TabNet.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pytorch_tabnet.tab_model import TabNetClassifier\n",
    "\n",
    "import os\n",
    "import torch\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    " \n",
    "from sklearn.model_selection import KFold, train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
    "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using mps\n"
     ]
    }
   ],
   "source": [
    "#Usamos el GPU si esta disponible, mps porque es apple silicon.\n",
    "device = 'mps' if torch.backends.mps.is_available() else 'cpu'\n",
    "print(f\"using {device}\")\n",
    "\n",
    "#también hay que mover los datos a la gpu."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cargar los datos y estructurarlos para entrenar el modelo.\n",
    "\n",
    "Voy a usar los datos que usó Diego."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "folder = \"Data\"\n",
    "match_file = \"estadisticas_partidos.csv\"\n",
    "\n",
    "path_matches = os.path.join(folder, match_file)\n",
    "matches = pd.read_csv(path_matches)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['team_home' 'team_away' 'score_home' 'score_away' 'expected_goals_home'\n",
      " 'expected_goals_away' 'posession_home' 'posession_away'\n",
      " 'total_shots_home' 'total_shots_away' 'shots_on_target_home'\n",
      " 'shots_on_target_away' 'shots_off_target_home' 'shots_off_target_away'\n",
      " 'blocked_shots_home' 'blocked_shots_away' 'corner_kicks_home'\n",
      " 'corner_kicks_away' 'offsides_home' 'offsides_away' 'fouls_home'\n",
      " 'fouls_away' 'yellow_cards_home' 'yellow_cards_away' 'red_cards_home'\n",
      " 'red_cards_away' 'free_kicks_home' 'free_kicks_away' 'throw_ins_home'\n",
      " 'throw_ins_away' 'goal_kicks_home' 'goal_kicks_away'\n",
      " 'shots_inside_box_home' 'shots_inside_box_away' 'shots_outside_box_home'\n",
      " 'shots_outside_box_away' 'goalkeeper_saves_home' 'goalkeeper_saves_away'\n",
      " 'passes_home' 'passes_away' 'accurate_passes_home' 'accurate_passes_away'\n",
      " 'long_balls_home' 'long_balls_away' 'crosses_home' 'crosses_away'\n",
      " 'dribbles_home' 'dribbles_away' 'possesion_lost_home'\n",
      " 'possesion_lost_away' 'duels_won_home' 'duels_won_away'\n",
      " 'aerials_won_home' 'aerials_won_away' 'tackles_home' 'tackles_away'\n",
      " 'interceptions_home' 'interceptions_away' 'clearences_home'\n",
      " 'clearences_away']\n",
      "60\n"
     ]
    }
   ],
   "source": [
    "print(matches.columns.values)\n",
    "print(len(matches.columns.values))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 240 entries, 0 to 239\n",
      "Data columns (total 60 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   team_home               240 non-null    object \n",
      " 1   team_away               240 non-null    object \n",
      " 2   score_home              240 non-null    int64  \n",
      " 3   score_away              240 non-null    int64  \n",
      " 4   expected_goals_home     240 non-null    float64\n",
      " 5   expected_goals_away     240 non-null    float64\n",
      " 6   posession_home          240 non-null    int64  \n",
      " 7   posession_away          240 non-null    int64  \n",
      " 8   total_shots_home        240 non-null    int64  \n",
      " 9   total_shots_away        240 non-null    int64  \n",
      " 10  shots_on_target_home    240 non-null    int64  \n",
      " 11  shots_on_target_away    240 non-null    int64  \n",
      " 12  shots_off_target_home   240 non-null    int64  \n",
      " 13  shots_off_target_away   240 non-null    int64  \n",
      " 14  blocked_shots_home      240 non-null    int64  \n",
      " 15  blocked_shots_away      240 non-null    int64  \n",
      " 16  corner_kicks_home       240 non-null    int64  \n",
      " 17  corner_kicks_away       240 non-null    int64  \n",
      " 18  offsides_home           240 non-null    int64  \n",
      " 19  offsides_away           240 non-null    int64  \n",
      " 20  fouls_home              240 non-null    int64  \n",
      " 21  fouls_away              240 non-null    int64  \n",
      " 22  yellow_cards_home       240 non-null    int64  \n",
      " 23  yellow_cards_away       240 non-null    int64  \n",
      " 24  red_cards_home          240 non-null    int64  \n",
      " 25  red_cards_away          240 non-null    int64  \n",
      " 26  free_kicks_home         240 non-null    int64  \n",
      " 27  free_kicks_away         240 non-null    int64  \n",
      " 28  throw_ins_home          240 non-null    int64  \n",
      " 29  throw_ins_away          240 non-null    int64  \n",
      " 30  goal_kicks_home         240 non-null    int64  \n",
      " 31  goal_kicks_away         240 non-null    int64  \n",
      " 32  shots_inside_box_home   240 non-null    int64  \n",
      " 33  shots_inside_box_away   240 non-null    int64  \n",
      " 34  shots_outside_box_home  240 non-null    int64  \n",
      " 35  shots_outside_box_away  240 non-null    int64  \n",
      " 36  goalkeeper_saves_home   240 non-null    int64  \n",
      " 37  goalkeeper_saves_away   240 non-null    int64  \n",
      " 38  passes_home             240 non-null    int64  \n",
      " 39  passes_away             240 non-null    int64  \n",
      " 40  accurate_passes_home    240 non-null    int64  \n",
      " 41  accurate_passes_away    240 non-null    int64  \n",
      " 42  long_balls_home         240 non-null    int64  \n",
      " 43  long_balls_away         240 non-null    int64  \n",
      " 44  crosses_home            240 non-null    int64  \n",
      " 45  crosses_away            240 non-null    int64  \n",
      " 46  dribbles_home           240 non-null    int64  \n",
      " 47  dribbles_away           240 non-null    int64  \n",
      " 48  possesion_lost_home     240 non-null    int64  \n",
      " 49  possesion_lost_away     240 non-null    int64  \n",
      " 50  duels_won_home          240 non-null    int64  \n",
      " 51  duels_won_away          240 non-null    int64  \n",
      " 52  aerials_won_home        240 non-null    int64  \n",
      " 53  aerials_won_away        240 non-null    int64  \n",
      " 54  tackles_home            240 non-null    int64  \n",
      " 55  tackles_away            240 non-null    int64  \n",
      " 56  interceptions_home      240 non-null    int64  \n",
      " 57  interceptions_away      240 non-null    int64  \n",
      " 58  clearences_home         240 non-null    int64  \n",
      " 59  clearences_away         240 non-null    int64  \n",
      "dtypes: float64(2), int64(56), object(2)\n",
      "memory usage: 112.6+ KB\n"
     ]
    }
   ],
   "source": [
    "matches.info(verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Aqui agrega la columna de resultados\n",
    "\n",
    "results = []\n",
    "\n",
    "for index in matches.index:\n",
    "\n",
    "    match = matches.loc[index]\n",
    "    if match[\"score_home\"] > match[\"score_away\"]:\n",
    "        results.append(1)\n",
    "    elif match[\"score_home\"] < match[\"score_away\"]:\n",
    "        results.append(2)\n",
    "    else:\n",
    "        results.append(0)\n",
    "\n",
    "matches[\"result\"] = results\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>team_home</th>\n",
       "      <th>team_away</th>\n",
       "      <th>score_home</th>\n",
       "      <th>score_away</th>\n",
       "      <th>expected_goals_home</th>\n",
       "      <th>expected_goals_away</th>\n",
       "      <th>posession_home</th>\n",
       "      <th>posession_away</th>\n",
       "      <th>total_shots_home</th>\n",
       "      <th>total_shots_away</th>\n",
       "      <th>...</th>\n",
       "      <th>duels_won_away</th>\n",
       "      <th>aerials_won_home</th>\n",
       "      <th>aerials_won_away</th>\n",
       "      <th>tackles_home</th>\n",
       "      <th>tackles_away</th>\n",
       "      <th>interceptions_home</th>\n",
       "      <th>interceptions_away</th>\n",
       "      <th>clearences_home</th>\n",
       "      <th>clearences_away</th>\n",
       "      <th>result</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Curicó Unido</td>\n",
       "      <td>Colo Colo</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.07</td>\n",
       "      <td>2.31</td>\n",
       "      <td>32</td>\n",
       "      <td>68</td>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>...</td>\n",
       "      <td>33</td>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "      <td>11</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>23</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Huachipato</td>\n",
       "      <td>Audax Italiano</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1.34</td>\n",
       "      <td>0.45</td>\n",
       "      <td>64</td>\n",
       "      <td>36</td>\n",
       "      <td>12</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>35</td>\n",
       "      <td>13</td>\n",
       "      <td>9</td>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>19</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Unión Española</td>\n",
       "      <td>Cobresal</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.09</td>\n",
       "      <td>0.54</td>\n",
       "      <td>65</td>\n",
       "      <td>35</td>\n",
       "      <td>26</td>\n",
       "      <td>9</td>\n",
       "      <td>...</td>\n",
       "      <td>66</td>\n",
       "      <td>18</td>\n",
       "      <td>14</td>\n",
       "      <td>18</td>\n",
       "      <td>19</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>24</td>\n",
       "      <td>24</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Deportes Copiapó</td>\n",
       "      <td>Everton de Viña del Mar</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1.21</td>\n",
       "      <td>0.59</td>\n",
       "      <td>36</td>\n",
       "      <td>64</td>\n",
       "      <td>10</td>\n",
       "      <td>12</td>\n",
       "      <td>...</td>\n",
       "      <td>35</td>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "      <td>13</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>28</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Deportes Magallanes</td>\n",
       "      <td>Coquimbo Unido</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1.62</td>\n",
       "      <td>1.82</td>\n",
       "      <td>57</td>\n",
       "      <td>43</td>\n",
       "      <td>13</td>\n",
       "      <td>15</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>20</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 61 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             team_home                team_away  score_home  score_away  \\\n",
       "0         Curicó Unido                Colo Colo           0           1   \n",
       "1           Huachipato           Audax Italiano           2           0   \n",
       "2       Unión Española                 Cobresal           1           0   \n",
       "3     Deportes Copiapó  Everton de Viña del Mar           2           0   \n",
       "4  Deportes Magallanes           Coquimbo Unido           2           3   \n",
       "\n",
       "   expected_goals_home  expected_goals_away  posession_home  posession_away  \\\n",
       "0                 0.07                 2.31              32              68   \n",
       "1                 1.34                 0.45              64              36   \n",
       "2                 1.09                 0.54              65              35   \n",
       "3                 1.21                 0.59              36              64   \n",
       "4                 1.62                 1.82              57              43   \n",
       "\n",
       "   total_shots_home  total_shots_away  ...  duels_won_away  aerials_won_home  \\\n",
       "0                 5                20  ...              33                 4   \n",
       "1                12                13  ...              35                13   \n",
       "2                26                 9  ...              66                18   \n",
       "3                10                12  ...              35                11   \n",
       "4                13                15  ...              32                 9   \n",
       "\n",
       "   aerials_won_away  tackles_home  tackles_away  interceptions_home  \\\n",
       "0                11            11             8                   5   \n",
       "1                 9            12            12                   7   \n",
       "2                14            18            19                   3   \n",
       "3                 9            13             9                   7   \n",
       "4                11            10             6                   5   \n",
       "\n",
       "   interceptions_away  clearences_home  clearences_away  result  \n",
       "0                   8               23                8       2  \n",
       "1                   4               19               28       1  \n",
       "2                  11               24               24       1  \n",
       "3                   2               28               13       1  \n",
       "4                   1               16               20       2  \n",
       "\n",
       "[5 rows x 61 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 240 entries, 0 to 239\n",
      "Data columns (total 61 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   team_home               240 non-null    object \n",
      " 1   team_away               240 non-null    object \n",
      " 2   score_home              240 non-null    int64  \n",
      " 3   score_away              240 non-null    int64  \n",
      " 4   expected_goals_home     240 non-null    float64\n",
      " 5   expected_goals_away     240 non-null    float64\n",
      " 6   posession_home          240 non-null    int64  \n",
      " 7   posession_away          240 non-null    int64  \n",
      " 8   total_shots_home        240 non-null    int64  \n",
      " 9   total_shots_away        240 non-null    int64  \n",
      " 10  shots_on_target_home    240 non-null    int64  \n",
      " 11  shots_on_target_away    240 non-null    int64  \n",
      " 12  shots_off_target_home   240 non-null    int64  \n",
      " 13  shots_off_target_away   240 non-null    int64  \n",
      " 14  blocked_shots_home      240 non-null    int64  \n",
      " 15  blocked_shots_away      240 non-null    int64  \n",
      " 16  corner_kicks_home       240 non-null    int64  \n",
      " 17  corner_kicks_away       240 non-null    int64  \n",
      " 18  offsides_home           240 non-null    int64  \n",
      " 19  offsides_away           240 non-null    int64  \n",
      " 20  fouls_home              240 non-null    int64  \n",
      " 21  fouls_away              240 non-null    int64  \n",
      " 22  yellow_cards_home       240 non-null    int64  \n",
      " 23  yellow_cards_away       240 non-null    int64  \n",
      " 24  red_cards_home          240 non-null    int64  \n",
      " 25  red_cards_away          240 non-null    int64  \n",
      " 26  free_kicks_home         240 non-null    int64  \n",
      " 27  free_kicks_away         240 non-null    int64  \n",
      " 28  throw_ins_home          240 non-null    int64  \n",
      " 29  throw_ins_away          240 non-null    int64  \n",
      " 30  goal_kicks_home         240 non-null    int64  \n",
      " 31  goal_kicks_away         240 non-null    int64  \n",
      " 32  shots_inside_box_home   240 non-null    int64  \n",
      " 33  shots_inside_box_away   240 non-null    int64  \n",
      " 34  shots_outside_box_home  240 non-null    int64  \n",
      " 35  shots_outside_box_away  240 non-null    int64  \n",
      " 36  goalkeeper_saves_home   240 non-null    int64  \n",
      " 37  goalkeeper_saves_away   240 non-null    int64  \n",
      " 38  passes_home             240 non-null    int64  \n",
      " 39  passes_away             240 non-null    int64  \n",
      " 40  accurate_passes_home    240 non-null    int64  \n",
      " 41  accurate_passes_away    240 non-null    int64  \n",
      " 42  long_balls_home         240 non-null    int64  \n",
      " 43  long_balls_away         240 non-null    int64  \n",
      " 44  crosses_home            240 non-null    int64  \n",
      " 45  crosses_away            240 non-null    int64  \n",
      " 46  dribbles_home           240 non-null    int64  \n",
      " 47  dribbles_away           240 non-null    int64  \n",
      " 48  possesion_lost_home     240 non-null    int64  \n",
      " 49  possesion_lost_away     240 non-null    int64  \n",
      " 50  duels_won_home          240 non-null    int64  \n",
      " 51  duels_won_away          240 non-null    int64  \n",
      " 52  aerials_won_home        240 non-null    int64  \n",
      " 53  aerials_won_away        240 non-null    int64  \n",
      " 54  tackles_home            240 non-null    int64  \n",
      " 55  tackles_away            240 non-null    int64  \n",
      " 56  interceptions_home      240 non-null    int64  \n",
      " 57  interceptions_away      240 non-null    int64  \n",
      " 58  clearences_home         240 non-null    int64  \n",
      " 59  clearences_away         240 non-null    int64  \n",
      " 60  result                  240 non-null    int64  \n",
      "dtypes: float64(2), int64(57), object(2)\n",
      "memory usage: 114.5+ KB\n"
     ]
    }
   ],
   "source": [
    "matches.info(verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resultado 0, hay: 59 partidos\n",
      "Resultado 1, hay: 97 partidos\n",
      "Resultado 2, hay: 84 partidos\n",
      "0.4041666666666667\n"
     ]
    }
   ],
   "source": [
    "#Cantidad de partidos por resultado (local, empate, visita)\n",
    "\n",
    "for i in range(0,3):\n",
    "    cantidad = matches[matches[\"result\"]==i].shape[0]\n",
    "    print(f\"Resultado {i}, hay: {cantidad} partidos\")\n",
    "num = 97/(59+97+84)\n",
    "print(num)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Columnas categóricas:\n",
      "[]\n"
     ]
    }
   ],
   "source": [
    "result = matches[\"result\"]\n",
    "matches = matches.drop(columns = [\"team_home\", \"team_away\", \"result\", \"score_home\", \"score_away\"])\n",
    "matches.head()\n",
    "\n",
    "columnas_categoricas = matches.select_dtypes(include=['object']).columns.tolist()\n",
    "\n",
    "print(\"Columnas categóricas:\")\n",
    "print(columnas_categoricas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convertir columnas categoricas a numericas\n",
    "\n",
    "#col = LabelEncoder().fit(matches[\"columna_tabla\"])\n",
    "#data[\"columna_tabla\"] = col.transform(matches[\"columna_tabla\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = matches.to_numpy()\n",
    "scaler = StandardScaler()\n",
    "X_escalados = scaler.fit_transform(X)\n",
    "y = result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "#X_train, X_test, y_train, y_test = train_test_split(X_escalados, y, test_size=0.3, random_state = 1, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/abstract_model.py:82: UserWarning: Device used : cpu\n",
      "  warnings.warn(f\"Device used : {self.device}\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 0  | loss: 1.15431 | train_accuracy: 0.34896 | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 1  | loss: 1.18652 | train_accuracy: 0.36458 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 2  | loss: 1.15968 | train_accuracy: 0.42708 | valid_accuracy: 0.33333 |  0:00:00s\n",
      "epoch 3  | loss: 1.20191 | train_accuracy: 0.40625 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 4  | loss: 1.15111 | train_accuracy: 0.41667 | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 5  | loss: 1.13387 | train_accuracy: 0.4375  | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 6  | loss: 1.16715 | train_accuracy: 0.43229 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 7  | loss: 1.1886  | train_accuracy: 0.45312 | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 8  | loss: 1.10109 | train_accuracy: 0.46354 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 9  | loss: 1.11418 | train_accuracy: 0.47396 | valid_accuracy: 0.47917 |  0:00:00s\n",
      "epoch 10 | loss: 1.07827 | train_accuracy: 0.5     | valid_accuracy: 0.47917 |  0:00:00s\n",
      "epoch 11 | loss: 1.1286  | train_accuracy: 0.5     | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 12 | loss: 1.14297 | train_accuracy: 0.5     | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 13 | loss: 1.09593 | train_accuracy: 0.52604 | valid_accuracy: 0.47917 |  0:00:00s\n",
      "epoch 14 | loss: 1.08891 | train_accuracy: 0.51562 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 15 | loss: 1.10324 | train_accuracy: 0.50521 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 16 | loss: 1.06947 | train_accuracy: 0.51042 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 17 | loss: 1.08179 | train_accuracy: 0.50521 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 18 | loss: 1.10758 | train_accuracy: 0.50521 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 19 | loss: 1.07706 | train_accuracy: 0.48958 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 20 | loss: 1.09385 | train_accuracy: 0.51042 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 21 | loss: 1.11253 | train_accuracy: 0.48958 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 22 | loss: 1.08632 | train_accuracy: 0.49479 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 23 | loss: 1.08302 | train_accuracy: 0.48958 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 24 | loss: 1.08578 | train_accuracy: 0.48438 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 25 | loss: 1.07713 | train_accuracy: 0.48958 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 26 | loss: 1.03409 | train_accuracy: 0.51562 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 27 | loss: 1.02594 | train_accuracy: 0.52083 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 28 | loss: 1.08117 | train_accuracy: 0.52604 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 29 | loss: 1.02035 | train_accuracy: 0.53125 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 30 | loss: 1.04941 | train_accuracy: 0.53125 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 31 | loss: 1.01207 | train_accuracy: 0.55729 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 32 | loss: 1.06847 | train_accuracy: 0.5625  | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 33 | loss: 1.01615 | train_accuracy: 0.53646 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 34 | loss: 1.04198 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 35 | loss: 1.01884 | train_accuracy: 0.54167 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 36 | loss: 1.02546 | train_accuracy: 0.55729 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 37 | loss: 1.03941 | train_accuracy: 0.5625  | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 38 | loss: 1.01404 | train_accuracy: 0.55729 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 39 | loss: 1.03712 | train_accuracy: 0.55208 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 40 | loss: 1.01846 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 41 | loss: 1.02081 | train_accuracy: 0.54167 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 42 | loss: 1.01521 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 43 | loss: 0.97763 | train_accuracy: 0.54688 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 44 | loss: 1.00515 | train_accuracy: 0.54167 | valid_accuracy: 0.41667 |  0:00:03s\n",
      "epoch 45 | loss: 0.99631 | train_accuracy: 0.55208 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 46 | loss: 1.00427 | train_accuracy: 0.55729 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 47 | loss: 1.02056 | train_accuracy: 0.55729 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 48 | loss: 0.99224 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 49 | loss: 0.99672 | train_accuracy: 0.54167 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 50 | loss: 0.99952 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 51 | loss: 0.99067 | train_accuracy: 0.57292 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 52 | loss: 1.01522 | train_accuracy: 0.5625  | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 53 | loss: 1.0194  | train_accuracy: 0.5625  | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 54 | loss: 0.96804 | train_accuracy: 0.56771 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 55 | loss: 1.0298  | train_accuracy: 0.55729 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 56 | loss: 0.99484 | train_accuracy: 0.56771 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 57 | loss: 1.01533 | train_accuracy: 0.56771 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 58 | loss: 1.01454 | train_accuracy: 0.5625  | valid_accuracy: 0.375   |  0:00:04s\n",
      "epoch 59 | loss: 0.99318 | train_accuracy: 0.5625  | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 60 | loss: 0.9845  | train_accuracy: 0.57812 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 61 | loss: 0.98623 | train_accuracy: 0.58854 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 62 | loss: 0.98494 | train_accuracy: 0.60417 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 63 | loss: 1.03119 | train_accuracy: 0.60417 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 64 | loss: 0.95782 | train_accuracy: 0.60938 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 65 | loss: 0.95668 | train_accuracy: 0.61458 | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 66 | loss: 0.89601 | train_accuracy: 0.60417 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 67 | loss: 0.95883 | train_accuracy: 0.60938 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 68 | loss: 0.96985 | train_accuracy: 0.60938 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 69 | loss: 0.96763 | train_accuracy: 0.60938 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 70 | loss: 0.9662  | train_accuracy: 0.625   | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 71 | loss: 0.9311  | train_accuracy: 0.625   | valid_accuracy: 0.39583 |  0:00:04s\n",
      "epoch 72 | loss: 0.98962 | train_accuracy: 0.63021 | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 73 | loss: 0.9555  | train_accuracy: 0.63021 | valid_accuracy: 0.375   |  0:00:05s\n",
      "epoch 74 | loss: 0.93659 | train_accuracy: 0.64062 | valid_accuracy: 0.375   |  0:00:05s\n",
      "epoch 75 | loss: 0.9487  | train_accuracy: 0.63542 | valid_accuracy: 0.375   |  0:00:05s\n",
      "epoch 76 | loss: 0.89193 | train_accuracy: 0.64062 | valid_accuracy: 0.375   |  0:00:05s\n",
      "epoch 77 | loss: 0.92966 | train_accuracy: 0.64583 | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 78 | loss: 0.93415 | train_accuracy: 0.66146 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 79 | loss: 0.95923 | train_accuracy: 0.64062 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 80 | loss: 0.92354 | train_accuracy: 0.64062 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 81 | loss: 0.9571  | train_accuracy: 0.65625 | valid_accuracy: 0.41667 |  0:00:05s\n",
      "epoch 82 | loss: 0.8962  | train_accuracy: 0.65625 | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 83 | loss: 0.96366 | train_accuracy: 0.66146 | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 84 | loss: 0.94436 | train_accuracy: 0.6875  | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 85 | loss: 0.95764 | train_accuracy: 0.67188 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 86 | loss: 0.93546 | train_accuracy: 0.67188 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 87 | loss: 0.92022 | train_accuracy: 0.67708 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 88 | loss: 0.92068 | train_accuracy: 0.6875  | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 89 | loss: 0.95469 | train_accuracy: 0.68229 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 90 | loss: 0.93676 | train_accuracy: 0.68229 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 91 | loss: 0.94457 | train_accuracy: 0.69271 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 92 | loss: 0.90912 | train_accuracy: 0.69792 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 93 | loss: 0.91024 | train_accuracy: 0.68229 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 94 | loss: 0.90884 | train_accuracy: 0.69271 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 95 | loss: 0.93325 | train_accuracy: 0.70312 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 96 | loss: 0.89061 | train_accuracy: 0.69792 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 97 | loss: 0.92439 | train_accuracy: 0.69792 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 98 | loss: 0.88516 | train_accuracy: 0.71875 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 99 | loss: 0.90095 | train_accuracy: 0.72396 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 100| loss: 0.89347 | train_accuracy: 0.71875 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 101| loss: 0.86341 | train_accuracy: 0.74479 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 102| loss: 0.8811  | train_accuracy: 0.74479 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 103| loss: 0.83696 | train_accuracy: 0.71875 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 104| loss: 0.85705 | train_accuracy: 0.69271 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 105| loss: 0.87504 | train_accuracy: 0.71354 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 106| loss: 0.88291 | train_accuracy: 0.72396 | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 107| loss: 0.87934 | train_accuracy: 0.73438 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 108| loss: 0.84187 | train_accuracy: 0.74479 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 109| loss: 0.82199 | train_accuracy: 0.73958 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 110| loss: 0.8275  | train_accuracy: 0.73958 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "epoch 111| loss: 0.89153 | train_accuracy: 0.73958 | valid_accuracy: 0.39583 |  0:00:07s\n",
      "epoch 112| loss: 0.84806 | train_accuracy: 0.73958 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "epoch 113| loss: 0.81767 | train_accuracy: 0.73438 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "epoch 114| loss: 0.87023 | train_accuracy: 0.71875 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "epoch 115| loss: 0.84856 | train_accuracy: 0.72917 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "epoch 116| loss: 0.80438 | train_accuracy: 0.73958 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 117| loss: 0.85865 | train_accuracy: 0.73958 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "epoch 118| loss: 0.8392  | train_accuracy: 0.73958 | valid_accuracy: 0.39583 |  0:00:08s\n",
      "epoch 119| loss: 0.83217 | train_accuracy: 0.73438 | valid_accuracy: 0.39583 |  0:00:08s\n",
      "epoch 120| loss: 0.82226 | train_accuracy: 0.74479 | valid_accuracy: 0.39583 |  0:00:08s\n",
      "epoch 121| loss: 0.84349 | train_accuracy: 0.74479 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "epoch 122| loss: 0.8685  | train_accuracy: 0.73958 | valid_accuracy: 0.39583 |  0:00:08s\n",
      "epoch 123| loss: 0.85466 | train_accuracy: 0.72917 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "epoch 124| loss: 0.77303 | train_accuracy: 0.71875 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "epoch 125| loss: 0.79265 | train_accuracy: 0.73958 | valid_accuracy: 0.39583 |  0:00:08s\n",
      "epoch 126| loss: 0.84202 | train_accuracy: 0.73438 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "epoch 127| loss: 0.81286 | train_accuracy: 0.73438 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "epoch 128| loss: 0.83505 | train_accuracy: 0.73958 | valid_accuracy: 0.41667 |  0:00:08s\n",
      "\n",
      "Early stopping occurred at epoch 128 with best_epoch = 28 and best_valid_accuracy = 0.5\n",
      "epoch 0  | loss: 1.1633  | train_accuracy: 0.35938 | valid_accuracy: 0.39583 |  0:00:00s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/callbacks.py:172: UserWarning: Best weights from best epoch are automatically used!\n",
      "  warnings.warn(wrn_msg)\n",
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/abstract_model.py:82: UserWarning: Device used : cpu\n",
      "  warnings.warn(f\"Device used : {self.device}\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 1  | loss: 1.15766 | train_accuracy: 0.38542 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 2  | loss: 1.11117 | train_accuracy: 0.41667 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 3  | loss: 1.11695 | train_accuracy: 0.43229 | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 4  | loss: 1.10333 | train_accuracy: 0.46354 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 5  | loss: 1.10079 | train_accuracy: 0.46354 | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 6  | loss: 1.12493 | train_accuracy: 0.47917 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 7  | loss: 1.12894 | train_accuracy: 0.45833 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 8  | loss: 1.0539  | train_accuracy: 0.44271 | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 9  | loss: 1.13036 | train_accuracy: 0.44792 | valid_accuracy: 0.375   |  0:00:00s\n",
      "epoch 10 | loss: 1.08577 | train_accuracy: 0.44271 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 11 | loss: 1.06219 | train_accuracy: 0.44792 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 12 | loss: 1.07027 | train_accuracy: 0.46875 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 13 | loss: 1.09947 | train_accuracy: 0.46875 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 14 | loss: 1.13141 | train_accuracy: 0.46354 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 15 | loss: 1.10324 | train_accuracy: 0.47917 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 16 | loss: 1.09003 | train_accuracy: 0.47396 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 17 | loss: 1.0397  | train_accuracy: 0.46875 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 18 | loss: 1.07399 | train_accuracy: 0.48958 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 19 | loss: 1.04481 | train_accuracy: 0.48958 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 20 | loss: 1.09767 | train_accuracy: 0.48438 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 21 | loss: 1.03539 | train_accuracy: 0.48438 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 22 | loss: 1.01898 | train_accuracy: 0.51042 | valid_accuracy: 0.35417 |  0:00:01s\n",
      "epoch 23 | loss: 1.08658 | train_accuracy: 0.53125 | valid_accuracy: 0.35417 |  0:00:01s\n",
      "epoch 24 | loss: 1.06261 | train_accuracy: 0.50521 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 25 | loss: 1.05095 | train_accuracy: 0.52083 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 26 | loss: 1.08447 | train_accuracy: 0.53125 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 27 | loss: 1.01524 | train_accuracy: 0.53646 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 28 | loss: 1.0452  | train_accuracy: 0.53646 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 29 | loss: 1.0302  | train_accuracy: 0.53646 | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 30 | loss: 1.02435 | train_accuracy: 0.53125 | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 31 | loss: 1.01327 | train_accuracy: 0.53125 | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 32 | loss: 1.00147 | train_accuracy: 0.54688 | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 33 | loss: 1.06052 | train_accuracy: 0.55208 | valid_accuracy: 0.375   |  0:00:02s\n",
      "epoch 34 | loss: 1.06117 | train_accuracy: 0.55729 | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 35 | loss: 1.04178 | train_accuracy: 0.54167 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 36 | loss: 1.00214 | train_accuracy: 0.54167 | valid_accuracy: 0.375   |  0:00:02s\n",
      "epoch 37 | loss: 0.99435 | train_accuracy: 0.5625  | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 38 | loss: 1.00714 | train_accuracy: 0.55729 | valid_accuracy: 0.35417 |  0:00:02s\n",
      "epoch 39 | loss: 1.00558 | train_accuracy: 0.5625  | valid_accuracy: 0.35417 |  0:00:02s\n",
      "epoch 40 | loss: 1.02101 | train_accuracy: 0.5625  | valid_accuracy: 0.375   |  0:00:02s\n",
      "epoch 41 | loss: 1.03555 | train_accuracy: 0.55729 | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 42 | loss: 1.00725 | train_accuracy: 0.56771 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 43 | loss: 1.02941 | train_accuracy: 0.59375 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 44 | loss: 1.02712 | train_accuracy: 0.57292 | valid_accuracy: 0.35417 |  0:00:03s\n",
      "epoch 45 | loss: 1.01988 | train_accuracy: 0.55729 | valid_accuracy: 0.35417 |  0:00:03s\n",
      "epoch 46 | loss: 0.9833  | train_accuracy: 0.59375 | valid_accuracy: 0.35417 |  0:00:03s\n",
      "epoch 47 | loss: 1.02221 | train_accuracy: 0.59375 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 48 | loss: 1.01085 | train_accuracy: 0.57292 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 49 | loss: 0.96239 | train_accuracy: 0.57812 | valid_accuracy: 0.33333 |  0:00:03s\n",
      "epoch 50 | loss: 0.9913  | train_accuracy: 0.58333 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 51 | loss: 0.97531 | train_accuracy: 0.58333 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 52 | loss: 1.02292 | train_accuracy: 0.56771 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 53 | loss: 0.99647 | train_accuracy: 0.5625  | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 54 | loss: 0.97897 | train_accuracy: 0.58333 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 55 | loss: 1.01093 | train_accuracy: 0.58333 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 56 | loss: 0.96429 | train_accuracy: 0.58854 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 57 | loss: 0.98444 | train_accuracy: 0.59375 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 58 | loss: 0.96218 | train_accuracy: 0.61979 | valid_accuracy: 0.39583 |  0:00:04s\n",
      "epoch 59 | loss: 0.98086 | train_accuracy: 0.63542 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 60 | loss: 0.9821  | train_accuracy: 0.63021 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 61 | loss: 0.94983 | train_accuracy: 0.625   | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 62 | loss: 0.96926 | train_accuracy: 0.625   | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 63 | loss: 0.97444 | train_accuracy: 0.625   | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 64 | loss: 0.97384 | train_accuracy: 0.63021 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 65 | loss: 0.98656 | train_accuracy: 0.63021 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 66 | loss: 0.96213 | train_accuracy: 0.64062 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 67 | loss: 0.98193 | train_accuracy: 0.63021 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 68 | loss: 0.9511  | train_accuracy: 0.64062 | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 69 | loss: 0.94132 | train_accuracy: 0.65104 | valid_accuracy: 0.47917 |  0:00:05s\n",
      "epoch 70 | loss: 0.90123 | train_accuracy: 0.64062 | valid_accuracy: 0.5     |  0:00:05s\n",
      "epoch 71 | loss: 0.94258 | train_accuracy: 0.64062 | valid_accuracy: 0.52083 |  0:00:05s\n",
      "epoch 72 | loss: 0.94669 | train_accuracy: 0.63542 | valid_accuracy: 0.5     |  0:00:05s\n",
      "epoch 73 | loss: 0.94388 | train_accuracy: 0.64583 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 74 | loss: 0.93007 | train_accuracy: 0.65625 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 75 | loss: 0.92578 | train_accuracy: 0.6875  | valid_accuracy: 0.52083 |  0:00:05s\n",
      "epoch 76 | loss: 0.93742 | train_accuracy: 0.68229 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 77 | loss: 0.92397 | train_accuracy: 0.66146 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 78 | loss: 0.94608 | train_accuracy: 0.66667 | valid_accuracy: 0.58333 |  0:00:05s\n",
      "epoch 79 | loss: 0.91424 | train_accuracy: 0.64062 | valid_accuracy: 0.5     |  0:00:05s\n",
      "epoch 80 | loss: 0.93305 | train_accuracy: 0.63542 | valid_accuracy: 0.52083 |  0:00:05s\n",
      "epoch 81 | loss: 0.89802 | train_accuracy: 0.64583 | valid_accuracy: 0.52083 |  0:00:05s\n",
      "epoch 82 | loss: 0.90724 | train_accuracy: 0.64583 | valid_accuracy: 0.52083 |  0:00:05s\n",
      "epoch 83 | loss: 0.93934 | train_accuracy: 0.66146 | valid_accuracy: 0.5     |  0:00:05s\n",
      "epoch 84 | loss: 0.92723 | train_accuracy: 0.65625 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 85 | loss: 0.88346 | train_accuracy: 0.66146 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 86 | loss: 0.88774 | train_accuracy: 0.67188 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 87 | loss: 0.91471 | train_accuracy: 0.68229 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 88 | loss: 0.91033 | train_accuracy: 0.67188 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 89 | loss: 0.9603  | train_accuracy: 0.67708 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 90 | loss: 0.8528  | train_accuracy: 0.67188 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 91 | loss: 0.98476 | train_accuracy: 0.66667 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 92 | loss: 0.90362 | train_accuracy: 0.6875  | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 93 | loss: 0.92002 | train_accuracy: 0.67708 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 94 | loss: 0.91726 | train_accuracy: 0.70833 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 95 | loss: 0.89262 | train_accuracy: 0.71354 | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 96 | loss: 0.922   | train_accuracy: 0.72396 | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 97 | loss: 0.90267 | train_accuracy: 0.72917 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 98 | loss: 0.88244 | train_accuracy: 0.72396 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 99 | loss: 0.8799  | train_accuracy: 0.72396 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 100| loss: 0.92366 | train_accuracy: 0.72396 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 101| loss: 0.86294 | train_accuracy: 0.72396 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 102| loss: 0.85263 | train_accuracy: 0.73438 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 103| loss: 0.89073 | train_accuracy: 0.73438 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 104| loss: 0.86161 | train_accuracy: 0.75521 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 105| loss: 0.8716  | train_accuracy: 0.75    | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 106| loss: 0.89736 | train_accuracy: 0.75    | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 107| loss: 0.85931 | train_accuracy: 0.76042 | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 108| loss: 0.84453 | train_accuracy: 0.76042 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 109| loss: 0.82151 | train_accuracy: 0.77083 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 110| loss: 0.87775 | train_accuracy: 0.76562 | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 111| loss: 0.84071 | train_accuracy: 0.75521 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 112| loss: 0.90742 | train_accuracy: 0.75521 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 113| loss: 0.8243  | train_accuracy: 0.75    | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 114| loss: 0.8541  | train_accuracy: 0.75    | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 115| loss: 0.85138 | train_accuracy: 0.73958 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 116| loss: 0.8128  | train_accuracy: 0.75    | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 117| loss: 0.85845 | train_accuracy: 0.75521 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 118| loss: 0.89494 | train_accuracy: 0.75    | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 119| loss: 0.85339 | train_accuracy: 0.73438 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 120| loss: 0.84968 | train_accuracy: 0.74479 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 121| loss: 0.80216 | train_accuracy: 0.73958 | valid_accuracy: 0.54167 |  0:00:08s\n",
      "epoch 122| loss: 0.88995 | train_accuracy: 0.77083 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 123| loss: 0.90012 | train_accuracy: 0.75521 | valid_accuracy: 0.54167 |  0:00:08s\n",
      "epoch 124| loss: 0.84008 | train_accuracy: 0.76562 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 125| loss: 0.81948 | train_accuracy: 0.78125 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 126| loss: 0.85342 | train_accuracy: 0.78125 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 127| loss: 0.79626 | train_accuracy: 0.77604 | valid_accuracy: 0.54167 |  0:00:08s\n",
      "epoch 128| loss: 0.89165 | train_accuracy: 0.78125 | valid_accuracy: 0.5625  |  0:00:09s\n",
      "epoch 129| loss: 0.83704 | train_accuracy: 0.76042 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 130| loss: 0.8048  | train_accuracy: 0.75    | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 131| loss: 0.82513 | train_accuracy: 0.75    | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 132| loss: 0.79363 | train_accuracy: 0.76042 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 133| loss: 0.8311  | train_accuracy: 0.77083 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 134| loss: 0.82412 | train_accuracy: 0.76562 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 135| loss: 0.847   | train_accuracy: 0.77604 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 136| loss: 0.81254 | train_accuracy: 0.78646 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 137| loss: 0.81372 | train_accuracy: 0.78646 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 138| loss: 0.81577 | train_accuracy: 0.78125 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 139| loss: 0.79532 | train_accuracy: 0.77083 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 140| loss: 0.786   | train_accuracy: 0.78125 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 141| loss: 0.86023 | train_accuracy: 0.79167 | valid_accuracy: 0.5     |  0:00:09s\n",
      "epoch 142| loss: 0.83524 | train_accuracy: 0.78125 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 143| loss: 0.80777 | train_accuracy: 0.77604 | valid_accuracy: 0.5     |  0:00:10s\n",
      "epoch 144| loss: 0.77936 | train_accuracy: 0.78646 | valid_accuracy: 0.5     |  0:00:10s\n",
      "epoch 145| loss: 0.77262 | train_accuracy: 0.77083 | valid_accuracy: 0.5     |  0:00:10s\n",
      "epoch 146| loss: 0.77087 | train_accuracy: 0.78125 | valid_accuracy: 0.5     |  0:00:10s\n",
      "epoch 147| loss: 0.84294 | train_accuracy: 0.77604 | valid_accuracy: 0.47917 |  0:00:10s\n",
      "epoch 148| loss: 0.71512 | train_accuracy: 0.77604 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 149| loss: 0.75127 | train_accuracy: 0.79167 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 150| loss: 0.8035  | train_accuracy: 0.79167 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 151| loss: 0.77526 | train_accuracy: 0.78646 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 152| loss: 0.75303 | train_accuracy: 0.79167 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 153| loss: 0.76881 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 154| loss: 0.77841 | train_accuracy: 0.80208 | valid_accuracy: 0.47917 |  0:00:10s\n",
      "epoch 155| loss: 0.76793 | train_accuracy: 0.80208 | valid_accuracy: 0.5     |  0:00:10s\n",
      "epoch 156| loss: 0.77339 | train_accuracy: 0.80208 | valid_accuracy: 0.52083 |  0:00:10s\n",
      "epoch 157| loss: 0.76387 | train_accuracy: 0.80208 | valid_accuracy: 0.5     |  0:00:10s\n",
      "epoch 158| loss: 0.77206 | train_accuracy: 0.80729 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 159| loss: 0.75676 | train_accuracy: 0.80208 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 160| loss: 0.75496 | train_accuracy: 0.80729 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 161| loss: 0.74243 | train_accuracy: 0.8125  | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 162| loss: 0.75168 | train_accuracy: 0.8125  | valid_accuracy: 0.52083 |  0:00:11s\n",
      "epoch 163| loss: 0.76581 | train_accuracy: 0.8125  | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 164| loss: 0.79513 | train_accuracy: 0.81771 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 165| loss: 0.73699 | train_accuracy: 0.82292 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 166| loss: 0.75819 | train_accuracy: 0.81771 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 167| loss: 0.74879 | train_accuracy: 0.81771 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 168| loss: 0.73784 | train_accuracy: 0.81771 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 169| loss: 0.75042 | train_accuracy: 0.81771 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 170| loss: 0.7266  | train_accuracy: 0.82292 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 171| loss: 0.71787 | train_accuracy: 0.81771 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 172| loss: 0.77455 | train_accuracy: 0.81771 | valid_accuracy: 0.5     |  0:00:12s\n",
      "epoch 173| loss: 0.79629 | train_accuracy: 0.81771 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 174| loss: 0.8055  | train_accuracy: 0.80208 | valid_accuracy: 0.45833 |  0:00:12s\n",
      "epoch 175| loss: 0.70955 | train_accuracy: 0.8125  | valid_accuracy: 0.45833 |  0:00:12s\n",
      "epoch 176| loss: 0.75239 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:12s\n",
      "epoch 177| loss: 0.73289 | train_accuracy: 0.79688 | valid_accuracy: 0.4375  |  0:00:12s\n",
      "epoch 178| loss: 0.73361 | train_accuracy: 0.80208 | valid_accuracy: 0.4375  |  0:00:12s\n",
      "\n",
      "Early stopping occurred at epoch 178 with best_epoch = 78 and best_valid_accuracy = 0.58333\n",
      "epoch 0  | loss: 1.21665 | train_accuracy: 0.375   | valid_accuracy: 0.3125  |  0:00:00s\n",
      "epoch 1  | loss: 1.12691 | train_accuracy: 0.41146 | valid_accuracy: 0.3125  |  0:00:00s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/callbacks.py:172: UserWarning: Best weights from best epoch are automatically used!\n",
      "  warnings.warn(wrn_msg)\n",
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/abstract_model.py:82: UserWarning: Device used : cpu\n",
      "  warnings.warn(f\"Device used : {self.device}\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 2  | loss: 1.14949 | train_accuracy: 0.4375  | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 3  | loss: 1.20038 | train_accuracy: 0.42708 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 4  | loss: 1.09375 | train_accuracy: 0.41667 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 5  | loss: 1.19197 | train_accuracy: 0.41667 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 6  | loss: 1.111   | train_accuracy: 0.42188 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 7  | loss: 1.15288 | train_accuracy: 0.40625 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 8  | loss: 1.10676 | train_accuracy: 0.41667 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 9  | loss: 1.13055 | train_accuracy: 0.41667 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 10 | loss: 1.0639  | train_accuracy: 0.42708 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 11 | loss: 1.07731 | train_accuracy: 0.4375  | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 12 | loss: 1.10195 | train_accuracy: 0.42188 | valid_accuracy: 0.4375  |  0:00:00s\n",
      "epoch 13 | loss: 1.13944 | train_accuracy: 0.41667 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 14 | loss: 1.0542  | train_accuracy: 0.41667 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 15 | loss: 1.10484 | train_accuracy: 0.42708 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 16 | loss: 1.12424 | train_accuracy: 0.45312 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 17 | loss: 1.08649 | train_accuracy: 0.45833 | valid_accuracy: 0.52083 |  0:00:01s\n",
      "epoch 18 | loss: 1.0745  | train_accuracy: 0.46875 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 19 | loss: 1.0755  | train_accuracy: 0.48958 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 20 | loss: 1.05158 | train_accuracy: 0.49479 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 21 | loss: 1.08302 | train_accuracy: 0.5     | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 22 | loss: 1.08416 | train_accuracy: 0.52083 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 23 | loss: 1.06304 | train_accuracy: 0.53646 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 24 | loss: 1.0796  | train_accuracy: 0.52604 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 25 | loss: 1.03085 | train_accuracy: 0.51042 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 26 | loss: 1.01353 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 27 | loss: 1.07844 | train_accuracy: 0.55729 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 28 | loss: 1.00238 | train_accuracy: 0.56771 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 29 | loss: 1.06126 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 30 | loss: 1.05393 | train_accuracy: 0.54688 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 31 | loss: 1.04196 | train_accuracy: 0.5625  | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 32 | loss: 1.04554 | train_accuracy: 0.57292 | valid_accuracy: 0.52083 |  0:00:02s\n",
      "epoch 33 | loss: 1.03876 | train_accuracy: 0.57292 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 34 | loss: 1.02237 | train_accuracy: 0.57812 | valid_accuracy: 0.54167 |  0:00:02s\n",
      "epoch 35 | loss: 1.06258 | train_accuracy: 0.58854 | valid_accuracy: 0.54167 |  0:00:02s\n",
      "epoch 36 | loss: 1.00489 | train_accuracy: 0.57812 | valid_accuracy: 0.54167 |  0:00:02s\n",
      "epoch 37 | loss: 1.01907 | train_accuracy: 0.57812 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 38 | loss: 0.99119 | train_accuracy: 0.57292 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 39 | loss: 1.062   | train_accuracy: 0.60938 | valid_accuracy: 0.52083 |  0:00:02s\n",
      "epoch 40 | loss: 1.00612 | train_accuracy: 0.60938 | valid_accuracy: 0.52083 |  0:00:02s\n",
      "epoch 41 | loss: 1.05042 | train_accuracy: 0.61458 | valid_accuracy: 0.52083 |  0:00:02s\n",
      "epoch 42 | loss: 0.98679 | train_accuracy: 0.60938 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 43 | loss: 1.00001 | train_accuracy: 0.60417 | valid_accuracy: 0.47917 |  0:00:03s\n",
      "epoch 44 | loss: 1.01902 | train_accuracy: 0.61979 | valid_accuracy: 0.47917 |  0:00:03s\n",
      "epoch 45 | loss: 0.98332 | train_accuracy: 0.625   | valid_accuracy: 0.47917 |  0:00:03s\n",
      "epoch 46 | loss: 0.96941 | train_accuracy: 0.61458 | valid_accuracy: 0.47917 |  0:00:03s\n",
      "epoch 47 | loss: 1.01272 | train_accuracy: 0.61458 | valid_accuracy: 0.5     |  0:00:03s\n",
      "epoch 48 | loss: 1.0019  | train_accuracy: 0.60938 | valid_accuracy: 0.5     |  0:00:03s\n",
      "epoch 49 | loss: 0.99743 | train_accuracy: 0.60417 | valid_accuracy: 0.5     |  0:00:03s\n",
      "epoch 50 | loss: 0.97053 | train_accuracy: 0.59375 | valid_accuracy: 0.5     |  0:00:03s\n",
      "epoch 51 | loss: 0.97672 | train_accuracy: 0.59896 | valid_accuracy: 0.47917 |  0:00:03s\n",
      "epoch 52 | loss: 1.00061 | train_accuracy: 0.61458 | valid_accuracy: 0.5     |  0:00:03s\n",
      "epoch 53 | loss: 1.01185 | train_accuracy: 0.61458 | valid_accuracy: 0.54167 |  0:00:03s\n",
      "epoch 54 | loss: 0.96291 | train_accuracy: 0.625   | valid_accuracy: 0.54167 |  0:00:03s\n",
      "epoch 55 | loss: 0.95902 | train_accuracy: 0.61979 | valid_accuracy: 0.54167 |  0:00:03s\n",
      "epoch 56 | loss: 0.94834 | train_accuracy: 0.63021 | valid_accuracy: 0.54167 |  0:00:03s\n",
      "epoch 57 | loss: 0.98494 | train_accuracy: 0.64583 | valid_accuracy: 0.5625  |  0:00:03s\n",
      "epoch 58 | loss: 0.97828 | train_accuracy: 0.65104 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 59 | loss: 0.97026 | train_accuracy: 0.64062 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 60 | loss: 0.94329 | train_accuracy: 0.65104 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 61 | loss: 0.939   | train_accuracy: 0.66667 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 62 | loss: 0.95628 | train_accuracy: 0.65104 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 63 | loss: 0.99624 | train_accuracy: 0.64583 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 64 | loss: 0.97708 | train_accuracy: 0.65104 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 65 | loss: 0.99155 | train_accuracy: 0.64062 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 66 | loss: 0.99843 | train_accuracy: 0.63542 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 67 | loss: 0.97037 | train_accuracy: 0.63021 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 68 | loss: 0.98285 | train_accuracy: 0.64062 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 69 | loss: 0.94102 | train_accuracy: 0.63021 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 70 | loss: 0.98471 | train_accuracy: 0.66146 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 71 | loss: 0.94494 | train_accuracy: 0.66146 | valid_accuracy: 0.54167 |  0:00:04s\n",
      "epoch 72 | loss: 0.97093 | train_accuracy: 0.66667 | valid_accuracy: 0.5625  |  0:00:04s\n",
      "epoch 73 | loss: 0.95598 | train_accuracy: 0.67188 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 74 | loss: 0.92137 | train_accuracy: 0.66667 | valid_accuracy: 0.5625  |  0:00:05s\n",
      "epoch 75 | loss: 0.95138 | train_accuracy: 0.67188 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 76 | loss: 0.94702 | train_accuracy: 0.6875  | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 77 | loss: 0.90639 | train_accuracy: 0.70312 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 78 | loss: 0.95692 | train_accuracy: 0.69792 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 79 | loss: 0.94739 | train_accuracy: 0.6875  | valid_accuracy: 0.5625  |  0:00:05s\n",
      "epoch 80 | loss: 0.96122 | train_accuracy: 0.6875  | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 81 | loss: 0.93458 | train_accuracy: 0.6875  | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 82 | loss: 0.92068 | train_accuracy: 0.6875  | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 83 | loss: 0.94617 | train_accuracy: 0.69792 | valid_accuracy: 0.54167 |  0:00:05s\n",
      "epoch 84 | loss: 0.94809 | train_accuracy: 0.70312 | valid_accuracy: 0.58333 |  0:00:05s\n",
      "epoch 85 | loss: 0.90693 | train_accuracy: 0.70312 | valid_accuracy: 0.5625  |  0:00:05s\n",
      "epoch 86 | loss: 0.97071 | train_accuracy: 0.72396 | valid_accuracy: 0.5625  |  0:00:05s\n",
      "epoch 87 | loss: 0.91472 | train_accuracy: 0.69792 | valid_accuracy: 0.5625  |  0:00:05s\n",
      "epoch 88 | loss: 0.89328 | train_accuracy: 0.6875  | valid_accuracy: 0.5625  |  0:00:06s\n",
      "epoch 89 | loss: 0.89412 | train_accuracy: 0.69792 | valid_accuracy: 0.54167 |  0:00:06s\n",
      "epoch 90 | loss: 0.87328 | train_accuracy: 0.69792 | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 91 | loss: 0.88889 | train_accuracy: 0.70833 | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 92 | loss: 0.88823 | train_accuracy: 0.69792 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 93 | loss: 0.9189  | train_accuracy: 0.68229 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 94 | loss: 0.91699 | train_accuracy: 0.69792 | valid_accuracy: 0.54167 |  0:00:06s\n",
      "epoch 95 | loss: 0.86409 | train_accuracy: 0.70312 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 96 | loss: 0.91721 | train_accuracy: 0.69792 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 97 | loss: 0.92196 | train_accuracy: 0.69792 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 98 | loss: 0.90443 | train_accuracy: 0.70312 | valid_accuracy: 0.54167 |  0:00:06s\n",
      "epoch 99 | loss: 0.87618 | train_accuracy: 0.70833 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 100| loss: 0.90979 | train_accuracy: 0.70833 | valid_accuracy: 0.52083 |  0:00:06s\n",
      "epoch 101| loss: 0.87744 | train_accuracy: 0.70312 | valid_accuracy: 0.54167 |  0:00:06s\n",
      "epoch 102| loss: 0.90413 | train_accuracy: 0.69271 | valid_accuracy: 0.54167 |  0:00:06s\n",
      "epoch 103| loss: 0.90139 | train_accuracy: 0.69792 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 104| loss: 0.85909 | train_accuracy: 0.69792 | valid_accuracy: 0.54167 |  0:00:07s\n",
      "epoch 105| loss: 0.87286 | train_accuracy: 0.70312 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 106| loss: 0.87055 | train_accuracy: 0.70833 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 107| loss: 0.87235 | train_accuracy: 0.70312 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 108| loss: 0.86083 | train_accuracy: 0.70312 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 109| loss: 0.84433 | train_accuracy: 0.69792 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 110| loss: 0.92438 | train_accuracy: 0.69271 | valid_accuracy: 0.52083 |  0:00:07s\n",
      "epoch 111| loss: 0.89657 | train_accuracy: 0.69271 | valid_accuracy: 0.52083 |  0:00:07s\n",
      "epoch 112| loss: 0.88852 | train_accuracy: 0.70833 | valid_accuracy: 0.54167 |  0:00:07s\n",
      "epoch 113| loss: 0.849   | train_accuracy: 0.73438 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 114| loss: 0.8329  | train_accuracy: 0.72396 | valid_accuracy: 0.54167 |  0:00:07s\n",
      "epoch 115| loss: 0.84126 | train_accuracy: 0.71875 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 116| loss: 0.8313  | train_accuracy: 0.72396 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 117| loss: 0.84187 | train_accuracy: 0.73958 | valid_accuracy: 0.5625  |  0:00:07s\n",
      "epoch 118| loss: 0.8642  | train_accuracy: 0.73438 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 119| loss: 0.79972 | train_accuracy: 0.73438 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 120| loss: 0.80527 | train_accuracy: 0.72917 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 121| loss: 0.82324 | train_accuracy: 0.72917 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 122| loss: 0.81811 | train_accuracy: 0.72396 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 123| loss: 0.84054 | train_accuracy: 0.74479 | valid_accuracy: 0.54167 |  0:00:08s\n",
      "epoch 124| loss: 0.78742 | train_accuracy: 0.74479 | valid_accuracy: 0.54167 |  0:00:08s\n",
      "epoch 125| loss: 0.83826 | train_accuracy: 0.73958 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 126| loss: 0.90112 | train_accuracy: 0.74479 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 127| loss: 0.83356 | train_accuracy: 0.74479 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 128| loss: 0.87815 | train_accuracy: 0.74479 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 129| loss: 0.84108 | train_accuracy: 0.72917 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 130| loss: 0.87441 | train_accuracy: 0.74479 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 131| loss: 0.86787 | train_accuracy: 0.73438 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 132| loss: 0.84344 | train_accuracy: 0.76042 | valid_accuracy: 0.52083 |  0:00:08s\n",
      "epoch 133| loss: 0.87545 | train_accuracy: 0.76562 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 134| loss: 0.83978 | train_accuracy: 0.76042 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 135| loss: 0.80633 | train_accuracy: 0.75    | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 136| loss: 0.80095 | train_accuracy: 0.75521 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 137| loss: 0.82639 | train_accuracy: 0.75521 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 138| loss: 0.78867 | train_accuracy: 0.76562 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 139| loss: 0.80514 | train_accuracy: 0.75521 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 140| loss: 0.75693 | train_accuracy: 0.76562 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 141| loss: 0.80374 | train_accuracy: 0.77083 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 142| loss: 0.85674 | train_accuracy: 0.76562 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 143| loss: 0.79668 | train_accuracy: 0.76562 | valid_accuracy: 0.52083 |  0:00:09s\n",
      "epoch 144| loss: 0.79775 | train_accuracy: 0.76562 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 145| loss: 0.85549 | train_accuracy: 0.76562 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 146| loss: 0.78489 | train_accuracy: 0.77083 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 147| loss: 0.81504 | train_accuracy: 0.76042 | valid_accuracy: 0.54167 |  0:00:09s\n",
      "epoch 148| loss: 0.81404 | train_accuracy: 0.75521 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 149| loss: 0.81926 | train_accuracy: 0.74479 | valid_accuracy: 0.5625  |  0:00:10s\n",
      "epoch 150| loss: 0.79056 | train_accuracy: 0.75521 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 151| loss: 0.85572 | train_accuracy: 0.75521 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 152| loss: 0.76731 | train_accuracy: 0.77083 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 153| loss: 0.781   | train_accuracy: 0.77083 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 154| loss: 0.78455 | train_accuracy: 0.77083 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 155| loss: 0.78174 | train_accuracy: 0.78125 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 156| loss: 0.82503 | train_accuracy: 0.78125 | valid_accuracy: 0.5625  |  0:00:10s\n",
      "epoch 157| loss: 0.79136 | train_accuracy: 0.78125 | valid_accuracy: 0.5625  |  0:00:10s\n",
      "epoch 158| loss: 0.76064 | train_accuracy: 0.79688 | valid_accuracy: 0.5625  |  0:00:10s\n",
      "epoch 159| loss: 0.80735 | train_accuracy: 0.79167 | valid_accuracy: 0.5625  |  0:00:10s\n",
      "epoch 160| loss: 0.81401 | train_accuracy: 0.78125 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 161| loss: 0.73901 | train_accuracy: 0.78125 | valid_accuracy: 0.54167 |  0:00:10s\n",
      "epoch 162| loss: 0.81695 | train_accuracy: 0.79167 | valid_accuracy: 0.5625  |  0:00:10s\n",
      "epoch 163| loss: 0.7144  | train_accuracy: 0.79688 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 164| loss: 0.75759 | train_accuracy: 0.80729 | valid_accuracy: 0.52083 |  0:00:11s\n",
      "epoch 165| loss: 0.70225 | train_accuracy: 0.79167 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 166| loss: 0.83684 | train_accuracy: 0.79167 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 167| loss: 0.74722 | train_accuracy: 0.80208 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 168| loss: 0.83016 | train_accuracy: 0.79167 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 169| loss: 0.74916 | train_accuracy: 0.78646 | valid_accuracy: 0.5625  |  0:00:11s\n",
      "epoch 170| loss: 0.73863 | train_accuracy: 0.79167 | valid_accuracy: 0.52083 |  0:00:11s\n",
      "epoch 171| loss: 0.73141 | train_accuracy: 0.78646 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 172| loss: 0.73764 | train_accuracy: 0.79167 | valid_accuracy: 0.54167 |  0:00:11s\n",
      "epoch 173| loss: 0.7759  | train_accuracy: 0.78125 | valid_accuracy: 0.5625  |  0:00:11s\n",
      "epoch 174| loss: 0.74498 | train_accuracy: 0.78125 | valid_accuracy: 0.5625  |  0:00:11s\n",
      "epoch 175| loss: 0.73319 | train_accuracy: 0.80729 | valid_accuracy: 0.5625  |  0:00:11s\n",
      "epoch 176| loss: 0.74698 | train_accuracy: 0.80208 | valid_accuracy: 0.5625  |  0:00:11s\n",
      "epoch 177| loss: 0.72665 | train_accuracy: 0.8125  | valid_accuracy: 0.5625  |  0:00:11s\n",
      "epoch 178| loss: 0.72894 | train_accuracy: 0.80729 | valid_accuracy: 0.5625  |  0:00:12s\n",
      "epoch 179| loss: 0.73936 | train_accuracy: 0.80729 | valid_accuracy: 0.5625  |  0:00:12s\n",
      "epoch 180| loss: 0.69955 | train_accuracy: 0.81771 | valid_accuracy: 0.5625  |  0:00:12s\n",
      "epoch 181| loss: 0.72794 | train_accuracy: 0.81771 | valid_accuracy: 0.5625  |  0:00:12s\n",
      "epoch 182| loss: 0.71919 | train_accuracy: 0.80729 | valid_accuracy: 0.5625  |  0:00:12s\n",
      "epoch 183| loss: 0.69597 | train_accuracy: 0.79167 | valid_accuracy: 0.54167 |  0:00:12s\n",
      "epoch 184| loss: 0.79131 | train_accuracy: 0.80208 | valid_accuracy: 0.54167 |  0:00:12s\n",
      "\n",
      "Early stopping occurred at epoch 184 with best_epoch = 84 and best_valid_accuracy = 0.58333\n",
      "epoch 0  | loss: 1.1709  | train_accuracy: 0.35417 | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 1  | loss: 1.18034 | train_accuracy: 0.38021 | valid_accuracy: 0.375   |  0:00:00s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/callbacks.py:172: UserWarning: Best weights from best epoch are automatically used!\n",
      "  warnings.warn(wrn_msg)\n",
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/abstract_model.py:82: UserWarning: Device used : cpu\n",
      "  warnings.warn(f\"Device used : {self.device}\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 2  | loss: 1.13193 | train_accuracy: 0.41146 | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 3  | loss: 1.1455  | train_accuracy: 0.42188 | valid_accuracy: 0.3125  |  0:00:00s\n",
      "epoch 4  | loss: 1.10844 | train_accuracy: 0.43229 | valid_accuracy: 0.3125  |  0:00:00s\n",
      "epoch 5  | loss: 1.16993 | train_accuracy: 0.43229 | valid_accuracy: 0.29167 |  0:00:00s\n",
      "epoch 6  | loss: 1.1317  | train_accuracy: 0.44792 | valid_accuracy: 0.33333 |  0:00:00s\n",
      "epoch 7  | loss: 1.13085 | train_accuracy: 0.44792 | valid_accuracy: 0.33333 |  0:00:00s\n",
      "epoch 8  | loss: 1.08409 | train_accuracy: 0.45833 | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 9  | loss: 1.14956 | train_accuracy: 0.47396 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 10 | loss: 1.08056 | train_accuracy: 0.5     | valid_accuracy: 0.375   |  0:00:00s\n",
      "epoch 11 | loss: 1.15096 | train_accuracy: 0.51042 | valid_accuracy: 0.375   |  0:00:00s\n",
      "epoch 12 | loss: 1.10588 | train_accuracy: 0.48958 | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 13 | loss: 1.08543 | train_accuracy: 0.47917 | valid_accuracy: 0.33333 |  0:00:00s\n",
      "epoch 14 | loss: 1.06082 | train_accuracy: 0.47917 | valid_accuracy: 0.33333 |  0:00:01s\n",
      "epoch 15 | loss: 1.0688  | train_accuracy: 0.48438 | valid_accuracy: 0.33333 |  0:00:01s\n",
      "epoch 16 | loss: 1.0836  | train_accuracy: 0.49479 | valid_accuracy: 0.35417 |  0:00:01s\n",
      "epoch 17 | loss: 1.08463 | train_accuracy: 0.48958 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 18 | loss: 1.01326 | train_accuracy: 0.48438 | valid_accuracy: 0.33333 |  0:00:01s\n",
      "epoch 19 | loss: 1.0978  | train_accuracy: 0.51042 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 20 | loss: 1.04685 | train_accuracy: 0.52083 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 21 | loss: 1.03941 | train_accuracy: 0.51562 | valid_accuracy: 0.375   |  0:00:01s\n",
      "epoch 22 | loss: 1.07186 | train_accuracy: 0.52604 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 23 | loss: 1.06284 | train_accuracy: 0.53125 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 24 | loss: 1.05977 | train_accuracy: 0.54167 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 25 | loss: 1.05056 | train_accuracy: 0.55729 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 26 | loss: 1.06008 | train_accuracy: 0.58333 | valid_accuracy: 0.41667 |  0:00:01s\n",
      "epoch 27 | loss: 1.05288 | train_accuracy: 0.55729 | valid_accuracy: 0.39583 |  0:00:01s\n",
      "epoch 28 | loss: 1.05183 | train_accuracy: 0.5625  | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 29 | loss: 1.05008 | train_accuracy: 0.54688 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 30 | loss: 1.04062 | train_accuracy: 0.53125 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 31 | loss: 1.02163 | train_accuracy: 0.53646 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 32 | loss: 1.02293 | train_accuracy: 0.56771 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 33 | loss: 1.0561  | train_accuracy: 0.5625  | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 34 | loss: 1.01717 | train_accuracy: 0.55208 | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 35 | loss: 1.01232 | train_accuracy: 0.55729 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 36 | loss: 1.00541 | train_accuracy: 0.54688 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 37 | loss: 1.01498 | train_accuracy: 0.55729 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 38 | loss: 0.99903 | train_accuracy: 0.55208 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 39 | loss: 0.97815 | train_accuracy: 0.57812 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 40 | loss: 0.98316 | train_accuracy: 0.59896 | valid_accuracy: 0.375   |  0:00:02s\n",
      "epoch 41 | loss: 0.99966 | train_accuracy: 0.58333 | valid_accuracy: 0.375   |  0:00:02s\n",
      "epoch 42 | loss: 0.99943 | train_accuracy: 0.58854 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 43 | loss: 0.9902  | train_accuracy: 0.58333 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 44 | loss: 1.0642  | train_accuracy: 0.59375 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 45 | loss: 1.00787 | train_accuracy: 0.61979 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 46 | loss: 1.02993 | train_accuracy: 0.60938 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 47 | loss: 1.03417 | train_accuracy: 0.60938 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 48 | loss: 1.0107  | train_accuracy: 0.625   | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 49 | loss: 0.98477 | train_accuracy: 0.63542 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 50 | loss: 0.96307 | train_accuracy: 0.65104 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 51 | loss: 1.03212 | train_accuracy: 0.65104 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 52 | loss: 0.96351 | train_accuracy: 0.65625 | valid_accuracy: 0.375   |  0:00:03s\n",
      "epoch 53 | loss: 0.97089 | train_accuracy: 0.66146 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 54 | loss: 0.96258 | train_accuracy: 0.65104 | valid_accuracy: 0.41667 |  0:00:03s\n",
      "epoch 55 | loss: 1.01405 | train_accuracy: 0.64583 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 56 | loss: 0.97093 | train_accuracy: 0.64583 | valid_accuracy: 0.39583 |  0:00:03s\n",
      "epoch 57 | loss: 0.94693 | train_accuracy: 0.65625 | valid_accuracy: 0.35417 |  0:00:04s\n",
      "epoch 58 | loss: 0.96947 | train_accuracy: 0.64583 | valid_accuracy: 0.375   |  0:00:04s\n",
      "epoch 59 | loss: 0.98244 | train_accuracy: 0.64062 | valid_accuracy: 0.375   |  0:00:04s\n",
      "epoch 60 | loss: 0.95954 | train_accuracy: 0.65104 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 61 | loss: 0.94441 | train_accuracy: 0.66667 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 62 | loss: 0.96809 | train_accuracy: 0.67188 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 63 | loss: 0.99466 | train_accuracy: 0.66146 | valid_accuracy: 0.39583 |  0:00:04s\n",
      "epoch 64 | loss: 0.97061 | train_accuracy: 0.66146 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 65 | loss: 0.97159 | train_accuracy: 0.66667 | valid_accuracy: 0.375   |  0:00:04s\n",
      "epoch 66 | loss: 0.93983 | train_accuracy: 0.67708 | valid_accuracy: 0.375   |  0:00:04s\n",
      "epoch 67 | loss: 0.95527 | train_accuracy: 0.67708 | valid_accuracy: 0.39583 |  0:00:04s\n",
      "epoch 68 | loss: 0.95201 | train_accuracy: 0.66146 | valid_accuracy: 0.41667 |  0:00:04s\n",
      "epoch 69 | loss: 0.8883  | train_accuracy: 0.67708 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 70 | loss: 0.93772 | train_accuracy: 0.67708 | valid_accuracy: 0.39583 |  0:00:04s\n",
      "epoch 71 | loss: 0.90855 | train_accuracy: 0.68229 | valid_accuracy: 0.375   |  0:00:05s\n",
      "epoch 72 | loss: 0.97122 | train_accuracy: 0.68229 | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 73 | loss: 0.95664 | train_accuracy: 0.68229 | valid_accuracy: 0.375   |  0:00:05s\n",
      "epoch 74 | loss: 0.91721 | train_accuracy: 0.69271 | valid_accuracy: 0.41667 |  0:00:05s\n",
      "epoch 75 | loss: 0.9141  | train_accuracy: 0.6875  | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 76 | loss: 0.89902 | train_accuracy: 0.69271 | valid_accuracy: 0.39583 |  0:00:05s\n",
      "epoch 77 | loss: 0.93757 | train_accuracy: 0.69792 | valid_accuracy: 0.41667 |  0:00:05s\n",
      "epoch 78 | loss: 0.93988 | train_accuracy: 0.71354 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 79 | loss: 0.91845 | train_accuracy: 0.71354 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 80 | loss: 0.94582 | train_accuracy: 0.72917 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 81 | loss: 0.90034 | train_accuracy: 0.71354 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 82 | loss: 0.91025 | train_accuracy: 0.71875 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 83 | loss: 0.90131 | train_accuracy: 0.71875 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 84 | loss: 0.92526 | train_accuracy: 0.72396 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 85 | loss: 0.90667 | train_accuracy: 0.71875 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 86 | loss: 0.8747  | train_accuracy: 0.71875 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 87 | loss: 0.88043 | train_accuracy: 0.70312 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 88 | loss: 0.92702 | train_accuracy: 0.69792 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 89 | loss: 0.93705 | train_accuracy: 0.70312 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 90 | loss: 0.86784 | train_accuracy: 0.70312 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 91 | loss: 0.9157  | train_accuracy: 0.70833 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 92 | loss: 0.90513 | train_accuracy: 0.71875 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 93 | loss: 0.88894 | train_accuracy: 0.71875 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 94 | loss: 0.87341 | train_accuracy: 0.71875 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 95 | loss: 0.86428 | train_accuracy: 0.71875 | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 96 | loss: 0.87288 | train_accuracy: 0.72917 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 97 | loss: 0.88972 | train_accuracy: 0.75    | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 98 | loss: 0.88462 | train_accuracy: 0.76562 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 99 | loss: 0.83531 | train_accuracy: 0.75521 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 100| loss: 0.91086 | train_accuracy: 0.75521 | valid_accuracy: 0.5     |  0:00:06s\n",
      "epoch 101| loss: 0.81856 | train_accuracy: 0.76562 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 102| loss: 0.92752 | train_accuracy: 0.79167 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 103| loss: 0.86683 | train_accuracy: 0.79167 | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 104| loss: 0.83667 | train_accuracy: 0.80729 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 105| loss: 0.8418  | train_accuracy: 0.8125  | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 106| loss: 0.86266 | train_accuracy: 0.78646 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 107| loss: 0.8609  | train_accuracy: 0.79688 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 108| loss: 0.81552 | train_accuracy: 0.80208 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 109| loss: 0.83649 | train_accuracy: 0.77604 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 110| loss: 0.83892 | train_accuracy: 0.77604 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 111| loss: 0.84951 | train_accuracy: 0.79688 | valid_accuracy: 0.5     |  0:00:07s\n",
      "epoch 112| loss: 0.86887 | train_accuracy: 0.80208 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 113| loss: 0.81131 | train_accuracy: 0.80729 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 114| loss: 0.82532 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 115| loss: 0.819   | train_accuracy: 0.78646 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 116| loss: 0.84051 | train_accuracy: 0.79688 | valid_accuracy: 0.47917 |  0:00:07s\n",
      "epoch 117| loss: 0.80627 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:08s\n",
      "epoch 118| loss: 0.8124  | train_accuracy: 0.80208 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 119| loss: 0.84503 | train_accuracy: 0.80729 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 120| loss: 0.80063 | train_accuracy: 0.78646 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 121| loss: 0.8618  | train_accuracy: 0.8125  | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 122| loss: 0.81824 | train_accuracy: 0.80729 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 123| loss: 0.821   | train_accuracy: 0.80729 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 124| loss: 0.80759 | train_accuracy: 0.80208 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 125| loss: 0.83352 | train_accuracy: 0.8125  | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 126| loss: 0.80364 | train_accuracy: 0.81771 | valid_accuracy: 0.47917 |  0:00:08s\n",
      "epoch 127| loss: 0.82398 | train_accuracy: 0.8125  | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 128| loss: 0.81935 | train_accuracy: 0.80208 | valid_accuracy: 0.45833 |  0:00:08s\n",
      "epoch 129| loss: 0.81841 | train_accuracy: 0.79688 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 130| loss: 0.83034 | train_accuracy: 0.80729 | valid_accuracy: 0.5     |  0:00:08s\n",
      "epoch 131| loss: 0.79288 | train_accuracy: 0.80729 | valid_accuracy: 0.5     |  0:00:09s\n",
      "epoch 132| loss: 0.80724 | train_accuracy: 0.79167 | valid_accuracy: 0.47917 |  0:00:09s\n",
      "epoch 133| loss: 0.76993 | train_accuracy: 0.80208 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 134| loss: 0.86742 | train_accuracy: 0.80208 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 135| loss: 0.84491 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 136| loss: 0.79852 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 137| loss: 0.78254 | train_accuracy: 0.79688 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 138| loss: 0.77185 | train_accuracy: 0.80729 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 139| loss: 0.76681 | train_accuracy: 0.8125  | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 140| loss: 0.80123 | train_accuracy: 0.83333 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 141| loss: 0.81986 | train_accuracy: 0.82292 | valid_accuracy: 0.4375  |  0:00:09s\n",
      "epoch 142| loss: 0.7814  | train_accuracy: 0.82292 | valid_accuracy: 0.4375  |  0:00:09s\n",
      "epoch 143| loss: 0.81411 | train_accuracy: 0.83854 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 144| loss: 0.74936 | train_accuracy: 0.83333 | valid_accuracy: 0.45833 |  0:00:09s\n",
      "epoch 145| loss: 0.79574 | train_accuracy: 0.84375 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 146| loss: 0.72155 | train_accuracy: 0.84375 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 147| loss: 0.82315 | train_accuracy: 0.84375 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 148| loss: 0.7697  | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 149| loss: 0.75117 | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 150| loss: 0.7411  | train_accuracy: 0.85417 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 151| loss: 0.76149 | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:10s\n",
      "epoch 152| loss: 0.7355  | train_accuracy: 0.84896 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 153| loss: 0.77341 | train_accuracy: 0.84375 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 154| loss: 0.74194 | train_accuracy: 0.84375 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 155| loss: 0.74206 | train_accuracy: 0.84375 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 156| loss: 0.77182 | train_accuracy: 0.84375 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 157| loss: 0.73354 | train_accuracy: 0.84896 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 158| loss: 0.73661 | train_accuracy: 0.85417 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 159| loss: 0.74271 | train_accuracy: 0.85417 | valid_accuracy: 0.4375  |  0:00:10s\n",
      "epoch 160| loss: 0.81205 | train_accuracy: 0.85938 | valid_accuracy: 0.4375  |  0:00:11s\n",
      "epoch 161| loss: 0.72508 | train_accuracy: 0.85417 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 162| loss: 0.74992 | train_accuracy: 0.85417 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 163| loss: 0.7404  | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 164| loss: 0.74877 | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 165| loss: 0.76019 | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 166| loss: 0.76273 | train_accuracy: 0.85938 | valid_accuracy: 0.45833 |  0:00:11s\n",
      "epoch 167| loss: 0.74983 | train_accuracy: 0.85938 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 168| loss: 0.69587 | train_accuracy: 0.86458 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 169| loss: 0.78836 | train_accuracy: 0.86979 | valid_accuracy: 0.5     |  0:00:11s\n",
      "epoch 170| loss: 0.67505 | train_accuracy: 0.85938 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 171| loss: 0.67112 | train_accuracy: 0.85938 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 172| loss: 0.74085 | train_accuracy: 0.84896 | valid_accuracy: 0.47917 |  0:00:11s\n",
      "epoch 173| loss: 0.71446 | train_accuracy: 0.85938 | valid_accuracy: 0.4375  |  0:00:11s\n",
      "epoch 174| loss: 0.73047 | train_accuracy: 0.85938 | valid_accuracy: 0.4375  |  0:00:12s\n",
      "epoch 175| loss: 0.73419 | train_accuracy: 0.86979 | valid_accuracy: 0.4375  |  0:00:12s\n",
      "epoch 176| loss: 0.72542 | train_accuracy: 0.84896 | valid_accuracy: 0.45833 |  0:00:12s\n",
      "epoch 177| loss: 0.71563 | train_accuracy: 0.86458 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 178| loss: 0.76208 | train_accuracy: 0.86458 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 179| loss: 0.78148 | train_accuracy: 0.86458 | valid_accuracy: 0.45833 |  0:00:12s\n",
      "epoch 180| loss: 0.72298 | train_accuracy: 0.86458 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 181| loss: 0.68712 | train_accuracy: 0.85938 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 182| loss: 0.65344 | train_accuracy: 0.86979 | valid_accuracy: 0.5     |  0:00:12s\n",
      "epoch 183| loss: 0.68727 | train_accuracy: 0.86458 | valid_accuracy: 0.5     |  0:00:12s\n",
      "epoch 184| loss: 0.72843 | train_accuracy: 0.86458 | valid_accuracy: 0.5     |  0:00:12s\n",
      "epoch 185| loss: 0.68772 | train_accuracy: 0.88021 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 186| loss: 0.73963 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 187| loss: 0.73038 | train_accuracy: 0.88542 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 188| loss: 0.61391 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:12s\n",
      "epoch 189| loss: 0.73968 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:13s\n",
      "epoch 190| loss: 0.63749 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:13s\n",
      "epoch 191| loss: 0.70332 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:13s\n",
      "epoch 192| loss: 0.68347 | train_accuracy: 0.90104 | valid_accuracy: 0.47917 |  0:00:13s\n",
      "epoch 193| loss: 0.65771 | train_accuracy: 0.88542 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 194| loss: 0.65169 | train_accuracy: 0.875   | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 195| loss: 0.64308 | train_accuracy: 0.88542 | valid_accuracy: 0.52083 |  0:00:13s\n",
      "epoch 196| loss: 0.75609 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 197| loss: 0.72948 | train_accuracy: 0.89062 | valid_accuracy: 0.52083 |  0:00:13s\n",
      "epoch 198| loss: 0.71359 | train_accuracy: 0.88542 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 199| loss: 0.60872 | train_accuracy: 0.88542 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 200| loss: 0.61389 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 201| loss: 0.63588 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 202| loss: 0.7143  | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 203| loss: 0.70406 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:13s\n",
      "epoch 204| loss: 0.6104  | train_accuracy: 0.89583 | valid_accuracy: 0.5     |  0:00:14s\n",
      "epoch 205| loss: 0.69318 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:14s\n",
      "epoch 206| loss: 0.64728 | train_accuracy: 0.88542 | valid_accuracy: 0.45833 |  0:00:14s\n",
      "epoch 207| loss: 0.70164 | train_accuracy: 0.89062 | valid_accuracy: 0.45833 |  0:00:14s\n",
      "epoch 208| loss: 0.67242 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:14s\n",
      "epoch 209| loss: 0.63371 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:14s\n",
      "epoch 210| loss: 0.63458 | train_accuracy: 0.90104 | valid_accuracy: 0.47917 |  0:00:14s\n",
      "epoch 211| loss: 0.62827 | train_accuracy: 0.89583 | valid_accuracy: 0.5     |  0:00:14s\n",
      "epoch 212| loss: 0.60756 | train_accuracy: 0.90625 | valid_accuracy: 0.52083 |  0:00:14s\n",
      "epoch 213| loss: 0.68213 | train_accuracy: 0.91146 | valid_accuracy: 0.5     |  0:00:14s\n",
      "epoch 214| loss: 0.63069 | train_accuracy: 0.91146 | valid_accuracy: 0.47917 |  0:00:14s\n",
      "epoch 215| loss: 0.63181 | train_accuracy: 0.91146 | valid_accuracy: 0.5     |  0:00:14s\n",
      "epoch 216| loss: 0.5701  | train_accuracy: 0.90625 | valid_accuracy: 0.47917 |  0:00:14s\n",
      "epoch 217| loss: 0.63199 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:14s\n",
      "epoch 218| loss: 0.64908 | train_accuracy: 0.89583 | valid_accuracy: 0.5     |  0:00:15s\n",
      "epoch 219| loss: 0.63466 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:15s\n",
      "epoch 220| loss: 0.623   | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 221| loss: 0.69829 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:15s\n",
      "epoch 222| loss: 0.62293 | train_accuracy: 0.89062 | valid_accuracy: 0.52083 |  0:00:15s\n",
      "epoch 223| loss: 0.68453 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 224| loss: 0.65597 | train_accuracy: 0.89583 | valid_accuracy: 0.5     |  0:00:15s\n",
      "epoch 225| loss: 0.6749  | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 226| loss: 0.62655 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 227| loss: 0.64215 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 228| loss: 0.57731 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 229| loss: 0.65126 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:15s\n",
      "epoch 230| loss: 0.65322 | train_accuracy: 0.88542 | valid_accuracy: 0.45833 |  0:00:15s\n",
      "epoch 231| loss: 0.65253 | train_accuracy: 0.88021 | valid_accuracy: 0.45833 |  0:00:15s\n",
      "epoch 232| loss: 0.56456 | train_accuracy: 0.88542 | valid_accuracy: 0.45833 |  0:00:15s\n",
      "epoch 233| loss: 0.63073 | train_accuracy: 0.88542 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 234| loss: 0.63265 | train_accuracy: 0.88021 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 235| loss: 0.57059 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 236| loss: 0.72827 | train_accuracy: 0.89062 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 237| loss: 0.61901 | train_accuracy: 0.88542 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 238| loss: 0.63173 | train_accuracy: 0.88542 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 239| loss: 0.59643 | train_accuracy: 0.89583 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 240| loss: 0.59454 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 241| loss: 0.62357 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 242| loss: 0.59107 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 243| loss: 0.57882 | train_accuracy: 0.90104 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 244| loss: 0.59981 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 245| loss: 0.6023  | train_accuracy: 0.88542 | valid_accuracy: 0.52083 |  0:00:16s\n",
      "epoch 246| loss: 0.68237 | train_accuracy: 0.88542 | valid_accuracy: 0.47917 |  0:00:16s\n",
      "epoch 247| loss: 0.64816 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:16s\n",
      "epoch 248| loss: 0.51481 | train_accuracy: 0.89062 | valid_accuracy: 0.5     |  0:00:17s\n",
      "epoch 249| loss: 0.56311 | train_accuracy: 0.90104 | valid_accuracy: 0.5     |  0:00:17s\n",
      "epoch 250| loss: 0.61544 | train_accuracy: 0.90625 | valid_accuracy: 0.47917 |  0:00:17s\n",
      "epoch 251| loss: 0.62225 | train_accuracy: 0.91146 | valid_accuracy: 0.5     |  0:00:17s\n",
      "epoch 252| loss: 0.56287 | train_accuracy: 0.90625 | valid_accuracy: 0.47917 |  0:00:17s\n",
      "epoch 253| loss: 0.5458  | train_accuracy: 0.90625 | valid_accuracy: 0.47917 |  0:00:17s\n",
      "epoch 254| loss: 0.59986 | train_accuracy: 0.90625 | valid_accuracy: 0.5     |  0:00:17s\n",
      "epoch 255| loss: 0.58192 | train_accuracy: 0.91146 | valid_accuracy: 0.5     |  0:00:17s\n",
      "epoch 256| loss: 0.59908 | train_accuracy: 0.91146 | valid_accuracy: 0.52083 |  0:00:17s\n",
      "epoch 257| loss: 0.60166 | train_accuracy: 0.91146 | valid_accuracy: 0.52083 |  0:00:17s\n",
      "epoch 258| loss: 0.62183 | train_accuracy: 0.92188 | valid_accuracy: 0.54167 |  0:00:17s\n",
      "epoch 259| loss: 0.50624 | train_accuracy: 0.91667 | valid_accuracy: 0.54167 |  0:00:17s\n",
      "epoch 260| loss: 0.64019 | train_accuracy: 0.92188 | valid_accuracy: 0.5     |  0:00:17s\n",
      "epoch 261| loss: 0.61298 | train_accuracy: 0.91146 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 262| loss: 0.51222 | train_accuracy: 0.91667 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 263| loss: 0.55032 | train_accuracy: 0.91667 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 264| loss: 0.64005 | train_accuracy: 0.91667 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 265| loss: 0.57104 | train_accuracy: 0.91667 | valid_accuracy: 0.52083 |  0:00:18s\n",
      "epoch 266| loss: 0.59102 | train_accuracy: 0.91667 | valid_accuracy: 0.47917 |  0:00:18s\n",
      "epoch 267| loss: 0.52083 | train_accuracy: 0.91667 | valid_accuracy: 0.47917 |  0:00:18s\n",
      "epoch 268| loss: 0.56556 | train_accuracy: 0.91667 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 269| loss: 0.58128 | train_accuracy: 0.92188 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 270| loss: 0.62    | train_accuracy: 0.92188 | valid_accuracy: 0.47917 |  0:00:18s\n",
      "epoch 271| loss: 0.57904 | train_accuracy: 0.92188 | valid_accuracy: 0.5     |  0:00:18s\n",
      "epoch 272| loss: 0.62448 | train_accuracy: 0.92188 | valid_accuracy: 0.47917 |  0:00:18s\n",
      "epoch 273| loss: 0.58072 | train_accuracy: 0.92188 | valid_accuracy: 0.4375  |  0:00:18s\n",
      "epoch 274| loss: 0.54983 | train_accuracy: 0.92708 | valid_accuracy: 0.45833 |  0:00:18s\n",
      "epoch 275| loss: 0.59588 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 276| loss: 0.52582 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 277| loss: 0.50183 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 278| loss: 0.60275 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 279| loss: 0.57763 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 280| loss: 0.56858 | train_accuracy: 0.93229 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 281| loss: 0.62109 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 282| loss: 0.58931 | train_accuracy: 0.93229 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 283| loss: 0.54042 | train_accuracy: 0.92708 | valid_accuracy: 0.45833 |  0:00:19s\n",
      "epoch 284| loss: 0.54649 | train_accuracy: 0.92708 | valid_accuracy: 0.45833 |  0:00:19s\n",
      "epoch 285| loss: 0.59386 | train_accuracy: 0.92708 | valid_accuracy: 0.45833 |  0:00:19s\n",
      "epoch 286| loss: 0.54034 | train_accuracy: 0.93229 | valid_accuracy: 0.45833 |  0:00:19s\n",
      "epoch 287| loss: 0.51311 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 288| loss: 0.52663 | train_accuracy: 0.92708 | valid_accuracy: 0.47917 |  0:00:19s\n",
      "epoch 289| loss: 0.50524 | train_accuracy: 0.94271 | valid_accuracy: 0.5     |  0:00:20s\n",
      "epoch 290| loss: 0.48752 | train_accuracy: 0.94792 | valid_accuracy: 0.5     |  0:00:20s\n",
      "epoch 291| loss: 0.48374 | train_accuracy: 0.94792 | valid_accuracy: 0.5     |  0:00:20s\n",
      "epoch 292| loss: 0.54449 | train_accuracy: 0.94792 | valid_accuracy: 0.47917 |  0:00:20s\n",
      "epoch 293| loss: 0.52508 | train_accuracy: 0.94271 | valid_accuracy: 0.47917 |  0:00:20s\n",
      "epoch 294| loss: 0.56597 | train_accuracy: 0.94271 | valid_accuracy: 0.5     |  0:00:20s\n",
      "epoch 295| loss: 0.51148 | train_accuracy: 0.94792 | valid_accuracy: 0.5     |  0:00:20s\n",
      "epoch 296| loss: 0.52548 | train_accuracy: 0.95833 | valid_accuracy: 0.45833 |  0:00:20s\n",
      "epoch 297| loss: 0.49584 | train_accuracy: 0.95833 | valid_accuracy: 0.47917 |  0:00:20s\n",
      "epoch 298| loss: 0.526   | train_accuracy: 0.95833 | valid_accuracy: 0.45833 |  0:00:20s\n",
      "epoch 299| loss: 0.53895 | train_accuracy: 0.95833 | valid_accuracy: 0.47917 |  0:00:20s\n",
      "epoch 300| loss: 0.58236 | train_accuracy: 0.96354 | valid_accuracy: 0.4375  |  0:00:20s\n",
      "epoch 301| loss: 0.56514 | train_accuracy: 0.95833 | valid_accuracy: 0.45833 |  0:00:20s\n",
      "epoch 302| loss: 0.45012 | train_accuracy: 0.95833 | valid_accuracy: 0.45833 |  0:00:20s\n",
      "epoch 303| loss: 0.52959 | train_accuracy: 0.95833 | valid_accuracy: 0.45833 |  0:00:20s\n",
      "epoch 304| loss: 0.50455 | train_accuracy: 0.95833 | valid_accuracy: 0.47917 |  0:00:21s\n",
      "epoch 305| loss: 0.46266 | train_accuracy: 0.95833 | valid_accuracy: 0.47917 |  0:00:21s\n",
      "epoch 306| loss: 0.52762 | train_accuracy: 0.96354 | valid_accuracy: 0.5     |  0:00:21s\n",
      "epoch 307| loss: 0.53741 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 308| loss: 0.57044 | train_accuracy: 0.97396 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 309| loss: 0.45546 | train_accuracy: 0.97396 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 310| loss: 0.52885 | train_accuracy: 0.97396 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 311| loss: 0.46558 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 312| loss: 0.44886 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:21s\n",
      "epoch 313| loss: 0.53933 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 314| loss: 0.56983 | train_accuracy: 0.97396 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 315| loss: 0.50378 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 316| loss: 0.46373 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 317| loss: 0.52539 | train_accuracy: 0.96354 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 318| loss: 0.43316 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:21s\n",
      "epoch 319| loss: 0.52902 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:22s\n",
      "epoch 320| loss: 0.47452 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:22s\n",
      "epoch 321| loss: 0.45256 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:22s\n",
      "epoch 322| loss: 0.48581 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 323| loss: 0.50755 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 324| loss: 0.42647 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 325| loss: 0.4669  | train_accuracy: 0.97396 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 326| loss: 0.49399 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 327| loss: 0.52956 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 328| loss: 0.44305 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:22s\n",
      "epoch 329| loss: 0.46315 | train_accuracy: 0.96875 | valid_accuracy: 0.45833 |  0:00:22s\n",
      "epoch 330| loss: 0.50832 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 331| loss: 0.48835 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:22s\n",
      "epoch 332| loss: 0.4848  | train_accuracy: 0.96875 | valid_accuracy: 0.52083 |  0:00:22s\n",
      "epoch 333| loss: 0.50828 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:23s\n",
      "epoch 334| loss: 0.51776 | train_accuracy: 0.96875 | valid_accuracy: 0.47917 |  0:00:23s\n",
      "epoch 335| loss: 0.50015 | train_accuracy: 0.96875 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 336| loss: 0.47516 | train_accuracy: 0.96875 | valid_accuracy: 0.52083 |  0:00:23s\n",
      "epoch 337| loss: 0.48482 | train_accuracy: 0.96875 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 338| loss: 0.5286  | train_accuracy: 0.97396 | valid_accuracy: 0.52083 |  0:00:23s\n",
      "epoch 339| loss: 0.45319 | train_accuracy: 0.97917 | valid_accuracy: 0.52083 |  0:00:23s\n",
      "epoch 340| loss: 0.48623 | train_accuracy: 0.97917 | valid_accuracy: 0.52083 |  0:00:23s\n",
      "epoch 341| loss: 0.43757 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 342| loss: 0.45891 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 343| loss: 0.48831 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 344| loss: 0.451   | train_accuracy: 0.97396 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 345| loss: 0.44024 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 346| loss: 0.49715 | train_accuracy: 0.97917 | valid_accuracy: 0.52083 |  0:00:23s\n",
      "epoch 347| loss: 0.4797  | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:23s\n",
      "epoch 348| loss: 0.46151 | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 349| loss: 0.42976 | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 350| loss: 0.52991 | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 351| loss: 0.54258 | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 352| loss: 0.42068 | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 353| loss: 0.4633  | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 354| loss: 0.55863 | train_accuracy: 0.97917 | valid_accuracy: 0.47917 |  0:00:24s\n",
      "epoch 355| loss: 0.46392 | train_accuracy: 0.98438 | valid_accuracy: 0.5     |  0:00:24s\n",
      "epoch 356| loss: 0.45195 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:24s\n",
      "epoch 357| loss: 0.44634 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:24s\n",
      "epoch 358| loss: 0.50432 | train_accuracy: 0.97917 | valid_accuracy: 0.5     |  0:00:24s\n",
      "\n",
      "Early stopping occurred at epoch 358 with best_epoch = 258 and best_valid_accuracy = 0.54167\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/callbacks.py:172: UserWarning: Best weights from best epoch are automatically used!\n",
      "  warnings.warn(wrn_msg)\n",
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/abstract_model.py:82: UserWarning: Device used : cpu\n",
      "  warnings.warn(f\"Device used : {self.device}\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 0  | loss: 1.1589  | train_accuracy: 0.39062 | valid_accuracy: 0.22917 |  0:00:00s\n",
      "epoch 1  | loss: 1.15907 | train_accuracy: 0.39062 | valid_accuracy: 0.3125  |  0:00:00s\n",
      "epoch 2  | loss: 1.13093 | train_accuracy: 0.4375  | valid_accuracy: 0.33333 |  0:00:00s\n",
      "epoch 3  | loss: 1.13752 | train_accuracy: 0.4375  | valid_accuracy: 0.35417 |  0:00:00s\n",
      "epoch 4  | loss: 1.11492 | train_accuracy: 0.44271 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 5  | loss: 1.22809 | train_accuracy: 0.43229 | valid_accuracy: 0.39583 |  0:00:00s\n",
      "epoch 6  | loss: 1.12207 | train_accuracy: 0.44271 | valid_accuracy: 0.41667 |  0:00:00s\n",
      "epoch 7  | loss: 1.13545 | train_accuracy: 0.44792 | valid_accuracy: 0.47917 |  0:00:00s\n",
      "epoch 8  | loss: 1.12034 | train_accuracy: 0.44271 | valid_accuracy: 0.52083 |  0:00:00s\n",
      "epoch 9  | loss: 1.12698 | train_accuracy: 0.44792 | valid_accuracy: 0.52083 |  0:00:00s\n",
      "epoch 10 | loss: 1.12247 | train_accuracy: 0.46875 | valid_accuracy: 0.52083 |  0:00:00s\n",
      "epoch 11 | loss: 1.11799 | train_accuracy: 0.48438 | valid_accuracy: 0.5     |  0:00:00s\n",
      "epoch 12 | loss: 1.07384 | train_accuracy: 0.48958 | valid_accuracy: 0.47917 |  0:00:00s\n",
      "epoch 13 | loss: 1.10369 | train_accuracy: 0.49479 | valid_accuracy: 0.45833 |  0:00:00s\n",
      "epoch 14 | loss: 0.98429 | train_accuracy: 0.48958 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 15 | loss: 1.10696 | train_accuracy: 0.5     | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 16 | loss: 1.05847 | train_accuracy: 0.50521 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 17 | loss: 1.06964 | train_accuracy: 0.50521 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 18 | loss: 1.07745 | train_accuracy: 0.51042 | valid_accuracy: 0.4375  |  0:00:01s\n",
      "epoch 19 | loss: 1.09643 | train_accuracy: 0.51042 | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 20 | loss: 1.04755 | train_accuracy: 0.52083 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 21 | loss: 1.036   | train_accuracy: 0.53125 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 22 | loss: 1.11058 | train_accuracy: 0.53646 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 23 | loss: 1.06658 | train_accuracy: 0.54688 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 24 | loss: 1.01802 | train_accuracy: 0.55208 | valid_accuracy: 0.5     |  0:00:01s\n",
      "epoch 25 | loss: 1.04127 | train_accuracy: 0.55208 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 26 | loss: 0.9955  | train_accuracy: 0.5625  | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 27 | loss: 1.03202 | train_accuracy: 0.5625  | valid_accuracy: 0.45833 |  0:00:01s\n",
      "epoch 28 | loss: 1.05542 | train_accuracy: 0.54167 | valid_accuracy: 0.47917 |  0:00:01s\n",
      "epoch 29 | loss: 1.06152 | train_accuracy: 0.5625  | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 30 | loss: 1.02351 | train_accuracy: 0.55208 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 31 | loss: 1.00058 | train_accuracy: 0.54688 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 32 | loss: 1.03122 | train_accuracy: 0.56771 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 33 | loss: 1.00708 | train_accuracy: 0.57292 | valid_accuracy: 0.5     |  0:00:02s\n",
      "epoch 34 | loss: 1.03187 | train_accuracy: 0.58333 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 35 | loss: 1.03164 | train_accuracy: 0.58854 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 36 | loss: 0.96889 | train_accuracy: 0.58854 | valid_accuracy: 0.47917 |  0:00:02s\n",
      "epoch 37 | loss: 0.99551 | train_accuracy: 0.59375 | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 38 | loss: 1.03034 | train_accuracy: 0.58854 | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 39 | loss: 0.9974  | train_accuracy: 0.60417 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 40 | loss: 0.9817  | train_accuracy: 0.61458 | valid_accuracy: 0.45833 |  0:00:02s\n",
      "epoch 41 | loss: 0.97461 | train_accuracy: 0.60417 | valid_accuracy: 0.4375  |  0:00:02s\n",
      "epoch 42 | loss: 0.96105 | train_accuracy: 0.61979 | valid_accuracy: 0.39583 |  0:00:02s\n",
      "epoch 43 | loss: 1.01421 | train_accuracy: 0.61979 | valid_accuracy: 0.41667 |  0:00:02s\n",
      "epoch 44 | loss: 1.00263 | train_accuracy: 0.625   | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 45 | loss: 0.95629 | train_accuracy: 0.64583 | valid_accuracy: 0.41667 |  0:00:03s\n",
      "epoch 46 | loss: 0.97866 | train_accuracy: 0.64062 | valid_accuracy: 0.41667 |  0:00:03s\n",
      "epoch 47 | loss: 0.99287 | train_accuracy: 0.64583 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 48 | loss: 1.01172 | train_accuracy: 0.65625 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 49 | loss: 0.99661 | train_accuracy: 0.66146 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 50 | loss: 0.97197 | train_accuracy: 0.66146 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 51 | loss: 0.96997 | train_accuracy: 0.66667 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 52 | loss: 0.95224 | train_accuracy: 0.65625 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 53 | loss: 0.99482 | train_accuracy: 0.66146 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 54 | loss: 0.95944 | train_accuracy: 0.65104 | valid_accuracy: 0.45833 |  0:00:03s\n",
      "epoch 55 | loss: 0.93095 | train_accuracy: 0.65104 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 56 | loss: 1.02015 | train_accuracy: 0.66667 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 57 | loss: 0.94671 | train_accuracy: 0.65625 | valid_accuracy: 0.4375  |  0:00:03s\n",
      "epoch 58 | loss: 0.91907 | train_accuracy: 0.64583 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 59 | loss: 0.99962 | train_accuracy: 0.66146 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 60 | loss: 0.96382 | train_accuracy: 0.67708 | valid_accuracy: 0.4375  |  0:00:04s\n",
      "epoch 61 | loss: 0.96374 | train_accuracy: 0.70833 | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 62 | loss: 0.94634 | train_accuracy: 0.70312 | valid_accuracy: 0.47917 |  0:00:04s\n",
      "epoch 63 | loss: 0.95021 | train_accuracy: 0.69792 | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 64 | loss: 0.94138 | train_accuracy: 0.69271 | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 65 | loss: 0.93087 | train_accuracy: 0.6875  | valid_accuracy: 0.45833 |  0:00:04s\n",
      "epoch 66 | loss: 0.94342 | train_accuracy: 0.6875  | valid_accuracy: 0.5     |  0:00:04s\n",
      "epoch 67 | loss: 0.91329 | train_accuracy: 0.70312 | valid_accuracy: 0.5     |  0:00:04s\n",
      "epoch 68 | loss: 0.90729 | train_accuracy: 0.70312 | valid_accuracy: 0.52083 |  0:00:04s\n",
      "epoch 69 | loss: 0.92119 | train_accuracy: 0.69792 | valid_accuracy: 0.5     |  0:00:04s\n",
      "epoch 70 | loss: 0.92353 | train_accuracy: 0.70833 | valid_accuracy: 0.52083 |  0:00:04s\n",
      "epoch 71 | loss: 0.89796 | train_accuracy: 0.68229 | valid_accuracy: 0.5     |  0:00:04s\n",
      "epoch 72 | loss: 0.95235 | train_accuracy: 0.70312 | valid_accuracy: 0.47917 |  0:00:04s\n",
      "epoch 73 | loss: 0.96296 | train_accuracy: 0.69792 | valid_accuracy: 0.47917 |  0:00:05s\n",
      "epoch 74 | loss: 0.9126  | train_accuracy: 0.71354 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 75 | loss: 0.89252 | train_accuracy: 0.71354 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 76 | loss: 0.9006  | train_accuracy: 0.72396 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 77 | loss: 0.91133 | train_accuracy: 0.73438 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 78 | loss: 0.91728 | train_accuracy: 0.73958 | valid_accuracy: 0.47917 |  0:00:05s\n",
      "epoch 79 | loss: 0.85891 | train_accuracy: 0.74479 | valid_accuracy: 0.47917 |  0:00:05s\n",
      "epoch 80 | loss: 0.89017 | train_accuracy: 0.73438 | valid_accuracy: 0.47917 |  0:00:05s\n",
      "epoch 81 | loss: 0.86993 | train_accuracy: 0.74479 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 82 | loss: 0.8734  | train_accuracy: 0.73438 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 83 | loss: 0.90514 | train_accuracy: 0.73438 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 84 | loss: 0.88863 | train_accuracy: 0.75    | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 85 | loss: 0.89706 | train_accuracy: 0.73958 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 86 | loss: 0.82557 | train_accuracy: 0.74479 | valid_accuracy: 0.4375  |  0:00:05s\n",
      "epoch 87 | loss: 0.85783 | train_accuracy: 0.74479 | valid_accuracy: 0.45833 |  0:00:05s\n",
      "epoch 88 | loss: 0.83533 | train_accuracy: 0.74479 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 89 | loss: 0.87704 | train_accuracy: 0.75521 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 90 | loss: 0.8287  | train_accuracy: 0.76562 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 91 | loss: 0.8863  | train_accuracy: 0.76042 | valid_accuracy: 0.41667 |  0:00:06s\n",
      "epoch 92 | loss: 0.85528 | train_accuracy: 0.77083 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 93 | loss: 0.85743 | train_accuracy: 0.77604 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 94 | loss: 0.8473  | train_accuracy: 0.77604 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 95 | loss: 0.85018 | train_accuracy: 0.78646 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 96 | loss: 0.81156 | train_accuracy: 0.79167 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 97 | loss: 0.84067 | train_accuracy: 0.77604 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 98 | loss: 0.80725 | train_accuracy: 0.78125 | valid_accuracy: 0.45833 |  0:00:06s\n",
      "epoch 99 | loss: 0.83734 | train_accuracy: 0.79688 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 100| loss: 0.84653 | train_accuracy: 0.80208 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 101| loss: 0.85935 | train_accuracy: 0.79688 | valid_accuracy: 0.47917 |  0:00:06s\n",
      "epoch 102| loss: 0.84    | train_accuracy: 0.79167 | valid_accuracy: 0.4375  |  0:00:06s\n",
      "epoch 103| loss: 0.81087 | train_accuracy: 0.78125 | valid_accuracy: 0.45833 |  0:00:07s\n",
      "epoch 104| loss: 0.82385 | train_accuracy: 0.78125 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 105| loss: 0.86622 | train_accuracy: 0.77604 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 106| loss: 0.84202 | train_accuracy: 0.78646 | valid_accuracy: 0.4375  |  0:00:07s\n",
      "epoch 107| loss: 0.81106 | train_accuracy: 0.78646 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "epoch 108| loss: 0.77749 | train_accuracy: 0.80208 | valid_accuracy: 0.41667 |  0:00:07s\n",
      "\n",
      "Early stopping occurred at epoch 108 with best_epoch = 8 and best_valid_accuracy = 0.52083\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gustavofernandez/Desktop/pytorch-test/env/lib/python3.8/site-packages/pytorch_tabnet/callbacks.py:172: UserWarning: Best weights from best epoch are automatically used!\n",
      "  warnings.warn(wrn_msg)\n"
     ]
    }
   ],
   "source": [
    "kfold = KFold(n_splits=5)\n",
    "\n",
    "y_true_all = []\n",
    "y_pred_all = []\n",
    "\n",
    "for train_index, test_index in kfold.split(X_escalados):\n",
    "    X_train, X_test = X_escalados[train_index], X_escalados[test_index]\n",
    "    y_train, y_test = y[train_index], y[test_index]\n",
    "    clf = TabNetClassifier(\n",
    "        optimizer_fn=torch.optim.Adam,\n",
    "        optimizer_params=dict(lr=1e-3),\n",
    "        scheduler_params={\"step_size\":50, # how to use learning rate scheduler\n",
    "                            \"gamma\":0.9},\n",
    "        scheduler_fn=torch.optim.lr_scheduler.StepLR,\n",
    "        mask_type='entmax', # \"sparsemax\" \"entmax\"\n",
    "        )\n",
    "        \n",
    "    clf.fit(X_train, y_train,\n",
    "        eval_set=[(X_train, y_train), (X_test, y_test)], \n",
    "        eval_name=['train', 'valid'], \n",
    "        eval_metric=['accuracy'], \n",
    "        max_epochs=1000 , patience=100, \n",
    "        batch_size=28, drop_last=False)\n",
    "\n",
    "    pred = clf.predict(X_test)\n",
    "    # Almacenar las predicciones y las etiquetas verdaderas de este fold\n",
    "    y_true_all.extend(y_test)\n",
    "    y_pred_all.extend(pred)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classification report\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.39      0.22      0.28        59\n",
      "           1       0.55      0.72      0.62        97\n",
      "           2       0.57      0.54      0.55        84\n",
      "\n",
      "    accuracy                           0.53       240\n",
      "   macro avg       0.50      0.49      0.49       240\n",
      "weighted avg       0.52      0.53      0.51       240\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(\"classification report\")\n",
    "print(classification_report(y_true_all, y_pred_all))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cnf_matrix = confusion_matrix(y_true_all, y_pred_all)\n",
    "np.set_printoptions(precision=2)\n",
    "# Plot normalized confusion matrix\n",
    "plt.figure(figsize=(5,5))\n",
    "plot_confusion_matrix(cnf_matrix, classes=classes, normalize=True, title='Normalized confusion matrix')\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}