Final_W2V_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.116106Z",
     "iopub.status.busy": "2021-12-11T09:30:33.115588Z",
     "iopub.status.idle": "2021-12-11T09:30:33.120567Z",
     "shell.execute_reply": "2021-12-11T09:30:33.119852Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.116069Z"
    },
    "id": "jp1XWVr9bHpu"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import re\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Embedding, Bidirectional, LSTM, GRU, Dense, GlobalMaxPool1D"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4aDYiH3wb0VE"
   },
   "source": [
    "## Loading and Preprocessing of the Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FE--8lGPbHp4"
   },
   "source": [
    "The dataset used for the project is then loaded from the csv file into a pandas dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.123235Z",
     "iopub.status.busy": "2021-12-11T09:30:33.122641Z",
     "iopub.status.idle": "2021-12-11T09:30:33.172354Z",
     "shell.execute_reply": "2021-12-11T09:30:33.171642Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.123197Z"
    },
    "id": "5Rh2U1x0bHp7"
   },
   "outputs": [],
   "source": [
    "counseldf = pd.read_csv('/content/counselchat-data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F_1njmSFbHp8"
   },
   "source": [
    "The below function prints the columns available within the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.173887Z",
     "iopub.status.busy": "2021-12-11T09:30:33.173539Z",
     "iopub.status.idle": "2021-12-11T09:30:33.179191Z",
     "shell.execute_reply": "2021-12-11T09:30:33.178475Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.173848Z"
    },
    "id": "zedw31o5bHp9",
    "outputId": "4898a6a8-a310-4b49-c8a3-538f8dfa7bf6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['questionID', 'questionTitle', 'questionText', 'questionUrl', 'topics',\n",
      "       'therapistName', 'therapistUrl', 'answerText', 'upvotes'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "print(counseldf.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "n9ccFoRHbHp-"
   },
   "source": [
    "Below the first 5 samples within the dataset are displayed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 539
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.181141Z",
     "iopub.status.busy": "2021-12-11T09:30:33.180708Z",
     "iopub.status.idle": "2021-12-11T09:30:33.204313Z",
     "shell.execute_reply": "2021-12-11T09:30:33.203456Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.181091Z"
    },
    "id": "b0VHcVK0bHqA",
    "outputId": "9744a9a7-b98f-40b1-cdee-fae7e0d1922f"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>questionID</th>\n",
       "      <th>questionTitle</th>\n",
       "      <th>questionText</th>\n",
       "      <th>questionUrl</th>\n",
       "      <th>topics</th>\n",
       "      <th>therapistName</th>\n",
       "      <th>therapistUrl</th>\n",
       "      <th>answerText</th>\n",
       "      <th>upvotes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5566fab2a64752d71ec3ca69</td>\n",
       "      <td>Escalating disagreements between mother and wife</td>\n",
       "      <td>My wife and mother are having tense disagreeme...</td>\n",
       "      <td>https://counselchat.com/questions/escalating-d...</td>\n",
       "      <td>Family Conflict</td>\n",
       "      <td>Kristi King-Morgan, LMSW</td>\n",
       "      <td>https://counselchat.com/therapists/kristi-king...</td>\n",
       "      <td>&lt;p&gt;What you are describing is something psycho...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5566f94fa64752d71ec3ca64</td>\n",
       "      <td>I'm addicted to smoking. How can I stop?</td>\n",
       "      <td>I'm planning to have baby, so I have to quit s...</td>\n",
       "      <td>https://counselchat.com/questions/i-m-addicted...</td>\n",
       "      <td>Substance Abuse,Addiction</td>\n",
       "      <td>Rebecca Duellman</td>\n",
       "      <td>https://counselchat.com/therapists/rebecca-due...</td>\n",
       "      <td>&lt;p&gt;Hi. Good for you in planning ahead to do wh...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5567d26887a1cc0c3f3d8f46</td>\n",
       "      <td>Keeping secrets from my family</td>\n",
       "      <td>I have secrets in my mind, and I don't know wh...</td>\n",
       "      <td>https://counselchat.com/questions/keeping-secr...</td>\n",
       "      <td>Family Conflict</td>\n",
       "      <td>Jeevna Bajaj</td>\n",
       "      <td>https://counselchat.com/therapists/jeevna-bajaj</td>\n",
       "      <td>&lt;p&gt;It sounds like keeping the secrets has beco...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>556bed15c969ba5861709df5</td>\n",
       "      <td>The Underlying Causes of Being Possessive</td>\n",
       "      <td>I am extremely possessive in my relationships ...</td>\n",
       "      <td>https://counselchat.com/questions/the-underlyi...</td>\n",
       "      <td>Behavioral Change,Social Relationships</td>\n",
       "      <td>Rebecca Duellman</td>\n",
       "      <td>https://counselchat.com/therapists/rebecca-due...</td>\n",
       "      <td>&lt;p&gt;Hi there. It's great you are able to realiz...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>556ba115c969ba5861709de6</td>\n",
       "      <td>Can I control anxiety without medication?</td>\n",
       "      <td>I had a head injury a few years ago and my min...</td>\n",
       "      <td>https://counselchat.com/questions/can-i-contro...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>Rebecca Duellman</td>\n",
       "      <td>https://counselchat.com/therapists/rebecca-due...</td>\n",
       "      <td>&lt;p&gt;You didn't say what or how many medications...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 questionID  ... upvotes\n",
       "0  5566fab2a64752d71ec3ca69  ...       0\n",
       "1  5566f94fa64752d71ec3ca64  ...       0\n",
       "2  5567d26887a1cc0c3f3d8f46  ...       0\n",
       "3  556bed15c969ba5861709df5  ...       0\n",
       "4  556ba115c969ba5861709de6  ...       0\n",
       "\n",
       "[5 rows x 9 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "e4IFgqkVbHqB"
   },
   "source": [
    "As the dataset contains some columns with personal information about the therapist, those columns are removed. Moreover, for the purposes of this project the only relevant columns are 'questionText' and 'topics'. Hence all other columns except those are removed before further processing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.207166Z",
     "iopub.status.busy": "2021-12-11T09:30:33.206883Z",
     "iopub.status.idle": "2021-12-11T09:30:33.218059Z",
     "shell.execute_reply": "2021-12-11T09:30:33.217017Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.207128Z"
    },
    "id": "sLj8AazqbHqE"
   },
   "outputs": [],
   "source": [
    "counseldf.drop(['questionID', 'questionTitle', 'questionUrl', 'therapistName', 'therapistUrl','answerText', 'upvotes'], axis = 1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.219993Z",
     "iopub.status.busy": "2021-12-11T09:30:33.219699Z",
     "iopub.status.idle": "2021-12-11T09:30:33.232682Z",
     "shell.execute_reply": "2021-12-11T09:30:33.231916Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.219954Z"
    },
    "id": "TjnVajClbHqH",
    "outputId": "b465481e-2aee-4a0c-cf47-84d6b95e8f2a",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>questionText</th>\n",
       "      <th>topics</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>My wife and mother are having tense disagreeme...</td>\n",
       "      <td>Family Conflict</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I'm planning to have baby, so I have to quit s...</td>\n",
       "      <td>Substance Abuse,Addiction</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>I have secrets in my mind, and I don't know wh...</td>\n",
       "      <td>Family Conflict</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I am extremely possessive in my relationships ...</td>\n",
       "      <td>Behavioral Change,Social Relationships</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I had a head injury a few years ago and my min...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        questionText                                  topics\n",
       "0  My wife and mother are having tense disagreeme...                         Family Conflict\n",
       "1  I'm planning to have baby, so I have to quit s...               Substance Abuse,Addiction\n",
       "2  I have secrets in my mind, and I don't know wh...                         Family Conflict\n",
       "3  I am extremely possessive in my relationships ...  Behavioral Change,Social Relationships\n",
       "4  I had a head injury a few years ago and my min...                                 Anxiety"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.234741Z",
     "iopub.status.busy": "2021-12-11T09:30:33.234146Z",
     "iopub.status.idle": "2021-12-11T09:30:33.242986Z",
     "shell.execute_reply": "2021-12-11T09:30:33.241427Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.234695Z"
    },
    "id": "UM4QBgs3bHqJ",
    "outputId": "754012a5-c1bb-4af3-cc61-3667d4fcee41"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1482, 2)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.245147Z",
     "iopub.status.busy": "2021-12-11T09:30:33.244649Z",
     "iopub.status.idle": "2021-12-11T09:30:33.254948Z",
     "shell.execute_reply": "2021-12-11T09:30:33.253970Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.245107Z"
    },
    "id": "s8bcAQj3bHqK",
    "outputId": "3e6c3031-fee1-4e4d-824f-028047e81864"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "questionText    99\n",
       "topics          10\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Ro2WO9Z9bHqL"
   },
   "source": [
    "The datset contains some null values in both columns. To avoid the problem created by null values, the samples containing them are simply removed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.256858Z",
     "iopub.status.busy": "2021-12-11T09:30:33.256543Z",
     "iopub.status.idle": "2021-12-11T09:30:33.264676Z",
     "shell.execute_reply": "2021-12-11T09:30:33.263180Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.256820Z"
    },
    "id": "Qn_TIqWTbHqN"
   },
   "outputs": [],
   "source": [
    "counseldf = counseldf.dropna(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.268437Z",
     "iopub.status.busy": "2021-12-11T09:30:33.267323Z",
     "iopub.status.idle": "2021-12-11T09:30:33.278831Z",
     "shell.execute_reply": "2021-12-11T09:30:33.276982Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.268378Z"
    },
    "id": "0MZp8eVCbHqO",
    "outputId": "c48090ff-8594-4462-bf82-39b1e11584fd"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "questionText    0\n",
       "topics          0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.285943Z",
     "iopub.status.busy": "2021-12-11T09:30:33.285720Z",
     "iopub.status.idle": "2021-12-11T09:30:33.300352Z",
     "shell.execute_reply": "2021-12-11T09:30:33.299607Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.285912Z"
    },
    "id": "UrwvBrdYbHqP",
    "outputId": "9b56b9f5-f43a-4218-a85d-63fe992b73a5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1376, 2)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "29tnwAK2bHqS"
   },
   "source": [
    "Finally, we have the clean dataset which can now be processed with NLP techniques. After removing rows with null values, the size of the dataset remains at 1376 samples.\n",
    "\n",
    "The 'topics' column is the target categories for the samples. As seen in the first few lines, some of the samples contain multiple labels. An LSTM network was tried to train using multiple labels but as the model failed to converge, it was decided that only the first label would be kept as the target label of the sample if it has multiple labels. The below code does just that."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.302112Z",
     "iopub.status.busy": "2021-12-11T09:30:33.301742Z",
     "iopub.status.idle": "2021-12-11T09:30:33.330337Z",
     "shell.execute_reply": "2021-12-11T09:30:33.329166Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.302071Z"
    },
    "id": "SAvC9hydbHqU"
   },
   "outputs": [],
   "source": [
    "counseldf['topics'] = counseldf['topics'].str.split(',')\n",
    "counseldf['topics'] = counseldf['topics'].apply(lambda x: x[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.332706Z",
     "iopub.status.busy": "2021-12-11T09:30:33.332077Z",
     "iopub.status.idle": "2021-12-11T09:30:33.351340Z",
     "shell.execute_reply": "2021-12-11T09:30:33.350450Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.332665Z"
    },
    "id": "hPKUTDl_bHqV",
    "outputId": "63369e8b-9c80-4a7a-bb2c-9eaea4e416de"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>questionText</th>\n",
       "      <th>topics</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>My wife and mother are having tense disagreeme...</td>\n",
       "      <td>Family Conflict</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I'm planning to have baby, so I have to quit s...</td>\n",
       "      <td>Substance Abuse</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>I have secrets in my mind, and I don't know wh...</td>\n",
       "      <td>Family Conflict</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I am extremely possessive in my relationships ...</td>\n",
       "      <td>Behavioral Change</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I had a head injury a few years ago and my min...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        questionText             topics\n",
       "0  My wife and mother are having tense disagreeme...    Family Conflict\n",
       "1  I'm planning to have baby, so I have to quit s...    Substance Abuse\n",
       "2  I have secrets in my mind, and I don't know wh...    Family Conflict\n",
       "3  I am extremely possessive in my relationships ...  Behavioral Change\n",
       "4  I had a head injury a few years ago and my min...            Anxiety"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.357466Z",
     "iopub.status.busy": "2021-12-11T09:30:33.356967Z",
     "iopub.status.idle": "2021-12-11T09:30:33.368600Z",
     "shell.execute_reply": "2021-12-11T09:30:33.367849Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.357395Z"
    },
    "id": "fik7KeuFbHqW",
    "outputId": "4149ec13-d36a-4a68-c8d5-8584063baed9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Relationships                246\n",
       "Anxiety                      178\n",
       "Family Conflict              113\n",
       "Depression                    97\n",
       "Marriage                      89\n",
       "Self-esteem                   64\n",
       "Parenting                     59\n",
       "Trauma                        51\n",
       "Human Sexuality               45\n",
       "Behavioral Change             43\n",
       "Intimacy                      42\n",
       "Relationship Dissolution      42\n",
       "Counseling Fundamentals       39\n",
       "Social Relationships          38\n",
       "Anger Management              31\n",
       "Professional Ethics           28\n",
       "Stress                        24\n",
       "LGBTQ                         24\n",
       "Spirituality                  19\n",
       "Addiction                     14\n",
       "Grief and Loss                14\n",
       "Substance Abuse               14\n",
       "Legal & Regulatory            11\n",
       "Diagnosis                     10\n",
       "Sleep Improvement             10\n",
       "Domestic Violence              9\n",
       "Workplace Relationships        8\n",
       "Eating Disorders               4\n",
       "Children & Adolescents         3\n",
       "Military Issues                3\n",
       "Alzheimer's                    2\n",
       "Career Counseling              2\n",
       "Name: topics, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf['topics'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 593
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.370860Z",
     "iopub.status.busy": "2021-12-11T09:30:33.370095Z",
     "iopub.status.idle": "2021-12-11T09:30:33.952091Z",
     "shell.execute_reply": "2021-12-11T09:30:33.951376Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.370820Z"
    },
    "id": "AVmDryFNbHqX",
    "outputId": "facfa1ce-d483-4fa9-af6c-fa16b206a3ed"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(20, 10))\n",
    "counseldf.groupby('topics').agg('count').sort_values('questionText', ascending=False).plot.bar(ax=ax)\n",
    "ax.set_title(\"Number of Patient Questions by Topic\", fontsize=25)\n",
    "ax.set_ylabel(\"Number of Questions\", fontsize=20)\n",
    "ax.set_xlabel(\"Topic\", fontsize=20)\n",
    "ax.set_xticklabels(ax.get_xticklabels(), fontsize=15)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2USbssAmbHqY"
   },
   "source": [
    "After this initial analysis to look at the overall dataset, the 'topics' column could be one-hot encoded so that it is ready for the learning model to take as targets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 261
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:30:33.953294Z",
     "iopub.status.busy": "2021-12-11T09:30:33.953062Z",
     "iopub.status.idle": "2021-12-11T09:30:33.974211Z",
     "shell.execute_reply": "2021-12-11T09:30:33.973456Z",
     "shell.execute_reply.started": "2021-12-11T09:30:33.953260Z"
    },
    "id": "24hmE7yQbHqZ",
    "outputId": "aa0c7578-c727-410f-eabf-9172c843ab08"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Addiction</th>\n",
       "      <th>Alzheimer's</th>\n",
       "      <th>Anger Management</th>\n",
       "      <th>Anxiety</th>\n",
       "      <th>Behavioral Change</th>\n",
       "      <th>Career Counseling</th>\n",
       "      <th>Children &amp; Adolescents</th>\n",
       "      <th>Counseling Fundamentals</th>\n",
       "      <th>Depression</th>\n",
       "      <th>Diagnosis</th>\n",
       "      <th>Domestic Violence</th>\n",
       "      <th>Eating Disorders</th>\n",
       "      <th>Family Conflict</th>\n",
       "      <th>Grief and Loss</th>\n",
       "      <th>Human Sexuality</th>\n",
       "      <th>Intimacy</th>\n",
       "      <th>LGBTQ</th>\n",
       "      <th>Legal &amp; Regulatory</th>\n",
       "      <th>Marriage</th>\n",
       "      <th>Military Issues</th>\n",
       "      <th>Parenting</th>\n",
       "      <th>Professional Ethics</th>\n",
       "      <th>Relationship Dissolution</th>\n",
       "      <th>Relationships</th>\n",
       "      <th>Self-esteem</th>\n",
       "      <th>Sleep Improvement</th>\n",
       "      <th>Social Relationships</th>\n",
       "      <th>Spirituality</th>\n",
       "      <th>Stress</th>\n",
       "      <th>Substance Abuse</th>\n",
       "      <th>Trauma</th>\n",
       "      <th>Workplace Relationships</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Addiction  Alzheimer's  ...  Trauma  Workplace Relationships\n",
       "0          0            0  ...       0                        0\n",
       "1          0            0  ...       0                        0\n",
       "2          0            0  ...       0                        0\n",
       "3          0            0  ...       0                        0\n",
       "4          0            0  ...       0                        0\n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "targetdf = pd.get_dummies(counseldf['topics'])\n",
    "targetdf.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dByCDyGqbHqb"
   },
   "source": [
    "Now, finally it is time to apply natural language processing techniques to each of the individual text samples.\n",
    "\n",
    "First, we define a pre-process function."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "W4YbdtsBcJIL"
   },
   "source": [
    "## Stopwords Removal and Normalisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:36:25.619045Z",
     "iopub.status.busy": "2021-12-11T09:36:25.618298Z",
     "iopub.status.idle": "2021-12-11T09:36:25.632658Z",
     "shell.execute_reply": "2021-12-11T09:36:25.631692Z",
     "shell.execute_reply.started": "2021-12-11T09:36:25.619007Z"
    },
    "id": "ebxywD-YbHqb",
    "outputId": "1ff6295e-61b1-47b7-943a-49651a4c3cd1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
      "[nltk_data]   Unzipping corpora/stopwords.zip.\n",
      "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
      "[nltk_data]   Unzipping tokenizers/punkt.zip.\n",
      "[nltk_data] Downloading package wordnet to /root/nltk_data...\n",
      "[nltk_data]   Unzipping corpora/wordnet.zip.\n"
     ]
    }
   ],
   "source": [
    "#Preprocess function\n",
    "\n",
    "\n",
    "import nltk, re\n",
    "nltk.download('stopwords')\n",
    "nltk.download('punkt')\n",
    "nltk.download('wordnet')\n",
    "from nltk.corpus import wordnet\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tokenize import word_tokenize\n",
    "from nltk.stem import WordNetLemmatizer\n",
    "from collections import Counter\n",
    "\n",
    "stop_words = stopwords.words('english')\n",
    "normalizer = WordNetLemmatizer()\n",
    "\n",
    "def get_part_of_speech(word):\n",
    "    probable_part_of_speech = wordnet.synsets(word)\n",
    "    pos_counts = Counter()\n",
    "    pos_counts[\"n\"] = len(  [ item for item in probable_part_of_speech if item.pos()==\"n\"]  )\n",
    "    pos_counts[\"v\"] = len(  [ item for item in probable_part_of_speech if item.pos()==\"v\"]  )\n",
    "    pos_counts[\"a\"] = len(  [ item for item in probable_part_of_speech if item.pos()==\"a\"]  )\n",
    "    pos_counts[\"r\"] = len(  [ item for item in probable_part_of_speech if item.pos()==\"r\"]  )\n",
    "    most_likely_part_of_speech = pos_counts.most_common(1)[0][0]\n",
    "    return most_likely_part_of_speech\n",
    "\n",
    "def preprocess_text(text):\n",
    "    cleaned = re.sub(r'\\W+', ' ', text).lower()\n",
    "    tokenized = word_tokenize(cleaned)\n",
    "    normalized = [normalizer.lemmatize(token, get_part_of_speech(token)) for token in tokenized]\n",
    "    return normalized"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "O6LynVkZbHqe"
   },
   "source": [
    "The few cells below show the NLP pre-processing techniques applied to the first samples of text."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 70
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:36:28.567969Z",
     "iopub.status.busy": "2021-12-11T09:36:28.567313Z",
     "iopub.status.idle": "2021-12-11T09:36:28.573200Z",
     "shell.execute_reply": "2021-12-11T09:36:28.572481Z",
     "shell.execute_reply.started": "2021-12-11T09:36:28.567933Z"
    },
    "id": "9pGkytUwbHqe",
    "outputId": "43bd3bb6-7d5a-4c30-8c16-c5b73ff417b9"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'My wife and mother are having tense disagreements. In the past, they’ve had minor differences. For example, my wife would complain to me my mother is too overbearing; my mother would complain my wife is lazy.\\r\\nHowever, it’s intensified lately. I think the cause is my wife talked back to her once. Now, any little disagreement is magnified, leading to major disagreements. What can I do?'"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf['questionText'][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:36:29.138562Z",
     "iopub.status.busy": "2021-12-11T09:36:29.138114Z",
     "iopub.status.idle": "2021-12-11T09:36:29.143658Z",
     "shell.execute_reply": "2021-12-11T09:36:29.142893Z",
     "shell.execute_reply.started": "2021-12-11T09:36:29.138524Z"
    },
    "id": "Dj7HhqkVbHqf",
    "outputId": "bc9ecbfa-3f02-46ee-81a6-c86fcb0bcc14"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "my wife and mother are having tense disagreements in the past they ve had minor differences for example my wife would complain to me my mother is too overbearing my mother would complain my wife is lazy however it s intensified lately i think the cause is my wife talked back to her once now any little disagreement is magnified leading to major disagreements what can i do \n"
     ]
    }
   ],
   "source": [
    "text = counseldf['questionText'][0]\n",
    "cleaned = re.sub(r'\\W+', ' ', text).lower()\n",
    "print(cleaned)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gBawgkUKcP5L"
   },
   "source": [
    "## Tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:36:29.977355Z",
     "iopub.status.busy": "2021-12-11T09:36:29.977105Z",
     "iopub.status.idle": "2021-12-11T09:36:29.982839Z",
     "shell.execute_reply": "2021-12-11T09:36:29.982162Z",
     "shell.execute_reply.started": "2021-12-11T09:36:29.977327Z"
    },
    "id": "lbOHsSGZbHqg",
    "outputId": "fb27ca98-7637-4d3c-97c0-3c33a537907f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['my', 'wife', 'and', 'mother', 'are', 'having', 'tense', 'disagreements', 'in', 'the', 'past', 'they', 've', 'had', 'minor', 'differences', 'for', 'example', 'my', 'wife', 'would', 'complain', 'to', 'me', 'my', 'mother', 'is', 'too', 'overbearing', 'my', 'mother', 'would', 'complain', 'my', 'wife', 'is', 'lazy', 'however', 'it', 's', 'intensified', 'lately', 'i', 'think', 'the', 'cause', 'is', 'my', 'wife', 'talked', 'back', 'to', 'her', 'once', 'now', 'any', 'little', 'disagreement', 'is', 'magnified', 'leading', 'to', 'major', 'disagreements', 'what', 'can', 'i', 'do']\n"
     ]
    }
   ],
   "source": [
    "tokenized = word_tokenize(cleaned)\n",
    "print(tokenized)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Xhqa-uZIcTN2"
   },
   "source": [
    "## normalizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:36:30.903959Z",
     "iopub.status.busy": "2021-12-11T09:36:30.903449Z",
     "iopub.status.idle": "2021-12-11T09:36:30.912322Z",
     "shell.execute_reply": "2021-12-11T09:36:30.911481Z",
     "shell.execute_reply.started": "2021-12-11T09:36:30.903918Z"
    },
    "id": "AjXsENnXbHqh",
    "outputId": "7945f9ad-680e-4344-f670-9f19785fb723"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['my', 'wife', 'and', 'mother', 'be', 'have', 'tense', 'disagreement', 'in', 'the', 'past', 'they', 've', 'have', 'minor', 'difference', 'for', 'example', 'my', 'wife', 'would', 'complain', 'to', 'me', 'my', 'mother', 'be', 'too', 'overbear', 'my', 'mother', 'would', 'complain', 'my', 'wife', 'be', 'lazy', 'however', 'it', 's', 'intensify', 'lately', 'i', 'think', 'the', 'cause', 'be', 'my', 'wife', 'talk', 'back', 'to', 'her', 'once', 'now', 'any', 'little', 'disagreement', 'be', 'magnify', 'lead', 'to', 'major', 'disagreement', 'what', 'can', 'i', 'do']\n"
     ]
    }
   ],
   "source": [
    "normalized = [normalizer.lemmatize(token, get_part_of_speech(token)) for token in tokenized]\n",
    "print(normalized)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "IAeX4r0cbHqh"
   },
   "source": [
    "It is clear from the above few cells what pre-processing does to each of the samples. Below the same process is applied to the entire dataset calling the pre-process function defined above."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qWEmf3Y-cWS6"
   },
   "source": [
    "## Stop words Removal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:36:32.487449Z",
     "iopub.status.busy": "2021-12-11T09:36:32.487171Z",
     "iopub.status.idle": "2021-12-11T09:36:36.256963Z",
     "shell.execute_reply": "2021-12-11T09:36:36.256274Z",
     "shell.execute_reply.started": "2021-12-11T09:36:32.487400Z"
    },
    "id": "9swA3aFUbHqi"
   },
   "outputs": [],
   "source": [
    "processed_questionText = counseldf['questionText'].apply(lambda x: preprocess_text(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tXVPGDAIbHqj"
   },
   "source": [
    "After pre-processing, stop words removal can be done to each of the samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:35:53.807502Z",
     "iopub.status.busy": "2021-12-11T09:35:53.807232Z",
     "iopub.status.idle": "2021-12-11T09:35:53.823241Z",
     "shell.execute_reply": "2021-12-11T09:35:53.822393Z",
     "shell.execute_reply.started": "2021-12-11T09:35:53.807471Z"
    },
    "id": "_DSmgdaFbHqj"
   },
   "outputs": [],
   "source": [
    "stop_words = set(stopwords.words('english'))\n",
    "\n",
    "questionText_nostops = []\n",
    "for title in processed_questionText:\n",
    "    text_no_stops = [word for word in title if word not in stop_words]\n",
    "    questionText_nostops.append(text_no_stops)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tsZ18By0bHqk"
   },
   "source": [
    "Below is the first sample with stop words removed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:35:55.528504Z",
     "iopub.status.busy": "2021-12-11T09:35:55.528229Z",
     "iopub.status.idle": "2021-12-11T09:35:55.533060Z",
     "shell.execute_reply": "2021-12-11T09:35:55.532349Z",
     "shell.execute_reply.started": "2021-12-11T09:35:55.528475Z"
    },
    "id": "G2gsNrP5bHqk",
    "outputId": "33999ee7-9f89-44f2-baa8-5970e39bd36c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['wife', 'mother', 'tense', 'disagreement', 'past', 'minor', 'difference', 'example', 'wife', 'would', 'complain', 'mother', 'overbear', 'mother', 'would', 'complain', 'wife', 'lazy', 'however', 'intensify', 'lately', 'think', 'cause', 'wife', 'talk', 'back', 'little', 'disagreement', 'magnify', 'lead', 'major', 'disagreement']\n"
     ]
    }
   ],
   "source": [
    "print(questionText_nostops[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gVeNsXDXbHql"
   },
   "source": [
    "Finally, with the tokens available for all samples, it is time to create the word-index dictionary for changing the data to numerical values. To do this, Tensorflow's tokenizer class will be used."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Da1o7XKJld6Q",
    "outputId": "bada5125-c68c-483a-a1e7-a866b4bba1d2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sample tokenized: [68, 101, 1760, 971, 49, 858, 972, 684, 68, 52, 477, 101, 1761, 101, 52, 477, 68, 859, 165, 1386, 150, 9, 216, 68, 18, 41, 166, 971, 1762, 353, 626, 971]\n",
      "==================================================================================================================\n",
      "\n",
      "Questions max length is 220 words\n",
      "==================================================================================================================\n",
      "\n",
      "Input data shape -> (1376, 220)\n",
      "\n",
      "Input data sample->\n",
      " [  68  101 1760  971   49  858  972  684   68   52  477  101 1761  101\n",
      "   52  477   68  859  165 1386  150    9  216   68   18   41  166  971\n",
      " 1762  353  626  971    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
      "    0    0    0    0    0    0    0    0    0    0]\n",
      "==================================================================================================================\n",
      "\n",
      "Number of Question tokens = 2417\n",
      "\n",
      "Dictionary: {'feel': 1, 'get': 2, 'want': 3, 'like': 4, 'know': 5, 'time': 6, 'go': 7, 'year': 8, 'think': 9, 'say': 10, 'make': 11, 'tell': 12, 'relationship': 13, 'love': 14, 'never': 15, 'really': 16, 'friend': 17, 'talk': 18, 'thing': 19, 'always': 20, 'even': 21, 'work': 22, 'help': 23, 'life': 24, 'boyfriend': 25, 'try': 26, 'find': 27, 'still': 28, 'see': 29, 'one': 30, 'day': 31, 'people': 32, 'need': 33, 'start': 34, 'sex': 35, 'anxiety': 36, 'take': 37, 'family': 38, 'lot': 39, 'month': 40, 'back': 41, 'child': 42, 'much': 43, 'every': 44, 'husband': 45, 'anything': 46, 'stop': 47, 'together': 48, 'past': 49, 'live': 50, 'something': 51, 'would': 52, 'someone': 53, 'leave': 54, 'move': 55, 'problem': 56, 'sometimes': 57, 'two': 58, 'bad': 59, 'school': 60, 'ago': 61, 'lose': 62, 'parent': 63, 'ask': 64, 'also': 65, 'everything': 66, 'girl': 67, 'wife': 68, 'come': 69, 'well': 70, 'keep': 71, 'issue': 72, 'seem': 73, 'break': 74, 'depression': 75, 'cheat': 76, 'dad': 77, 'use': 78, 'happen': 79, 'way': 80, 'wrong': 81, 'stress': 82, 'give': 83, 'away': 84, 'u': 85, 'marry': 86, 'last': 87, 'ex': 88, 'since': 89, 'call': 90, 'guy': 91, 'hurt': 92, 'right': 93, 'normal': 94, 'anymore': 95, 'recently': 96, 'good': 97, 'drink': 98, 'mom': 99, 'night': 100, 'mother': 101, 'end': 102, 'stay': 103, 'date': 104, 'girlfriend': 105, 'home': 106, 'trust': 107, 'woman': 108, 'lie': 109, 'first': 110, 'job': 111, 'cry': 112, 'let': 113, 'deal': 114, 'long': 115, 'believe': 116, 'else': 117, 'week': 118, 'care': 119, 'counseling': 120, 'depress': 121, 'late': 122, 'alone': 123, 'around': 124, 'change': 125, 'happy': 126, 'fight': 127, 'could': 128, 'thought': 129, 'person': 130, 'daughter': 131, 'nothing': 132, 'ever': 133, 'kid': 134, 'afraid': 135, 'angry': 136, 'decide': 137, 'listen': 138, 'without': 139, '20': 140, 'hard': 141, 'meet': 142, 'young': 143, 'couple': 144, 'almost': 145, 'though': 146, 'three': 147, 'felt': 148, 'enough': 149, 'lately': 150, 'win': 151, 'disorder': 152, 'therapy': 153, 'become': 154, 'self': 155, 'men': 156, 'everyone': 157, 'great': 158, 'worry': 159, 'scar': 160, 'therapist': 161, 'attack': 162, 'house': 163, 'understand': 164, 'however': 165, 'little': 166, 'act': 167, 'far': 168, 'new': 169, 'look': 170, 'upset': 171, 'anger': 172, 'sad': 173, 'man': 174, 'mean': 175, 'anyone': 176, 'constantly': 177, 'session': 178, 'mind': 179, 'put': 180, 'abuse': 181, 'forgive': 182, 'best': 183, 'another': 184, 'handle': 185, 'okay': 186, 'teenager': 187, 'voice': 188, 'sleep': 189, 'situation': 190, 'marriage': 191, 'son': 192, 'continue': 193, 'nervous': 194, 'control': 195, '3': 196, 'fiancé': 197, 'five': 198, 'able': 199, 'old': 200, 'phone': 201, 'fear': 202, 'experience': 203, 'counselor': 204, 'many': 205, 'walk': 206, 'decision': 207, 'sure': 208, 'eat': 209, 'reason': 210, 'hate': 211, 'father': 212, 'sexual': 213, 'early': 214, 'four': 215, 'cause': 216, 'miss': 217, 'diagnose': 218, 'head': 219, 'often': 220, 'close': 221, 'divorce': 222, 'spend': 223, 'brother': 224, 'place': 225, 'big': 226, 'plan': 227, 'actually': 228, 'sit': 229, 'nightmare': 230, 'along': 231, 'remember': 232, 'turn': 233, 'guilty': 234, 'word': 235, 'crazy': 236, 'partner': 237, 'tire': 238, 'drive': 239, 'hear': 240, 'personality': 241, 'room': 242, 'heart': 243, 'baby': 244, 'process': 245, 'whenever': 246, 'doctor': 247, 'frustrate': 248, 'others': 249, '30': 250, 'fix': 251, 'learn': 252, 'sister': 253, 'money': 254, 'interest': 255, 'message': 256, 'high': 257, 'client': 258, 'emotion': 259, 'either': 260, 'panic': 261, 'step': 262, 'currently': 263, 'forever': 264, 'gay': 265, 'face': 266, 'alcohol': 267, 'harm': 268, 'emotional': 269, 'addiction': 270, 'hour': 271, 'answer': 272, 'different': 273, 'physically': 274, 'apart': 275, 'kind': 276, 'dog': 277, 'already': 278, 'hit': 279, 'embarrass': 280, 'sleeping': 281, 'order': 282, 'forget': 283, 'boy': 284, 'mad': 285, 'car': 286, 'pain': 287, 'abusive': 288, 'question': 289, 'difficult': 290, 'female': 291, 'attention': 292, 'due': 293, 'show': 294, 'accept': 295, 'medication': 296, 'adult': 297, 'social': 298, 'whole': 299, 'less': 300, 'age': 301, 'terrible': 302, 'virgin': 303, 'everytime': 304, 'shaky': 305, 'emotionally': 306, 'run': 307, 'drug': 308, 'might': 309, 'fine': 310, 'several': 311, 'text': 312, 'fact': 313, 'random': 314, 'lack': 315, 'half': 316, 'matter': 317, 'hopeless': 318, 'saw': 319, 'confuse': 320, 'pretty': 321, 'split': 322, 'stupid': 323, 'push': 324, 'smoking': 325, 'trouble': 326, 'violent': 327, 'horrible': 328, 'instead': 329, 'negative': 330, 'respect': 331, 'enjoy': 332, 'terrify': 333, 'play': 334, 'struggle': 335, 'serious': 336, 'event': 337, 'childhood': 338, '5': 339, 'next': 340, 'throw': 341, 'honestly': 342, 'refuse': 343, 'tear': 344, 'kiss': 345, 'overwhelm': 346, 'attract': 347, 'mess': 348, 'pay': 349, 'weekend': 350, 'seven': 351, 'constant': 352, 'lead': 353, 'sick': 354, 'name': 355, 'straight': 356, 'pregnant': 357, 'kill': 358, '2': 359, 'argument': 360, 'point': 361, 'hang': 362, 'current': 363, 'figure': 364, 'especially': 365, 'die': 366, 'single': 367, 'quit': 368, 'wake': 369, 'low': 370, 'pass': 371, 'cope': 372, 'wall': 373, 'possible': 374, 'paranoid': 375, 'watch': 376, 'approach': 377, 'separate': 378, 'catch': 379, 'admit': 380, 'read': 381, 'agree': 382, 'class': 383, 'bed': 384, 'graduate': 385, 'afford': 386, 'freak': 387, 'terminate': 388, 'completely': 389, 'argue': 390, 'speak': 391, 'dream': 392, 'send': 393, 'twice': 394, 'small': 395, 'advice': 396, 'male': 397, 'comment': 398, 'rap': 399, 'suffer': 400, 'wear': 401, 'touch': 402, 'insecurity': 403, 'real': 404, 'pick': 405, 'distance': 406, 'yet': 407, 'create': 408, 'kick': 409, 'today': 410, 'cut': 411, 'hide': 412, 'grow': 413, 'honest': 414, 'worth': 415, 'fell': 416, 'begin': 417, 'support': 418, 'student': 419, 'promise': 420, 'addict': 421, 'bother': 422, 'yesterday': 423, 'mental': 424, 'body': 425, 'accuse': 426, 'forward': 427, 'please': 428, 'state': 429, 'soon': 430, 'contact': 431, 'second': 432, 'town': 433, 'usually': 434, 'overcome': 435, 'stuff': 436, 'present': 437, 'mid': 438, 'write': 439, 'involve': 440, 'binge': 441, 'suicidal': 442, 'wait': 443, 'daily': 444, 'scream': 445, 'gender': 446, 'weather': 447, 'extremely': 448, 'medicine': 449, 'behavior': 450, 'notice': 451, 'nowhere': 452, 'treat': 453, 'mine': 454, 'part': 455, 'truly': 456, 'communicate': 457, 'rather': 458, 'erection': 459, 'desire': 460, 'mistake': 461, 'college': 462, 'wonder': 463, 'fake': 464, 'idea': 465, 'bipolar': 466, 'summer': 467, 'travel': 468, 'yell': 469, 'true': 470, 'fault': 471, 'realize': 472, 'minute': 473, 'bring': 474, 'video': 475, 'personal': 476, 'complain': 477, 'physical': 478, 'calm': 479, 'unable': 480, 'quite': 481, 'drop': 482, '1': 483, 'christian': 484, 'explain': 485, 'confront': 486, 'disappoint': 487, 'ruin': 488, '4': 489, 'health': 490, 'stick': 491, 'affect': 492, 'hospital': 493, 'blame': 494, 'multiple': 495, 'express': 496, 'god': 497, 'suppose': 498, '0': 499, 'type': 500, 'follow': 501, 'awful': 502, 'road': 503, 'mg': 504, 'require': 505, 'friendship': 506, 'mouth': 507, 'effort': 508, 'comfortable': 509, 'visit': 510, 'maybe': 511, '11': 512, 'healthy': 513, 'death': 514, 'open': 515, 'whether': 516, 'nice': 517, 'hurtful': 518, 'side': 519, 'obsess': 520, 'raise': 521, 'hearing': 522, 'regard': 523, 'worker': 524, 'confident': 525, 'line': 526, 'tough': 527, 'rough': 528, 'country': 529, 'future': 530, 'deep': 531, 'within': 532, 'glad': 533, 'flashback': 534, 'rid': 535, 'race': 536, 'fall': 537, 'develop': 538, 'hand': 539, '15': 540, 'beautiful': 541, 'worthless': 542, 'previous': 543, 'hope': 544, 'effect': 545, 'result': 546, 'connect': 547, 'avoid': 548, 'heaven': 549, 'wonderful': 550, 'six': 551, 'wish': 552, 'grade': 553, 'basically': 554, 'concern': 555, 'distant': 556, 'breakup': 557, 'case': 558, 'maintain': 559, 'couch': 560, 'expect': 561, 'empty': 562, 'shake': 563, 'courage': 564, 'ignore': 565, 'porn': 566, 'main': 567, 'prescription': 568, 'suicide': 569, 'sibling': 570, 'habit': 571, 'shut': 572, 'apologize': 573, 'reach': 574, 'short': 575, 'attempt': 576, 'mentally': 577, 'important': 578, 'focus': 579, 'sudden': 580, 'mood': 581, 'pressure': 582, '40': 583, 'cousin': 584, 'communication': 585, 'abandon': 586, 'towards': 587, 'resolve': 588, 'bully': 589, 'attractive': 590, 'morning': 591, 'conversation': 592, 'everyday': 593, 'offer': 594, 'information': 595, 'naked': 596, 'general': 597, 'nine': 598, 'whatever': 599, 'marijuana': 600, 'extreme': 601, 'must': 602, 'may': 603, 'huge': 604, 'rarely': 605, 'teenage': 606, 'treatment': 607, 'force': 608, 'overreact': 609, 'behind': 610, 'game': 611, 'allow': 612, 'tend': 613, 'except': 614, 'count': 615, 'pornography': 616, 'badly': 617, 'posttraumatic': 618, 'facebook': 619, 'birth': 620, 'adopt': 621, 'enter': 622, 'target': 623, 'winter': 624, 'brief': 625, 'major': 626, 'secret': 627, 'reject': 628, 'intimacy': 629, 'hop': 630, 'suggestion': 631, 'business': 632, 'jealous': 633, 'claim': 634, 'movie': 635, 'although': 636, 'sociopath': 637, 'cop': 638, 'front': 639, 'safety': 640, 'unfortunately': 641, 'insecure': 642, 'character': 643, 'positive': 644, 'busy': 645, 'esteem': 646, 'space': 647, 'cycle': 648, 'sweet': 649, 'hold': 650, 'stand': 651, 'none': 652, 'teen': 653, 'connection': 654, 'sexually': 655, 'check': 656, 'hotel': 657, 'bisexual': 658, 'hat': 659, 'shock': 660, 'action': 661, 'toddler': 662, 'pee': 663, '50': 664, 'university': 665, 'finally': 666, 'neither': 667, 'ready': 668, 'public': 669, 'easy': 670, 'group': 671, 'insurance': 672, 'military': 673, 'coworker': 674, 'us': 675, 'number': 676, 'dose': 677, 'peer': 678, 'drift': 679, 'coexist': 680, 'invest': 681, 'uncared': 682, 'reconnected': 683, 'example': 684, 'annoying': 685, 'stomach': 686, 'easily': 687, 'breathe': 688, '8': 689, 'improve': 690, 'exercise': 691, 'sixty': 692, 'seriously': 693, 'pansexual': 694, 'judge': 695, 'finish': 696, 'texted': 697, 'picture': 698, 'strong': 699, 'sound': 700, 'disgust': 701, 'party': 702, 'etc': 703, 'discus': 704, 'imagine': 705, 'surround': 706, 'lonely': 707, 'online': 708, 'weight': 709, '14': 710, 'diminish': 711, 'trigger': 712, 'separation': 713, 'full': 714, 'twenty': 715, 'chance': 716, 'fail': 717, 'affair': 718, 'co': 719, 'fully': 720, 'literally': 721, 'memory': 722, 'cuss': 723, 'course': 724, 'medium': 725, '10': 726, 'swear': 727, 'form': 728, 'perfect': 729, '25': 730, 'stepdaughter': 731, 'reply': 732, 'amaze': 733, 'crush': 734, 'christmas': 735, 'moment': 736, 'weed': 737, 'excellent': 738, 'mostly': 739, 'influence': 740, 'team': 741, 'miserable': 742, 'fiancée': 743, 'rate': 744, 'guilt': 745, 'consider': 746, 'ptsd': 747, 'chest': 748, 'psychologist': 749, 'disrespectful': 750, 'obsessive': 751, 'transgender': 752, 'track': 753, 'ashamed': 754, 'degree': 755, 'gym': 756, '18': 757, 'uncomfortable': 758, 'slow': 759, 'program': 760, '12': 761, 'proper': 762, 'niece': 763, 'reminder': 764, 'roof': 765, 'rule': 766, 'anxious': 767, 'condition': 768, 'devastate': 769, 'hole': 770, 'compare': 771, 'accident': 772, 'base': 773, 'id': 774, 'card': 775, 'quickly': 776, 'clear': 777, 'choice': 778, 'severe': 779, 'boob': 780, 'unhappy': 781, 'ugly': 782, 'save': 783, 'tonight': 784, 'respond': 785, '100': 786, 'religious': 787, 'somehow': 788, 'world': 789, 'possibly': 790, 'food': 791, 'legal': 792, 'rest': 793, 'bit': 794, 'verbally': 795, 'affection': 796, 'hook': 797, 'sorry': 798, 'threaten': 799, 'unwanted': 800, 'toilet': 801, 'blood': 802, 'drama': 803, 'eye': 804, 'lust': 805, 'wine': 806, 'bear': 807, 'expression': 808, 'texting': 809, 'position': 810, 'ii': 811, 'funny': 812, 'partially': 813, 'basis': 814, 'inside': 815, 'virginity': 816, 'sport': 817, 'anybody': 818, 'add': 819, 'crave': 820, 'violence': 821, 'domestic': 822, 'offend': 823, 'afterwards': 824, 'compulsive': 825, 'sign': 826, 'aunt': 827, 'cancer': 828, 'forth': 829, 'favor': 830, 'beer': 831, 'fridge': 832, 'dead': 833, 'invite': 834, 'shame': 835, 'financial': 836, 'pill': 837, 'post': 838, 'color': 839, 'scary': 840, 'water': 841, 'instructor': 842, 'swimsuit': 843, 'top': 844, 'display': 845, 'dangerous': 846, 'finance': 847, 'daddy': 848, 'unfocused': 849, 'recovery': 850, 'pantyhose': 851, 'schizophrenia': 852, '00': 853, 'occur': 854, 'pawn': 855, 'dependent': 856, 'facilitate': 857, 'minor': 858, 'lazy': 859, 'secure': 860, 'manage': 861, 'asleep': 862, 'symptom': 863, 'custody': 864, 'visitation': 865, 'medical': 866, 'numerous': 867, 'period': 868, 'shortly': 869, 'original': 870, 'inappropriate': 871, 'exist': 872, 'selfish': 873, 'relative': 874, 'apartment': 875, 'regret': 876, 'company': 877, 'poor': 878, 'sense': 879, 'aware': 880, 'clean': 881, 'excite': 882, 'fast': 883, 'unsure': 884, 'role': 885, 'simply': 886, 'genuine': 887, 'belief': 888, 'lunch': 889, 'lock': 890, 'impact': 891, 'throughout': 892, 'convince': 893, 'block': 894, 'anti': 895, 'religion': 896, 'teach': 897, 'hell': 898, 'vehicle': 899, 'cat': 900, 'floor': 901, 'joke': 902, 'appointment': 903, 'least': 904, 'deserve': 905, 'rage': 906, 'household': 907, 'fill': 908, 'poop': 909, 'birthday': 910, 'research': 911, 'detail': 912, 'match': 913, 'laugh': 914, 'adolescent': 915, 'sober': 916, 'gift': 917, 'e': 918, 'elementary': 919, 'quiet': 920, 'trap': 921, 'function': 922, 'loud': 923, 'stranger': 924, 'unless': 925, 'truth': 926, 'oral': 927, 'dirty': 928, 'totally': 929, 'confess': 930, 'remain': 931, 'organization': 932, 'ten': 933, 'lash': 934, 'individual': 935, 'biological': 936, 'instagram': 937, 'transparent': 938, 'orgasmed': 939, 'ehh': 940, 'foreplay': 941, 'tickle': 942, 'pant': 943, 'hug': 944, 'urge': 945, 'uncle': 946, 'disappointment': 947, 'edge': 948, 'stressor': 949, 'middle': 950, 'survivor': 951, 'swallow': 952, 'intimate': 953, 'report': 954, 'resurface': 955, '17': 956, 'factor': 957, 'heel': 958, 'distract': 959, 'policy': 960, 'demean': 961, 'threesome': 962, 'complicate': 963, 'methamphetamine': 964, 'driveway': 965, 'responsibility': 966, 'mentalhealth': 967, 'facility': 968, 'background': 969, 'narrow': 970, 'disagreement': 971, 'difference': 972, 'injury': 973, 'anyway': 974, 'solve': 975, 'manner': 976, 'unacceptable': 977, 'lay': 978, 'reaction': 979, 'steal': 980, '7': 981, 'vacation': 982, 'wedding': 983, 'weird': 984, 'fetish': 985, 'false': 986, 'bos': 987, 'acknowledge': 988, 'professional': 989, 'confidence': 990, 'incident': 991, 'awfully': 992, 'hot': 993, 'sometime': 994, 'belittle': 995, 'include': 996, 'inconsiderate': 997, 'particular': 998, 'closure': 999, 'victim': 1000, 'schedule': 1001, 'obviously': 1002, 'complete': 1003, 'otherwise': 1004, 'transfer': 1005, 'upon': 1006, 'return': 1007, 'similar': 1008, '9': 1009, 'unfaithful': 1010, 'faithful': 1011, 'unsettle': 1012, 'comfort': 1013, 'book': 1014, 'nobody': 1015, 'bothersome': 1016, 'beg': 1017, 'shoot': 1018, 'serve': 1019, 'friday': 1020, 'emergency': 1021, 'financially': 1022, 'drain': 1023, 'episode': 1024, 'beside': 1025, 'sort': 1026, 'tv': 1027, 'challenge': 1028, 'pull': 1029, 'hair': 1030, 'law': 1031, 'psychiatric': 1032, 'sight': 1033, 'infant': 1034, 'human': 1035, 'accusation': 1036, 'burn': 1037, 'besides': 1038, 'parental': 1039, 'tension': 1040, 'despite': 1041, 'commit': 1042, 'suddenly': 1043, 'neglect': 1044, 'irritate': 1045, 'masturbate': 1046, 'smart': 1047, 'deny': 1048, 'demand': 1049, 'draw': 1050, 'fantasize': 1051, 'jealousy': 1052, 'likely': 1053, 'private': 1054, 'bright': 1055, 'gain': 1056, 'light': 1057, 'door': 1058, 'apology': 1059, 'store': 1060, 'file': 1061, 'suggest': 1062, 'strain': 1063, 'loathing': 1064, 'temptation': 1065, 'luckily': 1066, 'surprise': 1067, 'tall': 1068, 'socialize': 1069, 'thanksgiving': 1070, 'eight': 1071, 'level': 1072, 'value': 1073, 'excuse': 1074, 'term': 1075, 'worsen': 1076, 'explode': 1077, 'cheater': 1078, 'pretend': 1079, 'barely': 1080, 'tempt': 1081, 'remarry': 1082, 'animal': 1083, 'snap': 1084, 'ongoing': 1085, 'intercourse': 1086, 'internal': 1087, 'calorie': 1088, 'mechanism': 1089, 'intense': 1090, 'dwell': 1091, 'pride': 1092, 'sarcastic': 1093, 'normally': 1094, 'punishment': 1095, 'vivid': 1096, 'outcome': 1097, 'content': 1098, 'note': 1099, 'trauma': 1100, 'mirror': 1101, 'dysphoria': 1102, 'strategy': 1103, 'silly': 1104, 'fatigue': 1105, 'jumble': 1106, 'skip': 1107, 'beat': 1108, 'exactly': 1109, 'toe': 1110, 'deficit': 1111, 'protection': 1112, 'xanax': 1113, '5mg': 1114, 'perspective': 1115, 'accomplishment': 1116, 'clothe': 1117, 'coverage': 1118, 'hyper': 1119, 'history': 1120, 'shy': 1121, 'presentation': 1122, 'join': 1123, 'dont': 1124, 'st': 1125, 'scare': 1126, 'thearapist': 1127, 'available': 1128, 'autism': 1129, 'pedophile': 1130, 'logically': 1131, 'teacher': 1132, 'tech': 1133, 'overwork': 1134, 'underpay': 1135, 'contribution': 1136, 'demoralize': 1137, 'wreck': 1138, 'gravitate': 1139, 'toward': 1140, 'unavailable': 1141, 'dictate': 1142, 'beck': 1143, 'security': 1144, 'survive': 1145, 'monster': 1146, 'frighten': 1147, 'chronic': 1148, 'release': 1149, 'warn': 1150, 'increase': 1151, 'irritability': 1152, 'belong': 1153, 'pet': 1154, 'paper': 1155, 'certain': 1156, 'goodbye': 1157, 'propose': 1158, 'cook': 1159, 'somewhere': 1160, 'wallet': 1161, '2014': 1162, 'lend': 1163, 'narcissistic': 1164, 'sadly': 1165, 'ring': 1166, 'proposal': 1167, 'apply': 1168, 'strange': 1169, '6': 1170, 'swing': 1171, 'proof': 1172, 'employer': 1173, 'dinner': 1174, 'opinion': 1175, 'behave': 1176, 'service': 1177, 'highly': 1178, 'consume': 1179, 'member': 1180, 'confide': 1181, 'fat': 1182, 'helpless': 1183, 'stab': 1184, 'per': 1185, 'blow': 1186, 'unreasonable': 1187, 'lady': 1188, 'spiral': 1189, 'story': 1190, 'unhealthy': 1191, 'pleasure': 1192, 'sensitive': 1193, 'purpose': 1194, 'grandma': 1195, 'manic': 1196, 'verbal': 1197, 'lesson': 1198, 'rollercoaster': 1199, 'psychiatrist': 1200, 'cruel': 1201, 'homeless': 1202, 'manager': 1203, 'encourage': 1204, 'study': 1205, 'senior': 1206, 'set': 1207, 'rub': 1208, 'nearly': 1209, 'punch': 1210, 'inform': 1211, 'irrational': 1212, 'outgoing': 1213, 'shoulder': 1214, 'repeatedly': 1215, 'lingerie': 1216, 'seek': 1217, 'relate': 1218, 'quick': 1219, 'fulfil': 1220, 'nap': 1221, 'possibility': 1222, 'ideal': 1223, 'international': 1224, 'semester': 1225, 'unite': 1226, 'cultural': 1227, 'arrive': 1228, 'savior': 1229, 'special': 1230, 'appropriate': 1231, '13': 1232, 'dear': 1233, 'flirtation': 1234, 'progress': 1235, 'office': 1236, 'hugger': 1237, 'clearly': 1238, 'prefer': 1239, 'model': 1240, 'system': 1241, 'fathom': 1242, 'dread': 1243, 'dude': 1244, 'submissive': 1245, 'basketball': 1246, 'phrase': 1247, 'slowly': 1248, 'provide': 1249, 'worst': 1250, 'restroom': 1251, 'cover': 1252, 'illness': 1253, 'dispute': 1254, 'stutter': 1255, 'truck': 1256, 'driver': 1257, 'music': 1258, 'super': 1259, 'aspect': 1260, 'actual': 1261, 'site': 1262, 'resist': 1263, 'frequently': 1264, 'previously': 1265, 'properly': 1266, 'rude': 1267, 'absolutely': 1268, 'ton': 1269, 'lesbian': 1270, 'charge': 1271, 'felony': 1272, 'pander': 1273, 'admission': 1274, 'minimize': 1275, 'repent': 1276, 'opposite': 1277, 'defy': 1278, 'authoritative': 1279, 'entire': 1280, 'opportunity': 1281, 'third': 1282, 'stable': 1283, 'stage': 1284, 'probably': 1285, 'borderline': 1286, 'test': 1287, 'curse': 1288, 'rumor': 1289, 'graceful': 1290, 'insist': 1291, 'rehabilitation': 1292, 'brain': 1293, 'friendly': 1294, 'identity': 1295, 'eventually': 1296, 'happily': 1297, 'blend': 1298, 'wage': 1299, 'earner': 1300, 'sacrifice': 1301, 'core': 1302, 'resolution': 1303, 'limit': 1304, 'consumption': 1305, 'board': 1306, 'discover': 1307, 'vodka': 1308, 'snore': 1309, 'giveaway': 1310, 'slur': 1311, 'honesty': 1312, 'evil': 1313, 'football': 1314, 'player': 1315, 'indoor': 1316, 'league': 1317, 'prove': 1318, 'noise': 1319, 'grandson': 1320, 'internally': 1321, 'externally': 1322, 'unidentifiable': 1323, 'cage': 1324, 'loss': 1325, 'rape': 1326, 'fun': 1327, 'sexuality': 1328, 'bisexuality': 1329, 'dreaming': 1330, 'potential': 1331, 'wan': 1332, 'na': 1333, 'inexperienced': 1334, 'differently': 1335, 'stroke': 1336, 'participate': 1337, 'thereafter': 1338, 'overdose': 1339, 'penis': 1340, 'snit': 1341, 'silent': 1342, 'masturbation': 1343, 'star': 1344, 'motivation': 1345, 'heavy': 1346, 'awkward': 1347, 'popular': 1348, 'identify': 1349, 'variation': 1350, 'roughly': 1351, 'substance': 1352, 'skill': 1353, 'heterosexual': 1354, 'skirt': 1355, 'stock': 1356, 'fascination': 1357, 'kinky': 1358, 'vague': 1359, 'empathize': 1360, 'empathy': 1361, 'bra': 1362, 'skin': 1363, 'rent': 1364, 'suspicious': 1365, 'sweetheart': 1366, 'penny': 1367, 'imessage': 1368, 'snapchat': 1369, 'messy': 1370, 'crossdress': 1371, 'feminine': 1372, 'frequent': 1373, 'ego': 1374, 'bedroom': 1375, 'imposter': 1376, 'syndrome': 1377, 'account': 1378, 'phenomenal': 1379, 'publish': 1380, 'faculty': 1381, 'alike': 1382, 'fraud': 1383, 'paul': 1384, 'recluse': 1385, 'intensify': 1386, 'estimate': 1387, 'bankrupt': 1388, 'p': 1389, 'realization': 1390, 'missouri': 1391, 'reconnect': 1392, 'sicken': 1393, 'beyond': 1394, 'horror': 1395, 'pop': 1396, '2013': 1397, 'miscarriage': 1398, 'hardly': 1399, 'amount': 1400, 'former': 1401, 'apologetic': 1402, 'finger': 1403, 'successful': 1404, 'server': 1405, 'hundred': 1406, 'dollar': 1407, 'engage': 1408, 'cant': 1409, 'petty': 1410, 'anniversary': 1411, 'promiscuous': 1412, 'coworkers': 1413, 'statement': 1414, 'pair': 1415, 'doc': 1416, 'foot': 1417, 'hello': 1418, 'city': 1419, 'fuse': 1420, 'temperamental': 1421, 'coordinate': 1422, 'condescend': 1423, 'oblivious': 1424, 'openly': 1425, 'thoughtful': 1426, 'spatially': 1427, 'baker': 1428, 'gather': 1429, 'funeral': 1430, 'custom': 1431, 'status': 1432, 'horribly': 1433, 'register': 1434, 'offender': 1435, 'record': 1436, 'roommate': 1437, 'dump': 1438, 'skinny': 1439, 'libido': 1440, 'encounter': 1441, 'importance': 1442, 'spouse': 1443, 'reassurance': 1444, 'companion': 1445, 'priority': 1446, 'aversion': 1447, 'reconcile': 1448, 'linger': 1449, 'prevent': 1450, 'dessert': 1451, 'tone': 1452, 'jump': 1453, 'recur': 1454, 'blue': 1455, 'dismiss': 1456, 'mask': 1457, 'happiness': 1458, 'creative': 1459, 'passion': 1460, 'louisiana': 1461, 'doubt': 1462, 'dearly': 1463, 'holiday': 1464, 'loan': 1465, 'heavily': 1466, 'medicate': 1467, 'credit': 1468, 'owe': 1469, 'narcissist': 1470, 'spirit': 1471, 'pity': 1472, 'adhd': 1473, 'nut': 1474, 'extend': 1475, 'box': 1476, 'corner': 1477, 'perfectly': 1478, 'defend': 1479, 'deception': 1480, 'willingness': 1481, 'intimidate': 1482, 'silence': 1483, 'immigration': 1484, 'application': 1485, 'slap': 1486, 'supportive': 1487, 'guidance': 1488, 'intervene': 1489, 'wrongly': 1490, 'mommy': 1491, 'cold': 1492, 'lull': 1493, 'chair': 1494, 'diaper': 1495, 'pattern': 1496, 'recommend': 1497, 'phase': 1498, 'assume': 1499, 'erectile': 1500, 'dysfunction': 1501, 'viagra': 1502, 'cialis': 1503, 'discuss': 1504, 'frustration': 1505, 'cancel': 1506, 'smoke': 1507, 'negativity': 1508, 'adore': 1509, 'starve': 1510, 'sorority': 1511, 'patch': 1512, 'babysitter': 1513, 'escalate': 1514, 'gastric': 1515, 'sleeve': 1516, 'surgery': 1517, 'feedback': 1518, 'mail': 1519, 'appearance': 1520, 'facial': 1521, 'psychology': 1522, 'depressive': 1523, 'list': 1524, 'prohibit': 1525, 'depressant': 1526, 'clinic': 1527, 'tick': 1528, 'appetite': 1529, 'immediately': 1530, 'x': 1531, 'ray': 1532, 'spine': 1533, 'bickering': 1534, 'semi': 1535, 'bottom': 1536, 'nerve': 1537, 'independent': 1538, 'unemployed': 1539, 'verge': 1540, 'anytime': 1541, 'ipad': 1542, 'grandbaby': 1543, 'shelf': 1544, 'unit': 1545, 'bake': 1546, 'pie': 1547, 'wrecking': 1548, 'garbage': 1549, 'receive': 1550, 'flare': 1551, 'outside': 1552, 'license': 1553, 'sadness': 1554, 'funk': 1555, 'headache': 1556, 'attitude': 1557, 'somebody': 1558, 'unworthiness': 1559, 'motivate': 1560, 'build': 1561, 'large': 1562, 'debt': 1563, 'borrow': 1564, 'disappear': 1565, 'anywhere': 1566, '000': 1567, 'bill': 1568, 'antisocial': 1569, 'liar': 1570, 'pace': 1571, 'conduct': 1572, 'respectfully': 1573, 'sake': 1574, 'definite': 1575, 'attraction': 1576, 'misery': 1577, 'establish': 1578, 'share': 1579, 'implant': 1580, 'address': 1581, 'ability': 1582, 'fire': 1583, 'santa': 1584, 'desperate': 1585, 'magic': 1586, 'pushy': 1587, 'madly': 1588, 'mile': 1589, 'mutual': 1590, 'fly': 1591, 'sooner': 1592, 'aggression': 1593, 'environment': 1594, 'investigate': 1595, 'procrastination': 1596, 'c': 1597, 'f': 1598, 'freshman': 1599, 'tim': 1600, 'sally': 1601, 'endure': 1602, 'fair': 1603, 'instantly': 1604, 'stepdad': 1605, 'parking': 1606, 'depend': 1607, 'damage': 1608, 'penetration': 1609, 'supervision': 1610, 'alcoholic': 1611, 'goal': 1612, 'delete': 1613, 'definitely': 1614, 'career': 1615, 'ungrateful': 1616, 'exhaust': 1617, 'contract': 1618, 'hunt': 1619, 'creature': 1620, 'ill': 1621, 'halloween': 1622, 'diagnosis': 1623, 'agoraphobia': 1624, 'incapable': 1625, 'waste': 1626, 'resource': 1627, 'click': 1628, 'associate': 1629, 'failure': 1630, 'seat': 1631, 'belt': 1632, 'meaningless': 1633, 'conclusion': 1634, 'possessiveness': 1635, '29': 1636, 'rush': 1637, 'protect': 1638, 'interaction': 1639, 'july': 1640, 'grocery': 1641, 'alive': 1642, 'intimately': 1643, 'awhile': 1644, 'relieve': 1645, 'repress': 1646, 'oddly': 1647, 'jumpy': 1648, 'slight': 1649, 'injure': 1650, 'prison': 1651, 'blur': 1652, 'direct': 1653, 'resentment': 1654, 'war': 1655, 'butterfly': 1656, 'molest': 1657, 'pink': 1658, 'barbie': 1659, 'backpack': 1660, 'fit': 1661, 'abstain': 1662, 'guess': 1663, 'incredibly': 1664, 'escape': 1665, 'crowd': 1666, 'sweaty': 1667, 'subsequently': 1668, 'accountability': 1669, 'halfway': 1670, 'relapse': 1671, 'dress': 1672, 'apathetic': 1673, 'television': 1674, 'unconditionally': 1675, 'voicemail': 1676, '24': 1677, 'impress': 1678, 'hysterically': 1679, 'sweat': 1680, 'scribble': 1681, 'grief': 1682, 'hobby': 1683, 'hr': 1684, 'nourish': 1685, 'nasty': 1686, 'ive': 1687, 'penalize': 1688, 'prior': 1689, 'softball': 1690, 'obligate': 1691, 'runner': 1692, 'wrap': 1693, 'arise': 1694, 'tremendously': 1695, 'alzheimer': 1696, 'interview': 1697, 'tumblr': 1698, 'squat': 1699, 'unusually': 1700, 'wont': 1701, 'altercation': 1702, 'total': 1703, 'imaginary': 1704, 'choose': 1705, 'clue': 1706, 'carry': 1707, 'decrease': 1708, 'butt': 1709, 'peace': 1710, 'mainly': 1711, 'december': 1712, 'object': 1713, 'makeup': 1714, 'tuck': 1715, 'resemble': 1716, 'vagina': 1717, 'strap': 1718, 'anal': 1719, 'react': 1720, 'bash': 1721, 'various': 1722, 'overly': 1723, 'hypochondriac': 1724, 'chemistry': 1725, 'loathe': 1726, 'esd': 1727, 'disable': 1728, 'grandparent': 1729, 'automobile': 1730, 'response': 1731, 'payment': 1732, 'recent': 1733, 'photo': 1734, 'boundary': 1735, 'rock': 1736, 'community': 1737, 'comb': 1738, 'furious': 1739, 'attatchment': 1740, 'classify': 1741, 'characteristic': 1742, 'view': 1743, 'suppress': 1744, 'vibe': 1745, 'overall': 1746, 'discussion': 1747, 'purchase': 1748, 'personalize': 1749, 'reimburse': 1750, 'consistently': 1751, 'cancellation': 1752, 'shit': 1753, 'classmate': 1754, 'ache': 1755, 'peter': 1756, 'jennifer': 1757, 'buy': 1758, 'orient': 1759, 'tense': 1760, 'overbear': 1761, 'magnify': 1762, 'possessive': 1763, 'underlie': 1764, 'blister': 1765, 'recognize': 1766, 'hypocrite': 1767, 'tomorrow': 1768, 'dig': 1769, 'grind': 1770, 'r': 1771, 'practically': 1772, 'florida': 1773, 'ok': 1774, 'affectionate': 1775, 'internet': 1776, 'addition': 1777, 'originally': 1778, 'replace': 1779, 'tie': 1780, 'window': 1781, 'restaurant': 1782, 'adapt': 1783, 'astound': 1784, 'cruelty': 1785, 'faith': 1786, 'humanity': 1787, 'hint': 1788, 'dumb': 1789, 'fund': 1790, 'material': 1791, 'envious': 1792, 'plus': 1793, 'grieve': 1794, 'awake': 1795, 'police': 1796, 'bore': 1797, 'distraction': 1798, 'choke': 1799, 'spanking': 1800, 'toy': 1801, 'likewise': 1802, 'blatantly': 1803, 'angle': 1804, 'image': 1805, 'invalidate': 1806, 'location': 1807, 'isolation': 1808, 'genuinely': 1809, 'utter': 1810, 'discredit': 1811, 'plea': 1812, 'decade': 1813, 'unforgiving': 1814, 'granddaughter': 1815, 'ugliness': 1816, 'exaggerate': 1817, 'proportion': 1818, 'reality': 1819, 'reasonable': 1820, 'insensitive': 1821, 'postpartum': 1822, 'grandchild': 1823, 'sincerely': 1824, 'newly': 1825, 'particularly': 1826, 'partly': 1827, 'romance': 1828, 'center': 1829, 'hectic': 1830, 'orgasm': 1831, 'legitimately': 1832, 'angrily': 1833, 'denial': 1834, 'broach': 1835, 'subject': 1836, 'obsession': 1837, 'prioritize': 1838, 'grab': 1839, 'exchange': 1840, 'brunch': 1841, 'reciprocate': 1842, 'harmful': 1843, 'prophesy': 1844, 'accusatory': 1845, 'unbearably': 1846, 'cell': 1847, 'sanity': 1848, 'ethical': 1849, 'cod': 1850, 'ramification': 1851, 'thirty': 1852, 'till': 1853, 'kinda': 1854, 'brick': 1855, 'expressive': 1856, 'thru': 1857, 'strength': 1858, 'rapist': 1859, 'convict': 1860, 'texas': 1861, 'occasion': 1862, 'adoption': 1863, 'maintenance': 1864, 'evict': 1865, 'majorly': 1866, 'impossible': 1867, 'sunday': 1868, 'restrict': 1869, 'competition': 1870, 'planet': 1871, 'flip': 1872, 'upside': 1873, '2008': 1874, 'deeply': 1875, 'loner': 1876, 'management': 1877, 'apparently': 1878, 'odd': 1879, 'demeanor': 1880, 'shift': 1881, 'inappropriately': 1882, 'urinate': 1883, 'peeing': 1884, 'litter': 1885, 'bottle': 1886, 'directly': 1887, 'literature': 1888, 'incorrect': 1889, 'harsh': 1890, 'science': 1891, 'assignment': 1892, 'network': 1893, 'remote': 1894, 'elsewhere': 1895, 'fee': 1896, 'brag': 1897, 'bath': 1898, 'soap': 1899, 'loyal': 1900, 'gynecologist': 1901, 'attend': 1902, 'sonogram': 1903, 'transition': 1904, 'persist': 1905, '4th': 1906, 'loom': 1907, 'spurt': 1908, 'pure': 1909, 'hormone': 1910, 'destructive': 1911, 'bite': 1912, 'arm': 1913, 'scratch': 1914, 'thigh': 1915, 'earring': 1916, 'sickness': 1917, 'hospitalize': 1918, 'loneliness': 1919, 'search': 1920, 'void': 1921, 'conscious': 1922, 'train': 1923, 'gag': 1924, 'june': 1925, '2016': 1926, 'detach': 1927, 'dependency': 1928, 'bi': 1929, 'polar': 1930, 'detachment': 1931, 'sneaky': 1932, 'harass': 1933, 'stalk': 1934, 'presence': 1935, 'battle': 1936, 'intervention': 1937, 'casually': 1938, 'compromise': 1939, 'casual': 1940, 'power': 1941, 'wit': 1942, 'relatively': 1943, 'fraternity': 1944, 'regretful': 1945, 'flaunt': 1946, 'bury': 1947, 'jan': 1948, 'access': 1949, 'wedge': 1950, 'advise': 1951, 'consequence': 1952, 'dislocate': 1953, 'hip': 1954, 'electroshock': 1955, 'keen': 1956, 'mentor': 1957, 'spiritually': 1958, 'manipulation': 1959, 'leadership': 1960, 'suspect': 1961, 'bomb': 1962, 'skyped': 1963, 'skype': 1964, 'joy': 1965, 'assure': 1966, 'cyst': 1967, 'remove': 1968, 'testicle': 1969, 'season': 1970, 'unfair': 1971, 'vomit': 1972, 'joyous': 1973, 'massive': 1974, 'forum': 1975, 'thankful': 1976, 'academic': 1977, 'culture': 1978, 'relocate': 1979, 'approve': 1980, 'loose': 1981, 'tunnel': 1982, 'chaos': 1983, 'infest': 1984, 'rip': 1985, 'everywhere': 1986, 'solution': 1987, 'farm': 1988, 'piece': 1989, 'curl': 1990, 'mall': 1991, 'cling': 1992, 'conflict': 1993, 'cashier': 1994, 'pump': 1995, 'bet': 1996, 'poodle': 1997, 'disagree': 1998, 'tease': 1999, 'chase': 2000, 'undiagnosed': 2001, 'mass': 2002, 'infection': 2003, 'locker': 2004, 'smile': 2005, 'unconfident': 2006, 'doom': 2007, 'employee': 2008, 'outlet': 2009, 'asthma': 2010, 'weapon': 2011, 'entirely': 2012, 'tragic': 2013, 'twist': 2014, 'legitimate': 2015, 'borrowing': 2016, 'trail': 2017, 'sob': 2018, 'gamble': 2019, 'compulsively': 2020, 'epilepsy': 2021, 'simple': 2022, 'pathological': 2023, 'technology': 2024, 'fantasy': 2025, 'unwilling': 2026, 'participant': 2027, 'hypnosis': 2028, 'debilitate': 2029, 'wellbutrin': 2030, 'unwillingness': 2031, 'despair': 2032, 'complacent': 2033, 'fulfillment': 2034, 'regain': 2035, 'zombie': 2036, 'haze': 2037, 'concentrate': 2038, 'infuriate': 2039, 'gasoline': 2040, 'respectful': 2041, 'rebound': 2042, 'console': 2043, 'companionship': 2044, 'hysterectomy': 2045, 'knee': 2046, 'eve': 2047, 'alright': 2048, 'crap': 2049, 'golden': 2050, 'deviant': 2051, 'trip': 2052, 'unintentional': 2053, 'unbearable': 2054, 'august': 2055, 'signal': 2056, 'cross': 2057, 'recover': 2058, 'ridiculous': 2059, 'violently': 2060, 'interact': 2061, 'spark': 2062, 'acknowledgment': 2063, 'repetition': 2064, 'cheap': 2065, 'walmart': 2066, 'honeymoon': 2067, 'downward': 2068, 'famous': 2069, 'across': 2070, 'afternoon': 2071, 'table': 2072, 'assessment': 2073, 'abuser': 2074, 'invisible': 2075, 'romantically': 2076, 'platonic': 2077, 'free': 2078, 'kitty': 2079, 'gross': 2080, 'debit': 2081, 'punish': 2082, 'ground': 2083, 'betrayal': 2084, 'forgiveness': 2085, '3000': 2086, 'grateful': 2087, 'validation': 2088, 'haunt': 2089, 'newborn': 2090, 'pregnancy': 2091, 'abortion': 2092, 'temper': 2093, 'wheel': 2094, 'stupidly': 2095, 'climax': 2096, 'h': 2097, 'lovecraft': 2098, 'edgar': 2099, 'allen': 2100, 'poe': 2101, 'saturday': 2102, 'powerful': 2103, 'flesh': 2104, 'elephant': 2105, 'insinuate': 2106, 'babysit': 2107, 'refer': 2108, '16': 2109, 'aggravate': 2110, 'forbid': 2111, 'brainwash': 2112, 'deceive': 2113, 'slut': 2114, 'devil': 2115, 'worshipper': 2116, 'retail': 2117, 'passenger': 2118, 'ride': 2119, 'backseat': 2120, 'lean': 2121, 'degrade': 2122, 'warmly': 2123, 'agonize': 2124, 'distrust': 2125, 'temporary': 2126, 'strand': 2127, 'passive': 2128, 'aggressive': 2129, 'snoring': 2130, 'deliver': 2131, 'clot': 2132, 'bleed': 2133, 'transfusion': 2134, 'undergo': 2135, 'ordeal': 2136, 'unaccomplished': 2137, 'comfortably': 2138, 'clarity': 2139, 'dance': 2140, 'knob': 2141, 'paranoia': 2142, 'noisy': 2143, 'reproduce': 2144, 'hispanic': 2145, 'language': 2146, 'overstep': 2147, 'balance': 2148, 'reflect': 2149, 'intuition': 2150, 'guardianship': 2151, 'nonexistent': 2152, 'nowadays': 2153, 'interfere': 2154, 'automatically': 2155, '19': 2156, 'vastly': 2157, 'disproportionate': 2158, 'lucid': 2159, 'drown': 2160, 'crumble': 2161, 'adoptive': 2162, 'plug': 2163, 'ear': 2164, 'moody': 2165, 'clingy': 2166, 'info': 2167, 'pathetic': 2168, 'mature': 2169, 'bossy': 2170, 'randomly': 2171, 'compatible': 2172, 'criminal': 2173, 'justice': 2174, 'bdsm': 2175, 'unprotected': 2176, 'paralysis': 2177, 'mock': 2178, 'practice': 2179, 'piano': 2180, 'quinceañera': 2181, 'film': 2182, 'meditation': 2183, 'cute': 2184, 'reflection': 2185, 'prescribe': 2186, 'scold': 2187, 'button': 2188, 'sweating': 2189, 'conscience': 2190, 'annoy': 2191, 'im': 2192, 'charm': 2193, 'hissy': 2194, 'assist': 2195, 'joyless': 2196, 'motion': 2197, 'assault': 2198, 'blunt': 2199, 'crisis': 2200, 'twin': 2201, 'meltdown': 2202, 'camera': 2203, 'wherever': 2204, 'radio': 2205, 'whisper': 2206, 'hangout': 2207, 'wednesday': 2208, 'maze': 2209, 'socially': 2210, 'straightforward': 2211, 'unsettled': 2212, 'soul': 2213, 'codependent': 2214, 'manipulative': 2215, 'diabetes': 2216, 'horrendous': 2217, 'contribute': 2218, 'unrealistic': 2219, 'item': 2220, 'intellectual': 2221, 'disease': 2222, 'paralyze': 2223, 'performance': 2224, 'extensive': 2225, 'humiliate': 2226, 'temp': 2227, 'agency': 2228, 'provider': 2229, 'pursue': 2230, 'betray': 2231, 'isolate': 2232, 'curb': 2233, 'horrifically': 2234, 'unspeakable': 2235, 'stressful': 2236, 'solace': 2237, 'motherly': 2238, 'mentality': 2239, 'fruit': 2240, 'vegetable': 2241, 'junk': 2242, 'pointless': 2243, 'clown': 2244, 'uncontrollably': 2245, 'inflammatory': 2246, 'demyelinate': 2247, 'polyneuropathy': 2248, 'caregiver': 2249, 'uncontrollable': 2250, 'repetitive': 2251, 'somewhat': 2252, 'occasionally': 2253, 'benign': 2254, 'tumor': 2255, 'destroy': 2256, 'armed': 2257, 'leukemia': 2258, 'playfully': 2259, 'outfit': 2260, 'scan': 2261, 'diabetic': 2262, 'attachment': 2263, 'innocent': 2264, 'freedom': 2265, 'lover': 2266, 'actively': 2267, 'stretch': 2268, 'mark': 2269, 'heartbroken': 2270, 'quality': 2271, 'overweight': 2272, 'enormous': 2273, 'load': 2274, 'electrician': 2275, 'operate': 2276, 'machinery': 2277, 'november': 2278, '11pm': 2279, 'april': 2280, 'yes': 2281, 'danger': 2282, 'formally': 2283, 'lovingly': 2284, 'education': 2285, 'jail': 2286, 'overdramatic': 2287, 'option': 2288, 'proud': 2289, 'hateful': 2290, 'quote': 2291, 'appreciate': 2292, 'insure': 2293, 'illegible': 2294, 'envelope': 2295, 'contain': 2296, 'ersatz': 2297, 'atone': 2298, 'cowardly': 2299, 'kindness': 2300, 'karma': 2301, 'jade': 2302, 'convey': 2303, 'specifically': 2304, 'deteriorate': 2305, 'despise': 2306, 'absurdly': 2307, 'vulgar': 2308, 'upcoming': 2309, 'schooling': 2310, 'zero': 2311, 'prank': 2312, '60': 2313, 'asperger': 2314, 'flirty': 2315, 'risk': 2316, 'restrain': 2317, 'website': 2318, 'hack': 2319, 'misophonia': 2320, '99': 2321, 'avoidant': 2322, 'dsm': 2323, 'iv': 2324, 'peaceful': 2325, 'eventhough': 2326, 'discontinue': 2327, 'apparent': 2328, 'toxic': 2329, 'fishing': 2330, 'pond': 2331, 'frog': 2332, 'knife': 2333, 'million': 2334, 'profusely': 2335, 'paycheck': 2336, 'trash': 2337, 'validate': 2338, 'tendency': 2339, 'flight': 2340, 'stair': 2341, 'leg': 2342, 'curiosity': 2343, 'kidnap': 2344, 'fourteen': 2345, 'gunpoint': 2346, 'nineteen': 2347, 'careful': 2348, 'project': 2349, 'onto': 2350, 'hyperactivity': 2351, 'income': 2352, 'crappy': 2353, 'grand': 2354, 'strict': 2355, '48': 2356, 'tuesday': 2357, 'double': 2358, 'sink': 2359, 'useless': 2360, 'area': 2361, 'primary': 2362, 'physician': 2363, 'zoosk': 2364, 'luck': 2365, '5000': 2366, 'deductible': 2367, 'screw': 2368, 'promote': 2369, 'reoccurring': 2370, 'ninth': 2371, 'principal': 2372, 'spell': 2373, 'boot': 2374, 'traumatic': 2375, 'outburst': 2376, 'intil': 2377, 'yr': 2378, 'ireland': 2379, 'profanity': 2380, 'pertain': 2381, 'dramatic': 2382, 'tiptoe': 2383, 'neatly': 2384, 'abou': 2385, 'basic': 2386, 'regularly': 2387, 'reputation': 2388, 'longtime': 2389, '1000': 2390, 'invitation': 2391, 'mend': 2392, 'sneak': 2393, 'difficulty': 2394, 'heroin': 2395, 'manipulate': 2396, 'stepson': 2397, 'stepmom': 2398, 'illegally': 2399, 'opiate': 2400, 'unannounced': 2401, 'opinionated': 2402, 'unsolicited': 2403, 'rear': 2404, 'privacy': 2405, 'barge': 2406, 'key': 2407, 'permission': 2408, 'combative': 2409, 'opening': 2410, 'instigate': 2411, 'terminally': 2412, 'psychologically': 2413, 'foster': 2414, 'untill': 2415, 'pm': 2416}\n"
     ]
    }
   ],
   "source": [
    "tokenizer = preprocessing.text.Tokenizer()\n",
    "tokenizer.fit_on_texts( questionText_nostops ) \n",
    "tokenized_questions = tokenizer.texts_to_sequences( questionText_nostops ) \n",
    "word_index = tokenizer.word_index\n",
    "print('Sample tokenized: {}'.format(tokenized_questions[0]))\n",
    "print('==================================================================================================================\\n')\n",
    "\n",
    "length_list = list()\n",
    "for token_seq in tokenized_questions:\n",
    "    length_list.append( len( token_seq ))\n",
    "max_input_length = np.array( length_list ).max()\n",
    "print( 'Questions max length is {} words'.format( max_input_length ))\n",
    "print('==================================================================================================================\\n')\n",
    "\n",
    "padded_questions = preprocessing.sequence.pad_sequences( tokenized_questions , maxlen=max_input_length , padding='post' )\n",
    "input_data = np.array( padded_questions )\n",
    "print( 'Input data shape -> {}\\n'.format( input_data.shape ))\n",
    "print('Input data sample->\\n {}'.format(input_data[0]))\n",
    "print('==================================================================================================================\\n')\n",
    "\n",
    "question_word_dict = tokenizer.word_index\n",
    "num_question_tokens = len( question_word_dict )+1\n",
    "print( 'Number of Question tokens = {}\\n'.format( num_question_tokens))\n",
    "print('Dictionary: {}'.format(question_word_dict))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2qAqsYsgceCG"
   },
   "source": [
    "## Word Embedding - Glove Embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "bg1mJuJ8fsfr",
    "outputId": "891e1122-eac8-4350-d2aa-a5c084d2084d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<gensim.models.keyedvectors.Word2VecKeyedVectors at 0x7fb58eac05d0>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from gensim.test.utils import datapath, get_tmpfile\n",
    "from gensim.models import KeyedVectors\n",
    "from gensim.scripts.glove2word2vec import glove2word2vec\n",
    "\n",
    "glove_file = datapath('test_glove.txt')\n",
    "tmp_file = get_tmpfile(\"test_word2vec.txt\")\n",
    "\n",
    "_ = glove2word2vec(glove_file, tmp_file)\n",
    "\n",
    "model = KeyedVectors.load_word2vec_format(tmp_file)\n",
    "model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "9ziHAvOXkI7a",
    "outputId": "1a71d624-231f-472f-bea5-246b9b51fee1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('stupid', 0.8132887482643127),\n",
       " ('confused', 0.7932009100914001),\n",
       " ('birds', 0.7878215909004211),\n",
       " ('weird', 0.7833624482154846),\n",
       " ('annoying', 0.7818509936332703),\n",
       " ('sometimes', 0.780109167098999),\n",
       " ('mad', 0.7680050730705261),\n",
       " ('scared', 0.7669514417648315),\n",
       " ('bit', 0.7651330232620239),\n",
       " ('feel', 0.7628020644187927)]"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import gensim.downloader as api\n",
    "\n",
    "info = api.info()  # show info about available models/datasets\n",
    "model = api.load(\"glove-twitter-50\")  # download the model and return as object ready for use\n",
    "model.most_similar(\"angry\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.740932Z",
     "iopub.status.idle": "2021-12-11T09:31:57.741326Z",
     "shell.execute_reply": "2021-12-11T09:31:57.741131Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.741109Z"
    },
    "id": "ei8ZP3LnbHqo",
    "outputId": "8f592cd0-16c4-44a0-efc0-7ec16febd94d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('quiet', 0.9196479320526123),\n",
       " ('crazy', 0.8963231444358826),\n",
       " ('loud', 0.8955198526382446),\n",
       " ('feel', 0.8897919058799744),\n",
       " ('loving', 0.8833610415458679),\n",
       " ('sober', 0.882926344871521),\n",
       " ('sleeping', 0.8801181316375732),\n",
       " ('depressed', 0.879048764705658),\n",
       " ('scary', 0.8757081627845764),\n",
       " ('sad', 0.8738341927528381)]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "glove_vectors.most_similar('lonely')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "BOzszAE8d3MZ",
    "outputId": "e7c48571-70c6-4c81-a3e9-b827b8618c13"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('hate', 0.9336040019989014),\n",
       " ('feel', 0.9293274283409119),\n",
       " ('stupid', 0.9281623959541321),\n",
       " ('reason', 0.9275999665260315),\n",
       " ('confused', 0.9258189797401428)]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result = model.most_similar(positive=['angry', 'sad'], topn=5)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.762300Z",
     "iopub.status.idle": "2021-12-11T09:31:57.762748Z",
     "shell.execute_reply": "2021-12-11T09:31:57.762543Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.762519Z"
    },
    "id": "8f5l1-aobHqp",
    "outputId": "8c82bd36-7a9b-4ea8-f333-83d94d6def3d"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:1: DeprecationWarning: Call to deprecated `wv` (Attribute will be removed in 4.0.0, use self instead).\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1193514\n",
      "The no of key-value pairs :  1193514\n"
     ]
    }
   ],
   "source": [
    "vocab=list(model.wv.vocab)\n",
    "index_of_words = len(vocab)\n",
    "print(index_of_words)\n",
    "word_vec_dict={}\n",
    "for word in vocab:\n",
    "    word_vec_dict[word]=(glove_vectors.get_vector(word))\n",
    "print(\"The no of key-value pairs : \",len(word_vec_dict)) # should come equal to vocab size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.764083Z",
     "iopub.status.idle": "2021-12-11T09:31:57.764761Z",
     "shell.execute_reply": "2021-12-11T09:31:57.764541Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.764511Z"
    },
    "id": "vVQSuJLsbHqp"
   },
   "outputs": [],
   "source": [
    "embed_matrix=np.zeros(shape=(len(vocab),25))\n",
    "i=0\n",
    "for word,i in word_index.items():\n",
    "    embed_vector=word_vec_dict.get(word)\n",
    "    # print(embed_vector)\n",
    "    if embed_vector is not None:  # word is in the vocabulary learned by the w2v model\n",
    "        embed_matrix[i]=embed_vector\n",
    "  # if word is not found then embed_vector corressponding to that vector will stay zero."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.767942Z",
     "iopub.status.idle": "2021-12-11T09:31:57.768337Z",
     "shell.execute_reply": "2021-12-11T09:31:57.768142Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.768120Z"
    },
    "id": "IDuhRengbHqq",
    "outputId": "16f29caf-75dc-4416-aeae-1e15a8ad7567"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'2.7.0'"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras import preprocessing\n",
    "\n",
    "tf.random.set_seed(4)\n",
    "tf.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "s6m-HdirbHqq"
   },
   "source": [
    "Finally, with the input data ready with the numerical token indices, it is time to split the dataset into training and testing sets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.769651Z",
     "iopub.status.idle": "2021-12-11T09:31:57.770046Z",
     "shell.execute_reply": "2021-12-11T09:31:57.769851Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.769829Z"
    },
    "id": "7XtZwMwzbHqr"
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.771527Z",
     "iopub.status.idle": "2021-12-11T09:31:57.771920Z",
     "shell.execute_reply": "2021-12-11T09:31:57.771727Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.771705Z"
    },
    "id": "UGA656DgbHqr"
   },
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(input_data, targetdf, test_size = 0.2, random_state=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.773362Z",
     "iopub.status.idle": "2021-12-11T09:31:57.773789Z",
     "shell.execute_reply": "2021-12-11T09:31:57.773595Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.773571Z"
    },
    "id": "ExZeG56RbHqs",
    "outputId": "8a6617d0-6236-42cf-8cb0-91d167ea866e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1100, 220) (276, 220)\n",
      "(1100, 32) (276, 32)\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape, x_test.shape)\n",
    "print(y_train.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kJSEsIdmbHqs"
   },
   "source": [
    "Finally, in the below cell, the bi-directional LSTM model is defined."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KDCLwGM_JB8u"
   },
   "source": [
    "## Model Architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.775263Z",
     "iopub.status.idle": "2021-12-11T09:31:57.775702Z",
     "shell.execute_reply": "2021-12-11T09:31:57.775493Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.775470Z"
    },
    "id": "MB9pbT-ubHqs"
   },
   "outputs": [],
   "source": [
    "# vacab size is number of unique words + reserved 0 index for padding\n",
    "vocab_size = index_of_words\n",
    "# Embedding layer before the actaul LSTM \n",
    "embedd_layer = tf.keras.layers.Embedding(vocab_size,25,input_length = max_input_length,weights = [embed_matrix],trainable=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.778824Z",
     "iopub.status.idle": "2021-12-11T09:31:57.779225Z",
     "shell.execute_reply": "2021-12-11T09:31:57.779027Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.779005Z"
    },
    "id": "92Up13o-bHqt",
    "outputId": "6ee6624f-62d4-4e1d-f3d5-f9bce1d56509"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_17\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_7 (Embedding)     (None, 220, 25)           29837850  \n",
      "                                                                 \n",
      " bidir (Bidirectional)       (None, 200)               100800    \n",
      "                                                                 \n",
      " dense_20 (Dense)            (None, 500)               100500    \n",
      "                                                                 \n",
      " dense_21 (Dense)            (None, 32)                16032     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 30,055,182\n",
      "Trainable params: 217,332\n",
      "Non-trainable params: 29,837,850\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Embedding Layer, LSTM or biLSTM, Dense, softmax\n",
    "model = Sequential()\n",
    "model.add(embedd_layer)\n",
    "model.add(Bidirectional(LSTM(100, dropout=0.1,return_state=False), name='bidir'))\n",
    "model.add(Dense(500, activation=tf.keras.activations.relu))\n",
    "model.add(Dense(32, activation='softmax'))\n",
    "model.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy','Precision','Recall'])\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 466
    },
    "id": "b-X6vn6nJGE-",
    "outputId": "ffdeda01-d201-4f66-b253-ee729e42c2eb"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.utils.plot_model(\n",
    "    model, to_file='model.png', show_shapes=False, show_dtype=False,\n",
    "    show_layer_names=True, rankdir='TB', expand_nested=False, dpi=96,\n",
    "    layer_range=None, show_layer_activations=False\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.780353Z",
     "iopub.status.idle": "2021-12-11T09:31:57.781254Z",
     "shell.execute_reply": "2021-12-11T09:31:57.781043Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.781018Z"
    },
    "id": "flQyylNnbHqu",
    "outputId": "1b52eb37-b30e-40af-fbb2-d009330574d6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "9/9 [==============================] - 13s 986ms/step - loss: 3.1902 - accuracy: 0.1600 - precision: 0.0000e+00 - recall: 0.0000e+00 - val_loss: 2.9932 - val_accuracy: 0.1630 - val_precision: 0.0000e+00 - val_recall: 0.0000e+00\n",
      "Epoch 2/30\n",
      "9/9 [==============================] - 7s 819ms/step - loss: 2.8899 - accuracy: 0.2027 - precision: 0.0000e+00 - recall: 0.0000e+00 - val_loss: 2.9177 - val_accuracy: 0.1739 - val_precision: 0.0000e+00 - val_recall: 0.0000e+00\n",
      "Epoch 3/30\n",
      "9/9 [==============================] - 7s 822ms/step - loss: 2.8122 - accuracy: 0.1973 - precision: 0.0000e+00 - recall: 0.0000e+00 - val_loss: 2.8481 - val_accuracy: 0.2210 - val_precision: 0.0000e+00 - val_recall: 0.0000e+00\n",
      "Epoch 4/30\n",
      "9/9 [==============================] - 7s 824ms/step - loss: 2.7379 - accuracy: 0.2273 - precision: 1.0000 - recall: 9.0909e-04 - val_loss: 2.7783 - val_accuracy: 0.1920 - val_precision: 0.0000e+00 - val_recall: 0.0000e+00\n",
      "Epoch 5/30\n",
      "9/9 [==============================] - 7s 822ms/step - loss: 2.6270 - accuracy: 0.2636 - precision: 1.0000 - recall: 9.0909e-04 - val_loss: 2.6729 - val_accuracy: 0.2428 - val_precision: 1.0000 - val_recall: 0.0145\n",
      "Epoch 6/30\n",
      "9/9 [==============================] - 7s 827ms/step - loss: 2.5176 - accuracy: 0.2845 - precision: 1.0000 - recall: 0.0045 - val_loss: 2.5742 - val_accuracy: 0.2790 - val_precision: 0.0000e+00 - val_recall: 0.0000e+00\n",
      "Epoch 7/30\n",
      "9/9 [==============================] - 7s 822ms/step - loss: 2.3980 - accuracy: 0.3136 - precision: 0.8182 - recall: 0.0164 - val_loss: 2.4896 - val_accuracy: 0.2790 - val_precision: 0.8889 - val_recall: 0.0580\n",
      "Epoch 8/30\n",
      "9/9 [==============================] - 7s 818ms/step - loss: 2.2849 - accuracy: 0.3264 - precision: 0.8532 - recall: 0.0845 - val_loss: 2.3802 - val_accuracy: 0.3370 - val_precision: 0.8649 - val_recall: 0.1159\n",
      "Epoch 9/30\n",
      "9/9 [==============================] - 7s 820ms/step - loss: 2.1754 - accuracy: 0.3591 - precision: 0.8121 - recall: 0.1100 - val_loss: 2.3778 - val_accuracy: 0.3514 - val_precision: 0.8333 - val_recall: 0.1087\n",
      "Epoch 10/30\n",
      "9/9 [==============================] - 7s 825ms/step - loss: 2.0542 - accuracy: 0.3955 - precision: 0.8750 - recall: 0.1336 - val_loss: 2.2063 - val_accuracy: 0.3913 - val_precision: 0.7869 - val_recall: 0.1739\n",
      "Epoch 11/30\n",
      "9/9 [==============================] - 7s 824ms/step - loss: 2.0272 - accuracy: 0.4082 - precision: 0.6713 - recall: 0.1764 - val_loss: 2.3115 - val_accuracy: 0.4022 - val_precision: 0.7119 - val_recall: 0.1522\n",
      "Epoch 12/30\n",
      "9/9 [==============================] - 7s 827ms/step - loss: 1.9547 - accuracy: 0.4300 - precision: 0.8139 - recall: 0.1709 - val_loss: 2.2310 - val_accuracy: 0.3949 - val_precision: 0.7222 - val_recall: 0.1884\n",
      "Epoch 13/30\n",
      "9/9 [==============================] - 7s 827ms/step - loss: 1.8190 - accuracy: 0.4818 - precision: 0.8306 - recall: 0.1873 - val_loss: 2.0893 - val_accuracy: 0.4239 - val_precision: 0.7975 - val_recall: 0.2283\n",
      "Epoch 14/30\n",
      "9/9 [==============================] - 7s 826ms/step - loss: 1.6779 - accuracy: 0.5300 - precision: 0.8305 - recall: 0.2627 - val_loss: 2.0036 - val_accuracy: 0.5109 - val_precision: 0.8571 - val_recall: 0.2391\n",
      "Epoch 15/30\n",
      "9/9 [==============================] - 7s 819ms/step - loss: 1.6101 - accuracy: 0.5318 - precision: 0.8610 - recall: 0.2927 - val_loss: 2.0276 - val_accuracy: 0.5000 - val_precision: 0.7708 - val_recall: 0.2681\n",
      "Epoch 16/30\n",
      "9/9 [==============================] - 7s 821ms/step - loss: 1.4856 - accuracy: 0.5945 - precision: 0.8682 - recall: 0.3355 - val_loss: 1.9742 - val_accuracy: 0.4819 - val_precision: 0.6960 - val_recall: 0.3152\n",
      "Epoch 17/30\n",
      "9/9 [==============================] - 7s 832ms/step - loss: 1.4188 - accuracy: 0.5964 - precision: 0.8581 - recall: 0.3573 - val_loss: 1.8837 - val_accuracy: 0.5290 - val_precision: 0.7323 - val_recall: 0.3370\n",
      "Epoch 18/30\n",
      "9/9 [==============================] - 7s 824ms/step - loss: 1.3220 - accuracy: 0.6364 - precision: 0.8607 - recall: 0.4100 - val_loss: 1.8106 - val_accuracy: 0.5580 - val_precision: 0.7710 - val_recall: 0.3659\n",
      "Epoch 19/30\n",
      "9/9 [==============================] - 7s 831ms/step - loss: 1.2559 - accuracy: 0.6473 - precision: 0.8689 - recall: 0.4400 - val_loss: 1.8183 - val_accuracy: 0.5543 - val_precision: 0.7664 - val_recall: 0.3804\n",
      "Epoch 20/30\n",
      "9/9 [==============================] - 7s 833ms/step - loss: 1.1686 - accuracy: 0.6609 - precision: 0.8852 - recall: 0.4555 - val_loss: 1.7724 - val_accuracy: 0.5580 - val_precision: 0.7676 - val_recall: 0.3949\n",
      "Epoch 21/30\n",
      "9/9 [==============================] - 7s 832ms/step - loss: 1.0656 - accuracy: 0.7082 - precision: 0.8906 - recall: 0.5255 - val_loss: 1.7567 - val_accuracy: 0.5471 - val_precision: 0.7727 - val_recall: 0.4312\n",
      "Epoch 22/30\n",
      "9/9 [==============================] - 7s 824ms/step - loss: 1.0382 - accuracy: 0.7000 - precision: 0.9035 - recall: 0.5109 - val_loss: 1.7124 - val_accuracy: 0.5833 - val_precision: 0.7607 - val_recall: 0.4493\n",
      "Epoch 23/30\n",
      "9/9 [==============================] - 7s 820ms/step - loss: 0.9916 - accuracy: 0.7327 - precision: 0.8906 - recall: 0.5627 - val_loss: 1.6700 - val_accuracy: 0.5797 - val_precision: 0.8062 - val_recall: 0.4674\n",
      "Epoch 24/30\n",
      "9/9 [==============================] - 7s 831ms/step - loss: 0.9098 - accuracy: 0.7545 - precision: 0.8917 - recall: 0.5991 - val_loss: 1.7134 - val_accuracy: 0.5833 - val_precision: 0.7416 - val_recall: 0.4783\n",
      "Epoch 25/30\n",
      "9/9 [==============================] - 7s 833ms/step - loss: 0.8795 - accuracy: 0.7545 - precision: 0.8821 - recall: 0.6055 - val_loss: 1.6822 - val_accuracy: 0.6051 - val_precision: 0.7543 - val_recall: 0.4783\n",
      "Epoch 26/30\n",
      "9/9 [==============================] - 7s 826ms/step - loss: 0.8179 - accuracy: 0.7764 - precision: 0.9331 - recall: 0.6336 - val_loss: 1.6131 - val_accuracy: 0.6630 - val_precision: 0.7892 - val_recall: 0.5290\n",
      "Epoch 27/30\n",
      "9/9 [==============================] - 7s 819ms/step - loss: 0.7600 - accuracy: 0.7873 - precision: 0.9179 - recall: 0.6709 - val_loss: 1.6108 - val_accuracy: 0.6159 - val_precision: 0.7436 - val_recall: 0.5254\n",
      "Epoch 28/30\n",
      "9/9 [==============================] - 7s 819ms/step - loss: 0.7254 - accuracy: 0.7900 - precision: 0.9223 - recall: 0.6691 - val_loss: 1.5639 - val_accuracy: 0.6232 - val_precision: 0.7979 - val_recall: 0.5580\n",
      "Epoch 29/30\n",
      "9/9 [==============================] - 7s 828ms/step - loss: 0.6878 - accuracy: 0.8182 - precision: 0.9287 - recall: 0.6982 - val_loss: 1.6226 - val_accuracy: 0.6341 - val_precision: 0.7938 - val_recall: 0.5580\n",
      "Epoch 30/30\n",
      "9/9 [==============================] - 7s 825ms/step - loss: 0.6373 - accuracy: 0.8400 - precision: 0.9315 - recall: 0.7291 - val_loss: 1.5132 - val_accuracy: 0.6558 - val_precision: 0.8010 - val_recall: 0.5978\n"
     ]
    }
   ],
   "source": [
    "#history = LSTM_model.fit(x_train, y_train, validation_data=(x_test, y_test), batch_size=128, epochs=50, verbose=1)\n",
    "hist = model.fit(x_train, y_train, validation_data=(x_test, y_test), batch_size=128, epochs=30, verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ZFi8vcqiv2aE"
   },
   "source": [
    "## Model Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "_kg_hide-input": true,
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.784134Z",
     "iopub.status.idle": "2021-12-11T09:31:57.784562Z",
     "shell.execute_reply": "2021-12-11T09:31:57.784336Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.784315Z"
    },
    "id": "dR1tcek6bHqv"
   },
   "outputs": [],
   "source": [
    "history_df = pd.DataFrame(hist.history)\n",
    "history_df['f1_score'] = (2 * history_df['precision']*history_df['recall'])/(history_df['precision']+history_df['recall'])\n",
    "history_df['val_f1_score'] = (2 * history_df['val_precision']*history_df['val_recall'])/(history_df['val_precision']+history_df['val_recall'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.785846Z",
     "iopub.status.idle": "2021-12-11T09:31:57.786241Z",
     "shell.execute_reply": "2021-12-11T09:31:57.786047Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.786025Z"
    },
    "id": "bwtSmUktbHqw",
    "outputId": "756d0f4f-c01d-4f96-8116-ae6d9add6d6c"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(hist.history['loss'], label='training data')\n",
    "plt.plot(hist.history['val_loss'], label='validation data')\n",
    "plt.title('Categorical Crossentropy Loss Plot')\n",
    "plt.ylabel('loss value')\n",
    "plt.xlabel('No. of epoch')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 729
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.787578Z",
     "iopub.status.idle": "2021-12-11T09:31:57.787976Z",
     "shell.execute_reply": "2021-12-11T09:31:57.787780Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.787758Z"
    },
    "id": "Ts1-5afGbHqw",
    "outputId": "c5631d9e-536a-42d3-bfa5-4b2badf29318"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABC8AAALICAYAAABfINo9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd3hU1drG4d9KJ4QAIdRAaKF3CKAUKYIiCNhABBULICr2fvRYjr1/KiAqWJGmIiCCiEpvEqRX6R1CCSW9rO+PPWhEatqeJM99XXNNZs+evZ/J8TA776z1LmOtRURERERERETEW/m4HUBERERERERE5FxUvBARERERERERr6bihYiIiIiIiIh4NRUvRERERERERMSrqXghIiIiIiIiIl5NxQsRERERERER8WoqXohIgWWMqWKMscYYP7eziIiIiHcwxjxvjBntdg4RuTgqXogUcMaY2caYo8aYQLez5AZjTHtjTIYx5qQx5oQxZqMx5vYsHEcXMiIiIjnMGLPdGNPJhfN+boxJ8VwfHDHGzDTG1M7CcVzJLyL/puKFSAFmjKkCtAUs0COPz52Xox32WmtDgFDgCeATY0zdPDy/iIiIeJ83PNcHFYGDwOfuxhGR7FDxQqRguxVYjPNh3T/zE8aYSsaYicaYWGPMYWPM0EzPDTTGrPeMZFhnjGnq2W6NMVGZ9vvcGPOS5+f2xpjdxpgnjDH7gc+MMSWNMVM95zjq+blipteHGWM+M8bs9Tw/ybN9jTGme6b9/I0xh4wxTc71Zq1jEnAU+FfxwhhTwRgzxfMNzGZjzEDP9i7Af4AbPd/QrLzA36+IiIhkgTEm0Bjzf55rgL2enwM9z4V7rhniPJ/Z84wxPp7nnjDG7Mk02vLy853LWpsAjAHqnyVLD2PMWs/5Zhtj6ni2fwVEAj94rg8ez6n3LyIXT8ULkYLtVuBrz+1KY0xZAGOMLzAV2AFUASKAcZ7negHPe14bijNi4/AFnq8cEAZUBgbh/BvzmedxJJAIDM20/1dAMFAPKAO869n+JXBzpv26AvustcvPdXJjjI8x5lqgBLD6DLuMA3YDFYAbgFeMMR2ttT8BrwDjrbUh1tpGF/h+RUREJGueBi4BGgONgBbAM57nHsH5vC4NlMX5gsEaY2oBQ4Dm1tpiwJXA9vOdyBgTAvQD/nUdYYypCYwFHvScbxpOsSLAWnsLsBPo7rk+eCPL71ZEsk3FC5ECyhjTBqdoMMFauwzYAvT1PN0C5w/4x6y18dbaJGvtfM9zA3CGWS71jGTYbK3dcYGnzQCes9YmW2sTrbWHrbXfWWsTrLUngJeBdp585YGrgMHW2qPW2lRr7RzPcUYDXY0xoZ7Ht+AUOs6mgjEmDjgEPAfcYq3deNrvoxLQGnjC835XACNxijQiIiKSt/oB/7PWHrTWxgIv4HzeA6QC5YHKnuuDedZaC6QDgUBdY4y/tXa7tXbLOc7xqOf6YDMQAtx2hn1uBH601s601qYCbwFFgFY58B5FJAepeCFScPUHfrbWHvI8HsPfU0cqATustWlneF0lnEJHVsRaa5NOPTDGBBtjPjLG7DDGHAfmAiU8Iz8qAUestUdPP4i1di+wALjeGFMCp8jx9TnOu9daW8JaG2atbWytHXeGfSp4znci07YdOKNOREREJG9VwPkcPmWHZxvAmzgFh5+NMVuNMU8CWGs344yQeB44aIwZZ4ypwNm95bk+KGet7XGWQsc/clhrM4Bd6PpAxOuoeCFSABljigC9gXbGmP2eHhQPAY2MMY1wPpQjz9JUcxdQ/SyHTsCZ5nFKudOet6c9fgSoBbS01oYCl52K6DlPmKc4cSZf4Ewd6QUsstbuOct+F2qv53zFMm2LBE4d9/TsIiIiknv24owQPSXSsw1r7Qlr7SPW2mo401cfPtXbwlo7xlp7anSpBV7PyRzGGIPzBYuuD0S8jIoXIgXTNThDK+vizCVtDNQB5uFMk/gd2Ae8ZowpaowJMsa09rx2JM4wy2bGEWWMOfWhvgLoa4zx9TS5bHeeHMVw+lzEGWPCcKZ0AGCt3QdMB4Z7Gnv6G2Muy/TaSUBT4AGcHhjZYq3dBSwEXvW834bAnThTVAAOAFVONQQTERGRHOPv+ew9dfPD6TPxjDGmtDEmHHgWz2eyMeZqz/WHAY7hXNNkGGNqGWM6ehp7JuFcY2RkM9sEoJsx5nJjjD/OFy/JONcM4FwfVMvmOUQkB+giXaRg6g98Zq3daa3df+qG0yyzH87Ih+5AFE4jqt04cz6x1n6D05tiDHACp4gQ5jnuA57XxXmOM+k8Of4PZ97oIZxVT3467flbcOa1bsBZwuzBU09YaxOB74CqwMSLe/tndRNOg9K9wPc4/Tl+8Tz3jef+sDHmjxw6n4iIiDhNMBMz3Z4HXgJigFU4Tbb/8GwDqAH8ApwEFgHDrbWzcPpdvIZzXbEfp9n3U9kJ5umRdTPwgee43XEadKZ4dnkVp8gSZ4x5NDvnEpHsMU7vGxER72OMeRaoaa29+bw7i4iIiIhIgXWm+e4iIq7zTDO5k787j4uIiIiISCGlaSMi4nWMMQNxGnpOt9bOdTuPiIiIiIi4S9NGRERERERERMSraeSFiIiIiIiIiHi1fNfzIjw83FapUsXtGCIiIoXesmXLDllrS7udIzt0XSEiIuIdznddke+KF1WqVCEmJsbtGCIiIoWeMWaH2xmyS9cVIiIi3uF81xWaNiIiIiIiIiIiXk3FCxERERERERHxaipeiIiIiIiIiIhXU/FCRERERERERLyaihciIiIiIiIi4tVUvBARERERERERr6bihYiIiIiIiIh4NRUvRERERERERMSrqXghIiIiIiIiIl5NxQsRERERERER8WoqXoiIiBRwKWkZbkcodNLS9TsXERHJSSpeiIiIFFDxyWm8NWMjl78zm5PJaW7HKTRenLqOW0b97nYMERGRAkXFCxERkQImI8MyIWYX7d+azdBZm2kaWZLk1HS3YxUaESWKsGjrYZZuP+J2FBERkQJDxQsREZECZMnWw/QYNp/Hv11FRIkiTLynFe/1aUKpkEC3oxUaN7WIJKxoAMNmbXY7ioiISIHh53YAERERyb6dhxN4dfp6pq/ZT4XiQbzXpzE9GlXAGON2tEKnSIAvd7apypszNrJmzzHqRxR3O5KIiEi+p5EXIiIi+diJpFRenb6eTu/MYfbGWB7uXJNfH2lPz8YRKly46JZLK1MsyI/hszX6QkREJCdo5IWIiEg+lO7pa/H2zxs5dDKF65tW5LEra1GueJDb0QQIDfKn/6VVGDZ7M5sPniCqTDG3I4mIiORrGnkhIiKSzyzcfIhu78/jqYmrqVKqKJPvbc3bvRupcOFl7mhTlSA/X4bP3uJ2FBERkXxPIy9ERETyiW2H4nll2npmrjtARIkiDO3bhG4Nymt6iJcKKxrATS0i+WLRdh7qVJNKYcFuRxIREcm3NPJCRETEyx1LTOWlqeu44t05LNx8iMeurMWvj7Tj6oZqyOntBl1WDV9j+GiuRl+IiIhkh0ZeiIiIeLGY7Ud4YNwK9h5LpHezSjxyZU3KFNP0kPyiXPEgrm9WkQkxu7m/Yw3KhOp/OxERkazQyAsREREvlJ5hGTZrMzd+vBhfH8PEu1vx+g0NVbjIhwa3q0ZaegYj529zO4qIiEi+peKFiIiIlzl4PIlbP13CmzM20rVBeX68vw1NIku6HUuyqHKpovRoVIHRi3dwND7F7TgiIiL5kooXIiIiXmT2xoNc9d48lu04yhvXN+T9Po0pFuTvdizJpns6RJGQks5nC7e7HUVERCRfUvFCRETEC6SkZfDqtPXc9tlSwkMC+WFIG3o3r6SGnAVEzbLFuKJuWT5fsI0TSaluxxEREcl3crV4YYzpYozZaIzZbIx58gzPRxpjZhljlhtjVhljuuZmHhEREW+083ACvT5axEdzt9KvZSSTh7SmRtlibseSHDakYxTHk9L4eslOt6OIiIjkO7lWvDDG+ALDgKuAusBNxpi6p+32DDDBWtsE6AMMz608IiIi3ujHVfvo9v48tsaeZHi/prx8bQOC/H3djiW5oGHFErStEc7IedtISk13O46IiEi+kpsjL1oAm621W621KcA4oOdp+1gg1PNzcWBvLuYRERHxGokp6Tw1cTX3jvmDqLIhTLu/LV0blHc7luSyeztEcehkMuOX7nI7ioiISL6Sm8WLCCDzJ/Nuz7bMngduNsbsBqYB953pQMaYQcaYGGNMTGxsbG5kFRERyTObDpyg57D5jP19J3e3r86Euy6lUliw27EkD7SsGkZ05ZJ8NGcLKWkZbscRERHJN9xu2HkT8Lm1tiLQFfjKGPOvTNbaj6210dba6NKlS+d5SBERkZxgrWXs7zvpMXQ+R+JT+PKOFjzRpTb+vm5/HHs/Y0yQMeZ3Y8xKY8xaY8wLZ9gn0Bgz3tNra4kxpkreJz03Ywz3doxi77EkJq3Y43YcERGRfCM3r5b2AJUyPa7o2ZbZncAEAGvtIiAICM/FTCIiIq44npTKfWOX89TE1URXDmPaA225rKYK8hchGehorW0ENAa6GGMuOW2fO4Gj1too4F3g9TzOeEHa1yxNvQqhfDh7C+kZ1u04IiIi+YJfLh57KVDDGFMVp2jRB+h72j47gcuBz40xdXCKF5oXIiIiXis9w3IiKZW4hFSOJqQQl5jKsYRU4hJSOJqQyrHEv392nvPsk5iKjzE83qUWgy+rjo+PlkC9GNZaC5z0PPT33E7/y78nzpRUgG+BocYY43lt3ok/DAHB4F/kjE8bY7i3QxT3fP0H09fs4+qGFfI0noiISH6Ua8ULa22aMWYIMAPwBT611q41xvwPiLHWTgEeAT4xxjyEcwFyW55fYIiIiJzDkfgU5m6KZfbGgyzccpjYk8mc65MqNMiPEsEBlAj2p3gRfyLDgilRxJ8Swf50qlOWRpVK5F34AsazktkyIAoYZq1dctouf/Xb8lyHHANKAYdOO84gYBBAZGRkzgf9+WnYOhvaPQ5NbgFf/3/t0qVeOaqXLsqwWVvo1qA8xqiYJSIici65OfICa+00nEacmbc9m+nndUDr3MwgIiJyMdIzLCt3xzF7YyxzNsWyancc1kJY0QDa1gincqmifxUjSgYHUDzY3/M4gNAgP/zUvyLXWGvTgcbGmBLA98aY+tbaNVk4zsfAxwDR0dE5/6VJk5vhyFaY+hAseA86PA31rwefv5fA9fEx3NM+ike+WclvGw5yeZ2yOR5DRESkIMnV4oWIiEh+cOhksmd0RSzz/ozlaEIqxkDjSiV48PKatK9VmgYRxTXVw0tYa+OMMbOALkDm4sWpflu7jTF+OMuwH87zgFXawB0z4M+f4dcXYeJAmP8udPwv1LoKPKMsejSuwLu/bGLorM10rF1Goy9ERETOQcULEREpdNIzLCt2Hc00uuIYAOEhAXSoXYZ2NUtzWY3SlCwa4HJSOcUYUxpI9RQuigCd+XdDzilAf2ARcAPwm2vTUY2BmldCVGdY9z389jKMuwkiouHyZ6FaO/x9fbirXXX+O2kNi7YeplV19SwXERE5GxUvRESk0Fiz5xgfzd3K3E2xngaa0CSyJI90rkn7WmWoVyFUoyu8V3ngC0/fCx9ggrV26mm9tEbhLLu+GTiC0yzcXT4+zpSROj1gxRiY8zp82QOqtYeOz9KrWWPe//VPhs3arOKFiIjIOah4ISIiBd7B40m89fNGvlm2m9AgfzrXLUv7WqVpExVOiWCNrsgPrLWrgCZn2J65l1YS0Csvc10wX39o1h8a3ggxn8K8t2BkR4JqX83jTW7lsbmHWb7zKE0iS7qdVERExCupeCEiIgVWUmo6o+ZvY/iszaSkZzCgTVWGdKxB8SL/Xv1BJE/4B8Gl90DTW2Dxh7DwA25I/pEiQW0Y/3M6TQb0cDuhiIiIV1LxQkREChxrLVNX7eO16RvYE5fIFXXL8p+udagSXtTtaCKOwGLOUqrNB2AW/B9dFo3gyl0LOTqhLyW7PA2h5d1OKCIi4lVUvBARkQJl5a44Xpy6jpgdR6lTPpQ3ezVULwHxXsFh0Pl/xDccwPThD9Nr3RjY9A007Q+t74fiFd1OKCIi4hVUvBARkQJh/7Ek3vhpAxOX7yE8JIDXrmtAr+hK+KoBp+QDxctGsu2SF+k4tytT6yyhWMwopzdG45ugzUMQVs3tiCIiIq7ycTuAiIhIdiSmpPPeL3/S4a3ZTF21j7vbV2fWo+3p0yJShQvJVwa0qcZ+3/K85Hsv3L8cmt0GK8fDB81g4iA4uMHtiCIiIq7RyAsREcmXMjIsU1bu5fWfNrDvWBLdGpTnyatqUyks2O1oIllSulggfZpX4qvFO/D3M9x/+UuUuexRWPiBMwpj1QSo2wPaPgrlG7odV0REJE+peCEiIvnOHzuP8r8f1rFiVxz1I0J5r08TWlQNczuWSLY91qU2FhizZCffLdvDwLZVGdjueYq1eRgWD4ffP4Z1k6HGlXDZY1CpuduRRURE8oSx1rqd4aJER0fbmJgYt2OIiEgWrdodxyvT1rNq97EsHyMhJZ0yxQJ5vEttrmsSgY+mh7jCGLPMWhvtdo7s8Nbriu2H4nnz5438uGofpYoGcF/HKPq2rExA6nFY+gksGg6JR6BqO6eIUaUNGP3/QERE8q/zXVeoeCEiInli/7Ek3pixgYl/OA01uzeqgF8Wiw5ligXRt2UkRQM1gNBNKl7kvpW74nht+gYWbT1MZFgwj1xRk+4NK+CTGg/LPnOmlJw8AJVaOkWMqE4qYoiISL6k4oWIiLgqMSWdT+Zt5cPZW0jPsNzRpir3dqhOsSB/t6NJNql4kTestcz98xCvTd/A+n3HqR8RypNd6tCmRjikJsLy0TD//+D4bijfGC57FGp1Ax/1ZRcRkfzjfNcV+spKRERyhbVOQ83XpjsNNbs2KMeTXeoQWUoNNUUuhjGGdjVL0zYqnMkr9/DWjE3cPGoJbWuE80SX2tRvMRCa9odV42He2zD+ZihTF9o+AvWuBR9ft9+CiIhItmnkhYiI5Lg/dh7lxanrWL7Taaj53251aVmtlNuxJIdp5IU7ktPS+WrRDobO2kxcQio9G1fgkc61nMJgehqs/R7mvQWxGyCsulPEaNgbfDXaSUREvJemjYiISJ7ZG5fI6z9tYPKKvZQuFsjjV9bi+qYV1VCzgFLxwl3Hk1IZMXsLny7YRnqGpV/LytzXMYpSIYGQkQEbpsLcN2H/KigeCW0ehCY3g1+g29FFRET+RcULERHJdQkpaYyYs5WP527BWhjYthp3t6+uhpoFnIoX3mH/sSTe+3UT45fuIjjAjzvbVOXOtlUJDfIHa+HPn2HOG7AnBoqVh1b3Q7PbIEBTuERExHuoeCEiIrkmI8Py/fI9vDFjAweOJ9O9UQWe6FKLiiX1R1FhoOKFd9l88ARvzdjET2v3UyLYn7suq07/VpUJDvBzihjb5sDct2D7PAgOh1ZDIPpOCAp1O7qIiIiKFyIikjuW7TjKCz+sZdXuYzSqWJxnu9elWeUwt2NJHlLxwjut3n2Md2ZuZNbGWMJDArinfRR9W0YS5O9p3LljkTOdZMuvEFQCLrkbWt4FRUq6G1xERAo1FS9ERCTHjV68g+emrKV0SCBPXFWLno0i1NeiEFLxwrst23GEt3/exMIthykXGsR9l0fRq1klAvw8S6juWQZz34aNP0JAMWgxAFo/oCKGiIi4QsULERHJMekZllemrWfU/G10rF2G929qQoj6WhRaKl7kDws3H+Ktnzfyx844KoUV4cHLa3JNkwh8TxUc969xllhd+z2ERsD1I6Hype6GFhGRQud81xU+eRlGRETyr/jkNO76ahmj5m/jtlZV+OTWaBUuRPKBVlHhfHd3Kz67vTnFi/jzyDcrueLdOUxdtZeMDAvl6kOvz2Dgr85yqp93dRp8ZqS7HV1EROQvKl6IiMh57T+WRO+PFvHbhgO80KMez/eo9/e3tiLi9YwxdKhVhh+GtGHEzU3x9TEMGbOcbh/M55d1B7DWQkQzuGsu1L8eZr0MX/aE43vdji4iIgKoeCEiIuexdu8xrhm2gO2H4hnVvzn9W1VxO5KIZJExhi71yzP9gct4r09jElPSGPBlDNcMX8i8P2OxgcXguk+g53CnJ8aHrWHTDLdji4iIqHghIiJn9+v6A/QasQgfA9/e3YoOtcu4HUlEcoCvj6Fn4wh+ebgdb1zfkEMnkrll1O8MGbOcE8lp0KQfDJrj9MAY0xt+egrSkt2OLSIihZiKFyIi8i/WWj6dv42BX8ZQvXQIk+5tTZ3yoW7HEpEc5ufrQ+/mlfjt0XY8dmUtflq7n55DF7Bx/wkoXRMG/AIt7oLFw2FUZzi8xe3IIiJSSKl4ISIi/5CWnsFzU9byv6nr6Fy3LOPvuoQyoUFuxxKRXBTo58u9HaL4ekBLjielcc2wBUxavgf8g6DrG9BnDMTthI8ug5Xj3Y4rIiKFkIoXIiLylxNJqQz4MoYvF+3grsuq8WG/ZgQHaEURkcLikmqlmHZ/GxpULM6D41fwzKTVJKelQ+1uMHg+lGsI3w+C7wdD8km344qISCGi4oWISD528HgS78zcxA8r97LrSIKzYkAW7YlLpNeIRcz78xCvXteAp7rWwUcriogUOmVCgxgzoCV3XVaN0Yt30nvEInYfTYDiFaH/D9DuCVg13hmFsW+l23FFRKSQyNWv04wxXYD3AF9gpLX2tdOefxfo4HkYDJSx1pbIzUwiIgVFUmo6A76MYdXuY39tCw8JoFHFEjSuVILGkSVoWLEExYv4n/dYK3fFcecXMSSnpfPF7S1oUyM8N6OLiJfz8/Xhqa51aBJZkse+WcnVH8zn/25sTPtaZaDDf6BKW5g4EEZ2gs7/g5aDwajYKSIiuSfXihfGGF9gGNAZ2A0sNcZMsdauO7WPtfahTPvfBzTJrTwiIgWJtZanJq5m1e5jfNivKZXCglm+K44VO+NYuTuOXzcc/GvfaqWLOsUMz612uVAC/P4eePfTmn08OH4F4SGBjB3Ykhpli7nxlkTEC3WpX45a5Ypx9+hl3P75Uu7rWIMHLq+Bb9W2MHgBTL4HfnoSts6BnsOgaCm3I4uISAGVmyMvWgCbrbVbAYwx44CewLqz7H8T8Fwu5hERKTBGzd/G98v38EjnmlzVoDwA9SOKc8sllQE4npTKql3HWLk7juU745i76RAT/9gDQICfD/UqhNK4Ugn8fX34ZN5WGlcqwSe3RhMeEujaexIR71Q1vCjf39Oa/05ew/u//snynUd5r08TwoqWgpvGwZKPYOZ/4cNLocdQqHmF25FFRKQAys3iRQSwK9Pj3UDLM+1ojKkMVAV+O8vzg4BBAJGRkTmbUkQkn5m7KZZXpq3nqvrlGNIx6oz7hAb506ZG+F/TP6y17D2WxIqdcazYdZSVu44x9vedJKVmcHXD8rzVqxFB/r55+TZEJB8pEuDLmzc0JLpySZ6dspZu789jWL+mNI0sCZcMhsqt4Pu7YEwvaNofrnwZAjWKS0REco63tJDvA3xrrU0/05PW2o+BjwGio6Oz3o1ORCSf234onvvGLqdm2WK81asR5gLnmBtjiChRhIgSRejW0BmpkZaewYETyVQoHnTBxxGRwssYQ58WkdSPKM7g0cu48aNFPNOtLrdeWhlTviEMmg2zXoYF78PW2XDtCKeoISIikgNyc7WRPUClTI8reradSR9gbC5mERHJ904mpzHwyxiMgU9ujaZoYPbqz36+PkSUKKLChYhclPoRxfnxvrZcVqM0z01Zy/3jVhCfnAZ+gU7zztunO807P+sKPz8DqUluRxYRkQIgN4sXS4EaxpiqxpgAnALFlNN3MsbUBkoCi3Ixi4hIvpaRYXlo/Aq2HopneF+nQaeIiFuKB/vzya3RPHZlLX5ctZeewxaw+eAJ58nKlzrNPJvdBgs/gI/ba0lVERHJtlwrXlhr04AhwAxgPTDBWrvWGPM/Y0yPTLv2AcZZazUdRETkLN779U9mrjvAM93q0CpKy5iKiPt8fAz3dohi9J0tORqfQo+hC5i8wjPINjAEuv8f9PsWEo/CJx1hzpuQnuZuaBERybdMfqsZREdH25iYGLdjiIjkmZ/W7GPw6D+4oVlF3ryhoaZ5iNcwxiyz1ka7nSM7dF2RMw4cT+K+Mcv5ffsR+raM5Nmr6/7dBDjhCEx7FNZ8BxHRTi+M8BruBhYREa9zvuuK3Jw2IiIi2bRh/3EenrCSxpVK8NI19VW4EBGvVDY0iDEDWzK4XXXGLNnJ9R8uZMfheOfJ4DC44VPndmQLjGjrLK+akeFuaBERyVdUvBAR8VJH41MY+GUMIYF+fHRLMy1lKiJezc/Xhyevqs3IW6PZfTSRqz+Yz09r9v+9Q/3r4e5FUKUNTH8cvroGju12L7CIiOQrKl6IiHihtPQMhoz9gwPHkhlxSzPKhga5HUlE5IJ0qluWqfe1oVp4UQaPXsaLU9eRkuYZZRFaHvp9A93fgz3LYPilsGIs5LNpzCIikvdUvBAR8UKvTt/Ags2Heena+jSNLOl2HBGRi1IpLJgJgy/ltlZVGDV/G30+XsTeuETnSWOclUgGz4ey9WHSYBjXF47ucDWziIh4NxUvRES8zHfLdjNq/jZua1WF3tGV3I4jIpIlgX6+PN+jHkP7NmHTgZN0e38eszYe/HuHsKpw21S44iXYOhuGtYBZr0JKgmuZRUTEe6l4ISLiRVbsiuOp71fTqnopnu5Wx+04IiLZdnXDCkwZ0pqyoUHc/tlS3pqxkbR0zzQSH19odR8MWQq1u8Gc15wixrrJmkoiIiL/oOKFiIiXOHg8ibu+iqFMsUCG9m2Kv6/+iRaRgqFa6RAm3duaG6MrMXTWZm4etYSDJ5L+3qF4RWc1ktt+hMBQmHArfNkTDq53L7SIiHgVXRmLiHiB5LR07hq9jOOJaXxyazRhRQPcjiQikqOC/H15/YaGvNWrESt2xdH1vfks2nL4nztVaQN3zYWub8G+lfBha5j+JCTGuRNaRES8hooXIiIus9by7KS1LN8Zxzu9G1GnfKjbkQquEwfglxcg4R19x24AACAASURBVIjbSUQKrRuaVWTyvW0ILeJHv5GLGTZrMxkZmaaI+PpBi4Fw3x/QrD8sGQEfNIM/voSMDPeCi4iIq1S8EBFxSUaGZdGWwzw8YSXjY3ZxX8cormpQ3u1YBZe1MGUIzH8Hvr0d0tPcTiQXwRhTyRgzyxizzhiz1hjzwBn2aW+MOWaMWeG5PetGVjm/WuWKMWVIG7o1rMCbMzZy5xdLOZ6U+s+dipaCq9+Fu+ZAqSiYch+M7Ai7lroTWkREXKXihYhIHrLWsm7vcV6dtp7Wr//GTZ8sZsba/dzWqgoPdarpdryCbeU4+PNniOrsrGww879uJ8o7BaPxYRrwiLW2LnAJcK8xpu4Z9ptnrW3suf0vbyPKxQgJ9OP9Po15sWc95v15iGuHLWD7ofh/71i+EdzxE1z3CRzfB6M6wfd3OyOpRESk0PBzO4CISGGw+2gCk1fsZfKKPWw6cBI/H8NlNUvz5FW16Vy3LMEB+uc4V53YDz89AZUugb4TYMZ/YPFwKFsfmvRzO13uSU+FpaNg5Vi4Ywb4B7mdKMustfuAfZ6fTxhj1gMRwDpXg0m2GGO45dIqRJUpxt1fL+Oa4QsY3q8praqHn74jNOwNta6CeW/DwqGw/gdo9zi0HAx+6hMkIlLQGZvPvo2Jjo62MTExbscQETmvo/Ep/Lh6H5NX7GHp9qMANKtckmsaV6BbwwpqyplXrIVxfWHLbzB4AYRHOVNGRl8LOxfD7dOhYrTbKXOWtbBpBvz8DBz+E6q1h2s/gmLlcvQ0xphl1to8/+UZY6oAc4H61trjmba3B74DdgN7gUettWvP8PpBwCCAyMjIZjt27Mj90HJeOw7Hc+cXMWw/FM8LPevRr2Xls+98eAv89BT8OQPCa8L1I50RGiIikm+d77pCxQsRkRyUmJLOL+sPMHnFHuZsiiU13RJVJoRrGlegZ+MIKoUFux2x8Fn1DUwcAFe8BK3u+3t7whH4pAOkJsGg2RBaQPqNHFjnjCzZOsvpE3DFy1DzSueb6xzmRvHCGBMCzAFettZOPO25UCDDWnvSGNMVeM9aW+Ncx9N1hXc5npTK/WOXM3tjLLe1qsIz3ergd65lozfNgB8ehITDcNXr0Oy2XPlvXUREcp+KFyIiuSwxJZ3FWw8zddU+flqzj/iUdMqGBtKjkVOwqFchFKOLaXecPAjDWjh/xN8xA3x8//n8gXUwshOUqQ23TcvX0yo4GQuzXoY/voDAUGj/FDS/E3z9c+2UeV28MMb4A1OBGdbady5g/+1AtLX20Nn20XWF90nPsLw6bT0j52+jbY1wht7UlOLB5/jvOP4QTBzojK5q2AeufgcCiuZdYBERyRHnu67QJGsRkYtkrWXboXhmb4xl9qZYlmw9THJaBsUC/ejWsDzXNI6gZbVS+PqoYOEqa+HHhyElAXoO+3fhAqBsXbjuIxh/M0x9CK4Znv++tU1LhsUfOn0AUhOgxSBo9wQEh7mdLEcZpwI4Clh/tsKFMaYccMBaa40xLXAakx/Ow5iSA3x9DM9cXZcaZUN4ZtIarh2+gJH9o6lWOuTMLygaDv2+hblvwexXYd8K6PWFU5QUEZECQ8ULEZELkJCSxqIth5mzKZbZG2PZeSQBgGqli9KvZWXa1SpNy6phBPmf4Q9kccfa752Gfp2eh9K1zr5fne7OKIXZr0K5+nDpvXmVMHushfVTYOazcHQ71OziTI0JP+csifysNXALsNoYs8Kz7T9AJIC1dgRwA3C3MSYNSAT62Pw2xFT+cmPzSKqGhzB49DKuGbaA4f2a0aZG+Jl39vGF9k9AZEv4boAzJaz7e06TTxERKRA0bURE5AystWyJjWf2xoPM2RTLkm1HSEnLoIi/L62ql6J9rdK0r1VGPSy8VfwhZ7pIicpw50zwPU+tPiMDvrkVNvwIN38H1TvmTc6s2rscZjwNOxZAmbpw5cuuZHarYWdO0nWF99t1JIEBX8SwOfYkz3Wvy62XVjn3C47vg2/vgJ0LnR4YXV7P31PCREQKCU0bERG5QPHJzuiK2ZsOMntjLLuPJgIQVSaEWy+pTPtaZYiuUlKjK/KDaY9C8glnusj5ChcAPj5wzQgY1Rm+uR0G/galqud+zot1fB/8+j9n6dPgUnD1u9Dk1gt7jyL5VKWwYL67pxUPjlvOs5PXsunACZ7rXg//szXyDC0P/X+A316EBf8He5ZB7y8hrFreBhcRkRylqx0REWD9vuP0G7mEI/EpBAf40qp6OIPbVaddzdIaXZHfrJviTBnp+IzT0+JCBYZAnzHOcPOxN8GAXyAoNPdyXozkk7B4OMx/FzLSoPX90PYRCCrudjKRPBES6MdHt0TzxowNfDRnK1tj4xneryklgs+y5LSvH3R+ASIvhe/vgo/aOcXMuj3yNriIiOQYTRsRkUJv88GT3PjRIvx9fXizV0NaVA0j0E+jK/KlhCPOdJFi5Z3RE1lZaWPrHPjqWmd50Ru/dkZl5DVr4dCfsHkmbP4Fti+A9GSo2xM6vQBhVfM+0xlo2oi44dtlu/nPxNVUKBHEyP7NiSpzlkaepxzdAd/cBnv/gEvucf4/5HeWooeIiLhG00ZERM5h5+EE+o1cjDHw9cCWVD9bN3vJH6Y/AYlH4ZZJWV8itFo76PIqTH8cZr/ijODIC8knYdvcvwsWcTud7eG1oPkAqHctVGqeN1lEvNgNzSpSNTyYu75axrXDFzC0b1Pa1Sx99heUrOwslfzzM84Ipt1L4YbPoESlvAstIiLZpuKFiBRae+MS6TtyMclpGYwbdIkKF/ndhmmweoKzcki5+tk7VotBsH8VzH0TytZzCgc5zVo4uP7vYsWORZCRCgEhULUdtHkIql/u/OElIv/QrHIYk+5tzYAvYrj9s995vEtt7rqsGuZsSx37BUDXN6DypTD5PvioLVz3CdTonLfBRUQky1S8EJFC6eCJJPqNXMKxhFS+HtiS2uW8pLeBZE3iUZj6EJStD20ezv7xjIFu70DsJph0D5SKgnINsn/cpGOwdbZTrNj8Kxzf42wvUw8uudv5Q6rSJRrSLnIBKpYM5ru7W/H4t6t4bfoGVu6K481ejQgJPMflbb1roWwD+KY/fH2DUySscQX4+P3z5uvvLL/q459pW+Z9PM+frVgiIiI5Tj0vRKTQORqfQp+PF7PzSAJf3dmC6CphbkeS7Pr+blg1HgbNgvKNcu64J/bDxx2cP1YGzYKi4Rf+2tREOLgO9q2C/ath30pniVObDoGhUK29U6yofjkUj8i5zHlIPS/EG1hrGTlvG6/9tIEqpYL56Jbo8/fBSE10pob98WU2zmyg1X1wxYvZOIaIiJyinhciIpkcS0zllk+XsO1wPJ/f1lyFi4Jg0wxYOQYueyxnCxcAxcpBn9Hw6VUwoT/cepZeGglHnGkmpwoV+1fDoU1OoQKcYkW5BtD6AadgUbF51ntyiMg/GGMYeFk16kWEct+Y5fQcOp+3ezeiS/3yZ3+RfxHo8QE0HwgJhyEj3VnJJyPNmb516nF6aqbtmW7paXBgNSx8H0pEQouBefeGRUQKKRUvRKTQiE9O4/bPfmfj/hN8fEs0raIu4lt08U5Jx+CHB6F0Had4kRsimjl/5Hw/CH56Elrd7xQq9q/+u1hxfPff+4dGOIWKOt2d+3INoGQVDS8XyWWtqocz9f42DB79h3NrV51Hr6iJn+85Vgwq3zDrJ8xIh3F9nUbBYdUg6vKsH0tERM5LxQsRKRSSUtMZ8EUMK3bFMaxvUzrULuN2JMkJM56Gkwegz9fgF5h752l0o+db1g9g6Uhnm/GB8JpOA8ByDT2FioZQtFTu5RCRcypfvAgT7rqEF35Yx4g5W1i9J473+zShVEgu/Pvg4wvXj4RPu8A3t8OAX6B0zZw/j4iIACpeiEghkJyWzuDRy1i87TDv9G7EVQ3OMZRY8o/Nv8Dyr5yGexFNc/98nV6AklWdokW5hlC2rjP0XES8SqCfL69c24DGlUrwzKQ1dP9gPh/e3IxGlUrkwsmKwU1j4ZOOMKY3DPwNgjUdUUQkN5xjHJ2ISP6Xlp7BA2NXMHtjLK9c24Brm1R0O5LkhKTjMOUBCK8F7Z7Mm3P6+ELzOyH6dqjYTIULES/XO7oS3w1uhTGGXiMWMe73nblzohKR0GcMHN8L42+GtJTcOY+ISCGn4oWIFFjpGZZHv1nJT2v38+zVdbmpRaTbkSSnzHwWTuyFnsPAP8jtNCLipRpULM4P97WhZbUwnpy4mqcmriI5LT3nT1SpBfQcCjsWwI8PQT5bzU9EJD/I1eKFMaaLMWajMWazMeaMX40ZY3obY9YZY9YaY8bkZh4RKTystTz9/WomrdjLY1fW4o42Vd2OJDnBWlj+NSz7DC69Fyo1dzuRiHi5sKIBfH57C+5pX52xv++i94hF7I1LzPkTNeztNA5ePtrpj5Mbkk/ApHvg3fqw7AunaaiISCGRa8ULY4wvMAy4CqgL3GSMqXvaPjWAp4DW1tp6wIO5lUdECg9rLS/8sI5xS3cxpEMU93aIcjuS5IQ9y5zGeJPvcVYA6fC024lEJJ/w9TE83qU2I25uxpbYeK7+YD4LNx/K+RO1/w/U7emMDtswLWePvXcFfNQOVo6FgBD44X4Y0RY2/5qz5xER8VK5OfKiBbDZWrvVWpsCjAN6nrbPQGCYtfYogLX2YC7mEZFC4s0ZG/l84XbubFOVR65Q5/d879gemDjIaYh3ZCt0fx/unKmeEyJy0brUL8fkIa0pVTSAm0ctYcScLdicnOLh4wPXjIAKjeG7Ac5SytllLSwaDiM7QWoi9P8B7lkEvT6HlJMw+joYfT0cXJ/9c4mIeLHcLF5EALsyPd7t2ZZZTaCmMWaBMWaxMabLmQ5kjBlkjIkxxsTExsbmUlwRyY8yMizxyWkcPJHE9kPxvDtzE8Nnb6Fvy0ie6VYHY4zbESWrUuJh1ivwQTNYOwnaPAz3/wHN+jvNM0VEsqB66RAm3duaq+qX57XpG7jrq2UcS0zNuRMEBEOfsRBUHMb0gRMHsn6s+EPOKiYznoKoTnD3AqjSBoyBetfCkKVwxUuwayl82Ap+eABO6rtAESmY3F4q1Q+oAbQHKgJzjTENrLVxmXey1n4MfAwQHR2tDkgiBdixhFTGx+wkLiGVhJR0TiankZCSRnxy+j/vU9JJSE4jITX9X33RrmsSwUs966twkV9lZMCq8fDrC3Bin3OB3ukFKFnZ7WQiUkAUDfRjaN8mNJlfgtemb6D7B/MZ3q8p9SOK58wJQss7S6h+2gXG9YXbfrz45sLb5sJ3AyHxCFz1BrQY5BQtMvMLhFb3QeN+MOd1WDoSVn8LbR6ES4dohJqIFCi5WbzYA1TK9LiiZ1tmu4El1tpUYJsxZhNOMWNpLuYSES/22Lcr+XndAfx8DEUD/Sga4EvwqfsAPyqU8Cc4wO/fz3nuw4oG0LZGaXx8VLjIl3Yscr5h3LscKjR1hkVHXuJ2KhEpgIwxDGhbjSaRJRgyZjnXfbiQ57rXpW+LyJwpfldoDNd9DBNugcn3wvUj/118OJP0NJj9Ksx7G0pFQb9voHzDc78mOAyueh2aD4RfnoPfXoKYz+DyZ6FBb2c6i4hIPnfBxQtjTLC1NuEijr0UqGGMqYpTtOgD9D1tn0nATcBnxphwnGkkWy/iHCJSgPy6/gA/rzvAE11qM7hdNY2cKEyObnca3K2bDMUqwLUfQ4NeuuAWkVzXrHIYP97flgfHr+Dp79fw+7YjvHJtA4oG5sB3fHV7QMf/wm8vQula0O7xc+8ft9PplbFrCTS52RlxEVD0ws8XHgV9vobt82HGf+D7u2Dxh3Dly850ExGRfOy8V4XGmFbGmHXABs/jRsaY4ed7nbU2DRgCzADWAxOstWuNMf8zxvTw7DYDOOw5/izgMWvt4Sy+FxHJxxJT0nluylpqlAlhQNuqKlwUFknHYeZzMLQ5/DnT6dR/3zJodKMKFyKSZ8KKBvD5bc15pHNNfli5l57DFvDngRM5c/C2j0DDPjDrZVgz8ez7rZsMI9rAgXVw/SjoOeziCheZVWkDA2fDtR9BfCx83g3G9YNDm7N2PBERL2DO12HZGLMEuAGYYq1t4tm2xlpbPw/y/Ut0dLSNiYlx49QikovenLGBYbO2MH7QJbSsVsrtOIWHtRc2jDmnZaTDH186F/PxsdCoL1z+XwitkPdZJMuMMcustdFu58gOXVfI6RZuPsT945YTn5zOy9fW57qmFbN/0LRk+KI77FsJt09zlns+JTURfnoKln3mTJe7YRSEVcv+OU9JSYDFw2D+/0FakjO1pNPzF9+DQ0Qkl53vuuKCxsNZa3ed9i1oenaDiYicsvngST6eu5XrmkaocJGX/pzpDE+2FoqVhZCyEFLmtPtMt+BS5x8NkZoE8QedbvcnDzi3E577v7Z57tOTIfJS6DsBIprmzXsWETmPVlHhTLu/LUPGLufhCStZuv0Iz3WvR5B/NlY58guEG7+GkR1hbF8Y+BsUj3CWN/3mdohdD60fgA7PgF9Azr0ZcFY/uewxaHIrzH4FlnwIhzZCnzFq6Cki+cqFFC92GWNaAdYY4w88gDMNREQk26y1/HfSGor4+/KfrnXcjlN4rJ3kFC5K14bKrf4uKuxd4fyccvLfrzG+ULS0U9goVg6KlnG+xfurILEfko6d+XzB4X8XRcJrOPeVWkKtru6M/BAROYcyoUGMGdCSt2du4sPZW1i56xjD+zWlSngWp3EAhJSGm8bDqCtg7I1OMWHmfyGwGNw8EaIuz7k3cCbFykL39yAiGqbcB2NuhJvGOcUNEZF84EKKF4OB94AInMabPwP35mYoESk8pqzcy6Kth3npmvqEhwS6HadwWP41TBkCFVtAvwkQdIalAZNP/nMExV+jJzKNnNi/xvk2MaQslKkN1dqdecRG0XDw9c/79ykikg1+vj480aU2zauU5KHxK+n+wXze7NWQLvXLZ/2gZevCDZ86xYvpj0H1jk5fipAyORf8fJreAj5+MOluGNMb+o7Pem8NEZE8dN6eF95Gc1NFCo5jialc/vYcIkoWYeLdrfDV8qa5b8lHMP1xqNbB6UivC1bJBvW8kMJi99EE7h2znJW74rijdVWevKo2AX7ZaCq85jtIPArN7nCvOfGqCc5qJJUucQrZgcXcySEi4pHtnhfGmM+Af1U4rLV3ZDObiBRy7/y8kSPxyXx+e3MVLnKbtTDvbWe5vtpXO9/8+Wmki4jIhahYMphv7rqUV6at59MF21i+6yjD+jalQoks9oyof33OBsyKhr3Bxxe+Gwijb4B+30BQqNupRETO6kJKvVOBHz23X4FQ4AyToUVELtzq3cf4avEObr20CvUjzjBtQXKOtfDLc07houGN0OsLFS5ERC5SgJ8Pz/eox7C+TfnzwEm6vT+PuZti3Y6VPfWvd4rZe2Jg9HVn71skIuIFzlu8sNZ+l+n2NdAbyNdDREXEXekZlmcmraZUSCAPX1HT7TgFW0YG/PgwLHgPou+Ea0aA7wUtNCUiImfQrWF5pgxpTdnQIG7/fClfLdrudqTsqXcN9Poc9i6HL6+BxDi3E4mInFFWJtnVAPKwq5CIFDRjft/Jyt3HeKZbHUKD1Mgx16SnwaTBEPMptH4Qur3t3txqEZECpFrpEL69uxXta5bmv5PX8vyUtaSlZ7gdK+vqdIfeX8H+1fBlT0g44nYiEZF/Oe9VrDHmhDHm+Kl74AfgidyPJiIFUeyJZN74aQOtqpeiR6MKbscpuNKS4Zv+sGo8XP4sdH5BS5KKiOSgkEA/Pr41moFtq/L5wu3c8UUMx5NS3Y6VdbW7Oo2cD66DL3uogCEiXudCpo0Us9aGZrqvaa39Li/CiUjB8+q09SSlpvO/nvUx+mM6d6TEw5gbYcNUuOpNaPuI24lERAokXx/D093q8up1DVi4+RDXD1/IriMJbsfKuppXQp+xELsJvugO8YfcTiQi8pezTnw2xjQ91wuttX/kfBwRKcgWbz3MxOV7GNIhiqgyIW7H+addS2HKEEg+ASFlIKRspvvMN8+2gOALO25KPJzYDycPwskDme4z3w6Cjx9UawdRnaFaeyhSImvvIzHOKVzs/h16Docm/bJ2HBERuWA3tYikcqlg7h79Bz2HLeCjW5rRvEqY27GypkYn6DsOxt4En18N/ac4n30iIi4z1v5rFVTnCWNmneN11lrbMXcinZvWYxfJn1LSMuj6/jySUtOZ+VA7igT4uh3JYS0sHg4zn4XQCKjSFk7u/7uoEB8L9gzzmAND/13ksBmnFSgOQsoZFmcyPlC0jPO6YuWc+6TjsHUOJB8D4wuVWkBUJ+dWruGF9aqIPwRfXQsH18P1I50mbCK56HzrsZ/jdZWBGtbaX4wxRQA/a+2JnE94frqukJy0NfYkd34Rw56jibx2fQOua1rR7UhZt3WOUwwvEQn9f4BiZd1OJCIF3PmuK8468sJa2yF3IolIYTRq/jY2HzzJqP7R3lO4SDwKk4c40ytqXw09h/17xENGOiQcPm30xGkjKPavhhO/OEWJYp4RGhWaZCpslPvnKI7gMPA5w+8gPRV2x8DmmfDnTGdp099edAodUZ0g6nKo3tF5/emO73WarMXthJvGQo3OufM7E8kmY8xAYBAQBlQHKgIjgMvdzCWSE6qVDuH7e1px9+g/eHjCSrbGxvNw55r4+OTDaZLV2sHN38LXveHzbk4BI7S826lEpBA768iLf+xkTH2gLhB0apu19stczHVW+oZEJP/ZfTSBzu/MpW2NcD6+1UtWWt7zh9PQ8vhe6PwiXHK39zW0PHEAtvwKm3+BLb85xRbjAxHRTjGjRico3wTidvzdHb7veKjS2u3kUkhkZeSFMWYF0AJYYq1t4tm22lrbIDcyno+uKyQ3pKRl8OzkNYxbuouuDcrxdq/G3lO4v1g7FsLXvZzie/8foHiE24lEpIDK8siLTAd4DmiPU7yYBlwFzAdcKV6ISP7zwg/rAHiuRz2Xk+BME/n9E/j5aWdEw+0/QaXmbqc6s2JloXFf55aR7hRcNs90ihmzX4XZr0BwKWdfm+HMS444Z7siEW+QbK1NOdWw1xjjB5z/mxSRfCTAz4dXr2tAVJkQXp62nt1HF/HJrdGUDQ06/4u9TeVWcPNEGH09fN4V+n4DpWu6nUpECqELmETNDThDOfdba28HGgHFczWViBQYv6w7wMx1B3igUw0iShRxN0zScfjmNpj+GFTrAIPneW/h4nQ+vk7WDv+Bgb/BY5vhupHOCIwydeG2aSpcSH4xxxjzH6CIMaYz8A3OMuwiBYoxhgFtq/HJLdFsPniSnkMXsGbPMbdjZU1kS7h1ktMU+sNWTp+o5DP0dBIRyUUXUrxItNZmAGnGmFDgIFApd2OJSEGQmJLOc1PWUqNMCHe2qepumH2r4ON2sP4H6PQC3DTuzL0j8oui4dCwF1z3Mdw2FcrWdTuRyIV6EogFVgN34YzqfMbVRCK5qFPdsnw7uBU+BnqNWMSMtfvdjpQ1FaNhyFJo2BsWvAdDm8Pq/2fvvsOrqNY2Dv/WTg8kIUCAQOgt9BaqIEVQOiIooKgUBcSC7ViOnnO+ox57A0WlWRFBESxIl16l994Teu+EJOv7Y4KiAkkgO5Py3Nc1VzKzZ89+ogFmv3utd411RjSKiGSA1BQvlhpj8gDDgGXAcmChV1OJSLbwwYwtxB0/xyu3V8bPJzV/3XiBtbD0MxjeHC6ec97oN3w8dat3iEi6s9YmWWuHWWvvtNZ2Tv5e734kW6tYOJQfHrmJcoVC6DdyGZ/M3kaW/LXPXQBu/wh6T4PcEfB9b/iiHRxY73YyEckBUux5Ya3tn/ztJ8aYyUCotXa1d2OJSFa39eAphs3dTqeaUdQtlc+dEBdOw4QnYM23zjSRO4Y5N1si4hpjzA6u0OPCWlvKhTgiGaZASCBj+tTjqe9W8fqkjczYeJBaxcOpGBlKxcKhlMiXC5+ssipJ0Trw4ExY9jn8+hJ80hDq9oMmz0KgZpeLiHekpmHnT8Bo4Edr7U6vJxKRLM9ay4s/rCXIz4fnW0e7E+LAemc1kcNboOkL0OipKy9PKiIZ7fIu4oHAnTjLpopke4F+PnzQtQaVC4fx06q9DJuznYQkp5YX5OdDdGTI78WMipGhRBcKzbyrlHh8oHZvqHg7zHgJFn0Ea76DW1+Gql0y3wpeIpLlpbhUqjGmMdAFaAMswSlkTLDWnvd+vL/TkmYiGc9aS3xiEmcvJHImPoGz8YmcueB8PX0hgbPxCZy5kPj71z3HzjJueRyv3F6Z7vWKZ3zgFV/DL09BQAh0Gu6sVS8i6e56lkq9xnVqpUemtNJ9hbjpQkIiWw+eZv3ek6zfd/L3r6fOJwDgMVAyfy4qFg77U1EjIiTA5eRXELccJj4NccugaD1o/RZEVnU7lYhkITe8VKq1djZOZ3AfoBnwIPApEJpuKUUkU0lITOLZ79ewcNthziQXKi59MpQawf4+3FapIN3qFPNiyitIiHemiawcCSUaOYWLkEIZm0FErskYc/myOB6ckRipGQlaFGeZ9oI4006GWmsH/uUcAwwEWgNngR7W2uXpFF0k3QX4+lCpcBiVCv8x1cJaS+yxc38qZizfdYyfV+39/Zzi+YJ5qHFpOtWKcq+n1F8VqQm9pzv/Bk//P6dJdkxvaPYCBIWn/jrnT8DBjXDosu3gRvANcJqFVusKeTXLTCQnSnHkBYAxJghohzMCoybOyItHvZztivQJiYj3vTZxA0PmbKdNlUgiQgII9vchV4Avufx9CA7wJZe/L8EBPs7XvzwW7OeDx605uz8/Dss+g0ZPQ5PnwSfF90MicgOuZ+SFMWbmZbsJwE7gbWvtphSeFwlEWmuXG2NCcJqI326tXX/ZOa2BR3GKF3WBgdbaute6ru4rJKs4cfYi6/edweWpMwAAIABJREFUZN3eE/y8eh+r9hynaN4gHm1alo41i2SeIgbAuWMw81VYMtwpXDT/P6je/c/Nss8dh0ObLitQbHD2T/1RpME3CCLKQUQFOH0Ats8CLBSrD9W6QaXb1WNDJBtJ6b4iNdNGvgXqAJOBMcDs5KVTXaGbDBHvmrB6L4+MWsG99Yrz8u2V3Y6TekuGO1NFbnocWvzX7TQiOUJ6TRu5ztf+EfjQWjvtsmNDgFnW2m+S9zcBTay1+652Hd1XSFZkrWXWpkO8N30zq2NPUCxvMI80K8MdNYrgm5mKGPtWw8R/wJ5FUKQWFK2bcpGiQDREJG95iv+54HEiDlaPgVXfwOHN4BsI0W2hejenMbd6W4lkaelRvLgNmG6tTUzvcNdDNxki3rNx/0k6Dl5ApcKhjHqwHv6+megG6Fp2zoMvO0DpZtBttG5eRDJIWooXxpgnr/W4tfbdNLxuCWAOUNlae/Ky4xOA162185L3fwWetdZe9cZB9xWSlVlrmbHxIO9P38KauBMUzxfMo83Kcnv1wpmniGGtU3CY9m84fzJ1RYrUXDNuOawaBWvGwvnjkLuQM62k+t1QoIL3fh4R8Zr06HkxJX0jiUhmdPxsPH2+XEZIoC8f3VMz6xQuju+Gb++D8JJOjwsVLkQyq5D0uIgxJjfwPfD45YWLNF6jD9AHoFixDO7NI5KOjDHcUqEgzaILMH3DQd6fvpmnv1vF4JlbebRZGdpXywRFDGOcPhVV7nL201KkuNY1o2o5222vwubJsPIbZ8WTBYMgsrpTxKjcGXK5tFy7iKS7VPW8yEz0CYlI+ktMsvT6fAkLth1mdJ/61CqehsZaboo/AyNucwoYD/4K+cu6nUgkR8noaSPGGD9gAjDlSiM1NG1EcjprLdPWH+D96VtYv+8kpfLn4rFbytKuWmF83OpHlZFOH3KWa101CvavAY8flLsNat4HZW/V8q0imdwNj7wQkezvvWmbmb35EP/rWDnrFC6shR/6w4G1cM93KlyIZBHGmECgN1AJCLx03FrbK4XnGWAEsOEaU0x+Ah4xxozGadh54lqFC5HsxhjDrZUK0bxCQaauP8D70zfz+JiVDJqxhQG3lKVt1WxexMgdAfX7O9v+tU5vjNXfwsYJUPoWaPsuhJdwO6WIXKcUx20ZY8YZY9oYY7LIGHIRSYvJa/fx4cytdK1dlLsvX9r05F5nRENmNfdtWP+D08G8bAu304hI6n0FFAJuA2YDUcCpVDzvJuBeoJkxZmXy1toY088Y0y/5nInAdmArMAzon+7pRbIAj8fQsnIhJj7WiI/vqYmfx8OA0Su57f05/LxqL1lt5PV1KVQZbvsfPLkBWr0JexbD4How731IvOh2OhG5Dqlp2Nkc6AnUA74DPktpOTNv0vBOkfSz5cApbh88n7IFQxjTtx4Bvsn9IhIuwMBqcGof5C8HZZo7W/GbwC/w2hfNCBsnwuhuUOVOuGOYhoGKuOQ6l0pdYa2tYYxZba2tmjwVZK61tp6XYl6T7iskJ0hKskxau5+Bv25m84HT9GhQgn+3reje0uZuOBEHk55xRmEUrAztBkKUK4slichVpHRfkeJoCmvtdGvtPUBNnLXYpxtjFhhjeibfcIhIFnTy/EX6frWMIH8fPu5e84/CBcDacU7hom4/CCsKS0bAyDvgjRLw9Z2weAgc2eZO8IMbYVwfiKwG7T9Q4UIk67n0kedxY0xlIAwo4GIekWzP4zG0qRrJ5AE380DDkny+YCfPfL+ahMQkt6NlnLAi0PVr6DISzh6F4c3hl6edFVBEJEtIVc8LY0w+oDvOcM0VwNdAQ+B+oMk1ntcSGAj4AMOtta//5fEewFtAXPKhD621w9P0E4hImiUlWZ4cs5LdR8/y9QN1iQwL+uNBa2HRYGcZs5avO8WB+LOwaz5smQZbp8OWqc654SWdKRtlWkCJhuAf7N3g5445Iy78gqDrKOeriGQ1Q40x4cC/cHpU5E7+XkS8zOMxvNCmAiGBfrw3fTNnLiTwftfqf/4AI7ur0A5KNoYZr8BvQ52RGK3fco6LSKaWYvHCGDMeKI8zR7XdZY2vxhhjrjrO0hjjAwwGWgCxwBJjzE/W2vV/OXWMtfaR60ovItdl0IwtTN9wkP+2r0TdUn9ZQmznPKdDd7tBf4xq8A92ihSXeksc3Q5bpjuFjOVfOf/4+wRAiZuSp5i0cBpopueoiMQE+K4nHN8DPSZAWFT6XVtEMtJn1tpEnH4XpdwOI5LTGGMY0LwsuQN9eXnCes58uYwh3WsR5J+DChiBodD6TajaBX4eAGO6Q/k2zjHdX4hkWqlpwjnIWlvRWvvaXzt2pzDPtQ6w1Vq73VobD4wGOtxAVhFJB9OTl1DrVDOK++oX//sJiz6C4HxQ9a6rXyRvKajbB+75Fp7dCfeOh9oPOPNJp/wTBteGwXVhxUhIiE+n4P+B7TOhzTtQzJWp8SKSPnYYY4YaY25JXkFERFzQu2FJ3uhUhblbDnHfp4s5eT4HNrGMqgV9ZkKLl2DbDOfeZdHHkJTodjIRuYLUFC8qGmPyXNoxxoQbY1LTvbsIsOey/djkY3/VyRiz2hgz1hhT9EoXMsb0McYsNcYsPXToUCpeWkSuZPuh0zwxZiWVi4Tyv46V+dv7hiPbYNMkiOmV+ikZfoFQuhm0fBUe+Q0GrIY274KPP/z4MAyq4dwIxJ+5/uArv4GFH0KdPlDr/uu/johkBtHAdOBhYKcx5kNjTEOXM4nkSF1qF+ODbjVYsfs4dw9bxNEz6fSBQ1bi4wc3DYCHF0Ox+jD5ORh+C+xb5XYyEfmL1BQvHrTWHr+0Y609BjyYTq//M1DCWlsVmAZ8caWTrLVDrbUx1tqYiIiIdHppkZzl9IUE+ny1DD9fD590r0Wg3xWGhy7+BDy+ziiK6xVeHGr3hn5z4Z6xkKeYcyPwXmWY/abTtyItYpc5QzpLNILbXr3+XCKSKVhrz1prv7XW3gFUB0JxppCIiAvaVi3MsPti2HLgNHcNWcj+E+fdjuSO8OJwz3fQ+VNnJOnQpjDlhRv78EVE0lVqihc+lw/rTO5l4Z+K58UBl4+kiOKPxpwAWGuPWGsvJO8OB2ql4roikkZJSZanvl3JjsNn+PDuGkSFX6Gx5rnjsOJrqNIZQgrd+Isa4/TI6DUJek2BqNow839OEWPqi3Bqf8rXOLUfxtwDIQXhzi+cT0dEJMszxjQ2xnwELAMCgWvMUxMRb2saXYAvetVh3/Fz3DlkAbuPnHU7kjuMgcqdnJGkNe91Rn0OrgerRmsqiUgmkJrixWSc5py3GGNuAb5JPpaSJUBZY0xJY4w/0BWnq/jvjDGRl+22BzakLraIpMXHs7cxZd0Bnm8VTYPS+a980vIv4eIZqJeaWWFpVKye0x+j33wo1xIWDob3qzgjKo5uv/JzLp53GmidPwFdv4Fc+a58nohkKcaYncDjwFygirX2Lmvt9+6mEpF6pfIx6sF6nDqfQOdPFrD5wCm3I7knKBzaDYSekyEoD4zv6/TDWDNWRQwRF6WmePEsMBN4KHn7FXgmpSdZaxOAR4ApOEWJb62164wxLxlj2ief9pgxZp0xZhXwGNAj7T+CiFzLrE0HeXvqJtpXK0zvhiWvfFJigrNiSIlGEFnVe2EKVYbOI+DRZVD9Hlg5Cj6oBWN7OSucXGIt/PIkxC6Bjp84zxOR7KKqtbajtfYba63GY4tkItWK5mFMn/pYoMuQhayJPeF2JHcVrw9950CXkU4vr+97w8cNYN14SEpyO51IjmOstW5nSJOYmBi7dOlVV2gVkcvsOnKGdh/Mo0h4MOMeanD1ZdDWjoOxPZ0RDtGtMy7gqf3OKIyln0L8aSh7KzR80mmSNflZaPwsNP1nxuURkTQxxixLYeWxTE/3FSJ/t+vIGe4ZvpjjZy8y4v6Yvy+rnhMlJcGGH2HW63BoIxSoBE2fh+i26bs0vEgOltJ9RYojL4wxZZNXAllvjNl+aUvfmCKS3s7FJ9L3q2UYY1Jev33RRxBe0pnSkZFCCsGtL8MTa6HpixC3DD5r6RQuyreBxs9lbB4RERGheL5cfNevPgVDA7jv09+Ytemg25Hc5/FApY7w0ALoNAISLzjTW4fc7KzUlsU+EBbJilIzbeQz4GMgAWgKfAmM9GYoEblx//15HRv3n2Jg1+oUy3eFBp2X7FniTM+o95DzD7MbgsKh8T/g8bXQ8g2odjfcMcS9PCIiIjlcZFgQ3/atT5kCuXnwy6VMXLPP7UiZg8fHaW7efzF0HAIXTsE3XWFYU9g8VUUMES9KzTuDIGvtrzhTTHZZa/8PaOPdWCJyI35cGcfoJXvo36Q0TcoXuPbJiwZDQJjTg8Jt/sFQrx90/BgCQtxOIyJeYIwZYIwJNY4Rxpjlxphb3c4lIn+XL3cAox6sR7WoPDwyajnfLt3jdqTMw8cXqnWFR5ZCh8Fw9giMuhOGN4etv6qIIeIFvqk454IxxgNsMcY8grPcaW7vxhKR67Xz8Bn+OW4NtYqH82SLctc++fgeWP8T1O8PAfpjLSIZope1dqAx5jYgHLgX+AqY6m4sEbmSsCA/vuxdh75fLeOZsas5fzGR++qXcDtW5uHjCzW6Q5W7YNUomP0WjLwDitZzemJEVHBGZ8Sfcr5eOAUXTsOFk8nHT192/LIt/jSEFYUKbZ2ptCEF3f5JRVyXmuLFACAYZzWQl3GmjtzvzVAicn0uJCTyyDfL8fXxMKhbDXx9Uhhc9dtQ52udvt4PJyLiuNTZrjXwVfJKZOp2J5KJBfv7Mvz+GB4ZtYJ//7iOi4n26iuY5VS+/lCrB1Tr5iw/P/cd+LJDys8zPs5o04BQ54OkgBAIzgt5ijorsU14AiY86Sw7H93WKWaEl/D2TyOSKV2zeGGM8QG6WGufBk4DPTMklYhcl9cmbmRt3EmG3RdDkTxB1z75wmlY9gVUbO/8AykikjGWGWOmAiWB540xIYDWHBTJ5AJ8ffjonpo89s0KXp6wnoTEJPo2Lu12rMzHNwDqPAg17nWWVI0/nVyYCEnecv953zfw6quVWAsHN8CGn2HjzzD1BWcrVAUqtIcK7SAiWqudSI5xzeKFtTbRGNMwo8KIyPWbsm4/ny/YSc+bStCiYiqGFq4cBRdOQL2HvR9OROQPvYHqwHZr7VljTF704YhIluCXPLLziTEreW3SRhKSLA83LeN2rMzJLxCqd7uxaxgDBSs6W5Nn4egO2PiLU8yY+SrM/B/kLe0UMSq0g8I11excsrXUTBtZYYz5CfgOOHPpoLV2nNdSiUiaxB0/xzNjV1O5SCjPtYpO+QlJSbD4Y4iqDUVrez+giMgf6gMrrbVnjDHdgZrAQJcziUgq+fl4eL9LdXw9hrembCIh0TKgeVm3Y+UMeUtCg0ec7dQB2JRcyFj4Icx/H0IKO9NKKrSDYg2cfhwi2UhqfqMDgSNAs8uOWUDFC5FM4GJiEo99s4LEJMuH3WoS4OuT8pM2T4aj26HZv7wfUETkzz4GqhljqgFPAcNxlmFv7GoqEUk1Xx8P79xVHV8fD+9N30xCUhJPtiiH2tdkoJCCENPL2c4dc5Zp3fATLP/K6Wnm4w8hkRBaBEILJ29F/vw1dwFn6VeRLCLF4oW1VkM5RTKx96ZtZtmuYwzqVoMS+XOl7kmLPoLQKGe+pIhIxkqw1lpjTAfgQ2vtCGNMb7dDiUja+HgMb3aqiq/H8MGMrVxMtDzbsrwKGG4ICodqXZwt/ixs+xVil8LJvc4Wt8wZoZF44c/PMz7JBY6/FDfCS0DJRhAY5sqPI3I1KRYvjDGf4Yy0+BNrbS+vJBKRVJuz+RAfzdpGtzpFaV+tcOqetG817JwLLV7ScEIRccMpY8zzOEukNkpejt3P5Uwich08HsOrHavg62P4ZPY2EhKTeKFNBRUw3OQf/EcPjMtZC2ePwsm45KJG3B/FjZNxcGAdbJkKF88653v8oERDiG4D5VqqubtkCql55zLhsu8DgY7AXu/EEZHUOnjyPE9+u5JyBXPz77aVUv/ERR+DXy6oqRWPRcQVXYC7gV7W2v3GmGLAWy5nEpHr5PEYXu5QGV+Ph+HzdpCQZPlPu4oqYGQ2xkCufM4WWfXK51gL50/AwfWwaRJsmggTn3a2QlWhfGuIbu18r/+/4oLUTBv5/vJ9Y8w3wDyvJRKRFCUmWR4fs5LTFxL45sF6BPmncr7iqQOwdqyzDnlQHq9mFBG5kuSCxddAbWNMW+A3a+2XbucSketnjOE/7Sri6zHJBYwkXmpfGY9Hb3CzFGOc+8PiDZzt1pfh8BZnhZNNE2H2GzD7dWfqcflWTiGjeEPw9Xc7ueQQ1zNmvCxQIL2DiEjqfTRzKwu2HeHNTlUpWzAk9U9cMhwSL0Ldft4LJyJyDcaYu3BGWswCDPCBMeYf1tqxrgYTkRtijOGFNhXw9fEkTyGxvNqxigoYWV3+stDwcWc7fchp+r5pEqwYCUuGQUAolGnuTC8p01wfjolXpabnxSn+3PNiP/Cs1xKJyDUt3n6E96Zv5vbqhbkzJir1T7x4DpaOcCrl+Up7L6CIyLW9ANS21h4EMMZEANMBFS9EsjhjDM+2LI+fj9PEMyHJ8kanqviogJE95I6Amvc6W/xZ2D7LGZGxeTKsGwceXyh+k9Mjo9xtGXO/efYobJkGW6ZAwgUoUhOiakPhGhCQhg/4JEtIzbQR/V8XySSOnolnwOiVFM+Xi1c6VknbfNLV38LZI1Cvv/cCioikzHOpcJHsCOBxK4yIpC9jDE/dWh5fj7OMamKS5a3OVfH10R/zbMU/2Jk2Et0akhKdFU0uTS+Z8ryz5SvjFDLK3grF6qff9JLDW5zRH5snw+6FYJMgVwGnWLExuV2j8UBEBYiKSd5qQ/7y4NHvYVaWmpEXHYEZ1toTyft5gCbW2h+8HU5E/mCt5envVnH0TDzj7m9A7oA0zPqy1mnUWbCK0zlaRMQ9k40xU4Bvkve7ABNdzCMiXjCgeVl8fQxvTdlEQpLlvbuqqYCRXXl8oGgdZ2vxXzi6w1m5ZPNk+G0oLPzQmV5SuqlTzCjTwhnFkVqJCU6R4tKUlaPbnOMFq0Cjp6BcK2ekhcfjjMSIW+YsFRu7BNb/CMu/cM4PCHXOi6rtFDSKxKQth7guNe9+/mOtHX9px1p73BjzH0DFC5EMNGLeDmZsPMh/21eicpE0rru9fSYc2gC3f6zu0CLiKmvtP4wxnYCbkg8Nvfw+Q0Syj4eblsHXY3ht0kYSEpMY1K0GfipgZH95S0Ldvs524TTsmO0UHrZMc4oJGGd6x6VRGZHV/n5/eu44bJ3+x/POHwcffyjRCOo95ExLyVPs768dnBfKtnA2gKQkp9hxqZgRtxTmvQc20Xk8T3GnmFGyEVTtAn5BXv1PIzfGWGuvfYIxq621Vf9ybI21topXk11FTEyMXbp0qRsvLeKalXuOc+cnC2gWXYBPutdK+/JjIzvDvlXwxFrwDfBOSBHJcYwxy6y1MW7nuBG6rxDxvuFzt/PKLxuoGBnKw03L0LJyIfXByImshf2rYfMUZ4tbBlgIiXSKDWWaw4lYZ3TF7oWQlADB+Z1CRbmWzsiN9OhjEX/WuS++VMyIXQon4yBXBNR/GGJ6Q2Dojb+OpFlK9xWpKV58ChwHBicfehjIa63tkV4h00I3GZLTnDx/kTaD5pKUBBMfa0RYsF/aLnBoEwyuA01fgMbPeCekiORIaSleXKEB+O8PAdZa68qdou4rRDLGhNV7eXfqZrYfPkPJ/Lnoe3MpOtYsQoBvKpd7l+zn9CHYOs0pZGybARdOOscjKkD5llC+NRSp5UxL8bad82Hu206OwDBnZb66/ZyRHJJh0qN4kQv4F9Ac56ZjGvA/a+2Z9AyaWrrJkJzEWssjo1Ywed1+vu1bn1rFw9N+kZ8fh5Wj4Mn1kCt/+ocUkRxLIy9EJC0SkyxT1u3n41nbWBN3goKhATzQsBTd6hZLWy8vyX4SLzojMXIXdKaduCVuGcx912n86ZcLYnpCg0chpJB7mXKQGy5eZDa6yZCcYsO+k7zyy3rmbz3Cc62i6df4OpabOnsU3q0IVe+E9h+kf0gRydFUvBCR62GtZd7Ww3w8axsLth0hLMiP++sXp8dNJcmbK51WpBC5EQc3OEWMtWPB4wc1usNNAyC8uNvJsrWU7itS7JhjjJmWvMLIpf3w5C7hIuIFh09f4Plxa2gzaC7r9p7kv+0r0adRqeu72NJPIeGclkcVERGRTMMYQ6OyEYx6sB7j+zegXqm8DJqxlQav/8r//bSOuOPn3I4oOV2BCtBpGDy6DKp1heVfwgc1YfxDcGiz2+lyrNSMz8pvrT1+acdae8wYU8CLmURypAsJiXw2fycfztjK+YuJ9GhQkgG3lE17j4tLju1ylqcq3cz5C1hEREQkk6lRLJwh98aw9eApPpm9nZGLdjFy0S46VC/CQ01KUaZAOjRoFLleeUtB+0HQ+FlY8AEs+xxWfQMV20OjpyGyaoqXyJTiz4DxAb9At5OkSWqKF0nGmGLW2t0AxpjiXLnhlohcB2stk9fu57VJG9l99Cy3RBfgn20qUDoi9/VfdONE+KGf09W56QvpF1ZERETEC8oUCOHtO6vxRItyDJuzndFLdvP98lhurViQ/k3LUL1onpQvIuItYUWg1etw89Ow6CP4bZiz7GvZW50VSkIiwePrNBf1+DpTTf607/vHflpXDUxP+1bD0hGw+jvw9Yc7hv2xrGwWkJqGnS2BocBsnI7gjYA+1lpXpo5obqpkJ2vjTvDShPX8tuMo5QuG8GLbCjQqG3H9F0y8CL++BAsGOWtm3/m5UzEWEfEC9bwQEW85eiaez+fv4IuFuzhx7iItKhbk3buqERJ4nSNSRdLTueOwZBgs/AjOHU3bc40P+CQXN3z8oFh9qNAeyreCIC8U6S6eg3XjYckIZ2lY30Co3MlZLvbAWmj0FDT5J/i43zQ3XRp2GmPyA/WSdxdZaw+nU740002GZAcHT57nrSmbGLs8lvBgf55sUY6utYvi65NiG5qrOxEHY3vCnsVQ+wG49X9ZbiiYiGQtKl6IiLedvpDAlwt38u7UzZQpkJtPe9SmcJ4gt2OJOOLPwPbZTo+5pERISvhjS0z48/5fH09KgAunYOuvcDLWGa1RqjFU7ADl20CufDeW7fAWp//dylFw/jjkLwcxvZweHkHhTlFj0jNOP48SjaDTcNdXVUmv4kU4UBb4/Z2QtXZOuiRMI91kSFZ2/mIiI+btYPDMrVxMTKLnTSV5uGkZwoJu8FOELdNh3IOQGA/tBkKVzukTWETkGlS8EJGMMm/LYR4auYzgAB9G3F+bykXC3I4kkj6shbjlsP4H2PATHNvpjM4o0dDprRHdDkIKpu5aCfGw6RdnlMXOuU5BpEI7p2hRouGVp6ysHAUTnoSAEOg8AkrenK4/XlrccPHCGPMAMACIAlbijMBYaK1tlp5BU0s3GZIVWWuZsHofr0/aSNzxc9xWqSDPt6pAify5buzCiQkw61WY+w4UqAR3fQH5y6ZPaBGRFKh4ISIZadP+U/T87DeOn7vI4Ltr0jRaawhINmMt7F/t9NNY/yMc2QoYZ2pJxQ5OISKsyN+fd3w3LPvCGUVx5iCEFYOYHlDjXsidij8nB9bDd/c7r9f0n9DwKfDcwIjw65QexYs1QG2c6SLVjTHRwKvW2jvSN2rq6CZDspLzFxOZs/kQQ+dsZ+muY1SIDOVfbSvQoHT+G7/4qf0wtjfsmuf8xdT6LfDTMEoRyTgqXohIRjt48jy9vljC+uTl5O+tX8LtSCLeYS0c3OCMxlj/Ixxc7xyPqu30yKjQ1pkasmQEbJnqjKooe5szyqLMLU5z0LS4cBp+HgBrx0KZ5tBx6I1PXUmj9CheLLHW1jbGrATqWmsvGGPWWWsrpeLFWwIDAR9guLX29auc1wkYC9S21l7zDkI3GZLZnb+YyOzNh5i4Zh+/bjjI6QsJRIQE8PSt5ehcqyg+nnToMLx9Fnz/gDPPrs27UL3bjV9TRCSNVLwQETecuZDAgNErmL7hIA82KsnzrSrgSY/7K5HM7PAWp4ix4Sen2eYluQtCzfug5v2Qp+iNvYa1Tp+Myc9Brgin+X/ROjd2zTRI6b4iNS1FY40xeYAfgGnGmGPArlS8sA8wGGgBxAJLjDE/WWvX/+W8EJxpKYtTkUUkUzp/MZFZmy4VLA5wJj6RPMF+tKkSSevKETQo7MEvpMCNL42UlAhz3oJZrztNd+6fAAWi0+eHEBEREckCcgX4MuTeGF6esJ5hc3ew5+g53utSnSD/NH7SLJKV5C/rLNV689NwdAdsnuI02Ixu46xakh6Mgdq9oUgtZxrJZ62gxUtQr7+7S7wmS7F4Ya3tmPzt/xljZgJhwORUXLsOsNVaux3AGDMa6ACs/8t5LwNvAP9IbWiRzOBcfCKzNh1k4tr9zEguWIQH+9GuWmFuL22ISVyB77aRMH4WXDgBwfmhUBVni6zmfM1XJvVDuk4fdJpybp8FVbtC23fB/wZ7ZoiIiIhkQT4ew/+1r0SxvMG8/Mt6ug1bxPD7Y8ifO8DtaCLel7ck1OvnvesXrg59ZsOPD8OUf8LuhdBhMAS62yg3TYu5Wmtnp+H0IsCey/ZjgbqXn2CMqQkUtdb+Yoy5avHCGNMH6ANQrFixNEQQSV/n4hOZuekgv6zZx8yNBzkbn0jeXP50rFqAuwrFUenMXHy2z4A1a50nhERCxXYQEQ2HNsL+NbD4E2dVEADfIChY6c9FjQIVwT/4zy+8cz6M7eUsc9RukDM0LBNUP0VERETc1KthSYqEBzFg9Ao6fjSfz3rUoUyB3G7HEsn6gvJAl5Gw6COY9m8YcjPc9aXzfsUlaSpepCdjjAd4F+iR0rnW2qHAUHDmpno3mcjfLd99jBHzdjBjw0HOXUzOINriAAAgAElEQVQkXy5/7q/kQ6fQLZQ6vhDP5jmw9hR4fJ1uwM3/6zS6KVjp70WGhHg4vNnpJLx/jbOtHQfLPnMeNx5nREahqk5BI/60s5pIeEnoPtY5JiKSwxhjPgXaAgettZWv8HgT4EdgR/KhcdbalzIuoYi45bZKhRjTpz69v1jCHR/NZ+h9MdQrlbGNBkWyJWOg/sNOk9DvesDwFtDqdajV05UPUr1ZvIgDLu8YEpV87JIQoDIwyzg/eCHgJ2NM+5SadopkpGnrD/Dw18vJG2D5R9n9tAxYS+Sh+ZgNG50TQqOgSico08JZFzkw9NoX9PWHQpWd7RJrnSWOLhU09q2G3Yucbr8Ale6AdgNTvraISPb1OfAh8OU1zplrrW2bMXFEJDOpVjQP4/vfRM/Pl3DviMW82bkqHWtEuR1LJHsoWgf6zoXxfWDCE7BrIbR9DwIydpSTN4sXS4CyxpiSOEWLrsDdlx601p4Afl8v0hgzC3hahQvJTH5cGceT367kjbDxdEr4BbP9LPj4Q/EGULO7U7CIKH/jlUdjILy4s1Vo98fxs0fh9AFn2ommiYhIDmatnWOMKeF2DhHJvIrmDeb7hxrw0MhlPDFmFbuPnOOxW8pgdA8lcuNy5YO7v4N578CKkX9Mg89AXiteWGsTjDGPAFNwlkr91Fq7zhjzErDUWvuTt15bJD2M/m03z49fzaC842h35ntn9EPVLlCiYcZVGYPzOpuIiKRGfWPMKmAvzgci69wOJCIZKyzIj8971uH5cWt4b/pmdh89y2t3VMHf1+N2NJGsz+OBm/8B9R8Bv6AMf3mv9ryw1k4EJv7l2L+vcm4Tb2YRSYvhc7fzyi8beL/gJNqd+B5qPwit39LoBxGRzGs5UNxae9oY0xpnifeyVzpRjcBFsjd/Xw9v31mV4vmCeXfaZvYcPcvzraOpUSzc7Wgi2YMLhQsAlSBFLmOtZeD0LbzyywbeKzKD20+MhBrdodWbKlyIiGRi1tqT1trTyd9PBPyMMfmvcu5Qa22MtTYmIiIiQ3OKSMYwxvDYLWV5v0t1Nh04RcePFtBlyEJmbjyIter/L5IVubbaiEhmY63ltUkbGTpnO+8XX8jtB4ZD5c7O0qQe1flERDIzY0wh4IC11hpj6uB8QHPE5Vgi4rLbaxShecWCjP5tNyPm7aDn50uILhRC38alaFu1MH4+uscTySpUvBABkpIsL/64llGLdzOw7Eo67PkAottCx0/A4+N2PBGRHM8Y8w3QBMhvjIkF/gP4AVhrPwE6Aw8ZYxKAc0BXq49XRQTIHeDLA41KcV/9Evy8ai9D5mzjiTGreHvKZno3LEnXOkUJ9tfbIpHMzmS1f9djYmLs0qVakETST0JiEk9/t4ofVu7lw0qbaLPtJUzZFtDla2dZUxERuSJjzDJrbYzbOW6E7itEcp6kJMuszQf5ZNZ2ftt5lDzBftxXvwT31y9OvtwBbscTybFSuq9QiVFytAsJiTw6agVT1x/gk5q7abnhZSjZCO76UoULERERkWzI4zE0iy5Is+iCLNt1jE9mb2PQr1sYOmcbd8UU5YGGpSiWL9jtmCLyFypeSI51Nj6Bvl8tY+6Wwwyvd4jmq1+EqDrQbbRrHXRFREREJOPUKh7OsPti2HrwNEPnbOOb33YzctEu2lQtTN+bS1G5SJjbEUUkmTrUSI508vxF7v/0N+ZvPcwXN5+m+Zp/QKGqcM934J/L7XgiIiIikoHKFMjNm52rMfeZZjzYqBQzNx6k7QfzuHfEYjYfOOV2PBFBxQvJgY6eieeeYYtZsfs4I1sk0Hj5AMhfDrp/D4GhbscTEREREZcUCgvk+dYVmP9cM55tGc3auBO0HTSPj2dtIzEpa/UKFMluVLyQHOXgyfN0HbqQTQdOMbqNLw0W9Yc8xeDeHyA4r9vxRERERCQTCAvy46EmpZn6RGOaRkfwxuSNdP5kAdsOnXY7mkiOpeKF5Bixx85y55CFxB47x3cdgomZ+wDkioD7foTcEW7HExEREZFMJiIkgE+612Jg1+psP3SG1gPnMnzudo3CEHGBiheSI6zbe4LOHy/k2Jl4xnYKp9qMHhAQCvf/BKGRbscTERERkUzKGEOH6kWY9sTNNCqbn1d+2UDXoQvZefiM29FEchQVLyTb+3XDAe78ZCHGwPguBak4tTv4+DsjLvIUczueiIiIiGQBBUIDGXZfDO/cWY2N+0/RcuAcPp+/gySNwhDJECpeSLb22fwdPPjlUkpH5Obn7kUpPelusEnOiIt8pd2OJyIiIiJZiDGGTrWimPZEY+qVysf//byeu4cvYs/Rs25HE8n2VLyQbCkhMYn//LiW//68nvtLnWZc6QnkH3UbxJ+B+36AiPJuRxQRERGRLKpQWCCf9ajNG52qsDbuJLe9P4eRi3ZhrUZhiHiLr9sBRNLbqfMXeW7kbPLt+ImFeRcRGbcJ9vlB+VbQ+FkoVNntiCIiIiKSxRlj6FK7GA3LRvDs2NW8+MNaJq/dzxudq1IkT5Db8USyHRUvJPtITODw6klsmPgx78YvJsAvAUKrws1vQJU7IVc+txOKiIiISDZTJE8QX/Wuw6jfdvO/XzZw23tz+FfbCtwVUxRjjNvxRLINFS8k6zu0GVZ+zcUVo8h/9iCVbAiHorsT1fQBKFTF7XQiIiIiks0ZY7inbnFuLhvBP8au4tnv1zBxzX7+274SJfLncjueSLag4oVkTedPwNpxsPJriF1CkvFhblJ1pvn3pGfPvpQrrFEWIiIiIpKxiuYNZtQD9fhq0S5en7SRpu/MolXlQvS5uTTVi+ZxO55IlqbihWQtuxbA0k9hw8+QcB4bEc2i0o/z2PpyFIkqwbD7YogICXA7pYiIiIjkUB6P4f4GJWhVuRCfL9jJV4t2MXHNfuqUzEu/xqVoUq4AHo+mk4iklYoXknVsmQ5fd4LAMKh+DwnV7uZfv/nxzZJY2lSJ5J27qhHo5+N2ShERERERCoQG8kzLaPo3LcPo33YzYt4Oen2+lHIFc/Ngo1J0qF4Ef18t/iiSWvrTIlnD+RPw82MQEQ1PbuTELW/QY0oi3yyJ5eGmpfmgWw0VLkREREQk08kd4MsDjUox55mmvNelGh5j+MfY1dz85kyGzN7GyfMX3Y4okiVo5IVkDVP/Baf2wV1fsec09Pp8ATuPnOGtzlW5M6ao2+lERERERK7Jz8dDxxpR3F69CHO2HGbI7G28NmkjH87Yyt11i9HzppIUCgt0O6ZIpqXihWR+22bA8i/gpgEsSyxFn8HzSUiyfNmrLvVLqzGniIiIiGQdxhgal4ugcbkI1sSeYMicbQybu51P5++gQ/Ui9Lm5FOUKhrgdUyTTUfFCMrcLp0j88VHO5CrB47taMHvmQqLCg/i0R21KR+R2O52IiIiIyHWrEhXGh3fXZPeRs4yYt50xS/cwdlkszaIL8NgtZbVCichl1PNCMqUjpy8wavFupg/qhzkRR4+jPdh6NIE+N5difP+bVLgQERERkWyjWL5g/tuhMgueu4Unmpdj5Z7j3D54Ps+PW82xM/FuxxPJFDTyQjKNI6cvMHndfiau2cei7Uepw1q+8Z/AksJ381LbXlQqHIoxWlZKRERERLKnvLn8GdC8LL0blWTQr1sYMW8Hk9fu59mW0dwVU1RLrEqOpuKFuOrw6QtMXnupYHGEJAsl8+fi0YaF6L/hGaxfKWr3fAf8g92OKiIiIiKSIXIH+PLP1hXoVDOKf/2wlufGrWHM0j283KEylYuEuR1PxBUqXkiGO3QqeYTF6n0s3uEULErlz0X/JmVoXSWSCpEhmEnPwqlY6DlRhQsRERERyZHKFwphTN96jF8Rx6sTN9D+w3ncV78ET95ajtBAP7fjiWQoFS8kwyQmWQZO38zgWdtITLKUisjFw02dgkV0oZA/poTsnA+/DYG6/aB4A3dDi4iIiIi4yBjDHTWjuKVCQd6ZuokvFu5kwup9vNAmmturF9G0askxVLyQDHHw1HkGfLOShduP0LFGEfo2LkX5giF//8s2/iz8+DCEl4Bb/u1KVhERERGRzCYsyI+XOlTmzlpFefHHtTwxZhWjf9vDy7dX1tKqkiNotRHxukXbj9Bm0DyW7z7Gm52r8l6X6kQXukrzzRmvwLEd0P5D8M+V8WFFRERERDKxKlFhjH+oAa92rMLG/adoPXAur03cwJkLCW5HE/EqrxYvjDEtjTGbjDFbjTHPXeHxfsaYNcaYlcaYecaYit7MIxkrKckyeOZW7h62iJAAX354+Cbuiil69SfsXgSLPoLaD0DJRhkXVEREREQkC/F4DHfXLcaMpxpzR80iDJmznebvzmbSmn1Ya92OJ+IVXiteGGN8gMFAK6Ai0O0KxYlR1toq1trqwJvAu97KIxnr2Jl4en2xhLembKJ1lUh+erQhFSJDr/6Ei+ec6SJhRaH5fzMuqIiIiIhIFpUvdwBvdq7G9w/VJyzIj4e+Xs79ny1h5+EzbkcTSXfeHHlRB9hqrd1urY0HRgMdLj/BWnvyst1cgMqE2cDy3cdoM2guC7Ye4eUOlfigWw1yB6TQXmXmq3BkK3T4AAJyZ0xQEREREZFsoFbxvEx4tCH/bluR5buO0XLgHEYu2qVRGJKteLN4UQTYc9l+bPKxPzHGPGyM2YYz8uKxK13IGNPHGLPUGLP00KFDXgkrN85ay4h5O7jrk4V4PIaxD9Xn3volUu6AHLsUFn4ItXpAqSYZkFREREREJHvx9fHQq2FJfn2qMbVL5OXFH9bS56tlHD0T73Y0kXThesNOa+1ga21p4FngxaucM9RaG2OtjYmIiMjYgJIqJ89f5KGRy3l5wnqaRhfgl0cbUTUqT8pPvHgefugPIYWhxcveDyoiIiIiko0VDA3ki551eLFNBWZvOkTL9+cwb8tht2OJ3DBvFi/igMu7M0YlH7ua0cDtXswjXrI27gRtB81j2oYDvNC6AkPvrUVYsF/qnjz7DTi8CdoPhMBr9MQQEREREZFU8XgMDzQqxfiHGxAS6Ev3EYt5deIG4hOS3I4mct28WbxYApQ1xpQ0xvgDXYGfLj/BGFP2st02wBYv5pF0Zq3l68W7uOPjBcQnJDGmTz0evLlUytNELolbDvMHQo3uUKa5d8OKiIiIiOQwlQqHMeHRRtxTtxhD52yn40fz2XrwtNuxRK6L14oX1toE4BFgCrAB+NZau84Y85Ixpn3yaY8YY9YZY1YCTwL3eyuPpK8zFxJ4fMxKXhi/lnql8vHLYw2JKZE39RdIuOCsLpK7ANz6P+8FFRERERHJwYL8ffhfxyoMvbcWe4+fo+0Hcxm1eLeaeUqWk8ISEDfGWjsRmPiXY/++7PsB3nx98Y61cScYMHoFOw6f4akW5Xi4aRk8nlSOtrhkzttwcD3c/S0EpaI3hoiIiIiIXLdbKxWiWtE8PPXtKv45fg2zNx/k9TuqEp7L3+1oIqniesNOyTr2HD3L46NX0PaDeZw4l8DI3nV59JayaS9c7FsFc9+Bat2g3G3eCSsiIiIiIn9SMDSQL3vV4YXWFZix8SAtB85h/lY185SswasjLyR7OH42ng9nbOXLhbswBh5rVIS+ZY6Ry2c97LiOC07+J+TKD7e9mu5ZRURERETk6jwew4M3l6J+6XwMGL2C7iMW06dRKZ66tTz+vvpsWzIvFS/kqs5fTOSLBTsZPHMrpy4kcGetKJ66KZyCP9wNS9bc2MW7joLgNPTIEBERERGRdFO5iNPM8+Vf1jNkznYWbDvC+12rUzoit9vRRK5IxQv5m8Qkyw8r4nhn6ib2njhP0/IRPNsqmuigk/BlBzi5FzoOgbCo63uBXBEQUT59Q4uIiIiISJoE+fvwascqNC4XwbPfr6btoHk81yqaFhULEhkWmPpVBEUygIoX8jtrLXO2HOa1iRvYuP8UVaPCePuuajQonR+ObINPO8D5E3DveChWz+24IiIiIiKSDm6rVIjqRfPw5Lcr+c9P6/jPT+sID/ajUuEwKhYOpVLyVjJ/bnzS2u9OJJ2oeCGAs4LIa5M2MH/rEYrmDeKDbjVoUyXSacZ5YD18dTskJcD9P0Ph6m7HFRERERGRdFQwNJCvetVlxZ7jrNt7gnVxJ1m37wSfz99JfGISAEF+PkRHhlCpcCgVI8OoVDiU8oVCCPTzcTm95AQqXuRwe46e5Z2pm/hh5V7Cg/34T7uK3FO3+B/NeuKWwchO4BsIPSdpuoeIiIiISDbl8RhqFQ+nVvHw349dTExi68HTrNt70ilq7D3Jjyv2MnLRbgB8PIYyEbmdgkbhUJqUL0CZAuqbIelPxYsc6q8riPRvUpp+TUoTGuj3x0k758GoLhCcD+7/CcJLuJZXREREREQynp+PhwqRoVSIDKVzLafnXVKSJfbYud+LGev2nmDe1sOMWxHHK79soGyB3LSqXIiWlSOpEBmi3hmSLlS8yIF2Hj7DPcMXs+/EOTrXiuKJFuWIDAv680mbp8K390Ke4nDfDxBa2J2wIiIigDHmU6AtcNBaW/kKjxtgINAaOAv0sNYuz9iUIiI5g8djKJYvmGL5gmlVJfL343uPn2Pquv1MWrufD2duZdCMrRTPF0zLSoVoWdnpq6FChlwvFS9ymE37T9F9xGISkyzj+t9E9aJ5/n7SuvHw/QNQsBJ0Hw+58mV8UBERkT/7HPgQ+PIqj7cCyiZvdYGPk7+KiEgGKZwniB43laTHTSU5fPoC09YfYNLa/YyYt4Mhc7YTGRbIbZUK0apyIWJK5FXzT0kTFS9ykNWxx7nv098I8PUwqk89yhYM+ftJy7+Cnx+DonXh7jEQGJbxQUVERP7CWjvHGFPiGqd0AL601lpgkTEmjzEm0lq7L0MCiojIn+TPHUC3OsXoVqcYJ85eZPqGA0xet59Rv+3m8wU7yZ/bn1srFaJlpULUL50PPx+P25Elk1PxIof4bcdRen2+hDzBfox6oB7F8gX//aRFH8Pk56B0M+jyNfhf4RwREZHMqQiw57L92ORjfyteGGP6AH0AihUrliHhRERysrBgPzrViqJTrSjOXEhg5qaDTFq7nx9WxDFq8W7CgvxoXqEg99UvTrUrjQwXQcWLHGH25kP0/WophfME8fUDdf/e38JamPM2zHwFottC50/BN8CdsCIiIl5mrR0KDAWIiYmxLscREclRcgX40rZqYdpWLcz5i4nM3XKYSWv3MXX9fr5fHkvHGkV4pmX5v79nkRxPxYtsbvLafTz6zQrKFAjhq951yJ/7L0UJa2Hav2DBB1C1K3QYDD76tRARkSwnDih62X5U8jEREcmkAv18aFGxIC0qFuT0hQQ+mrmV4fN2MGntPvreXJq+jUsR7K/3JuLQxKJsbNzyWB4etYLKRcIY/WC9vxcukhJhwhNO4aL2A3D7xypciIhIVvUTcJ9x1ANOqN+FiEjWkTvAl2daRvPrk425pUJBBv66hWZvz2b8iliSkjRITlS8yLa+WrSLJ79dRd2SeRnZuy5hwX5/PiHxIozvC8s+g4ZPQOu3waNfBxERyZyMMd8AC4HyxphYY0xvY0w/Y0y/5FMmAtuBrcAwoL9LUUVE5AYUzRvM4Ltr8l2/+hQIDeCJMavo+NF8lu066nY0cZk+Zs+GhszexmuTNnJLdAEG31OTQD+fP59grVO4WPs93PJvaPSUO0FFRERSyVrbLYXHLfBwBsUREREvq10iLz/0v4nxK+J4c8pGOn28kLZVI3muVTRR4VpYICdS8SIbsdby3rTNDJqxlbZVI3mvS/UrLzm0eIhTuGj2LxUuREREREQkU/J4DJ1qRdGqSiE+mb2doXO2MXX9AR5sVJKHmpQhd4DezuYkmieQTVhreXnCBgbN2MpdMVEM7FrjyoWL2GUw9UUo10qFCxERERERyfSC/X15skU5ZjzVhNaVCzF45jaavj2Lb5fuUT+MHETFi2wgMcny/Lg1fDp/Bz1vKsHrd1TFx2P+fuLZo/BdDwiJhI4fg7nCOSIiIiIiIplQ4TxBvN+1BuP7NyAqPIhnxq6m3YfzWLT9iNvRJANonE0WdzExiSe/XcXPq/byaLMyPNmiHOZKRQlr4Yf+cGof9JoCQeEZH1ZEREREROQG1SgWzriHGvDz6n28PnEDXYcuolbxcGoVD6daVB6qFQ2jSJ6gK78vkixLxYss7PzFRB4ZtZzpGw7yXKto+jUuffWTF3wAmydByzcgqlbGhRQREREREUlnxhjaVyvMrRUL8un8HUxbf4DPF+wkPiEJgPy5/ZMLGclbVBh5gv1dTi03QsWLLOrYmXj6f72chduP8PLtlbm3XvGrn7x7EUz/P6jQHur2zbCMIiIiIiIi3hTo50P/JmXo36QM8QlJbNp/ipWxx1m1x9lmbDqITW6LUSJfcHIhwyloVCoc+veVGSXTUvEiC1oTe4J+I5dx6NQF3utSjY41oq5+8pnD8F1PyFMMOnyoPhciki4uXrxIbGws58+fdzuKZIDAwECioqLw8/NzO4qIiMhV+ft6qBIVRpWosN8/3D11/iJr4k6was8JNu09Tq18CRT0HOVo3DHmxYGfj8Hf14dcAT5XXvBA0t313leoeOG2w1shbynwpO4PyrdL9vDij2vJn8uf7/rVp1rRPFc/OSkJxvWBs0eg91QIDEun0CKS08XGxhISEkKJEiU0nzSbs9Zy5MgRYmNjKVmypNtxRERE0iQk0I8GpfPToHR+duzYQUhICKF5wjl3Men/2bvv+Kiq9I/jnycz6aEmoXdUQOmEoogdRcVeKDZQwLK6/ly769p3ZVcsa19UUFEEVkRdRSnKrg3poYhI7y0ECSQhbXJ+f8wQIySQQJKZJN/365VXZu69c+4zh3llHp57zrnsz8kjM8dHRo4P5xyRkV4S4iKpEeVVflNOjiWvUPEimOa9CZ/fDa3PgsvfgNiEYg/NzvPx2KfL+WDuRnofF8+LA7sQHxd5+Pa/exbWfAUXPgeNOpdx8CJSnWVlZalwUU2YGfHx8aSkpAQ7FBERkWNSOH+J8HqoFe2/8p/ry+fXjBxSM3JYn5pBhDeM+NhI6sSG4y3hRWYpmWPJK1S8CJb138EX90ODjrD+e3i9D1w5BpqffMihW/bs57b3FrB4cxq3ndGau89tU/StUAtb9y3M+hu0vxKSbiynNyEi1ZkKF9WH/q1FRKSqKOo7LdwTRr2aUSTUiGTv/lxS03PYlrafHXuzqBMTTnxcpNbGKENHm1eoeBEMv26ASdf7p4sM+cz//N83wNsXwtl/gVPuLJhG8v3qXdzxwSJy8vJ5/dpu9Gvf4Mjtp++EyTf527/oBa1zISIiIiIicgRhZtSOiaB2TAT7c/LYlZ7D7sxcUjNyiIv0Eh8XSU1NKQkaFS8qWk4GTLgGfHkw8AP/OhQNO8KI/8Gnd/jvCrJhNu7S13ht3q+MmvYLrRPj+Nd13WiVGHfk9vN9/sJFVhpcNwUia5T7WxIREREREalKoiO8NK3rpaEvn92BKSUbUjOI8IQRHxdBnZgIvFrgs0KptyuSc/DxbbDzJ/8UkYTjftsXVROuehvOfwa35mt2P9+LGdM+44IODfn4D71LVrgA+N/fYd03cMEoqH9SubwNEZFg27NnD6+++mqpX3fBBRewZ8+ewx7zyCOPMHPmzKMNTURERKoQb2BKSdsGNWheN4Zwbxjb0rJYsX0fm3/NJCvXV+K2Klv+EhdXwv+DVhAVLyrSt6Ng+cdwzuNw/DmH7jdjZYtB3Br5NJm5jslRT/JSy9nERpRwftWar+F//4BOg6HLtWUbu4hICCnuyz8vL++wr5s6dSq1ax/mLk3AE088wTnnFPE3uhI40vsXERGRo2Nm1IqJoHViHMfXq0Ht6HD2ZOaycsc+1uxM59fMHPLz3WHbUP5ybMp12oiZ9QP+CXiAN51zIw/a/ydgGJAHpAA3Ouc2lGdMQbNiKnz9FHS4Gk65o8hD/rN4K/d9uITYyJbsGDidpskPw7SHYMMPcMkrEH2YD+zebTB5OCS2gQtHaZ0LEakwj//nJ5Zv3VumbZ7YqCaPXlT86LEHHniANWvW0LlzZ8LDw4mKiqJOnTqsWLGClStXcumll7Jp0yaysrK48847GTFiBAAtWrRg/vz5pKenc/7553Pqqafyww8/0LhxYz755BOio6MZMmQI/fv358orr6RFixbccMMN/Oc//yE3N5d///vftG3blpSUFAYPHszWrVs5+eSTmTFjBgsWLCAhoei7RhUXz5dffslDDz2Ez+cjISGBr776ivT0dO644w7mz5+PmfHoo49yxRVXEBcXR3p6OgAffvghn332GW+//TZDhgwhKiqKRYsW0bt3bwYOHMidd95JVlYW0dHRjB07ljZt2uDz+bj//vv58ssvCQsLY/jw4Zx00km8+OKLfPzxxwDMmDGDV199lSlTppTlP6eIiEjIKYv8xQF5vnzyfI5852hdL477+7WlbmxEkQt8Vrb8peB9Osd9993HF198gZnx8MMPM2DAALZt28aAAQPYu3cveXl5vPbaa5xyyincdNNNBXnMjTfeyF133XVM/XxAuRUvzMwDvAL0BTYD88zsU+fc8kKHLQKSnHOZZnYr8A9gQHnFVB7enb2eUdN+IaFGJC3jY2keH0vLhJjA71ga1Y7Gs2sFfDQcGnWBi188pLCQ68tn5BcreOu7dXRrXodXr+lK/ZpR0PZ9+PFVmPEI/Os0uPodfxsH8+XBhzdCbiZc/S5ExFbMmxcRCZKRI0eybNkykpOT+e9//8uFF17IsmXLCu4XPmbMGOrWrcv+/fvp3r07V1xxBfHx8b9rY9WqVXzwwQe88cYbXH311UyePJlrrz101FpCQgILFy7k1VdfZdSoUbz55ps8/vjjnHXWWTz44IN8+eWXvPXWW4eNt6h48vPzGT58ON988w0tW7Zk9+7dADz55JPUqlWLpUuXAvDrr19zkaQAACAASURBVL8esT82b97MDz/8gMfjYe/evXz77bd4vV5mzpzJQw89xOTJkxk9ejTr168nOTkZr9fL7t27qVOnDrfddhspKSkkJiYyduxYbrxRd6gSEREpCcN/p5JwD/jyHeEeIzU9h13p2cRGeomPjaBmdDhhgf//Vbb85YCPPvqI5ORkFi9ezK5du+jevTunnXYa48eP57zzzuPPf/4zPp+PzMxMkpOT2bJlC8uWLQM44nSX0ijPkRc9gNXOubUAZjYBuAQoKF4452YVOv5HoFLNdfh08VYe+eQnureoQ3xsJOtTM/h+zS6ycvMLjknwZPBJxF+Is3DGxD9G/PwdtIiPpUV8LI1qR7E7M4fb31/E3PW7GXJKCx66oB0R3sBsHjM4+Q/QpDv8eyi8dS6c9zfoPuz3BZBZT8HGH+Cy0f6RFyIiFehwIyQqSo8ePQq++AFefPHFgtEDmzZtYtWqVYd8+bds2ZLOnTsD0K1bN9avX19k25dffnnBMR999BEA3333XUH7/fr1o06dOoeNr6h4UlJSOO200wrirlu3LgAzZ85kwoQJBa89UtsAV111FR6P/wpPWloaN9xwA6tWrcLMyM3NLWj3lltuwev1/u581113He+99x5Dhw5l9uzZvPvuu0c8n4iISGVXXvlLri+fXzNz2J2Rw8bdmXjDwqgTG07dmIhDjg31/OWA7777jkGDBuHxeKhfvz6nn3468+bNo3v37tx4443k5uZy6aWX0rlzZ1q1asXatWu54447uPDCCzn33HNLdI6SKM/iRWNgU6Hnm4Gehzn+JuCLonaY2QhgBECzZs3KKr5j8sPqXdw9KZkeLery7k09CoYFOefYsTeb9akZbNiZRo8fbqb+3l3cG/s3vkzOYv+8nwraCPcY4Z4w8p3j+QGduKxLk6JP1rQH3PItTLkZpt7jn0Zy0T/9i3yunA7fPQ9db4BOlWrQiohImYmN/W3E2X//+19mzpzJ7NmziYmJ4YwzziArK+uQ10RGRhY89ng87N+/v8i2Dxzn8XiOak2JksZzJIVvy3bw6wu//7/85S+ceeaZTJkyhfXr13PGGWcctt2hQ4dy0UUXERUVxVVXXVVQ3BAREZHSC/eEUa9GFIlxkaRn57E7I4dd+3JI2ZfNntQMfPn+KSYQ2vlLSZx22ml88803fP755wwZMoQ//elPXH/99SxevJhp06bx+uuvM2nSJMaMGVMm5wuJBTvN7FogCXimqP3OudHOuSTnXFJiYmLFBleEn7amMWLcAlomxPLG9Um/m89kZjSoFUWvVvEMSHuLlnvn4r34eZ6/52aWP3Eecx46m4kjevH3KzowrE8rLurYiCm39S6+cHFATF0YNBHOeQyWfwKjz4CV02DKCKjfAc7/e3m+ZRGRkFKjRg327dtX5L60tDTq1KlDTEwMK1as4Mcffyzz8/fu3ZtJkyYBMH369MNO7Sgunl69evHNN9+wbt06gIJpI3379uWVV14peP2BtuvXr8/PP/9Mfn7+YdekSEtLo3HjxgC8/fbbBdv79u3Lv/71r4IE5sD5GjVqRKNGjXjqqacYOnRoqfpBREREimZm1IgKp3l8LG0b1qBBzSjCo2JJ27uXFdv2kZqeXVDEgNDLXwrr06cPEydOxOfzkZKSwjfffEOPHj3YsGED9evXZ/jw4QwbNoyFCxeya9cu8vPzueKKK3jqqadYuHBhmcVfnpdXtgBNCz1vEtj2O2Z2DvBn4HTnXHY5xlMmNu3OZMjYedSI8vLOjT2oFRNe9IHJH8Dsl6HHzdD1esD/Aa5fM4r6NaPo2Sq+6NcdTlgYnHoXNOnhX+Ni/NUQUcN/i9Xw6KN/UyIilUx8fDy9e/emffv2REdHU79+/YJ9/fr14/XXX6ddu3a0adOGXr16lfn5H330UQYNGsS4ceM4+eSTadCgATVq1Cjy2OLiSUxMZPTo0Vx++eXk5+dTr149ZsyYwcMPP8wf/vAH2rdvj8fj4dFHH+Xyyy9n5MiR9O/fn8TERJKSkgoW7zzYfffdxw033MBTTz3FhRdeWLB92LBhrFy5ko4dOxIeHs7w4cO5/fbbAbjmmmtISUmhXbt2ZdxTIiIiEh643Wriic3p3bs3l53di/CIKOomJrJyxz5iIzz0Ou0sXnvttZDJXwq77LLLmD17Np06dcLM+Mc//kGDBg145513eOaZZwgPDycuLo53332XLVu2MHToUPLz/UspPP3002UWvzl3+Nu5HHXDZl5gJXA2/qLFPGCwc+6nQsd0AT4E+jnnVpWk3aSkJDd//vxyiPjIdmfkcOVrP5CakcOHt5zM8fWL+YfevADGnu+f7nHdFPAUU+A4Fukp8NVjcOKlcHzfsm9fROQwfv7552r9H93s7Gw8Hg9er5fZs2dz6623kpycHOywjtrtt99Oly5duOmmm4o9pqh/czNb4JxLKu/4ylMw8woREalYoZS/5Obl8+v+HDKyfWRm5+EL/L88whtGbISX2EgvsREeIrxhv5s6eixCKX85mryi3EZeOOfyzOx2YBr+W6WOcc79ZGZPAPOdc5/inyYSB/w78A+y0Tl3cXnFdCwyc/K48e15bNmzn/eG9Sy+cLF3G0wYDDUa+O/8UR6FC4C4RP/tU0VEpMJt3LiRq6++mvz8fCIiInjjjTeCHdJR69atG7GxsTz77LPBDkVERKTaCPf618aghn/dxKxcH+nZPjJz8tiXlcevmTkAeD1hxEZ4AsUML1HhR1/MqOz5S7muyuWcmwpMPWjbI4Uen1Oe5y8reb58bh+/iCWb9/Datd3o3qJu0QfmZsHEayF7H1z3kX+dChERqXKOP/54Fi1a9LttqampnH322Ycc+9VXXx2yUngoWbBgQbBDEBERqdbMjOgIL9ERXiAS5xzZeflkZOeRkeMjIzuPtP3+u4d5wiwwMsNf0IgO95S4mFHZ8xctKX4EzjkemrKUr1fs5KlL23PeSQ2KOxA+uwu2zIerx0H94N86UEREKk58fHylnjoiIiIiocHMiAr3EBXu4UD5ICfPR0a2j4ycPDKyfezN+n0xIy7KS1ykl8hSTjOpTPmLihdH8NyMlUyav5k/nnUc1/ZqXvyBc16HxePh9AfgxJCc+SIiIiIiIiKVUITXQ4TXQ53YCAByff6RGemBn717/MUMryeMuEhv4Mf/mqpCxYvDGPfjBl76ejUDkppyV98Tij9wzSyY9mdo2x9Ov7/iAhQREREREZFqJ9wTRu2YCGrH+IsZOXk+fyEjy0d6Vh57AmtmRHh/K2bERnoJ94QFM+xjouJFMb5cto1HPlnG2W3r8dfL2hc99CZjF6z+Cr64DxLbwGWv+29nKiIiIiIiIlJBIrwe6no91I2lYM0MfzEjj7TMXHZn+IsZUeEe4iK91AhMMymrO5lUBBUvijB33W7+OCGZzk1r8/LgrngPVKfyfbBlAayeCatmwNZFgINaTWHgeIg88j1yRUTk2O3Zs4fx48dz2223lfq1L7zwAiNGjCAmJqYcIhMREREJrsJrZiTE+RcA3Z/rH5GRnp3H7owcdqVn4w0Lo1ZMOLWjw4mJKPnCn8GiYQIHWbljH8PemUeTOtG8dUN3orN3QfJ4+PBGeKY1vNUXvnkGwrxw5kMw/Gu4cwnUbRns0EVEqo09e/bw6quvHtVrX3jhBTIzM8s4oqOTl5cX7BBERESkgpR3/tKiRQt27dp1yHYzIybCS72aUbRKjOPEhjVpHh9LbKSHXzNyWJOSzi/b97E9bT9Zub6jiq8iqHhRyNY9+7nxrdn08vzCx22+ou64s+HZE+DjW2Hdt3DC+XDlGLh3DQybAaffB427aaqIiEgFe+CBB1izZg2dO3fm3nvv5ZlnnqF79+507NiRRx99FICMjAwuvPBCOnXqRPv27Zk4cSIvvvgiW7du5cwzz+TMM88stv1bb72VpKQkTjrppIL2AObNm8cpp5xCp06d6NGjB/v27cPn83HPPffQvn17OnbsyEsvvQT8PoGYP38+Z5xxBgCPPfYY1113Hb179+a6665j/fr19OnTh65du9K1a1d++OGHgvP9/e9/p0OHDnTq1KngPXft2rVg/6pVq373XEREREJXqFx8CQszakWH0zw+lnYNa9C0TgyR4R5S9uWwcsc+Vu7Yx859WeTkhVYhQ9NGAPZuJXP5NFbPnMTU3GRqWiYs8EDTnnD2I3DcOVC/g4oUIiJF+eIB2L60bNts0AHOH1ns7pEjR7Js2TKSk5OZPn06H374IXPnzsU5x8UXX8w333xDSkoKjRo14vPPPwcgLS2NWrVq8dxzzzFr1iwSEhKKbf+vf/0rdevWxefzcfbZZ7NkyRLatm3LgAEDmDhxIt27d2fv3r1ER0czevRo1q9fT3JyMl6vl927dx/x7S1fvpzvvvuO6OhoMjMzmTFjBlFRUaxatYpBgwYxf/58vvjiCz755BPmzJlDTEwMu3fvpm7dutSqVYvk5GQ6d+7M2LFjGTp0aOn7V0REpLoLQv5S+OJL3759qVevHpMmTSI7O5vLLruMxx9/nIyMDK6++mo2b96Mz+fjL3/5Czt27Ci4+JKQkMCsWbOOGMpzzz3HmDFjABg2bBj/93//V2TbAwYM4O9PPsKnn36Kx+vltDPO5q6Hn2R7Whbb07KIjfBSOyacWtHhvy2nECQqXgC+aQ8T89NkTnB12X98f2p27Q+tzoCoWsEOTUREjmD69OlMnz6dLl26AJCens6qVavo06cPd999N/fffz/9+/enT58+JW5z0qRJjB49mry8PLZt28by5csxMxo2bEj37t0BqFmzJgAzZ87klltuwev1f6XWrVv3iO1ffPHFREdHA5Cbm8vtt99OcnIyHo+HlStXFrQ7dOjQgrU5DrQ7bNgwxo4dy3PPPcfEiROZO3duid+XiIiIBE95X3w5YMGCBYwdO5Y5c+bgnKNnz56cfvrprF279pC2U1NTmTJlCitWrMDM2LNnD7Vrx5GT52NPZi579ueyZc9+tu7JIi7KS+3ocGpGh+MJq/j1MVS8AEZl9mdWTnduH3Ax/Ts1DnY4IiKVy2GuMFQE5xwPPvggN9988yH7Fi5cyNSpU3n44Yc5++yzeeSRR47Y3rp16xg1ahTz5s2jTp06DBkyhKysrFLH5fV6yc/PBzjk9bGxsQWPn3/+eerXr8/ixYvJz88nKirqsO1eccUVPP7445x11ll069aN+Pj4UscmIiJS7QU5fymPiy8HfPfdd1x22WUF+cbll1/Ot99+S79+/Q5pOy8vj6ioKG666Sb69+9P//79Af/dS+rV9FCvZhT7c32kZeawJzOXTVm5hO0x2jSoUeG3XdU8COD8s8/i+ksuVOFCRKSSqFGjBvv27QPgvPPOY8yYMaSnpwOwZcsWdu7cydatW4mJieHaa6/l3nvvZeHChYe8tih79+4lNjaWWrVqsWPHDr744gsA2rRpw7Zt25g3bx4A+/btIy8vj759+/Kvf/2rYPHNA9NGWrRowYIFCwCYPHlysedLS0ujYcOGhIWFMW7cOHw+//zSvn37Mnbs2IL5rQfajYqK4rzzzuPWW2/VlBEREZFK6sDFl+TkZJKTk1m9ejU33XQTJ5xwAgsXLqRDhw48/PDDPPHEE2V2zqLa9nq9zJ07lyuvvJLPPvuMfv36HfK66HAPDWpF06ZBDVonxlGvRmSFFy5AxQsAOjapzeCezYIdhoiIlFB8fDy9e/emffv2zJgxg8GDB3PyySfToUMHrrzySvbt28fSpUvp0aMHnTt35vHHH+fhhx8GYMSIEfTr16/YBTs7depEly5daNu2LYMHD6Z3794AREREMHHiRO644w46depE3759ycrKYtiwYTRr1oyOHTvSqVMnxo8fD8Cjjz7KnXfeSVJSEh6Pp9j3ctttt/HOO+/QqVMnVqxYUXCVpF+/flx88cUkJSXRuXNnRo0aVfCaa665hrCwMM4999wy6U8REREpf+V58aWwPn368PHHH5OZmUlGRgZTpkyhT58+Rbadnp5OWloaF1xwAc8//zyLFy8utl0zIzbSf9eSYDDnXFBOfLSSkpLc/Pnzgx2GiEi19vPPP9OuXbtgh1FtjRo1irS0NJ588skKO2dR/+ZmtsA5l1RhQZQD5RUiItVHKOQvgwcPZsmSJZx//vk0adKEN998E4C4uDjee+89Vq9ezb333ktYWBjh4eG89tprJCUl8dJLL/Hyyy/TqFGjYhfsbNGiBfPnzychIaHIBTunTZt2SNuNGzfmkksuISsrC+cc99xzDzfccEO598PR5BUqXoiISKmFwpd/dXXZZZexZs0avv766xIt2lVWVLwQEZHKTvlL6DiavEILdoqISLXVs2dPsrOzf7dt3LhxdOjQIUgRHdmUKVOCHYKIiIhIhVPxQkREqq05c+YEOwQRERGRUqmMF1/KgooXIiJyVJxzmFX8Pb6l4lW2KaYiIiLFqQr5S2W/+HK0eYXuNiIiIqUWFRVFamqq/lNbDTjnSE1NJSoqOCuLi4iIlBXlL8F3LHmFRl6IiEipNWnShM2bN5OSkhLsUKQCREVF0aRJk2CHISIickyUv4SGo80rVLwQEZFSCw8Pp2XLlsEOQ0RERKTElL9Ubpo2IiIiIiIiIiIhTcULEREREREREQlpKl6IiIiIiIiISEizyrbSqpmlABvKoekEYFc5tFtVqb9KR/1VOuqv0lF/lY76q/SK67PmzrnEig6mLCmvCBnqr9JRf5WO+qt01F+lo/4qncP112HzikpXvCgvZjbfOZcU7DgqC/VX6ai/Skf9VTrqr9JRf5We+qz01Gelo/4qHfVX6ai/Skf9VTrqr9I5lv7StBERERERERERCWkqXoiIiIiIiIhISFPx4jejgx1AJaP+Kh31V+mov0pH/VU66q/SU5+VnvqsdNRfpaP+Kh31V+mov0pH/VU6R91fWvNCREREREREREKaRl6IiIiIiIiISEhT8UJEREREREREQpqKF4CZ9TOzX8xstZk9EOx4Qp2ZrTezpWaWbGbzgx1PqDGzMWa208yWFdpW18xmmNmqwO86wYwxlBTTX4+Z2ZbAZyzZzC4IZoyhxMyamtksM1tuZj+Z2Z2B7fqMFeEw/aXPWBHMLMrM5prZ4kB/PR7Y3tLM5gS+JyeaWUSwYw1lyitKR3nF4SmvKB3lFaWjvKJ0lFeUTlnnFdV+zQsz8wArgb7AZmAeMMg5tzyogYUwM1sPJDnndgU7llBkZqcB6cC7zrn2gW3/AHY750YGEtk6zrn7gxlnqCimvx4D0p1zo4IZWygys4ZAQ+fcQjOrASwALgWGoM/YIQ7TX1ejz9ghzMyAWOdcupmFA98BdwJ/Aj5yzk0ws9eBxc6514IZa6hSXlF6yisOT3lF6SivKB3lFaWjvKJ0yjqv0MgL6AGsds6tdc7lABOAS4Ick1RizrlvgN0Hbb4EeCfw+B38f+SEYvtLiuGc2+acWxh4vA/4GWiMPmNFOkx/SRGcX3rgaXjgxwFnAR8GtuvzdXjKK6RMKa8oHeUVpaO8onSUV5ROWecVKl74P2ybCj3fjD6AR+KA6Wa2wMxGBDuYSqK+c25b4PF2oH4wg6kkbjezJYHhnxqqWAQzawF0Aeagz9gRHdRfoM9YkczMY2bJwE5gBrAG2OOcywscou/Jw1NeUXrKK0pPf/NLT3/zj0B5RekoryiZsswrVLyQo3Gqc64rcD7wh8DwPCkh55+rVb3nax3Za0BroDOwDXg2uOGEHjOLAyYD/+ec21t4nz5jhyqiv/QZK4Zzzuec6ww0wT+KoG2QQ5KqT3nFMdDf/BLR3/wjUF5ROsorSq4s8woVL2AL0LTQ8yaBbVIM59yWwO+dwBT8H0I5vB2BOXIH5srtDHI8Ic05tyPwhy4feAN9xn4nMGdwMvC+c+6jwGZ9xopRVH/pM3Zkzrk9wCzgZKC2mXkDu/Q9eXjKK0pJecVR0d/8UtDf/MNTXlE6yiuOTlnkFSpe+BfSOj6w4mkEMBD4NMgxhSwziw0sToOZxQLnAssO/yrB/5m6IfD4BuCTIMYS8g58WQZchj5jBQILH70F/Oyce67QLn3GilBcf+kzVjQzSzSz2oHH0fgXnfwZf7JxZeAwfb4OT3lFKSivOGr6m18K+ptfPOUVpaO8onTKOq+o9ncbAQjcyuYFwAOMcc79NcghhSwza4X/qgiAFxiv/vo9M/sAOANIAHYAjwIfA5OAZsAG4GrnnBaTotj+OgP/sDsHrAduLjTvslozs1OBb4GlQH5g80P451vqM3aQw/TXIPQZO4SZdcS/cJYH/wWOSc65JwJ/+ycAdYFFwLXOuezgRRralFeUnPKKI1NeUTrKK0pHeUXpKK8onbLOK1S8EBEREREREZGQpmkjIiIiIiIiIhLSVLwQERERERERkZCm4oWIiIiIiIiIhDQVL0REREREREQkpKl4ISIiIiIiIiIhTcULkWrMzJyZPVvo+T1m9lg5nOcDM1tiZneVddtHOO9/zSypIs8pIiJSXSmvEJHy5A12ACISVNnA5Wb2tHNuV3mcwMwaAN2dc8eVR/siIiISMpRXiEi50cgLkeotDxgNHHLlwsxamNnXgSsbX5lZs8M1ZGZRZjbWzJaa2SIzOzOwazrQ2MySzazPQa9JNLPJZjYv8NM7sP0xMxtnZrPNbJWZDQ9sNzN7xsyWBc4zoFBb9we2LTazkYVOc5WZzTWzlQefX0RERMqU8goRKTcaeSEirwBLzOwfB21/CXjHOfeOmd0IvAhceph2/gA451wHM2sLTDezE4CLgc+cc52LeM0/geedc98FkphpQLvAvo5ALyAWWGRmnwMnA52BTkACMM/MvglsuwTo6ZzLNLO6hc7hdc71MLMLgEeBc0rUKyIiInI0lFeISLlQ8UKkmnPO7TWzd4E/AvsL7ToZuDzweBxwcBJysFPxJyY451aY2QbgBGDvYV5zDnCimR14XtPM4gKPP3HO7Qf2m9ksoEfgHB8453zADjP7H9AdOB0Y65zLDJx/d6FzfBT4vQBocYT3ICIiIsdAeYWIlBcVL0QE4AVgITC2gs8bBvRyzmUV3hhIOtxBxx78vKSyA7996G+eiIhIRVBeISJlTmteiMiBKwqTgJsKbf4BGBh4fA3w7RGa+TZwHIFhnc2AX47wmunAHQeemFnhIaCXBOa7xgNnAPMC5xhgZh4zSwROA+YCM4ChZhYTaKfw8E4RERGpQMorRKQ8qHghIgc8i3++5wF34P/iXgJcB9wJYGa3mNktRbz+VSDMzJYCE4EhzrnsIo4r7I9AUmDxruVA4XaXALOAH4EnnXNbgSmB7YuBr4H7nHPbnXNfAp8C880sGbinNG9cREREypzyChEpU+bc0Y6YEhEpH+a/J3y6c25UsGMRERGRyk15hUjVoJEXIiIiIiIiIhLSNPJCREREREREREKaRl6IiIiIiIiISEhT8UJEREREREREQpqKFyIiIiIiIiIS0lS8EBEREREREZGQpuKFiIiIiIiIiIQ0FS9EREREREREJKSpeCEiIiIiIiIiIU3FCxEREREREREJaSpeiIiIiIiIiEhIU/FCREREREREREKaihciUuHM7L9mNizweIiZfXeU7TgzO65soxMRERH5PTM7w8w2BzsOkepMxQuRas7M1pvZfjNLN7PtZva2mcUFOy4AM2sRKFCkB37Wm9kDR9HOURdIRERE5DcH5Q0HfhoF9o02s1/MLN/MhhyhnSZmNtnMdplZmpktO9JrylsgX/AF3tNeM0s2s/5H0c7bZvZUecQoUp2peCEiABc55+KAzkAX4MEgx3Ow2oH4BgGPmFm/YAckIiJSjV3knIsr9LM1sH0xcBuwsARtjAM2Ac2BeOA6YEdZBmlm3qN42exAzlEbeAuYZGZ1yjIuETk6Kl6ISAHn3HZgGv4iBgBm1svMfjCzPWa22MzOKLSvrpmNNbOtZvarmX0c2F7HzD4zs5TA9s/MrEkZxDcb+Alof/A+M6tlZu8GzrnBzB42szAzawe8DpwcuJKy51jjEBERkUM5515xzn0FZJXg8O7A2865DOdcnnNukXPuiwM7zezUQvnHpgOjMor7vg/sG2Jm35vZ82aWCjxmZpFmNsrMNprZDjN73cyiS/Be8oExQDTQ+uD9ZtYuMA12j5n9ZGYXB7aPAK4B7gvkHf8pQV+ISAmoeCEiBQIFhvOB1YHnjYHPgaeAusA9wGQzSwy8ZBwQA5wE1AOeD2wPA8biv5rSDNgPvHyMsZmZ9Q6ca1ERh7wE1AJaAacD1wNDnXM/A7cQuJLinKt9LHGIiIhImfgReMXMBppZs8I7zKw58AX+7/ZE/BdVkgO7i/y+L/TynsBaoD7wV2AkcEKgjeOAxsAjRwouMGpjGJAOrDpoXzjwH2A6/vznDuB9M2vjnBsNvA/8I5B3XFSSzhCRI1PxQkQAPjazffiHb+4EHg1svxaY6pyb6pzLd87NAOYDF5hZQ/yFjlucc78653Kdc/8DcM6lOucmO+cynXP78CcPpx9DfLuA3cCbwAOBqzoFzMwDDAQedM7tc86tB57FPwRVREREytbHgREHew6MujwKVwHfAn8B1gXWl+ge2DcYmOmc+yCQX6Q655JL+H2/1Tn3knMuD/8IkBHAXc653YGc5G+BNorTKzBKczv+6aqXOefSDj4GiANGOudynHNfA58FjheRcnI088BEpOq51Dk308xOB8YDCcAe/CMnrjKzwlcNwoFZQFNgt3Pu14MbM7MY/KMw+gEH5onWMDOPc853FPElBJKQYvcH4tpQaNsG/FdXREREpGxd6pybeSwNBPKHB4AHzCwBGIW/KNIEf46xpoiXleT7flOhx4n4R4guMLMD2wzwHCa0H51zpx4h/EbApsDUkuLiEJEyppEXIlIgMHLibfwJBPgTgHHOudqFfmKdcyMD++qaOfpkgQAAIABJREFUWVHTMO4G2gA9nXM1gdMC262IY8vCLiAXf7HlgGbAlsBjV07nFRERkWPknNuFP/dohH+a6iaKWGeCI3/fw++/83fhn7p6UqE8plZgQc5jsRVoemCtjSLiUN4hUg5UvBCRg70A9DWzTsB7wEVmdp6Zecwsyvz3OW/inNuGfz7qq4EFOsPN7ECRogb+ZGGPmdXlt2ko5SIwmmMS8FczqxGYK/unQPzgX728iZlFlGccIiIi1ZmZRZhZFP6LFeGBvKHI/2+Y2d/NrL2Zec2sBnArsNo5l4p/zYhzzOzqwP54M+tcgu/73wmMjHgDeN7M6gXO29jMzjvGtzoHyMS/KGd4YDHzi4AJgf078K/JISJlSMULEfkd51wK8C7wiHNuE3AJ8BCQgv9KyL389rfjOvxXQFbgXyvj/wLbX8C/Ovcu/AtyfVkBod8BZOBfpOs7/NNfxgT2fY3/LiXbzWxXBcQiIiJSHU3Hf/HiFGB04PFpxRwbA0zBP011Lf7RFBcDOOc2AhfgH8m5G/9inZ0Crzvc931R7se/EPmPZrYXmIl/dOhRc87l4C9WnI8/13kVuN45tyJwyFvAice4JoiIHMSc06gmEREREREREQldGnkhIiIiIiIiIiFNxQsRERERERERCWkqXoiIiIiIiIhISFPxQkRERERERERCmjfYAZRWQkKCa9GiRbDDEBERqfYWLFiwyzmXGOw4joXyChERkdBwpLyi0hUvWrRowfz584MdhoiISLVnZhuCHcOxUl4hIiISGo6UV2jaiIiIiIiIiIiENBUvRERERERERCSkqXghIiIiIiIiIiFNxQsRERERERERCWkqXoiIiIiIiIhISFPxQkRERERERERCmooXIiIiIiIiIhLSVLwQERERERERkZCm4oWIiIiIiIiIhDQVL0REREREREQkpKl4ISIiUoVl5fr46ucdwQ5DREREqoD07DxmrdgZlHN7g3JWERERKXfrdmVw2/sLWbljH1/ffTrN42ODHZKIiIhUQsu2pDF+7kY+WbSF/bk+fnzwbOrVjKrQGFS8EBERqYL+s3grD0xeQrg3jDeu76bChYiIiJTK/hwfny3ZyvtzNpK8aQ9R4WFc3KkR1/RsTmKNyAqPR8ULERGRKiQr18eTny3n/Tkb6dqsNi8P7kqj2tHBDktEREQqiVU79vH+nI18tHAze7PyOL5eHI9ddCKXdW1CrejwoMWl4oWIiEgVsT4wTWT5tr3cfFor7jmvDeGeqrO8lZn1A/4JeIA3nXMjD9rfDHgHqB045gHn3NQKD1RERKSSyc7z8eWy7bw/ZyNz1+0mwhPG+R0acE3P5nRvUQczC3aIKl6IiIhUBZ8v2cb9k5fgCTPeuiGJs9vVD3ZIZcrMPMArQF9gMzDPzD51zi0vdNjDwCTn3GtmdiIwFWhR4cGKiIhUEhtSMxg/dyP/nr+Z3Rk5NI+P4cHz23JltybEx1X81JDDUfFCRESkEsvK9fG3qT/z7uwNdAlME2lcNaeJ9ABWO+fWApjZBOASoHDxwgE1A49rAVsrNEIREZFKIM+Xz8yfd/L+nA18u2oXnjDjnHb1uKZnc049LoGwsOCPsiiKihciIiKV1IbUDP4wfiHLtuxleJ+W3NevbZWaJnKQxsCmQs83Az0POuYxYLqZ3QHEAucU1ZCZjQBGADRr1qzMAxUREQlF29L288HcTUyct5Ede7NpWCuKu845gQHdm9KgVsXeOeRoqHghIiJSCU1duo37P1xCWJjxxvVJ9D2xak0TOUqDgLedc8+a2cnAODNr75zLL3yQc240MBogKSnJBSFOERGRCuGcY/baVMbN3sD05TvId47TT0jkqUubc2abRLyV6KJHuRYvSrCw1vPAmYGnMUA951zt8oxJRESkMsvO8/G3z3/mndkb6Ny0Ni8P7kKTOjHBDqsibAGaFnreJLCtsJuAfgDOudlmFgUkADsrJEIREZEQkZ6dx5SFm3l39gZW7Uyndkw4w05tyTU9m9MsvnLmDeVWvCjJwlrOubsKHX8H0KW84hEREansNqZm8ofxC1m6JY1hp/qniUR4K88Vk2M0DzjezFriL1oMBAYfdMxG4GzgbTNrB0QBKRUapYiISBCt3pnOuNnrmbxwC+nZebRvXJN/XNmRizs1IircE+zwjkl5jrwoycJahQ0CHi3HeERERCqtL5dt494Pl2DA6Ou6ce5JDYIdUoVyzuWZ2e3ANPwjOsc4534ysyeA+c65T4G7gTfM7C78i3cOcc5pWoiIiFRpeb58vlqxk3dnr+f71alEeMK4sGNDrju5OV2a1g6J25yWhfIsXpRkYS0AzKw50BL4upj9WlhLRESqpaxcH09P9U8T6dS0Ni8P6kLTupVzuOexcs5NxX/708LbHin0eDnQu6LjEhERCYbU9GwmzNvE+z9uYGtaFg1rRXHveW0Y0L0pCSF2m9OyECoLdg4EPnTO+YraqYW1RESkOvppaxr/NyGZVTvTuenUltxfvaaJiIiIyEGccyRv2sO42Rv4bMk2cnz5nNI6nkcuOolz2tWrVAtwllZ5Fi9KsrDWAQOBP5RjLCIiIpVGfr7jre/W8cy0X6gdE867N/bgtBMSgx2WiIiIHAXnHFvTssjNy8cB+c7hnMM5yHf+5/mB567w88Br8wPb1+/K4L05G1iyOY3YCA8DezTlul7NOb5+jWC/xQpRnsWLkiyshZm1BeoAs8sxFhERkUphe1oWd/87me9Xp3LuifUZeUVH6sZGBDssERERKaVfM3KYvHAzH8zdyJqUjDJp87h6cTx5yUlc1rUJcZGhMpGiYpTbuy3hwlrgL2pM0IJaIiJS3U1duo0HP1pKTl4+Iy/vwIDuTavMIlsiIiLVgXOOuet288HcjUxdtp2cvHy6NKvNYxedSM3ocMLMMAMzI8zwP+eg54V+F95eI8pLh8a1qm1uUK6lmiMtrBV4/lh5xiAiIhLq0rPzeOzTn/hwwWY6NanFCwO70DIhNthhiYiISAntzsjho4WbGT93I2tTMqgR5WVQ96YM7NGMdg1rBju8KqF6jTMREREJMQs2/MpdE5PZ/Gsmt595HHeeczzhVXixLRERkarCOcePa/2jLL5ctp0cXz7dmtdh1FXHcWGHhkRHeIIdYpWi4oWIiEgQ5PnyeXnWal76ejUNakYx8eaT6d6ibrDDEhERkSNITc9m8sLNTJi7ibW7MqgZ5WVwz2YM6tGMNg2qx+KZwaDihYiISAXbmJrJ/01cxMKNe7isS2Mev+QkakaFBzssERERKYZzjtlrUhk/dyPTftpOrs/RvUUdbj/rOC7o0JCocI2yKG8qXoiIiFQQ5xwfLtjMY5/+RFiY8eKgLlzcqVGwwxIREZFiZObkMW72Bj6Yu5H1qZnUig7nul4tGNSjabW5RWmoUPFCRESkBHalZ/O3qT+zd38u0RFeYiM8REd4iInwEBPhDfz+7XF04PGB4wCenrqCz5duo0fLujw/oDONa0cH+V2JiIhIcZZuTuPOCYtYuyuDHi3qcuc5x3N+e42yCBYVL0RERI5g0+5Mrh8zl6179tM6MY79uT4ysvPYn+MjIyeP/BLe7NsbZtzXrw03n9YaT1j1vM2ZiIhIqMvPd4z+di3PTv+F+NhIxg/vySmtE4IdVrWn4oWIiMhhLN+6lxvGziUnL5/xw3vSrfnvF9V0zpGdl19QyNif4yPzoMeZgcc9W8XrdmkiIiIhbHtaFn+alMwPa1I5v30Dnr68A7VjIoIdlqDihYiISLF+XJvK8HfmExflZfwtJxc5t9XMiAr3EBXuoU6skhsREZHK6stl23ngoyVk5+bz9ys6cHVSU8w0UjJUqHghIiJShC+XbeOPE5JpVjeGd2/sQSOtTyEiIlIlZebk8eRny/lg7iY6NK7FPwd2plViXLDDkoOoeCEiInKQ9+ds4C8fL6NT09qMuaG7RlSIiIhUUcu2pPHHCYtYtyuDW05vzZ/6nkCENyzYYUkRVLwQEREJcM7x4lereX7mSs5qW49XBnctuFOIiIiIVB35+Y43vl3LqMCinO8P06KcoU7FCxEREcCX73j002W89+NGrujahJFXdCDcoysvIiIiVc32tCzu/ncy369Opd9J/kU5Ncoy9Kl4ISIi1V5Wro+7JibzxbLt3HJ6a+7v10YLdImIiFRB037azv2T/Ytyjry8AwO6a1HOykLFCxERqdb2ZuUy4t35/Lh2Nw9f2I5hfVoFOyQREREpY/5FOX/mg7kbad+4Jv8c2IXWWpSzUlHxQkREqq2de7O4Yew8Vu3YxwsDOnNpl8bBDklERETKWOFFOW8+vRV3922jRTkrIRUvRESkWlq3K4Prx8whNT2Ht4Z05/QTEoMdkoiIiJSCc44cXz65PkdOXv5vPz4fOXn+fbPXpPLcjF+oGxvB+zf15JTjtChnZaXihYiIVDtLN6cxZOxcHDB+eC86N60d7JBEREQkwJfvmLJoC5PmbWJfdh45eT5/kSJQkPitSJFfovbOO6k+Iy/vqEU5KzkVL0REpFr5dlUKt4xbQO2YCN69qYfmu4qIiIQI5xzTftrOqOkrWb0znTb1a9AsPoYIbxgRnsCP1/8THngc6f1te3ih/REe/75aMeF0aVpbi3JWASpeiIhItZCf7xjz/Tr+/uUKWifG8c6NPahfMyrYYYmIiAjw3apdPDNtBYs3p9E6MZbXrulKv/YNVHSQAipeiIhIlbdpdyb3/Hsxc9bt5px29Xn26k7Uig4PdlhSSmbWD/gn4AHedM6NPGj/88CZgacxQD3nnOYEiYiEsIUbf2XUtF/4YU0qjWtH88yVHbmsS2O8Hi2oKb+n4oWIiFRZzjkmzd/EE/9ZjpnxzJUdubJbE13FqYTMzAO8AvQFNgPzzOxT59zyA8c45+4qdPwdQJcKD1RERErkl+37GDX9F2Ys30FCXASPXXQig3o2I9LrCXZoEqJUvBARkSpp574sHpy8lK9W7KRXq7qMuqoTTerEBDssOXo9gNXOubUAZjYBuARYXszxg4BHKyg2EREpoY2pmTw/cyUfJ28hLtLLPeeewNDeLYmN1H9N5fD0CRERkSpn6tJt/HnKUjJzfDzS/0SGnNKCsDCNtqjkGgObCj3fDPQs6kAzaw60BL4uZv8IYARAs2bNyjZKEREp0s69Wbz49SomzN2E12PcfFprbjm9FbVjdAcQKRkVL0REpMpIy8zlkU+X8UnyVjo2qcVzV3fmuHq6m0g1NBD40DnnK2qnc240MBogKSnJVWRgIiLVzZ7MHF7/31re/mEdeT7HwB5N+eNZx1NPi2ZLKal4ISIiVcI3K1O478Ml7ErP5q5zTuC2M1sTrsW+qpItQNNCz5sEthVlIPCHco9IRESKlZGdx9jv1/Gvb9aSnp3HpZ0bc9c5J9AsXlM45eioeCEiIpVaZk4ef5v6M+/9uJHj68XxxvVJdGhSK9hhSdmbBxxvZi3xFy0GAoMPPsjM2gJ1gNkVG56IiBww7aft/HnKMnalZ9P3xPrcc24b2jSoEeywpJJT8UJERCqtBRt286dJi9m4O5PhfVpy97ltiArXKuVVkXMuz8xuB6bhv1XqGOfcT2b2BDDfOfdp4NCBwATnnKaDiIhUsP05Pp78fDnj52ykQ+NajL6+G12b1Ql2WFKWfLmwZQE061Xhp1bxQkREKp3sPB8vzFzFv/63hka1o/lgeC96tYoPdlhSzpxzU4GpB2175KDnj1VkTCIi4rd8617+OGERq3emc/Pprbi7bxsivJq+WaWsmgHTHoLd6+DOxVCrcYWevlyLF2bWD/gn/iskbzrnRhZxzNXAY4ADFjvnDhkCKiIi1Ztzjl3pOaxJSWdtSgbvzl7Piu37GNi9KQ/3P5E43V5NREQkKJxzvP3Dep6euoLaMeG8d1NPTj0+IdhhSVnauQKm/xlWz4S6rWHAe1CzUYWHUW7Znpl5gFeAvvhvZzbPzD51zi0vdMzxwINAb+fcr2ZWr7ziERGR0JeTl8+G1AzWpGQUFCrWpKSzJiWdfVl5BcfVrxnJmCFJnNW2fhCjFRERqd52pWdz778XM+uXFM5pV4+/X9GR+LjIYIclZSVzN/x3JMx7EyLi4Ly/Qffh4A3O7W3L81JVD2C1c24tgJlNAC4Blhc6ZjjwinPuVwDn3M5yjEdEREJEVq6PpVvSWLMznbW7Mgp+b9ydiS//t6UK6teMpHViHJd0bkTrxDhaJ8bRKjGWRrWiCQuzIL4DERGR6u1/K1O4e9Ji9mbl8sQlJ3Fdr+aY6bu5SvDlwry34L9PQ/ZeSLoRzngIYoM7Rbc8ixeNgU2Fnm8Geh50zAkAZvY9/qkljznnvjy4ITMbAYwAaNasWbkEKyIiFWNDagY3jJnL+tRMACK8YbRKiKVdwxr079iQVomxtE6Mo2VCLDWiwoMcbRWw/nv4/gW46m2IiA12NCIiUsll5/l45stfePO7dZxQP473h/XUnUSqkgPrWuxaCa3OgPOehvonBjsqIPgLdnqB44Ez8N+v/Rsz6+Cc21P4IOfcaGA0QFJSklYPFxGppJI37eGmt+eR7xwvD+5Cx8a1aVwnGo9GUZSPNbPgg0FQuylkp6t4ISIix2RNSjp//GARP23dy/UnN+ehC9rpLl9VxcHrWgyaACf0gxAaTVOexYstQNNCz5sEthW2GZjjnMsF1pnZSvzFjHnlGJeIiATBVz/v4Pbxi0ioEcE7Q3vQKjEu2CFVbSunw8RrIf44uP4TiEsMdkQiIlJJOeeYOG8Tj/9nOVHhYbx5fRLnnKh1p6qEEFvX4nDKs3gxDzjezFriL1oMBA6+k8jHwCBgrJkl4J9GsrYcYxIRkSAYP2cjD3+8lJMa1WLMkO4k1tBiXuVqxecw6Qb/MM/rPoaYusGOSEREKoBzjtlrU4kK93BcvThqlsH0y7TMXB6csoSpS7fT+7h4nru6M/VrRpVBtBJUB69r0W0onPkQxIbunWLKrXjhnMszs9uBafjXsxjjnPvJzJ4A5jvnPg3sO9fMlgM+4F7nXGp5xSQiIhXLOcdzM1by0terObNNIi8P7kqsbmtavpZ9BB8Nh4ad4drJEF072BGJiEgFeXb6Sl6etbrgeb0akRxXL+63n0T/78QakSVaXHPO2lTumpjMzn3ZPHh+W4b3aaUFs6uCEF7X4nDKNYN0zk0Fph607ZFCjx3wp8CPiIhUIbm+fB6YvJTJCzczsHtTnrq0PV5PWLDDqtoWT4SPb4GmPWHwJIiqGeyIRESkgkyat4mXZ63mym5NOO+kBqzeme7/SUnno4VbSM/+7ZbjNaO8tC5UzDjw06RODJ4wI8+Xzz+/WsUrs1bTrG4MH912Ch2bqBgeNPn58O0omP2Kfw0KT6R/WocnEryR4Iko5nfh4wK/ty6CNV+F7LoWh6PLXyIiUubSs/O49b0FfLtqF3edcwJ/PPs43T6tvC18Fz79I7Ts409GtDiniEi18f3qXTw0ZSl9jk/g6cs7EO4Jo2+hNSmcc+zYmx0oaOxjdYq/sDHrlxT+vWBzwXGR3jBaJsSS7xwrd6RzVbcmPHbxSRo1GUxZe2HKLfDL5/5CQ62m4MuGvJyDfmdDXhZkpYEvx/+84PeB43Igsgac+1foMeL/2bvv6Kiq9Y3j353QIfReQi/SS2hWQFQQBQREQBAEQRTs/drLz4v12gsIFpAmIk0EpQoondAh9BQg1ISSPrN/f5ygEYEMYSaT8nzWyiIzs8+Zh7uuZPKefd43S/a1uBT9v1BERLzqyKkEBn29hp3Rp3m7V2N6h1RJ/yC5MqvHwNwnoVZHuGsC5C3o70QiIpJJdkWfZviEddQsU4RP725O3gvscjTGUL5YAcoXK8C1tf/Z0yAmLok9qcWMc19HzyTyUd9mdG1SMbP+GnIhx3bB5H5wfA90egta33/luySszTY7Lc6n4oWIiHjN7iOnGThuDSfjkhg7MIR2dcv6O1LO9+enzn2rdTpD72+dbaIiIpIrHDntXDAokDeQcfe2zFCDzuKF8tGiaklaVFVz5yxlx1z46X7nFpCBs6Datd45bzYtXICKFyIi4iWr951g6HdryRsYwNT729KwUjF/R8r5lr0HC1+D+t2gx1fZbvuniIhkXHySi6HfruXE2SSm3t+WSsW16y5HcLvh97edKSAVmjo7KotrFyuoeCEiIl4wd/MhHp0SSuUSBfn23lZUKVnI35FyNmudDzVL34JGd0L3LyBQP9JFRHILl9vy6JQNbIqKZfSAEBpV1gWDHCEhNrW/xVxo0g9ue1+3gqahTzoiInJFxi7fxxs/b6N5cAm+uieEEoV19d+nrIUFL8OKD6Fpf+j6EQQE+juViIhkov/O3c78rdG8fHv9fzTmlGzsaJjT3+LkPuj8DrQamq1v8fAFFS9ERCRD3G7Lm3O389XyfXRqUJ4P+jSlQF79Eu1T1sK8Z2HVFxAyGG59DwI0flZEJDcZ/+d+vlq+j0FXV+Pea6r7O07O53bBkW1Qtr7vLhbs+Bmm3+/0rbpnFlS7xjfvk82peCEiIpfN7bY8NjWUmaEHGXR1NV68rT6BAbo64FNuN/z8OKz7Gto8CLe8qSsyIiK5zKId0bw8aysdryrLi7fV93ecnC8lEaYPhW0zoXBZaHAHNOwJlVt65+KB2w1LRzm3gVZsDneNh2KVr/y8OZSKFyIictk+WrSLmaEHeeqWujzYriZGv0T7ltsFsx6C0O/h2sfgxpdVuBARyWW2Hoxl5MQN1K9YlA/7NNNFA19LOgtTBsCehdB2JMSEw7pvYPWXUKzK34WMCk0y9jM5IRamD4OwedD0bujyPuQt4PW/Rk6i4oWIiFyWhduj+WDBLno0r6TChadOHnDGnR3ZlrHj3S5IOgPtnoMbnlHhQkQklzkUG8/gb9ZQvGBexg1sSeH8+jXOp+JjYOJdELkaun4Mze9xnk84BTt/gS0/wsrP4I+PoFQtp4jRsCeUqevZ+Y/sgCl3w8n9cOu70PI+/Wz3gP5fLyIiHtt37CyPTgmlYaWivHlHIxUuPLFnEUwb7GwNbdw74/fLVmwOTe7ybjYREcnyziSmMPibtZxNdDHtgbaULaqr8z515ihMuMMpMPT6Ghp0//u1AkWdn8VN7oK4E7B9llPIWPq2c+tHuUbQsIfzVaLahc+/fY5zQSNvQRg4G6penSl/rZxAxQsREfHI2cQUhn23ljwBhi/6t1BzzvRY60wEWfgqlK4Lfb6HUjX9nUpERLKRFJebEd+vJyz6NF8Pakm98kX9HSlni4mA8d0hNgr6TYZaHS++tlBJaDHI+Tp9GLbOcAoZC191viqFQKNeUL87FK3gXMRY8ib8/g5UagG9x0OxSpn1N8sRVLwQEZF0WWt5atpG9hw9w3eDW1O5RCF/R8raEs/AzBGwbYZzT2zXTyB/EX+nEhGRbMRay8uztrI07Cj/7dGI6+uU8XeknO3YLviuOySehgE/QdW2nh8bVB7aDHe+Th6ArdOdQsa8Z2Hec1DtWjABsG8pNOvvTAtTf4vLpuKFiIik68vf9zJ382Ge61yPa2uX9necrO34Hph8NxzbCTe9Blc/rPtYRUTksn21bB/frwpn+A016dsq2N9xcrZDG2F8D+f7QXOgQuOMn6tEVae59rWPwdEwp4ixZZrT8FP9La6IihciInJJy3Yd5e15O+jSuALDrq/h7zhZ2855TufwgEDoPx1qtvd3IhERyYZ+2XyIN3/ZTpfGFXj6Fg+bQErGhK+E73tD/iC4ZwaUru29c5epA+2fg3bPgjsFAvN679y5kBeG04qISE4VcSKOhyZtoHbZIN7u2VgNOi/G7YYlo2DSXc4Vl2FLVLjwAWNMJ2PMTmPMbmPMsxdZ09sYs80Ys9UYMzGzM4qIXKkN4Sd5dEoozaoU5707mxCgkai+s2uBc6tIkTIweJ53CxdpGaPChRdo54WIiFxQQrKL4RPW4XZbvhzQQmPZLiYhFqbfD2G/QJO+cNv/nA7i4lXGmEDgU+AmIBJYY4yZZa3dlmZNbeA54Bpr7UljTFn/pBURyZiIE3Hc9+1ayhUtwJh7QtQc25e2/gQ/DoWy9aD/T04BQ7I0fRIVEZF/sdbyn+mb2XboFOMGtqRa6cL+jpQ1pZ3T3vkdaDVU97H6Titgt7V2L4AxZjLQDdiWZs1Q4FNr7UkAa+2RTE8pIpIBZxJTWLg9mg8X7CLFbfn63paUKpLf37FyrvXfwexHoHIr6DcFChb3dyLxgIoXIiLyL9/+sZ/pG6J4/KY6tK+ni9cXtHUGzHgQ8hXWnPbMUQmISPM4Emh93po6AMaYFUAg8Iq1dl7mxBMRuTxnE1NYuOMIP286yOKdR0lKcVOhmLPjomYZTajymT8+gV+fh5o3wl3jnZ/jki2oeCEiIv+wau9x3vh5Ox2vKsfI9rX8Hcf7TuyFgDxQuGzGxpS5XbDodVj+P6jcEnp/B0Urej+nZEQeoDbQDqgM/G6MaWStjUm7yBgzDBgGEBysDv4iknniklJYtOMIP286xKIdR0hMcVM2KD/9WgVzW+MKNA8uoR4XvmItLP4/+P0dqN8deoyBPPn8nUoug4oXIiLyl0Ox8YyYuJ7gkoV4/64c2CRs2fuw8NW/H+cLcu5xLVwWCpeGImWhcBnn66/vU18rUAziT8KPQ2DPImhxL3R+C/JoW28miQKqpHlcOfW5tCKBVdbaZGCfMSYMp5ixJu0ia+1oYDRASEiI9VliEREgPsnF4p1OwWLhjmgSkt2UCcpPn5ZV6NK4IiFVc2DB4nQ0bJ8FxapAxaYQVN6/edxumPcMrB4NzQbA7R86k8EkW1HxQkQkG1u59zgvzdzCdbXL0L1pJRpWKprhiSCJKS4emLCe+CQXk4e1oWiBHNYVO2w+LHwN6t0GtW+Gs0f//jpzBI7vhvA/Ie4EcIHfZwPzOzs23Mlw+0fe+gefAAAgAElEQVTQYmCm/xVyuTVAbWNMdZyiRR+g33lrZgB9ga+NMaVxbiPZm6kpRURwml4v2XmEOZsOsXD7EeKTXZQuko87W1ShS+MKtKxWksCcVrAAp8i/4iNY9QUkx/39fFAFqNDUKWSc+zOzChquFJg5AjZNhrYj4eY31J8qm1LxQkQkm0pKcfOfnzZz7HQi3/25n7HL91GjTGG6NalEt6YVL7vJ5iuzthIaEcMX/ZtTq2yQb0L7y9Ew+PE+KN/I2Saar9DF17pSIO44nD2SWtg4V+Q4AgmnnCs2lVtkXnYBwFqbYowZCczH6Wcxzlq71RjzGrDWWjsr9bWbjTHbABfwlLX2uP9Si0hukpDsYmnYUWeHxfZozia5KFU4Hz2aV6JL4wq0rl4qZxYsABLPwKrPYcXHkHgKGvWCax9zJnIdDIVDoXBwA4TN468LBEXKpylmNPNOQcPtct4z7oTzszz+hNOcc+dc6PACXPekChfZmIoXIiLZ1Hd/7mfv0bOMGxRCi+CSzN1yiJmhUXywMIz/LQijSZXidG9akS6NK1A26NK9HSatDmfS6ggebFeTTg0rZM5fILPEx8DkvhCYD/pMvHThAiAwDwSVc74kS7HWzgXmnvfcS2m+t8DjqV8iIpkiMcXFt3/s55NFuzmVkEKJQnnp2rQStzWuQOvqJckTGODviL6TnADrvobf34W4Y1C3C3R4Hso1+HtN2obWiafh8OY0BY1QZ2fkBQsaTaFMPUiOd4oQaQsScSecXR5xx1O/P/daDP/ePWmciWCth/n4fwzxNRUvRESyoSOnE/hgwS7a1y1Dh3rOL9l9WwXTt1Uwh2Ljmb3xIDM2HOTV2dt4fc42rqlVmm5NK3FLg3IEnXc7yIbwk7w8cyvX1ynDEzfX9cdfx3fcLpg+1BllOnA2FK+S7iEiIiKesNYyd/NhRs3bTsSJeNrXLcPga6vTpkYp8ubkggU4uxRDv4elb8GpKKh+A9z4ElQOufRx+YOcYsY/ChpnnILGuWLGwQ3/LGhcSJ6CUKgUFCoBBUtCsUbO44IloVDJ1D9TXy9ayf89N8QrVLwQEcmG3p63k8QUFy/eVv9fr1UoVpBh19dk2PU12RV9mpmhB5m5MYonf9jI8z8F0PGqcnRrWpEb6pbhVHwKD0xYT7li+fmoT9Oct5110euw61fo8r5GmYqIiNdsCD/JGz9vZ92Bk9QrH8SEIa25tnZpf8fyPbcbtk6HxW/CiT1QKQS6fwY12mX8nPmLQNW2ztc55woax3dBviLnFSRKQt6CV/o3kWxIxQsRkWxmQ/hJpq2LZPgNNamRzhz42uWCePKWujxxcx3Wh8cwMzSKOZsO8fPmQxQrmJcShfISE5/E9AeuoXihHDYubMuPzjjTFvdCyyH+TiMiIjlAxIk43p6/k9kbD1ImKD9v9WxErxZVcl7x/3zWOrshFr0O0VugbAPoOxnqdPJND4kLFTQk11PxQkQkG3G7La/M2krZoPyM7FDL4+OMMbSoWoIWVUvw4m31Wb77GLNCD/J72FHe7tWE+hWL+jC1HxzaCDNGQHBb6Py2v9OIiEg2dyohmc8W72Hcin0EGHi4Qy3uv6EmhfPngl+n9v3uTOuKXAMla0DPsdCgBwTk8FtjJMvx6X9txphOwIc4XcG/staOOu/1QcA7/D2n/RNr7Ve+zCQikp1NWxfJxshYPrirKUUy+IEpb2AA7euWpX3dsl5Ol0WcPQaT73a2lfb+DvLksB0lIiKSaVJcbiavieB/v4Vx/GwSPZpX4qlb6lKhWC64bSFyHSx6DfYucfpG3P4RNO0HgTlslLpkGz4rXhhjAoFPgZuASGCNMWaWtXbbeUunWGtH+iqHiEhOERufzFvzdtCiagm6Na3o7zhZkysZpt7jjDYdPA+K5NACjYiI+JS1liU7j/Lm3O3sOnKG1tVL8k2X+jSqXMzf0XzH7YKjOyBitXOLSNgvTo+JW/4LIYMh76Unl4n4mi93XrQCdltr9wIYYyYD3YDzixciIuKBDxfs4kRcEt92bYXRjPILm/ccHFgBPcY4M+NFREQu0/ZDp3hz7naW7TpG9dKFGT2gBTfVL5fzfvbGn4TItU6xInI1RK2HxFPOa4XLQPsXoM1wZ0KISBbgy+JFJSAizeNIoPUF1vU0xlwPhAGPWWsjzl9gjBkGDAMIDg72QVQRkaxtV/Rpvv1zP31bBdOwUg6+6nMl1n0La8bA1Q9B497+TiMiItnMkdMJvP9rGFPXRhBUIC8v316fu1tXJV+eHNDbwe12dlVEroaINc6fx8Kc10wAlGsAje6EKq2gckunt0VOK9ZItufvDjOzgUnW2kRjzP3At0CH8xdZa0cDowFCQkIuMfBXRCTnsdbyyuytFM4XyJM31/V3nKwpfBX8/ATU7AAdX/V3GhERyUbcbsvnS/fw6eLdJLvc3HtNdR7qUCt7T+GKj3F2VUSudnZWRK37e1dFwZJOgaJxb6jcCiq1cKZ7iGRxvixeRAFV0jyuzN+NOQGw1h5P8/ArQC3hRUTOM3/rYVbsPs6rXRtQsnA2/iDlK7FRMKU/FK8CvcZBQKC/E4mISDbhdluenb6JqWsjuaVBOZ7rfBXVShf2d6yM2/kLLHrDGWcKzq6KsvWhYc/UXRWtoFRN7aqQbMmXxYs1QG1jTHWcokUfoF/aBcaYCtbaQ6kPuwLbfZhHRCTbSUh28fqc7dQrH8TdrXXb3L8kJziFi+Q4GDgbCpbwdyIREckmXG7LMz9uYtq6SB65sTaP3VTH35EyLjYKfnkadsyB0nWdfhVVWqbuqlDPCskZfFa8sNamGGNGAvNxRqWOs9ZuNca8Bqy11s4CHjbGdAVSgBPAIF/lERHJjr5cupeomHgmD2tDnsAccM+tN1kLsx+Bg+uhz0QoW8/fiUREJJtwuS1PT9vEj+sjebRjbR7tmE0LF24XrB7t7LZwp8CNL0HbhzQmXHIkn/a8sNbOBeae99xLab5/DnjOlxlERLKryJNxfLZkN7c1rkCbGqX8HSfrWfkZbJoM7Z+Hel38nUZERLIJl9vy1A8bmb4hisdvqsPDN9b2d6SMObjBKeIf2gg1b4Qu70HJ6v5OJeIz/m7YKSIiF/Hm3O0YA/+59Sp/R8l69iyGX1+Aq26H6570dxoREckmXG7LE1NDmRF6kCdvrsPIDtmwcJFwChb/n7PjonAZp99Tgx7qYyE5nooXIiJZ0B+7jzF382GeuKkOFYsX9Hcc74o7AfuXw6mo9NdeiNsFv78DZepB9y8gQLfTiIhI+lJcbh6fupFZGw/y1C11GdG+lr8jXR5rYfts+OUZOH0IWg6BDi9CweL+TiaSKVS8EBHJYpJdbl6ZvZUqJQsy9Poa/o5z5ZLjIXwl7F0C+5bCwVDgCqdeFynv9LnQaDcREfFAisvNo1NCmbPpEM90qscD7Wr6O9LliQmHuU9B2Dwo1xDuGg+VQ/ydSiRTqXghIpLFTFh5gLDoM4we0IICebPh2E+3Cw6FOsWKvUsgfBW4EiEgjzNXvt2zUKMdlK6T8S2ueQurGZmIiHgk2eXm0cmh/Lz5EM91rsf9N2SjwoUrGVZ+Dkv+6zy+6XVo8wAE5vVvLhE/UPFCRCQLOXYmkfd/C+O62qW5qX45f8fxjLVwfA/sXewUK/Yvg4RY57VyDaHlfU6xourV2ikhIiKZKtnl5pHJG5i7+TDP33pV9trRGLEG5jwK0VugTie49R0orrHpknupeCEikoW8O38n8UkuXr69ASYrN95yJcOOn2HXr07B4lz/imJV4KquTrGi+vVQpKwfQ4qISG6W7HLz0MQNzNt6mBe6XMV912WTwkV8DCx8DdaOg6AK0Hu806A6K38uEMkEKl6IiGQRmyJjmLI2gvuurU6tsll0h0L8SVj3Dawe4xQsChSHGjdA9SecgkXJGvpwJSIifpeU4uahSeuZvzWal26rz+Brs8EI0aQ42DHHmaZ19ii0Hg4dnof8Qf5OJpIlqHghIpIFuN2Wl2dtpVTh/Flz3vyx3bDqcwidCMlxzq6KLu9B7ZshIBv25RARkRwrKcXNiInr+W1bNK/cXp9B12TRwkVKIkSugX2/w75lELUWXElQoQn0mwIVm/k7oUiWouKFiEgW8NOGKDaEx/DunU0IKpBFmnBZ60wH+fMz2DUfAvNBozudRmHlG/k7nYiIyL8kprgY8f16Fmw/wmvdGnBP22r+jvS3lCQ4uN4pVOz/HSJWQ0oCmAAo3xha3w/Vb4Aa7SFQv6aJnE//VYiI+NnphGRGzdtBs+Di9GhWyd9xIDkBtkxzuptHb4FCpeGGZyBkCARlkyaiIiKS6ySmuHhwwnoW7jjC690aMMDfhQtXChza6BQq9i2D8D+d3YsA5RpByGCodp3T0Lpgcf9mFckGVLwQEfGTmLgkZm88yMTVERw7k8hX94QQEODHfhFnjsCasbB2rHOvbdn60PUTZ7dF3gL+yyUiIpKOhGQXD0xYx+KdR3mje0P6t6ma+SHcboje7BQq9v3uFCsSTzmvlakHTe+G6tdB1WuhcKnMzyeSzal4ISKSiVJcbpbtOsa0dZH8ti2aJJeb+hWK8n7vJjSp4qerLoe3OLssNk917rWtfTO0edBpwKnmm+JlxphywJtARWttZ2NMfaCttXasB8d2Aj4EAoGvrLWjznt9EPAOkDr+hk+stV95M7+IZD0JyS6GT1jHkp1HefOORvRrnUnjRFOS4OAGp0gR/ieEr4SEGOe1kjWhYU+nWFHtOk3fEvECFS9ERDLBrujTTFsfyU/rozhyOpGShfNxd5tgerWoTIOKxTI/kNvtjDld+alzdShPQWg2wOlnUToLNgyVnOQb4Gvg+dTHYcAU4JLFC2NMIPApcBMQCawxxsyy1m47b+kUa+1IryYWkSzJWsvGyFjenreDP/YcZ1SPRvRp5cPCRcIpp0/FuWJF1DqnZwVAqdrOONNq1zkFi6IVfZdDJJdS8UJExEdi45KZtekg09ZFsjEihjwBhvb1ytKrRWXa1y1LvjwBmR8q6awzMWTVF3B8NwRVhBtfhhaDoFDJzM8juVFpa+1UY8xzANbaFGOMy4PjWgG7rbV7AYwxk4FuwPnFCxHJ4U4lJDMz9CATV4Wz/dApCuUL5O1ejekdUsW7b3T6sFOkOJBarIjeAtYNJhAqNHZ6QVVtC1XaQJEy3n1vEfkXFS9ERLzI5bYs23WUaesi+XVbNEkpbuqVD+KFLlfRvVklShfJ759gsVGwejSs+8bZ0lqxOfQcC/W7QWAWmW4iucVZY0wpwAIYY9oAsR4cVwmISPM4Emh9gXU9jTHX4+zoeMxaG3GBNSKSzZzbZTFpVTizNh4kPtlF/QpFeaN7Q7o1rXjlk7qsheN7IPwP5/aPA3/AyX3Oa3kLQeUQuP5pCG4DlVtC/iJX/pcSkcui4oWIiBdEnIhj4upwpq+PJPpUIiUK5aVfq3O3hRTF+Kt3ROQ659aQrTMAC/Vug7YjoEpr9bMQf3kcmAXUNMasAMoAvbx07tnAJGttojHmfuBboMP5i4wxw4BhAMHBmXRvvIhkyOmEZGaEHmTSqnC2pe6y6Na0In1bBdO4cjHv/Hzd+QvMfQpiU2udBUtCcFtoOQSCr3Z2WajQL+J3Kl6IiFwBay1T10bwyqxtJLnctK9bhle7VqZ9vbLkzxPon1CuFNgxB1Z+BhGrIH9Rp5dFq6FQopp/MomkstauN8bcANQFDLDTWpvswaFRQNo94ZX5uzHnuXMfT/PwK+Dti2QYDYwGCAkJsZ6nF5HMYK1lU2Qsk1Y7uyzikry8y+KchFiY9xyEfg9lG8BtHzhjS0vXUYFfJAvyqHhhjKkK1LbWLjDGFATyWGtP+zaaiEjWdiohmf9M38ycTYe4plYp3unVhIrFC/ovUEIsrB8Pq76E2HAoXhU6jXJGsxUo6r9cImkYY+4576nmxhistd+lc+gaoLYxpjpO0aIP0O+8c1ew1h5KfdgV2O6NzCKSOc4kpjBjQxSTVoez9eApCuYNpGuTivRr7cVdFufsWQQzH4LTB+G6J+CGZyCPn27tFBGPpFu8MMYMxdlaWRKoiXOl4wvgRt9GExHJujaEn+ThyRs4GJPA053qMvz6mgQE+OkqzYl9TsFiw3hIOgNVr4FOb0LdWyHAT7s/RC6uZZrvC+B8nlgPXLJ4kdrYcyQwH2dU6jhr7VZjzGvAWmvtLOBhY0xXIAU4AQzyQX4R8bJNkTFMWh3OzFBnl8VVFYryeuoui6Le2mVxTuIZ+O0lWDvW2WExZAFUbuHd9xARn/Bk58UInA7fqwCstbuMMRpULCK5kttt+eL3Pbz/axjlixXgh+FtaR5cwj9hrIXl/4NFr4MJcObJt3kAKjbzTx4RD1hrH0r72BhTHJjs4bFzgbnnPfdSmu+fA57zQkwRyQSHYxMYMXE96w6cpGDeQG5vUoF+ravSxNu7LM7ZvwJmPggnD0DbkdDhBcjrxx2TInJZPCleJFprk879A2KMyUNqh3ARkdzkyKkEHp+6keW7j9GlcQXevKMRxQr6qYGXKwV+eQrWjoMGPeCWN6FoBf9kEbkyZ4Hq/g4hIplr79EzDBi7mtj4ZF7r1oDuzSp5f5fFOcnxsPB1pxdUiapw71ynt4WIZCueFC+WGmP+AxQ0xtwEPIjTzVtEJNdYsvMIT0zdyNmkFEb1aMRdLav4b4JI0lmYNgTCfoFrH4MOL0FAgH+yiFwmY8xs/r4IEgDUB6b6L5GIZLbNkbEM/Ho1Bpg8rA0NKxXz3ZtFroWfhsPxXdDyPuj4qsacimRTnhQvngWGAJuB+3G2a37ly1AiIllFUoqbd+bvYMyyfdQrH8Tkvm2oXS7If4HOHIWJveFQKNz6rjNBRCR7eTfN9ynAAWttpL/CiEjm+mP3MYZ+t5bihfIxfkgrapTxUSEhJRGWjIIVH0BQRRgwA2q29817iUimSLd4Ya11A2NSv0REco39x87y0KQNbI6KZUCbqjzf5SoK5PVjA8zje2BCDzgdDXdNgHpd/JdFJIOstUv9nUFE/OOXzYd4ZHIo1UsX5rshrShXtIBv3ujQJme3xZGt0LS/08S6gA93d4hIpvBk2sg+LtDjwlpbwyeJRESygJ82RPLCT1vIExjAlwNacEuD8v4NFLEaJt7lzJ0fOBuqtEz/GJEsxBhzmgv3zDKAtdZqnq9IDjZxVTjPz9hM8+ASjBvYkmKFfNDfwpXsNLJe+hYUKgV9p0DdTt5/HxHxC09uGwlJ830B4E6csakiIjnO2cQUXpy5henro2hVrSQf9GlKxeJ+7kS+fQ78OASCKkD/H6FUTf/mEckAa60f77cSEX+x1vLp4t28+2sY7euW4bO7W1Awnw92MR7Z7uy2OBQKDXvBre9AIf3KIpKTeHLbyPHznvrAGLMOeOlC60VEsqstUbE8NGkDB46f5ZEba/NQh1rkCfRzI8zVY2DuU1CpOfSbCoVL+zePiJekjl3/a8+4tTbcj3FExAfcbstrc7bxzR/7uaNZJd7u1Zi83v65mpIIKz+HxW9CvsJw5zfQ4A7vvoeIZAme3DbSPM3DAJydGJ7s2BARyTa2HTxFry/+oHjBfEwc2oY2NUr5N5DbDQtehj8+grq3Qs+xkK+QfzOJeIExpivwHlAROAJUBbYDDfyZS0S8K9nl5qkfNjIj9CBDrq3O87deRUCAF6d0ud2w5UdY9DrEHIC6XeD2D6BIWe+9h4hkKZ4UId5L830KsB/o7cnJjTGdgA+BQOAra+2oi6zrCUwDWlpr13pybhERb4mNT+aB79dRrGBeZo28hrK+aiDmqZREmPGA86EsZIiz9TXAj41CRbzrdaANsMBa28wY0x7o7+dMIuJFcUkpPPj9epbsPMrTnerywA01vTde3FrYsxAWvAKHN0O5Rs4tlTVvdPpCiUiO5cltIxmaKWSMCQQ+BW4CIoE1xphZ1tpt560LAh4BVmXkfUREroTbbXliaihRJ+OZcn8b/xcu4k/C5P5wYDl0fAWueVQfxiSnSbbWHjfGBBhjAqy1i40xH/g7lIh4R0xcEoO/WUNoRAyjejSiT6tg7508ap1TtNj3OxQPhh5jnP4WAX6+xVNEMsVFixfGmMcvdaC19v10zt0K2G2t3Zt6vslAN2DbeeteB94Cnko3rYiIl322ZDcLth/hldvr06Kqnxt7xUTA972ckag9voLGd/o3j4hvxBhjigC/A98bY44AZ/2cSUS84FBsPPeMXc2B43F8dndzOjWs4J0TH98DC1+DbTOcKSKd3oKQeyFPfu+cX0SyhUvtvLjSruCVgIg0jyOB1mkXpPbTqGKt/dkYc9HihTFmGDAMIDjYi9VbEcnVfg87ynu/hdGtaUUGXl3Nv2EObYLv74TkOBgwHapf7988Ir7TDYgHHgPuBooBr/k1kYhcsT1Hz3DP2NXExifzzeCWXF3TCw2mT0c7Y0/XfwuB+eH6p+Hqh6CAJiuL5EYXLV5Ya1/15RsbYwKA94FB6a211o4GRgOEhIRcaEa8iMhliTwZxyOTN1CnbBD/7dHIe/fiXq64E05viwWvOh/GBs+HcvX9k0Ukc9wPTLHWRgHf+juMiFy5TZExDPp6DQaYPKwNDSsVu7ITJpyCPz6GPz8BVxK0GOQULoLKeSOuiGRTnkwbKQAMwekCnnak2eB0Do0CqqR5XDn1uXOCgIbAktRfGsoDs4wxXdW0U0R8KSHZxYPfryfFZfm8f3MK5cvkAUpJZ2HnL7D5B9i9ANwpUKkF3DUBilbM3CwimS8I+NUYcwKYAvxgrY32cyYRyaAVu48x7Lu1lCicj/FDWlO9dOGMnywlEdaOg9/fgbjjzsjTDi9CqZreCywi2ZYnn9jHAzuAW3C2dd6NM9IsPWuA2saY6jhFiz5Av3MvWmtjgb/2kxljlgBPqnAhIr726uxtbIqM5csBLahRpkjmvKkrGfYsdgoWO36G5LMQVBHaPACNekP5RmrMKblC6s7OV40xjYG7gKXGmEhrbUc/RxORy/TzpkM8NiWU6qUL892QVpTLaNNrtxu2TEsdexru3DrZ8VWo1Ny7gUUkW/OkeFHLWnunMaabtfZbY8xEYFl6B1lrU4wxI4H5OKNSx1lrtxpjXgPWWmtnXVl0EZHLN3VtBJNWh/NAu5rc0qC8b9/M7YbI1U7BYutPzlWkAsWdRpyN7oTgq9UhXXKzI8Bh4DhQ1s9ZROQyJKW4effXnYz+fS8hVUswdmBLihXKm7GT7VkEv74E0ZudQn7/6VCzgwr6IvIvnhQvklP/jDHGNMT5oOHRhwxr7Vxg7nnPvXSRte08OaeISEZtiYrlhRlbuKZWKZ64qY7v3ih6G2yeCpt/hNhwyFMQ6nZ2Cha1blR3dMnVjDEPAr2BMsAPwNDzx6iLSNYVfjyOhyZvYGNEDP3bBPNCl/oUyBt4+SdKPAPzn4P130HxqtBzLDTooaK+iFyUJ8WL0caYEsCLwCygSOr3IiLZRkxcEsMnrKNU4Xx81KcZeQK9/OEoJhw2T3O+jmwFEwg120OH56FeF8h/pQOcRHKMKsCj1tpQfwcRkcsze+NB/jN9Mxj4/O7mdG6UwVGokWth+lA4sQ+ufQzaPafCvoiky5PixdfWWhewFKjh4zwiIl7ndlsenRJK9KkEpt7fllJFvPQB6exx2PaTU7AI/9N5rkpruPVdqN8dipTxzvuI5CDW2uf8nUFELk98kovX5mxl0uoImgcX58M+zahSstDln8iVAsvehaVvOw2qB/0M1a7xfmARyZE8KV7sM8bMw+kIvshaq1GlIpKtfLRoF0t2HuX17g1pFlziyk6WeObvSSF7FjqTQsrUc7qhN+wJJat7J7SIiEgWsPPwaUZOXM/uo2d4sF1NHrupDnkzsnvx+B6YPgyi1kLju+DWd6DAFY5UFZFcxZPiRT3gNmAEMM4YMxuYbK1d7tNkIiJesHjnET5cuIsezSvRv3Vwxk7iSnYaim2aCjvnQnIcFK0MbUc4fSzKNVRjMRERyVGstUxeE8Ers7YSVCAP3w1uxXW1M7Cj0FrYMB5+eRYC8zi9LRr18n5gEcnx0i1eWGvjgKnA1NTeFx/i3EKSgc48IiKZJ+JEHI9ODqVe+aL8X/dGmMspMLjdELEydVLIDIg/AQVLOFeLGveGKm3UVEwkg4wxVYHa1toFxpiCQB5r7Wl/5xIRx6mEZJ6bvpmfNx3iutqlea93E8oGZWAM6tnjMPth2DHHGX/a/XMoVtn7gUUkV/Bk5wXGmBtwZrF3AtbidAkXEcmyEpJdDJ+wDre1fNG/OQXzeVhvPbzFKVhs+RFiIyBvIah7q7PDomYHyJPPt8FFcjhjzFBgGFASqAlUBr4AbvRnLhFxbAg/ycOTN3AwJoGnO9Vl+PU1CQjIwO7CXb/BzBEQfxJu/j9o86CK/iJyRdItXhhj9gMbcHZfPGWtPevrUCIiV8Jay4sztrD14CnGDgyhaqnC6R90eDPMHAmHQp1JIbVuhBtfcgoX+Yv4PrRI7jECaAWsArDW7jLGeDSCXUR8x+22jFm2l3fm76Rc0QJMvb8tLapmoE9UUhz89hKsGQNl60P/6VC+ofcDi0iu48nOi8bW2lM+TyIi4iWT10Tww7pIHupQixuvKnfpxW43/PkJLHrduS3k1nehwR1QuHTmhBXJfRKttUnnbuMyxuQBPGoGbozphHP7aiDwlbV21EXW9QSmAS2ttWu9klokBzt2JpEnpm5kadhROjcsz6gejSlWKO/ln+hgqNOU89hOaDPCuQiQNwO3m4iIXIAnPS9UuBCRbGNjRAwvz9zKdbVL82jHOpdeHBsJPw2H/cug3m1w+0dQuFTmBBXJvZYaY/4DFDTG3AQ8CMxO7yBjTCDwKXATEAmsMcbMstZuO29dEPAIqTs7ROTSVuw+xqNTQomNT+aN7g25u3Xw5fWIAopqm/YAACAASURBVHC7YMWHsPhNp/g/YAbUbO+bwCKSa3nU80JEJDsIjYhhxPfrKROUn4/6NCPwUvfobpkOcx51Zs53/RiaDdDEEJHM8QxwH7AZuB+YC3zlwXGtgN3W2r0AxpjJQDdg23nrXgfeAp7yVmCRnCjF5eaDBbv4dMluapQuzHeDW3FVhaKXf6KTB5wLAeF/QP1ucNsHUKik9wOLSK6n4oWIZHu7j5zmnfk7mb81mlKF8/HNva0oUfgijTUTTsHcp2DTZKjUAnqMgVI1MzewSC6Vuntiq7W2HjDmMg+vBESkeRwJtD7v/M2BKtban40xKl6IXITLbRn49WpW7D5O75DKvNK1AYXyXeavBW4XbJwM8551xqF2/wKa9NGFABHxmYv+K2WMefxSB1pr3/d+HBERz0XFxPPBb2H8uD6SQvny8PhNdRh8bXWK5L/IP20H/oSfhjm3i9zwDFz/FARm4J5eEckQa63LGLPTGBNsrQ335rmNMQHA+8AgD9YOw5l4QnBwsDdjiGQL09dHsmL3cV7t2oCBV1e7vIOTE2DjJPjjIzix1xkd3uNLKHGZ5xERuUyXKrEGZVoKEZHLcOJsEp8u3s34Pw+AgcHXVOfB9rUoebHdFq5kWDIKlr8PxYNh8Hyo0ipzQ4vIOSWArcaY1cBfE8ystV3TOS4KqJLmceXU584JAhoCS1Lv1y8PzDLGdD2/aae1djQwGiAkJMSjZqEiOUVCsov3fg2jSeVi3NO26mUcGAtrxsLKz+HsEajYHHp/5/SMCvBwHLmIyBW4aPHCWvtqZgYREUnPmcQUxi7bx5hle4lLSqFXi8o80rEOlYoXvPhBx3bD9Pvg4AZo2h86j4L8qs2K+NGLGTxuDVDbGFMdp2jRB+h37kVrbSzw15ggY8wS4ElNGxH5p7HL93H4VAIf9mnqWWPOU4dg5Wew9mtIOg01O8A1j0L163WLiIhkqkvdNvLRpQ601j7s/TgiIv+WmOJi4qpwPlm0m+Nnk+jUoDxP3lKHWmUvUYSwFtZ9A/P/A3nyw53fQoPumZZZRC7MWrvUGFMOaJn61Gpr7REPjksxxowE5uOMSh1nrd1qjHkNWGutneW71CI5w/EziXy+ZA8drypH6xrpTNc6thv++NDpa+FOccaIX/MIVGiSOWFFRM5zqdtG1mVaChGRC3C5LT9tiOJ/v4URFRNP2xqleKZzPZpWKX7pA88eg1kPwc65UKMddP8cilbMjMgikg5jTG/gHWAJYICPjTFPWWunpXestXYuznSStM+9dJG17a44rEgO8/Gi3cQnu3i2c92LL4pcByv+B9vnOMX/ZgPg6pFQskbmBRURuYBL3TbybWYGERE5x1rLb9uieffXnYRFn6FRpWKM6tmIa2uVTn+La9ivMHOEc2/uLf+F1sMhICBzgouIJ54HWp7bbWGMKQMsANItXohIxu0/dpYJKw9wV8sq/965aC3sWQjLP4D9y6BAMbjucednaJGy/gksInKedGcipX6oeAaoDxQ497y1toMPc4lILrX7yBmemraRDeEx1ChdmM/ubk7nhuXTL1pEb4NFrzu7Lco2gHtmQLkGmRNaRC5HwHm3iRwHVGEU8bG35+8gX54AHu1Y++8nXSmwbQas+AAOb4agCnDzG9BikPpDiUiW48lA5++BKUAXYDgwEDjqy1AikjvtO3aWvmNW4nZbRvVoRK8WlckTmM7vNCf2wZL/wqapzgetDi9A24cgb4FLHyci/jLPGDMfmJT6+C7gFz/mEcnx1oefZO7mwzzasTZlgwpA3AnY/AP8+SnEHIBStaHrJ9C4t3OriIhIFuRJ8aKUtXasMeYRa+1SYKkxZo2vg4lI7hJ5Mo67x6zE5bZMHtaGOuXSueJzOhp+f8dpyhkQCNc87HQ/L1QyU/KKSMZYa58yxvQArk19arS19id/ZhLJyay1/HfudsoXCWR4+Z0w5XUImweuJKgUAre8CXVv1S2WIpLleVK8SE7985AxpgtwENBvByLiNYdjE+g3ZhVnElOYODSdwkV8DPzxkTNnPiURmt8DNzythpwi2UTqqNO51trpqY8LGmOqWWv3+zeZSA5kLX+uWEznyC/pV2g1BaadgEKloeV90KQvlG+kcacikm14Urx4wxhTDHgC+BgoCjzm01QikmscPZ1Iv69WcvxMIhPua03DSsUuvDApDlaPhuX/g4QYaNgT2j8PpWpmbmARuVI/AFeneexKfa7lhZeLyGU7fRg2/4ANncjVR7YRkicPeWrcCs36Qa2OEJjX3wlFRC5busULa+2c1G9jgfa+jSMiuUlMXBIDxq7iYEw8397bimbBJf69yJUMG8bDkrfgzGGodRPc+KLmzItkX3mstUnnHlhrk4wx+fwZSCRHSE6AnT9D6CRncoh1c6xYIz5Mvpcbez1A+2aXGI8qIpINeDJt5FvgEWttTOrjEsB71trBvg4nIjnXqYRk7hm3mr3HzjJ2YAita5T65wK3G7ZOh8X/Byf2QpXW0GscVLvGP4FFxFuOGmO6WmtnARhjugHH/JxJJHuyFiJWQehE2DoDEmOhaCW45lHO1r+TzuOiqBFchNeb1vF3UhGRK+bJbSONzxUuAKy1J40xzXyYSURyuLOJKdz79Rq2HTzFlwNacF3tMn+/aC3s+g0WvgbRm6FcQ+g7BercovtyRXKG4cD3xphPAANEAPf4N5JINhMb6RQsNk5yCvx5C8FVXaFpX6h2HQQE8uWvOzl2JomvBl6V/rhxEZFswJPiRYAxpoS19iSAMaakh8eJiPxLQrKL+75dy4bwk3zSrzk3XlXunwsWvOLMmy9RDXp85fS2UAd0kRzDWrsHaGOMKZL6+IyfI4lkL4c2wje3QeIpp1Bx3ZNQv6szLjxV9KkExizbx22NK9C0SnE/hhUR8R5PihDvAX8aY35IfXwn8H++iyQiOVViiov7x69j5b7jvN+7Cbc2qvDPBavHOIWLFoOg8zuQR7fBi+Q0xphHgK+B08AYY0xz4Flr7a/+TSaSDRzfAxN6QoFiMHQxlK51wWX/+y2MFLebp2+pl8kBRUR8J93Lmdba74AeQHTqVw9r7XhfBxORnCXZ5ebhSRtYGnaUN+9oxB3NKv9zwc5f4JenoU4nuPU9FS5Ecq7B1tpTwM1AKWAAMMq/kUSygdOHYfwdYN0w4KeLFi7Cok8zdW0EA9pUI7hUoUwOKSLiO57uxS4JnLXWfoLTaKu6JwcZYzoZY3YaY3YbY569wOvDjTGbjTGhxpjlxpj6l5FdRLIJl9vy+NSNzN8azcu316dvq+B/LohaD9MGQ/nGTlPOQN2ZJpKDnbv5/lbgO2vt1jTPiciFxMfA+B4Qdxzungala1906ahfdlA4fx4e6nDh4oaISHaVbvHCGPMy8AzwXOpTeYEJHhwXCHwKdAbqA30vUJyYaK1tZK1tCrwNvH8Z2UUkG3C7Lc/8uInZGw/yTKd63HvNebXPkwdg4l1QqDT0mwr5CvsnqIhklnXGmF9xihfzjTFBgNvPmUSyruR4mNQHjoXBXROgUvOLLv1jzzEW7TjCiPa1KFFYOxhFJGfx5PLmHUAzYD2AtfZg6geN9LQCdltr9wIYYyYD3YBt5xakbhs9pzBgPcwtItmAtZaXZ21l2rpIHr6xNg+0q/nPBfEn4fs7wZUIg+ZAULkLn0hEcpIhQFNgr7U2zhhTCrjXz5lEsiZXCvxwL4SvdHYm1mx/0aVut+W/c3dQsVgBBl1dLfMyiohkEk+KF0nWWmuMsQDGGE8vi1bCGX92TiTQ+vxFxpgRwONAPqDDhU5kjBkGDAMIDg6+0BIRyWKstbw5dzvjVx7g/utr8FjH87a4piTC5P7OiLd7ZkCZuv4JKiKZylrrJvWCSOrj48Bx/yUSyaKshVkPQdgv0OU9aNjjkstnbzrI5qhY3u/dhAJ5AzMppIhI5vGk58VUY8yXQHFjzFBgATDGWwGstZ9aa2vi3JrywkXWjLbWhlhrQ8qUKeOttxYRH/rfb2GMWbaPe9pW5dnO9f45Y97thpkj4MBy6P45VLvWf0FFRESyot9ego0Tod1/oOV9l1yamOLinfk7qV+hKN2bVsqkgCIimeuSOy+M89vGFKAecAqoC7xkrf3Ng3NHAVXSPK6c+tzFTAY+9+C8IpKFpbjcvPdbGJ8v2UPvkMq8cnuDfxYuABa/AZt/gA4vQuM7/RNUREQkq1rxIfzxEbQcCjc8ne7y8X8eIPJkPBOGNCYgQP1vRSRnumTxIvV2kbnW2kaAJwWLtNYAtVMnk0QBfYB+aRcYY2pba3elPuwC7EJEsq2omHgenbyBNftP0rdVFd7o3ujfH6LWfQPL3oPmA+G6J/ySU0SyFmNMEWvtGX/nEMkSNnzv7Lpo0AM6vw3nXwA4T2xcMh8v2s31dcpwbe3SmRRSRCTzedLzYr0xpqW1ds3lnNham2KMGQnMBwKBcdbarcaY14C11tpZwEhjTEcgGTgJDLzM/CKSRczbcphnftyEy235sE9Tul1o2+qu32DO41CrI3R5P90PZCKSa2wD1NRKZMdcp89FzQ5wx5cQkP4d3p8u2c2phGSe61wvEwKKiPiPJ8WL1sDdxpgDwFmcWezWWts4vQOttXOBuec991Ka7x+5vLgiktUkJLt44+dtTFgZTuPKxfi4bzOqlrpAX99DG+GHQVCuPtz5DQR68s+PiOQUxpjHL/YSUCQzs4hkSQf+gGn3QsWm0Hs85El/1GnEiTi+WbGfns0rc1WFopkQUkTEfzz57eEWn6cQkWwpLPo0D03cwM7o09x/fQ2euLku+fJc4CpRTAR83xsKFId+P0B+T6Yti0gO8ybwDpBygdc8aSAuknMd3gwT+0Dx4NSfk57V8977dSfGwBM31/FxQBER/0u3eGGtPZAZQUQk+7DWMml1BK/N2UqR/Hn4dnArbqhzkUlACbEwsTckx8Hg+VC0QuaGFZGsYj0ww1q77vwXjDGXHqUgkpOd2AcTejoFi/7ToXApjw7bEhXLjNCDPNiuJhWKFfRxSBER/9O+bRG5LLHxyTw3fRNzNx/mutqlea93E8oGFbjw4pQkmDIAjoVB/x+dW0ZEJLe6Fzh+kddCMjOISJZx5giMvwNcSXDPLCheJf1jcC4ivDl3OyUL52N4u5o+DikikjVom6aIeGzdgRPc+uEyft0azbOd6/Htva0uXriwFmY/DPuWQtePoUa7zIwqIlnPC9baY8aYf/W7stZG+yOQiF8lxMKEHnAm2rlVpKznDTeXhB3ljz3HebhDLYoWyOvDkCIiWYeKFyKSLpfb8smiXfT+ciUBAfDD8LYMv6HmpWfJLxkFGydBu/9A034XXyciuUULY0xFYLAxpoQxpmTaL09OYIzpZIzZaYzZbYx59gKvDzfGbDbGhBpjlhtjtN1LsqbkBJjUD45sd5pzVmnp0WFJKW6mr4/khZ+2UK1UIfq1rurjoCIiWYduGxGRS4o+lcCjk0P5c+9xujapyBt3NLz0VR5rYeXnsHQUNL0bbng688KKSFb2BbAQqAGsw5kyco5Nff6ijDGBwKfATUAksMYYM8tauy3NsonW2i9S13cF3gc6ee1vIOINrhT4cQgcWA49x0LtjukeEhufzKTV4Xy9Yh/RpxKpU64I/+3R6MJNskVEcigVL0TkohZuj+bJHzaSkOzm7V6NubNFZYy5xG6LM0dh1kgImwd1OsPtH8Kl1otIrmGt/Qj4yBjzubX2gQycohWw21q7F8AYMxnoBvxVvLDWnkqzvjBOUUQk63C7Yc4jsGMOdHoLGvW65PKIE3GMW7GPqWsiOJvk4tpapXm7VxOur1360j+PRURyIBUvROSC3v8tjI8W7uKqCkX5uG8zapVNZ2xb2HyYOQISTkGnUdDqfgjQFSER+acMFi4AKgERaR5HAq3PX2SMGQE8DuQDOlzoRMaYYcAwgODg4AzGEblMrmSY8QBs/gFueAbaDL/o0tCIGMYs28svmw8RYAxdm1RkyHXVaVCxWCYGFhHJWlS8EJF/2RIVy0cLd9G9aUVG9WxMgbyBF1+cFAe/vgBrx0LZBnDPTCjXIPPCioikYa39FPjUGNMPeAEYeIE1o4HRACEhIdqdIb6XHA8/DHJ2Jt74Mlz3+L+WuN2WBduj+WrZPlbvP0FQgTwMvb4Gg66uplGoIiKoeCEiFzDqlx2UKJSX17o3vHTh4mAoTB/qjEJtOxI6vAh5LzJ9RETkykQBaedIVk597mImA5/7NJGIJxJPw6S+sH85dHkfWg75x8vxSS5+XB/JuOX72HvsLJWKF+TF2+pzV8sqFMmvj+oiIufoX0QR+Yffw46yfPcxXryt/sUbc7pdsOJDWPx/ULiss9uiRrvMjCkiuc8aoLYxpjpO0aIP8I9RRsaY2tbaXakPuwC7EPGnuBMwoScc2gg9xkDjO/966diZRL778wDj/9zPybhkGlcuxsd9m9G5YXnyBOq2SxGR86l4ISJ/cbvt/7d339FVVGsfx787hSQQakJoIRSldwglgiACChZAmooIelHsr9fewXaveMV+RUUsgIoCAoIiUgQEpTfphBIgoYYQQhLS9/vHidwASSBwSsrvs1YW58zsmfOczazMzjO78MYv26lZKYAhHfIYBx6/H2bcD/v+gMZ94Kb3oPRFrXIoInLJrLUZxpiHgV8Bb+ALa+0WY8yrwBpr7SzgYWNMdyAdOEEuQ0akhMrKggk3g81yTCZdub7rPzPhIEy6BRu3l/QBk4gP60bC0UTiktKYsT6aH9bFkJaRRfdGVbj36jq0q1NJk3CKiORDyQsROWPmhhi2HUrg/dta4ueTy3CRv6bCz0+AzYS+H0OL27WaiIi4jbV2DjDnnG0jc7x+1O1BSdHw1/eOpUl9AuCTTnDtixDxEHjlMzQyF3FJaew+lkjC6XQSUtI5lZKR/drx76mUDBJS0glI3M/L8S9Q3p7kvoynWTYJHCsFO/j5eDGgTSjDO9XhisoXmBBbREQAJS9EJFtKeiZvz9tJsxrlubl59bN3no53JC02T4Oa7eGWT6FSHc8EKiIiUhBpyfDba1C9Ndw+2XE/m/8SbJsFfcZedC+MNVFx/OOr1SSkZJy3z8/Hi3IBvpT196GpTwyvJLxIKZPBV1d8SJOgZkT4+1IuwJdy/j6U8/eleWh5ggL9nP1NRUSKNSUvRASAicujiIk/zVsDmuPllaM3RdQymH4fnDoEXV+ETo+Bt351iIhIEbFiLCTEOOacKFsVbv0aNv8Ac5509MLo+jxc9Ui+vTAW7TjKA1+vpWo5f967rSWVyvhRzt+Hsv6OhMWZya2j18I3L0CAHwydw0Mhjdz0JUVEij/9BSIixCen8d/fdnFNg8pcdWWwY2NGmmNCzj/eh0p1Yfh8CG3j2UBFREQKIvEoLHsXGt4EtTs6thkDzQZAnc7w02OwYBRsmw19x0LlBued4scNMTwxZSP1q5Rlwj/aUblsHj0m9v7uWFWkTDDcOVM9FEVEnEzJCxFh7OLdnErN4PkuIbBlBuxZDLsWwskD0HoYXP9v8NOYXBERKWIWj4aMFOj+yvn7AkMcvTC2TIefn4RProauz0HEI2d6GE5aHsXIWVtoW7sS44eF570K145fYMowR7L/zhlQrprrvpOISAml5IVISZaWzLGti6m84luWlt9O6MRdgIVSZaF2J7hhDDTo6ekoRURECu7YDlj7FbQdDsFX5l7GGGjaH2pfDT8/Dgtehq2zsH3H8uEmH96Zv5PujUL47+DW/xsacq6/pjhW4arWAob8oBW4RERcRMkLkZIkMwMOroe9i2HPEjiwksqZaQwz3tigdlB/ENS9xjGpmea1EBGRomz+KChVBro8c+GygSEwaBJsmYGd8ySZH1/N6bT+DGh5L28MbI2vt1fux636DOY85Uj43z4Z/Mo69zuIiMgZ+utEpDizFmJ3OoaB7FkCUUshNcGxr2ozYpvcxRNrKtLsqp48eVNrj4YqIiLiNHt/h52/QPeXHXNQXAxjSG/Ul9c2VqTD9n/zjO932IQdmOMfQ24Tby59Gxa+CvV7wcAvwTfAmd9ARETOoeSFSHEVGwmTb4PjuxzvK4RBk1ugbheo0wXKBPPY5yvZ5H+SD7o182ysIiIizpKVBfNehPI1of39F31YSnomD3+7jgXbkgnuMRZbZTNmzhPwaWdH742O/3T0SrTWMbzkj/eg2UDo+zF45zEXhoiIOI2SFyLF0ZGtMLEPYOGmd6Fu1/NmPV8aeYylkbG8eGMjygeo0SUiIsXEpqlwaKNjadSL7A2RkJLOPRPWsDoqjtf6NOHOiNpAPcdwkDlPwm+vOVYk6fMRrB4Pa7+E8OGOuaG88hhSIiIiTqXkhUhxc2gjTOwLPn4wdBZUrn9ekawsy+hfthNaMYA7I2p5IEgREREXSD/tGMpRrSU0HXBRh8QmpjLsi1XsOHyK925tSZ+WNf63M7AyDJrgWInr5yfhk+zlVjs9Dt1GOib8FBERt1DyQqQ4iVkLk24Bv3IwbJZjybZczNp4kC0HE3jv1pb4+eQxe7qIiEhRs+JjSIiGWz65qB4RB+KSGfrFKg6dPM34YeFc0yAk94JNbnGsSLLwVajSBNrf5+TARUTkQpS8ECku9q+ArwdAmSAYNtsxx0UuUtIzeevXHTSpXo7eLaq7OUgREREXSYqFpe9AgxugztUXLB555BR3fr6K5LQMvh7envDaF1jitEww9P7AScGKiEhBaZCeSHGwdylM6gdlq8Bdc/JMXAB8vWIfMfGnea5XI7y81N1VRESKiSVvQnoydH/lgkU3HIhn4KfLybSW7++LuHDiQkREPE49L0SKul0L4Ls7oGJtxxwXZavkWfRkcjof/raLzvUr06neRS4dJyIiUtjFRsKaLyD87lznesppWWQsIyatITjQj0nD21ErqIybghQRkcuh5IVIUbZjLky5E4IbwNCZF1zLfuySXSSkpPNsz4ZuClBERMQNFrwMPgHQ5dl8i/2y6RCPfreBupXLMPEf7Qgp5++e+ERE5LK5dNiIMaanMWaHMWaXMea8u4kx5nFjzFZjzF/GmIXGGC17IHKxtv4I39/hmDhs2KwLJi5i4k/z5R9R3NKqBo2rl3NTkCIiIi4W9Qds/wk6/dOxOkgevlm5j4e+XUez0PJ8PyJCiQsRkSLGZckLY4w38BHQC2gM3G6MaXxOsfVAuLW2OTAN+I+r4hEpVjZNg6l3Q/XWMPRHKH3hsbrvzNsJwBPXNXB1dCIiIu6RlQXzXoRyNaDDg7kWSUrN4PHvN/DCjM10rl+ZScPbUb60r5sDFRGRy+XKYSPtgF3W2j0AxpjvgD7A1r8LWGsX5Si/AhjiwnhEiof138CPD0GtjjD4O/Are8FDth1KYPr6aEZcXZcaFQLcEKSIiIgbbJkOB9dB30+gVOnzdm89mMDDk9cRFZvEY93r8/C1V+KtyapFRIokVyYvagAHcryPBtrnU3448EtuO4wxI4ARAGFhea+iIFLsrfkSfvon1O0Kt32ba0MtN6N/2U45f18evOZKFwcoIiLiJukpsOAVqNoMmt961i5rLd+u2s8rs7dSIcCXb+7pQMQVQR4KVEREnKFQTNhpjBkChANdcttvrR0HjAMIDw+3bgxNpPBY8QnMfQbqXQeDJoHvxY3V/WNXLEt2HuP5Gxqqm6yIiBQfqz6Fk/uhzyzw+t9I6ISUdJ6bvomf/zpEl/qVeXtQC4ID/TwYqIiIOIMrkxcxQM0c70Ozt53FGNMdeAHoYq1NdWE8IkXXH+/D/JHQ8CYY8CX4lLqow7KyLG/8so0aFQIYGlHbtTGKiIi4S9Jx+P1tqHc91P3fs6+/ouN5+Nv1xMSf5pmeDbmvc128NExERKRYcGXyYjVQzxhTB0fS4jZgcM4CxphWwKdAT2vtURfGIlJ0LfkPLPoXNOkH/caB98X3npj910E2xyTwzqAW+Pt6uzBIERERN/r9P5B2Cnq8CjiGiXz5RxRv/LKNyoF+TLmvA21qXXgyaxERKTpclryw1mYYYx4GfgW8gS+stVuMMa8Ca6y1s4C3gEBgqjEGYL+1trerYhIpUrKyYN4LsGIstLgd+nwEXheXgEjLyGL+1iOM/mU7jaqVo2/LGi4OVkRExE2O74bV46H1MAhpSHxyGk9N+4v5W4/QvVEVxgxsToXSF9dDUUREig6XznlhrZ0DzDln28gcr7u78vNFiqz00zB9BGybBe3vh+vfOGs8b172xibx3ar9TFsbzfGkNGpUCODftzRVl1kRESk+FowCH3+45jnW7jvB/01ez9FTKYy8qTF3d6xN9gMxEREpZgrFhJ0ikkNyHEy+HQ6sgOv/DREP5Vs8NSOTX7ccYfLK/SzfcxxvL0P3RiHc3i6Mq+tV1pJwIlJsGGN6Au/j6NE53lo7+pz9jwP3ABnAMeAf1tp9bg9UzhazDk7shbAIKFf98s61bzlsm03WNS/w6dpExszbQY0KAfzwwFU0D63gnHhFRKRQUvJCpDA5EQVfD4D4/TDwK2hyS55Fdx9LPNPL4kRyOqEVA3jq+gYMbBNKSLmLW4lERKSoMMZ4Ax8BPXAsv77aGDPLWrs1R7H1QLi1NtkY8wDwH+DW888mbpNyEr7uD6fjHO8r1IJaV0FYBwi7CoLrwcX2lLAW5r1AVmBV7tvVgfm7tnNjs2q80b8Z5fy1mpaISHGn5IWIi/2xK5bJq/ZTpZw/YZVKE1apNDUrlSa0YsDZk2geXA/fDILMNBg609G4O0dKeiZzNx9m8qr9rNwbh4+XoUfjKtzeLoxOVwZreIiIFGftgF3W2j0AxpjvgD7AmeSFtXZRjvIrgCFujVDO9+eHjsRF/88h8Sjs/xMi58PGyY79pYMcPTLCIqBWBFRtAd55NE+3TIeYtbzu/RBL4pP41y1NGdwuTMNERERKCCUvRC7Gnx/C4c3Q+UnH7XVBeQAAIABJREFUU6KL9MeuWO7+ajWlS3mTkp5JSnrWWfurZic0evhuZFjMy6T7VSSq1zdUrtScytaeaZBFHjnF5FUHmL4+mvjkdMIqleaZng0Z0CaUymW1dr2IlAg1gAM53kcD7fMpPxz4JbcdxpgRwAiAsLAwZ8Un5zp1BJZ/5Fgtq9kAx7aIBx09KI7vgv3LHcNA9v8J239y7PctA6Hh2b0zIhyvS5UhMy2FxNkvEZNVi98DuzFzeFsaVy/nue8mIiJup+SFyIXErIP5I8Fmwaap0OoO6PIslM9/BY81UXHcM2ENdYLK8N2IDlQo7cuxxFQOxCWzPy6Z/cdPsz8umXrRP3B3wodsywrjH8lPcWzyMWAh/r5ehFUqTSkfLzbHJODrbbiuSVUGtwsjom6QelmIiOTBGDMECAe65LbfWjsOGAcQHh5u3RhaybLkTUdvwmtfPHu7MY4HAcH1oPVQx7aEQ45kxt8/i0cDFrx8SA9pzpZ4b1qmxvBt7bf58Y4ulPFTE1ZEpKTRb36R/GRmwOz/gzIhcPccWDUO1nwBG7+HdvdCp8ehTNB5h208EM9dX66mWnl/Jt3TjoplHEu2hZT1J6Ssv2PteWth8Ruw9X24sjv1+n7O5GSf/yU3sn9OJKXxXK+G9G8TSnCgelmISIkVA9TM8T40e9tZjDHdgReALtbaVDfFJuc6vhvWTXAsZxp0xYXLl6sGTfs5fsAxV8aBVRze9BuHNi2mcVYkMVW78sA/7nFt3CIiUmgpeSGSn5Ufw+FNMHCCo/HV603o8KDjidCKsbB2Alz1iKMbrF9ZALYeTGDoF6uoWMaXb+5tT0jZXCbPzEyH2Y/Chm+g1Z1w07v4e/tyZSBcGRLo5i8pIlIkrAbqGWPq4Eha3AYMzlnAGNMK+BToaa096v4Q5YzfXgPvUtDlmUs63PqVY1JsPV5bm0618tfxye1NaVzj/IcFIiJScnh5OgCRQuvEPlj0b6jfCxr3+d/2irXglo/hgeVQtwss/je83xJWfMyuQ7Hc+flKSpfy5tt7OlCtfMD5501JgG8GOhIX1zwPvT8Eb82SLiKSH2ttBvAw8CuwDZhird1ijHnVGNM7u9hbQCAw1RizwRgzy0PhlmwH18OWGY6lvstWKfDhyWkZPD5lIyN/3ELnepWZ/XAnGtesDF5qtoqIlGTqeSGSG2vh5ycAAze8lfsybiEN4bZvIHoNLHgZ5j5LGcZws9etDP3HM9SsVPr8YxIOORIXx7ZBn4+glSbCFxG5WNbaOcCcc7aNzPG6u9uDkvMteBkCKsFV/1fgQ/fGJnH/pLXsPHqKJ3rU56GuV2qOJxERAdTzQiR3W6bDrvmOScYq1My/bGg40b2/55++ozhBOV7O+oi6U3vAttmOJMjfjm6D8d3hxF4Y/L0SFyIiUvzs/g32LIbOT4F/wVYD+XXLYXp/uIyjp1KYcHc7HulWT4kLERE5Qz0vRM51+gT88gxUbwXt77tg8SMJKdzx+Sri0ptwzz2LIGEJ/PY6fD8EarSBbqPAyxu+Gww+AY6JP6u1cMMXERERcaOsLEevi/Jh0Hb4RR+WkZnFmHk7+WTJbpqHlmfsHa0JrZhL70URESnRlLwQOdf8UZAcB0N+cCQd8hGbmMrgz1YQeyqVSfe0p2loBaAPNLgRNk52TOw5sTcYLwiqB0OmQYUw93wPERERd9oyHQ5thFs+BZ+LWx0rNjGV/5u8nj93H2dw+zBG3dwYP5/8770iIlIyKXkhktO+Px1Lu0U8fMHeEfHJaQwZv5KY+NN8dXc7WodV/N9Obx9ofSc0G+hYWvXIZrj+XxBQMe8TioiIFFUZaY5ehyFNHPe+i7B23wke+mYdJ5LTGDOwBQPahLo4SBERKcqUvBD5W0YqzP6no7tr1+fzLZqQks7QL1ax51gS44eF06FuHsu3+fo7llEVEREpztZNyJ7TaeoFey1aa5m4fB+v/7yVauUDmP7gVTSpXt5NgYqISFGl5IXI35a9B7E74I5pUKpMnsWS0zL4x5er2XowgU+GtKFz/cpuDFJERKSQSU2EJW9CrY5Qr0e+RZPTMnh++iZmbjhIt4YhvDOoJeVLa7lwERG5MCUvRABiI2HpGGjSL9+GV0p6JvdMWMO6/Sf48PbWdG9c8PXrRUREipUVYyHpGNw2OfelxbPtOZbIA1+v0zKoIiJySZS8ELHWMVzENwB6js6zWFpGFg98vZble47z9sAW3Ni8mhuDFBERKYSSYuGP96HhTVCzbZ7F5m4+zFNTN+LjbZhwdzv1WhQRkQJT8kJk/dewbxnc/D6Uzb0nRUZmFo9MXseiHcf49y3N6Ndak4qJiIjw+xhIT4ZuI/Ms8umS3bzxy3YtgyoiIpdFyQsp2RKPwbwXISwCWg3NtcieY4mMmrWFpZGxjLypMYPba6lTERERTkTB6vHQaghUbpBrkfjkNN5dsJPujUL46I7WWgZVREQumZIXUrL9+jykJTl6XXh5nbXr5Ol0PlwYyYTlUfj5ePN636YM6VDLM3GKiIgUNov+7VhZ5Jrn8iwydU00KelZPNajvhIXIiJyWZS8kJJr1wLYNAW6PHPWE6PMLMv3qw/w9rwdxCWnMahNTZ68vgGVy/p5MFgREZFC5PAm+GsKdPw/KFc91yJZWZZJK/bRtnZFLYUqIiKXTckLKZnSkuGnxyHoSuj0+JnNy3cf59WftrLtUALtaldiws2NaVpDDS4REZGzLHgF/MtBp8fyLLJk5zH2xyXz1PW5DykREREpCCUvpGRa8ibE74O7fgZff/YfT+bfc7Yxd8thalQI4KPBrbmhWVVMPku+iYiIlEh7l8Ku+dD9FQiomGexCcujCCnrx/VNqrovNhERKbaUvJCS5/Am+PNDaDWExGodGDt3O+OX7sXby/BEj/rc27ku/r4alysiInIea2HBKChbHdrfl2exqNgkFu84xj+716OUj1ee5URERC6WkhdSdFgLe5eAX1mo3AhKXcJSa1mZMPtRbEBFfqz8AP8as5hjp1Lp16oGT/dsSNXy/s6PW0REpLjYNhti1kLvD8E3IM9ik1bsw8fLMLidVugSERHnUPJCio6N38HM+7PfGKhUB0IaQ5Um2f82dWzzyqfXxOrPIWYtbwc+xX9n7aNlzQqMu7MNrcLy7vYqIiIiQGYGLHwVgutDi8F5FktOy2DKmgP0alaNkHJ6KCAiIs6h5IUUDXF7Yc6TEHYVRDwIR7bCkc1wdCvsmAM2y1HOJ8CxcsiZhEZjCGkCgSEcjt5DhbkjWZXZjKmp7Xn31kb0aVEDLy/NayEiInJBG76G45Fw6zfgnXcTcub6g5xKyWBYhJYXFxER51HyQgq/zAyYPgKMN/QbBxVqQqOb/7c//TQc2+5IaBzdCke2OJZB3fDNmSKnfSuQleaDJZPItq+yqGdnSpfS5S8iInJR0pJh8WgIbQcNb8yzmLWWicujaFytHG1qqVejiIg4j0v/ejPG9ATeB7yB8dba0efs7wy8BzQHbrPWTnNlPFJELR0D0aug/+eOxMW5fAOgeivHT05JsezYtJLfliyi4qlIwsscxb/jMwzvfK174hYRESkuVn4Cpw457sX5rMS1am8c2w+f4s3+zbRil4iIOJXLkhfGGG/gI6AHEA2sNsbMstZuzVFsP3AX8KSr4pAi7sAqWPIfaH4rNBtw0YfFJ6fx5q+HmLwKqpfvxcu3P8mVWqpNRESk4JLjYNl7UO96qN0x36ITl++jfIAvvVvUcFNwIiJSUriy50U7YJe1dg+AMeY7oA9wJnlhrY3K3pflwjikqEo9BdPvhfI14Ia3LuoQay0z1sfwr5+3EX86nXs61eGxHvUp46chIiIiIpdk2TuQmgDdR+Vb7PDJFH7dcpi7O9YmoJSWHBcREedy5V90NYADOd5HA+0v5UTGmBHACICwMC25VWL88gzE74e7fwH/8hcsvvtYIi/N3Myfu4/TsmYFJt7SlCbVL3yciIiI5OFkNKwcBy1uc0yGnY9vV+0n01qGdNBEnSIi4nxF4nG0tXYcMA4gPDzcejgccYctMxwTbnZ+GsI65Fs0JT2TsYt388ni3fj5evF636YMbhemVUREREQu1+I3AAtdn8+3WFpGFt+u3E/XBiHUCirjnthERKRE8XLhuWOAnLMrhmZvE8nfyWiY/SjUaANdns636LLIWHq+9zsfLIykV7Oq/PbENQzpUEuJCxGRYsgY09MYs8MYs8sY82wu+zsbY9YZYzKMMRc/UZLkzlqoUAs6PgoV8u/5+svmQ8QmpjJUy6OKiIiLuLLnxWqgnjGmDo6kxW3AYBd+nhQHWVkw437H8qj9PgNv31yLHT2Vwr9+3saPGw5SJ7gMXw9vT6d6wW4OVkRE3EUTgXuAMRd8iPC3icv3UTuoNJ3rVXZxUCIiUlK5LHlhrc0wxjwM/IpjqdQvrLVbjDGvAmustbOMMW2BGUBF4GZjzCvW2vwHVErxtvxDiFoKvf8LQVectzsry/Ltqv28OXc7qelZPNqtHg9ccwX+vpoYTESkmNNE4IXU5piTrN13gpduaqyejyIi4jIunfPCWjsHmHPOtpE5Xq/GMZxEBA5ugIWvQaPe0GrIebsTUzMY9sUq1u47wVVXBPFa36ZcUTnQA4GKiIgHaCLwQmri8igCfL0Z0EZNOhERcZ0iMWGnlABpyY5lUcsEw83vO7qqnuODhZGs3XeCtwY0Z0CbUEwuZURERC5EE4E7z4mkNH7ccJD+bUIpH5D7UE8RERFnUPJCCod5L0LsThj6I5SudN7unUdO8cWyvdzWtiYDw2vmcgIRESnmNBF4ITRlzQFSM7I0UaeIiLicK1cbEbk4O+bCms8h4mGoe815u621vDRzM4H+Pjzds6HbwxMRkULhzETgxphSOCYCn+XhmEq0zCzLpBX7aF+nEg2rlvN0OCIiUswpeSGelXgUfnwIqjSDbiNzLTJr40FW7o3j6esbUqlMKTcHKCIihYG1NgP4eyLwbcCUvycCN8b0BjDGtDXGRAMDgU+NMVs8F3Hxt3jHUaJPnGbYVbU9HYqIiJQAGjYinmMtzHwQ0hKh/3jw8TuvSEJKOq//vI0WoeW5ta2Gi4iIlGSaCLxwmbB8H1XL+dOjcRVPhyIiIiWAkhfiOas+g13z4YYxEJL7cJD35kcSm5jK58PC8dbyayLFTnp6OtHR0aSkpHg6FMmHv78/oaGh+PpqQkZx2HMskd93HuOJHvXx9VZHXhEpPtQ2cb1LbVcoeSGecXQbzH8J6l0Hbe/Jtci2QwlMWB7F4HZhNA+t4N74RMQtoqOjKVu2LLVr19YKQoWUtZbjx48THR1NnTp1PB2OFBKTVuzD19twWzstNSsixYvaJq51Oe0KpcrF/TJS4Yd7oFQg9Pko12VRrbWM/HEz5fx9eOr6Bh4IUkTcISUlhaCgIDUOCjFjDEFBQXoCJWckpWYwbU00NzSrRuWy5w/5FBEpytQ2ca3LaVcoeSHut/BVOLIZ+o6FwJBci0xfF8PqqBM826shFUprkk6R4kyNg8JP/0eS04z1MZxKzWBoRG1PhyIi4hK677nWpdavkhfiXrsXwfL/OoaK1L8+1yInT6fzxi/baBVWgYFtNEmniIhIYWGtZeLyKJrWKEfrMA3pFBER91HyQtwnOQ5mPgDBDaDHa3kWe2feDuKS0nitT1O8NEmniIhIobFiTxw7jyQyNEJjwUVExL2UvBD3sBZmPQJJsY5lUUuVzrXY5piTTFqxjyEdatG0Rnk3BykiJU18fDxjx44t8HE33HAD8fHx+ZYZOXIkCxYsuNTQ3Oarr77i4YcfBuDll19mzJgxHo5ICrOJy6OoUNqX3i2qezoUEZFiq6i1Tz744AMaNWrEHXfcwfbt24mIiMDPz8/pbQqtNiLusX4SbP/J0eOiWvNci2RlOSbprFi6FE9cp0k6RUqaV2ZvYevBBKees3H1coy6uUme+/9uHDz44INnbc/IyMDHJ+9b5Jw5cy742a+++urFB1pA1lqstXh56RmEuM/B+NPM23qEe66ug7+vt6fDERFxOU+0TaDotU/Gjh3LggULCA0N5ejRo3zwwQfMnDnT6Z+jVo+43vHd8MuzUKczRDycZ7Fpa6NZtz+e525oRPmAgq35KyJyKZ599ll2795Ny5Ytadu2LVdffTW9e/emcePGAPTt25c2bdrQpEkTxo0bd+a42rVrExsbS1RUFI0aNeLee++lSZMmXHfddZw+fRqAu+66i2nTpp0pP2rUKFq3bk2zZs3Yvn07AMeOHaNHjx40adKEe+65h1q1ahEbG5trrFFRUTRo0IChQ4fStGlTDhw4wFtvvUXbtm1p3rw5o0aNOlN24sSJNG/enBYtWnDnnXcCMHv2bNq3b0+rVq3o3r07R44ccX6FSrH27cr9ZFnLkPa1PB2KiEixVpTaJ/fffz979uyhV69evPvuu4SEhNC2bVt8fZ3/95x6XohrZaY7lkX19oW+n0AeTwnjk9MYPXc74bUq0q9VDTcHKSKFwYWeQrjC6NGj2bx5Mxs2bGDx4sXceOONbN68+cy641988QWVKlXi9OnTtG3blv79+xMUFHTWOSIjI5k8eTKfffYZgwYN4ocffmDIkCHnfVZwcDDr1q1j7NixjBkzhvHjx/PKK69w7bXX8txzzzF37lw+//zzfOONjIxkwoQJdOjQgXnz5hEZGcmqVauw1tK7d29+//13goKCeP311/nzzz8JDg4mLi4OgE6dOrFixQqMMYwfP57//Oc/vP32206qSSnuUjMy+W71fro1rELNSrkP/RQRKW480TaBotU++eSTT5g7dy6LFi0iODjYuRVxDiUvxLWWvAkH18HACVA+76TEmHk7iE9O49U+7TVJp4h4TLt27c40DMAxhnPGjBkAHDhwgMjIyPMaB3Xq1KFly5YAtGnThqioqFzP3a9fvzNlpk+fDsCyZcvOnL9nz55UrFgx3/hq1apFhw4dAJg3bx7z5s2jVatWACQmJhIZGcnGjRsZOHDgmQZEpUqVAIiOjubWW2/l0KFDpKWlnfU9RS7kl02HiU1MY9hV6nUhIuJuhb194i4aNiKus285LH0bWg6BJn3zLPZXdDzfrNzPsKtq07h6OTcGKCJytjJlypx5vXjxYhYsWMDy5cvZuHEjrVq1IiUl5bxj/Pz8zrz29vYmIyMj13P/XS6/MgWJz1rLc889x4YNG9iwYQO7du1i+PDheR77yCOP8PDDD7Np0yY+/fTTXL+LSF4mLI+ibuUydLzCtU/VRETkfIW9feIuSl6Ia6SchOkjoEIt6DU6z2JZWZaXZm4mqIwfj/Wo78YARUSgbNmynDp1Ktd9J0+epGLFipQuXZrt27ezYsUKp39+x44dmTJlCuDoSXHixImLPvb666/niy++IDExEYCYmBiOHj3Ktddey9SpUzl+/DjAmWEjJ0+epEYNRw+4CRMmOPNrSDH3V3Q86/fHM7RDLfWOFBFxg6LcPnElDRsR1/j5SUiIgX/8Cn5l8yz2/ZoDbIw+ybu3tqCcvybpFBH3CgoKomPHjjRt2pSAgACqVKlyZl/Pnj355JNPaNSoEQ0aNDgzXMOZRo0axe23386kSZOIiIigatWqlC2b9+/MnK677jq2bdtGREQEAIGBgXz99dc0adKEF154gS5duuDt7U2rVq346quvePnllxk4cCAVK1bk2muvZe/evU7/PlI8TVy+jzKlvOnfJtTToYiIlAhFuX1y+PBhwsPDSUhIwMvLi/fee4+tW7dSrtzl97A31trLPok7hYeH2zVr1ng6DMnPX1Nh+j3Q9QXo8nSexeKS0rj27cXUr1KW70d0wBg9zREpabZt20ajRo08HYbHpKam4u3tjY+PD8uXL+eBBx5gw4YNng4rV7n9Xxlj1lprwz0UklOoXZG/uKQ0OryxkEHhobzet5mnwxERcbmS3jYB97RPLqVdoZ4X4lzx++Hnx6Fme+j0eL5F3/p1O6dSMnitT1MlLkSkRNq/fz+DBg0iKyuLUqVK8dlnn3k6JJEzsrIs45fuIS0ji6ERtT0djoiIuElhbZ8oeSHOk5UJ0+8Da6HfOPDO+/Jav/8E360+wPCOdWhQ9eK6IImIFDf16tVj/fr1Z207fvw43bp1O6/swoULz5tJXMQVUtIzmb4uhs+X7WH3sSS6NQyhfhXdq0VESorC2j5R8kKcZ9m7sP9PuOVTqFg7z2KZWZaXftxMSFk//qlJOkVEzhIUFFRoh45I8RabmMqk5fv4esU+jiel0bRGOd6/rSU3NKvm6dBERMTDCkP7RMkLcY6YtbD4DWjaH5rfmm/Rb1ftZ3NMAh/c3opAP12CIiIinrTraCKfL9vDD+tiSMvIolvDEO65ui4d6lbSsE4RESk09JejXL7URPjhXgisCje+A3k0dFIzMlm5J4635m4nom4QNzfXkxwRERFPsNayfM9xxi/dy2/bj+Ln40X/1qEM71SHK0MCPR2eiIjIeZS8KGnST8POX2HLdDBejp4SV/YAX/9LP+evz0HcHrjrJwiocNau6BPJLN5xjMU7jvLn7uMkp2VSzt+H1/o20dMcERERN0vPzOLnvw7x2dI9bDmYQFCZUjzWvT5DOoQRFOjn6fBERETypORFSZCZAXuXwKZpsG02pJ2CwCpgs2DLDPArD41vhmYDofbV4OV98efeNhvWTXSsLFK7E2kZWayJimPxzmMs2n6UyKOJAIRWDGBAm1CuaVCZiLrBBJQqwGeIiIjIZTl5Op3vVu3nqz+jOHQyhStDAhndrxl9W9XA31f3ZBERKfyUvCiurHXMQ7FpKmyeDklHwa8cNO4DzbOTFNbC3sWOpMaWmbD+a8fQj6b9odkAqN4qzyEgACQcglmPkBbSnBmBQ/ht0hqWRcaSlJaJr7ehfZ0gbm1bk2sahHBF5TLqaSEihU58fDzffvstDz74YIGPfe+99xgxYgSlS5d2QWR5W7x4MWPGjOGnn37iq6++Ys2aNfz3v/91awxSdByIS+bLP6L4fvV+ktIyueqKIP59SzO61K+Ml5fuyyIihZGr2ydTp05l5MiRVK1alWnTpjFgwABWr17NXXfdVajbFC5NXhhjegLvA97AeGvt6HP2+wETgTbAceBWa22UK2Mq9o7tdCQsNk2FE3vB2w/qX+/oVVHvuvOHh1zZ3fFz07uwc64jkbH6M1jxEQRd6Tiu2UAIugJwjJFNOJ3B9kPxhPw4lGqnk7khfih79m+nenl/+rSqQdcGIVx1RRBlNBmniBTEL8/C4U3OPWfVZtBrdJ674+PjGTt27CU3DoYMGXLRyYvMzEy8vfWEW1zj5Ol09h9PJup4EvuOJxF1PJm9sUms338CL2O4uUV1hneqQ9Ma5T0dqohI0eGBtgm4vn3y+eef89lnn9GpUyeSkpJ47bXX2Lx5M5s3by7w57mTy/66NMZ4Ax8BPYBoYLUxZpa1dmuOYsOBE9baK40xtwFvAvkvVSHnOxkDm39wJCwO/+WYy6JOZ+j8JDS86bx5KM6VkZlFXIohtmI3jrfpxMk6Rym39xfCDs4hbPFovBa/wU7vevxMJ6amtOVgZgWGe8/hJd9VfFb+UW5tfS1dG4ZQLyRQvStEpEh59tln2b17Ny1btqRHjx6EhIQwZcoUUlNTueWWW3jllVdISkpi0KBBREdHk5mZyUsvvcSRI0c4ePAgXbt2JTg4mEWLFuV6/sDAQO677z4WLFjARx99RFRUFB988AFpaWm0b9+esWPH4u3tzdy5c3n++efJzMwkODiYhQsXsmrVKh599FFSUlIICAjgyy+/pEGDBm6uocKlJD8UsdYSn5yenZxIPu/fuKS0s8pXLedPraDS3N/lCu6MqEW18gEeilxERArKle2TV199lWXLljF8+HB69+7NW2+9RadOndi1a5cHvmnBuPLReDtgl7V2D4Ax5jugD5AzedEHeDn79TTgv8YYY621LozrPJuWTCcp2nNZpswsS6a1ZGVZMrMgMyuLLGvJyHLsy7LWUeac115Z6bRMW0uT9E14YYn0qc/vgSP4w68zJxIrwXJg+ZZcP9NaSE7LIDYxjRPJaZxf440o5d2EhmVO0dtnJd0zlvBY2pc86juBwyFtqHpyA+lX9OLewa/kP7RERORiXeAphCuMHj2azZs3s2HDBubNm8e0adNYtWoV1lp69+7N77//zrFjx6hevTo///wzACdPnqR8+fK88847LFq0iODg4DzPn5SURPv27Xn77bfZtm0bb775Jn/88Qe+vr48+OCDfPPNN/Tq1Yt7772X33//nTp16hAXFwdAw4YNWbp0KT4+PixYsIDnn3+eH374wS31UhgVpYcii3YcZXf2nE+X4+TpdKKOJzt6UsQmkZCScWafMVC9fAC1g0vTs2lVageVplZQGWoHlSGsUmnNLSUi4gweaJuAa9snI0eO5LfffmPMmDGEh4e782tdNlcmL2oAB3K8jwba51XGWpthjDkJBAGxOQsZY0YAIwDCwsKcHujp9VPpED/H6ed1h0M+ocwodyfLS3fliG+NM9svZpGzKuX8aFu7EkGBflQOLEVwoB9BgX4EB5YiuKwfZf18sntSDHIccGwnXpunUf2vKRBYBa++HylxISLFxrx585g3bx6tWrUCIDExkcjISK6++mqeeOIJnnnmGW666Sauvvrqiz6nt7c3/fv3B2DhwoWsXbuWtm3bAnD69GlCQkJYsWIFnTt3pk6dOgBUqlQJcDRChg0bRmRkJMYY0tPTnfl1i6Ii81Dkx/UxzNxw8LLP42UgtGJpagWVpk/LGtQKKk3toDLUDi5NaMXSmmhTRKQEcEX7pKgqEpMSWGvHAeMAwsPDnd4AaXbvOBLS0y5c0EVKeXvh4wXeXqaAwy4M1fzK0t8Y+rssuhwq14euz8M1zzlWKinIqiQiIoWctZbnnnuO++5VF43FAAAL3ElEQVS777x969atY86cObz44ot069aNkSNHXtQ5/f39z8xzYa1l2LBhvPHGG2eVmT17dq7HvvTSS3Tt2pUZM2YQFRXFNddcU7AvVPwUmYcib/Rrzqt9m172eQJ8vfH19nJCRCIiUlS5on1SVLkyeRED1MzxPjR7W25loo0xPkB5HGNU3SqgTFk0ErQAjAGjxIWIFH1ly5bl1KlTAFx//fW89NJL3HHHHQQGBhITE4Ovry8ZGRlUqlSJIUOGUKFCBcaPH3/WsfkNG8mpW7du9OnTh8cee4yQkBDi4uI4deoUHTp04MEHH2Tv3r1nho1UqlSJkydPUqOGo0fdV1995ZLvX1K5+qFIQClvAtB9UkRELo072ydFiSuTF6uBesaYOjiSFLcBg88pMwsYhmN2hgHAb+7u2ikiIiVXUFAQHTt2pGnTpvTq1YvBgwcTEREBOCbb/Prrr9m1axdPPfUUXl5e+Pr68vHHHwMwYsQIevbsSfXq1fOcsDOnxo0b8/rrr3PdddeRlZWFr68vH330ER06dGDcuHH069ePrKwsQkJCmD9/Pk8//TTDhg3j9ddf58Ybb3RpPRQRReahiIiIyOVwZ/sEoHbt2iQkJJCWlsbMmTOZN28ejRs3dtn3u1TGlbkCY8wNwHs4ZgX/wlr7L2PMq8Aaa+0sY4w/MAloBcQBt/09ljUv4eHhds2aNS6LWURE3Gfbtm00atTI02HIRcjt/8oYs9Za65bZvrKTETuBbjiSFKuBwdbaLTnKPAQ0s9benz1hZz9r7aD8zqt2hYiI5KS2iXtcSrvCpXNeWGvnAHPO2TYyx+sUYKArYxAREZGiL3sOi4eBX/nfQ5EtOR+KAJ8Dk4wxu8h+KOK5iEVERMSZisSEnSIiIoVZ+/btSU1NPWvbpEmTaNasmYciKp70UEREROTiFbf2iZIXIiLiUdbaAq60VPisXLnS0yG4lKajEhGRkqQ4tE2g8LZPLrVdofW3RETEY/z9/Tl+/Lj+OC7ErLUcP34cf39/T4ciIiLicmqbuNbltCvU80JERDwmNDSU6Ohojh075ulQJB/+/v6EhoZ6OgwRERGXU9vE9S61XaHkhYiIeIyvry916tTxdBgiIiIigNomhZmGjYiIiIiIiIhIoabkhYiIiIiIiIgUakpeiIiIiIiIiEihZoraLKrGmGPAPhecOhiIdcF5iyvVV8GovgpG9VUwqq+CUX0VXF51VstaW9ndwTiTC9sVhYWud+dTnTqf6tT5VKfOpfp0vtzqNN92RZFLXriKMWaNtTbc03EUFaqvglF9FYzqq2BUXwWj+io41VnRpf8751OdOp/q1PlUp86l+nS+S6lTDRsRERERERERkUJNyQsRERERERERKdSUvPifcZ4OoIhRfRWM6qtgVF8Fo/oqGNVXwanOii793zmf6tT5VKfOpzp1LtWn8xW4TjXnhYiIiIiIiIgUaup5ISIiIiIiIiKFmpIXIiIiIiIiIlKoKXkBGGN6GmN2GGN2GWOe9XQ8hZ0xJsoYs8kYs8EYs8bT8RQ2xpgvjDFHjTGbc2yrZIyZb4yJzP63oidjLEzyqK+XjTEx2dfYBmPMDZ6MsTAxxtQ0xiwyxmw1xmwxxjyavV3XWC7yqS9dY7kwxvgbY1YZYzZm19cr2dvrGGNWZt8nvzfGlPJ0rHJhul9fPt3TnUv3fOdTu8D51HZwPme1L0r8nBfGGG9gJ9ADiAZWA7dba7d6NLBCzBgTBYRba2M9HUthZIzpDCQCE621TbO3/QeIs9aOzk6QVbTWPuPJOAuLPOrrZSDRWjvGk7EVRsaYakA1a+06Y0xZYC3QF7gLXWPnyae+BqFr7DzGGAOUsdYmGmN8gWXAo8DjwHRr7XfGmE+Ajdbajz0Zq1yY7teXT/d059I93/nULnA+tR2cz1ntC/W8gHbALmvtHmttGvAd0MfDMUkRZq39HYg7Z3MfYEL26wk4fgEKedaX5MFae8hauy779SlgG1ADXWO5yqe+JBfWITH7rW/2jwWuBaZlb9f1JSWG7unOpXu+86ld4HxqOzifs9oXSl44LsQDOd5Ho4vzQiwwzxiz1hgzwtPBFBFVrLWHsl8fBqp4Mpgi4mFjzF/ZXUzV1TEXxpjaQCtgJbrGLuic+gJdY7kyxngbYzYAR4H5wG4g3lqbkV1E98miQ/dr19DvW+fT72MnULvA+dR2cB5ntC+UvJBL0cla2xroBTyU3QVQLpJ1jNUq2eO1Luxj4AqgJXAIeNuz4RQ+xphA4Afgn9bahJz7dI2dL5f60jWWB2ttprW2JRCKo3diQw+HJJdO92sX0+9bp9DvYydQu8D51HZwLme0L5S8gBigZo73odnbJA/W2pjsf48CM3BcfJK/I9nj5/4eR3fUw/EUatbaI9m/4LKAz9A1dpbssYI/AN9Ya6dnb9Y1lofc6kvX2IVZa+OBRUAEUMEY45O9S/fJIkL3a5fR71sn0u/jy6d2gfOp7eA6l9O+UPLCMUFnveyZTksBtwGzPBxToWWMKZM9cQ3GmDLAdcDm/I8SHNfUsOzXw4AfPRhLoff3zTbbLegaOyN7wqPPgW3W2ndy7NI1lou86kvXWO6MMZWNMRWyXwfgmMx6G45GxoDsYrq+igDdr11Kv2+dSL+PL4/aBc6ntoPzOat9UeJXGwHIXubmPcAb+MJa+y8Ph1RoGWPq4nh6A+ADfKv6OpsxZjJwDRAMHAFGATOBKUAYsA8YZK3VhFXkWV/X4OiSZ4Eo4L4c4zZLNGNMJ2ApsAnIyt78PI6xmLrGzpFPfd2OrrHzGGOa45gwyxvHA44p1tpXs3/3fwdUAtYDQ6y1qZ6LVC5E92vn0D3duXTPdz61C5xPbQfnc1b7QskLERERERERESnUNGxERERERERERAo1JS9EREREREREpFBT8kJERERERERECjUlL0RERERERESkUFPyQkREREREREQKNSUvREowY4w1xryd4/2TxpiXXfA5k40xfxljHnP2uS/wuYuNMeHu/EwREZGSSu0KEXElH08HICIelQr0M8a8Ya2NdcUHGGOqAm2ttVe64vwiIiJSaKhdISIuo54XIiVbBjAOOO/JhTGmtjHmt+wnGwuNMWH5ncgY42+M+dIYs8kYs94Y0zV71zyghjFmgzHm6nOOqWyM+cEYszr7p2P29peNMZOMMcuNMZHGmHuztxtjzFvGmM3Zn3NrjnM9k71tozFmdI6PGWiMWWWM2Xnu54uIiIhTqV0hIi6jnhci8hHwlzHmP+ds/xCYYK2dYIz5B/AB0Def8zwEWGttM2NMQ2CeMaY+0Bv4yVrbMpdj3gfetdYuy27E/Ao0yt7XHOgAlAHWG2N+BiKAlkALIBhYbYz5PXtbH6C9tTbZGFMpx2f4WGvbGWNuAEYB3S+qVkRERORSqF0hIi6h5IVICWetTTDGTAT+DzidY1cE0C/79STg3EbIuTrhaJhgrd1ujNkH1AcS8jmmO9DYGPP3+3LGmMDs1z9aa08Dp40xi4B22Z8x2VqbCRwxxiwB2gJdgC+ttcnZnx+X4zOmZ/+7Fqh9ge8gIiIil0HtChFxFSUvRATgPWAd8KWbP9cL6GCtTcm5MbvRYc8pe+77i5Wa/W8m+p0nIiLiDmpXiIjTac4LEfn7icIUYHiOzX8Ct2W/vgNYeoHTLM0uR3a3zjBgxwWOmQc88vcbY0zOLqB9sse7BgHXAKuzP+NWY4y3MaYy0BlYBcwH7jbGlM4+T87unSIiIuJGaleIiCsoeSEif3sbx3jPvz2C48b9F3An8CiAMeZ+Y8z9uRw/FvAyxmwCvgfustam5lIup/8DwrMn79oK5DzvX8AiYAXwmrX2IDAje/tG4DfgaWvtYWvtXGAWsMYYswF4siBfXERERJxO7QoRcSpj7aX2mBIRcQ3jWBM+0Vo7xtOxiIiISNGmdoVI8aCeFyIiIiIiIiJSqKnnhYiIiIiIiIgUaup5ISIiIiIiIiKFmpIXIiIiIiIiIlKoKXkhIiIiIiIiIoWakhciIiIiIiIiUqgpeSEiIiIiIiIihdr/A6rF9avzoLRtAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplots(figsize=(15,10))\n",
    "ax = plt.subplot(2,2,1)\n",
    "plt.plot(hist.history['accuracy'], label='training_accuracy')\n",
    "plt.plot(hist.history['val_accuracy'], label='test_accuracy')\n",
    "plt.title('Accuracy Plot')\n",
    "plt.ylabel('accuracy value')\n",
    "plt.xlabel('No. of epoch')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "ax = plt.subplot(2,2,2)\n",
    "plt.plot(hist.history['loss'], label='training_loss')\n",
    "plt.plot(hist.history['val_loss'], label='test_loss')\n",
    "plt.title('Loss Plot')\n",
    "plt.ylabel('loss value')\n",
    "plt.xlabel('No. of epoch')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "ax = plt.subplot(2,2,3)\n",
    "plt.plot(hist.history['recall'], label='training_recall')\n",
    "plt.plot(hist.history['val_recall'], label='test_recall')\n",
    "plt.title('Recall Plot')\n",
    "plt.ylabel('recall value')\n",
    "plt.xlabel('No. of epoch')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "ax = plt.subplot(2,2,4)\n",
    "plt.plot(history_df['f1_score'], label='training_f1')\n",
    "plt.plot(history_df['val_f1_score'], label='test_f1')\n",
    "plt.title('F1 Score Plot')\n",
    "plt.ylabel('f1 score value')\n",
    "plt.xlabel('No. of epoch')\n",
    "plt.legend(loc=\"lower right\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {
    "id": "LBprGBfhw9ks"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YbP3LjD4wES0"
   },
   "source": [
    "## Prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Z6tiagbqbHqx"
   },
   "source": [
    "Below, a predict function is defined that will take as input natural language text and try to classify it into the categories of mental health issues defined from the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {
    "id": "Uw1eIn7EFLt2"
   },
   "outputs": [],
   "source": [
    "model = tf.keras.models.load_model('/content/Final_LSTM_model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-12-11T09:35:29.589470Z",
     "iopub.status.busy": "2021-12-11T09:35:29.589141Z",
     "iopub.status.idle": "2021-12-11T09:35:29.599894Z",
     "shell.execute_reply": "2021-12-11T09:35:29.599159Z",
     "shell.execute_reply.started": "2021-12-11T09:35:29.589434Z"
    },
    "id": "HXKjAM-pbHqy"
   },
   "outputs": [],
   "source": [
    "def predict(text):\n",
    "    processed = preprocess_text(text)\n",
    "    text_no_stops = [word for word in processed if word not in stop_words]\n",
    "    tokenized = tokenizer.texts_to_sequences( text_no_stops )\n",
    "    padded = preprocessing.sequence.pad_sequences( tokenized , maxlen=max_input_length , padding='post' )\n",
    "    input_data = np.array( padded )\n",
    "    output = model.predict(input_data)\n",
    "    index = np.argmax(output[0])\n",
    "    return (targetdf.columns[index])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {
    "id": "V1XBI-fhHl1K"
   },
   "outputs": [],
   "source": [
    "sent = ['I break out in cold sweat. I have panic attaks and bad dreams at night.',\n",
    "        'I am constantly worried about my work future.',\n",
    "        'I am very sad and depressed. I want to commit suicide.',\n",
    "        'My husband and I have constant arguments. I want to get a divorce',\n",
    "        'I am very angry all the time. I smash things in anger']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2021-12-11T09:35:30.568698Z",
     "iopub.status.busy": "2021-12-11T09:35:30.567967Z",
     "iopub.status.idle": "2021-12-11T09:35:30.624762Z",
     "shell.execute_reply": "2021-12-11T09:35:30.623796Z",
     "shell.execute_reply.started": "2021-12-11T09:35:30.568660Z"
    },
    "id": "qvShTZfObHqy",
    "outputId": "d04222c7-7387-4c9a-fd86-8d4d68bdb701"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sleep Improvement\n",
      "Stress\n",
      "Depression\n",
      "Marriage\n",
      "Anger Management\n"
     ]
    }
   ],
   "source": [
    "for i in sent:\n",
    "    print(predict(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.792630Z",
     "iopub.status.idle": "2021-12-11T09:31:57.793028Z",
     "shell.execute_reply": "2021-12-11T09:31:57.792833Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.792810Z"
    },
    "id": "FsgEH9RrbHqz"
   },
   "outputs": [],
   "source": [
    "pred = []\n",
    "for index,rows in counseldf.iterrows():\n",
    "    pred.append(predict(rows['questionText']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.794305Z",
     "iopub.status.idle": "2021-12-11T09:31:57.794748Z",
     "shell.execute_reply": "2021-12-11T09:31:57.794541Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.794518Z"
    },
    "id": "TioY0cmibHqz"
   },
   "outputs": [],
   "source": [
    "counseldf['predict'] = pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.796014Z",
     "iopub.status.idle": "2021-12-11T09:31:57.796411Z",
     "shell.execute_reply": "2021-12-11T09:31:57.796216Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.796194Z"
    },
    "id": "lGS6Zq-wbHq0",
    "outputId": "9ab0376d-9b17-4cbc-eb0a-bb3561439da0"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>questionText</th>\n",
       "      <th>topics</th>\n",
       "      <th>predict</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>My wife and mother are having tense disagreeme...</td>\n",
       "      <td>Family Conflict</td>\n",
       "      <td>Marriage</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I'm planning to have baby, so I have to quit s...</td>\n",
       "      <td>Substance Abuse</td>\n",
       "      <td>Substance Abuse</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>I have secrets in my mind, and I don't know wh...</td>\n",
       "      <td>Family Conflict</td>\n",
       "      <td>Social Relationships</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I am extremely possessive in my relationships ...</td>\n",
       "      <td>Behavioral Change</td>\n",
       "      <td>Behavioral Change</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I had a head injury a few years ago and my min...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>Behavioral Change</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        questionText  ...               predict\n",
       "0  My wife and mother are having tense disagreeme...  ...              Marriage\n",
       "1  I'm planning to have baby, so I have to quit s...  ...       Substance Abuse\n",
       "2  I have secrets in my mind, and I don't know wh...  ...  Social Relationships\n",
       "3  I am extremely possessive in my relationships ...  ...     Behavioral Change\n",
       "4  I had a head injury a few years ago and my min...  ...     Behavioral Change\n",
       "\n",
       "[5 rows x 3 columns]"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.797735Z",
     "iopub.status.idle": "2021-12-11T09:31:57.798204Z",
     "shell.execute_reply": "2021-12-11T09:31:57.797945Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.797922Z"
    },
    "id": "Rjw1hDIYbHq1",
    "outputId": "f04a345a-200e-48df-9c9c-4928fd0729ba"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Relationships                246\n",
       "Anxiety                      178\n",
       "Family Conflict              113\n",
       "Depression                    97\n",
       "Marriage                      89\n",
       "Self-esteem                   64\n",
       "Parenting                     59\n",
       "Trauma                        51\n",
       "Human Sexuality               45\n",
       "Behavioral Change             43\n",
       "Intimacy                      42\n",
       "Relationship Dissolution      42\n",
       "Counseling Fundamentals       39\n",
       "Social Relationships          38\n",
       "Anger Management              31\n",
       "Professional Ethics           28\n",
       "Stress                        24\n",
       "LGBTQ                         24\n",
       "Spirituality                  19\n",
       "Addiction                     14\n",
       "Grief and Loss                14\n",
       "Substance Abuse               14\n",
       "Legal & Regulatory            11\n",
       "Diagnosis                     10\n",
       "Sleep Improvement             10\n",
       "Domestic Violence              9\n",
       "Workplace Relationships        8\n",
       "Eating Disorders               4\n",
       "Children & Adolescents         3\n",
       "Military Issues                3\n",
       "Alzheimer's                    2\n",
       "Career Counseling              2\n",
       "Name: topics, dtype: int64"
      ]
     },
     "execution_count": 222,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.topics.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.799325Z",
     "iopub.status.idle": "2021-12-11T09:31:57.800006Z",
     "shell.execute_reply": "2021-12-11T09:31:57.799801Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.799777Z"
    },
    "id": "-kaKnJsgbHq1",
    "outputId": "5310766d-38a8-4207-f839-fbca759599b0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Relationships                194\n",
       "Social Relationships         172\n",
       "Anxiety                       97\n",
       "Counseling Fundamentals       96\n",
       "Marriage                      91\n",
       "Parenting                     87\n",
       "Self-esteem                   76\n",
       "Behavioral Change             68\n",
       "Sleep Improvement             68\n",
       "LGBTQ                         63\n",
       "Spirituality                  63\n",
       "Human Sexuality               41\n",
       "Stress                        35\n",
       "Grief and Loss                31\n",
       "Depression                    29\n",
       "Trauma                        25\n",
       "Anger Management              23\n",
       "Relationship Dissolution      20\n",
       "Diagnosis                     19\n",
       "Career Counseling             16\n",
       "Family Conflict               15\n",
       "Domestic Violence             13\n",
       "Substance Abuse                7\n",
       "Addiction                      6\n",
       "Professional Ethics            6\n",
       "Intimacy                       6\n",
       "Eating Disorders               3\n",
       "Workplace Relationships        2\n",
       "Children & Adolescents         2\n",
       "Legal & Regulatory             1\n",
       "Military Issues                1\n",
       "Name: predict, dtype: int64"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counseldf.predict.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {
    "execution": {
     "iopub.status.busy": "2021-12-11T09:31:57.801486Z",
     "iopub.status.idle": "2021-12-11T09:31:57.801889Z",
     "shell.execute_reply": "2021-12-11T09:31:57.801696Z",
     "shell.execute_reply.started": "2021-12-11T09:31:57.801675Z"
    },
    "id": "pZfJdMZLbHq2"
   },
   "outputs": [],
   "source": [
    "model.save('glove_model2.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ajISf_XmbHq2"
   },
   "source": [
    "We saved the word2vec model which was trained for 30 epochs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "SIDc_pk8bHrI",
    "outputId": "9250cd54-0fec-4e4a-9e82-c4a97a6b7f3a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           precision    recall  f1-score   support\n",
      "\n",
      "                Addiction       0.00      0.00      0.00        14\n",
      "              Alzheimer's       0.00      0.00      0.00         2\n",
      "         Anger Management       0.17      0.13      0.15        31\n",
      "                  Anxiety       0.47      0.26      0.33       178\n",
      "        Behavioral Change       0.04      0.07      0.05        43\n",
      "        Career Counseling       0.00      0.00      0.00         2\n",
      "   Children & Adolescents       0.00      0.00      0.00         3\n",
      " Counseling Fundamentals        0.27      0.67      0.39        39\n",
      "               Depression       0.34      0.10      0.16        97\n",
      "                Diagnosis       0.42      0.80      0.55        10\n",
      "        Domestic Violence       0.31      0.44      0.36         9\n",
      "         Eating Disorders       0.67      0.50      0.57         4\n",
      "          Family Conflict       0.47      0.06      0.11       113\n",
      "           Grief and Loss       0.00      0.00      0.00        14\n",
      "          Human Sexuality       0.07      0.07      0.07        45\n",
      "                 Intimacy       0.00      0.00      0.00        42\n",
      "                    LGBTQ       0.06      0.17      0.09        24\n",
      "       Legal & Regulatory       0.00      0.00      0.00        11\n",
      "                 Marriage       0.45      0.46      0.46        89\n",
      "         Military Issues        0.00      0.00      0.00         3\n",
      "                Parenting       0.30      0.44      0.36        59\n",
      "      Professional Ethics       0.33      0.07      0.12        28\n",
      "Relationship Dissolution        0.50      0.24      0.32        42\n",
      "            Relationships       0.45      0.36      0.40       246\n",
      "              Self-esteem       0.12      0.14      0.13        64\n",
      "        Sleep Improvement       0.01      0.10      0.03        10\n",
      "     Social Relationships       0.09      0.39      0.14        38\n",
      "             Spirituality       0.19      0.63      0.29        19\n",
      "                   Stress       0.14      0.21      0.17        24\n",
      "          Substance Abuse       1.00      0.50      0.67        14\n",
      "                   Trauma       0.60      0.29      0.39        51\n",
      "  Workplace Relationships       1.00      0.25      0.40         8\n",
      "\n",
      "                 accuracy                           0.25      1376\n",
      "                macro avg       0.27      0.23      0.21      1376\n",
      "             weighted avg       0.34      0.25      0.26      1376\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1308: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1308: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1308: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(counseldf.topics, counseldf.predict))"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "Final W2V model.ipynb",
   "provenance": []
  },
  "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}