09_baseline_lstm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from math import sqrt\n",
    "\n",
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from keras.preprocessing.sequence import pad_sequences\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Embedding\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Activation\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Input\n",
    "from keras.layers import GlobalMaxPooling1D\n",
    "from keras.callbacks import ReduceLROnPlateau\n",
    "from keras.callbacks import EarlyStopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Run Local\n",
    "df = pd.read_csv(\"csv/book_info_complete.csv\")\n",
    "#Run on COLAB\n",
    "#df = pd.read_csv(\"/content/drive/My Drive/final_project/book_info_complete.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5816 entries, 0 to 5815\n",
      "Data columns (total 9 columns):\n",
      "title               5816 non-null object\n",
      "description         5816 non-null object\n",
      "plot                5816 non-null object\n",
      "csm_review          5816 non-null object\n",
      "need_to_know        5816 non-null object\n",
      "csm_rating          5816 non-null int64\n",
      "Genre               5816 non-null object\n",
      "Book type           5816 non-null object\n",
      "Publication date    5816 non-null int64\n",
      "dtypes: int64(2), object(7)\n",
      "memory usage: 409.1+ KB\n"
     ]
    }
   ],
   "source": [
    "df = df.dropna(axis=1)\n",
    "df[\"Publication date\"] = df[\"Publication date\"].str[-4:].astype(int)\n",
    "df = df.drop(\"Last updated\", axis=1)\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Luminous story of pregnant teen's summer in Spain.\""
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[1, \"description\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create the splits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def splitter(df):\n",
    "    split = StratifiedShuffleSplit(n_splits=1, test_size=0.2, random_state=999)\n",
    "    for train_index, test_index in split.split(df, df['csm_rating']):\n",
    "        train_data= df.loc[train_index]\n",
    "        test_data = df.loc[test_index]\n",
    "    \n",
    "    \n",
    "    return train_data, test_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data, test_data = splitter(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_tr, y_tr = train_data['description'].values, train_data[\"csm_rating\"].values\n",
    "x_val, y_val = test_data[\"description\"].values, test_data[\"csm_rating\"].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4652,) (4652,)\n"
     ]
    }
   ],
   "source": [
    "print(x_tr.shape, y_tr.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1164,) (1164,)\n"
     ]
    }
   ],
   "source": [
    "print(x_val.shape, y_val.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14\n",
      "4\n"
     ]
    }
   ],
   "source": [
    "#Tokenize the sentences\n",
    "tokenizer = Tokenizer()\n",
    "\n",
    "#preparing vocabulary\n",
    "tokenizer.fit_on_texts(list(x_tr))\n",
    "\n",
    "#converting text into integer sequences\n",
    "x_tr_seq  = tokenizer.texts_to_sequences(x_tr) \n",
    "x_val_seq = tokenizer.texts_to_sequences(x_val)\n",
    "\n",
    "print(len(max(x_tr_seq, key=len)))\n",
    "print(len(min(x_tr_seq, key=len)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14\n",
      "14\n"
     ]
    }
   ],
   "source": [
    "#padding to prepare sequences of same length\n",
    "x_tr_seq  = pad_sequences(x_tr_seq, maxlen=14)\n",
    "x_val_seq = pad_sequences(x_val_seq, maxlen=14)\n",
    "print(len(max(x_tr_seq, key=len)))\n",
    "print(len(min(x_tr_seq, key=len)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([   0,    0,    0,    0,    0,    0,    0,    5,  114,    3,   15,\n",
       "         29,   33, 1039], dtype=int32)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_tr_seq[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_length = len(max(x_tr_seq, key=len))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5799\n"
     ]
    }
   ],
   "source": [
    "size_of_vocabulary=len(tokenizer.word_index) + 1 #+1 for padding\n",
    "print(size_of_vocabulary)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5798 unique tokens.\n"
     ]
    }
   ],
   "source": [
    "word_index = tokenizer.word_index\n",
    "print(\"Found %s unique tokens.\" % len(word_index))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_1 (Embedding)      (None, 14, 300)           1739700   \n",
      "_________________________________________________________________\n",
      "lstm_1 (LSTM)                (None, 14, 128)           219648    \n",
      "_________________________________________________________________\n",
      "global_max_pooling1d_1 (Glob (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                8256      \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 1,967,669\n",
      "Trainable params: 1,967,669\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "model=Sequential()\n",
    "\n",
    "#embedding layer\n",
    "model.add(Embedding(size_of_vocabulary,300,input_length=max_length,trainable=True)) \n",
    "\n",
    "#lstm layer\n",
    "model.add(LSTM(128,return_sequences=True,dropout=0.2))\n",
    "\n",
    "#Global Maxpooling\n",
    "model.add(GlobalMaxPooling1D())\n",
    "\n",
    "#Dense Layer\n",
    "model.add(Dense(64,activation='relu')) \n",
    "model.add(Dense(1,activation='relu')) \n",
    "\n",
    "#Add loss function, metrics, optimizer\n",
    "#optimizer = tf.keras.optimizers.RMSprop(0.001)\n",
    "\n",
    "model.compile(optimizer=\"RMSprop\", loss='mean_squared_error', metrics=[\"mae\"]) \n",
    "\n",
    "#Print summary of model\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Callbacks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "earlystop = EarlyStopping(patience=10)\n",
    "learning_rate_reduction = ReduceLROnPlateau(monitor='val_mae', \n",
    "                                            patience=2, \n",
    "                                            verbose=1, \n",
    "                                            factor=0.5, \n",
    "                                            min_lr=0.00001)\n",
    "\n",
    "callbacks = [earlystop, learning_rate_reduction]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fit the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/evan/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/framework/indexed_slices.py:433: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n",
      "  \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 4652 samples, validate on 1164 samples\n",
      "Epoch 1/1000\n",
      "4652/4652 [==============================] - 3s 738us/step - loss: 24.3774 - mae: 3.9477 - val_loss: 13.8948 - val_mae: 3.1549\n",
      "Epoch 2/1000\n",
      "4652/4652 [==============================] - 2s 484us/step - loss: 11.4385 - mae: 2.8053 - val_loss: 17.1394 - val_mae: 3.3137\n",
      "Epoch 3/1000\n",
      "4652/4652 [==============================] - 3s 576us/step - loss: 6.7773 - mae: 2.0764 - val_loss: 8.8575 - val_mae: 2.4038\n",
      "Epoch 4/1000\n",
      "4652/4652 [==============================] - 2s 507us/step - loss: 5.0800 - mae: 1.7737 - val_loss: 6.9498 - val_mae: 2.1248\n",
      "Epoch 5/1000\n",
      "4652/4652 [==============================] - 2s 498us/step - loss: 4.1994 - mae: 1.6004 - val_loss: 6.9846 - val_mae: 2.1129\n",
      "Epoch 6/1000\n",
      "4652/4652 [==============================] - 2s 482us/step - loss: 3.5799 - mae: 1.4683 - val_loss: 6.7294 - val_mae: 2.0509\n",
      "Epoch 7/1000\n",
      "4652/4652 [==============================] - 2s 467us/step - loss: 3.0312 - mae: 1.3504 - val_loss: 7.1089 - val_mae: 2.1000\n",
      "Epoch 8/1000\n",
      "4652/4652 [==============================] - 2s 482us/step - loss: 2.6479 - mae: 1.2535 - val_loss: 6.4742 - val_mae: 1.9810\n",
      "Epoch 9/1000\n",
      "4652/4652 [==============================] - 2s 474us/step - loss: 2.5371 - mae: 1.2222 - val_loss: 7.6355 - val_mae: 2.1613\n",
      "Epoch 10/1000\n",
      "4652/4652 [==============================] - 2s 479us/step - loss: 2.1911 - mae: 1.1217 - val_loss: 7.5215 - val_mae: 2.1421\n",
      "\n",
      "Epoch 00010: ReduceLROnPlateau reducing learning rate to 0.0005000000237487257.\n",
      "Epoch 11/1000\n",
      "4652/4652 [==============================] - 2s 478us/step - loss: 1.5764 - mae: 0.9282 - val_loss: 6.6115 - val_mae: 1.9938\n",
      "Epoch 12/1000\n",
      "4652/4652 [==============================] - 2s 469us/step - loss: 1.4588 - mae: 0.8850 - val_loss: 6.8631 - val_mae: 2.0355\n",
      "\n",
      "Epoch 00012: ReduceLROnPlateau reducing learning rate to 0.0002500000118743628.\n",
      "Epoch 13/1000\n",
      "4652/4652 [==============================] - 2s 468us/step - loss: 1.2406 - mae: 0.7994 - val_loss: 6.7648 - val_mae: 2.0207\n",
      "Epoch 14/1000\n",
      "4652/4652 [==============================] - 2s 479us/step - loss: 1.1882 - mae: 0.7747 - val_loss: 6.8071 - val_mae: 2.0209\n",
      "\n",
      "Epoch 00014: ReduceLROnPlateau reducing learning rate to 0.0001250000059371814.\n",
      "Epoch 15/1000\n",
      "4652/4652 [==============================] - 2s 477us/step - loss: 1.0980 - mae: 0.7323 - val_loss: 7.0609 - val_mae: 2.0647\n",
      "Epoch 16/1000\n",
      "4652/4652 [==============================] - 2s 488us/step - loss: 1.0818 - mae: 0.7230 - val_loss: 6.9302 - val_mae: 2.0434\n",
      "\n",
      "Epoch 00016: ReduceLROnPlateau reducing learning rate to 6.25000029685907e-05.\n",
      "Epoch 17/1000\n",
      "4652/4652 [==============================] - 2s 528us/step - loss: 1.0310 - mae: 0.6974 - val_loss: 6.9612 - val_mae: 2.0458\n",
      "Epoch 18/1000\n",
      "4652/4652 [==============================] - 2s 466us/step - loss: 1.0125 - mae: 0.6891 - val_loss: 6.9493 - val_mae: 2.0452\n",
      "\n",
      "Epoch 00018: ReduceLROnPlateau reducing learning rate to 3.125000148429535e-05.\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(np.array(x_tr_seq),\n",
    "                    np.array(y_tr),\n",
    "                    batch_size=128,\n",
    "                    epochs=1000,\n",
    "                    validation_data=(np.array(x_val_seq),np.array(y_val)),\n",
    "                    verbose=1,\n",
    "                    callbacks=callbacks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1164/1164 [==============================] - 0s 194us/step\n",
      "The val_mae is 2.045.\n"
     ]
    }
   ],
   "source": [
    "#evaluation \n",
    "val_loss, val_mae = model.evaluate(x_val_seq, y_val)\n",
    "\n",
    "print(\"The val_mae is %.3f.\" % val_mae)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAD4CAYAAAATpHZ6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3deXxcdb3/8dd3JpN1kjZ70y1J29CFVlooS4GGTdYrIILIYkHkByKK4lUuer0qF/W6XnxclYsiIBVQC4KKgixKL6VAC21taUsL3dukaba2WdpsM/P9/XEmaRIyyWSZmczk/Xw85jFnzpyZ+WQ6fc93vud7vsdYaxERkcTginUBIiIychTqIiIJRKEuIpJAFOoiIglEoS4ikkCSovlieXl5tqSkJJovKSIS99auXVtnrc0PZ9uohnpJSQlr1qyJ5kuKiMQ9Y8yecLcdsPvFGDPFGLPcGLPFGLPZGPPF4Pp7jDGVxpj1wcslwylaRESGL5yWug/4srV2nTEmE1hrjHk5eN9PrLU/jlx5IiIyGAOGurW2CqgKLjcZY7YAkyJdmIiIDN6g+tSNMSXAAmA1cAbweWPMDcAanNb8oT4ecytwK8DUqVOHWa6IxKOOjg4qKipobW2NdSmjWmpqKpMnT8bj8Qz5OUy4c78YY7zAq8B3rbXPGGMKgTrAAt8Giqy1n+7vORYuXGi1o1Rk7Nm1axeZmZnk5uZijIl1OaOStZb6+nqampooLS3tcZ8xZq21dmE4zxPWOHVjjAd4GnjCWvtMsIBqa63fWhsAfgWcMqi/QETGjNbWVgX6AIwx5ObmDvvXTDijXwzwMLDFWntft/VF3Ta7Atg0rEpEJKEp0Ac2Eu9ROH3qZwBLgI3GmPXBdf8OXGuMmY/T/bIb+Mywqwlh+dYathxo5PazZ0TqJUREEkI4o19WAn19fTw/8uX07fXtdTy2ag+fKZ+O26VvexEZHK/XS3Nzc6zLiIq4mPulrNBLmy9A5aGWWJciIjKqxUWozyjIBGBbTVOMKxGReGat5a677mLu3LnMmzePZcuWAVBVVUV5eTnz589n7ty5vPbaa/j9fj71qU91bfuTn/wkxtWHJ6pzvwzVjAIvANtqmjlvdmGMqxGR4fjPv2zm3f2NI/qccyZm8a1Ljx9wu2eeeYb169ezYcMG6urqOPnkkykvL+e3v/0tF154IV//+tfx+/0cPXqU9evXU1lZyaZNzhiQw4cPj2jNkRIXLfVxaR4Ks1LYVj02+sREJDJWrlzJtddei9vtprCwkLPOOou3336bk08+mV//+tfcc889bNy4kczMTKZNm8bOnTu54447eOGFF8jKyop1+WGJi5Y6OK317ep+EYl74bSoIyXUwZbl5eWsWLGC5557jiVLlnDXXXdxww03sGHDBl588UXuv/9+nnzySR555JEoVzx4cdFSBygryGR7TXPIfxQRkYGUl5ezbNky/H4/tbW1rFixglNOOYU9e/ZQUFDALbfcws0338y6deuoq6sjEAhw5ZVX8u1vf5t169bFuvywxFVL/Ui7n6qGViaOT4t1OSISh6644grefPNNTjjhBIwx/PCHP2TChAksXbqUH/3oR3g8HrxeL7/5zW+orKzkpptuIhAIAPC9730vxtWHJ+y5X0bCcOZ+Wb2znk88uIqlnz6Fs44L6wQgIjJKbNmyhdmzZ8e6jLjQ13s14nO/jAZlhcFhjdXqVxcRCSVuQj0nI5ncjGS212gEjIhIKHET6uD0q29TqIuIhBR/oV7dpBEwIiIhxFWolxV4aWz1UdvUFutSRERGpfgK9eDOUvWri4j0Lb5CvdscMCIi8kFxFer5mSlkpSZptkYRiSiv1xvyvt27dzN37twoVjM4cRXqxhjKCjM1sZeISAhxM01Ap7ICLy+/Wx3rMkRkqP72VTiwcWSfc8I8uPj7Ie++++67KS4u5vbbbwfgnnvuwRjDihUrOHToEB0dHXznO9/h8ssvH9TLtra28tnPfpY1a9aQlJTEfffdxznnnMPmzZu56aabaG9vJxAI8PTTTzNx4kSuvvpqKioq8Pv9fOMb3+ATn/jEsP7svsRdqM8o8PL7t/dR39xGrjcl1uWISBy45ppruPPOO7tC/cknn+SFF17gS1/6EllZWdTV1XHaaadx2WWXDerkz/fffz8AGzduZOvWrVxwwQW8//77/OIXv+CLX/wi119/Pe3t7fj9fp5//nkmTpzIc889B0BDQ8PI/6HEaaiDMwJGoS4Sh/ppUUfKggULqKmpYf/+/dTW1pKdnU1RURFf+tKXWLFiBS6Xi8rKSqqrq5kwYULYz7ty5UruuOMOAGbNmkVxcTHvv/8+ixYt4rvf/S4VFRV87GMfo6ysjHnz5vGVr3yFu+++m4985CMsXrw4In9rXPWpQ7c5YDQCRkQG4aqrruIPf/gDy5Yt45prruGJJ56gtraWtWvXsn79egoLC2ltbR3Uc4Y6EPK6667j2WefJS0tjQsvvJBXXnmF4447jrVr1zJv3jy+9rWvce+9947En/UBcddSnzgulYxkt8aqi8igXHPNNdxyyy3U1dXx6quv8uSTT1JQUIDH42H58uXs2bNn0M9ZXl7OE088wbnnnsv777/P3r17mTlzJjt37mTatGl84QtfYOfOnbzzzjvMmjWLnJwcPvnJT+L1enn00UdH/o8kDkPdGBM8C5JCXUTCd/zxx9PU1MSkSZMoKiri+uuv59JLL2XhwoXMnz+fWbNmDfo5b7/9dm677TbmzZtHUlISjz76KCkpKSxbtozHH38cj8fDhAkT+OY3v8nbb7/NXXfdhcvlwuPx8MADD0Tgr4yj+dS7+/KTG1i5vZbV//7hEahKRCJN86mHb8zMp95dWaGX6sY2Glo6Yl2KiMioEnfdL3BsuoDtNc2cVJwd42pEJBFt3LiRJUuW9FiXkpLC6tWrY1RReOIy1I8Na2xSqIvECWvtoMaAx9q8efNYv359VF9zJLrD47L7ZXJ2OilJLk0XIBInUlNTqa+v17kQ+mGtpb6+ntTU1GE9T1y21N0uw/R8L9trFeoi8WDy5MlUVFRQW1sb61JGtdTUVCZPnjys54jLUAdnZ+ma3YdiXYaIhMHj8VBaWhrrMsaEuOx+AWdnaeXhFo60+WJdiojIqBG3oT6jwJkuYIe6YEREusRtqJcVBs+CpJ2lIiJdBgx1Y8wUY8xyY8wWY8xmY8wXg+tzjDEvG2O2Ba+jOrawOCcdj9toYi8RkW7Caan7gC9ba2cDpwGfM8bMAb4K/MNaWwb8I3g7apLcLkrzMtiuU9uJiHQZMNSttVXW2nXB5SZgCzAJuBxYGtxsKfDRSBUZSllBplrqIiLdDKpP3RhTAiwAVgOF1toqcIIfKAjxmFuNMWuMMWtGeozqjAIv+w4epbXDP6LPKyISr8IOdWOMF3gauNNa2xju46y1D1prF1prF+bn5w+lxpDKCr0ELOysPTKizysiEq/CCnVjjAcn0J+w1j4TXF1tjCkK3l8E1ESmxNDKCjrPgqR+dRERCG/0iwEeBrZYa+/rdtezwI3B5RuBP498ef0ryUvH7TI6YYaISFA40wScASwBNhpjOqcs+3fg+8CTxpibgb3AxyNTYmgpSW6Kc9M1Vl1EJGjAULfWrgRCzZd53siWM3gz8r3qfhERCYrbI0o7lRV62V1/lHZfINaliIjEXPyHekEm/oBld71GwIiIxH2oz+h2ajsRkbEu7kN9er4XYzSxl4gIJECopyW7mZKdrp2lIiIkQKiDc8IMdb+IiCRIqM8o9LKz9gg+v0bAiMjYlhihnu+l3R9g78GjsS5FRCSmEiLUywo754BRF4yIjG0JEeoa1igi4kiIUPemJDFxXKpCXUTGvIQIdYAZhZka1igiY17ChHrnsMZAwMa6FBGRmEmoUG/tCFB5uCXWpYiIxEzChHrnzlJ1wYjIWJZ4oa45YERkDEuYUB+fnkx+ZorGqovImJYwoQ5hzAFz9CA07o9eQSIiUZaQoW5tHyNgfO2w9FJ47IroFyYiEiUJFeozCjNpbvNxoLH1g3eu/AlUb4LarWqti0jCSqhQLwu1s7T6XVjxI5h4onN712tRrkxEJDoSM9S796sH/PDs5yE1C657EtKyYdeKGFUoIhJZSbEuYCTlelPITvewvftY9VX/C5Vr4cqHwZsPJWfCboW6iCSmhGqpA5QVZB7rfqnfAa98B2ZeAnOvdNaVlMPhvXBod8xqFBGJlIQL9RmFXrbVNGMDfnj2C+BOgX+5D4xxNigtd67Vry4iCSjhQr2swEtDSwfNbzwEe1bChd+BrKJjG+TPhIwC2K1QF5HEk4ChnkkR9aS9ei9MOxsWLOm5gTFQutjZWdrXeHYRkTiWgKGewX95HsIGfHDp/xzrdumuZDE0VTl97iIiCSThQr1g1584x72BlybcBtklfW/U2a+uUTAikmASK9SbazAvfo0tntk8Hrgg9HY50yBrksari0jCSaxQf/4r0H6Uv5b8O9tqj4bezhinC2bXa+pXF5GEkjih/u6z8O6f4ey7GT9lLnXN7Rw60h56+9JyOFoHNVuiV6OISIQlRqgfPQjPfRkmfAhO/wIzCp3pArbX9jMNb+li51pDG0UkgQwY6saYR4wxNcaYTd3W3WOMqTTGrA9eLolsmQN48evQchAuvx/cntATe3U3fqqzI1X96iKSQMJpqT8KXNTH+p9Ya+cHL8+PbFmDsO3vsOG3cMadUPQhACaOSyM92T3w+UpLFsPulRAIRKFQEZHIGzDUrbUrgINRqGXw2prgr3dC3nFw1r91rXa5DDMGOgsSOP3qrYehemOECxURiY7h9Kl/3hjzTrB7JnvEKhqMv98DDRVOt0tSSo+7ZhR4Bz4JdUmwX11dMCKSIIYa6g8A04H5QBXw36E2NMbcaoxZY4xZU1tbO8SX68Pu1+Hth+DU22DKKR+4e0aBlwONrTS2doR+jqwiyC3T5F4ikjCGFOrW2mprrd9aGwB+BXwwVY9t+6C1dqG1dmF+fv5Q6+ypowWevQPGF8N53+hzk7KCTIDwumD2vAF+38jUJiISQ0MKdWNMt2kPuQLYFGrbiFj+X3BwB1z2U0jO6HOTzhEwA4f6Ymhvgqr1I12liEjUDXjmI2PM74CzgTxjTAXwLeBsY8x8wAK7gc9EsMaeKtfCmz+HE29wZmEMYUpOOslJroFDvXu/+uSFI1amiEgsDBjq1tpr+1j9cARqGZivHf58B3gL4YLv9Lup22WYnu9lW/UAwxoz8qDgeCfUF//rCBYrIhJ98XVE6cr7oGYzfOQnkDpuwM3LCrw9T0IdSuli2LvK+dIQEYlj8RPq1e/Cih/D3Ktg5sVhPaSswEvFoRaOtg+wE7S0HHwtULlmBAoVEYmd+Ah1vw/+/DlIzYKLfxD2w2YEd5buqDnS/4bFpwNGQxtFJO7FR6iv+l/Yvw4u/qHTBx6msuDEXgNOF5CW7UwxoIOQRCTOxUeoZ+TBCdfC3CsH9bDi3AySXCbMfvVyqHjLGQMvIhKn4iPU518HV/yi7/ON9sPjdlGalzHwsEaAknLwt8O+t4ZYpIhI7MVHqA9DWWEYE3sBFC8C41YXjIjEtYQP9RkFmeypP0Jrh7//DVMyYdKJOmmGiMS1hA/1sgIvAQu76gYYAQNOv3rlWmgLo2UvIjIKJX6od42ACadffTEEfM6BSCIicSjhQ700LwOXge0DTRcAMOVUcHlgt/rVRSQ+JXyopyS5Kc7NCK+lnpzuzM2unaUiEqcSPtSB8E5t16lkMVRtgJbDkS1KRCQCxkSolxV42VV3hA5/GCeYLi0HG3BOnCEiEmfGRqgXevEFLHvqwxgBM3khJKVqaKOIxKWxEerBU9sNeCJqcE5gPeVUTe4lInFpTIT69HwvxoQ5rBGcLpjqjXCkPrKFiYiMsDER6mnJbiaNTxtcqAPsWRm5okREImBMhDoEz4IUzlh1gIkLINmroY0iEnfGTqgXZrKz7gj+gB14Y7cHpi5Sv7qIxJ0xE+ozCry0+wLsO3g0vAeULoa696DpQGQLExEZQWMm1MsKBjEHDBzrV9+tfnURiR9jJtRnFIR5artOEz4EqePUry4icWXMhHpmqoeicalsD2esOoDLDcVnKtRFJK6MmVAHp7X+TmUD1oaxsxScfvVDu+DwvsgWJiIyQsZUqF8yr4jtNc28uTPMg4q6+tU1CkZE4sOYCvUrFkwiz5vMr1bsDO8B+bMhPVdDG0UkboypUE/1uLlxUQnL36vl/XAORHK5oORMp6UebpeNiEgMjalQB/jkacWkelw89FqYrfXScmjY5/Sti4iMcmMu1LMzkrl64RT+9M/91DS2DvyAkmC/urpgRCQOjLlQB7j5zFJ8gQCPvrF74I3zysA7QUMbRSQujMlQL87N4KK5E3h81R6OtPn639gYZ2ij+tVFJA6MyVAHuGXxNBpbfTy5Jowx6CWLobka6t6PfGEiIsMwZkN9wdRsTi7J5uGVu/ANdO7SzvHq6oIRkVFuwFA3xjxijKkxxmzqti7HGPOyMWZb8Do7smVGxi2Lp1FxqIW/bRpgJsbsEhg3RQchicioF05L/VHgol7rvgr8w1pbBvwjeDvufHh2IdPyMnhwxc7+pw4wxmmt73oNAgO06kVEYmjAULfWrgAO9lp9ObA0uLwU+OgI1xUVLpfh/y2exsbKBlbv6v0n9lKyGFoOQs270SlORGQIhtqnXmitrQIIXheE2tAYc6sxZo0xZk1tbe0QXy5yPnbiJHIzknlwoKkDShc71+pXF5FRLOI7Sq21D1prF1prF+bn50f65QYt1ePmhkUlvLK1pv9zmI6bDDnT1K8uIqPaUEO92hhTBBC8rhm5kqJvyaLOqQMGmAqgZDHsfh0C/ugUJiIySEMN9WeBG4PLNwJ/HplyYiMnI5mPnzSFP/6zkpqmfqYOKC2Htgao2hC94kREBiGcIY2/A94EZhpjKowxNwPfB843xmwDzg/ejms3n1lKRyDAb97YE3qjkmC/urpgRGSUCmf0y7XW2iJrrcdaO9la+7C1tt5ae561tix4PcDQkdGvJC+DC+dM4LH+pg7ILIS8mdpZKiKj1pg9orQvt5RPo6Glg6f6mzqgtBz2vAn+jugVJiISJoV6NycVZ3NScTYPv97P1AGli6HjCOz/Z3SLExEJg0K9l1vLp7HvYAsvbq7ue4POfvVdr0avKBGRMCnUe/nw7EJK8zJ4cMWOvqcOSM+BohPg7UegZmv0CxQR6YdCvRe3y3DzmaVsqGjgrVBTB1x+P1g/PHIh7F0d3QJFRPqhUO/DVSdNJicjmV+FOo/phHlw80uQngu/uQy2Ph/dAkVEQlCo98GZOqCYv2+pYXtNc98bZZc4wV4wB5ZdD2uX9r2diEgUKdRDWHJaMSlJLh4K1VoHyMiDG/8C08+Fv3wBXv2RTnknIjGlUA8h15vCVSdN5pl1A0wdkOKFa38PJ1wLy78Dz39Fc8OISMwo1PvROXXAY2/2M3UAgNsDH30AzvgivP0QPPUp6Ojni0BEJEIU6v2Ylu/l/NmFPLZqD0fbQ0wd0MkYOP9euPB7sOVZePxKaDkcnUJFRIIU6gP4zFnTOHy0g6fWVIT3gEW3w5UPw77V8OtLoLEqsgWKiHSjUB/AScU5nDh1PA+t3Ik/EOZO0HlXwfVPweE98PD5UPt+ZIsUEQlKinUB8eDW8mnc9vg6Xtx8gEvmFYX3oOnnwKeegyeucg5Suu5JmHJyZAuVntqPwqr74Z2nICkZUrIg2QspmaEvyV5nu651Xmedyx3rv0YkLAr1MJw/ZwIluen8csVOLp47AWNMeA+cON8Zy/7Yx2DppXD1UjjuwsgWK87oo/W/heXfhaYqZ2bNZC+0NUHzAajf5iy3NYOvJbznTPZC1kRYsAROXAJp2ZH9G0SGyPQ5v0mELFy40K5ZsyZqrzeSHlu1h2/8aRNP3baIk0tyBvfg5lqnxX5gI1z2U1jwycgUOdZZC9tehr9/C2rehcknw/nfhuJFoR/j7wgGfBO0Nx9bbms8Fvyd66rWw57XwZMOH/oEnPoZKJgdvb9PxixjzFpr7cJwtlVLPUxXnTiZ+156j1++unPwoe7Nh0/9FZYtgT9/Dpqr4cx/dUbMyMjYvx5e/oZzApPsUvj4Uphz+cDvsdvjTNKWHua/6YGNsPqXsOF3sPbXUHoWnHqb8wtMXTQyCmhHaZjSkt0sWVTC37dUs6M2xNQB/UnJdPrV530c/nEv/O1uCISYsz1etTbC3lVO10f1u9E5uvbQHnj6FnjwLKjeDBf/ED73Fhz/0ch8aU6YB5f/HL70Lpz3LajfDr+/Fn66AN74uYaxDpeOyB42db8MQl1zG6d//xWuPHEy3/vYvKE9SSDgtCjf/Dkcf4UTQtZCwAeBDqc/OOA7dvH7et7u8+J3+njHTYHxUyA5Y2T/8A/8DX44uBOqNzlBWr3ZWT68t+d246fCcRc5l5IzISll5GpoOQSv/bfTajYuOO12OPNOSB03cq8RDr8Ptv7VqWPvG+DJgBOucbpm8mdGt5Z44GuHxgpoqIDD+6Bh37Hrhn3QUOlslzquj0tWr9vjneuUXus9af1/oVvrfIatv+f/od7/92zAuQbnM4ZxnteYbrddvW73dZ9xbidnOL8Mh2Aw3S8K9UH62jMbeXpdBW989VzyvMMIqTd+Bi/9x8gV1l1ajhPu46Y4wdoZ9p2307LDb8UeqT8W3jXBAK/ZAr7gEbPGDXllUHi8cyk4HrKLnXH6770AO//P2RmZ7HVGBB13MZRd4HRJDYWvDd76Faz4EbQ2wPzr4Jyvw7hJQ3u+kbR/Pbz1IGx8CvztMO0cOO2zMON8cI3Qj+KOFif4GvZC+xEnKJK9Tj9/53JyxsDBFimtjR8M6q7lCmg6APTKHO+EY5/PcZOdulsbul0auy0fdt7b/rg8zheAcXUL7G7XNkbTeFz/NJR9eEgPVahH0I7aZj5836vcfEYp//GROcN7sl0rnBNtuNzgSnK+xV1Jx267kpwPaI/bSeBO6nnbuOBoffA/z94PtoA6jvZ8XU9G8D/R5G6BPxWyipzA6N4Cbz5w7HHpeTBhLhTOPRbieTPBkxr6b+xocf7O9/4G778ITfsB4+zEPO5CmHmxM9PlQAEUCMDmZ+Af/+n8Iph+Hpz/n053yGjTXAvrHoW3H3ZG32SXOi33+df1/0vCWjh60Pk37GrJVvS8fbQuzCLMsYDvunTeTu95nyvJ2WHs73B+LXZd+7rd9vW8v69tjtY5wdudOzn4OZvsfMa6h/f4KZA1afC/4DpanddpazwW9H19CWC7/T9xOw2Q7re7/78yvW53LptuX8Y24Pwb2YDz3KFuY4Pre9036yNOg2cIFOoR9pWnNvCHtRVcd+pU7rn0eJKTRvGuCWudrorDe3sGfefthgrnC6E7dzLkzwqG95xggM8Fb8Hwa6naAO+/4Fw6z/M6birMvMgJ+ZLFH/xPvmsFvPQNZ/TJhHnOdAzTzx1eLdHg73CmjFj9S+eXS7LXCfbjLoLmmg8GdkPFB4dYJqX1/ALuHogpmc5Y/PYjznlz2zsvzX0sH+21vtt9AZ/zb97ZqHB7nMaEO8lZ37ns8vS8z+UJPi64nJbdLbSDjYWMgpH7lTKGKdQjzB+w/Pil93jg/3ZwUnE2D1x/IgVZ/bRWR7v2I06gNFZC5kTInT7kvr9BaayCbS+G6Ka5CHLLnH7zbS9C1mQ47xsw7+r4DInKdU7XzKane3YfZOT3DOzeAZ6eo1FSolCPlufeqeIrT20gMzWJXyw5iROn6oCUIeuzmwZIGQeL/9XpvvCkxbbGkdBc4+yTyJrk7AdIhL9JIk6hHkVbqhq59bE1VDe0ce/lx3PNKVNjXVL8sxYOvANV78DMSyAjN9YVicTUYEI9Dn/Hji6zi7L4y+fP5NRpOXz1mY38x5820u5LsPHn0WYMFJ3gHI6vQBcZFIX6CBifnsyvP3UynymfxuOr9nL9Q6v6P1uSiEiEKNRHSJLbxdcumc1Pr13AxsoGLvvZ66zfp6MLRSS6FOoj7LITJvLMZ88gyW24+hdv8uSafbEuSUTGEIV6BMyZ6PSzn1yazb/94R2++edNdPjVzy4ikadQj5DsjGSW3nQKtywu5Tdv7uH6h1ZT19wW67JEJMEp1CMoye3i6/8yh/+5Zj4b9h3m0p+t5J0K9bOLSOQo1KPg8vmTePqzp+Myhqt+8SZPrw3zJNYiIoM0rFA3xuw2xmw0xqw3xiTWUUUjbO6kcTz7+TM4aWo2X35qA/c8u1n97CIy4kaipX6OtXZ+uEc7jWW53hQeu/kUPn1GKY++sZtPPrSamkaNZxeRkaPulyhLcrv45qVzuO/qE1i/7zBn/mA5X1q2nrV7DhHNKRtEJDENa+4XY8wu4BDOrPe/tNY+2Mc2twK3AkydOvWkPXv2DPn1Es3O2maWvrGbp9dV0tzm4/iJWdywqJjLTphEWrLOdykijqhN6GWMmWit3W+MKQBeBu6w1q4ItX0iTug1EprbfPzpn5U89uYe3qtuIis1iY8vnMInTyumNC/Cp6YTkVEvJrM0GmPuAZqttT8OtY1CvX/WWt7efYjfvLmbFzYdwBewLC7LY8lpxZw3uxC3S/Nqi4xFgwn1pGG8SAbgstY2BZcvAO4d6vMJGGM4pTSHU0pzqGlq5fdv7eO3q/dy62NrmTQ+jetOnconTp4yvHOjikhCG3JL3RgzDfhj8GYS8Ftr7Xf7e4xa6oPn8wf4+5ZqHlu1h9e31+NxGy6ZV8QNi4o5cWo2RmfFEUl4OklGgtpe08zjq/bw9NoKmtp8zCnKYsmiYi6fP5H05CH/6BKRUU6hnuCOtPn403pnx+rWA01kpiZx4fETOHdWAWeW5ZGVGoXzi4pI1CjUxwhrLWv2HOK3q/fyjy3VNLb6SHIZTi7J4dxZBZwzq4Dp+RnqohGJcwr1McjnD7Bu72Fe2VrD8q01vFfdBMDUnPSugD+1NIdUj8a/i8QbhbpQcegoy9+rZfnWGt7YUUdrR4A0j5szZuQFQz6fonE6k71IPFCoSw+tHX7e3FHPK1treGVrDZWHWwCYNSGTcz8yjS8AAAwbSURBVGcVcO6sAhZMzdY4eJFRSqEuIVlr2V7T3BXwa/Ycwh+wjE/3cNZx+Swuy+f06blMHK9WvMhooVCXsDW0dPDatlpe2VrDq+/VUn+kHYDSvAwWTc/ljOl5nDYth1wd8CQSMwp1GZJAwPJedRNv7Kjnje11rN51kOY2HwCzi7I4fXoup0/P5ZTSHDI1bFIkahTqMiJ8/gAbKxuckN9Rx5rdh2jzBXC7DB+aPI4zpudx+vRcTizO1qgakQhSqEtEtHb4Wbf3EG/uqOf17XVsqGjAH7AkJ7lYWJzttORn5PGhSeNIcmuqfpGRolCXqGhu8/HWrnre2F7PGzvqebeqEYCMZDdzJmYxa0IWs4uymFWUyawJmZrKQGSIFOoSEwePtLNqZz2rdtazpaqRrVVNNAX75I2B4pz0HkE/pyiLSePTcGkopUi/ojL1rkhvORnJXDKviEvmFQHO8MmKQy1OwB9oYuuBRrZUNfHiuwfobEt4U5KYOSGT2UWZwcDPZOaELLwp+miKDIVa6hJ1R9t9vHegyQn6KifotxxopKnV17XN1Jx0jivMZHy6B29KknNJTSIjJQlvihtvioeMFDeZwevO+9M8bs11IwlHLXUZ1dKTk1gwNZsFU7O71llrqTzcwtaqYy367TXNvLu/g+Y2H81tPgJhtD9chmDwJ3Vdj0vzUJqXQVmhl7KCTMoKvGRnJEfwLxSJHYW6jArGGCZnpzM5O50Pzyn8wP3WWlo7AjS1dXCkzc+RNh9NrT6OBAO/89J1u9XHkXZnm/ojbby9+yBH2/1dz5fnTWZGgRPyzrWXGYVe8r0paulLXFOoS1wwxpCW7CYt2Q2Zg398IGDZ39DCtppmtlc3s72mmW01TfxpfWWPbp9xaR7KCryUFXqZnu+lrNBp2ReNS1XYS1xQqMuY4HId+yVwzsyCrvXWWmqa2thW3cz2mia21TSzraaZFzYd4NDRjq7tOnfolpflc86sfOZOHKdROzIqaUepSAj1zW3BFr3Tsv/nvsO8U3EYayHPm8LZM/N1timJCu0oFRkBud4Ucr0pnDott2tdfXMbr75fy/L3anlp8wH+sLaCJJfhpOLsrpORlBV41VUjMaOWusgQ+fwB/rnvMMuD0xhvPeCcbWrS+DTOmeW04hdNy3P2A4gMg44oFYmBqoYWlm+tZfl7Nby+vY6j7X6Sk1wsmpbrtOJnFjA1Nz3WZUocUqiLxFibz89buw52hfyuuiMATM/P4OSSHMale8gMjqPPTPXgTU1ybqceO9AqK9VDSpJLXTmiUBcZbXbVHeH/3nO6ad7d30hTm492X2DAxyW5zLGgT3GCvvP2+HQPhVmpwUsKE7JSKchKJSs1SV8ECUY7SkVGmdK8DErzSrnpjNKudW0+P0fa/DS3+mhq66Cp1TloqrnNR1PwAKqm1o6ug6k619U0tbKz1sfBI+00dhtj3ynN46YwK6VH4HcuTxiXSmFmKgVZKZoDP0Ep1EViJCXJTUqSm5xhTFnQ0u6npqmVAw2tVDe1Ud3QSnXjseUNFYc50NBKWx+/CsaneyjMTGVcuodkt4vkJBcet8ETXE52u7qWPW4XyW7TtdxjmyRDsttNSpKL9OABYunJSd2W3aQmuTWuP0oU6iJxLC3ZTXFuBsW5GSG3sdbS2OKjujP8GzsvbRxobKWptYOWDj8NLR10+AO0+wK0+wNdyx1+S3tweThSPS7Sk51J19KDYd/5BZCW7O5a73G7cBlwGYMxpmvZZYBet537Tbftj92X5O75peNxGzzdvqw6v8BSOr+ogl9qx+53bsdbV5ZCXSTBGWMYl+5hXLqH4wqHMMdCkLUWX8D2Cn7rLPsCtHb4aenw09Lu52i7n6PtPlo6Opf9tLT7nOtu23T+0uhcPtruxx+wBGznxXndgIWAtURxF2AXlwG3q+cXjLvzC8TV+aXS88vm2PrgOpfhv66YxymlORGvV6EuImExxnS1btNjNMmlDQa7ha7gt5auL4CAtdgA+K3FFzj2S+PYr46etzt/kXT4A3T4bK9fKAHa/ZZAoOcXjPOl0/klY/F3//IJvnb3uvwBZzkjJTr7MBTqIhI3TLCFDOAmvrpFokVnBxYRSSAKdRGRBKJQFxFJIMMKdWPMRcaY94wx240xXx2pokREZGiGHOrGGDdwP3AxMAe41hgzZ6QKExGRwRtOS/0UYLu1dqe1th34PXD5yJQlIiJDMZxQnwTs63a7IriuB2PMrcaYNcaYNbW1tcN4ORERGchwQr2vQaIfON7LWvugtXahtXZhfn7+MF5OREQGMpyDjyqAKd1uTwb29/eAtWvX1hlj9gzx9fKAuiE+NlZUc+TFW72gmqMl3mrur97icJ9kyPOpG2OSgPeB84BK4G3gOmvt5iE94cCvtybc+YRHC9UcefFWL6jmaIm3mkeq3iG31K21PmPM54EXATfwSKQCXUREwjOsuV+stc8Dz49QLSIiMkzxdETpg7EuYAhUc+TFW72gmqMl3moekXqjeo5SERGJrHhqqYuIyAAU6iIiCWTUhfpAk4QZY1KMMcuC9682xpREv8oe9Uwxxiw3xmwxxmw2xnyxj23ONsY0GGPWBy/fjEWtvWrabYzZGKxnTR/3G2PMT4Pv8zvGmBNjUWewlpnd3rv1xphGY8ydvbaJ+XtsjHnEGFNjjNnUbV2OMeZlY8y24HV2iMfeGNxmmzHmxhjX/CNjzNbgv/sfjTHjQzy2389QlGu+xxhT2e3f/5IQj436JIQh6l3Wrdbdxpj1IR47+PfYBk/JNBouOEMjdwDTgGRgAzCn1za3A78ILl8DLItxzUXAicHlTJyx+71rPhv4a6zf31417Qby+rn/EuBvOEcOnwasjnXN3T4jB4Di0fYeA+XAicCmbut+CHw1uPxV4Ad9PC4H2Bm8zg4uZ8ew5guApODyD/qqOZzPUJRrvgf4ShifnX7zJVr19rr/v4FvjtR7PNpa6uFMEnY5sDS4/AfgPBPD031ba6usteuCy03AFvqYAycOXQ78xjpWAeONMUWxLgrnYLcd1tqhHpkcMdbaFcDBXqu7f16XAh/t46EXAi9baw9aaw8BLwMXRazQbvqq2Vr7krXWF7y5Cudo8VEjxPscjphMQthfvcHsuhr43Ui93mgL9XAmCevaJvjBawByo1LdAIJdQQuA1X3cvcgYs8EY8zdjzPFRLaxvFnjJGLPWGHNrH/eHNWFbDFxD6P8Ao+09Bii01laB0wAACvrYZrS+1wCfxvnF1peBPkPR9vlgl9EjIbq5RuP7vBiottZuC3H/oN/j0Rbq4UwSFtZEYtFmjPECTwN3Wmsbe929Dqe74ATgZ8Cfol1fH86w1p6IMx/+54wx5b3uH3XvszEmGbgMeKqPu0fjexyuUfdeAxhjvg74gCdCbDLQZyiaHgCmA/OBKpwujd5G4/t8Lf230gf9Ho+2UA9nkrCubYLzz4xjaD/FRowxxoMT6E9Ya5/pfb+1ttFa2xxcfh7wGGPyolxm75r2B69rgD/i/DTtbtATtkXBxcA6a2117ztG43scVN3ZbRW8ruljm1H3Xgd31n4EuN4GO3d7C+MzFDXW2mprrd9aGwB+FaKWUfU+B/PrY8CyUNsM5T0ebaH+NlBmjCkNtsquAZ7ttc2zQOfogKuAV0J96KIh2Cf2MLDFWntfiG0mdPb7G2NOwXnf66NX5QfqyTDGZHYu4+wY29Rrs2eBG4KjYE4DGjq7EWIoZKtmtL3H3XT/vN4I/LmPbV4ELjDGZAe7DS4IrosJY8xFwN3AZdbaoyG2CeczFDW99vdcEaKWcPIlmj4MbLXWVvR155Df40jv+R3CnuJLcEaQ7AC+Hlx3L84HDCAV5+f3duAtYFqM6z0T5yfcO8D64OUS4DbgtuA2nwc24+xtXwWcHuOapwVr2RCsq/N97l6zwTld4Q5gI7AwxjWn44T0uG7rRtV7jPOFUwV04LQKb8bZ3/MPYFvwOie47ULgoW6P/XTwM70duCnGNW/H6Xvu/Dx3jjabCDzf32cohjU/FvycvoMT1EW9aw7e/kC+xKLe4PpHOz+/3bYd9nusaQJERBLIaOt+ERGRYVCoi4gkEIW6iEgCUaiLiCQQhbqISAJRqIuIJBCFuohIAvn/j1cMiv0omDwAAAAASUVORK5CYII=\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": [
    "plt.plot(model.history.history[\"loss\"], label=\"loss\");\n",
    "plt.plot(model.history.history[\"val_loss\"], label=\"val_loss\");\n",
    "plt.legend();\n",
    "plt.show();\n",
    "plt.close();\n",
    "\n",
    "plt.plot(model.history.history[\"mae\"], label=\"mae\");\n",
    "plt.plot(model.history.history[\"val_mae\"], label=\"val_mae\");\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [Use Transfer Learning](https://www.analyticsvidhya.com/blog/2020/03/pretrained-word-embeddings-nlp/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded 400000 word vectors.\n"
     ]
    }
   ],
   "source": [
    "# load the whole embedding into memory\n",
    "embeddings_index = dict()\n",
    "\n",
    "with open(\"glove/glove.6B.300d.txt\") as f:\n",
    "\n",
    "    for line in f:\n",
    "        values = line.split()\n",
    "        word = values[0]\n",
    "        coefs = np.asarray(values[1:], dtype='float32')\n",
    "        embeddings_index[word] = coefs\n",
    "\n",
    "print('Loaded %s word vectors.' % len(embeddings_index))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a weight matrix for words in training docs\n",
    "embedding_matrix = np.zeros((size_of_vocabulary, 300))\n",
    "\n",
    "for word, i in tokenizer.word_index.items():\n",
    "    embedding_vector = embeddings_index.get(word)\n",
    "    if embedding_vector is not None:\n",
    "        embedding_matrix[i] = embedding_vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_2 (Embedding)      (None, 14, 300)           1739700   \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, 14, 128)           219648    \n",
      "_________________________________________________________________\n",
      "global_max_pooling1d_2 (Glob (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 64)                8256      \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 1,967,669\n",
      "Trainable params: 227,969\n",
      "Non-trainable params: 1,739,700\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "model=Sequential()\n",
    "\n",
    "#embedding layer\n",
    "model.add(Embedding(size_of_vocabulary,\n",
    "                    300,\n",
    "                    weights=[embedding_matrix],\n",
    "                    input_length=max_length,\n",
    "                    trainable=False)) \n",
    "\n",
    "#lstm layer\n",
    "model.add(LSTM(128,return_sequences=True,dropout=0.2))\n",
    "\n",
    "#Global Maxpooling\n",
    "model.add(GlobalMaxPooling1D())\n",
    "\n",
    "#Dense Layer\n",
    "model.add(Dense(64,activation='relu')) \n",
    "model.add(Dense(1,activation='relu')) \n",
    "\n",
    "model.compile(optimizer=\"RMSprop\", loss='mean_squared_error', metrics=[\"mae\"]) \n",
    "\n",
    "#Print summary of model\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 4652 samples, validate on 1164 samples\n",
      "Epoch 1/1000\n",
      "4652/4652 [==============================] - 2s 514us/step - loss: 19.6867 - mae: 3.4965 - val_loss: 8.5701 - val_mae: 2.3782\n",
      "Epoch 2/1000\n",
      "4652/4652 [==============================] - 1s 285us/step - loss: 7.7170 - mae: 2.2242 - val_loss: 7.6070 - val_mae: 2.1973\n",
      "Epoch 3/1000\n",
      "4652/4652 [==============================] - 2s 360us/step - loss: 7.0111 - mae: 2.0968 - val_loss: 7.8980 - val_mae: 2.1984\n",
      "Epoch 4/1000\n",
      "4652/4652 [==============================] - 2s 352us/step - loss: 6.5527 - mae: 2.0195 - val_loss: 7.1523 - val_mae: 2.0933\n",
      "Epoch 5/1000\n",
      "4652/4652 [==============================] - 2s 331us/step - loss: 6.2012 - mae: 1.9522 - val_loss: 7.8042 - val_mae: 2.2389\n",
      "Epoch 6/1000\n",
      "4652/4652 [==============================] - 1s 299us/step - loss: 5.9721 - mae: 1.9210 - val_loss: 6.4697 - val_mae: 1.9965\n",
      "Epoch 7/1000\n",
      "4652/4652 [==============================] - 1s 298us/step - loss: 5.5651 - mae: 1.8556 - val_loss: 6.7753 - val_mae: 2.0436\n",
      "Epoch 8/1000\n",
      "4652/4652 [==============================] - 1s 304us/step - loss: 5.5614 - mae: 1.8435 - val_loss: 6.6136 - val_mae: 2.0282\n",
      "\n",
      "Epoch 00008: ReduceLROnPlateau reducing learning rate to 0.0005000000237487257.\n",
      "Epoch 9/1000\n",
      "4652/4652 [==============================] - 1s 292us/step - loss: 4.9058 - mae: 1.7378 - val_loss: 6.1703 - val_mae: 1.9540\n",
      "Epoch 10/1000\n",
      "4652/4652 [==============================] - 1s 289us/step - loss: 4.7331 - mae: 1.6931 - val_loss: 7.0706 - val_mae: 2.0845\n",
      "Epoch 11/1000\n",
      "4652/4652 [==============================] - 1s 292us/step - loss: 4.6993 - mae: 1.6797 - val_loss: 6.4881 - val_mae: 2.0031\n",
      "\n",
      "Epoch 00011: ReduceLROnPlateau reducing learning rate to 0.0002500000118743628.\n",
      "Epoch 12/1000\n",
      "4652/4652 [==============================] - 1s 310us/step - loss: 4.3697 - mae: 1.6324 - val_loss: 6.0440 - val_mae: 1.9232\n",
      "Epoch 13/1000\n",
      "4652/4652 [==============================] - 1s 297us/step - loss: 4.2990 - mae: 1.6199 - val_loss: 6.1135 - val_mae: 1.9355\n",
      "Epoch 14/1000\n",
      "4652/4652 [==============================] - 1s 310us/step - loss: 4.2293 - mae: 1.6003 - val_loss: 5.9969 - val_mae: 1.9137\n",
      "Epoch 15/1000\n",
      "4652/4652 [==============================] - 1s 292us/step - loss: 4.1560 - mae: 1.5877 - val_loss: 6.0385 - val_mae: 1.9187\n",
      "Epoch 16/1000\n",
      "4652/4652 [==============================] - 1s 296us/step - loss: 4.0929 - mae: 1.5747 - val_loss: 6.0013 - val_mae: 1.9110\n",
      "Epoch 17/1000\n",
      "4652/4652 [==============================] - 2s 338us/step - loss: 4.0298 - mae: 1.5670 - val_loss: 5.9626 - val_mae: 1.9081\n",
      "Epoch 18/1000\n",
      "4652/4652 [==============================] - 1s 290us/step - loss: 3.9636 - mae: 1.5464 - val_loss: 5.9491 - val_mae: 1.9046\n",
      "Epoch 19/1000\n",
      "4652/4652 [==============================] - 1s 291us/step - loss: 3.8961 - mae: 1.5381 - val_loss: 6.0595 - val_mae: 1.9196\n",
      "Epoch 20/1000\n",
      "4652/4652 [==============================] - 1s 292us/step - loss: 3.7866 - mae: 1.5160 - val_loss: 5.9267 - val_mae: 1.9009\n",
      "Epoch 21/1000\n",
      "4652/4652 [==============================] - 1s 289us/step - loss: 3.7740 - mae: 1.5130 - val_loss: 5.9491 - val_mae: 1.9016\n",
      "Epoch 22/1000\n",
      "4652/4652 [==============================] - 1s 297us/step - loss: 3.6469 - mae: 1.4899 - val_loss: 5.9650 - val_mae: 1.8963\n",
      "Epoch 23/1000\n",
      "4652/4652 [==============================] - 1s 291us/step - loss: 3.6456 - mae: 1.4863 - val_loss: 5.9707 - val_mae: 1.9045\n",
      "Epoch 24/1000\n",
      "4652/4652 [==============================] - 1s 291us/step - loss: 3.5677 - mae: 1.4737 - val_loss: 5.9651 - val_mae: 1.9003\n",
      "\n",
      "Epoch 00024: ReduceLROnPlateau reducing learning rate to 0.0001250000059371814.\n",
      "Epoch 25/1000\n",
      "4652/4652 [==============================] - 1s 299us/step - loss: 3.4613 - mae: 1.4499 - val_loss: 5.9356 - val_mae: 1.8946\n",
      "Epoch 26/1000\n",
      "4652/4652 [==============================] - 1s 301us/step - loss: 3.4263 - mae: 1.4376 - val_loss: 6.0432 - val_mae: 1.9119\n",
      "Epoch 27/1000\n",
      "4652/4652 [==============================] - 1s 291us/step - loss: 3.4209 - mae: 1.4440 - val_loss: 5.9282 - val_mae: 1.8963\n",
      "\n",
      "Epoch 00027: ReduceLROnPlateau reducing learning rate to 6.25000029685907e-05.\n",
      "Epoch 28/1000\n",
      "4652/4652 [==============================] - 1s 294us/step - loss: 3.3281 - mae: 1.4302 - val_loss: 5.9377 - val_mae: 1.8977\n",
      "Epoch 29/1000\n",
      "4652/4652 [==============================] - 2s 334us/step - loss: 3.2681 - mae: 1.4055 - val_loss: 5.9255 - val_mae: 1.8955\n",
      "\n",
      "Epoch 00029: ReduceLROnPlateau reducing learning rate to 3.125000148429535e-05.\n",
      "Epoch 30/1000\n",
      "4652/4652 [==============================] - 1s 285us/step - loss: 3.2599 - mae: 1.4052 - val_loss: 5.9291 - val_mae: 1.8959\n",
      "Epoch 31/1000\n",
      "4652/4652 [==============================] - 1s 288us/step - loss: 3.2614 - mae: 1.4118 - val_loss: 5.9366 - val_mae: 1.8958\n",
      "\n",
      "Epoch 00031: ReduceLROnPlateau reducing learning rate to 1.5625000742147677e-05.\n",
      "Epoch 32/1000\n",
      "4652/4652 [==============================] - 1s 290us/step - loss: 3.2435 - mae: 1.4088 - val_loss: 5.9333 - val_mae: 1.8956\n",
      "Epoch 33/1000\n",
      "4652/4652 [==============================] - 1s 302us/step - loss: 3.2444 - mae: 1.4048 - val_loss: 5.9310 - val_mae: 1.8954\n",
      "\n",
      "Epoch 00033: ReduceLROnPlateau reducing learning rate to 1e-05.\n",
      "Epoch 34/1000\n",
      "4652/4652 [==============================] - 1s 296us/step - loss: 3.1859 - mae: 1.3967 - val_loss: 5.9325 - val_mae: 1.8957\n",
      "Epoch 35/1000\n",
      "4652/4652 [==============================] - 1s 297us/step - loss: 3.1934 - mae: 1.3883 - val_loss: 5.9344 - val_mae: 1.8959\n",
      "Epoch 36/1000\n",
      "4652/4652 [==============================] - 1s 304us/step - loss: 3.2792 - mae: 1.4119 - val_loss: 5.9301 - val_mae: 1.8952\n",
      "Epoch 37/1000\n",
      "4652/4652 [==============================] - 1s 289us/step - loss: 3.2101 - mae: 1.4029 - val_loss: 5.9318 - val_mae: 1.8952\n",
      "Epoch 38/1000\n",
      "4652/4652 [==============================] - 1s 294us/step - loss: 3.1879 - mae: 1.3949 - val_loss: 5.9332 - val_mae: 1.8957\n",
      "Epoch 39/1000\n",
      "4652/4652 [==============================] - 1s 289us/step - loss: 3.2674 - mae: 1.4138 - val_loss: 5.9270 - val_mae: 1.8945\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(np.array(x_tr_seq),\n",
    "                    np.array(y_tr),\n",
    "                    batch_size=128,\n",
    "                    epochs=1000,\n",
    "                    validation_data=(np.array(x_val_seq),np.array(y_val)),\n",
    "                    verbose=1,\n",
    "                    callbacks=callbacks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1164/1164 [==============================] - 0s 233us/step\n",
      "The val_mae is 1.894.\n"
     ]
    }
   ],
   "source": [
    "#evaluation \n",
    "_, val_mae = model.evaluate(x_val_seq, y_val)\n",
    "\n",
    "print(\"The val_mae is %.3f.\" % val_mae)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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"
    }
   ],
   "source": [
    "plt.plot(model.history.history[\"loss\"], label=\"loss\");\n",
    "plt.plot(model.history.history[\"val_loss\"], label=\"val_loss\");\n",
    "plt.legend();\n",
    "plt.show();\n",
    "plt.close();\n",
    "\n",
    "plt.plot(model.history.history[\"mae\"], label=\"mae\");\n",
    "plt.plot(model.history.history[\"val_mae\"], label=\"val_mae\");\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "base"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}