AoxjW2sc26NmfUGZpvZu8655zIPaA1BUwAGDhxYgBLbp/vsiIiIiIh0H3kf2XHOrWl93AA8CnyynWNmOOfqnHN1NTU1+S5xj7IaFKj1tIiIiIhIUctr2DGzUjMr3/EcOB14K5817JOM++yo9bSIiIiISHHL9zS2WuBRM9tx7fudc0/muYZOM29mNzaFHRERERGRYpbXsOOc+xA4Kp/X7EpqPS0iIiIi0n2o9XRHqPW0iIiIiEi3obDTAZmtpzWNTURERESkuCnsdEDmNDbUelpEREREpKjtc9gxs8PN7AtmdlBXFFTM1HpaRERERKT76FDYMbNbzOz2jNfnAv8CHgLeMbNju7i+omL+zJuKKuyIiIiIiBSzjo7snAG8mPH6B8DjpDusvQLc1EV1FaeM1tNozY6IiIiISFHraNjpAywDMLP+wBHAT51zbwI3Awf2yI5P99kREREREekuOhp2IkBZ6/OTgQZgfuvrRqC8i+oqSrrPjoiIiIhI99HRm4q+BlxhZiuAK4DZzrlU674hwNquLK7o6D47IiIiIiLdRkfDzg3Ak6SbEmwFLs/YN4H0up0DVuZ9dtR6WkRERESkuHUo7DjnXjWzgcAIYLFzriFj9wxgcVcWV2zUelpEREREpPvo6MgOzrkmYEHmNjPr5Zx7osuqKlJqPS0iIiIi0n109D47/2Fm38p4faSZrQI2mNl8M+vT5RUWE3VjExERERHpNjraje0q0h3ZdphOeu3ONUAF8MMuqqsome6zIyIiIiLSbXR0GttA4F0AM6sg3X56gnPu72a2GfhpF9dXVNR6WkRERESk++joyI4X2NFq+kTAAf9ofb0S6N01ZRUny2o9rZEdEREREZFi1tGwsxg4s/X5+cCLzrnm1tcHAfVdVVhRyhjZQQ0KRERERESKWkensf0SuNfMLgaqgC9m7Ps0sLCrCitGmffZUetpEREREZHi1tH77NxvZiuA44BXnXPPZexeD/y1K4srNprGJiIiIiLSfXTmPjsvAC+0s/2mLqmoiGU1KNA0NhERERGRotbhsGNmJcAk0p3YegKbSTcpuCtj/c7ezuEF5gOrnXOf72gNBaPW0yIiIiIi3UZHbyraB3gNuBmoA0qAY4FbgAVmVpvjqa4GFnXk2sVAradFRERERLqPjnZj+znpxgTjnHNDnHPHO+eGkG5DXQn8195OYGb9SXd0+2NHiy00rdkREREREek+Ohp2zgCud879M3Ojc+5F4EZ2tqX+KL8Bvs3O+/XsxsymmNl8M5u/cePGDpa4H2WN7MQLWIiIiIiIiOxNR8NOGbBmD/tWte7fIzP7PLDBObfgo45zzs1wztU55+pqamo6WOL+k9l6Gk1jExEREREpah0NO+8BF+5h30Tg3b28fyxwtpktAx4ATjGz/+5gDQWjaWwiIiIiIt1HZ24qek9rI4L7gbVAH+B84FT2HIQAcM5dD1wPYGafAq51zk3sYA0Fk916WiM7IiIiIiLFrKM3Ff3v1tbTPyS7wcB64DLn3P1dWVzR8WaO7GjNjoiIiIhIMevoNDacczOAg4AjgHGtj/2AZWa2sAPn+Ue3uscOWrMjIiIiItKddPimogDOuRS73CfHzCpIB58Dlvkzu7FpzY6IiIiISDHr8MjOx5l51aBARERERKS7UNjpiIwGBSjsiIiIiIgUNYWdDsjqxqawIyIiIiJS1Pa6ZsfMhuZ4rj77WEvRU+tpEREREZHuI5cGBUsAl8NxluNx3ZZGdkREREREuo9cws6l+72K7iJzzU5c99kRERERESlmew07zrm781FId6BpbCIiIiIi3YcaFHSABQJtz1MtLTh3QM/aExERERHp1hR2OsATDOIpL0+/SCRIbt1a2IJERERERGSPFHY6yFfbu+15Yv36AlYiIiIiIiIfRWGng/y9M8LOhg0FrERERERERD6Kwk4H+WoUdkREREREugOFnQ7y1da2PY9rGpuIiIiISNFS2OkgX9Y0to0FrERERERERD6Kwk4HqUGBiIiIiEj3oLDTQWpQICIiIiLSPSjsdFDmNLb4RoUdEREREZFipbDTQb7qajADILlpMy4eL3BFIiIiIiLSnryGHTMLmdkrZvYvM3vbzH6Qz+t3BfP78fbqlX7hHInNmwtbkIiIiIiItCvfIztR4BTn3FHAaOCzZjYmzzXss6x1O2pSICIiIiJSlPIadlxaY+tLf+uPy2cNXSFr3Y6aFIiIiIiIFKW8r9kxM6+ZvQFsAGY7517Odw37yqeObCIiIiIiRS/vYcc5l3TOjQb6A580s5G7HmNmU8xsvpnN37ix+G7cmX2vHYUdEREREZFiVLBubM65rcA/gM+2s2+Gc67OOVdXU1OT99r2RiM7IiIiIiLFL9/d2GrMrLL1eRg4FXg3nzV0Bd1YVERERESk+PnyfL2+wN1m5iUdtB50zj2e5xr2ma+2tu15fIO6sYmIiIiIFKO8hh3n3ELg6Hxec3/InsZWfGuKRERERESkgGt2ujNvVRX4/QCkGhpIRSIFrkhERERERHalsNMJZoY/o3GC1u2IiIiIiBQfhZ1OUkc2EREREZHiprDTSVlNCtapSYGIiIiISLFR2Okkf5/MsLO2gJWIiIiIiEh7FHY6ydenb9vzhEZ2RERERESKjsJOJ/n79ml7Hl+3roCViIiIiIhIexR2OsnfZ2fYSazVNDYRERERkWKjsNNJvr47p7FpZEdEREREpPgo7HSSr7oafD4AkvX1pKLRAlckIiIiIiKZFHY6ybxefL0zbiyq0R0RERERkaKisLMP/Bkd2eJrFXZERERERIqJws4+0L12RERERESKl8LOPsi+145GdkREREREionCzj7IbD+taWwiIiIiIsVFYWcf+LJuLKppbCIiIiIixURhZx9kNihIaGRHRERERKSoKOzsA3/WyI7CjoiIiIhIMVHY2Qfenj0xvx+AVEMDqaamAlckIiIiIiI75DXsmNkAM5trZovM7G0zuzqf1+9q5vHgy2xSsH59AasREREREZFM+R7ZSQDfdM4dBowBrjCzw/NcQ5fK7simJgUiIiIiIsUir2HHObfWOfda6/PtwCKgXz5r6GqZHdl0rx0RERERkeJRsDU7ZjYYOBp4uVA1dAV/re61IyIiIiJSjAoSdsysDHgYuMY519DO/ilmNt/M5m/cuDH/BXaA7rUjIiIiIlKc8h52zMxPOujc55x7pL1jnHMznHN1zrm6mpqa/BbYQf6+O++1E1+1uoCViIiIiIhIpnx3YzPgTmCRc256Pq+9vwSHDm17Hv1gSQErERERERGRTPke2RkLXAicYmZvtP58Ls81dCl///5YMAhAcuMmklu3FrgiEREREREB8OXzYs65FwDL5zX3N/N6CRw8lOg7iwCILllCSV1dgasSEREREZGCdWM7kASHDWt7Hl2iqWwiIiIiIsVAYacLBIcd0vY8ulhhR0RERESkGCjsdIHgsIPbnmtkR0RERESkOCjsdAFNYxMRERERKT4KO13A378/FgoBkNy8mcSWLQWuSEREREREFHa6gHk82ffbWby4gNWIiIiIiAgo7HSZ4CGayiYiIiIiUkwUdrpIIGPdTkxhR0RERESk4BR2ukh2k4IPCliJiIiIiIiAwk6XCR6Sca8djeyIiIiIiBScwk4X8R90EFZSAkCyvp7Epk0FrkhERERE5ONNYaeLmMdDaPjwttctb79dwGpERERERERhpwuFRo5sex5R2BERERERKSiFnS4UOuKItuctbynsiIiIiIgUksJOFwodcXjbc01jExEREREpLIWdLhQcOhQLhwFIrF9PYuPGAlckIiIiIvLxpbDThcznIzRiRNtrrdsRERERESkchZ0ulrVuR2FHRERERKRgFHY6KRJL8sArK/jnkuz76YRGZoadd/JdloiIiIiItPIVuoDuqL4pxqV3vcq/Vm7FY/DXK09kZL8KAMJZHdneKlSJIiIiIiIfe3kd2TGzP5nZBjPrtilg7bYIX7rjJf61cisAKQfPLFrftj+Q2aRgwwbiGzYUpE4RERERkY+7fE9juwv4bJ6v2WWcc0y5ZwFLNjRmbZ+/bEvbc/N6CR12WNvrhr/9Dedc3moUEREREZG0vIYd59xzQH0+r9mVPtzUxJurtwHg9Vjb9tdXbCGRTLW9Dh85su35hl/8khWTJhFfty5/hYqIiIiISHE2KDCzKWY238zmbyyie9W8sWJr2/OTh9fQtyIEQFMsybvrtrftq/ryl/H16dP2uvmleay87HJcPJ6/YkVEREREPuaKMuw452Y45+qcc3U1NTWFLqfN6yt3TlcbPaCSusE9216/umzngFVg8GCGPv44PS++GDzprzj63ntsnnlX3moVEREREfm4K8qwU6zeWLlzZOfogZUcO7iq7fX85VuyjvWWlVJ7/XX0vvbatm2bbr2V2PLl+79QERERERFR2MlVJJZk0dr0VDUzOGpAJXWDdo7szF9W324jgp4XXUjw8HTDAheNsvbG75Fqbs5P0SIiIiIiH2P5bj09C3gJONTMVpnZ5Hxef1+8tWYbyVQ6zBxcU0aPkJ9D+5RTHkzfqmh9Q5RVWyK7vc98Pvr+8Edt09maX32VpeecS2ThwvwVLyIiIiLyMZTvbmxfds71dc75nXP9nXN35vP6++L1FTunqR09oBJId2Q7etDOqWyZ63YyhUceQfXll7W9ji1fzrIv/TuLP30Kq6Zezda//IVUU9N+qlxERERE5ONJ09hylLleZ/TAyrbnx2aEnbnv7blzXPVVV9H3Jz/BU1LSti2xdi3bn36atdddz/snjmPV17/O1ocfIb5eNyIVEREREdlXCjs5ej2j7fTRA3YGnBOGVbc9/9u/1vDMO+vbfb+ZUXnuOQx57C+UffrTWCiUtd9FImz/3ydZe8MNLDn5ZFZMmkTD3/9O82uv0fj8C8RXr+7iTyQiIiIicmDzFbqA7mDdthbWbmsBoCTgZXhtWdu+YwZW8tkj+vDk2+mbhn774YU8OWAcvctDbNjewux31rOivpnD+/bgk0N64qmoYfN3f8LWxhaaV62h8e13iL4yD1u1En8qiT+ZwJdK4H9zKYF//YRQIkYoGSOUiFF98UXUfOPreIJBAJxzJNavx7W04D/oICwQaLd+l0jQsnoNgepeeEtL9/O3JSIiIiJSHBR2cvBGxv11juxXgc+7c0DMzPjpuUfyxsqtrGtoob4pxkk/n0vPkgBrG1pop0HbLqpg4BkwcO91hDZHCX33r5jXC87hkglobZqAgXk8eH0+fAEfXjM8iTjxeIIG56XZlx5J8rokfsDvNfw+L4Ggn4Dfl37t9RDyeykP+SgP+egR8rc+Tz+G/V7iKUcimSKRdMRTKZJJt3NbypFIOkJ+D2U73hf0URb0URZKP2aez+/VwKKIiIiI7D8KOznImsI2sGq3/VWlAaZ/6SguuPNlnIOWeIo1rSNBXanFFyTrrN49HBjf8SQI3mDWrqR5SQItKSDmIBYDYl1caW6qwj5qeoQI+Dx4zYgnHZF4klgiRdDvIez3pn8CXkKtz4M+Dw5IOUfQ56Ek4CPg8+AcOBxBr4egP318yO8h5Nv53OMxPGZ4DIz0IwYe2xH0PPg8HlriSaKJFB4Dn9eDz2N4PYbPY9mvvYbP42nb50iPtqUc4NIN+MJ+b1Y4FhEREZH8UdjJwelH1OL3enh95RaOG9Kz3WNOGFbN9WeMYPrs92mJpwDwGHxySE+OGlDJwpXb+NeqrYT8Xvr0CNGzNEDQl/6LciLliCVS6Z/kzsdoIkkklqIpmiAST+7TZzDncGb7dI6utiWSYEuksdBl7HcBn4fSgJeSgA+PBxLJ9K93PJkikUzhaw1ovtb/FpKtI2U7Wp17zAj6PVSVBOhVFqBXaZBeZQF8Hg/NsQSxRIrykI+KsD9rtGzHL7ft8uvu9xqVJQEqwn7iyRTN0STNsQSNTS2s27ydNzfHWLyhkd7lQcYOq2b0gErKgr7W0LgjfKbDaDyZYvnmZtZsjRD0eekR9gOOhpZ0XT1LA/QqDZBy0BxL4DGjV1mAynCARCpFSzzFuoYIyzY1s3ZbhM1NMRoiCfpUBDmkdzm1PUIEfR4CPg9+745HI5jxOplyfLixiQ83NuL1eujbI0TPskA6nHo8DOgZpjzkxznHivpmPtzUBK0DojXlQYbWlFIS8JFIpmiKJmloidMYTeAcBP0eSgJeKsMBYokUj7y+iicWrgXgjCP7MmH0QfQqy/4HhY5Iplw6fO/D781ILEnI7+n0ObZF4kTjSWrKg/tUh4iISDGy9m6EWUzq6urc/PnzC11GzlIpx/Zogq3NMSrDASpK/F1y3mTKsX3tera89i/w+fCEw/gPOghfnz6YGaloCy0fLiPy1ls0vfceSY8PT58+BPv1o+/RI6kaPACiLTQueo+md99j+7vv07h4Cc1LlxOLJ0l4vCQ9Xlq8AZr8ofSPL0SzP0yjP/0Y9frwpVJ4XRJfKol3x3OXwptKb/PgiHl8NPtDNPtCNPuCNPnDNPuCO7f509tSphEPyY8BPcNEYkk2NbY/ilkS8NIc6/g/KHg9xuBeJQzrXUbQ520bFWyJJ2mJJ9kaibO1OY4ZlAZ8hANeSgNeAj4P6xpaWLu1hYDPw1H9Kzm0TznbInE2NUapKglwSO8yAj4P769vZNnmJrY0x2iIxKkqCXDUgEpKA16eX7KJDzc2UV0WZOywXpQFfby9poGN26P0qwwzpLqUoTWlDKkupTToY2V9M+sboqRcOlS/sqye+cvqSbn0jZIv+ORAFm/Yzux31hNLpBg9sJLD+vSgMZqgvilGz7IAh9aWE/J7WbKhkbXbIvTpEWZoTSlBn4ftLQmSzlHbI0Tv8iDRRIqtzTG2tX4PLfEk/arCDOxZQiSWZF1DCx4zPjGoir4VId5f38iLH2xiWySOc1Ae8vGZI/owoGe6i2UimcJBW6h3Lh2saQ2mQV869DnnWLa5mTdWbqFPjzBjhvZsC3LxZCo9zdeze7BzzinwiYh0U2a2wDlXt9t2hZ2PN5dIEFu5EvP78VZWYoEALholtX07sVWriK9cRaolkl4b5FK4VGrn82QK83own4/k9kYiby4k+s4iLBTCX1uLr08f/H1q8fSoILFuLbEVK8FjeCsraV7yARs+WMnWYBkJj5eUefClkoSSMXypJDGPj6jXT9QXSD96/US9AWJeP+YchiPu8RHxBYl7vHha/zuOt74v1vqemNdP1JN+TJmRsvTctcxHhxH3eIl6/STNSzAVJ5CM47B0CDQPKfOQ9HhImid7m3lJeNKPhmurzYAURosvgFOoE9mr8qCP7dHEbtvN4PihvYjEk7y9poF4MkV1WZAeIR/rtrXQlBFSvR6jqsSPmbFxe7Rt++gBlZx11EHMfmcdLy+tx2tGz9IAZSEfAa+HlHNsboxR3xyjIuxnaHUpfSvC+Lw7p7B6PUZDS4KNDVGa4wl6lgapKQtSXR6gpixIOOAlnkgRTzpiyfTI7Y4/Xv1eozzkpyTgxTmIp9LrHhMpR6p1BNcM+lakA2p5yEdDS5x40tG/KkxVSYAXP9jE3/61hq3NcY4ZVMWYob3oXxWmMuxvd6qsc47mWJJlm5t4Z00D2yJxjj+4F0ccVAFASzzJ+oYWmqJJ4skUw3qXURrcfbJHewGwM6EwlXJsakoH+Xyt10ylHI70fxcicuBT2JGi4pyj5a23aZr3Et7KSkKHHoq3qopUcwQXaSYViZDcvp3okiVEFy3CpRwldXWERx9FcssWYstXkIo0QzKF+f34+/fH36eW+Nq1RJd8gGuJYOEwnnAJnnAYC4dw8TguEgHAU16O+fzEViwntnQZ5vXg7VWNr1cvfNW98JSW0vL++7S8+RYuGsXXtw+eYJDo+4uJLlmSDoc9e+IJh3GJBC6ZwFNaire0lFSkhcTmzcRXrsS59EhXxBekxRckheF3CbwGAZ8PrwcSzRFiHh/J1sDncSm8LoXHpTDAYUS9frYFy9gaKGVbsIxtwTKS5iGciOJ1qdbRuDDJ1mC183d1+g95l/FnfczjZ3ughEZ/GH8qQTgRI5SMEk5EKU3FGVq/kqHb1rCyvDdv1BzCupKexFqDY0tr6NwRJMHRt6me3s1bSHo8NPrDGFASb8GfSrCttV6PSxFOxEiah63BMhoDJemug6kkldFGDmrcSG3zFiqj2ylJRFlfUsWK8loaAyXEPT7iHi9xj4+Ex0ui9fmObQ6jT3M9A7an70+1KVxBQ6AUZ+nvbU1pNUlPeoFbaTzCsK2r8SfjpMzD+pIq1pb2IuXxYi6V/g7iLZQkophLEff6iXgDNARKSXh9DN26ms8tm4fHpZg98FgW9Rq8v3+r7HfmHF4cCYXyohXweoglU3vc3yPko6o0QNjvpTGaYHtLgsZoom0qbKYRfcoxM95fvz1rv89jjOpfQc/SIPVNUeqbYmxuirG9JUG/yjCj+qdD0sJV21izLULPkgA15UFSzrG9JUHI76VuUBVHDahk7bYI769vpLElQco5tkXiLN3URDSRwu81hlaX0bcylJ6i6/NwcE0Zh/YpJ5ZIsXprhEQyxcBepfSvCuMxI5FMtTXCAQi3TgtOr5F0JFOtsx9a4izZ2MiS9Y0s3tDIkg2NJFIpRvTpwch+FQzsWcJBlelmPVuaYkQTKapKAlSW+An5vfi8xqr6CK+v3MKGhihH9q/g5OE1+Dweltc3sW5bC1uaYzRF09M++1WG6VcVpl9lOpBub4mzNRJvG8kM+T0MrSmjb48QHo+lZ360JKhvjmFAv6rwXoNfcyzB++sb8XmMw/r26PLglkylp/eWBryayirdnsKOSJ7FN2xg++zZRBYswFNWjr9/fwIDBxIcPpzAoIHprnqkR9eSW7aQ3N6IS8QhmcRTVoa3vJxkYxOJdWtJbK7HRVtItbSAGebx4pIJXEsLqZYoriVCqiWKJxzGW1kJHiNZv4Xk1q24WBQXi+GtriZ85JF4wmGa5r1My1tv4enRA/9BBxE8+GDKThqHp0cPtv3lMRqeeILE+vUkt24l1dzc9pksHCZ81FFYwE981WpSjY34+vbB17MXsZUriC1dBsmM6WA+H4HBg3DRGPF16yAeZ08sEEi3T/d4cNEoLhrd47EdEfN4WV3WG28qSf/GjXjI/n9e3LzEvD7Cidhu+3ZwQNzjJZDKnuoW8QZYWd6b1WU1OIxA66hgMJl+LI81UxaPYDhafEEi3iAtraOVldFGapvr2R4oYVHVYNaXVFERa6Iy2sjmUA9WlNcS9/gYuH09A7avp2d0O6XxCOtKevJez0E0+4IcvnkZR9QvY1VZDQurDyZpHoY2rKG2eQvrS6pYVVbD6rIa/n979x4c130ddvx77t2778XiQRAAAT5AkRRF60E5tmJHjmPZcSw/ZtS0cWs3TuOMW7utPZPMJOm47kzTZJw0TadN2xnHHWfs+pk6SmonHtup5Wfs2okoWZRFURQlii+ABPFavPa9997TP+4FCYKASIogQKzOZwaze+9e7P6whz9wD36/3/mN5ntpOh59tRl6q7O4GlxMEl89fgyArw3/DE/37GZLbZafG32S7sY8x7p2MpbrodCq0tGsMpUpcqbQT8tNsH1hgq3VGSYznZzLbyEUh2yrjqCU0kVKqQLpoEner1F0Id8o49TrjKeLjGe7SQVNemuzlL0Mx7p30UgkybVq3DvxHDsWov3Kjnft4Imt+y4bHRU0HjuNpAlJiNJUh+aS85lWnX1zozzTPUzLWa2aizE3XyJOUPxlyafrCIOdGZKJlROeph8yMlO9OELYkU5w33A3/cU0nZkkrTCkHK+PXBzdmyo3mFxo4DpCTy5JynOZXKgzU2mxtSNaC5lLuUwuNDg9XeHI6NzF0dF8KsFtvTl29+bZ0Z1FNSoa5IjECaZLJpkg6y3ej5LObNJFBMZm61yYr5P2osSpI50gVKXRChmdrTE6UyOfcnnFtiK39eYRiYoNuY7gOQ6JuCqs5zoXk7r5eotvPH2B8fk6Q11ZilmPH5+e4dFT0wShMtCZoSPtMVuNEvP9/QXedvcAB4c6qfsBjVa44vMuUtW4IJFcjEO54VMqN0kno7WyU+UGPxmZ5cJcnTsHi9yzvfOmj06GoXJyqkw+5dFfvLQvY70VRO+X69AKQkZKVabKTfo6UmzrvHry3O4s2THGvCRhs0kwM4vWqniDg4i3+jq0pddqqCSHBi/u/6RBgD81Tev8OfzxCcJKmbBSxe3pJnPXXXjbt1/2V0UNAsJajbBaJSyXCUol/Klp/NI0wdQ02mriZLM42SySzeKkM4TlBVpjFwhmZ0GjvwJ7g4Ok9u7FyRcI5ucIpkv4k5P4ExP4ExO0JiYQxyF94ACpffsgDAgWyoQL81EC2mjg5PM4uRxOPhq9U1XC+QX8yQkaJ0/RPHWKsFZDDuq2bQAAFh1JREFUWy3wL03FknSaRHc3YaNBMD19Te+3eB4aBBCu/pf8dtMSl+lMkS21WRJ6+c99IdvNU1tuo7OxwL6ZEfKtGqV0B2UvQ099jo5m9WKK0xKX+VSOmpukv1oioSETmU7+fN+bGCls5acvPMMbRg9TaFaZS+WouSkCx0WBYrNCR6PCbCrPaL6XuVT+sumrgbhk/Drd9QXSQZO5ZI6ZdIGZVIGZdIGWk8CLRysToY8bhheT56aToJpIU08kcVQvrXtcMoLbclzGs92cy/fSdBLkWjUclLFsDzUvTUejzAOjT7Bzfpynt+zmuc7tzKVylL3MqlNlU0GT7maZ4ZlzJMKAR/sP0Ehc2o+ttzpDrlUjdBOczW+9CZG9XCEBC1fOVDTmMo7Azp4c2zrT/PjMzMWiT2v13AnXIek6OALVZnAxCU0lHByRqxaEyqcS7O3Ls7WQopD2Lo48tvxoC45WsFiAaPF+dOuHStpzGezM0NeRwg+Uuh/EazxDEo4w0Jkm4Th885lxzs1GM1EeuL2X+/ds4ZGj4xw6XQIgl3Sp++Flo7OOQH9HmqGuLJmky8RCg9lqk65skoFimv5imoFimlwqwfh8g4mFelw5NiocdM9QJ7t6ctHo6ESZvo4Ub9y/lf6ONGdLVY6en+fo+TmOXyhTSCd47e4efmZPD0Nd2TWLz42yZMcYY9aBhiHq+xAESDp9MYELm80owbpwAX9qGrezE2+gn7Bep/H8CYJSifSBO0jffTcCNM+do37kCNXHHqPxwkm8/n6St+3GSaUI5uYASL/iFaTvvBOCAH9igtpTT1H50d/RGh3F27WT9P47ojV18/MQ+Dj5AuJ5tM6do3n2LG6xSPa++0hs6aHy6KPUnnwSrdZQ38fJ50nu3Ely1y6Su3aR6O2l8v9+wPzXvk4wN4dbLOL29BAszBOUZqIRPcfByeXI3HUXqX37qB5+gvpPnrrs/XG3bCE5NIS2WviTk+C6ZO68k9Qd+wkrFfyJSZx0Cm9wiLBSZuGRb9I8c+a6YpAcHiasVvHHx9ckphtJgbKXIduq464w8hggVLwM86kcddcj22qQ82tkWw08vfxDWyWR4siW28j4DfbMniPnX9rMYC6Z5Vj3LnzHpdioUGyUKTbLZPwGI4U+TnQOgSp75s4xtDBBOZmllCrgakjOrzOd7uCpLbdxptBPb22WXfNjdNUXcFGSQYvB8iSFVo1KIsXZQj8LySyhCBUvzenCAGc7+kj7DbZWZ3E1ZCzXw2SmE0HjxDAgEf8BoJ5IUnejpM3VAEcVR0OSGjK4MM6O+XF2xiOiroacKA5yujjAZKaTqUwnokqhWcULfRaSWRaSWVqpDGGhSIcbcvvUKYoXzvJkcSdHe4ZJBi0GKtNsrc1QbFTI+A1K6Q7Gs11M5HuYKPRSdpIUtEW+WSFfmSffrFJNpBjp6Gc+eenDYC4RrRdrBiHj5dVHuhc5KDvCCmV1mHJvzofKHk+p+yEVtRFQc6Vc0r1sbeRyu3tzfOM3Xn9LjCpZsmOMMeaGaRiizSZOOn3ZOYg2Nl7OL5UI5uaiNW2FAk4mc32vp4o/NhYlitPT+FPTBKVpgvkFJJnESadQVbTVwi10UHjjAyR37QIgKJepHjpE+fvfp3V2BEmncdJpvMFBksPDuJ2daKNOMDtL49QpmqdPQ6hIOoU2mwRT01Fi191Nom8rbkcR8bx4yqWHOC5+aRp/7AKt8XH8sTHCahWnUMDt7opGQUPFyefJHLyH1G17aBx/lurhJ0EVb9s2JJmkefIkzZER3K5OkoNDOIUC2mgQzM/TOnuWYG6ORG8vHW9/O97gIOUffJ/6kaejkc1iB61z5wnjBBjA7e4mrFbR+qVkxsnncYtFnGIH3tY+vMFBwnKZyqFD+GNj1xUTc30aTgJHlYQGLJ1EVXeTTGaKq1YmFVW21mZIBy0UOFvo44XitnjNZRYv9Mn6DbzQj4rmIBSbFbrq84TiMJftIOwboGN8lMJ86eJayJaToLs+T099nj1zo2ypRf92ZlIFRvO9jBa2MpHpwgujtViaTFJr+NRdD79vgFbPViqVOpX5CnVxaaRzBAmPnuoM3bMTtBIes/keask0Eldq7UsEDOaTzLopjjeTjKsX7XvnSFT4B8EPiUZIuPz92L0wxsHKGKW+HcwUurljsJOfvXs7RXzOPj/CQrnGlm29pLf28p0nTvGt02XmQodk6JMMWgRxtdmWOJdNgV2UdB2CuEIlQNKB7oTSwGG2pWQ8l1cUhN6gypP1FOfr6/O5uZBKUG76l21Ovzj5YfHctmKa3kKK8fkG4wvXspH92ju4vZO/+uD96//CK7BkxxhjjLnJ1PeRxNpuYRdWq0gms+ricVWldeYM/sxMNArX1YWqRtM5wxC3o2PV6aeqin/hAvVnn6Xx3POgitvVhSST+FOTBNMlcBzE80h0d5G64w7cYieVH/6QyqN/Dy0fyWZwO4rRSOD2ISSVApFoDWGxiD81xcK3v0P96FGSu3bR8Y63k33VqyEMqD/7LHNf+jIL3/rWxRFFr6+P/JveSPbee1n45reY/9rXotFMEXDd6H1wXZxMBieXi0YJp6ejKaSOg1sokL7nbjrf+U7S+/dT+eGPqD7+OM0zZ2idPQuOQ6K3FyebJZibwx8fv2xt4iLJZvHiqqKJ/gHEdWiOjuKPXUCbTcJ6naBUuvJNFcHbvp2wWiWYmrqh2L8c1V2Pc/leLmS7GahMs3v++pJxJaqGutpI6GKhm1AcUo6S7e4Cx6FabdAqV8i2ahdTIs3mEA0hLm6kwESmi8lsJ6VUB7VEkkQ8LXVxaqoXT2P1kh654Z3k9u3FSyZwalUqocNYIsesmyaV9EgnXbyFeZypqNz/dE+UzO45+RQHfvhVxtNF/ubONzOV7+Gu2TP87MnHyE+NUUukSIQ+uf5e0rfvJ9G3lbCrm8l8LxdSHbTyHWzbOUDvUB8z9YCxmQpnjr7AyPNnWBifojh2hu5yKdoofc9exofv4KjTyVg9ZHtYZUdplBecAoedLnyEojbZU5/idq/J/q05St39PN7I8PiFKv/8dbv5rbfcfmNBXyOW7BhjjDFm01n8nPJilcI0DNFaVIVzpRHGF33+MKR58iS1n/wE9QNSe24juXt3tB3DVaqTNU6dovy9v8UfHyc5PEz69n3RGsF4bV/z1GnqR56icfo0zVOnaS6OIKqSuftuMj/1Uzj5HChI0sNJZwhmZ6geOkT9+HMkd+yg8OY3k75jP8H8AmGtGq0fzGTxJyZonjqFPzVFWKkQNuq4hQ7cjgK1I09TP3LkYjudXA4SCbReR5JJEn1bo4Qvk8VJp0gMDJDetw/xPOrPPEPt6FHqzxy7bMTQbEJxIaTLCgddh7qbpJpI0dVYWGFMDFqZHM5d93DnZz/50tu4hizZMcYYY4zZYC827XMttSYm8MfG8IaGcLu7r7ustKrinz+PhiGJ7m6CSoXqo4eoHzuG199P9r5X43Z30xoZwS+VSA4Nkdy1C/X9aLSsVotG+YDW6CjNU6fRIMDJZZGEFxWBmZsnmJ8nmJ9DPI/k9h0kd2zH27GD5Pbt0fYOYUjz1CkqP/o7aocP0zx9mua5czipFKm9e3E7O2m88AKtkRG8bdvIvvpVZO69l/T+/XhDQ7QujNM8fZr60aPUjjyFf36MsNVEa3XCcvnyH9p1Se0eJrlrF/Vjz9IaHQUguXMnmYP30JqYoHXmLHiJaN1l/wC5174mGkE89Bjl736X1thYVAl1yTTSNSUSTVVdWECbK2+UfS3cri6CmZkbbk76rrsY/ouHb/h51oIlO8YYY4wxZtPTMIy2YVhawTMMrzuBDGu1qFCKSLSuLZe7WEEUoBWvZ/MGBq67ja3xcWqHD1N/5hi4Dm4uB+IQ1mtorUZYqxPWayS6uvB27kQ8j+bJU/gTE6QPHKDw82/C7eqieeYMQamE29NDoqcnmmbqumirFVUDfeEEfmmGoDQdrZGcLtGaGKc1MnrZNMvEtgHy97+O3OteR/aV95Lo7aV59izl732P6mOPUz18mGBqiszBgxQefAtuR5HmyFnwfZLDu/G2baN1bpTGc89Rf+65qLDO1BTFf/QP2fb7v3/d78/NYMmOMcYYY4wxLxPabF5a63aVRFBVIQwv7gF4LfxSCW028fr7b7Spa2K1ZGdtV1EaY4wxxhhjNtzSUaqrXhsnRdcj0d19vU3aEOteFFtEHhSR4yJyQkQ+vN6vb4wxxhhjjHl5WNdkR0Rc4GPAW4EDwLtF5MB6tsEYY4wxxhjz8rDeIzv3ASdU9aSqNoEvAg+tcxuMMcYYY4wxLwPrnewMAiNLjkfjc5cRkfeLyOMi8vjk5OS6Nc4YY4wxxhjTPtY72VmpyPsV5eBU9ROq+ipVfVVvb+86NMsYY4wxxhjTbtY72RkFti85HgLOr3MbjDHGGGOMMS8D653sPAbsFZFhEUkC7wK+ss5tMMYYY4wxxrwMrOs+O6rqi8iHgG8ALvApVT26nm0wxhhjjDHGvDyI6hVLZm4pIjIJnNnodsS2AFMb3QhzU1hs25fFtj1ZXNuXxbZ9WWzb160Q252qesVi/1s+2bmViMjjqvqqjW6HWXsW2/ZlsW1PFtf2ZbFtXxbb9nUrx3a91+wYY4wxxhhjzLqwZMcYY4wxxhjTlizZuT6f2OgGmJvGYtu+LLbtyeLaviy27cti275u2djamh1jjDHGGGNMW7KRHWOMMcYYY0xbsmTnGonIgyJyXEROiMiHN7o95qUTkdMickREnhSRx+Nz3SLyTRF5Pr7t2uh2mqsTkU+JyISIPL3k3IqxlMj/iPvwUyLyyo1rubmaVWL7H0TkXNx3nxSRty157N/GsT0uIm/ZmFabqxGR7SLyXRE5JiJHReTX4/PWbze5F4mt9dtNTkTSInJIRH4Sx/Z34/PDIvJo3G//XESS8flUfHwifnzXRrbfkp1rICIu8DHgrcAB4N0icmBjW2Vu0AOqenBJmcQPA99W1b3At+Njc+v7NPDgsnOrxfKtwN746/3Ax9epjeal+TRXxhbgj+O+e1BVvw4Q/z5+F/CK+Hv+JP69bW49PvCbqnoH8Brgg3H8rN9ufqvFFqzfbnYN4I2qeg9wEHhQRF4D/Cei2O4FZoD3xde/D5hR1T3AH8fXbRhLdq7NfcAJVT2pqk3gi8BDG9wms7YeAj4T3/8M8A82sC3mGqnq94HSstOrxfIh4LMa+XugU0QG1qel5nqtEtvVPAR8UVUbqnoKOEH0e9vcYlR1TFWfiO8vAMeAQazfbnovEtvVWL/dJOL+V44PvfhLgTcCfxmfX95vF/vzXwJvEhFZp+ZewZKdazMIjCw5HuXFO7C5tSnwiIj8WETeH5/rU9UxiH5hA1s3rHXmRq0WS+vH7eFD8XSmTy2Zbmqx3YTiqS33Ao9i/batLIstWL/d9ETEFZEngQngm8ALwKyq+vElS+N3Mbbx43NAz/q2+BJLdq7NStmolbHbvO5X1VcSTY/4oIi8fqMbZNaF9ePN7+PAbUTTKMaA/xKft9huMiKSB/4P8BuqOv9il65wzmJ7C1shttZv24CqBqp6EBgiGoG7Y6XL4ttbKraW7FybUWD7kuMh4PwGtcXcIFU9H99OAF8m6rTji1Mj4tuJjWuhuUGrxdL68SanquPxf7gh8KdcmvJisd1ERMQj+jD8BVX9Unza+m0bWCm21m/bi6rOAt8jWpfVKSKJ+KGl8bsY2/jxItc+LXnNWbJzbR4D9sZVJ5JEC+q+ssFtMi+BiOREpLB4H/gF4GmieP5qfNmvAn+9MS00a2C1WH4F+GdxdafXAHOL02bM5rBsrcYvEvVdiGL7rrgC0DDRYvZD690+c3XxvP1PAsdU9b8uecj67Sa3Wmyt325+ItIrIp3x/Qzw80Rrsr4L/FJ82fJ+u9iffwn4jm7gxp6Jq19iVNUXkQ8B3wBc4FOqenSDm2Vemj7gy/E6uQTwZ6r6f0XkMeBhEXkfcBZ45wa20VwjEfnfwBuALSIyCvwO8IesHMuvA28jWgRbBX5t3RtsrtkqsX2DiBwkmg5xGvgAgKoeFZGHgWeIKkJ9UFWDjWi3uar7gV8BjsTz/wE+gvXbdrBabN9t/XbTGwA+E1fLc4CHVfWrIvIM8EUR+ShwmCjZJb79nIicIBrReddGNHqRbGCiZYwxxhhjjDE3jU1jM8YYY4wxxrQlS3aMMcYYY4wxbcmSHWOMMcYYY0xbsmTHGGOMMcYY05Ys2THGGGOMMca0JUt2jDHGvCQi8l4R0VW+Zje4bZ+OS1YbY4x5GbN9dowxxtyodxLtmL2UvxENMcYYY5ayZMcYY8yNelJVT2x0I4wxxpjlbBqbMcaYm2rJdLfXi8hfiUhZRKZF5GMikll27YCIfFZEpkSkISJPich7VnjOYRH5nIhciK87KSL/fYXr7hWRH4hIVUSeF5F/uezxfhH5jIicj59nTES+KiJb1/6dMMYYs95sZMcYY8yNckVk+f8noaqGy859HngY+BPgPuDfAzngvQAikgP+FugCPgKMAO8BPiciWVX9RHzdMHAIqAK/AzwPbAd+YdnrdQB/Bvw34PeAXwM+LiLHVfW78TWfA3YCvx2/Xh/wJiD7Ut4IY4wxtxZLdowxxtyoZ1c49zXgHcvOfV1Vfyu+/4iIKPB7IvIHqvocUTKyF3hAVb8XX/c3ItIHfFREPqmqAfC7QAa4R1XPL3n+zyx7vQLwrxcTGxH5PlFC9G5gMdl5LfARVf3Cku/7i2v6qY0xxtzyLNkxxhhzo36RKwsUrFSN7eFlx18EPko0yvMc8Hrg3JJEZ9Hngf8FHACOECUsX12W6KykumQEB1VtiMjzwI4l1zwG/LaICPAd4GlV1as8rzHGmE3Ckh1jjDE36ulrLFAwvsrxYHzbDYyt8H0XljwO0MOVydVKZlY41wDSS47/CdFUuH9DNN1tTET+J/DRFabhGWOM2WSsQIExxpj10rfK8bn4tgT0r/B9i+em49spLiVIN0RVJ1T1g6o6COwHPk00Te4Da/H8xhhjNpYlO8YYY9bLP152/C4gJCo2AFFxgiERuX/Zdf8UmACOxcePAO8QkYG1bJyqHlfVjxCNCN25ls9tjDFmY9g0NmOMMTfqoIhsWeH846q6dHPRt4nIfyZKVu4jmj722bg4AUSjKr8OfElE/h3RVLVfBt4MfCAuTkD8fW8HfiQifwCcIBrpeVBVryhTvRoRKQLfAr5AVGShBTxEVA3ukWt9HmOMMbcuS3aMMcbcqNWql/USTTlb9B7gN4F/BTSBPwUWq7OhqhUR+Tngj4A/JKqmdhz4FVX9/JLrTovITxMVN/iP8XXngL++znbXgSeAf0FUfjqMX++XVfV6n8sYY8wtSKzojDHGmJtJRN5LVE1t7zUWMjDGGGPWhK3ZMcYYY4wxxrQlS3aMMcYYY4wxbcmmsRljjDHGGGPako3sGGOMMcYYY9qSJTvGGGOMMcaYtmTJjjHGGGOMMaYtWbJjjDHGGGOMaUuW7BhjjDHGGGPakiU7xhhjjDHGmLb0/wGhMcByNC2CIQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1008x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0UAAAE7CAYAAAASMCotAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdd3iTVfvA8e9J994tUAplKSICylBAkOEAXhVFwQHKELegILh/+rpfleECBUVBZakoiDJEFGTJ3nuVLrr3SNIk5/fHkz5JmrQFLUM4n+viIs951nkClNy5z7mPkFKiKIqiKIqiKIpysTKc6w4oiqIoiqIoiqKcSyooUhRFURRFURTloqaCIkVRFEVRFEVRLmoqKFIURVEURVEU5aKmgiJFURRFURRFUS5qKihSFEVRFEVRFOWipoIiRVGUC5QQ4nMhhBRCTDrXffm3EJrBQoiVQohcIUSFECJVCDFPCNHzXPdPURRFOTOEWqdIURTlwiOECAAygFAgC4iXUlrOba/Ob0IIL2AecDswC1gM5AEJwEDgViBCSll4zjqpKIqinBHe57oDiqIoyhlxO1pAtAToB/QBfj6nPfJACOEnpTSd637YPQ/cCdwppVxQZd9sIcSNQMU/vcl59syKoigKavicoijKhWookA8MA8qB+z0dJIRoK4T40T5UrFwIcVAI8XyVY24XQqwTQpQIIYqEEJuEELfa9yXah+gNq3JOD3t7D6e2VUKItUKIW4QQ24UQJuAx+74nhBAbhBB5QogCIcRfQoj/eOhvkBDif0KIo0IIkxAiQwixQAgRJ4Rob79nfw/nzbQPg/Oq5n3wBZ4GfvEQEAEgpfxVSlnm9CyrPFwnSQgx02l7mL1P3YUQ3wkhCoCNQohnhBBmIUSUh2vsE0IsdNoOFEK8I4Q4bj/nuBDiRSGEwemYYCHER0KIZPv7kimE+E0I0dLTsyiKoiiuVKZIURTlAiOEaABcD0yXUmbbP2APEEJESCnznY7rBKwCjgBjgFSgBdDG6ZhRwIfAQrRAqwS4Ckj8m927xH6914FjaMPTsF/vcyAJ7f+mW4CfhRD9pJRL7X3xBVYA7YC3gb+AMOAmtGFtW4UQm4GHgUVOzxAODALelVJaq+lXByAc+OlvPldtZgNz0TJR3sAu+zPcBUx16mt74DLg/+zb3sByoBXae7YbuMa+PxItkAOYjDa87wXgMBAFdLU/k6IoilILFRQpiqJceO5DGwnwlX17FnAP2gfwT52OmwDkAtdUZkCA3yt3CiFCgbeAH6WUA5zOW/4P+hYN3Cil3OHcKKUc53RfA7ASLYB6BFhq3zUE6Az0l1I6By/fO72eCswQQjSWUp6wt90P+KIFXdVJsP9+ooZj/onvpZTPODcIIX5H+7Oa6tR8H1qGr3Ko4z3AtcB1Uso/7W0rhRAArwgh3pFSZqG9L7OllDOcrvVj3T+GoijKhUkNn1MURbnw3A8cllJusG//BqTjNIROCBGIlkmY7RQQVdUFCAam12HfkqoGRPb+tBdC/CyEyAQsaHN3bgAudTrsRiCjSkBU1TygAHjQqe1htGFxqf+493+fpwDla+AaIUQL0LNCdwPfOs056oMWqK0XQnhX/gJ+BXzQskYAm4FhQogXhBAdqhsmqCiKonimgiJFUZQLiBCiI9pQqx+EEOH2oWMhwA9AZyHEJfZDI9D+D6gpUKic71KXwcTJqg1CiAS0zFAkMAotGOsILAP8q/QnraaLSymNwJfAA/YAohva+/FpTecBKfbfG5/CM/wdbs8NLABK0TJgoAV9cWjBUqVYe58qqvzaZN9f+Wc0CpgGjEALkLKEEJPtwa+iKIpSCxUUKYqiXFiG2n9/Fm0YVuWvJ+ztldmifMAGxNdwrRz77zUdY7T/7lul3a2AgJ2ndSD6oM0NGiSl/FZK+ZeUcgtQ9QN9Ti19qfQJWnDRHy1LlETtQ/62oGWYbjmF64P23FWfGbTAzhO355ZSlqJlkAbbm4YAx6SU65wOywWOowWJnn4ttl+rREr5vJSyOdr8rLfQ/sxfOcXnURRFuaipoEhRFOUCYS9EcDewEejp4dcO4D4hhLAPmVsLDLGvaeTJerTCCg/VcNtMwAS0rtLuVjmuBpXBj17u2p7R6lrluF+BekKIGgMXKeVR+7Hj0QobfCaltNVyjhmYCNwshLjD0zFCiBucMi8ngEvs73nl/u5oWbnT8TXQTAhxE1oQ93WV/cvQ5juVSCm3ePiVU/WCUsoTUsqJaEUZqv65KIqiKB6oQguKoigXjpvRMjRPSylXVd0phJiGlkXpAfwBjANWAxuEEBPRhsk1BdpJKUdJKYvt5bk/EkIsQKugVoxW/c0opfxISimFEPPRhqsdAg6iBUQ9TqPfv6HNI/rK3o/6wKtAMq5f3n2DNldorhDibbTgLwSt+tz7UsoDTsdORatAVwF8cYr9eBtoC8y3l9WuXLy1IXAHMABt2CFoc5ceAr6wH9sEGAuc7sKulfO9ZqAFh99U2T8bGI5WXGEisBMtQ9UMrdrcbVLKMiHEBrTKebvRAtnr7M8y6zT7oyiKclFSmSJFUZQLx1C0oOW7avbPRVuzaCiAlHIzWjYmBfgIbaHX8TjNIZJSfgwMRAsMZqPNg7kTbUhXpSfR5iz9F5iPNg9o1Kl2Wkq5F20IWWO0D/bPAM8Bf1Y5rgJt3s0naAHJErTgJxpHae9KvwBlwCIpZcYp9sOKVrp7GFrQMROtGt97aMHVdVLKQvuxf6BVxrsaLXgajjb8reBUn9t+HRswB21Y4AYp5ZEq+yvQgr7PcDzzbLQ/w/WA2X7on/a+z7Y/+53AGCnlB6fTH0VRlIuVkNLT8G5FURRF+fcSQtyANoTueinlynPdH0VRFOX8poIiRVEU5YIhhGiGNgRwMmCSUrY/x11SFEVR/gXU8DlFURTlQvJ/aIu9mnBal0lRFEVRaqIyRYqiKIqiKIqiXNRUpkhRFEVRFEVRlIuaCooURVEURVEURbmoXRDrFEVHR8vExMRz3Q1FURRFURRFUc5TW7duzZFSxnjad0EERYmJiWzZsuVcd0NRFEVRFEVRlPOUEOJEdfvU8DlFURRFURRFUS5qKihSFEVRFEVRFOWipoIiRVEURVEURVEuaiooUhRFURRFURTlonbWgyIhRB8hxEEhxBEhxHMe9jcWQqwUQuwSQqwSQjQ8231UFEVRFEVRFOXicVaDIiGEFzAF6Au0Au4RQrSqctgE4CspZRvgNeDts9lHRVEURVEURVEuLmc7U9QJOCKlPCalNAPzgP5VjmkFrLS//sPDfkVRFEVRFEVRlDpztoOieCDFaTvV3uZsJ3CH/fXtQIgQIqrqhYQQDwkhtgghtmRnZ5+RziqKoiiKoijKhUjabBQsWED+/G+RFRXnujvn3NkOioSHNlllexxwnRBiO3AdkAZY3E6ScrqUsoOUskNMjMeFaRVFURRFURTlvCWlJHfGDDJefwNrQcFZvXfOx1M4+eJLZLzyCukvvIi02Wo9x5yUhOU0khFSVv2Yf/4620FRKpDgtN0QSHc+QEqZLqUcIKW8EnjR3lZ49rqoKIqiKIqiKGde8fJfyXpvAvmzZ5M1afJZu2/5rl3kTJumbxctXkzm//5XYxCT/+23HO3TlyO9r6do6dJa72EzmUh7agyFv/xSJ30+0852ULQZaCGEaCKE8AXuBn5yPkAIES2EqOzX88AXZ7mPiqIoiqIoinLGFf7k+BhctHQpNrO5zu8hKyqQVqu+bSsvJ/3Z58CpDSD/q6852LYdB9pdScrjT2AzmfR9pmPHyXzzLe16ZjNpY58mb/bsau9pLS4m5cGHKF6+nPTnnqdk3bo6fqq6d1aDIimlBXgCWA7sB76VUu4VQrwmhLjVflgP4KAQ4hAQB7x5NvuoKIqiKIqiKGeataiI0jVr9G1bcTGla9fW6T3Ktm7l0NXXcPTGmyjftQtptZLx31cxHz8OgCEwkKDu3fTjpdmMNBopWbmS3M8+19osFtKfew7pFCQhJZmvv0HGW2+5BXKW7GxO3D+Usk2btIaKCso2ba7T5zoTxL9prF91OnToILds2XKuu6EoiqIoiqJcAGzl5WS8+SZYbcQ+Mx7viIg6v0fBDz9y8oUXXNpC//MfYsaMIfONN6jIygTAKziE2HFPE9CmzWnfwzk4EYGBBLRpQ9lff+n767/xOqG33kr6009TvOI3l3OFry9Nf1pEwYIfyP3sM63Rxwe/5s0x7d+vH+fX6jLiJ07Er0kTpMXC8TvuxHTwoL4/dtzTRD7wAEJ4Ki1wdgkhtkopO3jcp4IiRVEURVEU5UJmKy0l7elxVGRk0ODdd/C/5JIaj8/64ANyP/kUgMBOnWj0xQyEt3ed9il55INumSEREIBvw3hMh4+4tBtCQmj8zdf4X3rpKV+/Ii2NI72vr3Z/2O23U/+tN/VgxWY2g8XCiaHDMO7erd/XVlysnxMzdiyRQwaTNv4ZSlau1Nu94+Jo+ssvlG/fTsqDD2qNXl7Uf/11wgfcfsp9PtNqCorO9pwiRVEURVEURTmrcmfOpGTVKkwHDujDwqojbTYKFy3St8s2bSJr4iSPx5bv3kPGW29RvmvXKfWjZM1aMl57nYIFP1C6YYPe7h0bq927vNwtIAJtaF3yyJHkzZlDxptvkfPptFrLaBcuXlztvshhw6j/5hsu2RuDry+GwEDqvfIKGAz6fSsFXn01UQ+MwBAYSMOPPyLuxRcRPj4AWDIzKVryC0VLljjuMWTweRUQ1aZuQ15FURRFURRFOccKvv+e4pW/E/XACPxbtyZ/9hx9n7GWAKZs8xYs6Sdd2vK+/JKAtm0I7dNHb5NWK6mPP44lK4uiRT/R/PeVGIKCqr2uJT+f1FGjkEajS3vAVVcR1KULOR9/7NIe9/xz+DZpQtrYp7GVlGDNziHztdf1/daCAuKeexYpJaVr11G+fRvGg4cQBgMxY8dQuNAR2MWOH0fJqtUYDx4k+tFHiRw2tNrhbAGtLyfinnvIdyqkEHH/fcSOG4fw8gJACEHkfUPAZiXz7f8BkD93HhWpqfo5oTffXO17cT5SQZGiKIqiKIpywSjdtImTL/2f9nrjRsLvvANrXp6+33ziBNaiIrxCQz2e75wlEn5+eoGBjNdeJ7h7dwyBgQAY9x/AkpUFgLWwkMLFi4m4++5q+1W2ZYtbQAQQ2q8fQV06uwRF4XfdReTQoQAkfDKV5AdGIqsUNMj7+mvC+t9K7mefUbTEtUR26bp12MrKAK2YQsS99xL1wANIq1UPbGoSM+YpKtLTqTh5kpgnRxPSs6fH40JvvZWsCRORFRUu84x8EhLwb9261vucT9TwOUVRFEVRFOWCIM1mMl57zbFdVkb+V1+7HWfcu9fj+bbycoqXLdO3G06Zgne9egBY8/LInzdf36dXV7PLnz27xnV+yrds1V8LPz9AGzYXdvN/8GvalJC+WhYqqEtn4l50FGAI7NiRhOnTCOrSmdB+ffG77DJth9XKicFD3AIiQA+IAEL69MEQEKDd9xQCIgCv4GASPplK04U/VhsQAXhHRBByww1u7aH9+p0XhRVOhwqKFEVRFEVRlAtC3ldfYT5ytNbjyvfs8dhevPJ3PaDwTUwkqGsXoh95WN+fO2OGvr9qUGQ6fISyja5tzsq2OoKi+MmTabZsKU2XLMErPFxrmzCBZr+tIGHGDAy+vi7nBl1zDY2++IL4SZOIn/Ae2OfyVA1+4l56Ca+wMJdzw/r3r7ZPdSF80EC3ttB+/c7oPc8EFRQpiqIoiqIo55SUkrJt2yhc/DO20tK/dY2KkyfJnjJV3w66rrvLfp/4eP21cY/nTFHBgu/112G39UcIQdiAAY5sUW4u+fPmI61WyjxUPs6ePJmUhx/hWP/byPtmtj7kzVZairFyeJkQBLa/Ct/ERLyCHXOQhJcXvg0b1pph8WvWjKjhw13aQm+5hfhJE4kcMpjGc2bjXb8+AP6tWhHY0WOxtToT2KkTPgkJ+rZv82b4XdLijN7zTFBBkaIoiqIoinJGWYuLPVZLk1KS++VMjt5wIyfuHUz6+PFaMQL7MLTCxYvJmjiJgh8XYjx4sMbhaQXfL0CWlwPgd+mlJEyZQtRDDwHg3aA+9f77X/1Yo4dMUelff1G2wb6GjxCE3XILoFVli374If243BkzKN+2DVtJibbfqbhC+c6dlKxejengQTLfeIOjt9xC6fr1lO3YAVar1rcWLdyyOacr+tFH8LOX5w7u3ZsGb72JsFeM82vWjKaLF9Nw6lStlLjhzH7cFwYDEXffpW+H3dr/Xzd0DlShBUVRFEVRFOUMKvhxIRmvvorw8iJq5ANEDh2qFyvInf4Z2ZMnuxxfun4D5Vu3UpGRSfr4Z1z2BXXpQvwH7+MVEuJ2n7LNm/XXUQ89iPD2JnbsGMLvvAOvyEgMvr4IX1+k2UxFWhqW/Hx9UVZps5H17nv6+WG33eaSWQq74w5ypk3HkpGBNTeXtGee1fcF9+iBJS/XEVA5qTiRTPLDjxDSq5feFtjhn2duDAEBJM6bS0VqKr7Nm7sFIV7BQYT0qn4uUF2LvP9+bKWlSIuVyGFDz9p965IKihRFURRFUc4xKSXZk9+ndP16fBIa4n9pS0L79cW3UaNz3bV/pGjFCk6++CLYbEgg+4MPyZ8zlwYTJuDXvBm506Y5DhYC7JmgnE+nYT52zO16pevXc+K++0mYPg0f+9o+oC08Wr5zp74d2LGj/tr5PfRr2VIvyW3cs5fgbtdq/fxlCcZ9+7Ru+PkR8+Rol/safH2JGT2aky9oBRAsJx0luwM7dSKwYwfSnnwKhCC0bx+Ejw850z/DVlQEFRUUL1/uOL5D+1N782phCAjAr8X5MUxN+PgQM3p07Qeex1RQpCiKoiiKco6VbdxE7vTpgDa0q3jpMvJmzqTpkl/wjow8I/e0lpRSMG8ufi0vI/jarnV23aIlS7ThYhYrBd9/Dzaby35LdjYpjz5KYPv2jqIGzZoRP+E9jt8+AIDStWv1473Cwwlo146SVasAMB04wIl7B9N49mx84rTAyLhnr14626dRI5eAyVlA68udgqLdBHe7ltJNm8iaMEE/JnLoUHzsc4ichd3Wn4IFCyh3KpgAENipI35NmtB08U8u7UHdu2vPYx82p/eh/Zmd46P8PWpOkaIoiqIoyjlW8uefbm3WggJypn5yRu4npSRt9CiyJkwkZeRIipb/qrVXVGDct89tTZzqFP7yC+nPPY/xwAFAywyljX2a/K++Jn/OHP06Po0baZXRoqK0+5SVUbpmjX6d2HFP43/ZZQQ7DTOrFP3EEyR8+gn1334b7CWlK1JTSRszRr9+2VZH0YPA9tVnYvxbX+HS9+QRD5B8/1AsmZkAeEVEEPXgSI/nCoOBei+/rPcBwDsmBt/ERM/3uuQSfa2hSj4JCXogp5xfVFCkKIqiKIpyjpWuW6e/rlyvBiB/3jzMSUkezzHu20fJmrXIKpmYU1H0yxJK12/Qt08+/zxFS5dy7JZbOT7gDlIeeaTGogZSSrImv0/60+MoXLiQ5AcfxFpURM6HH7kd6x0TQ6MZM7TKaF9/pZegrhTYqRPBPXoAEDXyAZd9vo0bE3HXIADCb7+Nhh99CPbCAeXbtpFpnwfkXAmupjk7/q0v11+bjxyldP16fVv4+lL/jdc9zlfSz7/0EiLvv1/fDurSucaiAjGPP6ZXroOaAzbl3FJBkaIoiqIoyjlUkZWF6eBBbcPHhwZvvOH4YG+xkDXJtRCBzWQi4/U3tODlwQfJ/+ab2u9x8iTHbrudY7fcSv5335H5zv9cr1lWRtqYsXoAVrp+A0anOTrOpNVKxssvu8wHsmbncGLYMEyHDwMgAgOJfe5Z4l7+PxK//x7fhg0B8GvalITp0xD2QgsAsePH64FF4FVXEXDVVfq+mKfHIuxr8gCE9OpF7Ngx+nb+N99QsHAh5du26201zdnxa9pUz1bphCDstttotmwpIb17V3uu3qfRowgfeCdB3bsRM3ZsjccagoKo/+p/Eb6+YDAQPmhQrddXzg1R07cA/xYdOnSQWzzUilcURVEURTnfFSxcyMnnnge0rEnjr2ZRvmsXSYMcZY4T580loF07KjIySHn4EUcQhZZNabpsaY0Zi/QXX6RwwQ9u7V7R0dhKSpBGo9u+sDvvoMEbb1C4aBHle/YS/dCDeMfEkPfNbDLfeKPGZ4p68EFin64+YCjftYu8WV8R3KMHYbfc7LKvIjOT7A8/JKB1a8LvvtvtubShf09SvGKF/SG89Hk7XjHRtPjzzxrfi9JNmyj49ju8QkPxa3kpgR21OUFnkjk1FSGES0U75ewTQmyVUnpMJapCC4qiKIqiKHaW3FyMe/YQ2KkThoAAl33WoiLKd+0m8Kor9ZLSdaF0nWMIV1BXreBBQJs2hPbrS9GSpYAWOAW0a0fG62+4BEQA5hMnMO3fj3+rVp6fKT+fop9/8biv3ksvIs1m0u0lpn0bN8Z84gQARUuWEtCmDRkvvwJA+a6dJH7zDblfzNDPD+t/KxVZWS7lqEVgIJEjXBcXrSqgTRviJ07wuM8nLo4Gb75Z7blCCOq//Ramo0e1CnVOhQwC23eodY2coE6dCOrUqcZj6lplpkw5f6nhc4qiKIqiKIDNaOT4wIGkPPwI6c+/4LJPSknKI4+SMnIkSYOHYDvFQgSVKrKyyP3iS44PuovjA+7AePCQdl2bzWVeS2VQBBBxzz3665I//8RaUuJSkMG3eTP9ddGSJdXeu3DBAr0ym3dMDMIe7IXceCMhN91E2K23kvjtfBpOnUqTxT/h16K51reyMj0gAjDu3EXauPFY0rVy1F4REdR79VXqvfyyyxC3yMGD9fV/zhSv4GAafvShW3Cq5uwof5cKihRFURRFuahYi4upSEtzay/bulX/wF+8fDkV6en6PuOuXZRv2waAaf9+8mbMcDu/OkW//srR3teT9e67GHftwrhvH+njnkZWVGA6cABrbi6gBRn+rS7Tzwto1w6DfdK/Jf0keV/OhIoKAPwuu4zYceP0YwuXLPFYGEFareTPmatvx4wZQ/PfV9J47hzi35+sZ1UC2rQhpFdPDL6+hN95Z7XP4rzeTsSQwRj8/fFr0oTY8eMB8G3atNYsUV3xa9aM+m+95dJWV2sAKRcfFRQpiqIoinLBk1JSvGoVqaNGc7hLV470vp68r75yOaZs02bnEyhc/LO+WTUTk/PpNMwpKbXe11ZaSsbLryDtwUwl0+Ej5M2apa3jYxfUpQvC4PhoJnx8XDJHuZ99pr8O6dmT4C5dMISFAVrQVL5jh9v9S1av1oM7r/BwQvv1xTsigsArr3S5l7PQW291yfwghFtGRvj7E3Hvvfp25P330WLdWpr8+MMZzxI5C+1zE7HPPoshMJCw227Dr2XLs3ZvpWZWmyS/9PQyqueSCooURVEURfnXsOTnk//dd5iOHfe4X9psHjMm+V99Reojj1K8YoUeoOTP/9blmLJNm1y2CxctQkqJtFr1uT36fUwmMt54o9Zy2Hlz5mAtKADAOzaWsP636vuyJk5yyeIEX9fd7fzg7o4257WDgnv1Qvj6EnLD9Xpb+nPPceSmm0h++GEs+fnIigpynCrEhQ+8E4O/f439BfCOiCDkppv07cihQ4kePcrlmPABA9yCH++oKAx+frVev65FDR/GJVs20+B/b9c6n0g5e1bsy6Tz/1byyqI9pOSVnevu1EoVWlAURVEU5ZSU/rWR0rVrCB80CN9Gjc76/U1Hj5L8wEgsGRl4RUfTfPkyDEFBAFhycsiZOpWChYvwb9mS+EkT8bGvDyMtFnK/+NLteuajR7EWFOAVHo6trIzyPXtc9x87hnHPHmzl5ViyswGtxLKtrAykpHT1nxy/807ixo0jsLP7ejW20lLyZnyhb8eMHkVY//4Y9+3XSlc7BW9BXbsS2q+fWx+Du13r1uYdF4f/5VpRhbB+/Sj8fgEAFSeS9d9THnoY/9aXY9y5SzvJy4vwu+6u4d11FffiC2AQeEfHEPPUkwDkz55DRUoKwseHyOHDTvlaZ0N1WS/l3JBSMu3PoxgrbMzacIJAP2+e7XN+Z/HU3yBFURRFUWpVkZlFysMPk/v5DNKeHue2v2TNWlKffIqSNWv+9j3MqWmU79nrMdNTvmMHJ+4djCUjAwBrTg5l9jk+hYsWcfTGm8ifMxdZVkb5tm0k3XsvpmPHACj+bSWWzEwAvKKi9EICAGX2IWflO3bo83WcFS5c5DJ0Lqz/rUTc4wguTPv2kzziAY7dcgs5n32Gcf9+bGYzNrOZ3Bkz9CyRT3w8Yf37I3x8qPffV1zuEdqvHwmfTEV4u39X7R0Tg//ll7u0Bffq6VjXp1MnfBs3djvPuHs3BXPn6dsxo0bh2/DUy0F7R0QQ/+67xD0zHoOvLwZfXxrP/JLIYcNImD4N34SEU76WcvHZciKf7cna331fLwPDuySe2w6dApUpUhRFURSlVkU/L9YrmBl378aclIRvYiIA1oICUkePRpaXU7xyJY2/mkXgVVdRvmMHxv378W3SBL9LL61xronx4CGSBg5Ems3Uf/NNwu8YoO+ryMwkeeSD2EpKXM4p27IVv2bNSH/xJbBYXPZZ0k9y4t7BJEyfRv7s2Xp7xF2DsJWWYjp8BIDy7TsI6dGDUqehc36XXILpkFYdrnDhQnDKAIX260dA27YYQkPJmzlLX9/HfOQo2RMnkT1xElQGN059inrkYX2eTmD79sSOe5q8OXMIu/VWYkaPrjHTEXxdd4x79+rbIb166a+FtzcJMz6nZOVKvCIisGRlkTVhouv5vXsT9dCD1V7/VPnExxP33LP/+DpK3ZJSsiu1kPrh/sSG1D488myYtvqo/nrAVfHEhp4f/aqJyhQpiqIoilIjKSWFCxe5tBX9ukJ/XbBgAbK8XNuwWEh78iky3niTpLvvIePV10geNpzDnbuQ/tzzHrNAAHmzZulzZpyDGID8OXMdAZGXl95etnULJatX68GHT3w8sePHIexFAawFBZy4fyhlm+0FFLy9Cb/rbgKuvFK/RmVFOeciC9GPPYp3/fqANgSu8t7ecXEEXHUVwl11ETwAACAASURBVMeH2KeeotnyZYQPGqSXuNZZLC4BkU98POH9+7scEjVyJC1+/53Yp56qdeiX87wiQ2AggVdf7bLft2FDIocOJezWW4kaOdJl/o9v48baXBs1vOyCNXXVUfpPWUef99dwsrD8XHeHw5nF/LY/S98e2a3pOezNqVP/QhRFURRFqZHpwAFtDoyT4l9/BdxLPgNYsrPJ/+Ybt+sULlyoZ2CcWUtKKFrqKGRg3LcPc2oqADaTiYJvHQUR6r30ouO4XbspXvGbvh05bBhRDzxA45lf4hUervXPnskBCLnhenziYgm48iq9rXz3bqzFxZTv3q23BXbqRMyoUY6Mj134HXe4BBc+cXHUf+1VLlm7hgbv/I+QG27Ax2mulU+jRoT06UPDKR8jfH3dnvtU+bdpQ3CvXmAwEPXIIxhquVb0o49S77VXCb/nbhp9+QVe9rLeyr+LlJIlu08yYflB8mqo4vbrPm1oaF6pmc/XeC5AcjZ9tuaY/vqGVnE0jw0+h705dWr4nKIoiqKcQ8WrVpE7/TNC+/Qh8v77zth9ZEUFJWvWaEPZmjQ5rXOrZokAjHv2UJGWhvHgIX3NH0NgIDajEZwqsgW0bUtFZqY+F6h07Vr8L73U5VpFP//syDTZFf+6gqgRwylauhRrfj4A3vXrEz5wIHkzZ2E+cQJpNrssfBrcvZt2zzZtaDxnNskjR+rrDoG2qCiAT1wsPvHxVKSlIY1GLairXP+nRQu8IyMJH3A7wT17YNy7D9PBgxiCg12G9DkzBAUR1r8/YfZskK20VG+vC8JgIGHqFGzl5RiqZqU8HS8EEYMG1cm9lXOjsKyC53/cxZLd2r+bNUdyWPBIZ7y93PMZJwsc/3bmbkpmdK8WhAX6uB13Nuw/WcSP2x1rgD1y3b8jSwQqU6QoiqIo50z5zp2kjhpN+bZtZL79NuZU9wVF60rWhImkPvY4SXcO1AsQnAppsVD4s2O9Hq+oKP110YoVrvN17r2HuGefAYMBfHyIe/45Gs+bq2Vd7ErXrUNKycn/+z8Ode5C1qTJ5H/rWhoboHiFNjwvf/Ycx/XvuQfh7U2AhwU6fRMTXQoO+DVtSuKcOfhdcgmgVXcLaO84z3kIXc7UqfrrwI4d9dfeEREEX9uVqAdGEHHXII+FEDwxBAXVWUDkct1TCIiU88fawzkMmLqOKX8cqXbYqCdHskro9+EaPSAC2JlSwJQ/jroda7bYyC4x6dtlZivfbDzxzzr+N5ksVsbM30GFVXvWTk0iad848pz05e9QmSJFURRFOQXWkhKtJHFqCtGPPIJPvFbJy3ziBKWbNmE6eAhbSQlRD4zAr0WLWq9nyc0ldfSTjopnUlK6YT2+AwfWed9tJhMF332nvS4tJWvCRBKmTqF0wwYKFy4i7I4BBHXq5PHc0nXrsObmAloltOjHHyPjv68CkDttup7FwWAg/O578G0YT3CvXniFheEVGgpAUNcu+vXKtmylZNUqCr7TFi3NnT5d3yd8fZFWK1itlG/fTtHyXzHah7UJX1/C77wDgMD2HShc8INLPz2t8eNTrx5NfliA6dAhfJs3dymZHXDVlRTZg73KAhJ4eRE2wHM2SDm/FZZXEOrv/bfXKbLaJCUmC2EBdZNhKTdbGTV3G/llFWxLLqB+mD8Drmro8ViTxYqxwkZYgA8mi5Un5mwjrcB9btBHvx+mZ8sY2jQM19syi4xUjbe+XJfETZfXI9jPm3phZ77AQU6JiWKjha83nOBARjEAft4G3h5wxRm/d11SQZGiKIpywTEnJZH5zrv41K9H3AsvnPI3/FJKjDt3Yi0uJqhrV4TBgJSSgnnzyP7oY6x5eQBY8vJJmPIxpX/9RfLwES7rzZRu2kizpUtrXMTSnJpK+jPP6mWiK5Vt+IuIgQOpyMjAuG8fwddei/D1xWYykfnW21SkphJ8XXdC//MfvJ0yNubkZMzJKQR1vgbhVIhA79P69draOnYlv/9O1gcfkDv9M7BaKfr1V5r98jM+DRq4nVu0/Ff9dejNNxNyww1kvPY62GyOgAgI7tFDL/lctVyzT716+DZvhvnIUaTZTMYr//X4voTceCOWnBzK/voLgLSnnnLcu18/vCO1b50DPWSKgrq7B0WgVWfzb9XKrT3QKVNUKXbcOAJaX+7Wrpy/CsrMvPDjbpbszuDa5tF8NaITBsPpBUZlZgu3T1nPwcxi3ritNUOucS9xfrq+25pCfpmjxPsri/ZyddMo4sNds31JOaUMmraBvFIzj/ZoRrnZqgcW/j4GJg9qx4y1x9lyIh+LTfL4nG1MG9KBVg20Lxw8BU85JSaun7QagDuuasjEQW3/8fNUZ8ofR3hv+UG39uf7tqRZzL9jLlElNXxOURRFuaBYi4pIfuhhSv74g/w5cyn4/nt9n81orHYYS/nevSSPGEHS3feQ8uBDevGA/LlzyXj1NT0gAm1ejK28nPx586n6Na0l/ST5X3/t8R42o5HMd97lWN9+etUzZ6V//YW1sJCku+8h9bHHSXnkUaSU5E6bRsH8+ZSuW0fmW29zuPt1+mKk5qQkjt9xJykjR5L9wYce71vsVCmuUu4nn4LVCoAsLyfz7f+5HSOtVkpWrdK3Q/vchHdUFIFXu2aV/C67jNjx7msXOQvu6liE1JLlqEzlXMwg4q5BhNx4g1MHtPdW+PkROWK43uyTkIB3TIy+LQICXIa9nQq/Sy5xGeIW2q8vkcOGntY1lHNrc1IefT9wDDNbeySHPemFp32dn3ee5GCmFoi8t/wgpSZLLWfUzGK1uRU8KDZZGD13O7PWJ/HDtlQKyyuwWG2M+XYHWcUmLDbJR78f4fO1jvOe73sZfa+oz8RBbQn01b7sSMkr57ap6/hqQxKAS7U5Xw/zjRZsSyXdQ+BUVWp+GbM3nmDW+iS+2pDEQXtgBlqQNX9zMrPWJzFrfRLbk7UvQyqsNj5d5T6kr2vzKO7vnFjrPc83KlOkKIqiXDCkzUb6M89SkZystxV8+x0Rd99N9kcfkzNlCsHX96bhRx+5DLMpXLSI9Gefc7lWwQ8LiBwymPw5c6hKmkyUrF1LqdNCpcG9e1OyciUAOdOmE3bHHS7r8kgpSR8/3qVaGkDM2LHkffEF1oICrHl5ZL77rqMowfr15H72OblfznTtgNVK1rvvEnh1J/JmzcJWrH2AyZ8zh+gnHkcIQc6n00AIIocNpfj33x3nGgwuhRAqFa9YQfGqVfg1bYq0WvFr0oTynbv0YNArOhr/K7ThMPVfeYXsj6fgHRtL2C0349eyZa3DloKu7UrerFmubV06EzNmLPnffI1/mzYEduyIT+PGZL7+hh4Q+TRsSPyE9/C3zw0CrZBAYMcOFC3RKtYFde5ca0W2qoSXF5EPjCDnw48I7NCB+q+//reHXilnX2aRkeFfbqakSgCzJSnfZXjZqVi8K11/XVhewfzNKYy49vSKkThbtjeD5DwtMxvs502Z2YJNwtYT+Ww9oQUUcaF+XNs8Rl/gtKpuLaK5z56xahwVxOS72vHUvB2UV1gxW2y8vGgvCZGBpBc4qivee3UjSkwWtiTlkVNi1t+bNYezuatjI4/3ASgyVnD71PVkFzvmJvn7GPhldDcSo4IY8vlGPXsF2rJd3z/SBZuUFNvv4ettoEGYP4nRQbx7Z5vTztadD1SmSFEURalTNpOJktWrseTknPV753z6qUtmA7TyzgULF5LzyScAlPy2EvORI/p+m9lM5jvvul3LtG8/pRs3YT6ifRMq/PwIv+sufX/2+x/oVcZ8EhJo+P5kfTFTW3ExuZ9+6nK93M8+dwmIAtq1o/Gc2UQ/9KDLujNV58pkT5qkV2bzbdwY3+bN9H0nX3yJop9/cTxLSQmla9aQ8+k0cqZMIefjj0kaOAhbofbtuXf9+kQOH6Yf7xUVRXDPnvp26iOPcvTGmzjWt5/2Xv7hCKZCevbQy1H7JiYSP+E94p4Zj/9ll51SMBHYoYNbWeqIwYMJuKI1Dd55x1EZLjaWqEce1iq63XYbTX78gYB27dyuF3JTH/11WJU1gE5VzGOPccmmjTT6+qszUhhBOXOW783QP/Q7//WrDDpOVU6JiXVHXH9WzVh7nAqr+xcHnmw6nsf473ay4ag2705KybTVjkImI7om8miPZm7nZRaZWLAtVd++tnk0XvZAItTf2y2wuOnyevw8+louqx+qty3fk+GSBWoUGciEgW1ZNb4no3s319v/PKw937xNyTz7/S72pRe59GXuxmSXgAjAWGFj+upjLNuT4RIQac+oVbn781C23nbHVfGsGt+TmcM7nTcLyJ4ulSlSFEVR6lT6s89RvGwZPgkJNF38Ewb/s/MfZMmff5Lz0cf6tiEsTA8GTr74kkt2pHzXbr0YQvGyZXo2xDs2Fu969TDu2gVAxmuv6ecEdelCaN++FMyfD4D5qGPYSEivnggfH2KeHkvaqNEA5M2ZS/jdd+PXpAml69eT/f77+vERQ4YQ9+ILejAR1Pkaipcvr/UZ417+P3zq1+fYLbeC1YrpwAG3Ywq+X+AyNM983DEcJ+SG64l57DEq0tOxZGdT76WX8I6L41ifvlgLXYcdZX88xSXTFdyzV639q4khIIDADu0pXb8BAO8G9Qnu0cPjsbFPPknM6NE1BlshN95AwrRPQQiXxU1PV2UxCOXf5c9DjkDmrg4JzNucAsCWE3lIKU8567d090lsVUbUphWUs2T3Sfq3i6/xXGOFlYe/3kJ+WQW/H8hi4wu9OZBRzO407d+Sn7eB+7skEhHoS2yIP0ezS5ASftl90mXdoSsbhTNzeEcOZBSzfG8G/a6oT/0w90qDzWKCeaFfS+6bsQmAg5nFhDsVhmjgNF+pW4sYQPv5sPZwDuuP5PDcD1rRkh+3p/Fc35YM75qI2Wrji3WOnxHXNo9mrT1I/HF7GluTHUFm24RwdqZoma3lezJoGBlY5X7/bipTpCiKotQZabVSYh+qVZGSQum6dWflvuaUFNLGP6MPuQrs1ImG7092HGCfO1OpfPcu/XXeN84lpe8lrP+tjus6BT7BvXoSeNWVHjMKlQFDyPXXE3CVfWHQigoyX3+dirQ00sY+rQdlAe3bE/fsMy4f2oKuucbtmlWLAwRdey3BXbvi17Qp4QPv9PxGACV//OEW4FQKveEGDEFBNJw8mcRvvsG/ZUu8IyKo999XoLJAQ+UcH4sFS7b2TbDw9yeos3sfT1fw9dfrryOH3OexKESl2j7UCiEIvu66fxQQXUyyio3c/NEabv5oDVlFxtpPOA8YK6wM+XwjvSeuYv9JR3bDbLGx4agjKBrZrSnBftrf28wik8fiA86m/HGEDm/8xksLd7Ngm6MMftNox7/tj38/Qm6JydPpulUHs/RiCrmlZk7klbHXaU5Tr5axRAf74WUQDO2SyGv9W/P6ba1Z9mQ3ujbXCqWEBfgwaVA7vL0MtI4P4+kbL3XJBlV1aT3HQryHMopdhs81CHd8AdWyXggxIVqxl8LyCp79wfEzz2y18drP+3hg1hZmrksis0h7zpgQP2YM68BVjcL1445klQDa8LjP7+9AI3sgVGyy6H8mBgFdm0XX+F79G6igSFEURakz5uRkpNnxDWjxr7/WcHTdsJlMpI5+0jFErF494idPIvCaa1zWrXFm3KV9Y1q+a5eeFRI+PoQPvJMQp+FkzkJ69ED4+hLUpYtLuyE0lMD2WiAkhKDeSy9q83aA0vUbSLr7HqwF2rerXjHRxE+ehPBxLfvr07gx3vXrOxqEIH7yJIJ799Y2fX1dChnEPP44ItDxLW1Q1674NnMfouPc5hUd7QjYqgjt25cWa9fQYs2fNP3xB5fiB6BlyepijZyIQYOIHfc0sc88Q+TQ+//x9ZRTN3NdEnvSitiTVsSsDUnVHme22NibXojlFIePnUk/7Uhn7ZEcjmaX8onThP5tyfmUmrUvOhIiA2gWE8SVjRzziGoaQncsu4QJvx4kp8TEN38ls8Oe+fAyCKbf354AHy1QP5xVQt8P1rgNrTuaXUKhPRBavPOky75DGcUczCjRt6sLbmJD/fl6xNUserwrK8Z2p0n0qQ/djAn2I8K+MGup2crhLMfQNudMkRCCbi0cgUpKnnug+PuBLN5e6sg2j+jaBD9vLx7q7v6z5M72DYkJ8eOWtvXd9rVNCD9ni8XWpbMeFAkh+gghDgohjgghnvOwv5EQ4g8hxHYhxC4hRL+z3UdFUZQLmbRaKVq2jLJt2+v82qbDh122i3//wyVIOhOKly/HtH8/oAU2DT94H++oKIQQLhkV4VQi23joEDaj0WVh0NB+/fCOisKnQQP8LrvM5R7+bdvo1c6qrocT3L27S5Dj36oVEfb5MYCebcHbm4YffIBPbKzbMwghXLJFQd274du4MfGTJ1H/rbdoPHcO/pdequ/3jokh9kltmJ4ICCBmzBhC+/V1uaZXeDiJ8+cTaV83qf5rr9aYmfGOiMA7Jga/Fi2IqlKFLaT3Pxs6V0l4exM1ciRRI4bX2Bel7m1OclRP3JXqOZNYYbVx/xcb+c+Ha3nwqy1nq2vV2pHqKEKwy+n1msOOuSzdWsQghKB9Y8dQzy1J1QdFn6057rauD0CXZlE0jw3hub4t9basYhNDZmzk171a4ZNJKw7Re+Jqbpi8mgMZRfy237Wk/sHMYg5lOoKUS+JCqI7BIGibEH7a82+EEC7XrRz65+ttICrIdc7edZe4D2m7/rI4HuzmXkQi2M+be6/WijHc0CrOJWsmBDzYrSkAt7R1L9vf/QIYOgdnOSgSQngBU4C+QCvgHiFE1cUDXgK+lVJeCdwNTEVRFEWpM3kzZ5H21BhODBlCcZWiBP9U1aDIVlxM6cZNNZ5j3L+frAkT3M6tTumGDWRNmIA5RZtDULZlq74v8oERBLR1rMkRfuederYo9pnx+DaxfxiwWChdt46iJUv0YyOGOAKZqtmiEKf5NEHdXIOikF7umaWY0aPwinEdThL37LMEVpOpAYi4+y7w8dHmJj3+OAAGX1/CB9xOwOXua+dEDh1K4nff0nTRQgJaX05oX9fvECMGD8YrOIi48eNpuvgnQnqdemAT/eijeDfQvhEWgYHVzv1R/h1MFis7nQKhPWmFSCkpMlbw5brjrLVPxJ/yxxH+OqYFT38czKawvMLj9Ww2yY/bU/lxe2q1Je7rwt40R5+TcssoMmr9cZ5PVPmBvEPjSL1tSzWZouxi1+IGvt6Oj8G32ecPDe2SyMzhHYkO1gIMKeG5H3azeGc6H/2u/YzKKjZxz/S/MFlcs2mHMov10t7gOtStLnm6bv0wf7chp12buw9pe7xnM178TyuXZwStcl3lwrVeBsFD3Zvq+/pcXk/PZrWsF8olca7rD3W/5N8/dA7OfqGFTsARKeUxACHEPKA/sM/pGAlU5hvDgHQURVGUOqMHAjYbma+9TtDVV1O+Ywf5c+bi3+YKIgcPxuA0NOt0mA4fcWsr/vVXgrtd6+FosOTlcWLoMGxFReR/9z3NlvzisihpVRWZmdraPSYTZVu3kTh3DuU7d+r7g67p7HK8V3g4TRb+iLW4GJ/YWIy7dumFBzLfehtZoX3I8m/ThoArHKuvB/fuRc5Ux3dyzpkSn7hYAjt1omzTJrzCwwnq1s2tn14hIdR74QXSxowFIPSWW1yCLk8C2ralxepVCIMBr/BTKyns3Ge/pk0I6NCe8i1bMYSEEDH43lO6hieGoCASv/6avDlzCO7WvcY/E+X8tyetCLPTB/j8sgrSC418+Nth5m/Rvly46fI4ftuf5XLesewSrmwUQVWLd6UzZr72767CIhnUMcHtmH+qwmpjf5WqZ/vSi2gRG6yvReRlEHSxz81p1ygcg9AyJwcziig2VhDi7zqka9b6JP19aNswjPfvvpKpfxwhLtSf2690FFXocWksS0Z3o/+UdZwsNJJXambUXNfMuvPCrJU2Hc8nxz4Pyc/boM+/qWueMlANPBRmiA72o3V8KHvStLk/VzeJ1P88e1way9Inu/PhysNIJGOuv8Tl3Ls6JpBbaiatoJxnb2rpsu+WNg2YuOIQACH+3rQ9zRLo56uzHRTFAylO26nA1VWO+S/wqxBiFBAEXI+iKIpSJ2xlZRidKpZVpKeT8uhjlG3eDFYrxStWkP/V18Q89SRhAwac9rotnrI9xStXUu+/r3gcLpUz9RNsRdp/2LbCQrImTKTB229Ve/2ipUuRJu1DR/n27ZhPnHDc02DAv3Vrt3MMAQH6fBj/K9pQuOgn7dnTHBOsI6sEEP6tWuHfujXGPXsIaNcO3+bNXfbHT5xA0bLlBF7dCa8Qz98Gh/btiyEoCEtuHmG33HxK76V3ZGStx9Sk4eTJFC1dSuDVV//ja/nExxM3frxL25yNyfy6L4PHezanY+I/u/75as7GZJbv1Z6xU5O6ecaDGcW8u+wA7RLCGdW7RZ1c83RsPZHn1rYzpYAlexxzYpbvzXQ75lh2qcegaPVBx/C1eZuTz0hQdDizxCWQAy3DlVlk1Ie/XZkQTqg98An286ZlvVD2nSzCJmF7cgHdnYaP5ZSY+PqvE/r2w9c1o0l0EO8NbIsnsaH+vHdnW4bM2OjSLoTbes0u96jUIi5YL7Fd1zxmisI9D8Pr3zaePWlFCAGjern+3YsJ8eP129x/ZoI2TO/xns097rvtynimrjpKeYWVm9s0wNvDorH/Rmc7KPL0t6PqX617gJlSyolCiM7A10KI1lJKl38ZQoiHgIcAGjWqfkEqRVGUfxtLTg4pDz2MJTeXBv97m6DOnWs/6RSV79rtVomt7K+/XO+fnc3JF1/CfCKZ2LFj9HYpJQXffYclK5uQ3r3cFuy0mc2Yk5L0ba/ISKx5eVjz8kgadBf+V7QmpHdvgjp3Rnh7Y05KIn/ePJd7F/74I+F3DCCwQwf9ntb8fLzCwhBeXvpinZVypk3Xq7r5tWiBV3DNE5YD2lzh1uYVGUlInz4ubUIIGn0xg7KtW7X1daoENN4xMUTeN6TGewFnvTKad0wMkfefmQIG2cUm/m/RHqw2SUahkWVPnbuqb2aLzWXoU13JKjLqz5hWUM5vY6+rk+u+tWQ/qw9ls/JAFj0ujeWKhmEu+80WGz5eotrA2WqT/+gDtqc5Nt/8dYJio8XD0Q7Hcko8tu92Gta2LbmAlLwyEqrJipgsVvy8Xb8Qqfo8VpvEJiU+Th+u96S7z3vak1bo8qGxe5U5Mx0SI9hnr4i2cn+mvn/dkRyemr9DHw7YOCqQmy6v57G/zq5tEc2wLonMXJ+kt00a1JYZa4/r2Ze2CeEUlJk5kVvmcu6lcWeu1LunTFF8uOdCKMO7JlI/3J/oYD+uaVo3Gd+EyEC+f7QzBzOK6dO69vfx3+Jsh3apgPPXCQ1xHx73APAtgJRyA+APuA1WlFJOl1J2kFJ2iIm5MCZ4KYqiAGRNnIRx3z4smZmkPP6Ey/AwT2xlZWS89hrJI0aQPGIEqaNGUb57j8djy7dv89gOWqUyb6efp7nTp5M7c6a+XbJyJRkvv0LOxx9z/PYBHO9/G8W/Oxb3NB8/rgdcPgkJhNx0o77PuHcvBfPmk/LgQxzu2ZP0518g/YUXwWL/UOb0YTD9hRfJ/fxzsj+ewtGb+nC4S1eShw3HdOSIXimuUuGiRfprTwt8VuXXsiVUqfwWPnAgBqciDJW8QkMJ6dmz2kzQxeZ4TilW+6zuQ5nFmCzWWs44Mz5fc4zLX1nGY7O31vl8lj3phfozHslyVBn7J2w2yTanOS5VP+xvOp5H+zdW0GviarKKXUtlJ+WUcvVbv3HtO7+zJ81zcYTaSCk9VmNbb19sFODyBqGE+HljENC7paMQyNGsUrfzyswWjma7Bks/7zrpdlxOiYkHv9rCZf+3jJcW7tbPHfjpei57eRkLtmpze/akFdLxzd/o9OZvLNzuyN56et6tyfn8ts+R0erV0rVoyY2tHB/Qv92SSn6pmXmbkhkyY6PL4qTP3NTylIPMZ/u01Cvb3XdNY26/siGTB7WjXqg/BgFP9W7hMUi5tF6wW1tdCQvwoX6Ya2bI07pGAN5eBm5u06DOAqJKlzcIY8BVDQn0vXCWPD3bQdFmoIUQookQwhetkMJPVY5JBnoDCCEuQwuKslEURbkIGA8coHDhQn1blpWR8vAjmJzWy6kq6/33yZ8zl9L1Gyhdv4HiFb+R+thjWEvcv+Ut2+4YFx85fDjCvrBqYMeOJM6bS7Ply1wm1Wf97x0Kf/kFgAKnfgGYDh0i9fEnyPv6G/u2Y+icX4sWRI8c6bFMtDU7h8Iff3RZYDT+g/f1EtMVyclkTZhIzscfU5GcrPV782aSH3rI/eGdsl7OBRaqY/D1xb+l0/h4Ly+twIFSq5OFjpK+NgnJVb4ZLzVZWHUwi2LjPw8kqlNhtfH+b4epsEqW7M7Qv6335EhWMZuO52GrujJnDZzLKQMua87Uptxs5Y8DWS5DqACS88ooNjkyMseqBBSzNiRRbLRwPKeUqX+4/jv/+I8jZBaZOFloZPTc7ZSbTz8QPZFbRq59odDqAoFHezRjwwu92fjC9S7D+zxlivafLHZb7HTxTtfvt9cczqbP+2tYsS8Tm4Rv/kpmT1ohczelsDkpH7PFxvM/7GZnSgGj520nr9RMflkFT83fwdj5OygxWTwGRSl55Xop7qYxQVzewDUb07V5lF4Gu7zCyms/7+Pln/bqw92ig32ZObwj/2njXla6OgG+Xnz7cGc2PN+L1/prBU9axIWw8unr2PnKjfRsGculHoKimirP1YWq169u+Jxy6s5qUCSltABPAMuB/WhV5vYKIV4TQlSulvc08KAQYicwFxgmz2RpE0VRlPNI1nsT3AasWwsKtOF0+e7f9lZkZVEw/1u3dkt2NjkffeTSJm02ync4sk4Rd99FkwXf03DqFBp9MQOvkBAMgYHET55EQPv2qnHCSgAAIABJREFU+nGZb76FJTub0tV/6m2VwRRSkvnmm2S9/z6mQ4f0/X7Nm+MTH0+zX36mxZo/SfjsMyKHDsUr2r1KUWi/voTeeCOxT4+t8b2xpLt/G+0soF3tQRG4FicI6d0bn/qn/gHpYua8SCTA0WxHFkFKyX0zNjLsy82MmLn5jFUk255cQIlTgPHnYffvTM0WG28t2c/1k/5k0LQNDJ+52SVLUBPncsrgOkysNk/N387wmZvp//E6l+ClamboWLZr9uWgUzGB+ZtTyLcHMCaLleV7Mhzn5ZTyzrIDnC7nSmzXNo8myNd1KJsQWnuwnzcxIX40jXEMQU3KLdMzZ/rzeHhP9p0s0hf5XHckh6FfbHILDqf8cYQZa47p22arjYHTNri9Hz9sT+OBmZv1YXAAsSHumdxb2jRwG24ohOBhp6ppP25P0+cltawXwpInu9HjUveS+LXx8TJQPyzAdcFlP2+9kMMlHub4nKnKc9Vdv7rhc8qpO+s5LynlEmBJlbaXnV7vA7qe7X4pinJxsuTmYis34tswvtpjpNWK6chRfBMa/u2qbKeiZO06Stet0zYMBuq/+SYZr7+OLCujIi2N9PHPkDDtU5eCBbmff64XHvBv1YrQfn3JmjARgLyvvyHsttvwt6+5YzpyRC9q4BUVhU+jRggh8KuSzTEEBJDwyVSO9b8Ny8mTWPPySBsz1lGprVUrEmZ8Tuqjj1G+Y4fWj0+nYXAaZubXwvFts3dMDMExMQR3u5bY8eMo374d4/4DGA8dxBAQSIx9vZ3IwYMJvPJKynfuxHjgINJsJrh7Nwq++57S9ev16wk/PwKvudolSDOEhuKbmHhK73P4XXdRuHgxCEH0E4+f0jkKpBe4Lv7onEXILjGxLVlbR2ZzUj7HckppFlP3w4fWVAmCVh/K5vGezVl1MIufd53EYrVxIKOYA06BxupD2fT9YA3X2quUtU+MZMjVjTzO3zlYpdrZnvTqM1HOdqcW6oUK0grKWXkgk5vbaOu5VA2sjuU4ggCTxcpxp+3yCivf/HWCUb1bsPpgtkuGCWDm+iSyi034eDn63iA8gMd6NifYz5sSk4VPVh0hOtiPwVc3xtfb4FJkoWNiBGVmC5ud5hi1aRhOeKCjNHOovw/RwX7klJgwW2yk5ZfTKMrxc885KPI2CCz2oOmnnek8cG0Txn23U88kRQT66FXaljoFeJWcCym0b/z/7J11eBVn2ofvOZaTE3d3AUIIBNdQnGKFlpa6bL2lUN16t7JfZbvbbnXr3a2w1IUq0iLB3SEQISEJxF2OzffHyZkzk3MSQptS2bmvi+vKzLwz8x5JeH/zPM/vCZLS/DYXuuYc6W9kdGoIn+4oVZzrqV8OwMysKJ7+/jClsu+rQavhuQuzT7snUE/pHCnyM+qI9P9lIzdukaIANVL0c/njJAKqqKioyLC3tNCUm4spO1tRJyOnac0ajt96G2JrK7EvvYjfpEkex5U/8CD1n32GV0Y/EpcsQWP0/J9P28GD2OrqMI0ciSAIiHY7TWvWYD3pWCzZamtpO3QYc2EhYkctjSE5iahHH0Xr50fFU09K1wo87zwC581F6+/H8ZsXAtCcm8vJp54i+NJL0cfGYq2qUkSJQhfdgu/48TStX0/Lxk1gt3N88a34TZqE96BBWKtdvT1Mg7O7dUPT+vsTcuUVnHzCMaeWba5Gjv4zZ6ALCiL+7bc4fuutkjixN7oWlF7pnh22BJ0O07BhmIYN83jcmJGBMUPZvs47O5uCOedIgs53/Hh8x49XiCLvrCwETc+SH4x90klbuwZBp1M0XVXpHnn6HCjrTfI6pZ2ty6v8RUTR2jylKNpxrJYDZQ1c+842LLauo1NVTe18vsuR4vX5rjLC/bzcCu2tNjtHO6W27e9hpOjVtcq0t2W7yyRRtL9Til9xTYtkFFFQ2ewWifn3hiKuzUlmmaxOx1uvpdXiiD59vdc9YtpitvHwnP48szyPt9Y7LOc/3VHKxL7hfLTN1ZdnSEIw1c1mhSgan+YevU0J85EiPflVTQpRJBd5F4+I552NDke3l388yg+HTlJe74goBpn0fLs4h7s+3s26I1XI6RvppxCuk/qG88YVQ/nH8jxe/FFp658Z409mdIBCFGVE+ZMa7vn7pddquHpsEo9+5er2cue09F80cpMU6qMQiH0i/E7bufN0kQsxP6POzX5c5fT5Y3joqaioqMgQRZHjCxdSumgxBeeei6Wiwm1My46dHF98K2KrY6FX8957Hq/Vuns39Z99BkD7gYPUvPuux3EN33xD4XnzKb7qT9R1OKpVv/Emx2+8iRMPP8KJhx+h8rnnafz+e9rz8jAXFGAuKKBp5SpKF99K9VtvSz1+NCYToQsdQshv0iRCrr9euk/tO++SP3UahwZkcXTCRFeUKDMT3/HjEQSByAcfkswELMXF1Lz9NqWLF3Py/1xW196Dsk/5PgbOn4/G391Byb/DqU3j7U3s88+722Brta4mqb2APjKSqEcfdbwmvZ7gK6/EZ6wyoaAnJgtyNN7eqiA6TUo7pc/JI0WHO6Wdre20CHZS22zm0jc2c9mbmzlR3+ZxTFfUNpvZ00mkWO0it/x3h5sgMmg1PDQrg3f+NJwwD6lXziJ/Occ6xIqcgqrmU9ZIFVe38E0nofLj4Uoa2iyIouiWPmezixTXOARl53Q9gOpmMy/9eFRhKPD65UM9ppA5+WJXKW0WG5/vcgmHvaX1PLfqiLRQTwr1ITs+kMxopfNdZwc3gGSZoJWnt7VZbBypcH3uCyemEhfsSNuy2kVFjdfj8wYQGWBUNAEFh8B7/5oR0n0j/L144rwBCILA4slpDIhRzi8zJoDMTvu6ihI5WTAsjoQOITcuLZSrxyZ3O/7nYtBpFGmHntLpepu+UX7SPSf1Pf2UQBV31EiRiorKH47m3PU0b9gIOIr6Tzz4ELGv/Et6ctd+5AglN9yA2OZalLVu2469tVXqZ+Ok8qWXFNvVr75G4Pz56IJcvTva8vIou/8ByRq6+s23CDzvPGrefadH823ZutXRJ6iD0FtuQR/h+k8ubNEttO3b50qtAzdb7dCFN0uvzys5ifBbFzvqk+TITQkGn1oUaXx8CLrwQqpfe8113qBB6GNcqYYaLy9in3+OwvPmY+uoeTIkJqIxGNyu93Pwnz4NY/8MBK0WfbRjQeSVlioJSSEzS0rrMeo1pIT5/uQntU3tVgQcNQOng90uUtXc/rNSdOpbLGi1Ar6nee+fiiiKlNW3ER1gPOX71TlSVFDZjCiKCIJAXqe0s4351R6tmJ9bdYTcow7BtGjpTv577cgeu4DlHq3y2B9GXtt059R04oJNDEsMJrqjxuKHO8az/mgVFY3tPPTFfgBWd4gWf9nT9c6vwcmBsgZGdDh3iaJIeX0bUbL3683cAjfjAbPVzvL9JxmRFEydBwe7/MpmUsP9FOl6YX5eUu3TCz+4oiVp4b6MSQ3hq0Vj2VRQg83uEm5PfnuIkw3t1LZY+Nt3h6npqEfqzMC4QF64MBujXkuWzA7cz0vHwDj3xpspsgW+3Bji0IlGKbKVFOpDuJ+RJdeMZNHSnezsSJ8EODc7hrMHOGr1xqaGkhHlL9UILRgWR4ivF29eMZSN+dVkRPsT6usQfHqthmcXDGTm87m0dwjUATEBZET7S41ZAWadwijBx0vHpzeO5tCJRkYkBf9i/YLk9I30J++k473qdwZEkV6r4bMbx7CvrL7X+mn9r6NGilRUVP5QiKJI1YsvKvY1rVlD/SefSMfL7n9ASsWSzrNYFCliAK179tC8dp1in72piZOPPUb5Q3+h6JJLKbv7bo4vvEWKOAFYjh/n5JNPYat0LP60gYEELlhA8BWXE/XXx0j88AOSv/6KkBuupzNe6ekEX3qJYp+g1RL74guELV6EadRItDJBhkZDwLx5+I5X9lMJufpqUlauJOa55wi55mo0vq4nv4K3N8b+/T2+f50JuvQSRUTFf8YMtzH66GhinvkH6BwLeb+JE3t07dPFEBcnCSKAsNtuQxsURNvE6Uxe08qsF3KZ9UIuk59Zy43vdW093h2HTjQw6R+rGfDw92zI9xzt8ITVZmfBaxsZ/n+reEyWtnM67CutZ8QTKxny2IqfbL98OuRXNjH7xVzGPPkDC5fs7HZsi9nqtrivb7VIi/DOkaJWi83NBtpmFxWpX1sKa3grt7DH85Wnzg330Dg2I8qfmyekcs6gGEkQAfgZ9UzPjOLyUYmSW5nZ5hAtcjq/BifOuqL8yibmvLie0U/+wOKljlq6uhYzH2xz9aSXR12W7S7r0r3OGX2RR4rumJIuOafJmT3QYSgQ7mdkzsBo5mXHSv9mZ7l+H5xpcwBTMiLIiPLHqNdw01kpfHzDKCkFLjXcl9kDozFoNdw6JV3RG8iJPOoht9+Wfy+d72VcsIkPrx/FzRNSMOg0DIgJ4C9zXH9fBEHgsbmZBJr09InwkxqC6rUactLDJEHkJDXcj7/Nz8LXS8folBDGp4fh66Xj2nHJ6DQCV41J7LInkpwQXy/GpIaescaiV41JJMikJyXMhzmDuq5R7U0CTHrGpIZ6/AxVTh81UqSionJGqX7rbWr/+18C5s0l9IYbelwH0lOac9d77Otz8vEnMI0YgbmgQOp1IxgMmEaMoHnduo5zc/EdNw5wiKfKF1ziypCYKDUmlTcQbd2+3eM8apcskX4OvOACRRNUJ2GLFtF+OI+mH3+U9kX+5SGPaV0ab29Cb7yR0BtvdMzP3PFEWKNB0Hn+U26IjcEQG4P/tKkEX3011a++RvOmTQRffnmPIzn68HCCLruMmrfeQhsUhP+smR7H+YwaRfIXn9N+NB/fs3qn4eWp8Js4Ed8N6/n3hiLqlimFyHf7T7CtqIahHhbPXdFutXHr0l2cbHA8rf/3+iJGp7jXW3hia1GtVKfxZm4hQxOCpCflPeXj7cdpsziejr++roDnLjx1NK8zVU3trDxwkqGJQaSGK59Wt5itfLGrjJpmM83tVt5eX6SoU7mjskmRNiWns/Ock4KqZoJMBo54EBTrjlQp3r/NhdVuLnBPf3+YFrMNg07D+PQwMqI9N7wURVFRl3LblHT+9O+t0vwBrh+ffMpo1+yB0ezvEDnLdpcxf0isdEwuUAbEBEi1M/tK6/lwWwl/+WK/dL8vd5fx6Dn92VRQLX1mfSP9eHxeJmOfcvw+5x6t6iTOdFKzVGf0RS7EBsUHMmdQNI98eUASWoLQfarY7IHRvOFBWC6ckMrAuECPTW4FQeCFi7Kx2OxdLqZTOqXP7S6pY0N+NasOuoSkPM1Nr9Vw17S+LJ6U7rEJ7ZCEIHY9NJWecs6gGGZlRSsiPPfO6Med0/r8ZgVAdnwQW+6fjE7TdRNeld82v81vloqKyh8Sa00NFX//O5aSEqqef4Hye++VHM26onHVKk48+pgkSLqjc5Qo4LxzpdoWe0sLxxctpvJ5l0114IIFBF9+mbTd1JGe1rJtG0UXXiiJJQSB2JdeVNhUeyL0phs97g+cf57H/YJGQ/RTT0rGAsFX/wnTKe4hnWswOP51IYg6owsKIuKeu0n+/DMCz53Xo3OchN9xOwnvv0fyV8vQBXctMrxSUvCfNtVjI9RfCkEQFJ3k5QvAV9cWeDqlS55Zkaco/t7fQ+cxcHdFu++zvW6NOE+FPJVqxYGTp92TZtXBk0x5Zg33fLqX6f9cx79W5yt69Nz/2T7u/XQvT39/mJdX5ysEBcCy3V1bnndOnXOSX9FEaZ2rd4yczqYI8us714xmm51nV+bx1HeHOPdf66lo8PyeHShv4ETHMX+jjmGJQYxMdn0XYwK9mdkDESpPu8o9WqVIN5O//+cOdj3p/3J3GX/+eI/b+7WvtEFhOpCTHkZskIkhCY5Irs0u8sHWYun42ZkuY4eCqmaa2q2U1DjeV51GIDnUF5NBx1Pzs3j5ksHkpIfx2DmZJIW6ojadyYoNkGpnnMQHm6QUuc6CSE534iI2yISh43hFYzvnvLSep747pLD37lzn47xfbwkCTylvv1VB5ESv7b3Xr3LmUSNFKioqZ4zmdeukuhuA+i++pL2wCJ/Ro/DOysI3J0exyK/77HPK770XgMaVK0n65OMuneQAGleskKJEgl5P2KJFWCsqOXbxxYgWC+0HD0pjBYOBkGuuQevvh6DXI1osmI/mc/KJJ6n5z38U1w2YOxevlBSiHnmYkutvwN7ait+UKfiOz8FSXo752DFMQ4biP20qTevX07Z7j3SuaeRIDPHxXc5Z6+9P4gdLsdbUKuqIfksIWm2PxVpPsdtF/r78MJWN7dw7ox/BPj+9Bqmo2lVTcsuEVP6xwtEvaeXBkxwoa2DVwZMcrWzitsnpJHZaYH6y/bijl4nNztaiGsWx0rpWaprNPZpb5345tS0W7vlkL29eMRRBEFh18CT/2XiM9k4L63B/I3dMccxLHqloMdv44VAF6RG+PP39YepbHQ8PIgOM3D+jH+H+Rk42tPHXrw9S0dCGxWaXLLHBUfT+1HeHWH+0in9dOpjqJrOiCN9JgLdeuvaXu0tZNCmVN3ML2V/WwB1T04kNciy4O9txOymoalakP2XG+HOovBGrXWR/WQMXvLqRMD8vrhydyLf7XKLoyXMH8NAX+6W6EYA2i50tRTWSa5uct3KLpJ/H9wlHp9UwMyuaHw873vebJqT0KE3KKVq2H6vFZhf5Zm85l45MoM1io6hDXAsCzBkYzWNfHcAuonCHEwRXG7F9ZfUKYwGnSFgwLE5KHZTXGp0zKIYPO5zg8iubFNG15DAfhYCZMSCKGT0QeYIgMDsrWuHYNntg1M9emGs1AgkhJoWpgpwgk95jLZKKyu8ZVRSpqKh0SfOGDVhra/GfNq3HEYnuaJJZKDtp27NHSmfzzs4m5u9Po4+JofGHHyl/4AFpnLWigtLbbif+7bc8ppfZm5s5+fgT0nbghReij4hAHxFBxP33c+LhhxXjAxcskESI99AhDgtrUAgiQa8n6NJLCbt1MeBoSJq6amW3rzFw/nxOyERR4Pnzux3vvM9vVRD9Uny5u4yXVztsjE0GLY+ck3mKMxw4C/vlFMsiRZP6RbD7eB0rD1YgijD3pfWYbY6Fd4vZxuuXD5XGFlU1c9fHu92K5OXsL6tnXFrXQhwcKWvOxbG8GPyHQxWsPVJF/2h/bl6yQ0qz6kyr2caT5w2gulOR/NKtxRyrbqG4pkWxv7yujfeuGcGN721XCCEncmvg3KNVPPj5Pny8dNJiPiPKn/F9wogN8mZ6/0jGPvUjrRYb+ZXNPPjFPt7b5IhuFFU38+mNoxEEQZE+Fx9skuZUUNlEgLfr93FIfBAmvY4tHQJzS0e/ma9l9tIxgd6cPySOflH+rDhwkg351ZKIkNt8S6+3vpUvZILuqjGJgKOYXyM4FvBzTuFGJmd2VpR0v2/3OUSR3Bo7PthEiK8XKWG+ClFw0fA4UsJ8+evXjocre0vrFTU2mR2pf+cPiaWkpoUXfzwqvee+XjpGJodI1tp1LRY2FbhEeOeeM6fD7IGdRVHP34vuSA7zUbz+cWmhZMYEoNdqODsz8oyZgaionCl+23FIFRWVX42WHTsp/tPVlN1xJ7UffPCzrydarVJ6GoD/7NmuHJoOWnfupGDuPArOmcvxW25xc1hr2bZNakzamcqXX8Z6wtEcUBsSQpisKWfgggsImOdKGXNGiZz4jh3rdj3voUNI+e5bIu7+82mlgwXMmIEmwPHEWBsait+UKT0+93+JlbLaBIeA6UaZAIVVzeT87Ucm/H21oimjzS5SUusSDfEhJq7LcTWjdQoigE351Yqn/st2l7kJojA/Lyb3cwnUzo03PbH+qKvWZXB8EBcNd0UGX1ubzzsbiroURACbCqo5WO6eqrfuSJWbIALYUlTD/Fc2eBREUzMi2HDvRG46y/UefL6rjKVbXWYAD8zqx93T+3LJiARCfL2YkhEhHXMKIoCdxXVSnZQ8UjRW1temoLJZEeFKj/TjitGJXb5WcKSwaTQCWbGB3DG1D+cNdtX1yG2+nby9vkgSecMTgxkc70hP02gEzh0cyzmDYk4rMjI90xWB2VpYS4vZqnwNHQJlwbA4wJGu99LFg3ni3CyGyWrUco9USULW10tHYogjCikIAndM7cP714yQGnjOHxKLViMoUuG+k0XOOjf/PB36RPoxoY9DuI9PD/tZ15Jz7uBYBMGREvfInP6886fh3D29L7d3YQihovJ7R5X5KioqHmnOdbmuNa1dS/AlSkc0y4kTWEpKMGZl9Ug0tO7Zg73escDURUQQ/benCFt0C627dtG6c5dDeNls2BsbaT98WDpPHxeH3+TJ1Lz9NuCI5ATOPw+vtDQaf/iB2vfex97erjBXiPjzXWgDXPnugiAQ+ZeHsNZU07x+A+F33aWIzPiMGQMy+2pjRgZxr7yC1vf0m09qfHyIe+VfNHz9DQHnnNPr1tS/Byw2O3tL6+kb6YfJ4P7fjM0uSrbM4EhTK6hq7rbZ55PfHpQEwrsbj3HP2X0BRxTB2aMm1NcLXy9HvUl2fKDCIhigsd2x+HUu6JbtKZOO3Tk1nWGJwfSN8uebveWsPOjobdW58aYn1shqZ3LSw5iXHcOH20qw2UXWH61WzOOes/syMNaRdrRo6U4qG9tpareybHeZ23Xl3Ht2X040tPH2+iIA9hx3ibVrxiYxqV8EYX5eUkPLP0/vy8mGdj7Z4UjXcorBATEBjOqwl3Yye2A0X3Zx/1fX5DM8KVhqyAkwJiWUJZsd4ulYTYvi2UbfSD+GJASTFTuB47WtWGx2/rkyTyHgOqfHJSvsn5WRovpWi3QvwK3nzU8hMsBIeoQveSebMNvsbC6sYXNhtXTcKSquGZfM5H4RRPgb8TY4rMX7RPpJkThn2iHgsIzuVAMzOiWUtX+eQHFNi2RxnRLuK1lT75Z9hj+3r83rlw+loKqZ5FCfXqtpmdY/kty7J+Jn1Cmsy1VU/qiookhFRcUj7UeOuH7OO6I4Zq2upmDGTOwtLWj8/fGfNg19fByIYCkvo/3QYSxlZY7ke60WvwkTEPSuPze+OeMQBAFDXByGuDgCZs8mYM5sSu+4E0upK03GOzub6KeeRB8bS/vhwzRv2ABAw7ffERweTuntdyh6DQGYhg7Ff84ct9ejMRqJe+UVEEU3xzuvPn0wDhhA2969GBITiXv9tZ8kiKQ5ZGdjyj5957A/AnknG7llyU4On2wkLdyXbxaPcyuO3lda72bvvC6vsktRVFDZxHJZI8vtx1xpR3KTBWfBuSAIPHZOJguX7CDcz4jV7qq32Xasln5R/hw60SD1FPHWa/nT2CRJwMmbW54qUtTZFW1cWihxwSZmDoiShEZLhwlBfLCJa8YmSbUvwxKD+GavI7opNyGICfRWRMNmDIjkupxkrHaR7cdqFYIoOz6Qe87u67Ge5i9zMthUUK24lieHtpz0UIUzmk4jYBNFRBFWHargyMlGymRGCynhPkQHGCmrb8NmFxV9gtI6BEVcsEmyTR6dEsILPxzlw20lnJ0ZRWaMMsqgdDprUqRIvpVbSFO7Y16p4b5M7KUmlTlpYdLn/8PBCr7dd0I6dlYfV7pk5xo0o15LWoSfW2Svc0NUJwadRhKqAMldmCb83OiOTqv5WSl4XRET6H3qQSoqfxBUUaSiouIRuRCylpdjq6+Xoi8tW7dib3EsRu0NDdR99FG315LbUwP45OS4jfEeNIikL76gOTcXXVgoXunpaP1c/8kHXnCBJIoaVyxHFx7uJog0fn5EPvJwl09KBUFwS9lz7o9/43Va9+zBNGwYGuNPb755JimpaeHj7cc5q08Y2fFBpxx/sLyBb/eWM2tgtNsC6mhFE1/uKmV8n3DJPet0+XBrCQ9+sU8qnj9S0cTe0nop3clJZ1cygLVHqrhyTJLH676+rlDRtHP38XqpMahCFMl6l2TGBLD6rgkAvLGuQBJFO47VctnIBEVkZlK/cEVEKz3SF71WwGITKa5pob7FQoDJ85PyQycaJZvpQJOerI4o0HU5yW7Rl2vHJSnEy+B4lyiSO5vdMjGVh77Yj9lmJ8zPi7/OHYAgCOi1As9c4Gps6a3X8swFg7o0GPA36nn6/Cwufn2z4/0JMTG9f6TbOC+dltkDo6WIzK2T09hzvF4Soq+uLVCkz0UFeJMRHUBZvfL3LzrA6DGioNNquG1KOrdNSfc4z1BfgyTKms02KhrbifA3sq+0npdktTLX5SS7RWN+KuPSwyQr66Vbi6VoY3SA0e372pkBMf5uomhAbM/SybLj3c0Jwvy8etR3R0VF5ZdFFUUqKipu2NvaMBcXK/a1HzmCaaijSF0ezTlt9Hp8Ro3yeEjr64P/9Gkej/nmjEMwGhHb2mg/cpTqN96QjoVcew2+48djSElBF/TTFvTagACpR9HvAVEUueG97ewva+Ct9YXk3j1RUfDemXarjcvf2kJlYzsfbCthzV0TMOq1iKLIe5uL+etXB2i32nl9XSFr7jqLcP/TE4Y7imv58yd73PZvL6p1W2TKIytONuZXS0JHTkVjm5QC5sRstbOvtIEhCUEcq3FFKeJDPC8s5b2Kth2rQRRFRWSmc2G6l05LeoSfZMm9v7y+y35FH8hqdcakhko2wpkxAYxLC5Vea7CPgflD4rqcl5yJfcMJ9fXix8MV/GlsksL9LjXcj3f+NJyPtx/ngmFx3do1gyOF67XLhrA6r5I/jUnqUkDdN6MfPgYtEf5GrhqTxK6SWkkUfbzd9f77eunwN+p4YGY/jlU3Kwrxf2oKmCAIJIf5srvEIVzzO8wb7vhwt1RLlB0fyLnZvdcQc0RSMAadBrPVLgkigFkDo08pvDJjAiQXOWlfF5GizoxPD+P+Gf2kdD1vg47LRyV4tJ9WUVE5s6iiSEXlV8Le2krTmjV49emDV5LnJ+QTdvmNAAAgAElEQVSnwlxSQuvOnZiGDEEf89MXDKIoUvPmmzStWUvY4kUI3t7QqfC9LS/PoygKmH8e+vAIRLPjabk2MBCvPn0wJCUh6PU0fPstFU8+JY03DRny02p1TCZ8x42lcYXD/c1yvGNRotMRfNVV3fbP+SNQUNnEX77cT2yQN4/MyeRoRZO0aG9ss7LhaFW3zULXHK6UIhrOWpMLhsZx69JdfL3XJRBaLTbe3lDE3dP7ul2j3Wrjnk/2UtXUzt/mZxEV4Eqt+VeHkxw43OScKWPbj9VyrewajW0WdhS7ep2E+npR1dROq8XG9mO1jE4JRRRF/u/rg6w6VEFjmxWz1d2kYPuxGoYkBCmc55yF7p3pH+2PUa+hzWKnpKaVlQcrpPokP6NOkS7lJDM6QHp/95V6FkUb86v5z8YiabtzFGbxpDQ2FVRjsYksnJAq1aV4mpeTQJPeYfaQEcFkmQGCnBHJIYzoVBfUHVP7RzLVQ4RIjq+XjvtnZkjbQxKCGZ4ULLnHOYkKMCIIAomhPny5cCyPfX1AijBN6ud5vj0hJdRHJoqaWXO4UmpseqqI2E/BqNcyIinYTaD3xMWufycB5K3Xdtn0tjOCIHBtTjLX9kJtlIqKSu+iiiIVlV+JiqefpnbJf0GnI/qJxwmYPfu0zm8/coSiBRdKaWymYcMI/tNV+E2Y4HG8ramZ0ltvpe3gQXxGjyZgzhx8Ro9C0GppWrNGcnUru+9+Qm90b0LafjhP+tksE0W+OTn4T+26U3nIlVeiCw2j/L77EM1mgi+9pMuxp8Jv6lRJFEn3HzfudyeIGtos+HnpelwQ3Waxce0726TajZQwX6qalPbNa49UuomiNosNnUZAp9WwbI+yMefraws42dCuEERO3tt0jJsnpLpZ7n637wSf7XR89k9+e4jnLnTUTR2taGKFrObn6fkDuXnJDsBRwyOvEdmYXy09/c+M8WdQXKDkeLY2r4rRKaFsLKiWUpvkyCMv24pquS4HqbcMdB0p0ms1ZMUGSgv8u2URrWn9I92iUwCZsQF8sM0RBdrnwWyhsc3CnR/tlp4d5KSHKRqDgiMS9PnNY6hvsTA61V1U6bUaBsYGsrlQac38W2n++Pi8Acx8fp2il1CUrMbE26Dl8XkDuGhYPDUtZsZ5eI09JUVWd7OzuFaR3njfjL6njIj9FHLSwhSiKCnUh/7Rp06Dy4jyV1ivZ0T7q5EeFZU/AKolt4rKr4AoijR8861jw2ql7K4/U/3mm9iaPDfKAxBtNlr37sVWV4etsZHjC2+RBBE46nyO33gTdZ997vH8iqeepDk3F1t1NQ3LllFy7bUcu/wKbA0NVPzd5bxmKS6m8fvv3c5vz3OJInmkqCcRqoBZM0ld/SMpK1fiN3nyKcd3he/48dCpR1HAOe6mCr9VaprN3PDudrIeXs6cF9dztKLx1CcBT313SFHM/mZuoZtb2dq8KoWt9YdbS+j74HfMfXk95fWtrJSJFnCIiedXuerGLhruSsVqbLOydIsyfRIc9TNOlu8/SYvZUQD/xroCaf/kfuGcnRmJn9EhqKqa2qWoTKvZxhLZdcelhSl6AK3raIAq79/iZExqCPfN6Cdtb+8QW8Wyxq0J3dRlDJXVSdV02ChrBLhsZILH8ZmyxfG+TmYLtc1mFi7ZKRkYBHjreXp+lkcx0z86wKMgctK5fqu37JR7g9RwX+49WxkxDPSQojkgNoDx6WE/q95HbkDw5a4yKaUtM8afS7v4jH4u49KVn8vsrJ41PfU2aBXmCZk9EFIqKiq/fVRRpPKHwNbURMPy5Vir3GsVfouYC4uw1Sntgiue/jt5Q4dxdOo06pctUxyz1dVx7LLLKTr/AvLG5VB43nzMx44BjsafyNzUyh94gMYff1Sc37h6NXUffew2j9bt2ymYcw7mo/mK/U2rV7uNbc/LQxRFRFHEUupakBt6mLanCw7GEPvzagK0/v74jBopbWv8/PDtIjLWm7SYrXy1p4wTnQrLT4eN+dWc/dxavtvvKKzfW1rPrBdy+efKPN7dWMTnO0slly05649WSTbMTsrr2xSOYuCytQZoarfy2NcHAEeU44JXNyoK+TszLDGIv84dwLXjXCk9b+YWuqWtyVPVWi02fjhUQUVDG5/ucInk63JS0GgERR3RtqJaDp1oYPaLuaw+7DJZOCs9jNEpIdJT9v1lDVQ2tivc5R6enUHu3RN47+oR9Inww79DbFU3m9lRXEtzR5qer5dOUXvTmaGJ7rVm149PYWCce+E7QL8of/Rax7wKqprZc9zx+7q5oJqzn1unsOH+69xMIk6zBquref1ca+be5vJRiQyIcaWLdXaO6y3k6WdWWfOouafZg+h06BPhR1SA63M7naan8u93T0xOVFRUfvuo6XMqfwjK7ryLptWr0cfHk/L1Vw6h0EuIosjJ/3uchu++Qx8RgVffPgTMOQefEcN/8jVbd+7s8piluJjy+x/ANHwE+ohwLCdOUHzNNS7hYrFgkZkgRD/1JN5DhlJy/fW0HzoENhulixbjPWgQXmmpCF5GhcjyHT8efUy0I3UPpIanXaLRgN2OvbkZS2kZGpM3YqtjQa7x8ZEalZ4p/GfMoHmto4eS/6yZp9VY9adgsdm5+PXN7CqpI9TXwNeLxp32Aji/sokr3t7iJjLaLHb+udIVrZncL5w3rhgmbS/bXcZ9n+2VtuW2yZ5w2lov3VKsGFdS4xJQl41M4IOtJVJTU5NByz/OH4RWI3Du4BieWXGYqiYz5fVtLN1azOWjEqVz5aYGzvltLayRrpUdH8iwjkX+0IQgSTgs21PGI8v20yCb04XD4hieFIwgCAyOD5SahK7Jq2SXrKfNlP6Rki2wIMDghCBJWH0iE2MJIaZuF8+dzR76Rflz22TPbmjgqDmZOSCKz3c5HgC8uraA63OSuezNLYqGsDeelXJai+lTzeu3FCkCR4PUN68cyn2f7kMQ4PxOZhG9hePzU5YyCoJ7T6PeRBAEnjh3AM+vOsL0zEjJTrwn3DwhldK6ViL8jW5pkyoqKr9PVFGk8rvH1tBA05o1gENQtBcWYkzverFzurTt20/te+857lVVRdv+/dR/uYzU5d+jj/pp/xm27Nwh/Rx6042IZjNNa9bSXlgIViui2Uz1G28Qet21HLv4EkfPHw8EX3EF/jNmABD/+msUXXwJlpISRIuFlq1badm6VTFeGxZK1JNPoAsKQhceTuU/n3MdCw7GVqNMWxL0eowDBtC6wzHf9rw8dGGulBN9TO8+xd1XWs87G4sYkxrKnIHRHq8dMGcOluJirNU1hN92W6/de8WBk6w4cIIZA6I4q4+rF8oLq46wq6MAvKrJzN2f7OHtK4dJcyupcaSh1Xb03YkL9ub2Ken4yayJX12TLwmiYB8Dt09J552NRVKfFCcrD1aw93g9/aL8ePCLffx3i8vZLMik58PrRzHnxfWKqM+0/hF8v9+RGrf2SBWXjEzgTQ/1OE6uGpMIwLubHJHGB2dlSLU4Rr2Wq8Yk8fT3jua5T3xziJy0MBJDfRBFkWNVLYprrTpYoXiqf31OivS+DJFFQOTRIW+9lkfm9Of8obHS2Jy0MEkUvbGuQIr+RAUY3fqkDJWJoi92KkVRdwSaDJJxgEGr4ZkLBmLQdZ8scV1OiiSKvt1bzp7jdZIgCjLpeXr+wC7NEHpKoMnAsMQgthbV4m/UkfEbTMUK9zPyxhVDf9F7GPVaYoO8FQJ+eGIwkQG/rD3+WX3CFb/vPSUu2MS7V4/4BWakoqLya6GKIpXfPa07dyoeL5oLi3pVFDVv3Oi+02KhedNmAufN/UnXbN25S/rZZ9QoTMOGEX7nnTStWUPJ9TcAUPfBB7Tu3u0SRHo90U88gTGjH40rV6H19yfw/PnSdXRhYcS/+QbHF99K+8GD7jcVBKIee0yyrA65/nqsNTXUvvMuABH33E312/9WnGtITsaYkSETRYcllznoWT1RT7HbRW5esoNj1S18uO043+8/wRPzstz6wwgaDWGLFvXafVvMVh7+cr9ksfvhtuNcPiqB+2b042B5Ay+tVqYWrj5cyZItxVwyIgFRFFm0dCc7i5WpkHqtRqp/OdnQJpkTALx62RCGJQYzf0gsS7cUc7SyiV0ldVIx/6tr80kO9VEIorhgb16+eAhpEX5cODxOSqcL9TVw9/S+kijamF/NJ9uPU96R5hfqa8BLp5VS7TJj/EkO8+WBWf1ICDERFeDNzE5Pua8em8TnO0s5UtFEq8XGHR/t5sPrR1HfaqGxU3qfXBBN7BvOtP4ugTAoLhCtRsAmG6MR4N2rh7tZUY9LD+MfKxw1a/K6JU/9kgbL9jnFE0B88KkL8Z9dMIjPd5ZyVp8w+kX1oKA+2l8yd7CLroibyaDl05vG9Frx/7MLBvHJ9lJy0kPdzC3+l0gJ81WIop8TgVNRUVE5Xf53//qq/G5oXLUKW10dAXPnImjdXaJatm1TbJsL3Z+Si1YraLXSk2nzsWM0/vAj/mdPRx/ZvVVtyyaXKDIkJUnXb921q0tR1LJ9O+1H8wmcNxfBoKxzsNXVYc7vWGjrdBgzM6VjPjk5GAcMoG3vXkSzmbY9HS5ZGg2xLzyP31lnAeB1nWc7V0N8PEmffoK1vJy2Q4cxFx+Djifb3oMGYhoyRBorCAIR996L75gxCHo9PqNH055fIImiowExvNb3AposftzjG058UwXteXmK1MTeFEU7S2oVjTi/2XuC3SX1LL1uZK82NhRFkX9vKOK1tQVUN5ux20XF4h7gnY3HWLK5GLsoSg5TPgattAj/61cHGZMSysmGNjdBBI6ok1MUvbW+UCoaH5IQxLAOQWDUa6Vmpfs66oscr7tcESGblRXF4+cOkJpiXjsumc92llLXYuGKUYkkh/mSEGLiWHULrRYb98rS7a4cnciwxGCufHsrrRYbN4xPARx9eK4Z5/k7ZNRreXbBIOa+tB6rXWT7sVreWFfA8KSuHf6CTHqePG+AYt4mg47+0f7sOe4yKbjprFSPvXkGxAQQaNJT1xFtc+JJFA1JCCIlzEdhPAGQeIpIEUBMoDc3T0g95Tg51+ekuNk2Pzgro1fd0GKDTCyenNZr1/u9khzqK0UBtRqBszO7/9usoqKi0puookjlN03zpk0cv3khANbqGkKvu9ZtTMu27YrtzqKo/uuvKX/gQbyzsoh98QXsLa0ULbgQW10d1W+9SdInn6AP95w+YW9vp2W7K9Ut5LrrKL/3XgBad+/2eE7rnj0cu+JKsFpp2bKFmH/8XXG8RVZPZMzIQOPtSg8SBIGwhTdL0SInYbfeKgkiT4iiSH5lk8ym2QsSsxz/5BRUE+7nJRU1C4JA/cDhaAUBH8B3fA6Vr77GFyljeav/TKyCDtrhnYzpPLDlHdry8tAGuhaqpxJFjnk1ExlgPOUTcHkzTSelda3c9sEuPrh+lEfLW7PVTkFVE2nhfm7HbXaR/WX1Ur8cALso8vraAn6UpXPJSQ7zoaBjsS0XSj4GLV8sHMuN721XRFCcDmvgSGNbk1dJm8VOYVUzJTUtBJr0LNnkqv+6voveJJkxAYxNDSX3qCMi4Yx8Dk8K5rkLsxWvLTrQm+W35VBe1yYVwI9LC+VYteM+zqCpyaDl0pEJBJoMrLpjPC1mm8IxqzsyYwJYPClNit78e0ORoo5qQEwAe2WObP83bwDhfu5pTkMTgiVR1D/an0WTPC/8tRqBsamhfNXJNnxogruA8tJpee+aEdy6dJfCyvqXsGwGh+td/2h/qWfRxL7hXDjsl6mr+V8nLcL1/RybGkqI7y9bL6iioqIiRxVFKr9pmnNzpZ/rli4l5JqrEWROa/a2Nlr37VOc016kFEVVL7yI2NpKy+bNlN1zL6LFLDm/2SqrKL3tdhL+/bYUAbHW1tK2bz/eAzJpO5yH2O5IFzMkJOA3ZQrl990Hokh7Xh725mY0PsrFWOXzL4DVkWbU8PXXBF92Kd6DBknH5alzpuxBdEYeLQLwmzKZkGuv6fZ9enl1vlQH0hPOzozk0XMyeSO3gNfWFqDXaHjrymGMycrijSHn8VncSMX4LRH9aNZ5QX4BGpPr9epjuk5vqWpq586PdrP6cCUxgd58vWgsgSbP7mA2u6hYEF+Xk8xbuYVY7SLbjtXy2toCbjwrRXGO3S5y+Vub2VRQQ58IP56/KJs+Hc5doihyy3938M3eU5hIdBDh78Wfp/Xl3MExLNlSzLMrjlDV5PjcfQxa/n7+QFLDfd0iKE4EAf48vS9tFrtkLrD2SCXN7VYp5Sw5zIfJ3TS3vH58MrlHXREJH4OWf5w/0KMYDPczKkTI9TkprM2rkqyvvfVa7pvRT3q/ozvV5fSEG85K4ZU1+TSbbZTXt7Exv1o6NjolhAl9wvj3hiKuHJPEjC6axl6Xk8zaI5XoNALPX5TdbQ1PTlqY4jtgMmjpF+W58D0qwJsl147kX6uP8tb6IgbHB0oRuN5GEAT+b94AbnxvOxH+RreImErvMSsrivc3H6OysZ07p/b5taejoqLyP4YqilR+07QdcNW3WMrKaN2+HdMwlztX6549YFGm3JgLi6RmkdbKSsxFRdKxplWr3O7Run07x2+7DWPffrTt30/TunVgtaKPicF3fI40zjRqJFpfH7zS0hw9e+x2Wvfuw2ekq9i2ddcuhZADOPm3p0l4/z1pIeWszwHwzh7sNh9BEIh+4nHK7rsffXQ0UX99rNtFmDMd7HT4dt8JVhw4KUVDzDY7t3+4i/tn9nMTRAAWrZ5NUf2ZVLLDldKHMlJkttr5cncZBZVN2EX4ePtxSViU1rXyzsZjLJqUxtGKJpbtLsPSkdY3OD4Ib4NWGhvq68Xd0/viY9Dx7EpHpOKZFYfd6kBW51VI/WwOn2xkzou5PDgrg0tGxLN0a8kpBdG145JYPDkdnUbAoNVIPVYuGZHARcPipYJ6vVYjCZPOERQnU/pFkBLmS056mCSKfjhYwb4yVzTlunHJ3fZxGZsaqohI/GV2/x6nDcYFm1hz11lSk01nw9afg16rITs+SBJq8iav8SEmLhmRwO2nWLhGBhhZefv4Ht2vc8+YQXGB3b4GrUZg4cQ0bp6Q+ouLlEFxgWy8d9Iveg8V8DPqWbZwLIAqPFVUVM44qihS6VUsJyvQBQf1iiW2KIq0HTqk2Ff3xRdKUbR9e+fTsDc0YKupQRcSokh964xXWirtR44C0LRyFU0rlYLJUloq2VYD+IwcBYD3wIFSI9PW3bsVoqjypZfd7tO6YwdNq1bhN3ky1tpaRWTLOzvb89xSU0n68IMu5y7n0IlGKhsdgsKo15AV67nvCjiEi9NJrXMdTUVjO4uXuqJYE1ICGZQULgmTNTHZTCpRvp/OHkVFVc0sWrpTUT/SmX9vKGLuoBjOf2WD5NTmJFBmpjArKwqtRuDmCSn8cLiC3SV1WGwi9322l09vHC0tll5dU6C4RrvVzgOf7+OHQxVsKnBFNdLCfQmS9a8J8NZz+agERdPQzmg0AkaNe/0aOCyYVx5yzMvJ9R21OuPTQ3msY9+qQxXS8VBfL+Zmd59qKAgCL18ymGdX5DEwLpDzh8Z2O97T+Ua95zn/VIYkuESRvIdSYkjvp6pFBXiTFu7LkYom6d49QV08/7FQP08VFZVfC1UUqfQaNf/5DyefeBJDSgpJH32IxvTziuOtFRVuFtGN336H/YEH0BgdqUMtW7d5OhVzYWGHKJKJJr1eiioZUlJI/Ogjyv58N43Ll596MoKAqaMvkfegQdR99BGgrCtq3bWL5nWO/jloNPhOmCBFpsoffgRrZSU1776H2OZwBtPHxaGPOH0r2M6slTWRnNwvghcvdo8+yfl+/wnu/mQPdS0W9FqBedkxkvOak1BfA09fNJRWs00SRTvC02nQm/C3OFK0nD2K1h2p5IZ3tyucwFzX8UIQoLKxnZpmM+f+a72bIAIUBfZOxyldh23y2c+tw2y1s7O4ji2FNYxIDmFXSZ1UT6LTCCSH+Uj21j/IxEhKmA/Lbhnbq2JBp9Xw7AUDmf1CLs1mG+PSQqUFfEqYL1EBRsn9zclVYxJ7NIeEEB/+eaFnofxr4KnhKUB8LxpfyJmbHcPT3x9GEGBaf7XIXkVFRUXlzKGKIpVeo7ZDKJjz82lYvpzAuZ6d2Vr37sNcVIjf1KndNt5s82ArbW9upvqNN9EGBmJvbaFllyuy4Z2dLTVFbS8sxDR0KC3bXaIp6rFHaVj2Fba6OqKffAKN0Uj035+m7qPh2KodkQWNrx9+kyZS9frr1H/8iXSuV7++kpW196CBrteyaxeiKGKtqKD09juk/f6zZhJx773kb9mCvbERW1UVJx55VDpu1WiJue1WaVsURWx2sccpTxabHa0goNEIrD3iEkU53UQ/nEzrH0l2fCA/HKxgWFIwKR2mC3Jh9Pi8AYR2FDlnxweys7gOm0bLD3GDmXB8B/7mFil17u6P90iCSK8VuHxUIsE+Bvy99Y7mlztLefSrAwAyIwhHxKWioZ1PdrjuGxPozeB4V6QrJcyX8wbHSPbUr64tYERyCK+tddlkzxkYzePnDuDJbw8p0gi1GoFnFwzq9egJQHKYL18tGse2ohqmyxyyBEEgJy2MD7a57LR9DFouHZHQ63M4EwyKC0QjgDyoqNcKP6lGqSdcl5NMbJA3UQHeZMac2abAKioqKir/26iiSKVXEG02LMdcLlsN33zjURSZi4s5dvHFiBYLwYcOEXHXXV1eU94vR/D2Rmx19K+oevFFt7G6yEh8xoyRRJG5sAhbUxPthzrMBzQa/CZPdpuTxmAg+JJL3K4Xcc89tGzchKXU0V/GmToHDltujZ+fQ+zU1NC2bx/l990n9RMSvL0Ju+kmdEFBxDzzDGV33y1FvOwIPDbyKrZHZbDAHMf9ZhubCqt54LN9NLVbuXNqOpeOTOgyhUQURT7dUcpjXx9Ar9Xw+uVD2VroKvjvXJfRFeF+Ri4cHi9tPzgrg32lDRwob+DqsUlMlT2ln50VLdlOv5o1l1ez5hLfcIJ/sYdDJxop64iK+Bl1/PfakW6L2QXD4nhu1RHqW13RoKvHJnH39L4ATM+M5J5P9lDdbOamCSlur/2accks3VqCKDqiQG/lFvLtPle90LU5yRj1Wh6e05+xqaH8+ZM91DSbuWd6325TCX8uSaE+Hh3PctKVoujC4fFuvZZ+L/gZ9fSJ9OdgeYO0Ly7I5NH8oTfQazWcM6j3bN5VVFRUVFR6iiqKVHoFS3k5oszwoHnDRqy1tVJ0xUnTmrXSuIZlXxF+552KRbCtoQHL8eN49etH20FXPVHIVVdR9bJ7vY4Tv8mT8UpOkrbNhYUOgWR3FJ579e2D1rdndsQAWl9fYp75B8cX3gIaDcGXXSodEzQavLOyaF6/HoCi8y9wnajTEfv8cxgSEwHwHTeWlOXfU/PWW9T85x0ORqSxKTIDRHhvUzErD1RwosGVavXgF/tZe6SKv52XpaiDAWhss/DA5/v4YleZtO+yNzdLhgDpEb5EBfy0J/h+Rj2f3Tya6iazWxRgZlYUf/36gCJaUOwfyTcGO/6y1L3x6WEen+77eOm4bGQCL/7oqN9KDfflrmmuAv0pGRHkpE+krsWisH12khLmy5R+ESw/4GhQ6ow6Oe8pN1+YnBHBhrSJVDebifmFohmnYmxqqNS0VKcR+NPYpFOf9BtmaEKQQhTF96AfkIqKioqKyu8NVRSp9ApuDVOtVhpXrCDoggsUu+U1ONaKCtqPHMGYng6AvbWVoosuxpyfT+CCBYr0Ob/JkxCMRmrffx9dSAheffqgC3FY8OrCIwg491wsJa5IlbmwUNG/yDRk6Gm/Ju+BA0n9YRXodG7RC++BAyVRJCf6iSfwHTdOsU/r60vYokWE3nwzy5YfhjWu90ouiJysOHCSi2s38/nNo/HSOVK/dhbXsmjpTkW3d4DGNlfxe3fGAT3BS6f1mBYV4W/k/pkZvLG2gPqaOlp0DuGyUggjTNbUsrvUvRvOSiHvZCM1zWaePG+AW0qbl05LhH/XaW7Xj0+RRJGTMD8vHpqd4TbWqNf+aoIIIMCk54GZ/Xhv0zGuHJP0q86lNxiaGMS7m45J2wm/UD2RioqKiorKr0mPRJEgCIIoiuKpR6r8r2IuLHLb1/DNt+6iSFYDBNC8foMkiprWrsOc76gVqftA5rym02FITSU0I8Nj81YnQoKrbsN8/DgtmzZJ26YhQ3r8Wo5WNLJ0Swlj0kKZ0MezEULA3HOofvttR0qfIGBITsbruhv5ly6Fig8crzHM34vLRiYQG+RYRApaLevyazxeb2LfcOKCvPnPRsfi82B5A8+syOPuaX15ZW0+zyzPU7jFOet85OSk/zxR1B1Xj03i6rFJHHngEWa0D8Si1XO4Xc9Rmctbd6l7vl46Xrv89IWpkyEJQQxPCmZLh7nCxL7hPD0/6zfb3PGqMUlcNeb3HSFy0tkFLuEXcJ5TUVFRUVH5telppOiYIAivA2+Kolh2ytEq/3PIewE5adm8GUtFBfpwh7CwVlVhOa50OWtev56Qq64EoOmHHzxe2ys1FY3Bc9NPORqTCV1UFNbycrBaFVEp09Cei6KFS3Zy6EQjb+QWcuGwOB6anYHJoPxVMcTHk7b6RyxlZRgSEtCYTNz+4S4+3ZivGLdkczFPnDuAWVnR1Dab2VPqsKzWCPDJjaP5Zm85GdH+zB0UgyAIJIT4SOlhr60tYHNBjWShDeDnpePxcwcwc0AUV7y9hXUdkRqDTsPwX6h5pZzkOxcz+u9fs8bqqNWxdQi1tPCfnrrXU56/MJvX1xXQP9qfedkxqnXvGSIm0JtIf6MU1UxQ0+dUVFRUVP6A9LS73w/APUCRIAifCoIw9Reck8rvELkoEpyOcqJI43ffS/tbZU0/nbRs3Yq9rQ3RaqVpzRqP1zb27dvjeXglJbrtM+jvKIcAACAASURBVCQnowv1HMU4fKKRmc+v46b3t9NutVFU1cyhE43S8aVbS5j9Qi4HyhoQRZEPt5Uw9dk1DP3rCka8sIXzV1ZT0GTjeG2LotbHSWOblYVLdvLIsv3kHq3CGW8dFBdIdnwQ98/MYF52rLTAv3J0IuPSHHMVRRSCaHB8IN8sHsfsgdFoNAJPzx9ImJ/jvT5vcCzeht53WeuMNjCQCy6Y6Lb/l4xSOYkMMPLgrAzOHRyrCqIziCAIUo8lf6OOoWdAfKuoqKioqJxpehQpEkXxSkEQFgNXANcB3wmCUAi8BrwtimJFtxdQ+cMjF0WBCy6g9p13Aahftozgyy8DoHXXbrfzxPZ2WrZvR+Plha2uzu04gDGjX4/nYUhKpnnDRmlbGxhIxD13U9HQRrCPwc3y+r7P9rK/rIH9ZQ2MTS3F2mHMICe/spm5L61nSEIQG2XpYuCwmL75/Z0MTQySoiYD4wK5cFgcL68+KtUAvb2+iFUHXb8mXdX/aDQCf5ufxbRn19LQUS8kCLBwQiqLJ6Up5h8ZYGT5rTkcPNHA4PieNbrsDSb2Dcdk0NIi60vkFHIqf0zumJpOTlooqeG+BHj/Pp30VFRUVFRUuqOnkSJEUawXRfF5URQzgfHABuBhoFgQhKWCIJz1y0xR5beOva0NS3m5Y0OjIeSaaxyNUoG2vXtpLygAlPVE+uho6efm3PU0rnKlzvlNm4bGz0/aNvbruSgKmDcPjY8PgtFIyHXXkbJiOf8lluGPr2LsUz+y4ajLGGBbUQ3bj7nsrJftLmNtnuv4pI7FP4DZZncTRE4On2zk/c0uk4fbJqdx0fB4vl40jsn9IqT9xTUt0s/dRVaiArz5xwWD8NJpiAv2Zsk1I7ljah+PPYyCfAyMTgn9RXrxdIW3QcuUDNfrMug0jEgKOWP3Vznz6LUaRqeGEu7BHVBFRUVFReWPQI9FUSfWA58BuwADMAtYJQjCFkEQul3BCoIwXRCEw4IgHBUE4R4Px58VBGFXx788QRA8hw9UfjOYi4tx5oXpY2LQh4fjd9ZZ0vH6L75EtFpp3bcPgLzAWDadfxMWwbGQb/rxRxpXrZLGB547j6jHHkXj74/vpEl4n4ZJgndmf1LXrCF9y2bCb78Nu8mHl1c7rKBPNLRxyZub+dt3h7DY7LyypkBx7qbCanKPuiym75/Zj69uGUv/aH/FuIuGx7Pp3kk8MNP9q9430o/xHYLH36jnnxcOIi5YWWvjb9QxMLb7xpRTMiLY+/A01tw5gVEpvz3Bcc4gl6gdlRxyRlL3VFRUVFRUVFR+KU7LklsQhDjgWuBqIAJYAZwDfA1MAp4B/gMM7+J8LfASMAU4DmwVBOFLURSlxiOiKN4mG38LkH06c1Q588hT55z9eQLmnkPjihUA1H/5Jf7TpiK2tHDSFMRt4xdhP6Zh1sC53LzrE2U9ksmEaeRINF5e+E+f/pPmo/V1uWOtP1pFbYurf5Iowsur81l9uJIDst4rzmNtFkf6XGyQN0mhPgiCwKc3jeZfq/PZfqyWS0YkMD3T0dj06rFJrMmrlMwOAK7LSVbUu/h66fjH+YNY8NpGqZ5oTGqox6hPZwy6n/rM4pdnQp9wFk5I5UB5A/fN6HnNl4qKioqKiorKb5GeWnLPBq4HpgH1wNvAv0RRlD9qXyEIwu04BFJXDAeOOs8TBGEpDlF1oIvxFwF/6ckcVX49zEWuHiaupqXj0AYGYqurw1peTuVLjsare0OSsQuOxf53iSNYcGgFRpuZN/vPotwnFG1wEF7/2QmAViMwNi2Uq8cmoZeJiCMnG3npx6PEB5tY1KnOpjNf7naZH/gbdVKdjlwQ+XnpaGy3Ks7LSQ+TxI2XTsutk9Pdri0IDrOD6c+tpa7FQmKIidkDo93GDU8K5vqcFF5Z43Cmk0dZfq8IgsCdsgasKioqKioqKiq/Z3oaKfoC2ApcAywVRbG9i3H5wPvdXCcGKJFtHwdGeBooCEICkITD+U7lN4wyUuToFSQYDPjPmEHtkiUANHWkx5X7uFLBrGhYdc1DHNl7lPVBqa4Lymp3co9W8e2+E/xzwSBiAr35ePtxHv1qvxTR0Ws13DIpTTEfZ0utdqud5ftdDT/fuXoEmwuqefr7w4qeP/+4YCA3vb9DsS+nh8YBkQFGli0cy/f7T3D2gCiFeJNz9/Q+9Ivyw6DVMD0zqkfXVlFRUVFRUVFROTP0VBQNFUVxx6kGdUSArupmiCcf3a6awl4IfCyKos3TQUEQrsPhhEd8fPyppqbyMxDtdsofeJD2Q4eIfOhBvAcNUhz3lD4HjhQ6pyhyUu6nNBh4t0zAJhdEHthdUseEv6/2eOy5VUc4q084AzpqdL7bd4L/++YAeq2GiX3CaeqIACWEmBgYG8CguEBGJoewaOlOjlW3MLlfOFMyIhiXFsqPhx31RFqNwKiUnrupxQWbuGZccrdjBEHgnEExPb6mioqKioqKiorKmaOnoqhEEIR0URTzOh8QBCEdqBFFscrDeZ05DsTJtmOBrprBXgjc3NWFRFF8DYclOEOHDu1KWKn0Ao0rV1L/6acAlCy8heRlX6ILcllA15SUUxoQQ0p9KV4yUWQcMAD9lGns3XaQvpoWgmfNoMZrOFS0SWNssujMuYNjmD84VtreWVLHP1fmYbG5f7xajYDNLmK1i9z24S7umd6XHw9XKFzgCioLpZ9nZ0VL6XAD4wJZeft4jlY0kRruiyAIzB4YLYmi7LhA1XZYRUVFRUVFReV/iJ6KopeBGhx1RZ25DQgBLujBdbYCaYIgJAGlOITPxZ0HCYLQBwgCNnY+pnLmaV6XK/1sq6rixF8eJvTGG6h67TVqdu5lcdallPhFMPn4Dl6LjJTGiiLc3e98tvvUMalPGG9eNZySx1Z4vEdiiIm/zs3EZHB9JUenhjIuLZS/fLmfvcfrATAZtMwfEseCYXHMe3k9LWYbRyuauOadbd2+hs61Pnqthn5RLle5WVnRfLvvBAfLG7hLrZVRUVFRUVFRUfmfoqeiaCxdR22WAy/25CKiKFoFQVgIfA9ogbdEUdwvCMKjwDZRFL/sGHoRjtolNQL0KyOKIs3r1yv2NS5fTuPy5QDkxmZT4ufoWbMydjCf7SrnvCGOaM/Oklq2Fzsc1VcdrqSsrpWaZjPgcFYLNhk40dCGRoBnFgxSCCInWbGBfHbTGI9zu39mP+7/bJ/b/rMzI2kx21iT54j8ZET50yfSz22cHINOw+uXD+12jIqKioqKioqKyh+TnoqiIByuc55owBEp6hGiKH4DfNNp30Odth/u6fVUfh6i1Yr15Em0YWFoDAa34+bCIixlXWU4wpoYZX3Rw1/uZ2RKCDGB3izbXa449s1e13ZckDcvXDSY/2woYnpmJIPjgzhdLh4eT6vZxsb8akTAoNVw9oBI5gyMRhTho+0lbCuq5dqc7ut9VFRUVFRUVFRU/rfpqShyusSt8nBsBFDuYb/KbxR7czM1775L44qV7Kls4/X0aQy01/LYK/e4CSN5lMg0aiSW0jIsxY66HfvEaewI7A921/jGdit3fribd64ezld7lF+Lr2WiKDHE5//bu+/4qKr0j+Ofk2QmjTQgCTUUpQhYQEQXBFEWRPkpFmRFUBEUUFnFtqJr3UXXsrLqulgwgCBlFdHFBoqCiCwqWKlLEaUTCBBCSD+/P2YymUkmMMAkk5Dv+/Xixdx7z73zzFxH7+M55zm0axTP0wPOOO7PYYzh5u4t/RY5MAb+cE4afzhHRThERERE5MgCTYpmAw8aY3601nrWITLG9APGAi9XRnByfHILioiMCPNZRBRcQ+H2v/02GS/+k6I9eyjC8Lfe97Mztj4rgWsXfs3pF3f3OackKSrGENGrN437XULW++8T1eF03i+qR8E7PwPQKCGKnVm5FFv476a93DHze/Zk+1Zu/949lA4grV5MJXxyEREREZFjV/Gql77+AvwEzDXGbDPGfGOM2QbMBX4GHq+sAOXYfPzzDjr99VN6jf+Cg7kFPscOvPseOx95lKI9rkKBSxudzs7Y0tLTmzdu9Wlv8/M59M035IdFcPuFd3PRmkQWbMuj7g03ENOpo8/wuGHnt2DUBaeUxrFy5xHjbF4v9rg/o4iIiIhIMAXUU2StzTHGXABcD/TGNYdoA64iC29aawsrL0QJ1G97c7j37R/JyS9iU8Yh5v64ncHnNvMcz/l6med1eGoq7/UcAjml5+/d5lro9NB//8uB997DFhZhc3JY0aA9mxMaQpFlytJf6NuhARkH81i60ZVcGeOq3lY31snCdRms2ZF11FjVUyQiIiIi1UWgw+ew1hYAk9x/pJopKrbc+/aPHMovXet2xeZ9PklR3voNntfb73+C1Yv3+1xjz+79FB8+zNYxd1F8oLSuxs7Yup7XGzMOATBv5Q5Klhg6p3ldGiREAfD8H87isn8uIb/INdGocWI0sZHh/G9Xts97NaurpEhEREREqodAh89JNff6l5v4ZnOmz77lv+7zvLbFxeRt2uTZnuw7Ug6AzP2HOPzzzz4JEcCumNKkKONgHlm5BT7zg/q2L12bqE2DOO69uLVne2DnpnRonOBzvTADTZKUFImIiIhI9RBwT5Ex5mJgFNAGiCpz2FprTyl/llSF/MJi/vn5hnL7f8vMYffBXFLioijYtg2bm+van3YaX2zaX679AeMk68PSaulR7drhbNGC/UldwKujZ1PGIdbtOujZbt8o3vsy3NK9JclxkezPKeD685oxbdmvzPlum+d4o8RonBHKx0VERESkegjoydQYcymutYVigLbAWuA3oCmugsyLKytAObpf9x4iO881rate/kFO27vZc2zFZldvkffQuTlte3lex0WW5sVZkbFkvf8+s0/tydDeD/JZ7+tp/Nzf2VWntBgDwPpdB1m/uzRLap3quzCqMYYrOzbhpm4tiAgPK9dT1EzziURERESkGgn0f9c/DPwLuNS9/ZC1tifQHggHPg5+aCe/4mLLL3sOYa09oeuUzPMBaL5vG6fv3ejZXv7rPoqKLStXb6bAhJMRncCCmNJ5RiO8FjbNcsaQk1fIG+36siu2LuO3OSkqtmzbd9jn/RatyyC/0DVnKCUukqTY8ou+ejutYTze1cGbqfKciIiIiFQjgSZFbYH3cfUKWdzD7qy1/wMew5U0yTEa/sa3XPj3Rdw3+6cTus6mPaW9Nk2yM2jn1VP07eZMRs/4jkFbU7jl9/fz8hlXUoQrQ+nSoi4Xtk3xtM1yxrI3Kp7CMFfvUU6h5aet+zmY51tccOG63Z7XbRr49hL5Uycyghb1SxMhFVkQERERkeok0KSoGCi0ri6NDCDN69h2QPOJjlHmoXwWrssA4N3vt1FUfPy9RZu8eoqaZGdwWuavnu2fth7wrBm0K7Yu/23YwXNs1AUtqevVy5PljCUzynd+0GdrdlNWjleFuzapR0+KALo0Ly3W0L5RwhFaioiIiIhUrUALLawDmrtfLwfGGGO+AgqBe4DNQY/sJLd9f+mQtKJiy57sPFLjy9avCMzGjNKeosYHdxNfkEPTg7vYEpda4TmtU+vQs3UKee5hcOAaPpcZ5ZvkfLa2fFLkc50AeooA7urdmvyiYprVjaXbqfUCOkdEREREpCoEmhRNB05zv34UWACUFHUuAq4LclwnvW37fefp7DyQe9Sk6NUvNvLxyp0UW0uYMfQ7vSE3d2/h01PUNNuVxLTbu9knKWqatYv4/EOsqu+aQzT6olaEhRmineFEOcLILSimINzB9ljfogpHW4g10J6i1Pgoxg88K6C2IiIiIiJVKaCkyFr7L6/XK4wxpwN9cVWjW2CtXV1J8Z20dpRNirJyOROw1lJUbIkI9x3ZOPfH7fzt47U++37Ysp82DeI4cLgAgOiCXOrlZhHdqRPt9mxmfvNzAYgw8KcVM2iRtYNve1xJ2pg/0uu00oQpKcbJjgOuct2b4xse0+dolVrnmNqLiIiIiFQ3R51TZIxxGmPuNMZ4JqNYa7daa1+31r6ohOj4lCQhJXZl5ZJXWMTVLy+l418/ZZFXMYOdB3J56N2f/V5n4pelC7I2PrQHAySPvp1uO1aSeiiTsOIixsRlcOqBbYTbYi5pHOmTEIErKSrxyxGSoqQYh892Wt0YYpwBL3UlIiIiIlItHTUpstbmA08BdY/WVgJXdvjcjgO5fLEug+9+28/B3EImLHKV1bbWct/sH8nKdVWAa5IUzX0Xt/Gc9+X6PZ7XTQ7uJjwhgZjzzqNu8ya8vuAp3pz/Vy7+dKqnTWSrU8vF4l1sYVuZNYm8nd8q2We77PpEIiIiIiI1UaDV59YALY/aSgJWrqfoQC6/7CmdG7R6exbFxZa5P273JD7GwHPXnMkfzmnq95pNsjOIPvtsTFgY0WedSYQtJikvm8IdOzxtIk8tnxQlevUAFYeFVxhzj1a+CVObBho6JyIiIiI1X6BJ0SPAw+65RBIE2/3MKfo1M8eznZ1XyOa9h/hk9S7Pvpu6tuDclvWoXyeSDo19S2cDNM7OILqjq5hB9Fl+iho4HES2alVud92jLL4KEOMMp3Nz387CNg3KxyAiIiIiUtMEOiHkfqAO8L0xZjOwA9ciriWstfaCIMd20iosKmZXlm9P0c6sXMKM8dn387YDrNi8z7N9VafGntc9WiWzcptvZbim2buJ6dQJgLhevch4/gUKM1xrITmaNqXezTcTnphYLp7EmKMnRU2SommaFI0j3FBQ5Lr1gVaeExERERGpzgJNiooAFVQIkt0H8yi7VuuuA7kUFBX77Ptk9S52upOnWGc4bb3WBOreKtkz76hE47z9RLVvD0B4YiKnzJ9H/m+/4WjSlPA6sRXGU7dMAQVwDdWzXjE2SYohIjyMXm1TmbdqJ6em1OGU5IqvKSIiIiJSUwRakrtnJcdRq5QdOgdwKL+InH2+++et3Ol53TEtyadM99nNkohxhpOTXwRAcs4+ktq0IiyqdK2jsJgYotq2PWo8SX6Gz7VtEM/6XQcpdGdvTZKiAXj+2rNYtmkvZzVNLFc2XERERESkJtJTbQhsL1NkoYQt03tU5NWddHazJJ9jzogwfteynme7SXYG0e6hc8cqyc/wucaJ0bT06gkqSYqiHOH0bJMS0JA7EREREZGaIKCeImNMj6O1sdYuPvFwagd/PUVH07l5Url9F7ZN4bO1rvWMTjmwjeiOfY4rHn+FFlLiI0mMcfC/XdmAym+LiIiIyMkr0DlFi/AtrOBPxbWcxceOY0yKwgyc1bR8gYSBZ6by1YuTyAqP4soNi4npNPa44vE3fC4lLpKrOzUh63ABaXVj6FFmjSIRERERkZNFoEnRhX721QP+D7gAGB20iGoB7+FzTZKi2VpmLlGUI4zcgtKiC20axBMX5VsMIX/rNvbPmskd370FgCMtjYj6FS+8eiRJfgotpMRF0bRuDK/d0Pm4rikiIiIiUlMEWmjhiwoOzTHG/AO4DPg4aFGd5LyHz3VMSyqXFF3YJoWPvYosdC4zn2jfzJns/MtffSYhxXT0sy5RgKId4URGhJFXWJqIpcRFHvf1RERERERqkmAUWvgQGBiE69QaO7x6ijqllR8Wd+npDX22vecT2YICMl76l09CFJ6YSNL1Nxx3PMaYcsUWUuOjKmgtIiIiInJyCXT43JG0AYqP2koAOJxfROahfAAiwgynN04o16ZHq2QSoh0cOFwA+Faey/7qK4r27gUgvG5dUh98kLheFxEWHX1CcSXFOj1rIoGr0IKIiIiISG0QaPU5f90QTqADMByYE8ygTmY7DpQOlUuNj6Jhom8ykxTjICHGwQOXtGX8p//j6rOb0CQpxnP8wH/+43mdcOUVJPxfv6DEVTe2dF6RMVDPT/EFEREREZGTUaA9RVMq2J8H/Bu4MyjR1ALeQ+caJUaREheJMaWj4dLqudYGurZLGtd2SfM5tygri+zPPvdsJ/TvH7S4vNcdqhcbqYVZRURERKTWCDQpauFnX661dlcwg6kNtnkVWWiUGI0jPIx6sZHsyc4DoHm9mIpOJWv+fGy+a+hd5GmnEdW6ddDiquuVFKnIgoiIiIjUJoFWn/u1sgOpLQ7kFHhe14t1JR8NEkqTomZ1K06KfIbO9b88qHF5l+XWfCIRERERqU0CGiNljPk/Y4zftYiMMbcbYy4Nblgnl8P5RZ7XB/MKPa/rRLly0oYJpfOKmrmHz5WVu24dh5evcG2EhZHQLzhziUrU9+odaqDKcyIiIiJSiwQ6ceRhwP/TOkS7j0sZ1lqGTv6GMx6fz5vLXJ1th7yToshwAK7rkoYzIowmSdH0aZ/q91q7//5c6Xk9exKRnBzUWPt2aED9OpHEOMO5pnPToF5bRERERKQ6C3ROUVvguwqO/QA8FJxwTi6/7DnEonUZAEz972aGnNeM7FzvpMg1ZO3CtimseOj3xDgjCA8zgCuhsocPYyIjyfn6aw59+aXrpLAwku8Mfl2LlLgolo69iMLiYmKcwajULiIiIiJSMwT69BsG1KngWBzgqOBYrZbjNWxun3suUbaf4XMAcVG+X2HG8y+w99VXCa9fH+MoPZZw5RVEtQlegQVvzogwnEFZz1dEREREpOYI9An4R2BwBccGAz8F+obGmL7GmHXGmA3GmLEVtBlojFltjFlljJkR6LVD7d63f2TktOUMm/ItOfmF5BeVrmlb0kOU7Wf4XFkFu3axd+JEAIr27KFwxw4ATFQUyXfcUVnhi4iIiIjUSoH2FD0HvGOMeRuYCGwFGgMjgCuBawK5iDEmHPgX0Nt9jW+NMXOttau92rQCHgC6WWv3GWNSAv0wobZgzS72u3uE1l07mN/iUqBhXwAOFxRRUFTskxQ5d+8k4/N3SLjsMpxNS+fxHHj3XSgupqy6Q2/Ekep/zpGIiIiIiByfQEtyv2uMuRN4ArjKvdsA2cAd1to5Ab5fF2CDtXYTgDFmFtAfWO3V5hbgX9bafe733h3gtUPO4bXg6aFNm8mJOwwNS49n5xb6FFo4+OQ49mxeyb6ZM2k5dy4RSUnY4mL2z37H06bu8GEU5+QQkVSX+reOqpLPISIiIiJSmwQ8o95a+09jzBSgK1AP2AMstdZmH8P7NQa2eG1vBc4t06Y1gDHmKyAceMxaO6/shYwxI3D1VJGWlnYMIVQep1dSVBgWTkGY79ebnVfIQa9CC5G7twNQlLGHnY//hcb/GE/OsmUUbN0KQFhCAsl33EFYpNYNEhERERGpLMdUZsxaexCYfwLvZ/xdtsx2BNAK6Ak0Ab40xnSw1u4vE8trwGsAnTt3LnuNkHBGlCZFBWERFIT5zhnKyi3wGT4XU5jreX1w3jwOnN+N7C8We/YlXH65EiIRERERkUoWUFJkjLkfaGKt/aOfYy8CW6y1zwZwqa2A9yI4TYDtftoss9YWAL8YY9bhSpK+DSTWUHKEl+Z8BWHhFJZJig6WGT4XXZjnc3zHQ77LPSUOGFAJUYqIiIiIiLdAq8/dRMUV5n5wHw/Et0ArY0wLY4wTuBaYW6bNe8CFAMaY+riG020K8PohVbanqGxStCc7j8JiV6eWg2KcxUVUJOrMMyqt9LaIiIiIiJQKdPhcGrC+gmObgGaBXMRaW2iMGY1rCF44MMlau8oY8xdgubV2rvtYH2PMaqAIuM9auzfAOEPKUWZOUWGZOUU7D5QOl4spLu0xqjdiBNlffknRnj0AhCfXp8HDj1RytCIiIiIiAoEnRTm4iiT40wTIq+BYOdbaj4CPyux7xOu1Be52/6lRfAstRJQrtLB9v1dSVHDY8zr+kr6k3H1X5QcoIiIiIiLlBDp87kvgPmOMz6x/9/Y97uO1nu/wufByhRZ2ZpUmQtG5hzyvHU2rR/U8EREREZHaKNCeoseApcD/jDFvAttw9RwNwVWee2hlBFfTeA+fKwiLoND4JkU7vIbPRRe4XofXr094ndiqCVBERERERMoJdPHWH40xFwJ/B+7H1cNUDCwBrrbW/lh5IdYczrJJUdmeIu85Re7Kc85qssaSiIiIiEhtFejwOay131hrewBxuOYRxVlrewKxxphJlRRfjeKIOPLirbuyvHqK3GsUKSkSEREREQmtgJOiEtbaw0AM8IAx5hdgITAw2IHVREfrKSr2WmLW01PUTEmRiIiIiEgoBZwUGWMSjDEjjDFLgHXAn4F9wK1Ao0qKr0ZxRpQu3uqvp8hbSVLkUE+RiIiIiEhIHXFOkTEmDOgL3ABcDkQB24F/AbcDY6y1iys7yJrCt6covFxPkbeYgpLhcwEt8SQiIiIiIpWkwqTIGPN3YDCQAuQC7wJvAAuAeGB0VQRYk5SrPnekpMgzp6hppcclIiIiIiIVO1JP0d2AxbXQ6lBr7d6SA8YYW+FZtdjRCi14iy7MIzwxkfCEhKoITUREREREKnCkOUWTgINAP2CdMeYlY0yXqgmrZjpaoQVvMQV5OFRkQUREREQk5CpMiqy1NwMNcC3QugIYBfzXGLMG11pF6i0qw3mMPUWaTyQiIiIiEnpHrD5nrc211s6w1l4MNAUeBIqAsYABnjLGDDHGRFV+qNWfd09RYQBziuJ6XVQVYYmIiIiIyBEcy+KtO6y1T1trOwDnAhOAVsBUYEclxVejOMJLS3IXhEVQcISkqNX4Z4jv27cqwhIRERERkSM45sVbAay131prR+Nan2gA8EVQo6qhvAstuEpyVzx8rm7rU6oiJBEREREROYojrlN0NNbaAmCO+0+tV67Qgqm4pygu6oS+ehERERERCZLj6ikS/8oVWgivOPGJjVRSJCIiIiJSHSgpCqJAS3JHRoT5LPQqIiIiIiKhoyfzIHKE+/YUFToi/bbT0DkRERERkepDSVEQOcsWWnA4/bbT0DkRERERkepDSVEQOUzperYFYREURjj8tqujpEhEREREpNpQUhRE4YX5ntdFEU4KKqg+p54ifZ1nBgAAIABJREFUEREREZHqQ0lREDny8jyvCyKcFFbw9cYpKRIRERERqTaUFAVRWG6O53VhhIMCjN926ikSEREREak+lBQFkSMv1/O6IDyCAus/Kaqj6nMiIiIiItWGkqIgCvfqKcoLc2AraKfhcyIiIiIi1YeSoiCKyMn2vD4cVnHio+FzIiIiIiLVh5KiIAo7fMjzOseUJj6O4kKfdirJLSIiIiJSfSgpCqLwnNLhc9aryEJ8XAwRYaXbSopERERERKoPJUVBFHEoy+9+Z0SYT3EFFVoQEREREak+lBQFkcnO9rvfER5GnHdSpJ4iEREREZFqQ0/nQWQPHSKiqJDCcN+v1RFuiHZGsIXDAD4JkoiIiIROXl4emZmZHDx4kKKiolCHIyLHyOl0Ur9+fRISEk7oOno6D6LigwdxOMsnRc6IcK7s2ISV21ZzakodTm98YjdNRERETlxeXh6//fYbSUlJNG/eHIfDgTH+1xgUkerHWsvhw4fZunUrkZGRREVFHfe1lBQFUfGhbCIc5f8vkzPcMPz8FlzSoQEpcZFEhGvUooiISKhlZmaSlJRE/fr1Qx2KiBwHYwwxMTHUr1+fjIwMmjZtetzX0tN5EBVlH8JRVFhuv8OdBDVKjFZCJCIiUk0cPHiQ+Pj4UIchIicoLi6O3NzcE7qGntCDqPjgQSJs+Z4ihxIhERGRaqeoqAiHwxHqMETkBEVERFBYWL5j4lhU+dO6MaavMWadMWaDMWasn+NDjTEZxpgf3H9uruoYj1dxdrb/nqIIJUUiIiLVkeYQidR8wfgdV+mcImNMOPAvoDewFfjWGDPXWru6TNN/W2tHV2VswVB06BCOYn9zipQUiYiIiIhUV1X9tN4F2GCt3WStzQdmAf2rOIZK0+ipvxHTMLXcfmeE/i+UiIiIiEh1VdVJUWNgi9f2Vve+sq42xvxkjJltjPFbRsIYM8IYs9wYszwjI6MyYj1m8X36EFW/brn9mlMkIiIitdHYsWMxxrBz587jOj83NxdjDKNGjQpyZCK+qvpp3V+XiS2z/T7Q3Fp7BrAAeMPfhay1r1lrO1trOycnJwc5zOPnLwFSUiQiIiKhYowJ+M/mzZtDHW619/3333u+r+XLl4c6HAmSql6naCvg3fPTBNju3cBau9drcyLwdBXEFTROP0UVlBSJiIhIqEybNs1n+8svv+S1115jxIgRdO/e3edYsP9H87hx43jssceOe1HNqKgoDh8+TERE9VlaMz09naSkJM/rzp07hzgiCYaq/ifsW6CVMaYFsA24FrjOu4ExpqG1dod783JgTdWGeGL8FVWIVPU5ERERCZEhQ4b4bBcWFvLaa6/xu9/9rtyxilhrycnJITY29pjeOyIi4oQTmuNNqCpDbm4uM2bMYNCgQVhrmTFjBuPHjyc6OjrUoR3VwYMHiYuLC3UY1VaVPq1bawuB0cB8XMnOW9baVcaYvxhjLnc3u8MYs8oY8yNwBzC0KmM8Uf57ilRoQURERGqGefPmYYxh5syZvPDCC7Rt25bIyEj++c9/ArB06VJuuOEGWrVqRUxMDPHx8fTo0YMPPvig3LX8zSkq2ffLL79w33330bhxY6KioujUqROffvqpz/n+5hR571u8eDHnn38+MTExJCcnM2rUKHJycsrFsWDBAs4991yioqJo2LAh9957r2cY3FNPPRXwdzNnzhz27dvHjTfeyNChQzlw4ADvvPNOhe1nzZpFjx49SEhIICYmhrZt2zJmzBiKikqrFRcXFzNhwgTOOecc6tSpQ1xcHGeeeSbjxo074vdYokGDBvTt29fv9zNv3jy6du1KbGws11xzDQBbtmzhrrvu4swzzyQxMZHo6Gg6dOjAc889R3Fxcbnr5+bm8uSTT3LGGWcQHR1NYmIiXbp04dVXXwXgySefxBjDkiVLyp176NAh4uPj6devXwDfbmhVeV+ktfYj4KMy+x7xev0A8EBVxxUsmlMkIiIiJ4Onn36aAwcOMGzYMFJSUmjZsiUAb7/9Nhs3buTaa68lLS2NjIwMpkyZwmWXXcY777zDVVddFdD1Bw0aRHR0NH/60584fPgw//jHP7j88svZsGEDjRv7q8Pl65tvvuHtt9/m5ptvZsiQIXz22We8+uqrOJ1OXnzxRU+7zz77jEsuuYSUlBQefPBB4uLimDVrFosWLTrm7yQ9PZ22bdvSpUsXAE477TQmTZrkt8ftnnvuYfz48Zx++uncc889pKamsmHDBmbPns1TTz1FeHg41lr+8Ic/MHv2bLp168ZDDz1EQkICq1evZvbs2Tz00EPHHGOJr776ihkzZjBixAhuuukmwsPDAVixYgXvv/8+/fv355RTTiEvL48PP/yQe++9l99++40XXnjBc43c3Fx69erF0qVLueSSS7jxxhtxOp389NNPvPfee4wcOZJhw4bx6KOPkp6ezvnnn+8Tw9tvv83BgwcZPnz4cX+OqlJ9BmieJDSnSEREpOZb0/a0UIdQodPWVs3Mgu3bt7N27Vrq1vWtrDtu3Lhyw+juuOMOzjjjDMaNGxdwUtS4cWNmz57tWXizW7du9OjRg9dff51HH330qOf/9NNPfPvtt3Ts2BGAUaNG0atXL1577TWeffZZIiMjAbj77rtxOp0sW7aMpk1dU9tvv/12unbtGlCcJTZv3szChQt58sknPftuvPFGHnjgATZt2uRJGgEWL17M+PHjufjii5k7dy5Op9Nz7JlnnvF85mnTpjF79myGDx/OxIkTfRYh9ddrcyxWrlzJ4sWLy80b69OnD+vXr/d5r7vuuouBAwfy8ssv88gjj1CvXj1PrEuXLuXxxx/nkUce8blOSXwNGjTgsssu4+233+bFF1/0GaKXnp5OSkoKl1122Ql9lqqgp/Ug85cA+UuURERERKqzYcOGlUuIAJ+EKCcnh71795Kbm8sFF1zADz/8QF5eXkDXHzNmjM+D+fnnn4/T6WT9+vUBnX/BBRd4EqISF110EXl5eWzZ4loB5tdff+Wnn35iwIABnoQIwOl0cscddwT0PiUmTZqEMcanV+j6668nLCyMyZMn+7SdPn064Opt806IAJ/PPH36dMLDw30SpRJhYSf2/HjuueeWS4gAYmJiPO+Vl5dHZmYme/bsoU+fPhQUFPDdd9/5xJeSksIDD5QfxOUd34gRIzh06BCzZs3y7Pvf//7HkiVLuOGGG3A4HCf0WaqCntaDzOln/pC/4gsiIiIi1Vnr1q397t+xYwfDhg0jOTmZ2NhY6tevT3JyMlOmTMFay4EDBwK6vnfPCriShaSkJPbu3VvBGUc+H/D0cJRc45dffgGgTZs25dr621eR4uJipkyZQufOncnNzWXDhg1s2LCBnJwcunTpwpQpU3x6dtavX4/D4aBDhw5HvO769etJS0vzm3yeqIruX35+Po899hinnnoq0dHR1KtXj+TkZG655RYA9u3bB7iKa2zcuJH27dsfNanp06cPzZs3Jz093bOv5PXNN98cjI9T6TR8LshUaEFERKTmq6ohatVZTExMuX1FRUX06tWLX375hTvvvJOzzz6bhIQEwsLCePXVV5k9e3bAw75K5riUZW3ZJSyP7XzvawR6raP55JNP2LJlC1u2bKFVq1YVtikpeBDo+1prA+oRKtuL5K2wsNDvfn/3D2D06NFMnDiRwYMH88gjj5CcnIzD4WDZsmU8/PDD5e7fkd67RFhYGMOHD+fhhx9m1apVtGnThqlTp3L++ecfU/IZSkqKgsxvoQUNnxMREZGTwPLly1mzZg1PPvlkuSFVL730UoiiqliLFi0AWLduXblj/vZVZNKkScTGxjJlyhS/x4cNG0Z6eronKWrTpg2LFi1i1apVnHHGGRVet02bNixYsIDMzMwj9haVHMvMzKRBgwae/VlZWQH3rJV488036dOnD2+++abP/pUrV/psG2M49dRTWblyJQUFBUftLRo2bBiPPfYY6enpXHDBBezcuZO//e1vxxRbKOlpPchUaEFEREROViW9M2V7Qr777js+/PDDUIR0RM2bN6dDhw7Mnj3bM88IXEPIvCvUHcnevXv5z3/+w6WXXsqAAQP8/unXrx9z585lz549AFx3nWsZzrFjx1JQUOBzPe/vbvDgwRQVFTF27Nhy36n3dslQuAULFvi0ee655wL6DN7XjIiIKPdeWVlZPlXnvOPbvXs3zzzzjN9reWvUqBH9+vVj2rRpvPzyy8THxzNw4MBjii+U1FMUZH4LLSgpEhERkZPAGWecQevWrRk3bhz79++nVatWrFmzhokTJ3LGGWf4TNKvLsaPH88ll1zCeeedx6hRo4iLi2PmzJmeYWFHGx42bdo08vPzufrqqytsc/XVVzNr1izefPNNxowZQ48ePbjzzjt54YUX6Ny5M9dccw2pqals2rSJt956i1WrVhEVFcWQIUOYM2cOEydOZM2aNVx22WXEx8ezbt06vvjiC8/3eemll9KiRQvuv/9+duzYQVpaGl988QU//PADCQkJAX8Xxhiuuuoq3njjDQYPHkzPnj3ZuXMnr7/+OikpKWzevNmn/X333ceHH37IQw89xH//+1969eqF0+nk559/5rfffuOjj3xW2WHEiBHMnTuX+fPnM3LkyAqH8FVHSoqCzF8CpOpzIiIicjJwOp189NFH3HfffUyaNInDhw9z+umnM3PmTJYsWVItk6LevXt7HuyfeOIJkpKSuO6667jiiivo0aMH0dHRRzx/0qRJREZGcumll1bY5pJLLiE6OppJkyYxZswYAJ5//nnOPvtsJkyYwFNPPYW1lrS0NPr37+8ZimaMYfbs2bz00ktMnjyZRx99FIfDQcuWLX16WRwOBx988IEn0SqJZ9GiRZx11lnH9H289NJLJCYmMmfOHN555x2aNWvGH//4R9q1a1dukdWoqCgWLlzIM888w6xZs/j000+JiYmhdevWfgsoXHLJJaSlpfHbb7/ViLWJvJlgTUALpc6dO9vly5eHOgwA3li6mUfnrvLZN/GGzvRulxqiiERERMSfNWvWcNpp1Xc9Iqlc06dPZ8iQIbz77rtcccUVoQ7npGCtpVWrVsTGxvLjjz9W6XsH8ns2xqyw1nb2d0xdGEHmt9CCqs+JiIiIhERxcTH5+fk++/Ly8nj++eeJjIz0u5aPHJ+PP/6YjRs3MnLkyFCHcsw0fC7I/A2V05wiERERkdDIysritNNOY/DgwbRu3ZqMjAxmzpzJqlWrePTRRz1rG8nxW7BgARs3buSJJ56gUaNGDB06NNQhHTMlRUHmr1dIJblFREREQiM6Opo+ffowZ84cdu7cCUDbtm159dVXGTFiRIijOzk89NBDrFixgg4dOjBhwoQaVWChhJKiIItUT5GIiIhItREZGckbb7wR6jBOasuWLQt1CCdMT+tB5n9Okb5mEREREZHqSk/rQeZ3naIIFVoQEREREamulBQFmb9CC+opEhERERGpvvS0HmQaPiciIiIiUrPoaT3I/BZaUPU5EREREZFqS0/rQaaeIhERERGRmkVP60GmxVtFRERERGoWPa0Hmd/FW/3sExERERGR6kFJUZCV7RUyBsLDlBSJiIjIyev888/n1FNP9dk3ZMgQIiIiAjp/w4YNGGMYN25c0GMrLCzEGMPNN98c9GvLyUNJUZCVHT7nDA/DGCVFIiIiEhrXXHMNxhh++OGHCttYa2nRogWJiYkcPny4CqMLjszMTB577DEWL14c6lACcvfdd2OMoW3btqEORdyUFAVZ2aIKmk8kIiIioTR8+HAAJk+eXGGbhQsXsnnzZq699lqio6OD8r6TJ0/m0KFDQbnW0WRmZvL444/7TYoiIiI4fPgwr7zySpXEcjQFBQW8+eabnHLKKaxbt46vvvoq1CEJSoqCrmxPkUPluEVERCSE+vTpQ9OmTZk+fTr5+fl+25QkTCUJVDA4HA4iIyODdr0TERUVFfBQvso2d+5cMjIySE9Pp379+kyaNCnUIQWkqKiInJycUIdRafTEHmQRZeYPqciCiIiIhFJYWBhDhw5l7969zJ07t9zxrKws5syZQ4cOHTjnnHM8+2fMmMFll11GWloakZGRJCcnc9VVV7Fy5cqA3reiOUWLFy+ma9euREdH06BBA+644w6/PUqFhYWMGzeO7t27k5qaitPppFmzZtx+++1kZmZ62i1YsIBWrVoB8PDDD2OMwRjjmeN0pDlFr776Kh07diQ6OprExEQuvvhili5dWi6OkvOXLFlC9+7diYmJoX79+owYMeKYe8PS09Np1aoVF1xwAYMGDeKtt94iOzvbb9sDBw7w4IMP0rZtW6KioqhXrx7du3fnrbfe8mm3Y8cORo8eTcuWLYmMjCQ1NZU+ffrw+eefe9o0adKE3//+9+XeY8GCBRhjePPNNz37Xn/9dYwxLFy4kMcff9xz3Tlz5gAwb948Bg4cSIsWLYiKiiIpKYmLL76YL7/80u/nWL9+PTfeeCNNmjTB6XTSqFEjrrjiCr7//nsA2rdvT4sWLbDWljt3xowZGGOYOXPmUb7ZE1M9UuaTiDEGZ3gY+UXFgNYoEhERkdC76aabGDduHJMnT2bAgAE+x2bNmkVOTk65XqKXXnqJ1NRURo4cSWpqKhs2bOC1116ja9eufP/995xyyinHHMfSpUvp3bs3iYmJjB07lvj4eGbOnMmSJUvKtc3NzeW5557j6quv5oorriA2NpZvvvmG1157ja+++opvv/0Wh8NBhw4d+Pvf/869997LgAED6N+/PwBxcXFHjOWee+5h/PjxnHfeefztb3/jwIEDvPrqq/Ts2ZMPPviAPn36+LRfsWIF7777LsOHD2fIkCF8/vnnTJw4kYiICCZMmBDQ59+2bRuffPIJjz/+OABDhw7ln//8J2+99RbDhg3zaZuZmUm3bt1Yu3YtAwcO5LbbbqOoqIgVK1bw4YcfMnDgQAA2bdpEt27dyMjIYOjQoXTq1Ins7GyWLVvGggULuOiiiwKKzZ+77rqLoqIiRowYQXx8vCf5nDRpEvv372fo0KE0btyYrVu38vrrr3PRRRfxxRdf0LVrV881vv76a3r37k1RURHDhw+nffv27N27l0WLFrFs2TI6duzILbfcwl133cXnn39Or169fGKYNGkSSUlJXHnllcf9OQKhpKgSOCNKkyJ/6xaJiIhI9dZ87IehDqFCm5/qd8zntGjRggsvvJD58+ezfft2GjVq5Dk2efJknE4nQ4YM8Tnn008/JTY21mffkCFD6NSpEy+88AIvvvjiMccxZswYAL766itPT85tt93m8xBdIiYmhu3bt/vMcRo1ahTnnnsuo0aN4v333+eqq66iQYMG9O/fn3vvvZczzzyz3OfwZ/Xq1YwfP54ePXqwYMECHA4H4Bo+2K5dO2699VbWr19PWFjpc9xPP/3E119/TefOnQEYOXIk+/fvJz09neeeey6guVhTpkyhuLiY66+/HoBOnTpx+umnk56eXi4pGjt2LGvXrvV7rLi42Oc72bVrF59++mm5hMK73fHIz89nxYoV5T7b5MmTy/2zMXLkSNq3b89TTz3l6ZEsLi5m6NChFBQUsHz5ctq3b+9p/+CDD3riu+GGG3jggQdIT0/3+QybN2/m888/Z/To0URFRZ3QZzkaPbFXAu8hcyq0ICIiItXB8OHDKSoqYtq0aZ59a9euZdmyZVx++eXUr1/fp33JQ6+1lqysLPbs2UODBg049dRT+frrr4/5/bdv3863337LVVdd5VO+OzIy0pMseQsLC/M8jBcVFbF//3727Nnj6fk4nhhKvPfeewDcf//9noQIXEPMbrjhBjZt2sRPP/3kc87555/vSYhKXHTRReTn5/Prr78e9T2ttUyePJkLL7yQtLQ0z/4bb7yRpUuXsnbtWs++oqIi/v3vf3P66aeXS4gAT7KWkZHBggUL6NevX7mEyLvd8brtttv8JnveCVF2djZ79+7F4XDQpUsXn/uyYsUK1q5dy8033+yTEJWNr27dulx99dXMmTOHffv2eY5PmjQJa21Q57pVRE/slcC7d0jD50RERKQ6uOqqq0hMTPSpQlcyyd/fg/eKFSu49NJLiYuLIyEhgeTkZJKTk1mzZo3Pg2ugNm3aBOC3DHW7du38njNr1izOOeccoqOjSUpKIjk5mdatWwMcVwwlfvnlFwC/D+odOnTwibdEy5Yty7WtV68eAHv37j3qey5atIiNGzfSq1cvNmzY4Plz3nnnYYzxKbiwa9cusrKyOOuss454zfXr12OtpWPHjkd9/+NR8l2XtWHDBv7whz+QmJhIXFwc9evXJzk5mfnz5/vcl/Xr1wMEFN+IESPIy8tj+vTpgKuXacqUKXTu3JkzzzwzCJ/myDR8rhJ4J0IqtCAiIlLzHM8QteouKiqK6667jgkTJrB06VLOPfdcpk2bRpMmTcrNn9m8eTM9evSgbt26PPLII7Ru3ZrY2FiMMfzxj3+koKDgmN+/ZBK9v/Ub/U2wf+uttxg0aBDnnXceL774Ik2aNCEqKor8/Hz69et3QkPD/L3f0YSHh5/Q9dLT0wH485//zJ///Odyx6dOncqTTz5JRETEEb8rf+8byJqYFbUpLCys8JyYmJhy+7KysujevTu5ubncdddddOjQgbi4OMLCwhg3bpzP/LBjia9Hjx60bduW9PR0Ro8ezSeffMKWLVv8fleVQUlRJVBPkYiIiFRHw4cPZ8KECUyePJnMzEx27tzJn//853IP/O+88w45OTnMmzeP7t27e/Zba9mzZw8JCQnH/N4lhRnWrFlT7pi/fdOmTSMmJoaFCxf6zCfxV/0ukIduf7GsWrWKZs2a+RxbvXo14L9n6HgdOHCAOXPm0LdvX79DwX744QeeeOIJPvzwQ/r370+DBg2Ij4/3VGerSOvWrTHGHLUduIaoeVftK1G2R+xoPv30U3bu3MnUqVM9c6NKjB071me7TZs2AHz//ffcdNNNR732Lbfcwj333MN3331Heno6MTExDBo06JjiO156Yq8E3vOIVGhBREREqotOnTpx1lln8e9//5uXXnoJY4zfh9WSJKlsD8grr7zCnj17juu9GzVqROfOnZkzZw4bNmzw7M/Ly+P555/3G0NYWJhPj5C1lnHjxpVrW6dOHQC/D/3+lFSoe/bZZ316SrZt28Ybb7xBy5YtOeOMMwL7YAGYMWMGhw8f5tZbb2XAgAHl/owdO5aoqCjPELrw8HCuvfZafv75Z954441y1yu5L8nJyfTu3ZsPPviAhQsXVtgOXAnU6tWr2bFjh2dfbm5uwJXzSlT0z8bHH3/MihUrfPZ16tSJtm3b8vrrr/tNfMte48YbbyQyMpKnn36auXPncs011xAfH39M8R0v9RRVAu/eIRVaEBERkepk+PDh/PGPf2T+/Pn07NnTb2ntfv368eCDDzJ48GBuv/12EhIS+Oqrr5g3bx4tWrQ47vceP348vXr1olu3btx2220kJCQwY8YMv8PPBgwYwH/+8x8uuugirr/+evLy8nj33XfJzc0t1zY1NZXmzZszffp0mjdvTkpKCnFxcfTr538YZLt27bj77rsZP348F1xwAQMHDiQrK4tXXnmFw4cPM2HChBMuUuAtPT2dOnXqlBumWKJOnTpcfPHFfPjhh+zcuZMGDRrw5JNPsmjRIoYOHcq8efPo2rUrxcXFnl6hKVOmADBhwgS6du1Knz59PCW5c3JyWLZsGa1bt+aJJ54AYPTo0cyePZtevXoxcuRI8vLymDp1qiehDFSPHj1ISUlhzJgxbNy4kcaNG/Pdd98xffp0OnTo4JP8hIWFMXnyZH7/+99zzjnncPPNN9OuXTv27dvHF198wWWXXcatt97qaV+vXj2uvPJKZs2aBeB3balKY62t8X/OPvtsW51cNeEr2+z+D2yz+z+wI6cuD3U4IiIi4sfq1atDHUJIZGZm2qioKAvYqVOnVthu4cKFtmvXrrZOnTo2MTHR9uvXz65atcp269bNnnLKKT5t/e0bPHiwDQ8P93vd8847z0ZGRtqUlBQ7evRo+8MPP1jA/vWvf/Vp+/LLL9u2bdvayMhI27BhQzty5Ei7e/duC9jhw4f7tF26dKn93e9+Z2NiYizgiaegoMBve2utfeWVV+yZZ55pIyMjbVxcnO3du7ddsmSJT5sjnT9x4kQL2C+//LLC7/HHH3+0gB04cGCFbay1durUqRawTz/9tGdfZmamveeee2zLli2t0+m09erVs927d7ezZ8/2OXfLli12xIgRtkmTJtbhcNiUlBR78cUX288//9ynXXp6um3VqpV1OBy2RYsW9tlnn7Xz58+3gJ02bVrAn+v777+3vXv3tgkJCbZOnTq2Z8+edsmSJRXe89WrV9tBgwbZ1NRU63A4bMOGDe0VV1xhv//++3JtP//8cwvYNm3aHPH78vceRwMstxXkE8Yex0SzE2GM6Qu8AIQDr1trn6qg3QDgbeAca+3yI12zc+fOdvnyIzapUte+9l+WbXJ13152ZiP+OahyKoKIiIjI8VuzZg2nnXZaqMMQES9Lly6lW7duPPPMM9x3330BnxfI79kYs8Ja29nfsSod22WMCQf+BVwCtAMGGWPK1WA0xsQBdwDHX4A+hJwRpZMVVX1ORERERCQwL730Ek6nk6FDh1bp+1b1nKIuwAZr7SYAY8wsoD+wuky7vwLPAPdWbXjB4fRKhCJVaEFEREREpELZ2dl88MEH/Pzzz8yaNYvbbruN5OTkKo2hqpOixsAWr+2twLneDYwxHYGm1toPjDE1MylSSW4RERERkYDs3LmTQYMGUadOHQYOHMhTT/mdXVOpqjop8jeWzDOpyRgTBvwDGHrUCxkzAhgBkJaWFqTwgsN38VYlRSIiIiIiFTn11FOPa0HdYKrqJ/atQFOv7SbAdq/tOKADsMgYsxk4D5hrjCk3Icpa+5q1trO1tnOqbMrXAAAKZ0lEQVRVd68djZIiEREREZGao6qf2L8FWhljWhhjnMC1wNySg9baA9ba+tba5tba5sAy4PKjVZ+rbryHzzlVaEFEREREpFqr0qTIWlsIjAbmA2uAt6y1q4wxfzHGXF6VsVQm7wVbnSq0ICIiUm2FesiOiJy4YPyOq3pOEdbaj4CPyux7pIK2PasipmCLiyr9WmMjq/wrFhERkQCEh4dTUFCA0+kMdSgicgIKCwuJiDixZ251Y1SCKzs2pnFiNC2TY+l3esNQhyMiIiJ+xMXFkZWVFeowROQEHTx4kKioqBO6hroxKkHL5Dos/tOFhBkwRnOKREREqqO6devy22+/ARAfH4/D4dB/t0VqEGsthw8fZs+ePSdcjVpJUSUJD9O/VEVERKqzyMhI0tLSyMzMZPPmzRQVFYU6JBE5RpGRkaSmpqqnSEREROR4RUZG0rBhQxo21HB3kdpMc4pERERERKRWU1IkIiIiIiK1mpIiERERERGp1ZQUiYiIiIhIraakSEREREREajUlRSIiIiIiUqspKRIRERERkVrNWGtDHcMJM8ZkAL+GOg4v9YE9oQ5Cgk739eSle3vy0r09eenenpx0X09e1eHeNrPWJvs7cFIkRdWNMWa5tbZzqOOQ4NJ9PXnp3p68dG9PXrq3Jyfd15NXdb+3Gj4nIiIiIiK1mpIiERERERGp1ZQUVY7XQh2AVArd15OX7u3JS/f25KV7e3LSfT15Vet7qzlFIiIiIiJSq6mnSEREREREajUlRUFkjOlrjFlnjNlgjBkb6njkxBhjNhtjfjbG/GCMWe7eV9cY86kxZr3776RQxylHZ4yZZIzZbYxZ6bXP7700Li+6f8c/GWM6hS5yOZIK7utjxpht7t/tD8aYS72OPeC+r+uMMReHJmoJhDGmqTFmoTFmjTFmlTHmTvd+/W5ruCPcW/12azhjTJQx5htjzI/ue/u4e38LY8zX7t/tv40xTvf+SPf2Bvfx5qGMX0lRkBhjwoF/AZcA7YBBxph2oY1KguBCa+1ZXiUkxwKfWWtbAZ+5t6X6mwL0LbOvont5CdDK/WcE8HIVxSjHbgrl7yvAP9y/27OstR8BuP99fC3Q3n3OBPe/t6V6KgTusdaeBpwH3O6+h/rd1nwV3VvQb7emywMustaeCZwF9DXGnAc8jevetgL2AcPd7YcD+6y1pwL/cLcLGSVFwdMF2GCt3WStzQdmAf1DHJMEX3/gDffrN4ArQhiLBMhauxjILLO7onvZH5hqXZYBicaYhlUTqRyLCu5rRfoDs6y1edbaX4ANuP69LdWQtXaHtfY79+uDwBqgMfrd1nhHuLcV0W+3hnD//rLdmw73HwtcBMx27y/7uy35Pc8GehljTBWFW46SouBpDGzx2t7KkX/kUv1Z4BNjzApjzAj3vlRr7Q5w/YsdSAlZdHKiKrqX+i3XfKPdQ6gmeQ1x1X2todxDajoCX6Pf7UmlzL0F/XZrPGNMuDHmB2A38CmwEdhvrS10N/G+f5576z5+AKhXtRGXUlIUPP4yW5X2q9m6WWs74RqWcbsxpkeoA5Iqod9yzfYycAquoRs7gOfc+3VfayBjTB3gHWCMtTbrSE397NP9rcb83Fv9dk8C1toia+1ZQBNcPXqn+Wvm/rta3VslRcGzFWjqtd0E2B6iWCQIrLXb3X/vBt7F9ePeVTIkw/337tBFKCeoonup33INZq3d5f6PcjEwkdJhNrqvNYwxxoHroXm6tXaOe7d+tycBf/dWv92Ti7V2P7AI17yxRGNMhPuQ9/3z3Fv38QQCHxIddEqKgudboJW7woYT16TAuSGOSY6TMSbWGBNX8hroA6zEdU9vdDe7EfhPaCKUIKjoXs4FbnBXszoPOFAyXEeqvzLzSK7E9bsF13291l3tqAWuCfnfVHV8Ehj3vIJ0YI21drzXIf1ua7iK7q1+uzWfMSbZGJPofh0N/B7XnLGFwAB3s7K/25Lf8wDgcxvCBVQjjt5EAmGtLTTGjAbmA+HAJGvtqhCHJccvFXjXPd8vAphhrZ1njPkWeMsYMxz4DbgmhDFKgIwxM4GeQH1jzFbgUeAp/N/Lj4BLcU3mzQFuqvKAJSAV3NeexpizcA3B2AyMBLDWrjLGvAWsxlX96nZrbVEo4paAdAOuB352z08AeBD9bk8GFd3bQfrt1ngNgTfc1QHDgLestR8YY1YDs4wx44DvcSXFuP+eZozZgKuH6NpQBF3ChDAhExERERERCTkNnxMRERERkVpNSZGIiIiIiNRqSopERERERKRWU1IkIiIiIiK1mpIiERERERGp1ZQUiYhIpTHGDDXG2Ar+7A9xbFPcpbxFRKSW0zpFIiJSFa7BtXq5t8JQBCIiIlKWkiIREakKP1hrN4Q6CBEREX80fE5ERELOa5hdD2PMe8aYbGPMXmPMv4wx0WXaNjTGTDXG7DHG5BljfjLGDPFzzRbGmGnGmJ3udpuMMS/4adfRGPOlMSbHGLPeGDOqzPEGxpg3jDHb3dfZYYz5wBiTEvxvQkREQkE9RSIiUhXCjTFl/5tTbK0tLrPvTeAtYALQBXgEiAWGAhhjYoEvgCTgQWALMASYZoyJsda+5m7XAvgGyAEeBdYDTYE+Zd4vHpgBPA/8BbgJeNkYs85au9DdZhrQDLjP/X6pQC8g5ni+CBERqX6UFImISFVY62ffh8D/ldn3kbX2XvfrT4wxFviLMeZJa+3/cCUtrYALrbWL3O0+NsakAuOMMenW2iLgcSAaONNau93r+m+Ueb844LaSBMgYsxhX4jQIKEmKfgc8aK2d7nXe2wF9ahERqRGUFImISFW4kvKFFvxVn3urzPYsYByuXqP/AT2AbV4JUYk3gclAO+BnXInNB2USIn9yvHqEsNbmGWPWA2lebb4F7jPGGOBzYKW11h7luiIiUoMoKRIRkaqwMsBCC7sq2G7s/rsusMPPeTu9jgPUo3wS5s8+P/vygCiv7T/gGoL3J1zD7HYYY14BxvkZ/iciIjWQCi2IiEh1klrB9jb335lAAz/nlezb6/57D6WJ1Amx1u621t5urW0MtAWm4BqeNzIY1xcRkdBTUiQiItXJwDLb1wLFuIomgKvIQhNjTLcy7a4DdgNr3NufAP9njGkYzOCsteustQ/i6mHqEMxri4hI6Gj4nIiIVIWzjDH1/exfbq31XsT1UmPMs7iSmi64hq1NdRdZAFcvzZ3AHGPMn3ENkRsM9AZGuoss4D6vH7DUGPMksAFXz1Ffa2258t0VMcYkAAuA6biKRRQA/XFVv/sk0OuIiEj1pqRIRESqQkXV2pJxDXUrMQS4B7gVyAcmAiXV6LDWHjLGXAA8AzyFq3rcOuB6a+2bXu02G2POxVWk4W/udtuA/xxj3LnAd8AtuMpyF7vfb7C19livJSIi1ZRRAR0REQk1Y8xQXNXjWgVYkEFERCRoNKdIRERERERqNSVFIiIiIiJSq2n4nIiIiIiI1GrqKRIRERERkVpNSZGIiIiIiNRqSopERERERKRWU1IkIiIiIiK1mpIiERERERGp1ZQUiYiIiIhIrfb/PYsLXSL3epsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1008x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Loss Curves\n",
    "plt.figure(figsize=[14,10])\n",
    "plt.subplot(211)\n",
    "plt.plot(history.history['loss'],'#d62728',linewidth=3.0)\n",
    "plt.plot(history.history['val_loss'],'#1f77b4',linewidth=3.0)\n",
    "plt.legend(['Training loss', 'Validation Loss'],fontsize=18)\n",
    "plt.xlabel('Epochs ',fontsize=16)\n",
    "plt.ylabel('Loss',fontsize=16)\n",
    "plt.title('Loss Curves',fontsize=16)\n",
    " \n",
    "# Accuracy Curves\n",
    "plt.figure(figsize=[14,10])\n",
    "plt.subplot(212)\n",
    "plt.plot(history.history['accuracy'],'#d62728',linewidth=3.0)\n",
    "plt.plot(history.history['val_accuracy'],'#1f77b4',linewidth=3.0)\n",
    "plt.legend(['Training Accuracy', 'Validation Accuracy'],fontsize=18)\n",
    "plt.xlabel('Epochs ',fontsize=16)\n",
    "plt.ylabel('Accuracy',fontsize=16)\n",
    "plt.title('Accuracy Curves',fontsize=16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "257/257 [==============================] - 0s 228us/step\n",
      "Test loss: 0.35465695793062796\n",
      "Test accuracy: 0.8404669165611267\n"
     ]
    }
   ],
   "source": [
    "scores = model.evaluate(x_test, y_test, verbose=1)\n",
    "print('Test loss:', scores[0])\n",
    "print('Test accuracy:', scores[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "257/257 [==============================] - 0s 444us/step\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    abnormal       0.78      0.80      0.79        97\n",
      "      normal       0.88      0.86      0.87       160\n",
      "\n",
      "    accuracy                           0.84       257\n",
      "   macro avg       0.83      0.83      0.83       257\n",
      "weighted avg       0.84      0.84      0.84       257\n",
      "\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict(x_test, verbose=1)\n",
    "\n",
    "y_true, y_pred = [],[]\n",
    "classes = encoder.classes_\n",
    "for idx, prediction in enumerate(predictions): \n",
    "    y_true.append(classes[np.argmax(y_test[idx])])\n",
    "    y_pred.append(classes[np.argmax(prediction)])\n",
    "    \n",
    "print(classification_report(y_pred, y_true))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save Model and Weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_name = \"heartbeat_classifier (normalised).h5\"\n",
    "model.save(model_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1 sec overlap\n",
    "# Trained for 3sec for files >3sec"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Classifying an heart beat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 89ms/step\n",
      "[[0.03617344 0.9638266 ]]\n"
     ]
    }
   ],
   "source": [
    "# # load and evaluate a saved model\n",
    "# from keras.models import load_model\n",
    " \n",
    "# # load model\n",
    "# model = load_model(\"trained_heartbeat_classifier.h5\")\n",
    "\n",
    "# # File to be classified\n",
    "# classify_file = \"my_heartbeat.wav\"\n",
    "# x_test = []\n",
    "# x_test.append(extract_features(classify_file,0.5))\n",
    "# x_test = np.asarray(x_test)\n",
    "# x_test = x_test.reshape(x_test.shape[0], x_test.shape[1], x_test.shape[2], 1)\n",
    "# pred = model.predict(x_test,verbose=1)\n",
    "\n",
    "# print(pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normal heartbeat\n",
      "confidence: 0.9638266\n"
     ]
    }
   ],
   "source": [
    "# pred_class = model.predict_classes(x_test)\n",
    "# if pred_class[0]:\n",
    "#     print(\"Normal heartbeat\")\n",
    "#     print(\"confidence:\",pred[0][1])\n",
    "# else:\n",
    "#     print(\"Abnormal heartbeat\")\n",
    "#     print(\"confidence:\",pred[0][0])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}