nlp with embedding.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Science Task - NLP with Tensorflow\n",
    "\n",
    "\n",
    "Use Tensorflow to compare performances of a pretrained model vs. a basic model from scratch by you, then offer your opinions (not to build) how you can improve your custom-built model performance.\n",
    "\n",
    "\n",
    " __Requirements__:\n",
    "\n",
    "1. Pick a text (NLP) dataset or image/text (CNN/NLP) dataset of your choice from here:\n",
    "https://github.com/tensorflow/datasets\n",
    "2. Pick one of the pretrained models to do a transfer learning on your dataset: https://tfhub.dev/\n",
    "3. Describe the purpose of the model, show the model architecture and performance (two metrics of your choice)\n",
    "4. Please choose a problem directly related to your work experience\n",
    "5. Now, build your own model in tensorflow from scratch to solve the same problem based on your work experience, show the model architecture and performance (two same metrics used from step 3) – note that a model from scratch should have significantly poorer model performance due to time/computing resource constraints, and limited training dataset \n",
    "6. I am more interested in how you would design the model architect and why, not the model performance\n",
    "7. Offer your opinions, in a few bullet points or short sentences, steps you can take (you do not have to build the model) to improve your own model to get to a closer power as the pretrained model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sentiment analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import tensorflow_datasets as tfds\n",
    "import tensorflow as tf\n",
    "import keras\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "import tensorflow_hub as hub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import os\n",
    "# os.chdir(\"/Volumes/Work/DS projects/Sentiment Analysis\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Data__:\n",
    "Movie Review dataset was from \"http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz\".  The compiled .csv data file was downloaded and saved in the working directory of this jupyter notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25000 positive reviews, 25000 negative reviews, 50000 in total\n"
     ]
    }
   ],
   "source": [
    "dataset = pd.read_csv('./data/IMDB Dataset.csv')\n",
    "negative = len(dataset[dataset['sentiment']=='positive'])\n",
    "positive = len(dataset) - negative\n",
    "print(\"%d positive reviews, %d negative reviews, %d in total\" %(positive, negative, len(dataset)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Pre-Processing the text__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "le = LabelEncoder()\n",
    "training_reviews,testing_reviews,training_labels,testing_labels  = train_test_split(dataset['review'].values,dataset['sentiment'].values,test_size = 0.2)\n",
    "training_labels = le.fit_transform(training_labels)\n",
    "testing_labels = le.fit_transform(testing_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Unique Words : 112146\n"
     ]
    }
   ],
   "source": [
    "tokenizer = Tokenizer(num_words=10000,oov_token='OOV')\n",
    "tokenizer.fit_on_texts(training_reviews)\n",
    "word_index = tokenizer.word_index\n",
    "training_sequence = tokenizer.texts_to_sequences(training_reviews)\n",
    "testing_sequence = tokenizer.texts_to_sequences(testing_reviews)\n",
    "train_pad_sequence = pad_sequences(training_sequence,maxlen = 200,truncating= 'post',padding = 'pre')\n",
    "test_pad_sequence = pad_sequences(testing_sequence,maxlen = 200,truncating= 'post',padding = 'pre')\n",
    "print('Total Unique Words : {}'.format(len(word_index)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Build a model of sentiment analysis by use of the pre-trained wikipedia Word2Vec model__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "embed = hub.load(\"https://tfhub.dev/google/Wiki-words-250/2\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedding_matrix1 = np.zeros((len(word_index) + 1,250))\n",
    "for word, i in word_index.items():\n",
    "    embedding_vector = np.array(embed([word]))[0]\n",
    "    if embedding_vector is not None:\n",
    "        embedding_matrix1[i] = embedding_vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(112147, 250)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "embedding_matrix1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "\n",
    "def recall_score(y_true, y_pred):\n",
    "    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "    possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))\n",
    "    recall = true_positives / (possible_positives + K.epsilon())\n",
    "    return recall\n",
    "\n",
    "def precision_score(y_true, y_pred):\n",
    "    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "    predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))\n",
    "    precision = true_positives / (predicted_positives + K.epsilon())\n",
    "    return precision\n",
    "\n",
    "def f1_score(y_true, y_pred):\n",
    "    precision = precision_score(y_true, y_pred)\n",
    "    recall = recall_score(y_true, y_pred)\n",
    "    return 2*((precision*recall)/(precision+recall+K.epsilon()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "model1 = tf.keras.Sequential([tf.keras.layers.Embedding(len(word_index) + 1, 250,\n",
    "                                                        weights=[embedding_matrix1],\n",
    "                                                        input_length=200,\n",
    "                                                        trainable=False),\n",
    "                             tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(64)),\n",
    "                             tf.keras.layers.Dropout(0.5),\n",
    "                             tf.keras.layers.Dense(256,activation = 'relu',),\n",
    "                             tf.keras.layers.Dense(128,activation = 'relu'),\n",
    "                             tf.keras.layers.Dropout(0.5),\n",
    "                             tf.keras.layers.Dense(1,activation = tf.nn.sigmoid)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding (Embedding)        (None, 200, 250)          28036750  \n",
      "_________________________________________________________________\n",
      "bidirectional (Bidirectional (None, 128)               161280    \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 256)               33024     \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 128)               32896     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 28,264,079\n",
      "Trainable params: 227,329\n",
      "Non-trainable params: 28,036,750\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model1.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "1250/1250 [==============================] - 74s 57ms/step - loss: 0.5863 - accuracy: 0.6895 - auc: 0.7584 - recall_score: 0.6554 - precision_score: 0.7173 - f1_score: 0.6556 - val_loss: 0.5579 - val_accuracy: 0.7151 - val_auc: 0.8152 - val_recall_score: 0.5391 - val_precision_score: 0.8314 - val_f1_score: 0.6457\n",
      "Epoch 2/20\n",
      "1250/1250 [==============================] - 73s 58ms/step - loss: 0.4941 - accuracy: 0.7728 - auc: 0.8424 - recall_score: 0.7722 - precision_score: 0.7829 - f1_score: 0.7646 - val_loss: 0.4484 - val_accuracy: 0.7947 - val_auc: 0.8730 - val_recall_score: 0.7828 - val_precision_score: 0.8028 - val_f1_score: 0.7871\n",
      "Epoch 3/20\n",
      "1250/1250 [==============================] - 80s 64ms/step - loss: 0.4981 - accuracy: 0.7624 - auc: 0.8395 - recall_score: 0.7730 - precision_score: 0.7695 - f1_score: 0.7573 - val_loss: 0.4547 - val_accuracy: 0.7823 - val_auc: 0.8846 - val_recall_score: 0.9066 - val_precision_score: 0.7272 - val_f1_score: 0.8024\n",
      "Epoch 4/20\n",
      "1250/1250 [==============================] - 80s 64ms/step - loss: 0.4375 - accuracy: 0.8059 - auc: 0.8804 - recall_score: 0.8074 - precision_score: 0.8109 - f1_score: 0.7997 - val_loss: 0.4432 - val_accuracy: 0.7854 - val_auc: 0.8973 - val_recall_score: 0.9182 - val_precision_score: 0.7253 - val_f1_score: 0.8061\n",
      "Epoch 5/20\n",
      "1250/1250 [==============================] - 78s 62ms/step - loss: 0.4088 - accuracy: 0.8202 - auc: 0.8964 - recall_score: 0.8228 - precision_score: 0.8231 - f1_score: 0.8140 - val_loss: 0.4024 - val_accuracy: 0.8219 - val_auc: 0.9074 - val_recall_score: 0.8545 - val_precision_score: 0.8003 - val_f1_score: 0.8225\n",
      "Epoch 6/20\n",
      "1250/1250 [==============================] - 76s 61ms/step - loss: 0.3816 - accuracy: 0.8361 - auc: 0.9105 - recall_score: 0.8437 - precision_score: 0.8360 - f1_score: 0.8326 - val_loss: 0.3658 - val_accuracy: 0.8448 - val_auc: 0.9226 - val_recall_score: 0.8348 - val_precision_score: 0.8517 - val_f1_score: 0.8386\n",
      "Epoch 7/20\n",
      "1250/1250 [==============================] - 77s 61ms/step - loss: 0.3623 - accuracy: 0.8443 - auc: 0.9195 - recall_score: 0.8491 - precision_score: 0.8444 - f1_score: 0.8401 - val_loss: 0.3728 - val_accuracy: 0.8302 - val_auc: 0.9247 - val_recall_score: 0.7489 - val_precision_score: 0.8926 - val_f1_score: 0.8096\n",
      "Epoch 8/20\n",
      "1250/1250 [==============================] - 77s 61ms/step - loss: 0.3501 - accuracy: 0.8510 - auc: 0.9252 - recall_score: 0.8616 - precision_score: 0.8470 - f1_score: 0.8479 - val_loss: 0.3587 - val_accuracy: 0.8424 - val_auc: 0.9315 - val_recall_score: 0.7776 - val_precision_score: 0.8942 - val_f1_score: 0.8270\n",
      "Epoch 9/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.3371 - accuracy: 0.8569 - auc: 0.9306 - recall_score: 0.8654 - precision_score: 0.8531 - f1_score: 0.8536 - val_loss: 0.3315 - val_accuracy: 0.8554 - val_auc: 0.9340 - val_recall_score: 0.8570 - val_precision_score: 0.8540 - val_f1_score: 0.8515\n",
      "Epoch 10/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.3285 - accuracy: 0.8631 - auc: 0.9343 - recall_score: 0.8730 - precision_score: 0.8582 - f1_score: 0.8598 - val_loss: 0.3458 - val_accuracy: 0.8450 - val_auc: 0.9328 - val_recall_score: 0.9291 - val_precision_score: 0.7962 - val_f1_score: 0.8538\n",
      "Epoch 11/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.3175 - accuracy: 0.8665 - auc: 0.9386 - recall_score: 0.8717 - precision_score: 0.8646 - f1_score: 0.8629 - val_loss: 0.3402 - val_accuracy: 0.8492 - val_auc: 0.9369 - val_recall_score: 0.7844 - val_precision_score: 0.9021 - val_f1_score: 0.8346\n",
      "Epoch 12/20\n",
      "1250/1250 [==============================] - 78s 63ms/step - loss: 0.3120 - accuracy: 0.8672 - auc: 0.9409 - recall_score: 0.8692 - precision_score: 0.8684 - f1_score: 0.8633 - val_loss: 0.3202 - val_accuracy: 0.8632 - val_auc: 0.9391 - val_recall_score: 0.8859 - val_precision_score: 0.8478 - val_f1_score: 0.8626\n",
      "Epoch 13/20\n",
      "1250/1250 [==============================] - 80s 64ms/step - loss: 0.3007 - accuracy: 0.8748 - auc: 0.9450 - recall_score: 0.8778 - precision_score: 0.8746 - f1_score: 0.8710 - val_loss: 0.3178 - val_accuracy: 0.8683 - val_auc: 0.9393 - val_recall_score: 0.8723 - val_precision_score: 0.8648 - val_f1_score: 0.8650\n",
      "Epoch 14/20\n",
      "1250/1250 [==============================] - 79s 63ms/step - loss: 0.2930 - accuracy: 0.8781 - auc: 0.9478 - recall_score: 0.8861 - precision_score: 0.8749 - f1_score: 0.8758 - val_loss: 0.3166 - val_accuracy: 0.8672 - val_auc: 0.9387 - val_recall_score: 0.8680 - val_precision_score: 0.8665 - val_f1_score: 0.8633\n",
      "Epoch 15/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.2865 - accuracy: 0.8816 - auc: 0.9502 - recall_score: 0.8862 - precision_score: 0.8797 - f1_score: 0.8780 - val_loss: 0.3095 - val_accuracy: 0.8720 - val_auc: 0.9413 - val_recall_score: 0.8789 - val_precision_score: 0.8667 - val_f1_score: 0.8691\n",
      "Epoch 16/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.2771 - accuracy: 0.8860 - auc: 0.9534 - recall_score: 0.8893 - precision_score: 0.8838 - f1_score: 0.8821 - val_loss: 0.3144 - val_accuracy: 0.8693 - val_auc: 0.9413 - val_recall_score: 0.8727 - val_precision_score: 0.8663 - val_f1_score: 0.8658\n",
      "Epoch 17/20\n",
      "1250/1250 [==============================] - 78s 62ms/step - loss: 0.2712 - accuracy: 0.8886 - auc: 0.9552 - recall_score: 0.8939 - precision_score: 0.8870 - f1_score: 0.8860 - val_loss: 0.3210 - val_accuracy: 0.8644 - val_auc: 0.9377 - val_recall_score: 0.8428 - val_precision_score: 0.8814 - val_f1_score: 0.8578\n",
      "Epoch 18/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.2611 - accuracy: 0.8916 - auc: 0.9586 - recall_score: 0.8957 - precision_score: 0.8899 - f1_score: 0.8887 - val_loss: 0.3176 - val_accuracy: 0.8641 - val_auc: 0.9398 - val_recall_score: 0.9057 - val_precision_score: 0.8372 - val_f1_score: 0.8664\n",
      "Epoch 19/20\n",
      "1250/1250 [==============================] - 78s 62ms/step - loss: 0.2539 - accuracy: 0.8946 - auc: 0.9607 - recall_score: 0.8977 - precision_score: 0.8936 - f1_score: 0.8914 - val_loss: 0.3144 - val_accuracy: 0.8657 - val_auc: 0.9417 - val_recall_score: 0.9083 - val_precision_score: 0.8375 - val_f1_score: 0.8678\n",
      "Epoch 20/20\n",
      "1250/1250 [==============================] - 79s 64ms/step - loss: 0.2467 - accuracy: 0.9002 - auc: 0.9629 - recall_score: 0.9003 - precision_score: 0.9012 - f1_score: 0.8968 - val_loss: 0.3202 - val_accuracy: 0.8631 - val_auc: 0.9411 - val_recall_score: 0.8112 - val_precision_score: 0.9057 - val_f1_score: 0.8517\n"
     ]
    }
   ],
   "source": [
    "model1.compile(loss = tf.keras.losses.BinaryCrossentropy() , \n",
    "               optimizer=tf.keras.optimizers.Adam(learning_rate=0.001) , \n",
    "               metrics = ['accuracy',\n",
    "                           tf.keras.metrics.AUC(), \n",
    "#                            tf.keras.metrics.Precision(), \n",
    "#                            tf.keras.metrics.Recall(),\n",
    "                           recall_score,\n",
    "                           precision_score,\n",
    "                           f1_score])\n",
    "\n",
    "history1 = model1.fit(train_pad_sequence,training_labels,\n",
    "                      epochs = 20,\n",
    "                      validation_data=(test_pad_sequence,testing_labels),\n",
    "                      verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_items([('loss', [0.5862730145454407, 0.49413740634918213, 0.49811679124832153, 0.437511682510376, 0.40879666805267334, 0.38156658411026, 0.36229628324508667, 0.3500630557537079, 0.33714431524276733, 0.32848426699638367, 0.31750884652137756, 0.31204771995544434, 0.3006991744041443, 0.2930067181587219, 0.2864946722984314, 0.2770780622959137, 0.271226167678833, 0.26107627153396606, 0.25394749641418457, 0.24669799208641052]), ('accuracy', [0.6894500255584717, 0.7728000283241272, 0.7623999714851379, 0.8059499859809875, 0.8202250003814697, 0.8360999822616577, 0.8443499803543091, 0.8510000109672546, 0.8569250106811523, 0.8630750179290771, 0.8664500117301941, 0.8672000169754028, 0.8747749924659729, 0.8780500292778015, 0.8816249966621399, 0.8860250115394592, 0.8886250257492065, 0.8915749788284302, 0.8946250081062317, 0.9001500010490417]), ('auc', [0.7583937644958496, 0.8424305319786072, 0.839506208896637, 0.8803722858428955, 0.8964293599128723, 0.9105474948883057, 0.9195422530174255, 0.9252190589904785, 0.9306194186210632, 0.934299111366272, 0.9385864734649658, 0.9408552050590515, 0.9450334310531616, 0.9477646350860596, 0.9502125978469849, 0.9533529877662659, 0.9551748037338257, 0.9586243629455566, 0.9606640934944153, 0.9628773331642151]), ('recall_score', [0.6554011702537537, 0.7721747159957886, 0.7729983329772949, 0.8073676228523254, 0.8227817416191101, 0.8436853289604187, 0.8490504026412964, 0.8615637421607971, 0.8654216527938843, 0.8729580044746399, 0.8716691136360168, 0.8691813349723816, 0.8777604699134827, 0.8860985636711121, 0.8861842751502991, 0.8893367052078247, 0.8939396739006042, 0.8957326412200928, 0.8976524472236633, 0.9003294706344604]), ('precision_score', [0.7172579169273376, 0.7828596234321594, 0.7694911360740662, 0.8108627796173096, 0.8230735063552856, 0.8359898328781128, 0.8443644642829895, 0.847011148929596, 0.8531122207641602, 0.8581651449203491, 0.8645853400230408, 0.868389368057251, 0.8746155500411987, 0.8749269247055054, 0.8796936273574829, 0.883803129196167, 0.8870362043380737, 0.8898886442184448, 0.8935702443122864, 0.9011854529380798]), ('f1_score', [0.6555891633033752, 0.7645565271377563, 0.7572784423828125, 0.7996772527694702, 0.8139938712120056, 0.8326407074928284, 0.8401270508766174, 0.8479124307632446, 0.8535705208778381, 0.8597780466079712, 0.8628702163696289, 0.8633224368095398, 0.8710331320762634, 0.8758148550987244, 0.8779901266098022, 0.8821412324905396, 0.8859848380088806, 0.8887355327606201, 0.891428530216217, 0.8967799544334412]), ('val_loss', [0.5578751564025879, 0.44837716221809387, 0.4547056257724762, 0.4431750476360321, 0.40244802832603455, 0.36580824851989746, 0.37278276681900024, 0.3587116301059723, 0.3315412104129791, 0.34577804803848267, 0.3402152955532074, 0.3201713263988495, 0.31779253482818604, 0.3166187107563019, 0.3094528913497925, 0.3143988847732544, 0.3209669291973114, 0.3176453411579132, 0.3143998682498932, 0.3202047646045685]), ('val_accuracy', [0.7150999903678894, 0.794700026512146, 0.7822999954223633, 0.7853999733924866, 0.8219000101089478, 0.8447999954223633, 0.8302000164985657, 0.8424000144004822, 0.855400025844574, 0.8450000286102295, 0.8492000102996826, 0.8632000088691711, 0.8683000206947327, 0.8672000169754028, 0.871999979019165, 0.8693000078201294, 0.8644000291824341, 0.8640999794006348, 0.8657000064849854, 0.863099992275238]), ('val_auc', [0.8152073621749878, 0.8729510307312012, 0.8845611214637756, 0.8973416090011597, 0.9074004888534546, 0.9226483106613159, 0.9246726632118225, 0.9315229654312134, 0.934036910533905, 0.9327558279037476, 0.9369078278541565, 0.939123809337616, 0.9392943978309631, 0.9387443661689758, 0.9412939548492432, 0.9412873983383179, 0.9377160668373108, 0.9398221373558044, 0.9417194128036499, 0.9411397576332092]), ('val_recall_score', [0.5390664935112, 0.7828019857406616, 0.9065606594085693, 0.9182078242301941, 0.8544718027114868, 0.8347873091697693, 0.7488765716552734, 0.777603268623352, 0.8570424318313599, 0.9290754795074463, 0.784355103969574, 0.8858600854873657, 0.8722880482673645, 0.8680384755134583, 0.8788695931434631, 0.8727107048034668, 0.8427509069442749, 0.9056518077850342, 0.9082618951797485, 0.8112008571624756]), ('val_precision_score', [0.8314340710639954, 0.8028166890144348, 0.7271778583526611, 0.7253068685531616, 0.8002657890319824, 0.8516638278961182, 0.8925684690475464, 0.894243061542511, 0.8539994955062866, 0.7962125539779663, 0.902099609375, 0.8477975130081177, 0.8647767901420593, 0.8664861917495728, 0.8666577935218811, 0.8662843704223633, 0.8814059495925903, 0.83721923828125, 0.8375463485717773, 0.9057286381721497]), ('val_f1_score', [0.6456575393676758, 0.7870603203773499, 0.8023953437805176, 0.8061305284500122, 0.8224501013755798, 0.8386290669441223, 0.8095747232437134, 0.8270460367202759, 0.8515258431434631, 0.8537725210189819, 0.8346191048622131, 0.8626083731651306, 0.8650293350219727, 0.8633373379707336, 0.8691380620002747, 0.8657810688018799, 0.8577888607978821, 0.8663791418075562, 0.8677952289581299, 0.851726770401001])])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history1.history.items()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_graphs(history, metric):\n",
    "  x=[int(i+1) for i in range(len(history.history[metric]))]\n",
    "  plt.plot(x, history.history[metric], 'r', label='training_'+metric)\n",
    "  plt.plot(x, history.history['val_'+metric], 'b', label='validation_'+metric)\n",
    "  plt.xlabel(\"Epochs\")\n",
    "  plt.ylabel(metric)\n",
    "  plt.legend(loc=[0.49, 0.55])\n",
    "  plt.show()\n",
    "  \n",
    "#check your metrics  \n",
    "plot_graphs(history1, 'accuracy')\n",
    "plot_graphs(history1, 'loss')\n",
    "plot_graphs(history1, 'recall_score')\n",
    "plot_graphs(history1, 'precision_score')\n",
    "plot_graphs(history1, 'f1_score')\n",
    "plot_graphs(history1, 'auc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 score is 0.863 for training, and 0.863 for validation \n",
      "AUC-ROC is 0.941 for training, and 0.939 for validation \n"
     ]
    }
   ],
   "source": [
    "print('F1 score is %.3f for training, and %.3f for validation ' \n",
    "      % (history1.history['f1_score'][11], history1.history['val_f1_score'][11]))\n",
    "\n",
    "print('AUC-ROC is %.3f for training, and %.3f for validation ' \n",
    "      % (history1.history['auc'][11], history1.history['val_auc'][11]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Build a model of sentiment analysi by use of the pre-trained GloVe model__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedded_words = {}\n",
    "with open('./data/glove.6B.200d.txt') as file:\n",
    "    for line in file:\n",
    "        words, coeff = line.split(maxsplit=1)\n",
    "        coeff = np.array(coeff.split(),dtype = float)\n",
    "        embedded_words[words] = coeff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedding_matrix2 = np.zeros((len(word_index) + 1,200))\n",
    "for word, i in word_index.items():\n",
    "    embedding_vector = embedded_words.get(word)\n",
    "    if embedding_vector is not None:\n",
    "        embedding_matrix2[i] = embedding_vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "model2 = tf.keras.Sequential([tf.keras.layers.Embedding(len(word_index) + 1, 200,\n",
    "                                                        weights=[embedding_matrix2],\n",
    "                                                        input_length=200,\n",
    "                                                        trainable=False),\n",
    "                             tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(64)),\n",
    "                             tf.keras.layers.Dropout(0.5),\n",
    "                             tf.keras.layers.Dense(256,activation = 'relu',),\n",
    "                             tf.keras.layers.Dense(128,activation = 'relu'),\n",
    "                             tf.keras.layers.Dropout(0.5),\n",
    "                             tf.keras.layers.Dense(1,activation = tf.nn.sigmoid)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_1 (Embedding)      (None, 200, 200)          22455600  \n",
      "_________________________________________________________________\n",
      "bidirectional_1 (Bidirection (None, 128)               135680    \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 256)               33024     \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 128)               32896     \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 22,657,329\n",
      "Trainable params: 201,729\n",
      "Non-trainable params: 22,455,600\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model2.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "1250/1250 [==============================] - 77s 60ms/step - loss: 0.5569 - accuracy: 0.6985 - auc_1: 0.7820 - recall_score: 0.6959 - precision_score: 0.7082 - f1_score: 0.6786 - val_loss: 0.3920 - val_accuracy: 0.8185 - val_auc_1: 0.9104 - val_recall_score: 0.9006 - val_precision_score: 0.7722 - val_f1_score: 0.8275\n",
      "Epoch 2/20\n",
      "1250/1250 [==============================] - 77s 61ms/step - loss: 0.3710 - accuracy: 0.8396 - auc_1: 0.9161 - recall_score: 0.8422 - precision_score: 0.8398 - f1_score: 0.8352 - val_loss: 0.3548 - val_accuracy: 0.8475 - val_auc_1: 0.9328 - val_recall_score: 0.7781 - val_precision_score: 0.9025 - val_f1_score: 0.8311\n",
      "Epoch 3/20\n",
      "1250/1250 [==============================] - 76s 61ms/step - loss: 0.3336 - accuracy: 0.8597 - auc_1: 0.9322 - recall_score: 0.8645 - precision_score: 0.8577 - f1_score: 0.8561 - val_loss: 0.3173 - val_accuracy: 0.8670 - val_auc_1: 0.9409 - val_recall_score: 0.8879 - val_precision_score: 0.8515 - val_f1_score: 0.8658\n",
      "Epoch 4/20\n",
      "1250/1250 [==============================] - 77s 61ms/step - loss: 0.3086 - accuracy: 0.8730 - auc_1: 0.9422 - recall_score: 0.8810 - precision_score: 0.8688 - f1_score: 0.8701 - val_loss: 0.3143 - val_accuracy: 0.8659 - val_auc_1: 0.9435 - val_recall_score: 0.8226 - val_precision_score: 0.9001 - val_f1_score: 0.8558\n",
      "Epoch 5/20\n",
      "1250/1250 [==============================] - 78s 62ms/step - loss: 0.2872 - accuracy: 0.8787 - auc_1: 0.9499 - recall_score: 0.8830 - precision_score: 0.8785 - f1_score: 0.8758 - val_loss: 0.2955 - val_accuracy: 0.8743 - val_auc_1: 0.9472 - val_recall_score: 0.8956 - val_precision_score: 0.8577 - val_f1_score: 0.8731\n",
      "Epoch 6/20\n",
      "1250/1250 [==============================] - 79s 63ms/step - loss: 0.2649 - accuracy: 0.8911 - auc_1: 0.9574 - recall_score: 0.8987 - precision_score: 0.8873 - f1_score: 0.8890 - val_loss: 0.3011 - val_accuracy: 0.8797 - val_auc_1: 0.9477 - val_recall_score: 0.8868 - val_precision_score: 0.8734 - val_f1_score: 0.8768\n",
      "Epoch 7/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.2474 - accuracy: 0.8991 - auc_1: 0.9627 - recall_score: 0.9050 - precision_score: 0.8969 - f1_score: 0.8971 - val_loss: 0.3007 - val_accuracy: 0.8775 - val_auc_1: 0.9496 - val_recall_score: 0.8574 - val_precision_score: 0.8925 - val_f1_score: 0.8712\n",
      "Epoch 8/20\n",
      "1250/1250 [==============================] - 79s 63ms/step - loss: 0.2259 - accuracy: 0.9081 - auc_1: 0.9689 - recall_score: 0.9143 - precision_score: 0.9045 - f1_score: 0.9057 - val_loss: 0.3077 - val_accuracy: 0.8761 - val_auc_1: 0.9499 - val_recall_score: 0.9147 - val_precision_score: 0.8483 - val_f1_score: 0.8771\n",
      "Epoch 9/20\n",
      "1250/1250 [==============================] - 76s 61ms/step - loss: 0.2044 - accuracy: 0.9179 - auc_1: 0.9743 - recall_score: 0.9241 - precision_score: 0.9140 - f1_score: 0.9161 - val_loss: 0.3383 - val_accuracy: 0.8773 - val_auc_1: 0.9480 - val_recall_score: 0.9134 - val_precision_score: 0.8508 - val_f1_score: 0.8777\n",
      "Epoch 10/20\n",
      "1250/1250 [==============================] - 78s 62ms/step - loss: 0.1850 - accuracy: 0.9270 - auc_1: 0.9788 - recall_score: 0.9334 - precision_score: 0.9225 - f1_score: 0.9251 - val_loss: 0.3234 - val_accuracy: 0.8752 - val_auc_1: 0.9456 - val_recall_score: 0.8937 - val_precision_score: 0.8605 - val_f1_score: 0.8737\n",
      "Epoch 11/20\n",
      "1250/1250 [==============================] - 76s 61ms/step - loss: 0.1633 - accuracy: 0.9373 - auc_1: 0.9833 - recall_score: 0.9421 - precision_score: 0.9351 - f1_score: 0.9363 - val_loss: 0.3480 - val_accuracy: 0.8750 - val_auc_1: 0.9428 - val_recall_score: 0.8588 - val_precision_score: 0.8867 - val_f1_score: 0.8689\n",
      "Epoch 12/20\n",
      "1250/1250 [==============================] - 75s 60ms/step - loss: 0.1480 - accuracy: 0.9430 - auc_1: 0.9861 - recall_score: 0.9474 - precision_score: 0.9397 - f1_score: 0.9413 - val_loss: 0.3576 - val_accuracy: 0.8751 - val_auc_1: 0.9418 - val_recall_score: 0.8628 - val_precision_score: 0.8835 - val_f1_score: 0.8696\n",
      "Epoch 13/20\n",
      "1250/1250 [==============================] - 75s 60ms/step - loss: 0.1295 - accuracy: 0.9512 - auc_1: 0.9890 - recall_score: 0.9545 - precision_score: 0.9494 - f1_score: 0.9501 - val_loss: 0.3956 - val_accuracy: 0.8721 - val_auc_1: 0.9389 - val_recall_score: 0.9006 - val_precision_score: 0.8514 - val_f1_score: 0.8719\n",
      "Epoch 14/20\n",
      "1250/1250 [==============================] - 75s 60ms/step - loss: 0.1143 - accuracy: 0.9577 - auc_1: 0.9912 - recall_score: 0.9610 - precision_score: 0.9554 - f1_score: 0.9566 - val_loss: 0.4404 - val_accuracy: 0.8777 - val_auc_1: 0.9379 - val_recall_score: 0.8697 - val_precision_score: 0.8807 - val_f1_score: 0.8722\n",
      "Epoch 15/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.1011 - accuracy: 0.9625 - auc_1: 0.9931 - recall_score: 0.9651 - precision_score: 0.9603 - f1_score: 0.9614 - val_loss: 0.4573 - val_accuracy: 0.8746 - val_auc_1: 0.9345 - val_recall_score: 0.8789 - val_precision_score: 0.8701 - val_f1_score: 0.8712\n",
      "Epoch 16/20\n",
      "1250/1250 [==============================] - 79s 63ms/step - loss: 0.0882 - accuracy: 0.9668 - auc_1: 0.9946 - recall_score: 0.9684 - precision_score: 0.9658 - f1_score: 0.9659 - val_loss: 0.5632 - val_accuracy: 0.8657 - val_auc_1: 0.9291 - val_recall_score: 0.8300 - val_precision_score: 0.8932 - val_f1_score: 0.8567\n",
      "Epoch 17/20\n",
      "1250/1250 [==============================] - 76s 61ms/step - loss: 0.0771 - accuracy: 0.9718 - auc_1: 0.9956 - recall_score: 0.9736 - precision_score: 0.9704 - f1_score: 0.9709 - val_loss: 0.5672 - val_accuracy: 0.8753 - val_auc_1: 0.9296 - val_recall_score: 0.8953 - val_precision_score: 0.8604 - val_f1_score: 0.8740\n",
      "Epoch 18/20\n",
      "1250/1250 [==============================] - 77s 62ms/step - loss: 0.0719 - accuracy: 0.9746 - auc_1: 0.9961 - recall_score: 0.9758 - precision_score: 0.9733 - f1_score: 0.9736 - val_loss: 0.4993 - val_accuracy: 0.8685 - val_auc_1: 0.9292 - val_recall_score: 0.8592 - val_precision_score: 0.8733 - val_f1_score: 0.8628\n",
      "Epoch 19/20\n",
      "1250/1250 [==============================] - 78s 63ms/step - loss: 0.0673 - accuracy: 0.9762 - auc_1: 0.9966 - recall_score: 0.9769 - precision_score: 0.9757 - f1_score: 0.9754 - val_loss: 0.5809 - val_accuracy: 0.8693 - val_auc_1: 0.9270 - val_recall_score: 0.8812 - val_precision_score: 0.8597 - val_f1_score: 0.8667\n",
      "Epoch 20/20\n",
      "1250/1250 [==============================] - 77s 61ms/step - loss: 0.0596 - accuracy: 0.9793 - auc_1: 0.9972 - recall_score: 0.9808 - precision_score: 0.9784 - f1_score: 0.9789 - val_loss: 0.6265 - val_accuracy: 0.8686 - val_auc_1: 0.9208 - val_recall_score: 0.8679 - val_precision_score: 0.8676 - val_f1_score: 0.8646\n"
     ]
    }
   ],
   "source": [
    "model2.compile(loss = tf.keras.losses.BinaryCrossentropy() , \n",
    "               optimizer=tf.keras.optimizers.Adam(learning_rate=0.001) , \n",
    "               metrics = ['accuracy',\n",
    "                           tf.keras.metrics.AUC(), \n",
    "#                            tf.keras.metrics.Precision(), \n",
    "#                            tf.keras.metrics.Recall(),\n",
    "                           recall_score,\n",
    "                           precision_score,\n",
    "                           f1_score])\n",
    "\n",
    "history2 = model2.fit(train_pad_sequence,training_labels,\n",
    "                      epochs = 20,\n",
    "                      validation_data=(test_pad_sequence,testing_labels),\n",
    "                      verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_items([('loss', [0.5568873286247253, 0.37103545665740967, 0.3335936665534973, 0.3086208701133728, 0.28715792298316956, 0.26489579677581787, 0.24737079441547394, 0.22585712373256683, 0.20437242090702057, 0.18502791225910187, 0.16328510642051697, 0.14801697432994843, 0.12945814430713654, 0.11429859697818756, 0.10112541168928146, 0.08816491067409515, 0.07706194370985031, 0.07191339135169983, 0.06725174933671951, 0.05956510454416275]), ('accuracy', [0.6984999775886536, 0.8396250009536743, 0.8597249984741211, 0.8729749917984009, 0.8786500096321106, 0.8910999894142151, 0.8991000056266785, 0.9081249833106995, 0.9178749918937683, 0.9270250201225281, 0.9373499751091003, 0.9429749846458435, 0.9511749744415283, 0.9577249884605408, 0.9625499844551086, 0.9667750000953674, 0.9718499779701233, 0.9746249914169312, 0.9761999845504761, 0.9793499708175659]), ('auc_1', [0.782046377658844, 0.9160861968994141, 0.9322227835655212, 0.9421548843383789, 0.9498872756958008, 0.9574301242828369, 0.9626523852348328, 0.968936562538147, 0.9743216633796692, 0.9788113832473755, 0.9832861423492432, 0.9860643744468689, 0.9890462756156921, 0.9912227392196655, 0.9930837750434875, 0.9945635199546814, 0.9956039190292358, 0.9961244463920593, 0.9965901374816895, 0.9971963763237]), ('recall_score', [0.6959330439567566, 0.8422475457191467, 0.8645443320274353, 0.8809967041015625, 0.8829940557479858, 0.8986895680427551, 0.9050202369689941, 0.9142795205116272, 0.9241194128990173, 0.9333528280258179, 0.9421230554580688, 0.9473538994789124, 0.9544867873191833, 0.9609604477882385, 0.9651274681091309, 0.9684395790100098, 0.9735826253890991, 0.9757794737815857, 0.9768759608268738, 0.9808178544044495]), ('precision_score', [0.7081915736198425, 0.8398033976554871, 0.8576616048812866, 0.8688201904296875, 0.8784898519515991, 0.887299120426178, 0.8969110250473022, 0.9045253992080688, 0.913986325263977, 0.9224992394447327, 0.9350500106811523, 0.9397319555282593, 0.9493662118911743, 0.9554312229156494, 0.9603378772735596, 0.9657678604125977, 0.9703770279884338, 0.9732667207717896, 0.9757370352745056, 0.97842937707901]), ('f1_score', [0.6785714030265808, 0.8351807594299316, 0.856134295463562, 0.8701121211051941, 0.875774621963501, 0.8889891505241394, 0.8970699906349182, 0.9057288765907288, 0.9161167144775391, 0.925079882144928, 0.936333179473877, 0.9413271546363831, 0.9500657320022583, 0.9566186666488647, 0.9614274501800537, 0.9658727645874023, 0.9709382057189941, 0.9736063480377197, 0.9754011034965515, 0.9789491891860962]), ('val_loss', [0.3920314908027649, 0.35477298498153687, 0.31733307242393494, 0.31428131461143494, 0.29545217752456665, 0.3010841906070709, 0.30067873001098633, 0.3076881468296051, 0.33834323287010193, 0.32338765263557434, 0.3480181396007538, 0.3576401472091675, 0.3956393897533417, 0.4403500556945801, 0.4573335349559784, 0.5631855726242065, 0.5671882033348083, 0.49929726123809814, 0.5808500051498413, 0.6264825463294983]), ('val_accuracy', [0.8184999823570251, 0.8475000262260437, 0.8669999837875366, 0.8658999800682068, 0.8743000030517578, 0.8797000050544739, 0.8774999976158142, 0.8761000037193298, 0.8773000240325928, 0.8751999735832214, 0.875, 0.8751000165939331, 0.8720999956130981, 0.8776999711990356, 0.8745999932289124, 0.8657000064849854, 0.8752999901771545, 0.8684999942779541, 0.8693000078201294, 0.8686000108718872]), ('val_auc_1', [0.9103906750679016, 0.932833731174469, 0.9409473538398743, 0.9434992671012878, 0.9472312927246094, 0.9476523399353027, 0.9496468901634216, 0.9499102234840393, 0.9479884505271912, 0.9456104636192322, 0.9428055286407471, 0.9418394565582275, 0.9388582110404968, 0.9378865361213684, 0.9345093965530396, 0.9291040897369385, 0.9296130537986755, 0.9291865825653076, 0.9270239472389221, 0.920793890953064]), ('val_recall_score', [0.9006332159042358, 0.7780930399894714, 0.8878790140151978, 0.8226096034049988, 0.895630955696106, 0.8868409991264343, 0.8573983907699585, 0.9147118330001831, 0.9133797287940979, 0.8937410116195679, 0.858778178691864, 0.8627509474754333, 0.9005734324455261, 0.8697106838226318, 0.8789252638816833, 0.8299726247787476, 0.8952509760856628, 0.8592029213905334, 0.8811860084533691, 0.8679176568984985]), ('val_precision_score', [0.7722494602203369, 0.9025090336799622, 0.8514767289161682, 0.9000884890556335, 0.8576664924621582, 0.8734341263771057, 0.8924981951713562, 0.848326563835144, 0.850764274597168, 0.8605354428291321, 0.8866945505142212, 0.8835131525993347, 0.8514355421066284, 0.8806754946708679, 0.8700613975524902, 0.8931728005409241, 0.8603687286376953, 0.873317539691925, 0.8596723675727844, 0.8676462173461914]), ('val_f1_score', [0.8275022506713867, 0.8311088681221008, 0.8658257126808167, 0.8557786345481873, 0.8731249570846558, 0.8768188953399658, 0.8711955547332764, 0.8770546317100525, 0.8777287602424622, 0.8736613392829895, 0.868938684463501, 0.8696471452713013, 0.8718916773796082, 0.8721899390220642, 0.8711868524551392, 0.8567284345626831, 0.8740206956863403, 0.8627966642379761, 0.8666636347770691, 0.8645787835121155])])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history2.history.items()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#check your metrics  \n",
    "plot_graphs(history2, 'accuracy')\n",
    "plot_graphs(history2, 'loss')\n",
    "plot_graphs(history2, 'recall_score')\n",
    "plot_graphs(history2, 'precision_score')\n",
    "plot_graphs(history2, 'f1_score')\n",
    "plot_graphs(history2, 'auc_1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 score is 0.876 for training, and 0.873 for validation \n",
      "AUC-ROC is 0.950 for training, and 0.947 for validation \n"
     ]
    }
   ],
   "source": [
    "print('F1 score is %.3f for training, and %.3f for validation ' \n",
    "      % (history2.history['f1_score'][4], history2.history['val_f1_score'][4]))\n",
    "\n",
    "print('AUC-ROC is %.3f for training, and %.3f for validation ' \n",
    "      % (history2.history['auc_1'][4], history2.history['val_auc_1'][4]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Build a model of sentiment analysis by use of custome embedding (i.e. not pre-trained)__\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "model3 = tf.keras.Sequential([tf.keras.layers.Embedding(len(word_index) + 1, 250, input_length=200),\n",
    "                             tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(64)),\n",
    "                             tf.keras.layers.Dropout(0.5),\n",
    "                             tf.keras.layers.Dense(256,activation = 'relu',),\n",
    "                             tf.keras.layers.Dense(128,activation = 'relu'),\n",
    "                             tf.keras.layers.Dropout(0.5),\n",
    "                             tf.keras.layers.Dense(1,activation = tf.nn.sigmoid)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_2 (Embedding)      (None, 200, 250)          28069500  \n",
      "_________________________________________________________________\n",
      "bidirectional_2 (Bidirection (None, 128)               161280    \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 256)               33024     \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 128)               32896     \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 28,296,829\n",
      "Trainable params: 28,296,829\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model3.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "1250/1250 [==============================] - 454s 361ms/step - loss: 0.4505 - accuracy: 0.7894 - auc_2: 0.8723 - recall_score: 0.8133 - precision_score: 0.7910 - f1_score: 0.7867 - val_loss: 0.3429 - val_accuracy: 0.8494 - val_auc_2: 0.9386 - val_recall_score: 0.7578 - val_precision_score: 0.9306 - val_f1_score: 0.8298\n",
      "Epoch 2/20\n",
      "1250/1250 [==============================] - 449s 359ms/step - loss: 0.2869 - accuracy: 0.8893 - auc_2: 0.9495 - recall_score: 0.8946 - precision_score: 0.8873 - f1_score: 0.8864 - val_loss: 0.3260 - val_accuracy: 0.8777 - val_auc_2: 0.9467 - val_recall_score: 0.9073 - val_precision_score: 0.8564 - val_f1_score: 0.8778\n",
      "Epoch 3/20\n",
      "1250/1250 [==============================] - 447s 358ms/step - loss: 0.2208 - accuracy: 0.9168 - auc_2: 0.9697 - recall_score: 0.9225 - precision_score: 0.9139 - f1_score: 0.9148 - val_loss: 0.3189 - val_accuracy: 0.8764 - val_auc_2: 0.9439 - val_recall_score: 0.8874 - val_precision_score: 0.8684 - val_f1_score: 0.8744\n",
      "Epoch 4/20\n",
      "1250/1250 [==============================] - 448s 358ms/step - loss: 0.1650 - accuracy: 0.9402 - auc_2: 0.9825 - recall_score: 0.9430 - precision_score: 0.9384 - f1_score: 0.9385 - val_loss: 0.3472 - val_accuracy: 0.8708 - val_auc_2: 0.9362 - val_recall_score: 0.8777 - val_precision_score: 0.8658 - val_f1_score: 0.8679\n",
      "Epoch 5/20\n",
      "1250/1250 [==============================] - 445s 356ms/step - loss: 0.1312 - accuracy: 0.9535 - auc_2: 0.9885 - recall_score: 0.9588 - precision_score: 0.9505 - f1_score: 0.9527 - val_loss: 0.3974 - val_accuracy: 0.8655 - val_auc_2: 0.9305 - val_recall_score: 0.8789 - val_precision_score: 0.8570 - val_f1_score: 0.8639\n",
      "Epoch 6/20\n",
      "1250/1250 [==============================] - 444s 355ms/step - loss: 0.0903 - accuracy: 0.9687 - auc_2: 0.9942 - recall_score: 0.9718 - precision_score: 0.9663 - f1_score: 0.9679 - val_loss: 0.4364 - val_accuracy: 0.8534 - val_auc_2: 0.9240 - val_recall_score: 0.8641 - val_precision_score: 0.8447 - val_f1_score: 0.8507\n",
      "Epoch 7/20\n",
      "1250/1250 [==============================] - 439s 351ms/step - loss: 0.0672 - accuracy: 0.9772 - auc_2: 0.9962 - recall_score: 0.9780 - precision_score: 0.9765 - f1_score: 0.9764 - val_loss: 0.5214 - val_accuracy: 0.8730 - val_auc_2: 0.9281 - val_recall_score: 0.8824 - val_precision_score: 0.8679 - val_f1_score: 0.8715\n",
      "Epoch 8/20\n",
      "1250/1250 [==============================] - 433s 346ms/step - loss: 0.0531 - accuracy: 0.9822 - auc_2: 0.9974 - recall_score: 0.9830 - precision_score: 0.9817 - f1_score: 0.9817 - val_loss: 0.7048 - val_accuracy: 0.8717 - val_auc_2: 0.9184 - val_recall_score: 0.8743 - val_precision_score: 0.8709 - val_f1_score: 0.8692\n",
      "Epoch 9/20\n",
      "1250/1250 [==============================] - 441s 353ms/step - loss: 0.0463 - accuracy: 0.9853 - auc_2: 0.9979 - recall_score: 0.9864 - precision_score: 0.9846 - f1_score: 0.9849 - val_loss: 0.6521 - val_accuracy: 0.8659 - val_auc_2: 0.9180 - val_recall_score: 0.8634 - val_precision_score: 0.8688 - val_f1_score: 0.8621\n",
      "Epoch 10/20\n",
      "1250/1250 [==============================] - 450s 360ms/step - loss: 0.0339 - accuracy: 0.9891 - auc_2: 0.9986 - recall_score: 0.9893 - precision_score: 0.9889 - f1_score: 0.9887 - val_loss: 0.7807 - val_accuracy: 0.8639 - val_auc_2: 0.9101 - val_recall_score: 0.8513 - val_precision_score: 0.8735 - val_f1_score: 0.8586\n",
      "Epoch 11/20\n",
      "1250/1250 [==============================] - 449s 359ms/step - loss: 0.0304 - accuracy: 0.9900 - auc_2: 0.9988 - recall_score: 0.9900 - precision_score: 0.9904 - f1_score: 0.9898 - val_loss: 0.7882 - val_accuracy: 0.8659 - val_auc_2: 0.9136 - val_recall_score: 0.8564 - val_precision_score: 0.8725 - val_f1_score: 0.8607\n",
      "Epoch 12/20\n",
      "1250/1250 [==============================] - 440s 352ms/step - loss: 0.0333 - accuracy: 0.9892 - auc_2: 0.9988 - recall_score: 0.9908 - precision_score: 0.9878 - f1_score: 0.9888 - val_loss: 0.7346 - val_accuracy: 0.8636 - val_auc_2: 0.9150 - val_recall_score: 0.8620 - val_precision_score: 0.8659 - val_f1_score: 0.8601\n",
      "Epoch 13/20\n",
      "1250/1250 [==============================] - 442s 354ms/step - loss: 0.0233 - accuracy: 0.9933 - auc_2: 0.9991 - recall_score: 0.9932 - precision_score: 0.9932 - f1_score: 0.9930 - val_loss: 0.8793 - val_accuracy: 0.8447 - val_auc_2: 0.8998 - val_recall_score: 0.8118 - val_precision_score: 0.8698 - val_f1_score: 0.8350\n",
      "Epoch 14/20\n",
      "1250/1250 [==============================] - 444s 355ms/step - loss: 0.0247 - accuracy: 0.9931 - auc_2: 0.9990 - recall_score: 0.9933 - precision_score: 0.9929 - f1_score: 0.9929 - val_loss: 0.7188 - val_accuracy: 0.8500 - val_auc_2: 0.9090 - val_recall_score: 0.8453 - val_precision_score: 0.8542 - val_f1_score: 0.8451\n",
      "Epoch 15/20\n",
      "1250/1250 [==============================] - 442s 354ms/step - loss: 0.0185 - accuracy: 0.9948 - auc_2: 0.9992 - recall_score: 0.9946 - precision_score: 0.9947 - f1_score: 0.9945 - val_loss: 0.9202 - val_accuracy: 0.8633 - val_auc_2: 0.9080 - val_recall_score: 0.8700 - val_precision_score: 0.8599 - val_f1_score: 0.8608\n",
      "Epoch 16/20\n",
      "1250/1250 [==============================] - 442s 354ms/step - loss: 0.0172 - accuracy: 0.9954 - auc_2: 0.9993 - recall_score: 0.9953 - precision_score: 0.9955 - f1_score: 0.9952 - val_loss: 0.8673 - val_accuracy: 0.8579 - val_auc_2: 0.9057 - val_recall_score: 0.8536 - val_precision_score: 0.8615 - val_f1_score: 0.8536\n",
      "Epoch 17/20\n",
      "1250/1250 [==============================] - 443s 354ms/step - loss: 0.0172 - accuracy: 0.9950 - auc_2: 0.9994 - recall_score: 0.9954 - precision_score: 0.9945 - f1_score: 0.9947 - val_loss: 0.9443 - val_accuracy: 0.8536 - val_auc_2: 0.9016 - val_recall_score: 0.8611 - val_precision_score: 0.8494 - val_f1_score: 0.8509\n",
      "Epoch 18/20\n",
      "1250/1250 [==============================] - 450s 360ms/step - loss: 0.0153 - accuracy: 0.9953 - auc_2: 0.9995 - recall_score: 0.9953 - precision_score: 0.9951 - f1_score: 0.9950 - val_loss: 0.8765 - val_accuracy: 0.8596 - val_auc_2: 0.9084 - val_recall_score: 0.8699 - val_precision_score: 0.8534 - val_f1_score: 0.8579\n",
      "Epoch 19/20\n",
      "1250/1250 [==============================] - 452s 362ms/step - loss: 0.0153 - accuracy: 0.9955 - auc_2: 0.9995 - recall_score: 0.9956 - precision_score: 0.9954 - f1_score: 0.9953 - val_loss: 0.8980 - val_accuracy: 0.8588 - val_auc_2: 0.9065 - val_recall_score: 0.8289 - val_precision_score: 0.8836 - val_f1_score: 0.8512\n",
      "Epoch 20/20\n",
      "1250/1250 [==============================] - 447s 358ms/step - loss: 0.0114 - accuracy: 0.9965 - auc_2: 0.9997 - recall_score: 0.9964 - precision_score: 0.9967 - f1_score: 0.9964 - val_loss: 0.8171 - val_accuracy: 0.8605 - val_auc_2: 0.9098 - val_recall_score: 0.8915 - val_precision_score: 0.8408 - val_f1_score: 0.8617\n"
     ]
    }
   ],
   "source": [
    "model3.compile(loss = tf.keras.losses.BinaryCrossentropy() , \n",
    "               optimizer=tf.keras.optimizers.Adam(learning_rate=0.001) , \n",
    "               metrics = ['accuracy',\n",
    "                           tf.keras.metrics.AUC(), \n",
    "#                            tf.keras.metrics.Precision(), \n",
    "#                            tf.keras.metrics.Recall(),\n",
    "                           recall_score,\n",
    "                           precision_score,\n",
    "                           f1_score])\n",
    "\n",
    "history3 = model3.fit(train_pad_sequence,training_labels,\n",
    "                      epochs = 20 ,\n",
    "                      validation_data=(test_pad_sequence,testing_labels),\n",
    "                      verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deZzN9f7A8dfb2JeQGYylqKhUDIYU0XKvpCIly21BlptIVEq/SrTctFw3SrQJ3UoqotJGulKUoSFRWSJbGrKvM+b9++NzhjHOzJyZOd9zzsx5Px+P72PO+X6/5/t9zzHO+3x2UVWMMcZEr2LhDsAYY0x4WSIwxpgoZ4nAGGOinCUCY4yJcpYIjDEmyhUPdwB5FRsbq3Xq1Al3GMYYU6gsWbJku6rG+TtW6BJBnTp1SEpKCncYxhhTqIjIhuyOWdWQMcZEOUsExhgT5SwRGGNMlLNEYIwxUc4SgTHGRDlLBMYYE+UsERhjTJSzRGCMKfIOH4aXXoJt28IdSWSyRGCMKdKOHIGuXeH22+Gii+DXX8MdUeSxRGCMKbJSU6F7d5g5E+65B/buhZYt4bvvwh1ZZLFEYIwpktLS4OabYfp0+M9/4Nln4dtvoUIFuPxymD073BFGDksExpgi5+hR6NEDpk2DZ56BwYPd/nr1XDI4+2zo0AFefz28cUYKSwTGGM+pwscfw5493t/r6FHo1QveeguefBLuvffE49Wrw//+50oFt90Gjz/u4otmlgiMMZ776CO45hpo1Ai+/tq7+6SnQ9++8MYb8OijMGyY//MqVHAx3XQTPPwwDBjgEki0skRgjPHc669DbCwUKwZt2sD//Z/rzRNM6emuZ9Drr8Pw4e4DPiclS8KUKTB0KIwfDzfeCAcPBjemwsISgTHGUykp8OGHrs4+OdlV2zz5pOvKuWpVcO6hCgMHwiuvwAMPwIgRgb2uWDF4+ml47jn44ANo2xZ27gxOTIWJJQJjjKfeftv14OnRw1XJvPaa68mzYQM0aQLPP1+wOnpV1xg8frz7dv/EEyCSt2vcdZeL8/vvoVUr2Lgx//EURpYIjDGemjQJmjaFCy44vq9TJ/jxR7jsMhg0CK66CrZsyfu1VV1j8NixLhk89VTek0CGrl3h009h0yZXWlmxIn/XyWr9epgwAW65BZYsCc41g80SgTHGM8uWwQ8/QM+eJx+Lj3c9icaNg/nzXaJ4//3Ar63qGoNHj3bVQqNH5z8JZLjsMhdLerorGcyfn/drHDoEn38OQ4bAuedC3brQvz9MneqqnlauLFiMXrBEYIzxzOTJUKKEG93rjwjccYdLFmecAZ07u6SRWzdTVdcY/PTTroF47NiCJ4EMjRrBwoWum2nbtoElpzVrXBXX1VfDqafClVe6qqrTT3eD2X7+2W0lS8Lf/w6//RacWINGVQvV1rRpUzXGRL4jR1Tj4lRvuCHw8x96SLVYMdU6dVTnz8/+3BEjVEG1Tx/Vo0eDE29W27erXnSRqojq88+feGz/ftWPP1YdOFD1rLNcLOAe33mn6uzZ7pysli9XrVxZ9cwzVbds8Sbu7ABJms3natg/2PO6WSIwJu/++EO1e3fVjRtDd8+ZM90nzIcf5u1133yjesYZ7gP4gQdUDx8+8fhjj7nr9uzpXRLIsH+/6rXXuvvde6/q6NGqbduqlirl9pUpo9q+vUsUq1cHds2FC1XLlVM9/3zVHTu8jT8zSwTGRLk77nD/2/v3D909O3VSrVbNfdPPqz17VG+7zcXcpInqypVu/6hRbt/NN6umpQU33uykpqr27Xv8W/8556gOGaL62WeqBw/m75pz5qiWLKl64YWqe/cGN97sWCIwJoqtW6davLj7FlqqlCsdeO3PP90977mnYNeZMUM1Nla1dGlXogHVbt1ClwQypKerLlig+ttvwbvm9OmuGuyKK1QPHQredbOTUyKwxmJjiriRI6F4cddD58gRN3jKa5nHDhTEddcd72b69ttu9O8bb0BMTHDiDJSIm766Tp3gXbNTJ5g4EebOdY3paWnBu3ZeWSIwpghbudJ9cA4c6KZ26NwZXnwRdu/29r7+xg7kV/XqLol9/72bSK548YJfM1L06AFjxsCMGdCnj+u2Gg6WCIwpwoYPh3Ll4P773fNhw1zXzPHjvbtnTmMH8ksEmjUrWkkgw6BBbkqMyZPh7rvDMxOqJQJjiqglS1wf+HvucRO+gZvSoW1bVz3k1QRrkye7/vLZjR0wJxs+3E1zMWaMmzU11CwRGFNEPfQQVKniRrhm9sADbhH3SZOCf8/UVPjvf92iL1WqBP/6RZWIGxnds6crHYwZE9r7WyIwpgiaP9/NmzNsGJxyyonH2rSBCy90K3cFu4Fy9mw322gwq4WiRbFibvbU66938yZ5kaizvXfobmWMCQVVePBBqFHDLbiSlYgrFfz2m1vKMZgmTYJq1dwUCybvihd3DeJ/+xv07u0akUPBEoExRcxnn8GCBa5qqEwZ/+dcey00aACjRgWvcTIlxa36dcstRbNRN1RKlXIJoHlz6NYN5szx/p6WCIwpQtLT3epfdeu6b5TZKVbM9ST68UdXnRMMb70VnLEDBsqXd/8uZ5/txlIsWuTt/SwRGFOETJ/uum6OHOl67uSke3c47TS3WlgwTJoEiYlw/vnBuV60q1zZTWddvbpbr+HHH727lyUCY4qIo0fd1MwNGsA//pH7+SVKuEVdvvnGVSUVRHKy26yROLiqV3dVQ2XLum6/a9Z4cx9LBMYUEf/9r5vz/rHHAp+CoXdvN8agoKWCjLED3boV7DrmZHXqwBdfuET/9dfe3MPTRCAi7UTkFxFZIyLD/Bw/TUTmicgPIrJcRNp7GY8xRdXhw/DII25ah06dAn9d2bJuINPs2bB8ef7ufeSIjR3wWoMG8Ouv0KuXN9f3LBGISAwwDrgKaAB0F5EGWU57CJimqo2BbsCLXsVjTFH26qtuMfh//SvvK3UNGOAaJ0eNyt+9P/kEtm+3aiGvVark3bW9LBE0B9ao6jpVPQJMBTpmOUeBjOEuFYF8LF9tTHQ7cAAefxxat3bLIOZV5cpuucd33oF16/L+ehs7UPh5mQhqAhszPd/k25fZCOBmEdkEzAbu9HchEeknIkkikpSSkuJFrMYUWi+8AH/8AU88kf91e4cMcX3/n3kmb6+zsQNFg5eJwN+fZNahK92BSapaC2gPvCEiJ8Wkqi+raqKqJsbFxXkQqjHBt28frF7tpnuYOtV9c96/P7j32L3bVem0bw+tWuX/OjVquP7/r7/ukkqgbOxA0eBlDt8E1M70vBYnV/30BtoBqOpCESkNxAJ/ehiXMQWydy9s2QJbtx7/mflxxs+9e09+7ZNPug/Ppk2DE8u//w07d7qqoYK67z547TU3M2mg7QU2dqBo8DIRLAbqiUhdYDOuMThr7+bfgSuASSJyLlAasLofE3FWrnR989es8f+tvkwZiI9336wbNXIDgOLjj++Lj4dNm1x3zRYtXBfPoUMLttJWSgr85z9u1a7GjfN/nQxnnXV84Zphw3JvnMwYO/DCCwW/twkvzxKBqqaJyEDgMyAGmKiqP4nIo7i1M2cB9wCviMgQXLVRT9/amsZEDFXo3x82boS+fU/8cM/4WbFi7vXz553numjefrub9O3TT2HKFDe6Nz9GjXINxcGcv37YMDcR3fjxLsac2NiBIiS7xYwjdbPF602ovf22WzT9pZeCc730dNVJk1TLl1etWFF16tS8X2PjRrcQfa9ewYkps3btVKtWVT1wIPtzDh92i8p37hz8+xtvYIvXG5M/+/e7aRiaNMl5Ere8EHGNq8nJcO657hv1rbe6JSQD9fjjboK54cODE1Nmw4bBn3+6huPszJ5tYweKEksExuTgX/+CzZvh+ecLVp/vz5lnuikDHnkE3nwTEhLg229zf92aNa5R95//dNMPBFvr1nDRRTkvXDNpkpsHx8YOFA2WCIzJxpo18Oyzro/8xRd7c4/ixd3ShBlzyFxyiUsMOa0cNmKEmzDuwQe9iUnElQrWr3eDzLL680/4+GMbO1CUWCIwJht33+0aQ596yvt7XXyxqyq6+WbX+NuqFaxde/J5K1a47qd33eW+kXvlmmtc4/aoUa4KKjMbO1D0WCIwxo9PPoEPP3R18PHxobnnKae4njhTp7pZRBMSXBVM5n50Dz8MFSq4rqdeyli4ZsUK9+0/s0mToFkzlyhM0WCJwJgsjhxx37jr13c/Q61rV9fNtGlTN9tkly7w11/w/ffwwQcuCZx6qvdxdOsGp5/uBsFlJKPkZFi2zBqJixpLBMZk8dxzbmqIMWNyX+XLK6edBnPnug/hDz6Ahg3d+IO4uNAlp4yFaxYuPN6GMWmSjR0oiiwRGJPJli1u1O+110K7duGNJSbGNdouWgTlyrklKB94wFUNhcptt7nkM2qUKym9+SZ07BiaEokJHUsExmRy//2QmuqmbogUTZvC0qXw/vtwp9/5eb2TsXDNJ5+4rrQ2dqBoEi1kMzokJiZqUlJSuMMwRdA337jeOg8+GJxJ3IqKXbtcVdXeva6n0saN1m20MBKRJaqa6O+YlQiMwa0He+edUKtW7nPsRJtKldxcS2BjB4oq+yc1BrfU4w8/uK6b5cqFO5rIc889bvWyAQPCHYnxglUNmaj311+uq+j558O8eflf5cuYSGZVQ6bQW7vWzbv//ffBv/bw4W5xl7FjLQmY6GSJwBQKTz0F773npmIYOTLnuXjyYvlyN/d+//6ur74x0cgSgYl4u3e7/utdu0L37m7StZYt3aCvglB1DcSVKwd3cRdjChtrLDYRb8oUtxLX0KGuT/2117pRtgkJMHo09OuXvyqdd95xC8tPmJD/AVKpqals2rSJQ4cO5e8CJuRKly5NrVq1KFGiRLhDiRjWWGwimqqb3KxCBfjuu+P7N2928/B88QVcfbXr9ZOX2Tj374ezz4aqVWHx4vyvNfDbb79RoUIFqlSpglgDQ8RTVXbs2MHevXupW7duuMMJKWssNoXW//4Hq1Yd78eeoWZNt+bv2LFuTp4LLnBz8gQqWAvOHDp0yJJAISIiVKlSxUpwWVgiMBHtxRddHX7XricfK1bM1fEvWQK1a0OnTm45yb17c77m2rVuwZmbb3ZtDQVlSaBwsX+vk1kiMBFr61aYMcNVAZUpk/15DRq4idn+7//c7JiNGsGCBdmfP2RI6BacMaYwsERgItarr7puorffnvu5JUvCE0+4xl+ANm1cYjhy5MTzMhacefhhqFEj+DEbUxhZIjARKS0NXn4Z/v53qFcv8Ne1bHl84ZQnn4QWLWDlSnfsyBEYPNiNIh482JOwQ27Xrl28+OKLeX5d+/bt2bVrV47nDB8+nDlz5uQ3NL/Kly8f1OuZ4LDuoyYiffQRbNrkGnPzqkIFeO011820b1/X5fSpp+DgQfj1V5g9O3wLzgRbRiK44447Tth/9OhRYnJoBZ89e3au137UBldEDUsEJiKNH+9mAr3mmvxf47rrXImgT5/jq3pdey1cdVVwYjzJ4MFuLcdgSkhwS6ZlY9iwYaxdu5aEhARKlChB+fLliY+PJzk5mZUrV3LdddexceNGDh06xF133UW/fv0AqFOnDklJSezbt4+rrrqKVq1a8e2331KzZk1mzpxJmTJl6NmzJ9dccw2dO3emTp069OjRgw8//JDU1FTeffddzjnnHFJSUvjHP/7Bjh07aNasGZ9++ilLliwhNjY2x19LVbnvvvv45JNPEBEeeughunbtytatW+natSt79uwhLS2N8ePHc/HFF9O7d2+SkpIQEW677TaGDBkS1Lc52lnVkIk4q1fD55+7gWIFnfK4enXXJjBhgltwPZIWnAmGUaNGceaZZ5KcnMwzzzzD999/zxNPPMFKX33YxIkTWbJkCUlJSYwdO5YdO3acdI3Vq1czYMAAfvrpJypVqsT777/v916xsbEsXbqU/v378+yzzwIwcuRILr/8cpYuXUqnTp34/fffA4p7+vTpJCcns2zZMubMmcPQoUPZunUrb731FldeeeWxYwkJCSQnJ7N582ZWrFjBjz/+SK9evfL5bpnsWInARJyXXnIJoE+f4FxPBP75T7d5Kodv7qHSvHnzEwZKjR07lhkzZgCwceNGVq9eTZUqVU54Td26dUlISACgadOmrF+/3u+1r7/++mPnTJ8+HYAFCxYcu367du2oXLlyQHEuWLCA7t27ExMTQ7Vq1WjTpg2LFy+mWbNm3HbbbaSmpnLdddeRkJDAGWecwbp167jzzju5+uqradu2beBviAmIlQhMRDl4ECZOdGMC4uPDHU3hUy7TYgpfffUVc+bMYeHChSxbtozGjRv7HUhVqlSpY49jYmJIy2ZGv4zzMp+T35kJsntd69atmT9/PjVr1uSWW25hypQpVK5cmWXLlnHppZcybtw4+gTrG4I5xhKBiSjTprkpobOOJDb+VahQgb3ZjKDbvXs3lStXpmzZsvz8888sWrQo6Pdv1aoV06ZNA+Dzzz9n586dAb2udevWvPPOOxw9epSUlBTmz59P8+bN2bBhA1WrVqVv37707t2bpUuXsn37dtLT07nhhht47LHHWLp0adB/j2hnVUMmorz4IpxzDlx6abgjKRyqVKlCy5YtOf/88ylTpgzVqlU7dqxdu3ZMmDCBhg0bcvbZZ9OiRYug3/+RRx6he/fuvPPOO7Rp04b4+HgqVKiQ6+s6derEwoULadSoESLC008/TfXq1Zk8eTLPPPPMsYbvKVOmsHnzZnr16kV6ejoATz75ZNB/j2hnk86ZiLFkCSQmwpgxMGhQuKMJzKpVqzj33HPDHUbYHD58mJiYGIoXL87ChQvp378/ycHuOeWBaPx3y2nSOSsRmIgxfjyULQu33hruSEygfv/9d7p06UJ6ejolS5bklVdeCXdIJh8sEZiIsGsXvPUW3HQTVKoU7mhMoOrVq8cPP/xwwr4dO3ZwxRVXnHTu3LlzT+qxZCKDp4lARNoBY4AY4FVVHeXnnC7ACECBZar6Dy9jMpFp8mTXY8gaiQu/KlWqFIrqIXOcZ4lARGKAccDfgU3AYhGZpaorM51TD3gAaKmqO0WkqlfxmMil6qqFLrwQmjQJdzTGRB8vu482B9ao6jpVPQJMBTpmOacvME5VdwKo6p8exmMi1Lx58MsvkGW6HGNMiHiZCGoCGzM93+Tbl1l9oL6IfCMii3xVSScRkX4ikiQiSSkpKR6Fa8Jl/Hi3ZnCXLuGOxJjo5GUi8LcMUNa+qsWBesClQHfgVRE5qalQVV9W1URVTYyLiwt6oCZ8tmxxi8/cdhuULh3uaIyJTl4mgk1A7UzPawFb/JwzU1VTVfU34BdcYjBR4tVX4ejREMwDZIDj6wFs2bKFzp07+z3n0ksvJbexOs899xwHDhw49jyQ9Q3yomfPnrz33ntBu57JmZeJYDFQT0TqikhJoBswK8s5HwCXAYhILK6qaJ2HMZkIkrH4zJVXwllnhTua6FKjRo0CfdBmTQSzZ8+mkvX7LbQ86zWkqmkiMhD4DNd9dKKq/iQijwJJqjrLd6ytiKwEjgJDVfXkeXJNkfThh7B5M4wbF+5IgiMMyxFw//33c/rppx9bmGbEiBGICPPnz2fnzp2kpqby+OOP07Hjif001q9fzzXXXMOKFSs4ePAgvXr1YuXKlZx77rkcPHjw2Hn9+/dn8eLFHDx4kM6dOzNy5EjGjh3Lli1buOyyy4iNjWXevHnH1jeIjY1l9OjRTJw4EYA+ffowePBg1q9fn+26B7mZO3cu9957L2lpaTRr1ozx48dTqlQphg0bxqxZsyhevDht27bl2Wef5d1332XkyJHExMRQsWJF5mesXWpy5Ok4AlWdDczOsm94pscK3O3bTJR58UWoXRuuvjrckRRe3bp1Y/DgwccSwbRp0/j0008ZMmQIp5xyCtu3b6dFixZ06NABEX/NdjB+/HjKli3L8uXLWb58OU0y9eF94oknOPXUUzl69ChXXHEFy5cvZ9CgQYwePZp58+adtADNkiVLeP311/nuu+9QVS688ELatGlD5cqVWb16NW+//TavvPIKXbp04f333+fmm2/O8fc7dOgQPXv2ZO7cudSvX59bb72V8ePHc+uttzJjxgx+/vlnRORYtdSjjz7KZ599Rs2aNYNaVVXU2chiExa//gpz5sBjjxV88ZlIEY7lCBo3bsyff/7Jli1bSElJoXLlysTHxzNkyBDmz59PsWLF2Lx5M9u2baN69ep+rzF//nwG+SZ3atiwIQ0bNjx2bNq0abz88sukpaWxdetWVq5cecLxrBYsWECnTp2OTYd9/fXX8/XXX9OhQ4eA1z3I7JdffqFu3brUr18fgB49ejBu3DgGDhxI6dKl6dOnD1dffTXX+Jaya9myJT179qRLly7H1k8wubNpqE1YTJgQ3MVnolnnzp157733eOedd+jWrRtvvvkmKSkpLFmyhOTkZKpVq+Z3HYLM/JUWfvvtN5599lnmzp3L8uXLufrqq3O9Tk6TWAa67kEg1ytevDjff/89N9xwAx988AHt2rme5xMmTODxxx9n48aNJCQk+F2RzZzMEoEJuQMHYNIkuP56t5SkKZhu3boxdepU3nvvPTp37szu3bupWrUqJUqUYN68eWzYsCHH17du3Zo333wTgBUrVrB8+XIA9uzZQ7ly5ahYsSLbtm3jk08+Ofaa7NZBaN26NR988AEHDhxg//79zJgxg0suuSTfv9s555zD+vXrWbNmDQBvvPEGbdq0Yd++fezevZv27dvz3HPPHZvSYu3atVx44YU8+uijxMbGsnHjxpwub3yKSKHcFCbvvOMWn7GRxMFx3nnnsXfvXmrWrEl8fDw33XQT1157LYmJiSQkJHDOOefk+Pr+/fvTq1cvGjZsSEJCAs2bNwegUaNGNG7cmPPOO48zzjiDli1bHntNv379uOqqq4iPj2fevHnH9jdp0oSePXseu0afPn1o3LhxQNVA/pQuXZrXX3+dG2+88Vhj8e23385ff/1Fx44dOXToEKrKf3yLUQ8dOpTVq1ejqlxxxRU0atQoX/eNNrYegQm55s1h/35YscKtJ1yYReO89kVBNP675bQegVUNmZBKSoLFi90so4U9CRhTVFjVkAmpjMVnbrkl3JGYSDBgwAC++eabE/bddddd9OrVK0wRRSdLBCZkdu6Et992SaBixXBHEzyqmm0ffZOzcWEYTVjYqsNDwaqGTMgUxcVnSpcuzY4dO+zDpZBQVXbs2EFpm+HwBFYiMCGRsfjMRRe5aROKilq1arFp0yZsevTCo3Tp0tSqVSvcYUQUSwQmJN59140mnjIl3JEEV4kSJahbt264wzCmQKxqyHhu0SLo2ROaNbPFZ4yJRJYIjKd+/RWuuQZq1ICPPoJMswwYYyKEJQLjmW3boF07KFYMPv0UqlYNd0TGGH8CSgQicpeInCLOayKyVETaeh2cKbz27XPTS2/b5koCtvCMMZEr0BLBbaq6B2gLxAG9gFGeRWUKtdRUuPFGt0jLtGluSgljTOQKtNdQxmiZ9sDrqrpMbASN8UMV+vVzVUGvvmqLzhhTGARaIlgiIp/jEsFnIlIBSPcuLFNYDR/uppgeMQJ69w53NMaYQARaIugNJADrVPWAiJyKqx4y5piXXoLHH3eLzQwfnvv5xpjIEGiJ4CLgF1XdJSI3Aw8Bu70LyxQ2s2a59QXat3cjiK3i0JjCI9BEMB44ICKNgPuADUARGyNq8mvRIujWDZo2dY3DRWUNYmOiRaCJIE3drFodgTGqOgao4F1YprDIOmDMt2a5MaYQCfS7214ReQC4BbhERGKAEt6FZQqDP/6wAWPGFAWBlgi6Aodx4wn+AGoCz3gWlYl4e/ceHzD28cc2YMyYwiygROD78H8TqCgi1wCHVLXwtRGkpoY7giIhY8DYsmVuVtFmzcIdkTGmIAKdYqIL8D1wI9AF+E5EOnsZWNBNmAD16rmVUUy+qULfvvDZZ667aPv24Y7IGFNQgVYNPQg0U9Ueqnor0Bx42LuwPHDOObBhg1smy+Tbww+7t3DkSBswZkxREWgiKKaqf2Z6viMPr40Mbdq4SW+efRaOHg13NIXShAnwxBOuRPBw4foaYIzJQaAf5p+KyGci0lNEegIfA7O9C8sDInDffbB2LUyfHu5oCpW//oIHHoABA1wD8Ysv2oAxY4oSCXTRbRG5AWiJm4BuvqrO8DKw7CQmJmpSUlL+Xnz0KJx7LpxyCixebJ9mudizB557Dv79b9dLqHt3ePllGytgTGEkIktUNdHfsYCrd1T1fVW9W1WHhCsJFFhMDNx7LyxZAl99Fe5oItb+/fD001C3LjzyCFxxhesh9OablgSMKYpyTAQisldE9vjZ9orInlAFGVS33upGPj39dLgjiTiHDsHYsXDmmXD//XDhha7gNH06XHBBuKMzxnglx0SgqhVU9RQ/WwVVPSVUQQZV6dJw111uKOyyZeGOJiKkproqn3r13Ftz7rmwYAHMng2JfguSxpiixNOePyLSTkR+EZE1IjIsh/M6i4iKSGg+dvr3h/Ll4ZnoHhx99ChMmeJ61v7zn1C7NsydC/PmQcuW4Y7OGBMqniUC33xE44CrgAZAdxFp4Oe8CsAg4DuvYjlJ5cpuGa2pU93YgiiTnu5mCT3/fOjRAypWdNNEfPMNXH55uKMzxoSalyWC5sAaVV2nqkeAqbjZS7N6DHgaOORhLCcbPNj1Gho9OqS3DSdVt25A48bQtatrO58+3bWdt29vnaiMiVZeJoKawMZMzzf59h0jIo2B2qr6kYdx+Fe7Ntx0k1tYd8eOkN8+1BYtco2/HTu6WTbeess1kXTqZAnAmGjnZSLw9/FybNCCiBQD/gPck+uFRPqJSJKIJKWkpAQvwnvvhQMH3AipIurgQRg61NX5b90KEyfCypVuTEBMTLijM8ZEAi8TwSagdqbntYAtmZ5XAM4HvhKR9UALYJa/BmNVfVlVE1U1MS4uLngRnn++Gyo7dqxLCEXMt99CQoKbVaNPH/jpJ+jVy1YQM8acyMtEsBioJyJ1RaQk0A2YlXFQVXeraqyq1lHVOsAioIOq5tikL7YAABSNSURBVHPYcD7ddx9s3w6TJoX0tl46cADuuQdatXJjA774ws0Uekrh7PBrjPGYZ4lAVdOAgcBnwCpgmqr+JCKPikgHr+6bZ5dcAi1auHkU0tLCHU2BLVjgSgGjR7suoStWwN/+Fu6ojDGRzNNxBKo6W1Xrq+qZqvqEb99wVZ3l59xLQ14agOOT0a1bV6gnoztwAIYMgdat3QCxuXNh/HioYCtLG2NyUbimkvZKhw5Qvz489ZTrY1nIfP01NGrkJojr3x9+/NHGAxhjAmeJAFz3maFDYelS+PLLcEcTsP373ZQQbdq4UcJffgnjxrlB08YYEyhLBBluvhmqVy80k9HNn+9KAWPHunUCli+Hyy4Ld1TGmMLIEkGGjMnoPv8cfvgh3NFka98+uPNOVwpQdbNpP/+8lQKMMflniSCz2293rasROhndV19Bw4bwwgswaJArBbRpE+6ojDGFXdQkgt27YdeuXE6qVMn1uZw2DX77LSRx5ebAATcdRNu2ruonJsZVC40ZY4vEGGOCI2oSweuvw6mnusHEffu65z//7GbiPMHgwVCsWFgno1N1M4H27euaLW66CVavhkcfdfMDXXJJ2EIzxhRBAa9ZHCnyu2bx8uVu5s2FC922c6fbX7kyXHSR2y6+GJo3h/KDbnNTVP/+O8TGBvk3yN7vv8Mbb8Dkye6Dv1w56NwZevZ04wOKRU3aNsYEW05rFkfNrDMNG7oNXCng11/dXDwZiWH2bHesWDG4oN6LXHwwkYv6fc1FT3fizDO9m6HzwAE3jm3yZDcITBUuvRQefBBuuMEagY0x3ouaEkFudu2C775zSeHbb+G7efvZk+Yq4ePiXIkhMRFq1nTVNRlbXByUKJG3e2VU/Uya5Joj9u51C8X36OGWVK5bN+i/njEmylmJIACVKsGVV7oN4Oj8ZFa2uZ2F3caysPRlfPutq1ryJzb2xOSQsVWrduLzffvgv/91CWDtWlf106WLq/pp1cqqfowx4WElgpy0bAlbtrgK++LFOXgQtm2DP/44/jO77VAO661ddpn78L/+eqv6McaEhpUI8uu+++C66+C996BbN8qUgTp13JYTVdiz53hSyEgaaWnuwz+31xtjTChZiSAn6elw3nlu1PHSpbamozGm0MqpRGC10jkpVsxNRpecDHPmhDsaY4zxhCWC3Nx0E9SoUWgmozPGmLyyRJCbUqXcaOM5c1z1kDHGFDGWCALRr59b8NdKBcaYIsgSQSAqVnRLf02b5lZ+McaYIsS6jwZq+HBYtQoGDoT1692yljYCzBhTBNgnWaDKlnWTAg0YAM8+C9265TxqzBhjCgkrEeRFTIxbDqxuXbj3XjfqeOZMqFIl3JEZY0y+WYkgr0Tgnntce0FSkpu7eu3acEdljDH5Zokgv2680c0bvX27m5r0u+/CHZExxuSLJYKCaNnSzVtdoYKbSW7mzHBHZIwxeWaJoKDq13fJoGFD6NTJtSEYY0whYokgGKpWhS+/hI4dYdAg14Zw0mLIxhgTmSwRBEvZsm666kGD3ML3XbrAwYPhjsoYY3Jl3UeDKSYGxoxxCw7ccw9s3eraDWJjwx2ZMcZky0oEXhgyxHUvXbLEupcaYyKeJQKvdO7s2g3++gtatIBFi8IdkTHG+GWJwEsXX+x6FFWs6LqXzpgR7oiMMeYklgi8Vq+eSwYJCW7B4gcegNTUcEdljDHHWCIIhbg4V03Urx+MGgVt2sCGDeGOyhhjAI8TgYi0E5FfRGSNiAzzc/xuEVkpIstFZK6InO5lPGFVpgy89BJMnQo//eRKCFZVZIyJAJ4lAhGJAcYBVwENgO4i0iDLaT8AiaraEHgPKPpLgHXtCj/8AGed5aqKBg606ayNMWHlZYmgObBGVdep6hFgKtAx8wmqOk9VD/ieLgJqeRhP5DjjDPjmG7j7brfi2UUXwa+/hjsqY0yU8jIR1AQ2Znq+ybcvO72BT/wdEJF+IpIkIkkpKSlBDDGMSpaEf/8bPvoINm6EJk3gjTfCHZUxJgp5mQjEzz71e6LIzUAi8Iy/46r6sqomqmpiXFxcEEOMAFdfDcnJ0LQp3Hor9OwJ+/aFOypjTBTxMhFsAmpnel4L2JL1JBH5G/Ag0EFVD3sYT+SqVcutbTB8OEyZAomJsHx5uKMyxkQJLxPBYqCeiNQVkZJAN2BW5hNEpDHwEi4J/OlhLJGveHEYOdIlhD17oHlzGD8e1G8hyhhjgsazRKCqacBA4DNgFTBNVX8SkUdFpIPvtGeA8sC7IpIsIrOyuVz0uOwyV1V02WVwxx1uJbRdu8IdlTGmCBMtZN84ExMTNSkpKdxheC893U1n/cADrupo6lS48MJwR2WMKaREZImqJvo7ZiOLI1WxYnDvvbBggXveqpUblWxjDowxQWaJINJdeKEbgNaxoysdnHaaa0soKt1ojTFhZ4mgMKhUCd591zUkN28OI0ZA7dpu7qJVq8IdnTGmkLNEUFiIwOWXuwFoq1ZBjx5uAFqDBm4swpdfWg8jY0y+WCIojM45x01g9/vvrpooKQmuuOL46OQjR8IdoTGmELFEUJjFxblBaBs2wKuvugRw661Qt65rWN65M9wRGmMKAUsERUHp0tC7N6xYAZ98Auedd7zb6Z132prJxpgcWSIoSkSgXTv4/HNYtswNRnvpJbdK2vXXw7ffhjtCY0wEskRQVDVsCJMmuWqjBx6Ar76Cli1dw7LNY2SMycQSQVEXHw9PPOGmun76aVcqSEhws5z+/nu4ozPGRABLBNGiXDkYOhTWrXMjlqdOhfr13b6//gp3dMaYMLJEEG0qV3Ylg9Wr4R//cIvjnHmm23fwYLijM8aEgSWCaFW7Nkyc6NoLWrWC++93jcoTJ8LRo+GOzhgTQpYIot3558OHH7rG5Fq1XDfURo3cPhupbExUsERgnDZtYOFCeO89SE2FDh2O7zPGFGmWCMxxInDDDW5g2vjx8OuvcPHFbt8vv4Q7OmOMRywRmJOVKAG33w5r1sBjj8EXX7jRyn37wv/+B2lp4Y7QGBNElghM9sqXh4ceclNUDBjgJrS79FKoWhVuucVVI+3dG+4ojTEFZInA5C4uDsaMge3b4f334dpr3ZxGN94IsbFuWosXX4RNm8IdqTEmH2zNYpM/aWmuIXnmTLetWeP2N2niGpo7dHAjmEXCG6cxBsh5zWJLBKbgVF1j8syZMGuWSxCqbqxCRlK49FIoWTLckRoTtSwRmND680+3ktqsWW4m1IMHoUIFaNsWmjVzJYVGjaB69XBHakzUsERgwufgQbfW8syZrvfRhg3Hj1WrdjwpJCS4rV49KF48fPEaU0RZIjCR46+/3LQWycluW7YMfvrJDWIDt8jOBRecmCAaNnQlCmNMvlkiMJHtyBFYtcolhYwEkZx84lKbZ57pEkSNGq5KqVq1k7eyZcP3OxgT4SwRmMJH1XVHzSg1JCfDypXwxx/Zr8VcvrxLCNklivh4OPtsNwOrMVEmp0RglbEmMom4Xke1a7txC5kdOeIapP/4A7Zt87/9/LMbBb1jx8nXjo+Hc8+FBg1O3OLiQvO7GRNhLBGYwqdkSTdTaq1auZ+bmgopKS45bNrkEsTKlW6bPPnEkdGxsceTQuZEER9v4yFMkWZVQyZ6qcLmzccTQ8b200+wa9fx8ypWdAmhfn1XrVS+/IlbuXIn78vYypaFYtkM4E9LgwMH/G8HD568LzXVJb8zznBbXJwlKBMwqxoyxh+R4yWLtm2P71d1JYiMxLBqlfs5Z44rQezbB+npgd8nI1GUKQOHDh3/kM/oKZVf5codTwpZtzp1XA8sYwJgicCYrERcg3P16nD55ScfV3Uf6Pv2+d/27/e//+BB9+FcpowrKWS3ZXc8JgY2bnTrTv/2m/uZsc2Z4+6bWY0aJyaH005zpZPUVLelpfl/nN2x9HRXIoqLc9VocXEnbqee6mI0hY5VDRlTFKi6tpDMySHztmlTYCvOxcS4AX0lSrgt43Hx4i6J7NwJu3f7f62ISwZZE0RG0oiNhVKlTrxmxpbT88yPs1aF5eW5iLtWyZJROWjRqoaMKepE3PTgVatCixYnHz982LWHiGT/QZ/xYZ+bI0fcTLQpKW7L/Djzvp9/hq+/dj238lKVFgrFirmEkLGVKpX745Ilc39/cmqzEXEluwoV4JRT3M+sW9b95coF9m9SQJ4mAhFpB4wBYoBXVXVUluOlgClAU2AH0FVV13sZkzFRqVQpVz0UDCVLumqnGjUCO//oUVeS2LHDJZG0tONbRtVTIM+ztqlkLeHk9jw93V3n8GEXx5Ej2T/O/PzAgeOPcypV5VbiSk931Xd797otkBKaiEsGGUlixAjo1i331+WRZ4lARGKAccDfgU3AYhGZpaorM53WG9ipqmeJSDfgKaCrVzEZY8IgJsZVC8XGhjuSyKHqEkxGUsjY9uw5eV/mY1WqeBKOlyWC5sAaVV0HICJTgY5A5kTQERjhe/we8IKIiBa2hgtjjMmLjG/65cpFxCy8XlY+1QQ2Znq+ybfP7zmqmgbsBk5KeSLST0SSRCQpJSXFo3CNMSY6eZkI/LWaZP2mH8g5qOrLqpqoqolxNg2AMcYElZeJYBNQO9PzWsCW7M4RkeJAReAvD2MyxhiThZeJYDFQT0TqikhJoBswK8s5s4AevsedgS+tfcAYY0LLs8ZiVU0TkYHAZ7juoxNV9ScReRRIUtVZwGvAGyKyBlcSCH6/KGOMMTnydByBqs4GZmfZNzzT40PAjV7GYIwxJmfeD1kzxhgT0SwRGGNMlCt0k86JSAqwIdxxZCMW2B7uIHJg8RVMpMcHkR+jxVcwBYnvdFX12/++0CWCSCYiSdnN7hcJLL6CifT4IPJjtPgKxqv4rGrIGGOinCUCY4yJcpYIguvlcAeQC4uvYCI9Poj8GC2+gvEkPmsjMMaYKGclAmOMiXKWCIwxJspZIsgjEaktIvNEZJWI/CQid/k551IR2S0iyb5tuL9reRjjehH50XfvJD/HRUTGisgaEVkuIk1CGNvZmd6XZBHZIyKDs5wT8vdPRCaKyJ8isiLTvlNF5AsRWe37WTmb1/bwnbNaRHr4O8eD2J4RkZ99/34zRKRSNq/N8W/B4xhHiMjmTP+O7bN5bTsR+cX39zgshPG9kym29SKSnM1rPX0Ps/tMCenfn6ralocNiAea+B5XAH4FGmQ551LgozDGuB6IzeF4e+AT3HoQLYDvwhRnDPAHbqBLWN8/oDXQBFiRad/TwDDf42HAU35edyqwzvezsu9x5RDE1hYo7nv8lL/YAvlb8DjGEcC9AfwNrAXOAEoCy7L+f/IqvizH/w0MD8d7mN1nSij//qxEkEequlVVl/oe7wVWcfLKa5GuIzBFnUVAJRGJD0McVwBrVTXsI8VVdT4nr4XREZjsezwZuM7PS68EvlDVv1R1J/AF0M7r2FT1c3Wr+gEswq33ETbZvH+BOLakraoeATKWtA2qnOITEQG6AG8H+76ByOEzJWR/f5YICkBE6gCNge/8HL5IRJaJyCcicl5IA3OrvH0uIktEpJ+f44EsIxoK3cj+P184378M1VR1K7j/rEBVP+dEwnt5G66E509ufwteG+irvpqYTdVGJLx/lwDbVHV1NsdD9h5m+UwJ2d+fJYJ8EpHywPvAYFXdk+XwUlx1RyPgeeCDEIfXUlWbAFcBA0SkdZbjAS0R6iVxixV1AN71czjc719ehPW9FJEHgTTgzWxOye1vwUvjgTOBBGArrvolq7D/LQLdybk0EJL3MJfPlGxf5mdfnt8/SwT5ICIlcP9gb6rq9KzHVXWPqu7zPZ4NlBCR2FDFp6pbfD//BGbgit+ZBbKMqNeuApaq6rasB8L9/mWyLaPKzPfzTz/nhO299DUMXgPcpL4K46wC+FvwjKpuU9WjqpoOvJLNvcP6tyhuidzrgXeyOycU72E2nykh+/uzRJBHvvrE14BVqjo6m3Oq+85DRJrj3ucdIYqvnIhUyHiMa1RckeW0WcCtvt5DLYDdGUXQEMr2W1g4378sMi+l2gOY6eecz4C2IlLZV/XR1rfPUyLSDrgf6KCqB7I5J5C/BS9jzNzu1CmbeweypK2X/gb8rKqb/B0MxXuYw2dK6P7+vGoJL6ob0ApX9FoOJPu29sDtwO2+cwYCP+F6QCwCLg5hfGf47rvMF8ODvv2Z4xNgHK63xo9AYojfw7K4D/aKmfaF9f3DJaWtQCruW1ZvoAowF1jt+3mq79xE4NVMr70NWOPbeoUotjW4uuGMv8EJvnNrALNz+lsI4fv3hu/vaznuQy0+a4y+5+1xPWXWehWjv/h8+ydl/N1lOjek72EOnykh+/uzKSaMMSbKWdWQMcZEOUsExhgT5SwRGGNMlLNEYIwxUc4SgTHGRDlLBMb4iMhROXFm1KDNhCkidTLPfGlMJCke7gCMiSAHVTUh3EEYE2pWIjAmF7756J8Ske9921m+/aeLyFzfpGpzReQ03/5q4tYIWObbLvZdKkZEXvHNOf+5iJTxnT9IRFb6rjM1TL+miWKWCIw5rkyWqqGumY7tUdXmwAvAc759L+Cm826Im/RtrG//WOB/6ibNa4IbkQpQDxinqucBu4AbfPuHAY1917ndq1/OmOzYyGJjfERkn6qW97N/PXC5qq7zTQ72h6pWEZHtuGkTUn37t6pqrIikALVU9XCma9TBzRtfz/f8fqCEqj4uIp8C+3CzrH6gvgn3jAkVKxEYExjN5nF25/hzONPjoxxvo7saN/dTU2CJb0ZMY0LGEoExgema6edC3+NvcbNlAtwELPA9ngv0BxCRGBE5JbuLikgxoLaqzgPuAyoBJ5VKjPGSffMw5rgycuIC5p+qakYX0lIi8h3uy1N3375BwEQRGQqkAL18++8CXhaR3rhv/v1xM1/6EwP8V0Qq4maF/Y+q7grab2RMAKyNwJhc+NoIElV1e7hjMcYLVjVkjDFRzkoExhgT5axEYIwxUc4SgTHGRDlLBMYYE+UsERhjTJSzRGCMMVHu/wELAo7nyGhIegAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEGCAYAAACQO2mwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3dd3xUZfb48c8hoYN0FOmyIAhSI8qqFF0RBQHFRRR2FURWVMSCIj9d+9e1F2yshRUVKaKgq4gIIlHEldAkoUiXAEKkh5Z2fn88EzKEmZCQmdyZ5Lxfr0tmbpl7mEzumafc5xFVxRhjjAmklNcBGGOMiVyWJIwxxgRlScIYY0xQliSMMcYEZUnCGGNMULFeBxBKNWvW1EaNGnkdhjHGRJXFixf/oaq1Am0rVkmiUaNGJCQkeB2GMcZEFRHZHGybVTcZY4wJypKEMcaYoMKaJERkvIjsFJHEINtFRMaKyDoR+UVE2vttu1FE1vqWG8MZpzHGmMDCXZJ4D+iRx/YrgKa+ZRjwJoCIVAceAc4HOgKPiEi1sEZqjDHmBGFNEqoaD+zOY5c+wPvq/ARUFZE6wOXAN6q6W1X3AN+Qd7IxxhgTBl63SdQFtvg9T/atC7b+BCIyTEQSRCQhJSUlbIEaY0xJ5HWSkADrNI/1J65UfUtV41Q1rlatgN18jTHGnCKv75NIBur7Pa8HbPOt75pr/XdFFpUxxkSaw4dh71637Nlz4uPGjeGGG0J+Wq+TxOfAHSIyGddIvU9Vt4vI18BTfo3V3YExXgVpjPGQKqSlQWqqWw4eDPxYBMqVy1nKlg382P95mTLuuLzOnZnployMvH+mp7s409OPf5z7Z7BtqaknXvj9k8HRo3m/T337Rl+SEJFJuBJBTRFJxvVYKg2gquOAmcCVwDrgEDDYt223iDwBLPK91OOqmlcDuDEm0mVlwe7dsH07/P67W/wf790bPAlkZoYvruykIXLixT8rK3znzS02FqpVg6pV3VKtGjRokPPYf33ux1WquP9DGEhxmpkuLi5ObVgOYwooIwMSE2HfPihVyl0sS5U6/nHun7nXibiLfKAEkP1zxw73rTm3ihXhjDPcBa9SJfe8UqXjH+e1rmJFt4D7tn3kiFsK+ljVXahjYgr+MyYGSpd2JZNT+Zn9OK9STRiJyGJVjQu0zevqJmNMUUtLg4QEiI93yw8/wIEDoT2HCNSu7S7+depAq1bu5xln5KzL/lmpUmjPbULKkoQxxd2hQ/C//7mEMH8+/PSTawQFOOccGDgQLr7YXbRVXRVL9k//xyfbVqVKzsW/Vi33LdtEPfstGuMFVfftfd8+V02zb5+rcqhSBU47zS0VK55a9cP+/fDjjy4hxMfDokWumkcE2raFYcOgc2eXGKzbuDkJSxLGFNaRI7B8eU7ja+4lOxHkXneyRtFSpXIShn/yCPS4YkVISnJJYelS99qxsRAXB3ffDV26wJ//7Bo5jSkASxLGFFRyMixc6L6tL1wIS5YEbpCtXDmnF0rVqlC3LrRsefy6KlVyfqanu1LAvn3up//j7J8pKbBuXc727GojcL1bLrgAHnzQJYULLshp0DXmFFmSMCYvaWnum/nChTmJITnZbStXLueb+gUXQKNGORf8004rmjr5tDRXbbV/P5x5Zti6QZqSy5KEMf62b89JCAsXul5A2TcxNWgAF17oqm06dYI2bVy3RS+VKQM1arjFmDCwJGFKrpQUWLbMlRSWLnW9fjZtctvKlIEOHeD2211C6NTJVRcZU8JYkjDFX1YWbNzoEsGyZTmJYdu2nH3q14fzz4cRI1xCaN/eqm6MwZKEKW6OHoWVK3MSwtKlrudR9s1iMTHQogVceqnrDpq9VK/ubdzGRChLEia6HToEs2fDF1+4+wFWrnTDTIC7k7dNG/j7310iaNfO9S4qV87bmI2JIpYkTPTZtcslhRkz4OuvXTfQqlVdNVHPni4ZtG0LTZq4ew2MMafMkoSJDps3w2efucQQH+9G6axXD26+2Q2R3Lmzu2PZGBNSliRMZFJ1I5POmAHTp7u2BXBjDT3wgEsMHTp4NmqmMSWFJQkTOTIz3b0JM2a4Zf16lwQ6dYJnn4U+faBZM6+jNKZEsSRhvLduHbzwAnzyibt3oUwZ1/vo/vuhd283qqgxxhOWJIx31q2D//s/+OAD157Qt69brrjCDWthjPGcJQlT9NavhyefzEkOd97pSg1WYjAm4liSMEVn/XpXcnj/fZccRoxwyaFOHa8jM8YEYUnChJ8lB2OiliUJEz4bNrjkMGGCJQdjopQlCRN6/skhNhbuuANGj7bkYEwUsiRhQmfjxpzkEBPjksP997vJcIwxUcmShCm8gwfhnntg/HiXHG67zZUcLDkYE/UsSZjCSUmBq65yI7DefrsbMsOSgzHFhiUJc+o2bIAePWDLFne3dN++XkdkjAkxSxLm1CxeDFde6eZumDvXzftsjCl2bLB9U3CzZkGXLlC+PCxYYAnCmGLMkoQpmAkTXBtE06ZuxNbmzb2OyBgTRpYkTP6owlNPwU03QdeuMH++3fdgTAlgbRLm5DIz3T0P48bBoEHw7rtuOG9jTLFnJQmTt0OHoF8/lyBGj3bVTZYgjCkxrCRhgtu1y7U//PQTvPqqK00YY0oUSxImsE2b3D0QmzbBtGlwzTVeR2SM8YAlCXOipUvdPRBHj8KcOXDRRV5HZIzxiLVJmON98w107uzaHRYssARhTAlnScLk+OADV4I46yx3D0SLFl5HZIzxmCUJ4+6BeOYZ+PvfXSkiPt4G6TPGAEWQJESkh4isEZF1IvJAgO0NRWSuiPwiIt+JSD2/bZkissy3fB7uWEuktDS45RY3euv118NXX0GVKl5HZYyJEGFtuBaRGOB14DIgGVgkIp+r6kq/3Z4H3lfVCSJyCfAv4G++bYdVtW04YyzRdu1y90DMnw///Cc8+iiUssKlMSZHuHs3dQTWqeoGABGZDPQB/JPEOcDdvsfzgBlhjskArF4NvXpBcjJMnAg33OB1RMaYCBTur411gS1+z5N96/wtB/r5Hl8NVBaRGr7n5UQkQUR+EpGAkxWIyDDfPgkpKSmhjL34mj0bLrgADhyAefMsQRhjggp3kpAA6zTX81FAFxFZCnQBtgIZvm0NVDUOuAF4WUSanPBiqm+papyqxtWqVSuEoRdTr7/uejA1aAA//wydOnkdkTEmgoW7uikZqO/3vB6wzX8HVd0GXAMgIpWAfqq6z28bqrpBRL4D2gHrwxxz8ZSRAXfd5ZJEr17w0UdQubLXURljIly4SxKLgKYi0lhEygADgON6KYlITRHJjmMMMN63vpqIlM3eB7iQ49syTH7t3etKD6+/DqNGwYwZliCMMfkS1pKEqmaIyB3A10AMMF5Vk0TkcSBBVT8HugL/EhEF4oHbfYe3AP4tIlm4ZPZ0rl5RJj/WrXOD9K1bB++8Azff7HVExpgoIqq5mwiiV1xcnCYkJHgdRuSYPz9nYL5PP3VTjhpjTC4istjX/nsC6xRfXL37Llx2GdSuDf/7nyUIY8wpsSRR3GRmunaHoUOhWzc3BtOf/uR1VMaYKGVDhRcnBw64ex6++AJuvx1efhli7VdsjDl1dgUpLjZvdg3UK1fCa6+5JGE8lZ6eTnJyMkeOHPE6FBMBypUrR7169ShdurTXoRSIJYni4Lff4Pzz4cgRmDkTunf3OiIDJCcnU7lyZRo1aoRIoPtKTUmhquzatYvk5GQaN27sdTgFYm0S0S4zEwYNgoMH4YcfLEFEkCNHjlCjRg1LEAYRoUaNGlFZqrSSRLR7+mn4/nuYMAFatfI6GpOLJQiTLVo/C1aSiGY//QSPPOLmgfjb306+vzHGFJAliWi1f7/ryVS/Prz5JkTptxRjTGSzJBGtbr/d9Wj68EObSc4EtXfvXt54440CH3fllVeyd+/ePPd5+OGHmTNnzqmGFtDYsWNp0aIFAwcOZPXq1XTq1ImyZcvy/PPPh/Q8Jv+sTSIaTZzoksOjj8KFF3odjYlg2UnitttuO259ZmYmMTExQY+bOXPmSV/78ccfL3R8ub3xxht89dVXNG7cmJ07dzJ27FhmzCj6ecgyMjKItXuMAEsS0WfDBhg+3CWHBx/0OhqTX3fdBcuWhfY127Z1N0zm4YEHHmD9+vW0bduW0qVLU6lSJerUqcOyZctYuXIlffv2ZcuWLRw5coSRI0cybNgwABo1akRCQgKpqalcccUVXHTRRfz444/UrVuXzz77jPLly3PTTTfRq1cvrr32Who1asSNN97If//7X9LT0/n4449p3rw5KSkp3HDDDezatYvzzjuPWbNmsXjxYmrWrHlCrLfeeisbNmygd+/eDBkyhLvvvpvatWvz5ZdfnvStOHjwIP379yc5OZnMzEz++c9/ct1117Fo0SJGjhzJwYMHKVu2LHPnzqV06dIMHz6chIQEYmNjefHFF+nWrRvvvfceX375JUeOHOHgwYN8++23PPfcc0ydOpWjR49y9dVX89hjj53a7yqKWZKIJhkZrrtrqVKuNGHfdMxJPP300yQmJrJs2TK+++47evbsSWJi4rG++uPHj6d69eocPnyY8847j379+lGjRo3jXmPt2rVMmjSJt99+m/79+/PJJ58waNCgE85Vs2ZNlixZwhtvvMHzzz/PO++8w2OPPcYll1zCmDFjmDVrFm+99VbQWMeNG8esWbOYN29ewCSSl1mzZnHmmWceSyj79u0jLS2N6667jilTpnDeeeexf/9+ypcvzyuvvALAihUrWL16Nd27d+fXX38FYOHChfzyyy9Ur16d2bNns3btWn7++WdUld69exMfH0/nzp0LFFu0s6tMNHniCTcW06RJ0LCh19GYgjjJN/6i0rFjx+Nu5ho7dizTp08HYMuWLaxdu/aEJNG4cWPatm0LQIcOHdi0aVPA177GN+Jwhw4d+PTTTwH44Ycfjr1+jx49qFatWkj/P9nOPfdcRo0axejRo+nVqxcXX3wxK1asoE6dOpx33nkAnHbaacdiGjFiBADNmzenYcOGx5LEZZddRvXq1QGYPXs2s2fPpl27dgCkpqaydu1aSxImQn3/PTz5JNx4IwwY4HU0JkpVrFjx2OPvvvuOOXPmsHDhQipUqEDXrl0D3uxVtmzZY49jYmI4fPhwwNfO3i8mJoaMDDcDcVFNRdCsWTMWL17MzJkzGTNmDN27d6dv374B703IKyb/90dVGTNmDP/4xz/CEnO0sN5N0WDPHhg4EBo3hldf9ToaE0UqV67MgQMHAm7bt28f1apVo0KFCqxevZqffvop5Oe/6KKLmDp1KuC+me/Zsyfk5wDYtm0bFSpUYNCgQYwaNYolS5bQvHlztm3bxqJFiwA4cOAAGRkZdO7cmYkTJwLw66+/8ttvv3H22Wef8JqXX34548ePJzU1FYCtW7eyc+fOsMQfyawkEelU4dZbYft2WLDAph01BVKjRg0uvPBCWrVqRfny5Tn99NOPbevRowfjxo2jdevWnH322VxwwQUhP/8jjzzC9ddfz5QpU+jSpQt16tShcj4/w7///jtxcXHs37+fUqVK8fLLL7Ny5cpj1Ub+VqxYwX333UepUqUoXbo0b775JmXKlGHKlCmMGDGCw4cPU758eebMmcNtt93GrbfeyrnnnktsbCzvvffecaWlbN27d2fVqlV06tQJgEqVKvHhhx9Su3btwr0pUcZmpot0770HgwfDU0/BmDFeR2MKYNWqVbRo0cLrMDx19OhRYmJiiI2NZeHChQwfPpxloe7lFUUi9TOR18x0VpKIZGvXwh13QNeucP/9XkdjTIH99ttv9O/fn6ysLMqUKcPbb7/tdUimgCxJRKq0NDfsRpky8MEHkMeNT8ZEqqZNm7J06dLj1u3atYtLL730hH3nzp17Qs+q3ApzrDk1liQi1cMPQ0ICfPIJ1KvndTTGhEyNGjVOucqpMMeaU2O9myLRt9/Cs8/CLbeAr++5McZ4wZJEpNm1yw373awZvPSS19EYY0o4q26KJKowdCikpMAXX4DfjT3GGOMFSxKR5K23YMYMeOEF8A0FYIwxXrLqpkixahXcfbebo/quu7yOxpRQlSpVAtwdzNdee23Afbp27crJ7kd6+eWXOXTo0LHn+ZmfoiBWr15N27ZtadeuHevXr2fIkCHUrl2bVjaFb8hZkogEaWluCtKKFd3Nc6Xs12K8deaZZzJt2rRTPj53kpg5cyZVq1YNRWgAzJgxgz59+rB06VKaNGnCTTfdxKxZs0L2+gWRmZnpyXmLilU3RYKXXoLly+Gzz6BOHa+jMWHg0XQSjB49moYNGx6bdOjRRx9FRIiPj2fPnj2kp6fz5JNP0qdPn+OO27RpE7169SIxMZHDhw8zePBgVq5cSYsWLY4b4G/48OEsWrSIw4cPc+211/LYY48xduxYtm3bRrdu3ahZsybz5s07Nj9FzZo1efHFFxk/fjwAQ4cO5a677mLTpk1B563IbebMmbz88svExMQQHx/PvHnz6Ny5c9DRaXMbO3Ys48aNIzY2lnPOOYfJkyeTmprKiBEjSEhIQER45JFH6NevH5MmTeKpp55CVenZsyfPPPMM4Epc99xzD19//TUvvPAC5cuX55577iE1NZWaNWvy3nvvUae4/C2rarFZOnTooFHnt99UK1ZU7dPH60hMiK1cufLY45EjVbt0Ce0ycuTJY1iyZIl27tz52PMWLVro5s2bdd++faqqmpKSok2aNNGsrCxVVa1YsaKqqm7cuFFbtmypqqovvPCCDh48WFVVly9frjExMbpo0SJVVd21a5eqqmZkZGiXLl10+fLlqqrasGFDTUlJOXbe7OcJCQnaqlUrTU1N1QMHDug555yjS5Ys0Y0bN2pMTIwuXbpUVVX/+te/6gcffBD0//XII4/oc889d9w6/5jzUqdOHT1y5Iiqqu7Zs0dVVe+//34d6feG7t69W7du3ar169fXnTt3anp6unbr1k2nT5+uqqqATpkyRVVV09LStFOnTrpz505VVZ08efKx9ys3/89EJAESNMh11UoSXrv3XsjMjJj5Bkx4ePXrbdeuHTt37mTbtm2kpKRQrVo16tSpw9133018fDylSpVi69at7NixgzPOOCPga8THx3PnnXcC0Lp1a1q3bn1s29SpU3nrrbfIyMhg+/btrFy58rjtuf3www9cffXVx4bkvuaaa/j+++/p3bt3vuetKKzWrVszcOBA+vbtS9++fQGYM2cOkydPPrZPtWrViI+Pp2vXrtSqVQuAgQMHEh8fT9++fYmJiaFfv34ArFmzhsTERC677DLAVT8Vm1IEVt3krTlz4OOP4fHHoVEjr6MxxdS1117LtGnT+P333xkwYAATJ04kJSWFxYsXU7p0aRo1ahRwHgl/geZl2LhxI88//zyLFi2iWrVq3HTTTSd9Hc1jQNH8zltRWF9++SXx8fF8/vnnPPHEEyQlJaGqJ/wf84q1XLlyx+YIV1VatmzJwoULwxKv1/LdQioiF4nIYN/jWiLS+GTHmDykpcGIEdCkCdx3n9fRmGJswIABTJ48mWnTpnHttdeyb98+ateuTenSpZk3bx6bN2/O83j/+RcSExP55ZdfANi/fz8VK1akSpUq7Nixg6+++urYMcHmsejcuTMzZszg0KFDHDx4kOnTp3PxxReH8H+bt6ysLLZs2UK3bt149tln2bt3L6mpqXTv3p3XXnvt2H579uzh/PPPZ/78+fzxxx9kZmYyadIkunTpcsJrnn322aSkpBxLEunp6SQlJRXZ/ync8pUkROQRYDSQPVZ1aeDDcAVVIrz8MqxeDa+8AuXKeR2NKcZatmzJgQMHqFu3LnXq1GHgwIEkJCQQFxfHxIkTad68eZ7HDx8+nNTUVFq3bs2zzz5Lx44dAWjTpg3t2rWjZcuWDBkyhAsvvPDYMcOGDeOKK66gW7dux71W+/btuemmm+jYsSPnn38+Q4cOPTY9aGFcf/31dOrUiTVr1lCvXj3efffdgPtlZmYyaNAgzj33XNq1a8fdd99N1apVeeihh9izZw+tWrWiTZs2zJs3jzp16vCvf/2Lbt260aZNG9q3b39CAz9AmTJlmDZtGqNHj6ZNmza0bduWH3/8sdD/p0iRr/kkRGQZ0A5YoqrtfOt+UdXglY8eiJr5JJKToXlzuPRS16PJFEuROneA8U6kfibymk8iv9VNab4WcPW9oI0XURijRlljtTEmKuS34XqqiPwbqCoitwBDAJs95FR8+y1MmQKPPurmrDbGBHX77bezYMGC49aNHDmSwYMHh/VYkyPf05eKyGVAd0CAr1X1m3AGdioivropLc3dAXX0KCQmQoAbhUzxsWrVKpo3bx6wZ5ApeVSV1atXR11100lLEiISg0sKfwEiLjFElbFj3RhN//2vJYgSoFy5cuzatYsaNWpYoijhVJVdu3ZRLgo7qZw0SahqpogcEpEqqrqvoCcQkR7AK0AM8I6qPp1re0NgPFAL2A0MUtVk37YbgYd8uz6pqhMKev6IsXUrPPYY9OrlFlPs1atXj+TkZFJSUrwOxUSAcuXKUS8KZ5nMb5vEEWCFiHwDHMxeqap35nWQrxTyOnAZkAwsEpHPVXWl327PA++r6gQRuQT4F/A3EakOPALE4RrMF/uO3ZPPmCPLqFGQnu66vJoSoXTp0jS2dicT5fKbJL70LQXVEVinqhsARGQy0AfwTxLnAHf7Hs8DZvgeXw58o6q7fcd+A/QAJp1CHN6aNw8mT4ZHHoGzzvI6GmOMybd8JQnft/wyQDPfqjWqmp6PQ+sCW/yeJwPn59pnOdAPVyV1NVBZRGoEObZu7hOIyDBgGECDBg3yEVIRS0+HO+5wPZlGj/Y6GmOMKZD83nHdFViLqzp6A/hVRDrn59AA63J3pxoFdBGRpUAXYCuQkc9jUdW3VDVOVeOyB+KKKK++CitXunsirLHaGBNl8lvd9ALQXVXXAIhIM1y1T4eTHJcM1Pd7Xg/Y5r+Dqm4DrvG9biWgn6ruE5FkoGuuY7/LZ7yRYds2V8XUsydcdZXX0RhjTIHl947r0tkJAkBVf8WN33Qyi4CmItLYV101APjcfwcRqSki2XGMwfV0Avga6C4i1USkGu4eja/zGW9kuO++nMZq6wJpjIlC+U0SCSLyroh09S1vA4tPdpCqZgB34C7uq4CpqpokIo+LSG/fbl2BNSLyK3A68H++Y3cDT+ASzSLg8exG7Kgwfz589BHcf78b6dUYY6JQfgf4KwvcDlyEayuIB95Q1aPhDa9gIuaO6/R0aNcOUlNde0SFCl5HZIwxQRXqjmu//V5R1Rd9LxgDlM37kBLstdcgKQlmzLAEYYyJavmtbpoL+HfNKQ/MCX04xcD27a6x+ooroHfvk+9vjDERLL9JopyqpmY/8T22r8iB3H+/G8Bv7FhrrDbGRL38JomDItI++4mIdADCMwFtNIuPhw8/dIniT3/yOhpjjCm0/LZJ3AV8LCLZ9zjUAa4LT0hRKiPD3VndsCGMGXPy/Y0xJgrkd1iORSLSHDgb17tpdT6H5Sg5Xn8dVqyATz+1xmpjTLGR32E5/oprl0jEDdA3xb/6qcTbuxcefhguvxz69vU6GmOMCZn8tkn8U1UPiMhFuNFZJwBvhi+sKLNoEezf7+6wtsZqY0wxkt8kken72RN4U1U/A8qEJ6QolJTkfrZu7W0cxhgTYvlNEltF5N9Af2Cm7w7s/B5b/CUmQq1abjHGmGIkvxf6/rjxl3qo6l6gOnBf9kbfAHwlV1IStGrldRTGGBNy+UoSqnpIVT9V1bW+59tVdbbfLnPDEl00UHUlCUsSxphiKFRVRiW3tfa339xAfi1beh2JMcaEXKiSxMmHki2ushutrSRhjCmG8nvHtQkmMZEjlOX+9+OoMQfat3fLmWdab1hjTPQLVZIouZfDxEQ+rTaUV986fuT02rWhQ4ecpNG+vRuxwxKHMSaanHKSEJFKfiPDXhqieKJPUhIfxb5B/fqu5mnFCliyxC2LF8Ps2ZDpu8ukWjWXLPyTR5MmUMo6ExtjIlRhShIrgQZwbKrRkiczkz+SdvB1ehz3jILKleHPf3ZLtiNHchLH4sXu58svQ1qa2165spvErkMHuOYauPBCK20YYyJHnklCRO4JtgmoFPpwoszGjUw92psMYhg4MPAu5crBeee5JVtampvVNDtpLFkCb74JL70EzZrBkCHw979DnTpF898wxphgTlbR8RRQDaica6mUj2OLv8REJjKQlmcd4txz839YmTLQti3cfLMbPHbhQkhJgfHjXVvGAw9A/fpuYrvPPnNTZhtjjBdOVt20BJihqotzbxCRoeEJKXps/D6ZH+nLU387WugqokqVYPBgt6xZA//5D0yYAP/9L5x+uitZDBkCzZuHJvZIsm8fDB/uquaaNHHzNWUv9epBTIzXERpTcolq8FscRORjYDgwUFVfybXtdFXdEeb4CiQuLk4TEhKK7HxPtZ7MgysGsHEjNGoU+tfPyICvvnIljC++cM///GeXLPr3d+0Z0S4z05WYZs92SWHDhpz2GnClrsaN3Tb/BNKkiXvPy9gwk8YUmogsVtW4gNtOkiSSgCuBz4Gu5OrqGmkN1kWZJFShVfl1VKuQxg+7zwn7+XbsgA8+gHffhdWroWJFlyiGDInuxu4HHoBnnnFtMrfe6pLG1q2wfj2sW+cW/8cHD+YcW6qU61bcpAk0bQpXXeWm9LDeYsYUTGGSxJ24ksRZwFaOTxKqqmeFMtDCKsoksTwhnbbnleaN7jMY/nXRTTSkCj/95JLFlCluRJDsxu4rr3Sjg0TLRfKjj2DgQPjHP2DcuJPvrwo7dwZOHmvWuCk9Gjd2yWbIEKhZM/z/B2OKg7ySBKp60gU3h0S+9vVy6dChgxaV+wanaCxpmvLmx0V2ztwOHFAdP171ootU3SVUtVo11V69VJ9+WnXBAtUjRzwLL0+LFqmWK6d68cWqR48W/vWOHlWdMkW1Sxf3PpQtqzpokOqPP6pmZRX+9Y0pzoAEDXJdzbMkEW2KqiSRlQUNax+iza5v+WJpPddVyWObNsH8+fDDD/D99+6bNbguuB07wh+JUrUAABdwSURBVEUXwcUXQ6dOUKWKp6Hy+++uS3CpUm5Sv9q1Q/v6SUmuZDJhAhw4AG3awG23wQ03uA4CxpjjnXJ1U7QpqiQxfz507QofyUCuP/SuuxJHmJQUlzCyk8aSJa6+v1QpN4FedtK4+OKivR/j6FG45BJYtgwWLAhvfk1NdVVab7wBy5fDaae5XmLDh8M54W9GMiZqWJIIsWHD4KPxh9lx1p+p+OvSsJ8vFA4edG0Z33/vEsfChXDokNt21lnQuTOMHh3eLraqcMstrj1l6lT461/Dd67c5/3pJ5cspk51vae6dnXJom9f6yFljCWJEDp6FM44A3plfsYHl70Pn3wS1vOFS3q6+zafnTS+/daVNN591/WaCodXX4U774SHHoInngjPOU4mJcXdgzJuHGzc6H6XQ4e6xF+/vjcxGeM1SxIhNGMGXH01zJSeXPHPOHjssbCer6hs3eq+2S9cCHfdBc8+C6VLh+71v/0WuneHnj1h+nTve2BlZsLXX7uut19+6boQt2vn2ko6dnQ/W7SwG/lM5FN13eMrVoR+/U7tNSxJhFD//vDdnHS27SlP7NRJRVdnUgTS0uC++2DsWHfvxdSpbl6MwtqwwV10zzjDJaHTTiv8a4bSpk3w3nuuRLVoketKC+6PrkOH4xNHo0aFvydFFfbscefNXjZvdh0iKlQo+FK+PJQtG733yqSlwZgx7nMxbJiNWVYQq1e7atPvvoM+fdyX2FNhSSJE9u93Q2QMvXg1r37TwnWjKYYtoJMnuyqYSpXcvRhdupz6ax044O4S37rVXYCbNAldnOGQlQVr18LPP7t4Fy2CpUtdNSO4ey/8k8Z5553YO0sVdu8+PgH4J4RNm9z74q9SJYiNhcOHc85VELGx8K9/wahRBT/WS6mp7tvv7NnueenS7nvXiBFw/vnRm/jC7fBh9/t++mn3ZeaZZ9zf7KmW0PNKEjYzXQF8+qkbX+iGM+a5T3PTpl6HFBYDBsC557o/3ksvdR/Ee+8t+B9sVpbrTbRqFcyaFfkJAtwf2dlnu+Vvf3Pr0tIgMTEncfz8s6uqyspy2xs2dCWOtLScJJCaevzrVq7sbvRr3Bi6dXMlEv+latWc9zcz010EDh06cQm2fu5cd/d6164QF/iWqIize7erfvz5Zzf0zMUXuwEvx493vdLi4lwbVv/+rqRknNmzXZfu9eth0CB4/nn35TVsgt1AEY1LuG+mu+wy1bPOUs3q2Uu1VauwnisS7Nuneu217ua0a65xzwvi4YfdsS+/HJ74vHTggOr8+arPP6963XWqTZuqtmmj2qeP6siRqi++qPrpp6pLlqju3h3+G/p271Y980zVc85RPXw4vOcKha1b3Z9QmTLuffK3f7/q66+rNm/uPj+1a6s+9JBqcrI3seZlzx73+W7d2t3U+t57qgcPhudc27erDhjg3pOmTVXnzAnda5PHzXSeX9hDuYQzSWzfrlqqlPuwauPG7rdVAmRlqb7wgmpMjPtgrliRv+OmTXOfrsGD7Y7novLVV+49v+8+ryPJ29q1qo0aqVaqpDp3bvD9srJUZ89WveoqVRHV2FiXkBcs8P4zlZCgevPNquXLu/e8Y0fVZs3c4ypVVO+4Q/WXX0JzrowMlzSrVHFJ9dFHQ/9FwJJECLz0knu3Vi5KdQ+efDJs54pE8fGqZ5yhWqGC6sSJee+7fLnb74ILIndYkOLqllvcBXXBAq8jCWzZMtXTT1etUUP155/zf9z69ar33OMulKDavr3qf/5TtKWmQ4dcSaFjRxdDhQqqQ4eqLl7stmdlqX73neoNN7hhYcD9Dbz7rmpq6qmdc+nSnPNdeqnqmjWh+//4syQRAnFxqu3aqer//ufetunTw3auSLVtmxtrCdw3pUBjLqWkuG+Jdeu6/U3R2r9ftWFD1T/96dQvTOHy/ffuIl+vnurKlaf2GqmpquPGuWo1UK1ZU/X//T/VLVtCG6u/tWtV771XtXp1d87mzVVfecVVNQXzxx/ui2WLFu6Y005THT7cXfTz48AB1bvvdrUXtWurfvhheEtPliQKac0a9049/7y6EfXAfXJKoLQ09wcDqp06Hf/HmZam2rWr+xZVkG+JJrS+/db9fkaM8DqSHF9+6apmmjVT3by58K+XleWqqvr2dRfSmBj3jfvGG93gljNmuL/b9PRTe/30dPca3bu79zI21rXPffttwS7WWVkuOf79725AS1A97zzVt992iSCQ6dNdIgXVf/zDtTeFm6dJAugBrAHWAQ8E2N4AmAcsBX4BrvStbwQcBpb5lnEnO1e4ksQjj7gifHKyujJv+fKuorAE+/hjV6dcq1ZOvfLtt7tP1IcfehubcQkC3EXNaxMnuots+/aqO3eG/vU3blQdM8ZVx5x5ph4bERlUS5d2pY5rrnEljg8+cCMQB7tAb9+u+sQTORfpunVVH3vMNbQX1u7dqmPHugZ7cH8/w4a59g1V1U2bVHv3dttat3YjGBcVz5IEEAOsx81HUQZYDpyTa5+3gOG+x+cAmzQnSSQW5HzhSBJZWa7o3q2bb0X37u7TbnTVKlecLlVKtV8/jYpG05Li4EHX0aBhQ1cF5ZXXXnNfsLp0KXjvuFO1b5+rFZ4wQfWBB1xp4+yzXWnDP4HUq+d6LI4Y4S7e/fu7ZAaqf/mL63V1qiWRvGRlqS5c6Dp1ZDd8t27t2jgqVFB97jlXKi9KXiaJTsDXfs/HAGNy7fNvYLTf/j9qBCWJn39279I77/hW1K3ryo5GVd03suuuc+9Rjx4lvoAVURYscAn8lluK/txZWe4bOLhvx5HQLffoUdcW8sknqv/3f26+kbg4940eVKtWde0A4WocDmTvXtdz6fzzXXVWKKriToWXSeJa4B2/538DXsu1Tx1gBZAM7AE6aE6SOOirhpoPXBzkHMOABCChQYMGIX/zRo503c727FH3D6g+80zIzxPNsrJU582LvIZS40p24LrHFpXMTNU773TnvfHG8HwbD6WsLFeVfOiQ15F4J68kEe5h1gLdo5t7HJDrgfdUtR5uPu0PRKQUsB1ooKrtgHuAj0TkhFF/VPUtVY1T1bhatWqFNPiMDDdERc+e7o5YkpLchlatQnqeaCfi7vStWNHrSExujz/uRo4ZOtSNFxVu6elw441u/K+77nJ3T8dG+LgOIlC3rhsDy5wo3EkiGfAfgLkesC3XPjcDUwFUdSFQDqipqkdVdZdv/WJc20azMMd7nHnzYMcONw8z4MZmAEsSJmqUKwfvv+9mAxw5MrznOnwYrrkGPvwQnnwSXnzR+9F+TeGF+1e4CGgqIo1FpAwwAPg81z6/AZcCiEgLXJJIEZFaIhLjW38W0BTYEOZ4jzNxopvqs2dP34rERDcIj008YKJIhw7w4INuOOnPPgvPOfbtg8svd8Ouv/GGO58Nzlc8hDVJqGoGcAfwNbAKmKqqSSLyuIj09u12L3CLiCwHJgE3+erIOgO/+NZPA25V1d3hjNff4cNuQL9+/fxmJ01KgpYt7dNvos6DD7qpYocNgz/+CO1rL1jghpZfuNANzDd8eGhf33gr7LWFqjoTmJlr3cN+j1cCFwY47hPAs2nf/vtfN5zzDTf4rUxMdIO2GxNlypRx1U4dOrgRRKdOLfxr7tgB99/vXrdePZg5Ey67rPCvayKL1RgG8dFHbvKTrl19K3budHNftmzpZVjGnLJzz3UTKX78sZsn5FRlZMArr0CzZjBpkhuifPVqSxDFlSWJAHbvdt+Krr/eb/pK69lkioH77nMTJt12m2vMLqj4eDfN6113wQUXwIoVbvIb69lWfFmSCGDaNNeV74SqJrCShIlqsbEwYYKbqGjYMHfvcX5s2+Z6+XXp4mZo/PRTN5HU2WeHN17jPUsSAXz0kfvwt2/vtzIpCapXdxM1GxPFmjeHp55y7W7vv5/3vunp8MIL7u9h2jR46CE30+DVV1v/jZLCkkQuW7bA/PnuW9NxfwSJia6qyf4yTDEwcqSbLnTkSPeZD2TePNcjatQoV4JISoInnoAKFYo2VuMtSxK5TJrkfl5/vd9KVZckrKrJFBOlSsF//uMaoYcOPb7aKTnZzXN+ySWuK/jnn8MXX8Cf/uRdvMY7liRymTgRzj8/1x/Etm3ubiFrtDbFSJMm8NxzMHs2vPUWpKXBs8+66qjPPnM9oZKS4KqrvI7UeCnCR1UpWomJ8MsvbtyZEzaAJQlT7Nx6K0yfDvfeCy+9BGvWuFuBXnoJGjf2OjoTCawk4eejj1yX1/79c22wnk2mmBKBd9+F0qUhM9MNqzFjhiUIk8NKEj5ZWS5J/OUvcPrpuTYmJbleTTVqeBKbMeFUvz5s2ODudShTxutoTKSxkoTPwoWwebPfiK/+rNHaFHPVqlmCMIFZkvCZONGNJ9+3b64NWVmwcqW1RxhjSiRLErgbhqZOhd693Ujgx9m8GQ4etCRhjCmRLEkAixe7WbuCVjWBVTcZY0oka7jGDVS2bZurlz1B9sB+liSMMSWQJQmfE3o0ZUtMhAYN4LQTptc2xphiz6qbTsZ6NhljSjBLEnnJyHCzqVijtTGmhLIkkZf16+HoUStJGGNKLEsSebHZ6IwxJZwlibwkJrrBbVq08DoSY4zxhCWJvCQmwlln2SwrxpgSy5JEXpKSrKrJGFOiWZII5uhR+PVXSxLGmBLNkkQwv/7qusBazyZjTAlmSSIY69lkjDGWJIJKTHTT1DVr5nUkxhjjGUsSwSQluQRRtqzXkRhjjGcsSQSTmGhVTcaYEs+SRCCHDrkhOazR2hhTwlmSCGT1alC1koQxpsSzJBFI9mx0liSMMSWcJYlAEhOhTBlo0sTrSIwxxlOWJAJJSnKD+sXaxH3GmJLNkkQg1rPJGGMASxIn2r8ffvvNejYZYwyWJE60cqX7aSUJY4yxJHGC7J5NVpIwxpjwJwkR6SEia0RknYg8EGB7AxGZJyJLReQXEbnSb9sY33FrROTycMcKuEbrChWgUaMiOZ0xxkSysHbfEZEY4HXgMiAZWCQin6vqSr/dHgKmquqbInIOMBNo5Hs8AGgJnAnMEZFmqpoZzphJTHSliFJWyDLGmHBfCTsC61R1g6qmAZOBPrn2UeA03+MqwDbf4z7AZFU9qqobgXW+1wuv7CRhjDEm7EmiLrDF73myb52/R4FBIpKMK0WMKMCxiMgwEUkQkYSUlJTCRbtrF/z+uzVaG2OMT7iThARYp7meXw+8p6r1gCuBD0SkVD6PRVXfUtU4VY2rVatW4aK1iYaMMeY44b6lOBmo7/e8HjnVSdluBnoAqOpCESkH1MznsaGVnSSsuskYY4DwlyQWAU1FpLGIlME1RH+ea5/fgEsBRKQFUA5I8e03QETKikhjoCnwc1ijTUyEKlWg7gm1WsYYUyKFtSShqhkicgfwNRADjFfVJBF5HEhQ1c+Be4G3ReRuXHXSTaqqQJKITAVWAhnA7UXWs0kC1XQZY0zJI+56XDzExcVpQkLCqR2sCrVqQb9+8O9/hzYwY4yJYCKyWFXjAm2zmwGy7djhejdZo7UxxhxjSSKbDcdhjDEnsCSRzbq/GmPMCSxJZEtMdG0StWt7HYkxxkQMSxLZkpKsqskYY3KxJAGuZ5PNRmeMMSewJAGwZQscOGAlCWOMycWSBEB6Ovz1r9Ax/IPMGmNMNAn32E3RoUkTmDrV6yiMMSbiWEnCGGNMUJYkjDHGBGVJwhhjTFCWJIwxxgRlScIYY0xQliSMMcYEZUnCGGNMUJYkjDHGBFWsZqYTkRRgs9dx5KEm8IfXQeTB4isci69wLL7CKUx8DVW1VqANxSpJRDoRSQg2RWAksPgKx+IrHIuvcMIVn1U3GWOMCcqShDHGmKAsSRStt7wO4CQsvsKx+ArH4iucsMRnbRLGGGOCspKEMcaYoCxJGGOMCcqSRAiJSH0RmSciq0QkSURGBtinq4jsE5FlvuVhD+LcJCIrfOdPCLBdRGSsiKwTkV9EpH0Rxna233uzTET2i8hdufYp0vdQRMaLyE4RSfRbV11EvhGRtb6f1YIce6Nvn7UicmMRxveciKz2/f6mi0jVIMfm+VkIY3yPishWv9/hlUGO7SEia3yfxQeKML4pfrFtEpFlQY4tivcv4HWlyD6DqmpLiBagDtDe97gy8CtwTq59ugJfeBznJqBmHtuvBL4CBLgA+J9HccYAv+Nu9PHsPQQ6A+2BRL91zwIP+B4/ADwT4LjqwAbfz2q+x9WKKL7uQKzv8TOB4svPZyGM8T0KjMrH7389cBZQBlie++8pXPHl2v4C8LCH71/A60pRfQatJBFCqrpdVZf4Hh8AVgF1vY3qlPQB3lfnJ6CqiNTxII5LgfWq6uld9KoaD+zOtboPMMH3eALQN8ChlwPfqOpuVd0DfAP0KIr4VHW2qmb4nv4E1Av1efMryPuXHx2Bdaq6QVXTgMm49z2k8opPRAToD0wK9XnzK4/rSpF8Bi1JhImINALaAf8LsLmTiCwXka9EpGWRBuYoMFtEFovIsADb6wJb/J4n402yG0DwP06v38PTVXU7uD9ioHaAfSLlfRyCKxkGcrLPQjjd4asOGx+kqiQS3r+LgR2qujbI9iJ9/3JdV4rkM2hJIgxEpBLwCXCXqu7PtXkJrvqkDfAqMKOo4wMuVNX2wBXA7SLSOdd2CXBMkfaVFpEyQG/g4wCbI+E9zI9IeB8fBDKAiUF2OdlnIVzeBJoAbYHtuCqd3Dx//4DrybsUUWTv30muK0EPC7CuQO+hJYkQE5HSuF/kRFX9NPd2Vd2vqqm+xzOB0iJSsyhjVNVtvp87gem4Yr2/ZKC+3/N6wLaiie6YK4Alqroj94ZIeA+BHdlVcL6fOwPs4+n76Guk7AUMVF8FdW75+CyEharuUNVMVc0C3g5yXq/fv1jgGmBKsH2K6v0Lcl0pks+gJYkQ8tVfvgusUtUXg+xzhm8/RKQj7newqwhjrCgilbMf4xo4E3Pt9jnwd18vpwuAfdnF2iIU9Buc1++hz+dAdk+RG4HPAuzzNdBdRKr5qlO6+9aFnYj0AEYDvVX1UJB98vNZCFd8/m1cVwc57yKgqYg09pUsB+De96LyF2C1qiYH2lhU718e15Wi+QyGs1W+pC3ARbii3C/AMt9yJXArcKtvnzuAJFxPjZ+APxdxjGf5zr3cF8eDvvX+MQrwOq5nyQogrohjrIC76FfxW+fZe4hLVtuBdNw3s5uBGsBcYK3vZ3XfvnHAO37HDgHW+ZbBRRjfOlxddPbncJxv3zOBmXl9Fooovg98n61fcBe7Ornj8z2/EtebZ31Rxudb/172Z85vXy/ev2DXlSL5DNqwHMYYY4Ky6iZjjDFBWZIwxhgTlCUJY4wxQVmSMMYYE5QlCWOMMUFZkjAmH0QkU44fnTZkI5KKSCP/EUiNiSSxXgdgTJQ4rKptvQ7CmKJmJQljCsE3n8AzIvKzb/mTb31DEZnrG8Burog08K0/Xdz8Dst9y599LxUjIm/75guYLSLlffvfKSIrfa8z2aP/pinBLEkYkz/lc1U3Xee3bb+qdgReA172rXsNN9x6a9zgemN968cC89UNTtged6cuQFPgdVVtCewF+vnWPwC0873OreH6zxkTjN1xbUw+iEiqqlYKsH4TcImqbvANwva7qtYQkT9wQ02k+9ZvV9WaIpIC1FPVo36v0Qg35n9T3/PRQGlVfVJEZgGpuJFuZ6hvYENjioqVJIwpPA3yONg+gRz1e5xJTnthT9w4Wh2Axb6RSY0pMpYkjCm86/x+LvQ9/hE3ainAQOAH3+O5wHAAEYkRkdOCvaiIlALqq+o84H6gKnBCacaYcLJvJcbkT3kRWeb3fJaqZneDLSsi/8N96bret+5OYLyI3AekAIN960cCb4nIzbgSw3DcCKSBxAAfikgV3Mi8L6nq3pD9j4zJB2uTMKYQfG0Scar6h9exGBMOVt1kjDEmKCtJGGOMCcpKEsYYY4KyJGGMMSYoSxLGGGOCsiRhjDEmKEsSxhhjgvr/mPtbLjvgxcgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#check your metrics  \n",
    "plot_graphs(history3, 'accuracy')\n",
    "plot_graphs(history3, 'loss')\n",
    "plot_graphs(history3, 'recall_score')\n",
    "plot_graphs(history3, 'precision_score')\n",
    "plot_graphs(history3, 'f1_score')\n",
    "plot_graphs(history3, 'auc_2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 score is 0.886 for training, and 0.878 for validation\n",
      "AUC-ROC is 0.949 for training, and 0.947 for validation\n"
     ]
    }
   ],
   "source": [
    "print('F1 score is %.3f for training, and %.3f for validation' \n",
    "      % (history3.history['f1_score'][1], history3.history['val_f1_score'][1]))\n",
    "\n",
    "print('AUC-ROC is %.3f for training, and %.3f for validation' \n",
    "      % (history3.history['auc_2'][1], history3.history['val_auc_2'][1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}