Project Important Factors in Heart Disease.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "47ddd4ec",
   "metadata": {},
   "source": [
    "### Import & Data Load"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c89c6aaa",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "plt.style.use(\"seaborn\")\n",
    "%matplotlib inline\n",
    "plt.rc('font', family='Malgun Gothic')\n",
    "\n",
    "seed = 0 # fixed seed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "1c71936e",
   "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>HeartDisease</th>\n",
       "      <th>BMI</th>\n",
       "      <th>Smoking</th>\n",
       "      <th>AlcoholDrinking</th>\n",
       "      <th>Stroke</th>\n",
       "      <th>PhysicalHealth</th>\n",
       "      <th>MentalHealth</th>\n",
       "      <th>DiffWalking</th>\n",
       "      <th>Sex</th>\n",
       "      <th>AgeCategory</th>\n",
       "      <th>Race</th>\n",
       "      <th>Diabetic</th>\n",
       "      <th>PhysicalActivity</th>\n",
       "      <th>GenHealth</th>\n",
       "      <th>SleepTime</th>\n",
       "      <th>Asthma</th>\n",
       "      <th>KidneyDisease</th>\n",
       "      <th>SkinCancer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>No</td>\n",
       "      <td>16.60</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>3.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Female</td>\n",
       "      <td>55-59</td>\n",
       "      <td>White</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Very good</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>No</td>\n",
       "      <td>20.34</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Female</td>\n",
       "      <td>80 or older</td>\n",
       "      <td>White</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Very good</td>\n",
       "      <td>7.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>No</td>\n",
       "      <td>26.58</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>20.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Male</td>\n",
       "      <td>65-69</td>\n",
       "      <td>White</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Fair</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>No</td>\n",
       "      <td>24.21</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Female</td>\n",
       "      <td>75-79</td>\n",
       "      <td>White</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Good</td>\n",
       "      <td>6.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>No</td>\n",
       "      <td>23.71</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Female</td>\n",
       "      <td>40-44</td>\n",
       "      <td>White</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Very good</td>\n",
       "      <td>8.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319790</th>\n",
       "      <td>Yes</td>\n",
       "      <td>27.41</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Male</td>\n",
       "      <td>60-64</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Fair</td>\n",
       "      <td>6.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319791</th>\n",
       "      <td>No</td>\n",
       "      <td>29.84</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Male</td>\n",
       "      <td>35-39</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Very good</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319792</th>\n",
       "      <td>No</td>\n",
       "      <td>24.24</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Female</td>\n",
       "      <td>45-49</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Good</td>\n",
       "      <td>6.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319793</th>\n",
       "      <td>No</td>\n",
       "      <td>32.81</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Female</td>\n",
       "      <td>25-29</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Good</td>\n",
       "      <td>12.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319794</th>\n",
       "      <td>No</td>\n",
       "      <td>46.56</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Female</td>\n",
       "      <td>80 or older</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Good</td>\n",
       "      <td>8.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>319795 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       HeartDisease    BMI Smoking AlcoholDrinking Stroke  PhysicalHealth  \\\n",
       "0                No  16.60     Yes              No     No             3.0   \n",
       "1                No  20.34      No              No    Yes             0.0   \n",
       "2                No  26.58     Yes              No     No            20.0   \n",
       "3                No  24.21      No              No     No             0.0   \n",
       "4                No  23.71      No              No     No            28.0   \n",
       "...             ...    ...     ...             ...    ...             ...   \n",
       "319790          Yes  27.41     Yes              No     No             7.0   \n",
       "319791           No  29.84     Yes              No     No             0.0   \n",
       "319792           No  24.24      No              No     No             0.0   \n",
       "319793           No  32.81      No              No     No             0.0   \n",
       "319794           No  46.56      No              No     No             0.0   \n",
       "\n",
       "        MentalHealth DiffWalking     Sex  AgeCategory      Race Diabetic  \\\n",
       "0               30.0          No  Female        55-59     White      Yes   \n",
       "1                0.0          No  Female  80 or older     White       No   \n",
       "2               30.0          No    Male        65-69     White      Yes   \n",
       "3                0.0          No  Female        75-79     White       No   \n",
       "4                0.0         Yes  Female        40-44     White       No   \n",
       "...              ...         ...     ...          ...       ...      ...   \n",
       "319790           0.0         Yes    Male        60-64  Hispanic      Yes   \n",
       "319791           0.0          No    Male        35-39  Hispanic       No   \n",
       "319792           0.0          No  Female        45-49  Hispanic       No   \n",
       "319793           0.0          No  Female        25-29  Hispanic       No   \n",
       "319794           0.0          No  Female  80 or older  Hispanic       No   \n",
       "\n",
       "       PhysicalActivity  GenHealth  SleepTime Asthma KidneyDisease SkinCancer  \n",
       "0                   Yes  Very good        5.0    Yes            No        Yes  \n",
       "1                   Yes  Very good        7.0     No            No         No  \n",
       "2                   Yes       Fair        8.0    Yes            No         No  \n",
       "3                    No       Good        6.0     No            No        Yes  \n",
       "4                   Yes  Very good        8.0     No            No         No  \n",
       "...                 ...        ...        ...    ...           ...        ...  \n",
       "319790               No       Fair        6.0    Yes            No         No  \n",
       "319791              Yes  Very good        5.0    Yes            No         No  \n",
       "319792              Yes       Good        6.0     No            No         No  \n",
       "319793               No       Good       12.0     No            No         No  \n",
       "319794              Yes       Good        8.0     No            No         No  \n",
       "\n",
       "[319795 rows x 18 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.read_csv(\"C:/Users/dongjae/[3-2_Class_Data_mining]/heart_2020_cleaned.csv\", encoding='cp949')\n",
    "display(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ab98cf1c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "6d42b948",
   "metadata": {},
   "source": [
    "### 데이터 탐색"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0921f7d6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 319795 entries, 0 to 319794\n",
      "Data columns (total 18 columns):\n",
      " #   Column            Non-Null Count   Dtype  \n",
      "---  ------            --------------   -----  \n",
      " 0   HeartDisease      319795 non-null  object \n",
      " 1   BMI               319795 non-null  float64\n",
      " 2   Smoking           319795 non-null  object \n",
      " 3   AlcoholDrinking   319795 non-null  object \n",
      " 4   Stroke            319795 non-null  object \n",
      " 5   PhysicalHealth    319795 non-null  float64\n",
      " 6   MentalHealth      319795 non-null  float64\n",
      " 7   DiffWalking       319795 non-null  object \n",
      " 8   Sex               319795 non-null  object \n",
      " 9   AgeCategory       319795 non-null  object \n",
      " 10  Race              319795 non-null  object \n",
      " 11  Diabetic          319795 non-null  object \n",
      " 12  PhysicalActivity  319795 non-null  object \n",
      " 13  GenHealth         319795 non-null  object \n",
      " 14  SleepTime         319795 non-null  float64\n",
      " 15  Asthma            319795 non-null  object \n",
      " 16  KidneyDisease     319795 non-null  object \n",
      " 17  SkinCancer        319795 non-null  object \n",
      "dtypes: float64(4), object(14)\n",
      "memory usage: 43.9+ MB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "None"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(df.info())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "90e16bc6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BMI의 범위는 12.02 - 94.85 이다.\n",
      "PhysicalHealth의 범위는 0.0 - 30.0 이다.\n",
      "MentalHealth의 범위는 0.0 - 30.0 이다.\n",
      "SleepTime의 범위는 1.0 - 24.0 이다.\n",
      "HeartDisease: ['No' 'Yes']\n",
      "Smoking: ['Yes' 'No']\n",
      "AlcoholDrinking: ['No' 'Yes']\n",
      "Stroke: ['No' 'Yes']\n",
      "DiffWalking: ['No' 'Yes']\n",
      "Sex: ['Female' 'Male']\n",
      "AgeCategory: ['55-59' '80 or older' '65-69' '75-79' '40-44' '70-74' '60-64' '50-54'\n",
      " '45-49' '18-24' '35-39' '30-34' '25-29']\n",
      "Race: ['White' 'Black' 'Asian' 'American Indian/Alaskan Native' 'Other'\n",
      " 'Hispanic']\n",
      "Diabetic: ['Yes' 'No' 'No, borderline diabetes' 'Yes (during pregnancy)']\n",
      "PhysicalActivity: ['Yes' 'No']\n",
      "GenHealth: ['Very good' 'Fair' 'Good' 'Poor' 'Excellent']\n",
      "Asthma: ['Yes' 'No']\n",
      "KidneyDisease: ['No' 'Yes']\n",
      "SkinCancer: ['Yes' 'No']\n"
     ]
    }
   ],
   "source": [
    "# numeric(4)\n",
    "numeric_col = ['BMI','PhysicalHealth','MentalHealth','SleepTime']\n",
    "for i in numeric_col:\n",
    "    print('{}의 범위는 {} - {} 이다.'.format(i, min(df[i]), max(df[i]),))\n",
    "    \n",
    "# categorical(14)\n",
    "categorical_col = ['HeartDisease','Smoking','AlcoholDrinking','Stroke','DiffWalking', 'Sex',\n",
    "                   'AgeCategory','Race','Diabetic','PhysicalActivity','GenHealth',\n",
    "                   'Asthma','KidneyDisease','SkinCancer']\n",
    "for i in categorical_col:\n",
    "    print('{}: {}'.format(i, df[i].unique()))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72c3c64e",
   "metadata": {},
   "source": [
    "#### 종속변수와 독립변수 확인"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "17aacbba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "종속변수의 분포는 No: 91.44%, Yes: 8.56% 으로 불균형\n"
     ]
    }
   ],
   "source": [
    "plt.figure(figsize=(10,8))\n",
    "df['HeartDisease'].value_counts().plot(kind='bar')\n",
    "plt.title('종속변수의 분포')\n",
    "# plt.savefig('종속변수의 분포.png')\n",
    "plt.show()\n",
    "rate_yes = df['HeartDisease'].value_counts()['Yes']/df['HeartDisease'].value_counts().sum()\n",
    "rate_no = 1 - rate_yes\n",
    "print('종속변수의 분포는 No: {:.2%}, Yes: {:.2%} 으로 불균형'.format(rate_no,rate_yes))\n",
    "# imbalanced data의 경우 oversampling통해 해결할 예정"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "041aeb66",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "범주형 변수들\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1440 with 15 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "수치형 변수들\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1728x864 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 범주형 변수들\n",
    "print('범주형 변수들')\n",
    "f, ax = plt.subplots(3,5,figsize=(30,20))\n",
    "for n, i  in enumerate(categorical_col):\n",
    "    sns.barplot(x=df.loc[:,i].value_counts().index,y=df.loc[:,i].value_counts().values, ax= ax[n//5,n%5])\n",
    "    ax[n//5,n%5].set_title(i)\n",
    "    if i =='AgeCategory':\n",
    "        ax[n//5,n%5].tick_params(labelrotation=45)\n",
    "# plt.savefig('범주형 변수들.png')\n",
    "plt.show()\n",
    "# 수치형 변수들\n",
    "print('수치형 변수들')\n",
    "f, ax = plt.subplots(2,4, figsize=(24,12))\n",
    "for n, i in enumerate(numeric_col):\n",
    "    ax[0,n%4].set_title(i)\n",
    "    sns.boxplot(data = df.loc[:,i], ax = ax[0,n%4])\n",
    "for n, i in enumerate(numeric_col):\n",
    "    sns.kdeplot(df.loc[:,i], ax = ax[1,n%4])\n",
    "# plt.savefig('수치형 변수들.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "22ae345f",
   "metadata": {},
   "source": [
    "#### GenHealth columns 관계"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3f26d86f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# GenHealth \n",
    "plt.figure(figsize=(10,8))\n",
    "df['GenHealth'].value_counts().plot(kind='bar').tick_params(labelrotation=45)\n",
    "plt.title('GenHealth의 분포')\n",
    "# plt.savefig('GenHealth의 분포.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "22814073",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# GenHealth와 SleepTime의 관계\n",
    "plt.figure(figsize=(10,8))\n",
    "df.groupby(['GenHealth'])['SleepTime'].mean().sort_values(ascending=False).plot()\n",
    "plt.title('GenHealth와 SleepTime의 관계')\n",
    "# plt.savefig('GenHealth와 SleepTime의 관계.png')\n",
    "plt.show()\n",
    "# 자신이 생각하는 몸상태와 평균수면시간과 반비례하는 경향을 띄었다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "09c02c6c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# GenHealth과 BMI의 관계\n",
    "plt.figure(figsize=(10,8))\n",
    "df.groupby(['GenHealth'])['BMI'].mean().reindex(['Excellent','Very good','Good','Fair','Poor']).plot()\n",
    "plt.title('GenHealth와 BMI의 관계')\n",
    "# plt.savefig('GenHealth와 BMI의 관계.png')\n",
    "plt.show()\n",
    "# 자신이 생각하는 몸상태와 MBI와 비례하는 경향을 띄었다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1807ce00",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "62301498",
   "metadata": {},
   "source": [
    "### 모델링"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e03dba3",
   "metadata": {},
   "source": [
    "#### 전처리 - 범주형 데이터 라벨 인코딩"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "edf03d1c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import OrdinalEncoder\n",
    "\n",
    "oe = OrdinalEncoder()\n",
    "\n",
    "df[categorical_col] = oe.fit_transform(df[categorical_col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "7482c43e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BMI의 범위는 12.02 - 94.85 이다.\n",
      "PhysicalHealth의 범위는 0.0 - 30.0 이다.\n",
      "MentalHealth의 범위는 0.0 - 30.0 이다.\n",
      "SleepTime의 범위는 1.0 - 24.0 이다.\n",
      "HeartDisease: [0. 1.]\n",
      "Smoking: [1. 0.]\n",
      "AlcoholDrinking: [0. 1.]\n",
      "Stroke: [0. 1.]\n",
      "DiffWalking: [0. 1.]\n",
      "Sex: [0. 1.]\n",
      "AgeCategory: [ 7. 12.  9. 11.  4. 10.  8.  6.  5.  0.  3.  2.  1.]\n",
      "Race: [5. 2. 1. 0. 4. 3.]\n",
      "Diabetic: [2. 0. 1. 3.]\n",
      "PhysicalActivity: [1. 0.]\n",
      "GenHealth: [4. 1. 2. 3. 0.]\n",
      "Asthma: [1. 0.]\n",
      "KidneyDisease: [0. 1.]\n",
      "SkinCancer: [1. 0.]\n"
     ]
    }
   ],
   "source": [
    "# numeric: 독립변수(4)\n",
    "numeric_col = ['BMI','PhysicalHealth','MentalHealth','SleepTime']\n",
    "for i in numeric_col:\n",
    "    print('{}의 범위는 {} - {} 이다.'.format(i, min(df[i]), max(df[i]),))\n",
    "    \n",
    "# categorical: 독립변수(13), 종속변수(1)\n",
    "categorical_col = ['HeartDisease','Smoking','AlcoholDrinking','Stroke','DiffWalking',\n",
    "       'Sex','AgeCategory','Race','Diabetic','PhysicalActivity',\n",
    "       'GenHealth','Asthma','KidneyDisease','SkinCancer']\n",
    "for i in categorical_col:\n",
    "    print('{}: {}'.format(i, df[i].unique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bdbdb7fa",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "b28ceaa3",
   "metadata": {},
   "source": [
    "#### 준비 - 스케일링, 데이터셋 나누기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "95dac572",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\dongjae\\AppData\\Local\\Temp\\ipykernel_20028\\818629280.py:8: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  test.drop('HeartDisease', axis=1, inplace=True)\n"
     ]
    }
   ],
   "source": [
    "# 데이터 비율 6:2:2\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "\n",
    "train = df.iloc[:256000,:] # 학습/테스트 = 8:2\n",
    "test = df.iloc[256000:,:]\n",
    "test_y = df.iloc[256000:,0]\n",
    "test.drop('HeartDisease', axis=1, inplace=True)\n",
    "y = train.pop('HeartDisease')\n",
    "\n",
    "rc = RobustScaler()\n",
    "train = rc.fit_transform(train)\n",
    "train_x, valid_x, train_y, valid_y = train_test_split(train, y, test_size=0.25) # 학습/검증 = 3:1\n",
    "test = rc.transform(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f5db92a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(192000, 17) (64000, 17) (63795, 17)\n"
     ]
    }
   ],
   "source": [
    "print(train_x.shape, valid_x.shape, test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52b7a15e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "3514820c",
   "metadata": {},
   "source": [
    "#### 준비 - 오버샘플링"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "49ab1580",
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE\n",
    "sm = SMOTE(random_state=seed)\n",
    "resampled_x, resampled_y = sm.fit_resample(train_x, train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "85404f26",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(192000, 17)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "1327fddd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(350934, 17)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resampled_x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "d8908831",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 다운샘플링의 결과 정확도, 재현율  각 70%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "67280d66",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "1aac97ee",
   "metadata": {},
   "source": [
    "#### 하이퍼 파라미터 튜닝\n",
    "- 해당 코드 출처: github.com/optuna"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "c312d168",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m[I 2022-11-12 15:23:12,387]\u001b[0m A new study created in memory with name: no-name-347b0167-9208-469e-bbe9-9ad172590e1d\u001b[0m\n",
      "\n",
      "  0%|                                                                                            | 0/7 [00:00<?, ?it/s]\u001b[A\n",
      "feature_fraction, val_score: inf:   0%|                                                          | 0/7 [00:00<?, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.019376 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.179456\tvalid_1's binary_logloss: 0.249898\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.244539:   0%|                                                     | 0/7 [00:03<?, ?it/s]\u001b[A\n",
      "feature_fraction, val_score: 0.244539:  14%|######4                                      | 1/7 [00:03<00:18,  3.16s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:15,561]\u001b[0m Trial 0 finished with value: 0.2445386291038646 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.2445386291038646.\u001b[0m\n",
      "\n",
      "feature_fraction, val_score: 0.244539:  14%|######4                                      | 1/7 [00:03<00:18,  3.16s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.165013\tvalid_1's binary_logloss: 0.244539\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.165013\tvalid_1's binary_logloss: 0.244539\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.019032 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.18031\tvalid_1's binary_logloss: 0.250526\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.244232:  14%|######4                                      | 1/7 [00:06<00:18,  3.16s/it]\u001b[A\n",
      "feature_fraction, val_score: 0.244232:  29%|############8                                | 2/7 [00:06<00:16,  3.34s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:19,020]\u001b[0m Trial 1 finished with value: 0.24423163336642237 and parameters: {'feature_fraction': 0.5}. Best is trial 1 with value: 0.24423163336642237.\u001b[0m\n",
      "\n",
      "feature_fraction, val_score: 0.244232:  29%|############8                                | 2/7 [00:06<00:16,  3.34s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.165548\tvalid_1's binary_logloss: 0.244232\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.165548\tvalid_1's binary_logloss: 0.244232\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018496 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.180101\tvalid_1's binary_logloss: 0.250749\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.244232:  29%|############8                                | 2/7 [00:09<00:16,  3.34s/it]\u001b[A\n",
      "feature_fraction, val_score: 0.244232:  43%|###################2                         | 3/7 [00:09<00:13,  3.31s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:22,287]\u001b[0m Trial 2 finished with value: 0.2449779720094361 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 1 with value: 0.24423163336642237.\u001b[0m\n",
      "\n",
      "feature_fraction, val_score: 0.244232:  43%|###################2                         | 3/7 [00:09<00:13,  3.31s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.165391\tvalid_1's binary_logloss: 0.244978\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.165391\tvalid_1's binary_logloss: 0.244978\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016840 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.180424\tvalid_1's binary_logloss: 0.250257\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.243658:  43%|###################2                         | 3/7 [00:15<00:13,  3.31s/it]\u001b[A\n",
      "feature_fraction, val_score: 0.243658:  57%|#########################7                   | 4/7 [00:15<00:12,  4.14s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:27,714]\u001b[0m Trial 3 finished with value: 0.24365767799633972 and parameters: {'feature_fraction': 0.4}. Best is trial 3 with value: 0.24365767799633972.\u001b[0m\n",
      "\n",
      "feature_fraction, val_score: 0.243658:  57%|#########################7                   | 4/7 [00:15<00:12,  4.14s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.165478\tvalid_1's binary_logloss: 0.243658\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.165478\tvalid_1's binary_logloss: 0.243658\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.023583 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.179583\tvalid_1's binary_logloss: 0.250364\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.243658:  57%|#########################7                   | 4/7 [00:19<00:12,  4.14s/it]\u001b[A\n",
      "feature_fraction, val_score: 0.243658:  71%|################################1            | 5/7 [00:19<00:08,  4.02s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:31,522]\u001b[0m Trial 4 finished with value: 0.24478792777867991 and parameters: {'feature_fraction': 0.8}. Best is trial 3 with value: 0.24365767799633972.\u001b[0m\n",
      "\n",
      "feature_fraction, val_score: 0.243658:  71%|################################1            | 5/7 [00:19<00:08,  4.02s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.165267\tvalid_1's binary_logloss: 0.244788\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.165267\tvalid_1's binary_logloss: 0.244788\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.022179 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.180225\tvalid_1's binary_logloss: 0.250853\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.243658:  71%|################################1            | 5/7 [00:22<00:08,  4.02s/it]\u001b[A\n",
      "feature_fraction, val_score: 0.243658:  86%|######################################5      | 6/7 [00:22<00:03,  3.94s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:35,296]\u001b[0m Trial 5 finished with value: 0.24504208526071186 and parameters: {'feature_fraction': 1.0}. Best is trial 3 with value: 0.24365767799633972.\u001b[0m\n",
      "\n",
      "feature_fraction, val_score: 0.243658:  86%|######################################5      | 6/7 [00:22<00:03,  3.94s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.164884\tvalid_1's binary_logloss: 0.245042\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.164884\tvalid_1's binary_logloss: 0.245042\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.017061 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.179835\tvalid_1's binary_logloss: 0.250162\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction, val_score: 0.243658:  86%|######################################5      | 6/7 [00:27<00:03,  3.94s/it]\u001b[A\n",
      "feature_fraction, val_score: 0.243658: 100%|#############################################| 7/7 [00:27<00:00,  4.14s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:39,837]\u001b[0m Trial 6 finished with value: 0.24437622833094425 and parameters: {'feature_fraction': 0.6}. Best is trial 3 with value: 0.24365767799633972.\u001b[0m\n",
      "feature_fraction, val_score: 0.243658: 100%|#############################################| 7/7 [00:27<00:00,  3.92s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.16551\tvalid_1's binary_logloss: 0.244376\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.16551\tvalid_1's binary_logloss: 0.244376\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                           | 0/20 [00:00<?, ?it/s]\u001b[A\n",
      "num_leaves, val_score: 0.243658:   0%|                                                          | 0/20 [00:00<?, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018510 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.447816\tvalid_1's binary_logloss: 0.473612\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.243658:   0%|                                                          | 0/20 [00:02<?, ?it/s]\u001b[A\n",
      "num_leaves, val_score: 0.243658:   5%|##5                                               | 1/20 [00:02<00:41,  2.21s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:42,059]\u001b[0m Trial 7 finished with value: 0.4469013554100348 and parameters: {'num_leaves': 2}. Best is trial 7 with value: 0.4469013554100348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.243658:   5%|##5                                               | 1/20 [00:02<00:41,  2.21s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.420368\tvalid_1's binary_logloss: 0.446901\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.420368\tvalid_1's binary_logloss: 0.446901\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018686 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.165645\tvalid_1's binary_logloss: 0.244628\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242363:   5%|##5                                               | 1/20 [00:07<00:41,  2.21s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242363:  10%|#####                                             | 2/20 [00:07<01:11,  3.95s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:47,233]\u001b[0m Trial 8 finished with value: 0.24236283113463397 and parameters: {'num_leaves': 90}. Best is trial 8 with value: 0.24236283113463397.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242363:  10%|#####                                             | 2/20 [00:07<01:11,  3.95s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.152521\tvalid_1's binary_logloss: 0.242363\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.152521\tvalid_1's binary_logloss: 0.242363\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014216 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169149\tvalid_1's binary_logloss: 0.24534\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  10%|#####                                             | 2/20 [00:12<01:11,  3.95s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  15%|#######5                                          | 3/20 [00:12<01:13,  4.32s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:51,989]\u001b[0m Trial 9 finished with value: 0.24217719291433348 and parameters: {'num_leaves': 67}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  15%|#######5                                          | 3/20 [00:12<01:13,  4.32s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156163\tvalid_1's binary_logloss: 0.242177\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156163\tvalid_1's binary_logloss: 0.242177\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.013820 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.165576\tvalid_1's binary_logloss: 0.244722\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  15%|#######5                                          | 3/20 [00:17<01:13,  4.32s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  20%|##########                                        | 4/20 [00:17<01:15,  4.70s/it]\u001b[A\u001b[32m[I 2022-11-12 15:23:57,278]\u001b[0m Trial 10 finished with value: 0.24226629721878432 and parameters: {'num_leaves': 91}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  20%|##########                                        | 4/20 [00:17<01:15,  4.70s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.152256\tvalid_1's binary_logloss: 0.242266\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.152256\tvalid_1's binary_logloss: 0.242266\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.023500 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.165528\tvalid_1's binary_logloss: 0.244663\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  20%|##########                                        | 4/20 [00:22<01:15,  4.70s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  25%|############5                                     | 5/20 [00:22<01:14,  4.96s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:02,690]\u001b[0m Trial 11 finished with value: 0.242207482005304 and parameters: {'num_leaves': 92}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  25%|############5                                     | 5/20 [00:22<01:14,  4.96s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.152188\tvalid_1's binary_logloss: 0.242207\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.152188\tvalid_1's binary_logloss: 0.242207\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016157 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.15124\tvalid_1's binary_logloss: 0.243182\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  25%|############5                                     | 5/20 [00:32<01:14,  4.96s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  30%|###############                                   | 6/20 [00:32<01:32,  6.62s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:12,539]\u001b[0m Trial 12 finished with value: 0.24353138609579994 and parameters: {'num_leaves': 253}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  30%|###############                                   | 6/20 [00:32<01:32,  6.62s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.133876\tvalid_1's binary_logloss: 0.243531\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.133876\tvalid_1's binary_logloss: 0.243531\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.017777 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.155399\tvalid_1's binary_logloss: 0.243205\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  30%|###############                                   | 6/20 [00:40<01:32,  6.62s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  35%|#################5                                | 7/20 [00:40<01:30,  6.98s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:20,268]\u001b[0m Trial 13 finished with value: 0.24276791984965554 and parameters: {'num_leaves': 194}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  35%|#################5                                | 7/20 [00:40<01:30,  6.98s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.139683\tvalid_1's binary_logloss: 0.242768\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.139683\tvalid_1's binary_logloss: 0.242768\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.019818 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.163629\tvalid_1's binary_logloss: 0.244188\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  35%|#################5                                | 7/20 [00:45<01:30,  6.98s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  40%|####################                              | 8/20 [00:45<01:17,  6.42s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:25,482]\u001b[0m Trial 14 finished with value: 0.24232756710846973 and parameters: {'num_leaves': 107}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  40%|####################                              | 8/20 [00:45<01:17,  6.42s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.15008\tvalid_1's binary_logloss: 0.242328\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.15008\tvalid_1's binary_logloss: 0.242328\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.020491 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.156373\tvalid_1's binary_logloss: 0.243414\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  40%|####################                              | 8/20 [00:52<01:17,  6.42s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  45%|######################5                           | 9/20 [00:52<01:13,  6.70s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:32,788]\u001b[0m Trial 15 finished with value: 0.24301710001776694 and parameters: {'num_leaves': 182}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  45%|######################5                           | 9/20 [00:52<01:13,  6.70s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.14054\tvalid_1's binary_logloss: 0.243017\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.14054\tvalid_1's binary_logloss: 0.243017\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.020136 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.159826\tvalid_1's binary_logloss: 0.243589\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  45%|######################5                           | 9/20 [01:01<01:13,  6.70s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  50%|########################5                        | 10/20 [01:01<01:11,  7.16s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:40,981]\u001b[0m Trial 16 finished with value: 0.24277147414468075 and parameters: {'num_leaves': 141}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  50%|########################5                        | 10/20 [01:01<01:11,  7.16s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.145367\tvalid_1's binary_logloss: 0.242771\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.145367\tvalid_1's binary_logloss: 0.242771\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.021353 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.179107\tvalid_1's binary_logloss: 0.24952\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  50%|########################5                        | 10/20 [01:05<01:11,  7.16s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  55%|##########################9                      | 11/20 [01:05<00:56,  6.32s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:45,406]\u001b[0m Trial 17 finished with value: 0.24309549710435657 and parameters: {'num_leaves': 34}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  55%|##########################9                      | 11/20 [01:05<00:56,  6.32s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.164346\tvalid_1's binary_logloss: 0.243095\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.164346\tvalid_1's binary_logloss: 0.243095\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016863 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.172523\tvalid_1's binary_logloss: 0.246626\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  55%|##########################9                      | 11/20 [01:09<00:56,  6.32s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  60%|#############################4                   | 12/20 [01:09<00:45,  5.69s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:49,659]\u001b[0m Trial 18 finished with value: 0.2425817588020773 and parameters: {'num_leaves': 52}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  60%|#############################4                   | 12/20 [01:09<00:45,  5.69s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.159448\tvalid_1's binary_logloss: 0.242582\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.159448\tvalid_1's binary_logloss: 0.242582\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018176 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.171759\tvalid_1's binary_logloss: 0.246322\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  60%|#############################4                   | 12/20 [01:14<00:45,  5.69s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  65%|###############################8                 | 13/20 [01:14<00:38,  5.45s/it]\u001b[A\u001b[32m[I 2022-11-12 15:24:54,538]\u001b[0m Trial 19 finished with value: 0.24239938841900574 and parameters: {'num_leaves': 56}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  65%|###############################8                 | 13/20 [01:14<00:38,  5.45s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.158572\tvalid_1's binary_logloss: 0.242399\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.158572\tvalid_1's binary_logloss: 0.242399\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.017294 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.159515\tvalid_1's binary_logloss: 0.243736\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242177:  65%|###############################8                 | 13/20 [01:21<00:38,  5.45s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242177:  70%|##################################3              | 14/20 [01:21<00:34,  5.76s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:01,033]\u001b[0m Trial 20 finished with value: 0.24267985126746844 and parameters: {'num_leaves': 145}. Best is trial 9 with value: 0.24217719291433348.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242177:  70%|##################################3              | 14/20 [01:21<00:34,  5.76s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.144733\tvalid_1's binary_logloss: 0.24268\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.144733\tvalid_1's binary_logloss: 0.24268\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016585 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242070:  70%|##################################3              | 14/20 [01:25<00:34,  5.76s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242070:  75%|####################################7            | 15/20 [01:25<00:27,  5.47s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:05,836]\u001b[0m Trial 21 finished with value: 0.2420696001863824 and parameters: {'num_leaves': 66}. Best is trial 21 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242070:  75%|####################################7            | 15/20 [01:25<00:27,  5.47s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.013963 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.242481\tvalid_1's binary_logloss: 0.291476\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242070:  75%|####################################7            | 15/20 [01:29<00:27,  5.47s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242070:  80%|#######################################2         | 16/20 [01:29<00:19,  4.86s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:09,260]\u001b[0m Trial 22 finished with value: 0.2614541465207301 and parameters: {'num_leaves': 7}. Best is trial 21 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242070:  80%|#######################################2         | 16/20 [01:29<00:19,  4.86s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.199145\tvalid_1's binary_logloss: 0.261454\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.199145\tvalid_1's binary_logloss: 0.261454\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.021388 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.171188\tvalid_1's binary_logloss: 0.246163\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242070:  80%|#######################################2         | 16/20 [01:33<00:19,  4.86s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242070:  85%|#########################################6       | 17/20 [01:33<00:14,  4.67s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:13,494]\u001b[0m Trial 23 finished with value: 0.24261139856228808 and parameters: {'num_leaves': 58}. Best is trial 21 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242070:  85%|#########################################6       | 17/20 [01:33<00:14,  4.67s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.158224\tvalid_1's binary_logloss: 0.242611\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.158224\tvalid_1's binary_logloss: 0.242611\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.022073 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.161949\tvalid_1's binary_logloss: 0.243815\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242070:  85%|#########################################6       | 17/20 [01:39<00:14,  4.67s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242070:  90%|############################################1    | 18/20 [01:39<00:09,  4.96s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:19,119]\u001b[0m Trial 24 finished with value: 0.2422647180063992 and parameters: {'num_leaves': 120}. Best is trial 21 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242070:  90%|############################################1    | 18/20 [01:39<00:09,  4.96s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.148266\tvalid_1's binary_logloss: 0.242265\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.148266\tvalid_1's binary_logloss: 0.242265\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.024489 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.180722\tvalid_1's binary_logloss: 0.250328\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242070:  90%|############################################1    | 18/20 [01:43<00:09,  4.96s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242070:  95%|##############################################5  | 19/20 [01:43<00:04,  4.77s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:23,457]\u001b[0m Trial 25 finished with value: 0.24390320338811622 and parameters: {'num_leaves': 30}. Best is trial 21 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "num_leaves, val_score: 0.242070:  95%|##############################################5  | 19/20 [01:43<00:04,  4.77s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.165809\tvalid_1's binary_logloss: 0.243903\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.165809\tvalid_1's binary_logloss: 0.243903\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014050 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.167955\tvalid_1's binary_logloss: 0.244966\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "num_leaves, val_score: 0.242070:  95%|##############################################5  | 19/20 [01:48<00:04,  4.77s/it]\u001b[A\n",
      "num_leaves, val_score: 0.242070: 100%|#################################################| 20/20 [01:48<00:00,  4.78s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:28,257]\u001b[0m Trial 26 finished with value: 0.2420707091577252 and parameters: {'num_leaves': 75}. Best is trial 21 with value: 0.2420696001863824.\u001b[0m\n",
      "num_leaves, val_score: 0.242070: 100%|#################################################| 20/20 [01:48<00:00,  5.42s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.154882\tvalid_1's binary_logloss: 0.242071\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.154882\tvalid_1's binary_logloss: 0.242071\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                           | 0/10 [00:00<?, ?it/s]\u001b[A\n",
      "bagging, val_score: 0.242070:   0%|                                                             | 0/10 [00:00<?, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.019602 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169934\tvalid_1's binary_logloss: 0.245854\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:   0%|                                                             | 0/10 [00:06<?, ?it/s]\u001b[A\n",
      "bagging, val_score: 0.242070:  10%|#####3                                               | 1/10 [00:06<00:55,  6.12s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:34,389]\u001b[0m Trial 27 finished with value: 0.24248410265450654 and parameters: {'bagging_fraction': 0.8906902430249992, 'bagging_freq': 5}. Best is trial 27 with value: 0.24248410265450654.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  10%|#####3                                               | 1/10 [00:06<00:55,  6.12s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156532\tvalid_1's binary_logloss: 0.242484\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156532\tvalid_1's binary_logloss: 0.242484\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.020253 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170315\tvalid_1's binary_logloss: 0.24622\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  10%|#####3                                               | 1/10 [00:11<00:55,  6.12s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  20%|##########6                                          | 2/10 [00:11<00:46,  5.86s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:40,058]\u001b[0m Trial 28 finished with value: 0.24263951986643964 and parameters: {'bagging_fraction': 0.5826908973695131, 'bagging_freq': 5}. Best is trial 27 with value: 0.24248410265450654.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  20%|##########6                                          | 2/10 [00:11<00:46,  5.86s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156727\tvalid_1's binary_logloss: 0.24264\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156727\tvalid_1's binary_logloss: 0.24264\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016235 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170245\tvalid_1's binary_logloss: 0.246164\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  20%|##########6                                          | 2/10 [00:16<00:46,  5.86s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  30%|###############9                                     | 3/10 [00:16<00:37,  5.40s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:44,918]\u001b[0m Trial 29 finished with value: 0.24241387677151746 and parameters: {'bagging_fraction': 0.6059134741283667, 'bagging_freq': 4}. Best is trial 29 with value: 0.24241387677151746.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  30%|###############9                                     | 3/10 [00:16<00:37,  5.40s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156508\tvalid_1's binary_logloss: 0.242414\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156508\tvalid_1's binary_logloss: 0.242414\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016017 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170009\tvalid_1's binary_logloss: 0.245938\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  30%|###############9                                     | 3/10 [00:21<00:37,  5.40s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  40%|#####################2                               | 4/10 [00:21<00:31,  5.19s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:49,793]\u001b[0m Trial 30 finished with value: 0.24246801166990228 and parameters: {'bagging_fraction': 0.7492688445574169, 'bagging_freq': 7}. Best is trial 29 with value: 0.24241387677151746.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  40%|#####################2                               | 4/10 [00:21<00:31,  5.19s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156682\tvalid_1's binary_logloss: 0.242468\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156682\tvalid_1's binary_logloss: 0.242468\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014030 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.16964\tvalid_1's binary_logloss: 0.245719\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  40%|#####################2                               | 4/10 [00:27<00:31,  5.19s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  50%|##########################5                          | 5/10 [00:27<00:26,  5.38s/it]\u001b[A\u001b[32m[I 2022-11-12 15:25:55,494]\u001b[0m Trial 31 finished with value: 0.2423789317342574 and parameters: {'bagging_fraction': 0.9683276073306413, 'bagging_freq': 2}. Best is trial 31 with value: 0.2423789317342574.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  50%|##########################5                          | 5/10 [00:27<00:26,  5.38s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156643\tvalid_1's binary_logloss: 0.242379\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156643\tvalid_1's binary_logloss: 0.242379\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018064 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170234\tvalid_1's binary_logloss: 0.246143\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  50%|##########################5                          | 5/10 [00:32<00:26,  5.38s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  60%|###############################8                     | 6/10 [00:32<00:20,  5.20s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:00,352]\u001b[0m Trial 32 finished with value: 0.24253903085035877 and parameters: {'bagging_fraction': 0.8032062335125877, 'bagging_freq': 5}. Best is trial 31 with value: 0.2423789317342574.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  60%|###############################8                     | 6/10 [00:32<00:20,  5.20s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156573\tvalid_1's binary_logloss: 0.242539\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156573\tvalid_1's binary_logloss: 0.242539\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016272 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169557\tvalid_1's binary_logloss: 0.245855\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  60%|###############################8                     | 6/10 [00:37<00:20,  5.20s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  70%|#####################################                | 7/10 [00:38<00:16,  5.43s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:06,272]\u001b[0m Trial 33 finished with value: 0.24228436072309747 and parameters: {'bagging_fraction': 0.8461344832088241, 'bagging_freq': 3}. Best is trial 33 with value: 0.24228436072309747.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  70%|#####################################                | 7/10 [00:38<00:16,  5.43s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156495\tvalid_1's binary_logloss: 0.242284\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156495\tvalid_1's binary_logloss: 0.242284\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.019866 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170064\tvalid_1's binary_logloss: 0.246131\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  70%|#####################################                | 7/10 [00:42<00:16,  5.43s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  80%|##########################################4          | 8/10 [00:42<00:10,  5.25s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:11,128]\u001b[0m Trial 34 finished with value: 0.24269382356674993 and parameters: {'bagging_fraction': 0.7192786692654867, 'bagging_freq': 3}. Best is trial 33 with value: 0.24228436072309747.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  80%|##########################################4          | 8/10 [00:42<00:10,  5.25s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156377\tvalid_1's binary_logloss: 0.242694\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156377\tvalid_1's binary_logloss: 0.242694\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.013422 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.16964\tvalid_1's binary_logloss: 0.245661\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  80%|##########################################4          | 8/10 [00:48<00:10,  5.25s/it]\u001b[A\n",
      "bagging, val_score: 0.242070:  90%|###############################################7     | 9/10 [00:48<00:05,  5.47s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:17,085]\u001b[0m Trial 35 finished with value: 0.2423387120613582 and parameters: {'bagging_fraction': 0.9743188621821203, 'bagging_freq': 6}. Best is trial 33 with value: 0.24228436072309747.\u001b[0m\n",
      "\n",
      "bagging, val_score: 0.242070:  90%|###############################################7     | 9/10 [00:48<00:05,  5.47s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156654\tvalid_1's binary_logloss: 0.242339\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156654\tvalid_1's binary_logloss: 0.242339\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016472 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169848\tvalid_1's binary_logloss: 0.245976\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "bagging, val_score: 0.242070:  90%|###############################################7     | 9/10 [00:53<00:05,  5.47s/it]\u001b[A\n",
      "bagging, val_score: 0.242070: 100%|####################################################| 10/10 [00:53<00:00,  5.31s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:22,047]\u001b[0m Trial 36 finished with value: 0.24242840444931313 and parameters: {'bagging_fraction': 0.7983754459400675, 'bagging_freq': 7}. Best is trial 33 with value: 0.24228436072309747.\u001b[0m\n",
      "bagging, val_score: 0.242070: 100%|####################################################| 10/10 [00:53<00:00,  5.38s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156666\tvalid_1's binary_logloss: 0.242428\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156666\tvalid_1's binary_logloss: 0.242428\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                            | 0/3 [00:00<?, ?it/s]\u001b[A\n",
      "feature_fraction_stage2, val_score: 0.242070:   0%|                                              | 0/3 [00:00<?, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018398 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction_stage2, val_score: 0.242070:   0%|                                              | 0/3 [00:04<?, ?it/s]\u001b[A\n",
      "feature_fraction_stage2, val_score: 0.242070:  33%|############6                         | 1/3 [00:04<00:08,  4.06s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:26,117]\u001b[0m Trial 37 finished with value: 0.2420696001863824 and parameters: {'feature_fraction': 0.41600000000000004}. Best is trial 37 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "feature_fraction_stage2, val_score: 0.242070:  33%|############6                         | 1/3 [00:04<00:08,  4.06s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.015759 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.168613\tvalid_1's binary_logloss: 0.245437\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction_stage2, val_score: 0.242070:  33%|############6                         | 1/3 [00:08<00:08,  4.06s/it]\u001b[A\n",
      "feature_fraction_stage2, val_score: 0.242070:  67%|#########################3            | 2/3 [00:08<00:04,  4.13s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:30,300]\u001b[0m Trial 38 finished with value: 0.2429165710150791 and parameters: {'feature_fraction': 0.48000000000000004}. Best is trial 37 with value: 0.2420696001863824.\u001b[0m\n",
      "\n",
      "feature_fraction_stage2, val_score: 0.242070:  67%|#########################3            | 2/3 [00:08<00:04,  4.13s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156398\tvalid_1's binary_logloss: 0.242917\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156398\tvalid_1's binary_logloss: 0.242917\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016752 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.168613\tvalid_1's binary_logloss: 0.245437\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "feature_fraction_stage2, val_score: 0.242070:  67%|#########################3            | 2/3 [00:12<00:04,  4.13s/it]\u001b[A\n",
      "feature_fraction_stage2, val_score: 0.242070: 100%|######################################| 3/3 [00:12<00:00,  4.11s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:34,377]\u001b[0m Trial 39 finished with value: 0.2429165710150791 and parameters: {'feature_fraction': 0.44800000000000006}. Best is trial 37 with value: 0.2420696001863824.\u001b[0m\n",
      "feature_fraction_stage2, val_score: 0.242070: 100%|######################################| 3/3 [00:12<00:00,  4.11s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156398\tvalid_1's binary_logloss: 0.242917\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156398\tvalid_1's binary_logloss: 0.242917\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                           | 0/20 [00:00<?, ?it/s]\u001b[A\n",
      "regularization_factors, val_score: 0.242070:   0%|                                              | 0/20 [00:00<?, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016442 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:   0%|                                              | 0/20 [00:04<?, ?it/s]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:   5%|#9                                    | 1/20 [00:04<01:22,  4.36s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:38,747]\u001b[0m Trial 40 finished with value: 0.24206656470870536 and parameters: {'lambda_l1': 6.236668171116407e-08, 'lambda_l2': 6.487825418792685e-05}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:   5%|#9                                    | 1/20 [00:04<01:22,  4.36s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.242067\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.242067\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016865 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:   5%|#9                                    | 1/20 [00:08<01:22,  4.36s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:  10%|###8                                  | 2/20 [00:08<01:20,  4.47s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:43,286]\u001b[0m Trial 41 finished with value: 0.24206657697066766 and parameters: {'lambda_l1': 1.1923774006295502e-07, 'lambda_l2': 2.806153445809285e-05}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:  10%|###8                                  | 2/20 [00:08<01:20,  4.47s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.242067\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.242067\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014122 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.17173\tvalid_1's binary_logloss: 0.246141\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:  10%|###8                                  | 2/20 [00:13<01:20,  4.47s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:  15%|#####7                                | 3/20 [00:13<01:15,  4.45s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:47,723]\u001b[0m Trial 42 finished with value: 0.24225506767900584 and parameters: {'lambda_l1': 7.529392267583384, 'lambda_l2': 2.2753487283381475e-08}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:  15%|#####7                                | 3/20 [00:13<01:15,  4.45s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[200]\tvalid_0's binary_logloss: 0.159524\tvalid_1's binary_logloss: 0.242255\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.159524\tvalid_1's binary_logloss: 0.242255\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016360 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:  15%|#####7                                | 3/20 [00:18<01:15,  4.45s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:  20%|#######6                              | 4/20 [00:18<01:17,  4.82s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:53,114]\u001b[0m Trial 43 finished with value: 0.24206960017341897 and parameters: {'lambda_l1': 2.428119738168349e-07, 'lambda_l2': 1.253197419654777e-08}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:  20%|#######6                              | 4/20 [00:18<01:17,  4.82s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018698 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169534\tvalid_1's binary_logloss: 0.245447\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:  20%|#######6                              | 4/20 [00:25<01:17,  4.82s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:  25%|#########5                            | 5/20 [00:25<01:23,  5.54s/it]\u001b[A\u001b[32m[I 2022-11-12 15:26:59,917]\u001b[0m Trial 44 finished with value: 0.24215898078669995 and parameters: {'lambda_l1': 3.469058708101267e-06, 'lambda_l2': 0.0029530722130802684}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:  25%|#########5                            | 5/20 [00:25<01:23,  5.54s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156524\tvalid_1's binary_logloss: 0.242159\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156524\tvalid_1's binary_logloss: 0.242159\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018701 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170463\tvalid_1's binary_logloss: 0.245686\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:  25%|#########5                            | 5/20 [00:31<01:23,  5.54s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:  30%|###########4                          | 6/20 [00:31<01:20,  5.73s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:06,010]\u001b[0m Trial 45 finished with value: 0.24208258446927536 and parameters: {'lambda_l1': 3.294714198743081, 'lambda_l2': 0.9572274845839805}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:  30%|###########4                          | 6/20 [00:31<01:20,  5.73s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[200]\tvalid_0's binary_logloss: 0.158108\tvalid_1's binary_logloss: 0.242083\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.158108\tvalid_1's binary_logloss: 0.242083\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.017666 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169495\tvalid_1's binary_logloss: 0.245504\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.242067:  30%|###########4                          | 6/20 [00:37<01:20,  5.73s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.242067:  35%|#############3                        | 7/20 [00:37<01:14,  5.71s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:11,680]\u001b[0m Trial 46 finished with value: 0.24218402709501052 and parameters: {'lambda_l1': 0.020820326578973394, 'lambda_l2': 6.838249073412626e-08}. Best is trial 40 with value: 0.24206656470870536.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.242067:  35%|#############3                        | 7/20 [00:37<01:14,  5.71s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156711\tvalid_1's binary_logloss: 0.242184\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156711\tvalid_1's binary_logloss: 0.242184\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.021907 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169726\tvalid_1's binary_logloss: 0.2455\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  35%|#############3                        | 7/20 [00:43<01:14,  5.71s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  40%|###############2                      | 8/20 [00:43<01:10,  5.89s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:17,952]\u001b[0m Trial 47 finished with value: 0.24183433913467367 and parameters: {'lambda_l1': 1.2884203097000726, 'lambda_l2': 0.00010898758830735766}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  40%|###############2                      | 8/20 [00:43<01:10,  5.89s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156929\tvalid_1's binary_logloss: 0.241834\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156929\tvalid_1's binary_logloss: 0.241834\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014040 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169548\tvalid_1's binary_logloss: 0.24553\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  40%|###############2                      | 8/20 [00:48<01:10,  5.89s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  45%|#################1                    | 9/20 [00:48<01:00,  5.50s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:22,609]\u001b[0m Trial 48 finished with value: 0.24190030274587326 and parameters: {'lambda_l1': 0.1624408715267981, 'lambda_l2': 5.033490794455589e-07}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  45%|#################1                    | 9/20 [00:48<01:00,  5.50s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156547\tvalid_1's binary_logloss: 0.2419\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156547\tvalid_1's binary_logloss: 0.2419\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018087 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  45%|#################1                    | 9/20 [00:53<01:00,  5.50s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  50%|##################5                  | 10/20 [00:53<00:54,  5.43s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:27,868]\u001b[0m Trial 49 finished with value: 0.24206959969028843 and parameters: {'lambda_l1': 1.81380482651296e-07, 'lambda_l2': 1.4833599479910268e-06}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  50%|##################5                  | 10/20 [00:53<00:54,  5.43s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156493\tvalid_1's binary_logloss: 0.24207\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016951 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169435\tvalid_1's binary_logloss: 0.245425\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  50%|##################5                  | 10/20 [00:57<00:54,  5.43s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  55%|####################3                | 11/20 [00:57<00:46,  5.14s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:32,348]\u001b[0m Trial 50 finished with value: 0.24207387661436908 and parameters: {'lambda_l1': 0.0003285559549004932, 'lambda_l2': 0.046332349962920585}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  55%|####################3                | 11/20 [00:57<00:46,  5.14s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156724\tvalid_1's binary_logloss: 0.242074\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156724\tvalid_1's binary_logloss: 0.242074\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.021602 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169559\tvalid_1's binary_logloss: 0.245519\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  55%|####################3                | 11/20 [01:02<00:46,  5.14s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  60%|######################2              | 12/20 [01:02<00:39,  4.90s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:36,710]\u001b[0m Trial 51 finished with value: 0.24212837446626143 and parameters: {'lambda_l1': 0.047724991216974585, 'lambda_l2': 2.9936946495930516e-06}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  60%|######################2              | 12/20 [01:02<00:39,  4.90s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156683\tvalid_1's binary_logloss: 0.242128\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156683\tvalid_1's binary_logloss: 0.242128\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016634 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169401\tvalid_1's binary_logloss: 0.245436\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  60%|######################2              | 12/20 [01:06<00:39,  4.90s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  65%|########################             | 13/20 [01:06<00:33,  4.73s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:41,047]\u001b[0m Trial 52 finished with value: 0.2420087819309935 and parameters: {'lambda_l1': 0.10821196669437584, 'lambda_l2': 0.0017489072779733935}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  65%|########################             | 13/20 [01:06<00:33,  4.73s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156447\tvalid_1's binary_logloss: 0.242009\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156447\tvalid_1's binary_logloss: 0.242009\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014563 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169533\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  65%|########################             | 13/20 [01:11<00:33,  4.73s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  70%|#########################9           | 14/20 [01:11<00:27,  4.63s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:45,453]\u001b[0m Trial 53 finished with value: 0.24212757905870982 and parameters: {'lambda_l1': 0.0018984491806957671, 'lambda_l2': 1.2799325620964944e-06}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  70%|#########################9           | 14/20 [01:11<00:27,  4.63s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156593\tvalid_1's binary_logloss: 0.242128\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156593\tvalid_1's binary_logloss: 0.242128\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.017148 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169653\tvalid_1's binary_logloss: 0.245655\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  70%|#########################9           | 14/20 [01:15<00:27,  4.63s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  75%|###########################7         | 15/20 [01:15<00:23,  4.63s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:50,074]\u001b[0m Trial 54 finished with value: 0.24230054292011535 and parameters: {'lambda_l1': 0.31779080530496073, 'lambda_l2': 0.0002908483196257079}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  75%|###########################7         | 15/20 [01:15<00:23,  4.63s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156679\tvalid_1's binary_logloss: 0.242301\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156679\tvalid_1's binary_logloss: 0.242301\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018994 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169532\tvalid_1's binary_logloss: 0.245446\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241834:  75%|###########################7         | 15/20 [01:20<00:23,  4.63s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241834:  80%|#############################6       | 16/20 [01:20<00:18,  4.64s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:54,753]\u001b[0m Trial 55 finished with value: 0.24205139919582047 and parameters: {'lambda_l1': 1.5919461763128976e-05, 'lambda_l2': 1.1746210375791305e-05}. Best is trial 47 with value: 0.24183433913467367.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241834:  80%|#############################6       | 16/20 [01:20<00:18,  4.64s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.15654\tvalid_1's binary_logloss: 0.242051\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.15654\tvalid_1's binary_logloss: 0.242051\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.017775 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169635\tvalid_1's binary_logloss: 0.245405\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241717:  80%|#############################6       | 16/20 [01:24<00:18,  4.64s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241717:  85%|###############################4     | 17/20 [01:24<00:13,  4.64s/it]\u001b[A\u001b[32m[I 2022-11-12 15:27:59,381]\u001b[0m Trial 56 finished with value: 0.24171678915738617 and parameters: {'lambda_l1': 0.8290885023746439, 'lambda_l2': 1.718942996506061e-07}. Best is trial 56 with value: 0.24171678915738617.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241717:  85%|###############################4     | 17/20 [01:25<00:13,  4.64s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156765\tvalid_1's binary_logloss: 0.241717\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156765\tvalid_1's binary_logloss: 0.241717\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.019008 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169431\tvalid_1's binary_logloss: 0.245424\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241717:  85%|###############################4     | 17/20 [01:29<00:13,  4.64s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241717:  90%|#################################3   | 18/20 [01:29<00:09,  4.57s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:03,782]\u001b[0m Trial 57 finished with value: 0.24220801741924175 and parameters: {'lambda_l1': 0.0025731750353149767, 'lambda_l2': 0.03914135527498656}. Best is trial 56 with value: 0.24171678915738617.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241717:  90%|#################################3   | 18/20 [01:29<00:09,  4.57s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156558\tvalid_1's binary_logloss: 0.242208\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156558\tvalid_1's binary_logloss: 0.242208\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.018832 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169842\tvalid_1's binary_logloss: 0.245485\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241717:  90%|#################################3   | 18/20 [01:34<00:09,  4.57s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241717:  95%|###################################1 | 19/20 [01:34<00:04,  4.59s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:08,420]\u001b[0m Trial 58 finished with value: 0.24184734603859326 and parameters: {'lambda_l1': 1.5282163310444257, 'lambda_l2': 2.1022628673011167e-07}. Best is trial 56 with value: 0.24171678915738617.\u001b[0m\n",
      "\n",
      "regularization_factors, val_score: 0.241717:  95%|###################################1 | 19/20 [01:34<00:04,  4.59s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.157206\tvalid_1's binary_logloss: 0.241847\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.157206\tvalid_1's binary_logloss: 0.241847\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.020453 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.170037\tvalid_1's binary_logloss: 0.245605\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "regularization_factors, val_score: 0.241717:  95%|###################################1 | 19/20 [01:38<00:04,  4.59s/it]\u001b[A\n",
      "regularization_factors, val_score: 0.241717: 100%|#####################################| 20/20 [01:38<00:00,  4.54s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:12,832]\u001b[0m Trial 59 finished with value: 0.24226846742866032 and parameters: {'lambda_l1': 0.006341479231822191, 'lambda_l2': 1.587488642541311}. Best is trial 56 with value: 0.24171678915738617.\u001b[0m\n",
      "regularization_factors, val_score: 0.241717: 100%|#####################################| 20/20 [01:38<00:00,  4.92s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.157464\tvalid_1's binary_logloss: 0.242268\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.157464\tvalid_1's binary_logloss: 0.242268\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "min_data_in_leaf, val_score: 0.241717:   0%|                                                     | 0/5 [00:00<?, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.015486 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169775\tvalid_1's binary_logloss: 0.24542\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "min_data_in_leaf, val_score: 0.241717:   0%|                                                     | 0/5 [00:04<?, ?it/s]\u001b[A\n",
      "min_data_in_leaf, val_score: 0.241717:  20%|#########                                    | 1/5 [00:04<00:17,  4.35s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:17,195]\u001b[0m Trial 60 finished with value: 0.24191122257967382 and parameters: {'min_child_samples': 50}. Best is trial 60 with value: 0.24191122257967382.\u001b[0m\n",
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  20%|#########                                    | 1/5 [00:04<00:17,  4.35s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[200]\tvalid_0's binary_logloss: 0.156784\tvalid_1's binary_logloss: 0.241911\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156784\tvalid_1's binary_logloss: 0.241911\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016434 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169599\tvalid_1's binary_logloss: 0.245435\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  20%|#########                                    | 1/5 [00:08<00:17,  4.35s/it]\u001b[A\n",
      "min_data_in_leaf, val_score: 0.241717:  40%|##################                           | 2/5 [00:08<00:13,  4.40s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:21,633]\u001b[0m Trial 61 finished with value: 0.2417791004536601 and parameters: {'min_child_samples': 5}. Best is trial 61 with value: 0.2417791004536601.\u001b[0m\n",
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  40%|##################                           | 2/5 [00:08<00:13,  4.40s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156661\tvalid_1's binary_logloss: 0.241779\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156661\tvalid_1's binary_logloss: 0.241779\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016665 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169667\tvalid_1's binary_logloss: 0.245423\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  40%|##################                           | 2/5 [00:13<00:13,  4.40s/it]\u001b[A\n",
      "min_data_in_leaf, val_score: 0.241717:  60%|###########################                  | 3/5 [00:13<00:08,  4.37s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:25,954]\u001b[0m Trial 62 finished with value: 0.24209427008322146 and parameters: {'min_child_samples': 10}. Best is trial 61 with value: 0.2417791004536601.\u001b[0m\n",
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  60%|###########################                  | 3/5 [00:13<00:08,  4.37s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156697\tvalid_1's binary_logloss: 0.242094\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156697\tvalid_1's binary_logloss: 0.242094\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.016106 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169598\tvalid_1's binary_logloss: 0.245482\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  60%|###########################                  | 3/5 [00:17<00:08,  4.37s/it]\u001b[A\n",
      "min_data_in_leaf, val_score: 0.241717:  80%|####################################         | 4/5 [00:17<00:04,  4.49s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:30,627]\u001b[0m Trial 63 finished with value: 0.24223128500391303 and parameters: {'min_child_samples': 100}. Best is trial 61 with value: 0.2417791004536601.\u001b[0m\n",
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  80%|####################################         | 4/5 [00:17<00:04,  4.49s/it]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156954\tvalid_1's binary_logloss: 0.242231\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156954\tvalid_1's binary_logloss: 0.242231\n",
      "[LightGBM] [Info] Number of positive: 175371, number of negative: 175371\n",
      "[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.020255 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 4335\n",
      "[LightGBM] [Info] Number of data points in the train set: 350742, number of used features: 17\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=0.000000\n",
      "Training until validation scores don't improve for 100 rounds\n",
      "[100]\tvalid_0's binary_logloss: 0.169668\tvalid_1's binary_logloss: 0.245425\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "min_data_in_leaf, val_score: 0.241717:  80%|####################################         | 4/5 [00:21<00:04,  4.49s/it]\u001b[A\n",
      "min_data_in_leaf, val_score: 0.241717: 100%|#############################################| 5/5 [00:21<00:00,  4.38s/it]\u001b[A\u001b[32m[I 2022-11-12 15:28:34,820]\u001b[0m Trial 64 finished with value: 0.24192259767524904 and parameters: {'min_child_samples': 25}. Best is trial 61 with value: 0.2417791004536601.\u001b[0m\n",
      "min_data_in_leaf, val_score: 0.241717: 100%|#############################################| 5/5 [00:21<00:00,  4.40s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[200]\tvalid_0's binary_logloss: 0.156838\tvalid_1's binary_logloss: 0.241923\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[200]\tvalid_0's binary_logloss: 0.156838\tvalid_1's binary_logloss: 0.241923\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best params: {'objective': 'binary', 'metric': 'binary_logloss', 'random_state': 0, 'n_jobs': -1, 'feature_pre_filter': False, 'bagging_freq': 1, 'lambda_l1': 0.8290885023746439, 'lambda_l2': 1.718942996506061e-07, 'num_leaves': 66, 'feature_fraction': 0.4, 'bagging_fraction': 1.0, 'min_child_samples': 20, 'num_iterations': 200, 'early_stopping_round': 100}\n",
      "Accuracy = 0.9055\n",
      "Params:\n",
      "   objective:  binary\n",
      "   metric:  binary_logloss\n",
      "   random_state:  0\n",
      "   n_jobs:  -1\n",
      "   feature_pre_filter:  False\n",
      "   bagging_freq:  1\n",
      "   lambda_l1:  0.8290885023746439\n",
      "   lambda_l2:  1.718942996506061e-07\n",
      "   num_leaves:  66\n",
      "   feature_fraction:  0.4\n",
      "   bagging_fraction:  1.0\n",
      "   min_child_samples:  20\n",
      "   num_iterations:  200\n",
      "   early_stopping_round:  100\n"
     ]
    }
   ],
   "source": [
    "# # (10m)\n",
    "# import optuna.integration.lightgbm as lgb\n",
    "# from sklearn.metrics import accuracy_score\n",
    "\n",
    "# params = {\n",
    "#     'objective':'binary',\n",
    "#     'metric':'binary_logloss',\n",
    "#     'n_estimators': 200,\n",
    "#     'subsample_freq':1,\n",
    "#     'random_state': seed,\n",
    "#     'n_jobs': -1,\n",
    "# }\n",
    "# dtrain = lgb.Dataset(resampled_x, label = resampled_y)\n",
    "# dval = lgb.Dataset(valid_x, label = valid_y)\n",
    "# model = lgb.train(params, dtrain, valid_sets=[dtrain, dval],\n",
    "#                  verbose_eval=100, early_stopping_rounds=100)\n",
    "# prediction = np.rint(model.predict(valid_x, num_iteration=model.best_iteration))\n",
    "# accuracy = accuracy_score(valid_y, prediction)\n",
    "# params = model.params\n",
    "# print('Best params:', params)\n",
    "# print('Accuracy =', accuracy)\n",
    "# print('Params:')\n",
    "# for key, value in params.items():\n",
    "#     print('   {}:  {}'.format(key, value))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "e5abee1a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'objective': 'binary',\n",
       " 'metric': 'binary_logloss',\n",
       " 'random_state': 0,\n",
       " 'n_jobs': -1,\n",
       " 'feature_pre_filter': False,\n",
       " 'bagging_freq': 3,\n",
       " 'lambda_l1': 0.0,\n",
       " 'lambda_l2': 0.0,\n",
       " 'num_leaves': 7,\n",
       " 'feature_fraction': 0.4,\n",
       " 'bagging_fraction': 0.8772503866491428,\n",
       " 'min_child_samples': 20,\n",
       " 'num_iterations': 1000,\n",
       " 'early_stopping_round': 100}"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 저장된 하이퍼 파라미터\n",
    "import json\n",
    "# json.dump(params, open('params.json','w'))\n",
    "params = json.load(open('params.json'))\n",
    "params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "0becf128",
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {'objective': 'binary',\n",
    " 'metric': 'binary_logloss',\n",
    " 'random_state': 0,\n",
    " 'n_jobs': -1,\n",
    " 'feature_pre_filter': False,\n",
    " 'bagging_freq': 3,\n",
    " 'lambda_l1': 0.0,\n",
    " 'lambda_l2': 0.0,\n",
    " 'num_leaves': 7,\n",
    " 'feature_fraction': 0.4,\n",
    " 'bagging_fraction': 0.8772503866491428,\n",
    " 'min_child_samples': 20,\n",
    " 'num_iterations': 1000,\n",
    " 'early_stopping_round': 100}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a05bb78",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "965255db",
   "metadata": {},
   "source": [
    "#### 학습 & 검증 - stratifiedKFold\n",
    "- 해당 코드 출처: github.com/kaggler-tv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "fd388d52",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training model for CV #1\n",
      "[LightGBM] [Warning] feature_fraction is set=0.4, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.4\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.0, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.0\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8772503866491428, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8772503866491428\n",
      "[LightGBM] [Warning] lambda_l2 is set=0.0, reg_lambda=0.0 will be ignored. Current value: lambda_l2=0.0\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\engine.py:177: UserWarning: Found `num_iterations` in params. Will use it instead of argument\n",
      "  _log_warning(f\"Found `{alias}` in params. Will use it instead of argument\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tvalid_0's binary_logloss: 0.662082\n",
      "[2]\tvalid_0's binary_logloss: 0.639422\n",
      "[3]\tvalid_0's binary_logloss: 0.624755\n",
      "[4]\tvalid_0's binary_logloss: 0.607741\n",
      "[5]\tvalid_0's binary_logloss: 0.597293\n",
      "[6]\tvalid_0's binary_logloss: 0.57334\n",
      "[7]\tvalid_0's binary_logloss: 0.554678\n",
      "[8]\tvalid_0's binary_logloss: 0.546324\n",
      "[9]\tvalid_0's binary_logloss: 0.527743\n",
      "[10]\tvalid_0's binary_logloss: 0.514379\n",
      "[11]\tvalid_0's binary_logloss: 0.507063\n",
      "[12]\tvalid_0's binary_logloss: 0.491\n",
      "[13]\tvalid_0's binary_logloss: 0.479692\n",
      "[14]\tvalid_0's binary_logloss: 0.473895\n",
      "[15]\tvalid_0's binary_logloss: 0.463928\n",
      "[16]\tvalid_0's binary_logloss: 0.455153\n",
      "[17]\tvalid_0's binary_logloss: 0.444096\n",
      "[18]\tvalid_0's binary_logloss: 0.437109\n",
      "[19]\tvalid_0's binary_logloss: 0.430106\n",
      "[20]\tvalid_0's binary_logloss: 0.426437\n",
      "[21]\tvalid_0's binary_logloss: 0.417476\n",
      "[22]\tvalid_0's binary_logloss: 0.413667\n",
      "[23]\tvalid_0's binary_logloss: 0.410868\n",
      "[24]\tvalid_0's binary_logloss: 0.405523\n",
      "[25]\tvalid_0's binary_logloss: 0.398358\n",
      "[26]\tvalid_0's binary_logloss: 0.393918\n",
      "[27]\tvalid_0's binary_logloss: 0.386782\n",
      "[28]\tvalid_0's binary_logloss: 0.381266\n",
      "[29]\tvalid_0's binary_logloss: 0.378266\n",
      "[30]\tvalid_0's binary_logloss: 0.375234\n",
      "[31]\tvalid_0's binary_logloss: 0.372414\n",
      "[32]\tvalid_0's binary_logloss: 0.368589\n",
      "[33]\tvalid_0's binary_logloss: 0.365666\n",
      "[34]\tvalid_0's binary_logloss: 0.363554\n",
      "[35]\tvalid_0's binary_logloss: 0.361481\n",
      "[36]\tvalid_0's binary_logloss: 0.359646\n",
      "[37]\tvalid_0's binary_logloss: 0.356992\n",
      "[38]\tvalid_0's binary_logloss: 0.351262\n",
      "[39]\tvalid_0's binary_logloss: 0.349112\n",
      "[40]\tvalid_0's binary_logloss: 0.346909\n",
      "[41]\tvalid_0's binary_logloss: 0.345696\n",
      "[42]\tvalid_0's binary_logloss: 0.344374\n",
      "[43]\tvalid_0's binary_logloss: 0.343088\n",
      "[44]\tvalid_0's binary_logloss: 0.33851\n",
      "[45]\tvalid_0's binary_logloss: 0.336962\n",
      "[46]\tvalid_0's binary_logloss: 0.335496\n",
      "[47]\tvalid_0's binary_logloss: 0.330813\n",
      "[48]\tvalid_0's binary_logloss: 0.329638\n",
      "[49]\tvalid_0's binary_logloss: 0.328378\n",
      "[50]\tvalid_0's binary_logloss: 0.326576\n",
      "[51]\tvalid_0's binary_logloss: 0.323738\n",
      "[52]\tvalid_0's binary_logloss: 0.320968\n",
      "[53]\tvalid_0's binary_logloss: 0.317231\n",
      "[54]\tvalid_0's binary_logloss: 0.313524\n",
      "[55]\tvalid_0's binary_logloss: 0.312093\n",
      "[56]\tvalid_0's binary_logloss: 0.308735\n",
      "[57]\tvalid_0's binary_logloss: 0.307492\n",
      "[58]\tvalid_0's binary_logloss: 0.30486\n",
      "[59]\tvalid_0's binary_logloss: 0.303798\n",
      "[60]\tvalid_0's binary_logloss: 0.302099\n",
      "[61]\tvalid_0's binary_logloss: 0.300957\n",
      "[62]\tvalid_0's binary_logloss: 0.300201\n",
      "[63]\tvalid_0's binary_logloss: 0.297071\n",
      "[64]\tvalid_0's binary_logloss: 0.296064\n",
      "[65]\tvalid_0's binary_logloss: 0.294478\n",
      "[66]\tvalid_0's binary_logloss: 0.293057\n",
      "[67]\tvalid_0's binary_logloss: 0.291369\n",
      "[68]\tvalid_0's binary_logloss: 0.290153\n",
      "[69]\tvalid_0's binary_logloss: 0.289314\n",
      "[70]\tvalid_0's binary_logloss: 0.288574\n",
      "[71]\tvalid_0's binary_logloss: 0.287595\n",
      "[72]\tvalid_0's binary_logloss: 0.285552\n",
      "[73]\tvalid_0's binary_logloss: 0.284447\n",
      "[74]\tvalid_0's binary_logloss: 0.281868\n",
      "[75]\tvalid_0's binary_logloss: 0.278205\n",
      "[76]\tvalid_0's binary_logloss: 0.277435\n",
      "[77]\tvalid_0's binary_logloss: 0.274249\n",
      "[78]\tvalid_0's binary_logloss: 0.272996\n",
      "[79]\tvalid_0's binary_logloss: 0.271931\n",
      "[80]\tvalid_0's binary_logloss: 0.269814\n",
      "[81]\tvalid_0's binary_logloss: 0.267801\n",
      "[82]\tvalid_0's binary_logloss: 0.266972\n",
      "[83]\tvalid_0's binary_logloss: 0.265154\n",
      "[84]\tvalid_0's binary_logloss: 0.264093\n",
      "[85]\tvalid_0's binary_logloss: 0.261482\n",
      "[86]\tvalid_0's binary_logloss: 0.260542\n",
      "[87]\tvalid_0's binary_logloss: 0.259036\n",
      "[88]\tvalid_0's binary_logloss: 0.258309\n",
      "[89]\tvalid_0's binary_logloss: 0.25761\n",
      "[90]\tvalid_0's binary_logloss: 0.256583\n",
      "[91]\tvalid_0's binary_logloss: 0.255918\n",
      "[92]\tvalid_0's binary_logloss: 0.253513\n",
      "[93]\tvalid_0's binary_logloss: 0.251659\n",
      "[94]\tvalid_0's binary_logloss: 0.249568\n",
      "[95]\tvalid_0's binary_logloss: 0.248864\n",
      "[96]\tvalid_0's binary_logloss: 0.247611\n",
      "[97]\tvalid_0's binary_logloss: 0.24601\n",
      "[98]\tvalid_0's binary_logloss: 0.244541\n",
      "[99]\tvalid_0's binary_logloss: 0.24423\n",
      "[100]\tvalid_0's binary_logloss: 0.243109\n",
      "[101]\tvalid_0's binary_logloss: 0.242522\n",
      "[102]\tvalid_0's binary_logloss: 0.241402\n",
      "[103]\tvalid_0's binary_logloss: 0.240775\n",
      "[104]\tvalid_0's binary_logloss: 0.23989\n",
      "[105]\tvalid_0's binary_logloss: 0.23913\n",
      "[106]\tvalid_0's binary_logloss: 0.238439\n",
      "[107]\tvalid_0's binary_logloss: 0.23723\n",
      "[108]\tvalid_0's binary_logloss: 0.236634\n",
      "[109]\tvalid_0's binary_logloss: 0.236245\n",
      "[110]\tvalid_0's binary_logloss: 0.235503\n",
      "[111]\tvalid_0's binary_logloss: 0.234814\n",
      "[112]\tvalid_0's binary_logloss: 0.234353\n",
      "[113]\tvalid_0's binary_logloss: 0.232928\n",
      "[114]\tvalid_0's binary_logloss: 0.232494\n",
      "[115]\tvalid_0's binary_logloss: 0.232034\n",
      "[116]\tvalid_0's binary_logloss: 0.231578\n",
      "[117]\tvalid_0's binary_logloss: 0.231185\n",
      "[118]\tvalid_0's binary_logloss: 0.230666\n",
      "[119]\tvalid_0's binary_logloss: 0.229806\n",
      "[120]\tvalid_0's binary_logloss: 0.228971\n",
      "[121]\tvalid_0's binary_logloss: 0.228535\n",
      "[122]\tvalid_0's binary_logloss: 0.227622\n",
      "[123]\tvalid_0's binary_logloss: 0.226875\n",
      "[124]\tvalid_0's binary_logloss: 0.22646\n",
      "[125]\tvalid_0's binary_logloss: 0.226095\n",
      "[126]\tvalid_0's binary_logloss: 0.225881\n",
      "[127]\tvalid_0's binary_logloss: 0.225237\n",
      "[128]\tvalid_0's binary_logloss: 0.224683\n",
      "[129]\tvalid_0's binary_logloss: 0.224071\n",
      "[130]\tvalid_0's binary_logloss: 0.223747\n",
      "[131]\tvalid_0's binary_logloss: 0.22297\n",
      "[132]\tvalid_0's binary_logloss: 0.222482\n",
      "[133]\tvalid_0's binary_logloss: 0.221937\n",
      "[134]\tvalid_0's binary_logloss: 0.221432\n",
      "[135]\tvalid_0's binary_logloss: 0.220498\n",
      "[136]\tvalid_0's binary_logloss: 0.22028\n",
      "[137]\tvalid_0's binary_logloss: 0.220063\n",
      "[138]\tvalid_0's binary_logloss: 0.219764\n",
      "[139]\tvalid_0's binary_logloss: 0.219558\n",
      "[140]\tvalid_0's binary_logloss: 0.219279\n",
      "[141]\tvalid_0's binary_logloss: 0.218961\n",
      "[142]\tvalid_0's binary_logloss: 0.21811\n",
      "[143]\tvalid_0's binary_logloss: 0.217525\n",
      "[144]\tvalid_0's binary_logloss: 0.217269\n",
      "[145]\tvalid_0's binary_logloss: 0.217018\n",
      "[146]\tvalid_0's binary_logloss: 0.21686\n",
      "[147]\tvalid_0's binary_logloss: 0.216653\n",
      "[148]\tvalid_0's binary_logloss: 0.216395\n",
      "[149]\tvalid_0's binary_logloss: 0.215824\n",
      "[150]\tvalid_0's binary_logloss: 0.215128\n",
      "[151]\tvalid_0's binary_logloss: 0.214669\n",
      "[152]\tvalid_0's binary_logloss: 0.214396\n",
      "[153]\tvalid_0's binary_logloss: 0.214028\n",
      "[154]\tvalid_0's binary_logloss: 0.213831\n",
      "[155]\tvalid_0's binary_logloss: 0.213447\n",
      "[156]\tvalid_0's binary_logloss: 0.212935\n",
      "[157]\tvalid_0's binary_logloss: 0.212327\n",
      "[158]\tvalid_0's binary_logloss: 0.211946\n",
      "[159]\tvalid_0's binary_logloss: 0.211822\n",
      "[160]\tvalid_0's binary_logloss: 0.21147\n",
      "[161]\tvalid_0's binary_logloss: 0.211216\n",
      "[162]\tvalid_0's binary_logloss: 0.210996\n",
      "[163]\tvalid_0's binary_logloss: 0.210886\n",
      "[164]\tvalid_0's binary_logloss: 0.21026\n",
      "[165]\tvalid_0's binary_logloss: 0.20979\n",
      "[166]\tvalid_0's binary_logloss: 0.209473\n",
      "[167]\tvalid_0's binary_logloss: 0.209279\n",
      "[168]\tvalid_0's binary_logloss: 0.208783\n",
      "[169]\tvalid_0's binary_logloss: 0.208667\n",
      "[170]\tvalid_0's binary_logloss: 0.208147\n",
      "[171]\tvalid_0's binary_logloss: 0.208029\n",
      "[172]\tvalid_0's binary_logloss: 0.20777\n",
      "[173]\tvalid_0's binary_logloss: 0.207632\n",
      "[174]\tvalid_0's binary_logloss: 0.207142\n",
      "[175]\tvalid_0's binary_logloss: 0.206981\n",
      "[176]\tvalid_0's binary_logloss: 0.206728\n",
      "[177]\tvalid_0's binary_logloss: 0.206643\n",
      "[178]\tvalid_0's binary_logloss: 0.206253\n",
      "[179]\tvalid_0's binary_logloss: 0.205843\n",
      "[180]\tvalid_0's binary_logloss: 0.205605\n",
      "[181]\tvalid_0's binary_logloss: 0.205438\n",
      "[182]\tvalid_0's binary_logloss: 0.205173\n",
      "[183]\tvalid_0's binary_logloss: 0.204998\n",
      "[184]\tvalid_0's binary_logloss: 0.204649\n",
      "[185]\tvalid_0's binary_logloss: 0.204408\n",
      "[186]\tvalid_0's binary_logloss: 0.204205\n",
      "[187]\tvalid_0's binary_logloss: 0.204044\n",
      "[188]\tvalid_0's binary_logloss: 0.203452\n",
      "[189]\tvalid_0's binary_logloss: 0.203084\n",
      "[190]\tvalid_0's binary_logloss: 0.202969\n",
      "[191]\tvalid_0's binary_logloss: 0.202892\n",
      "[192]\tvalid_0's binary_logloss: 0.202782\n",
      "[193]\tvalid_0's binary_logloss: 0.202669\n",
      "[194]\tvalid_0's binary_logloss: 0.202089\n",
      "[195]\tvalid_0's binary_logloss: 0.201781\n",
      "[196]\tvalid_0's binary_logloss: 0.201644\n",
      "[197]\tvalid_0's binary_logloss: 0.201505\n",
      "[198]\tvalid_0's binary_logloss: 0.201313\n",
      "[199]\tvalid_0's binary_logloss: 0.201045\n",
      "[200]\tvalid_0's binary_logloss: 0.200924\n",
      "[201]\tvalid_0's binary_logloss: 0.200702\n",
      "[202]\tvalid_0's binary_logloss: 0.200539\n",
      "[203]\tvalid_0's binary_logloss: 0.200427\n",
      "[204]\tvalid_0's binary_logloss: 0.200204\n",
      "[205]\tvalid_0's binary_logloss: 0.200093\n",
      "[206]\tvalid_0's binary_logloss: 0.199709\n",
      "[207]\tvalid_0's binary_logloss: 0.199561\n",
      "[208]\tvalid_0's binary_logloss: 0.199062\n",
      "[209]\tvalid_0's binary_logloss: 0.198862\n",
      "[210]\tvalid_0's binary_logloss: 0.19854\n",
      "[211]\tvalid_0's binary_logloss: 0.198391\n",
      "[212]\tvalid_0's binary_logloss: 0.19809\n",
      "[213]\tvalid_0's binary_logloss: 0.197881\n",
      "[214]\tvalid_0's binary_logloss: 0.197547\n",
      "[215]\tvalid_0's binary_logloss: 0.197348\n",
      "[216]\tvalid_0's binary_logloss: 0.19724\n",
      "[217]\tvalid_0's binary_logloss: 0.196809\n",
      "[218]\tvalid_0's binary_logloss: 0.196689\n",
      "[219]\tvalid_0's binary_logloss: 0.196613\n",
      "[220]\tvalid_0's binary_logloss: 0.196496\n",
      "[221]\tvalid_0's binary_logloss: 0.196336\n",
      "[222]\tvalid_0's binary_logloss: 0.196149\n",
      "[223]\tvalid_0's binary_logloss: 0.196107\n",
      "[224]\tvalid_0's binary_logloss: 0.19604\n",
      "[225]\tvalid_0's binary_logloss: 0.195959\n",
      "[226]\tvalid_0's binary_logloss: 0.195554\n",
      "[227]\tvalid_0's binary_logloss: 0.195488\n",
      "[228]\tvalid_0's binary_logloss: 0.195382\n",
      "[229]\tvalid_0's binary_logloss: 0.195278\n",
      "[230]\tvalid_0's binary_logloss: 0.195034\n",
      "[231]\tvalid_0's binary_logloss: 0.19494\n",
      "[232]\tvalid_0's binary_logloss: 0.194875\n",
      "[233]\tvalid_0's binary_logloss: 0.194668\n",
      "[234]\tvalid_0's binary_logloss: 0.194548\n",
      "[235]\tvalid_0's binary_logloss: 0.194206\n",
      "[236]\tvalid_0's binary_logloss: 0.194124\n",
      "[237]\tvalid_0's binary_logloss: 0.193905\n",
      "[238]\tvalid_0's binary_logloss: 0.19373\n",
      "[239]\tvalid_0's binary_logloss: 0.193551\n",
      "[240]\tvalid_0's binary_logloss: 0.193363\n",
      "[241]\tvalid_0's binary_logloss: 0.193297\n",
      "[242]\tvalid_0's binary_logloss: 0.193044\n",
      "[243]\tvalid_0's binary_logloss: 0.192686\n",
      "[244]\tvalid_0's binary_logloss: 0.192606\n",
      "[245]\tvalid_0's binary_logloss: 0.192501\n",
      "[246]\tvalid_0's binary_logloss: 0.192352\n",
      "[247]\tvalid_0's binary_logloss: 0.19227\n",
      "[248]\tvalid_0's binary_logloss: 0.192037\n",
      "[249]\tvalid_0's binary_logloss: 0.192008\n",
      "[250]\tvalid_0's binary_logloss: 0.191958\n",
      "[251]\tvalid_0's binary_logloss: 0.19178\n",
      "[252]\tvalid_0's binary_logloss: 0.191731\n",
      "[253]\tvalid_0's binary_logloss: 0.191675\n",
      "[254]\tvalid_0's binary_logloss: 0.191625\n",
      "[255]\tvalid_0's binary_logloss: 0.191553\n",
      "[256]\tvalid_0's binary_logloss: 0.191475\n",
      "[257]\tvalid_0's binary_logloss: 0.191349\n",
      "[258]\tvalid_0's binary_logloss: 0.191268\n",
      "[259]\tvalid_0's binary_logloss: 0.190921\n",
      "[260]\tvalid_0's binary_logloss: 0.190774\n",
      "[261]\tvalid_0's binary_logloss: 0.190714\n",
      "[262]\tvalid_0's binary_logloss: 0.190449\n",
      "[263]\tvalid_0's binary_logloss: 0.190379\n",
      "[264]\tvalid_0's binary_logloss: 0.190332\n",
      "[265]\tvalid_0's binary_logloss: 0.190051\n",
      "[266]\tvalid_0's binary_logloss: 0.189802\n",
      "[267]\tvalid_0's binary_logloss: 0.189746\n",
      "[268]\tvalid_0's binary_logloss: 0.189667\n",
      "[269]\tvalid_0's binary_logloss: 0.1894\n",
      "[270]\tvalid_0's binary_logloss: 0.189319\n",
      "[271]\tvalid_0's binary_logloss: 0.189198\n",
      "[272]\tvalid_0's binary_logloss: 0.189098\n",
      "[273]\tvalid_0's binary_logloss: 0.189062\n",
      "[274]\tvalid_0's binary_logloss: 0.189026\n",
      "[275]\tvalid_0's binary_logloss: 0.188745\n",
      "[276]\tvalid_0's binary_logloss: 0.188692\n",
      "[277]\tvalid_0's binary_logloss: 0.188478\n",
      "[278]\tvalid_0's binary_logloss: 0.188427\n",
      "[279]\tvalid_0's binary_logloss: 0.188299\n",
      "[280]\tvalid_0's binary_logloss: 0.188104\n",
      "[281]\tvalid_0's binary_logloss: 0.18787\n",
      "[282]\tvalid_0's binary_logloss: 0.187712\n",
      "[283]\tvalid_0's binary_logloss: 0.187672\n",
      "[284]\tvalid_0's binary_logloss: 0.187549\n",
      "[285]\tvalid_0's binary_logloss: 0.187416\n",
      "[286]\tvalid_0's binary_logloss: 0.187194\n",
      "[287]\tvalid_0's binary_logloss: 0.187124\n",
      "[288]\tvalid_0's binary_logloss: 0.186994\n",
      "[289]\tvalid_0's binary_logloss: 0.186867\n",
      "[290]\tvalid_0's binary_logloss: 0.186643\n",
      "[291]\tvalid_0's binary_logloss: 0.186592\n",
      "[292]\tvalid_0's binary_logloss: 0.186429\n",
      "[293]\tvalid_0's binary_logloss: 0.186315\n",
      "[294]\tvalid_0's binary_logloss: 0.186161\n",
      "[295]\tvalid_0's binary_logloss: 0.185944\n",
      "[296]\tvalid_0's binary_logloss: 0.185848\n",
      "[297]\tvalid_0's binary_logloss: 0.185802\n",
      "[298]\tvalid_0's binary_logloss: 0.185709\n",
      "[299]\tvalid_0's binary_logloss: 0.185671\n",
      "[300]\tvalid_0's binary_logloss: 0.185557\n",
      "[301]\tvalid_0's binary_logloss: 0.185429\n",
      "[302]\tvalid_0's binary_logloss: 0.185389\n",
      "[303]\tvalid_0's binary_logloss: 0.18527\n",
      "[304]\tvalid_0's binary_logloss: 0.185234\n",
      "[305]\tvalid_0's binary_logloss: 0.185122\n",
      "[306]\tvalid_0's binary_logloss: 0.185089\n",
      "[307]\tvalid_0's binary_logloss: 0.185058\n",
      "[308]\tvalid_0's binary_logloss: 0.184975\n",
      "[309]\tvalid_0's binary_logloss: 0.18486\n",
      "[310]\tvalid_0's binary_logloss: 0.184705\n",
      "[311]\tvalid_0's binary_logloss: 0.18458\n",
      "[312]\tvalid_0's binary_logloss: 0.184536\n",
      "[313]\tvalid_0's binary_logloss: 0.184397\n",
      "[314]\tvalid_0's binary_logloss: 0.184339\n",
      "[315]\tvalid_0's binary_logloss: 0.184253\n",
      "[316]\tvalid_0's binary_logloss: 0.184129\n",
      "[317]\tvalid_0's binary_logloss: 0.184054\n",
      "[318]\tvalid_0's binary_logloss: 0.183973\n",
      "[319]\tvalid_0's binary_logloss: 0.183864\n",
      "[320]\tvalid_0's binary_logloss: 0.183781\n",
      "[321]\tvalid_0's binary_logloss: 0.183772\n",
      "[322]\tvalid_0's binary_logloss: 0.183687\n",
      "[323]\tvalid_0's binary_logloss: 0.183603\n",
      "[324]\tvalid_0's binary_logloss: 0.183538\n",
      "[325]\tvalid_0's binary_logloss: 0.183494\n",
      "[326]\tvalid_0's binary_logloss: 0.183412\n",
      "[327]\tvalid_0's binary_logloss: 0.183262\n",
      "[328]\tvalid_0's binary_logloss: 0.183071\n",
      "[329]\tvalid_0's binary_logloss: 0.182907\n",
      "[330]\tvalid_0's binary_logloss: 0.182903\n",
      "[331]\tvalid_0's binary_logloss: 0.182868\n",
      "[332]\tvalid_0's binary_logloss: 0.182825\n",
      "[333]\tvalid_0's binary_logloss: 0.18281\n",
      "[334]\tvalid_0's binary_logloss: 0.182648\n",
      "[335]\tvalid_0's binary_logloss: 0.182535\n",
      "[336]\tvalid_0's binary_logloss: 0.182514\n",
      "[337]\tvalid_0's binary_logloss: 0.182463\n",
      "[338]\tvalid_0's binary_logloss: 0.182439\n",
      "[339]\tvalid_0's binary_logloss: 0.182352\n",
      "[340]\tvalid_0's binary_logloss: 0.182244\n",
      "[341]\tvalid_0's binary_logloss: 0.182223\n",
      "[342]\tvalid_0's binary_logloss: 0.182147\n",
      "[343]\tvalid_0's binary_logloss: 0.182117\n",
      "[344]\tvalid_0's binary_logloss: 0.182077\n",
      "[345]\tvalid_0's binary_logloss: 0.181999\n",
      "[346]\tvalid_0's binary_logloss: 0.181902\n",
      "[347]\tvalid_0's binary_logloss: 0.181872\n",
      "[348]\tvalid_0's binary_logloss: 0.181821\n",
      "[349]\tvalid_0's binary_logloss: 0.181691\n",
      "[350]\tvalid_0's binary_logloss: 0.181679\n",
      "[351]\tvalid_0's binary_logloss: 0.181618\n",
      "[352]\tvalid_0's binary_logloss: 0.181556\n",
      "[353]\tvalid_0's binary_logloss: 0.18153\n",
      "[354]\tvalid_0's binary_logloss: 0.181484\n",
      "[355]\tvalid_0's binary_logloss: 0.181362\n",
      "[356]\tvalid_0's binary_logloss: 0.181333\n",
      "[357]\tvalid_0's binary_logloss: 0.181339\n",
      "[358]\tvalid_0's binary_logloss: 0.181262\n",
      "[359]\tvalid_0's binary_logloss: 0.181194\n",
      "[360]\tvalid_0's binary_logloss: 0.181197\n",
      "[361]\tvalid_0's binary_logloss: 0.181187\n",
      "[362]\tvalid_0's binary_logloss: 0.181099\n",
      "[363]\tvalid_0's binary_logloss: 0.181005\n",
      "[364]\tvalid_0's binary_logloss: 0.180993\n",
      "[365]\tvalid_0's binary_logloss: 0.180926\n",
      "[366]\tvalid_0's binary_logloss: 0.180793\n",
      "[367]\tvalid_0's binary_logloss: 0.180641\n",
      "[368]\tvalid_0's binary_logloss: 0.180575\n",
      "[369]\tvalid_0's binary_logloss: 0.180503\n",
      "[370]\tvalid_0's binary_logloss: 0.180467\n",
      "[371]\tvalid_0's binary_logloss: 0.180467\n",
      "[372]\tvalid_0's binary_logloss: 0.180371\n",
      "[373]\tvalid_0's binary_logloss: 0.180312\n",
      "[374]\tvalid_0's binary_logloss: 0.180279\n",
      "[375]\tvalid_0's binary_logloss: 0.180243\n",
      "[376]\tvalid_0's binary_logloss: 0.180148\n",
      "[377]\tvalid_0's binary_logloss: 0.180119\n",
      "[378]\tvalid_0's binary_logloss: 0.180036\n",
      "[379]\tvalid_0's binary_logloss: 0.179931\n",
      "[380]\tvalid_0's binary_logloss: 0.179872\n",
      "[381]\tvalid_0's binary_logloss: 0.179831\n",
      "[382]\tvalid_0's binary_logloss: 0.179721\n",
      "[383]\tvalid_0's binary_logloss: 0.179625\n",
      "[384]\tvalid_0's binary_logloss: 0.179616\n",
      "[385]\tvalid_0's binary_logloss: 0.179572\n",
      "[386]\tvalid_0's binary_logloss: 0.17954\n",
      "[387]\tvalid_0's binary_logloss: 0.179484\n",
      "[388]\tvalid_0's binary_logloss: 0.179472\n",
      "[389]\tvalid_0's binary_logloss: 0.17947\n",
      "[390]\tvalid_0's binary_logloss: 0.17947\n",
      "[391]\tvalid_0's binary_logloss: 0.179421\n",
      "[392]\tvalid_0's binary_logloss: 0.179363\n",
      "[393]\tvalid_0's binary_logloss: 0.179264\n",
      "[394]\tvalid_0's binary_logloss: 0.179259\n",
      "[395]\tvalid_0's binary_logloss: 0.179179\n",
      "[396]\tvalid_0's binary_logloss: 0.179159\n",
      "[397]\tvalid_0's binary_logloss: 0.179051\n",
      "[398]\tvalid_0's binary_logloss: 0.179025\n",
      "[399]\tvalid_0's binary_logloss: 0.179012\n",
      "[400]\tvalid_0's binary_logloss: 0.178943\n",
      "[401]\tvalid_0's binary_logloss: 0.178934\n",
      "[402]\tvalid_0's binary_logloss: 0.17893\n",
      "[403]\tvalid_0's binary_logloss: 0.178925\n",
      "[404]\tvalid_0's binary_logloss: 0.178874\n",
      "[405]\tvalid_0's binary_logloss: 0.178811\n",
      "[406]\tvalid_0's binary_logloss: 0.178732\n",
      "[407]\tvalid_0's binary_logloss: 0.178666\n",
      "[408]\tvalid_0's binary_logloss: 0.17864\n",
      "[409]\tvalid_0's binary_logloss: 0.178592\n",
      "[410]\tvalid_0's binary_logloss: 0.178555\n",
      "[411]\tvalid_0's binary_logloss: 0.17853\n",
      "[412]\tvalid_0's binary_logloss: 0.178457\n",
      "[413]\tvalid_0's binary_logloss: 0.17841\n",
      "[414]\tvalid_0's binary_logloss: 0.178379\n",
      "[415]\tvalid_0's binary_logloss: 0.178268\n",
      "[416]\tvalid_0's binary_logloss: 0.17821\n",
      "[417]\tvalid_0's binary_logloss: 0.178173\n",
      "[418]\tvalid_0's binary_logloss: 0.178127\n",
      "[419]\tvalid_0's binary_logloss: 0.178005\n",
      "[420]\tvalid_0's binary_logloss: 0.177952\n",
      "[421]\tvalid_0's binary_logloss: 0.177875\n",
      "[422]\tvalid_0's binary_logloss: 0.177803\n",
      "[423]\tvalid_0's binary_logloss: 0.177777\n",
      "[424]\tvalid_0's binary_logloss: 0.177752\n",
      "[425]\tvalid_0's binary_logloss: 0.177679\n",
      "[426]\tvalid_0's binary_logloss: 0.177659\n",
      "[427]\tvalid_0's binary_logloss: 0.177551\n",
      "[428]\tvalid_0's binary_logloss: 0.17748\n",
      "[429]\tvalid_0's binary_logloss: 0.177473\n",
      "[430]\tvalid_0's binary_logloss: 0.177411\n",
      "[431]\tvalid_0's binary_logloss: 0.177311\n",
      "[432]\tvalid_0's binary_logloss: 0.177261\n",
      "[433]\tvalid_0's binary_logloss: 0.177206\n",
      "[434]\tvalid_0's binary_logloss: 0.177144\n",
      "[435]\tvalid_0's binary_logloss: 0.177112\n",
      "[436]\tvalid_0's binary_logloss: 0.177013\n",
      "[437]\tvalid_0's binary_logloss: 0.17698\n",
      "[438]\tvalid_0's binary_logloss: 0.176979\n",
      "[439]\tvalid_0's binary_logloss: 0.176919\n",
      "[440]\tvalid_0's binary_logloss: 0.176839\n",
      "[441]\tvalid_0's binary_logloss: 0.176768\n",
      "[442]\tvalid_0's binary_logloss: 0.176703\n",
      "[443]\tvalid_0's binary_logloss: 0.176652\n",
      "[444]\tvalid_0's binary_logloss: 0.176603\n",
      "[445]\tvalid_0's binary_logloss: 0.176587\n",
      "[446]\tvalid_0's binary_logloss: 0.17657\n",
      "[447]\tvalid_0's binary_logloss: 0.176521\n",
      "[448]\tvalid_0's binary_logloss: 0.176489\n",
      "[449]\tvalid_0's binary_logloss: 0.176469\n",
      "[450]\tvalid_0's binary_logloss: 0.176424\n",
      "[451]\tvalid_0's binary_logloss: 0.176404\n",
      "[452]\tvalid_0's binary_logloss: 0.17637\n",
      "[453]\tvalid_0's binary_logloss: 0.17633\n",
      "[454]\tvalid_0's binary_logloss: 0.176251\n",
      "[455]\tvalid_0's binary_logloss: 0.176225\n",
      "[456]\tvalid_0's binary_logloss: 0.17614\n",
      "[457]\tvalid_0's binary_logloss: 0.176095\n",
      "[458]\tvalid_0's binary_logloss: 0.17608\n",
      "[459]\tvalid_0's binary_logloss: 0.176018\n",
      "[460]\tvalid_0's binary_logloss: 0.175991\n",
      "[461]\tvalid_0's binary_logloss: 0.175934\n",
      "[462]\tvalid_0's binary_logloss: 0.175926\n",
      "[463]\tvalid_0's binary_logloss: 0.175874\n",
      "[464]\tvalid_0's binary_logloss: 0.175843\n",
      "[465]\tvalid_0's binary_logloss: 0.175832\n",
      "[466]\tvalid_0's binary_logloss: 0.17579\n",
      "[467]\tvalid_0's binary_logloss: 0.175774\n",
      "[468]\tvalid_0's binary_logloss: 0.175729\n",
      "[469]\tvalid_0's binary_logloss: 0.1757\n",
      "[470]\tvalid_0's binary_logloss: 0.175676\n",
      "[471]\tvalid_0's binary_logloss: 0.175588\n",
      "[472]\tvalid_0's binary_logloss: 0.175579\n",
      "[473]\tvalid_0's binary_logloss: 0.175567\n",
      "[474]\tvalid_0's binary_logloss: 0.175564\n",
      "[475]\tvalid_0's binary_logloss: 0.175552\n",
      "[476]\tvalid_0's binary_logloss: 0.175519\n",
      "[477]\tvalid_0's binary_logloss: 0.175488\n",
      "[478]\tvalid_0's binary_logloss: 0.175484\n",
      "[479]\tvalid_0's binary_logloss: 0.175447\n",
      "[480]\tvalid_0's binary_logloss: 0.175438\n",
      "[481]\tvalid_0's binary_logloss: 0.175353\n",
      "[482]\tvalid_0's binary_logloss: 0.175331\n",
      "[483]\tvalid_0's binary_logloss: 0.175272\n",
      "[484]\tvalid_0's binary_logloss: 0.175208\n",
      "[485]\tvalid_0's binary_logloss: 0.1752\n",
      "[486]\tvalid_0's binary_logloss: 0.175149\n",
      "[487]\tvalid_0's binary_logloss: 0.175108\n",
      "[488]\tvalid_0's binary_logloss: 0.175048\n",
      "[489]\tvalid_0's binary_logloss: 0.174997\n",
      "[490]\tvalid_0's binary_logloss: 0.174968\n",
      "[491]\tvalid_0's binary_logloss: 0.174952\n",
      "[492]\tvalid_0's binary_logloss: 0.174946\n",
      "[493]\tvalid_0's binary_logloss: 0.17491\n",
      "[494]\tvalid_0's binary_logloss: 0.174878\n",
      "[495]\tvalid_0's binary_logloss: 0.174834\n",
      "[496]\tvalid_0's binary_logloss: 0.174754\n",
      "[497]\tvalid_0's binary_logloss: 0.174653\n",
      "[498]\tvalid_0's binary_logloss: 0.174633\n",
      "[499]\tvalid_0's binary_logloss: 0.174599\n",
      "[500]\tvalid_0's binary_logloss: 0.174579\n",
      "[501]\tvalid_0's binary_logloss: 0.174535\n",
      "[502]\tvalid_0's binary_logloss: 0.174514\n",
      "[503]\tvalid_0's binary_logloss: 0.174499\n",
      "[504]\tvalid_0's binary_logloss: 0.174487\n",
      "[505]\tvalid_0's binary_logloss: 0.174459\n",
      "[506]\tvalid_0's binary_logloss: 0.174418\n",
      "[507]\tvalid_0's binary_logloss: 0.174389\n",
      "[508]\tvalid_0's binary_logloss: 0.174329\n",
      "[509]\tvalid_0's binary_logloss: 0.17423\n",
      "[510]\tvalid_0's binary_logloss: 0.174141\n",
      "[511]\tvalid_0's binary_logloss: 0.174141\n",
      "[512]\tvalid_0's binary_logloss: 0.174134\n",
      "[513]\tvalid_0's binary_logloss: 0.17409\n",
      "[514]\tvalid_0's binary_logloss: 0.174073\n",
      "[515]\tvalid_0's binary_logloss: 0.174015\n",
      "[516]\tvalid_0's binary_logloss: 0.173959\n",
      "[517]\tvalid_0's binary_logloss: 0.173907\n",
      "[518]\tvalid_0's binary_logloss: 0.17386\n",
      "[519]\tvalid_0's binary_logloss: 0.173825\n",
      "[520]\tvalid_0's binary_logloss: 0.173806\n",
      "[521]\tvalid_0's binary_logloss: 0.173793\n",
      "[522]\tvalid_0's binary_logloss: 0.173764\n",
      "[523]\tvalid_0's binary_logloss: 0.173718\n",
      "[524]\tvalid_0's binary_logloss: 0.173665\n",
      "[525]\tvalid_0's binary_logloss: 0.173626\n",
      "[526]\tvalid_0's binary_logloss: 0.173578\n",
      "[527]\tvalid_0's binary_logloss: 0.173546\n",
      "[528]\tvalid_0's binary_logloss: 0.173541\n",
      "[529]\tvalid_0's binary_logloss: 0.17351\n",
      "[530]\tvalid_0's binary_logloss: 0.173479\n",
      "[531]\tvalid_0's binary_logloss: 0.173443\n",
      "[532]\tvalid_0's binary_logloss: 0.17341\n",
      "[533]\tvalid_0's binary_logloss: 0.173384\n",
      "[534]\tvalid_0's binary_logloss: 0.173366\n",
      "[535]\tvalid_0's binary_logloss: 0.173357\n",
      "[536]\tvalid_0's binary_logloss: 0.173351\n",
      "[537]\tvalid_0's binary_logloss: 0.173304\n",
      "[538]\tvalid_0's binary_logloss: 0.173275\n",
      "[539]\tvalid_0's binary_logloss: 0.173257\n",
      "[540]\tvalid_0's binary_logloss: 0.173195\n",
      "[541]\tvalid_0's binary_logloss: 0.173149\n",
      "[542]\tvalid_0's binary_logloss: 0.173098\n",
      "[543]\tvalid_0's binary_logloss: 0.173056\n",
      "[544]\tvalid_0's binary_logloss: 0.173052\n",
      "[545]\tvalid_0's binary_logloss: 0.173047\n",
      "[546]\tvalid_0's binary_logloss: 0.173041\n",
      "[547]\tvalid_0's binary_logloss: 0.172996\n",
      "[548]\tvalid_0's binary_logloss: 0.172943\n",
      "[549]\tvalid_0's binary_logloss: 0.172905\n",
      "[550]\tvalid_0's binary_logloss: 0.172857\n",
      "[551]\tvalid_0's binary_logloss: 0.172828\n",
      "[552]\tvalid_0's binary_logloss: 0.172807\n",
      "[553]\tvalid_0's binary_logloss: 0.172753\n",
      "[554]\tvalid_0's binary_logloss: 0.172711\n",
      "[555]\tvalid_0's binary_logloss: 0.172682\n",
      "[556]\tvalid_0's binary_logloss: 0.172593\n",
      "[557]\tvalid_0's binary_logloss: 0.172576\n",
      "[558]\tvalid_0's binary_logloss: 0.172548\n",
      "[559]\tvalid_0's binary_logloss: 0.172502\n",
      "[560]\tvalid_0's binary_logloss: 0.172499\n",
      "[561]\tvalid_0's binary_logloss: 0.172459\n",
      "[562]\tvalid_0's binary_logloss: 0.172457\n",
      "[563]\tvalid_0's binary_logloss: 0.172409\n",
      "[564]\tvalid_0's binary_logloss: 0.172405\n",
      "[565]\tvalid_0's binary_logloss: 0.172407\n",
      "[566]\tvalid_0's binary_logloss: 0.172383\n",
      "[567]\tvalid_0's binary_logloss: 0.172365\n",
      "[568]\tvalid_0's binary_logloss: 0.172314\n",
      "[569]\tvalid_0's binary_logloss: 0.172273\n",
      "[570]\tvalid_0's binary_logloss: 0.172231\n",
      "[571]\tvalid_0's binary_logloss: 0.172217\n",
      "[572]\tvalid_0's binary_logloss: 0.172214\n",
      "[573]\tvalid_0's binary_logloss: 0.172211\n",
      "[574]\tvalid_0's binary_logloss: 0.172206\n",
      "[575]\tvalid_0's binary_logloss: 0.172198\n",
      "[576]\tvalid_0's binary_logloss: 0.172163\n",
      "[577]\tvalid_0's binary_logloss: 0.17213\n",
      "[578]\tvalid_0's binary_logloss: 0.172128\n",
      "[579]\tvalid_0's binary_logloss: 0.172115\n",
      "[580]\tvalid_0's binary_logloss: 0.172086\n",
      "[581]\tvalid_0's binary_logloss: 0.172078\n",
      "[582]\tvalid_0's binary_logloss: 0.172024\n",
      "[583]\tvalid_0's binary_logloss: 0.17195\n",
      "[584]\tvalid_0's binary_logloss: 0.171921\n",
      "[585]\tvalid_0's binary_logloss: 0.171885\n",
      "[586]\tvalid_0's binary_logloss: 0.171871\n",
      "[587]\tvalid_0's binary_logloss: 0.171824\n",
      "[588]\tvalid_0's binary_logloss: 0.171794\n",
      "[589]\tvalid_0's binary_logloss: 0.171754\n",
      "[590]\tvalid_0's binary_logloss: 0.17174\n",
      "[591]\tvalid_0's binary_logloss: 0.171719\n",
      "[592]\tvalid_0's binary_logloss: 0.17169\n",
      "[593]\tvalid_0's binary_logloss: 0.171675\n",
      "[594]\tvalid_0's binary_logloss: 0.171666\n",
      "[595]\tvalid_0's binary_logloss: 0.171669\n",
      "[596]\tvalid_0's binary_logloss: 0.171642\n",
      "[597]\tvalid_0's binary_logloss: 0.171627\n",
      "[598]\tvalid_0's binary_logloss: 0.171623\n",
      "[599]\tvalid_0's binary_logloss: 0.171596\n",
      "[600]\tvalid_0's binary_logloss: 0.171583\n",
      "[601]\tvalid_0's binary_logloss: 0.171583\n",
      "[602]\tvalid_0's binary_logloss: 0.171549\n",
      "[603]\tvalid_0's binary_logloss: 0.171533\n",
      "[604]\tvalid_0's binary_logloss: 0.171527\n",
      "[605]\tvalid_0's binary_logloss: 0.171517\n",
      "[606]\tvalid_0's binary_logloss: 0.171502\n",
      "[607]\tvalid_0's binary_logloss: 0.171447\n",
      "[608]\tvalid_0's binary_logloss: 0.171424\n",
      "[609]\tvalid_0's binary_logloss: 0.171397\n",
      "[610]\tvalid_0's binary_logloss: 0.171388\n",
      "[611]\tvalid_0's binary_logloss: 0.171357\n",
      "[612]\tvalid_0's binary_logloss: 0.17135\n",
      "[613]\tvalid_0's binary_logloss: 0.171313\n",
      "[614]\tvalid_0's binary_logloss: 0.171278\n",
      "[615]\tvalid_0's binary_logloss: 0.17126\n",
      "[616]\tvalid_0's binary_logloss: 0.171261\n",
      "[617]\tvalid_0's binary_logloss: 0.17121\n",
      "[618]\tvalid_0's binary_logloss: 0.171217\n",
      "[619]\tvalid_0's binary_logloss: 0.171192\n",
      "[620]\tvalid_0's binary_logloss: 0.171171\n",
      "[621]\tvalid_0's binary_logloss: 0.171156\n",
      "[622]\tvalid_0's binary_logloss: 0.171135\n",
      "[623]\tvalid_0's binary_logloss: 0.171117\n",
      "[624]\tvalid_0's binary_logloss: 0.171113\n",
      "[625]\tvalid_0's binary_logloss: 0.171112\n",
      "[626]\tvalid_0's binary_logloss: 0.171078\n",
      "[627]\tvalid_0's binary_logloss: 0.171063\n",
      "[628]\tvalid_0's binary_logloss: 0.171038\n",
      "[629]\tvalid_0's binary_logloss: 0.171026\n",
      "[630]\tvalid_0's binary_logloss: 0.170983\n",
      "[631]\tvalid_0's binary_logloss: 0.170962\n",
      "[632]\tvalid_0's binary_logloss: 0.170967\n",
      "[633]\tvalid_0's binary_logloss: 0.170955\n",
      "[634]\tvalid_0's binary_logloss: 0.170942\n",
      "[635]\tvalid_0's binary_logloss: 0.17092\n",
      "[636]\tvalid_0's binary_logloss: 0.17091\n",
      "[637]\tvalid_0's binary_logloss: 0.170905\n",
      "[638]\tvalid_0's binary_logloss: 0.17085\n",
      "[639]\tvalid_0's binary_logloss: 0.170843\n",
      "[640]\tvalid_0's binary_logloss: 0.170839\n",
      "[641]\tvalid_0's binary_logloss: 0.170811\n",
      "[642]\tvalid_0's binary_logloss: 0.170806\n",
      "[643]\tvalid_0's binary_logloss: 0.170804\n",
      "[644]\tvalid_0's binary_logloss: 0.170768\n",
      "[645]\tvalid_0's binary_logloss: 0.170758\n",
      "[646]\tvalid_0's binary_logloss: 0.170759\n",
      "[647]\tvalid_0's binary_logloss: 0.170708\n",
      "[648]\tvalid_0's binary_logloss: 0.170663\n",
      "[649]\tvalid_0's binary_logloss: 0.170634\n",
      "[650]\tvalid_0's binary_logloss: 0.170619\n",
      "[651]\tvalid_0's binary_logloss: 0.170598\n",
      "[652]\tvalid_0's binary_logloss: 0.170591\n",
      "[653]\tvalid_0's binary_logloss: 0.170576\n",
      "[654]\tvalid_0's binary_logloss: 0.170555\n",
      "[655]\tvalid_0's binary_logloss: 0.170511\n",
      "[656]\tvalid_0's binary_logloss: 0.17051\n",
      "[657]\tvalid_0's binary_logloss: 0.170509\n",
      "[658]\tvalid_0's binary_logloss: 0.170491\n",
      "[659]\tvalid_0's binary_logloss: 0.170489\n",
      "[660]\tvalid_0's binary_logloss: 0.170476\n",
      "[661]\tvalid_0's binary_logloss: 0.170467\n",
      "[662]\tvalid_0's binary_logloss: 0.170459\n",
      "[663]\tvalid_0's binary_logloss: 0.170457\n",
      "[664]\tvalid_0's binary_logloss: 0.170428\n",
      "[665]\tvalid_0's binary_logloss: 0.170411\n",
      "[666]\tvalid_0's binary_logloss: 0.170386\n",
      "[667]\tvalid_0's binary_logloss: 0.170374\n",
      "[668]\tvalid_0's binary_logloss: 0.170358\n",
      "[669]\tvalid_0's binary_logloss: 0.170337\n",
      "[670]\tvalid_0's binary_logloss: 0.170328\n",
      "[671]\tvalid_0's binary_logloss: 0.170322\n",
      "[672]\tvalid_0's binary_logloss: 0.17033\n",
      "[673]\tvalid_0's binary_logloss: 0.170322\n",
      "[674]\tvalid_0's binary_logloss: 0.170303\n",
      "[675]\tvalid_0's binary_logloss: 0.170292\n",
      "[676]\tvalid_0's binary_logloss: 0.170289\n",
      "[677]\tvalid_0's binary_logloss: 0.170266\n",
      "[678]\tvalid_0's binary_logloss: 0.170266\n",
      "[679]\tvalid_0's binary_logloss: 0.170227\n",
      "[680]\tvalid_0's binary_logloss: 0.170198\n",
      "[681]\tvalid_0's binary_logloss: 0.170154\n",
      "[682]\tvalid_0's binary_logloss: 0.170134\n",
      "[683]\tvalid_0's binary_logloss: 0.170116\n",
      "[684]\tvalid_0's binary_logloss: 0.170089\n",
      "[685]\tvalid_0's binary_logloss: 0.170077\n",
      "[686]\tvalid_0's binary_logloss: 0.170052\n",
      "[687]\tvalid_0's binary_logloss: 0.17004\n",
      "[688]\tvalid_0's binary_logloss: 0.170006\n",
      "[689]\tvalid_0's binary_logloss: 0.169984\n",
      "[690]\tvalid_0's binary_logloss: 0.169964\n",
      "[691]\tvalid_0's binary_logloss: 0.169944\n",
      "[692]\tvalid_0's binary_logloss: 0.169924\n",
      "[693]\tvalid_0's binary_logloss: 0.169903\n",
      "[694]\tvalid_0's binary_logloss: 0.169904\n",
      "[695]\tvalid_0's binary_logloss: 0.169892\n",
      "[696]\tvalid_0's binary_logloss: 0.169878\n",
      "[697]\tvalid_0's binary_logloss: 0.169874\n",
      "[698]\tvalid_0's binary_logloss: 0.169825\n",
      "[699]\tvalid_0's binary_logloss: 0.169787\n",
      "[700]\tvalid_0's binary_logloss: 0.169781\n",
      "[701]\tvalid_0's binary_logloss: 0.169778\n",
      "[702]\tvalid_0's binary_logloss: 0.169765\n",
      "[703]\tvalid_0's binary_logloss: 0.169734\n",
      "[704]\tvalid_0's binary_logloss: 0.169724\n",
      "[705]\tvalid_0's binary_logloss: 0.169711\n",
      "[706]\tvalid_0's binary_logloss: 0.169696\n",
      "[707]\tvalid_0's binary_logloss: 0.169677\n",
      "[708]\tvalid_0's binary_logloss: 0.169642\n",
      "[709]\tvalid_0's binary_logloss: 0.169644\n",
      "[710]\tvalid_0's binary_logloss: 0.16964\n",
      "[711]\tvalid_0's binary_logloss: 0.169642\n",
      "[712]\tvalid_0's binary_logloss: 0.169629\n",
      "[713]\tvalid_0's binary_logloss: 0.169615\n",
      "[714]\tvalid_0's binary_logloss: 0.169581\n",
      "[715]\tvalid_0's binary_logloss: 0.16956\n",
      "[716]\tvalid_0's binary_logloss: 0.16955\n",
      "[717]\tvalid_0's binary_logloss: 0.16953\n",
      "[718]\tvalid_0's binary_logloss: 0.169521\n",
      "[719]\tvalid_0's binary_logloss: 0.169518\n",
      "[720]\tvalid_0's binary_logloss: 0.169501\n",
      "[721]\tvalid_0's binary_logloss: 0.1695\n",
      "[722]\tvalid_0's binary_logloss: 0.16948\n",
      "[723]\tvalid_0's binary_logloss: 0.169471\n",
      "[724]\tvalid_0's binary_logloss: 0.169461\n",
      "[725]\tvalid_0's binary_logloss: 0.169425\n",
      "[726]\tvalid_0's binary_logloss: 0.16942\n",
      "[727]\tvalid_0's binary_logloss: 0.169389\n",
      "[728]\tvalid_0's binary_logloss: 0.169365\n",
      "[729]\tvalid_0's binary_logloss: 0.169356\n",
      "[730]\tvalid_0's binary_logloss: 0.169343\n",
      "[731]\tvalid_0's binary_logloss: 0.169326\n",
      "[732]\tvalid_0's binary_logloss: 0.169301\n",
      "[733]\tvalid_0's binary_logloss: 0.16928\n",
      "[734]\tvalid_0's binary_logloss: 0.169279\n",
      "[735]\tvalid_0's binary_logloss: 0.169242\n",
      "[736]\tvalid_0's binary_logloss: 0.16923\n",
      "[737]\tvalid_0's binary_logloss: 0.169189\n",
      "[738]\tvalid_0's binary_logloss: 0.169165\n",
      "[739]\tvalid_0's binary_logloss: 0.169145\n",
      "[740]\tvalid_0's binary_logloss: 0.169139\n",
      "[741]\tvalid_0's binary_logloss: 0.16912\n",
      "[742]\tvalid_0's binary_logloss: 0.169098\n",
      "[743]\tvalid_0's binary_logloss: 0.169081\n",
      "[744]\tvalid_0's binary_logloss: 0.169068\n",
      "[745]\tvalid_0's binary_logloss: 0.169039\n",
      "[746]\tvalid_0's binary_logloss: 0.169012\n",
      "[747]\tvalid_0's binary_logloss: 0.168987\n",
      "[748]\tvalid_0's binary_logloss: 0.168983\n",
      "[749]\tvalid_0's binary_logloss: 0.168975\n",
      "[750]\tvalid_0's binary_logloss: 0.168969\n",
      "[751]\tvalid_0's binary_logloss: 0.168938\n",
      "[752]\tvalid_0's binary_logloss: 0.168919\n",
      "[753]\tvalid_0's binary_logloss: 0.168911\n",
      "[754]\tvalid_0's binary_logloss: 0.168907\n",
      "[755]\tvalid_0's binary_logloss: 0.168906\n",
      "[756]\tvalid_0's binary_logloss: 0.168909\n",
      "[757]\tvalid_0's binary_logloss: 0.168886\n",
      "[758]\tvalid_0's binary_logloss: 0.16887\n",
      "[759]\tvalid_0's binary_logloss: 0.168847\n",
      "[760]\tvalid_0's binary_logloss: 0.168836\n",
      "[761]\tvalid_0's binary_logloss: 0.168826\n",
      "[762]\tvalid_0's binary_logloss: 0.16882\n",
      "[763]\tvalid_0's binary_logloss: 0.168808\n",
      "[764]\tvalid_0's binary_logloss: 0.1688\n",
      "[765]\tvalid_0's binary_logloss: 0.168781\n",
      "[766]\tvalid_0's binary_logloss: 0.168756\n",
      "[767]\tvalid_0's binary_logloss: 0.168755\n",
      "[768]\tvalid_0's binary_logloss: 0.168741\n",
      "[769]\tvalid_0's binary_logloss: 0.168715\n",
      "[770]\tvalid_0's binary_logloss: 0.168699\n",
      "[771]\tvalid_0's binary_logloss: 0.168682\n",
      "[772]\tvalid_0's binary_logloss: 0.16867\n",
      "[773]\tvalid_0's binary_logloss: 0.168667\n",
      "[774]\tvalid_0's binary_logloss: 0.168643\n",
      "[775]\tvalid_0's binary_logloss: 0.168646\n",
      "[776]\tvalid_0's binary_logloss: 0.168648\n",
      "[777]\tvalid_0's binary_logloss: 0.168652\n",
      "[778]\tvalid_0's binary_logloss: 0.168653\n",
      "[779]\tvalid_0's binary_logloss: 0.168641\n",
      "[780]\tvalid_0's binary_logloss: 0.168634\n",
      "[781]\tvalid_0's binary_logloss: 0.168627\n",
      "[782]\tvalid_0's binary_logloss: 0.168613\n",
      "[783]\tvalid_0's binary_logloss: 0.168601\n",
      "[784]\tvalid_0's binary_logloss: 0.168598\n",
      "[785]\tvalid_0's binary_logloss: 0.168602\n",
      "[786]\tvalid_0's binary_logloss: 0.168584\n",
      "[787]\tvalid_0's binary_logloss: 0.168583\n",
      "[788]\tvalid_0's binary_logloss: 0.168552\n",
      "[789]\tvalid_0's binary_logloss: 0.168547\n",
      "[790]\tvalid_0's binary_logloss: 0.168534\n",
      "[791]\tvalid_0's binary_logloss: 0.168504\n",
      "[792]\tvalid_0's binary_logloss: 0.168487\n",
      "[793]\tvalid_0's binary_logloss: 0.168472\n",
      "[794]\tvalid_0's binary_logloss: 0.16846\n",
      "[795]\tvalid_0's binary_logloss: 0.168452\n",
      "[796]\tvalid_0's binary_logloss: 0.168419\n",
      "[797]\tvalid_0's binary_logloss: 0.168405\n",
      "[798]\tvalid_0's binary_logloss: 0.168369\n",
      "[799]\tvalid_0's binary_logloss: 0.168359\n",
      "[800]\tvalid_0's binary_logloss: 0.168337\n",
      "[801]\tvalid_0's binary_logloss: 0.168341\n",
      "[802]\tvalid_0's binary_logloss: 0.168309\n",
      "[803]\tvalid_0's binary_logloss: 0.168276\n",
      "[804]\tvalid_0's binary_logloss: 0.168262\n",
      "[805]\tvalid_0's binary_logloss: 0.168258\n",
      "[806]\tvalid_0's binary_logloss: 0.168201\n",
      "[807]\tvalid_0's binary_logloss: 0.168194\n",
      "[808]\tvalid_0's binary_logloss: 0.168197\n",
      "[809]\tvalid_0's binary_logloss: 0.168182\n",
      "[810]\tvalid_0's binary_logloss: 0.16817\n",
      "[811]\tvalid_0's binary_logloss: 0.168161\n",
      "[812]\tvalid_0's binary_logloss: 0.168153\n",
      "[813]\tvalid_0's binary_logloss: 0.168144\n",
      "[814]\tvalid_0's binary_logloss: 0.168132\n",
      "[815]\tvalid_0's binary_logloss: 0.168111\n",
      "[816]\tvalid_0's binary_logloss: 0.168105\n",
      "[817]\tvalid_0's binary_logloss: 0.168085\n",
      "[818]\tvalid_0's binary_logloss: 0.168071\n",
      "[819]\tvalid_0's binary_logloss: 0.168043\n",
      "[820]\tvalid_0's binary_logloss: 0.168041\n",
      "[821]\tvalid_0's binary_logloss: 0.168041\n",
      "[822]\tvalid_0's binary_logloss: 0.168023\n",
      "[823]\tvalid_0's binary_logloss: 0.167995\n",
      "[824]\tvalid_0's binary_logloss: 0.167991\n",
      "[825]\tvalid_0's binary_logloss: 0.167974\n",
      "[826]\tvalid_0's binary_logloss: 0.167956\n",
      "[827]\tvalid_0's binary_logloss: 0.167948\n",
      "[828]\tvalid_0's binary_logloss: 0.167944\n",
      "[829]\tvalid_0's binary_logloss: 0.167928\n",
      "[830]\tvalid_0's binary_logloss: 0.167912\n",
      "[831]\tvalid_0's binary_logloss: 0.167902\n",
      "[832]\tvalid_0's binary_logloss: 0.167885\n",
      "[833]\tvalid_0's binary_logloss: 0.167873\n",
      "[834]\tvalid_0's binary_logloss: 0.167857\n",
      "[835]\tvalid_0's binary_logloss: 0.167856\n",
      "[836]\tvalid_0's binary_logloss: 0.167833\n",
      "[837]\tvalid_0's binary_logloss: 0.167828\n",
      "[838]\tvalid_0's binary_logloss: 0.167808\n",
      "[839]\tvalid_0's binary_logloss: 0.167798\n",
      "[840]\tvalid_0's binary_logloss: 0.167782\n",
      "[841]\tvalid_0's binary_logloss: 0.167775\n",
      "[842]\tvalid_0's binary_logloss: 0.16776\n",
      "[843]\tvalid_0's binary_logloss: 0.167742\n",
      "[844]\tvalid_0's binary_logloss: 0.167741\n",
      "[845]\tvalid_0's binary_logloss: 0.167711\n",
      "[846]\tvalid_0's binary_logloss: 0.167706\n",
      "[847]\tvalid_0's binary_logloss: 0.16769\n",
      "[848]\tvalid_0's binary_logloss: 0.167688\n",
      "[849]\tvalid_0's binary_logloss: 0.167679\n",
      "[850]\tvalid_0's binary_logloss: 0.16766\n",
      "[851]\tvalid_0's binary_logloss: 0.167663\n",
      "[852]\tvalid_0's binary_logloss: 0.167662\n",
      "[853]\tvalid_0's binary_logloss: 0.167649\n",
      "[854]\tvalid_0's binary_logloss: 0.16765\n",
      "[855]\tvalid_0's binary_logloss: 0.167639\n",
      "[856]\tvalid_0's binary_logloss: 0.167641\n",
      "[857]\tvalid_0's binary_logloss: 0.167628\n",
      "[858]\tvalid_0's binary_logloss: 0.167611\n",
      "[859]\tvalid_0's binary_logloss: 0.167588\n",
      "[860]\tvalid_0's binary_logloss: 0.167583\n",
      "[861]\tvalid_0's binary_logloss: 0.16758\n",
      "[862]\tvalid_0's binary_logloss: 0.167571\n",
      "[863]\tvalid_0's binary_logloss: 0.167558\n",
      "[864]\tvalid_0's binary_logloss: 0.167562\n",
      "[865]\tvalid_0's binary_logloss: 0.167548\n",
      "[866]\tvalid_0's binary_logloss: 0.167531\n",
      "[867]\tvalid_0's binary_logloss: 0.167527\n",
      "[868]\tvalid_0's binary_logloss: 0.167493\n",
      "[869]\tvalid_0's binary_logloss: 0.167483\n",
      "[870]\tvalid_0's binary_logloss: 0.167476\n",
      "[871]\tvalid_0's binary_logloss: 0.167464\n",
      "[872]\tvalid_0's binary_logloss: 0.167458\n",
      "[873]\tvalid_0's binary_logloss: 0.167442\n",
      "[874]\tvalid_0's binary_logloss: 0.167442\n",
      "[875]\tvalid_0's binary_logloss: 0.167431\n",
      "[876]\tvalid_0's binary_logloss: 0.167427\n",
      "[877]\tvalid_0's binary_logloss: 0.167408\n",
      "[878]\tvalid_0's binary_logloss: 0.167395\n",
      "[879]\tvalid_0's binary_logloss: 0.167371\n",
      "[880]\tvalid_0's binary_logloss: 0.167367\n",
      "[881]\tvalid_0's binary_logloss: 0.167339\n",
      "[882]\tvalid_0's binary_logloss: 0.167329\n",
      "[883]\tvalid_0's binary_logloss: 0.167318\n",
      "[884]\tvalid_0's binary_logloss: 0.167304\n",
      "[885]\tvalid_0's binary_logloss: 0.167284\n",
      "[886]\tvalid_0's binary_logloss: 0.167258\n",
      "[887]\tvalid_0's binary_logloss: 0.167242\n",
      "[888]\tvalid_0's binary_logloss: 0.167225\n",
      "[889]\tvalid_0's binary_logloss: 0.167199\n",
      "[890]\tvalid_0's binary_logloss: 0.167197\n",
      "[891]\tvalid_0's binary_logloss: 0.167201\n",
      "[892]\tvalid_0's binary_logloss: 0.167201\n",
      "[893]\tvalid_0's binary_logloss: 0.167188\n",
      "[894]\tvalid_0's binary_logloss: 0.167184\n",
      "[895]\tvalid_0's binary_logloss: 0.167181\n",
      "[896]\tvalid_0's binary_logloss: 0.167153\n",
      "[897]\tvalid_0's binary_logloss: 0.167132\n",
      "[898]\tvalid_0's binary_logloss: 0.167128\n",
      "[899]\tvalid_0's binary_logloss: 0.167097\n",
      "[900]\tvalid_0's binary_logloss: 0.167098\n",
      "[901]\tvalid_0's binary_logloss: 0.167094\n",
      "[902]\tvalid_0's binary_logloss: 0.167079\n",
      "[903]\tvalid_0's binary_logloss: 0.167066\n",
      "[904]\tvalid_0's binary_logloss: 0.16706\n",
      "[905]\tvalid_0's binary_logloss: 0.167051\n",
      "[906]\tvalid_0's binary_logloss: 0.167024\n",
      "[907]\tvalid_0's binary_logloss: 0.167014\n",
      "[908]\tvalid_0's binary_logloss: 0.166987\n",
      "[909]\tvalid_0's binary_logloss: 0.166988\n",
      "[910]\tvalid_0's binary_logloss: 0.166978\n",
      "[911]\tvalid_0's binary_logloss: 0.166971\n",
      "[912]\tvalid_0's binary_logloss: 0.166968\n",
      "[913]\tvalid_0's binary_logloss: 0.166944\n",
      "[914]\tvalid_0's binary_logloss: 0.166933\n",
      "[915]\tvalid_0's binary_logloss: 0.166935\n",
      "[916]\tvalid_0's binary_logloss: 0.16692\n",
      "[917]\tvalid_0's binary_logloss: 0.166919\n",
      "[918]\tvalid_0's binary_logloss: 0.166915\n",
      "[919]\tvalid_0's binary_logloss: 0.166898\n",
      "[920]\tvalid_0's binary_logloss: 0.166884\n",
      "[921]\tvalid_0's binary_logloss: 0.166883\n",
      "[922]\tvalid_0's binary_logloss: 0.166868\n",
      "[923]\tvalid_0's binary_logloss: 0.166853\n",
      "[924]\tvalid_0's binary_logloss: 0.166831\n",
      "[925]\tvalid_0's binary_logloss: 0.166826\n",
      "[926]\tvalid_0's binary_logloss: 0.166819\n",
      "[927]\tvalid_0's binary_logloss: 0.166818\n",
      "[928]\tvalid_0's binary_logloss: 0.166801\n",
      "[929]\tvalid_0's binary_logloss: 0.166783\n",
      "[930]\tvalid_0's binary_logloss: 0.166787\n",
      "[931]\tvalid_0's binary_logloss: 0.166791\n",
      "[932]\tvalid_0's binary_logloss: 0.16678\n",
      "[933]\tvalid_0's binary_logloss: 0.16677\n",
      "[934]\tvalid_0's binary_logloss: 0.166762\n",
      "[935]\tvalid_0's binary_logloss: 0.166747\n",
      "[936]\tvalid_0's binary_logloss: 0.166737\n",
      "[937]\tvalid_0's binary_logloss: 0.166723\n",
      "[938]\tvalid_0's binary_logloss: 0.16671\n",
      "[939]\tvalid_0's binary_logloss: 0.1667\n",
      "[940]\tvalid_0's binary_logloss: 0.166681\n",
      "[941]\tvalid_0's binary_logloss: 0.166686\n",
      "[942]\tvalid_0's binary_logloss: 0.166688\n",
      "[943]\tvalid_0's binary_logloss: 0.166667\n",
      "[944]\tvalid_0's binary_logloss: 0.166641\n",
      "[945]\tvalid_0's binary_logloss: 0.166629\n",
      "[946]\tvalid_0's binary_logloss: 0.166623\n",
      "[947]\tvalid_0's binary_logloss: 0.166617\n",
      "[948]\tvalid_0's binary_logloss: 0.166615\n",
      "[949]\tvalid_0's binary_logloss: 0.166607\n",
      "[950]\tvalid_0's binary_logloss: 0.16661\n",
      "[951]\tvalid_0's binary_logloss: 0.166612\n",
      "[952]\tvalid_0's binary_logloss: 0.166598\n",
      "[953]\tvalid_0's binary_logloss: 0.166582\n",
      "[954]\tvalid_0's binary_logloss: 0.166579\n",
      "[955]\tvalid_0's binary_logloss: 0.166556\n",
      "[956]\tvalid_0's binary_logloss: 0.166548\n",
      "[957]\tvalid_0's binary_logloss: 0.166528\n",
      "[958]\tvalid_0's binary_logloss: 0.166511\n",
      "[959]\tvalid_0's binary_logloss: 0.166502\n",
      "[960]\tvalid_0's binary_logloss: 0.166499\n",
      "[961]\tvalid_0's binary_logloss: 0.166481\n",
      "[962]\tvalid_0's binary_logloss: 0.166471\n",
      "[963]\tvalid_0's binary_logloss: 0.166467\n",
      "[964]\tvalid_0's binary_logloss: 0.166459\n",
      "[965]\tvalid_0's binary_logloss: 0.166456\n",
      "[966]\tvalid_0's binary_logloss: 0.166448\n",
      "[967]\tvalid_0's binary_logloss: 0.166446\n",
      "[968]\tvalid_0's binary_logloss: 0.166414\n",
      "[969]\tvalid_0's binary_logloss: 0.166389\n",
      "[970]\tvalid_0's binary_logloss: 0.16637\n",
      "[971]\tvalid_0's binary_logloss: 0.16636\n",
      "[972]\tvalid_0's binary_logloss: 0.166356\n",
      "[973]\tvalid_0's binary_logloss: 0.166343\n",
      "[974]\tvalid_0's binary_logloss: 0.166334\n",
      "[975]\tvalid_0's binary_logloss: 0.166328\n",
      "[976]\tvalid_0's binary_logloss: 0.166312\n",
      "[977]\tvalid_0's binary_logloss: 0.166298\n",
      "[978]\tvalid_0's binary_logloss: 0.166272\n",
      "[979]\tvalid_0's binary_logloss: 0.166259\n",
      "[980]\tvalid_0's binary_logloss: 0.166248\n",
      "[981]\tvalid_0's binary_logloss: 0.166247\n",
      "[982]\tvalid_0's binary_logloss: 0.166241\n",
      "[983]\tvalid_0's binary_logloss: 0.166238\n",
      "[984]\tvalid_0's binary_logloss: 0.166232\n",
      "[985]\tvalid_0's binary_logloss: 0.166219\n",
      "[986]\tvalid_0's binary_logloss: 0.166215\n",
      "[987]\tvalid_0's binary_logloss: 0.166202\n",
      "[988]\tvalid_0's binary_logloss: 0.166197\n",
      "[989]\tvalid_0's binary_logloss: 0.166196\n",
      "[990]\tvalid_0's binary_logloss: 0.166195\n",
      "[991]\tvalid_0's binary_logloss: 0.16619\n",
      "[992]\tvalid_0's binary_logloss: 0.166167\n",
      "[993]\tvalid_0's binary_logloss: 0.166151\n",
      "[994]\tvalid_0's binary_logloss: 0.166133\n",
      "[995]\tvalid_0's binary_logloss: 0.166129\n",
      "[996]\tvalid_0's binary_logloss: 0.166115\n",
      "[997]\tvalid_0's binary_logloss: 0.166113\n",
      "[998]\tvalid_0's binary_logloss: 0.166105\n",
      "[999]\tvalid_0's binary_logloss: 0.16609\n",
      "[1000]\tvalid_0's binary_logloss: 0.166083\n",
      "training model for CV #2\n",
      "[LightGBM] [Warning] feature_fraction is set=0.4, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.4\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.0, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.0\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8772503866491428, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8772503866491428\n",
      "[LightGBM] [Warning] lambda_l2 is set=0.0, reg_lambda=0.0 will be ignored. Current value: lambda_l2=0.0\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\engine.py:177: UserWarning: Found `num_iterations` in params. Will use it instead of argument\n",
      "  _log_warning(f\"Found `{alias}` in params. Will use it instead of argument\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tvalid_0's binary_logloss: 0.662532\n",
      "[2]\tvalid_0's binary_logloss: 0.639989\n",
      "[3]\tvalid_0's binary_logloss: 0.625283\n",
      "[4]\tvalid_0's binary_logloss: 0.608414\n",
      "[5]\tvalid_0's binary_logloss: 0.597939\n",
      "[6]\tvalid_0's binary_logloss: 0.57399\n",
      "[7]\tvalid_0's binary_logloss: 0.555683\n",
      "[8]\tvalid_0's binary_logloss: 0.547421\n",
      "[9]\tvalid_0's binary_logloss: 0.527829\n",
      "[10]\tvalid_0's binary_logloss: 0.514709\n",
      "[11]\tvalid_0's binary_logloss: 0.50754\n",
      "[12]\tvalid_0's binary_logloss: 0.491943\n",
      "[13]\tvalid_0's binary_logloss: 0.480558\n",
      "[14]\tvalid_0's binary_logloss: 0.474778\n",
      "[15]\tvalid_0's binary_logloss: 0.465083\n",
      "[16]\tvalid_0's binary_logloss: 0.456446\n",
      "[17]\tvalid_0's binary_logloss: 0.447583\n",
      "[18]\tvalid_0's binary_logloss: 0.437652\n",
      "[19]\tvalid_0's binary_logloss: 0.430811\n",
      "[20]\tvalid_0's binary_logloss: 0.427157\n",
      "[21]\tvalid_0's binary_logloss: 0.418112\n",
      "[22]\tvalid_0's binary_logloss: 0.414397\n",
      "[23]\tvalid_0's binary_logloss: 0.411676\n",
      "[24]\tvalid_0's binary_logloss: 0.406467\n",
      "[25]\tvalid_0's binary_logloss: 0.39919\n",
      "[26]\tvalid_0's binary_logloss: 0.394866\n",
      "[27]\tvalid_0's binary_logloss: 0.387775\n",
      "[28]\tvalid_0's binary_logloss: 0.380835\n",
      "[29]\tvalid_0's binary_logloss: 0.377723\n",
      "[30]\tvalid_0's binary_logloss: 0.373302\n",
      "[31]\tvalid_0's binary_logloss: 0.370222\n",
      "[32]\tvalid_0's binary_logloss: 0.364719\n",
      "[33]\tvalid_0's binary_logloss: 0.361823\n",
      "[34]\tvalid_0's binary_logloss: 0.359675\n",
      "[35]\tvalid_0's binary_logloss: 0.357693\n",
      "[36]\tvalid_0's binary_logloss: 0.355896\n",
      "[37]\tvalid_0's binary_logloss: 0.353394\n",
      "[38]\tvalid_0's binary_logloss: 0.348124\n",
      "[39]\tvalid_0's binary_logloss: 0.345874\n",
      "[40]\tvalid_0's binary_logloss: 0.343836\n",
      "[41]\tvalid_0's binary_logloss: 0.342697\n",
      "[42]\tvalid_0's binary_logloss: 0.341313\n",
      "[43]\tvalid_0's binary_logloss: 0.340075\n",
      "[44]\tvalid_0's binary_logloss: 0.336009\n",
      "[45]\tvalid_0's binary_logloss: 0.334497\n",
      "[46]\tvalid_0's binary_logloss: 0.333124\n",
      "[47]\tvalid_0's binary_logloss: 0.328799\n",
      "[48]\tvalid_0's binary_logloss: 0.327415\n",
      "[49]\tvalid_0's binary_logloss: 0.326421\n",
      "[50]\tvalid_0's binary_logloss: 0.324541\n",
      "[51]\tvalid_0's binary_logloss: 0.320898\n",
      "[52]\tvalid_0's binary_logloss: 0.319493\n",
      "[53]\tvalid_0's binary_logloss: 0.316407\n",
      "[54]\tvalid_0's binary_logloss: 0.313216\n",
      "[55]\tvalid_0's binary_logloss: 0.312188\n",
      "[56]\tvalid_0's binary_logloss: 0.309617\n",
      "[57]\tvalid_0's binary_logloss: 0.308442\n",
      "[58]\tvalid_0's binary_logloss: 0.305306\n",
      "[59]\tvalid_0's binary_logloss: 0.304241\n",
      "[60]\tvalid_0's binary_logloss: 0.302565\n",
      "[61]\tvalid_0's binary_logloss: 0.301279\n",
      "[62]\tvalid_0's binary_logloss: 0.30051\n",
      "[63]\tvalid_0's binary_logloss: 0.297893\n",
      "[64]\tvalid_0's binary_logloss: 0.297236\n",
      "[65]\tvalid_0's binary_logloss: 0.295824\n",
      "[66]\tvalid_0's binary_logloss: 0.294369\n",
      "[67]\tvalid_0's binary_logloss: 0.292359\n",
      "[68]\tvalid_0's binary_logloss: 0.291111\n",
      "[69]\tvalid_0's binary_logloss: 0.290229\n",
      "[70]\tvalid_0's binary_logloss: 0.289241\n",
      "[71]\tvalid_0's binary_logloss: 0.288271\n",
      "[72]\tvalid_0's binary_logloss: 0.286738\n",
      "[73]\tvalid_0's binary_logloss: 0.285988\n",
      "[74]\tvalid_0's binary_logloss: 0.28295\n",
      "[75]\tvalid_0's binary_logloss: 0.279214\n",
      "[76]\tvalid_0's binary_logloss: 0.278519\n",
      "[77]\tvalid_0's binary_logloss: 0.27527\n",
      "[78]\tvalid_0's binary_logloss: 0.273496\n",
      "[79]\tvalid_0's binary_logloss: 0.272355\n",
      "[80]\tvalid_0's binary_logloss: 0.271161\n",
      "[81]\tvalid_0's binary_logloss: 0.270012\n",
      "[82]\tvalid_0's binary_logloss: 0.268843\n",
      "[83]\tvalid_0's binary_logloss: 0.266473\n",
      "[84]\tvalid_0's binary_logloss: 0.265587\n",
      "[85]\tvalid_0's binary_logloss: 0.264574\n",
      "[86]\tvalid_0's binary_logloss: 0.263147\n",
      "[87]\tvalid_0's binary_logloss: 0.261466\n",
      "[88]\tvalid_0's binary_logloss: 0.260637\n",
      "[89]\tvalid_0's binary_logloss: 0.259761\n",
      "[90]\tvalid_0's binary_logloss: 0.259056\n",
      "[91]\tvalid_0's binary_logloss: 0.258388\n",
      "[92]\tvalid_0's binary_logloss: 0.255843\n",
      "[93]\tvalid_0's binary_logloss: 0.254649\n",
      "[94]\tvalid_0's binary_logloss: 0.252437\n",
      "[95]\tvalid_0's binary_logloss: 0.251076\n",
      "[96]\tvalid_0's binary_logloss: 0.249383\n",
      "[97]\tvalid_0's binary_logloss: 0.248526\n",
      "[98]\tvalid_0's binary_logloss: 0.246419\n",
      "[99]\tvalid_0's binary_logloss: 0.246032\n",
      "[100]\tvalid_0's binary_logloss: 0.244512\n",
      "[101]\tvalid_0's binary_logloss: 0.244003\n",
      "[102]\tvalid_0's binary_logloss: 0.243307\n",
      "[103]\tvalid_0's binary_logloss: 0.242786\n",
      "[104]\tvalid_0's binary_logloss: 0.241727\n",
      "[105]\tvalid_0's binary_logloss: 0.240408\n",
      "[106]\tvalid_0's binary_logloss: 0.239742\n",
      "[107]\tvalid_0's binary_logloss: 0.239114\n",
      "[108]\tvalid_0's binary_logloss: 0.238432\n",
      "[109]\tvalid_0's binary_logloss: 0.238132\n",
      "[110]\tvalid_0's binary_logloss: 0.237526\n",
      "[111]\tvalid_0's binary_logloss: 0.236969\n",
      "[112]\tvalid_0's binary_logloss: 0.235993\n",
      "[113]\tvalid_0's binary_logloss: 0.234896\n",
      "[114]\tvalid_0's binary_logloss: 0.234487\n",
      "[115]\tvalid_0's binary_logloss: 0.233931\n",
      "[116]\tvalid_0's binary_logloss: 0.233353\n",
      "[117]\tvalid_0's binary_logloss: 0.232989\n",
      "[118]\tvalid_0's binary_logloss: 0.232486\n",
      "[119]\tvalid_0's binary_logloss: 0.231891\n",
      "[120]\tvalid_0's binary_logloss: 0.231518\n",
      "[121]\tvalid_0's binary_logloss: 0.230919\n",
      "[122]\tvalid_0's binary_logloss: 0.229967\n",
      "[123]\tvalid_0's binary_logloss: 0.229557\n",
      "[124]\tvalid_0's binary_logloss: 0.229318\n",
      "[125]\tvalid_0's binary_logloss: 0.228206\n",
      "[126]\tvalid_0's binary_logloss: 0.227864\n",
      "[127]\tvalid_0's binary_logloss: 0.22723\n",
      "[128]\tvalid_0's binary_logloss: 0.226708\n",
      "[129]\tvalid_0's binary_logloss: 0.226199\n",
      "[130]\tvalid_0's binary_logloss: 0.225891\n",
      "[131]\tvalid_0's binary_logloss: 0.22529\n",
      "[132]\tvalid_0's binary_logloss: 0.224778\n",
      "[133]\tvalid_0's binary_logloss: 0.224298\n",
      "[134]\tvalid_0's binary_logloss: 0.223897\n",
      "[135]\tvalid_0's binary_logloss: 0.223526\n",
      "[136]\tvalid_0's binary_logloss: 0.22331\n",
      "[137]\tvalid_0's binary_logloss: 0.222934\n",
      "[138]\tvalid_0's binary_logloss: 0.222573\n",
      "[139]\tvalid_0's binary_logloss: 0.221858\n",
      "[140]\tvalid_0's binary_logloss: 0.221586\n",
      "[141]\tvalid_0's binary_logloss: 0.221327\n",
      "[142]\tvalid_0's binary_logloss: 0.221066\n",
      "[143]\tvalid_0's binary_logloss: 0.220712\n",
      "[144]\tvalid_0's binary_logloss: 0.220545\n",
      "[145]\tvalid_0's binary_logloss: 0.220281\n",
      "[146]\tvalid_0's binary_logloss: 0.220041\n",
      "[147]\tvalid_0's binary_logloss: 0.219799\n",
      "[148]\tvalid_0's binary_logloss: 0.218989\n",
      "[149]\tvalid_0's binary_logloss: 0.21883\n",
      "[150]\tvalid_0's binary_logloss: 0.218001\n",
      "[151]\tvalid_0's binary_logloss: 0.21773\n",
      "[152]\tvalid_0's binary_logloss: 0.217138\n",
      "[153]\tvalid_0's binary_logloss: 0.217005\n",
      "[154]\tvalid_0's binary_logloss: 0.216569\n",
      "[155]\tvalid_0's binary_logloss: 0.216201\n",
      "[156]\tvalid_0's binary_logloss: 0.215784\n",
      "[157]\tvalid_0's binary_logloss: 0.215462\n",
      "[158]\tvalid_0's binary_logloss: 0.215103\n",
      "[159]\tvalid_0's binary_logloss: 0.214924\n",
      "[160]\tvalid_0's binary_logloss: 0.214797\n",
      "[161]\tvalid_0's binary_logloss: 0.214466\n",
      "[162]\tvalid_0's binary_logloss: 0.213496\n",
      "[163]\tvalid_0's binary_logloss: 0.213328\n",
      "[164]\tvalid_0's binary_logloss: 0.212732\n",
      "[165]\tvalid_0's binary_logloss: 0.212283\n",
      "[166]\tvalid_0's binary_logloss: 0.211646\n",
      "[167]\tvalid_0's binary_logloss: 0.211491\n",
      "[168]\tvalid_0's binary_logloss: 0.211118\n",
      "[169]\tvalid_0's binary_logloss: 0.210986\n",
      "[170]\tvalid_0's binary_logloss: 0.210505\n",
      "[171]\tvalid_0's binary_logloss: 0.210349\n",
      "[172]\tvalid_0's binary_logloss: 0.21008\n",
      "[173]\tvalid_0's binary_logloss: 0.209944\n",
      "[174]\tvalid_0's binary_logloss: 0.209849\n",
      "[175]\tvalid_0's binary_logloss: 0.209574\n",
      "[176]\tvalid_0's binary_logloss: 0.209162\n",
      "[177]\tvalid_0's binary_logloss: 0.208942\n",
      "[178]\tvalid_0's binary_logloss: 0.208596\n",
      "[179]\tvalid_0's binary_logloss: 0.208301\n",
      "[180]\tvalid_0's binary_logloss: 0.208061\n",
      "[181]\tvalid_0's binary_logloss: 0.207701\n",
      "[182]\tvalid_0's binary_logloss: 0.207563\n",
      "[183]\tvalid_0's binary_logloss: 0.20744\n",
      "[184]\tvalid_0's binary_logloss: 0.207313\n",
      "[185]\tvalid_0's binary_logloss: 0.207194\n",
      "[186]\tvalid_0's binary_logloss: 0.207074\n",
      "[187]\tvalid_0's binary_logloss: 0.206818\n",
      "[188]\tvalid_0's binary_logloss: 0.206612\n",
      "[189]\tvalid_0's binary_logloss: 0.206321\n",
      "[190]\tvalid_0's binary_logloss: 0.20623\n",
      "[191]\tvalid_0's binary_logloss: 0.206111\n",
      "[192]\tvalid_0's binary_logloss: 0.205408\n",
      "[193]\tvalid_0's binary_logloss: 0.204934\n",
      "[194]\tvalid_0's binary_logloss: 0.204761\n",
      "[195]\tvalid_0's binary_logloss: 0.204271\n",
      "[196]\tvalid_0's binary_logloss: 0.203906\n",
      "[197]\tvalid_0's binary_logloss: 0.203824\n",
      "[198]\tvalid_0's binary_logloss: 0.20338\n",
      "[199]\tvalid_0's binary_logloss: 0.202909\n",
      "[200]\tvalid_0's binary_logloss: 0.202715\n",
      "[201]\tvalid_0's binary_logloss: 0.202535\n",
      "[202]\tvalid_0's binary_logloss: 0.202474\n",
      "[203]\tvalid_0's binary_logloss: 0.202363\n",
      "[204]\tvalid_0's binary_logloss: 0.202219\n",
      "[205]\tvalid_0's binary_logloss: 0.202042\n",
      "[206]\tvalid_0's binary_logloss: 0.201688\n",
      "[207]\tvalid_0's binary_logloss: 0.201542\n",
      "[208]\tvalid_0's binary_logloss: 0.201406\n",
      "[209]\tvalid_0's binary_logloss: 0.201271\n",
      "[210]\tvalid_0's binary_logloss: 0.200873\n",
      "[211]\tvalid_0's binary_logloss: 0.200821\n",
      "[212]\tvalid_0's binary_logloss: 0.200426\n",
      "[213]\tvalid_0's binary_logloss: 0.200321\n",
      "[214]\tvalid_0's binary_logloss: 0.200247\n",
      "[215]\tvalid_0's binary_logloss: 0.200034\n",
      "[216]\tvalid_0's binary_logloss: 0.199942\n",
      "[217]\tvalid_0's binary_logloss: 0.199665\n",
      "[218]\tvalid_0's binary_logloss: 0.199562\n",
      "[219]\tvalid_0's binary_logloss: 0.199262\n",
      "[220]\tvalid_0's binary_logloss: 0.199024\n",
      "[221]\tvalid_0's binary_logloss: 0.198772\n",
      "[222]\tvalid_0's binary_logloss: 0.198709\n",
      "[223]\tvalid_0's binary_logloss: 0.198586\n",
      "[224]\tvalid_0's binary_logloss: 0.198495\n",
      "[225]\tvalid_0's binary_logloss: 0.198199\n",
      "[226]\tvalid_0's binary_logloss: 0.198094\n",
      "[227]\tvalid_0's binary_logloss: 0.19799\n",
      "[228]\tvalid_0's binary_logloss: 0.197771\n",
      "[229]\tvalid_0's binary_logloss: 0.197601\n",
      "[230]\tvalid_0's binary_logloss: 0.197206\n",
      "[231]\tvalid_0's binary_logloss: 0.197114\n",
      "[232]\tvalid_0's binary_logloss: 0.197053\n",
      "[233]\tvalid_0's binary_logloss: 0.19691\n",
      "[234]\tvalid_0's binary_logloss: 0.196789\n",
      "[235]\tvalid_0's binary_logloss: 0.196611\n",
      "[236]\tvalid_0's binary_logloss: 0.196183\n",
      "[237]\tvalid_0's binary_logloss: 0.196124\n",
      "[238]\tvalid_0's binary_logloss: 0.195947\n",
      "[239]\tvalid_0's binary_logloss: 0.195799\n",
      "[240]\tvalid_0's binary_logloss: 0.195657\n",
      "[241]\tvalid_0's binary_logloss: 0.195418\n",
      "[242]\tvalid_0's binary_logloss: 0.195381\n",
      "[243]\tvalid_0's binary_logloss: 0.195284\n",
      "[244]\tvalid_0's binary_logloss: 0.195077\n",
      "[245]\tvalid_0's binary_logloss: 0.195045\n",
      "[246]\tvalid_0's binary_logloss: 0.19501\n",
      "[247]\tvalid_0's binary_logloss: 0.194893\n",
      "[248]\tvalid_0's binary_logloss: 0.194678\n",
      "[249]\tvalid_0's binary_logloss: 0.194413\n",
      "[250]\tvalid_0's binary_logloss: 0.194326\n",
      "[251]\tvalid_0's binary_logloss: 0.194181\n",
      "[252]\tvalid_0's binary_logloss: 0.193998\n",
      "[253]\tvalid_0's binary_logloss: 0.193762\n",
      "[254]\tvalid_0's binary_logloss: 0.193475\n",
      "[255]\tvalid_0's binary_logloss: 0.193439\n",
      "[256]\tvalid_0's binary_logloss: 0.193348\n",
      "[257]\tvalid_0's binary_logloss: 0.193249\n",
      "[258]\tvalid_0's binary_logloss: 0.193178\n",
      "[259]\tvalid_0's binary_logloss: 0.193101\n",
      "[260]\tvalid_0's binary_logloss: 0.192986\n",
      "[261]\tvalid_0's binary_logloss: 0.192854\n",
      "[262]\tvalid_0's binary_logloss: 0.192498\n",
      "[263]\tvalid_0's binary_logloss: 0.192439\n",
      "[264]\tvalid_0's binary_logloss: 0.192399\n",
      "[265]\tvalid_0's binary_logloss: 0.192228\n",
      "[266]\tvalid_0's binary_logloss: 0.192121\n",
      "[267]\tvalid_0's binary_logloss: 0.192009\n",
      "[268]\tvalid_0's binary_logloss: 0.191953\n",
      "[269]\tvalid_0's binary_logloss: 0.191888\n",
      "[270]\tvalid_0's binary_logloss: 0.191779\n",
      "[271]\tvalid_0's binary_logloss: 0.191729\n",
      "[272]\tvalid_0's binary_logloss: 0.191565\n",
      "[273]\tvalid_0's binary_logloss: 0.191353\n",
      "[274]\tvalid_0's binary_logloss: 0.191259\n",
      "[275]\tvalid_0's binary_logloss: 0.191135\n",
      "[276]\tvalid_0's binary_logloss: 0.191086\n",
      "[277]\tvalid_0's binary_logloss: 0.190736\n",
      "[278]\tvalid_0's binary_logloss: 0.190643\n",
      "[279]\tvalid_0's binary_logloss: 0.190614\n",
      "[280]\tvalid_0's binary_logloss: 0.190548\n",
      "[281]\tvalid_0's binary_logloss: 0.190251\n",
      "[282]\tvalid_0's binary_logloss: 0.190069\n",
      "[283]\tvalid_0's binary_logloss: 0.189955\n",
      "[284]\tvalid_0's binary_logloss: 0.189798\n",
      "[285]\tvalid_0's binary_logloss: 0.189674\n",
      "[286]\tvalid_0's binary_logloss: 0.189423\n",
      "[287]\tvalid_0's binary_logloss: 0.189374\n",
      "[288]\tvalid_0's binary_logloss: 0.189319\n",
      "[289]\tvalid_0's binary_logloss: 0.189238\n",
      "[290]\tvalid_0's binary_logloss: 0.189016\n",
      "[291]\tvalid_0's binary_logloss: 0.188993\n",
      "[292]\tvalid_0's binary_logloss: 0.188927\n",
      "[293]\tvalid_0's binary_logloss: 0.188805\n",
      "[294]\tvalid_0's binary_logloss: 0.188543\n",
      "[295]\tvalid_0's binary_logloss: 0.188372\n",
      "[296]\tvalid_0's binary_logloss: 0.188323\n",
      "[297]\tvalid_0's binary_logloss: 0.188254\n",
      "[298]\tvalid_0's binary_logloss: 0.188122\n",
      "[299]\tvalid_0's binary_logloss: 0.188103\n",
      "[300]\tvalid_0's binary_logloss: 0.188026\n",
      "[301]\tvalid_0's binary_logloss: 0.187826\n",
      "[302]\tvalid_0's binary_logloss: 0.187751\n",
      "[303]\tvalid_0's binary_logloss: 0.187576\n",
      "[304]\tvalid_0's binary_logloss: 0.187528\n",
      "[305]\tvalid_0's binary_logloss: 0.187393\n",
      "[306]\tvalid_0's binary_logloss: 0.187323\n",
      "[307]\tvalid_0's binary_logloss: 0.187196\n",
      "[308]\tvalid_0's binary_logloss: 0.187093\n",
      "[309]\tvalid_0's binary_logloss: 0.186955\n",
      "[310]\tvalid_0's binary_logloss: 0.186905\n",
      "[311]\tvalid_0's binary_logloss: 0.186753\n",
      "[312]\tvalid_0's binary_logloss: 0.186606\n",
      "[313]\tvalid_0's binary_logloss: 0.1865\n",
      "[314]\tvalid_0's binary_logloss: 0.186479\n",
      "[315]\tvalid_0's binary_logloss: 0.18646\n",
      "[316]\tvalid_0's binary_logloss: 0.186375\n",
      "[317]\tvalid_0's binary_logloss: 0.18637\n",
      "[318]\tvalid_0's binary_logloss: 0.186312\n",
      "[319]\tvalid_0's binary_logloss: 0.186172\n",
      "[320]\tvalid_0's binary_logloss: 0.186097\n",
      "[321]\tvalid_0's binary_logloss: 0.186024\n",
      "[322]\tvalid_0's binary_logloss: 0.185876\n",
      "[323]\tvalid_0's binary_logloss: 0.185739\n",
      "[324]\tvalid_0's binary_logloss: 0.185677\n",
      "[325]\tvalid_0's binary_logloss: 0.185589\n",
      "[326]\tvalid_0's binary_logloss: 0.185525\n",
      "[327]\tvalid_0's binary_logloss: 0.185468\n",
      "[328]\tvalid_0's binary_logloss: 0.185457\n",
      "[329]\tvalid_0's binary_logloss: 0.185306\n",
      "[330]\tvalid_0's binary_logloss: 0.185302\n",
      "[331]\tvalid_0's binary_logloss: 0.185274\n",
      "[332]\tvalid_0's binary_logloss: 0.185156\n",
      "[333]\tvalid_0's binary_logloss: 0.185072\n",
      "[334]\tvalid_0's binary_logloss: 0.184878\n",
      "[335]\tvalid_0's binary_logloss: 0.184818\n",
      "[336]\tvalid_0's binary_logloss: 0.184595\n",
      "[337]\tvalid_0's binary_logloss: 0.184535\n",
      "[338]\tvalid_0's binary_logloss: 0.184514\n",
      "[339]\tvalid_0's binary_logloss: 0.184377\n",
      "[340]\tvalid_0's binary_logloss: 0.184293\n",
      "[341]\tvalid_0's binary_logloss: 0.184277\n",
      "[342]\tvalid_0's binary_logloss: 0.184253\n",
      "[343]\tvalid_0's binary_logloss: 0.184169\n",
      "[344]\tvalid_0's binary_logloss: 0.184035\n",
      "[345]\tvalid_0's binary_logloss: 0.184003\n",
      "[346]\tvalid_0's binary_logloss: 0.183997\n",
      "[347]\tvalid_0's binary_logloss: 0.183962\n",
      "[348]\tvalid_0's binary_logloss: 0.183938\n",
      "[349]\tvalid_0's binary_logloss: 0.183822\n",
      "[350]\tvalid_0's binary_logloss: 0.183727\n",
      "[351]\tvalid_0's binary_logloss: 0.183646\n",
      "[352]\tvalid_0's binary_logloss: 0.183634\n",
      "[353]\tvalid_0's binary_logloss: 0.183618\n",
      "[354]\tvalid_0's binary_logloss: 0.18351\n",
      "[355]\tvalid_0's binary_logloss: 0.183493\n",
      "[356]\tvalid_0's binary_logloss: 0.183425\n",
      "[357]\tvalid_0's binary_logloss: 0.183359\n",
      "[358]\tvalid_0's binary_logloss: 0.18325\n",
      "[359]\tvalid_0's binary_logloss: 0.18323\n",
      "[360]\tvalid_0's binary_logloss: 0.183191\n",
      "[361]\tvalid_0's binary_logloss: 0.183183\n",
      "[362]\tvalid_0's binary_logloss: 0.183128\n",
      "[363]\tvalid_0's binary_logloss: 0.183088\n",
      "[364]\tvalid_0's binary_logloss: 0.183044\n",
      "[365]\tvalid_0's binary_logloss: 0.183042\n",
      "[366]\tvalid_0's binary_logloss: 0.183004\n",
      "[367]\tvalid_0's binary_logloss: 0.182898\n",
      "[368]\tvalid_0's binary_logloss: 0.182803\n",
      "[369]\tvalid_0's binary_logloss: 0.182666\n",
      "[370]\tvalid_0's binary_logloss: 0.182661\n",
      "[371]\tvalid_0's binary_logloss: 0.182582\n",
      "[372]\tvalid_0's binary_logloss: 0.182487\n",
      "[373]\tvalid_0's binary_logloss: 0.182417\n",
      "[374]\tvalid_0's binary_logloss: 0.182323\n",
      "[375]\tvalid_0's binary_logloss: 0.182187\n",
      "[376]\tvalid_0's binary_logloss: 0.182122\n",
      "[377]\tvalid_0's binary_logloss: 0.182099\n",
      "[378]\tvalid_0's binary_logloss: 0.182084\n",
      "[379]\tvalid_0's binary_logloss: 0.181932\n",
      "[380]\tvalid_0's binary_logloss: 0.181895\n",
      "[381]\tvalid_0's binary_logloss: 0.181856\n",
      "[382]\tvalid_0's binary_logloss: 0.181822\n",
      "[383]\tvalid_0's binary_logloss: 0.181746\n",
      "[384]\tvalid_0's binary_logloss: 0.181709\n",
      "[385]\tvalid_0's binary_logloss: 0.181681\n",
      "[386]\tvalid_0's binary_logloss: 0.181672\n",
      "[387]\tvalid_0's binary_logloss: 0.181647\n",
      "[388]\tvalid_0's binary_logloss: 0.181611\n",
      "[389]\tvalid_0's binary_logloss: 0.181598\n",
      "[390]\tvalid_0's binary_logloss: 0.181561\n",
      "[391]\tvalid_0's binary_logloss: 0.181525\n",
      "[392]\tvalid_0's binary_logloss: 0.181495\n",
      "[393]\tvalid_0's binary_logloss: 0.18141\n",
      "[394]\tvalid_0's binary_logloss: 0.18141\n",
      "[395]\tvalid_0's binary_logloss: 0.181365\n",
      "[396]\tvalid_0's binary_logloss: 0.181358\n",
      "[397]\tvalid_0's binary_logloss: 0.181334\n",
      "[398]\tvalid_0's binary_logloss: 0.181322\n",
      "[399]\tvalid_0's binary_logloss: 0.181258\n",
      "[400]\tvalid_0's binary_logloss: 0.181244\n",
      "[401]\tvalid_0's binary_logloss: 0.181166\n",
      "[402]\tvalid_0's binary_logloss: 0.181031\n",
      "[403]\tvalid_0's binary_logloss: 0.181029\n",
      "[404]\tvalid_0's binary_logloss: 0.181026\n",
      "[405]\tvalid_0's binary_logloss: 0.180995\n",
      "[406]\tvalid_0's binary_logloss: 0.180897\n",
      "[407]\tvalid_0's binary_logloss: 0.180852\n",
      "[408]\tvalid_0's binary_logloss: 0.180813\n",
      "[409]\tvalid_0's binary_logloss: 0.180757\n",
      "[410]\tvalid_0's binary_logloss: 0.180712\n",
      "[411]\tvalid_0's binary_logloss: 0.180712\n",
      "[412]\tvalid_0's binary_logloss: 0.180676\n",
      "[413]\tvalid_0's binary_logloss: 0.180529\n",
      "[414]\tvalid_0's binary_logloss: 0.180488\n",
      "[415]\tvalid_0's binary_logloss: 0.180431\n",
      "[416]\tvalid_0's binary_logloss: 0.180353\n",
      "[417]\tvalid_0's binary_logloss: 0.180324\n",
      "[418]\tvalid_0's binary_logloss: 0.18027\n",
      "[419]\tvalid_0's binary_logloss: 0.180252\n",
      "[420]\tvalid_0's binary_logloss: 0.180231\n",
      "[421]\tvalid_0's binary_logloss: 0.180215\n",
      "[422]\tvalid_0's binary_logloss: 0.180192\n",
      "[423]\tvalid_0's binary_logloss: 0.180156\n",
      "[424]\tvalid_0's binary_logloss: 0.180135\n",
      "[425]\tvalid_0's binary_logloss: 0.180087\n",
      "[426]\tvalid_0's binary_logloss: 0.180079\n",
      "[427]\tvalid_0's binary_logloss: 0.179989\n",
      "[428]\tvalid_0's binary_logloss: 0.179883\n",
      "[429]\tvalid_0's binary_logloss: 0.179876\n",
      "[430]\tvalid_0's binary_logloss: 0.179853\n",
      "[431]\tvalid_0's binary_logloss: 0.179835\n",
      "[432]\tvalid_0's binary_logloss: 0.179803\n",
      "[433]\tvalid_0's binary_logloss: 0.179735\n",
      "[434]\tvalid_0's binary_logloss: 0.179688\n",
      "[435]\tvalid_0's binary_logloss: 0.179623\n",
      "[436]\tvalid_0's binary_logloss: 0.17952\n",
      "[437]\tvalid_0's binary_logloss: 0.179481\n",
      "[438]\tvalid_0's binary_logloss: 0.179446\n",
      "[439]\tvalid_0's binary_logloss: 0.17941\n",
      "[440]\tvalid_0's binary_logloss: 0.179316\n",
      "[441]\tvalid_0's binary_logloss: 0.179251\n",
      "[442]\tvalid_0's binary_logloss: 0.179177\n",
      "[443]\tvalid_0's binary_logloss: 0.17908\n",
      "[444]\tvalid_0's binary_logloss: 0.179054\n",
      "[445]\tvalid_0's binary_logloss: 0.179052\n",
      "[446]\tvalid_0's binary_logloss: 0.179001\n",
      "[447]\tvalid_0's binary_logloss: 0.17894\n",
      "[448]\tvalid_0's binary_logloss: 0.178927\n",
      "[449]\tvalid_0's binary_logloss: 0.178868\n",
      "[450]\tvalid_0's binary_logloss: 0.17886\n",
      "[451]\tvalid_0's binary_logloss: 0.17879\n",
      "[452]\tvalid_0's binary_logloss: 0.178726\n",
      "[453]\tvalid_0's binary_logloss: 0.178725\n",
      "[454]\tvalid_0's binary_logloss: 0.17862\n",
      "[455]\tvalid_0's binary_logloss: 0.178594\n",
      "[456]\tvalid_0's binary_logloss: 0.178529\n",
      "[457]\tvalid_0's binary_logloss: 0.178523\n",
      "[458]\tvalid_0's binary_logloss: 0.178494\n",
      "[459]\tvalid_0's binary_logloss: 0.178492\n",
      "[460]\tvalid_0's binary_logloss: 0.178452\n",
      "[461]\tvalid_0's binary_logloss: 0.178415\n",
      "[462]\tvalid_0's binary_logloss: 0.178369\n",
      "[463]\tvalid_0's binary_logloss: 0.178321\n",
      "[464]\tvalid_0's binary_logloss: 0.178238\n",
      "[465]\tvalid_0's binary_logloss: 0.178216\n",
      "[466]\tvalid_0's binary_logloss: 0.17816\n",
      "[467]\tvalid_0's binary_logloss: 0.178106\n",
      "[468]\tvalid_0's binary_logloss: 0.178006\n",
      "[469]\tvalid_0's binary_logloss: 0.17795\n",
      "[470]\tvalid_0's binary_logloss: 0.177899\n",
      "[471]\tvalid_0's binary_logloss: 0.177849\n",
      "[472]\tvalid_0's binary_logloss: 0.177755\n",
      "[473]\tvalid_0's binary_logloss: 0.177713\n",
      "[474]\tvalid_0's binary_logloss: 0.177654\n",
      "[475]\tvalid_0's binary_logloss: 0.177651\n",
      "[476]\tvalid_0's binary_logloss: 0.177624\n",
      "[477]\tvalid_0's binary_logloss: 0.177594\n",
      "[478]\tvalid_0's binary_logloss: 0.177565\n",
      "[479]\tvalid_0's binary_logloss: 0.177483\n",
      "[480]\tvalid_0's binary_logloss: 0.17748\n",
      "[481]\tvalid_0's binary_logloss: 0.177469\n",
      "[482]\tvalid_0's binary_logloss: 0.177442\n",
      "[483]\tvalid_0's binary_logloss: 0.177444\n",
      "[484]\tvalid_0's binary_logloss: 0.177376\n",
      "[485]\tvalid_0's binary_logloss: 0.177347\n",
      "[486]\tvalid_0's binary_logloss: 0.177324\n",
      "[487]\tvalid_0's binary_logloss: 0.177297\n",
      "[488]\tvalid_0's binary_logloss: 0.177296\n",
      "[489]\tvalid_0's binary_logloss: 0.177243\n",
      "[490]\tvalid_0's binary_logloss: 0.177236\n",
      "[491]\tvalid_0's binary_logloss: 0.17719\n",
      "[492]\tvalid_0's binary_logloss: 0.177146\n",
      "[493]\tvalid_0's binary_logloss: 0.177114\n",
      "[494]\tvalid_0's binary_logloss: 0.177067\n",
      "[495]\tvalid_0's binary_logloss: 0.177054\n",
      "[496]\tvalid_0's binary_logloss: 0.177012\n",
      "[497]\tvalid_0's binary_logloss: 0.176986\n",
      "[498]\tvalid_0's binary_logloss: 0.176956\n",
      "[499]\tvalid_0's binary_logloss: 0.176908\n",
      "[500]\tvalid_0's binary_logloss: 0.1769\n",
      "[501]\tvalid_0's binary_logloss: 0.176892\n",
      "[502]\tvalid_0's binary_logloss: 0.176866\n",
      "[503]\tvalid_0's binary_logloss: 0.176845\n",
      "[504]\tvalid_0's binary_logloss: 0.176831\n",
      "[505]\tvalid_0's binary_logloss: 0.1768\n",
      "[506]\tvalid_0's binary_logloss: 0.176786\n",
      "[507]\tvalid_0's binary_logloss: 0.176778\n",
      "[508]\tvalid_0's binary_logloss: 0.176757\n",
      "[509]\tvalid_0's binary_logloss: 0.176695\n",
      "[510]\tvalid_0's binary_logloss: 0.176676\n",
      "[511]\tvalid_0's binary_logloss: 0.176628\n",
      "[512]\tvalid_0's binary_logloss: 0.176559\n",
      "[513]\tvalid_0's binary_logloss: 0.176549\n",
      "[514]\tvalid_0's binary_logloss: 0.176484\n",
      "[515]\tvalid_0's binary_logloss: 0.176467\n",
      "[516]\tvalid_0's binary_logloss: 0.176436\n",
      "[517]\tvalid_0's binary_logloss: 0.176345\n",
      "[518]\tvalid_0's binary_logloss: 0.176308\n",
      "[519]\tvalid_0's binary_logloss: 0.176284\n",
      "[520]\tvalid_0's binary_logloss: 0.176205\n",
      "[521]\tvalid_0's binary_logloss: 0.176201\n",
      "[522]\tvalid_0's binary_logloss: 0.176168\n",
      "[523]\tvalid_0's binary_logloss: 0.176154\n",
      "[524]\tvalid_0's binary_logloss: 0.176139\n",
      "[525]\tvalid_0's binary_logloss: 0.176122\n",
      "[526]\tvalid_0's binary_logloss: 0.176065\n",
      "[527]\tvalid_0's binary_logloss: 0.176033\n",
      "[528]\tvalid_0's binary_logloss: 0.175978\n",
      "[529]\tvalid_0's binary_logloss: 0.175969\n",
      "[530]\tvalid_0's binary_logloss: 0.17596\n",
      "[531]\tvalid_0's binary_logloss: 0.175951\n",
      "[532]\tvalid_0's binary_logloss: 0.17592\n",
      "[533]\tvalid_0's binary_logloss: 0.175915\n",
      "[534]\tvalid_0's binary_logloss: 0.175836\n",
      "[535]\tvalid_0's binary_logloss: 0.17576\n",
      "[536]\tvalid_0's binary_logloss: 0.175724\n",
      "[537]\tvalid_0's binary_logloss: 0.175651\n",
      "[538]\tvalid_0's binary_logloss: 0.175631\n",
      "[539]\tvalid_0's binary_logloss: 0.175628\n",
      "[540]\tvalid_0's binary_logloss: 0.175627\n",
      "[541]\tvalid_0's binary_logloss: 0.175591\n",
      "[542]\tvalid_0's binary_logloss: 0.17556\n",
      "[543]\tvalid_0's binary_logloss: 0.175527\n",
      "[544]\tvalid_0's binary_logloss: 0.175503\n",
      "[545]\tvalid_0's binary_logloss: 0.175486\n",
      "[546]\tvalid_0's binary_logloss: 0.175486\n",
      "[547]\tvalid_0's binary_logloss: 0.175463\n",
      "[548]\tvalid_0's binary_logloss: 0.175423\n",
      "[549]\tvalid_0's binary_logloss: 0.175389\n",
      "[550]\tvalid_0's binary_logloss: 0.175386\n",
      "[551]\tvalid_0's binary_logloss: 0.175345\n",
      "[552]\tvalid_0's binary_logloss: 0.175308\n",
      "[553]\tvalid_0's binary_logloss: 0.175252\n",
      "[554]\tvalid_0's binary_logloss: 0.175252\n",
      "[555]\tvalid_0's binary_logloss: 0.175247\n",
      "[556]\tvalid_0's binary_logloss: 0.175188\n",
      "[557]\tvalid_0's binary_logloss: 0.175167\n",
      "[558]\tvalid_0's binary_logloss: 0.175165\n",
      "[559]\tvalid_0's binary_logloss: 0.175111\n",
      "[560]\tvalid_0's binary_logloss: 0.17511\n",
      "[561]\tvalid_0's binary_logloss: 0.175091\n",
      "[562]\tvalid_0's binary_logloss: 0.175089\n",
      "[563]\tvalid_0's binary_logloss: 0.175083\n",
      "[564]\tvalid_0's binary_logloss: 0.175082\n",
      "[565]\tvalid_0's binary_logloss: 0.175087\n",
      "[566]\tvalid_0's binary_logloss: 0.175056\n",
      "[567]\tvalid_0's binary_logloss: 0.174996\n",
      "[568]\tvalid_0's binary_logloss: 0.174973\n",
      "[569]\tvalid_0's binary_logloss: 0.174953\n",
      "[570]\tvalid_0's binary_logloss: 0.174943\n",
      "[571]\tvalid_0's binary_logloss: 0.174939\n",
      "[572]\tvalid_0's binary_logloss: 0.174922\n",
      "[573]\tvalid_0's binary_logloss: 0.174915\n",
      "[574]\tvalid_0's binary_logloss: 0.17488\n",
      "[575]\tvalid_0's binary_logloss: 0.174881\n",
      "[576]\tvalid_0's binary_logloss: 0.174834\n",
      "[577]\tvalid_0's binary_logloss: 0.174828\n",
      "[578]\tvalid_0's binary_logloss: 0.17483\n",
      "[579]\tvalid_0's binary_logloss: 0.174812\n",
      "[580]\tvalid_0's binary_logloss: 0.174816\n",
      "[581]\tvalid_0's binary_logloss: 0.174812\n",
      "[582]\tvalid_0's binary_logloss: 0.174804\n",
      "[583]\tvalid_0's binary_logloss: 0.174751\n",
      "[584]\tvalid_0's binary_logloss: 0.174731\n",
      "[585]\tvalid_0's binary_logloss: 0.174697\n",
      "[586]\tvalid_0's binary_logloss: 0.174669\n",
      "[587]\tvalid_0's binary_logloss: 0.17462\n",
      "[588]\tvalid_0's binary_logloss: 0.174623\n",
      "[589]\tvalid_0's binary_logloss: 0.17462\n",
      "[590]\tvalid_0's binary_logloss: 0.174596\n",
      "[591]\tvalid_0's binary_logloss: 0.174578\n",
      "[592]\tvalid_0's binary_logloss: 0.174548\n",
      "[593]\tvalid_0's binary_logloss: 0.174544\n",
      "[594]\tvalid_0's binary_logloss: 0.17454\n",
      "[595]\tvalid_0's binary_logloss: 0.174519\n",
      "[596]\tvalid_0's binary_logloss: 0.174512\n",
      "[597]\tvalid_0's binary_logloss: 0.174474\n",
      "[598]\tvalid_0's binary_logloss: 0.174459\n",
      "[599]\tvalid_0's binary_logloss: 0.174434\n",
      "[600]\tvalid_0's binary_logloss: 0.17441\n",
      "[601]\tvalid_0's binary_logloss: 0.174393\n",
      "[602]\tvalid_0's binary_logloss: 0.17435\n",
      "[603]\tvalid_0's binary_logloss: 0.174329\n",
      "[604]\tvalid_0's binary_logloss: 0.174311\n",
      "[605]\tvalid_0's binary_logloss: 0.174268\n",
      "[606]\tvalid_0's binary_logloss: 0.174241\n",
      "[607]\tvalid_0's binary_logloss: 0.174204\n",
      "[608]\tvalid_0's binary_logloss: 0.174142\n",
      "[609]\tvalid_0's binary_logloss: 0.174121\n",
      "[610]\tvalid_0's binary_logloss: 0.174065\n",
      "[611]\tvalid_0's binary_logloss: 0.174064\n",
      "[612]\tvalid_0's binary_logloss: 0.174059\n",
      "[613]\tvalid_0's binary_logloss: 0.174064\n",
      "[614]\tvalid_0's binary_logloss: 0.17406\n",
      "[615]\tvalid_0's binary_logloss: 0.17404\n",
      "[616]\tvalid_0's binary_logloss: 0.174017\n",
      "[617]\tvalid_0's binary_logloss: 0.173969\n",
      "[618]\tvalid_0's binary_logloss: 0.173918\n",
      "[619]\tvalid_0's binary_logloss: 0.173893\n",
      "[620]\tvalid_0's binary_logloss: 0.173881\n",
      "[621]\tvalid_0's binary_logloss: 0.173846\n",
      "[622]\tvalid_0's binary_logloss: 0.173844\n",
      "[623]\tvalid_0's binary_logloss: 0.17383\n",
      "[624]\tvalid_0's binary_logloss: 0.17379\n",
      "[625]\tvalid_0's binary_logloss: 0.173755\n",
      "[626]\tvalid_0's binary_logloss: 0.173726\n",
      "[627]\tvalid_0's binary_logloss: 0.173708\n",
      "[628]\tvalid_0's binary_logloss: 0.173703\n",
      "[629]\tvalid_0's binary_logloss: 0.173677\n",
      "[630]\tvalid_0's binary_logloss: 0.173657\n",
      "[631]\tvalid_0's binary_logloss: 0.173658\n",
      "[632]\tvalid_0's binary_logloss: 0.173617\n",
      "[633]\tvalid_0's binary_logloss: 0.173611\n",
      "[634]\tvalid_0's binary_logloss: 0.173593\n",
      "[635]\tvalid_0's binary_logloss: 0.173582\n",
      "[636]\tvalid_0's binary_logloss: 0.173549\n",
      "[637]\tvalid_0's binary_logloss: 0.173519\n",
      "[638]\tvalid_0's binary_logloss: 0.173489\n",
      "[639]\tvalid_0's binary_logloss: 0.173435\n",
      "[640]\tvalid_0's binary_logloss: 0.173404\n",
      "[641]\tvalid_0's binary_logloss: 0.173367\n",
      "[642]\tvalid_0's binary_logloss: 0.173326\n",
      "[643]\tvalid_0's binary_logloss: 0.173307\n",
      "[644]\tvalid_0's binary_logloss: 0.173291\n",
      "[645]\tvalid_0's binary_logloss: 0.173279\n",
      "[646]\tvalid_0's binary_logloss: 0.173261\n",
      "[647]\tvalid_0's binary_logloss: 0.173244\n",
      "[648]\tvalid_0's binary_logloss: 0.173236\n",
      "[649]\tvalid_0's binary_logloss: 0.173236\n",
      "[650]\tvalid_0's binary_logloss: 0.173227\n",
      "[651]\tvalid_0's binary_logloss: 0.173205\n",
      "[652]\tvalid_0's binary_logloss: 0.173197\n",
      "[653]\tvalid_0's binary_logloss: 0.173186\n",
      "[654]\tvalid_0's binary_logloss: 0.173179\n",
      "[655]\tvalid_0's binary_logloss: 0.17313\n",
      "[656]\tvalid_0's binary_logloss: 0.173131\n",
      "[657]\tvalid_0's binary_logloss: 0.173111\n",
      "[658]\tvalid_0's binary_logloss: 0.1731\n",
      "[659]\tvalid_0's binary_logloss: 0.173081\n",
      "[660]\tvalid_0's binary_logloss: 0.173062\n",
      "[661]\tvalid_0's binary_logloss: 0.173052\n",
      "[662]\tvalid_0's binary_logloss: 0.173045\n",
      "[663]\tvalid_0's binary_logloss: 0.173029\n",
      "[664]\tvalid_0's binary_logloss: 0.173014\n",
      "[665]\tvalid_0's binary_logloss: 0.172995\n",
      "[666]\tvalid_0's binary_logloss: 0.172991\n",
      "[667]\tvalid_0's binary_logloss: 0.17298\n",
      "[668]\tvalid_0's binary_logloss: 0.172971\n",
      "[669]\tvalid_0's binary_logloss: 0.172954\n",
      "[670]\tvalid_0's binary_logloss: 0.172923\n",
      "[671]\tvalid_0's binary_logloss: 0.172892\n",
      "[672]\tvalid_0's binary_logloss: 0.172893\n",
      "[673]\tvalid_0's binary_logloss: 0.172861\n",
      "[674]\tvalid_0's binary_logloss: 0.172832\n",
      "[675]\tvalid_0's binary_logloss: 0.172809\n",
      "[676]\tvalid_0's binary_logloss: 0.17276\n",
      "[677]\tvalid_0's binary_logloss: 0.172732\n",
      "[678]\tvalid_0's binary_logloss: 0.17273\n",
      "[679]\tvalid_0's binary_logloss: 0.172728\n",
      "[680]\tvalid_0's binary_logloss: 0.172724\n",
      "[681]\tvalid_0's binary_logloss: 0.172714\n",
      "[682]\tvalid_0's binary_logloss: 0.172692\n",
      "[683]\tvalid_0's binary_logloss: 0.172641\n",
      "[684]\tvalid_0's binary_logloss: 0.172607\n",
      "[685]\tvalid_0's binary_logloss: 0.172589\n",
      "[686]\tvalid_0's binary_logloss: 0.172574\n",
      "[687]\tvalid_0's binary_logloss: 0.172547\n",
      "[688]\tvalid_0's binary_logloss: 0.172539\n",
      "[689]\tvalid_0's binary_logloss: 0.172531\n",
      "[690]\tvalid_0's binary_logloss: 0.172509\n",
      "[691]\tvalid_0's binary_logloss: 0.172503\n",
      "[692]\tvalid_0's binary_logloss: 0.172487\n",
      "[693]\tvalid_0's binary_logloss: 0.172471\n",
      "[694]\tvalid_0's binary_logloss: 0.172458\n",
      "[695]\tvalid_0's binary_logloss: 0.172453\n",
      "[696]\tvalid_0's binary_logloss: 0.172423\n",
      "[697]\tvalid_0's binary_logloss: 0.172417\n",
      "[698]\tvalid_0's binary_logloss: 0.172392\n",
      "[699]\tvalid_0's binary_logloss: 0.172367\n",
      "[700]\tvalid_0's binary_logloss: 0.172359\n",
      "[701]\tvalid_0's binary_logloss: 0.17236\n",
      "[702]\tvalid_0's binary_logloss: 0.172349\n",
      "[703]\tvalid_0's binary_logloss: 0.172312\n",
      "[704]\tvalid_0's binary_logloss: 0.172294\n",
      "[705]\tvalid_0's binary_logloss: 0.172282\n",
      "[706]\tvalid_0's binary_logloss: 0.172245\n",
      "[707]\tvalid_0's binary_logloss: 0.172213\n",
      "[708]\tvalid_0's binary_logloss: 0.172189\n",
      "[709]\tvalid_0's binary_logloss: 0.172192\n",
      "[710]\tvalid_0's binary_logloss: 0.172192\n",
      "[711]\tvalid_0's binary_logloss: 0.172172\n",
      "[712]\tvalid_0's binary_logloss: 0.172171\n",
      "[713]\tvalid_0's binary_logloss: 0.172155\n",
      "[714]\tvalid_0's binary_logloss: 0.172126\n",
      "[715]\tvalid_0's binary_logloss: 0.172124\n",
      "[716]\tvalid_0's binary_logloss: 0.172106\n",
      "[717]\tvalid_0's binary_logloss: 0.172085\n",
      "[718]\tvalid_0's binary_logloss: 0.172051\n",
      "[719]\tvalid_0's binary_logloss: 0.172027\n",
      "[720]\tvalid_0's binary_logloss: 0.172027\n",
      "[721]\tvalid_0's binary_logloss: 0.172016\n",
      "[722]\tvalid_0's binary_logloss: 0.172004\n",
      "[723]\tvalid_0's binary_logloss: 0.172005\n",
      "[724]\tvalid_0's binary_logloss: 0.172002\n",
      "[725]\tvalid_0's binary_logloss: 0.171988\n",
      "[726]\tvalid_0's binary_logloss: 0.171986\n",
      "[727]\tvalid_0's binary_logloss: 0.171976\n",
      "[728]\tvalid_0's binary_logloss: 0.171948\n",
      "[729]\tvalid_0's binary_logloss: 0.171934\n",
      "[730]\tvalid_0's binary_logloss: 0.171917\n",
      "[731]\tvalid_0's binary_logloss: 0.171899\n",
      "[732]\tvalid_0's binary_logloss: 0.171883\n",
      "[733]\tvalid_0's binary_logloss: 0.171888\n",
      "[734]\tvalid_0's binary_logloss: 0.171895\n",
      "[735]\tvalid_0's binary_logloss: 0.171869\n",
      "[736]\tvalid_0's binary_logloss: 0.171868\n",
      "[737]\tvalid_0's binary_logloss: 0.171845\n",
      "[738]\tvalid_0's binary_logloss: 0.171822\n",
      "[739]\tvalid_0's binary_logloss: 0.171791\n",
      "[740]\tvalid_0's binary_logloss: 0.171784\n",
      "[741]\tvalid_0's binary_logloss: 0.171765\n",
      "[742]\tvalid_0's binary_logloss: 0.171759\n",
      "[743]\tvalid_0's binary_logloss: 0.171752\n",
      "[744]\tvalid_0's binary_logloss: 0.171746\n",
      "[745]\tvalid_0's binary_logloss: 0.171735\n",
      "[746]\tvalid_0's binary_logloss: 0.171727\n",
      "[747]\tvalid_0's binary_logloss: 0.171697\n",
      "[748]\tvalid_0's binary_logloss: 0.171696\n",
      "[749]\tvalid_0's binary_logloss: 0.171691\n",
      "[750]\tvalid_0's binary_logloss: 0.171668\n",
      "[751]\tvalid_0's binary_logloss: 0.171645\n",
      "[752]\tvalid_0's binary_logloss: 0.171618\n",
      "[753]\tvalid_0's binary_logloss: 0.171605\n",
      "[754]\tvalid_0's binary_logloss: 0.171577\n",
      "[755]\tvalid_0's binary_logloss: 0.171573\n",
      "[756]\tvalid_0's binary_logloss: 0.171553\n",
      "[757]\tvalid_0's binary_logloss: 0.171547\n",
      "[758]\tvalid_0's binary_logloss: 0.17151\n",
      "[759]\tvalid_0's binary_logloss: 0.171511\n",
      "[760]\tvalid_0's binary_logloss: 0.171503\n",
      "[761]\tvalid_0's binary_logloss: 0.171489\n",
      "[762]\tvalid_0's binary_logloss: 0.171449\n",
      "[763]\tvalid_0's binary_logloss: 0.171438\n",
      "[764]\tvalid_0's binary_logloss: 0.17141\n",
      "[765]\tvalid_0's binary_logloss: 0.171413\n",
      "[766]\tvalid_0's binary_logloss: 0.171404\n",
      "[767]\tvalid_0's binary_logloss: 0.171401\n",
      "[768]\tvalid_0's binary_logloss: 0.17138\n",
      "[769]\tvalid_0's binary_logloss: 0.171358\n",
      "[770]\tvalid_0's binary_logloss: 0.171342\n",
      "[771]\tvalid_0's binary_logloss: 0.171322\n",
      "[772]\tvalid_0's binary_logloss: 0.171313\n",
      "[773]\tvalid_0's binary_logloss: 0.171298\n",
      "[774]\tvalid_0's binary_logloss: 0.171269\n",
      "[775]\tvalid_0's binary_logloss: 0.171268\n",
      "[776]\tvalid_0's binary_logloss: 0.171266\n",
      "[777]\tvalid_0's binary_logloss: 0.171264\n",
      "[778]\tvalid_0's binary_logloss: 0.171262\n",
      "[779]\tvalid_0's binary_logloss: 0.171258\n",
      "[780]\tvalid_0's binary_logloss: 0.171238\n",
      "[781]\tvalid_0's binary_logloss: 0.171228\n",
      "[782]\tvalid_0's binary_logloss: 0.171223\n",
      "[783]\tvalid_0's binary_logloss: 0.171213\n",
      "[784]\tvalid_0's binary_logloss: 0.171211\n",
      "[785]\tvalid_0's binary_logloss: 0.171204\n",
      "[786]\tvalid_0's binary_logloss: 0.171197\n",
      "[787]\tvalid_0's binary_logloss: 0.171192\n",
      "[788]\tvalid_0's binary_logloss: 0.17118\n",
      "[789]\tvalid_0's binary_logloss: 0.171183\n",
      "[790]\tvalid_0's binary_logloss: 0.171174\n",
      "[791]\tvalid_0's binary_logloss: 0.171165\n",
      "[792]\tvalid_0's binary_logloss: 0.171146\n",
      "[793]\tvalid_0's binary_logloss: 0.171136\n",
      "[794]\tvalid_0's binary_logloss: 0.171122\n",
      "[795]\tvalid_0's binary_logloss: 0.171103\n",
      "[796]\tvalid_0's binary_logloss: 0.171053\n",
      "[797]\tvalid_0's binary_logloss: 0.171047\n",
      "[798]\tvalid_0's binary_logloss: 0.171034\n",
      "[799]\tvalid_0's binary_logloss: 0.171019\n",
      "[800]\tvalid_0's binary_logloss: 0.171014\n",
      "[801]\tvalid_0's binary_logloss: 0.171016\n",
      "[802]\tvalid_0's binary_logloss: 0.171008\n",
      "[803]\tvalid_0's binary_logloss: 0.17101\n",
      "[804]\tvalid_0's binary_logloss: 0.170993\n",
      "[805]\tvalid_0's binary_logloss: 0.170959\n",
      "[806]\tvalid_0's binary_logloss: 0.170932\n",
      "[807]\tvalid_0's binary_logloss: 0.170916\n",
      "[808]\tvalid_0's binary_logloss: 0.170904\n",
      "[809]\tvalid_0's binary_logloss: 0.170901\n",
      "[810]\tvalid_0's binary_logloss: 0.170873\n",
      "[811]\tvalid_0's binary_logloss: 0.170867\n",
      "[812]\tvalid_0's binary_logloss: 0.170862\n",
      "[813]\tvalid_0's binary_logloss: 0.170828\n",
      "[814]\tvalid_0's binary_logloss: 0.170821\n",
      "[815]\tvalid_0's binary_logloss: 0.170814\n",
      "[816]\tvalid_0's binary_logloss: 0.170788\n",
      "[817]\tvalid_0's binary_logloss: 0.170757\n",
      "[818]\tvalid_0's binary_logloss: 0.17073\n",
      "[819]\tvalid_0's binary_logloss: 0.170716\n",
      "[820]\tvalid_0's binary_logloss: 0.170697\n",
      "[821]\tvalid_0's binary_logloss: 0.170694\n",
      "[822]\tvalid_0's binary_logloss: 0.170684\n",
      "[823]\tvalid_0's binary_logloss: 0.170671\n",
      "[824]\tvalid_0's binary_logloss: 0.170654\n",
      "[825]\tvalid_0's binary_logloss: 0.170648\n",
      "[826]\tvalid_0's binary_logloss: 0.170625\n",
      "[827]\tvalid_0's binary_logloss: 0.170605\n",
      "[828]\tvalid_0's binary_logloss: 0.17059\n",
      "[829]\tvalid_0's binary_logloss: 0.170573\n",
      "[830]\tvalid_0's binary_logloss: 0.170549\n",
      "[831]\tvalid_0's binary_logloss: 0.170521\n",
      "[832]\tvalid_0's binary_logloss: 0.170521\n",
      "[833]\tvalid_0's binary_logloss: 0.1705\n",
      "[834]\tvalid_0's binary_logloss: 0.170482\n",
      "[835]\tvalid_0's binary_logloss: 0.170476\n",
      "[836]\tvalid_0's binary_logloss: 0.170478\n",
      "[837]\tvalid_0's binary_logloss: 0.170467\n",
      "[838]\tvalid_0's binary_logloss: 0.170453\n",
      "[839]\tvalid_0's binary_logloss: 0.170441\n",
      "[840]\tvalid_0's binary_logloss: 0.170433\n",
      "[841]\tvalid_0's binary_logloss: 0.170432\n",
      "[842]\tvalid_0's binary_logloss: 0.170401\n",
      "[843]\tvalid_0's binary_logloss: 0.170399\n",
      "[844]\tvalid_0's binary_logloss: 0.170401\n",
      "[845]\tvalid_0's binary_logloss: 0.170388\n",
      "[846]\tvalid_0's binary_logloss: 0.170392\n",
      "[847]\tvalid_0's binary_logloss: 0.170369\n",
      "[848]\tvalid_0's binary_logloss: 0.170362\n",
      "[849]\tvalid_0's binary_logloss: 0.170353\n",
      "[850]\tvalid_0's binary_logloss: 0.17033\n",
      "[851]\tvalid_0's binary_logloss: 0.170313\n",
      "[852]\tvalid_0's binary_logloss: 0.170311\n",
      "[853]\tvalid_0's binary_logloss: 0.170292\n",
      "[854]\tvalid_0's binary_logloss: 0.170283\n",
      "[855]\tvalid_0's binary_logloss: 0.170272\n",
      "[856]\tvalid_0's binary_logloss: 0.170267\n",
      "[857]\tvalid_0's binary_logloss: 0.17027\n",
      "[858]\tvalid_0's binary_logloss: 0.170254\n",
      "[859]\tvalid_0's binary_logloss: 0.170239\n",
      "[860]\tvalid_0's binary_logloss: 0.170209\n",
      "[861]\tvalid_0's binary_logloss: 0.170187\n",
      "[862]\tvalid_0's binary_logloss: 0.170165\n",
      "[863]\tvalid_0's binary_logloss: 0.170164\n",
      "[864]\tvalid_0's binary_logloss: 0.170153\n",
      "[865]\tvalid_0's binary_logloss: 0.170132\n",
      "[866]\tvalid_0's binary_logloss: 0.170124\n",
      "[867]\tvalid_0's binary_logloss: 0.170122\n",
      "[868]\tvalid_0's binary_logloss: 0.170104\n",
      "[869]\tvalid_0's binary_logloss: 0.170087\n",
      "[870]\tvalid_0's binary_logloss: 0.17008\n",
      "[871]\tvalid_0's binary_logloss: 0.170067\n",
      "[872]\tvalid_0's binary_logloss: 0.170058\n",
      "[873]\tvalid_0's binary_logloss: 0.170046\n",
      "[874]\tvalid_0's binary_logloss: 0.170029\n",
      "[875]\tvalid_0's binary_logloss: 0.170018\n",
      "[876]\tvalid_0's binary_logloss: 0.169983\n",
      "[877]\tvalid_0's binary_logloss: 0.169972\n",
      "[878]\tvalid_0's binary_logloss: 0.169959\n",
      "[879]\tvalid_0's binary_logloss: 0.169932\n",
      "[880]\tvalid_0's binary_logloss: 0.16993\n",
      "[881]\tvalid_0's binary_logloss: 0.169916\n",
      "[882]\tvalid_0's binary_logloss: 0.169915\n",
      "[883]\tvalid_0's binary_logloss: 0.169905\n",
      "[884]\tvalid_0's binary_logloss: 0.1699\n",
      "[885]\tvalid_0's binary_logloss: 0.169894\n",
      "[886]\tvalid_0's binary_logloss: 0.16988\n",
      "[887]\tvalid_0's binary_logloss: 0.169849\n",
      "[888]\tvalid_0's binary_logloss: 0.16985\n",
      "[889]\tvalid_0's binary_logloss: 0.169839\n",
      "[890]\tvalid_0's binary_logloss: 0.169834\n",
      "[891]\tvalid_0's binary_logloss: 0.169832\n",
      "[892]\tvalid_0's binary_logloss: 0.169815\n",
      "[893]\tvalid_0's binary_logloss: 0.16982\n",
      "[894]\tvalid_0's binary_logloss: 0.169818\n",
      "[895]\tvalid_0's binary_logloss: 0.16982\n",
      "[896]\tvalid_0's binary_logloss: 0.169811\n",
      "[897]\tvalid_0's binary_logloss: 0.169807\n",
      "[898]\tvalid_0's binary_logloss: 0.169793\n",
      "[899]\tvalid_0's binary_logloss: 0.169792\n",
      "[900]\tvalid_0's binary_logloss: 0.169763\n",
      "[901]\tvalid_0's binary_logloss: 0.169755\n",
      "[902]\tvalid_0's binary_logloss: 0.169743\n",
      "[903]\tvalid_0's binary_logloss: 0.169716\n",
      "[904]\tvalid_0's binary_logloss: 0.169699\n",
      "[905]\tvalid_0's binary_logloss: 0.169696\n",
      "[906]\tvalid_0's binary_logloss: 0.169689\n",
      "[907]\tvalid_0's binary_logloss: 0.169682\n",
      "[908]\tvalid_0's binary_logloss: 0.169678\n",
      "[909]\tvalid_0's binary_logloss: 0.169681\n",
      "[910]\tvalid_0's binary_logloss: 0.169672\n",
      "[911]\tvalid_0's binary_logloss: 0.169666\n",
      "[912]\tvalid_0's binary_logloss: 0.169665\n",
      "[913]\tvalid_0's binary_logloss: 0.169666\n",
      "[914]\tvalid_0's binary_logloss: 0.169652\n",
      "[915]\tvalid_0's binary_logloss: 0.169651\n",
      "[916]\tvalid_0's binary_logloss: 0.169642\n",
      "[917]\tvalid_0's binary_logloss: 0.16965\n",
      "[918]\tvalid_0's binary_logloss: 0.16964\n",
      "[919]\tvalid_0's binary_logloss: 0.169622\n",
      "[920]\tvalid_0's binary_logloss: 0.169616\n",
      "[921]\tvalid_0's binary_logloss: 0.169606\n",
      "[922]\tvalid_0's binary_logloss: 0.169597\n",
      "[923]\tvalid_0's binary_logloss: 0.169588\n",
      "[924]\tvalid_0's binary_logloss: 0.169578\n",
      "[925]\tvalid_0's binary_logloss: 0.169563\n",
      "[926]\tvalid_0's binary_logloss: 0.169557\n",
      "[927]\tvalid_0's binary_logloss: 0.16953\n",
      "[928]\tvalid_0's binary_logloss: 0.169528\n",
      "[929]\tvalid_0's binary_logloss: 0.169524\n",
      "[930]\tvalid_0's binary_logloss: 0.169516\n",
      "[931]\tvalid_0's binary_logloss: 0.169517\n",
      "[932]\tvalid_0's binary_logloss: 0.169511\n",
      "[933]\tvalid_0's binary_logloss: 0.169511\n",
      "[934]\tvalid_0's binary_logloss: 0.169498\n",
      "[935]\tvalid_0's binary_logloss: 0.169496\n",
      "[936]\tvalid_0's binary_logloss: 0.169483\n",
      "[937]\tvalid_0's binary_logloss: 0.169469\n",
      "[938]\tvalid_0's binary_logloss: 0.169467\n",
      "[939]\tvalid_0's binary_logloss: 0.169466\n",
      "[940]\tvalid_0's binary_logloss: 0.16946\n",
      "[941]\tvalid_0's binary_logloss: 0.169465\n",
      "[942]\tvalid_0's binary_logloss: 0.169458\n",
      "[943]\tvalid_0's binary_logloss: 0.169438\n",
      "[944]\tvalid_0's binary_logloss: 0.169431\n",
      "[945]\tvalid_0's binary_logloss: 0.169405\n",
      "[946]\tvalid_0's binary_logloss: 0.169382\n",
      "[947]\tvalid_0's binary_logloss: 0.169361\n",
      "[948]\tvalid_0's binary_logloss: 0.169327\n",
      "[949]\tvalid_0's binary_logloss: 0.169331\n",
      "[950]\tvalid_0's binary_logloss: 0.169315\n",
      "[951]\tvalid_0's binary_logloss: 0.169317\n",
      "[952]\tvalid_0's binary_logloss: 0.169316\n",
      "[953]\tvalid_0's binary_logloss: 0.169312\n",
      "[954]\tvalid_0's binary_logloss: 0.169299\n",
      "[955]\tvalid_0's binary_logloss: 0.169303\n",
      "[956]\tvalid_0's binary_logloss: 0.1693\n",
      "[957]\tvalid_0's binary_logloss: 0.169261\n",
      "[958]\tvalid_0's binary_logloss: 0.169245\n",
      "[959]\tvalid_0's binary_logloss: 0.169226\n",
      "[960]\tvalid_0's binary_logloss: 0.169205\n",
      "[961]\tvalid_0's binary_logloss: 0.169201\n",
      "[962]\tvalid_0's binary_logloss: 0.16919\n",
      "[963]\tvalid_0's binary_logloss: 0.169184\n",
      "[964]\tvalid_0's binary_logloss: 0.169159\n",
      "[965]\tvalid_0's binary_logloss: 0.169158\n",
      "[966]\tvalid_0's binary_logloss: 0.169151\n",
      "[967]\tvalid_0's binary_logloss: 0.169137\n",
      "[968]\tvalid_0's binary_logloss: 0.16912\n",
      "[969]\tvalid_0's binary_logloss: 0.169112\n",
      "[970]\tvalid_0's binary_logloss: 0.169096\n",
      "[971]\tvalid_0's binary_logloss: 0.169092\n",
      "[972]\tvalid_0's binary_logloss: 0.169083\n",
      "[973]\tvalid_0's binary_logloss: 0.169065\n",
      "[974]\tvalid_0's binary_logloss: 0.169058\n",
      "[975]\tvalid_0's binary_logloss: 0.16905\n",
      "[976]\tvalid_0's binary_logloss: 0.169033\n",
      "[977]\tvalid_0's binary_logloss: 0.169029\n",
      "[978]\tvalid_0's binary_logloss: 0.169015\n",
      "[979]\tvalid_0's binary_logloss: 0.169005\n",
      "[980]\tvalid_0's binary_logloss: 0.168993\n",
      "[981]\tvalid_0's binary_logloss: 0.168977\n",
      "[982]\tvalid_0's binary_logloss: 0.168979\n",
      "[983]\tvalid_0's binary_logloss: 0.168977\n",
      "[984]\tvalid_0's binary_logloss: 0.168977\n",
      "[985]\tvalid_0's binary_logloss: 0.168963\n",
      "[986]\tvalid_0's binary_logloss: 0.168956\n",
      "[987]\tvalid_0's binary_logloss: 0.16893\n",
      "[988]\tvalid_0's binary_logloss: 0.168921\n",
      "[989]\tvalid_0's binary_logloss: 0.168908\n",
      "[990]\tvalid_0's binary_logloss: 0.16889\n",
      "[991]\tvalid_0's binary_logloss: 0.168873\n",
      "[992]\tvalid_0's binary_logloss: 0.168871\n",
      "[993]\tvalid_0's binary_logloss: 0.168861\n",
      "[994]\tvalid_0's binary_logloss: 0.168861\n",
      "[995]\tvalid_0's binary_logloss: 0.168853\n",
      "[996]\tvalid_0's binary_logloss: 0.168847\n",
      "[997]\tvalid_0's binary_logloss: 0.168844\n",
      "[998]\tvalid_0's binary_logloss: 0.168847\n",
      "[999]\tvalid_0's binary_logloss: 0.168846\n",
      "[1000]\tvalid_0's binary_logloss: 0.168843\n",
      "training model for CV #3\n",
      "[LightGBM] [Warning] feature_fraction is set=0.4, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.4\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.0, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.0\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8772503866491428, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8772503866491428\n",
      "[LightGBM] [Warning] lambda_l2 is set=0.0, reg_lambda=0.0 will be ignored. Current value: lambda_l2=0.0\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\engine.py:177: UserWarning: Found `num_iterations` in params. Will use it instead of argument\n",
      "  _log_warning(f\"Found `{alias}` in params. Will use it instead of argument\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tvalid_0's binary_logloss: 0.66232\n",
      "[2]\tvalid_0's binary_logloss: 0.63979\n",
      "[3]\tvalid_0's binary_logloss: 0.625949\n",
      "[4]\tvalid_0's binary_logloss: 0.60899\n",
      "[5]\tvalid_0's binary_logloss: 0.598662\n",
      "[6]\tvalid_0's binary_logloss: 0.574758\n",
      "[7]\tvalid_0's binary_logloss: 0.556323\n",
      "[8]\tvalid_0's binary_logloss: 0.548115\n",
      "[9]\tvalid_0's binary_logloss: 0.528458\n",
      "[10]\tvalid_0's binary_logloss: 0.51519\n",
      "[11]\tvalid_0's binary_logloss: 0.508062\n",
      "[12]\tvalid_0's binary_logloss: 0.493139\n",
      "[13]\tvalid_0's binary_logloss: 0.481725\n",
      "[14]\tvalid_0's binary_logloss: 0.475906\n",
      "[15]\tvalid_0's binary_logloss: 0.466219\n",
      "[16]\tvalid_0's binary_logloss: 0.457442\n",
      "[17]\tvalid_0's binary_logloss: 0.448582\n",
      "[18]\tvalid_0's binary_logloss: 0.438657\n",
      "[19]\tvalid_0's binary_logloss: 0.431712\n",
      "[20]\tvalid_0's binary_logloss: 0.427977\n",
      "[21]\tvalid_0's binary_logloss: 0.419918\n",
      "[22]\tvalid_0's binary_logloss: 0.416089\n",
      "[23]\tvalid_0's binary_logloss: 0.413272\n",
      "[24]\tvalid_0's binary_logloss: 0.407985\n",
      "[25]\tvalid_0's binary_logloss: 0.399814\n",
      "[26]\tvalid_0's binary_logloss: 0.395435\n",
      "[27]\tvalid_0's binary_logloss: 0.388426\n",
      "[28]\tvalid_0's binary_logloss: 0.382607\n",
      "[29]\tvalid_0's binary_logloss: 0.379374\n",
      "[30]\tvalid_0's binary_logloss: 0.37721\n",
      "[31]\tvalid_0's binary_logloss: 0.373752\n",
      "[32]\tvalid_0's binary_logloss: 0.368141\n",
      "[33]\tvalid_0's binary_logloss: 0.365978\n",
      "[34]\tvalid_0's binary_logloss: 0.363647\n",
      "[35]\tvalid_0's binary_logloss: 0.361719\n",
      "[36]\tvalid_0's binary_logloss: 0.359869\n",
      "[37]\tvalid_0's binary_logloss: 0.357185\n",
      "[38]\tvalid_0's binary_logloss: 0.35211\n",
      "[39]\tvalid_0's binary_logloss: 0.349796\n",
      "[40]\tvalid_0's binary_logloss: 0.34806\n",
      "[41]\tvalid_0's binary_logloss: 0.346878\n",
      "[42]\tvalid_0's binary_logloss: 0.345434\n",
      "[43]\tvalid_0's binary_logloss: 0.344325\n",
      "[44]\tvalid_0's binary_logloss: 0.340478\n",
      "[45]\tvalid_0's binary_logloss: 0.338197\n",
      "[46]\tvalid_0's binary_logloss: 0.336662\n",
      "[47]\tvalid_0's binary_logloss: 0.332756\n",
      "[48]\tvalid_0's binary_logloss: 0.331371\n",
      "[49]\tvalid_0's binary_logloss: 0.330284\n",
      "[50]\tvalid_0's binary_logloss: 0.328485\n",
      "[51]\tvalid_0's binary_logloss: 0.323842\n",
      "[52]\tvalid_0's binary_logloss: 0.320908\n",
      "[53]\tvalid_0's binary_logloss: 0.317462\n",
      "[54]\tvalid_0's binary_logloss: 0.313576\n",
      "[55]\tvalid_0's binary_logloss: 0.312601\n",
      "[56]\tvalid_0's binary_logloss: 0.309891\n",
      "[57]\tvalid_0's binary_logloss: 0.308502\n",
      "[58]\tvalid_0's binary_logloss: 0.305597\n",
      "[59]\tvalid_0's binary_logloss: 0.302979\n",
      "[60]\tvalid_0's binary_logloss: 0.301175\n",
      "[61]\tvalid_0's binary_logloss: 0.299684\n",
      "[62]\tvalid_0's binary_logloss: 0.298601\n",
      "[63]\tvalid_0's binary_logloss: 0.295419\n",
      "[64]\tvalid_0's binary_logloss: 0.294643\n",
      "[65]\tvalid_0's binary_logloss: 0.293261\n",
      "[66]\tvalid_0's binary_logloss: 0.291699\n",
      "[67]\tvalid_0's binary_logloss: 0.289786\n",
      "[68]\tvalid_0's binary_logloss: 0.288226\n",
      "[69]\tvalid_0's binary_logloss: 0.287526\n",
      "[70]\tvalid_0's binary_logloss: 0.286482\n",
      "[71]\tvalid_0's binary_logloss: 0.285636\n",
      "[72]\tvalid_0's binary_logloss: 0.282989\n",
      "[73]\tvalid_0's binary_logloss: 0.28244\n",
      "[74]\tvalid_0's binary_logloss: 0.28087\n",
      "[75]\tvalid_0's binary_logloss: 0.277345\n",
      "[76]\tvalid_0's binary_logloss: 0.276666\n",
      "[77]\tvalid_0's binary_logloss: 0.273391\n",
      "[78]\tvalid_0's binary_logloss: 0.270804\n",
      "[79]\tvalid_0's binary_logloss: 0.26957\n",
      "[80]\tvalid_0's binary_logloss: 0.267692\n",
      "[81]\tvalid_0's binary_logloss: 0.266494\n",
      "[82]\tvalid_0's binary_logloss: 0.265817\n",
      "[83]\tvalid_0's binary_logloss: 0.264201\n",
      "[84]\tvalid_0's binary_logloss: 0.263256\n",
      "[85]\tvalid_0's binary_logloss: 0.260662\n",
      "[86]\tvalid_0's binary_logloss: 0.258581\n",
      "[87]\tvalid_0's binary_logloss: 0.257208\n",
      "[88]\tvalid_0's binary_logloss: 0.256466\n",
      "[89]\tvalid_0's binary_logloss: 0.255476\n",
      "[90]\tvalid_0's binary_logloss: 0.25481\n",
      "[91]\tvalid_0's binary_logloss: 0.253989\n",
      "[92]\tvalid_0's binary_logloss: 0.251813\n",
      "[93]\tvalid_0's binary_logloss: 0.250898\n",
      "[94]\tvalid_0's binary_logloss: 0.249357\n",
      "[95]\tvalid_0's binary_logloss: 0.248314\n",
      "[96]\tvalid_0's binary_logloss: 0.246881\n",
      "[97]\tvalid_0's binary_logloss: 0.245916\n",
      "[98]\tvalid_0's binary_logloss: 0.244498\n",
      "[99]\tvalid_0's binary_logloss: 0.243932\n",
      "[100]\tvalid_0's binary_logloss: 0.24315\n",
      "[101]\tvalid_0's binary_logloss: 0.242549\n",
      "[102]\tvalid_0's binary_logloss: 0.241092\n",
      "[103]\tvalid_0's binary_logloss: 0.240472\n",
      "[104]\tvalid_0's binary_logloss: 0.239257\n",
      "[105]\tvalid_0's binary_logloss: 0.238568\n",
      "[106]\tvalid_0's binary_logloss: 0.237984\n",
      "[107]\tvalid_0's binary_logloss: 0.237413\n",
      "[108]\tvalid_0's binary_logloss: 0.236722\n",
      "[109]\tvalid_0's binary_logloss: 0.236463\n",
      "[110]\tvalid_0's binary_logloss: 0.235821\n",
      "[111]\tvalid_0's binary_logloss: 0.235016\n",
      "[112]\tvalid_0's binary_logloss: 0.23412\n",
      "[113]\tvalid_0's binary_logloss: 0.233404\n",
      "[114]\tvalid_0's binary_logloss: 0.232876\n",
      "[115]\tvalid_0's binary_logloss: 0.232409\n",
      "[116]\tvalid_0's binary_logloss: 0.232016\n",
      "[117]\tvalid_0's binary_logloss: 0.231631\n",
      "[118]\tvalid_0's binary_logloss: 0.231203\n",
      "[119]\tvalid_0's binary_logloss: 0.23046\n",
      "[120]\tvalid_0's binary_logloss: 0.229887\n",
      "[121]\tvalid_0's binary_logloss: 0.229459\n",
      "[122]\tvalid_0's binary_logloss: 0.228534\n",
      "[123]\tvalid_0's binary_logloss: 0.22765\n",
      "[124]\tvalid_0's binary_logloss: 0.227233\n",
      "[125]\tvalid_0's binary_logloss: 0.226851\n",
      "[126]\tvalid_0's binary_logloss: 0.226438\n",
      "[127]\tvalid_0's binary_logloss: 0.225783\n",
      "[128]\tvalid_0's binary_logloss: 0.224436\n",
      "[129]\tvalid_0's binary_logloss: 0.223505\n",
      "[130]\tvalid_0's binary_logloss: 0.223283\n",
      "[131]\tvalid_0's binary_logloss: 0.222811\n",
      "[132]\tvalid_0's binary_logloss: 0.221912\n",
      "[133]\tvalid_0's binary_logloss: 0.221506\n",
      "[134]\tvalid_0's binary_logloss: 0.221025\n",
      "[135]\tvalid_0's binary_logloss: 0.220408\n",
      "[136]\tvalid_0's binary_logloss: 0.220161\n",
      "[137]\tvalid_0's binary_logloss: 0.219901\n",
      "[138]\tvalid_0's binary_logloss: 0.219629\n",
      "[139]\tvalid_0's binary_logloss: 0.218811\n",
      "[140]\tvalid_0's binary_logloss: 0.218507\n",
      "[141]\tvalid_0's binary_logloss: 0.218266\n",
      "[142]\tvalid_0's binary_logloss: 0.217804\n",
      "[143]\tvalid_0's binary_logloss: 0.217252\n",
      "[144]\tvalid_0's binary_logloss: 0.217089\n",
      "[145]\tvalid_0's binary_logloss: 0.216317\n",
      "[146]\tvalid_0's binary_logloss: 0.215605\n",
      "[147]\tvalid_0's binary_logloss: 0.21529\n",
      "[148]\tvalid_0's binary_logloss: 0.214719\n",
      "[149]\tvalid_0's binary_logloss: 0.214328\n",
      "[150]\tvalid_0's binary_logloss: 0.21368\n",
      "[151]\tvalid_0's binary_logloss: 0.213415\n",
      "[152]\tvalid_0's binary_logloss: 0.213123\n",
      "[153]\tvalid_0's binary_logloss: 0.212943\n",
      "[154]\tvalid_0's binary_logloss: 0.212525\n",
      "[155]\tvalid_0's binary_logloss: 0.212124\n",
      "[156]\tvalid_0's binary_logloss: 0.211632\n",
      "[157]\tvalid_0's binary_logloss: 0.210681\n",
      "[158]\tvalid_0's binary_logloss: 0.210284\n",
      "[159]\tvalid_0's binary_logloss: 0.21011\n",
      "[160]\tvalid_0's binary_logloss: 0.209761\n",
      "[161]\tvalid_0's binary_logloss: 0.209195\n",
      "[162]\tvalid_0's binary_logloss: 0.208879\n",
      "[163]\tvalid_0's binary_logloss: 0.208739\n",
      "[164]\tvalid_0's binary_logloss: 0.208354\n",
      "[165]\tvalid_0's binary_logloss: 0.208003\n",
      "[166]\tvalid_0's binary_logloss: 0.207687\n",
      "[167]\tvalid_0's binary_logloss: 0.207541\n",
      "[168]\tvalid_0's binary_logloss: 0.207079\n",
      "[169]\tvalid_0's binary_logloss: 0.206593\n",
      "[170]\tvalid_0's binary_logloss: 0.206398\n",
      "[171]\tvalid_0's binary_logloss: 0.206244\n",
      "[172]\tvalid_0's binary_logloss: 0.20595\n",
      "[173]\tvalid_0's binary_logloss: 0.205643\n",
      "[174]\tvalid_0's binary_logloss: 0.205389\n",
      "[175]\tvalid_0's binary_logloss: 0.205267\n",
      "[176]\tvalid_0's binary_logloss: 0.205162\n",
      "[177]\tvalid_0's binary_logloss: 0.20502\n",
      "[178]\tvalid_0's binary_logloss: 0.204574\n",
      "[179]\tvalid_0's binary_logloss: 0.204414\n",
      "[180]\tvalid_0's binary_logloss: 0.20419\n",
      "[181]\tvalid_0's binary_logloss: 0.20398\n",
      "[182]\tvalid_0's binary_logloss: 0.203868\n",
      "[183]\tvalid_0's binary_logloss: 0.203713\n",
      "[184]\tvalid_0's binary_logloss: 0.203645\n",
      "[185]\tvalid_0's binary_logloss: 0.203501\n",
      "[186]\tvalid_0's binary_logloss: 0.20334\n",
      "[187]\tvalid_0's binary_logloss: 0.203239\n",
      "[188]\tvalid_0's binary_logloss: 0.202953\n",
      "[189]\tvalid_0's binary_logloss: 0.20275\n",
      "[190]\tvalid_0's binary_logloss: 0.202696\n",
      "[191]\tvalid_0's binary_logloss: 0.202651\n",
      "[192]\tvalid_0's binary_logloss: 0.202289\n",
      "[193]\tvalid_0's binary_logloss: 0.201822\n",
      "[194]\tvalid_0's binary_logloss: 0.201193\n",
      "[195]\tvalid_0's binary_logloss: 0.20092\n",
      "[196]\tvalid_0's binary_logloss: 0.200828\n",
      "[197]\tvalid_0's binary_logloss: 0.200709\n",
      "[198]\tvalid_0's binary_logloss: 0.200609\n",
      "[199]\tvalid_0's binary_logloss: 0.200378\n",
      "[200]\tvalid_0's binary_logloss: 0.200213\n",
      "[201]\tvalid_0's binary_logloss: 0.200038\n",
      "[202]\tvalid_0's binary_logloss: 0.199954\n",
      "[203]\tvalid_0's binary_logloss: 0.199869\n",
      "[204]\tvalid_0's binary_logloss: 0.199701\n",
      "[205]\tvalid_0's binary_logloss: 0.19939\n",
      "[206]\tvalid_0's binary_logloss: 0.199232\n",
      "[207]\tvalid_0's binary_logloss: 0.198744\n",
      "[208]\tvalid_0's binary_logloss: 0.198354\n",
      "[209]\tvalid_0's binary_logloss: 0.198151\n",
      "[210]\tvalid_0's binary_logloss: 0.198072\n",
      "[211]\tvalid_0's binary_logloss: 0.197977\n",
      "[212]\tvalid_0's binary_logloss: 0.197736\n",
      "[213]\tvalid_0's binary_logloss: 0.197559\n",
      "[214]\tvalid_0's binary_logloss: 0.197097\n",
      "[215]\tvalid_0's binary_logloss: 0.196945\n",
      "[216]\tvalid_0's binary_logloss: 0.196731\n",
      "[217]\tvalid_0's binary_logloss: 0.196638\n",
      "[218]\tvalid_0's binary_logloss: 0.196471\n",
      "[219]\tvalid_0's binary_logloss: 0.19639\n",
      "[220]\tvalid_0's binary_logloss: 0.196333\n",
      "[221]\tvalid_0's binary_logloss: 0.195838\n",
      "[222]\tvalid_0's binary_logloss: 0.195772\n",
      "[223]\tvalid_0's binary_logloss: 0.195661\n",
      "[224]\tvalid_0's binary_logloss: 0.195562\n",
      "[225]\tvalid_0's binary_logloss: 0.195341\n",
      "[226]\tvalid_0's binary_logloss: 0.195033\n",
      "[227]\tvalid_0's binary_logloss: 0.194933\n",
      "[228]\tvalid_0's binary_logloss: 0.194704\n",
      "[229]\tvalid_0's binary_logloss: 0.194651\n",
      "[230]\tvalid_0's binary_logloss: 0.19437\n",
      "[231]\tvalid_0's binary_logloss: 0.194293\n",
      "[232]\tvalid_0's binary_logloss: 0.194218\n",
      "[233]\tvalid_0's binary_logloss: 0.194121\n",
      "[234]\tvalid_0's binary_logloss: 0.193876\n",
      "[235]\tvalid_0's binary_logloss: 0.193682\n",
      "[236]\tvalid_0's binary_logloss: 0.193589\n",
      "[237]\tvalid_0's binary_logloss: 0.193341\n",
      "[238]\tvalid_0's binary_logloss: 0.193149\n",
      "[239]\tvalid_0's binary_logloss: 0.193014\n",
      "[240]\tvalid_0's binary_logloss: 0.192894\n",
      "[241]\tvalid_0's binary_logloss: 0.192734\n",
      "[242]\tvalid_0's binary_logloss: 0.19253\n",
      "[243]\tvalid_0's binary_logloss: 0.19247\n",
      "[244]\tvalid_0's binary_logloss: 0.192149\n",
      "[245]\tvalid_0's binary_logloss: 0.192103\n",
      "[246]\tvalid_0's binary_logloss: 0.191968\n",
      "[247]\tvalid_0's binary_logloss: 0.191841\n",
      "[248]\tvalid_0's binary_logloss: 0.191671\n",
      "[249]\tvalid_0's binary_logloss: 0.191541\n",
      "[250]\tvalid_0's binary_logloss: 0.191471\n",
      "[251]\tvalid_0's binary_logloss: 0.191097\n",
      "[252]\tvalid_0's binary_logloss: 0.190949\n",
      "[253]\tvalid_0's binary_logloss: 0.190793\n",
      "[254]\tvalid_0's binary_logloss: 0.190745\n",
      "[255]\tvalid_0's binary_logloss: 0.19069\n",
      "[256]\tvalid_0's binary_logloss: 0.190514\n",
      "[257]\tvalid_0's binary_logloss: 0.190301\n",
      "[258]\tvalid_0's binary_logloss: 0.190206\n",
      "[259]\tvalid_0's binary_logloss: 0.190008\n",
      "[260]\tvalid_0's binary_logloss: 0.18996\n",
      "[261]\tvalid_0's binary_logloss: 0.189914\n",
      "[262]\tvalid_0's binary_logloss: 0.189803\n",
      "[263]\tvalid_0's binary_logloss: 0.189748\n",
      "[264]\tvalid_0's binary_logloss: 0.189547\n",
      "[265]\tvalid_0's binary_logloss: 0.189264\n",
      "[266]\tvalid_0's binary_logloss: 0.189136\n",
      "[267]\tvalid_0's binary_logloss: 0.189112\n",
      "[268]\tvalid_0's binary_logloss: 0.189006\n",
      "[269]\tvalid_0's binary_logloss: 0.188846\n",
      "[270]\tvalid_0's binary_logloss: 0.188806\n",
      "[271]\tvalid_0's binary_logloss: 0.188769\n",
      "[272]\tvalid_0's binary_logloss: 0.188699\n",
      "[273]\tvalid_0's binary_logloss: 0.188668\n",
      "[274]\tvalid_0's binary_logloss: 0.188581\n",
      "[275]\tvalid_0's binary_logloss: 0.188256\n",
      "[276]\tvalid_0's binary_logloss: 0.188196\n",
      "[277]\tvalid_0's binary_logloss: 0.18802\n",
      "[278]\tvalid_0's binary_logloss: 0.187967\n",
      "[279]\tvalid_0's binary_logloss: 0.187928\n",
      "[280]\tvalid_0's binary_logloss: 0.187789\n",
      "[281]\tvalid_0's binary_logloss: 0.187563\n",
      "[282]\tvalid_0's binary_logloss: 0.187521\n",
      "[283]\tvalid_0's binary_logloss: 0.187382\n",
      "[284]\tvalid_0's binary_logloss: 0.187208\n",
      "[285]\tvalid_0's binary_logloss: 0.187184\n",
      "[286]\tvalid_0's binary_logloss: 0.187039\n",
      "[287]\tvalid_0's binary_logloss: 0.186978\n",
      "[288]\tvalid_0's binary_logloss: 0.186955\n",
      "[289]\tvalid_0's binary_logloss: 0.186894\n",
      "[290]\tvalid_0's binary_logloss: 0.186663\n",
      "[291]\tvalid_0's binary_logloss: 0.186597\n",
      "[292]\tvalid_0's binary_logloss: 0.186321\n",
      "[293]\tvalid_0's binary_logloss: 0.186235\n",
      "[294]\tvalid_0's binary_logloss: 0.186113\n",
      "[295]\tvalid_0's binary_logloss: 0.186008\n",
      "[296]\tvalid_0's binary_logloss: 0.185935\n",
      "[297]\tvalid_0's binary_logloss: 0.185703\n",
      "[298]\tvalid_0's binary_logloss: 0.185544\n",
      "[299]\tvalid_0's binary_logloss: 0.185505\n",
      "[300]\tvalid_0's binary_logloss: 0.185327\n",
      "[301]\tvalid_0's binary_logloss: 0.185153\n",
      "[302]\tvalid_0's binary_logloss: 0.185078\n",
      "[303]\tvalid_0's binary_logloss: 0.184964\n",
      "[304]\tvalid_0's binary_logloss: 0.184936\n",
      "[305]\tvalid_0's binary_logloss: 0.184799\n",
      "[306]\tvalid_0's binary_logloss: 0.184643\n",
      "[307]\tvalid_0's binary_logloss: 0.184581\n",
      "[308]\tvalid_0's binary_logloss: 0.184464\n",
      "[309]\tvalid_0's binary_logloss: 0.184303\n",
      "[310]\tvalid_0's binary_logloss: 0.184285\n",
      "[311]\tvalid_0's binary_logloss: 0.184244\n",
      "[312]\tvalid_0's binary_logloss: 0.184202\n",
      "[313]\tvalid_0's binary_logloss: 0.184081\n",
      "[314]\tvalid_0's binary_logloss: 0.184035\n",
      "[315]\tvalid_0's binary_logloss: 0.184001\n",
      "[316]\tvalid_0's binary_logloss: 0.183987\n",
      "[317]\tvalid_0's binary_logloss: 0.183966\n",
      "[318]\tvalid_0's binary_logloss: 0.183841\n",
      "[319]\tvalid_0's binary_logloss: 0.183768\n",
      "[320]\tvalid_0's binary_logloss: 0.183744\n",
      "[321]\tvalid_0's binary_logloss: 0.183699\n",
      "[322]\tvalid_0's binary_logloss: 0.183606\n",
      "[323]\tvalid_0's binary_logloss: 0.183449\n",
      "[324]\tvalid_0's binary_logloss: 0.18342\n",
      "[325]\tvalid_0's binary_logloss: 0.18336\n",
      "[326]\tvalid_0's binary_logloss: 0.183175\n",
      "[327]\tvalid_0's binary_logloss: 0.183099\n",
      "[328]\tvalid_0's binary_logloss: 0.183018\n",
      "[329]\tvalid_0's binary_logloss: 0.182781\n",
      "[330]\tvalid_0's binary_logloss: 0.182729\n",
      "[331]\tvalid_0's binary_logloss: 0.182712\n",
      "[332]\tvalid_0's binary_logloss: 0.182617\n",
      "[333]\tvalid_0's binary_logloss: 0.182537\n",
      "[334]\tvalid_0's binary_logloss: 0.182476\n",
      "[335]\tvalid_0's binary_logloss: 0.182442\n",
      "[336]\tvalid_0's binary_logloss: 0.18236\n",
      "[337]\tvalid_0's binary_logloss: 0.182201\n",
      "[338]\tvalid_0's binary_logloss: 0.182135\n",
      "[339]\tvalid_0's binary_logloss: 0.18208\n",
      "[340]\tvalid_0's binary_logloss: 0.181907\n",
      "[341]\tvalid_0's binary_logloss: 0.181793\n",
      "[342]\tvalid_0's binary_logloss: 0.181753\n",
      "[343]\tvalid_0's binary_logloss: 0.18166\n",
      "[344]\tvalid_0's binary_logloss: 0.181604\n",
      "[345]\tvalid_0's binary_logloss: 0.181575\n",
      "[346]\tvalid_0's binary_logloss: 0.181546\n",
      "[347]\tvalid_0's binary_logloss: 0.18144\n",
      "[348]\tvalid_0's binary_logloss: 0.181397\n",
      "[349]\tvalid_0's binary_logloss: 0.181382\n",
      "[350]\tvalid_0's binary_logloss: 0.181232\n",
      "[351]\tvalid_0's binary_logloss: 0.181149\n",
      "[352]\tvalid_0's binary_logloss: 0.181025\n",
      "[353]\tvalid_0's binary_logloss: 0.181009\n",
      "[354]\tvalid_0's binary_logloss: 0.180999\n",
      "[355]\tvalid_0's binary_logloss: 0.180894\n",
      "[356]\tvalid_0's binary_logloss: 0.180756\n",
      "[357]\tvalid_0's binary_logloss: 0.180733\n",
      "[358]\tvalid_0's binary_logloss: 0.180709\n",
      "[359]\tvalid_0's binary_logloss: 0.180656\n",
      "[360]\tvalid_0's binary_logloss: 0.1806\n",
      "[361]\tvalid_0's binary_logloss: 0.180489\n",
      "[362]\tvalid_0's binary_logloss: 0.180432\n",
      "[363]\tvalid_0's binary_logloss: 0.180351\n",
      "[364]\tvalid_0's binary_logloss: 0.180336\n",
      "[365]\tvalid_0's binary_logloss: 0.180325\n",
      "[366]\tvalid_0's binary_logloss: 0.180296\n",
      "[367]\tvalid_0's binary_logloss: 0.180156\n",
      "[368]\tvalid_0's binary_logloss: 0.18011\n",
      "[369]\tvalid_0's binary_logloss: 0.180086\n",
      "[370]\tvalid_0's binary_logloss: 0.180073\n",
      "[371]\tvalid_0's binary_logloss: 0.179992\n",
      "[372]\tvalid_0's binary_logloss: 0.179952\n",
      "[373]\tvalid_0's binary_logloss: 0.179907\n",
      "[374]\tvalid_0's binary_logloss: 0.179902\n",
      "[375]\tvalid_0's binary_logloss: 0.179863\n",
      "[376]\tvalid_0's binary_logloss: 0.179795\n",
      "[377]\tvalid_0's binary_logloss: 0.179735\n",
      "[378]\tvalid_0's binary_logloss: 0.179684\n",
      "[379]\tvalid_0's binary_logloss: 0.179527\n",
      "[380]\tvalid_0's binary_logloss: 0.179436\n",
      "[381]\tvalid_0's binary_logloss: 0.179427\n",
      "[382]\tvalid_0's binary_logloss: 0.179348\n",
      "[383]\tvalid_0's binary_logloss: 0.179221\n",
      "[384]\tvalid_0's binary_logloss: 0.179135\n",
      "[385]\tvalid_0's binary_logloss: 0.179119\n",
      "[386]\tvalid_0's binary_logloss: 0.179096\n",
      "[387]\tvalid_0's binary_logloss: 0.179038\n",
      "[388]\tvalid_0's binary_logloss: 0.178977\n",
      "[389]\tvalid_0's binary_logloss: 0.178955\n",
      "[390]\tvalid_0's binary_logloss: 0.178948\n",
      "[391]\tvalid_0's binary_logloss: 0.178937\n",
      "[392]\tvalid_0's binary_logloss: 0.178889\n",
      "[393]\tvalid_0's binary_logloss: 0.178816\n",
      "[394]\tvalid_0's binary_logloss: 0.178789\n",
      "[395]\tvalid_0's binary_logloss: 0.178725\n",
      "[396]\tvalid_0's binary_logloss: 0.17873\n",
      "[397]\tvalid_0's binary_logloss: 0.178648\n",
      "[398]\tvalid_0's binary_logloss: 0.178625\n",
      "[399]\tvalid_0's binary_logloss: 0.178608\n",
      "[400]\tvalid_0's binary_logloss: 0.178545\n",
      "[401]\tvalid_0's binary_logloss: 0.178495\n",
      "[402]\tvalid_0's binary_logloss: 0.178379\n",
      "[403]\tvalid_0's binary_logloss: 0.178279\n",
      "[404]\tvalid_0's binary_logloss: 0.178167\n",
      "[405]\tvalid_0's binary_logloss: 0.178144\n",
      "[406]\tvalid_0's binary_logloss: 0.178042\n",
      "[407]\tvalid_0's binary_logloss: 0.178035\n",
      "[408]\tvalid_0's binary_logloss: 0.177952\n",
      "[409]\tvalid_0's binary_logloss: 0.177941\n",
      "[410]\tvalid_0's binary_logloss: 0.177881\n",
      "[411]\tvalid_0's binary_logloss: 0.177851\n",
      "[412]\tvalid_0's binary_logloss: 0.177765\n",
      "[413]\tvalid_0's binary_logloss: 0.177671\n",
      "[414]\tvalid_0's binary_logloss: 0.177593\n",
      "[415]\tvalid_0's binary_logloss: 0.177509\n",
      "[416]\tvalid_0's binary_logloss: 0.177407\n",
      "[417]\tvalid_0's binary_logloss: 0.177397\n",
      "[418]\tvalid_0's binary_logloss: 0.177365\n",
      "[419]\tvalid_0's binary_logloss: 0.177289\n",
      "[420]\tvalid_0's binary_logloss: 0.177229\n",
      "[421]\tvalid_0's binary_logloss: 0.177202\n",
      "[422]\tvalid_0's binary_logloss: 0.177192\n",
      "[423]\tvalid_0's binary_logloss: 0.177163\n",
      "[424]\tvalid_0's binary_logloss: 0.177097\n",
      "[425]\tvalid_0's binary_logloss: 0.177031\n",
      "[426]\tvalid_0's binary_logloss: 0.176977\n",
      "[427]\tvalid_0's binary_logloss: 0.176936\n",
      "[428]\tvalid_0's binary_logloss: 0.176838\n",
      "[429]\tvalid_0's binary_logloss: 0.176815\n",
      "[430]\tvalid_0's binary_logloss: 0.176783\n",
      "[431]\tvalid_0's binary_logloss: 0.176776\n",
      "[432]\tvalid_0's binary_logloss: 0.176746\n",
      "[433]\tvalid_0's binary_logloss: 0.176721\n",
      "[434]\tvalid_0's binary_logloss: 0.176717\n",
      "[435]\tvalid_0's binary_logloss: 0.176696\n",
      "[436]\tvalid_0's binary_logloss: 0.176627\n",
      "[437]\tvalid_0's binary_logloss: 0.176554\n",
      "[438]\tvalid_0's binary_logloss: 0.176536\n",
      "[439]\tvalid_0's binary_logloss: 0.176468\n",
      "[440]\tvalid_0's binary_logloss: 0.176393\n",
      "[441]\tvalid_0's binary_logloss: 0.176366\n",
      "[442]\tvalid_0's binary_logloss: 0.176361\n",
      "[443]\tvalid_0's binary_logloss: 0.176314\n",
      "[444]\tvalid_0's binary_logloss: 0.176267\n",
      "[445]\tvalid_0's binary_logloss: 0.176191\n",
      "[446]\tvalid_0's binary_logloss: 0.17617\n",
      "[447]\tvalid_0's binary_logloss: 0.176166\n",
      "[448]\tvalid_0's binary_logloss: 0.176093\n",
      "[449]\tvalid_0's binary_logloss: 0.176019\n",
      "[450]\tvalid_0's binary_logloss: 0.175973\n",
      "[451]\tvalid_0's binary_logloss: 0.175893\n",
      "[452]\tvalid_0's binary_logloss: 0.17589\n",
      "[453]\tvalid_0's binary_logloss: 0.175785\n",
      "[454]\tvalid_0's binary_logloss: 0.175725\n",
      "[455]\tvalid_0's binary_logloss: 0.175685\n",
      "[456]\tvalid_0's binary_logloss: 0.175627\n",
      "[457]\tvalid_0's binary_logloss: 0.175548\n",
      "[458]\tvalid_0's binary_logloss: 0.175543\n",
      "[459]\tvalid_0's binary_logloss: 0.175439\n",
      "[460]\tvalid_0's binary_logloss: 0.17538\n",
      "[461]\tvalid_0's binary_logloss: 0.175282\n",
      "[462]\tvalid_0's binary_logloss: 0.175262\n",
      "[463]\tvalid_0's binary_logloss: 0.175218\n",
      "[464]\tvalid_0's binary_logloss: 0.175153\n",
      "[465]\tvalid_0's binary_logloss: 0.175075\n",
      "[466]\tvalid_0's binary_logloss: 0.175049\n",
      "[467]\tvalid_0's binary_logloss: 0.174941\n",
      "[468]\tvalid_0's binary_logloss: 0.174899\n",
      "[469]\tvalid_0's binary_logloss: 0.174856\n",
      "[470]\tvalid_0's binary_logloss: 0.174816\n",
      "[471]\tvalid_0's binary_logloss: 0.174787\n",
      "[472]\tvalid_0's binary_logloss: 0.174753\n",
      "[473]\tvalid_0's binary_logloss: 0.1747\n",
      "[474]\tvalid_0's binary_logloss: 0.17468\n",
      "[475]\tvalid_0's binary_logloss: 0.17465\n",
      "[476]\tvalid_0's binary_logloss: 0.174599\n",
      "[477]\tvalid_0's binary_logloss: 0.174561\n",
      "[478]\tvalid_0's binary_logloss: 0.1745\n",
      "[479]\tvalid_0's binary_logloss: 0.174468\n",
      "[480]\tvalid_0's binary_logloss: 0.174448\n",
      "[481]\tvalid_0's binary_logloss: 0.174398\n",
      "[482]\tvalid_0's binary_logloss: 0.174355\n",
      "[483]\tvalid_0's binary_logloss: 0.174283\n",
      "[484]\tvalid_0's binary_logloss: 0.174266\n",
      "[485]\tvalid_0's binary_logloss: 0.174265\n",
      "[486]\tvalid_0's binary_logloss: 0.174252\n",
      "[487]\tvalid_0's binary_logloss: 0.174219\n",
      "[488]\tvalid_0's binary_logloss: 0.174192\n",
      "[489]\tvalid_0's binary_logloss: 0.174135\n",
      "[490]\tvalid_0's binary_logloss: 0.174085\n",
      "[491]\tvalid_0's binary_logloss: 0.174079\n",
      "[492]\tvalid_0's binary_logloss: 0.174045\n",
      "[493]\tvalid_0's binary_logloss: 0.174042\n",
      "[494]\tvalid_0's binary_logloss: 0.173956\n",
      "[495]\tvalid_0's binary_logloss: 0.173947\n",
      "[496]\tvalid_0's binary_logloss: 0.173938\n",
      "[497]\tvalid_0's binary_logloss: 0.173887\n",
      "[498]\tvalid_0's binary_logloss: 0.173842\n",
      "[499]\tvalid_0's binary_logloss: 0.173826\n",
      "[500]\tvalid_0's binary_logloss: 0.173814\n",
      "[501]\tvalid_0's binary_logloss: 0.173802\n",
      "[502]\tvalid_0's binary_logloss: 0.17377\n",
      "[503]\tvalid_0's binary_logloss: 0.173753\n",
      "[504]\tvalid_0's binary_logloss: 0.173738\n",
      "[505]\tvalid_0's binary_logloss: 0.173714\n",
      "[506]\tvalid_0's binary_logloss: 0.173717\n",
      "[507]\tvalid_0's binary_logloss: 0.173708\n",
      "[508]\tvalid_0's binary_logloss: 0.1737\n",
      "[509]\tvalid_0's binary_logloss: 0.173627\n",
      "[510]\tvalid_0's binary_logloss: 0.173593\n",
      "[511]\tvalid_0's binary_logloss: 0.173513\n",
      "[512]\tvalid_0's binary_logloss: 0.173498\n",
      "[513]\tvalid_0's binary_logloss: 0.173491\n",
      "[514]\tvalid_0's binary_logloss: 0.173478\n",
      "[515]\tvalid_0's binary_logloss: 0.173428\n",
      "[516]\tvalid_0's binary_logloss: 0.173411\n",
      "[517]\tvalid_0's binary_logloss: 0.173388\n",
      "[518]\tvalid_0's binary_logloss: 0.173329\n",
      "[519]\tvalid_0's binary_logloss: 0.173334\n",
      "[520]\tvalid_0's binary_logloss: 0.173276\n",
      "[521]\tvalid_0's binary_logloss: 0.173243\n",
      "[522]\tvalid_0's binary_logloss: 0.173188\n",
      "[523]\tvalid_0's binary_logloss: 0.173187\n",
      "[524]\tvalid_0's binary_logloss: 0.173136\n",
      "[525]\tvalid_0's binary_logloss: 0.173123\n",
      "[526]\tvalid_0's binary_logloss: 0.173077\n",
      "[527]\tvalid_0's binary_logloss: 0.173073\n",
      "[528]\tvalid_0's binary_logloss: 0.173044\n",
      "[529]\tvalid_0's binary_logloss: 0.173001\n",
      "[530]\tvalid_0's binary_logloss: 0.172977\n",
      "[531]\tvalid_0's binary_logloss: 0.172926\n",
      "[532]\tvalid_0's binary_logloss: 0.172903\n",
      "[533]\tvalid_0's binary_logloss: 0.172893\n",
      "[534]\tvalid_0's binary_logloss: 0.172868\n",
      "[535]\tvalid_0's binary_logloss: 0.172828\n",
      "[536]\tvalid_0's binary_logloss: 0.172796\n",
      "[537]\tvalid_0's binary_logloss: 0.172756\n",
      "[538]\tvalid_0's binary_logloss: 0.1727\n",
      "[539]\tvalid_0's binary_logloss: 0.172697\n",
      "[540]\tvalid_0's binary_logloss: 0.172685\n",
      "[541]\tvalid_0's binary_logloss: 0.172672\n",
      "[542]\tvalid_0's binary_logloss: 0.172602\n",
      "[543]\tvalid_0's binary_logloss: 0.172568\n",
      "[544]\tvalid_0's binary_logloss: 0.172505\n",
      "[545]\tvalid_0's binary_logloss: 0.172493\n",
      "[546]\tvalid_0's binary_logloss: 0.17244\n",
      "[547]\tvalid_0's binary_logloss: 0.172385\n",
      "[548]\tvalid_0's binary_logloss: 0.172353\n",
      "[549]\tvalid_0's binary_logloss: 0.1723\n",
      "[550]\tvalid_0's binary_logloss: 0.172249\n",
      "[551]\tvalid_0's binary_logloss: 0.172229\n",
      "[552]\tvalid_0's binary_logloss: 0.172164\n",
      "[553]\tvalid_0's binary_logloss: 0.172157\n",
      "[554]\tvalid_0's binary_logloss: 0.172158\n",
      "[555]\tvalid_0's binary_logloss: 0.172155\n",
      "[556]\tvalid_0's binary_logloss: 0.172109\n",
      "[557]\tvalid_0's binary_logloss: 0.172083\n",
      "[558]\tvalid_0's binary_logloss: 0.172063\n",
      "[559]\tvalid_0's binary_logloss: 0.172023\n",
      "[560]\tvalid_0's binary_logloss: 0.171991\n",
      "[561]\tvalid_0's binary_logloss: 0.171929\n",
      "[562]\tvalid_0's binary_logloss: 0.171901\n",
      "[563]\tvalid_0's binary_logloss: 0.171873\n",
      "[564]\tvalid_0's binary_logloss: 0.171847\n",
      "[565]\tvalid_0's binary_logloss: 0.171804\n",
      "[566]\tvalid_0's binary_logloss: 0.171788\n",
      "[567]\tvalid_0's binary_logloss: 0.171744\n",
      "[568]\tvalid_0's binary_logloss: 0.171729\n",
      "[569]\tvalid_0's binary_logloss: 0.171709\n",
      "[570]\tvalid_0's binary_logloss: 0.171694\n",
      "[571]\tvalid_0's binary_logloss: 0.17168\n",
      "[572]\tvalid_0's binary_logloss: 0.17166\n",
      "[573]\tvalid_0's binary_logloss: 0.171645\n",
      "[574]\tvalid_0's binary_logloss: 0.171613\n",
      "[575]\tvalid_0's binary_logloss: 0.171574\n",
      "[576]\tvalid_0's binary_logloss: 0.171571\n",
      "[577]\tvalid_0's binary_logloss: 0.17155\n",
      "[578]\tvalid_0's binary_logloss: 0.171511\n",
      "[579]\tvalid_0's binary_logloss: 0.171488\n",
      "[580]\tvalid_0's binary_logloss: 0.17147\n",
      "[581]\tvalid_0's binary_logloss: 0.171468\n",
      "[582]\tvalid_0's binary_logloss: 0.171414\n",
      "[583]\tvalid_0's binary_logloss: 0.171357\n",
      "[584]\tvalid_0's binary_logloss: 0.171318\n",
      "[585]\tvalid_0's binary_logloss: 0.171304\n",
      "[586]\tvalid_0's binary_logloss: 0.171266\n",
      "[587]\tvalid_0's binary_logloss: 0.171271\n",
      "[588]\tvalid_0's binary_logloss: 0.171234\n",
      "[589]\tvalid_0's binary_logloss: 0.171225\n",
      "[590]\tvalid_0's binary_logloss: 0.171198\n",
      "[591]\tvalid_0's binary_logloss: 0.171159\n",
      "[592]\tvalid_0's binary_logloss: 0.171103\n",
      "[593]\tvalid_0's binary_logloss: 0.171081\n",
      "[594]\tvalid_0's binary_logloss: 0.171041\n",
      "[595]\tvalid_0's binary_logloss: 0.171039\n",
      "[596]\tvalid_0's binary_logloss: 0.170997\n",
      "[597]\tvalid_0's binary_logloss: 0.170976\n",
      "[598]\tvalid_0's binary_logloss: 0.170956\n",
      "[599]\tvalid_0's binary_logloss: 0.170947\n",
      "[600]\tvalid_0's binary_logloss: 0.170929\n",
      "[601]\tvalid_0's binary_logloss: 0.170934\n",
      "[602]\tvalid_0's binary_logloss: 0.170908\n",
      "[603]\tvalid_0's binary_logloss: 0.170855\n",
      "[604]\tvalid_0's binary_logloss: 0.170833\n",
      "[605]\tvalid_0's binary_logloss: 0.170827\n",
      "[606]\tvalid_0's binary_logloss: 0.17077\n",
      "[607]\tvalid_0's binary_logloss: 0.170735\n",
      "[608]\tvalid_0's binary_logloss: 0.170717\n",
      "[609]\tvalid_0's binary_logloss: 0.17069\n",
      "[610]\tvalid_0's binary_logloss: 0.170674\n",
      "[611]\tvalid_0's binary_logloss: 0.170663\n",
      "[612]\tvalid_0's binary_logloss: 0.170658\n",
      "[613]\tvalid_0's binary_logloss: 0.170654\n",
      "[614]\tvalid_0's binary_logloss: 0.170616\n",
      "[615]\tvalid_0's binary_logloss: 0.170598\n",
      "[616]\tvalid_0's binary_logloss: 0.170595\n",
      "[617]\tvalid_0's binary_logloss: 0.170556\n",
      "[618]\tvalid_0's binary_logloss: 0.170553\n",
      "[619]\tvalid_0's binary_logloss: 0.170544\n",
      "[620]\tvalid_0's binary_logloss: 0.170519\n",
      "[621]\tvalid_0's binary_logloss: 0.170479\n",
      "[622]\tvalid_0's binary_logloss: 0.170442\n",
      "[623]\tvalid_0's binary_logloss: 0.170422\n",
      "[624]\tvalid_0's binary_logloss: 0.170415\n",
      "[625]\tvalid_0's binary_logloss: 0.170404\n",
      "[626]\tvalid_0's binary_logloss: 0.170396\n",
      "[627]\tvalid_0's binary_logloss: 0.170361\n",
      "[628]\tvalid_0's binary_logloss: 0.170345\n",
      "[629]\tvalid_0's binary_logloss: 0.170327\n",
      "[630]\tvalid_0's binary_logloss: 0.170315\n",
      "[631]\tvalid_0's binary_logloss: 0.170294\n",
      "[632]\tvalid_0's binary_logloss: 0.170275\n",
      "[633]\tvalid_0's binary_logloss: 0.170245\n",
      "[634]\tvalid_0's binary_logloss: 0.170237\n",
      "[635]\tvalid_0's binary_logloss: 0.170209\n",
      "[636]\tvalid_0's binary_logloss: 0.17019\n",
      "[637]\tvalid_0's binary_logloss: 0.170192\n",
      "[638]\tvalid_0's binary_logloss: 0.170182\n",
      "[639]\tvalid_0's binary_logloss: 0.170167\n",
      "[640]\tvalid_0's binary_logloss: 0.170166\n",
      "[641]\tvalid_0's binary_logloss: 0.170134\n",
      "[642]\tvalid_0's binary_logloss: 0.170085\n",
      "[643]\tvalid_0's binary_logloss: 0.170065\n",
      "[644]\tvalid_0's binary_logloss: 0.170025\n",
      "[645]\tvalid_0's binary_logloss: 0.169993\n",
      "[646]\tvalid_0's binary_logloss: 0.169964\n",
      "[647]\tvalid_0's binary_logloss: 0.16993\n",
      "[648]\tvalid_0's binary_logloss: 0.16991\n",
      "[649]\tvalid_0's binary_logloss: 0.169898\n",
      "[650]\tvalid_0's binary_logloss: 0.169875\n",
      "[651]\tvalid_0's binary_logloss: 0.169839\n",
      "[652]\tvalid_0's binary_logloss: 0.169824\n",
      "[653]\tvalid_0's binary_logloss: 0.169819\n",
      "[654]\tvalid_0's binary_logloss: 0.169822\n",
      "[655]\tvalid_0's binary_logloss: 0.169783\n",
      "[656]\tvalid_0's binary_logloss: 0.169776\n",
      "[657]\tvalid_0's binary_logloss: 0.169739\n",
      "[658]\tvalid_0's binary_logloss: 0.169738\n",
      "[659]\tvalid_0's binary_logloss: 0.169746\n",
      "[660]\tvalid_0's binary_logloss: 0.169746\n",
      "[661]\tvalid_0's binary_logloss: 0.169731\n",
      "[662]\tvalid_0's binary_logloss: 0.169699\n",
      "[663]\tvalid_0's binary_logloss: 0.169706\n",
      "[664]\tvalid_0's binary_logloss: 0.169691\n",
      "[665]\tvalid_0's binary_logloss: 0.169654\n",
      "[666]\tvalid_0's binary_logloss: 0.169643\n",
      "[667]\tvalid_0's binary_logloss: 0.169602\n",
      "[668]\tvalid_0's binary_logloss: 0.169594\n",
      "[669]\tvalid_0's binary_logloss: 0.169568\n",
      "[670]\tvalid_0's binary_logloss: 0.169566\n",
      "[671]\tvalid_0's binary_logloss: 0.169509\n",
      "[672]\tvalid_0's binary_logloss: 0.169478\n",
      "[673]\tvalid_0's binary_logloss: 0.169477\n",
      "[674]\tvalid_0's binary_logloss: 0.169429\n",
      "[675]\tvalid_0's binary_logloss: 0.169425\n",
      "[676]\tvalid_0's binary_logloss: 0.1694\n",
      "[677]\tvalid_0's binary_logloss: 0.169377\n",
      "[678]\tvalid_0's binary_logloss: 0.169359\n",
      "[679]\tvalid_0's binary_logloss: 0.169359\n",
      "[680]\tvalid_0's binary_logloss: 0.169352\n",
      "[681]\tvalid_0's binary_logloss: 0.169347\n",
      "[682]\tvalid_0's binary_logloss: 0.169336\n",
      "[683]\tvalid_0's binary_logloss: 0.169333\n",
      "[684]\tvalid_0's binary_logloss: 0.169303\n",
      "[685]\tvalid_0's binary_logloss: 0.169298\n",
      "[686]\tvalid_0's binary_logloss: 0.169266\n",
      "[687]\tvalid_0's binary_logloss: 0.169246\n",
      "[688]\tvalid_0's binary_logloss: 0.16922\n",
      "[689]\tvalid_0's binary_logloss: 0.169209\n",
      "[690]\tvalid_0's binary_logloss: 0.169213\n",
      "[691]\tvalid_0's binary_logloss: 0.169199\n",
      "[692]\tvalid_0's binary_logloss: 0.169192\n",
      "[693]\tvalid_0's binary_logloss: 0.169178\n",
      "[694]\tvalid_0's binary_logloss: 0.169168\n",
      "[695]\tvalid_0's binary_logloss: 0.169154\n",
      "[696]\tvalid_0's binary_logloss: 0.169128\n",
      "[697]\tvalid_0's binary_logloss: 0.169135\n",
      "[698]\tvalid_0's binary_logloss: 0.169109\n",
      "[699]\tvalid_0's binary_logloss: 0.169078\n",
      "[700]\tvalid_0's binary_logloss: 0.16904\n",
      "[701]\tvalid_0's binary_logloss: 0.169028\n",
      "[702]\tvalid_0's binary_logloss: 0.16903\n",
      "[703]\tvalid_0's binary_logloss: 0.168998\n",
      "[704]\tvalid_0's binary_logloss: 0.168981\n",
      "[705]\tvalid_0's binary_logloss: 0.168981\n",
      "[706]\tvalid_0's binary_logloss: 0.168959\n",
      "[707]\tvalid_0's binary_logloss: 0.168953\n",
      "[708]\tvalid_0's binary_logloss: 0.168929\n",
      "[709]\tvalid_0's binary_logloss: 0.168932\n",
      "[710]\tvalid_0's binary_logloss: 0.168902\n",
      "[711]\tvalid_0's binary_logloss: 0.168885\n",
      "[712]\tvalid_0's binary_logloss: 0.168873\n",
      "[713]\tvalid_0's binary_logloss: 0.168862\n",
      "[714]\tvalid_0's binary_logloss: 0.168849\n",
      "[715]\tvalid_0's binary_logloss: 0.16883\n",
      "[716]\tvalid_0's binary_logloss: 0.168815\n",
      "[717]\tvalid_0's binary_logloss: 0.168791\n",
      "[718]\tvalid_0's binary_logloss: 0.168757\n",
      "[719]\tvalid_0's binary_logloss: 0.16876\n",
      "[720]\tvalid_0's binary_logloss: 0.16873\n",
      "[721]\tvalid_0's binary_logloss: 0.168724\n",
      "[722]\tvalid_0's binary_logloss: 0.168696\n",
      "[723]\tvalid_0's binary_logloss: 0.168663\n",
      "[724]\tvalid_0's binary_logloss: 0.168631\n",
      "[725]\tvalid_0's binary_logloss: 0.168615\n",
      "[726]\tvalid_0's binary_logloss: 0.168613\n",
      "[727]\tvalid_0's binary_logloss: 0.168601\n",
      "[728]\tvalid_0's binary_logloss: 0.168591\n",
      "[729]\tvalid_0's binary_logloss: 0.168589\n",
      "[730]\tvalid_0's binary_logloss: 0.168588\n",
      "[731]\tvalid_0's binary_logloss: 0.168587\n",
      "[732]\tvalid_0's binary_logloss: 0.168584\n",
      "[733]\tvalid_0's binary_logloss: 0.168568\n",
      "[734]\tvalid_0's binary_logloss: 0.168572\n",
      "[735]\tvalid_0's binary_logloss: 0.16855\n",
      "[736]\tvalid_0's binary_logloss: 0.168535\n",
      "[737]\tvalid_0's binary_logloss: 0.168534\n",
      "[738]\tvalid_0's binary_logloss: 0.168528\n",
      "[739]\tvalid_0's binary_logloss: 0.168515\n",
      "[740]\tvalid_0's binary_logloss: 0.168505\n",
      "[741]\tvalid_0's binary_logloss: 0.168489\n",
      "[742]\tvalid_0's binary_logloss: 0.168453\n",
      "[743]\tvalid_0's binary_logloss: 0.168422\n",
      "[744]\tvalid_0's binary_logloss: 0.16841\n",
      "[745]\tvalid_0's binary_logloss: 0.168381\n",
      "[746]\tvalid_0's binary_logloss: 0.168353\n",
      "[747]\tvalid_0's binary_logloss: 0.168348\n",
      "[748]\tvalid_0's binary_logloss: 0.168312\n",
      "[749]\tvalid_0's binary_logloss: 0.16831\n",
      "[750]\tvalid_0's binary_logloss: 0.168289\n",
      "[751]\tvalid_0's binary_logloss: 0.168284\n",
      "[752]\tvalid_0's binary_logloss: 0.168272\n",
      "[753]\tvalid_0's binary_logloss: 0.168227\n",
      "[754]\tvalid_0's binary_logloss: 0.168183\n",
      "[755]\tvalid_0's binary_logloss: 0.168178\n",
      "[756]\tvalid_0's binary_logloss: 0.168175\n",
      "[757]\tvalid_0's binary_logloss: 0.168159\n",
      "[758]\tvalid_0's binary_logloss: 0.168137\n",
      "[759]\tvalid_0's binary_logloss: 0.168125\n",
      "[760]\tvalid_0's binary_logloss: 0.168115\n",
      "[761]\tvalid_0's binary_logloss: 0.16812\n",
      "[762]\tvalid_0's binary_logloss: 0.168088\n",
      "[763]\tvalid_0's binary_logloss: 0.168084\n",
      "[764]\tvalid_0's binary_logloss: 0.168075\n",
      "[765]\tvalid_0's binary_logloss: 0.168053\n",
      "[766]\tvalid_0's binary_logloss: 0.168027\n",
      "[767]\tvalid_0's binary_logloss: 0.168013\n",
      "[768]\tvalid_0's binary_logloss: 0.168017\n",
      "[769]\tvalid_0's binary_logloss: 0.168003\n",
      "[770]\tvalid_0's binary_logloss: 0.16799\n",
      "[771]\tvalid_0's binary_logloss: 0.167981\n",
      "[772]\tvalid_0's binary_logloss: 0.167985\n",
      "[773]\tvalid_0's binary_logloss: 0.16799\n",
      "[774]\tvalid_0's binary_logloss: 0.167964\n",
      "[775]\tvalid_0's binary_logloss: 0.167953\n",
      "[776]\tvalid_0's binary_logloss: 0.16793\n",
      "[777]\tvalid_0's binary_logloss: 0.167915\n",
      "[778]\tvalid_0's binary_logloss: 0.1679\n",
      "[779]\tvalid_0's binary_logloss: 0.167872\n",
      "[780]\tvalid_0's binary_logloss: 0.167847\n",
      "[781]\tvalid_0's binary_logloss: 0.167849\n",
      "[782]\tvalid_0's binary_logloss: 0.167835\n",
      "[783]\tvalid_0's binary_logloss: 0.167808\n",
      "[784]\tvalid_0's binary_logloss: 0.167791\n",
      "[785]\tvalid_0's binary_logloss: 0.167796\n",
      "[786]\tvalid_0's binary_logloss: 0.167771\n",
      "[787]\tvalid_0's binary_logloss: 0.167756\n",
      "[788]\tvalid_0's binary_logloss: 0.167748\n",
      "[789]\tvalid_0's binary_logloss: 0.167749\n",
      "[790]\tvalid_0's binary_logloss: 0.167732\n",
      "[791]\tvalid_0's binary_logloss: 0.167693\n",
      "[792]\tvalid_0's binary_logloss: 0.167663\n",
      "[793]\tvalid_0's binary_logloss: 0.167659\n",
      "[794]\tvalid_0's binary_logloss: 0.167661\n",
      "[795]\tvalid_0's binary_logloss: 0.167645\n",
      "[796]\tvalid_0's binary_logloss: 0.167622\n",
      "[797]\tvalid_0's binary_logloss: 0.167611\n",
      "[798]\tvalid_0's binary_logloss: 0.167614\n",
      "[799]\tvalid_0's binary_logloss: 0.167609\n",
      "[800]\tvalid_0's binary_logloss: 0.167606\n",
      "[801]\tvalid_0's binary_logloss: 0.167602\n",
      "[802]\tvalid_0's binary_logloss: 0.167593\n",
      "[803]\tvalid_0's binary_logloss: 0.167598\n",
      "[804]\tvalid_0's binary_logloss: 0.167574\n",
      "[805]\tvalid_0's binary_logloss: 0.167568\n",
      "[806]\tvalid_0's binary_logloss: 0.167557\n",
      "[807]\tvalid_0's binary_logloss: 0.167553\n",
      "[808]\tvalid_0's binary_logloss: 0.167542\n",
      "[809]\tvalid_0's binary_logloss: 0.167532\n",
      "[810]\tvalid_0's binary_logloss: 0.167525\n",
      "[811]\tvalid_0's binary_logloss: 0.167511\n",
      "[812]\tvalid_0's binary_logloss: 0.167507\n",
      "[813]\tvalid_0's binary_logloss: 0.167506\n",
      "[814]\tvalid_0's binary_logloss: 0.167492\n",
      "[815]\tvalid_0's binary_logloss: 0.16749\n",
      "[816]\tvalid_0's binary_logloss: 0.167471\n",
      "[817]\tvalid_0's binary_logloss: 0.167436\n",
      "[818]\tvalid_0's binary_logloss: 0.167425\n",
      "[819]\tvalid_0's binary_logloss: 0.167399\n",
      "[820]\tvalid_0's binary_logloss: 0.16739\n",
      "[821]\tvalid_0's binary_logloss: 0.167371\n",
      "[822]\tvalid_0's binary_logloss: 0.167348\n",
      "[823]\tvalid_0's binary_logloss: 0.167329\n",
      "[824]\tvalid_0's binary_logloss: 0.167312\n",
      "[825]\tvalid_0's binary_logloss: 0.167302\n",
      "[826]\tvalid_0's binary_logloss: 0.167294\n",
      "[827]\tvalid_0's binary_logloss: 0.167279\n",
      "[828]\tvalid_0's binary_logloss: 0.167265\n",
      "[829]\tvalid_0's binary_logloss: 0.167267\n",
      "[830]\tvalid_0's binary_logloss: 0.167258\n",
      "[831]\tvalid_0's binary_logloss: 0.167259\n",
      "[832]\tvalid_0's binary_logloss: 0.167251\n",
      "[833]\tvalid_0's binary_logloss: 0.167237\n",
      "[834]\tvalid_0's binary_logloss: 0.16723\n",
      "[835]\tvalid_0's binary_logloss: 0.167223\n",
      "[836]\tvalid_0's binary_logloss: 0.167213\n",
      "[837]\tvalid_0's binary_logloss: 0.167201\n",
      "[838]\tvalid_0's binary_logloss: 0.167179\n",
      "[839]\tvalid_0's binary_logloss: 0.167167\n",
      "[840]\tvalid_0's binary_logloss: 0.167139\n",
      "[841]\tvalid_0's binary_logloss: 0.167116\n",
      "[842]\tvalid_0's binary_logloss: 0.167085\n",
      "[843]\tvalid_0's binary_logloss: 0.167064\n",
      "[844]\tvalid_0's binary_logloss: 0.167055\n",
      "[845]\tvalid_0's binary_logloss: 0.167029\n",
      "[846]\tvalid_0's binary_logloss: 0.167021\n",
      "[847]\tvalid_0's binary_logloss: 0.167022\n",
      "[848]\tvalid_0's binary_logloss: 0.167017\n",
      "[849]\tvalid_0's binary_logloss: 0.16701\n",
      "[850]\tvalid_0's binary_logloss: 0.167004\n",
      "[851]\tvalid_0's binary_logloss: 0.166988\n",
      "[852]\tvalid_0's binary_logloss: 0.166975\n",
      "[853]\tvalid_0's binary_logloss: 0.166962\n",
      "[854]\tvalid_0's binary_logloss: 0.166956\n",
      "[855]\tvalid_0's binary_logloss: 0.166938\n",
      "[856]\tvalid_0's binary_logloss: 0.166936\n",
      "[857]\tvalid_0's binary_logloss: 0.166911\n",
      "[858]\tvalid_0's binary_logloss: 0.166877\n",
      "[859]\tvalid_0's binary_logloss: 0.166882\n",
      "[860]\tvalid_0's binary_logloss: 0.166878\n",
      "[861]\tvalid_0's binary_logloss: 0.166868\n",
      "[862]\tvalid_0's binary_logloss: 0.166874\n",
      "[863]\tvalid_0's binary_logloss: 0.166878\n",
      "[864]\tvalid_0's binary_logloss: 0.166844\n",
      "[865]\tvalid_0's binary_logloss: 0.166819\n",
      "[866]\tvalid_0's binary_logloss: 0.166808\n",
      "[867]\tvalid_0's binary_logloss: 0.166795\n",
      "[868]\tvalid_0's binary_logloss: 0.166779\n",
      "[869]\tvalid_0's binary_logloss: 0.166775\n",
      "[870]\tvalid_0's binary_logloss: 0.166773\n",
      "[871]\tvalid_0's binary_logloss: 0.16677\n",
      "[872]\tvalid_0's binary_logloss: 0.166768\n",
      "[873]\tvalid_0's binary_logloss: 0.16676\n",
      "[874]\tvalid_0's binary_logloss: 0.166749\n",
      "[875]\tvalid_0's binary_logloss: 0.166742\n",
      "[876]\tvalid_0's binary_logloss: 0.16674\n",
      "[877]\tvalid_0's binary_logloss: 0.166744\n",
      "[878]\tvalid_0's binary_logloss: 0.166736\n",
      "[879]\tvalid_0's binary_logloss: 0.166735\n",
      "[880]\tvalid_0's binary_logloss: 0.166719\n",
      "[881]\tvalid_0's binary_logloss: 0.166706\n",
      "[882]\tvalid_0's binary_logloss: 0.166694\n",
      "[883]\tvalid_0's binary_logloss: 0.166677\n",
      "[884]\tvalid_0's binary_logloss: 0.166671\n",
      "[885]\tvalid_0's binary_logloss: 0.166656\n",
      "[886]\tvalid_0's binary_logloss: 0.166659\n",
      "[887]\tvalid_0's binary_logloss: 0.166647\n",
      "[888]\tvalid_0's binary_logloss: 0.166643\n",
      "[889]\tvalid_0's binary_logloss: 0.166617\n",
      "[890]\tvalid_0's binary_logloss: 0.166602\n",
      "[891]\tvalid_0's binary_logloss: 0.166605\n",
      "[892]\tvalid_0's binary_logloss: 0.166599\n",
      "[893]\tvalid_0's binary_logloss: 0.166592\n",
      "[894]\tvalid_0's binary_logloss: 0.166585\n",
      "[895]\tvalid_0's binary_logloss: 0.166591\n",
      "[896]\tvalid_0's binary_logloss: 0.166573\n",
      "[897]\tvalid_0's binary_logloss: 0.166574\n",
      "[898]\tvalid_0's binary_logloss: 0.16657\n",
      "[899]\tvalid_0's binary_logloss: 0.166557\n",
      "[900]\tvalid_0's binary_logloss: 0.166554\n",
      "[901]\tvalid_0's binary_logloss: 0.166557\n",
      "[902]\tvalid_0's binary_logloss: 0.16653\n",
      "[903]\tvalid_0's binary_logloss: 0.166531\n",
      "[904]\tvalid_0's binary_logloss: 0.166503\n",
      "[905]\tvalid_0's binary_logloss: 0.166489\n",
      "[906]\tvalid_0's binary_logloss: 0.16648\n",
      "[907]\tvalid_0's binary_logloss: 0.166468\n",
      "[908]\tvalid_0's binary_logloss: 0.166462\n",
      "[909]\tvalid_0's binary_logloss: 0.166464\n",
      "[910]\tvalid_0's binary_logloss: 0.166464\n",
      "[911]\tvalid_0's binary_logloss: 0.166466\n",
      "[912]\tvalid_0's binary_logloss: 0.166469\n",
      "[913]\tvalid_0's binary_logloss: 0.166466\n",
      "[914]\tvalid_0's binary_logloss: 0.166456\n",
      "[915]\tvalid_0's binary_logloss: 0.16645\n",
      "[916]\tvalid_0's binary_logloss: 0.166448\n",
      "[917]\tvalid_0's binary_logloss: 0.166447\n",
      "[918]\tvalid_0's binary_logloss: 0.166448\n",
      "[919]\tvalid_0's binary_logloss: 0.166425\n",
      "[920]\tvalid_0's binary_logloss: 0.166392\n",
      "[921]\tvalid_0's binary_logloss: 0.166385\n",
      "[922]\tvalid_0's binary_logloss: 0.166374\n",
      "[923]\tvalid_0's binary_logloss: 0.166382\n",
      "[924]\tvalid_0's binary_logloss: 0.166376\n",
      "[925]\tvalid_0's binary_logloss: 0.166363\n",
      "[926]\tvalid_0's binary_logloss: 0.166351\n",
      "[927]\tvalid_0's binary_logloss: 0.166347\n",
      "[928]\tvalid_0's binary_logloss: 0.166345\n",
      "[929]\tvalid_0's binary_logloss: 0.166346\n",
      "[930]\tvalid_0's binary_logloss: 0.16634\n",
      "[931]\tvalid_0's binary_logloss: 0.166333\n",
      "[932]\tvalid_0's binary_logloss: 0.166311\n",
      "[933]\tvalid_0's binary_logloss: 0.166306\n",
      "[934]\tvalid_0's binary_logloss: 0.166296\n",
      "[935]\tvalid_0's binary_logloss: 0.166272\n",
      "[936]\tvalid_0's binary_logloss: 0.166265\n",
      "[937]\tvalid_0's binary_logloss: 0.166262\n",
      "[938]\tvalid_0's binary_logloss: 0.166256\n",
      "[939]\tvalid_0's binary_logloss: 0.166255\n",
      "[940]\tvalid_0's binary_logloss: 0.16625\n",
      "[941]\tvalid_0's binary_logloss: 0.16624\n",
      "[942]\tvalid_0's binary_logloss: 0.166206\n",
      "[943]\tvalid_0's binary_logloss: 0.16618\n",
      "[944]\tvalid_0's binary_logloss: 0.166188\n",
      "[945]\tvalid_0's binary_logloss: 0.166188\n",
      "[946]\tvalid_0's binary_logloss: 0.166175\n",
      "[947]\tvalid_0's binary_logloss: 0.166169\n",
      "[948]\tvalid_0's binary_logloss: 0.166171\n",
      "[949]\tvalid_0's binary_logloss: 0.166163\n",
      "[950]\tvalid_0's binary_logloss: 0.166161\n",
      "[951]\tvalid_0's binary_logloss: 0.166156\n",
      "[952]\tvalid_0's binary_logloss: 0.166142\n",
      "[953]\tvalid_0's binary_logloss: 0.166149\n",
      "[954]\tvalid_0's binary_logloss: 0.166138\n",
      "[955]\tvalid_0's binary_logloss: 0.166135\n",
      "[956]\tvalid_0's binary_logloss: 0.166132\n",
      "[957]\tvalid_0's binary_logloss: 0.166103\n",
      "[958]\tvalid_0's binary_logloss: 0.166095\n",
      "[959]\tvalid_0's binary_logloss: 0.166074\n",
      "[960]\tvalid_0's binary_logloss: 0.166052\n",
      "[961]\tvalid_0's binary_logloss: 0.166052\n",
      "[962]\tvalid_0's binary_logloss: 0.166048\n",
      "[963]\tvalid_0's binary_logloss: 0.16605\n",
      "[964]\tvalid_0's binary_logloss: 0.166049\n",
      "[965]\tvalid_0's binary_logloss: 0.166041\n",
      "[966]\tvalid_0's binary_logloss: 0.166036\n",
      "[967]\tvalid_0's binary_logloss: 0.166016\n",
      "[968]\tvalid_0's binary_logloss: 0.166013\n",
      "[969]\tvalid_0's binary_logloss: 0.166012\n",
      "[970]\tvalid_0's binary_logloss: 0.166003\n",
      "[971]\tvalid_0's binary_logloss: 0.165992\n",
      "[972]\tvalid_0's binary_logloss: 0.165983\n",
      "[973]\tvalid_0's binary_logloss: 0.165962\n",
      "[974]\tvalid_0's binary_logloss: 0.16594\n",
      "[975]\tvalid_0's binary_logloss: 0.165935\n",
      "[976]\tvalid_0's binary_logloss: 0.165929\n",
      "[977]\tvalid_0's binary_logloss: 0.165901\n",
      "[978]\tvalid_0's binary_logloss: 0.165885\n",
      "[979]\tvalid_0's binary_logloss: 0.165891\n",
      "[980]\tvalid_0's binary_logloss: 0.165876\n",
      "[981]\tvalid_0's binary_logloss: 0.165865\n",
      "[982]\tvalid_0's binary_logloss: 0.165862\n",
      "[983]\tvalid_0's binary_logloss: 0.165854\n",
      "[984]\tvalid_0's binary_logloss: 0.165857\n",
      "[985]\tvalid_0's binary_logloss: 0.165849\n",
      "[986]\tvalid_0's binary_logloss: 0.165854\n",
      "[987]\tvalid_0's binary_logloss: 0.165853\n",
      "[988]\tvalid_0's binary_logloss: 0.165843\n",
      "[989]\tvalid_0's binary_logloss: 0.165829\n",
      "[990]\tvalid_0's binary_logloss: 0.165811\n",
      "[991]\tvalid_0's binary_logloss: 0.165792\n",
      "[992]\tvalid_0's binary_logloss: 0.165794\n",
      "[993]\tvalid_0's binary_logloss: 0.16578\n",
      "[994]\tvalid_0's binary_logloss: 0.165774\n",
      "[995]\tvalid_0's binary_logloss: 0.16575\n",
      "[996]\tvalid_0's binary_logloss: 0.165748\n",
      "[997]\tvalid_0's binary_logloss: 0.16574\n",
      "[998]\tvalid_0's binary_logloss: 0.165727\n",
      "[999]\tvalid_0's binary_logloss: 0.165712\n",
      "[1000]\tvalid_0's binary_logloss: 0.165713\n",
      "training model for CV #4\n",
      "[LightGBM] [Warning] feature_fraction is set=0.4, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.4\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.0, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.0\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8772503866491428, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8772503866491428\n",
      "[LightGBM] [Warning] lambda_l2 is set=0.0, reg_lambda=0.0 will be ignored. Current value: lambda_l2=0.0\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\engine.py:177: UserWarning: Found `num_iterations` in params. Will use it instead of argument\n",
      "  _log_warning(f\"Found `{alias}` in params. Will use it instead of argument\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tvalid_0's binary_logloss: 0.662062\n",
      "[2]\tvalid_0's binary_logloss: 0.639396\n",
      "[3]\tvalid_0's binary_logloss: 0.625458\n",
      "[4]\tvalid_0's binary_logloss: 0.608388\n",
      "[5]\tvalid_0's binary_logloss: 0.598024\n",
      "[6]\tvalid_0's binary_logloss: 0.574237\n",
      "[7]\tvalid_0's binary_logloss: 0.555755\n",
      "[8]\tvalid_0's binary_logloss: 0.547474\n",
      "[9]\tvalid_0's binary_logloss: 0.528168\n",
      "[10]\tvalid_0's binary_logloss: 0.514808\n",
      "[11]\tvalid_0's binary_logloss: 0.507786\n",
      "[12]\tvalid_0's binary_logloss: 0.492716\n",
      "[13]\tvalid_0's binary_logloss: 0.481497\n",
      "[14]\tvalid_0's binary_logloss: 0.475785\n",
      "[15]\tvalid_0's binary_logloss: 0.466257\n",
      "[16]\tvalid_0's binary_logloss: 0.457621\n",
      "[17]\tvalid_0's binary_logloss: 0.448759\n",
      "[18]\tvalid_0's binary_logloss: 0.440352\n",
      "[19]\tvalid_0's binary_logloss: 0.433434\n",
      "[20]\tvalid_0's binary_logloss: 0.429871\n",
      "[21]\tvalid_0's binary_logloss: 0.420601\n",
      "[22]\tvalid_0's binary_logloss: 0.416955\n",
      "[23]\tvalid_0's binary_logloss: 0.414258\n",
      "[24]\tvalid_0's binary_logloss: 0.409\n",
      "[25]\tvalid_0's binary_logloss: 0.400881\n",
      "[26]\tvalid_0's binary_logloss: 0.396573\n",
      "[27]\tvalid_0's binary_logloss: 0.389337\n",
      "[28]\tvalid_0's binary_logloss: 0.382721\n",
      "[29]\tvalid_0's binary_logloss: 0.379449\n",
      "[30]\tvalid_0's binary_logloss: 0.377385\n",
      "[31]\tvalid_0's binary_logloss: 0.375061\n",
      "[32]\tvalid_0's binary_logloss: 0.369321\n",
      "[33]\tvalid_0's binary_logloss: 0.36716\n",
      "[34]\tvalid_0's binary_logloss: 0.365028\n",
      "[35]\tvalid_0's binary_logloss: 0.363109\n",
      "[36]\tvalid_0's binary_logloss: 0.361508\n",
      "[37]\tvalid_0's binary_logloss: 0.359113\n",
      "[38]\tvalid_0's binary_logloss: 0.354458\n",
      "[39]\tvalid_0's binary_logloss: 0.352252\n",
      "[40]\tvalid_0's binary_logloss: 0.349684\n",
      "[41]\tvalid_0's binary_logloss: 0.347269\n",
      "[42]\tvalid_0's binary_logloss: 0.345843\n",
      "[43]\tvalid_0's binary_logloss: 0.344477\n",
      "[44]\tvalid_0's binary_logloss: 0.339695\n",
      "[45]\tvalid_0's binary_logloss: 0.337859\n",
      "[46]\tvalid_0's binary_logloss: 0.336654\n",
      "[47]\tvalid_0's binary_logloss: 0.332592\n",
      "[48]\tvalid_0's binary_logloss: 0.331313\n",
      "[49]\tvalid_0's binary_logloss: 0.330193\n",
      "[50]\tvalid_0's binary_logloss: 0.328346\n",
      "[51]\tvalid_0's binary_logloss: 0.32387\n",
      "[52]\tvalid_0's binary_logloss: 0.32102\n",
      "[53]\tvalid_0's binary_logloss: 0.317721\n",
      "[54]\tvalid_0's binary_logloss: 0.314122\n",
      "[55]\tvalid_0's binary_logloss: 0.313281\n",
      "[56]\tvalid_0's binary_logloss: 0.310684\n",
      "[57]\tvalid_0's binary_logloss: 0.309313\n",
      "[58]\tvalid_0's binary_logloss: 0.306034\n",
      "[59]\tvalid_0's binary_logloss: 0.304932\n",
      "[60]\tvalid_0's binary_logloss: 0.303209\n",
      "[61]\tvalid_0's binary_logloss: 0.301457\n",
      "[62]\tvalid_0's binary_logloss: 0.300608\n",
      "[63]\tvalid_0's binary_logloss: 0.29788\n",
      "[64]\tvalid_0's binary_logloss: 0.296897\n",
      "[65]\tvalid_0's binary_logloss: 0.295657\n",
      "[66]\tvalid_0's binary_logloss: 0.2942\n",
      "[67]\tvalid_0's binary_logloss: 0.292359\n",
      "[68]\tvalid_0's binary_logloss: 0.291109\n",
      "[69]\tvalid_0's binary_logloss: 0.290296\n",
      "[70]\tvalid_0's binary_logloss: 0.289212\n",
      "[71]\tvalid_0's binary_logloss: 0.288322\n",
      "[72]\tvalid_0's binary_logloss: 0.285895\n",
      "[73]\tvalid_0's binary_logloss: 0.285107\n",
      "[74]\tvalid_0's binary_logloss: 0.283647\n",
      "[75]\tvalid_0's binary_logloss: 0.280003\n",
      "[76]\tvalid_0's binary_logloss: 0.279266\n",
      "[77]\tvalid_0's binary_logloss: 0.276103\n",
      "[78]\tvalid_0's binary_logloss: 0.274436\n",
      "[79]\tvalid_0's binary_logloss: 0.27342\n",
      "[80]\tvalid_0's binary_logloss: 0.271429\n",
      "[81]\tvalid_0's binary_logloss: 0.268923\n",
      "[82]\tvalid_0's binary_logloss: 0.267757\n",
      "[83]\tvalid_0's binary_logloss: 0.266117\n",
      "[84]\tvalid_0's binary_logloss: 0.265198\n",
      "[85]\tvalid_0's binary_logloss: 0.262603\n",
      "[86]\tvalid_0's binary_logloss: 0.261069\n",
      "[87]\tvalid_0's binary_logloss: 0.260133\n",
      "[88]\tvalid_0's binary_logloss: 0.259339\n",
      "[89]\tvalid_0's binary_logloss: 0.258302\n",
      "[90]\tvalid_0's binary_logloss: 0.25769\n",
      "[91]\tvalid_0's binary_logloss: 0.256883\n",
      "[92]\tvalid_0's binary_logloss: 0.254694\n",
      "[93]\tvalid_0's binary_logloss: 0.253434\n",
      "[94]\tvalid_0's binary_logloss: 0.25128\n",
      "[95]\tvalid_0's binary_logloss: 0.249541\n",
      "[96]\tvalid_0's binary_logloss: 0.248434\n",
      "[97]\tvalid_0's binary_logloss: 0.247711\n",
      "[98]\tvalid_0's binary_logloss: 0.246319\n",
      "[99]\tvalid_0's binary_logloss: 0.245604\n",
      "[100]\tvalid_0's binary_logloss: 0.244579\n",
      "[101]\tvalid_0's binary_logloss: 0.243223\n",
      "[102]\tvalid_0's binary_logloss: 0.242416\n",
      "[103]\tvalid_0's binary_logloss: 0.241824\n",
      "[104]\tvalid_0's binary_logloss: 0.240256\n",
      "[105]\tvalid_0's binary_logloss: 0.239438\n",
      "[106]\tvalid_0's binary_logloss: 0.23882\n",
      "[107]\tvalid_0's binary_logloss: 0.238201\n",
      "[108]\tvalid_0's binary_logloss: 0.236912\n",
      "[109]\tvalid_0's binary_logloss: 0.236229\n",
      "[110]\tvalid_0's binary_logloss: 0.235833\n",
      "[111]\tvalid_0's binary_logloss: 0.235245\n",
      "[112]\tvalid_0's binary_logloss: 0.234527\n",
      "[113]\tvalid_0's binary_logloss: 0.234121\n",
      "[114]\tvalid_0's binary_logloss: 0.233677\n",
      "[115]\tvalid_0's binary_logloss: 0.233186\n",
      "[116]\tvalid_0's binary_logloss: 0.232577\n",
      "[117]\tvalid_0's binary_logloss: 0.232295\n",
      "[118]\tvalid_0's binary_logloss: 0.231827\n",
      "[119]\tvalid_0's binary_logloss: 0.230753\n",
      "[120]\tvalid_0's binary_logloss: 0.230221\n",
      "[121]\tvalid_0's binary_logloss: 0.229801\n",
      "[122]\tvalid_0's binary_logloss: 0.229223\n",
      "[123]\tvalid_0's binary_logloss: 0.228813\n",
      "[124]\tvalid_0's binary_logloss: 0.22855\n",
      "[125]\tvalid_0's binary_logloss: 0.228096\n",
      "[126]\tvalid_0's binary_logloss: 0.22768\n",
      "[127]\tvalid_0's binary_logloss: 0.227382\n",
      "[128]\tvalid_0's binary_logloss: 0.226812\n",
      "[129]\tvalid_0's binary_logloss: 0.226479\n",
      "[130]\tvalid_0's binary_logloss: 0.225897\n",
      "[131]\tvalid_0's binary_logloss: 0.224815\n",
      "[132]\tvalid_0's binary_logloss: 0.22392\n",
      "[133]\tvalid_0's binary_logloss: 0.223013\n",
      "[134]\tvalid_0's binary_logloss: 0.222672\n",
      "[135]\tvalid_0's binary_logloss: 0.222312\n",
      "[136]\tvalid_0's binary_logloss: 0.222072\n",
      "[137]\tvalid_0's binary_logloss: 0.221748\n",
      "[138]\tvalid_0's binary_logloss: 0.221514\n",
      "[139]\tvalid_0's binary_logloss: 0.221101\n",
      "[140]\tvalid_0's binary_logloss: 0.220734\n",
      "[141]\tvalid_0's binary_logloss: 0.220274\n",
      "[142]\tvalid_0's binary_logloss: 0.21979\n",
      "[143]\tvalid_0's binary_logloss: 0.219162\n",
      "[144]\tvalid_0's binary_logloss: 0.219005\n",
      "[145]\tvalid_0's binary_logloss: 0.218685\n",
      "[146]\tvalid_0's binary_logloss: 0.218489\n",
      "[147]\tvalid_0's binary_logloss: 0.218193\n",
      "[148]\tvalid_0's binary_logloss: 0.217476\n",
      "[149]\tvalid_0's binary_logloss: 0.217094\n",
      "[150]\tvalid_0's binary_logloss: 0.216718\n",
      "[151]\tvalid_0's binary_logloss: 0.215737\n",
      "[152]\tvalid_0's binary_logloss: 0.215389\n",
      "[153]\tvalid_0's binary_logloss: 0.214965\n",
      "[154]\tvalid_0's binary_logloss: 0.214593\n",
      "[155]\tvalid_0's binary_logloss: 0.214009\n",
      "[156]\tvalid_0's binary_logloss: 0.213625\n",
      "[157]\tvalid_0's binary_logloss: 0.213276\n",
      "[158]\tvalid_0's binary_logloss: 0.213073\n",
      "[159]\tvalid_0's binary_logloss: 0.21293\n",
      "[160]\tvalid_0's binary_logloss: 0.212739\n",
      "[161]\tvalid_0's binary_logloss: 0.212497\n",
      "[162]\tvalid_0's binary_logloss: 0.21231\n",
      "[163]\tvalid_0's binary_logloss: 0.212127\n",
      "[164]\tvalid_0's binary_logloss: 0.211809\n",
      "[165]\tvalid_0's binary_logloss: 0.211383\n",
      "[166]\tvalid_0's binary_logloss: 0.211068\n",
      "[167]\tvalid_0's binary_logloss: 0.210878\n",
      "[168]\tvalid_0's binary_logloss: 0.210549\n",
      "[169]\tvalid_0's binary_logloss: 0.210289\n",
      "[170]\tvalid_0's binary_logloss: 0.209757\n",
      "[171]\tvalid_0's binary_logloss: 0.209566\n",
      "[172]\tvalid_0's binary_logloss: 0.209374\n",
      "[173]\tvalid_0's binary_logloss: 0.20924\n",
      "[174]\tvalid_0's binary_logloss: 0.208684\n",
      "[175]\tvalid_0's binary_logloss: 0.20853\n",
      "[176]\tvalid_0's binary_logloss: 0.208211\n",
      "[177]\tvalid_0's binary_logloss: 0.208073\n",
      "[178]\tvalid_0's binary_logloss: 0.207558\n",
      "[179]\tvalid_0's binary_logloss: 0.207367\n",
      "[180]\tvalid_0's binary_logloss: 0.206994\n",
      "[181]\tvalid_0's binary_logloss: 0.20666\n",
      "[182]\tvalid_0's binary_logloss: 0.206529\n",
      "[183]\tvalid_0's binary_logloss: 0.206443\n",
      "[184]\tvalid_0's binary_logloss: 0.206289\n",
      "[185]\tvalid_0's binary_logloss: 0.206115\n",
      "[186]\tvalid_0's binary_logloss: 0.205787\n",
      "[187]\tvalid_0's binary_logloss: 0.205683\n",
      "[188]\tvalid_0's binary_logloss: 0.205581\n",
      "[189]\tvalid_0's binary_logloss: 0.204886\n",
      "[190]\tvalid_0's binary_logloss: 0.204779\n",
      "[191]\tvalid_0's binary_logloss: 0.204692\n",
      "[192]\tvalid_0's binary_logloss: 0.204061\n",
      "[193]\tvalid_0's binary_logloss: 0.203506\n",
      "[194]\tvalid_0's binary_logloss: 0.203121\n",
      "[195]\tvalid_0's binary_logloss: 0.202536\n",
      "[196]\tvalid_0's binary_logloss: 0.202316\n",
      "[197]\tvalid_0's binary_logloss: 0.202177\n",
      "[198]\tvalid_0's binary_logloss: 0.201917\n",
      "[199]\tvalid_0's binary_logloss: 0.201703\n",
      "[200]\tvalid_0's binary_logloss: 0.201464\n",
      "[201]\tvalid_0's binary_logloss: 0.201306\n",
      "[202]\tvalid_0's binary_logloss: 0.201193\n",
      "[203]\tvalid_0's binary_logloss: 0.201084\n",
      "[204]\tvalid_0's binary_logloss: 0.200861\n",
      "[205]\tvalid_0's binary_logloss: 0.200793\n",
      "[206]\tvalid_0's binary_logloss: 0.200542\n",
      "[207]\tvalid_0's binary_logloss: 0.200279\n",
      "[208]\tvalid_0's binary_logloss: 0.19988\n",
      "[209]\tvalid_0's binary_logloss: 0.199758\n",
      "[210]\tvalid_0's binary_logloss: 0.199491\n",
      "[211]\tvalid_0's binary_logloss: 0.199379\n",
      "[212]\tvalid_0's binary_logloss: 0.198843\n",
      "[213]\tvalid_0's binary_logloss: 0.19864\n",
      "[214]\tvalid_0's binary_logloss: 0.198299\n",
      "[215]\tvalid_0's binary_logloss: 0.19812\n",
      "[216]\tvalid_0's binary_logloss: 0.198031\n",
      "[217]\tvalid_0's binary_logloss: 0.197813\n",
      "[218]\tvalid_0's binary_logloss: 0.197749\n",
      "[219]\tvalid_0's binary_logloss: 0.197576\n",
      "[220]\tvalid_0's binary_logloss: 0.197486\n",
      "[221]\tvalid_0's binary_logloss: 0.197374\n",
      "[222]\tvalid_0's binary_logloss: 0.197342\n",
      "[223]\tvalid_0's binary_logloss: 0.197282\n",
      "[224]\tvalid_0's binary_logloss: 0.197222\n",
      "[225]\tvalid_0's binary_logloss: 0.196786\n",
      "[226]\tvalid_0's binary_logloss: 0.196712\n",
      "[227]\tvalid_0's binary_logloss: 0.196663\n",
      "[228]\tvalid_0's binary_logloss: 0.196299\n",
      "[229]\tvalid_0's binary_logloss: 0.196139\n",
      "[230]\tvalid_0's binary_logloss: 0.196067\n",
      "[231]\tvalid_0's binary_logloss: 0.195909\n",
      "[232]\tvalid_0's binary_logloss: 0.195841\n",
      "[233]\tvalid_0's binary_logloss: 0.195651\n",
      "[234]\tvalid_0's binary_logloss: 0.195538\n",
      "[235]\tvalid_0's binary_logloss: 0.195433\n",
      "[236]\tvalid_0's binary_logloss: 0.195347\n",
      "[237]\tvalid_0's binary_logloss: 0.195299\n",
      "[238]\tvalid_0's binary_logloss: 0.195132\n",
      "[239]\tvalid_0's binary_logloss: 0.194824\n",
      "[240]\tvalid_0's binary_logloss: 0.194455\n",
      "[241]\tvalid_0's binary_logloss: 0.194216\n",
      "[242]\tvalid_0's binary_logloss: 0.194063\n",
      "[243]\tvalid_0's binary_logloss: 0.193881\n",
      "[244]\tvalid_0's binary_logloss: 0.19379\n",
      "[245]\tvalid_0's binary_logloss: 0.193645\n",
      "[246]\tvalid_0's binary_logloss: 0.193497\n",
      "[247]\tvalid_0's binary_logloss: 0.193412\n",
      "[248]\tvalid_0's binary_logloss: 0.19335\n",
      "[249]\tvalid_0's binary_logloss: 0.193233\n",
      "[250]\tvalid_0's binary_logloss: 0.193142\n",
      "[251]\tvalid_0's binary_logloss: 0.192904\n",
      "[252]\tvalid_0's binary_logloss: 0.192766\n",
      "[253]\tvalid_0's binary_logloss: 0.192574\n",
      "[254]\tvalid_0's binary_logloss: 0.192499\n",
      "[255]\tvalid_0's binary_logloss: 0.192444\n",
      "[256]\tvalid_0's binary_logloss: 0.192383\n",
      "[257]\tvalid_0's binary_logloss: 0.192308\n",
      "[258]\tvalid_0's binary_logloss: 0.192247\n",
      "[259]\tvalid_0's binary_logloss: 0.191926\n",
      "[260]\tvalid_0's binary_logloss: 0.19186\n",
      "[261]\tvalid_0's binary_logloss: 0.191691\n",
      "[262]\tvalid_0's binary_logloss: 0.191511\n",
      "[263]\tvalid_0's binary_logloss: 0.191475\n",
      "[264]\tvalid_0's binary_logloss: 0.191414\n",
      "[265]\tvalid_0's binary_logloss: 0.191269\n",
      "[266]\tvalid_0's binary_logloss: 0.191045\n",
      "[267]\tvalid_0's binary_logloss: 0.190768\n",
      "[268]\tvalid_0's binary_logloss: 0.19069\n",
      "[269]\tvalid_0's binary_logloss: 0.190529\n",
      "[270]\tvalid_0's binary_logloss: 0.190492\n",
      "[271]\tvalid_0's binary_logloss: 0.19045\n",
      "[272]\tvalid_0's binary_logloss: 0.190345\n",
      "[273]\tvalid_0's binary_logloss: 0.190269\n",
      "[274]\tvalid_0's binary_logloss: 0.190147\n",
      "[275]\tvalid_0's binary_logloss: 0.1901\n",
      "[276]\tvalid_0's binary_logloss: 0.190083\n",
      "[277]\tvalid_0's binary_logloss: 0.189705\n",
      "[278]\tvalid_0's binary_logloss: 0.189634\n",
      "[279]\tvalid_0's binary_logloss: 0.189523\n",
      "[280]\tvalid_0's binary_logloss: 0.189315\n",
      "[281]\tvalid_0's binary_logloss: 0.189068\n",
      "[282]\tvalid_0's binary_logloss: 0.188913\n",
      "[283]\tvalid_0's binary_logloss: 0.188872\n",
      "[284]\tvalid_0's binary_logloss: 0.188801\n",
      "[285]\tvalid_0's binary_logloss: 0.188671\n",
      "[286]\tvalid_0's binary_logloss: 0.188519\n",
      "[287]\tvalid_0's binary_logloss: 0.188419\n",
      "[288]\tvalid_0's binary_logloss: 0.188294\n",
      "[289]\tvalid_0's binary_logloss: 0.188148\n",
      "[290]\tvalid_0's binary_logloss: 0.187905\n",
      "[291]\tvalid_0's binary_logloss: 0.187898\n",
      "[292]\tvalid_0's binary_logloss: 0.187803\n",
      "[293]\tvalid_0's binary_logloss: 0.187756\n",
      "[294]\tvalid_0's binary_logloss: 0.18771\n",
      "[295]\tvalid_0's binary_logloss: 0.187492\n",
      "[296]\tvalid_0's binary_logloss: 0.18744\n",
      "[297]\tvalid_0's binary_logloss: 0.187286\n",
      "[298]\tvalid_0's binary_logloss: 0.187168\n",
      "[299]\tvalid_0's binary_logloss: 0.187145\n",
      "[300]\tvalid_0's binary_logloss: 0.187025\n",
      "[301]\tvalid_0's binary_logloss: 0.186923\n",
      "[302]\tvalid_0's binary_logloss: 0.186884\n",
      "[303]\tvalid_0's binary_logloss: 0.186669\n",
      "[304]\tvalid_0's binary_logloss: 0.186594\n",
      "[305]\tvalid_0's binary_logloss: 0.186484\n",
      "[306]\tvalid_0's binary_logloss: 0.186386\n",
      "[307]\tvalid_0's binary_logloss: 0.186284\n",
      "[308]\tvalid_0's binary_logloss: 0.186225\n",
      "[309]\tvalid_0's binary_logloss: 0.186139\n",
      "[310]\tvalid_0's binary_logloss: 0.18607\n",
      "[311]\tvalid_0's binary_logloss: 0.185947\n",
      "[312]\tvalid_0's binary_logloss: 0.185819\n",
      "[313]\tvalid_0's binary_logloss: 0.185748\n",
      "[314]\tvalid_0's binary_logloss: 0.185722\n",
      "[315]\tvalid_0's binary_logloss: 0.1857\n",
      "[316]\tvalid_0's binary_logloss: 0.185612\n",
      "[317]\tvalid_0's binary_logloss: 0.185508\n",
      "[318]\tvalid_0's binary_logloss: 0.185477\n",
      "[319]\tvalid_0's binary_logloss: 0.185387\n",
      "[320]\tvalid_0's binary_logloss: 0.185291\n",
      "[321]\tvalid_0's binary_logloss: 0.185232\n",
      "[322]\tvalid_0's binary_logloss: 0.185084\n",
      "[323]\tvalid_0's binary_logloss: 0.184991\n",
      "[324]\tvalid_0's binary_logloss: 0.184953\n",
      "[325]\tvalid_0's binary_logloss: 0.184848\n",
      "[326]\tvalid_0's binary_logloss: 0.184618\n",
      "[327]\tvalid_0's binary_logloss: 0.184513\n",
      "[328]\tvalid_0's binary_logloss: 0.184476\n",
      "[329]\tvalid_0's binary_logloss: 0.18427\n",
      "[330]\tvalid_0's binary_logloss: 0.18424\n",
      "[331]\tvalid_0's binary_logloss: 0.184207\n",
      "[332]\tvalid_0's binary_logloss: 0.184067\n",
      "[333]\tvalid_0's binary_logloss: 0.184053\n",
      "[334]\tvalid_0's binary_logloss: 0.184033\n",
      "[335]\tvalid_0's binary_logloss: 0.183874\n",
      "[336]\tvalid_0's binary_logloss: 0.183832\n",
      "[337]\tvalid_0's binary_logloss: 0.183745\n",
      "[338]\tvalid_0's binary_logloss: 0.183688\n",
      "[339]\tvalid_0's binary_logloss: 0.183656\n",
      "[340]\tvalid_0's binary_logloss: 0.183495\n",
      "[341]\tvalid_0's binary_logloss: 0.183414\n",
      "[342]\tvalid_0's binary_logloss: 0.183332\n",
      "[343]\tvalid_0's binary_logloss: 0.183239\n",
      "[344]\tvalid_0's binary_logloss: 0.183189\n",
      "[345]\tvalid_0's binary_logloss: 0.183114\n",
      "[346]\tvalid_0's binary_logloss: 0.183082\n",
      "[347]\tvalid_0's binary_logloss: 0.18306\n",
      "[348]\tvalid_0's binary_logloss: 0.183007\n",
      "[349]\tvalid_0's binary_logloss: 0.1829\n",
      "[350]\tvalid_0's binary_logloss: 0.182761\n",
      "[351]\tvalid_0's binary_logloss: 0.182736\n",
      "[352]\tvalid_0's binary_logloss: 0.182677\n",
      "[353]\tvalid_0's binary_logloss: 0.18262\n",
      "[354]\tvalid_0's binary_logloss: 0.182602\n",
      "[355]\tvalid_0's binary_logloss: 0.182552\n",
      "[356]\tvalid_0's binary_logloss: 0.182527\n",
      "[357]\tvalid_0's binary_logloss: 0.18252\n",
      "[358]\tvalid_0's binary_logloss: 0.182455\n",
      "[359]\tvalid_0's binary_logloss: 0.182451\n",
      "[360]\tvalid_0's binary_logloss: 0.182376\n",
      "[361]\tvalid_0's binary_logloss: 0.182286\n",
      "[362]\tvalid_0's binary_logloss: 0.182249\n",
      "[363]\tvalid_0's binary_logloss: 0.182233\n",
      "[364]\tvalid_0's binary_logloss: 0.182213\n",
      "[365]\tvalid_0's binary_logloss: 0.182167\n",
      "[366]\tvalid_0's binary_logloss: 0.182051\n",
      "[367]\tvalid_0's binary_logloss: 0.181979\n",
      "[368]\tvalid_0's binary_logloss: 0.181838\n",
      "[369]\tvalid_0's binary_logloss: 0.181786\n",
      "[370]\tvalid_0's binary_logloss: 0.181773\n",
      "[371]\tvalid_0's binary_logloss: 0.181735\n",
      "[372]\tvalid_0's binary_logloss: 0.181643\n",
      "[373]\tvalid_0's binary_logloss: 0.18156\n",
      "[374]\tvalid_0's binary_logloss: 0.181492\n",
      "[375]\tvalid_0's binary_logloss: 0.181421\n",
      "[376]\tvalid_0's binary_logloss: 0.181388\n",
      "[377]\tvalid_0's binary_logloss: 0.181388\n",
      "[378]\tvalid_0's binary_logloss: 0.181358\n",
      "[379]\tvalid_0's binary_logloss: 0.181286\n",
      "[380]\tvalid_0's binary_logloss: 0.181264\n",
      "[381]\tvalid_0's binary_logloss: 0.181197\n",
      "[382]\tvalid_0's binary_logloss: 0.181157\n",
      "[383]\tvalid_0's binary_logloss: 0.181112\n",
      "[384]\tvalid_0's binary_logloss: 0.181069\n",
      "[385]\tvalid_0's binary_logloss: 0.181038\n",
      "[386]\tvalid_0's binary_logloss: 0.181013\n",
      "[387]\tvalid_0's binary_logloss: 0.180988\n",
      "[388]\tvalid_0's binary_logloss: 0.180853\n",
      "[389]\tvalid_0's binary_logloss: 0.180761\n",
      "[390]\tvalid_0's binary_logloss: 0.180703\n",
      "[391]\tvalid_0's binary_logloss: 0.180636\n",
      "[392]\tvalid_0's binary_logloss: 0.180622\n",
      "[393]\tvalid_0's binary_logloss: 0.180611\n",
      "[394]\tvalid_0's binary_logloss: 0.180562\n",
      "[395]\tvalid_0's binary_logloss: 0.180506\n",
      "[396]\tvalid_0's binary_logloss: 0.18046\n",
      "[397]\tvalid_0's binary_logloss: 0.180358\n",
      "[398]\tvalid_0's binary_logloss: 0.180327\n",
      "[399]\tvalid_0's binary_logloss: 0.180307\n",
      "[400]\tvalid_0's binary_logloss: 0.180261\n",
      "[401]\tvalid_0's binary_logloss: 0.180255\n",
      "[402]\tvalid_0's binary_logloss: 0.18013\n",
      "[403]\tvalid_0's binary_logloss: 0.179985\n",
      "[404]\tvalid_0's binary_logloss: 0.179898\n",
      "[405]\tvalid_0's binary_logloss: 0.17984\n",
      "[406]\tvalid_0's binary_logloss: 0.179753\n",
      "[407]\tvalid_0's binary_logloss: 0.179618\n",
      "[408]\tvalid_0's binary_logloss: 0.17959\n",
      "[409]\tvalid_0's binary_logloss: 0.179528\n",
      "[410]\tvalid_0's binary_logloss: 0.179478\n",
      "[411]\tvalid_0's binary_logloss: 0.179432\n",
      "[412]\tvalid_0's binary_logloss: 0.179339\n",
      "[413]\tvalid_0's binary_logloss: 0.17923\n",
      "[414]\tvalid_0's binary_logloss: 0.179211\n",
      "[415]\tvalid_0's binary_logloss: 0.17913\n",
      "[416]\tvalid_0's binary_logloss: 0.17907\n",
      "[417]\tvalid_0's binary_logloss: 0.179053\n",
      "[418]\tvalid_0's binary_logloss: 0.178988\n",
      "[419]\tvalid_0's binary_logloss: 0.178894\n",
      "[420]\tvalid_0's binary_logloss: 0.178864\n",
      "[421]\tvalid_0's binary_logloss: 0.178791\n",
      "[422]\tvalid_0's binary_logloss: 0.178773\n",
      "[423]\tvalid_0's binary_logloss: 0.178742\n",
      "[424]\tvalid_0's binary_logloss: 0.178676\n",
      "[425]\tvalid_0's binary_logloss: 0.178669\n",
      "[426]\tvalid_0's binary_logloss: 0.178644\n",
      "[427]\tvalid_0's binary_logloss: 0.178531\n",
      "[428]\tvalid_0's binary_logloss: 0.178437\n",
      "[429]\tvalid_0's binary_logloss: 0.178411\n",
      "[430]\tvalid_0's binary_logloss: 0.178352\n",
      "[431]\tvalid_0's binary_logloss: 0.178323\n",
      "[432]\tvalid_0's binary_logloss: 0.178292\n",
      "[433]\tvalid_0's binary_logloss: 0.178219\n",
      "[434]\tvalid_0's binary_logloss: 0.178156\n",
      "[435]\tvalid_0's binary_logloss: 0.178087\n",
      "[436]\tvalid_0's binary_logloss: 0.178039\n",
      "[437]\tvalid_0's binary_logloss: 0.177987\n",
      "[438]\tvalid_0's binary_logloss: 0.177986\n",
      "[439]\tvalid_0's binary_logloss: 0.177947\n",
      "[440]\tvalid_0's binary_logloss: 0.177904\n",
      "[441]\tvalid_0's binary_logloss: 0.177864\n",
      "[442]\tvalid_0's binary_logloss: 0.177831\n",
      "[443]\tvalid_0's binary_logloss: 0.177811\n",
      "[444]\tvalid_0's binary_logloss: 0.177781\n",
      "[445]\tvalid_0's binary_logloss: 0.177756\n",
      "[446]\tvalid_0's binary_logloss: 0.177683\n",
      "[447]\tvalid_0's binary_logloss: 0.177648\n",
      "[448]\tvalid_0's binary_logloss: 0.177613\n",
      "[449]\tvalid_0's binary_logloss: 0.177539\n",
      "[450]\tvalid_0's binary_logloss: 0.177514\n",
      "[451]\tvalid_0's binary_logloss: 0.177453\n",
      "[452]\tvalid_0's binary_logloss: 0.177437\n",
      "[453]\tvalid_0's binary_logloss: 0.177372\n",
      "[454]\tvalid_0's binary_logloss: 0.177311\n",
      "[455]\tvalid_0's binary_logloss: 0.177295\n",
      "[456]\tvalid_0's binary_logloss: 0.177285\n",
      "[457]\tvalid_0's binary_logloss: 0.177216\n",
      "[458]\tvalid_0's binary_logloss: 0.177214\n",
      "[459]\tvalid_0's binary_logloss: 0.177151\n",
      "[460]\tvalid_0's binary_logloss: 0.177148\n",
      "[461]\tvalid_0's binary_logloss: 0.177133\n",
      "[462]\tvalid_0's binary_logloss: 0.177131\n",
      "[463]\tvalid_0's binary_logloss: 0.177115\n",
      "[464]\tvalid_0's binary_logloss: 0.17709\n",
      "[465]\tvalid_0's binary_logloss: 0.17707\n",
      "[466]\tvalid_0's binary_logloss: 0.177037\n",
      "[467]\tvalid_0's binary_logloss: 0.177024\n",
      "[468]\tvalid_0's binary_logloss: 0.17694\n",
      "[469]\tvalid_0's binary_logloss: 0.176895\n",
      "[470]\tvalid_0's binary_logloss: 0.176894\n",
      "[471]\tvalid_0's binary_logloss: 0.176887\n",
      "[472]\tvalid_0's binary_logloss: 0.17682\n",
      "[473]\tvalid_0's binary_logloss: 0.176777\n",
      "[474]\tvalid_0's binary_logloss: 0.176736\n",
      "[475]\tvalid_0's binary_logloss: 0.176713\n",
      "[476]\tvalid_0's binary_logloss: 0.17672\n",
      "[477]\tvalid_0's binary_logloss: 0.176681\n",
      "[478]\tvalid_0's binary_logloss: 0.176667\n",
      "[479]\tvalid_0's binary_logloss: 0.176631\n",
      "[480]\tvalid_0's binary_logloss: 0.176592\n",
      "[481]\tvalid_0's binary_logloss: 0.176565\n",
      "[482]\tvalid_0's binary_logloss: 0.176559\n",
      "[483]\tvalid_0's binary_logloss: 0.176503\n",
      "[484]\tvalid_0's binary_logloss: 0.176474\n",
      "[485]\tvalid_0's binary_logloss: 0.17646\n",
      "[486]\tvalid_0's binary_logloss: 0.17646\n",
      "[487]\tvalid_0's binary_logloss: 0.176456\n",
      "[488]\tvalid_0's binary_logloss: 0.17645\n",
      "[489]\tvalid_0's binary_logloss: 0.176383\n",
      "[490]\tvalid_0's binary_logloss: 0.176374\n",
      "[491]\tvalid_0's binary_logloss: 0.17637\n",
      "[492]\tvalid_0's binary_logloss: 0.176329\n",
      "[493]\tvalid_0's binary_logloss: 0.176288\n",
      "[494]\tvalid_0's binary_logloss: 0.176232\n",
      "[495]\tvalid_0's binary_logloss: 0.176201\n",
      "[496]\tvalid_0's binary_logloss: 0.176165\n",
      "[497]\tvalid_0's binary_logloss: 0.176132\n",
      "[498]\tvalid_0's binary_logloss: 0.176122\n",
      "[499]\tvalid_0's binary_logloss: 0.176086\n",
      "[500]\tvalid_0's binary_logloss: 0.176069\n",
      "[501]\tvalid_0's binary_logloss: 0.176065\n",
      "[502]\tvalid_0's binary_logloss: 0.176012\n",
      "[503]\tvalid_0's binary_logloss: 0.17597\n",
      "[504]\tvalid_0's binary_logloss: 0.17595\n",
      "[505]\tvalid_0's binary_logloss: 0.175924\n",
      "[506]\tvalid_0's binary_logloss: 0.175924\n",
      "[507]\tvalid_0's binary_logloss: 0.175896\n",
      "[508]\tvalid_0's binary_logloss: 0.175883\n",
      "[509]\tvalid_0's binary_logloss: 0.175807\n",
      "[510]\tvalid_0's binary_logloss: 0.175724\n",
      "[511]\tvalid_0's binary_logloss: 0.175667\n",
      "[512]\tvalid_0's binary_logloss: 0.175595\n",
      "[513]\tvalid_0's binary_logloss: 0.17555\n",
      "[514]\tvalid_0's binary_logloss: 0.175504\n",
      "[515]\tvalid_0's binary_logloss: 0.175462\n",
      "[516]\tvalid_0's binary_logloss: 0.175411\n",
      "[517]\tvalid_0's binary_logloss: 0.175381\n",
      "[518]\tvalid_0's binary_logloss: 0.175344\n",
      "[519]\tvalid_0's binary_logloss: 0.175343\n",
      "[520]\tvalid_0's binary_logloss: 0.175272\n",
      "[521]\tvalid_0's binary_logloss: 0.17526\n",
      "[522]\tvalid_0's binary_logloss: 0.175196\n",
      "[523]\tvalid_0's binary_logloss: 0.17517\n",
      "[524]\tvalid_0's binary_logloss: 0.175145\n",
      "[525]\tvalid_0's binary_logloss: 0.175122\n",
      "[526]\tvalid_0's binary_logloss: 0.175079\n",
      "[527]\tvalid_0's binary_logloss: 0.175076\n",
      "[528]\tvalid_0's binary_logloss: 0.175015\n",
      "[529]\tvalid_0's binary_logloss: 0.174989\n",
      "[530]\tvalid_0's binary_logloss: 0.174971\n",
      "[531]\tvalid_0's binary_logloss: 0.174945\n",
      "[532]\tvalid_0's binary_logloss: 0.174909\n",
      "[533]\tvalid_0's binary_logloss: 0.174892\n",
      "[534]\tvalid_0's binary_logloss: 0.174887\n",
      "[535]\tvalid_0's binary_logloss: 0.174877\n",
      "[536]\tvalid_0's binary_logloss: 0.17487\n",
      "[537]\tvalid_0's binary_logloss: 0.17479\n",
      "[538]\tvalid_0's binary_logloss: 0.174767\n",
      "[539]\tvalid_0's binary_logloss: 0.174753\n",
      "[540]\tvalid_0's binary_logloss: 0.174728\n",
      "[541]\tvalid_0's binary_logloss: 0.174709\n",
      "[542]\tvalid_0's binary_logloss: 0.174658\n",
      "[543]\tvalid_0's binary_logloss: 0.174628\n",
      "[544]\tvalid_0's binary_logloss: 0.174621\n",
      "[545]\tvalid_0's binary_logloss: 0.174594\n",
      "[546]\tvalid_0's binary_logloss: 0.174578\n",
      "[547]\tvalid_0's binary_logloss: 0.174582\n",
      "[548]\tvalid_0's binary_logloss: 0.174511\n",
      "[549]\tvalid_0's binary_logloss: 0.174491\n",
      "[550]\tvalid_0's binary_logloss: 0.174463\n",
      "[551]\tvalid_0's binary_logloss: 0.174461\n",
      "[552]\tvalid_0's binary_logloss: 0.174441\n",
      "[553]\tvalid_0's binary_logloss: 0.174433\n",
      "[554]\tvalid_0's binary_logloss: 0.174421\n",
      "[555]\tvalid_0's binary_logloss: 0.174412\n",
      "[556]\tvalid_0's binary_logloss: 0.174353\n",
      "[557]\tvalid_0's binary_logloss: 0.174331\n",
      "[558]\tvalid_0's binary_logloss: 0.174329\n",
      "[559]\tvalid_0's binary_logloss: 0.174296\n",
      "[560]\tvalid_0's binary_logloss: 0.174241\n",
      "[561]\tvalid_0's binary_logloss: 0.174211\n",
      "[562]\tvalid_0's binary_logloss: 0.174171\n",
      "[563]\tvalid_0's binary_logloss: 0.174105\n",
      "[564]\tvalid_0's binary_logloss: 0.174066\n",
      "[565]\tvalid_0's binary_logloss: 0.17406\n",
      "[566]\tvalid_0's binary_logloss: 0.174056\n",
      "[567]\tvalid_0's binary_logloss: 0.174023\n",
      "[568]\tvalid_0's binary_logloss: 0.173966\n",
      "[569]\tvalid_0's binary_logloss: 0.173938\n",
      "[570]\tvalid_0's binary_logloss: 0.173913\n",
      "[571]\tvalid_0's binary_logloss: 0.173892\n",
      "[572]\tvalid_0's binary_logloss: 0.173877\n",
      "[573]\tvalid_0's binary_logloss: 0.173848\n",
      "[574]\tvalid_0's binary_logloss: 0.173827\n",
      "[575]\tvalid_0's binary_logloss: 0.173798\n",
      "[576]\tvalid_0's binary_logloss: 0.173756\n",
      "[577]\tvalid_0's binary_logloss: 0.173728\n",
      "[578]\tvalid_0's binary_logloss: 0.173715\n",
      "[579]\tvalid_0's binary_logloss: 0.173702\n",
      "[580]\tvalid_0's binary_logloss: 0.17367\n",
      "[581]\tvalid_0's binary_logloss: 0.173666\n",
      "[582]\tvalid_0's binary_logloss: 0.17364\n",
      "[583]\tvalid_0's binary_logloss: 0.173611\n",
      "[584]\tvalid_0's binary_logloss: 0.173592\n",
      "[585]\tvalid_0's binary_logloss: 0.173598\n",
      "[586]\tvalid_0's binary_logloss: 0.173597\n",
      "[587]\tvalid_0's binary_logloss: 0.173597\n",
      "[588]\tvalid_0's binary_logloss: 0.173552\n",
      "[589]\tvalid_0's binary_logloss: 0.173524\n",
      "[590]\tvalid_0's binary_logloss: 0.17351\n",
      "[591]\tvalid_0's binary_logloss: 0.173486\n",
      "[592]\tvalid_0's binary_logloss: 0.173469\n",
      "[593]\tvalid_0's binary_logloss: 0.173456\n",
      "[594]\tvalid_0's binary_logloss: 0.173449\n",
      "[595]\tvalid_0's binary_logloss: 0.173444\n",
      "[596]\tvalid_0's binary_logloss: 0.173406\n",
      "[597]\tvalid_0's binary_logloss: 0.173373\n",
      "[598]\tvalid_0's binary_logloss: 0.173371\n",
      "[599]\tvalid_0's binary_logloss: 0.173328\n",
      "[600]\tvalid_0's binary_logloss: 0.17332\n",
      "[601]\tvalid_0's binary_logloss: 0.173296\n",
      "[602]\tvalid_0's binary_logloss: 0.173261\n",
      "[603]\tvalid_0's binary_logloss: 0.173209\n",
      "[604]\tvalid_0's binary_logloss: 0.173173\n",
      "[605]\tvalid_0's binary_logloss: 0.173157\n",
      "[606]\tvalid_0's binary_logloss: 0.173154\n",
      "[607]\tvalid_0's binary_logloss: 0.173101\n",
      "[608]\tvalid_0's binary_logloss: 0.173084\n",
      "[609]\tvalid_0's binary_logloss: 0.173053\n",
      "[610]\tvalid_0's binary_logloss: 0.173042\n",
      "[611]\tvalid_0's binary_logloss: 0.17301\n",
      "[612]\tvalid_0's binary_logloss: 0.172985\n",
      "[613]\tvalid_0's binary_logloss: 0.172968\n",
      "[614]\tvalid_0's binary_logloss: 0.172938\n",
      "[615]\tvalid_0's binary_logloss: 0.172937\n",
      "[616]\tvalid_0's binary_logloss: 0.172903\n",
      "[617]\tvalid_0's binary_logloss: 0.17286\n",
      "[618]\tvalid_0's binary_logloss: 0.172831\n",
      "[619]\tvalid_0's binary_logloss: 0.172805\n",
      "[620]\tvalid_0's binary_logloss: 0.172757\n",
      "[621]\tvalid_0's binary_logloss: 0.172728\n",
      "[622]\tvalid_0's binary_logloss: 0.172687\n",
      "[623]\tvalid_0's binary_logloss: 0.172665\n",
      "[624]\tvalid_0's binary_logloss: 0.17263\n",
      "[625]\tvalid_0's binary_logloss: 0.172609\n",
      "[626]\tvalid_0's binary_logloss: 0.172602\n",
      "[627]\tvalid_0's binary_logloss: 0.172596\n",
      "[628]\tvalid_0's binary_logloss: 0.172578\n",
      "[629]\tvalid_0's binary_logloss: 0.172561\n",
      "[630]\tvalid_0's binary_logloss: 0.17255\n",
      "[631]\tvalid_0's binary_logloss: 0.172547\n",
      "[632]\tvalid_0's binary_logloss: 0.172528\n",
      "[633]\tvalid_0's binary_logloss: 0.172505\n",
      "[634]\tvalid_0's binary_logloss: 0.172494\n",
      "[635]\tvalid_0's binary_logloss: 0.172481\n",
      "[636]\tvalid_0's binary_logloss: 0.172464\n",
      "[637]\tvalid_0's binary_logloss: 0.172458\n",
      "[638]\tvalid_0's binary_logloss: 0.172435\n",
      "[639]\tvalid_0's binary_logloss: 0.172437\n",
      "[640]\tvalid_0's binary_logloss: 0.172423\n",
      "[641]\tvalid_0's binary_logloss: 0.172392\n",
      "[642]\tvalid_0's binary_logloss: 0.172383\n",
      "[643]\tvalid_0's binary_logloss: 0.172368\n",
      "[644]\tvalid_0's binary_logloss: 0.172358\n",
      "[645]\tvalid_0's binary_logloss: 0.172348\n",
      "[646]\tvalid_0's binary_logloss: 0.172328\n",
      "[647]\tvalid_0's binary_logloss: 0.172283\n",
      "[648]\tvalid_0's binary_logloss: 0.172258\n",
      "[649]\tvalid_0's binary_logloss: 0.172215\n",
      "[650]\tvalid_0's binary_logloss: 0.172186\n",
      "[651]\tvalid_0's binary_logloss: 0.172161\n",
      "[652]\tvalid_0's binary_logloss: 0.172144\n",
      "[653]\tvalid_0's binary_logloss: 0.172115\n",
      "[654]\tvalid_0's binary_logloss: 0.172092\n",
      "[655]\tvalid_0's binary_logloss: 0.17207\n",
      "[656]\tvalid_0's binary_logloss: 0.172053\n",
      "[657]\tvalid_0's binary_logloss: 0.172023\n",
      "[658]\tvalid_0's binary_logloss: 0.171991\n",
      "[659]\tvalid_0's binary_logloss: 0.171979\n",
      "[660]\tvalid_0's binary_logloss: 0.171967\n",
      "[661]\tvalid_0's binary_logloss: 0.171968\n",
      "[662]\tvalid_0's binary_logloss: 0.171956\n",
      "[663]\tvalid_0's binary_logloss: 0.171962\n",
      "[664]\tvalid_0's binary_logloss: 0.171945\n",
      "[665]\tvalid_0's binary_logloss: 0.1719\n",
      "[666]\tvalid_0's binary_logloss: 0.171865\n",
      "[667]\tvalid_0's binary_logloss: 0.171866\n",
      "[668]\tvalid_0's binary_logloss: 0.171859\n",
      "[669]\tvalid_0's binary_logloss: 0.17182\n",
      "[670]\tvalid_0's binary_logloss: 0.171808\n",
      "[671]\tvalid_0's binary_logloss: 0.171796\n",
      "[672]\tvalid_0's binary_logloss: 0.171787\n",
      "[673]\tvalid_0's binary_logloss: 0.17179\n",
      "[674]\tvalid_0's binary_logloss: 0.17177\n",
      "[675]\tvalid_0's binary_logloss: 0.171776\n",
      "[676]\tvalid_0's binary_logloss: 0.171751\n",
      "[677]\tvalid_0's binary_logloss: 0.171724\n",
      "[678]\tvalid_0's binary_logloss: 0.171708\n",
      "[679]\tvalid_0's binary_logloss: 0.171674\n",
      "[680]\tvalid_0's binary_logloss: 0.17165\n",
      "[681]\tvalid_0's binary_logloss: 0.171636\n",
      "[682]\tvalid_0's binary_logloss: 0.171616\n",
      "[683]\tvalid_0's binary_logloss: 0.171603\n",
      "[684]\tvalid_0's binary_logloss: 0.171568\n",
      "[685]\tvalid_0's binary_logloss: 0.171573\n",
      "[686]\tvalid_0's binary_logloss: 0.171551\n",
      "[687]\tvalid_0's binary_logloss: 0.171521\n",
      "[688]\tvalid_0's binary_logloss: 0.171511\n",
      "[689]\tvalid_0's binary_logloss: 0.171493\n",
      "[690]\tvalid_0's binary_logloss: 0.171491\n",
      "[691]\tvalid_0's binary_logloss: 0.171491\n",
      "[692]\tvalid_0's binary_logloss: 0.171493\n",
      "[693]\tvalid_0's binary_logloss: 0.171495\n",
      "[694]\tvalid_0's binary_logloss: 0.171479\n",
      "[695]\tvalid_0's binary_logloss: 0.171436\n",
      "[696]\tvalid_0's binary_logloss: 0.1714\n",
      "[697]\tvalid_0's binary_logloss: 0.171384\n",
      "[698]\tvalid_0's binary_logloss: 0.171365\n",
      "[699]\tvalid_0's binary_logloss: 0.171334\n",
      "[700]\tvalid_0's binary_logloss: 0.171331\n",
      "[701]\tvalid_0's binary_logloss: 0.171312\n",
      "[702]\tvalid_0's binary_logloss: 0.171302\n",
      "[703]\tvalid_0's binary_logloss: 0.171265\n",
      "[704]\tvalid_0's binary_logloss: 0.171245\n",
      "[705]\tvalid_0's binary_logloss: 0.171235\n",
      "[706]\tvalid_0's binary_logloss: 0.171208\n",
      "[707]\tvalid_0's binary_logloss: 0.171183\n",
      "[708]\tvalid_0's binary_logloss: 0.171185\n",
      "[709]\tvalid_0's binary_logloss: 0.171175\n",
      "[710]\tvalid_0's binary_logloss: 0.171146\n",
      "[711]\tvalid_0's binary_logloss: 0.171127\n",
      "[712]\tvalid_0's binary_logloss: 0.171117\n",
      "[713]\tvalid_0's binary_logloss: 0.171102\n",
      "[714]\tvalid_0's binary_logloss: 0.171084\n",
      "[715]\tvalid_0's binary_logloss: 0.171078\n",
      "[716]\tvalid_0's binary_logloss: 0.171063\n",
      "[717]\tvalid_0's binary_logloss: 0.171028\n",
      "[718]\tvalid_0's binary_logloss: 0.171013\n",
      "[719]\tvalid_0's binary_logloss: 0.171003\n",
      "[720]\tvalid_0's binary_logloss: 0.170996\n",
      "[721]\tvalid_0's binary_logloss: 0.170996\n",
      "[722]\tvalid_0's binary_logloss: 0.170992\n",
      "[723]\tvalid_0's binary_logloss: 0.170957\n",
      "[724]\tvalid_0's binary_logloss: 0.170925\n",
      "[725]\tvalid_0's binary_logloss: 0.170926\n",
      "[726]\tvalid_0's binary_logloss: 0.170924\n",
      "[727]\tvalid_0's binary_logloss: 0.170895\n",
      "[728]\tvalid_0's binary_logloss: 0.170873\n",
      "[729]\tvalid_0's binary_logloss: 0.170854\n",
      "[730]\tvalid_0's binary_logloss: 0.170821\n",
      "[731]\tvalid_0's binary_logloss: 0.170812\n",
      "[732]\tvalid_0's binary_logloss: 0.170804\n",
      "[733]\tvalid_0's binary_logloss: 0.170788\n",
      "[734]\tvalid_0's binary_logloss: 0.170771\n",
      "[735]\tvalid_0's binary_logloss: 0.170747\n",
      "[736]\tvalid_0's binary_logloss: 0.170731\n",
      "[737]\tvalid_0's binary_logloss: 0.170711\n",
      "[738]\tvalid_0's binary_logloss: 0.170712\n",
      "[739]\tvalid_0's binary_logloss: 0.170702\n",
      "[740]\tvalid_0's binary_logloss: 0.170679\n",
      "[741]\tvalid_0's binary_logloss: 0.170661\n",
      "[742]\tvalid_0's binary_logloss: 0.170632\n",
      "[743]\tvalid_0's binary_logloss: 0.170609\n",
      "[744]\tvalid_0's binary_logloss: 0.170595\n",
      "[745]\tvalid_0's binary_logloss: 0.170596\n",
      "[746]\tvalid_0's binary_logloss: 0.170572\n",
      "[747]\tvalid_0's binary_logloss: 0.170569\n",
      "[748]\tvalid_0's binary_logloss: 0.170541\n",
      "[749]\tvalid_0's binary_logloss: 0.170536\n",
      "[750]\tvalid_0's binary_logloss: 0.170526\n",
      "[751]\tvalid_0's binary_logloss: 0.170486\n",
      "[752]\tvalid_0's binary_logloss: 0.170478\n",
      "[753]\tvalid_0's binary_logloss: 0.170449\n",
      "[754]\tvalid_0's binary_logloss: 0.170419\n",
      "[755]\tvalid_0's binary_logloss: 0.170406\n",
      "[756]\tvalid_0's binary_logloss: 0.17039\n",
      "[757]\tvalid_0's binary_logloss: 0.170381\n",
      "[758]\tvalid_0's binary_logloss: 0.170362\n",
      "[759]\tvalid_0's binary_logloss: 0.170338\n",
      "[760]\tvalid_0's binary_logloss: 0.170328\n",
      "[761]\tvalid_0's binary_logloss: 0.170301\n",
      "[762]\tvalid_0's binary_logloss: 0.170274\n",
      "[763]\tvalid_0's binary_logloss: 0.170242\n",
      "[764]\tvalid_0's binary_logloss: 0.170219\n",
      "[765]\tvalid_0's binary_logloss: 0.170202\n",
      "[766]\tvalid_0's binary_logloss: 0.170188\n",
      "[767]\tvalid_0's binary_logloss: 0.170191\n",
      "[768]\tvalid_0's binary_logloss: 0.170192\n",
      "[769]\tvalid_0's binary_logloss: 0.170164\n",
      "[770]\tvalid_0's binary_logloss: 0.170143\n",
      "[771]\tvalid_0's binary_logloss: 0.170134\n",
      "[772]\tvalid_0's binary_logloss: 0.170136\n",
      "[773]\tvalid_0's binary_logloss: 0.17013\n",
      "[774]\tvalid_0's binary_logloss: 0.170118\n",
      "[775]\tvalid_0's binary_logloss: 0.170108\n",
      "[776]\tvalid_0's binary_logloss: 0.170101\n",
      "[777]\tvalid_0's binary_logloss: 0.170093\n",
      "[778]\tvalid_0's binary_logloss: 0.170064\n",
      "[779]\tvalid_0's binary_logloss: 0.170043\n",
      "[780]\tvalid_0's binary_logloss: 0.170041\n",
      "[781]\tvalid_0's binary_logloss: 0.170032\n",
      "[782]\tvalid_0's binary_logloss: 0.17003\n",
      "[783]\tvalid_0's binary_logloss: 0.170018\n",
      "[784]\tvalid_0's binary_logloss: 0.170007\n",
      "[785]\tvalid_0's binary_logloss: 0.17\n",
      "[786]\tvalid_0's binary_logloss: 0.169992\n",
      "[787]\tvalid_0's binary_logloss: 0.169979\n",
      "[788]\tvalid_0's binary_logloss: 0.169975\n",
      "[789]\tvalid_0's binary_logloss: 0.169956\n",
      "[790]\tvalid_0's binary_logloss: 0.169937\n",
      "[791]\tvalid_0's binary_logloss: 0.169914\n",
      "[792]\tvalid_0's binary_logloss: 0.169895\n",
      "[793]\tvalid_0's binary_logloss: 0.169871\n",
      "[794]\tvalid_0's binary_logloss: 0.169861\n",
      "[795]\tvalid_0's binary_logloss: 0.169833\n",
      "[796]\tvalid_0's binary_logloss: 0.169826\n",
      "[797]\tvalid_0's binary_logloss: 0.169814\n",
      "[798]\tvalid_0's binary_logloss: 0.1698\n",
      "[799]\tvalid_0's binary_logloss: 0.169787\n",
      "[800]\tvalid_0's binary_logloss: 0.169769\n",
      "[801]\tvalid_0's binary_logloss: 0.169765\n",
      "[802]\tvalid_0's binary_logloss: 0.169752\n",
      "[803]\tvalid_0's binary_logloss: 0.169735\n",
      "[804]\tvalid_0's binary_logloss: 0.169715\n",
      "[805]\tvalid_0's binary_logloss: 0.169709\n",
      "[806]\tvalid_0's binary_logloss: 0.16972\n",
      "[807]\tvalid_0's binary_logloss: 0.169709\n",
      "[808]\tvalid_0's binary_logloss: 0.169711\n",
      "[809]\tvalid_0's binary_logloss: 0.169688\n",
      "[810]\tvalid_0's binary_logloss: 0.169664\n",
      "[811]\tvalid_0's binary_logloss: 0.169655\n",
      "[812]\tvalid_0's binary_logloss: 0.169634\n",
      "[813]\tvalid_0's binary_logloss: 0.169617\n",
      "[814]\tvalid_0's binary_logloss: 0.169601\n",
      "[815]\tvalid_0's binary_logloss: 0.169594\n",
      "[816]\tvalid_0's binary_logloss: 0.169569\n",
      "[817]\tvalid_0's binary_logloss: 0.169548\n",
      "[818]\tvalid_0's binary_logloss: 0.169526\n",
      "[819]\tvalid_0's binary_logloss: 0.169501\n",
      "[820]\tvalid_0's binary_logloss: 0.169494\n",
      "[821]\tvalid_0's binary_logloss: 0.169495\n",
      "[822]\tvalid_0's binary_logloss: 0.169477\n",
      "[823]\tvalid_0's binary_logloss: 0.169464\n",
      "[824]\tvalid_0's binary_logloss: 0.169461\n",
      "[825]\tvalid_0's binary_logloss: 0.169424\n",
      "[826]\tvalid_0's binary_logloss: 0.169408\n",
      "[827]\tvalid_0's binary_logloss: 0.169402\n",
      "[828]\tvalid_0's binary_logloss: 0.169381\n",
      "[829]\tvalid_0's binary_logloss: 0.169362\n",
      "[830]\tvalid_0's binary_logloss: 0.169348\n",
      "[831]\tvalid_0's binary_logloss: 0.169347\n",
      "[832]\tvalid_0's binary_logloss: 0.169337\n",
      "[833]\tvalid_0's binary_logloss: 0.169321\n",
      "[834]\tvalid_0's binary_logloss: 0.169313\n",
      "[835]\tvalid_0's binary_logloss: 0.169301\n",
      "[836]\tvalid_0's binary_logloss: 0.169293\n",
      "[837]\tvalid_0's binary_logloss: 0.169288\n",
      "[838]\tvalid_0's binary_logloss: 0.169277\n",
      "[839]\tvalid_0's binary_logloss: 0.169244\n",
      "[840]\tvalid_0's binary_logloss: 0.16923\n",
      "[841]\tvalid_0's binary_logloss: 0.169232\n",
      "[842]\tvalid_0's binary_logloss: 0.169216\n",
      "[843]\tvalid_0's binary_logloss: 0.169196\n",
      "[844]\tvalid_0's binary_logloss: 0.169187\n",
      "[845]\tvalid_0's binary_logloss: 0.169179\n",
      "[846]\tvalid_0's binary_logloss: 0.169179\n",
      "[847]\tvalid_0's binary_logloss: 0.169172\n",
      "[848]\tvalid_0's binary_logloss: 0.169159\n",
      "[849]\tvalid_0's binary_logloss: 0.169154\n",
      "[850]\tvalid_0's binary_logloss: 0.169144\n",
      "[851]\tvalid_0's binary_logloss: 0.169143\n",
      "[852]\tvalid_0's binary_logloss: 0.169145\n",
      "[853]\tvalid_0's binary_logloss: 0.16914\n",
      "[854]\tvalid_0's binary_logloss: 0.169137\n",
      "[855]\tvalid_0's binary_logloss: 0.169132\n",
      "[856]\tvalid_0's binary_logloss: 0.169124\n",
      "[857]\tvalid_0's binary_logloss: 0.169123\n",
      "[858]\tvalid_0's binary_logloss: 0.169121\n",
      "[859]\tvalid_0's binary_logloss: 0.169101\n",
      "[860]\tvalid_0's binary_logloss: 0.169085\n",
      "[861]\tvalid_0's binary_logloss: 0.169061\n",
      "[862]\tvalid_0's binary_logloss: 0.169063\n",
      "[863]\tvalid_0's binary_logloss: 0.169064\n",
      "[864]\tvalid_0's binary_logloss: 0.169059\n",
      "[865]\tvalid_0's binary_logloss: 0.169054\n",
      "[866]\tvalid_0's binary_logloss: 0.169056\n",
      "[867]\tvalid_0's binary_logloss: 0.169056\n",
      "[868]\tvalid_0's binary_logloss: 0.169048\n",
      "[869]\tvalid_0's binary_logloss: 0.169042\n",
      "[870]\tvalid_0's binary_logloss: 0.169042\n",
      "[871]\tvalid_0's binary_logloss: 0.169036\n",
      "[872]\tvalid_0's binary_logloss: 0.169034\n",
      "[873]\tvalid_0's binary_logloss: 0.169014\n",
      "[874]\tvalid_0's binary_logloss: 0.168992\n",
      "[875]\tvalid_0's binary_logloss: 0.168986\n",
      "[876]\tvalid_0's binary_logloss: 0.16898\n",
      "[877]\tvalid_0's binary_logloss: 0.168967\n",
      "[878]\tvalid_0's binary_logloss: 0.168945\n",
      "[879]\tvalid_0's binary_logloss: 0.168946\n",
      "[880]\tvalid_0's binary_logloss: 0.168938\n",
      "[881]\tvalid_0's binary_logloss: 0.168933\n",
      "[882]\tvalid_0's binary_logloss: 0.168927\n",
      "[883]\tvalid_0's binary_logloss: 0.168907\n",
      "[884]\tvalid_0's binary_logloss: 0.168903\n",
      "[885]\tvalid_0's binary_logloss: 0.168889\n",
      "[886]\tvalid_0's binary_logloss: 0.168866\n",
      "[887]\tvalid_0's binary_logloss: 0.168856\n",
      "[888]\tvalid_0's binary_logloss: 0.16884\n",
      "[889]\tvalid_0's binary_logloss: 0.168829\n",
      "[890]\tvalid_0's binary_logloss: 0.168832\n",
      "[891]\tvalid_0's binary_logloss: 0.168823\n",
      "[892]\tvalid_0's binary_logloss: 0.168809\n",
      "[893]\tvalid_0's binary_logloss: 0.168802\n",
      "[894]\tvalid_0's binary_logloss: 0.168794\n",
      "[895]\tvalid_0's binary_logloss: 0.168785\n",
      "[896]\tvalid_0's binary_logloss: 0.168775\n",
      "[897]\tvalid_0's binary_logloss: 0.168764\n",
      "[898]\tvalid_0's binary_logloss: 0.168755\n",
      "[899]\tvalid_0's binary_logloss: 0.168741\n",
      "[900]\tvalid_0's binary_logloss: 0.168739\n",
      "[901]\tvalid_0's binary_logloss: 0.168744\n",
      "[902]\tvalid_0's binary_logloss: 0.168741\n",
      "[903]\tvalid_0's binary_logloss: 0.16874\n",
      "[904]\tvalid_0's binary_logloss: 0.168726\n",
      "[905]\tvalid_0's binary_logloss: 0.168706\n",
      "[906]\tvalid_0's binary_logloss: 0.168685\n",
      "[907]\tvalid_0's binary_logloss: 0.168669\n",
      "[908]\tvalid_0's binary_logloss: 0.168647\n",
      "[909]\tvalid_0's binary_logloss: 0.168646\n",
      "[910]\tvalid_0's binary_logloss: 0.168625\n",
      "[911]\tvalid_0's binary_logloss: 0.168618\n",
      "[912]\tvalid_0's binary_logloss: 0.168608\n",
      "[913]\tvalid_0's binary_logloss: 0.168583\n",
      "[914]\tvalid_0's binary_logloss: 0.168573\n",
      "[915]\tvalid_0's binary_logloss: 0.168555\n",
      "[916]\tvalid_0's binary_logloss: 0.168544\n",
      "[917]\tvalid_0's binary_logloss: 0.168541\n",
      "[918]\tvalid_0's binary_logloss: 0.168534\n",
      "[919]\tvalid_0's binary_logloss: 0.168522\n",
      "[920]\tvalid_0's binary_logloss: 0.168506\n",
      "[921]\tvalid_0's binary_logloss: 0.168494\n",
      "[922]\tvalid_0's binary_logloss: 0.168489\n",
      "[923]\tvalid_0's binary_logloss: 0.168476\n",
      "[924]\tvalid_0's binary_logloss: 0.168476\n",
      "[925]\tvalid_0's binary_logloss: 0.168462\n",
      "[926]\tvalid_0's binary_logloss: 0.168462\n",
      "[927]\tvalid_0's binary_logloss: 0.168459\n",
      "[928]\tvalid_0's binary_logloss: 0.168448\n",
      "[929]\tvalid_0's binary_logloss: 0.168439\n",
      "[930]\tvalid_0's binary_logloss: 0.16843\n",
      "[931]\tvalid_0's binary_logloss: 0.168434\n",
      "[932]\tvalid_0's binary_logloss: 0.1684\n",
      "[933]\tvalid_0's binary_logloss: 0.168404\n",
      "[934]\tvalid_0's binary_logloss: 0.168395\n",
      "[935]\tvalid_0's binary_logloss: 0.16838\n",
      "[936]\tvalid_0's binary_logloss: 0.168381\n",
      "[937]\tvalid_0's binary_logloss: 0.168378\n",
      "[938]\tvalid_0's binary_logloss: 0.168358\n",
      "[939]\tvalid_0's binary_logloss: 0.16835\n",
      "[940]\tvalid_0's binary_logloss: 0.168344\n",
      "[941]\tvalid_0's binary_logloss: 0.168345\n",
      "[942]\tvalid_0's binary_logloss: 0.168337\n",
      "[943]\tvalid_0's binary_logloss: 0.168313\n",
      "[944]\tvalid_0's binary_logloss: 0.1683\n",
      "[945]\tvalid_0's binary_logloss: 0.168297\n",
      "[946]\tvalid_0's binary_logloss: 0.168288\n",
      "[947]\tvalid_0's binary_logloss: 0.168275\n",
      "[948]\tvalid_0's binary_logloss: 0.168266\n",
      "[949]\tvalid_0's binary_logloss: 0.168243\n",
      "[950]\tvalid_0's binary_logloss: 0.168236\n",
      "[951]\tvalid_0's binary_logloss: 0.168233\n",
      "[952]\tvalid_0's binary_logloss: 0.168234\n",
      "[953]\tvalid_0's binary_logloss: 0.168227\n",
      "[954]\tvalid_0's binary_logloss: 0.16823\n",
      "[955]\tvalid_0's binary_logloss: 0.168227\n",
      "[956]\tvalid_0's binary_logloss: 0.168218\n",
      "[957]\tvalid_0's binary_logloss: 0.168206\n",
      "[958]\tvalid_0's binary_logloss: 0.168205\n",
      "[959]\tvalid_0's binary_logloss: 0.168189\n",
      "[960]\tvalid_0's binary_logloss: 0.168175\n",
      "[961]\tvalid_0's binary_logloss: 0.168169\n",
      "[962]\tvalid_0's binary_logloss: 0.168172\n",
      "[963]\tvalid_0's binary_logloss: 0.168169\n",
      "[964]\tvalid_0's binary_logloss: 0.168157\n",
      "[965]\tvalid_0's binary_logloss: 0.168147\n",
      "[966]\tvalid_0's binary_logloss: 0.168134\n",
      "[967]\tvalid_0's binary_logloss: 0.168125\n",
      "[968]\tvalid_0's binary_logloss: 0.168111\n",
      "[969]\tvalid_0's binary_logloss: 0.16811\n",
      "[970]\tvalid_0's binary_logloss: 0.168101\n",
      "[971]\tvalid_0's binary_logloss: 0.168084\n",
      "[972]\tvalid_0's binary_logloss: 0.168087\n",
      "[973]\tvalid_0's binary_logloss: 0.168074\n",
      "[974]\tvalid_0's binary_logloss: 0.168068\n",
      "[975]\tvalid_0's binary_logloss: 0.168062\n",
      "[976]\tvalid_0's binary_logloss: 0.168057\n",
      "[977]\tvalid_0's binary_logloss: 0.168054\n",
      "[978]\tvalid_0's binary_logloss: 0.168036\n",
      "[979]\tvalid_0's binary_logloss: 0.16801\n",
      "[980]\tvalid_0's binary_logloss: 0.168002\n",
      "[981]\tvalid_0's binary_logloss: 0.167988\n",
      "[982]\tvalid_0's binary_logloss: 0.167977\n",
      "[983]\tvalid_0's binary_logloss: 0.167969\n",
      "[984]\tvalid_0's binary_logloss: 0.167961\n",
      "[985]\tvalid_0's binary_logloss: 0.167947\n",
      "[986]\tvalid_0's binary_logloss: 0.167927\n",
      "[987]\tvalid_0's binary_logloss: 0.167902\n",
      "[988]\tvalid_0's binary_logloss: 0.167894\n",
      "[989]\tvalid_0's binary_logloss: 0.167885\n",
      "[990]\tvalid_0's binary_logloss: 0.167878\n",
      "[991]\tvalid_0's binary_logloss: 0.167869\n",
      "[992]\tvalid_0's binary_logloss: 0.167866\n",
      "[993]\tvalid_0's binary_logloss: 0.167854\n",
      "[994]\tvalid_0's binary_logloss: 0.167837\n",
      "[995]\tvalid_0's binary_logloss: 0.167809\n",
      "[996]\tvalid_0's binary_logloss: 0.1678\n",
      "[997]\tvalid_0's binary_logloss: 0.167785\n",
      "[998]\tvalid_0's binary_logloss: 0.167776\n",
      "[999]\tvalid_0's binary_logloss: 0.167766\n",
      "[1000]\tvalid_0's binary_logloss: 0.167766\n",
      "training model for CV #5\n",
      "[LightGBM] [Warning] feature_fraction is set=0.4, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.4\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.0, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.0\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8772503866491428, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8772503866491428\n",
      "[LightGBM] [Warning] lambda_l2 is set=0.0, reg_lambda=0.0 will be ignored. Current value: lambda_l2=0.0\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\engine.py:177: UserWarning: Found `num_iterations` in params. Will use it instead of argument\n",
      "  _log_warning(f\"Found `{alias}` in params. Will use it instead of argument\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tvalid_0's binary_logloss: 0.662414\n",
      "[2]\tvalid_0's binary_logloss: 0.639877\n",
      "[3]\tvalid_0's binary_logloss: 0.626009\n",
      "[4]\tvalid_0's binary_logloss: 0.609063\n",
      "[5]\tvalid_0's binary_logloss: 0.59879\n",
      "[6]\tvalid_0's binary_logloss: 0.575428\n",
      "[7]\tvalid_0's binary_logloss: 0.557087\n",
      "[8]\tvalid_0's binary_logloss: 0.548918\n",
      "[9]\tvalid_0's binary_logloss: 0.528846\n",
      "[10]\tvalid_0's binary_logloss: 0.515701\n",
      "[11]\tvalid_0's binary_logloss: 0.508686\n",
      "[12]\tvalid_0's binary_logloss: 0.492766\n",
      "[13]\tvalid_0's binary_logloss: 0.481526\n",
      "[14]\tvalid_0's binary_logloss: 0.475835\n",
      "[15]\tvalid_0's binary_logloss: 0.466227\n",
      "[16]\tvalid_0's binary_logloss: 0.45782\n",
      "[17]\tvalid_0's binary_logloss: 0.446257\n",
      "[18]\tvalid_0's binary_logloss: 0.435844\n",
      "[19]\tvalid_0's binary_logloss: 0.429278\n",
      "[20]\tvalid_0's binary_logloss: 0.42579\n",
      "[21]\tvalid_0's binary_logloss: 0.41733\n",
      "[22]\tvalid_0's binary_logloss: 0.414426\n",
      "[23]\tvalid_0's binary_logloss: 0.411886\n",
      "[24]\tvalid_0's binary_logloss: 0.406824\n",
      "[25]\tvalid_0's binary_logloss: 0.398468\n",
      "[26]\tvalid_0's binary_logloss: 0.394247\n",
      "[27]\tvalid_0's binary_logloss: 0.387796\n",
      "[28]\tvalid_0's binary_logloss: 0.383497\n",
      "[29]\tvalid_0's binary_logloss: 0.380267\n",
      "[30]\tvalid_0's binary_logloss: 0.378145\n",
      "[31]\tvalid_0's binary_logloss: 0.375865\n",
      "[32]\tvalid_0's binary_logloss: 0.369749\n",
      "[33]\tvalid_0's binary_logloss: 0.367661\n",
      "[34]\tvalid_0's binary_logloss: 0.365436\n",
      "[35]\tvalid_0's binary_logloss: 0.363511\n",
      "[36]\tvalid_0's binary_logloss: 0.361995\n",
      "[37]\tvalid_0's binary_logloss: 0.359592\n",
      "[38]\tvalid_0's binary_logloss: 0.354987\n",
      "[39]\tvalid_0's binary_logloss: 0.352717\n",
      "[40]\tvalid_0's binary_logloss: 0.35022\n",
      "[41]\tvalid_0's binary_logloss: 0.347927\n",
      "[42]\tvalid_0's binary_logloss: 0.346526\n",
      "[43]\tvalid_0's binary_logloss: 0.345338\n",
      "[44]\tvalid_0's binary_logloss: 0.340731\n",
      "[45]\tvalid_0's binary_logloss: 0.338621\n",
      "[46]\tvalid_0's binary_logloss: 0.33726\n",
      "[47]\tvalid_0's binary_logloss: 0.332562\n",
      "[48]\tvalid_0's binary_logloss: 0.331188\n",
      "[49]\tvalid_0's binary_logloss: 0.330135\n",
      "[50]\tvalid_0's binary_logloss: 0.328248\n",
      "[51]\tvalid_0's binary_logloss: 0.32466\n",
      "[52]\tvalid_0's binary_logloss: 0.323517\n",
      "[53]\tvalid_0's binary_logloss: 0.320491\n",
      "[54]\tvalid_0's binary_logloss: 0.316783\n",
      "[55]\tvalid_0's binary_logloss: 0.315945\n",
      "[56]\tvalid_0's binary_logloss: 0.313389\n",
      "[57]\tvalid_0's binary_logloss: 0.312002\n",
      "[58]\tvalid_0's binary_logloss: 0.308127\n",
      "[59]\tvalid_0's binary_logloss: 0.30702\n",
      "[60]\tvalid_0's binary_logloss: 0.305317\n",
      "[61]\tvalid_0's binary_logloss: 0.303673\n",
      "[62]\tvalid_0's binary_logloss: 0.302846\n",
      "[63]\tvalid_0's binary_logloss: 0.300027\n",
      "[64]\tvalid_0's binary_logloss: 0.299321\n",
      "[65]\tvalid_0's binary_logloss: 0.298102\n",
      "[66]\tvalid_0's binary_logloss: 0.296717\n",
      "[67]\tvalid_0's binary_logloss: 0.294856\n",
      "[68]\tvalid_0's binary_logloss: 0.293672\n",
      "[69]\tvalid_0's binary_logloss: 0.292282\n",
      "[70]\tvalid_0's binary_logloss: 0.291127\n",
      "[71]\tvalid_0's binary_logloss: 0.29042\n",
      "[72]\tvalid_0's binary_logloss: 0.288441\n",
      "[73]\tvalid_0's binary_logloss: 0.287316\n",
      "[74]\tvalid_0's binary_logloss: 0.285125\n",
      "[75]\tvalid_0's binary_logloss: 0.281251\n",
      "[76]\tvalid_0's binary_logloss: 0.280544\n",
      "[77]\tvalid_0's binary_logloss: 0.277188\n",
      "[78]\tvalid_0's binary_logloss: 0.275593\n",
      "[79]\tvalid_0's binary_logloss: 0.27468\n",
      "[80]\tvalid_0's binary_logloss: 0.272193\n",
      "[81]\tvalid_0's binary_logloss: 0.270427\n",
      "[82]\tvalid_0's binary_logloss: 0.269297\n",
      "[83]\tvalid_0's binary_logloss: 0.268184\n",
      "[84]\tvalid_0's binary_logloss: 0.267139\n",
      "[85]\tvalid_0's binary_logloss: 0.264391\n",
      "[86]\tvalid_0's binary_logloss: 0.262911\n",
      "[87]\tvalid_0's binary_logloss: 0.261082\n",
      "[88]\tvalid_0's binary_logloss: 0.260238\n",
      "[89]\tvalid_0's binary_logloss: 0.259327\n",
      "[90]\tvalid_0's binary_logloss: 0.258501\n",
      "[91]\tvalid_0's binary_logloss: 0.257633\n",
      "[92]\tvalid_0's binary_logloss: 0.255315\n",
      "[93]\tvalid_0's binary_logloss: 0.254458\n",
      "[94]\tvalid_0's binary_logloss: 0.252163\n",
      "[95]\tvalid_0's binary_logloss: 0.250557\n",
      "[96]\tvalid_0's binary_logloss: 0.249468\n",
      "[97]\tvalid_0's binary_logloss: 0.248709\n",
      "[98]\tvalid_0's binary_logloss: 0.248112\n",
      "[99]\tvalid_0's binary_logloss: 0.247783\n",
      "[100]\tvalid_0's binary_logloss: 0.24681\n",
      "[101]\tvalid_0's binary_logloss: 0.246118\n",
      "[102]\tvalid_0's binary_logloss: 0.245248\n",
      "[103]\tvalid_0's binary_logloss: 0.244565\n",
      "[104]\tvalid_0's binary_logloss: 0.243376\n",
      "[105]\tvalid_0's binary_logloss: 0.242054\n",
      "[106]\tvalid_0's binary_logloss: 0.24116\n",
      "[107]\tvalid_0's binary_logloss: 0.240014\n",
      "[108]\tvalid_0's binary_logloss: 0.238432\n",
      "[109]\tvalid_0's binary_logloss: 0.238143\n",
      "[110]\tvalid_0's binary_logloss: 0.237417\n",
      "[111]\tvalid_0's binary_logloss: 0.236848\n",
      "[112]\tvalid_0's binary_logloss: 0.235991\n",
      "[113]\tvalid_0's binary_logloss: 0.234754\n",
      "[114]\tvalid_0's binary_logloss: 0.23433\n",
      "[115]\tvalid_0's binary_logloss: 0.233818\n",
      "[116]\tvalid_0's binary_logloss: 0.232702\n",
      "[117]\tvalid_0's binary_logloss: 0.232286\n",
      "[118]\tvalid_0's binary_logloss: 0.231833\n",
      "[119]\tvalid_0's binary_logloss: 0.231028\n",
      "[120]\tvalid_0's binary_logloss: 0.23065\n",
      "[121]\tvalid_0's binary_logloss: 0.230197\n",
      "[122]\tvalid_0's binary_logloss: 0.229557\n",
      "[123]\tvalid_0's binary_logloss: 0.229046\n",
      "[124]\tvalid_0's binary_logloss: 0.22851\n",
      "[125]\tvalid_0's binary_logloss: 0.227543\n",
      "[126]\tvalid_0's binary_logloss: 0.227103\n",
      "[127]\tvalid_0's binary_logloss: 0.226773\n",
      "[128]\tvalid_0's binary_logloss: 0.226265\n",
      "[129]\tvalid_0's binary_logloss: 0.2257\n",
      "[130]\tvalid_0's binary_logloss: 0.225228\n",
      "[131]\tvalid_0's binary_logloss: 0.224729\n",
      "[132]\tvalid_0's binary_logloss: 0.224226\n",
      "[133]\tvalid_0's binary_logloss: 0.223879\n",
      "[134]\tvalid_0's binary_logloss: 0.223566\n",
      "[135]\tvalid_0's binary_logloss: 0.223156\n",
      "[136]\tvalid_0's binary_logloss: 0.222678\n",
      "[137]\tvalid_0's binary_logloss: 0.222466\n",
      "[138]\tvalid_0's binary_logloss: 0.221999\n",
      "[139]\tvalid_0's binary_logloss: 0.221655\n",
      "[140]\tvalid_0's binary_logloss: 0.220854\n",
      "[141]\tvalid_0's binary_logloss: 0.220674\n",
      "[142]\tvalid_0's binary_logloss: 0.219926\n",
      "[143]\tvalid_0's binary_logloss: 0.219386\n",
      "[144]\tvalid_0's binary_logloss: 0.219244\n",
      "[145]\tvalid_0's binary_logloss: 0.219018\n",
      "[146]\tvalid_0's binary_logloss: 0.218747\n",
      "[147]\tvalid_0's binary_logloss: 0.21854\n",
      "[148]\tvalid_0's binary_logloss: 0.217827\n",
      "[149]\tvalid_0's binary_logloss: 0.217394\n",
      "[150]\tvalid_0's binary_logloss: 0.2166\n",
      "[151]\tvalid_0's binary_logloss: 0.216159\n",
      "[152]\tvalid_0's binary_logloss: 0.215672\n",
      "[153]\tvalid_0's binary_logloss: 0.215461\n",
      "[154]\tvalid_0's binary_logloss: 0.215058\n",
      "[155]\tvalid_0's binary_logloss: 0.214626\n",
      "[156]\tvalid_0's binary_logloss: 0.214163\n",
      "[157]\tvalid_0's binary_logloss: 0.213701\n",
      "[158]\tvalid_0's binary_logloss: 0.213344\n",
      "[159]\tvalid_0's binary_logloss: 0.213209\n",
      "[160]\tvalid_0's binary_logloss: 0.213043\n",
      "[161]\tvalid_0's binary_logloss: 0.212801\n",
      "[162]\tvalid_0's binary_logloss: 0.212461\n",
      "[163]\tvalid_0's binary_logloss: 0.211927\n",
      "[164]\tvalid_0's binary_logloss: 0.211598\n",
      "[165]\tvalid_0's binary_logloss: 0.211441\n",
      "[166]\tvalid_0's binary_logloss: 0.211241\n",
      "[167]\tvalid_0's binary_logloss: 0.211136\n",
      "[168]\tvalid_0's binary_logloss: 0.210942\n",
      "[169]\tvalid_0's binary_logloss: 0.210807\n",
      "[170]\tvalid_0's binary_logloss: 0.210153\n",
      "[171]\tvalid_0's binary_logloss: 0.209648\n",
      "[172]\tvalid_0's binary_logloss: 0.20944\n",
      "[173]\tvalid_0's binary_logloss: 0.209314\n",
      "[174]\tvalid_0's binary_logloss: 0.209028\n",
      "[175]\tvalid_0's binary_logloss: 0.208297\n",
      "[176]\tvalid_0's binary_logloss: 0.208073\n",
      "[177]\tvalid_0's binary_logloss: 0.207938\n",
      "[178]\tvalid_0's binary_logloss: 0.207663\n",
      "[179]\tvalid_0's binary_logloss: 0.207097\n",
      "[180]\tvalid_0's binary_logloss: 0.206735\n",
      "[181]\tvalid_0's binary_logloss: 0.206506\n",
      "[182]\tvalid_0's binary_logloss: 0.206361\n",
      "[183]\tvalid_0's binary_logloss: 0.206155\n",
      "[184]\tvalid_0's binary_logloss: 0.205985\n",
      "[185]\tvalid_0's binary_logloss: 0.205859\n",
      "[186]\tvalid_0's binary_logloss: 0.205747\n",
      "[187]\tvalid_0's binary_logloss: 0.205603\n",
      "[188]\tvalid_0's binary_logloss: 0.204976\n",
      "[189]\tvalid_0's binary_logloss: 0.204569\n",
      "[190]\tvalid_0's binary_logloss: 0.204437\n",
      "[191]\tvalid_0's binary_logloss: 0.204306\n",
      "[192]\tvalid_0's binary_logloss: 0.20412\n",
      "[193]\tvalid_0's binary_logloss: 0.203907\n",
      "[194]\tvalid_0's binary_logloss: 0.203581\n",
      "[195]\tvalid_0's binary_logloss: 0.203443\n",
      "[196]\tvalid_0's binary_logloss: 0.203235\n",
      "[197]\tvalid_0's binary_logloss: 0.203088\n",
      "[198]\tvalid_0's binary_logloss: 0.202813\n",
      "[199]\tvalid_0's binary_logloss: 0.202633\n",
      "[200]\tvalid_0's binary_logloss: 0.202461\n",
      "[201]\tvalid_0's binary_logloss: 0.202303\n",
      "[202]\tvalid_0's binary_logloss: 0.202115\n",
      "[203]\tvalid_0's binary_logloss: 0.201968\n",
      "[204]\tvalid_0's binary_logloss: 0.20183\n",
      "[205]\tvalid_0's binary_logloss: 0.201605\n",
      "[206]\tvalid_0's binary_logloss: 0.201373\n",
      "[207]\tvalid_0's binary_logloss: 0.200982\n",
      "[208]\tvalid_0's binary_logloss: 0.200486\n",
      "[209]\tvalid_0's binary_logloss: 0.200173\n",
      "[210]\tvalid_0's binary_logloss: 0.200047\n",
      "[211]\tvalid_0's binary_logloss: 0.199867\n",
      "[212]\tvalid_0's binary_logloss: 0.199782\n",
      "[213]\tvalid_0's binary_logloss: 0.199691\n",
      "[214]\tvalid_0's binary_logloss: 0.199494\n",
      "[215]\tvalid_0's binary_logloss: 0.19942\n",
      "[216]\tvalid_0's binary_logloss: 0.199362\n",
      "[217]\tvalid_0's binary_logloss: 0.199154\n",
      "[218]\tvalid_0's binary_logloss: 0.199026\n",
      "[219]\tvalid_0's binary_logloss: 0.198817\n",
      "[220]\tvalid_0's binary_logloss: 0.198733\n",
      "[221]\tvalid_0's binary_logloss: 0.198645\n",
      "[222]\tvalid_0's binary_logloss: 0.198577\n",
      "[223]\tvalid_0's binary_logloss: 0.198462\n",
      "[224]\tvalid_0's binary_logloss: 0.198421\n",
      "[225]\tvalid_0's binary_logloss: 0.198348\n",
      "[226]\tvalid_0's binary_logloss: 0.198065\n",
      "[227]\tvalid_0's binary_logloss: 0.198008\n",
      "[228]\tvalid_0's binary_logloss: 0.197709\n",
      "[229]\tvalid_0's binary_logloss: 0.197625\n",
      "[230]\tvalid_0's binary_logloss: 0.197252\n",
      "[231]\tvalid_0's binary_logloss: 0.197218\n",
      "[232]\tvalid_0's binary_logloss: 0.197107\n",
      "[233]\tvalid_0's binary_logloss: 0.196872\n",
      "[234]\tvalid_0's binary_logloss: 0.196711\n",
      "[235]\tvalid_0's binary_logloss: 0.196622\n",
      "[236]\tvalid_0's binary_logloss: 0.196319\n",
      "[237]\tvalid_0's binary_logloss: 0.196112\n",
      "[238]\tvalid_0's binary_logloss: 0.196043\n",
      "[239]\tvalid_0's binary_logloss: 0.195982\n",
      "[240]\tvalid_0's binary_logloss: 0.1958\n",
      "[241]\tvalid_0's binary_logloss: 0.195742\n",
      "[242]\tvalid_0's binary_logloss: 0.195618\n",
      "[243]\tvalid_0's binary_logloss: 0.195363\n",
      "[244]\tvalid_0's binary_logloss: 0.195259\n",
      "[245]\tvalid_0's binary_logloss: 0.195122\n",
      "[246]\tvalid_0's binary_logloss: 0.195064\n",
      "[247]\tvalid_0's binary_logloss: 0.19495\n",
      "[248]\tvalid_0's binary_logloss: 0.19469\n",
      "[249]\tvalid_0's binary_logloss: 0.19459\n",
      "[250]\tvalid_0's binary_logloss: 0.194465\n",
      "[251]\tvalid_0's binary_logloss: 0.194082\n",
      "[252]\tvalid_0's binary_logloss: 0.193925\n",
      "[253]\tvalid_0's binary_logloss: 0.193646\n",
      "[254]\tvalid_0's binary_logloss: 0.193625\n",
      "[255]\tvalid_0's binary_logloss: 0.193568\n",
      "[256]\tvalid_0's binary_logloss: 0.193507\n",
      "[257]\tvalid_0's binary_logloss: 0.193395\n",
      "[258]\tvalid_0's binary_logloss: 0.193263\n",
      "[259]\tvalid_0's binary_logloss: 0.192922\n",
      "[260]\tvalid_0's binary_logloss: 0.192823\n",
      "[261]\tvalid_0's binary_logloss: 0.192705\n",
      "[262]\tvalid_0's binary_logloss: 0.192485\n",
      "[263]\tvalid_0's binary_logloss: 0.192466\n",
      "[264]\tvalid_0's binary_logloss: 0.192402\n",
      "[265]\tvalid_0's binary_logloss: 0.192243\n",
      "[266]\tvalid_0's binary_logloss: 0.191976\n",
      "[267]\tvalid_0's binary_logloss: 0.191781\n",
      "[268]\tvalid_0's binary_logloss: 0.191714\n",
      "[269]\tvalid_0's binary_logloss: 0.191421\n",
      "[270]\tvalid_0's binary_logloss: 0.191397\n",
      "[271]\tvalid_0's binary_logloss: 0.191289\n",
      "[272]\tvalid_0's binary_logloss: 0.191207\n",
      "[273]\tvalid_0's binary_logloss: 0.19114\n",
      "[274]\tvalid_0's binary_logloss: 0.191102\n",
      "[275]\tvalid_0's binary_logloss: 0.190831\n",
      "[276]\tvalid_0's binary_logloss: 0.190791\n",
      "[277]\tvalid_0's binary_logloss: 0.190726\n",
      "[278]\tvalid_0's binary_logloss: 0.190646\n",
      "[279]\tvalid_0's binary_logloss: 0.190591\n",
      "[280]\tvalid_0's binary_logloss: 0.190498\n",
      "[281]\tvalid_0's binary_logloss: 0.190418\n",
      "[282]\tvalid_0's binary_logloss: 0.190245\n",
      "[283]\tvalid_0's binary_logloss: 0.190218\n",
      "[284]\tvalid_0's binary_logloss: 0.190116\n",
      "[285]\tvalid_0's binary_logloss: 0.190035\n",
      "[286]\tvalid_0's binary_logloss: 0.189875\n",
      "[287]\tvalid_0's binary_logloss: 0.189853\n",
      "[288]\tvalid_0's binary_logloss: 0.1898\n",
      "[289]\tvalid_0's binary_logloss: 0.189597\n",
      "[290]\tvalid_0's binary_logloss: 0.189306\n",
      "[291]\tvalid_0's binary_logloss: 0.189243\n",
      "[292]\tvalid_0's binary_logloss: 0.188953\n",
      "[293]\tvalid_0's binary_logloss: 0.188807\n",
      "[294]\tvalid_0's binary_logloss: 0.188654\n",
      "[295]\tvalid_0's binary_logloss: 0.188498\n",
      "[296]\tvalid_0's binary_logloss: 0.188438\n",
      "[297]\tvalid_0's binary_logloss: 0.188289\n",
      "[298]\tvalid_0's binary_logloss: 0.188101\n",
      "[299]\tvalid_0's binary_logloss: 0.188092\n",
      "[300]\tvalid_0's binary_logloss: 0.188075\n",
      "[301]\tvalid_0's binary_logloss: 0.187997\n",
      "[302]\tvalid_0's binary_logloss: 0.187953\n",
      "[303]\tvalid_0's binary_logloss: 0.187825\n",
      "[304]\tvalid_0's binary_logloss: 0.187692\n",
      "[305]\tvalid_0's binary_logloss: 0.187505\n",
      "[306]\tvalid_0's binary_logloss: 0.187446\n",
      "[307]\tvalid_0's binary_logloss: 0.18732\n",
      "[308]\tvalid_0's binary_logloss: 0.187288\n",
      "[309]\tvalid_0's binary_logloss: 0.18713\n",
      "[310]\tvalid_0's binary_logloss: 0.187077\n",
      "[311]\tvalid_0's binary_logloss: 0.186986\n",
      "[312]\tvalid_0's binary_logloss: 0.18687\n",
      "[313]\tvalid_0's binary_logloss: 0.186814\n",
      "[314]\tvalid_0's binary_logloss: 0.186757\n",
      "[315]\tvalid_0's binary_logloss: 0.186685\n",
      "[316]\tvalid_0's binary_logloss: 0.186515\n",
      "[317]\tvalid_0's binary_logloss: 0.186483\n",
      "[318]\tvalid_0's binary_logloss: 0.186354\n",
      "[319]\tvalid_0's binary_logloss: 0.186251\n",
      "[320]\tvalid_0's binary_logloss: 0.186212\n",
      "[321]\tvalid_0's binary_logloss: 0.186071\n",
      "[322]\tvalid_0's binary_logloss: 0.185916\n",
      "[323]\tvalid_0's binary_logloss: 0.185868\n",
      "[324]\tvalid_0's binary_logloss: 0.185816\n",
      "[325]\tvalid_0's binary_logloss: 0.18575\n",
      "[326]\tvalid_0's binary_logloss: 0.185724\n",
      "[327]\tvalid_0's binary_logloss: 0.185699\n",
      "[328]\tvalid_0's binary_logloss: 0.185545\n",
      "[329]\tvalid_0's binary_logloss: 0.185435\n",
      "[330]\tvalid_0's binary_logloss: 0.185404\n",
      "[331]\tvalid_0's binary_logloss: 0.18533\n",
      "[332]\tvalid_0's binary_logloss: 0.185153\n",
      "[333]\tvalid_0's binary_logloss: 0.185031\n",
      "[334]\tvalid_0's binary_logloss: 0.185003\n",
      "[335]\tvalid_0's binary_logloss: 0.184894\n",
      "[336]\tvalid_0's binary_logloss: 0.184835\n",
      "[337]\tvalid_0's binary_logloss: 0.184788\n",
      "[338]\tvalid_0's binary_logloss: 0.184714\n",
      "[339]\tvalid_0's binary_logloss: 0.184601\n",
      "[340]\tvalid_0's binary_logloss: 0.18449\n",
      "[341]\tvalid_0's binary_logloss: 0.184461\n",
      "[342]\tvalid_0's binary_logloss: 0.184379\n",
      "[343]\tvalid_0's binary_logloss: 0.184269\n",
      "[344]\tvalid_0's binary_logloss: 0.184239\n",
      "[345]\tvalid_0's binary_logloss: 0.184235\n",
      "[346]\tvalid_0's binary_logloss: 0.184157\n",
      "[347]\tvalid_0's binary_logloss: 0.184101\n",
      "[348]\tvalid_0's binary_logloss: 0.184069\n",
      "[349]\tvalid_0's binary_logloss: 0.184044\n",
      "[350]\tvalid_0's binary_logloss: 0.183994\n",
      "[351]\tvalid_0's binary_logloss: 0.183907\n",
      "[352]\tvalid_0's binary_logloss: 0.183856\n",
      "[353]\tvalid_0's binary_logloss: 0.183778\n",
      "[354]\tvalid_0's binary_logloss: 0.183669\n",
      "[355]\tvalid_0's binary_logloss: 0.183544\n",
      "[356]\tvalid_0's binary_logloss: 0.183479\n",
      "[357]\tvalid_0's binary_logloss: 0.183423\n",
      "[358]\tvalid_0's binary_logloss: 0.183401\n",
      "[359]\tvalid_0's binary_logloss: 0.183397\n",
      "[360]\tvalid_0's binary_logloss: 0.183368\n",
      "[361]\tvalid_0's binary_logloss: 0.183343\n",
      "[362]\tvalid_0's binary_logloss: 0.18323\n",
      "[363]\tvalid_0's binary_logloss: 0.183138\n",
      "[364]\tvalid_0's binary_logloss: 0.183113\n",
      "[365]\tvalid_0's binary_logloss: 0.183051\n",
      "[366]\tvalid_0's binary_logloss: 0.18295\n",
      "[367]\tvalid_0's binary_logloss: 0.182817\n",
      "[368]\tvalid_0's binary_logloss: 0.182698\n",
      "[369]\tvalid_0's binary_logloss: 0.182539\n",
      "[370]\tvalid_0's binary_logloss: 0.182496\n",
      "[371]\tvalid_0's binary_logloss: 0.18241\n",
      "[372]\tvalid_0's binary_logloss: 0.182342\n",
      "[373]\tvalid_0's binary_logloss: 0.182242\n",
      "[374]\tvalid_0's binary_logloss: 0.182215\n",
      "[375]\tvalid_0's binary_logloss: 0.182119\n",
      "[376]\tvalid_0's binary_logloss: 0.182013\n",
      "[377]\tvalid_0's binary_logloss: 0.181989\n",
      "[378]\tvalid_0's binary_logloss: 0.181941\n",
      "[379]\tvalid_0's binary_logloss: 0.181843\n",
      "[380]\tvalid_0's binary_logloss: 0.181815\n",
      "[381]\tvalid_0's binary_logloss: 0.181767\n",
      "[382]\tvalid_0's binary_logloss: 0.181706\n",
      "[383]\tvalid_0's binary_logloss: 0.181629\n",
      "[384]\tvalid_0's binary_logloss: 0.18155\n",
      "[385]\tvalid_0's binary_logloss: 0.181501\n",
      "[386]\tvalid_0's binary_logloss: 0.181469\n",
      "[387]\tvalid_0's binary_logloss: 0.181419\n",
      "[388]\tvalid_0's binary_logloss: 0.181399\n",
      "[389]\tvalid_0's binary_logloss: 0.1814\n",
      "[390]\tvalid_0's binary_logloss: 0.181308\n",
      "[391]\tvalid_0's binary_logloss: 0.18129\n",
      "[392]\tvalid_0's binary_logloss: 0.181224\n",
      "[393]\tvalid_0's binary_logloss: 0.181225\n",
      "[394]\tvalid_0's binary_logloss: 0.181142\n",
      "[395]\tvalid_0's binary_logloss: 0.181083\n",
      "[396]\tvalid_0's binary_logloss: 0.18108\n",
      "[397]\tvalid_0's binary_logloss: 0.180992\n",
      "[398]\tvalid_0's binary_logloss: 0.180897\n",
      "[399]\tvalid_0's binary_logloss: 0.180843\n",
      "[400]\tvalid_0's binary_logloss: 0.18082\n",
      "[401]\tvalid_0's binary_logloss: 0.180816\n",
      "[402]\tvalid_0's binary_logloss: 0.180697\n",
      "[403]\tvalid_0's binary_logloss: 0.180631\n",
      "[404]\tvalid_0's binary_logloss: 0.18055\n",
      "[405]\tvalid_0's binary_logloss: 0.180481\n",
      "[406]\tvalid_0's binary_logloss: 0.180416\n",
      "[407]\tvalid_0's binary_logloss: 0.180406\n",
      "[408]\tvalid_0's binary_logloss: 0.180305\n",
      "[409]\tvalid_0's binary_logloss: 0.180233\n",
      "[410]\tvalid_0's binary_logloss: 0.180186\n",
      "[411]\tvalid_0's binary_logloss: 0.180142\n",
      "[412]\tvalid_0's binary_logloss: 0.180065\n",
      "[413]\tvalid_0's binary_logloss: 0.17998\n",
      "[414]\tvalid_0's binary_logloss: 0.179916\n",
      "[415]\tvalid_0's binary_logloss: 0.179906\n",
      "[416]\tvalid_0's binary_logloss: 0.179901\n",
      "[417]\tvalid_0's binary_logloss: 0.179873\n",
      "[418]\tvalid_0's binary_logloss: 0.179805\n",
      "[419]\tvalid_0's binary_logloss: 0.179748\n",
      "[420]\tvalid_0's binary_logloss: 0.179727\n",
      "[421]\tvalid_0's binary_logloss: 0.179678\n",
      "[422]\tvalid_0's binary_logloss: 0.179609\n",
      "[423]\tvalid_0's binary_logloss: 0.179582\n",
      "[424]\tvalid_0's binary_logloss: 0.179524\n",
      "[425]\tvalid_0's binary_logloss: 0.17947\n",
      "[426]\tvalid_0's binary_logloss: 0.179463\n",
      "[427]\tvalid_0's binary_logloss: 0.1794\n",
      "[428]\tvalid_0's binary_logloss: 0.179333\n",
      "[429]\tvalid_0's binary_logloss: 0.179314\n",
      "[430]\tvalid_0's binary_logloss: 0.179248\n",
      "[431]\tvalid_0's binary_logloss: 0.179216\n",
      "[432]\tvalid_0's binary_logloss: 0.179174\n",
      "[433]\tvalid_0's binary_logloss: 0.179097\n",
      "[434]\tvalid_0's binary_logloss: 0.179064\n",
      "[435]\tvalid_0's binary_logloss: 0.178991\n",
      "[436]\tvalid_0's binary_logloss: 0.178925\n",
      "[437]\tvalid_0's binary_logloss: 0.178871\n",
      "[438]\tvalid_0's binary_logloss: 0.178821\n",
      "[439]\tvalid_0's binary_logloss: 0.178791\n",
      "[440]\tvalid_0's binary_logloss: 0.178762\n",
      "[441]\tvalid_0's binary_logloss: 0.178695\n",
      "[442]\tvalid_0's binary_logloss: 0.178661\n",
      "[443]\tvalid_0's binary_logloss: 0.17863\n",
      "[444]\tvalid_0's binary_logloss: 0.178614\n",
      "[445]\tvalid_0's binary_logloss: 0.178604\n",
      "[446]\tvalid_0's binary_logloss: 0.178559\n",
      "[447]\tvalid_0's binary_logloss: 0.17855\n",
      "[448]\tvalid_0's binary_logloss: 0.178483\n",
      "[449]\tvalid_0's binary_logloss: 0.178442\n",
      "[450]\tvalid_0's binary_logloss: 0.178389\n",
      "[451]\tvalid_0's binary_logloss: 0.178331\n",
      "[452]\tvalid_0's binary_logloss: 0.17831\n",
      "[453]\tvalid_0's binary_logloss: 0.17824\n",
      "[454]\tvalid_0's binary_logloss: 0.178197\n",
      "[455]\tvalid_0's binary_logloss: 0.178186\n",
      "[456]\tvalid_0's binary_logloss: 0.178173\n",
      "[457]\tvalid_0's binary_logloss: 0.178142\n",
      "[458]\tvalid_0's binary_logloss: 0.17811\n",
      "[459]\tvalid_0's binary_logloss: 0.178111\n",
      "[460]\tvalid_0's binary_logloss: 0.178067\n",
      "[461]\tvalid_0's binary_logloss: 0.177948\n",
      "[462]\tvalid_0's binary_logloss: 0.177948\n",
      "[463]\tvalid_0's binary_logloss: 0.177906\n",
      "[464]\tvalid_0's binary_logloss: 0.177842\n",
      "[465]\tvalid_0's binary_logloss: 0.17782\n",
      "[466]\tvalid_0's binary_logloss: 0.177786\n",
      "[467]\tvalid_0's binary_logloss: 0.177735\n",
      "[468]\tvalid_0's binary_logloss: 0.177723\n",
      "[469]\tvalid_0's binary_logloss: 0.177693\n",
      "[470]\tvalid_0's binary_logloss: 0.177652\n",
      "[471]\tvalid_0's binary_logloss: 0.177585\n",
      "[472]\tvalid_0's binary_logloss: 0.177557\n",
      "[473]\tvalid_0's binary_logloss: 0.17756\n",
      "[474]\tvalid_0's binary_logloss: 0.177537\n",
      "[475]\tvalid_0's binary_logloss: 0.17753\n",
      "[476]\tvalid_0's binary_logloss: 0.177495\n",
      "[477]\tvalid_0's binary_logloss: 0.177485\n",
      "[478]\tvalid_0's binary_logloss: 0.177475\n",
      "[479]\tvalid_0's binary_logloss: 0.177389\n",
      "[480]\tvalid_0's binary_logloss: 0.177365\n",
      "[481]\tvalid_0's binary_logloss: 0.177324\n",
      "[482]\tvalid_0's binary_logloss: 0.177322\n",
      "[483]\tvalid_0's binary_logloss: 0.177298\n",
      "[484]\tvalid_0's binary_logloss: 0.177292\n",
      "[485]\tvalid_0's binary_logloss: 0.177257\n",
      "[486]\tvalid_0's binary_logloss: 0.177241\n",
      "[487]\tvalid_0's binary_logloss: 0.177215\n",
      "[488]\tvalid_0's binary_logloss: 0.177191\n",
      "[489]\tvalid_0's binary_logloss: 0.177145\n",
      "[490]\tvalid_0's binary_logloss: 0.177147\n",
      "[491]\tvalid_0's binary_logloss: 0.177096\n",
      "[492]\tvalid_0's binary_logloss: 0.177068\n",
      "[493]\tvalid_0's binary_logloss: 0.177068\n",
      "[494]\tvalid_0's binary_logloss: 0.177014\n",
      "[495]\tvalid_0's binary_logloss: 0.176963\n",
      "[496]\tvalid_0's binary_logloss: 0.176915\n",
      "[497]\tvalid_0's binary_logloss: 0.176863\n",
      "[498]\tvalid_0's binary_logloss: 0.176851\n",
      "[499]\tvalid_0's binary_logloss: 0.176816\n",
      "[500]\tvalid_0's binary_logloss: 0.176788\n",
      "[501]\tvalid_0's binary_logloss: 0.176742\n",
      "[502]\tvalid_0's binary_logloss: 0.176754\n",
      "[503]\tvalid_0's binary_logloss: 0.176689\n",
      "[504]\tvalid_0's binary_logloss: 0.176671\n",
      "[505]\tvalid_0's binary_logloss: 0.176662\n",
      "[506]\tvalid_0's binary_logloss: 0.176655\n",
      "[507]\tvalid_0's binary_logloss: 0.176601\n",
      "[508]\tvalid_0's binary_logloss: 0.176565\n",
      "[509]\tvalid_0's binary_logloss: 0.176518\n",
      "[510]\tvalid_0's binary_logloss: 0.176441\n",
      "[511]\tvalid_0's binary_logloss: 0.176361\n",
      "[512]\tvalid_0's binary_logloss: 0.176305\n",
      "[513]\tvalid_0's binary_logloss: 0.176275\n",
      "[514]\tvalid_0's binary_logloss: 0.176247\n",
      "[515]\tvalid_0's binary_logloss: 0.176212\n",
      "[516]\tvalid_0's binary_logloss: 0.176183\n",
      "[517]\tvalid_0's binary_logloss: 0.176182\n",
      "[518]\tvalid_0's binary_logloss: 0.176183\n",
      "[519]\tvalid_0's binary_logloss: 0.176151\n",
      "[520]\tvalid_0's binary_logloss: 0.176121\n",
      "[521]\tvalid_0's binary_logloss: 0.176097\n",
      "[522]\tvalid_0's binary_logloss: 0.176026\n",
      "[523]\tvalid_0's binary_logloss: 0.176016\n",
      "[524]\tvalid_0's binary_logloss: 0.175996\n",
      "[525]\tvalid_0's binary_logloss: 0.175964\n",
      "[526]\tvalid_0's binary_logloss: 0.17596\n",
      "[527]\tvalid_0's binary_logloss: 0.175905\n",
      "[528]\tvalid_0's binary_logloss: 0.175856\n",
      "[529]\tvalid_0's binary_logloss: 0.175812\n",
      "[530]\tvalid_0's binary_logloss: 0.175783\n",
      "[531]\tvalid_0's binary_logloss: 0.175761\n",
      "[532]\tvalid_0's binary_logloss: 0.175752\n",
      "[533]\tvalid_0's binary_logloss: 0.175739\n",
      "[534]\tvalid_0's binary_logloss: 0.175715\n",
      "[535]\tvalid_0's binary_logloss: 0.175693\n",
      "[536]\tvalid_0's binary_logloss: 0.175666\n",
      "[537]\tvalid_0's binary_logloss: 0.175613\n",
      "[538]\tvalid_0's binary_logloss: 0.175578\n",
      "[539]\tvalid_0's binary_logloss: 0.175556\n",
      "[540]\tvalid_0's binary_logloss: 0.175494\n",
      "[541]\tvalid_0's binary_logloss: 0.175474\n",
      "[542]\tvalid_0's binary_logloss: 0.17541\n",
      "[543]\tvalid_0's binary_logloss: 0.175409\n",
      "[544]\tvalid_0's binary_logloss: 0.17535\n",
      "[545]\tvalid_0's binary_logloss: 0.17533\n",
      "[546]\tvalid_0's binary_logloss: 0.1753\n",
      "[547]\tvalid_0's binary_logloss: 0.175293\n",
      "[548]\tvalid_0's binary_logloss: 0.175248\n",
      "[549]\tvalid_0's binary_logloss: 0.175197\n",
      "[550]\tvalid_0's binary_logloss: 0.175121\n",
      "[551]\tvalid_0's binary_logloss: 0.175114\n",
      "[552]\tvalid_0's binary_logloss: 0.175093\n",
      "[553]\tvalid_0's binary_logloss: 0.175089\n",
      "[554]\tvalid_0's binary_logloss: 0.175084\n",
      "[555]\tvalid_0's binary_logloss: 0.175066\n",
      "[556]\tvalid_0's binary_logloss: 0.175063\n",
      "[557]\tvalid_0's binary_logloss: 0.175059\n",
      "[558]\tvalid_0's binary_logloss: 0.175026\n",
      "[559]\tvalid_0's binary_logloss: 0.174991\n",
      "[560]\tvalid_0's binary_logloss: 0.174953\n",
      "[561]\tvalid_0's binary_logloss: 0.174924\n",
      "[562]\tvalid_0's binary_logloss: 0.174929\n",
      "[563]\tvalid_0's binary_logloss: 0.174912\n",
      "[564]\tvalid_0's binary_logloss: 0.174914\n",
      "[565]\tvalid_0's binary_logloss: 0.174897\n",
      "[566]\tvalid_0's binary_logloss: 0.174874\n",
      "[567]\tvalid_0's binary_logloss: 0.174852\n",
      "[568]\tvalid_0's binary_logloss: 0.174799\n",
      "[569]\tvalid_0's binary_logloss: 0.174783\n",
      "[570]\tvalid_0's binary_logloss: 0.174774\n",
      "[571]\tvalid_0's binary_logloss: 0.174752\n",
      "[572]\tvalid_0's binary_logloss: 0.17474\n",
      "[573]\tvalid_0's binary_logloss: 0.174734\n",
      "[574]\tvalid_0's binary_logloss: 0.174719\n",
      "[575]\tvalid_0's binary_logloss: 0.174705\n",
      "[576]\tvalid_0's binary_logloss: 0.174698\n",
      "[577]\tvalid_0's binary_logloss: 0.174669\n",
      "[578]\tvalid_0's binary_logloss: 0.174647\n",
      "[579]\tvalid_0's binary_logloss: 0.174629\n",
      "[580]\tvalid_0's binary_logloss: 0.174615\n",
      "[581]\tvalid_0's binary_logloss: 0.174598\n",
      "[582]\tvalid_0's binary_logloss: 0.174542\n",
      "[583]\tvalid_0's binary_logloss: 0.174478\n",
      "[584]\tvalid_0's binary_logloss: 0.174449\n",
      "[585]\tvalid_0's binary_logloss: 0.174431\n",
      "[586]\tvalid_0's binary_logloss: 0.174417\n",
      "[587]\tvalid_0's binary_logloss: 0.174375\n",
      "[588]\tvalid_0's binary_logloss: 0.17431\n",
      "[589]\tvalid_0's binary_logloss: 0.17425\n",
      "[590]\tvalid_0's binary_logloss: 0.174214\n",
      "[591]\tvalid_0's binary_logloss: 0.174216\n",
      "[592]\tvalid_0's binary_logloss: 0.174218\n",
      "[593]\tvalid_0's binary_logloss: 0.174211\n",
      "[594]\tvalid_0's binary_logloss: 0.174174\n",
      "[595]\tvalid_0's binary_logloss: 0.174155\n",
      "[596]\tvalid_0's binary_logloss: 0.174144\n",
      "[597]\tvalid_0's binary_logloss: 0.174136\n",
      "[598]\tvalid_0's binary_logloss: 0.174132\n",
      "[599]\tvalid_0's binary_logloss: 0.174099\n",
      "[600]\tvalid_0's binary_logloss: 0.174094\n",
      "[601]\tvalid_0's binary_logloss: 0.174072\n",
      "[602]\tvalid_0's binary_logloss: 0.174064\n",
      "[603]\tvalid_0's binary_logloss: 0.174021\n",
      "[604]\tvalid_0's binary_logloss: 0.174016\n",
      "[605]\tvalid_0's binary_logloss: 0.173985\n",
      "[606]\tvalid_0's binary_logloss: 0.173969\n",
      "[607]\tvalid_0's binary_logloss: 0.173928\n",
      "[608]\tvalid_0's binary_logloss: 0.173903\n",
      "[609]\tvalid_0's binary_logloss: 0.17388\n",
      "[610]\tvalid_0's binary_logloss: 0.17386\n",
      "[611]\tvalid_0's binary_logloss: 0.173831\n",
      "[612]\tvalid_0's binary_logloss: 0.17381\n",
      "[613]\tvalid_0's binary_logloss: 0.173791\n",
      "[614]\tvalid_0's binary_logloss: 0.173732\n",
      "[615]\tvalid_0's binary_logloss: 0.173717\n",
      "[616]\tvalid_0's binary_logloss: 0.173707\n",
      "[617]\tvalid_0's binary_logloss: 0.173643\n",
      "[618]\tvalid_0's binary_logloss: 0.173598\n",
      "[619]\tvalid_0's binary_logloss: 0.173583\n",
      "[620]\tvalid_0's binary_logloss: 0.173552\n",
      "[621]\tvalid_0's binary_logloss: 0.173529\n",
      "[622]\tvalid_0's binary_logloss: 0.173458\n",
      "[623]\tvalid_0's binary_logloss: 0.173454\n",
      "[624]\tvalid_0's binary_logloss: 0.173434\n",
      "[625]\tvalid_0's binary_logloss: 0.173424\n",
      "[626]\tvalid_0's binary_logloss: 0.173396\n",
      "[627]\tvalid_0's binary_logloss: 0.173373\n",
      "[628]\tvalid_0's binary_logloss: 0.17334\n",
      "[629]\tvalid_0's binary_logloss: 0.173329\n",
      "[630]\tvalid_0's binary_logloss: 0.17329\n",
      "[631]\tvalid_0's binary_logloss: 0.17329\n",
      "[632]\tvalid_0's binary_logloss: 0.173265\n",
      "[633]\tvalid_0's binary_logloss: 0.173265\n",
      "[634]\tvalid_0's binary_logloss: 0.173237\n",
      "[635]\tvalid_0's binary_logloss: 0.173235\n",
      "[636]\tvalid_0's binary_logloss: 0.173191\n",
      "[637]\tvalid_0's binary_logloss: 0.173165\n",
      "[638]\tvalid_0's binary_logloss: 0.173159\n",
      "[639]\tvalid_0's binary_logloss: 0.173118\n",
      "[640]\tvalid_0's binary_logloss: 0.173095\n",
      "[641]\tvalid_0's binary_logloss: 0.173057\n",
      "[642]\tvalid_0's binary_logloss: 0.173043\n",
      "[643]\tvalid_0's binary_logloss: 0.173028\n",
      "[644]\tvalid_0's binary_logloss: 0.173002\n",
      "[645]\tvalid_0's binary_logloss: 0.172982\n",
      "[646]\tvalid_0's binary_logloss: 0.172941\n",
      "[647]\tvalid_0's binary_logloss: 0.172924\n",
      "[648]\tvalid_0's binary_logloss: 0.172919\n",
      "[649]\tvalid_0's binary_logloss: 0.172907\n",
      "[650]\tvalid_0's binary_logloss: 0.172884\n",
      "[651]\tvalid_0's binary_logloss: 0.172858\n",
      "[652]\tvalid_0's binary_logloss: 0.172849\n",
      "[653]\tvalid_0's binary_logloss: 0.172847\n",
      "[654]\tvalid_0's binary_logloss: 0.172837\n",
      "[655]\tvalid_0's binary_logloss: 0.172836\n",
      "[656]\tvalid_0's binary_logloss: 0.172824\n",
      "[657]\tvalid_0's binary_logloss: 0.17281\n",
      "[658]\tvalid_0's binary_logloss: 0.172783\n",
      "[659]\tvalid_0's binary_logloss: 0.172778\n",
      "[660]\tvalid_0's binary_logloss: 0.172754\n",
      "[661]\tvalid_0's binary_logloss: 0.172742\n",
      "[662]\tvalid_0's binary_logloss: 0.172722\n",
      "[663]\tvalid_0's binary_logloss: 0.172704\n",
      "[664]\tvalid_0's binary_logloss: 0.172659\n",
      "[665]\tvalid_0's binary_logloss: 0.172623\n",
      "[666]\tvalid_0's binary_logloss: 0.172586\n",
      "[667]\tvalid_0's binary_logloss: 0.17257\n",
      "[668]\tvalid_0's binary_logloss: 0.172557\n",
      "[669]\tvalid_0's binary_logloss: 0.172522\n",
      "[670]\tvalid_0's binary_logloss: 0.172519\n",
      "[671]\tvalid_0's binary_logloss: 0.172469\n",
      "[672]\tvalid_0's binary_logloss: 0.172465\n",
      "[673]\tvalid_0's binary_logloss: 0.172465\n",
      "[674]\tvalid_0's binary_logloss: 0.172448\n",
      "[675]\tvalid_0's binary_logloss: 0.172439\n",
      "[676]\tvalid_0's binary_logloss: 0.172401\n",
      "[677]\tvalid_0's binary_logloss: 0.172401\n",
      "[678]\tvalid_0's binary_logloss: 0.172401\n",
      "[679]\tvalid_0's binary_logloss: 0.172394\n",
      "[680]\tvalid_0's binary_logloss: 0.17239\n",
      "[681]\tvalid_0's binary_logloss: 0.172386\n",
      "[682]\tvalid_0's binary_logloss: 0.172373\n",
      "[683]\tvalid_0's binary_logloss: 0.172355\n",
      "[684]\tvalid_0's binary_logloss: 0.172354\n",
      "[685]\tvalid_0's binary_logloss: 0.172311\n",
      "[686]\tvalid_0's binary_logloss: 0.172276\n",
      "[687]\tvalid_0's binary_logloss: 0.172247\n",
      "[688]\tvalid_0's binary_logloss: 0.172251\n",
      "[689]\tvalid_0's binary_logloss: 0.17222\n",
      "[690]\tvalid_0's binary_logloss: 0.172205\n",
      "[691]\tvalid_0's binary_logloss: 0.172191\n",
      "[692]\tvalid_0's binary_logloss: 0.172178\n",
      "[693]\tvalid_0's binary_logloss: 0.172156\n",
      "[694]\tvalid_0's binary_logloss: 0.172154\n",
      "[695]\tvalid_0's binary_logloss: 0.172122\n",
      "[696]\tvalid_0's binary_logloss: 0.172114\n",
      "[697]\tvalid_0's binary_logloss: 0.17211\n",
      "[698]\tvalid_0's binary_logloss: 0.172081\n",
      "[699]\tvalid_0's binary_logloss: 0.172058\n",
      "[700]\tvalid_0's binary_logloss: 0.172056\n",
      "[701]\tvalid_0's binary_logloss: 0.172053\n",
      "[702]\tvalid_0's binary_logloss: 0.172054\n",
      "[703]\tvalid_0's binary_logloss: 0.172028\n",
      "[704]\tvalid_0's binary_logloss: 0.172014\n",
      "[705]\tvalid_0's binary_logloss: 0.172004\n",
      "[706]\tvalid_0's binary_logloss: 0.172004\n",
      "[707]\tvalid_0's binary_logloss: 0.171998\n",
      "[708]\tvalid_0's binary_logloss: 0.171985\n",
      "[709]\tvalid_0's binary_logloss: 0.171982\n",
      "[710]\tvalid_0's binary_logloss: 0.171967\n",
      "[711]\tvalid_0's binary_logloss: 0.171934\n",
      "[712]\tvalid_0's binary_logloss: 0.171911\n",
      "[713]\tvalid_0's binary_logloss: 0.171871\n",
      "[714]\tvalid_0's binary_logloss: 0.171836\n",
      "[715]\tvalid_0's binary_logloss: 0.171807\n",
      "[716]\tvalid_0's binary_logloss: 0.171794\n",
      "[717]\tvalid_0's binary_logloss: 0.171793\n",
      "[718]\tvalid_0's binary_logloss: 0.171748\n",
      "[719]\tvalid_0's binary_logloss: 0.171743\n",
      "[720]\tvalid_0's binary_logloss: 0.171724\n",
      "[721]\tvalid_0's binary_logloss: 0.171707\n",
      "[722]\tvalid_0's binary_logloss: 0.171696\n",
      "[723]\tvalid_0's binary_logloss: 0.171661\n",
      "[724]\tvalid_0's binary_logloss: 0.171645\n",
      "[725]\tvalid_0's binary_logloss: 0.171617\n",
      "[726]\tvalid_0's binary_logloss: 0.171608\n",
      "[727]\tvalid_0's binary_logloss: 0.17159\n",
      "[728]\tvalid_0's binary_logloss: 0.171549\n",
      "[729]\tvalid_0's binary_logloss: 0.171521\n",
      "[730]\tvalid_0's binary_logloss: 0.171495\n",
      "[731]\tvalid_0's binary_logloss: 0.171494\n",
      "[732]\tvalid_0's binary_logloss: 0.171487\n",
      "[733]\tvalid_0's binary_logloss: 0.171478\n",
      "[734]\tvalid_0's binary_logloss: 0.171484\n",
      "[735]\tvalid_0's binary_logloss: 0.17146\n",
      "[736]\tvalid_0's binary_logloss: 0.171449\n",
      "[737]\tvalid_0's binary_logloss: 0.171421\n",
      "[738]\tvalid_0's binary_logloss: 0.171409\n",
      "[739]\tvalid_0's binary_logloss: 0.171387\n",
      "[740]\tvalid_0's binary_logloss: 0.171366\n",
      "[741]\tvalid_0's binary_logloss: 0.171361\n",
      "[742]\tvalid_0's binary_logloss: 0.17135\n",
      "[743]\tvalid_0's binary_logloss: 0.171316\n",
      "[744]\tvalid_0's binary_logloss: 0.171297\n",
      "[745]\tvalid_0's binary_logloss: 0.171278\n",
      "[746]\tvalid_0's binary_logloss: 0.171251\n",
      "[747]\tvalid_0's binary_logloss: 0.171206\n",
      "[748]\tvalid_0's binary_logloss: 0.171187\n",
      "[749]\tvalid_0's binary_logloss: 0.171187\n",
      "[750]\tvalid_0's binary_logloss: 0.17118\n",
      "[751]\tvalid_0's binary_logloss: 0.171122\n",
      "[752]\tvalid_0's binary_logloss: 0.171108\n",
      "[753]\tvalid_0's binary_logloss: 0.171099\n",
      "[754]\tvalid_0's binary_logloss: 0.171067\n",
      "[755]\tvalid_0's binary_logloss: 0.17106\n",
      "[756]\tvalid_0's binary_logloss: 0.171052\n",
      "[757]\tvalid_0's binary_logloss: 0.171039\n",
      "[758]\tvalid_0's binary_logloss: 0.171035\n",
      "[759]\tvalid_0's binary_logloss: 0.17103\n",
      "[760]\tvalid_0's binary_logloss: 0.171021\n",
      "[761]\tvalid_0's binary_logloss: 0.171011\n",
      "[762]\tvalid_0's binary_logloss: 0.171001\n",
      "[763]\tvalid_0's binary_logloss: 0.17099\n",
      "[764]\tvalid_0's binary_logloss: 0.170984\n",
      "[765]\tvalid_0's binary_logloss: 0.170985\n",
      "[766]\tvalid_0's binary_logloss: 0.170985\n",
      "[767]\tvalid_0's binary_logloss: 0.170972\n",
      "[768]\tvalid_0's binary_logloss: 0.170962\n",
      "[769]\tvalid_0's binary_logloss: 0.17094\n",
      "[770]\tvalid_0's binary_logloss: 0.170926\n",
      "[771]\tvalid_0's binary_logloss: 0.170919\n",
      "[772]\tvalid_0's binary_logloss: 0.170902\n",
      "[773]\tvalid_0's binary_logloss: 0.170905\n",
      "[774]\tvalid_0's binary_logloss: 0.170904\n",
      "[775]\tvalid_0's binary_logloss: 0.170887\n",
      "[776]\tvalid_0's binary_logloss: 0.170874\n",
      "[777]\tvalid_0's binary_logloss: 0.170875\n",
      "[778]\tvalid_0's binary_logloss: 0.17088\n",
      "[779]\tvalid_0's binary_logloss: 0.170848\n",
      "[780]\tvalid_0's binary_logloss: 0.170833\n",
      "[781]\tvalid_0's binary_logloss: 0.170826\n",
      "[782]\tvalid_0's binary_logloss: 0.17082\n",
      "[783]\tvalid_0's binary_logloss: 0.170808\n",
      "[784]\tvalid_0's binary_logloss: 0.170807\n",
      "[785]\tvalid_0's binary_logloss: 0.170791\n",
      "[786]\tvalid_0's binary_logloss: 0.170785\n",
      "[787]\tvalid_0's binary_logloss: 0.170784\n",
      "[788]\tvalid_0's binary_logloss: 0.170782\n",
      "[789]\tvalid_0's binary_logloss: 0.170774\n",
      "[790]\tvalid_0's binary_logloss: 0.170761\n",
      "[791]\tvalid_0's binary_logloss: 0.170726\n",
      "[792]\tvalid_0's binary_logloss: 0.170725\n",
      "[793]\tvalid_0's binary_logloss: 0.170716\n",
      "[794]\tvalid_0's binary_logloss: 0.170717\n",
      "[795]\tvalid_0's binary_logloss: 0.170697\n",
      "[796]\tvalid_0's binary_logloss: 0.170665\n",
      "[797]\tvalid_0's binary_logloss: 0.170642\n",
      "[798]\tvalid_0's binary_logloss: 0.170627\n",
      "[799]\tvalid_0's binary_logloss: 0.170595\n",
      "[800]\tvalid_0's binary_logloss: 0.170591\n",
      "[801]\tvalid_0's binary_logloss: 0.170593\n",
      "[802]\tvalid_0's binary_logloss: 0.17057\n",
      "[803]\tvalid_0's binary_logloss: 0.170571\n",
      "[804]\tvalid_0's binary_logloss: 0.170546\n",
      "[805]\tvalid_0's binary_logloss: 0.170535\n",
      "[806]\tvalid_0's binary_logloss: 0.170521\n",
      "[807]\tvalid_0's binary_logloss: 0.170518\n",
      "[808]\tvalid_0's binary_logloss: 0.170499\n",
      "[809]\tvalid_0's binary_logloss: 0.170492\n",
      "[810]\tvalid_0's binary_logloss: 0.170463\n",
      "[811]\tvalid_0's binary_logloss: 0.170457\n",
      "[812]\tvalid_0's binary_logloss: 0.170457\n",
      "[813]\tvalid_0's binary_logloss: 0.17044\n",
      "[814]\tvalid_0's binary_logloss: 0.170425\n",
      "[815]\tvalid_0's binary_logloss: 0.170408\n",
      "[816]\tvalid_0's binary_logloss: 0.170411\n",
      "[817]\tvalid_0's binary_logloss: 0.170412\n",
      "[818]\tvalid_0's binary_logloss: 0.170406\n",
      "[819]\tvalid_0's binary_logloss: 0.1704\n",
      "[820]\tvalid_0's binary_logloss: 0.170365\n",
      "[821]\tvalid_0's binary_logloss: 0.170366\n",
      "[822]\tvalid_0's binary_logloss: 0.170336\n",
      "[823]\tvalid_0's binary_logloss: 0.170319\n",
      "[824]\tvalid_0's binary_logloss: 0.170304\n",
      "[825]\tvalid_0's binary_logloss: 0.170288\n",
      "[826]\tvalid_0's binary_logloss: 0.170281\n",
      "[827]\tvalid_0's binary_logloss: 0.170273\n",
      "[828]\tvalid_0's binary_logloss: 0.170264\n",
      "[829]\tvalid_0's binary_logloss: 0.17025\n",
      "[830]\tvalid_0's binary_logloss: 0.170242\n",
      "[831]\tvalid_0's binary_logloss: 0.17024\n",
      "[832]\tvalid_0's binary_logloss: 0.170242\n",
      "[833]\tvalid_0's binary_logloss: 0.170241\n",
      "[834]\tvalid_0's binary_logloss: 0.170231\n",
      "[835]\tvalid_0's binary_logloss: 0.170228\n",
      "[836]\tvalid_0's binary_logloss: 0.170198\n",
      "[837]\tvalid_0's binary_logloss: 0.170183\n",
      "[838]\tvalid_0's binary_logloss: 0.17017\n",
      "[839]\tvalid_0's binary_logloss: 0.17015\n",
      "[840]\tvalid_0's binary_logloss: 0.170126\n",
      "[841]\tvalid_0's binary_logloss: 0.170116\n",
      "[842]\tvalid_0's binary_logloss: 0.170114\n",
      "[843]\tvalid_0's binary_logloss: 0.170112\n",
      "[844]\tvalid_0's binary_logloss: 0.170094\n",
      "[845]\tvalid_0's binary_logloss: 0.170091\n",
      "[846]\tvalid_0's binary_logloss: 0.170092\n",
      "[847]\tvalid_0's binary_logloss: 0.17009\n",
      "[848]\tvalid_0's binary_logloss: 0.170073\n",
      "[849]\tvalid_0's binary_logloss: 0.170063\n",
      "[850]\tvalid_0's binary_logloss: 0.170037\n",
      "[851]\tvalid_0's binary_logloss: 0.170018\n",
      "[852]\tvalid_0's binary_logloss: 0.170017\n",
      "[853]\tvalid_0's binary_logloss: 0.170005\n",
      "[854]\tvalid_0's binary_logloss: 0.169979\n",
      "[855]\tvalid_0's binary_logloss: 0.169974\n",
      "[856]\tvalid_0's binary_logloss: 0.169962\n",
      "[857]\tvalid_0's binary_logloss: 0.16994\n",
      "[858]\tvalid_0's binary_logloss: 0.169919\n",
      "[859]\tvalid_0's binary_logloss: 0.169898\n",
      "[860]\tvalid_0's binary_logloss: 0.169877\n",
      "[861]\tvalid_0's binary_logloss: 0.169875\n",
      "[862]\tvalid_0's binary_logloss: 0.169876\n",
      "[863]\tvalid_0's binary_logloss: 0.169856\n",
      "[864]\tvalid_0's binary_logloss: 0.169852\n",
      "[865]\tvalid_0's binary_logloss: 0.169845\n",
      "[866]\tvalid_0's binary_logloss: 0.169837\n",
      "[867]\tvalid_0's binary_logloss: 0.169835\n",
      "[868]\tvalid_0's binary_logloss: 0.169816\n",
      "[869]\tvalid_0's binary_logloss: 0.169806\n",
      "[870]\tvalid_0's binary_logloss: 0.169796\n",
      "[871]\tvalid_0's binary_logloss: 0.169786\n",
      "[872]\tvalid_0's binary_logloss: 0.169782\n",
      "[873]\tvalid_0's binary_logloss: 0.169781\n",
      "[874]\tvalid_0's binary_logloss: 0.169784\n",
      "[875]\tvalid_0's binary_logloss: 0.16978\n",
      "[876]\tvalid_0's binary_logloss: 0.169771\n",
      "[877]\tvalid_0's binary_logloss: 0.169754\n",
      "[878]\tvalid_0's binary_logloss: 0.169729\n",
      "[879]\tvalid_0's binary_logloss: 0.169723\n",
      "[880]\tvalid_0's binary_logloss: 0.169713\n",
      "[881]\tvalid_0's binary_logloss: 0.169707\n",
      "[882]\tvalid_0's binary_logloss: 0.169709\n",
      "[883]\tvalid_0's binary_logloss: 0.169702\n",
      "[884]\tvalid_0's binary_logloss: 0.169694\n",
      "[885]\tvalid_0's binary_logloss: 0.169682\n",
      "[886]\tvalid_0's binary_logloss: 0.169678\n",
      "[887]\tvalid_0's binary_logloss: 0.169668\n",
      "[888]\tvalid_0's binary_logloss: 0.169662\n",
      "[889]\tvalid_0's binary_logloss: 0.169652\n",
      "[890]\tvalid_0's binary_logloss: 0.169648\n",
      "[891]\tvalid_0's binary_logloss: 0.16965\n",
      "[892]\tvalid_0's binary_logloss: 0.16963\n",
      "[893]\tvalid_0's binary_logloss: 0.169618\n",
      "[894]\tvalid_0's binary_logloss: 0.169613\n",
      "[895]\tvalid_0's binary_logloss: 0.169604\n",
      "[896]\tvalid_0's binary_logloss: 0.169606\n",
      "[897]\tvalid_0's binary_logloss: 0.169605\n",
      "[898]\tvalid_0's binary_logloss: 0.1696\n",
      "[899]\tvalid_0's binary_logloss: 0.169596\n",
      "[900]\tvalid_0's binary_logloss: 0.169595\n",
      "[901]\tvalid_0's binary_logloss: 0.169595\n",
      "[902]\tvalid_0's binary_logloss: 0.169583\n",
      "[903]\tvalid_0's binary_logloss: 0.169579\n",
      "[904]\tvalid_0's binary_logloss: 0.169563\n",
      "[905]\tvalid_0's binary_logloss: 0.16956\n",
      "[906]\tvalid_0's binary_logloss: 0.16954\n",
      "[907]\tvalid_0's binary_logloss: 0.169535\n",
      "[908]\tvalid_0's binary_logloss: 0.16952\n",
      "[909]\tvalid_0's binary_logloss: 0.169508\n",
      "[910]\tvalid_0's binary_logloss: 0.169486\n",
      "[911]\tvalid_0's binary_logloss: 0.169482\n",
      "[912]\tvalid_0's binary_logloss: 0.169485\n",
      "[913]\tvalid_0's binary_logloss: 0.169473\n",
      "[914]\tvalid_0's binary_logloss: 0.169478\n",
      "[915]\tvalid_0's binary_logloss: 0.169482\n",
      "[916]\tvalid_0's binary_logloss: 0.169473\n",
      "[917]\tvalid_0's binary_logloss: 0.169465\n",
      "[918]\tvalid_0's binary_logloss: 0.169444\n",
      "[919]\tvalid_0's binary_logloss: 0.169426\n",
      "[920]\tvalid_0's binary_logloss: 0.169398\n",
      "[921]\tvalid_0's binary_logloss: 0.16939\n",
      "[922]\tvalid_0's binary_logloss: 0.16938\n",
      "[923]\tvalid_0's binary_logloss: 0.169377\n",
      "[924]\tvalid_0's binary_logloss: 0.169364\n",
      "[925]\tvalid_0's binary_logloss: 0.169358\n",
      "[926]\tvalid_0's binary_logloss: 0.169351\n",
      "[927]\tvalid_0's binary_logloss: 0.169344\n",
      "[928]\tvalid_0's binary_logloss: 0.169343\n",
      "[929]\tvalid_0's binary_logloss: 0.169325\n",
      "[930]\tvalid_0's binary_logloss: 0.16932\n",
      "[931]\tvalid_0's binary_logloss: 0.169317\n",
      "[932]\tvalid_0's binary_logloss: 0.169317\n",
      "[933]\tvalid_0's binary_logloss: 0.169301\n",
      "[934]\tvalid_0's binary_logloss: 0.169285\n",
      "[935]\tvalid_0's binary_logloss: 0.169285\n",
      "[936]\tvalid_0's binary_logloss: 0.169269\n",
      "[937]\tvalid_0's binary_logloss: 0.169263\n",
      "[938]\tvalid_0's binary_logloss: 0.16924\n",
      "[939]\tvalid_0's binary_logloss: 0.169234\n",
      "[940]\tvalid_0's binary_logloss: 0.169231\n",
      "[941]\tvalid_0's binary_logloss: 0.16923\n",
      "[942]\tvalid_0's binary_logloss: 0.169226\n",
      "[943]\tvalid_0's binary_logloss: 0.169211\n",
      "[944]\tvalid_0's binary_logloss: 0.169205\n",
      "[945]\tvalid_0's binary_logloss: 0.169198\n",
      "[946]\tvalid_0's binary_logloss: 0.169182\n",
      "[947]\tvalid_0's binary_logloss: 0.169185\n",
      "[948]\tvalid_0's binary_logloss: 0.16918\n",
      "[949]\tvalid_0's binary_logloss: 0.169153\n",
      "[950]\tvalid_0's binary_logloss: 0.169153\n",
      "[951]\tvalid_0's binary_logloss: 0.169148\n",
      "[952]\tvalid_0's binary_logloss: 0.169135\n",
      "[953]\tvalid_0's binary_logloss: 0.16912\n",
      "[954]\tvalid_0's binary_logloss: 0.16911\n",
      "[955]\tvalid_0's binary_logloss: 0.169108\n",
      "[956]\tvalid_0's binary_logloss: 0.169108\n",
      "[957]\tvalid_0's binary_logloss: 0.169104\n",
      "[958]\tvalid_0's binary_logloss: 0.169097\n",
      "[959]\tvalid_0's binary_logloss: 0.169091\n",
      "[960]\tvalid_0's binary_logloss: 0.169086\n",
      "[961]\tvalid_0's binary_logloss: 0.16907\n",
      "[962]\tvalid_0's binary_logloss: 0.16905\n",
      "[963]\tvalid_0's binary_logloss: 0.169047\n",
      "[964]\tvalid_0's binary_logloss: 0.16904\n",
      "[965]\tvalid_0's binary_logloss: 0.169017\n",
      "[966]\tvalid_0's binary_logloss: 0.16901\n",
      "[967]\tvalid_0's binary_logloss: 0.169008\n",
      "[968]\tvalid_0's binary_logloss: 0.169\n",
      "[969]\tvalid_0's binary_logloss: 0.168994\n",
      "[970]\tvalid_0's binary_logloss: 0.168983\n",
      "[971]\tvalid_0's binary_logloss: 0.168977\n",
      "[972]\tvalid_0's binary_logloss: 0.168963\n",
      "[973]\tvalid_0's binary_logloss: 0.168951\n",
      "[974]\tvalid_0's binary_logloss: 0.168946\n",
      "[975]\tvalid_0's binary_logloss: 0.16894\n",
      "[976]\tvalid_0's binary_logloss: 0.168916\n",
      "[977]\tvalid_0's binary_logloss: 0.168907\n",
      "[978]\tvalid_0's binary_logloss: 0.16889\n",
      "[979]\tvalid_0's binary_logloss: 0.168878\n",
      "[980]\tvalid_0's binary_logloss: 0.168869\n",
      "[981]\tvalid_0's binary_logloss: 0.168853\n",
      "[982]\tvalid_0's binary_logloss: 0.168845\n",
      "[983]\tvalid_0's binary_logloss: 0.16882\n",
      "[984]\tvalid_0's binary_logloss: 0.168819\n",
      "[985]\tvalid_0's binary_logloss: 0.168814\n",
      "[986]\tvalid_0's binary_logloss: 0.168791\n",
      "[987]\tvalid_0's binary_logloss: 0.168788\n",
      "[988]\tvalid_0's binary_logloss: 0.16876\n",
      "[989]\tvalid_0's binary_logloss: 0.168763\n",
      "[990]\tvalid_0's binary_logloss: 0.168747\n",
      "[991]\tvalid_0's binary_logloss: 0.168739\n",
      "[992]\tvalid_0's binary_logloss: 0.168746\n",
      "[993]\tvalid_0's binary_logloss: 0.168745\n",
      "[994]\tvalid_0's binary_logloss: 0.16874\n",
      "[995]\tvalid_0's binary_logloss: 0.168744\n",
      "[996]\tvalid_0's binary_logloss: 0.16875\n",
      "[997]\tvalid_0's binary_logloss: 0.168741\n",
      "[998]\tvalid_0's binary_logloss: 0.168722\n",
      "[999]\tvalid_0's binary_logloss: 0.168695\n",
      "[1000]\tvalid_0's binary_logloss: 0.168685\n"
     ]
    }
   ],
   "source": [
    "from lightgbm import LGBMClassifier\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.metrics import recall_score\n",
    "\n",
    "n_fold = 5\n",
    "n_class = 2\n",
    "cv = StratifiedKFold(n_splits=n_fold, shuffle=True, random_state=seed)\n",
    "p_val = np.zeros((resampled_x.shape[0], n_class)) # 0,1 각각 확률\n",
    "p_tst = np.zeros((test.shape[0], n_class)) # 0,1 각각 확률\n",
    "\n",
    "for i, (i_trn, i_val) in enumerate(cv.split(resampled_x, resampled_y),1):\n",
    "    print(f'training model for CV #{i}')\n",
    "    clf = LGBMClassifier(**params)\n",
    "    clf.fit(resampled_x[i_trn], resampled_y[i_trn],\n",
    "            eval_set=[(resampled_x[i_val], resampled_y[i_val])],\n",
    "            eval_metric='binary',\n",
    "            early_stopping_rounds=100)\n",
    "    \n",
    "    p_val[i_val, :] = clf.predict_proba(resampled_x[i_val])\n",
    "    p_tst += clf.predict_proba(test) / n_fold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "ef1bad7c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "정확도: 90.6309%\n",
      "재현율: 16.6791%\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "print(f'정확도: {accuracy_score(test_y, np.argmax(p_tst, axis=1)) * 100:.4f}%')\n",
    "print(f'재현율: {recall_score(test_y, np.argmax(p_tst, axis=1),pos_label=1) * 100:.4f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "0854e85e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\sklearn.py:726: UserWarning: 'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.\n",
      "  _log_warning(\"'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. \"\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\lightgbm\\engine.py:177: UserWarning: Found `num_iterations` in params. Will use it instead of argument\n",
      "  _log_warning(f\"Found `{alias}` in params. Will use it instead of argument\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] feature_fraction is set=0.4, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.4\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.0, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.0\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8772503866491428, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8772503866491428\n",
      "[LightGBM] [Warning] lambda_l2 is set=0.0, reg_lambda=0.0 will be ignored. Current value: lambda_l2=0.0\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n",
      "[1]\ttraining's auc: 0.813128\ttraining's binary_logloss: 0.662225\n",
      "[2]\ttraining's auc: 0.818533\ttraining's binary_logloss: 0.639601\n",
      "[3]\ttraining's auc: 0.841438\ttraining's binary_logloss: 0.625618\n",
      "[4]\ttraining's auc: 0.835395\ttraining's binary_logloss: 0.608628\n",
      "[5]\ttraining's auc: 0.845737\ttraining's binary_logloss: 0.598249\n",
      "[6]\ttraining's auc: 0.885891\ttraining's binary_logloss: 0.574265\n",
      "[7]\ttraining's auc: 0.899969\ttraining's binary_logloss: 0.555808\n",
      "[8]\ttraining's auc: 0.903746\ttraining's binary_logloss: 0.547536\n",
      "[9]\ttraining's auc: 0.913586\ttraining's binary_logloss: 0.527925\n",
      "[10]\ttraining's auc: 0.913463\ttraining's binary_logloss: 0.514684\n",
      "[11]\ttraining's auc: 0.914309\ttraining's binary_logloss: 0.507576\n",
      "[12]\ttraining's auc: 0.919154\ttraining's binary_logloss: 0.492303\n",
      "[13]\ttraining's auc: 0.917556\ttraining's binary_logloss: 0.481168\n",
      "[14]\ttraining's auc: 0.919031\ttraining's binary_logloss: 0.475435\n",
      "[15]\ttraining's auc: 0.920887\ttraining's binary_logloss: 0.465753\n",
      "[16]\ttraining's auc: 0.921244\ttraining's binary_logloss: 0.457247\n",
      "[17]\ttraining's auc: 0.923968\ttraining's binary_logloss: 0.448422\n",
      "[18]\ttraining's auc: 0.928194\ttraining's binary_logloss: 0.438059\n",
      "[19]\ttraining's auc: 0.928251\ttraining's binary_logloss: 0.431305\n",
      "[20]\ttraining's auc: 0.929007\ttraining's binary_logloss: 0.427642\n",
      "[21]\ttraining's auc: 0.931278\ttraining's binary_logloss: 0.418661\n",
      "[22]\ttraining's auc: 0.932157\ttraining's binary_logloss: 0.41491\n",
      "[23]\ttraining's auc: 0.932177\ttraining's binary_logloss: 0.412179\n",
      "[24]\ttraining's auc: 0.932185\ttraining's binary_logloss: 0.407029\n",
      "[25]\ttraining's auc: 0.934079\ttraining's binary_logloss: 0.399551\n",
      "[26]\ttraining's auc: 0.933938\ttraining's binary_logloss: 0.395295\n",
      "[27]\ttraining's auc: 0.93564\ttraining's binary_logloss: 0.388371\n",
      "[28]\ttraining's auc: 0.938416\ttraining's binary_logloss: 0.381406\n",
      "[29]\ttraining's auc: 0.939892\ttraining's binary_logloss: 0.378186\n",
      "[30]\ttraining's auc: 0.940267\ttraining's binary_logloss: 0.376064\n",
      "[31]\ttraining's auc: 0.940058\ttraining's binary_logloss: 0.373793\n",
      "[32]\ttraining's auc: 0.941114\ttraining's binary_logloss: 0.368334\n",
      "[33]\ttraining's auc: 0.941115\ttraining's binary_logloss: 0.366248\n",
      "[34]\ttraining's auc: 0.941946\ttraining's binary_logloss: 0.364014\n",
      "[35]\ttraining's auc: 0.942319\ttraining's binary_logloss: 0.362067\n",
      "[36]\ttraining's auc: 0.942538\ttraining's binary_logloss: 0.360351\n",
      "[37]\ttraining's auc: 0.943382\ttraining's binary_logloss: 0.357688\n",
      "[38]\ttraining's auc: 0.945306\ttraining's binary_logloss: 0.352252\n",
      "[39]\ttraining's auc: 0.946267\ttraining's binary_logloss: 0.350028\n",
      "[40]\ttraining's auc: 0.946743\ttraining's binary_logloss: 0.347513\n",
      "[41]\ttraining's auc: 0.946932\ttraining's binary_logloss: 0.346372\n",
      "[42]\ttraining's auc: 0.947247\ttraining's binary_logloss: 0.344952\n",
      "[43]\ttraining's auc: 0.947551\ttraining's binary_logloss: 0.343522\n",
      "[44]\ttraining's auc: 0.948317\ttraining's binary_logloss: 0.339666\n",
      "[45]\ttraining's auc: 0.948208\ttraining's binary_logloss: 0.338257\n",
      "[46]\ttraining's auc: 0.948617\ttraining's binary_logloss: 0.336808\n",
      "[47]\ttraining's auc: 0.949948\ttraining's binary_logloss: 0.332477\n",
      "[48]\ttraining's auc: 0.950403\ttraining's binary_logloss: 0.331105\n",
      "[49]\ttraining's auc: 0.950618\ttraining's binary_logloss: 0.330044\n",
      "[50]\ttraining's auc: 0.951233\ttraining's binary_logloss: 0.328224\n",
      "[51]\ttraining's auc: 0.951769\ttraining's binary_logloss: 0.324778\n",
      "[52]\ttraining's auc: 0.952786\ttraining's binary_logloss: 0.321856\n",
      "[53]\ttraining's auc: 0.953959\ttraining's binary_logloss: 0.317761\n",
      "[54]\ttraining's auc: 0.954359\ttraining's binary_logloss: 0.315113\n",
      "[55]\ttraining's auc: 0.954555\ttraining's binary_logloss: 0.314161\n",
      "[56]\ttraining's auc: 0.9555\ttraining's binary_logloss: 0.310769\n",
      "[57]\ttraining's auc: 0.955821\ttraining's binary_logloss: 0.309424\n",
      "[58]\ttraining's auc: 0.956267\ttraining's binary_logloss: 0.30725\n",
      "[59]\ttraining's auc: 0.957028\ttraining's binary_logloss: 0.304684\n",
      "[60]\ttraining's auc: 0.957678\ttraining's binary_logloss: 0.302964\n",
      "[61]\ttraining's auc: 0.958103\ttraining's binary_logloss: 0.301506\n",
      "[62]\ttraining's auc: 0.958086\ttraining's binary_logloss: 0.300697\n",
      "[63]\ttraining's auc: 0.958951\ttraining's binary_logloss: 0.296991\n",
      "[64]\ttraining's auc: 0.959012\ttraining's binary_logloss: 0.296218\n",
      "[65]\ttraining's auc: 0.959557\ttraining's binary_logloss: 0.294801\n",
      "[66]\ttraining's auc: 0.960027\ttraining's binary_logloss: 0.29328\n",
      "[67]\ttraining's auc: 0.960588\ttraining's binary_logloss: 0.290565\n",
      "[68]\ttraining's auc: 0.960951\ttraining's binary_logloss: 0.289273\n",
      "[69]\ttraining's auc: 0.961149\ttraining's binary_logloss: 0.28839\n",
      "[70]\ttraining's auc: 0.961264\ttraining's binary_logloss: 0.287684\n",
      "[71]\ttraining's auc: 0.961407\ttraining's binary_logloss: 0.28648\n",
      "[72]\ttraining's auc: 0.961785\ttraining's binary_logloss: 0.284896\n",
      "[73]\ttraining's auc: 0.961871\ttraining's binary_logloss: 0.284316\n",
      "[74]\ttraining's auc: 0.962149\ttraining's binary_logloss: 0.28297\n",
      "[75]\ttraining's auc: 0.963189\ttraining's binary_logloss: 0.279431\n",
      "[76]\ttraining's auc: 0.963352\ttraining's binary_logloss: 0.278695\n",
      "[77]\ttraining's auc: 0.964176\ttraining's binary_logloss: 0.275624\n",
      "[78]\ttraining's auc: 0.964535\ttraining's binary_logloss: 0.273644\n",
      "[79]\ttraining's auc: 0.964884\ttraining's binary_logloss: 0.272496\n",
      "[80]\ttraining's auc: 0.965237\ttraining's binary_logloss: 0.26991\n",
      "[81]\ttraining's auc: 0.965536\ttraining's binary_logloss: 0.268768\n",
      "[82]\ttraining's auc: 0.965748\ttraining's binary_logloss: 0.267898\n",
      "[83]\ttraining's auc: 0.965927\ttraining's binary_logloss: 0.266052\n",
      "[84]\ttraining's auc: 0.966155\ttraining's binary_logloss: 0.265148\n",
      "[85]\ttraining's auc: 0.966492\ttraining's binary_logloss: 0.263977\n",
      "[86]\ttraining's auc: 0.966722\ttraining's binary_logloss: 0.262373\n",
      "[87]\ttraining's auc: 0.966937\ttraining's binary_logloss: 0.261447\n",
      "[88]\ttraining's auc: 0.967142\ttraining's binary_logloss: 0.260602\n",
      "[89]\ttraining's auc: 0.967234\ttraining's binary_logloss: 0.259997\n",
      "[90]\ttraining's auc: 0.967416\ttraining's binary_logloss: 0.259251\n",
      "[91]\ttraining's auc: 0.967507\ttraining's binary_logloss: 0.258287\n",
      "[92]\ttraining's auc: 0.968055\ttraining's binary_logloss: 0.255893\n",
      "[93]\ttraining's auc: 0.968281\ttraining's binary_logloss: 0.253902\n",
      "[94]\ttraining's auc: 0.968784\ttraining's binary_logloss: 0.25157\n",
      "[95]\ttraining's auc: 0.968904\ttraining's binary_logloss: 0.250386\n",
      "[96]\ttraining's auc: 0.969038\ttraining's binary_logloss: 0.249673\n",
      "[97]\ttraining's auc: 0.969147\ttraining's binary_logloss: 0.248214\n",
      "[98]\ttraining's auc: 0.969457\ttraining's binary_logloss: 0.246732\n",
      "[99]\ttraining's auc: 0.969522\ttraining's binary_logloss: 0.246386\n",
      "[100]\ttraining's auc: 0.969659\ttraining's binary_logloss: 0.244726\n",
      "[101]\ttraining's auc: 0.969803\ttraining's binary_logloss: 0.244108\n",
      "[102]\ttraining's auc: 0.969906\ttraining's binary_logloss: 0.243263\n",
      "[103]\ttraining's auc: 0.970092\ttraining's binary_logloss: 0.242545\n",
      "[104]\ttraining's auc: 0.970153\ttraining's binary_logloss: 0.241531\n",
      "[105]\ttraining's auc: 0.970436\ttraining's binary_logloss: 0.240168\n",
      "[106]\ttraining's auc: 0.97058\ttraining's binary_logloss: 0.239544\n",
      "[107]\ttraining's auc: 0.970873\ttraining's binary_logloss: 0.237876\n",
      "[108]\ttraining's auc: 0.970956\ttraining's binary_logloss: 0.236491\n",
      "[109]\ttraining's auc: 0.971072\ttraining's binary_logloss: 0.2359\n",
      "[110]\ttraining's auc: 0.97109\ttraining's binary_logloss: 0.235244\n",
      "[111]\ttraining's auc: 0.971143\ttraining's binary_logloss: 0.234468\n",
      "[112]\ttraining's auc: 0.971297\ttraining's binary_logloss: 0.233546\n",
      "[113]\ttraining's auc: 0.971455\ttraining's binary_logloss: 0.232867\n",
      "[114]\ttraining's auc: 0.971591\ttraining's binary_logloss: 0.232126\n",
      "[115]\ttraining's auc: 0.97171\ttraining's binary_logloss: 0.231622\n",
      "[116]\ttraining's auc: 0.97177\ttraining's binary_logloss: 0.230472\n",
      "[117]\ttraining's auc: 0.971843\ttraining's binary_logloss: 0.230073\n",
      "[118]\ttraining's auc: 0.971918\ttraining's binary_logloss: 0.229548\n",
      "[119]\ttraining's auc: 0.971993\ttraining's binary_logloss: 0.229117\n",
      "[120]\ttraining's auc: 0.972071\ttraining's binary_logloss: 0.228731\n",
      "[121]\ttraining's auc: 0.972192\ttraining's binary_logloss: 0.228096\n",
      "[122]\ttraining's auc: 0.972267\ttraining's binary_logloss: 0.227594\n",
      "[123]\ttraining's auc: 0.972322\ttraining's binary_logloss: 0.227164\n",
      "[124]\ttraining's auc: 0.972341\ttraining's binary_logloss: 0.226941\n",
      "[125]\ttraining's auc: 0.9724\ttraining's binary_logloss: 0.226542\n",
      "[126]\ttraining's auc: 0.972461\ttraining's binary_logloss: 0.226138\n",
      "[127]\ttraining's auc: 0.972545\ttraining's binary_logloss: 0.225544\n",
      "[128]\ttraining's auc: 0.972723\ttraining's binary_logloss: 0.224317\n",
      "[129]\ttraining's auc: 0.972788\ttraining's binary_logloss: 0.223404\n",
      "[130]\ttraining's auc: 0.97284\ttraining's binary_logloss: 0.223108\n",
      "[131]\ttraining's auc: 0.972879\ttraining's binary_logloss: 0.222782\n",
      "[132]\ttraining's auc: 0.972991\ttraining's binary_logloss: 0.222071\n",
      "[133]\ttraining's auc: 0.973018\ttraining's binary_logloss: 0.22148\n",
      "[134]\ttraining's auc: 0.973081\ttraining's binary_logloss: 0.221104\n",
      "[135]\ttraining's auc: 0.973106\ttraining's binary_logloss: 0.220675\n",
      "[136]\ttraining's auc: 0.973173\ttraining's binary_logloss: 0.220126\n",
      "[137]\ttraining's auc: 0.973221\ttraining's binary_logloss: 0.21984\n",
      "[138]\ttraining's auc: 0.973295\ttraining's binary_logloss: 0.219384\n",
      "[139]\ttraining's auc: 0.973351\ttraining's binary_logloss: 0.218594\n",
      "[140]\ttraining's auc: 0.973412\ttraining's binary_logloss: 0.218348\n",
      "[141]\ttraining's auc: 0.973436\ttraining's binary_logloss: 0.218136\n",
      "[142]\ttraining's auc: 0.973503\ttraining's binary_logloss: 0.217756\n",
      "[143]\ttraining's auc: 0.973563\ttraining's binary_logloss: 0.217482\n",
      "[144]\ttraining's auc: 0.973611\ttraining's binary_logloss: 0.217308\n",
      "[145]\ttraining's auc: 0.973631\ttraining's binary_logloss: 0.216789\n",
      "[146]\ttraining's auc: 0.973663\ttraining's binary_logloss: 0.216449\n",
      "[147]\ttraining's auc: 0.973705\ttraining's binary_logloss: 0.216184\n",
      "[148]\ttraining's auc: 0.973748\ttraining's binary_logloss: 0.215506\n",
      "[149]\ttraining's auc: 0.973767\ttraining's binary_logloss: 0.215202\n",
      "[150]\ttraining's auc: 0.973871\ttraining's binary_logloss: 0.214455\n",
      "[151]\ttraining's auc: 0.97392\ttraining's binary_logloss: 0.213973\n",
      "[152]\ttraining's auc: 0.973942\ttraining's binary_logloss: 0.213586\n",
      "[153]\ttraining's auc: 0.973965\ttraining's binary_logloss: 0.213106\n",
      "[154]\ttraining's auc: 0.973977\ttraining's binary_logloss: 0.212783\n",
      "[155]\ttraining's auc: 0.974042\ttraining's binary_logloss: 0.212349\n",
      "[156]\ttraining's auc: 0.974081\ttraining's binary_logloss: 0.211841\n",
      "[157]\ttraining's auc: 0.974122\ttraining's binary_logloss: 0.211559\n",
      "[158]\ttraining's auc: 0.974162\ttraining's binary_logloss: 0.211238\n",
      "[159]\ttraining's auc: 0.974229\ttraining's binary_logloss: 0.210991\n",
      "[160]\ttraining's auc: 0.974227\ttraining's binary_logloss: 0.210501\n",
      "[161]\ttraining's auc: 0.974288\ttraining's binary_logloss: 0.210121\n",
      "[162]\ttraining's auc: 0.974309\ttraining's binary_logloss: 0.20996\n",
      "[163]\ttraining's auc: 0.974339\ttraining's binary_logloss: 0.209741\n",
      "[164]\ttraining's auc: 0.974357\ttraining's binary_logloss: 0.209406\n",
      "[165]\ttraining's auc: 0.974388\ttraining's binary_logloss: 0.209023\n",
      "[166]\ttraining's auc: 0.974425\ttraining's binary_logloss: 0.208841\n",
      "[167]\ttraining's auc: 0.974457\ttraining's binary_logloss: 0.208631\n",
      "[168]\ttraining's auc: 0.974475\ttraining's binary_logloss: 0.208458\n",
      "[169]\ttraining's auc: 0.974548\ttraining's binary_logloss: 0.207725\n",
      "[170]\ttraining's auc: 0.974564\ttraining's binary_logloss: 0.207251\n",
      "[171]\ttraining's auc: 0.974587\ttraining's binary_logloss: 0.206942\n",
      "[172]\ttraining's auc: 0.974629\ttraining's binary_logloss: 0.206628\n",
      "[173]\ttraining's auc: 0.974648\ttraining's binary_logloss: 0.206496\n",
      "[174]\ttraining's auc: 0.974661\ttraining's binary_logloss: 0.206179\n",
      "[175]\ttraining's auc: 0.974697\ttraining's binary_logloss: 0.206041\n",
      "[176]\ttraining's auc: 0.97473\ttraining's binary_logloss: 0.205726\n",
      "[177]\ttraining's auc: 0.974753\ttraining's binary_logloss: 0.205567\n",
      "[178]\ttraining's auc: 0.974782\ttraining's binary_logloss: 0.205359\n",
      "[179]\ttraining's auc: 0.974814\ttraining's binary_logloss: 0.205076\n",
      "[180]\ttraining's auc: 0.974842\ttraining's binary_logloss: 0.204751\n",
      "[181]\ttraining's auc: 0.974855\ttraining's binary_logloss: 0.204581\n",
      "[182]\ttraining's auc: 0.974876\ttraining's binary_logloss: 0.204091\n",
      "[183]\ttraining's auc: 0.974897\ttraining's binary_logloss: 0.203909\n",
      "[184]\ttraining's auc: 0.974923\ttraining's binary_logloss: 0.203601\n",
      "[185]\ttraining's auc: 0.974939\ttraining's binary_logloss: 0.203516\n",
      "[186]\ttraining's auc: 0.974957\ttraining's binary_logloss: 0.203243\n",
      "[187]\ttraining's auc: 0.974972\ttraining's binary_logloss: 0.203146\n",
      "[188]\ttraining's auc: 0.974985\ttraining's binary_logloss: 0.203053\n",
      "[189]\ttraining's auc: 0.974992\ttraining's binary_logloss: 0.202798\n",
      "[190]\ttraining's auc: 0.975009\ttraining's binary_logloss: 0.202692\n",
      "[191]\ttraining's auc: 0.975021\ttraining's binary_logloss: 0.202572\n",
      "[192]\ttraining's auc: 0.975037\ttraining's binary_logloss: 0.202436\n",
      "[193]\ttraining's auc: 0.975079\ttraining's binary_logloss: 0.202094\n",
      "[194]\ttraining's auc: 0.975102\ttraining's binary_logloss: 0.201868\n",
      "[195]\ttraining's auc: 0.975155\ttraining's binary_logloss: 0.201391\n",
      "[196]\ttraining's auc: 0.975217\ttraining's binary_logloss: 0.200773\n",
      "[197]\ttraining's auc: 0.975232\ttraining's binary_logloss: 0.200595\n",
      "[198]\ttraining's auc: 0.97524\ttraining's binary_logloss: 0.200383\n",
      "[199]\ttraining's auc: 0.975257\ttraining's binary_logloss: 0.200233\n",
      "[200]\ttraining's auc: 0.975267\ttraining's binary_logloss: 0.20016\n",
      "[201]\ttraining's auc: 0.975288\ttraining's binary_logloss: 0.200008\n",
      "[202]\ttraining's auc: 0.975303\ttraining's binary_logloss: 0.199904\n",
      "[203]\ttraining's auc: 0.975305\ttraining's binary_logloss: 0.199835\n",
      "[204]\ttraining's auc: 0.975322\ttraining's binary_logloss: 0.199732\n",
      "[205]\ttraining's auc: 0.975328\ttraining's binary_logloss: 0.199656\n",
      "[206]\ttraining's auc: 0.975342\ttraining's binary_logloss: 0.199521\n",
      "[207]\ttraining's auc: 0.975366\ttraining's binary_logloss: 0.1993\n",
      "[208]\ttraining's auc: 0.975385\ttraining's binary_logloss: 0.199115\n",
      "[209]\ttraining's auc: 0.975395\ttraining's binary_logloss: 0.198962\n",
      "[210]\ttraining's auc: 0.975422\ttraining's binary_logloss: 0.198828\n",
      "[211]\ttraining's auc: 0.975432\ttraining's binary_logloss: 0.198779\n",
      "[212]\ttraining's auc: 0.975458\ttraining's binary_logloss: 0.198389\n",
      "[213]\ttraining's auc: 0.975462\ttraining's binary_logloss: 0.198232\n",
      "[214]\ttraining's auc: 0.975483\ttraining's binary_logloss: 0.198016\n",
      "[215]\ttraining's auc: 0.975525\ttraining's binary_logloss: 0.197856\n",
      "[216]\ttraining's auc: 0.97554\ttraining's binary_logloss: 0.19777\n",
      "[217]\ttraining's auc: 0.975574\ttraining's binary_logloss: 0.197288\n",
      "[218]\ttraining's auc: 0.97558\ttraining's binary_logloss: 0.197218\n",
      "[219]\ttraining's auc: 0.975601\ttraining's binary_logloss: 0.19688\n",
      "[220]\ttraining's auc: 0.975633\ttraining's binary_logloss: 0.196597\n",
      "[221]\ttraining's auc: 0.975672\ttraining's binary_logloss: 0.196187\n",
      "[222]\ttraining's auc: 0.975688\ttraining's binary_logloss: 0.196116\n",
      "[223]\ttraining's auc: 0.975712\ttraining's binary_logloss: 0.195984\n",
      "[224]\ttraining's auc: 0.975724\ttraining's binary_logloss: 0.195918\n",
      "[225]\ttraining's auc: 0.975734\ttraining's binary_logloss: 0.195847\n",
      "[226]\ttraining's auc: 0.975735\ttraining's binary_logloss: 0.195774\n",
      "[227]\ttraining's auc: 0.975738\ttraining's binary_logloss: 0.195705\n",
      "[228]\ttraining's auc: 0.975757\ttraining's binary_logloss: 0.195548\n",
      "[229]\ttraining's auc: 0.97578\ttraining's binary_logloss: 0.195362\n",
      "[230]\ttraining's auc: 0.97581\ttraining's binary_logloss: 0.195067\n",
      "[231]\ttraining's auc: 0.975822\ttraining's binary_logloss: 0.194956\n",
      "[232]\ttraining's auc: 0.975847\ttraining's binary_logloss: 0.194815\n",
      "[233]\ttraining's auc: 0.975868\ttraining's binary_logloss: 0.19461\n",
      "[234]\ttraining's auc: 0.975871\ttraining's binary_logloss: 0.194534\n",
      "[235]\ttraining's auc: 0.975879\ttraining's binary_logloss: 0.194472\n",
      "[236]\ttraining's auc: 0.975897\ttraining's binary_logloss: 0.194175\n",
      "[237]\ttraining's auc: 0.975913\ttraining's binary_logloss: 0.193821\n",
      "[238]\ttraining's auc: 0.975922\ttraining's binary_logloss: 0.193636\n",
      "[239]\ttraining's auc: 0.97594\ttraining's binary_logloss: 0.193558\n",
      "[240]\ttraining's auc: 0.975954\ttraining's binary_logloss: 0.193418\n",
      "[241]\ttraining's auc: 0.975971\ttraining's binary_logloss: 0.193262\n",
      "[242]\ttraining's auc: 0.975986\ttraining's binary_logloss: 0.193086\n",
      "[243]\ttraining's auc: 0.975995\ttraining's binary_logloss: 0.192839\n",
      "[244]\ttraining's auc: 0.976012\ttraining's binary_logloss: 0.192726\n",
      "[245]\ttraining's auc: 0.976024\ttraining's binary_logloss: 0.192685\n",
      "[246]\ttraining's auc: 0.976046\ttraining's binary_logloss: 0.192513\n",
      "[247]\ttraining's auc: 0.976065\ttraining's binary_logloss: 0.192377\n",
      "[248]\ttraining's auc: 0.976076\ttraining's binary_logloss: 0.192294\n",
      "[249]\ttraining's auc: 0.976096\ttraining's binary_logloss: 0.192214\n",
      "[250]\ttraining's auc: 0.976115\ttraining's binary_logloss: 0.192071\n",
      "[251]\ttraining's auc: 0.976138\ttraining's binary_logloss: 0.191879\n",
      "[252]\ttraining's auc: 0.976148\ttraining's binary_logloss: 0.191781\n",
      "[253]\ttraining's auc: 0.976161\ttraining's binary_logloss: 0.191739\n",
      "[254]\ttraining's auc: 0.976178\ttraining's binary_logloss: 0.191572\n",
      "[255]\ttraining's auc: 0.976187\ttraining's binary_logloss: 0.191506\n",
      "[256]\ttraining's auc: 0.976197\ttraining's binary_logloss: 0.191472\n",
      "[257]\ttraining's auc: 0.976239\ttraining's binary_logloss: 0.191304\n",
      "[258]\ttraining's auc: 0.97625\ttraining's binary_logloss: 0.191198\n",
      "[259]\ttraining's auc: 0.976271\ttraining's binary_logloss: 0.190973\n",
      "[260]\ttraining's auc: 0.976279\ttraining's binary_logloss: 0.190922\n",
      "[261]\ttraining's auc: 0.976295\ttraining's binary_logloss: 0.19066\n",
      "[262]\ttraining's auc: 0.976312\ttraining's binary_logloss: 0.190482\n",
      "[263]\ttraining's auc: 0.976321\ttraining's binary_logloss: 0.190406\n",
      "[264]\ttraining's auc: 0.976332\ttraining's binary_logloss: 0.19028\n",
      "[265]\ttraining's auc: 0.976347\ttraining's binary_logloss: 0.190194\n",
      "[266]\ttraining's auc: 0.976353\ttraining's binary_logloss: 0.190069\n",
      "[267]\ttraining's auc: 0.976358\ttraining's binary_logloss: 0.190019\n",
      "[268]\ttraining's auc: 0.976379\ttraining's binary_logloss: 0.189879\n",
      "[269]\ttraining's auc: 0.976397\ttraining's binary_logloss: 0.189629\n",
      "[270]\ttraining's auc: 0.976406\ttraining's binary_logloss: 0.189587\n",
      "[271]\ttraining's auc: 0.976416\ttraining's binary_logloss: 0.189511\n",
      "[272]\ttraining's auc: 0.97644\ttraining's binary_logloss: 0.189406\n",
      "[273]\ttraining's auc: 0.976449\ttraining's binary_logloss: 0.189361\n",
      "[274]\ttraining's auc: 0.97646\ttraining's binary_logloss: 0.189269\n",
      "[275]\ttraining's auc: 0.976477\ttraining's binary_logloss: 0.189051\n",
      "[276]\ttraining's auc: 0.976481\ttraining's binary_logloss: 0.188981\n",
      "[277]\ttraining's auc: 0.9765\ttraining's binary_logloss: 0.188754\n",
      "[278]\ttraining's auc: 0.976524\ttraining's binary_logloss: 0.188653\n",
      "[279]\ttraining's auc: 0.976526\ttraining's binary_logloss: 0.188583\n",
      "[280]\ttraining's auc: 0.976543\ttraining's binary_logloss: 0.1884\n",
      "[281]\ttraining's auc: 0.976548\ttraining's binary_logloss: 0.188358\n",
      "[282]\ttraining's auc: 0.976555\ttraining's binary_logloss: 0.188278\n",
      "[283]\ttraining's auc: 0.976565\ttraining's binary_logloss: 0.188166\n",
      "[284]\ttraining's auc: 0.976584\ttraining's binary_logloss: 0.187896\n",
      "[285]\ttraining's auc: 0.976592\ttraining's binary_logloss: 0.18786\n",
      "[286]\ttraining's auc: 0.976602\ttraining's binary_logloss: 0.18773\n",
      "[287]\ttraining's auc: 0.976604\ttraining's binary_logloss: 0.187698\n",
      "[288]\ttraining's auc: 0.976612\ttraining's binary_logloss: 0.187645\n",
      "[289]\ttraining's auc: 0.976627\ttraining's binary_logloss: 0.18749\n",
      "[290]\ttraining's auc: 0.976635\ttraining's binary_logloss: 0.187273\n",
      "[291]\ttraining's auc: 0.976637\ttraining's binary_logloss: 0.187222\n",
      "[292]\ttraining's auc: 0.976653\ttraining's binary_logloss: 0.186992\n",
      "[293]\ttraining's auc: 0.976665\ttraining's binary_logloss: 0.186859\n",
      "[294]\ttraining's auc: 0.976672\ttraining's binary_logloss: 0.186747\n",
      "[295]\ttraining's auc: 0.97669\ttraining's binary_logloss: 0.186632\n",
      "[296]\ttraining's auc: 0.976694\ttraining's binary_logloss: 0.186596\n",
      "[297]\ttraining's auc: 0.976718\ttraining's binary_logloss: 0.186375\n",
      "[298]\ttraining's auc: 0.976732\ttraining's binary_logloss: 0.186184\n",
      "[299]\ttraining's auc: 0.976744\ttraining's binary_logloss: 0.186115\n",
      "[300]\ttraining's auc: 0.976757\ttraining's binary_logloss: 0.185954\n",
      "[301]\ttraining's auc: 0.976766\ttraining's binary_logloss: 0.185832\n",
      "[302]\ttraining's auc: 0.976772\ttraining's binary_logloss: 0.185789\n",
      "[303]\ttraining's auc: 0.976781\ttraining's binary_logloss: 0.185658\n",
      "[304]\ttraining's auc: 0.976784\ttraining's binary_logloss: 0.185622\n",
      "[305]\ttraining's auc: 0.976812\ttraining's binary_logloss: 0.185452\n",
      "[306]\ttraining's auc: 0.976832\ttraining's binary_logloss: 0.185284\n",
      "[307]\ttraining's auc: 0.976845\ttraining's binary_logloss: 0.185099\n",
      "[308]\ttraining's auc: 0.976852\ttraining's binary_logloss: 0.184988\n",
      "[309]\ttraining's auc: 0.976858\ttraining's binary_logloss: 0.184847\n",
      "[310]\ttraining's auc: 0.976863\ttraining's binary_logloss: 0.184818\n",
      "[311]\ttraining's auc: 0.976863\ttraining's binary_logloss: 0.184739\n",
      "[312]\ttraining's auc: 0.976866\ttraining's binary_logloss: 0.1847\n",
      "[313]\ttraining's auc: 0.976885\ttraining's binary_logloss: 0.184558\n",
      "[314]\ttraining's auc: 0.976898\ttraining's binary_logloss: 0.184446\n",
      "[315]\ttraining's auc: 0.976915\ttraining's binary_logloss: 0.184374\n",
      "[316]\ttraining's auc: 0.976919\ttraining's binary_logloss: 0.184348\n",
      "[317]\ttraining's auc: 0.976924\ttraining's binary_logloss: 0.184323\n",
      "[318]\ttraining's auc: 0.97693\ttraining's binary_logloss: 0.184295\n",
      "[319]\ttraining's auc: 0.976937\ttraining's binary_logloss: 0.18412\n",
      "[320]\ttraining's auc: 0.976952\ttraining's binary_logloss: 0.184055\n",
      "[321]\ttraining's auc: 0.976965\ttraining's binary_logloss: 0.183863\n",
      "[322]\ttraining's auc: 0.976977\ttraining's binary_logloss: 0.183815\n",
      "[323]\ttraining's auc: 0.976981\ttraining's binary_logloss: 0.183774\n",
      "[324]\ttraining's auc: 0.976985\ttraining's binary_logloss: 0.183738\n",
      "[325]\ttraining's auc: 0.976997\ttraining's binary_logloss: 0.183594\n",
      "[326]\ttraining's auc: 0.977004\ttraining's binary_logloss: 0.183461\n",
      "[327]\ttraining's auc: 0.977014\ttraining's binary_logloss: 0.183348\n",
      "[328]\ttraining's auc: 0.977027\ttraining's binary_logloss: 0.183227\n",
      "[329]\ttraining's auc: 0.977035\ttraining's binary_logloss: 0.183174\n",
      "[330]\ttraining's auc: 0.977039\ttraining's binary_logloss: 0.183138\n",
      "[331]\ttraining's auc: 0.977044\ttraining's binary_logloss: 0.183074\n",
      "[332]\ttraining's auc: 0.977049\ttraining's binary_logloss: 0.182968\n",
      "[333]\ttraining's auc: 0.977058\ttraining's binary_logloss: 0.182889\n",
      "[334]\ttraining's auc: 0.977066\ttraining's binary_logloss: 0.182733\n",
      "[335]\ttraining's auc: 0.977076\ttraining's binary_logloss: 0.182698\n",
      "[336]\ttraining's auc: 0.977084\ttraining's binary_logloss: 0.182593\n",
      "[337]\ttraining's auc: 0.977089\ttraining's binary_logloss: 0.182573\n",
      "[338]\ttraining's auc: 0.977094\ttraining's binary_logloss: 0.182482\n",
      "[339]\ttraining's auc: 0.977101\ttraining's binary_logloss: 0.182335\n",
      "[340]\ttraining's auc: 0.977107\ttraining's binary_logloss: 0.182215\n",
      "[341]\ttraining's auc: 0.977123\ttraining's binary_logloss: 0.182092\n",
      "[342]\ttraining's auc: 0.977128\ttraining's binary_logloss: 0.182067\n",
      "[343]\ttraining's auc: 0.977135\ttraining's binary_logloss: 0.182009\n",
      "[344]\ttraining's auc: 0.977143\ttraining's binary_logloss: 0.181968\n",
      "[345]\ttraining's auc: 0.977147\ttraining's binary_logloss: 0.181936\n",
      "[346]\ttraining's auc: 0.977162\ttraining's binary_logloss: 0.18184\n",
      "[347]\ttraining's auc: 0.97717\ttraining's binary_logloss: 0.18174\n",
      "[348]\ttraining's auc: 0.97718\ttraining's binary_logloss: 0.181686\n",
      "[349]\ttraining's auc: 0.977186\ttraining's binary_logloss: 0.181604\n",
      "[350]\ttraining's auc: 0.977192\ttraining's binary_logloss: 0.181517\n",
      "[351]\ttraining's auc: 0.977206\ttraining's binary_logloss: 0.181458\n",
      "[352]\ttraining's auc: 0.977216\ttraining's binary_logloss: 0.181333\n",
      "[353]\ttraining's auc: 0.977224\ttraining's binary_logloss: 0.181261\n",
      "[354]\ttraining's auc: 0.977235\ttraining's binary_logloss: 0.181156\n",
      "[355]\ttraining's auc: 0.977246\ttraining's binary_logloss: 0.181089\n",
      "[356]\ttraining's auc: 0.977258\ttraining's binary_logloss: 0.181\n",
      "[357]\ttraining's auc: 0.977264\ttraining's binary_logloss: 0.180954\n",
      "[358]\ttraining's auc: 0.977269\ttraining's binary_logloss: 0.180934\n",
      "[359]\ttraining's auc: 0.977274\ttraining's binary_logloss: 0.180909\n",
      "[360]\ttraining's auc: 0.977278\ttraining's binary_logloss: 0.180891\n",
      "[361]\ttraining's auc: 0.977289\ttraining's binary_logloss: 0.18078\n",
      "[362]\ttraining's auc: 0.977293\ttraining's binary_logloss: 0.180723\n",
      "[363]\ttraining's auc: 0.977301\ttraining's binary_logloss: 0.180693\n",
      "[364]\ttraining's auc: 0.977305\ttraining's binary_logloss: 0.180668\n",
      "[365]\ttraining's auc: 0.97731\ttraining's binary_logloss: 0.180638\n",
      "[366]\ttraining's auc: 0.977323\ttraining's binary_logloss: 0.180558\n",
      "[367]\ttraining's auc: 0.977333\ttraining's binary_logloss: 0.180497\n",
      "[368]\ttraining's auc: 0.977343\ttraining's binary_logloss: 0.180455\n",
      "[369]\ttraining's auc: 0.977347\ttraining's binary_logloss: 0.180401\n",
      "[370]\ttraining's auc: 0.977352\ttraining's binary_logloss: 0.180335\n",
      "[371]\ttraining's auc: 0.977353\ttraining's binary_logloss: 0.180226\n",
      "[372]\ttraining's auc: 0.977356\ttraining's binary_logloss: 0.180125\n",
      "[373]\ttraining's auc: 0.977364\ttraining's binary_logloss: 0.179983\n",
      "[374]\ttraining's auc: 0.977371\ttraining's binary_logloss: 0.179941\n",
      "[375]\ttraining's auc: 0.97738\ttraining's binary_logloss: 0.179844\n",
      "[376]\ttraining's auc: 0.977391\ttraining's binary_logloss: 0.179772\n",
      "[377]\ttraining's auc: 0.977399\ttraining's binary_logloss: 0.179728\n",
      "[378]\ttraining's auc: 0.977409\ttraining's binary_logloss: 0.179683\n",
      "[379]\ttraining's auc: 0.977414\ttraining's binary_logloss: 0.179584\n",
      "[380]\ttraining's auc: 0.97742\ttraining's binary_logloss: 0.1795\n",
      "[381]\ttraining's auc: 0.977422\ttraining's binary_logloss: 0.179472\n",
      "[382]\ttraining's auc: 0.977428\ttraining's binary_logloss: 0.179446\n",
      "[383]\ttraining's auc: 0.977437\ttraining's binary_logloss: 0.179296\n",
      "[384]\ttraining's auc: 0.977442\ttraining's binary_logloss: 0.179244\n",
      "[385]\ttraining's auc: 0.977447\ttraining's binary_logloss: 0.17922\n",
      "[386]\ttraining's auc: 0.977452\ttraining's binary_logloss: 0.179204\n",
      "[387]\ttraining's auc: 0.977455\ttraining's binary_logloss: 0.179191\n",
      "[388]\ttraining's auc: 0.97746\ttraining's binary_logloss: 0.179063\n",
      "[389]\ttraining's auc: 0.977471\ttraining's binary_logloss: 0.178938\n",
      "[390]\ttraining's auc: 0.977477\ttraining's binary_logloss: 0.178887\n",
      "[391]\ttraining's auc: 0.977484\ttraining's binary_logloss: 0.178834\n",
      "[392]\ttraining's auc: 0.977499\ttraining's binary_logloss: 0.178754\n",
      "[393]\ttraining's auc: 0.977508\ttraining's binary_logloss: 0.178715\n",
      "[394]\ttraining's auc: 0.977529\ttraining's binary_logloss: 0.178626\n",
      "[395]\ttraining's auc: 0.977533\ttraining's binary_logloss: 0.178571\n",
      "[396]\ttraining's auc: 0.977549\ttraining's binary_logloss: 0.178499\n",
      "[397]\ttraining's auc: 0.977556\ttraining's binary_logloss: 0.178476\n",
      "[398]\ttraining's auc: 0.977561\ttraining's binary_logloss: 0.17837\n",
      "[399]\ttraining's auc: 0.977566\ttraining's binary_logloss: 0.178315\n",
      "[400]\ttraining's auc: 0.977568\ttraining's binary_logloss: 0.178301\n",
      "[401]\ttraining's auc: 0.977575\ttraining's binary_logloss: 0.178253\n",
      "[402]\ttraining's auc: 0.977579\ttraining's binary_logloss: 0.178181\n",
      "[403]\ttraining's auc: 0.977581\ttraining's binary_logloss: 0.1781\n",
      "[404]\ttraining's auc: 0.977583\ttraining's binary_logloss: 0.178091\n",
      "[405]\ttraining's auc: 0.977587\ttraining's binary_logloss: 0.178011\n",
      "[406]\ttraining's auc: 0.977594\ttraining's binary_logloss: 0.177905\n",
      "[407]\ttraining's auc: 0.9776\ttraining's binary_logloss: 0.177871\n",
      "[408]\ttraining's auc: 0.977607\ttraining's binary_logloss: 0.177835\n",
      "[409]\ttraining's auc: 0.977611\ttraining's binary_logloss: 0.177808\n",
      "[410]\ttraining's auc: 0.977616\ttraining's binary_logloss: 0.177786\n",
      "[411]\ttraining's auc: 0.977623\ttraining's binary_logloss: 0.17771\n",
      "[412]\ttraining's auc: 0.977639\ttraining's binary_logloss: 0.17764\n",
      "[413]\ttraining's auc: 0.977644\ttraining's binary_logloss: 0.177548\n",
      "[414]\ttraining's auc: 0.977657\ttraining's binary_logloss: 0.177468\n",
      "[415]\ttraining's auc: 0.977665\ttraining's binary_logloss: 0.17742\n",
      "[416]\ttraining's auc: 0.977678\ttraining's binary_logloss: 0.177341\n",
      "[417]\ttraining's auc: 0.977684\ttraining's binary_logloss: 0.177256\n",
      "[418]\ttraining's auc: 0.977687\ttraining's binary_logloss: 0.177245\n",
      "[419]\ttraining's auc: 0.977693\ttraining's binary_logloss: 0.177142\n",
      "[420]\ttraining's auc: 0.977697\ttraining's binary_logloss: 0.177081\n",
      "[421]\ttraining's auc: 0.977706\ttraining's binary_logloss: 0.177012\n",
      "[422]\ttraining's auc: 0.977717\ttraining's binary_logloss: 0.176957\n",
      "[423]\ttraining's auc: 0.977721\ttraining's binary_logloss: 0.176933\n",
      "[424]\ttraining's auc: 0.977725\ttraining's binary_logloss: 0.176888\n",
      "[425]\ttraining's auc: 0.977733\ttraining's binary_logloss: 0.176834\n",
      "[426]\ttraining's auc: 0.977737\ttraining's binary_logloss: 0.176804\n",
      "[427]\ttraining's auc: 0.97774\ttraining's binary_logloss: 0.176711\n",
      "[428]\ttraining's auc: 0.977741\ttraining's binary_logloss: 0.176595\n",
      "[429]\ttraining's auc: 0.977745\ttraining's binary_logloss: 0.176575\n",
      "[430]\ttraining's auc: 0.977755\ttraining's binary_logloss: 0.176523\n",
      "[431]\ttraining's auc: 0.977757\ttraining's binary_logloss: 0.176484\n",
      "[432]\ttraining's auc: 0.977759\ttraining's binary_logloss: 0.176462\n",
      "[433]\ttraining's auc: 0.977762\ttraining's binary_logloss: 0.17645\n",
      "[434]\ttraining's auc: 0.977772\ttraining's binary_logloss: 0.176369\n",
      "[435]\ttraining's auc: 0.977783\ttraining's binary_logloss: 0.176331\n",
      "[436]\ttraining's auc: 0.977788\ttraining's binary_logloss: 0.17626\n",
      "[437]\ttraining's auc: 0.977796\ttraining's binary_logloss: 0.176233\n",
      "[438]\ttraining's auc: 0.977805\ttraining's binary_logloss: 0.176172\n",
      "[439]\ttraining's auc: 0.977814\ttraining's binary_logloss: 0.176098\n",
      "[440]\ttraining's auc: 0.977821\ttraining's binary_logloss: 0.176072\n",
      "[441]\ttraining's auc: 0.977825\ttraining's binary_logloss: 0.176054\n",
      "[442]\ttraining's auc: 0.97783\ttraining's binary_logloss: 0.176009\n",
      "[443]\ttraining's auc: 0.977832\ttraining's binary_logloss: 0.175948\n",
      "[444]\ttraining's auc: 0.977836\ttraining's binary_logloss: 0.175941\n",
      "[445]\ttraining's auc: 0.97784\ttraining's binary_logloss: 0.175924\n",
      "[446]\ttraining's auc: 0.977844\ttraining's binary_logloss: 0.175906\n",
      "[447]\ttraining's auc: 0.977848\ttraining's binary_logloss: 0.175889\n",
      "[448]\ttraining's auc: 0.977851\ttraining's binary_logloss: 0.175877\n",
      "[449]\ttraining's auc: 0.977855\ttraining's binary_logloss: 0.175808\n",
      "[450]\ttraining's auc: 0.977864\ttraining's binary_logloss: 0.175759\n",
      "[451]\ttraining's auc: 0.977869\ttraining's binary_logloss: 0.175685\n",
      "[452]\ttraining's auc: 0.977875\ttraining's binary_logloss: 0.175646\n",
      "[453]\ttraining's auc: 0.977882\ttraining's binary_logloss: 0.175586\n",
      "[454]\ttraining's auc: 0.977886\ttraining's binary_logloss: 0.175523\n",
      "[455]\ttraining's auc: 0.97789\ttraining's binary_logloss: 0.175449\n",
      "[456]\ttraining's auc: 0.977893\ttraining's binary_logloss: 0.175388\n",
      "[457]\ttraining's auc: 0.977899\ttraining's binary_logloss: 0.17533\n",
      "[458]\ttraining's auc: 0.977904\ttraining's binary_logloss: 0.175289\n",
      "[459]\ttraining's auc: 0.977907\ttraining's binary_logloss: 0.175241\n",
      "[460]\ttraining's auc: 0.977913\ttraining's binary_logloss: 0.175173\n",
      "[461]\ttraining's auc: 0.977915\ttraining's binary_logloss: 0.175142\n",
      "[462]\ttraining's auc: 0.97792\ttraining's binary_logloss: 0.175126\n",
      "[463]\ttraining's auc: 0.977924\ttraining's binary_logloss: 0.175106\n",
      "[464]\ttraining's auc: 0.977931\ttraining's binary_logloss: 0.175069\n",
      "[465]\ttraining's auc: 0.977938\ttraining's binary_logloss: 0.175039\n",
      "[466]\ttraining's auc: 0.97794\ttraining's binary_logloss: 0.175015\n",
      "[467]\ttraining's auc: 0.977943\ttraining's binary_logloss: 0.175005\n",
      "[468]\ttraining's auc: 0.977951\ttraining's binary_logloss: 0.174911\n",
      "[469]\ttraining's auc: 0.977954\ttraining's binary_logloss: 0.174846\n",
      "[470]\ttraining's auc: 0.977957\ttraining's binary_logloss: 0.174832\n",
      "[471]\ttraining's auc: 0.977965\ttraining's binary_logloss: 0.174753\n",
      "[472]\ttraining's auc: 0.977966\ttraining's binary_logloss: 0.17474\n",
      "[473]\ttraining's auc: 0.97797\ttraining's binary_logloss: 0.17471\n",
      "[474]\ttraining's auc: 0.977975\ttraining's binary_logloss: 0.174671\n",
      "[475]\ttraining's auc: 0.977979\ttraining's binary_logloss: 0.174655\n",
      "[476]\ttraining's auc: 0.977983\ttraining's binary_logloss: 0.174629\n",
      "[477]\ttraining's auc: 0.97799\ttraining's binary_logloss: 0.174597\n",
      "[478]\ttraining's auc: 0.977999\ttraining's binary_logloss: 0.174548\n",
      "[479]\ttraining's auc: 0.978003\ttraining's binary_logloss: 0.174457\n",
      "[480]\ttraining's auc: 0.97801\ttraining's binary_logloss: 0.174408\n",
      "[481]\ttraining's auc: 0.978013\ttraining's binary_logloss: 0.174397\n",
      "[482]\ttraining's auc: 0.978019\ttraining's binary_logloss: 0.174359\n",
      "[483]\ttraining's auc: 0.978025\ttraining's binary_logloss: 0.174344\n",
      "[484]\ttraining's auc: 0.978031\ttraining's binary_logloss: 0.174325\n",
      "[485]\ttraining's auc: 0.978034\ttraining's binary_logloss: 0.174272\n",
      "[486]\ttraining's auc: 0.978038\ttraining's binary_logloss: 0.17421\n",
      "[487]\ttraining's auc: 0.978043\ttraining's binary_logloss: 0.174154\n",
      "[488]\ttraining's auc: 0.978048\ttraining's binary_logloss: 0.174104\n",
      "[489]\ttraining's auc: 0.978051\ttraining's binary_logloss: 0.17407\n",
      "[490]\ttraining's auc: 0.978053\ttraining's binary_logloss: 0.174062\n",
      "[491]\ttraining's auc: 0.978057\ttraining's binary_logloss: 0.174046\n",
      "[492]\ttraining's auc: 0.978063\ttraining's binary_logloss: 0.174009\n",
      "[493]\ttraining's auc: 0.97807\ttraining's binary_logloss: 0.173973\n",
      "[494]\ttraining's auc: 0.978077\ttraining's binary_logloss: 0.173902\n",
      "[495]\ttraining's auc: 0.978089\ttraining's binary_logloss: 0.173856\n",
      "[496]\ttraining's auc: 0.978097\ttraining's binary_logloss: 0.173812\n",
      "[497]\ttraining's auc: 0.978101\ttraining's binary_logloss: 0.173755\n",
      "[498]\ttraining's auc: 0.978105\ttraining's binary_logloss: 0.173739\n",
      "[499]\ttraining's auc: 0.978109\ttraining's binary_logloss: 0.173703\n",
      "[500]\ttraining's auc: 0.978113\ttraining's binary_logloss: 0.173671\n",
      "[501]\ttraining's auc: 0.978117\ttraining's binary_logloss: 0.173656\n",
      "[502]\ttraining's auc: 0.978119\ttraining's binary_logloss: 0.17365\n",
      "[503]\ttraining's auc: 0.978121\ttraining's binary_logloss: 0.173622\n",
      "[504]\ttraining's auc: 0.978122\ttraining's binary_logloss: 0.173584\n",
      "[505]\ttraining's auc: 0.978125\ttraining's binary_logloss: 0.173554\n",
      "[506]\ttraining's auc: 0.978137\ttraining's binary_logloss: 0.17348\n",
      "[507]\ttraining's auc: 0.978142\ttraining's binary_logloss: 0.173442\n",
      "[508]\ttraining's auc: 0.978146\ttraining's binary_logloss: 0.173429\n",
      "[509]\ttraining's auc: 0.978149\ttraining's binary_logloss: 0.173353\n",
      "[510]\ttraining's auc: 0.978154\ttraining's binary_logloss: 0.173279\n",
      "[511]\ttraining's auc: 0.978158\ttraining's binary_logloss: 0.17321\n",
      "[512]\ttraining's auc: 0.978162\ttraining's binary_logloss: 0.173149\n",
      "[513]\ttraining's auc: 0.978169\ttraining's binary_logloss: 0.173098\n",
      "[514]\ttraining's auc: 0.978173\ttraining's binary_logloss: 0.17305\n",
      "[515]\ttraining's auc: 0.978179\ttraining's binary_logloss: 0.172982\n",
      "[516]\ttraining's auc: 0.97818\ttraining's binary_logloss: 0.172947\n",
      "[517]\ttraining's auc: 0.978184\ttraining's binary_logloss: 0.172936\n",
      "[518]\ttraining's auc: 0.97819\ttraining's binary_logloss: 0.172911\n",
      "[519]\ttraining's auc: 0.978193\ttraining's binary_logloss: 0.172898\n",
      "[520]\ttraining's auc: 0.978197\ttraining's binary_logloss: 0.172845\n",
      "[521]\ttraining's auc: 0.978199\ttraining's binary_logloss: 0.172818\n",
      "[522]\ttraining's auc: 0.978204\ttraining's binary_logloss: 0.172769\n",
      "[523]\ttraining's auc: 0.978208\ttraining's binary_logloss: 0.172756\n",
      "[524]\ttraining's auc: 0.978214\ttraining's binary_logloss: 0.172715\n",
      "[525]\ttraining's auc: 0.978217\ttraining's binary_logloss: 0.172708\n",
      "[526]\ttraining's auc: 0.978219\ttraining's binary_logloss: 0.172653\n",
      "[527]\ttraining's auc: 0.978223\ttraining's binary_logloss: 0.172603\n",
      "[528]\ttraining's auc: 0.978226\ttraining's binary_logloss: 0.172542\n",
      "[529]\ttraining's auc: 0.978229\ttraining's binary_logloss: 0.172526\n",
      "[530]\ttraining's auc: 0.978229\ttraining's binary_logloss: 0.172504\n",
      "[531]\ttraining's auc: 0.978233\ttraining's binary_logloss: 0.172462\n",
      "[532]\ttraining's auc: 0.978234\ttraining's binary_logloss: 0.172449\n",
      "[533]\ttraining's auc: 0.978242\ttraining's binary_logloss: 0.172425\n",
      "[534]\ttraining's auc: 0.978244\ttraining's binary_logloss: 0.172407\n",
      "[535]\ttraining's auc: 0.978246\ttraining's binary_logloss: 0.17238\n",
      "[536]\ttraining's auc: 0.978249\ttraining's binary_logloss: 0.172363\n",
      "[537]\ttraining's auc: 0.978255\ttraining's binary_logloss: 0.172336\n",
      "[538]\ttraining's auc: 0.978258\ttraining's binary_logloss: 0.172316\n",
      "[539]\ttraining's auc: 0.978262\ttraining's binary_logloss: 0.172306\n",
      "[540]\ttraining's auc: 0.978266\ttraining's binary_logloss: 0.172262\n",
      "[541]\ttraining's auc: 0.978269\ttraining's binary_logloss: 0.172251\n",
      "[542]\ttraining's auc: 0.978271\ttraining's binary_logloss: 0.172196\n",
      "[543]\ttraining's auc: 0.978281\ttraining's binary_logloss: 0.172153\n",
      "[544]\ttraining's auc: 0.978283\ttraining's binary_logloss: 0.172115\n",
      "[545]\ttraining's auc: 0.978285\ttraining's binary_logloss: 0.172059\n",
      "[546]\ttraining's auc: 0.978288\ttraining's binary_logloss: 0.172042\n",
      "[547]\ttraining's auc: 0.978293\ttraining's binary_logloss: 0.172009\n",
      "[548]\ttraining's auc: 0.978297\ttraining's binary_logloss: 0.171982\n",
      "[549]\ttraining's auc: 0.978301\ttraining's binary_logloss: 0.171963\n",
      "[550]\ttraining's auc: 0.978305\ttraining's binary_logloss: 0.17191\n",
      "[551]\ttraining's auc: 0.978311\ttraining's binary_logloss: 0.171884\n",
      "[552]\ttraining's auc: 0.978315\ttraining's binary_logloss: 0.171845\n",
      "[553]\ttraining's auc: 0.978318\ttraining's binary_logloss: 0.171832\n",
      "[554]\ttraining's auc: 0.978325\ttraining's binary_logloss: 0.171795\n",
      "[555]\ttraining's auc: 0.978327\ttraining's binary_logloss: 0.171775\n",
      "[556]\ttraining's auc: 0.978331\ttraining's binary_logloss: 0.171745\n",
      "[557]\ttraining's auc: 0.978334\ttraining's binary_logloss: 0.17171\n",
      "[558]\ttraining's auc: 0.978334\ttraining's binary_logloss: 0.171686\n",
      "[559]\ttraining's auc: 0.978338\ttraining's binary_logloss: 0.171653\n",
      "[560]\ttraining's auc: 0.978342\ttraining's binary_logloss: 0.171639\n",
      "[561]\ttraining's auc: 0.978346\ttraining's binary_logloss: 0.171625\n",
      "[562]\ttraining's auc: 0.978349\ttraining's binary_logloss: 0.171615\n",
      "[563]\ttraining's auc: 0.978352\ttraining's binary_logloss: 0.171593\n",
      "[564]\ttraining's auc: 0.978355\ttraining's binary_logloss: 0.171586\n",
      "[565]\ttraining's auc: 0.97836\ttraining's binary_logloss: 0.171572\n",
      "[566]\ttraining's auc: 0.978368\ttraining's binary_logloss: 0.171536\n",
      "[567]\ttraining's auc: 0.978372\ttraining's binary_logloss: 0.171503\n",
      "[568]\ttraining's auc: 0.978376\ttraining's binary_logloss: 0.171483\n",
      "[569]\ttraining's auc: 0.978378\ttraining's binary_logloss: 0.171477\n",
      "[570]\ttraining's auc: 0.978382\ttraining's binary_logloss: 0.171445\n",
      "[571]\ttraining's auc: 0.978391\ttraining's binary_logloss: 0.171407\n",
      "[572]\ttraining's auc: 0.978395\ttraining's binary_logloss: 0.171383\n",
      "[573]\ttraining's auc: 0.978397\ttraining's binary_logloss: 0.171369\n",
      "[574]\ttraining's auc: 0.978401\ttraining's binary_logloss: 0.171345\n",
      "[575]\ttraining's auc: 0.978404\ttraining's binary_logloss: 0.171332\n",
      "[576]\ttraining's auc: 0.978406\ttraining's binary_logloss: 0.171324\n",
      "[577]\ttraining's auc: 0.97841\ttraining's binary_logloss: 0.17131\n",
      "[578]\ttraining's auc: 0.978413\ttraining's binary_logloss: 0.171293\n",
      "[579]\ttraining's auc: 0.978419\ttraining's binary_logloss: 0.17126\n",
      "[580]\ttraining's auc: 0.978423\ttraining's binary_logloss: 0.171233\n",
      "[581]\ttraining's auc: 0.978426\ttraining's binary_logloss: 0.171224\n",
      "[582]\ttraining's auc: 0.978429\ttraining's binary_logloss: 0.171194\n",
      "[583]\ttraining's auc: 0.978435\ttraining's binary_logloss: 0.171144\n",
      "[584]\ttraining's auc: 0.978438\ttraining's binary_logloss: 0.171132\n",
      "[585]\ttraining's auc: 0.978446\ttraining's binary_logloss: 0.17109\n",
      "[586]\ttraining's auc: 0.978453\ttraining's binary_logloss: 0.171062\n",
      "[587]\ttraining's auc: 0.978462\ttraining's binary_logloss: 0.171001\n",
      "[588]\ttraining's auc: 0.978467\ttraining's binary_logloss: 0.170945\n",
      "[589]\ttraining's auc: 0.978469\ttraining's binary_logloss: 0.170898\n",
      "[590]\ttraining's auc: 0.978472\ttraining's binary_logloss: 0.170871\n",
      "[591]\ttraining's auc: 0.978481\ttraining's binary_logloss: 0.170817\n",
      "[592]\ttraining's auc: 0.978484\ttraining's binary_logloss: 0.170805\n",
      "[593]\ttraining's auc: 0.97849\ttraining's binary_logloss: 0.170789\n",
      "[594]\ttraining's auc: 0.978492\ttraining's binary_logloss: 0.170782\n",
      "[595]\ttraining's auc: 0.978495\ttraining's binary_logloss: 0.170772\n",
      "[596]\ttraining's auc: 0.9785\ttraining's binary_logloss: 0.170739\n",
      "[597]\ttraining's auc: 0.978507\ttraining's binary_logloss: 0.170696\n",
      "[598]\ttraining's auc: 0.978511\ttraining's binary_logloss: 0.170681\n",
      "[599]\ttraining's auc: 0.978513\ttraining's binary_logloss: 0.170654\n",
      "[600]\ttraining's auc: 0.978518\ttraining's binary_logloss: 0.170634\n",
      "[601]\ttraining's auc: 0.978524\ttraining's binary_logloss: 0.170602\n",
      "[602]\ttraining's auc: 0.978527\ttraining's binary_logloss: 0.170568\n",
      "[603]\ttraining's auc: 0.978531\ttraining's binary_logloss: 0.170546\n",
      "[604]\ttraining's auc: 0.978535\ttraining's binary_logloss: 0.170528\n",
      "[605]\ttraining's auc: 0.97854\ttraining's binary_logloss: 0.170488\n",
      "[606]\ttraining's auc: 0.978545\ttraining's binary_logloss: 0.17045\n",
      "[607]\ttraining's auc: 0.978549\ttraining's binary_logloss: 0.17043\n",
      "[608]\ttraining's auc: 0.978551\ttraining's binary_logloss: 0.170419\n",
      "[609]\ttraining's auc: 0.978554\ttraining's binary_logloss: 0.170409\n",
      "[610]\ttraining's auc: 0.978557\ttraining's binary_logloss: 0.170402\n",
      "[611]\ttraining's auc: 0.978561\ttraining's binary_logloss: 0.170374\n",
      "[612]\ttraining's auc: 0.978562\ttraining's binary_logloss: 0.170368\n",
      "[613]\ttraining's auc: 0.978567\ttraining's binary_logloss: 0.170344\n",
      "[614]\ttraining's auc: 0.978568\ttraining's binary_logloss: 0.170339\n",
      "[615]\ttraining's auc: 0.978571\ttraining's binary_logloss: 0.170329\n",
      "[616]\ttraining's auc: 0.978573\ttraining's binary_logloss: 0.170319\n",
      "[617]\ttraining's auc: 0.978577\ttraining's binary_logloss: 0.170272\n",
      "[618]\ttraining's auc: 0.978579\ttraining's binary_logloss: 0.170226\n",
      "[619]\ttraining's auc: 0.978581\ttraining's binary_logloss: 0.170205\n",
      "[620]\ttraining's auc: 0.978585\ttraining's binary_logloss: 0.170195\n",
      "[621]\ttraining's auc: 0.978589\ttraining's binary_logloss: 0.170172\n",
      "[622]\ttraining's auc: 0.97859\ttraining's binary_logloss: 0.170154\n",
      "[623]\ttraining's auc: 0.978592\ttraining's binary_logloss: 0.170148\n",
      "[624]\ttraining's auc: 0.978594\ttraining's binary_logloss: 0.170142\n",
      "[625]\ttraining's auc: 0.978595\ttraining's binary_logloss: 0.170135\n",
      "[626]\ttraining's auc: 0.978602\ttraining's binary_logloss: 0.170095\n",
      "[627]\ttraining's auc: 0.978604\ttraining's binary_logloss: 0.170071\n",
      "[628]\ttraining's auc: 0.978608\ttraining's binary_logloss: 0.170046\n",
      "[629]\ttraining's auc: 0.978611\ttraining's binary_logloss: 0.170031\n",
      "[630]\ttraining's auc: 0.978617\ttraining's binary_logloss: 0.170009\n",
      "[631]\ttraining's auc: 0.97862\ttraining's binary_logloss: 0.169983\n",
      "[632]\ttraining's auc: 0.978624\ttraining's binary_logloss: 0.169955\n",
      "[633]\ttraining's auc: 0.978626\ttraining's binary_logloss: 0.169945\n",
      "[634]\ttraining's auc: 0.978631\ttraining's binary_logloss: 0.16993\n",
      "[635]\ttraining's auc: 0.978635\ttraining's binary_logloss: 0.169917\n",
      "[636]\ttraining's auc: 0.978637\ttraining's binary_logloss: 0.169907\n",
      "[637]\ttraining's auc: 0.978639\ttraining's binary_logloss: 0.169884\n",
      "[638]\ttraining's auc: 0.978645\ttraining's binary_logloss: 0.169869\n",
      "[639]\ttraining's auc: 0.978646\ttraining's binary_logloss: 0.169833\n",
      "[640]\ttraining's auc: 0.978647\ttraining's binary_logloss: 0.169828\n",
      "[641]\ttraining's auc: 0.978651\ttraining's binary_logloss: 0.169806\n",
      "[642]\ttraining's auc: 0.978653\ttraining's binary_logloss: 0.169803\n",
      "[643]\ttraining's auc: 0.978654\ttraining's binary_logloss: 0.169753\n",
      "[644]\ttraining's auc: 0.978656\ttraining's binary_logloss: 0.169712\n",
      "[645]\ttraining's auc: 0.978659\ttraining's binary_logloss: 0.169693\n",
      "[646]\ttraining's auc: 0.978662\ttraining's binary_logloss: 0.169637\n",
      "[647]\ttraining's auc: 0.978665\ttraining's binary_logloss: 0.169592\n",
      "[648]\ttraining's auc: 0.978668\ttraining's binary_logloss: 0.169577\n",
      "[649]\ttraining's auc: 0.97867\ttraining's binary_logloss: 0.169568\n",
      "[650]\ttraining's auc: 0.978672\ttraining's binary_logloss: 0.169545\n",
      "[651]\ttraining's auc: 0.978677\ttraining's binary_logloss: 0.169514\n",
      "[652]\ttraining's auc: 0.978679\ttraining's binary_logloss: 0.1695\n",
      "[653]\ttraining's auc: 0.978681\ttraining's binary_logloss: 0.169493\n",
      "[654]\ttraining's auc: 0.978682\ttraining's binary_logloss: 0.169489\n",
      "[655]\ttraining's auc: 0.978686\ttraining's binary_logloss: 0.169469\n",
      "[656]\ttraining's auc: 0.978689\ttraining's binary_logloss: 0.169444\n",
      "[657]\ttraining's auc: 0.978691\ttraining's binary_logloss: 0.169428\n",
      "[658]\ttraining's auc: 0.978694\ttraining's binary_logloss: 0.169386\n",
      "[659]\ttraining's auc: 0.978702\ttraining's binary_logloss: 0.169361\n",
      "[660]\ttraining's auc: 0.978704\ttraining's binary_logloss: 0.169353\n",
      "[661]\ttraining's auc: 0.97871\ttraining's binary_logloss: 0.169325\n",
      "[662]\ttraining's auc: 0.978713\ttraining's binary_logloss: 0.169297\n",
      "[663]\ttraining's auc: 0.97872\ttraining's binary_logloss: 0.169281\n",
      "[664]\ttraining's auc: 0.978725\ttraining's binary_logloss: 0.16926\n",
      "[665]\ttraining's auc: 0.978732\ttraining's binary_logloss: 0.169218\n",
      "[666]\ttraining's auc: 0.978735\ttraining's binary_logloss: 0.16921\n",
      "[667]\ttraining's auc: 0.978738\ttraining's binary_logloss: 0.169197\n",
      "[668]\ttraining's auc: 0.978739\ttraining's binary_logloss: 0.169178\n",
      "[669]\ttraining's auc: 0.978742\ttraining's binary_logloss: 0.169167\n",
      "[670]\ttraining's auc: 0.978745\ttraining's binary_logloss: 0.169131\n",
      "[671]\ttraining's auc: 0.978748\ttraining's binary_logloss: 0.169097\n",
      "[672]\ttraining's auc: 0.978749\ttraining's binary_logloss: 0.169093\n",
      "[673]\ttraining's auc: 0.978753\ttraining's binary_logloss: 0.169082\n",
      "[674]\ttraining's auc: 0.97876\ttraining's binary_logloss: 0.169055\n",
      "[675]\ttraining's auc: 0.978762\ttraining's binary_logloss: 0.169052\n",
      "[676]\ttraining's auc: 0.978764\ttraining's binary_logloss: 0.169033\n",
      "[677]\ttraining's auc: 0.978767\ttraining's binary_logloss: 0.169014\n",
      "[678]\ttraining's auc: 0.978771\ttraining's binary_logloss: 0.169005\n",
      "[679]\ttraining's auc: 0.978777\ttraining's binary_logloss: 0.168983\n",
      "[680]\ttraining's auc: 0.978782\ttraining's binary_logloss: 0.168969\n",
      "[681]\ttraining's auc: 0.978784\ttraining's binary_logloss: 0.168943\n",
      "[682]\ttraining's auc: 0.978786\ttraining's binary_logloss: 0.168908\n",
      "[683]\ttraining's auc: 0.978789\ttraining's binary_logloss: 0.1689\n",
      "[684]\ttraining's auc: 0.978793\ttraining's binary_logloss: 0.16887\n",
      "[685]\ttraining's auc: 0.978795\ttraining's binary_logloss: 0.168861\n",
      "[686]\ttraining's auc: 0.978798\ttraining's binary_logloss: 0.168854\n",
      "[687]\ttraining's auc: 0.9788\ttraining's binary_logloss: 0.168847\n",
      "[688]\ttraining's auc: 0.978801\ttraining's binary_logloss: 0.168828\n",
      "[689]\ttraining's auc: 0.978804\ttraining's binary_logloss: 0.168795\n",
      "[690]\ttraining's auc: 0.978805\ttraining's binary_logloss: 0.168757\n",
      "[691]\ttraining's auc: 0.978806\ttraining's binary_logloss: 0.168727\n",
      "[692]\ttraining's auc: 0.978809\ttraining's binary_logloss: 0.16871\n",
      "[693]\ttraining's auc: 0.978811\ttraining's binary_logloss: 0.168695\n",
      "[694]\ttraining's auc: 0.978814\ttraining's binary_logloss: 0.168687\n",
      "[695]\ttraining's auc: 0.978819\ttraining's binary_logloss: 0.168662\n",
      "[696]\ttraining's auc: 0.97882\ttraining's binary_logloss: 0.168629\n",
      "[697]\ttraining's auc: 0.978824\ttraining's binary_logloss: 0.168608\n",
      "[698]\ttraining's auc: 0.978826\ttraining's binary_logloss: 0.168575\n",
      "[699]\ttraining's auc: 0.978829\ttraining's binary_logloss: 0.168557\n",
      "[700]\ttraining's auc: 0.978833\ttraining's binary_logloss: 0.168529\n",
      "[701]\ttraining's auc: 0.978838\ttraining's binary_logloss: 0.168509\n",
      "[702]\ttraining's auc: 0.978842\ttraining's binary_logloss: 0.168484\n",
      "[703]\ttraining's auc: 0.978844\ttraining's binary_logloss: 0.168463\n",
      "[704]\ttraining's auc: 0.978846\ttraining's binary_logloss: 0.168443\n",
      "[705]\ttraining's auc: 0.978848\ttraining's binary_logloss: 0.168428\n",
      "[706]\ttraining's auc: 0.978855\ttraining's binary_logloss: 0.168409\n",
      "[707]\ttraining's auc: 0.978858\ttraining's binary_logloss: 0.1684\n",
      "[708]\ttraining's auc: 0.978859\ttraining's binary_logloss: 0.168397\n",
      "[709]\ttraining's auc: 0.97886\ttraining's binary_logloss: 0.168393\n",
      "[710]\ttraining's auc: 0.978864\ttraining's binary_logloss: 0.16837\n",
      "[711]\ttraining's auc: 0.978867\ttraining's binary_logloss: 0.168363\n",
      "[712]\ttraining's auc: 0.978867\ttraining's binary_logloss: 0.168345\n",
      "[713]\ttraining's auc: 0.978872\ttraining's binary_logloss: 0.168325\n",
      "[714]\ttraining's auc: 0.978874\ttraining's binary_logloss: 0.168318\n",
      "[715]\ttraining's auc: 0.978879\ttraining's binary_logloss: 0.168304\n",
      "[716]\ttraining's auc: 0.97888\ttraining's binary_logloss: 0.168297\n",
      "[717]\ttraining's auc: 0.978881\ttraining's binary_logloss: 0.168289\n",
      "[718]\ttraining's auc: 0.978885\ttraining's binary_logloss: 0.168278\n",
      "[719]\ttraining's auc: 0.978889\ttraining's binary_logloss: 0.168257\n",
      "[720]\ttraining's auc: 0.978892\ttraining's binary_logloss: 0.168239\n",
      "[721]\ttraining's auc: 0.978896\ttraining's binary_logloss: 0.16822\n",
      "[722]\ttraining's auc: 0.9789\ttraining's binary_logloss: 0.168207\n",
      "[723]\ttraining's auc: 0.978904\ttraining's binary_logloss: 0.168182\n",
      "[724]\ttraining's auc: 0.978905\ttraining's binary_logloss: 0.168128\n",
      "[725]\ttraining's auc: 0.978907\ttraining's binary_logloss: 0.16812\n",
      "[726]\ttraining's auc: 0.978909\ttraining's binary_logloss: 0.168114\n",
      "[727]\ttraining's auc: 0.978917\ttraining's binary_logloss: 0.168095\n",
      "[728]\ttraining's auc: 0.978918\ttraining's binary_logloss: 0.168059\n",
      "[729]\ttraining's auc: 0.978922\ttraining's binary_logloss: 0.168048\n",
      "[730]\ttraining's auc: 0.978922\ttraining's binary_logloss: 0.168013\n",
      "[731]\ttraining's auc: 0.978924\ttraining's binary_logloss: 0.167982\n",
      "[732]\ttraining's auc: 0.978927\ttraining's binary_logloss: 0.16797\n",
      "[733]\ttraining's auc: 0.97893\ttraining's binary_logloss: 0.167957\n",
      "[734]\ttraining's auc: 0.978934\ttraining's binary_logloss: 0.167946\n",
      "[735]\ttraining's auc: 0.978938\ttraining's binary_logloss: 0.167928\n",
      "[736]\ttraining's auc: 0.978941\ttraining's binary_logloss: 0.167915\n",
      "[737]\ttraining's auc: 0.978944\ttraining's binary_logloss: 0.167864\n",
      "[738]\ttraining's auc: 0.978945\ttraining's binary_logloss: 0.167828\n",
      "[739]\ttraining's auc: 0.978947\ttraining's binary_logloss: 0.167821\n",
      "[740]\ttraining's auc: 0.978949\ttraining's binary_logloss: 0.167807\n",
      "[741]\ttraining's auc: 0.978952\ttraining's binary_logloss: 0.167798\n",
      "[742]\ttraining's auc: 0.978961\ttraining's binary_logloss: 0.167754\n",
      "[743]\ttraining's auc: 0.978963\ttraining's binary_logloss: 0.167748\n",
      "[744]\ttraining's auc: 0.978967\ttraining's binary_logloss: 0.167729\n",
      "[745]\ttraining's auc: 0.978973\ttraining's binary_logloss: 0.167691\n",
      "[746]\ttraining's auc: 0.978976\ttraining's binary_logloss: 0.167681\n",
      "[747]\ttraining's auc: 0.978977\ttraining's binary_logloss: 0.167653\n",
      "[748]\ttraining's auc: 0.97898\ttraining's binary_logloss: 0.167627\n",
      "[749]\ttraining's auc: 0.978986\ttraining's binary_logloss: 0.1676\n",
      "[750]\ttraining's auc: 0.978987\ttraining's binary_logloss: 0.167583\n",
      "[751]\ttraining's auc: 0.978991\ttraining's binary_logloss: 0.16757\n",
      "[752]\ttraining's auc: 0.978992\ttraining's binary_logloss: 0.167565\n",
      "[753]\ttraining's auc: 0.978998\ttraining's binary_logloss: 0.167547\n",
      "[754]\ttraining's auc: 0.978999\ttraining's binary_logloss: 0.167533\n",
      "[755]\ttraining's auc: 0.979001\ttraining's binary_logloss: 0.167528\n",
      "[756]\ttraining's auc: 0.979002\ttraining's binary_logloss: 0.167522\n",
      "[757]\ttraining's auc: 0.979007\ttraining's binary_logloss: 0.167512\n",
      "[758]\ttraining's auc: 0.979008\ttraining's binary_logloss: 0.167504\n",
      "[759]\ttraining's auc: 0.97901\ttraining's binary_logloss: 0.167477\n",
      "[760]\ttraining's auc: 0.979013\ttraining's binary_logloss: 0.167467\n",
      "[761]\ttraining's auc: 0.979016\ttraining's binary_logloss: 0.167459\n",
      "[762]\ttraining's auc: 0.979019\ttraining's binary_logloss: 0.167441\n",
      "[763]\ttraining's auc: 0.979024\ttraining's binary_logloss: 0.167418\n",
      "[764]\ttraining's auc: 0.979025\ttraining's binary_logloss: 0.167415\n",
      "[765]\ttraining's auc: 0.979027\ttraining's binary_logloss: 0.167402\n",
      "[766]\ttraining's auc: 0.97903\ttraining's binary_logloss: 0.167378\n",
      "[767]\ttraining's auc: 0.979034\ttraining's binary_logloss: 0.167362\n",
      "[768]\ttraining's auc: 0.979037\ttraining's binary_logloss: 0.167347\n",
      "[769]\ttraining's auc: 0.979039\ttraining's binary_logloss: 0.167329\n",
      "[770]\ttraining's auc: 0.979038\ttraining's binary_logloss: 0.167306\n",
      "[771]\ttraining's auc: 0.979041\ttraining's binary_logloss: 0.167287\n",
      "[772]\ttraining's auc: 0.979042\ttraining's binary_logloss: 0.167263\n",
      "[773]\ttraining's auc: 0.979044\ttraining's binary_logloss: 0.167242\n",
      "[774]\ttraining's auc: 0.979048\ttraining's binary_logloss: 0.167218\n",
      "[775]\ttraining's auc: 0.979051\ttraining's binary_logloss: 0.167197\n",
      "[776]\ttraining's auc: 0.979054\ttraining's binary_logloss: 0.167184\n",
      "[777]\ttraining's auc: 0.979055\ttraining's binary_logloss: 0.16718\n",
      "[778]\ttraining's auc: 0.979056\ttraining's binary_logloss: 0.167166\n",
      "[779]\ttraining's auc: 0.979058\ttraining's binary_logloss: 0.16713\n",
      "[780]\ttraining's auc: 0.97906\ttraining's binary_logloss: 0.167109\n",
      "[781]\ttraining's auc: 0.979063\ttraining's binary_logloss: 0.167086\n",
      "[782]\ttraining's auc: 0.979066\ttraining's binary_logloss: 0.167077\n",
      "[783]\ttraining's auc: 0.979069\ttraining's binary_logloss: 0.167071\n",
      "[784]\ttraining's auc: 0.979073\ttraining's binary_logloss: 0.167057\n",
      "[785]\ttraining's auc: 0.979074\ttraining's binary_logloss: 0.167054\n",
      "[786]\ttraining's auc: 0.979079\ttraining's binary_logloss: 0.167037\n",
      "[787]\ttraining's auc: 0.979081\ttraining's binary_logloss: 0.167031\n",
      "[788]\ttraining's auc: 0.979082\ttraining's binary_logloss: 0.167019\n",
      "[789]\ttraining's auc: 0.979084\ttraining's binary_logloss: 0.167009\n",
      "[790]\ttraining's auc: 0.979087\ttraining's binary_logloss: 0.167002\n",
      "[791]\ttraining's auc: 0.97909\ttraining's binary_logloss: 0.166992\n",
      "[792]\ttraining's auc: 0.97909\ttraining's binary_logloss: 0.166968\n",
      "[793]\ttraining's auc: 0.979094\ttraining's binary_logloss: 0.166952\n",
      "[794]\ttraining's auc: 0.979096\ttraining's binary_logloss: 0.166939\n",
      "[795]\ttraining's auc: 0.979098\ttraining's binary_logloss: 0.166909\n",
      "[796]\ttraining's auc: 0.979103\ttraining's binary_logloss: 0.166874\n",
      "[797]\ttraining's auc: 0.979105\ttraining's binary_logloss: 0.166844\n",
      "[798]\ttraining's auc: 0.979107\ttraining's binary_logloss: 0.16682\n",
      "[799]\ttraining's auc: 0.979108\ttraining's binary_logloss: 0.166816\n",
      "[800]\ttraining's auc: 0.979111\ttraining's binary_logloss: 0.166799\n",
      "[801]\ttraining's auc: 0.979113\ttraining's binary_logloss: 0.166785\n",
      "[802]\ttraining's auc: 0.979118\ttraining's binary_logloss: 0.166768\n",
      "[803]\ttraining's auc: 0.979121\ttraining's binary_logloss: 0.166753\n",
      "[804]\ttraining's auc: 0.979122\ttraining's binary_logloss: 0.166745\n",
      "[805]\ttraining's auc: 0.979124\ttraining's binary_logloss: 0.166738\n",
      "[806]\ttraining's auc: 0.979126\ttraining's binary_logloss: 0.166732\n",
      "[807]\ttraining's auc: 0.979128\ttraining's binary_logloss: 0.166714\n",
      "[808]\ttraining's auc: 0.979129\ttraining's binary_logloss: 0.166702\n",
      "[809]\ttraining's auc: 0.979134\ttraining's binary_logloss: 0.16669\n",
      "[810]\ttraining's auc: 0.979135\ttraining's binary_logloss: 0.166677\n",
      "[811]\ttraining's auc: 0.979137\ttraining's binary_logloss: 0.166672\n",
      "[812]\ttraining's auc: 0.979138\ttraining's binary_logloss: 0.16665\n",
      "[813]\ttraining's auc: 0.979139\ttraining's binary_logloss: 0.166645\n",
      "[814]\ttraining's auc: 0.97914\ttraining's binary_logloss: 0.166632\n",
      "[815]\ttraining's auc: 0.979144\ttraining's binary_logloss: 0.166613\n",
      "[816]\ttraining's auc: 0.979147\ttraining's binary_logloss: 0.166596\n",
      "[817]\ttraining's auc: 0.979147\ttraining's binary_logloss: 0.166562\n",
      "[818]\ttraining's auc: 0.979149\ttraining's binary_logloss: 0.166535\n",
      "[819]\ttraining's auc: 0.979151\ttraining's binary_logloss: 0.166504\n",
      "[820]\ttraining's auc: 0.979154\ttraining's binary_logloss: 0.166488\n",
      "[821]\ttraining's auc: 0.979155\ttraining's binary_logloss: 0.166484\n",
      "[822]\ttraining's auc: 0.979157\ttraining's binary_logloss: 0.166469\n",
      "[823]\ttraining's auc: 0.97916\ttraining's binary_logloss: 0.166454\n",
      "[824]\ttraining's auc: 0.979161\ttraining's binary_logloss: 0.166451\n",
      "[825]\ttraining's auc: 0.979162\ttraining's binary_logloss: 0.166441\n",
      "[826]\ttraining's auc: 0.979164\ttraining's binary_logloss: 0.166427\n",
      "[827]\ttraining's auc: 0.979166\ttraining's binary_logloss: 0.166398\n",
      "[828]\ttraining's auc: 0.979166\ttraining's binary_logloss: 0.166383\n",
      "[829]\ttraining's auc: 0.979168\ttraining's binary_logloss: 0.166377\n",
      "[830]\ttraining's auc: 0.97917\ttraining's binary_logloss: 0.16637\n",
      "[831]\ttraining's auc: 0.979171\ttraining's binary_logloss: 0.166365\n",
      "[832]\ttraining's auc: 0.979177\ttraining's binary_logloss: 0.166345\n",
      "[833]\ttraining's auc: 0.979179\ttraining's binary_logloss: 0.166331\n",
      "[834]\ttraining's auc: 0.979181\ttraining's binary_logloss: 0.166314\n",
      "[835]\ttraining's auc: 0.979184\ttraining's binary_logloss: 0.166301\n",
      "[836]\ttraining's auc: 0.979187\ttraining's binary_logloss: 0.166296\n",
      "[837]\ttraining's auc: 0.979188\ttraining's binary_logloss: 0.166292\n",
      "[838]\ttraining's auc: 0.97919\ttraining's binary_logloss: 0.16626\n",
      "[839]\ttraining's auc: 0.979192\ttraining's binary_logloss: 0.16624\n",
      "[840]\ttraining's auc: 0.979194\ttraining's binary_logloss: 0.166216\n",
      "[841]\ttraining's auc: 0.979194\ttraining's binary_logloss: 0.166193\n",
      "[842]\ttraining's auc: 0.979197\ttraining's binary_logloss: 0.166177\n",
      "[843]\ttraining's auc: 0.9792\ttraining's binary_logloss: 0.166152\n",
      "[844]\ttraining's auc: 0.979203\ttraining's binary_logloss: 0.166134\n",
      "[845]\ttraining's auc: 0.979206\ttraining's binary_logloss: 0.166124\n",
      "[846]\ttraining's auc: 0.979209\ttraining's binary_logloss: 0.166112\n",
      "[847]\ttraining's auc: 0.97921\ttraining's binary_logloss: 0.166107\n",
      "[848]\ttraining's auc: 0.979211\ttraining's binary_logloss: 0.166102\n",
      "[849]\ttraining's auc: 0.979214\ttraining's binary_logloss: 0.166085\n",
      "[850]\ttraining's auc: 0.979217\ttraining's binary_logloss: 0.166069\n",
      "[851]\ttraining's auc: 0.979219\ttraining's binary_logloss: 0.166052\n",
      "[852]\ttraining's auc: 0.979223\ttraining's binary_logloss: 0.166035\n",
      "[853]\ttraining's auc: 0.979223\ttraining's binary_logloss: 0.166006\n",
      "[854]\ttraining's auc: 0.979227\ttraining's binary_logloss: 0.165993\n",
      "[855]\ttraining's auc: 0.97923\ttraining's binary_logloss: 0.165979\n",
      "[856]\ttraining's auc: 0.979234\ttraining's binary_logloss: 0.165959\n",
      "[857]\ttraining's auc: 0.979237\ttraining's binary_logloss: 0.165944\n",
      "[858]\ttraining's auc: 0.979239\ttraining's binary_logloss: 0.165929\n",
      "[859]\ttraining's auc: 0.979241\ttraining's binary_logloss: 0.16592\n",
      "[860]\ttraining's auc: 0.979242\ttraining's binary_logloss: 0.165911\n",
      "[861]\ttraining's auc: 0.979245\ttraining's binary_logloss: 0.165905\n",
      "[862]\ttraining's auc: 0.979246\ttraining's binary_logloss: 0.165878\n",
      "[863]\ttraining's auc: 0.979247\ttraining's binary_logloss: 0.165873\n",
      "[864]\ttraining's auc: 0.97925\ttraining's binary_logloss: 0.165852\n",
      "[865]\ttraining's auc: 0.979252\ttraining's binary_logloss: 0.165842\n",
      "[866]\ttraining's auc: 0.979255\ttraining's binary_logloss: 0.165836\n",
      "[867]\ttraining's auc: 0.979257\ttraining's binary_logloss: 0.16583\n",
      "[868]\ttraining's auc: 0.979258\ttraining's binary_logloss: 0.165825\n",
      "[869]\ttraining's auc: 0.97926\ttraining's binary_logloss: 0.16582\n",
      "[870]\ttraining's auc: 0.979264\ttraining's binary_logloss: 0.165808\n",
      "[871]\ttraining's auc: 0.979265\ttraining's binary_logloss: 0.165802\n",
      "[872]\ttraining's auc: 0.979266\ttraining's binary_logloss: 0.16579\n",
      "[873]\ttraining's auc: 0.979268\ttraining's binary_logloss: 0.165783\n",
      "[874]\ttraining's auc: 0.979269\ttraining's binary_logloss: 0.165767\n",
      "[875]\ttraining's auc: 0.979274\ttraining's binary_logloss: 0.165746\n",
      "[876]\ttraining's auc: 0.979275\ttraining's binary_logloss: 0.165734\n",
      "[877]\ttraining's auc: 0.979279\ttraining's binary_logloss: 0.165723\n",
      "[878]\ttraining's auc: 0.97928\ttraining's binary_logloss: 0.16571\n",
      "[879]\ttraining's auc: 0.97928\ttraining's binary_logloss: 0.1657\n",
      "[880]\ttraining's auc: 0.979282\ttraining's binary_logloss: 0.165692\n",
      "[881]\ttraining's auc: 0.979284\ttraining's binary_logloss: 0.165679\n",
      "[882]\ttraining's auc: 0.97929\ttraining's binary_logloss: 0.165656\n",
      "[883]\ttraining's auc: 0.979291\ttraining's binary_logloss: 0.16564\n",
      "[884]\ttraining's auc: 0.979294\ttraining's binary_logloss: 0.16563\n",
      "[885]\ttraining's auc: 0.979295\ttraining's binary_logloss: 0.165625\n",
      "[886]\ttraining's auc: 0.979298\ttraining's binary_logloss: 0.165604\n",
      "[887]\ttraining's auc: 0.9793\ttraining's binary_logloss: 0.165591\n",
      "[888]\ttraining's auc: 0.979302\ttraining's binary_logloss: 0.165578\n",
      "[889]\ttraining's auc: 0.979306\ttraining's binary_logloss: 0.165557\n",
      "[890]\ttraining's auc: 0.979309\ttraining's binary_logloss: 0.165545\n",
      "[891]\ttraining's auc: 0.97931\ttraining's binary_logloss: 0.165533\n",
      "[892]\ttraining's auc: 0.979314\ttraining's binary_logloss: 0.165523\n",
      "[893]\ttraining's auc: 0.979315\ttraining's binary_logloss: 0.165507\n",
      "[894]\ttraining's auc: 0.979317\ttraining's binary_logloss: 0.165491\n",
      "[895]\ttraining's auc: 0.979319\ttraining's binary_logloss: 0.165484\n",
      "[896]\ttraining's auc: 0.979322\ttraining's binary_logloss: 0.16547\n",
      "[897]\ttraining's auc: 0.979322\ttraining's binary_logloss: 0.16546\n",
      "[898]\ttraining's auc: 0.979323\ttraining's binary_logloss: 0.16545\n",
      "[899]\ttraining's auc: 0.979326\ttraining's binary_logloss: 0.165443\n",
      "[900]\ttraining's auc: 0.979328\ttraining's binary_logloss: 0.165433\n",
      "[901]\ttraining's auc: 0.979332\ttraining's binary_logloss: 0.165413\n",
      "[902]\ttraining's auc: 0.979335\ttraining's binary_logloss: 0.165401\n",
      "[903]\ttraining's auc: 0.979337\ttraining's binary_logloss: 0.165391\n",
      "[904]\ttraining's auc: 0.979338\ttraining's binary_logloss: 0.165373\n",
      "[905]\ttraining's auc: 0.979339\ttraining's binary_logloss: 0.165362\n",
      "[906]\ttraining's auc: 0.979341\ttraining's binary_logloss: 0.165342\n",
      "[907]\ttraining's auc: 0.979342\ttraining's binary_logloss: 0.165331\n",
      "[908]\ttraining's auc: 0.979343\ttraining's binary_logloss: 0.165328\n",
      "[909]\ttraining's auc: 0.979347\ttraining's binary_logloss: 0.165309\n",
      "[910]\ttraining's auc: 0.97935\ttraining's binary_logloss: 0.165302\n",
      "[911]\ttraining's auc: 0.979352\ttraining's binary_logloss: 0.165292\n",
      "[912]\ttraining's auc: 0.979354\ttraining's binary_logloss: 0.165275\n",
      "[913]\ttraining's auc: 0.979358\ttraining's binary_logloss: 0.165264\n",
      "[914]\ttraining's auc: 0.979359\ttraining's binary_logloss: 0.165259\n",
      "[915]\ttraining's auc: 0.979362\ttraining's binary_logloss: 0.165243\n",
      "[916]\ttraining's auc: 0.979366\ttraining's binary_logloss: 0.165233\n",
      "[917]\ttraining's auc: 0.979367\ttraining's binary_logloss: 0.16523\n",
      "[918]\ttraining's auc: 0.97937\ttraining's binary_logloss: 0.165213\n",
      "[919]\ttraining's auc: 0.979369\ttraining's binary_logloss: 0.165199\n",
      "[920]\ttraining's auc: 0.979371\ttraining's binary_logloss: 0.16519\n",
      "[921]\ttraining's auc: 0.979372\ttraining's binary_logloss: 0.165181\n",
      "[922]\ttraining's auc: 0.979373\ttraining's binary_logloss: 0.165169\n",
      "[923]\ttraining's auc: 0.979377\ttraining's binary_logloss: 0.165158\n",
      "[924]\ttraining's auc: 0.979379\ttraining's binary_logloss: 0.165143\n",
      "[925]\ttraining's auc: 0.979381\ttraining's binary_logloss: 0.165134\n",
      "[926]\ttraining's auc: 0.979381\ttraining's binary_logloss: 0.165128\n",
      "[927]\ttraining's auc: 0.979383\ttraining's binary_logloss: 0.165117\n",
      "[928]\ttraining's auc: 0.979385\ttraining's binary_logloss: 0.165109\n",
      "[929]\ttraining's auc: 0.979389\ttraining's binary_logloss: 0.165099\n",
      "[930]\ttraining's auc: 0.979389\ttraining's binary_logloss: 0.165085\n",
      "[931]\ttraining's auc: 0.979392\ttraining's binary_logloss: 0.165076\n",
      "[932]\ttraining's auc: 0.979394\ttraining's binary_logloss: 0.165069\n",
      "[933]\ttraining's auc: 0.979396\ttraining's binary_logloss: 0.165052\n",
      "[934]\ttraining's auc: 0.979399\ttraining's binary_logloss: 0.16504\n",
      "[935]\ttraining's auc: 0.9794\ttraining's binary_logloss: 0.165022\n",
      "[936]\ttraining's auc: 0.979402\ttraining's binary_logloss: 0.16501\n",
      "[937]\ttraining's auc: 0.979405\ttraining's binary_logloss: 0.164995\n",
      "[938]\ttraining's auc: 0.979407\ttraining's binary_logloss: 0.164986\n",
      "[939]\ttraining's auc: 0.979409\ttraining's binary_logloss: 0.164979\n",
      "[940]\ttraining's auc: 0.97941\ttraining's binary_logloss: 0.164972\n",
      "[941]\ttraining's auc: 0.979412\ttraining's binary_logloss: 0.164959\n",
      "[942]\ttraining's auc: 0.979414\ttraining's binary_logloss: 0.164952\n",
      "[943]\ttraining's auc: 0.979416\ttraining's binary_logloss: 0.164942\n",
      "[944]\ttraining's auc: 0.979418\ttraining's binary_logloss: 0.164928\n",
      "[945]\ttraining's auc: 0.979421\ttraining's binary_logloss: 0.164914\n",
      "[946]\ttraining's auc: 0.979424\ttraining's binary_logloss: 0.164905\n",
      "[947]\ttraining's auc: 0.979426\ttraining's binary_logloss: 0.164886\n",
      "[948]\ttraining's auc: 0.979428\ttraining's binary_logloss: 0.164878\n",
      "[949]\ttraining's auc: 0.97943\ttraining's binary_logloss: 0.164873\n",
      "[950]\ttraining's auc: 0.979431\ttraining's binary_logloss: 0.164858\n",
      "[951]\ttraining's auc: 0.979432\ttraining's binary_logloss: 0.16485\n",
      "[952]\ttraining's auc: 0.979434\ttraining's binary_logloss: 0.164834\n",
      "[953]\ttraining's auc: 0.979436\ttraining's binary_logloss: 0.164826\n",
      "[954]\ttraining's auc: 0.97944\ttraining's binary_logloss: 0.164807\n",
      "[955]\ttraining's auc: 0.979444\ttraining's binary_logloss: 0.164788\n",
      "[956]\ttraining's auc: 0.979448\ttraining's binary_logloss: 0.164773\n",
      "[957]\ttraining's auc: 0.97945\ttraining's binary_logloss: 0.164764\n",
      "[958]\ttraining's auc: 0.979451\ttraining's binary_logloss: 0.164739\n",
      "[959]\ttraining's auc: 0.979453\ttraining's binary_logloss: 0.164729\n",
      "[960]\ttraining's auc: 0.979454\ttraining's binary_logloss: 0.164725\n",
      "[961]\ttraining's auc: 0.979456\ttraining's binary_logloss: 0.164715\n",
      "[962]\ttraining's auc: 0.979458\ttraining's binary_logloss: 0.164699\n",
      "[963]\ttraining's auc: 0.97946\ttraining's binary_logloss: 0.164688\n",
      "[964]\ttraining's auc: 0.979463\ttraining's binary_logloss: 0.164678\n",
      "[965]\ttraining's auc: 0.979467\ttraining's binary_logloss: 0.164667\n",
      "[966]\ttraining's auc: 0.979469\ttraining's binary_logloss: 0.164659\n",
      "[967]\ttraining's auc: 0.97947\ttraining's binary_logloss: 0.164639\n",
      "[968]\ttraining's auc: 0.979472\ttraining's binary_logloss: 0.164632\n",
      "[969]\ttraining's auc: 0.979473\ttraining's binary_logloss: 0.164627\n",
      "[970]\ttraining's auc: 0.979474\ttraining's binary_logloss: 0.164614\n",
      "[971]\ttraining's auc: 0.979477\ttraining's binary_logloss: 0.164595\n",
      "[972]\ttraining's auc: 0.979482\ttraining's binary_logloss: 0.164583\n",
      "[973]\ttraining's auc: 0.979485\ttraining's binary_logloss: 0.164574\n",
      "[974]\ttraining's auc: 0.979488\ttraining's binary_logloss: 0.164568\n",
      "[975]\ttraining's auc: 0.979489\ttraining's binary_logloss: 0.164553\n",
      "[976]\ttraining's auc: 0.979493\ttraining's binary_logloss: 0.16454\n",
      "[977]\ttraining's auc: 0.979494\ttraining's binary_logloss: 0.16452\n",
      "[978]\ttraining's auc: 0.979496\ttraining's binary_logloss: 0.164511\n",
      "[979]\ttraining's auc: 0.979498\ttraining's binary_logloss: 0.164507\n",
      "[980]\ttraining's auc: 0.979499\ttraining's binary_logloss: 0.164491\n",
      "[981]\ttraining's auc: 0.979502\ttraining's binary_logloss: 0.164478\n",
      "[982]\ttraining's auc: 0.979508\ttraining's binary_logloss: 0.164455\n",
      "[983]\ttraining's auc: 0.979511\ttraining's binary_logloss: 0.164447\n",
      "[984]\ttraining's auc: 0.979515\ttraining's binary_logloss: 0.164434\n",
      "[985]\ttraining's auc: 0.979519\ttraining's binary_logloss: 0.16442\n",
      "[986]\ttraining's auc: 0.979521\ttraining's binary_logloss: 0.164408\n",
      "[987]\ttraining's auc: 0.979523\ttraining's binary_logloss: 0.164399\n",
      "[988]\ttraining's auc: 0.979524\ttraining's binary_logloss: 0.164387\n",
      "[989]\ttraining's auc: 0.979524\ttraining's binary_logloss: 0.164385\n",
      "[990]\ttraining's auc: 0.979529\ttraining's binary_logloss: 0.164367\n",
      "[991]\ttraining's auc: 0.979531\ttraining's binary_logloss: 0.164355\n",
      "[992]\ttraining's auc: 0.979534\ttraining's binary_logloss: 0.164325\n",
      "[993]\ttraining's auc: 0.979536\ttraining's binary_logloss: 0.164317\n",
      "[994]\ttraining's auc: 0.979538\ttraining's binary_logloss: 0.164305\n",
      "[995]\ttraining's auc: 0.979543\ttraining's binary_logloss: 0.164287\n",
      "[996]\ttraining's auc: 0.979546\ttraining's binary_logloss: 0.164274\n",
      "[997]\ttraining's auc: 0.979546\ttraining's binary_logloss: 0.164253\n",
      "[998]\ttraining's auc: 0.97955\ttraining's binary_logloss: 0.164232\n",
      "[999]\ttraining's auc: 0.979551\ttraining's binary_logloss: 0.164227\n",
      "[1000]\ttraining's auc: 0.979553\ttraining's binary_logloss: 0.164216\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LGBMClassifier(bagging_fraction=0.8772503866491428, bagging_freq=3,\n",
       "               early_stopping_round=100, feature_fraction=0.4,\n",
       "               feature_pre_filter=False, lambda_l1=0.0, lambda_l2=0.0,\n",
       "               metric=&#x27;binary_logloss&#x27;, num_iterations=1000, num_leaves=7,\n",
       "               objective=&#x27;binary&#x27;, random_state=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LGBMClassifier</label><div class=\"sk-toggleable__content\"><pre>LGBMClassifier(bagging_fraction=0.8772503866491428, bagging_freq=3,\n",
       "               early_stopping_round=100, feature_fraction=0.4,\n",
       "               feature_pre_filter=False, lambda_l1=0.0, lambda_l2=0.0,\n",
       "               metric=&#x27;binary_logloss&#x27;, num_iterations=1000, num_leaves=7,\n",
       "               objective=&#x27;binary&#x27;, random_state=0)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "LGBMClassifier(bagging_fraction=0.8772503866491428, bagging_freq=3,\n",
       "               early_stopping_round=100, feature_fraction=0.4,\n",
       "               feature_pre_filter=False, lambda_l1=0.0, lambda_l2=0.0,\n",
       "               metric='binary_logloss', num_iterations=1000, num_leaves=7,\n",
       "               objective='binary', random_state=0)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from lightgbm import LGBMClassifier\n",
    "clf = LGBMClassifier(**params)\n",
    "clf.fit(resampled_x, resampled_y,\n",
    "        eval_set=[(resampled_x, resampled_y)],\n",
    "        eval_metric='auc',\n",
    "        early_stopping_rounds=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "2ae09ad5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.93      0.97      0.95     58417\n",
      "         1.0       0.38      0.17      0.23      5378\n",
      "\n",
      "    accuracy                           0.91     63795\n",
      "   macro avg       0.65      0.57      0.59     63795\n",
      "weighted avg       0.88      0.91      0.89     63795\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "preds = clf.predict(test)\n",
    "print(classification_report(test_y,preds ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "890ee1ba",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "db438b80",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAApwAAAJOCAYAAAAavdeBAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAA7NUlEQVR4nO3de7ymc73/8ddyyKGmnJZDR4refgrpgCRmCEOEECmlpNLODqlkd9plF6UD9UtKqN1BSpJDDjFDYZLK1o4+2CnJllU6TL9CM7N+f9zX7H1bzZpZY+Za973Wej0fD4913df1va7rc9/fHrz7XNd93QPDw8NIkiRJbVmh1wVIkiRpcjNwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWrVSrwvQ6ObNmz/8hz/8tddlTHlrrrk6zkPvOQ/9wXnoD85Df3AeHm5wcNrAaNvscPaxlVZasdclCOehXzgP/cF56A/OQ39wHsbOwClJkqRWDfhLQ/3rvk9e6eRIkqRlMnDQ1uNyHi+pS5IkqWcMnJIkSWqVgVOSJEmtMnBKkiSpVQZOSZIktcrAKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSq1bqdQGLk2Q/4CTg6VW14BHsPwicDGwKPATcXlWvHWXs44CNquqmR16xJEmSRur3DucrgYuBXZd2xyQrA98CzqqqbarqhcARi9llK+CgR1LkGGoZ9aeeJEmSJru+7XAmeTIwF/gYnS7lpUmeBHyeTlC+AdipqrZNsj5wOjANGAIOAV4CXFlVsxces6oebI79Gjphdo1mv4uBTwBrJXl8Vb0qyfuAGcAAcHRV/SjJDOCDwJ+AXwF/qKrjkmwH/Fsz9u/AG6rqF0lmA5cCOyQ5G9isqt7X1HA5cFBV3b98PzlJkqT+0s8dztfS6U7+Cli9CZUnAv9aVS8Czu8a+xHgfVW1E3A1cCCwCXDTKMe+qKp2BnYAjqiq3wBHAV9pwuaLgDWqakdgH+Bfm/1OBvasqpnAHV3HOxV4WVVNB44HPty17cdVtQfwTeBFSVZI8kzgl4ZNSZI0FfRlhzPJCsABwFZJ3gIMAq8BnlpV1zbDftS1yxbAx5MArAp8HbgLeNoop3h1knWBec34kZ4N7Nx0KAFWbMbfXVW/7zr/rs19ovdU1RBAVf0wyRO6jnVds35ekkuAFwF7AZ9a8ichSZI08fVrh3M34Nyq2ruq9gG2B/YF/phky2bMTl3jbwcObTqM29EJcxcCByfZfOGgJI9OsjawS1W9vRm3WrN5PrBKs3xbc/7pzTF3A+4HNk7ymBHn/x3wpOa4JHkO8F9dtc3rWv4scBjwpKq6eek+EkmSpImpXwPn4XS6lABU1UPAjXTuh/xcklnAtsADzZDjgTOTXAWcB6xVVX+m8yWgf0tyTbPtvXSC41+TXEvnMvp9zTF+SqereSZwAfCEJN9P8h06l9HnAScA1ya5DHgM8EBVDTfHuaA5x78Axy7qTVXV7+h85l9c1g9IkiRpohgYHh7udQ1jlmTlqvp7s7wvsGNVHdWj858CzKqqby3F/o+jE5q3r6r5Sxp/3yevnDiTI0mS+tLAQVuPy3kGB6eN+lSevryHczEOSnJYs3w/8PpxPv9JSbYCVgZ+QKcTOiZJZgIfAN4xlrApSZI0WUyoDudUY4dTkiQtq37ocPbrPZySJEmaJAyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVvlYpP42PDQ0t9c1THmDg9NwHnrPeegPzkN/cB76g/PwcD4WSZIkST1j4JQkSVKrDJySJElqlYFTkiRJrTJwSpIkqVUr9boAje7ej/2s1yUIuLfXBQhwHvqF89AfxjoPKx7y5FbrkMbKDqckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklrlLw0tgyR/Bn4MTAM+V1WfSbIh8AvgmVV1S9fYxzfrN62qXyaZU1Xb9qJuSZKk8WSHc9ncUlXTga2BI7rW/ww4ZsTYo4DbxqcsSZKk/mHgXD42AH7T9fpnwHpJ1gNI8jjgGc16SZKkKcXAuWw2S3IN8CPgghHbTgXe3Cy/ETgdGBjH2iRJkvqCgXPZ3FJVOwBPAF6QZLuFG6rqCmDbJGsCuwEX9qhGSZKknjJwLgdVNQ/4I/DoEZvOBM4DvlxVw+NdlyRJUj8wcC6bzZLMTvI94CHguyO2nwusCnxp3CuTJEnqEwPDwzbe+tW9H/uZkyNJesRWPOTJvS5hUhscnMbQ0Nxel9E3BgenjfpdFTuckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklpl4JQkSVKrfPB7fxv2gbK954N9+4Pz0B+ch/7gPPQH5+HhfPC7JEmSesbAKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSq1bqdQEa3e2f+m2vSxDwR/7a6xKE89AvejkPaxy4es/OLWnZ2OGUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklpl4JQkSVKrJs0vDSVZETgJ2ApYHbgC2AGYWVUPLOOxNwbOAFYFngrcAvwI+HfgmVX1pWU5viRJ0mQ2aQInMBOYX1U7AyRZhU7gXGZVdQcwPcmGwIlVdVDX5puWxzkkSZImq8kUOO8EjkwyWFVDVfVgEuB/up+fAjYFFgCvq6o7k7wEeBudWwvOqKqzkryPTod0S2Bd4JSqOntRJ0wynU4H9bgks4ELgT2BW4GbgZcBjwb2qqr7FnW+5f8xSJIk9ZdJcw9nVd1CJ8ydluQDTYdzodd0htQM4EjgXUnWAI4CdgK2B16ZZNVm/EpVtRvwfOCfk6w2xjJubM6xLrBmVe0EfAN46RLOJ0mSNGlNpg4nVfVTYP8kM4GzuzY9G3hekn2a10PA04FN6NzrCbAOsF6zfEVzvAeS3N5s+/UYSrih+XsHMKdZvhPYeDHn+9XY3p0kSdLENGk6nEnWT/Ko5uX3gA27Nt8GnFpV06tqOvBqOkHwZmBGs27rqloY/rZujvk4OiHxnjGWMTzKMks4nyRJ0qQ1aQIn8AzgB0mupnMv5Xu6tn0W2DPJ1UkuAbapqiHgW8D1SS4H3tk1/vFJLgUuA46rqvnLWtwSzidJkjRpDQwPj2zETW3Nl4bmVNWlva7l9k/91smRpMYaB67e6xL6xuDgNIaG5va6jCnPeXi4wcFpA6Ntm0wdTkmSJPWhSfWloeWhqt7X6xokSZImEzuckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1Cofi9THNnnzej5Qtg/4YN/+4Dz0B+dB0iNhh1OSJEmtMnBKkiSpVQZOSZIktcrAKUmSpFYZOCVJktQqv6Xex6768lCvSxAAD/S6AAHOw5JtvuuqvS5BkhbJDqckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklo1aQJnkrWSnJHkuiTfa/7ZYCmPMTvJql2vN01y9iOoZU7zd5Uk245cL0mSNJVMip+2TLIy8G3gXVU1u1m3Sk+L6tgAOAo4qMd1SJIk9cykCJzAPsB3F4ZNgKp6ECDJ+4AZwABwdFX9KMls4DvATGA1YK+qWuwPlydZHzgdmAYMAYdU1UNJzgHWa45zcFX9ohm/InAO8NQkl1fVrs36E4AdgQXAzKr62/L4ACRJkvrVZLmk/jTgZvifS+uzk9yU5L3AGlW1I51Q+q9d+9xUVTOAc3l4B/LyZv/ZwBe71n8EeF9V7QRcDRzYrD+yOc5pwMsXDq6q+c1xr1oYNoF1ga9W1QuB/wAWrpckSZq0JkuH85fAxgBVdT8wvelsDgM7N+ERYMWufa5p/t4KbN21fteqegA693ACxzXrtwA+ngRgVeDrSdYF3pPkL8DjgXuWUOfvqupnXeddc+xvUZIkaWKaLB3Oi4D9kmzRtW5F4Dbg3KqaXlXTgd26tg93/R0YwzluBw5tjrMd8CngEODaqjqOTsdypPlA972kCxZxfkmSpEltUgTOqvoLncvX701yTZIrgM2A64EnJPl+ku8Aey7DaY4HzkxyFXAesBbwXeD4JBfR+YLQSPcA6yS5bBnOK0mSNKENDA/baOtXV315yMmRNGab77rqkgcto8HBaQwNzW39PFo856E/OA8PNzg4bdQrxpOiwylJkqT+ZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1KrJ8lvqk9JOrxj0gbJ9wAf79gfnQZImLjuckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVvkt9T722W/e1+sSBMDfel2AgF7Nw74vXK0n55WkycQOpyRJklpl4JQkSVKrDJySJElqlYFTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWtV3gTPJcJLdR6xbLclvk0x/BMdbP0mWMGZO83d6khNHbDsnyYZLec5Dk7yxWd4yyRoj10uSJE0VfRc4gduBN49Y9xrgt4/weDOBGctU0bI5Gli/h+eXJEnqqX78LfX7gd8keVZV3ZRkRWA/4LKFA5K8BHgbncB8RlWdleR9wKOBZwJPBt5A58eXjwNWSPLUqnp7knOA9YDVgIOr6hdjLSzJ64FDmvOeUFXfSbIH8FbgMcBlVfWervGH0Qm8myX5RLN6iyQXAhsD76+qry7l5yNJkjSh9GOHE+CjdDqDAPsDFwDzAZrL00cBOwHbA69Msmozdl5V7U4nFB5dVT8CTgQ+VlVvb8YcWVUzgNOAly/i3Acnmb3wn+Y8NJfldwV2oNMxfUcz/odVtTOwHbBfkv/5TKvq88ClwKuq6ivN6rWrai9gR+CYR/DZSJIkTSj92OGkqirJqkmeCLwO2Ad4d7P56cAmwBXN63XodCwBrmn+3gqsNfK4SdYF3pPkL8DjgXsWcfqvVNVxXfuc0yxu2fwzq3m9XpKVgBcn2Rx4CFgdeNQS3t7VzXu8L8mCJYyVJEma8Pq1wwnwCeCLwJyq+n9d6+8EbgZmVNV0YOuq+lWzbbjr70CzPB9YpVk+BLi2CZT/sZT13AZcXVXTm/M+p6rm0emYvhX4t67zdJs/Yn13yBxGkiRpkuvbwFlV1wPzgE+OWD8EfAu4PsnlwDuXcKjrgSOSnAB8Fzg+yUXABktZz03AXUmuT3IZcFizaU6SG+ncBnDXInb9DnBOkv2X5nySJEmTxcDwsE22fvXZb97n5Eg9tu8LV+t1CX1lcHAaQ0Nze13GlOc89Afn4eEGB6cNjLatbzuckiRJmhwMnJIkSWqVgVOSJEmtMnBKkiSpVQZOSZIktcrAKUmSpFYZOCVJktQqA6ckSZJa1Ze/pa6O1790XR8o2wd8sG9/cB4kaeKywylJkqRWGTglSZLUKgOnJEmSWmXglCRJUqsMnJIkSWqV31LvYweed1uvSxAA/93rAgQs73n41A4bLNfjSZJGZ4dTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWmXglCRJUqsMnJIkSWqVgVOSJEmt6mngTPLnJLOT3JDkLc26Q5O88REe71lJXrmU+8xOsmqSDZOcM2LbiUmmL+Xxpic5sVl+WpInjlwvSZI0lfT6py1vqarpSVYELk3y7WU5WFXdBNy0PApbTg4B5gB397oQSZKkXul14ASgquYn+Qmw8MeNt0hyIbAx8H7gauALVbULQJITgO8D2wC70enUvgx4KjCzqo5LshXwETrv8caqOjbJKcAWwGOBI6rqhrHWmOQlwNuac51RVWcleS7wIWA14Laqem3X+BcDhwIvTbIZ8GPgiUnOA9K8n48s7WclSZI00fTFPZxJ1gG2Bn7arFq7qvYCdgSOqap7gN8l2TjJSsD2wGXAS4EXVNW2wK9HHPZ04NCqmg68vVl3QlXNAI4BDl9EKTs1l9hnJ5kNHNzUtwZwFLBTc+5XJlkVuJNO4H0h8JQkT1h4oKq6GDgbeHtVfaxZvRFwIPDcUc4vSZI06fS6w7lZE+z+Ary1quYmgU5Hk6q6L8mCZuwngdfS6RSeV1XDSd4MnJrk58BpCw/aBNh7q+ru5jgLkqwGHJ/kQeDRwLRF1HNVVR3UdZyF91w+HdgEuKJ5vQ6wHvAMYPem/rVGOWa3a6tqHjAvyR+WMFaSJGlS6HWH85aqml5Ve1bVj7rWL+haHgaoquvoBLxX0OkcQudS+ZHAE4EXd+1zP7BRkrUBkqwM7AHcV1XHAbOXss47gZuBGU3HdOuq+hXwXuBo4N0L6xxhPrDK4t6XJEnSZNfrwLm0Lgd+3XRCVwCuTDKLzn2ZcxYOqqoFdILgRc32dzbb909yGbDl0py0qoaAbwHXJ7m8OR7A+XQ6rmcDv1nErlcBH0nyT0tzPkmSpMlkYHh44jTami8SHVNVt/e6lvFw4Hm3TZzJkSaYT+2wwZIH6R8MDk5jaGhur8uY8pyH/uA8PNzg4LSB0bb1+h7OMUmyPnAB8PWpEjYlSZImiwkROKvqXjqPQJIkSdIEM9Hu4ZQkSdIEY+CUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1akI8Fmmq+tp+T/eBsn3AB/v2B+dBkiYuO5ySJElqlYFTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRW+S31PrbvebN6XYL0Pz67w3N7XYIkaYKywylJkqRWGTglSZLUKgOnJEmSWmXglCRJUqsMnJIkSWqVgVOSJEmtMnBKkiSpVQZOSZIktcrAKUmSpFb5S0NLKcmfgR8DqwJzquqo3lYkSZLU3+xwLr1bqmp6VW0LbJBky14XJEmS1M/scD5CSVYF1gHua16fAmwBPBY4oqpuSLIV8BE6n/ONVXVsktcDh9AJ+ydU1Xd68gYkSZLGiYFz6W2W5AfAE4CXV9V/N+tPqKqhJDsChwM3AKcDL62qu5OskCTArsAOwMrA5YCBU5IkTWpeUl96t1TVNsCHgJcAJFkNOD7JicD+wLQk6wD3VtXdAFW1ANiy+WcWnbC5XhJDvyRJmtQMnI9QVf1f4NlJNgf2AO6rquOA2c2Q+4GNkqwNkGRl4Dbg6uYe0OnAc6pq3rgXL0mSNI4MnMvmaOAUYA6wf5LL6HQwF3Y0jwYuSjILeGdV3QTcleT6ZuxhvSlbkiRp/AwMDw/3ugaNYt/zZjk56huf3eG5PT3/4OA0hobm9rQGOQ/9wnnoD87Dww0OThsYbZsdTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWmXglCRJUqsMnJIkSWqVgVOSJEmtMnBKkiSpVf6Odx87f78ZPlC2D/hgX0mSlo0dTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWmXglCRJUqv8lnof2+sb5/W6BLXozB137XUJkiSNCzuckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklpl4JQkSVKrDJySJElq1ZQLnEn+nGR28893k7y4a9vJi9lvwyTnLMV5npbkic3ys5K8ctkqlyRJmpim4k9b3lJV0wGSrAeck+RPVfX9qjp2OZ7nEGAOcHdV3QTctByPLUmSNGFMxcD5P6rqt0mOBd4AfD/JnKraNslGwKeB1YG5wL7NLo9J8mXgKcC9wMFV9VCS9wEzgAHgaGB94FDgpUk2A34MzKyq45JsBXyEzmd/43IOuZIkSX1nyl1SX4RfAk8ese73wN5VtSNwH7B1s/4ZwBFVtT3wn8DLk7wIWKMZuw/wr1V1MXA28Paq+tiIY58OHNp0Wd++3N+NJElSn5nSHc7G5sAdI9ZtCrw6yVxgI2Bas/6Gqvpzs/wD4LnAesDOSWY361cc7URJ1gHuraq7AapqwXJ5B5IkSX1sSnc4kzwN+ADw8RGb3g2cUFXH0bmkvtDmSVZtll9M577M24Bzq2p607Xcrdk+H1hlxHHvBzZKsnZz/pWX01uRJEnqW1Oxw7lZkll0AuF9dC5v/9eIMV8HrkxyC/CnrvW/Ab6Y5PF07r+8MMkAMDPJ9+mE07OAc4GrgLOab6r/DDodzSRHAxcleQCYBby/tXcqSZLUBwaGh4d7XYNGsdc3znNyJrEzd9y11yVMKIOD0xgamrvkgWqV89AfnIf+4Dw83ODgtIHRtk3pS+qSJElqn4FTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWmXglCRJUqum4i8NTRgX7r+fD5TtAz7YV5KkZWOHU5IkSa0ycEqSJKlVYw6cSdZJslWbxUiSJGnyGVPgTHIY8E3g7Ob1lkk+1mJdkiRJmiTG2uF8DbAjcD9AVf0H8Oy2ipIkSdLkMdZvqT9YVcNJhgGSrABMa68sAex53ud7XYKAi/Y7rNclSJI0oY21w3lBkpOBxyXZCzgfuLy9siRJkjRZjKnDWVWnJtkRGAZmAGdW1QWtViZJkqRJYUyBM8kVVbULcHXL9UiSJGmSGesl9Z8l2abVSiRJkjQpjfVLQy8Gjkjye+AvwAAwXFVPb60ySZIkTQpjvYdzk7YLkSRJ0uQ01ns4d1jU+qq6ZvmWI0mSpMlmrJfUD+laXoXOQ+B/CBg4JUmStFhjvaR+ePfrJI8DfCp5I8mKwEnAVsDqwBVV9Z7eViVJktQfxtrhfJiq+lOSJy3vYiawmcD8qtoZIMkqPa5HkiSpb4z1Hs5/ofPQd4CVgecDP2urqAnoTuDIJINVNVRVDyZ5PvAhOo+euryqTkhyPnA88EvgYmDvqprbs6olSZLGwVg7nHd3Lc+jc8n4uhbqmZCq6pYkbwNOS3IrcAJwMrB7Vf05yTlJngIcCXwM+DlwgmFTkiRNBWMNnKtW1endK5K8rqrOaKGmCamqfgrsn2Qm8C3g6cC3kwCsATyxqq5N8gvg2d7jKUmSporFBs4ka9L5EsyrknybzgPfAR4LvBUwcAJJ1gfur6qHgO8B76XTxdy1qh5KsnpV/TXJE4CNgLuSbFNVP+hh2ZIkSeNiSR3OlwMHAJsBX+Z/A+ffgPe3WNdE8wzg5CR/BuYD7wFWBa5JMhe4M8kbgdOBtwD30+l+7lZVf+1V0ZIkSeNhsYGzqj4NfDrJyVV17DjVNOFU1ZV0Hok00oUjXu/ZtfzC9iqSJEnqH2N9DuexSTYGNuB/u5z+0pAkSZKWaKyPRfq/wDPp3H94JZ1fGvoJ/tKQJEmSlmCFMY7bqqp2BGYBxwDP4n+fyylJkiSNaqyBc0Hz92bgeVX1Z8BfGpIkSdISjTVwnppkUzrfVD8pyY+BH7ZXliRJkiaLsX5p6NyFy0meC6xRVb9vrSpJkiRNGmPqcCbZOMn5Sb5bVfOBdZPs0nJtkiRJmgTGekn9NOBfgBWb17cD726lIkmSJE0qYw2cw1V1y8IXVTWvpXokSZI0yYzpHk5gKMmWNI9CSnIA8EBrVQmAi/Y7jKGhub0uQ5IkaZmM2uFM8oGul+cDRwKbJLmDzm+sH9ZybZIkSZoEFtfh3CXJ+cBvgaOBA4D3Ntt86LskSZLGZHGB85+A44BBYDPgS13bBuiEzp3aK02SJEmTwaiBs6p+BLwMIMlJVfWOcatKkiRJk8aYvqVu2JQkSdIjNdZvqasH9jj/pF6XMOV9Yfs39boESZImvLE+h1OSJEl6RAyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklrlg98XIcmKwEnAVsDqwBXAVVU1eymOsSFwYlUd1EaNkiRJE4WBc9FmAvOrameAJKsAVwPb9rQqSZKkCcjAuWh3AkcmGayqIeBkYLMks4E3AW8HfgnsDmwHHA/sSucWhR9W1VELD5RkADgd+CFwJvApYFNgAfC6qrpzfN6SJElSb3gP5yJU1S3A24DTknwAOBa4paqmN9sA7qmqbYCdgA2BHarqBcDKSfbqOtzJwI+q6nPAazqHrxnAkcC7xucdSZIk9Y4dzlFU1U+B/ZPMBM5exJDrmr/PAi6uquHm9XfpdDB/SucS/B1V9dZm27OB5yXZp3k9tPwrlyRJ6i92OBchyfpJHtW8/B6dDubIcD6v+fszYLeu9TsBP2mW5wBXJflE8/o24NSmUzodePXyrVySJKn/GDgX7RnAD5JcDVwIvAe4JskNSdI9sKouAX6X5PrmHs/fVtV3u7Z/EPhTklOAzwJ7Jrk6ySXANuP0fiRJknpmYHh4eMmj1BN7nH+Sk9NjX9j+TQwOTmNoaG6vS5nynIf+4Dz0B+ehPzgPDzc4OG1gtG12OCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1KqRvw+uPnLJvu/wFwwkSdKEZ4dTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKr+l3sd2v+DlvS5hUvjidp/tdQmSJE1pdjglSZLUKgOnJEmSWmXglCRJUqsMnJIkSWqVgVOSJEmtMnBKkiSpVQZOSZIktcrAKUmSpFaNe+BM8ucks5PckOQtzbpDk7zxER7vWUleuZT7zE6yatfrtya5agz7bZlkjWZ5tyS7jDJutSQfaJYfl+RZS1OfJEnSZNKLXxq6paqmJ1kRuDTJt5flYFV1E3DTMta0O/CbJJtW1c8XM+5o4ETgj1V12WJq+hvw7ublVsDM5VCjJEnShNSzn7asqvlJfgJs0KzaIsmFwMbA+4GrgS9U1S4ASU4Avg9sA+xGpzv7MuCpwMyqOi7JVsBH6LyvG6vq2CSnAFsAjwWOqKobuutIsj3wA+Ay4PXAMc36jYBPAo8B7gZm0QmOmyX5BPAoYFVgbeAXVfXVJCsD3wOeD1wP7Ad8AlgryeOBvwDnVtXsJI8Fzq+qnZfH5ylJktSvenYPZ5J1gK2Bnzar1q6qvYAdgWOq6h7gd0k2TrISsD2dUPhS4AVVtS3w6xGHPR04tKqmA29v1p1QVTPoBMnDF1HKYcDnq+oaYJskqzTrPw28qznWq6rq88ClzfJXuvY/Czi4Wd4TOK+qhgGq6jfAUcBXqupVwKnAa5uxrwLOXPInJUmSNLH1InBulmQ2cDbw1qqa26y/GqCq7gMWNOs+SSeg7cP/Brk3A6cm+SdgYOFBmwB7b1Xd3RxnQZLVgOOTnAjsD0zrLqTpMs4APpbkW8Dj6ARagDWay/VU1QJG0QTj+UnWBV7OYkJkc7n+sc19oHsDXx9trCRJ0mTRi8B5S1VNr6o9q+pHXeu7Q93CDuF1wDOAV9AJqNC5VH4k8ETgxV373A9slGRtgOby9h7AfVV1HDB7EbW8AnhvVe1TVfsAO/O/HcgFSTbuOhbAfGCVfzgKfAE4EphbVb8fsW3kPmfQ6XReW1UPLeJYkiRJk8pEeCzS5cCvq2pukhWAK5PMonNf5pyFg5ou5NHARc32dzbb909yGbDlIo79auCCrmP8FngoySZ0OqlnNt3YU5oh3wHOSbL/iONcRCe8fm4R5/gpsHOShZ3Pi4HpwGfG9vYlSZImtoHh4eFe17BYzReJjqmq23tdy/KQZFs6X1569ZLG7n7By/t7ciaIL2732WXaf3BwGkNDc5c8UK1yHvqD89AfnIf+4Dw83ODgtIHRtvXsW+pLkmR9Ot3Hr0+isPlOOo9gOqjXtUiSJI2Xvg2cVXUvnUcgTRpV9SHgQ72uQ5IkaTxNhHs4JUmSNIEZOCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVffscTsF39v6qv2AgSZImPDuckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVvkt9T52wtd263UJE9obdvpGr0uQJEnY4ZQkSVLLDJySJElqlYFTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWjXlA2eS/ZLckWTUzyLJ9K7ls5NsOi7FSZIkTQJTPnACrwQuBnZdzJgTx6kWSZKkSWdK/7RlkicDc4GPAScDlyY5HHhNM+QY4BXAZklmA29q1r88yXbAesDBVfWfzfYLgT2BW4GbgZcBjwb2qqr7krwT2AVYA3hvVV3Y+puUJEnqsane4XwtcFZV/QpYPcn6wGHAi6pqO+CGqjoSuKWqplfVLc1+v62qXYD3NeMXurGqZgDrAmtW1U7AN4CXNtvPaNbtC/xz229OkiSpH0zZDmdzz+YBwFZJ3gIM0ulsHg58MMm9wCeABxax++zm763AS7rW39D8vQOY0yzfCWzcnO+oJCsBfwemLbc3I0mS1MemcodzN+Dcqtq7qvYBtqfTebyjqo4C/kAnfMI/BvMFzd/hEeuHR1kG2ApYp6reAZy/bKVLkiRNHFM5cB4OfH3hi6p6CLgR+F1zP+YBwKXN5muS3JAky3C+nwObJpkFvGgZjiNJkjShDAwPj2zEqV+c8LXdnJxl8IadvrFcjjM4OI2hobnL5Vh65JyH/uA89AfnoT84Dw83ODhtYLRtU7nDKUmSpHFg4JQkSVKrDJySJElqlYFTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWjXyN8LVR9514GX+goEkSZrw7HBKkiSpVQZOSZIktcrAKUmSpFYZOCVJktQqA6ckSZJa5bfU+9iZX9i11yVMOHvtcV6vS5AkSSPY4ZQkSVKrDJySJElqlYFTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWjUlHvyeZEXgJGArYHXgiqp6zzIcb05VbTti3QnACVX1wDIVK0mSNMlMicAJzATmV9XOAElWWd4nqKp3Le9jSpIkTQZTJXDeCRyZZLCqhqrqwSSzgQuBPYFbgZuBlwGPBvaqqvuSvAR4G7AA+CPw2qr6/cKDJnkVsBPwWuAqOsF2W+B1dDqpmwBnVNUpSdYAzgLWAH4OPK+qntvy+5YkSeq5KXEPZ1XdQic4npbkA10dzhuragawLrBmVe0EfAN4aRMQ3wnsVlU7Al8Cjl94zCQvpQmbVbVgxCmfAhwAPBd4Y7PubcA3mvN9EFhn+b9TSZKk/jMlAidAVf20qvYHrgXOblbf0Py9A5jTLN9Jpwu5CfDDqvprs/67wKbN8mrAh4G3LCJsAlxXVfOr6kHgz826rYBLmlp+Ddy3HN6WJElS35sSgTPJ+kke1bz8HrBhszzcNWyYh/sFsHWS1ZrXOwE/aZb/BrwB+EaSxyzilIs67l3AC5t6NgUev5RvQ5IkaUKaEoETeAbwgyRX07lvc4nfUG/u1fwoMCvJlcDewIld269stn87ybQx1PBvwLFJrgEOA3691O9CkiRpAhoYHh7Z2FMbkqxcVX9vlp8DvKuq9l3cPmd+YVcnZynttcd5y/2Yg4PTGBqau9yPq6XjPPQH56E/OA/9wXl4uMHBaQOjbZsq31LvBzOSHE/nEvuDwJt6XI8kSdK4MHCOk6q6HLi813VIkiSNt6lyD6ckSZJ6xMApSZKkVhk4JUmS1CoDpyRJklpl4JQkSVKrDJySJElqlYFTkiRJrfI5nH3sta++3F8wkCRJE54dTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWmXglCRJUqv8lnofu+Tze/S6hL72vJd8rdclSJKkMbDDKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1akI++D3Jn4EfNy/nAR+vqoubbSdX1bFJBoCvAxsAewAnAM8F/ggcWVV3NOO/CZxXVV9uXn8GOLeqrlrEeWcDM4FtgZlVdVzXticAB1TVJ5b7G5YkSZrAJmqH85aqml5V04FXAMcm2R6gqo5txjwRWKWqXlBVfwJeWFXPBy6gE0BJ8ijgKcDuXcfeFvje0hZUVb8xbEqSJP2jCdnh7FZVv01yLPAG4PtJ5lTVtnS6m09N8mVgxWb5SuDVwOeAU4EdgfOAfZKsADwN+EVV/T3JKcAWwGOBI6rqhpHnbgLruc3xfgacWFUHJTkb+BWwHTAIHFhVlWQG8EHgT832P3R3SSVJkiajidrhHOmXwJNHrDsIuKqqXlFVB9Hpiu5cVXcDj0uyOvBi4BLgh8BzgF2Bi5v9T6iqGcAxwOGLOOeKwFnAGQsv549wT1XtArwfeH2z7mRgz6qaCdzxyN6qJEnSxDJZAufmLF2AuwLYCXhWVd1EJ2TuTidwfifJasDxSU4E9gemLeIYBwAPVNVFo5zjmubvrcBaSdYF7q6q3zfrf7QU9UqSJE1YEz5wJnka8AHg40ux24XAEcDtzetZdC6vr1FV99C5x/O+5nL37FGO8TVgXpKjR9k+3PV3ALgf2DjJY5r1Oy1FvZIkSRPWRA2cmyWZleS7dMLmoVX1X0ux/4+ALelcTqeq/gb8Dfh+s30OsH+Sy5pxizIMvBHYPMlbl3TCqppH55vy1zbHfQzwwFLULEmSNCENDA8PL3mUloskK1fV35vlU4BZVfWt0cZf8vk9nJzFeN5LvjYu5xkcnMbQ0NxxOZdG5zz0B+ehPzgP/cF5eLjBwWkDo22b8N9Sn2BOSrIVsDLwAzqPaJIkSZrUDJzjqKqO6XUNkiRJ422i3sMpSZKkCcLAKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa3yOZx9bI/DLvEXDCRJ0oRnh1OSJEmtMnBKkiSpVQZOSZIktcrAKUmSpFYZOCVJktQqv6Xex37ymb16XULfeuJ+X+l1CZIkaYzscEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklpl4JQkSVKrDJySJElqlYFTkiRJrZrQD35PsiJwErAVsDpwBbADMLOqHugadwxwTlXds5hjPRX4MPAEYD4wu6re1WL5kiRJU8KEDpzATGB+Ve0MkGQVOoHzYarqY4s7SJK1gHOA11XVzV3HkiRJ0jKa6IHzTuDIJINVNVRVDyYBIMmLgKOAA4DTgBOBB4BPA38ANgWuqqq3Aa8FPrMwbAJU1YPNcd4J7AKsAby3qi5McjbwK2A7YBA4sKoqyVbAR+h8rjdW1bFJXg8cQuf2hROq6jvN/r8Edge2q6r57Xw8kiRJvTeh7+GsqluAtwGnJflAV1dyO+BI4ICq+tuI3QIcDjwXeFGSxwKbADeNcpozqmonYF/gn7vW31NVuwDvB17frDsdOLSqpgNvTyf97kqn6zoDeMeI/bcxbEqSpMluonc4qaqfAvsnmQmc3aw+HdhtEWETOp3HvwIkuQ1YE7gLeBrw4+6BSVYAjkqyEvB3YFrX5muav7cCeyVZB7i3qu5u6lqQZEtgS2BWM3a95lgA1z3CtyxJkjShTOgOZ5L1kzyqefk9YMNm+WDgi0k2WMRuwyOWB4B/B45L8pSuYz+azpeR1qmqdwDnj3Kchce4H9goydrN/isDtwFXV9X0puv5nKqa1+w3D0mSpClgQgdO4BnAD5JcDVwIvKdZ/1M6l9TPT/L4JR2kqu4C3gSckeTqJFcBbwB+DmyaZBbwoiUcYwFwNHBRM/6dVXUTcFeS65NcBhz2SN6kJEnSRDYwPDy85FHqiZ98Zi8nZxRP3O8r43auwcFpDA3NHbfzadGch/7gPPQH56E/OA8PNzg4bWC0bRO9wylJkqQ+Z+CUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVhk4JUmS1CoDpyRJklq1Uq8L0Oi2euOF/oKBJEma8OxwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJEmSWuW31PvYf3/6X3pdwrhZ6YDjel2CJElqiR1OSZIktcrAKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVPX/we5I5VbVtkhWBrwGfA/4TOKCqPjFi7IbAiVV10HI476HA8cA9zapfAe+qql8neRbwzKr60rKeR5IkaarreeDs8kng61V1WfP6E+Nwzo9V1WcAkswAzkuyfVXdBNw0DueXJEma9PoicCZ5F3BzVX2teb0hTSczyTOA/ws8BNzWtc/ZdLqS2wGDwIFVVUmeD3yIzu0ClwMfBa4Btq6q4SSvo/O+H+iuoapmJfkB8NwkjwJmVtVxSU4DtmiOtwOwCvBZYH3gr8Crqur+JKc04x4LHFFVN4x13+X1OUqSJPWjfriHcwNgP+Dzo2w/BTisqnYFrh6x7Z6q2gV4P/D6JAPAycBLqmoH4JnAusCVwE7NPgcC/z7KuX4JPHnhiyRrAptV1QuA7arq78BxwLlVtRPwaeCfmuEnVNUM4Bjg8KXcV5IkadLqh8D533Qun5/ZBMaRHl1V/9Us3zhi2zXN31uBteh0Op8OfDvJbGBT4Il0OqSvTvIcOp3U/zdKLZsDdyx8UVV/AD6a5FPAwc3qZwNvbY5/PLBWktWA45OcCOwPTBvrvqN9KJIkSZNFPwROquoLdC6Xf3wRm1dKskGzPGPEtuGuvwPA74CfA7tW1XQ6ncVrq+rXdN7r2+iEz3+Q5CBgjaq6sWvdysAlVfVmYM8kmzd1Hl9V06tqezrBcQ/gvqo6Dpi9lPtKkiRNan0ROAGq6gPAY5KMDGHHARcn+S7wpCUcYwHwYeCaJFfw8C8efRVYvap+0bXumCRXNcd+Gp3L7d3WBq5NchWwInA78EHgnUlmJbmw2W8OsH+Sy4Atl3JfSZKkSW1geHh4yaMmgeYLPOdW1axe1zJW//3pf5kakwOsdMBxvS5hVIOD0xgamtvrMqY856E/OA/9wXnoD87Dww0OTlvUrZFAH3U425TkOuDBiRQ2JUmSJou+eCxS26pqu17XIEmSNFVNiQ6nJEmSesfAKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa2aEs/hnKg2eNO/+QsGkiRpwrPDKUmSpFYZOCVJktQqA6ckSZJaZeCUJElSqwyckiRJapXfUu9jQ6d9qdcltGv/vXtdgSRJGgd2OCVJktQqA6ckSZJaZeCUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVo3Lg9+T7AecBDy9qhYkeR8wp6ouXYpjTAdmVtVxYxx/KLBqVX0myWw64XoBsCLwTeDUqpo/Yp/VgOOr6t2LOe7ZwIlV9fOuda8Cbqqqm8f6fiRJkqaK8epwvhK4GNh1nM63KLtW1XTgRcBTgLePHFBVf1tc2BxNVX3RsClJkrRorXc4kzwZmAt8DDgZuHTE9pcD/0Sn+/gZ4GvAKcAzgVWBb1bVh5vhT0xyHhDgC1X1kSTTgNOAJwCrA5+qqn8frZ6qejDJscC1wIeajuUvgd2B7YBrq2rbpgv76KaOJwNvqKrvd9W9RVPn/sCRNB3bJDcCNwLPAu6sqpc34z8MbA/8tjnuP3d3SSVJkiar8ehwvhY4q6p+BayeZP2FG5I8FTgMeFFV7QCc04y/r+lGbgds34Q7gI2AA4HnAoc3644DLq+qGcAOwJuSrLO4gqpqHg9/7/dU1TYjL7ED86pqd+AQ4Oiu9U8HPgocUFW/H7HPxsB7qmpb4NFJNk+yC7BmVW3X1D+4uPokSZImk1YDZ5IVgAOAo5J8i07Qek3XkGcDl1TVAwBVtYBOZ/Cirtez6HQ0odN9nNeM/0Ozrnv8g8ANdILp4up6NJ2O6kLXjTL0mubvrcBaXetPBd5fVb9bxD5VVfeN2G8r4JJm40PATxdXnyRJ0mTSdodzN+Dcqtq7qvahc0l5X2Cg2X47sHOSlQCSrAz8DJjZvF4BeCGw8P7I7pA43PztHv8oYMvmuIuU5LHA5+hc3l9o3ijDh7v+DnStfx1wYpL/s5h9uve7q3kfJFkd2Ha0+iRJkiabtgPn4cDXF75ouns30vniDlX1H8CVwPVJrgT2oBMGN0ryPTrdzYurqhZzjg8C+ya5GrgcOLmq/riIcZc35zgHOKOqvr6IMWN1N3AwcHaSzcYw/hvA+kmuB84CfgE8sAznlyRJmjAGhoeHlzxKyyTJisCCqhpO8jg6Qfp5i7hn9GGGTvvS5J6c/ffudQVjMjg4jaGhub0uY8pzHvqD89AfnIf+4Dw83ODgtIHRto3LczjFusCXmlsEVgaOW1LYlCRJmiwMnOOgqv4b2LnXdUiSJPWCP20pSZKkVhk4JUmS1CoDpyRJklpl4JQkSVKrDJySJElqlYFTkiRJrTJwSpIkqVU+h7OPDR7xSn/BQJIkTXh2OCVJktQqA6ckSZJaNTA8PNzrGiRJkjSJ2eGUJElSqwyckiRJapWBU5IkSa0ycEqSJKlVBk5JkiS1ysApSZKkVvlLQ30qyQeAHejM0eur6mc9LmnSSrIG8BlgfTr/J+zVwKOATwOrAtdV1duasc7LOEjyY+B44E6ch55IsjVwMrAicEHzj3MxzpIcA+xN5/N9M/BXnIdxkWQQOApYUFXvThLG+NmPNnYqs8PZh5K8EFivqnYE3gB8pMclTXarA8dU1XTgJOBY4BPAYVX1AmDDJNs4L+Mjyf7A45qXn8B5GHdJVgbeA+xdVS+oqg/jXIy75v8MvwSYTuf/CL8f52E8fRR4EFi5ef0Jxv7Z/8PYca28Dxk4+9OuwFcBquo/gbV6W87kVlX3VNU9zcs/0PkXzKpV9ctm3XnA83FeWpdkGnAI8GU63QLnoTd2B34FfDXJlU2307kYf/Pp/Hf6UcA6wBDOw7ipqlcB1wAkGfO/jxYzdkozcPanden8i2WheUmcq5YleQKd7uZHgd93bfo9sCbOy3g4FTgBWABMw3nolU3ohJY9gcOAr+FcjLuqmksn8NwKfBs4C+ehVwYZ42cPrDfK2CnN/0H2pz/x8P9xLqiqBb0qZipIsiedS4iHA/cDa3RtXpPOv1CclxYleQVwV1X9sFn1R5yHXpkHXF5V85ouzf08/DN3LsZBkhfTuZz7NGBTOpfUnYfe+CNj/PcRo/83ZEozcPan7wH7AyTZDLi7t+VMbkm2APaqqjdU1e+r6m/AKk3HE+ClwJU4L207GNgsyTl0Pud3AM9wHnriejqX1UmyHjAXeJRzMe6eAvy2qoaBP9Pp+q/lPIy/pfnvwmLGTml+S70/XQzskeR7dP5F/4Ye1zPZzQRemGR28/ou4BjgG0keBL5dVbcmKZyX1lTVixcuJ3kfMIfOpSjnYZxV1Q1JKsm1dLqdx9BpUDgX4+ts4MwkVwOrAKcDN+E89MrS/HfhH8b2puT+MTA8PNzrGiRJkjSJeUldkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRWGTglSZLUKgOnJPWBJI9LclKv65CkNvhYJEmaopJsCWxYVRf0uhZJk5sdTkmaYpIMNItbAZv3shZJU4MdTknqA0k2BM4Avk/nd5g3BjYBjgQOAv4PcB/w0qqan+QO4EvADGB14MiqmpNkVeBk4Jl0fp3m4qo6oTnHHXR+yeyZwInAJ5sxN1fV3kneArwMeBRwflV9MMl04Ajg70CA24FXVNVwkq2BD9H5ve+7q+rg5qdiP9wcYwh4TVX9tZ1PTdJEYYdTkvrP46tqT+DlwLeBM6rq+cB8OgETYEPgB1W1I/AK4HPN+ncA91bVdOAFwFZJ9mi2PRn4WlXtXFVX0Amdn6+qvZvtV1bVC4BtgdckWb1Zvz3wlqp6HrA2sH2SxwKfBw6tqh2A1yVZCfgIcFBV7QRcA7x+uX4ykiYkA6ck9Z9Lm783Ab+vquua1zcDT2qW76uq7wBU1W3AX5KsDbyYTueSqloAnAXs3Oxzf9exFuXBJP8CfBFYCxhs1s+qqt83y3PohN3tgKuq6tfNuf4KbApsAXwryWzglcB6S/vmJU0+K/W6AEnSP3gIOoExyd+61s8DVuwe02V14G90/r2+YMS2ha//MtoJm7D6deCfgU8BFwAL7/V8oGvo35saVm2Wu60IzKmqfUc7j6SpyQ6nJE1MT0ryPIAkzwf+2HQZLwfe3KxfAXgVcMkox3gAmNYsbwT8qqquadZttYTzzwH2SLJOc67HAgVsnuSpzbrBJI9/hO9P0iRi4JSkieku4JVJrgJOAA5r1n8A2CTJ94FZwOyqmjXKMa6iExovoHP5fn6S64AP0rl8P6qquhd4G3BRkquBE6vqAeB1wNeSXAN8hc6XhyRNcX5LXZImoCR3VNXGva5DksbCDqckSZJaZeCUJElSq7ykLkmSpFbZ4ZQkSVKrDJySJElqlYFTkiRJrTJwSpIkqVUGTkmSJLXKwClJkqRW/X+AZwUEkKiIYQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "imp = pd.DataFrame({'feature': df.drop('HeartDisease', axis=1).columns, 'importance': clf.feature_importances_})\n",
    "imp = imp.sort_values('importance').set_index('feature')\n",
    "\n",
    "plt.figure(figsize=(10,10))\n",
    "sns.barplot(y=imp.index, x=imp['importance'], orient='h').invert_yaxis()\n",
    "# plt.savefig('importance.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "878881eb",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "97cb7531",
   "metadata": {},
   "source": [
    "## 한계점 및 피드백\n",
    "\n",
    "- 오버샘플링과정에서 발생하는 재현율 감소를 해결하지 못했다.\n",
    "\n",
    "=> 적절한 샘플링 수준을 찾으면 보완할 가능성이 존재한다.\n",
    "\n",
    "- 추가적인 모델링과정이 필요해보인다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "37cb242e",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}