Mental_Health_Issue detection.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "593c805a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#basic imports\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "0cd3818e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"/Users/aadityajoshi/Downloads/Combined Data.csv\",encoding='utf-8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c36ce1f6",
   "metadata": {},
   "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>Unnamed: 0</th>\n",
       "      <th>statement</th>\n",
       "      <th>status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>oh my gosh</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>trouble sleeping, confused mind, restless hear...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>All wrong, back off dear, forward doubt. Stay ...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>I've shifted my focus to something else but I'...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>I'm restless and restless, it's been a month n...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                                          statement   status\n",
       "0           0                                         oh my gosh  Anxiety\n",
       "1           1  trouble sleeping, confused mind, restless hear...  Anxiety\n",
       "2           2  All wrong, back off dear, forward doubt. Stay ...  Anxiety\n",
       "3           3  I've shifted my focus to something else but I'...  Anxiety\n",
       "4           4  I'm restless and restless, it's been a month n...  Anxiety"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1bdd971f",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop('Unnamed: 0',axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "42e25d3f",
   "metadata": {},
   "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>statement</th>\n",
       "      <th>status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oh my gosh</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>trouble sleeping, confused mind, restless hear...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>All wrong, back off dear, forward doubt. Stay ...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I've shifted my focus to something else but I'...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I'm restless and restless, it's been a month n...</td>\n",
       "      <td>Anxiety</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           statement   status\n",
       "0                                         oh my gosh  Anxiety\n",
       "1  trouble sleeping, confused mind, restless hear...  Anxiety\n",
       "2  All wrong, back off dear, forward doubt. Stay ...  Anxiety\n",
       "3  I've shifted my focus to something else but I'...  Anxiety\n",
       "4  I'm restless and restless, it's been a month n...  Anxiety"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3d6d361b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 53043 entries, 0 to 53042\n",
      "Data columns (total 2 columns):\n",
      " #   Column     Non-Null Count  Dtype \n",
      "---  ------     --------------  ----- \n",
      " 0   statement  52681 non-null  object\n",
      " 1   status     53043 non-null  object\n",
      "dtypes: object(2)\n",
      "memory usage: 828.9+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "599ca309",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "statement    362\n",
       "status         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "91bffe15",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop_duplicates(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7305d49d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping missing values\n",
    "df.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d80f7f0c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 51093 entries, 0 to 52840\n",
      "Data columns (total 2 columns):\n",
      " #   Column     Non-Null Count  Dtype \n",
      "---  ------     --------------  ----- \n",
      " 0   statement  51093 non-null  object\n",
      " 1   status     51093 non-null  object\n",
      "dtypes: object(2)\n",
      "memory usage: 1.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "10387634",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check for blanks\n",
    "blanks = []\n",
    "for i,txt,lbl in df.itertuples():\n",
    "    if type(txt) == str:\n",
    "        if txt.isspace() == True:\n",
    "            blanks.append(i)\n",
    "blanks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "9f543257",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['len'] = df['statement'].apply(len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bd5089d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#removing punctuation and stop words\n",
    "import spacy\n",
    "nlp = spacy.load('en_core_web_lg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "435812f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_text(text):\n",
    "    doc = nlp(text)\n",
    "    return ' '.join([token.text.lower() for token in doc if not token.is_punct and not token.is_stop ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "dd1b4e7a",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text'] = df['statement'].apply(clean_text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "6eb01e3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#stemming, lemmatizing\n",
    "from nltk.stem import PorterStemmer\n",
    "stemmer = PorterStemmer()\n",
    "df['stem_text'] = df['text'].apply(lambda x : stemmer.stem(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "1e2fb142",
   "metadata": {},
   "outputs": [],
   "source": [
    "def lemmatize(text):\n",
    "    doc = nlp(text)\n",
    "    return ' '.join([token.lemma_ for token in doc])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "d8eefe61",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemm_text'] = df['text'].apply(lemmatize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "84e33b53",
   "metadata": {},
   "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>statement</th>\n",
       "      <th>status</th>\n",
       "      <th>len</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oh my gosh</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>trouble sleeping, confused mind, restless hear...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>All wrong, back off dear, forward doubt. Stay ...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I've shifted my focus to something else but I'...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I'm restless and restless, it's been a month n...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>72</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           statement   status  len\n",
       "0                                         oh my gosh  Anxiety   10\n",
       "1  trouble sleeping, confused mind, restless hear...  Anxiety   64\n",
       "2  All wrong, back off dear, forward doubt. Stay ...  Anxiety   78\n",
       "3  I've shifted my focus to something else but I'...  Anxiety   61\n",
       "4  I'm restless and restless, it's been a month n...  Anxiety   72"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "id": "0326e992",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['len'] = df['lemm_text'].apply(len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "e48f40d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9416     25282\n",
       "46660    13834\n",
       "10834    12775\n",
       "51396    11147\n",
       "13293    10079\n",
       "         ...  \n",
       "32810        0\n",
       "32095        0\n",
       "31749        0\n",
       "31755        0\n",
       "31187        0\n",
       "Name: len, Length: 51093, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['len'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "1073761e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def tokenize(text):\n",
    "    doc = nlp(text)\n",
    "    return [token.text for token in doc]\n",
    "df['tokens'] = df['statement'].apply(tokenize)\n",
    "df['wrd_len'] = df['tokens'].apply(len)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "93101ab9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16425    125\n",
       "16923    125\n",
       "37793    125\n",
       "12646    125\n",
       "34476    125\n",
       "        ... \n",
       "1854       1\n",
       "5202       1\n",
       "4680       1\n",
       "5971       1\n",
       "3180       1\n",
       "Name: wrd_len, Length: 34519, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['wrd_len'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "3a1fc4b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df['wrd_len'] <= 125]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "93a460d1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x395e5f590>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(df['wrd_len'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "2870d9b9",
   "metadata": {},
   "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>statement</th>\n",
       "      <th>status</th>\n",
       "      <th>len</th>\n",
       "      <th>text</th>\n",
       "      <th>stem_text</th>\n",
       "      <th>lemm_text</th>\n",
       "      <th>tokens</th>\n",
       "      <th>wrd_len</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oh my gosh</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>10</td>\n",
       "      <td>oh gosh</td>\n",
       "      <td>oh gosh</td>\n",
       "      <td>oh gosh</td>\n",
       "      <td>[oh, my, gosh]</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>trouble sleeping, confused mind, restless hear...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>64</td>\n",
       "      <td>trouble sleeping confused mind restless heart ...</td>\n",
       "      <td>trouble sleeping confused mind restless heart tun</td>\n",
       "      <td>trouble sleep confuse mind restless heart tune</td>\n",
       "      <td>[trouble, sleeping, ,, confused, mind, ,, rest...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>All wrong, back off dear, forward doubt. Stay ...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>78</td>\n",
       "      <td>wrong dear forward doubt stay restless restles...</td>\n",
       "      <td>wrong dear forward doubt stay restless restles...</td>\n",
       "      <td>wrong dear forward doubt stay restless restles...</td>\n",
       "      <td>[All, wrong, ,, back, off, dear, ,, forward, d...</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I've shifted my focus to something else but I'...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>61</td>\n",
       "      <td>shifted focus worried</td>\n",
       "      <td>shifted focus worri</td>\n",
       "      <td>shift focus worry</td>\n",
       "      <td>[I, 've, shifted, my, focus, to, something, el...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I'm restless and restless, it's been a month n...</td>\n",
       "      <td>Anxiety</td>\n",
       "      <td>72</td>\n",
       "      <td>restless restless month boy mean</td>\n",
       "      <td>restless restless month boy mean</td>\n",
       "      <td>restless restless month boy mean</td>\n",
       "      <td>[I, 'm, restless, and, restless, ,, it, 's, be...</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           statement   status  len  \\\n",
       "0                                         oh my gosh  Anxiety   10   \n",
       "1  trouble sleeping, confused mind, restless hear...  Anxiety   64   \n",
       "2  All wrong, back off dear, forward doubt. Stay ...  Anxiety   78   \n",
       "3  I've shifted my focus to something else but I'...  Anxiety   61   \n",
       "4  I'm restless and restless, it's been a month n...  Anxiety   72   \n",
       "\n",
       "                                                text  \\\n",
       "0                                            oh gosh   \n",
       "1  trouble sleeping confused mind restless heart ...   \n",
       "2  wrong dear forward doubt stay restless restles...   \n",
       "3                              shifted focus worried   \n",
       "4                   restless restless month boy mean   \n",
       "\n",
       "                                           stem_text  \\\n",
       "0                                            oh gosh   \n",
       "1  trouble sleeping confused mind restless heart tun   \n",
       "2  wrong dear forward doubt stay restless restles...   \n",
       "3                                shifted focus worri   \n",
       "4                   restless restless month boy mean   \n",
       "\n",
       "                                           lemm_text  \\\n",
       "0                                            oh gosh   \n",
       "1     trouble sleep confuse mind restless heart tune   \n",
       "2  wrong dear forward doubt stay restless restles...   \n",
       "3                                  shift focus worry   \n",
       "4                   restless restless month boy mean   \n",
       "\n",
       "                                              tokens  wrd_len  \n",
       "0                                     [oh, my, gosh]        3  \n",
       "1  [trouble, sleeping, ,, confused, mind, ,, rest...       13  \n",
       "2  [All, wrong, ,, back, off, dear, ,, forward, d...       17  \n",
       "3  [I, 've, shifted, my, focus, to, something, el...       13  \n",
       "4  [I, 'm, restless, and, restless, ,, it, 's, be...       20  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#creating 3 pipelines\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "2f7b9b82",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 34519 entries, 0 to 52840\n",
      "Data columns (total 8 columns):\n",
      " #   Column     Non-Null Count  Dtype \n",
      "---  ------     --------------  ----- \n",
      " 0   statement  34519 non-null  object\n",
      " 1   status     34519 non-null  object\n",
      " 2   len        34519 non-null  int64 \n",
      " 3   text       34519 non-null  object\n",
      " 4   stem_text  34519 non-null  object\n",
      " 5   lemm_text  34519 non-null  object\n",
      " 6   tokens     34519 non-null  object\n",
      " 7   wrd_len    34519 non-null  int64 \n",
      "dtypes: int64(2), object(6)\n",
      "memory usage: 3.4+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "3f62fdb6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='status', ylabel='count'>"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmAAAAGtCAYAAACx9cgnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABGNklEQVR4nO3deViVdf7/8ZeIBogBpYGW5oyAzTg6IAhupWJkZW6ImZENaWpA02i5lZqmgjpTaliiUso00Wi4RmOlU5qkgkumLSOBLaImIIqyKsv9+6Mv5+cJ14T7kDwf19V1de73fd/n/bk9nPt17uWcBoZhGAIAAIBp7GzdAAAAQH1DAAMAADAZAQwAAMBkBDAAAACTEcAAAABMRgADAAAwGQEMAADAZAQwAAAAk9nbuoHfksrKSuXk5KhJkyZq0KCBrdsBAABXwTAMFRUV6bbbbpOdXd049kQAuwY5OTnq2bOnrdsAAAC/wqeffioPDw9btyGJAHZNmjRpIunnf0BnZ2cbdwMAAK5GYWGhevbsadmP1wUEsGtQddrR2dmZAAYAwG9MXbp8qG6cCAUAAKhHCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJrNpADt16pSCg4OVlpZmmXbo0CH95S9/ka+vr7p166a5c+eqvLzcUl+/fr2Cg4Pl4+OjkJAQ7d+/31KrqKjQ/Pnz1a1bN/n6+ioiIkI5OTmWel5eniIjI+Xv76/AwEBFR0dbrRsAAMAMNgtg+/bt07Bhw3TkyBHLtFOnTik8PFzdunXT7t279e6772rbtm365z//KUlKS0vT7NmzNW/ePO3Zs0cDBgxQRESESkpKJElxcXHasWOH1q5dq5SUFDk4OGjatGmW9Y8bN05OTk5KSUnRmjVrtGvXLiUkJJg6bgAAAJsEsPXr12vChAkaP3681fQNGzaoTZs2Gjt2rBo1aqQ77rhDK1as0AMPPCBJSkpKUr9+/eTn56dGjRopPDxcbm5u2rRpk6U+evRotWjRQs7Ozpo6daq2b9+urKws/fjjj9q9e7cmTpwoR0dHtWrVSpGRkUpMTDR9/AAAoH6zSQDr0aOHtmzZogcffNBq+sGDB+Xt7a0XX3xR3bt317333qv33nvP8rtNmZmZ8vb2tlrG09NThw4dUkFBgU6cOGFVb9asmVxcXJSenq6MjAy5urrK3d3dUm/btq2OHz+us2fP1uJoAQAArNkkgDVv3lz29tV/BenMmTNat26dOnbsqG3btum1117T6tWrtXLlSklSUVGRHB0drZZxcHBQcXGxioqKJElOTk7V6kVFRRddtupxcXFxjY0NAADgSurUXZCNGzdWhw4dFBoaqkaNGumuu+7SY489pg8++EDSz4GptLTUapnS0lI1adLEEqaqrgf7Zd3JyalarepxXfpxTgAAcOOrUwGsbdu2On/+vNW0yspKGYYhSfLy8lJGRoZVPTMzU15eXnJxcZG7u7syMzMttdzcXOXn58vb21teXl7Kz8/XyZMnLfXDhw/Lw8NDTZs2rcVRAQAAWKtTAWzIkCH69ttvFR8fr4qKCqWnp+vtt9/WwIEDJUmhoaFKTk5WamqqysrKlJCQoLy8PAUHB0uSQkJCFBcXp6ysLBUWFiomJkYBAQFq3bq12rRpIz8/P8XExKiwsFBZWVlasmSJQkNDbTlkAABQD1W/EMuG2rZtq7ffflt///vftXz5cjk4OGj48OEaMWKEJKlr166aMWOGZs6cqezsbHl6eio+Pl6urq6SpKioKJWXlyssLExFRUUKDAzUokWLLOuPjY3VrFmz1KdPH9nZ2WnQoEGKjIy0wUjrn4rKSjW0q1N5v8bVhzECAGpGA6Pq/B6uqLCwUH5+ftq3b5+cnZ1t3c5vzrzF63Tk2Mkrz/gb1Pr2Zpry1xBbtwEAuIi6uP+uU0fAcGM7cuykMr8/Yes2AACwOc6XAAAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMpsGsFOnTik4OFhpaWnVajk5OerWrZvWrVtnNX39+vUKDg6Wj4+PQkJCtH//fkutoqJC8+fPV7du3eTr66uIiAjl5ORY6nl5eYqMjJS/v78CAwMVHR2t8vLy2hsgAADARdgsgO3bt0/Dhg3TkSNHqtUqKys1YcIEnT592mp6WlqaZs+erXnz5mnPnj0aMGCAIiIiVFJSIkmKi4vTjh07tHbtWqWkpMjBwUHTpk2zLD9u3Dg5OTkpJSVFa9as0a5du5SQkFCr4wQAAPglmwSw9evXa8KECRo/fvxF66+//ro8PDzUokULq+lJSUnq16+f/Pz81KhRI4WHh8vNzU2bNm2y1EePHq0WLVrI2dlZU6dO1fbt25WVlaUff/xRu3fv1sSJE+Xo6KhWrVopMjJSiYmJtT5eAACAC9kkgPXo0UNbtmzRgw8+WK2Wmpqq//znP5oxY0a1WmZmpry9va2meXp66tChQyooKNCJEyes6s2aNZOLi4vS09OVkZEhV1dXubu7W+pt27bV8ePHdfbs2RocHQAAwOXZ2+JJmzdvftHpeXl5euGFFxQbG6smTZpUqxcVFcnR0dFqmoODg4qLi1VUVCRJcnJyqlavqv1y2arHxcXFuvnmm3/dYAAAAK6RTQLYxRiGoUmTJmnEiBH605/+dNF5HB0dVVpaajWttLRUbm5uljBVdT3YhfUmTZrIMIxqtarHFwt7AAAAtaXOfA3FTz/9pN27d+v111+Xv7+//P39dfz4cb300ksaO3asJMnLy0sZGRlWy2VmZsrLy0suLi5yd3dXZmampZabm6v8/Hx5e3vLy8tL+fn5OnnypKV++PBheXh4qGnTpuYMEgAAQHUogLVs2VJffvml9u7da/mvZcuWmjFjhpYtWyZJCg0NVXJyslJTU1VWVqaEhATl5eUpODhYkhQSEqK4uDhlZWWpsLBQMTExCggIUOvWrdWmTRv5+fkpJiZGhYWFysrK0pIlSxQaGmrLYQMAgHqozpyCvBpdu3bVjBkzNHPmTGVnZ8vT01Px8fFydXWVJEVFRam8vFxhYWEqKipSYGCgFi1aZFk+NjZWs2bNUp8+fWRnZ6dBgwYpMjLSNoMBAAD1VgPDMAxbN/FbUVhYKD8/P+3bt0/Ozs62buc3J3LKcmV+f8LWbdQKz995aMm8MbZuAwBwEXVx/11nTkECAADUFwQwAAAAkxHAAAAATEYAAwAAMBkBDAAAwGQEMAAAAJMRwAAAAExGAAMAADAZAQwAAMBkBDAAAACTEcAAAABMRgADAAAwGQEMAADAZAQwAAAAkxHAAAAATEYAAwAAMBkBDAAAwGQEMAAAAJMRwAAAAExGAAMAADAZAQwAAMBkBDAAAACTEcAAAABMRgADAAAwGQEMAADAZAQwAAAAkxHAAAAATEYAAwAAMBkBDAAAwGQEMAAAAJMRwAAAAExGAAMAADAZAQwAAMBkBDAAAACT2TSAnTp1SsHBwUpLS7NM++ijjzRw4EB16tRJQUFBeu2111RZWWmpr1+/XsHBwfLx8VFISIj2799vqVVUVGj+/Pnq1q2bfH19FRERoZycHEs9Ly9PkZGR8vf3V2BgoKKjo1VeXm7OYAEAAP6PzQLYvn37NGzYMB05csQy7auvvtKkSZM0btw47d27V/Hx8Vq3bp0SEhIkSWlpaZo9e7bmzZunPXv2aMCAAYqIiFBJSYkkKS4uTjt27NDatWuVkpIiBwcHTZs2zbL+cePGycnJSSkpKVqzZo127dplWTcAAIBZbBLA1q9frwkTJmj8+PFW048dO6ZHHnlEvXv3lp2dndq2bavg4GDt2bNHkpSUlKR+/frJz89PjRo1Unh4uNzc3LRp0yZLffTo0WrRooWcnZ01depUbd++XVlZWfrxxx+1e/duTZw4UY6OjmrVqpUiIyOVmJho+vgBAED9ZpMA1qNHD23ZskUPPvig1fS+ffvq+eeftzwuLS3Vtm3b1L59e0lSZmamvL29rZbx9PTUoUOHVFBQoBMnTljVmzVrJhcXF6WnpysjI0Ourq5yd3e31Nu2bavjx4/r7NmztTFMAACAi7JJAGvevLns7e0vO09hYaGioqLk4OCg8PBwSVJRUZEcHR2t5nNwcFBxcbGKiookSU5OTtXqRUVFF1226nFxcfH1DAcAAOCa1Mm7IL/77js98sgjKi8v11tvvSVnZ2dJPwem0tJSq3lLS0vVpEkTS5iquh7sl3UnJ6dqtarHTZo0qa2hAAAAVFPnAtinn36qoUOH6u6779abb74pFxcXS83Ly0sZGRlW82dmZsrLy0suLi5yd3dXZmampZabm6v8/Hx5e3vLy8tL+fn5OnnypKV++PBheXh4qGnTprU/MAAAgP9TpwLYF198oaioKD3//POaPHlytdOUoaGhSk5OVmpqqsrKypSQkKC8vDwFBwdLkkJCQhQXF6esrCwVFhYqJiZGAQEBat26tdq0aSM/Pz/FxMSosLBQWVlZWrJkiUJDQ20xVAAAUI9d/kIsky1dulTl5eWKjo5WdHS0Zbqfn5/eeOMNde3aVTNmzNDMmTOVnZ0tT09PxcfHy9XVVZIUFRWl8vJyhYWFqaioSIGBgVq0aJFlPbGxsZo1a5b69OkjOzs7DRo0SJGRkSaPEgAA1HcNDMMwbN3Eb0VhYaH8/Py0b98+y3VpuHqRU5Yr8/sTtm6jVnj+zkNL5o2xdRsAgIuoi/vvOnUKEgAAoD4ggAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDKbBrBTp04pODhYaWlplmkHDhzQ0KFD5evrq6CgICUlJVkts379egUHB8vHx0chISHav3+/pVZRUaH58+erW7du8vX1VUREhHJyciz1vLw8RUZGyt/fX4GBgYqOjlZ5eXntDxQAAOACNgtg+/bt07Bhw3TkyBHLtDNnzmjMmDEaNGiQ9uzZo+joaM2dO1cHDx6UJKWlpWn27NmaN2+e9uzZowEDBigiIkIlJSWSpLi4OO3YsUNr165VSkqKHBwcNG3aNMv6x40bJycnJ6WkpGjNmjXatWuXEhISTB03AACATQLY+vXrNWHCBI0fP95q+ubNm+Xq6qqwsDDZ29ura9eu6t+/vxITEyVJSUlJ6tevn/z8/NSoUSOFh4fLzc1NmzZtstRHjx6tFi1ayNnZWVOnTtX27duVlZWlH3/8Ubt379bEiRPl6OioVq1aKTIy0rJuAAAAs9gkgPXo0UNbtmzRgw8+aDU9IyND3t7eVtM8PT116NAhSVJmZuYl6wUFBTpx4oRVvVmzZnJxcVF6eroyMjLk6uoqd3d3S71t27Y6fvy4zp49W9NDBAAAuCR7Wzxp8+bNLzq9qKhIjo6OVtMcHBxUXFx8xXpRUZEkycnJqVq9qvbLZaseFxcX6+abb/6VowEAALg2deouSEdHR5WWllpNKy0tVZMmTa5YrwpTVdeD/bLu5ORUrVb1uGr9AAAAZqhTAczb21sZGRlW0zIzM+Xl5SVJ8vLyumTdxcVF7u7uyszMtNRyc3OVn58vb29veXl5KT8/XydPnrTUDx8+LA8PDzVt2rQWRwUAAGCtTgWw4OBgnTx5UgkJCSorK1NqaqqSk5M1ZMgQSVJoaKiSk5OVmpqqsrIyJSQkKC8vT8HBwZKkkJAQxcXFKSsrS4WFhYqJiVFAQIBat26tNm3ayM/PTzExMSosLFRWVpaWLFmi0NBQWw4ZAADUQza5BuxS3NzctGLFCkVHRys2Nla33HKLpk2bpi5dukiSunbtqhkzZmjmzJnKzs6Wp6en4uPj5erqKkmKiopSeXm5wsLCVFRUpMDAQC1atMiy/tjYWM2aNUt9+vSRnZ2dBg0apMjISBuMFAAA1GcNDMMwbN3Eb0VhYaH8/Py0b98+OTs727qd35zIKcuV+f0JW7dRKzx/56El88bYug0AwEXUxf13nToFCQAAUB8QwAAAAExGAAMAADAZAQwAAMBkBDAAAACTEcAAAABMRgADAAAwGQEMAADAZAQwAAAAkxHAAAAATEYAAwAAMBkBDAAAwGQEMAAAAJMRwAAAAExGAAMAADAZAQwAAMBkBDAAAACTXXMAi4iIuOj0xx577LqbAQAAqA/sr2amo0ePasOGDZKkzz77TK+99ppVvbCwUOnp6TXeHAAAwI3oqgJYy5YtlZGRoVOnTqmiokJpaWlW9ZtuukkzZsyolQYBAABuNFcVwOzs7PTqq69KkqZNm6Y5c+bUalMAAAA3sqsKYBeaM2eOzp8/r1OnTqmystKq1rJlyxprDAAA4EZ1zQHsww8/1PTp01VYWGiZZhiGGjRooP/973812hwAAMCN6JoDWGxsrMLCwjR48GDZ21/z4gAAAPXeNSeon376SU8//TThCwAA4Fe65u8Ba9++vTIzM2ujFwAAgHrhmg9jderUSeHh4br//vvVrFkzq9rTTz9dY40BAADcqK45gO3fv19eXl46fPiwDh8+bJneoEGDGm0MAADgRnXNAexf//pXbfQBAABQb1xzAKv6SaKLGTRo0HW0AgAAUD/8qq+huNCZM2dUUlIiPz8/AhgAAMBVuOYA9sknn1g9NgxD8fHxys/Pr6meAAAAbmjX/DUUv9SgQQONGjVKGzdurIl+AAAAbnjXHcAk6fvvv+cuSAAAgKt0zacgR4wYYRW2ysrKlJ6ergEDBtRYU19//bViYmKUnp4uBwcH3X///Zo0aZIaN26sAwcOaM6cOcrMzJSbm5siIiI0dOhQy7Lr16/XkiVLlJubq9///veaPn26fH19JUkVFRV6+eWXtXHjRpWUlKhLly566aWXdNttt9VY7wAAAFdyzQEsMDDQ6rGdnZ3Cw8N177331khDlZWVGjt2rMaMGaN//etfysnJUXh4uNzc3PTYY49pzJgxeuaZZzRs2DDt2bNHUVFRateunTp27Ki0tDTNnj1b8fHx6tixoxITExUREaGtW7fK0dFRcXFx2rFjh9auXaumTZtq+vTpmjZtmpYvX14jvQMAAFyNaw5gF37bfV5enlxcXGr0dyHPnDmj3NxcVVZWyjAMST+HPEdHR23evFmurq4KCwuTJHXt2lX9+/dXYmKiOnbsqKSkJPXr109+fn6SpPDwcK1evVqbNm3SkCFDlJSUpAkTJqhFixaSpKlTp6pHjx7KyspSq1atamwMAAAAl3PN14CVlZUpJiZGvr6+6tGjh/z8/DR9+nSdP3++Rhpyc3NTeHi45s+frw4dOqhnz55q06aNwsPDlZGRIW9vb6v5PT09dejQIUlSZmbmJesFBQU6ceKEVb1Zs2ZycXFRenp6jfQOAABwNa45gC1ZskRpaWlatGiR3n//fS1atEgHDhzQokWLaqShyspKOTg4aPr06friiy/0/vvv6/Dhw4qNjVVRUZEcHR2t5ndwcFBxcbEkXbZeVFQkSXJycqpWr6oBAACY4ZoDWHJysl577TX17NlTbdu2Ve/evfXaa68pOTm5RhrasmWLPvroIz366KNq3LixvLy8FBUVpX//+99ydHRUaWmp1fylpaVq0qSJJF22XhXMSkpKLrk8AACAGa45gJ05c8ZyDVWVFi1aVAs+v9ZPP/1U7XSmvb29GjVqJG9vb2VkZFjVMjMz5eXlJUny8vK6ZN3FxUXu7u7KzMy01HJzc5Wfn1/ttCUAAEBtuuYA1q5dO61atcpq2qpVq2osxPTo0UO5ublaunSpKioqlJWVpbi4OPXv31/BwcE6efKkEhISVFZWptTUVCUnJ2vIkCGSpNDQUCUnJys1NVVlZWVKSEhQXl6egoODJUkhISGKi4tTVlaWCgsLFRMTo4CAALVu3bpGegcAALga13z74rhx4zRy5Ei99957atWqlY4cOaLMzEy9+eabNdKQp6enli1bpkWLFumNN95Q06ZNNWDAAEVFRalx48ZasWKFoqOjFRsbq1tuuUXTpk1Tly5dJP18V+SMGTM0c+ZMZWdny9PTU/Hx8XJ1dZUkRUVFqby8XGFhYSoqKlJgYGCNXbsGAABwtRoYVd/1cA1Wr16tAwcOyN7eXh4eHmrevLnVl6HeqAoLC+Xn56d9+/bJ2dnZ1u385kROWa7M70/Yuo1a4fk7Dy2ZN8bWbQAALqIu7r+v+QhYbGys1q9fr5UrV6pNmzb6+OOPFRMTozNnzujJJ5+sjR4BAABuKNd8DdiaNWv01ltvqU2bNpKkPn36aOXKlUpMTKzp3gAAAG5I1xzACgsLL3oXZNV3cQG4NhWVlbZuodbVhzECwLW45lOQ7du31/LlyxUZGWmZtmLFCt1111012hhQXzS0s9OMt9brhxMnbd1KrWjj0UwvPT7Y1m0AQJ1yzQFsypQpGjlypN599115eHjoxIkTKi8v1xtvvFEb/QH1wg8nTurbozfmDQoAgOp+1RGwzZs3a+vWrcrJyVGLFi3Uq1cvNW3atDb6AwAAuOFccwCTJBcXFw0aNKiGWwEAAKgfrvkifAAAAFwfAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYrE4GsPz8fE2aNEmBgYHq3LmzIiMjlZOTI0k6cOCAhg4dKl9fXwUFBSkpKclq2fXr1ys4OFg+Pj4KCQnR/v37LbWKigrNnz9f3bp1k6+vryIiIizrBVD3VFRW2rqFWnWjjw/ApdnbuoGL+etf/yoXFxdt2bJFdnZ2ev755zV9+nT9/e9/15gxY/TMM89o2LBh2rNnj6KiotSuXTt17NhRaWlpmj17tuLj49WxY0clJiYqIiJCW7dulaOjo+Li4rRjxw6tXbtWTZs21fTp0zVt2jQtX77c1kMGcBEN7ez0woa1+u7kSVu3UuN+36yZYgYNsXUbAGykzgWwr776SgcOHNDOnTvl7OwsSZo9e7Zyc3O1efNmubq6KiwsTJLUtWtX9e/fX4mJierYsaOSkpLUr18/+fn5SZLCw8O1evVqbdq0SUOGDFFSUpImTJigFi1aSJKmTp2qHj16KCsrS61atbLNgAFc1ncnT+rQiZ9s3QYA1Kg6dwry4MGD8vT01Lvvvqvg4GD16NFD8+fPV/PmzZWRkSFvb2+r+T09PXXo0CFJUmZm5iXrBQUFOnHihFW9WbNmcnFxUXp6eu0PDAAA4P/UuQB25swZpaen64cfftD69eu1YcMGZWdna/LkySoqKpKjo6PV/A4ODiouLpaky9aLiookSU5OTtXqVTUAAAAz1LkA1rhxY0k/nx50dnZWs2bNNG7cOH366acyDEOlpaVW85eWlqpJkyaSJEdHx0vWq4JZSUnJJZcHAAAwQ50LYJ6enqqsrFRZWZllWuX/3Sn0hz/8QRkZGVbzZ2ZmysvLS5Lk5eV1ybqLi4vc3d2VmZlpqeXm5io/P7/aaUsAAIDaVOcCWLdu3dSqVSu98MILKioq0qlTp7Rw4ULde++9euihh3Ty5EklJCSorKxMqampSk5O1pAhP99JFBoaquTkZKWmpqqsrEwJCQnKy8tTcHCwJCkkJERxcXHKyspSYWGhYmJiFBAQoNatW9tyyAAAoJ6pc3dBNmrUSP/61780b9489e3bV+fOnVNQUJCmTp2qm2++WStWrFB0dLRiY2N1yy23aNq0aerSpYukn++KnDFjhmbOnKns7Gx5enoqPj5erq6ukqSoqCiVl5crLCxMRUVFCgwM1KJFi2w3WAAAUC/VuQAmSe7u7lq4cOFFax06dNCqVasuuezAgQM1cODAi9YaNWqkCRMmaMKECTXSJwAAwK9R505BAgAA3OgIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJquzAayiokIjRozQlClTLNMOHDigoUOHytfXV0FBQUpKSrJaZv369QoODpaPj49CQkK0f/9+q/XNnz9f3bp1k6+vryIiIpSTk2PaeAAAAKrU2QD22muvae/evZbHZ86c0ZgxYzRo0CDt2bNH0dHRmjt3rg4ePChJSktL0+zZszVv3jzt2bNHAwYMUEREhEpKSiRJcXFx2rFjh9auXauUlBQ5ODho2rRpNhkbAACo3+pkANu1a5c2b96s++67zzJt8+bNcnV1VVhYmOzt7dW1a1f1799fiYmJkqSkpCT169dPfn5+atSokcLDw+Xm5qZNmzZZ6qNHj1aLFi3k7OysqVOnavv27crKyrLJGAEAQP1V5wJYXl6epk6dqldeeUWOjo6W6RkZGfL29raa19PTU4cOHZIkZWZmXrJeUFCgEydOWNWbNWsmFxcXpaen1+JoAAAAqqtTAayyslITJ07UE088obvuusuqVlRUZBXIJMnBwUHFxcVXrBcVFUmSnJycqtWragAAAGapUwFs2bJlaty4sUaMGFGt5ujoqNLSUqtppaWlatKkyRXrVcGs6nqwiy0PAABgFntbN3ChjRs3KicnR/7+/pJkCVT//e9/NWnSJO3YscNq/szMTHl5eUmSvLy8lJGRUa1+zz33yMXFRe7u7lanKXNzc5Wfn1/ttCUAAEBtq1NHwD788EN9/vnn2rt3r/bu3auHHnpIDz30kPbu3avg4GCdPHlSCQkJKisrU2pqqpKTkzVkyBBJUmhoqJKTk5WamqqysjIlJCQoLy9PwcHBkqSQkBDFxcUpKytLhYWFiomJUUBAgFq3bm3LIQMAgHqoTh0Buxw3NzetWLFC0dHRio2N1S233KJp06apS5cukqSuXbtqxowZmjlzprKzs+Xp6an4+Hi5urpKkqKiolReXq6wsDAVFRUpMDBQixYtst2AAABAvVWnA9i8efOsHnfo0EGrVq265PwDBw7UwIEDL1pr1KiRJkyYoAkTJtRojwAAANeqTp2CBAAAqA8IYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmKxOBrBDhw7piSeeUEBAgLp3765Jkybp1KlTkqQDBw5o6NCh8vX1VVBQkJKSkqyWXb9+vYKDg+Xj46OQkBDt37/fUquoqND8+fPVrVs3+fr6KiIiQjk5OaaODQAAoM4FsNLSUj355JPy9fXVZ599pvfff1/5+fl64YUXdObMGY0ZM0aDBg3Snj17FB0drblz5+rgwYOSpLS0NM2ePVvz5s3Tnj17NGDAAEVERKikpESSFBcXpx07dmjt2rVKSUmRg4ODpk2bZsvhAgCAeqjOBbDjx4/rrrvuUlRUlBo3biw3NzcNGzZMe/bs0ebNm+Xq6qqwsDDZ29ura9eu6t+/vxITEyVJSUlJ6tevn/z8/NSoUSOFh4fLzc1NmzZtstRHjx6tFi1ayNnZWVOnTtX27duVlZVlyyEDAIB6ps4FsN///vd644031LBhQ8u0jz76SO3bt1dGRoa8vb2t5vf09NShQ4ckSZmZmZesFxQU6MSJE1b1Zs2aycXFRenp6bU4IgAAAGt1LoBdyDAMLVy4UFu3btXUqVNVVFQkR0dHq3kcHBxUXFwsSZetFxUVSZKcnJyq1atqAAAAZrC3dQOXUlhYqOeff15ff/213n77bbVr106Ojo4qKCiwmq+0tFRNmjSRJDk6Oqq0tLRa3c3NzRLMqq4Hu9jyAAAAZqiTR8COHDmiIUOGqLCwUGvWrFG7du0kSd7e3srIyLCaNzMzU15eXpIkLy+vS9ZdXFzk7u6uzMxMSy03N1f5+fnVTlv+WpUVlTWynrqsPowRAIDaVueOgJ05c0Z/+ctf1KVLF0VHR8vO7v9nxODgYP3jH/9QQkKCwsLCtG/fPiUnJ2vJkiWSpNDQUEVFRemBBx6Qn5+fEhMTlZeXp+DgYElSSEiI4uLi1KFDB7m5uSkmJkYBAQFq3bp1jfRu19BO/5jxtrJ+yK6R9dU1rdq4a+JLj9m6DQAAfvPqXABbt26djh8/rg8++EAffvihVW3//v1asWKFoqOjFRsbq1tuuUXTpk1Tly5dJEldu3bVjBkzNHPmTGVnZ8vT01Px8fFydXWVJEVFRam8vFxhYWEqKipSYGCgFi1aVKP9Z/2QrcPpx2p0nQAA4MZS5wLYE088oSeeeOKS9Q4dOmjVqlWXrA8cOFADBw68aK1Ro0aaMGGCJkyYcN19AgAA/Fp18howAACAGxkBDAAAwGQEMAAAAJMRwADgN6ii8sb+SpgbfXxAnbsIHwBwZQ3t7DTz0yT9cCbH1q3UuDYut2lmz6G2bgOoVQQwAPiN+uFMjr7N+8nWbQD4FTgFCQAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAAGAyAhgAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAIAbRqVx4/+Id30YY33Ab0ECAG4Ydg3s9M+v3tKJ4mxbt1IrPJzc9Zc/PW7rNlADCGAAgBvKieJsHS04aus2gMviFCQAAIDJCGAAAAAmI4ABAACYjAAGAABgMgIYAACAyQhgAAAAJiOAAQAAmIwABgAAYDICGAAAgMkIYAAAACYjgAEAAJiMAAYAQD1QaVTYuoVa91saIz/GDQBAPWDXoKE+y5insyVZtm6lVtzs2Eo9vKbYuo2rRgADAKCeOFuSpVPFmbZuA+IUJAAAgOnqXQDLy8tTZGSk/P39FRgYqOjoaJWXl9u6LQAAUI/UuwA2btw4OTk5KSUlRWvWrNGuXbuUkJBg67YAAEA9Uq8C2I8//qjdu3dr4sSJcnR0VKtWrRQZGanExERbtwYAAOqRenURfkZGhlxdXeXu7m6Z1rZtWx0/flxnz57VzTfffNnlDcOQJBUWFl5yHvfbXVVefr5mGq5j3G93vezYr8SjubPKy9xqsKO6w6O583Vtm9vdnFV+/sbcNre7Xd+2aeXsrAq3W2qwo7qhlfP1bRdJur3xzSpvcq6GOqo7bm9883VtGze56rz9jfk+7Kbrex9uZLjLSTfmZTeNDPdLbpuq6VX78bqggVGXuqllGzdu1MKFC7Vt2zbLtCNHjig4OFiffvqpPDw8Lrv8iRMn1LNnz1ruEgAA1Iar2debpV4dAXNyclJJSYnVtKrHTZo0ueLyt912mz799FM1adJEDRo0qJUeAQBAzTIMQ0VFRbrtttts3YpFvQpgXl5eys/P18mTJ9WsWTNJ0uHDh+Xh4aGmTZtecXk7O7s6k5wBAMDVu5r9vJnq1UX4bdq0kZ+fn2JiYlRYWKisrCwtWbJEoaGhtm4NAADUI/XqGjBJOnnypGbNmqW0tDTZ2dlp0KBBmjBhgho2bGjr1gAAQD1R7wIYAACArdWrU5AAAAB1AQEMAADAZAQwAAAAkxHAAKCe+/HHH23dAlDvEMDqoYKCAp06dcrWbVi0a9dOY8aMqfYTEevWrVNQUJCNurK2ePFijRgx4rrWERQUpA4dOsjX11e+vr7y8fHRwIEDlZSUVENdmmPp0qV68sknbfb8Z86c0cyZM9WzZ0/5+PioR48emjx5sk6cOHHFZY8fPy5fX18dP378uua7ltdmTbx2rseVttf8+fMVFxdns/7M1q5dO3Xs2NHyN9i5c2dFRETop59+sszTr18/vffee9f9XHXpPcxMFRUVysrKsnUbdR4BzAYSExPVrl07JSQk1Mj6rnanUiU4OFgZGRk18tw15dNPP9Ubb7xh6zZq3UsvvaT9+/dr//792r17t6KiojRv3jwtX77c1q1dtaeeesqm/1bjx4/X6dOntWbNGn3xxRfasGGDzp8/ryeeeELl5Zf/jbuWLVtq//79atmyZY3M91twpe11+vRpW7douvj4eO3fv19ffPGFtm7dKsMwNHHiREv9P//5jwYMGFDjz3uxD2E9evTQ/PnzVVlZWePPV5vS0tLUrl07SdX3QePHj9eGDRtq5HmmTJmiKVOmSDL3w58ZH5wIYDaQmJio4cOH66233rriDuNqXOvOoi6+4Y4YMUKvvvqqPv/880vOk56ertGjRysgIED33HOPZs6cqYKCAkk/f9IMCQnRyJEj5e/vr+TkZI0YMUKxsbEaPny4fHx8NGDAAB08eFDPPfecOnXqpKCgIKvfBV2zZo1CQkIUGBgoX19fjR07tlaPFDZu3Fj33XefJk+erNdee02FhYU6efKkJkyYoO7du6tHjx568cUXLT8im5aWpnvuuUevvvqqAgMDFRgYqOjoaJ0///OPDk+ZMkXPPPOMHnjgAXXp0kVHjhy57PrKy8s1c+ZMde/eXYGBgXr00Ue1b98+ST//cO348eMVGBio7t27a9SoUTp8+LCk6m9M//3vfxUSEqJOnTqpb9++SkhIsOxMpkyZohdffFFPPfWUfH191adPH7311lvXtd327dun4OBgNW/eXJLUrFkzvfDCC/rzn/+ss2fPKigoSOvWrbPMf+GO4ujRo2rXrp2OHj0qScrKytJTTz0lPz8/de3aVTNnztT58+erzXf48GGNGDFCvr6+6t+/v7755hurnsx+7VyLy22vZcuWKTk5WcnJyZbA0a5dO82ZM0eBgYF66qmnJEk7d+5UaGio/P39qx0dysjIUFhYmDp37qzevXtr8uTJltfYnj17FBISIn9/fwUHBys6OrpG3vNqkrOzsx5++GF99dVXlmkXvoZGjBihefPmKSQkRD4+PgoJCdHevXst817ufemXioqK1KxZMzVu3FiSFBgYqAULFmjDhg16/PHHNXLkSA0ZMkQBAQHas2dPLY66Zv1yH1Rb+xhbf/iraQQwk+3atUt5eXmaMmWKKisr9dFHH1lqQUFBWrZsmQYNGiRfX18NGjRIqampkn7+RPanP/1Jhw4dkiR988036tixo7Zv315tZ3G5nW7fvn0lSaNHj1Z8fLweeOABLV261KrH/v37a82aNbW+LS4UHBysYcOG6dlnn1V+fn61+unTp/X444/L09NT27dv19q1a/X9999r0qRJlnm+/vpr9e/fXzt37lRwcLAkafXq1Zo9e7Z2796tm2++WY8++qgeeOABpaWlqW/fvpo9e7Yk6eDBg5ozZ45mzpyptLQ0ffDBB/rhhx+uOyxcjV69euncuXPat2+fIiMjZWdnp48++kjJycnKycnRiy++aJk3Oztb33//vT7++GOtXr1a27Zt05IlSyz1lJQUvfrqq9q8ebPuuOOOy65v48aN2r9/vz744APt3LlTnTt31ksvvSRJWrFihQoLC/Xpp59q69atat68uV5++eVqvaempmrcuHF68skntXv3bi1YsEArV6602m7r1q3TiBEjtGfPHo0ePVrz5s1Tdnb2r95e/fr104wZMzRz5kxt2rRJx44dU/PmzTVv3jzdcsstV72e8vJyjRo1Ss2bN9f27dv1/vvv64svvtDixYut5isrK9PYsWPl5eWl1NRULViwQP/9738tdVu+dq7G5bZXVFSU+vfvr/79+1uFqiNHjmjbtm36+9//rkOHDikiIkJjxoxRWlqaZs+erZiYGKWkpEj6+ahu165dtXv3bq1du1bffPON5bT6pEmTNGLECO3du1crV67Uhx9+qI8//tgm2+FSzpw5o//85z+67777LjnP6tWrNWnSJO3evVvBwcGKiIjQ6dOnr+p9qcrBgwd15swZDRo0yOp1UvW3l5ubq127dmncuHF6+OGHNWXKFAUEBGj06NFW1+j9MiBf7sOSJCUlJalfv37q1KlTtX/nESNG6JVXXlFYWJh8fX31wAMPaNOmTZb6559/rscff1w9evRQhw4dFBISoi+++KLa2C7cB02dOlV79+7VsmXL9NRTT+nFF1/UyJEjreafNWvWRbeRJH388cfq16+ffHx8NHbsWKswd+GHv+sd95QpU9S7d2/16tVLhYWF+vzzzzVkyBD5+PjokUcesexPq1zuQ8jFPvxeFQOmioiIMF5++WXDMAxj2bJlRkhIiKXWu3dvIzg42Pjhhx+M4uJiY/LkyUbfvn0t9SlTphj9+/c38vPzjfvuu8945ZVXDMMwjKysLMPb29vIysoyKioqjKFDhxoTJ040CgoKjFOnThljx441xo8fb1mPt7e3kZqaahiGYSxfvty4//77LbUvv/zS8PHxMQoLC2t1O1yoqp9z584ZgwcPNsaOHWtUVlYaa9euNXr37m0YhmEkJSUZ3bt3N8rLyy3LZWRkGN7e3kZOTo6xdu1ao3379kZFRYWl/thjjxkvvfSS5fHChQuN/v37Wx7v3LnTaNeunWEYhlFSUmJkZWUZhmEY+fn5xoEDB4xHHnnEmDJlimEYhhEbG2s89thj1zXO3r17G2vXrq02/fz584a3t7exePFi449//KPVtv/uu+8Mb29v49SpU0ZqaqrRrl0749SpU5b6qlWrLNto8uTJRnh4uKV24MCBy67v/fffNzp16mS8+eabRkZGhtW2W7ZsmdGjRw/j3//+t3HkyBGr2oXbYuLEicYzzzxjNZ63337b8rqdPHmyMXLkyGpj3b1797VtvAtUVFQYGzZsMMaMGWP4+fkZ3t7exr333mts3LjRMIzq2zk1NdXw9vY2DMP6byUtLc344x//aBQXF1vmzcrKMn788ceLzldSUmKZLyEhwbLdzXjtXI8rba/JkycbkydPtszv7e1tqRmGYcyYMcPq/cMwDOOVV14xxo4daxiGYYwcOdJ4+OGHjU2bNhmnT5+2eq307dvXGD16tPHJJ58YBQUFVjVb8fb2Nnx8fAw/Pz/D19fX8Pb2Nvz8/Iy9e/da5rnwNfTYY48Zs2fPttQqKyuNu+++21i7du1VvS9d+DqpWq7qdTJs2DBj1KhRRufOnY1Ro0YZwcHBxrx584xBgwYZR44cMUpLS43FixcbQUFBRmlpqaX/0aNHG8XFxcaZM2eMV1991XjyySeNkpIS49y5c8bkyZONp556yjAMw1i7dq3RqVMnY+fOnUZ5ebmxc+dOo1OnTsbmzZstYwsICDC+/vpr49y5c8aCBQsMPz8/o7S01CgpKTECAgKMt99+26ioqDCKioqMv/3tb8bw4cMNw7j031XVemNjYw3D+Pl96K677jJOnDhhGIZhnDt3zggICDB27dpV7d/m8OHDRvv27Y2NGzcaZWVlxpYtW4w//OEPltfnhX9L1zvuu+++2zhx4oRx5swZ49SpU4a/v7+xbNky4/z588bevXuNTp06WZ7rf//7n9GxY0fjo48+MsrLy419+/YZgYGBxvbt2w3D+PlvyMfHx0hPTzfOnDlz1a/FevVj3LZ27NgxpaSkWI5APPzww3r99de1e/duBQQESJJCQ0N15513Svr5SNSF59GnT5+ukJAQDR48WC1bttTf/va3as/x1Vdf6euvv9bKlSvVpEkTSdLkyZN1//33a/r06XJzc7Oaf9CgQVq0aJG+/PJLdejQQRs2bND9999vWdZMjRs31qJFizR48GCtWLHCqte8vDy1bNnS6iej7rjjDkk/b1dJat68uezsrA/qurq6Wv6/YcOGcnFxsTy2s7OzXPhvZ2ent956S8nJyXJyclK7du1UWFhY7caA2lB1qur3v/+9Kioq1LNnT6t648aNLRe0uri4WG2XFi1aKCcnx/L4tttus/z/0aNHL7u+fv36qaysTElJSVqwYIFuvfVWPfXUUxo+fLhGjx6txo0ba82aNZo1a5ZatWql5557rtpRgry8PP3hD3+wmnbHHXdY/k0kWU59SVKjRo0k6bqud7Gzs9PAgQM1cOBAGYahw4cPa+PGjZo0aZLVc11Jbm6u3Nzc5OjoaNW7JKtPv9nZ2XJzc5ODg4NlWuvWra36sdVr52r8mu114evo2LFjSk1Nlb+/v2VaRUWFZRssWrRIixcv1sKFC/Xss8+qU6dOmjlzpry8vPTPf/5Tixcv1ksvvaTc3Fzdfffdmjlzpjw8PGp30FewdOlSBQYGSpJKS0uVmJiov/zlL1q9erXat29fbf42bdpY/r9Bgwby8PBQbm6uJF3xfamKnZ2dCgsL9fzzz6tBgwZq2LChKisr1aRJE40aNUplZWUqKSnRqlWrFBsbq1atWkmSoqKi9O6772rbtm2WMxgPPfSQHB0d5ejoKAcHBx06dEgbNmxQ9+7dFRMTY3kfXLt2rYYNG6auXbtKkrp27aphw4Zp1apVlrMEffv21R//+EdJ0uDBg7V06VLl5eXJ3d1dq1ev1p133qlz587p2LFjcnV11ZdffnlN27pjx45q27at3n//fY0aNUrbtm2Ts7OzZftfaNOmTfrTn/5kOR1+7733qnfv3hdd7/WO+5577pG7u7skaf369XJ0dNTo0aPVoEED+fn5aciQIfrf//4nSVq1apX69Oljef/r1KmTHn74YSUmJuruu++WJPn4+Mjb2/uatg0BzETvvPOOysvLNXDgQMu08vJyrVixwhLAmjVrZqnZ29tbvYk7OTlpyJAhevnllxUVFXXR36+80k73lwGsefPmuvvuu7Vx40bdddddev/996udgjFT69atNXv2bE2aNEkhISGW6bfffruOHz+uiooKy7irDvM2b95c3333nRo0aFBtfRebdjEJCQnasWOHkpOTLf8GVde/1LZPPvlETk5OatasmRwcHJSWlmYZ4/nz55WVlaU777xT+/btU0FBgUpKSiyh4ejRo1bX/l04Xg8Pj8uu7/vvv1f79u01aNAglZaW6sMPP9TkyZPl7++viooKBQUFKTw8XAUFBXrnnXc0fvx4yynxKrfffnu1w+1ZWVnXFISuRUpKip555hlt3bpVrq6uatCggTw9PfXcc89px44d+uabb2RnZ6eysjLLMpe6HsXDw0OnT5+22p579+7VV199pXvvvdcyX4sWLXTq1CkVFRVZPphceMelLV87V3I12+tifvk6Gjx4sGbNmmWZlpOTI8MwVFlZqW+++UZ//etf9cILL+inn37S3LlzNWXKFL3zzjvKzMzUzJkzZW9vr++//17Tpk1TTEyMYmNja33sV8vBwUGjRo3S8uXLtXPnzosGsAtPmVdWVur48eNq0aKF7Ozsrvi+VCUhIUHnzp3TCy+8oL/85S+Sfn6duLq6KiIiQosXL1ZFRYWKi4v1t7/9zerDZFlZmVWguzAgX+7D0smTJy1Brsodd9yhTz75xPL4wr9Ve3t7yxgbNmyotLQ0jR49WsXFxfL09Ky2T7paISEh2rBhg0aNGqV169Zp8ODBF31vzs7OrnYtc+vWrS/6N3y9475wG2ZnZ6tFixZWPbVu3doSwK70IeSX67taXANmknPnzmnNmjWKjo7Wxo0bLf/FxcVp27ZtVueuL+XIkSOKi4vT0KFD9fe///2it91fuNPdu3ev9u7dq507d2rDhg2WTzm/NGTIEH344Yf67LPP1LRpU3Xu3Pm6x3s9HnzwQQ0ZMkSrV6+2TKsKlC+//LJKS0uVm5ur6OhodenSRbfffvt1P2dhYaHs7e3VqFEjlZeXa+PGjUpJSbHakde08+fPa9OmTVqwYIHGjx+vTp066c4779S8efNUVFSk0tJSxcTEKDw8XBUVFZJ+/qOfP3++zp07p++++05vvvmmQkNDL7r+jh07XnZ9W7du1dNPP62jR4/KwcFBrq6usre3V9OmTZWUlKRJkyYpLy9Pzs7OcnZ2lpOTk+Xi4SpDhgzRJ598og8++EAVFRX65ptvFB8fryFDhtTKNuvcubNuvfVWPf/880pPT1dZWZkKCwv13nvv6YcfflCvXr3Utm1bffzxx5bXyaWuxerYsaPatGmj+fPnq6SkRCdPntTcuXOrXTzv6+ur3/3ud5ozZ45KSkr0448/asWKFZa6LV47V+tqtlfjxo0vedG49PNR+ffff1+fffaZKisr9cMPP+ixxx7TihUrZGdnpzlz5mjRokU6d+6cbrnlFt10001yc3NTgwYN9Oyzz2rFihUqLy9X8+bNZW9vX+1DoK2Vl5dr7dq1Onv2rPz8/C46T1JSkr766iudP39er7/+ugzDUO/eva/pfanqOlx7e/tLvk7s7e110003acWKFZb3771792r9+vUaNmyYZb4Lg0J6erqCgoK0Zs0apaWlKSQkROPHj1dBQYHuuOOOX/0B6cCBA5o9e7YWLlyoHTt26J///Ke6d+9+xeUuZuDAgfruu++0f/9+7dixw+rD9YU8PDyqfX3Fpb5e5nrH/csPGceOHbM6Mn/h81Z9CLnw3+Sjjz6yunv9aj/sX4gAZpLk5GQ1aNBA/fv3l4eHh+W/e+65R97e3lf8SoqysjI9++yz6tevn+bMmaPOnTtr4sSJ1U7lXGmnK6naG26vXr1UUVGh2NjYS/5hmO2FF16wOrXVtGlTrVy5Ut9++6169uyphx56SLfffrteffXVGnm+kSNHqkWLFurdu7fuvvtuvffee3r00Uf17bff1sj6q8yYMcNyC/o999yjt99+Wy+99JIef/xx2dvba9myZTp58qTuu+8+9ejRQ0eOHNHKlSt10003Wdbh4uKiPn366PHHH9fgwYMveVv2ldb3+OOPq1evXnrkkUfk4+Ojf/zjH1q4cKE8PDz07LPP6s4777RcxLpu3TotWbLEqg9J+vOf/6xXX31V8fHx8vf319NPP63hw4fX2hEgBwcHvfPOO2revLkiIiLk7++vXr166b333tPKlSvVtm1bTZgwQUVFRerevbsef/zxS36dQKNGjbR06VJlZ2erV69eGjhwoDp37qxnnnnGar6GDRtq+fLlysnJUbdu3fTkk0+qT58+lrpZr51f42q214MPPqjPP/9cvXr1uug6/vznP2vBggVasGCBOnfurMcee0xBQUF67rnnJP18CvLw4cPq0aOHunXrpoKCAs2ePVuNGzdWXFycPv74YwUGBiooKEjNmzfXhAkTTNwCFzd69GjL32FAQIASExO1YMECderU6aLzBwQEaNasWerSpYvS0tK0YsUKNW3a9Jrel0aOHCl7e3vNmzfvkq+TBg0aKDQ0VK+88opOnDihyspKrV+/Xg899NAlvyz3ch+WQkNDtXr1au3atUsVFRVKTU3V6tWrr+oDUkFBgezs7Cyn3r/44gu99dZblruuL+eX+5hbb71VPXv21KxZs+Tv73/JO/YHDBigb7/9Vu+++67Ky8v12WefacuWLbU+7qCgIBmGocWLF+v8+fP66quvrL6f8XIfQq7LVV8thusyePBgY86cORetvf3220aHDh2Mjh07XvLi4X/84x9GUFCQ5YLqvLw8o0uXLsaSJUuqXQD5008/GePGjTO6detm+Pn5GU888YSRkZFhWe/s2bONjh07GgsWLLBMi4mJMe666y7jp59+qvGxo2Zc+HoAYI4LLyi/Xpe6EedCpaWlxj/+8Q+jd+/ehq+vrzFgwABjy5YtlvqFN1EZhmEUFhYakyZNMgIDAw0fHx8jJCTE6iaXd99913jwwQcNHx8fo2/fvsa///3vS47twn1JZWWlMXfuXCMgIMDw9/c3Bg0aZLzxxhvGH/7wByM3N/eyF+G/9957RqdOnSwX7BuGYfz3v/81vL29jeTk5MuOf+fOncaAAQOMjh07GsOHDzfGjRt30Yvwa3LchvHzhfbDhg0z/vznPxsDBgwwpk+fbnXzzNatW43BgwcbnTp1Mrp3727MnTvXOHfunGEY1W9kuVoNDKOOXCkKm3rrrbe0ffv2G+o7Vm40aWlpevzxx5Wenm7rVoB6Y8SIEQoICNBf//pXW7fym3bo0CGNGDFCn332WbUj6fUVpyDrudzcXB08eFD//Oc/NXz4cFu3AwC4gRQWFurbb7/VokWLFBISQvi6AEfA6rmkpCTNmTNHAwcOtLrLCQCA65WZmamhQ4fqrrvu0tKlS62+Cqi+I4ABAACYjFOQAAAAJiOAAQAAmIwABgAAYDICGIAbWkFBQbVvt7+cH374ofaaAYD/QwADcEMLDg5WRkbGVc37ySefaNSoUbXcEQAQwADc4C71Y9wXk5+f/6t+bBgArhUBDMANYfHixerZs6cCAgI0ZMgQffzxx+rbt6+kn3/7Lz4+XoZhaPny5erfv7/8/f3VuXNnPffccyotLVVaWppmzJih48ePy9fXV9nZ2RoxYoQWL15seY6jR4+qXbt2Onr0qCTpnXfe0b333it/f3/179/f6vfjAOBy7G3dAABcr6of2123bp2aN2+u1atXa+rUqUpJSdGf/vQnxcfHKzAwUJs2bdJbb72lt99+W23atNHhw4f16KOPKjk5WUOHDtVLL72k1157TZ988skVnzMrK0tz587Vxo0b9fvf/14pKSmKiopSz549ddttt5kwagC/ZRwBA/Cbd9NNN+nMmTN699139c0332jo0KHatWuXGjVqZDXfPffcozVr1qhNmzY6deqUTp8+LVdXV2VnZ1/zczZs2FCGYWjVqlXat2+funbtqi+++ILwBeCqEMAA/Ob5+vpq8eLF2r9/v8LCwtS9e3ctWbJElZWVVvMZhqGFCxcqICBAjz76qBITE1VWVvarrvtq2bKl/vWvf+nYsWN66qmnFBAQoJiYGJ07d66mhgXgBsYpSAC/ecePH9ett96qN998U+fPn9euXbv09NNPq3379lbzvfzyyzp+/Lg++eQTOTs7S5L69+9/yfXa2dmprKzM8vjCC/rz8vJUUVGh119/XZWVlfr888/1zDPP6He/+53CwsJqeIQAbjQcAQPwm/fll1/qySef1KFDh9S4cWPdeuutkiQ3Nzc1btxYBQUFkqTCwkLddNNNatiwoc6dO6cVK1bo22+/tYSsm266SSUlJSovL5cktW3bVikpKTp79qwKCgoUHx9vec7jx49r5MiR2rVrl+zs7OTu7m55TgC4En6MG8ANYdmyZVq1apVOnz6tW2+9VU8++aSGDx+uOXPmKCkpSeHh4QoNDdXzzz+vr7/+Wk5OTvLz85ODg4POnj2rpUuXKicnRyNHjtSxY8e0atUqubq6aurUqTpw4ICaNm2qZ555RpMnT9bHH3+sO+64Q2vWrFF8fLxycnLUtGlTPfzww4qKilKDBg1svTkA1HEEMAAAAJNxChIAAMBkBDAAAACTEcAAAABMRgADAAAwGQEMAADAZAQwAAAAkxHAAAAATEYAAwAAMBkBDAAAwGQEMAAAAJMRwAAAAExGAAMAADDZ/wNBsfc25On8uwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style('white')\n",
    "sns.countplot(x='status',data=df,palette='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "79a4866e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "status\n",
       "Normal                  15836\n",
       "Depression               7652\n",
       "Suicidal                 6176\n",
       "Anxiety                  1900\n",
       "Stress                   1569\n",
       "Bipolar                  1009\n",
       "Personality disorder      377\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['status'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1c385702",
   "metadata": {},
   "outputs": [],
   "source": [
    "#we will be dividing all classes into 3 bins\n",
    "\"\"\"Normal: Doesn't need help (15,836 samples)\n",
    "Depression/Suicidal: Needs help with depression and suicide risk (13,828 samples)\n",
    "Other: Needs other kinds of assistance (4,855 samples)\n",
    "\n",
    "This grouping makes sense because:\n",
    "\n",
    "It maintains a clear distinction between those who likely don't need intervention and those who do.\n",
    "It prioritizes depression and suicide, which often co-occur and represent high-risk conditions requiring immediate attention.\n",
    "It groups other conditions that, while serious, might require different types of interventions.\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "68609858",
   "metadata": {},
   "outputs": [],
   "source": [
    "def bin_3(label):\n",
    "    if label == \"Normal\":\n",
    "        return label\n",
    "    elif label == \"Depression\" or label == \"Suicidal\":\n",
    "        return \"Depression + Suicidal\"\n",
    "    else:\n",
    "        return \"Other\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "a0a7bf66",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "status\n",
       "Normal                   15836\n",
       "Depression + Suicidal    13828\n",
       "Other                     4855\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['status'].apply(bin_3).value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "05affc57",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['status'] = df['status'].apply(bin_3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "89bf2c10",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "status\n",
       "Normal                   15836\n",
       "Depression + Suicidal    13828\n",
       "Other                     4855\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['status'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "881e1675",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_raw = df['text']\n",
    "text_stem = df['stem_text']\n",
    "text_lemm = df['lemm_text']\n",
    "label = df['status']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "91ec30ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "#train and test split\n",
    "from sklearn.model_selection import train_test_split\n",
    "text_raw_train,text_raw_test,label_train,label_test = train_test_split(text_raw,label,test_size=0.2,stratify=label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "83da478d",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_stem_train,text_stem_test,label_train,label_test = train_test_split(text_stem,label,test_size=0.2,stratify=label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "882a2f47",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_lemm_train,text_lemm_test,label_train,label_test = train_test_split(text_lemm,label,test_size=0.2,stratify=label)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b58979e",
   "metadata": {},
   "source": [
    "### countvect only and countvect + tfidf with 4 models each"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "b4d50a3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.metrics import classification_report,roc_auc_score,accuracy_score,f1_score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6498526a",
   "metadata": {},
   "source": [
    "### Dtree pipelines + Bow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "0e1ef051",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_dtree_raw = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"dtree\",DecisionTreeClassifier())])\n",
    "bow_dtree_stem = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"dtree\",DecisionTreeClassifier())])\n",
    "bow_dtree_lemm = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"dtree\",DecisionTreeClassifier())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "1f3229ed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree + BOW - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.40      0.44      0.42      2766\n",
      "               Normal       0.45      0.45      0.45      3167\n",
      "                Other       0.16      0.12      0.13       971\n",
      "\n",
      "             accuracy                           0.40      6904\n",
      "            macro avg       0.34      0.34      0.33      6904\n",
      "         weighted avg       0.39      0.40      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_dtree_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_dtree_raw = bow_dtree_raw.predict(text_raw_test)\n",
    "print(\"Decision Tree + BOW - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_dtree_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "6e5e1e07",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree + BOW -Stem Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.40      0.42      0.41      2766\n",
      "               Normal       0.45      0.48      0.47      3167\n",
      "                Other       0.14      0.09      0.11       971\n",
      "\n",
      "             accuracy                           0.40      6904\n",
      "            macro avg       0.33      0.33      0.33      6904\n",
      "         weighted avg       0.39      0.40      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_dtree_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_dtree_stem = bow_dtree_stem.predict(text_stem_test)\n",
    "print(\"Decision Tree + BOW -Stem Text\")\n",
    "print(classification_report(label_test,pred_bow_dtree_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "id": "b5ef096f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree + BOW -Lemm Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.77      0.78      0.78      2766\n",
      "               Normal       0.84      0.87      0.86      3167\n",
      "                Other       0.59      0.51      0.54       971\n",
      "\n",
      "             accuracy                           0.78      6904\n",
      "            macro avg       0.74      0.72      0.73      6904\n",
      "         weighted avg       0.78      0.78      0.78      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_dtree_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_dtree_lemm = bow_dtree_lemm.predict(text_lemm_test)\n",
    "print(\"Decision Tree + BOW -Lemm Text\")\n",
    "print(classification_report(label_test,pred_bow_dtree_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20d18483",
   "metadata": {},
   "source": [
    "### RFC pipelines + Bow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "908183bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_rfc_raw = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"RFC\",RandomForestClassifier())])\n",
    "bow_rfc_stem = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"RFC\",RandomForestClassifier())])\n",
    "bow_rfc_lemm = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"RFC\",RandomForestClassifier())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "fdcb3df7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RFC + BOW - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.36      0.38      2766\n",
      "               Normal       0.45      0.59      0.51      3167\n",
      "                Other       0.11      0.02      0.03       971\n",
      "\n",
      "             accuracy                           0.42      6904\n",
      "            macro avg       0.32      0.32      0.31      6904\n",
      "         weighted avg       0.38      0.42      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_rfc_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_rfc_raw = bow_rfc_raw.predict(text_raw_test)\n",
    "print(\"RFC + BOW - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_rfc_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "82f77642",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RFC + BOW - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.41      0.37      0.39      2766\n",
      "               Normal       0.46      0.62      0.53      3167\n",
      "                Other       0.11      0.02      0.03       971\n",
      "\n",
      "             accuracy                           0.43      6904\n",
      "            macro avg       0.33      0.34      0.32      6904\n",
      "         weighted avg       0.39      0.43      0.40      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_rfc_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_rfc_stem = bow_rfc_stem.predict(text_stem_test)\n",
    "print(\"RFC + BOW - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_rfc_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "b2b790a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RFC + BOW - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.76      0.90      0.82      2766\n",
      "               Normal       0.88      0.91      0.90      3167\n",
      "                Other       0.90      0.32      0.47       971\n",
      "\n",
      "             accuracy                           0.82      6904\n",
      "            macro avg       0.84      0.71      0.73      6904\n",
      "         weighted avg       0.83      0.82      0.81      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_rfc_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_rfc_lemm = bow_rfc_lemm.predict(text_lemm_test)\n",
    "print(\"RFC + BOW - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_rfc_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4276a518",
   "metadata": {},
   "source": [
    "### SVM + BOW pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "fce212e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_svm_raw = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"svm\",SVC())])\n",
    "bow_svm_stem = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"svm\",SVC())])\n",
    "bow_svm_lemm = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"svm\",SVC())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "35449eb4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM + BOW - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.37      0.15      0.21      2766\n",
      "               Normal       0.45      0.83      0.59      3167\n",
      "                Other       0.00      0.00      0.00       971\n",
      "\n",
      "             accuracy                           0.44      6904\n",
      "            macro avg       0.28      0.33      0.27      6904\n",
      "         weighted avg       0.36      0.44      0.36      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_svm_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_svm_raw = bow_svm_raw.predict(text_raw_test)\n",
    "print(\"SVM + BOW - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_svm_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "0935309e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM + BOW - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.15      0.21      2766\n",
      "               Normal       0.46      0.85      0.60      3167\n",
      "                Other       0.00      0.00      0.00       971\n",
      "\n",
      "             accuracy                           0.45      6904\n",
      "            macro avg       0.28      0.33      0.27      6904\n",
      "         weighted avg       0.37      0.45      0.36      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_svm_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_svm_stem = bow_svm_stem.predict(text_stem_test)\n",
    "print(\"SVM + BOW - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_svm_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "989d5e53",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM + BOW - LEMM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.83      0.90      0.86      2766\n",
      "               Normal       0.91      0.92      0.92      3167\n",
      "                Other       0.80      0.58      0.67       971\n",
      "\n",
      "             accuracy                           0.86      6904\n",
      "            macro avg       0.85      0.80      0.82      6904\n",
      "         weighted avg       0.86      0.86      0.86      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_svm_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_svm_lemm = bow_svm_lemm.predict(text_lemm_test)\n",
    "print(\"SVM + BOW - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_svm_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "562a34fe",
   "metadata": {},
   "source": [
    "### NB+BOW pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "id": "dd06be57",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_nb_raw = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"nb\",MultinomialNB())])\n",
    "bow_nb_stem = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"nb\",MultinomialNB())])\n",
    "bow_nb_lemm = Pipeline([(\"bow\",CountVectorizer()),\n",
    "                          (\"nb\",MultinomialNB())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "id": "385d1e92",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB + BOW - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.32      0.35      2766\n",
      "               Normal       0.45      0.64      0.53      3167\n",
      "                Other       0.19      0.02      0.04       971\n",
      "\n",
      "             accuracy                           0.43      6904\n",
      "            macro avg       0.34      0.33      0.31      6904\n",
      "         weighted avg       0.39      0.43      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_nb_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_nb_raw = bow_nb_raw.predict(text_raw_test)\n",
    "print(\"NB + BOW - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_nb_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "id": "fb228aae",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB + BOW - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.32      0.35      2766\n",
      "               Normal       0.46      0.65      0.54      3167\n",
      "                Other       0.11      0.01      0.02       971\n",
      "\n",
      "             accuracy                           0.43      6904\n",
      "            macro avg       0.32      0.33      0.30      6904\n",
      "         weighted avg       0.38      0.43      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_nb_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_nb_stem = bow_nb_stem.predict(text_stem_test)\n",
    "print(\"NB + BOW - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_nb_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "id": "f73f7083",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB + BOW - LEMM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.71      0.93      0.81      2766\n",
      "               Normal       0.95      0.74      0.83      3167\n",
      "                Other       0.72      0.62      0.67       971\n",
      "\n",
      "             accuracy                           0.80      6904\n",
      "            macro avg       0.80      0.76      0.77      6904\n",
      "         weighted avg       0.82      0.80      0.80      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_nb_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_nb_lemm = bow_nb_lemm.predict(text_lemm_test)\n",
    "print(\"NB + BOW - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_nb_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6712bcc7",
   "metadata": {},
   "source": [
    "### Best Accuracy - SVM+LEMM (86%),RFC+LEMM(82%)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b427804",
   "metadata": {},
   "source": [
    "### Dtree pipelines + Bow +Tfidf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "id": "64672249",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_tfidf_dtree_raw = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"dtree\",DecisionTreeClassifier())])\n",
    "bow_tfidf_dtree_stem = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"dtree\",DecisionTreeClassifier())])\n",
    "bow_tfidf_dtree_lemm = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"dtree\",DecisionTreeClassifier())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "id": "514ec932",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dtree + BOW + Tfidf - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.40      0.39      2766\n",
      "               Normal       0.46      0.51      0.48      3167\n",
      "                Other       0.13      0.08      0.10       971\n",
      "\n",
      "             accuracy                           0.40      6904\n",
      "            macro avg       0.33      0.33      0.33      6904\n",
      "         weighted avg       0.39      0.40      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_dtree_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_tfidf_dtree_raw = bow_tfidf_dtree_raw.predict(text_raw_test)\n",
    "print(\"Dtree + BOW + Tfidf - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_dtree_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "beef3285",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dtree + BOW + Tfidf - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.40      0.39      2766\n",
      "               Normal       0.45      0.50      0.47      3167\n",
      "                Other       0.15      0.10      0.12       971\n",
      "\n",
      "             accuracy                           0.40      6904\n",
      "            macro avg       0.33      0.33      0.33      6904\n",
      "         weighted avg       0.38      0.40      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_dtree_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_tfidf_dtree_stem = bow_tfidf_dtree_stem.predict(text_stem_test)\n",
    "print(\"Dtree + BOW + Tfidf - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_dtree_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "id": "c85f347d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dtree + BOW + Tfidf - LEMM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.78      0.77      0.77      2766\n",
      "               Normal       0.84      0.87      0.85      3167\n",
      "                Other       0.57      0.54      0.56       971\n",
      "\n",
      "             accuracy                           0.78      6904\n",
      "            macro avg       0.73      0.73      0.73      6904\n",
      "         weighted avg       0.78      0.78      0.78      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_dtree_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_tfidf_dtree_lemm = bow_tfidf_dtree_lemm.predict(text_lemm_test)\n",
    "print(\"Dtree + BOW + Tfidf - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_dtree_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "86e1f498",
   "metadata": {},
   "source": [
    "### RFC + BOW +Tfidf pipelines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "3fbd3500",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_tfidf_rfc_raw = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"RFC\",RandomForestClassifier())])\n",
    "bow_tfidf_rfc_stem = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"RFC\",RandomForestClassifier())])\n",
    "bow_tfidf_rfc_lemm = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"RFC\",RandomForestClassifier())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "8f07fe08",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RFC + BOW + Tfidf - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.40      0.33      0.36      2766\n",
      "               Normal       0.46      0.66      0.54      3167\n",
      "                Other       0.13      0.01      0.02       971\n",
      "\n",
      "             accuracy                           0.43      6904\n",
      "            macro avg       0.33      0.33      0.31      6904\n",
      "         weighted avg       0.39      0.43      0.40      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_rfc_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_tfidf_rfc_raw = bow_tfidf_rfc_raw.predict(text_raw_test)\n",
    "print(\"RFC + BOW + Tfidf - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_rfc_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "5e9c3064",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RFC + BOW + Tfidf - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.40      0.33      0.36      2766\n",
      "               Normal       0.46      0.67      0.55      3167\n",
      "                Other       0.20      0.02      0.03       971\n",
      "\n",
      "             accuracy                           0.44      6904\n",
      "            macro avg       0.35      0.34      0.31      6904\n",
      "         weighted avg       0.40      0.44      0.40      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_rfc_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_tfidf_rfc_stem = bow_tfidf_rfc_stem.predict(text_stem_test)\n",
    "print(\"RFC + BOW + Tfidf - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_rfc_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "bf2af48c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rfc + BOW + Tfidf - LEMM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.76      0.91      0.83      2766\n",
      "               Normal       0.90      0.92      0.91      3167\n",
      "                Other       0.88      0.34      0.49       971\n",
      "\n",
      "             accuracy                           0.83      6904\n",
      "            macro avg       0.85      0.72      0.74      6904\n",
      "         weighted avg       0.84      0.83      0.82      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_rfc_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_tfidf_rfc_lemm = bow_tfidf_rfc_lemm.predict(text_lemm_test)\n",
    "print(\"rfc + BOW + Tfidf - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_rfc_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "739c3c32",
   "metadata": {},
   "source": [
    "### SVM + BOW +Tfidf pipelines "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "21ce50c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_tfidf_svm_raw = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"SVM\",SVC())])\n",
    "bow_tfidf_svm_stem = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"SVM\",SVC())])\n",
    "bow_tfidf_svm_lemm = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"SVM\",SVC())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "e72e0aa6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM + BOW + Tfidf - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.39      0.28      0.32      2766\n",
      "               Normal       0.46      0.71      0.56      3167\n",
      "                Other       0.00      0.00      0.00       971\n",
      "\n",
      "             accuracy                           0.44      6904\n",
      "            macro avg       0.28      0.33      0.29      6904\n",
      "         weighted avg       0.37      0.44      0.39      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_svm_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_tfidf_svm_raw = bow_tfidf_svm_raw.predict(text_raw_test)\n",
    "print(\"SVM + BOW + Tfidf - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_svm_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "07195df4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM + BOW + Tfidf - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.38      0.26      0.31      2766\n",
      "               Normal       0.46      0.72      0.56      3167\n",
      "                Other       0.00      0.00      0.00       971\n",
      "\n",
      "             accuracy                           0.44      6904\n",
      "            macro avg       0.28      0.33      0.29      6904\n",
      "         weighted avg       0.36      0.44      0.38      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_svm_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_tfidf_svm_stem = bow_tfidf_svm_stem.predict(text_stem_test)\n",
    "print(\"SVM + BOW + Tfidf - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_svm_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "id": "d47732e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "svm + BOW + Tfidf - LEMM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.86      0.89      0.87      2766\n",
      "               Normal       0.90      0.93      0.91      3167\n",
      "                Other       0.82      0.60      0.69       971\n",
      "\n",
      "             accuracy                           0.87      6904\n",
      "            macro avg       0.86      0.81      0.83      6904\n",
      "         weighted avg       0.87      0.87      0.87      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_svm_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_tfidf_svm_lemm = bow_tfidf_svm_lemm.predict(text_lemm_test)\n",
    "print(\"svm + BOW + Tfidf - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_svm_lemm))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "fbd2324c",
   "metadata": {},
   "outputs": [],
   "source": [
    "bow_tfidf_nb_raw = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"nb\",MultinomialNB())])\n",
    "bow_tfidf_nb_stem = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"nb\",MultinomialNB())])\n",
    "bow_tfidf_nb_lemm = Pipeline([(\"bow+tfidf\",TfidfVectorizer()),\n",
    "                          (\"nb\",MultinomialNB())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "f834b6e5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB + BOW + Tfidf - Raw Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.38      0.16      0.23      2766\n",
      "               Normal       0.45      0.82      0.59      3167\n",
      "                Other       0.00      0.00      0.00       971\n",
      "\n",
      "             accuracy                           0.44      6904\n",
      "            macro avg       0.28      0.33      0.27      6904\n",
      "         weighted avg       0.36      0.44      0.36      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_nb_raw.fit(text_raw_train,label_train)\n",
    "pred_bow_tfidf_nb_raw = bow_tfidf_nb_raw.predict(text_raw_test)\n",
    "print(\"NB + BOW + Tfidf - Raw Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_nb_raw))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "4073f36d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB + BOW + Tfidf - STEM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.37      0.17      0.23      2766\n",
      "               Normal       0.45      0.81      0.58      3167\n",
      "                Other       1.00      0.00      0.00       971\n",
      "\n",
      "             accuracy                           0.44      6904\n",
      "            macro avg       0.61      0.33      0.27      6904\n",
      "         weighted avg       0.50      0.44      0.36      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_nb_stem.fit(text_stem_train,label_train)\n",
    "pred_bow_tfidf_nb_stem = bow_tfidf_nb_stem.predict(text_stem_test)\n",
    "print(\"NB + BOW + Tfidf - STEM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_nb_stem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "3f6fc9d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB + BOW + Tfidf - LEMM Text\n",
      "                       precision    recall  f1-score   support\n",
      "\n",
      "Depression + Suicidal       0.64      0.97      0.77      2766\n",
      "               Normal       0.93      0.74      0.82      3167\n",
      "                Other       0.93      0.17      0.28       971\n",
      "\n",
      "             accuracy                           0.75      6904\n",
      "            macro avg       0.83      0.62      0.62      6904\n",
      "         weighted avg       0.81      0.75      0.73      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bow_tfidf_nb_lemm.fit(text_lemm_train,label_train)\n",
    "pred_bow_tfidf_nb_lemm = bow_tfidf_nb_lemm.predict(text_lemm_test)\n",
    "print(\"NB + BOW + Tfidf - LEMM Text\")\n",
    "print(classification_report(label_test,pred_bow_tfidf_nb_lemm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0fc585f1",
   "metadata": {},
   "source": [
    "### Best accuracy SVM+BoW+tfidf - 87%"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5b2b08f",
   "metadata": {},
   "source": [
    "### Vectorizing Text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "id": "a95b497a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def vectorize(text):\n",
    "    doc = nlp(text)\n",
    "    return doc.vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "64f70ca6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#use this for now, forgot to add vect\n",
    "text_raw = df['text'].apply(vectorize)\n",
    "text_stem = df['stem_text'].apply(vectorize)\n",
    "text_lemm = df['lemm_text'].apply(vectorize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "id": "c6de206b",
   "metadata": {},
   "outputs": [],
   "source": [
    "label = df['status']\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "lbl = LabelEncoder()\n",
    "label_en = lbl.fit_transform(df['status'])\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "label_en = to_categorical(label_en,num_classes =3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "id": "1c964528",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_raw_train_vect,text_raw_test_vect,label_train_en,label_test_en = train_test_split(text_raw,label_en,test_size=0.2,stratify=label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "id": "cc4acdf2",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_stem_train_vect,text_stem_test_vect,_,_ = train_test_split(text_stem,label_en,test_size=0.2,stratify=label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "id": "71e3fac9",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_lemm_train_vect,text_lemm_test_vect,_,_ = train_test_split(text_lemm,label_en,test_size=0.2,stratify=label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "id": "174fe893",
   "metadata": {},
   "outputs": [],
   "source": [
    "text_raw_train_vect = np.array(text_raw_train_vect.tolist())\n",
    "text_raw_test_vect = np.array(text_raw_test_vect.tolist())\n",
    "text_stem_train_vect = np.array(text_stem_train_vect.tolist())\n",
    "text_stem_test_vect = np.array(text_stem_test_vect.tolist())\n",
    "text_lemm_train_vect = np.array(text_lemm_train_vect.tolist())\n",
    "text_lemm_test_vect = np.array(text_lemm_test_vect.tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "id": "5d608a54",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.1360738 ,  2.1963723 , -2.441451  , ...,  0.16634843,\n",
       "        -1.3382624 ,  1.059014  ],\n",
       "       [-0.04329304,  0.4880501 , -2.5585122 , ...,  1.5938971 ,\n",
       "        -3.1678433 ,  2.2137842 ],\n",
       "       [-0.6871018 ,  0.89500815, -1.0987227 , ...,  0.16608816,\n",
       "        -1.2401965 ,  0.43304366],\n",
       "       ...,\n",
       "       [ 0.43235362, -0.11997981, -0.81152797, ...,  0.2651369 ,\n",
       "        -2.7310343 ,  0.7777474 ],\n",
       "       [ 0.29954374,  0.93336254, -1.622493  , ...,  0.76761806,\n",
       "        -1.0794976 ,  0.16948687],\n",
       "       [ 2.2735848 ,  0.70357543, -2.344398  , ..., -1.3723776 ,\n",
       "        -4.410455  ,  0.93325996]], dtype=float32)"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text_raw_train_vect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "id": "72b3e286",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Raw text train shape: (27615, 300)\n",
      "Raw text test shape: (6904, 300)\n",
      "Stem text train shape: (27615, 300)\n",
      "Stem text test shape: (6904, 300)\n",
      "Lemm text train shape: (27615, 300)\n",
      "Lemm text test shape: (6904, 300)\n",
      "Labels train shape: (27615, 3)\n",
      "Labels test shape: (6904, 3)\n"
     ]
    }
   ],
   "source": [
    "print(\"Raw text train shape:\", text_raw_train_vect.shape)\n",
    "print(\"Raw text test shape:\", text_raw_test_vect.shape)\n",
    "print(\"Stem text train shape:\", text_stem_train_vect.shape)\n",
    "print(\"Stem text test shape:\", text_stem_test_vect.shape)\n",
    "print(\"Lemm text train shape:\", text_lemm_train_vect.shape)\n",
    "print(\"Lemm text test shape:\", text_lemm_test_vect.shape)\n",
    "print(\"Labels train shape:\", label_train_en.shape)\n",
    "print(\"Labels test shape:\", label_test_en.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1ab41e5",
   "metadata": {},
   "source": [
    "### Creating NN for Word2vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "id": "0262bfb1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense,Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "id": "f478802e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_NN():\n",
    "    #define model\n",
    "    model = Sequential()\n",
    "    #data shape = (-1,300) first layer -> 300\n",
    "    model.add(Dense(1200,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(600,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(300,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    #2nd ->150\n",
    "    model.add(Dense(150,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    #3rd - >75\n",
    "    model.add(Dense(75,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    #4th-> 37\n",
    "    model.add(Dense(37,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    # 5th -> 18\n",
    "    model.add(Dense(18,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    #6th -> 9\n",
    "    model.add(Dense(9,activation='relu'))\n",
    "    model.add(Dropout(0.2))\n",
    "    #output layer\n",
    "    model.add(Dense(3,activation='softmax'))\n",
    "    \n",
    "    model.compile(loss='categorical_crossentropy',optimizer='adam',metrics=['accuracy'])\n",
    "    return model\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4b52632",
   "metadata": {},
   "source": [
    "### Raw text model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "id": "1dbb91c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.backend import clear_session\n",
    "clear_session()\n",
    "raw_text_model = create_NN()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "id": "8a752cb0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-07-30 21:50:04.898359: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:117] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 13ms/step - accuracy: 0.6304 - loss: 0.8566\n",
      "Epoch 2/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.7593 - loss: 0.6226\n",
      "Epoch 3/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.7833 - loss: 0.5744\n",
      "Epoch 4/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.7962 - loss: 0.5368\n",
      "Epoch 5/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8092 - loss: 0.5101\n",
      "Epoch 6/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8165 - loss: 0.4944\n",
      "Epoch 7/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8297 - loss: 0.4660\n",
      "Epoch 8/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8266 - loss: 0.4558\n",
      "Epoch 9/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8352 - loss: 0.4431\n",
      "Epoch 10/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8365 - loss: 0.4356\n",
      "Epoch 11/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8377 - loss: 0.4305\n",
      "Epoch 12/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8487 - loss: 0.4081\n",
      "Epoch 13/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8483 - loss: 0.4085\n",
      "Epoch 14/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8504 - loss: 0.4020\n",
      "Epoch 15/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8544 - loss: 0.3975\n",
      "Epoch 16/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8530 - loss: 0.3863\n",
      "Epoch 17/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8597 - loss: 0.3798\n",
      "Epoch 18/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8636 - loss: 0.3729\n",
      "Epoch 19/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8632 - loss: 0.3678\n",
      "Epoch 20/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8616 - loss: 0.3669\n",
      "Epoch 21/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8658 - loss: 0.3541\n",
      "Epoch 22/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8673 - loss: 0.3501\n",
      "Epoch 23/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.8724 - loss: 0.3416\n",
      "Epoch 24/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8709 - loss: 0.3393\n",
      "Epoch 25/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8725 - loss: 0.3336\n",
      "Epoch 26/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8746 - loss: 0.3313\n",
      "Epoch 27/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8768 - loss: 0.3264\n",
      "Epoch 28/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8802 - loss: 0.3210\n",
      "Epoch 29/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8784 - loss: 0.3215\n",
      "Epoch 30/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8853 - loss: 0.3156\n",
      "Epoch 31/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8759 - loss: 0.3229\n",
      "Epoch 32/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.8826 - loss: 0.3122\n",
      "Epoch 33/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8839 - loss: 0.3091\n",
      "Epoch 34/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8819 - loss: 0.3168\n",
      "Epoch 35/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8825 - loss: 0.3125\n",
      "Epoch 36/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8888 - loss: 0.2936\n",
      "Epoch 37/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8950 - loss: 0.2817\n",
      "Epoch 38/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8871 - loss: 0.3066\n",
      "Epoch 39/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8922 - loss: 0.2869\n",
      "Epoch 40/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8906 - loss: 0.2923\n",
      "Epoch 41/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8925 - loss: 0.2879\n",
      "Epoch 42/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 16ms/step - accuracy: 0.8976 - loss: 0.2789\n",
      "Epoch 43/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8977 - loss: 0.2752\n",
      "Epoch 44/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8958 - loss: 0.2806\n",
      "Epoch 45/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8976 - loss: 0.2730\n",
      "Epoch 46/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 16ms/step - accuracy: 0.9000 - loss: 0.2651\n",
      "Epoch 47/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.9024 - loss: 0.2623\n",
      "Epoch 48/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.8949 - loss: 0.2818\n",
      "Epoch 49/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.9007 - loss: 0.2745\n",
      "Epoch 50/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.9056 - loss: 0.2665\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x3871b8f10>"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_text_model.fit(text_raw_train_vect,label_train_en,epochs=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "00ee9729",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m216/216\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step\n"
     ]
    }
   ],
   "source": [
    "prediction = model.predict(text_raw_test_vect)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "071dc015",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = np.argmax(prediction, axis=1)\n",
    "y_true = np.argmax(label_test_en, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "id": "95ffc3dd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.12      0.01      0.01      2766\n",
      "           1       0.44      0.91      0.59      3167\n",
      "           2       0.07      0.02      0.03       971\n",
      "\n",
      "    accuracy                           0.42      6904\n",
      "   macro avg       0.21      0.31      0.21      6904\n",
      "weighted avg       0.26      0.42      0.28      6904\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "report = classification_report(y_true, y_pred)\n",
    "print(report)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6235db8d",
   "metadata": {},
   "source": [
    "### Stem text model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "id": "5ace4756",
   "metadata": {},
   "outputs": [],
   "source": [
    "stem_text_model = create_NN()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "812d8ede",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 14ms/step - accuracy: 0.4325 - loss: 1.0440\n",
      "Epoch 2/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.4558 - loss: 1.0095\n",
      "Epoch 3/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4576 - loss: 1.0042\n",
      "Epoch 4/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 12ms/step - accuracy: 0.4618 - loss: 1.0003\n",
      "Epoch 5/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 12ms/step - accuracy: 0.4557 - loss: 1.0013\n",
      "Epoch 6/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4628 - loss: 0.9987\n",
      "Epoch 7/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4582 - loss: 1.0037\n",
      "Epoch 8/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4613 - loss: 1.0007\n",
      "Epoch 9/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4583 - loss: 0.9985\n",
      "Epoch 10/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.4564 - loss: 1.0045\n",
      "Epoch 11/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 15ms/step - accuracy: 0.4624 - loss: 0.9970\n",
      "Epoch 12/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.4599 - loss: 0.9986\n",
      "Epoch 13/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.4546 - loss: 1.0023\n",
      "Epoch 14/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4583 - loss: 1.0005\n",
      "Epoch 15/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4578 - loss: 1.0032\n",
      "Epoch 16/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4549 - loss: 1.0041\n",
      "Epoch 17/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.4574 - loss: 1.0000\n",
      "Epoch 18/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4559 - loss: 1.0014\n",
      "Epoch 19/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4591 - loss: 1.0001\n",
      "Epoch 20/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4574 - loss: 1.0034\n",
      "Epoch 21/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4608 - loss: 0.9999\n",
      "Epoch 22/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4514 - loss: 1.0027\n",
      "Epoch 23/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4593 - loss: 0.9997\n",
      "Epoch 24/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4643 - loss: 0.9960\n",
      "Epoch 25/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4560 - loss: 1.0000\n",
      "Epoch 26/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4581 - loss: 0.9994\n",
      "Epoch 27/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4660 - loss: 0.9946\n",
      "Epoch 28/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4550 - loss: 0.9982\n",
      "Epoch 29/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4585 - loss: 1.0000\n",
      "Epoch 30/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4569 - loss: 1.0017\n",
      "Epoch 31/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4627 - loss: 1.0016\n",
      "Epoch 32/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4595 - loss: 0.9955\n",
      "Epoch 33/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4557 - loss: 0.9992\n",
      "Epoch 34/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4612 - loss: 0.9990\n",
      "Epoch 35/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4598 - loss: 1.0022\n",
      "Epoch 36/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4578 - loss: 0.9992\n",
      "Epoch 37/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4603 - loss: 0.9964\n",
      "Epoch 38/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4580 - loss: 0.9981\n",
      "Epoch 39/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 15ms/step - accuracy: 0.4554 - loss: 0.9978\n",
      "Epoch 40/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 15ms/step - accuracy: 0.4549 - loss: 1.0001\n",
      "Epoch 41/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 16ms/step - accuracy: 0.4582 - loss: 0.9998\n",
      "Epoch 42/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 15ms/step - accuracy: 0.4613 - loss: 1.0001\n",
      "Epoch 43/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 17ms/step - accuracy: 0.4544 - loss: 1.0042\n",
      "Epoch 44/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 16ms/step - accuracy: 0.4582 - loss: 0.9990\n",
      "Epoch 45/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m13s\u001b[0m 15ms/step - accuracy: 0.4553 - loss: 1.0014\n",
      "Epoch 46/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4590 - loss: 0.9996\n",
      "Epoch 47/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4547 - loss: 1.0000\n",
      "Epoch 48/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4592 - loss: 0.9982\n",
      "Epoch 49/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4538 - loss: 1.0038\n",
      "Epoch 50/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4612 - loss: 0.9986\n",
      "Epoch 51/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4657 - loss: 0.9997\n",
      "Epoch 52/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4558 - loss: 1.0001\n",
      "Epoch 53/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4552 - loss: 1.0016\n",
      "Epoch 54/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4544 - loss: 1.0007\n",
      "Epoch 55/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4578 - loss: 0.9976\n",
      "Epoch 56/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4618 - loss: 0.9948\n",
      "Epoch 57/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4565 - loss: 1.0009\n",
      "Epoch 58/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4587 - loss: 1.0021\n",
      "Epoch 59/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4580 - loss: 0.9997\n",
      "Epoch 60/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4572 - loss: 0.9970\n",
      "Epoch 61/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4592 - loss: 1.0007\n",
      "Epoch 62/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4598 - loss: 1.0007\n",
      "Epoch 63/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4628 - loss: 0.9980\n",
      "Epoch 64/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4591 - loss: 0.9999\n",
      "Epoch 65/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4609 - loss: 1.0012\n",
      "Epoch 66/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4579 - loss: 0.9987\n",
      "Epoch 67/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4631 - loss: 1.0017\n",
      "Epoch 68/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4574 - loss: 0.9997\n",
      "Epoch 69/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4613 - loss: 0.9976\n",
      "Epoch 70/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4556 - loss: 1.0012\n",
      "Epoch 71/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4656 - loss: 0.9974\n",
      "Epoch 72/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 13ms/step - accuracy: 0.4583 - loss: 1.0021\n",
      "Epoch 73/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4587 - loss: 1.0039\n",
      "Epoch 74/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4596 - loss: 0.9994\n",
      "Epoch 75/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4625 - loss: 0.9964\n",
      "Epoch 76/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4591 - loss: 1.0010\n",
      "Epoch 77/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4575 - loss: 0.9997\n",
      "Epoch 78/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4588 - loss: 0.9997\n",
      "Epoch 79/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4556 - loss: 1.0010\n",
      "Epoch 80/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4574 - loss: 0.9993\n",
      "Epoch 81/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4546 - loss: 1.0013\n",
      "Epoch 82/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4617 - loss: 1.0018\n",
      "Epoch 83/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 14ms/step - accuracy: 0.4620 - loss: 0.9985\n",
      "Epoch 84/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4579 - loss: 1.0003\n",
      "Epoch 85/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4622 - loss: 0.9950\n",
      "Epoch 86/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4606 - loss: 0.9986\n",
      "Epoch 87/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4569 - loss: 1.0014\n",
      "Epoch 88/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4596 - loss: 0.9994\n",
      "Epoch 89/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4618 - loss: 0.9958\n",
      "Epoch 90/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4561 - loss: 1.0024\n",
      "Epoch 91/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4585 - loss: 1.0015\n",
      "Epoch 92/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4602 - loss: 0.9969\n",
      "Epoch 93/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4557 - loss: 1.0048\n",
      "Epoch 94/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4545 - loss: 1.0019\n",
      "Epoch 95/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4588 - loss: 0.9988\n",
      "Epoch 96/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4561 - loss: 1.0060\n",
      "Epoch 97/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4564 - loss: 0.9997\n",
      "Epoch 98/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4575 - loss: 0.9989\n",
      "Epoch 99/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4617 - loss: 0.9973\n",
      "Epoch 100/100\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 13ms/step - accuracy: 0.4574 - loss: 0.9974\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "tf.debugging.set_log_device_placement(True)\n",
    "with tf.device('/GPU:0'):    \n",
    "    stem_text_model.fit(text_stem_train_vect,label_train_en,epochs=100)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81f490c0",
   "metadata": {},
   "source": [
    "### Lemm text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "id": "fd40875b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.backend import clear_session\n",
    "clear_session()\n",
    "lemm_text_model = create_NN()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "id": "0e989d29",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 58ms/step - accuracy: 0.4215 - loss: 1.0693\n",
      "Epoch 2/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 56ms/step - accuracy: 0.4508 - loss: 1.0027\n",
      "Epoch 3/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4544 - loss: 1.0072\n",
      "Epoch 4/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4557 - loss: 1.0006\n",
      "Epoch 5/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4558 - loss: 1.0033\n",
      "Epoch 6/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4589 - loss: 1.0017\n",
      "Epoch 7/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 60ms/step - accuracy: 0.4597 - loss: 0.9999\n",
      "Epoch 8/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 60ms/step - accuracy: 0.4593 - loss: 0.9999\n",
      "Epoch 9/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 60ms/step - accuracy: 0.4547 - loss: 1.0004\n",
      "Epoch 10/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4511 - loss: 1.0016\n",
      "Epoch 11/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 56ms/step - accuracy: 0.4575 - loss: 0.9975\n",
      "Epoch 12/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 56ms/step - accuracy: 0.4570 - loss: 1.0045\n",
      "Epoch 13/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 56ms/step - accuracy: 0.4569 - loss: 1.0019\n",
      "Epoch 14/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4602 - loss: 1.0026\n",
      "Epoch 15/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4565 - loss: 1.0010\n",
      "Epoch 16/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4542 - loss: 1.0057\n",
      "Epoch 17/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 57ms/step - accuracy: 0.4619 - loss: 0.9974\n",
      "Epoch 18/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4587 - loss: 0.9999\n",
      "Epoch 19/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 56ms/step - accuracy: 0.4604 - loss: 0.9979\n",
      "Epoch 20/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 60ms/step - accuracy: 0.4586 - loss: 0.9987\n",
      "Epoch 21/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4632 - loss: 0.9973\n",
      "Epoch 22/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4567 - loss: 1.0024\n",
      "Epoch 23/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4574 - loss: 0.9961\n",
      "Epoch 24/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4632 - loss: 0.9991\n",
      "Epoch 25/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4589 - loss: 1.0004\n",
      "Epoch 26/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4510 - loss: 1.0038\n",
      "Epoch 27/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4590 - loss: 0.9945\n",
      "Epoch 28/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4634 - loss: 1.0000\n",
      "Epoch 29/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4571 - loss: 0.9998\n",
      "Epoch 30/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4578 - loss: 0.9995\n",
      "Epoch 31/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4549 - loss: 1.0024\n",
      "Epoch 32/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 55ms/step - accuracy: 0.4589 - loss: 1.0021\n",
      "Epoch 33/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 56ms/step - accuracy: 0.4630 - loss: 0.9985\n",
      "Epoch 34/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 60ms/step - accuracy: 0.4547 - loss: 1.0041\n",
      "Epoch 35/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 56ms/step - accuracy: 0.4604 - loss: 1.0020\n",
      "Epoch 36/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4554 - loss: 1.0010\n",
      "Epoch 37/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 57ms/step - accuracy: 0.4605 - loss: 0.9976\n",
      "Epoch 38/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4580 - loss: 0.9989\n",
      "Epoch 39/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4622 - loss: 0.9994\n",
      "Epoch 40/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 56ms/step - accuracy: 0.4604 - loss: 0.9990\n",
      "Epoch 41/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 57ms/step - accuracy: 0.4537 - loss: 1.0037\n",
      "Epoch 42/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4571 - loss: 0.9992\n",
      "Epoch 43/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4629 - loss: 0.9995\n",
      "Epoch 44/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4603 - loss: 1.0006\n",
      "Epoch 45/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4584 - loss: 0.9988\n",
      "Epoch 46/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4543 - loss: 1.0021\n",
      "Epoch 47/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4569 - loss: 1.0026\n",
      "Epoch 48/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 59ms/step - accuracy: 0.4541 - loss: 1.0022\n",
      "Epoch 49/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4537 - loss: 1.0004\n",
      "Epoch 50/50\n",
      "\u001b[1m863/863\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 58ms/step - accuracy: 0.4555 - loss: 1.0006\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "tf.debugging.set_log_device_placement(True)\n",
    "with tf.device('/GPU:0'):    \n",
    "    lemm_text_model.fit(text_lemm_train_vect,label_train_en,epochs=50)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}