WNS2019.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime as dt\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from dateutil.parser import parse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "train=pd.read_csv(\"./WNS//train.csv\",parse_dates=['impression_time'])\n",
    "test=pd.read_csv(\"./WNS//test.csv\",parse_dates=['impression_time'])\n",
    "item=pd.read_csv(\"./WNS/item_data.csv\")\n",
    "view=pd.read_csv(\"./WNS/view_log.csv\",parse_dates=['server_time'])\n",
    "submission=pd.read_csv(\"./WNS/sample_submission.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 237609 entries, 0 to 237608\n",
      "Data columns (total 7 columns):\n",
      "impression_id      237609 non-null object\n",
      "impression_time    237609 non-null datetime64[ns]\n",
      "user_id            237609 non-null int64\n",
      "app_code           237609 non-null int64\n",
      "os_version         237609 non-null object\n",
      "is_4G              237609 non-null int64\n",
      "is_click           237609 non-null int64\n",
      "dtypes: datetime64[ns](1), int64(4), object(2)\n",
      "memory usage: 12.7+ MB\n",
      "None \n",
      "\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 90675 entries, 0 to 90674\n",
      "Data columns (total 6 columns):\n",
      "impression_id      90675 non-null object\n",
      "impression_time    90675 non-null datetime64[ns]\n",
      "user_id            90675 non-null int64\n",
      "app_code           90675 non-null int64\n",
      "os_version         90675 non-null object\n",
      "is_4G              90675 non-null int64\n",
      "dtypes: datetime64[ns](1), int64(3), object(2)\n",
      "memory usage: 4.2+ MB\n",
      "None \n",
      "\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 132761 entries, 0 to 132760\n",
      "Data columns (total 6 columns):\n",
      "item_id         132761 non-null int64\n",
      "item_price      132761 non-null int64\n",
      "category_1      132761 non-null int64\n",
      "category_2      132761 non-null int64\n",
      "category_3      132761 non-null int64\n",
      "product_type    132761 non-null int64\n",
      "dtypes: int64(6)\n",
      "memory usage: 6.1 MB\n",
      "None \n",
      "\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 3118622 entries, 0 to 3118621\n",
      "Data columns (total 5 columns):\n",
      "server_time    datetime64[ns]\n",
      "device_type    object\n",
      "session_id     int64\n",
      "user_id        int64\n",
      "item_id        int64\n",
      "dtypes: datetime64[ns](1), int64(3), object(1)\n",
      "memory usage: 119.0+ MB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "print(train.info(),\"\\n\")\n",
    "print(test.info(),\"\\n\")\n",
    "print(item.info(),\"\\n\")\n",
    "print(view.info())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train data null check\n",
      " impression_id      0\n",
      "impression_time    0\n",
      "user_id            0\n",
      "app_code           0\n",
      "os_version         0\n",
      "is_4G              0\n",
      "is_click           0\n",
      "dtype: int64 \n",
      "\n",
      "\n",
      "test data null check\n",
      " impression_id      0\n",
      "impression_time    0\n",
      "user_id            0\n",
      "app_code           0\n",
      "os_version         0\n",
      "is_4G              0\n",
      "dtype: int64 \n",
      "\n",
      "\n",
      "item data null check\n",
      " item_id         0\n",
      "item_price      0\n",
      "category_1      0\n",
      "category_2      0\n",
      "category_3      0\n",
      "product_type    0\n",
      "dtype: int64 \n",
      "\n",
      "\n",
      "view data null check\n",
      " server_time    0\n",
      "device_type    0\n",
      "session_id     0\n",
      "user_id        0\n",
      "item_id        0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\"train data null check\\n\",train.isnull().sum(),\"\\n\\n\")\n",
    "print(\"test data null check\\n\",test.isnull().sum(),\"\\n\\n\")\n",
    "print(\"item data null check\\n\",item.isnull().sum(),\"\\n\\n\")\n",
    "print(\"view data null check\\n\",view.isnull().sum())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def unique_count(data,data_name):\n",
    "    print(\"Size of {} dataset\".format(data_name),len(data))\n",
    "    print(\"Checking unique count for each column in {} dataset\".format(data_name))\n",
    "    for name in data.columns:\n",
    "        print(\"Unique count\",name,len(data[str(name)].unique()))\n",
    "    print(\"\\n\")\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of train dataset 237609\n",
      "Checking unique count for each column in train dataset\n",
      "Unique count impression_id 237609\n",
      "Unique count impression_time 36461\n",
      "Unique count user_id 74723\n",
      "Unique count app_code 490\n",
      "Unique count os_version 3\n",
      "Unique count is_4G 2\n",
      "Unique count is_click 2\n",
      "\n",
      "\n",
      "Size of test dataset 90675\n",
      "Checking unique count for each column in test dataset\n",
      "Unique count impression_id 90675\n",
      "Unique count impression_time 9636\n",
      "Unique count user_id 34079\n",
      "Unique count app_code 373\n",
      "Unique count os_version 3\n",
      "Unique count is_4G 2\n",
      "\n",
      "\n",
      "Size of item dataset 132761\n",
      "Checking unique count for each column in item dataset\n",
      "Unique count item_id 132761\n",
      "Unique count item_price 13531\n",
      "Unique count category_1 17\n",
      "Unique count category_2 79\n",
      "Unique count category_3 335\n",
      "Unique count product_type 7959\n",
      "\n",
      "\n",
      "Size of view dataset 3118622\n",
      "Checking unique count for each column in view dataset\n",
      "Unique count server_time 82840\n",
      "Unique count device_type 3\n",
      "Unique count session_id 1014970\n",
      "Unique count user_id 89157\n",
      "Unique count item_id 126708\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "unique_count(train,'train')\n",
    "unique_count(test,'test')\n",
    "unique_count(item,'item')\n",
    "unique_count(view,'view')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['is_click']=0\n",
    "train['is_train']=1\n",
    "test['is_train']=0\n",
    "train_test=pd.concat([train,test],axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "#creating new features based on impression_time\n",
    "train_test['day']=train_test['impression_time'].apply(lambda x:x.day)\n",
    "train_test['month']=train_test['impression_time'].apply(lambda x:x.month)\n",
    "train_test['week_day']=train_test['impression_time'].apply(lambda x:x.weekday())\n",
    "train_test['time']=train_test['impression_time'].apply(lambda x:x.time().hour)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Counting no of site visits by user per session id\n",
    "site_visit_count=view.groupby(by=['user_id','session_id'],as_index=False)['server_time'].count()\n",
    "site_visit_count.rename(columns={'server_time':'n_visit'},inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "#combining view and item data set\n",
    "view1=pd.merge(view,site_visit_count,how='left',on=['user_id','session_id'])\n",
    "view_item_comb=pd.merge(view,item,how='left',on='item_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "#mergiing all the three dataset i.e train/test,view and item on user_id\n",
    "final=pd.merge(train_test,view_item_comb,how='left',on='user_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating new feature 'diff' based on time difference between site visit and ad impression\n",
    "final['diff']=final['impression_time']-final['server_time']\n",
    "final['diff']=final['diff'].dt.days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# replacing NA with their mean and max frequency value\n",
    "final.item_price.fillna(final.item_price.mean(),inplace=True)\n",
    "final.category_1.fillna(9,inplace=True)\n",
    "final.category_2.fillna(38,inplace=True)\n",
    "final.category_3.fillna(62,inplace=True)\n",
    "final.product_type.fillna(2874,inplace=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Removing rows where impression time<server time\n",
    "final1=final[final.impression_time>final.server_time]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "final1.drop(['server_time'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9800453"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(final1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "final1=final1.drop_duplicates(keep='first')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "os_version=pd.get_dummies(final1.os_version,prefix='os',drop_first=True)\n",
    "device_type=pd.get_dummies(final1.device_type,prefix='device',drop_first=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "final1=pd.concat([final1,os_version,device_type],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_train=final1[final1.is_train==1]\n",
    "final_test=final1[final1.is_train==0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    ",'category_1', 'category_2', 'category_3'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "y=final_train.is_click\n",
    "X=final_train.drop(['impression_id','impression_time','user_id','is_click','is_train','session_id','item_id',\\\n",
    "                    'device_type','os_version'],axis=1)\n",
    "\n",
    "x_impression_id=final_test[['impression_id']]\n",
    "x_impression_id.reset_index(drop=True,inplace=True)\n",
    "\n",
    "X_test=final_test.drop(['impression_id','impression_time','user_id','is_click','is_train','session_id','item_id',\\\n",
    "                       'device_type','os_version'],axis=1)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['app_code', 'is_4G', 'day', 'month', 'week_day', 'time', 'n_visit',\n",
       "       'item_price', 'category_1', 'category_2', 'category_3', 'product_type',\n",
       "       'diff', 'os_latest', 'os_old', 'device_iphone', 'device_web'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE\n",
    "\n",
    "smote = SMOTE(random_state=42)\n",
    "X_res, y_res = smote.fit_sample(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "# models\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from xgboost.sklearn import XGBClassifier\n",
    "import catboost as cb\n",
    "\n",
    "\n",
    "LR=LogisticRegression()\n",
    "\n",
    "XGB=XGBClassifier(\n",
    " learning_rate =0.08,\n",
    " n_estimators=200,\n",
    " max_depth=5,\n",
    " min_child_weight=1,\n",
    " gamma=0,\n",
    " subsample=0.8,\n",
    " colsample_bytree=0.8,\n",
    " objective= 'binary:logistic',\n",
    " nthread=4,\n",
    " scale_pos_weight=1,\n",
    " eval_metric = 'auc',\n",
    " seed=27)\n",
    "\n",
    "CB=cb.CatBoostClassifier()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix,f1_score,accuracy_score,classification_report\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prediction(model,X,y,X_test,submission_file):\n",
    "    clf=model.fit(X,y)\n",
    "    y_pred=clf.predict(X)\n",
    "    \n",
    "    print(\"Training Data metrics\")\n",
    "    print(\"accuracy_score={0} and f1_score={1}\".format(accuracy_score(y,y_pred),f1_score(y,y_pred)))\n",
    "    print(\"Confusion Matrix\")\n",
    "    print(confusion_matrix(y,y_pred))\n",
    "    print(\"\\n\")\n",
    "    print(\"classification report\")\n",
    "    print(classification_report(y,y_pred))\n",
    "    \n",
    "    test_predict=clf.predict_proba(X_test)\n",
    "    sol=pd.DataFrame(test_predict[:,1])\n",
    "    sol.columns=['prob']\n",
    "    sol1=pd.concat([x_impression_id,sol],axis=1)\n",
    "    sol2=sol1.groupby('impression_id',as_index=False)['prob'].mean()\n",
    "    sol2=pd.DataFrame(sol2)\n",
    "    submission=pd.merge(submission_file,sol2,how='left',on='impression_id').drop('is_click',axis=1)\n",
    "    submission.rename(columns={'prob':'is_click'},inplace=True)\n",
    "    submission.to_csv(\"./WNS/submission.csv\",index=False)\n",
    "    return clf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[15:48:28] Tree method is automatically selected to be 'approx' for faster speed. to use old behavior(exact greedy algorithm on single machine), set tree_method to 'exact'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Lalit\\Anaconda3\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Data metrics\n",
      "accuracy_score=0.9423875821071221 and f1_score=0.007441746357070936\n",
      "Confusion Matrix\n",
      "[[6351648       4]\n",
      " [ 388390    1456]]\n",
      "\n",
      "\n",
      "classification report\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.94      1.00      0.97   6351652\n",
      "          1       1.00      0.00      0.01    389846\n",
      "\n",
      "avg / total       0.95      0.94      0.91   6741498\n",
      "\n"
     ]
    }
   ],
   "source": [
    "classifier_xgb=prediction(XGB,X,y,X_test,submission)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "importance=classifier_xgb.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "imp=pd.DataFrame(importance,index=X.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x25b9c207668>"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "imp.plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Learning rate set to 0.203078\n",
      "0:\tlearn: 0.4808205\ttotal: 2.39s\tremaining: 39m 45s\n",
      "1:\tlearn: 0.3661879\ttotal: 5.01s\tremaining: 41m 41s\n",
      "2:\tlearn: 0.2998549\ttotal: 7.56s\tremaining: 41m 52s\n",
      "3:\tlearn: 0.2647542\ttotal: 9.77s\tremaining: 40m 31s\n",
      "4:\tlearn: 0.2449497\ttotal: 11.9s\tremaining: 39m 38s\n",
      "5:\tlearn: 0.2308526\ttotal: 14.4s\tremaining: 39m 38s\n",
      "6:\tlearn: 0.2246990\ttotal: 16.7s\tremaining: 39m 25s\n",
      "7:\tlearn: 0.2200383\ttotal: 18.9s\tremaining: 39m 9s\n",
      "8:\tlearn: 0.2165072\ttotal: 21s\tremaining: 38m 31s\n",
      "9:\tlearn: 0.2143297\ttotal: 23.5s\tremaining: 38m 41s\n",
      "10:\tlearn: 0.2133438\ttotal: 25.9s\tremaining: 38m 45s\n",
      "11:\tlearn: 0.2123169\ttotal: 28.2s\tremaining: 38m 43s\n",
      "12:\tlearn: 0.2113645\ttotal: 30.3s\tremaining: 38m 18s\n",
      "13:\tlearn: 0.2110694\ttotal: 32.6s\tremaining: 38m 15s\n",
      "14:\tlearn: 0.2108508\ttotal: 35.2s\tremaining: 38m 29s\n",
      "15:\tlearn: 0.2105519\ttotal: 37.6s\tremaining: 38m 35s\n",
      "16:\tlearn: 0.2101932\ttotal: 40s\tremaining: 38m 35s\n",
      "17:\tlearn: 0.2098785\ttotal: 42s\tremaining: 38m 13s\n",
      "18:\tlearn: 0.2096124\ttotal: 44.5s\tremaining: 38m 16s\n",
      "19:\tlearn: 0.2092940\ttotal: 47s\tremaining: 38m 23s\n",
      "20:\tlearn: 0.2088849\ttotal: 49.4s\tremaining: 38m 24s\n",
      "21:\tlearn: 0.2087361\ttotal: 51.5s\tremaining: 38m 8s\n",
      "22:\tlearn: 0.2085531\ttotal: 53.7s\tremaining: 38m\n",
      "23:\tlearn: 0.2084089\ttotal: 56.2s\tremaining: 38m 3s\n",
      "24:\tlearn: 0.2082459\ttotal: 58.6s\tremaining: 38m 5s\n",
      "25:\tlearn: 0.2080254\ttotal: 1m\tremaining: 37m 58s\n",
      "26:\tlearn: 0.2077295\ttotal: 1m 3s\tremaining: 37m 54s\n",
      "27:\tlearn: 0.2076043\ttotal: 1m 5s\tremaining: 37m 56s\n",
      "28:\tlearn: 0.2074540\ttotal: 1m 8s\tremaining: 38m 2s\n",
      "29:\tlearn: 0.2072743\ttotal: 1m 10s\tremaining: 38m 3s\n",
      "30:\tlearn: 0.2072087\ttotal: 1m 12s\tremaining: 37m 53s\n",
      "31:\tlearn: 0.2071054\ttotal: 1m 14s\tremaining: 37m 48s\n",
      "32:\tlearn: 0.2069930\ttotal: 1m 17s\tremaining: 37m 49s\n",
      "33:\tlearn: 0.2069174\ttotal: 1m 19s\tremaining: 37m 48s\n",
      "34:\tlearn: 0.2067728\ttotal: 1m 22s\tremaining: 37m 45s\n",
      "35:\tlearn: 0.2066858\ttotal: 1m 24s\tremaining: 37m 44s\n",
      "36:\tlearn: 0.2065648\ttotal: 1m 26s\tremaining: 37m 39s\n",
      "37:\tlearn: 0.2063080\ttotal: 1m 29s\tremaining: 37m 41s\n",
      "38:\tlearn: 0.2061734\ttotal: 1m 31s\tremaining: 37m 42s\n",
      "39:\tlearn: 0.2059908\ttotal: 1m 34s\tremaining: 37m 40s\n",
      "40:\tlearn: 0.2058779\ttotal: 1m 36s\tremaining: 37m 36s\n",
      "41:\tlearn: 0.2057127\ttotal: 1m 38s\tremaining: 37m 34s\n",
      "42:\tlearn: 0.2055847\ttotal: 1m 41s\tremaining: 37m 34s\n",
      "43:\tlearn: 0.2054663\ttotal: 1m 43s\tremaining: 37m 35s\n",
      "44:\tlearn: 0.2053925\ttotal: 1m 45s\tremaining: 37m 29s\n",
      "45:\tlearn: 0.2053100\ttotal: 1m 48s\tremaining: 37m 22s\n",
      "46:\tlearn: 0.2051623\ttotal: 1m 50s\tremaining: 37m 20s\n",
      "47:\tlearn: 0.2050879\ttotal: 1m 52s\tremaining: 37m 12s\n",
      "48:\tlearn: 0.2050178\ttotal: 1m 55s\tremaining: 37m 12s\n",
      "49:\tlearn: 0.2048559\ttotal: 1m 57s\tremaining: 37m 14s\n",
      "50:\tlearn: 0.2046507\ttotal: 2m\tremaining: 37m 13s\n",
      "51:\tlearn: 0.2045177\ttotal: 2m 2s\tremaining: 37m 6s\n",
      "52:\tlearn: 0.2043098\ttotal: 2m 4s\tremaining: 37m 6s\n",
      "53:\tlearn: 0.2042509\ttotal: 2m 7s\tremaining: 37m 13s\n",
      "54:\tlearn: 0.2041387\ttotal: 2m 9s\tremaining: 37m 6s\n",
      "55:\tlearn: 0.2040367\ttotal: 2m 11s\tremaining: 37m 4s\n",
      "56:\tlearn: 0.2039508\ttotal: 2m 14s\tremaining: 37m 10s\n",
      "57:\tlearn: 0.2038647\ttotal: 2m 17s\tremaining: 37m 6s\n",
      "58:\tlearn: 0.2037648\ttotal: 2m 19s\tremaining: 37m 3s\n",
      "59:\tlearn: 0.2036718\ttotal: 2m 21s\tremaining: 37m 2s\n",
      "60:\tlearn: 0.2034834\ttotal: 2m 24s\tremaining: 37m 1s\n",
      "61:\tlearn: 0.2033960\ttotal: 2m 26s\tremaining: 36m 54s\n",
      "62:\tlearn: 0.2033448\ttotal: 2m 28s\tremaining: 36m 51s\n",
      "63:\tlearn: 0.2032347\ttotal: 2m 31s\tremaining: 36m 57s\n",
      "64:\tlearn: 0.2031554\ttotal: 2m 33s\tremaining: 36m 52s\n",
      "65:\tlearn: 0.2030569\ttotal: 2m 36s\tremaining: 36m 49s\n",
      "66:\tlearn: 0.2029705\ttotal: 2m 38s\tremaining: 36m 48s\n",
      "67:\tlearn: 0.2028793\ttotal: 2m 41s\tremaining: 36m 48s\n",
      "68:\tlearn: 0.2027805\ttotal: 2m 43s\tremaining: 36m 47s\n",
      "69:\tlearn: 0.2026518\ttotal: 2m 46s\tremaining: 36m 47s\n",
      "70:\tlearn: 0.2025279\ttotal: 2m 48s\tremaining: 36m 47s\n",
      "71:\tlearn: 0.2024317\ttotal: 2m 51s\tremaining: 36m 45s\n",
      "72:\tlearn: 0.2022688\ttotal: 2m 53s\tremaining: 36m 44s\n",
      "73:\tlearn: 0.2021920\ttotal: 2m 55s\tremaining: 36m 41s\n",
      "74:\tlearn: 0.2021430\ttotal: 2m 58s\tremaining: 36m 36s\n",
      "75:\tlearn: 0.2019918\ttotal: 3m\tremaining: 36m 32s\n",
      "76:\tlearn: 0.2019372\ttotal: 3m 3s\tremaining: 36m 37s\n",
      "77:\tlearn: 0.2018747\ttotal: 3m 5s\tremaining: 36m 35s\n",
      "78:\tlearn: 0.2017741\ttotal: 3m 8s\tremaining: 36m 37s\n",
      "79:\tlearn: 0.2017041\ttotal: 3m 10s\tremaining: 36m 33s\n",
      "80:\tlearn: 0.2016314\ttotal: 3m 12s\tremaining: 36m 27s\n",
      "81:\tlearn: 0.2015499\ttotal: 3m 15s\tremaining: 36m 25s\n",
      "82:\tlearn: 0.2014553\ttotal: 3m 17s\tremaining: 36m 24s\n",
      "83:\tlearn: 0.2013844\ttotal: 3m 20s\tremaining: 36m 21s\n",
      "84:\tlearn: 0.2013119\ttotal: 3m 22s\tremaining: 36m 20s\n",
      "85:\tlearn: 0.2012401\ttotal: 3m 25s\tremaining: 36m 20s\n",
      "86:\tlearn: 0.2011671\ttotal: 3m 27s\tremaining: 36m 16s\n",
      "87:\tlearn: 0.2011262\ttotal: 3m 29s\tremaining: 36m 10s\n",
      "88:\tlearn: 0.2010473\ttotal: 3m 31s\tremaining: 36m 8s\n",
      "89:\tlearn: 0.2009926\ttotal: 3m 34s\tremaining: 36m 10s\n",
      "90:\tlearn: 0.2008647\ttotal: 3m 37s\tremaining: 36m 12s\n",
      "91:\tlearn: 0.2007846\ttotal: 3m 39s\tremaining: 36m 10s\n",
      "92:\tlearn: 0.2006602\ttotal: 3m 42s\tremaining: 36m 8s\n",
      "93:\tlearn: 0.2005836\ttotal: 3m 44s\tremaining: 36m 8s\n",
      "94:\tlearn: 0.2005106\ttotal: 3m 47s\tremaining: 36m 6s\n",
      "95:\tlearn: 0.2004268\ttotal: 3m 49s\tremaining: 36m 3s\n",
      "96:\tlearn: 0.2003355\ttotal: 3m 52s\tremaining: 36m 3s\n",
      "97:\tlearn: 0.2002696\ttotal: 3m 54s\tremaining: 36m 1s\n",
      "98:\tlearn: 0.2001849\ttotal: 3m 57s\tremaining: 35m 59s\n",
      "99:\tlearn: 0.2001060\ttotal: 3m 59s\tremaining: 35m 58s\n",
      "100:\tlearn: 0.2000233\ttotal: 4m 2s\tremaining: 35m 56s\n",
      "101:\tlearn: 0.1998577\ttotal: 4m 4s\tremaining: 35m 56s\n",
      "102:\tlearn: 0.1997160\ttotal: 4m 7s\tremaining: 35m 58s\n",
      "103:\tlearn: 0.1996154\ttotal: 4m 10s\tremaining: 35m 54s\n",
      "104:\tlearn: 0.1995560\ttotal: 4m 13s\tremaining: 35m 56s\n",
      "105:\tlearn: 0.1994892\ttotal: 4m 15s\tremaining: 35m 54s\n",
      "106:\tlearn: 0.1994283\ttotal: 4m 18s\tremaining: 35m 54s\n",
      "107:\tlearn: 0.1992674\ttotal: 4m 20s\tremaining: 35m 53s\n",
      "108:\tlearn: 0.1992048\ttotal: 4m 22s\tremaining: 35m 48s\n",
      "109:\tlearn: 0.1990708\ttotal: 4m 25s\tremaining: 35m 49s\n",
      "110:\tlearn: 0.1990038\ttotal: 4m 28s\tremaining: 35m 47s\n",
      "111:\tlearn: 0.1989557\ttotal: 4m 30s\tremaining: 35m 45s\n",
      "112:\tlearn: 0.1989214\ttotal: 4m 33s\tremaining: 35m 45s\n",
      "113:\tlearn: 0.1987858\ttotal: 4m 35s\tremaining: 35m 45s\n",
      "114:\tlearn: 0.1986651\ttotal: 4m 38s\tremaining: 35m 42s\n",
      "115:\tlearn: 0.1985418\ttotal: 4m 41s\tremaining: 35m 44s\n",
      "116:\tlearn: 0.1984683\ttotal: 4m 43s\tremaining: 35m 40s\n",
      "117:\tlearn: 0.1983974\ttotal: 4m 46s\tremaining: 35m 42s\n",
      "118:\tlearn: 0.1983620\ttotal: 4m 49s\tremaining: 35m 39s\n",
      "119:\tlearn: 0.1983043\ttotal: 4m 51s\tremaining: 35m 38s\n",
      "120:\tlearn: 0.1982828\ttotal: 4m 54s\tremaining: 35m 36s\n",
      "121:\tlearn: 0.1982151\ttotal: 4m 56s\tremaining: 35m 31s\n",
      "122:\tlearn: 0.1981231\ttotal: 4m 58s\tremaining: 35m 31s\n",
      "123:\tlearn: 0.1980300\ttotal: 5m 1s\tremaining: 35m 29s\n",
      "124:\tlearn: 0.1979671\ttotal: 5m 3s\tremaining: 35m 27s\n",
      "125:\tlearn: 0.1978862\ttotal: 5m 6s\tremaining: 35m 25s\n",
      "126:\tlearn: 0.1978221\ttotal: 5m 8s\tremaining: 35m 22s\n",
      "127:\tlearn: 0.1977566\ttotal: 5m 11s\tremaining: 35m 20s\n",
      "128:\tlearn: 0.1976625\ttotal: 5m 13s\tremaining: 35m 18s\n",
      "129:\tlearn: 0.1976007\ttotal: 5m 16s\tremaining: 35m 15s\n",
      "130:\tlearn: 0.1974483\ttotal: 5m 18s\tremaining: 35m 14s\n",
      "131:\tlearn: 0.1973555\ttotal: 5m 21s\tremaining: 35m 12s\n",
      "132:\tlearn: 0.1972770\ttotal: 5m 23s\tremaining: 35m 7s\n",
      "133:\tlearn: 0.1971723\ttotal: 5m 26s\tremaining: 35m 7s\n",
      "134:\tlearn: 0.1971224\ttotal: 5m 28s\tremaining: 35m 4s\n",
      "135:\tlearn: 0.1970197\ttotal: 5m 30s\tremaining: 35m 1s\n",
      "136:\tlearn: 0.1969272\ttotal: 5m 33s\tremaining: 35m 1s\n",
      "137:\tlearn: 0.1968568\ttotal: 5m 36s\tremaining: 34m 59s\n",
      "138:\tlearn: 0.1968101\ttotal: 5m 38s\tremaining: 34m 56s\n",
      "139:\tlearn: 0.1966771\ttotal: 5m 41s\tremaining: 34m 55s\n",
      "140:\tlearn: 0.1965913\ttotal: 5m 43s\tremaining: 34m 52s\n",
      "141:\tlearn: 0.1965102\ttotal: 5m 45s\tremaining: 34m 50s\n",
      "142:\tlearn: 0.1963552\ttotal: 5m 48s\tremaining: 34m 48s\n",
      "143:\tlearn: 0.1963094\ttotal: 5m 50s\tremaining: 34m 45s\n",
      "144:\tlearn: 0.1962678\ttotal: 5m 53s\tremaining: 34m 43s\n",
      "145:\tlearn: 0.1962283\ttotal: 5m 55s\tremaining: 34m 40s\n",
      "146:\tlearn: 0.1961675\ttotal: 5m 58s\tremaining: 34m 38s\n",
      "147:\tlearn: 0.1961056\ttotal: 6m\tremaining: 34m 37s\n",
      "148:\tlearn: 0.1960082\ttotal: 6m 3s\tremaining: 34m 34s\n",
      "149:\tlearn: 0.1959262\ttotal: 6m 6s\tremaining: 34m 35s\n",
      "150:\tlearn: 0.1958327\ttotal: 6m 8s\tremaining: 34m 31s\n",
      "151:\tlearn: 0.1956671\ttotal: 6m 11s\tremaining: 34m 31s\n",
      "152:\tlearn: 0.1955782\ttotal: 6m 13s\tremaining: 34m 29s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153:\tlearn: 0.1955032\ttotal: 6m 15s\tremaining: 34m 25s\n",
      "154:\tlearn: 0.1953974\ttotal: 6m 18s\tremaining: 34m 24s\n",
      "155:\tlearn: 0.1953566\ttotal: 6m 21s\tremaining: 34m 21s\n",
      "156:\tlearn: 0.1952920\ttotal: 6m 23s\tremaining: 34m 17s\n",
      "157:\tlearn: 0.1952768\ttotal: 6m 26s\tremaining: 34m 17s\n",
      "158:\tlearn: 0.1952001\ttotal: 6m 28s\tremaining: 34m 14s\n",
      "159:\tlearn: 0.1951426\ttotal: 6m 30s\tremaining: 34m 9s\n",
      "160:\tlearn: 0.1950889\ttotal: 6m 33s\tremaining: 34m 10s\n",
      "161:\tlearn: 0.1950287\ttotal: 6m 35s\tremaining: 34m 7s\n",
      "162:\tlearn: 0.1949720\ttotal: 6m 38s\tremaining: 34m 6s\n",
      "163:\tlearn: 0.1949140\ttotal: 6m 40s\tremaining: 34m 3s\n",
      "164:\tlearn: 0.1948424\ttotal: 6m 43s\tremaining: 34m\n",
      "165:\tlearn: 0.1948044\ttotal: 6m 45s\tremaining: 33m 59s\n",
      "166:\tlearn: 0.1947236\ttotal: 6m 48s\tremaining: 33m 56s\n",
      "167:\tlearn: 0.1946468\ttotal: 6m 50s\tremaining: 33m 54s\n",
      "168:\tlearn: 0.1946053\ttotal: 6m 53s\tremaining: 33m 52s\n",
      "169:\tlearn: 0.1945168\ttotal: 6m 55s\tremaining: 33m 48s\n",
      "170:\tlearn: 0.1944444\ttotal: 6m 58s\tremaining: 33m 46s\n",
      "171:\tlearn: 0.1943486\ttotal: 7m\tremaining: 33m 43s\n",
      "172:\tlearn: 0.1943023\ttotal: 7m 2s\tremaining: 33m 39s\n",
      "173:\tlearn: 0.1942860\ttotal: 7m 5s\tremaining: 33m 39s\n",
      "174:\tlearn: 0.1942588\ttotal: 7m 7s\tremaining: 33m 36s\n",
      "175:\tlearn: 0.1941898\ttotal: 7m 9s\tremaining: 33m 32s\n",
      "176:\tlearn: 0.1941818\ttotal: 7m 12s\tremaining: 33m 31s\n",
      "177:\tlearn: 0.1941357\ttotal: 7m 14s\tremaining: 33m 27s\n",
      "178:\tlearn: 0.1940590\ttotal: 7m 17s\tremaining: 33m 26s\n",
      "179:\tlearn: 0.1940015\ttotal: 7m 19s\tremaining: 33m 23s\n",
      "180:\tlearn: 0.1939729\ttotal: 7m 22s\tremaining: 33m 21s\n",
      "181:\tlearn: 0.1939270\ttotal: 7m 24s\tremaining: 33m 19s\n",
      "182:\tlearn: 0.1938802\ttotal: 7m 27s\tremaining: 33m 16s\n",
      "183:\tlearn: 0.1937794\ttotal: 7m 29s\tremaining: 33m 14s\n",
      "184:\tlearn: 0.1936612\ttotal: 7m 32s\tremaining: 33m 11s\n",
      "185:\tlearn: 0.1935669\ttotal: 7m 34s\tremaining: 33m 9s\n",
      "186:\tlearn: 0.1935008\ttotal: 7m 37s\tremaining: 33m 9s\n",
      "187:\tlearn: 0.1934252\ttotal: 7m 39s\tremaining: 33m 5s\n",
      "188:\tlearn: 0.1933628\ttotal: 7m 42s\tremaining: 33m 3s\n",
      "189:\tlearn: 0.1932730\ttotal: 7m 44s\tremaining: 33m\n",
      "190:\tlearn: 0.1932354\ttotal: 7m 46s\tremaining: 32m 56s\n",
      "191:\tlearn: 0.1932233\ttotal: 7m 49s\tremaining: 32m 55s\n",
      "192:\tlearn: 0.1931223\ttotal: 7m 51s\tremaining: 32m 52s\n",
      "193:\tlearn: 0.1930607\ttotal: 7m 54s\tremaining: 32m 50s\n",
      "194:\tlearn: 0.1930177\ttotal: 7m 56s\tremaining: 32m 48s\n",
      "195:\tlearn: 0.1929944\ttotal: 7m 59s\tremaining: 32m 45s\n",
      "196:\tlearn: 0.1929627\ttotal: 8m 1s\tremaining: 32m 43s\n",
      "197:\tlearn: 0.1929497\ttotal: 8m 3s\tremaining: 32m 40s\n",
      "198:\tlearn: 0.1928916\ttotal: 8m 6s\tremaining: 32m 39s\n",
      "199:\tlearn: 0.1928449\ttotal: 8m 8s\tremaining: 32m 35s\n",
      "200:\tlearn: 0.1927703\ttotal: 8m 11s\tremaining: 32m 32s\n",
      "201:\tlearn: 0.1926911\ttotal: 8m 13s\tremaining: 32m 29s\n",
      "202:\tlearn: 0.1926423\ttotal: 8m 16s\tremaining: 32m 27s\n",
      "203:\tlearn: 0.1926272\ttotal: 8m 18s\tremaining: 32m 24s\n",
      "204:\tlearn: 0.1925922\ttotal: 8m 20s\tremaining: 32m 22s\n",
      "205:\tlearn: 0.1925397\ttotal: 8m 23s\tremaining: 32m 19s\n",
      "206:\tlearn: 0.1925223\ttotal: 8m 25s\tremaining: 32m 17s\n",
      "207:\tlearn: 0.1924936\ttotal: 8m 28s\tremaining: 32m 15s\n",
      "208:\tlearn: 0.1924391\ttotal: 8m 30s\tremaining: 32m 12s\n",
      "209:\tlearn: 0.1923904\ttotal: 8m 33s\tremaining: 32m 11s\n",
      "210:\tlearn: 0.1923271\ttotal: 8m 35s\tremaining: 32m 9s\n",
      "211:\tlearn: 0.1922931\ttotal: 8m 38s\tremaining: 32m 7s\n",
      "212:\tlearn: 0.1922867\ttotal: 8m 40s\tremaining: 32m 3s\n",
      "213:\tlearn: 0.1922351\ttotal: 8m 42s\tremaining: 32m\n",
      "214:\tlearn: 0.1921226\ttotal: 8m 45s\tremaining: 31m 59s\n",
      "215:\tlearn: 0.1920781\ttotal: 8m 47s\tremaining: 31m 55s\n",
      "216:\tlearn: 0.1920144\ttotal: 8m 50s\tremaining: 31m 54s\n",
      "217:\tlearn: 0.1920049\ttotal: 8m 52s\tremaining: 31m 50s\n",
      "218:\tlearn: 0.1919845\ttotal: 8m 54s\tremaining: 31m 47s\n",
      "219:\tlearn: 0.1919328\ttotal: 8m 57s\tremaining: 31m 45s\n",
      "220:\tlearn: 0.1918943\ttotal: 8m 59s\tremaining: 31m 42s\n",
      "221:\tlearn: 0.1918251\ttotal: 9m 2s\tremaining: 31m 40s\n",
      "222:\tlearn: 0.1917795\ttotal: 9m 4s\tremaining: 31m 38s\n",
      "223:\tlearn: 0.1917581\ttotal: 9m 7s\tremaining: 31m 35s\n",
      "224:\tlearn: 0.1916820\ttotal: 9m 9s\tremaining: 31m 33s\n",
      "225:\tlearn: 0.1915927\ttotal: 9m 12s\tremaining: 31m 30s\n",
      "226:\tlearn: 0.1915520\ttotal: 9m 14s\tremaining: 31m 27s\n",
      "227:\tlearn: 0.1914687\ttotal: 9m 17s\tremaining: 31m 26s\n",
      "228:\tlearn: 0.1914617\ttotal: 9m 19s\tremaining: 31m 23s\n",
      "229:\tlearn: 0.1914535\ttotal: 9m 22s\tremaining: 31m 22s\n",
      "230:\tlearn: 0.1914140\ttotal: 9m 24s\tremaining: 31m 19s\n",
      "231:\tlearn: 0.1912729\ttotal: 9m 27s\tremaining: 31m 18s\n",
      "232:\tlearn: 0.1912570\ttotal: 9m 29s\tremaining: 31m 15s\n",
      "233:\tlearn: 0.1912210\ttotal: 9m 31s\tremaining: 31m 12s\n",
      "234:\tlearn: 0.1911772\ttotal: 9m 34s\tremaining: 31m 11s\n",
      "235:\tlearn: 0.1910686\ttotal: 9m 37s\tremaining: 31m 9s\n",
      "236:\tlearn: 0.1910009\ttotal: 9m 40s\tremaining: 31m 7s\n",
      "237:\tlearn: 0.1909455\ttotal: 9m 42s\tremaining: 31m 5s\n",
      "238:\tlearn: 0.1908653\ttotal: 9m 44s\tremaining: 31m 2s\n",
      "239:\tlearn: 0.1908072\ttotal: 9m 47s\tremaining: 31m 1s\n",
      "240:\tlearn: 0.1907588\ttotal: 9m 50s\tremaining: 30m 58s\n",
      "241:\tlearn: 0.1906982\ttotal: 9m 53s\tremaining: 30m 57s\n",
      "242:\tlearn: 0.1906603\ttotal: 9m 55s\tremaining: 30m 54s\n",
      "243:\tlearn: 0.1906497\ttotal: 9m 57s\tremaining: 30m 52s\n",
      "244:\tlearn: 0.1906099\ttotal: 10m\tremaining: 30m 50s\n",
      "245:\tlearn: 0.1905519\ttotal: 10m 2s\tremaining: 30m 46s\n",
      "246:\tlearn: 0.1904748\ttotal: 10m 5s\tremaining: 30m 45s\n",
      "247:\tlearn: 0.1904458\ttotal: 10m 7s\tremaining: 30m 43s\n",
      "248:\tlearn: 0.1904131\ttotal: 10m 10s\tremaining: 30m 40s\n",
      "249:\tlearn: 0.1903548\ttotal: 10m 12s\tremaining: 30m 38s\n",
      "250:\tlearn: 0.1903264\ttotal: 10m 14s\tremaining: 30m 34s\n",
      "251:\tlearn: 0.1902679\ttotal: 10m 17s\tremaining: 30m 32s\n",
      "252:\tlearn: 0.1901922\ttotal: 10m 19s\tremaining: 30m 30s\n",
      "253:\tlearn: 0.1901441\ttotal: 10m 22s\tremaining: 30m 27s\n",
      "254:\tlearn: 0.1900947\ttotal: 10m 24s\tremaining: 30m 25s\n",
      "255:\tlearn: 0.1900307\ttotal: 10m 27s\tremaining: 30m 23s\n",
      "256:\tlearn: 0.1899858\ttotal: 10m 29s\tremaining: 30m 19s\n",
      "257:\tlearn: 0.1899198\ttotal: 10m 32s\tremaining: 30m 18s\n",
      "258:\tlearn: 0.1898757\ttotal: 10m 34s\tremaining: 30m 15s\n",
      "259:\tlearn: 0.1898487\ttotal: 10m 37s\tremaining: 30m 13s\n",
      "260:\tlearn: 0.1898258\ttotal: 10m 39s\tremaining: 30m 10s\n",
      "261:\tlearn: 0.1897579\ttotal: 10m 41s\tremaining: 30m 8s\n",
      "262:\tlearn: 0.1897186\ttotal: 10m 44s\tremaining: 30m 5s\n",
      "263:\tlearn: 0.1896746\ttotal: 10m 46s\tremaining: 30m 2s\n",
      "264:\tlearn: 0.1895830\ttotal: 10m 49s\tremaining: 30m\n",
      "265:\tlearn: 0.1895217\ttotal: 10m 51s\tremaining: 29m 58s\n",
      "266:\tlearn: 0.1894300\ttotal: 10m 54s\tremaining: 29m 55s\n",
      "267:\tlearn: 0.1893601\ttotal: 10m 56s\tremaining: 29m 53s\n",
      "268:\tlearn: 0.1893100\ttotal: 10m 59s\tremaining: 29m 51s\n",
      "269:\tlearn: 0.1892499\ttotal: 11m 1s\tremaining: 29m 48s\n",
      "270:\tlearn: 0.1891639\ttotal: 11m 4s\tremaining: 29m 46s\n",
      "271:\tlearn: 0.1891326\ttotal: 11m 6s\tremaining: 29m 43s\n",
      "272:\tlearn: 0.1890470\ttotal: 11m 9s\tremaining: 29m 42s\n",
      "273:\tlearn: 0.1890115\ttotal: 11m 11s\tremaining: 29m 38s\n",
      "274:\tlearn: 0.1889844\ttotal: 11m 13s\tremaining: 29m 35s\n",
      "275:\tlearn: 0.1889683\ttotal: 11m 16s\tremaining: 29m 33s\n",
      "276:\tlearn: 0.1888903\ttotal: 11m 18s\tremaining: 29m 30s\n",
      "277:\tlearn: 0.1888774\ttotal: 11m 21s\tremaining: 29m 28s\n",
      "278:\tlearn: 0.1887894\ttotal: 11m 23s\tremaining: 29m 26s\n",
      "279:\tlearn: 0.1887769\ttotal: 11m 25s\tremaining: 29m 23s\n",
      "280:\tlearn: 0.1887584\ttotal: 11m 28s\tremaining: 29m 21s\n",
      "281:\tlearn: 0.1887508\ttotal: 11m 30s\tremaining: 29m 18s\n",
      "282:\tlearn: 0.1887121\ttotal: 11m 33s\tremaining: 29m 16s\n",
      "283:\tlearn: 0.1886857\ttotal: 11m 35s\tremaining: 29m 14s\n",
      "284:\tlearn: 0.1886386\ttotal: 11m 38s\tremaining: 29m 11s\n",
      "285:\tlearn: 0.1886112\ttotal: 11m 40s\tremaining: 29m 9s\n",
      "286:\tlearn: 0.1885690\ttotal: 11m 43s\tremaining: 29m 6s\n",
      "287:\tlearn: 0.1885622\ttotal: 11m 45s\tremaining: 29m 4s\n",
      "288:\tlearn: 0.1885548\ttotal: 11m 47s\tremaining: 29m 1s\n",
      "289:\tlearn: 0.1884766\ttotal: 11m 50s\tremaining: 28m 59s\n",
      "290:\tlearn: 0.1884384\ttotal: 11m 52s\tremaining: 28m 56s\n",
      "291:\tlearn: 0.1883962\ttotal: 11m 55s\tremaining: 28m 54s\n",
      "292:\tlearn: 0.1882987\ttotal: 11m 57s\tremaining: 28m 51s\n",
      "293:\tlearn: 0.1882546\ttotal: 12m\tremaining: 28m 49s\n",
      "294:\tlearn: 0.1882384\ttotal: 12m 2s\tremaining: 28m 47s\n",
      "295:\tlearn: 0.1882125\ttotal: 12m 5s\tremaining: 28m 44s\n",
      "296:\tlearn: 0.1882071\ttotal: 12m 7s\tremaining: 28m 42s\n",
      "297:\tlearn: 0.1881818\ttotal: 12m 10s\tremaining: 28m 39s\n",
      "298:\tlearn: 0.1881007\ttotal: 12m 12s\tremaining: 28m 37s\n",
      "299:\tlearn: 0.1880652\ttotal: 12m 15s\tremaining: 28m 35s\n",
      "300:\tlearn: 0.1880120\ttotal: 12m 17s\tremaining: 28m 32s\n",
      "301:\tlearn: 0.1879783\ttotal: 12m 20s\tremaining: 28m 31s\n",
      "302:\tlearn: 0.1879337\ttotal: 12m 22s\tremaining: 28m 27s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "303:\tlearn: 0.1878746\ttotal: 12m 24s\tremaining: 28m 25s\n",
      "304:\tlearn: 0.1878684\ttotal: 12m 27s\tremaining: 28m 22s\n",
      "305:\tlearn: 0.1878642\ttotal: 12m 29s\tremaining: 28m 20s\n",
      "306:\tlearn: 0.1878022\ttotal: 12m 32s\tremaining: 28m 17s\n",
      "307:\tlearn: 0.1877477\ttotal: 12m 34s\tremaining: 28m 15s\n",
      "308:\tlearn: 0.1876553\ttotal: 12m 37s\tremaining: 28m 12s\n",
      "309:\tlearn: 0.1876122\ttotal: 12m 39s\tremaining: 28m 11s\n",
      "310:\tlearn: 0.1875279\ttotal: 12m 41s\tremaining: 28m 8s\n",
      "311:\tlearn: 0.1874814\ttotal: 12m 44s\tremaining: 28m 5s\n",
      "312:\tlearn: 0.1874067\ttotal: 12m 46s\tremaining: 28m 3s\n",
      "313:\tlearn: 0.1873442\ttotal: 12m 49s\tremaining: 28m\n",
      "314:\tlearn: 0.1872990\ttotal: 12m 51s\tremaining: 27m 58s\n",
      "315:\tlearn: 0.1872391\ttotal: 12m 54s\tremaining: 27m 56s\n",
      "316:\tlearn: 0.1871978\ttotal: 12m 56s\tremaining: 27m 53s\n",
      "317:\tlearn: 0.1871918\ttotal: 12m 59s\tremaining: 27m 51s\n",
      "318:\tlearn: 0.1871829\ttotal: 13m 1s\tremaining: 27m 48s\n",
      "319:\tlearn: 0.1871608\ttotal: 13m 4s\tremaining: 27m 46s\n",
      "320:\tlearn: 0.1871018\ttotal: 13m 6s\tremaining: 27m 43s\n",
      "321:\tlearn: 0.1870632\ttotal: 13m 9s\tremaining: 27m 42s\n",
      "322:\tlearn: 0.1870145\ttotal: 13m 11s\tremaining: 27m 39s\n",
      "323:\tlearn: 0.1869833\ttotal: 13m 13s\tremaining: 27m 35s\n",
      "324:\tlearn: 0.1869464\ttotal: 13m 16s\tremaining: 27m 33s\n",
      "325:\tlearn: 0.1869111\ttotal: 13m 18s\tremaining: 27m 30s\n",
      "326:\tlearn: 0.1868899\ttotal: 13m 20s\tremaining: 27m 28s\n",
      "327:\tlearn: 0.1868591\ttotal: 13m 23s\tremaining: 27m 25s\n",
      "328:\tlearn: 0.1867783\ttotal: 13m 25s\tremaining: 27m 23s\n",
      "329:\tlearn: 0.1866998\ttotal: 13m 28s\tremaining: 27m 21s\n",
      "330:\tlearn: 0.1866535\ttotal: 13m 30s\tremaining: 27m 18s\n",
      "331:\tlearn: 0.1866036\ttotal: 13m 33s\tremaining: 27m 16s\n",
      "332:\tlearn: 0.1865336\ttotal: 13m 36s\tremaining: 27m 14s\n",
      "333:\tlearn: 0.1865031\ttotal: 13m 38s\tremaining: 27m 12s\n",
      "334:\tlearn: 0.1863584\ttotal: 13m 41s\tremaining: 27m 10s\n",
      "335:\tlearn: 0.1863409\ttotal: 13m 43s\tremaining: 27m 7s\n",
      "336:\tlearn: 0.1863036\ttotal: 13m 46s\tremaining: 27m 5s\n",
      "337:\tlearn: 0.1862699\ttotal: 13m 48s\tremaining: 27m 2s\n",
      "338:\tlearn: 0.1862101\ttotal: 13m 51s\tremaining: 27m\n",
      "339:\tlearn: 0.1861697\ttotal: 13m 53s\tremaining: 26m 57s\n",
      "340:\tlearn: 0.1861532\ttotal: 13m 56s\tremaining: 26m 55s\n",
      "341:\tlearn: 0.1861111\ttotal: 13m 58s\tremaining: 26m 53s\n",
      "342:\tlearn: 0.1860919\ttotal: 14m\tremaining: 26m 50s\n",
      "343:\tlearn: 0.1860541\ttotal: 14m 3s\tremaining: 26m 48s\n",
      "344:\tlearn: 0.1860422\ttotal: 14m 5s\tremaining: 26m 45s\n",
      "345:\tlearn: 0.1860175\ttotal: 14m 8s\tremaining: 26m 43s\n",
      "346:\tlearn: 0.1859860\ttotal: 14m 10s\tremaining: 26m 41s\n",
      "347:\tlearn: 0.1859763\ttotal: 14m 13s\tremaining: 26m 38s\n",
      "348:\tlearn: 0.1859496\ttotal: 14m 16s\tremaining: 26m 36s\n",
      "349:\tlearn: 0.1859086\ttotal: 14m 18s\tremaining: 26m 33s\n",
      "350:\tlearn: 0.1858712\ttotal: 14m 21s\tremaining: 26m 32s\n",
      "351:\tlearn: 0.1858305\ttotal: 14m 23s\tremaining: 26m 29s\n",
      "352:\tlearn: 0.1857412\ttotal: 14m 26s\tremaining: 26m 27s\n",
      "353:\tlearn: 0.1857180\ttotal: 14m 28s\tremaining: 26m 24s\n",
      "354:\tlearn: 0.1857055\ttotal: 14m 30s\tremaining: 26m 21s\n",
      "355:\tlearn: 0.1856443\ttotal: 14m 33s\tremaining: 26m 19s\n",
      "356:\tlearn: 0.1855869\ttotal: 14m 35s\tremaining: 26m 16s\n",
      "357:\tlearn: 0.1855369\ttotal: 14m 38s\tremaining: 26m 15s\n",
      "358:\tlearn: 0.1854876\ttotal: 14m 40s\tremaining: 26m 12s\n",
      "359:\tlearn: 0.1854036\ttotal: 14m 42s\tremaining: 26m 9s\n",
      "360:\tlearn: 0.1853254\ttotal: 14m 45s\tremaining: 26m 7s\n",
      "361:\tlearn: 0.1852497\ttotal: 14m 48s\tremaining: 26m 5s\n",
      "362:\tlearn: 0.1852071\ttotal: 14m 50s\tremaining: 26m 2s\n",
      "363:\tlearn: 0.1852009\ttotal: 14m 53s\tremaining: 26m\n",
      "364:\tlearn: 0.1851802\ttotal: 14m 55s\tremaining: 25m 57s\n",
      "365:\tlearn: 0.1851269\ttotal: 14m 58s\tremaining: 25m 55s\n",
      "366:\tlearn: 0.1850777\ttotal: 15m\tremaining: 25m 52s\n",
      "367:\tlearn: 0.1850737\ttotal: 15m 2s\tremaining: 25m 50s\n",
      "368:\tlearn: 0.1850569\ttotal: 15m 5s\tremaining: 25m 48s\n",
      "369:\tlearn: 0.1850306\ttotal: 15m 7s\tremaining: 25m 45s\n",
      "370:\tlearn: 0.1850157\ttotal: 15m 10s\tremaining: 25m 43s\n",
      "371:\tlearn: 0.1849860\ttotal: 15m 12s\tremaining: 25m 40s\n",
      "372:\tlearn: 0.1849478\ttotal: 15m 15s\tremaining: 25m 38s\n",
      "373:\tlearn: 0.1849073\ttotal: 15m 17s\tremaining: 25m 36s\n",
      "374:\tlearn: 0.1848964\ttotal: 15m 20s\tremaining: 25m 33s\n",
      "375:\tlearn: 0.1848757\ttotal: 15m 22s\tremaining: 25m 31s\n",
      "376:\tlearn: 0.1848307\ttotal: 15m 25s\tremaining: 25m 28s\n",
      "377:\tlearn: 0.1847793\ttotal: 15m 27s\tremaining: 25m 26s\n",
      "378:\tlearn: 0.1847341\ttotal: 15m 29s\tremaining: 25m 23s\n",
      "379:\tlearn: 0.1847239\ttotal: 15m 32s\tremaining: 25m 20s\n",
      "380:\tlearn: 0.1846669\ttotal: 15m 34s\tremaining: 25m 18s\n",
      "381:\tlearn: 0.1846192\ttotal: 15m 37s\tremaining: 25m 16s\n",
      "382:\tlearn: 0.1845943\ttotal: 15m 39s\tremaining: 25m 13s\n",
      "383:\tlearn: 0.1845733\ttotal: 15m 42s\tremaining: 25m 11s\n",
      "384:\tlearn: 0.1845242\ttotal: 15m 44s\tremaining: 25m 9s\n",
      "385:\tlearn: 0.1844949\ttotal: 15m 47s\tremaining: 25m 6s\n",
      "386:\tlearn: 0.1844583\ttotal: 15m 49s\tremaining: 25m 4s\n",
      "387:\tlearn: 0.1844388\ttotal: 15m 51s\tremaining: 25m 1s\n",
      "388:\tlearn: 0.1844079\ttotal: 15m 54s\tremaining: 24m 59s\n",
      "389:\tlearn: 0.1844008\ttotal: 15m 56s\tremaining: 24m 56s\n",
      "390:\tlearn: 0.1843767\ttotal: 15m 59s\tremaining: 24m 53s\n",
      "391:\tlearn: 0.1843339\ttotal: 16m 1s\tremaining: 24m 51s\n",
      "392:\tlearn: 0.1843283\ttotal: 16m 4s\tremaining: 24m 49s\n",
      "393:\tlearn: 0.1842938\ttotal: 16m 6s\tremaining: 24m 46s\n",
      "394:\tlearn: 0.1842634\ttotal: 16m 9s\tremaining: 24m 44s\n",
      "395:\tlearn: 0.1842314\ttotal: 16m 11s\tremaining: 24m 41s\n",
      "396:\tlearn: 0.1841938\ttotal: 16m 14s\tremaining: 24m 39s\n",
      "397:\tlearn: 0.1841540\ttotal: 16m 16s\tremaining: 24m 37s\n",
      "398:\tlearn: 0.1841284\ttotal: 16m 19s\tremaining: 24m 34s\n",
      "399:\tlearn: 0.1840717\ttotal: 16m 21s\tremaining: 24m 32s\n",
      "400:\tlearn: 0.1840508\ttotal: 16m 23s\tremaining: 24m 29s\n",
      "401:\tlearn: 0.1840078\ttotal: 16m 26s\tremaining: 24m 28s\n",
      "402:\tlearn: 0.1839859\ttotal: 16m 29s\tremaining: 24m 25s\n",
      "403:\tlearn: 0.1839559\ttotal: 16m 31s\tremaining: 24m 23s\n",
      "404:\tlearn: 0.1839247\ttotal: 16m 34s\tremaining: 24m 20s\n",
      "405:\tlearn: 0.1838915\ttotal: 16m 36s\tremaining: 24m 17s\n",
      "406:\tlearn: 0.1838678\ttotal: 16m 39s\tremaining: 24m 15s\n",
      "407:\tlearn: 0.1838278\ttotal: 16m 41s\tremaining: 24m 13s\n",
      "408:\tlearn: 0.1838182\ttotal: 16m 43s\tremaining: 24m 10s\n",
      "409:\tlearn: 0.1837687\ttotal: 16m 46s\tremaining: 24m 8s\n",
      "410:\tlearn: 0.1837191\ttotal: 16m 49s\tremaining: 24m 5s\n",
      "411:\tlearn: 0.1837028\ttotal: 16m 51s\tremaining: 24m 3s\n",
      "412:\tlearn: 0.1836787\ttotal: 16m 53s\tremaining: 24m 1s\n",
      "413:\tlearn: 0.1836113\ttotal: 16m 56s\tremaining: 23m 58s\n",
      "414:\tlearn: 0.1835623\ttotal: 16m 59s\tremaining: 23m 56s\n",
      "415:\tlearn: 0.1835006\ttotal: 17m 1s\tremaining: 23m 53s\n",
      "416:\tlearn: 0.1834722\ttotal: 17m 4s\tremaining: 23m 51s\n",
      "417:\tlearn: 0.1834336\ttotal: 17m 6s\tremaining: 23m 49s\n",
      "418:\tlearn: 0.1834019\ttotal: 17m 8s\tremaining: 23m 46s\n",
      "419:\tlearn: 0.1833603\ttotal: 17m 11s\tremaining: 23m 44s\n",
      "420:\tlearn: 0.1833538\ttotal: 17m 13s\tremaining: 23m 41s\n",
      "421:\tlearn: 0.1833157\ttotal: 17m 16s\tremaining: 23m 39s\n",
      "422:\tlearn: 0.1832792\ttotal: 17m 18s\tremaining: 23m 36s\n",
      "423:\tlearn: 0.1832342\ttotal: 17m 21s\tremaining: 23m 34s\n",
      "424:\tlearn: 0.1832295\ttotal: 17m 23s\tremaining: 23m 31s\n",
      "425:\tlearn: 0.1832083\ttotal: 17m 26s\tremaining: 23m 29s\n",
      "426:\tlearn: 0.1831583\ttotal: 17m 28s\tremaining: 23m 27s\n",
      "427:\tlearn: 0.1830797\ttotal: 17m 31s\tremaining: 23m 24s\n",
      "428:\tlearn: 0.1830752\ttotal: 17m 33s\tremaining: 23m 22s\n",
      "429:\tlearn: 0.1830519\ttotal: 17m 36s\tremaining: 23m 19s\n",
      "430:\tlearn: 0.1830235\ttotal: 17m 38s\tremaining: 23m 17s\n",
      "431:\tlearn: 0.1830191\ttotal: 17m 41s\tremaining: 23m 15s\n",
      "432:\tlearn: 0.1829705\ttotal: 17m 43s\tremaining: 23m 12s\n",
      "433:\tlearn: 0.1829257\ttotal: 17m 45s\tremaining: 23m 9s\n",
      "434:\tlearn: 0.1828940\ttotal: 17m 48s\tremaining: 23m 7s\n",
      "435:\tlearn: 0.1828759\ttotal: 17m 50s\tremaining: 23m 4s\n",
      "436:\tlearn: 0.1828166\ttotal: 17m 53s\tremaining: 23m 2s\n",
      "437:\tlearn: 0.1827913\ttotal: 17m 55s\tremaining: 22m 59s\n",
      "438:\tlearn: 0.1827437\ttotal: 17m 57s\tremaining: 22m 57s\n",
      "439:\tlearn: 0.1826903\ttotal: 18m\tremaining: 22m 54s\n",
      "440:\tlearn: 0.1826636\ttotal: 18m 2s\tremaining: 22m 52s\n",
      "441:\tlearn: 0.1826053\ttotal: 18m 5s\tremaining: 22m 50s\n",
      "442:\tlearn: 0.1825747\ttotal: 18m 7s\tremaining: 22m 47s\n",
      "443:\tlearn: 0.1825558\ttotal: 18m 10s\tremaining: 22m 45s\n",
      "444:\tlearn: 0.1825146\ttotal: 18m 12s\tremaining: 22m 42s\n",
      "445:\tlearn: 0.1824383\ttotal: 18m 14s\tremaining: 22m 40s\n",
      "446:\tlearn: 0.1824153\ttotal: 18m 17s\tremaining: 22m 37s\n",
      "447:\tlearn: 0.1823665\ttotal: 18m 19s\tremaining: 22m 34s\n",
      "448:\tlearn: 0.1823168\ttotal: 18m 22s\tremaining: 22m 32s\n",
      "449:\tlearn: 0.1823035\ttotal: 18m 24s\tremaining: 22m 30s\n",
      "450:\tlearn: 0.1822778\ttotal: 18m 27s\tremaining: 22m 27s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "451:\tlearn: 0.1822613\ttotal: 18m 29s\tremaining: 22m 25s\n",
      "452:\tlearn: 0.1821804\ttotal: 18m 31s\tremaining: 22m 22s\n",
      "453:\tlearn: 0.1821758\ttotal: 18m 34s\tremaining: 22m 20s\n",
      "454:\tlearn: 0.1821336\ttotal: 18m 37s\tremaining: 22m 18s\n",
      "455:\tlearn: 0.1821068\ttotal: 18m 39s\tremaining: 22m 15s\n",
      "456:\tlearn: 0.1820337\ttotal: 18m 42s\tremaining: 22m 13s\n",
      "457:\tlearn: 0.1819934\ttotal: 18m 44s\tremaining: 22m 10s\n",
      "458:\tlearn: 0.1819730\ttotal: 18m 46s\tremaining: 22m 8s\n",
      "459:\tlearn: 0.1819315\ttotal: 18m 49s\tremaining: 22m 6s\n",
      "460:\tlearn: 0.1818949\ttotal: 18m 51s\tremaining: 22m 3s\n",
      "461:\tlearn: 0.1818695\ttotal: 18m 54s\tremaining: 22m 1s\n",
      "462:\tlearn: 0.1818055\ttotal: 18m 56s\tremaining: 21m 58s\n",
      "463:\tlearn: 0.1817810\ttotal: 18m 58s\tremaining: 21m 55s\n",
      "464:\tlearn: 0.1817646\ttotal: 19m 1s\tremaining: 21m 53s\n",
      "465:\tlearn: 0.1817300\ttotal: 19m 3s\tremaining: 21m 50s\n",
      "466:\tlearn: 0.1816778\ttotal: 19m 6s\tremaining: 21m 48s\n",
      "467:\tlearn: 0.1816567\ttotal: 19m 8s\tremaining: 21m 46s\n",
      "468:\tlearn: 0.1816389\ttotal: 19m 11s\tremaining: 21m 43s\n",
      "469:\tlearn: 0.1815810\ttotal: 19m 14s\tremaining: 21m 41s\n",
      "470:\tlearn: 0.1815759\ttotal: 19m 16s\tremaining: 21m 38s\n",
      "471:\tlearn: 0.1815334\ttotal: 19m 18s\tremaining: 21m 36s\n",
      "472:\tlearn: 0.1815062\ttotal: 19m 21s\tremaining: 21m 33s\n",
      "473:\tlearn: 0.1814594\ttotal: 19m 23s\tremaining: 21m 31s\n",
      "474:\tlearn: 0.1814215\ttotal: 19m 26s\tremaining: 21m 28s\n",
      "475:\tlearn: 0.1813732\ttotal: 19m 28s\tremaining: 21m 26s\n",
      "476:\tlearn: 0.1813676\ttotal: 19m 31s\tremaining: 21m 24s\n",
      "477:\tlearn: 0.1813638\ttotal: 19m 33s\tremaining: 21m 21s\n",
      "478:\tlearn: 0.1813325\ttotal: 19m 36s\tremaining: 21m 19s\n",
      "479:\tlearn: 0.1812816\ttotal: 19m 38s\tremaining: 21m 16s\n",
      "480:\tlearn: 0.1812004\ttotal: 19m 41s\tremaining: 21m 14s\n",
      "481:\tlearn: 0.1811876\ttotal: 19m 43s\tremaining: 21m 11s\n",
      "482:\tlearn: 0.1811774\ttotal: 19m 45s\tremaining: 21m 9s\n",
      "483:\tlearn: 0.1811509\ttotal: 19m 48s\tremaining: 21m 7s\n",
      "484:\tlearn: 0.1811345\ttotal: 19m 50s\tremaining: 21m 4s\n",
      "485:\tlearn: 0.1810680\ttotal: 19m 53s\tremaining: 21m 2s\n",
      "486:\tlearn: 0.1810223\ttotal: 19m 56s\tremaining: 20m 59s\n",
      "487:\tlearn: 0.1810084\ttotal: 19m 58s\tremaining: 20m 57s\n",
      "488:\tlearn: 0.1809443\ttotal: 20m 1s\tremaining: 20m 55s\n",
      "489:\tlearn: 0.1808892\ttotal: 20m 3s\tremaining: 20m 52s\n",
      "490:\tlearn: 0.1808577\ttotal: 20m 6s\tremaining: 20m 50s\n",
      "491:\tlearn: 0.1808292\ttotal: 20m 8s\tremaining: 20m 48s\n",
      "492:\tlearn: 0.1808198\ttotal: 20m 11s\tremaining: 20m 45s\n",
      "493:\tlearn: 0.1808131\ttotal: 20m 13s\tremaining: 20m 43s\n",
      "494:\tlearn: 0.1808090\ttotal: 20m 16s\tremaining: 20m 40s\n",
      "495:\tlearn: 0.1808002\ttotal: 20m 18s\tremaining: 20m 38s\n",
      "496:\tlearn: 0.1807890\ttotal: 20m 21s\tremaining: 20m 35s\n",
      "497:\tlearn: 0.1807345\ttotal: 20m 23s\tremaining: 20m 33s\n",
      "498:\tlearn: 0.1807288\ttotal: 20m 25s\tremaining: 20m 30s\n",
      "499:\tlearn: 0.1807202\ttotal: 20m 28s\tremaining: 20m 28s\n",
      "500:\tlearn: 0.1806661\ttotal: 20m 30s\tremaining: 20m 25s\n",
      "501:\tlearn: 0.1806452\ttotal: 20m 33s\tremaining: 20m 23s\n",
      "502:\tlearn: 0.1806303\ttotal: 20m 35s\tremaining: 20m 20s\n",
      "503:\tlearn: 0.1806228\ttotal: 20m 38s\tremaining: 20m 18s\n",
      "504:\tlearn: 0.1806169\ttotal: 20m 40s\tremaining: 20m 15s\n",
      "505:\tlearn: 0.1805817\ttotal: 20m 42s\tremaining: 20m 13s\n",
      "506:\tlearn: 0.1805579\ttotal: 20m 45s\tremaining: 20m 11s\n",
      "507:\tlearn: 0.1805072\ttotal: 20m 47s\tremaining: 20m 8s\n",
      "508:\tlearn: 0.1804809\ttotal: 20m 50s\tremaining: 20m 6s\n",
      "509:\tlearn: 0.1804359\ttotal: 20m 52s\tremaining: 20m 3s\n",
      "510:\tlearn: 0.1804041\ttotal: 20m 55s\tremaining: 20m 1s\n",
      "511:\tlearn: 0.1803611\ttotal: 20m 57s\tremaining: 19m 58s\n",
      "512:\tlearn: 0.1803565\ttotal: 21m\tremaining: 19m 56s\n",
      "513:\tlearn: 0.1803471\ttotal: 21m 2s\tremaining: 19m 53s\n",
      "514:\tlearn: 0.1803430\ttotal: 21m 5s\tremaining: 19m 51s\n",
      "515:\tlearn: 0.1803262\ttotal: 21m 7s\tremaining: 19m 49s\n",
      "516:\tlearn: 0.1802907\ttotal: 21m 10s\tremaining: 19m 46s\n",
      "517:\tlearn: 0.1802598\ttotal: 21m 12s\tremaining: 19m 44s\n",
      "518:\tlearn: 0.1802549\ttotal: 21m 14s\tremaining: 19m 41s\n",
      "519:\tlearn: 0.1802333\ttotal: 21m 17s\tremaining: 19m 39s\n",
      "520:\tlearn: 0.1801866\ttotal: 21m 19s\tremaining: 19m 36s\n",
      "521:\tlearn: 0.1801285\ttotal: 21m 22s\tremaining: 19m 34s\n",
      "522:\tlearn: 0.1800829\ttotal: 21m 24s\tremaining: 19m 31s\n",
      "523:\tlearn: 0.1800439\ttotal: 21m 27s\tremaining: 19m 29s\n",
      "524:\tlearn: 0.1800338\ttotal: 21m 29s\tremaining: 19m 26s\n",
      "525:\tlearn: 0.1799878\ttotal: 21m 32s\tremaining: 19m 24s\n",
      "526:\tlearn: 0.1799674\ttotal: 21m 34s\tremaining: 19m 21s\n",
      "527:\tlearn: 0.1799533\ttotal: 21m 37s\tremaining: 19m 19s\n",
      "528:\tlearn: 0.1799477\ttotal: 21m 39s\tremaining: 19m 16s\n",
      "529:\tlearn: 0.1799047\ttotal: 21m 42s\tremaining: 19m 14s\n",
      "530:\tlearn: 0.1798953\ttotal: 21m 44s\tremaining: 19m 11s\n",
      "531:\tlearn: 0.1798513\ttotal: 21m 46s\tremaining: 19m 9s\n",
      "532:\tlearn: 0.1797771\ttotal: 21m 49s\tremaining: 19m 7s\n",
      "533:\tlearn: 0.1797330\ttotal: 21m 51s\tremaining: 19m 4s\n",
      "534:\tlearn: 0.1797147\ttotal: 21m 54s\tremaining: 19m 2s\n",
      "535:\tlearn: 0.1796560\ttotal: 21m 56s\tremaining: 18m 59s\n",
      "536:\tlearn: 0.1796341\ttotal: 21m 59s\tremaining: 18m 57s\n",
      "537:\tlearn: 0.1796015\ttotal: 22m 1s\tremaining: 18m 54s\n",
      "538:\tlearn: 0.1795840\ttotal: 22m 4s\tremaining: 18m 52s\n",
      "539:\tlearn: 0.1795571\ttotal: 22m 6s\tremaining: 18m 50s\n",
      "540:\tlearn: 0.1795445\ttotal: 22m 9s\tremaining: 18m 47s\n",
      "541:\tlearn: 0.1794750\ttotal: 22m 11s\tremaining: 18m 45s\n",
      "542:\tlearn: 0.1794721\ttotal: 22m 13s\tremaining: 18m 42s\n",
      "543:\tlearn: 0.1794472\ttotal: 22m 16s\tremaining: 18m 40s\n",
      "544:\tlearn: 0.1794258\ttotal: 22m 18s\tremaining: 18m 37s\n",
      "545:\tlearn: 0.1794081\ttotal: 22m 20s\tremaining: 18m 35s\n",
      "546:\tlearn: 0.1793317\ttotal: 22m 23s\tremaining: 18m 32s\n",
      "547:\tlearn: 0.1793010\ttotal: 22m 25s\tremaining: 18m 30s\n",
      "548:\tlearn: 0.1792840\ttotal: 22m 28s\tremaining: 18m 27s\n",
      "549:\tlearn: 0.1792248\ttotal: 22m 30s\tremaining: 18m 24s\n",
      "550:\tlearn: 0.1791512\ttotal: 22m 33s\tremaining: 18m 22s\n",
      "551:\tlearn: 0.1791472\ttotal: 22m 35s\tremaining: 18m 20s\n",
      "552:\tlearn: 0.1791069\ttotal: 22m 38s\tremaining: 18m 18s\n",
      "553:\tlearn: 0.1790904\ttotal: 22m 40s\tremaining: 18m 15s\n",
      "554:\tlearn: 0.1790592\ttotal: 22m 43s\tremaining: 18m 13s\n",
      "555:\tlearn: 0.1790109\ttotal: 22m 45s\tremaining: 18m 10s\n",
      "556:\tlearn: 0.1790051\ttotal: 22m 48s\tremaining: 18m 8s\n",
      "557:\tlearn: 0.1789667\ttotal: 22m 50s\tremaining: 18m 5s\n",
      "558:\tlearn: 0.1789121\ttotal: 22m 53s\tremaining: 18m 3s\n",
      "559:\tlearn: 0.1788935\ttotal: 22m 55s\tremaining: 18m 1s\n",
      "560:\tlearn: 0.1788888\ttotal: 22m 58s\tremaining: 17m 58s\n",
      "561:\tlearn: 0.1788359\ttotal: 23m\tremaining: 17m 56s\n",
      "562:\tlearn: 0.1788092\ttotal: 23m 3s\tremaining: 17m 53s\n",
      "563:\tlearn: 0.1787765\ttotal: 23m 6s\tremaining: 17m 51s\n",
      "564:\tlearn: 0.1787355\ttotal: 23m 8s\tremaining: 17m 49s\n",
      "565:\tlearn: 0.1786741\ttotal: 23m 10s\tremaining: 17m 46s\n",
      "566:\tlearn: 0.1786310\ttotal: 23m 13s\tremaining: 17m 44s\n",
      "567:\tlearn: 0.1786229\ttotal: 23m 15s\tremaining: 17m 41s\n",
      "568:\tlearn: 0.1785859\ttotal: 23m 17s\tremaining: 17m 38s\n",
      "569:\tlearn: 0.1785804\ttotal: 23m 20s\tremaining: 17m 36s\n",
      "570:\tlearn: 0.1785371\ttotal: 23m 22s\tremaining: 17m 34s\n",
      "571:\tlearn: 0.1785299\ttotal: 23m 25s\tremaining: 17m 31s\n",
      "572:\tlearn: 0.1785197\ttotal: 23m 28s\tremaining: 17m 29s\n",
      "573:\tlearn: 0.1784886\ttotal: 23m 30s\tremaining: 17m 26s\n",
      "574:\tlearn: 0.1784064\ttotal: 23m 33s\tremaining: 17m 24s\n",
      "575:\tlearn: 0.1783670\ttotal: 23m 35s\tremaining: 17m 22s\n",
      "576:\tlearn: 0.1783629\ttotal: 23m 38s\tremaining: 17m 19s\n",
      "577:\tlearn: 0.1783246\ttotal: 23m 40s\tremaining: 17m 17s\n",
      "578:\tlearn: 0.1783167\ttotal: 23m 42s\tremaining: 17m 14s\n",
      "579:\tlearn: 0.1782519\ttotal: 23m 45s\tremaining: 17m 12s\n",
      "580:\tlearn: 0.1782107\ttotal: 23m 48s\tremaining: 17m 9s\n",
      "581:\tlearn: 0.1781902\ttotal: 23m 50s\tremaining: 17m 7s\n",
      "582:\tlearn: 0.1781845\ttotal: 23m 53s\tremaining: 17m 5s\n",
      "583:\tlearn: 0.1781743\ttotal: 23m 55s\tremaining: 17m 2s\n",
      "584:\tlearn: 0.1781465\ttotal: 23m 57s\tremaining: 17m\n",
      "585:\tlearn: 0.1781367\ttotal: 24m\tremaining: 16m 57s\n",
      "586:\tlearn: 0.1781102\ttotal: 24m 2s\tremaining: 16m 55s\n",
      "587:\tlearn: 0.1780830\ttotal: 24m 5s\tremaining: 16m 52s\n",
      "588:\tlearn: 0.1780730\ttotal: 24m 7s\tremaining: 16m 50s\n",
      "589:\tlearn: 0.1780266\ttotal: 24m 10s\tremaining: 16m 47s\n",
      "590:\tlearn: 0.1780171\ttotal: 24m 12s\tremaining: 16m 45s\n",
      "591:\tlearn: 0.1779819\ttotal: 24m 14s\tremaining: 16m 42s\n",
      "592:\tlearn: 0.1779684\ttotal: 24m 17s\tremaining: 16m 40s\n",
      "593:\tlearn: 0.1779555\ttotal: 24m 20s\tremaining: 16m 37s\n",
      "594:\tlearn: 0.1779270\ttotal: 24m 22s\tremaining: 16m 35s\n",
      "595:\tlearn: 0.1779089\ttotal: 24m 25s\tremaining: 16m 33s\n",
      "596:\tlearn: 0.1778828\ttotal: 24m 27s\tremaining: 16m 30s\n",
      "597:\tlearn: 0.1778588\ttotal: 24m 30s\tremaining: 16m 28s\n",
      "598:\tlearn: 0.1778482\ttotal: 24m 32s\tremaining: 16m 25s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "599:\tlearn: 0.1778298\ttotal: 24m 35s\tremaining: 16m 23s\n",
      "600:\tlearn: 0.1777927\ttotal: 24m 37s\tremaining: 16m 21s\n",
      "601:\tlearn: 0.1777865\ttotal: 24m 39s\tremaining: 16m 18s\n",
      "602:\tlearn: 0.1777495\ttotal: 24m 42s\tremaining: 16m 16s\n",
      "603:\tlearn: 0.1777160\ttotal: 24m 45s\tremaining: 16m 13s\n",
      "604:\tlearn: 0.1776826\ttotal: 24m 47s\tremaining: 16m 11s\n",
      "605:\tlearn: 0.1776788\ttotal: 24m 50s\tremaining: 16m 8s\n",
      "606:\tlearn: 0.1776444\ttotal: 24m 52s\tremaining: 16m 6s\n",
      "607:\tlearn: 0.1776397\ttotal: 24m 55s\tremaining: 16m 3s\n",
      "608:\tlearn: 0.1776064\ttotal: 24m 57s\tremaining: 16m 1s\n",
      "609:\tlearn: 0.1776033\ttotal: 24m 59s\tremaining: 15m 58s\n",
      "610:\tlearn: 0.1775987\ttotal: 25m 2s\tremaining: 15m 56s\n",
      "611:\tlearn: 0.1775829\ttotal: 25m 4s\tremaining: 15m 54s\n",
      "612:\tlearn: 0.1775536\ttotal: 25m 7s\tremaining: 15m 51s\n",
      "613:\tlearn: 0.1775390\ttotal: 25m 9s\tremaining: 15m 49s\n",
      "614:\tlearn: 0.1775285\ttotal: 25m 12s\tremaining: 15m 46s\n",
      "615:\tlearn: 0.1775155\ttotal: 25m 14s\tremaining: 15m 44s\n",
      "616:\tlearn: 0.1775017\ttotal: 25m 17s\tremaining: 15m 41s\n",
      "617:\tlearn: 0.1774935\ttotal: 25m 19s\tremaining: 15m 39s\n",
      "618:\tlearn: 0.1774657\ttotal: 25m 22s\tremaining: 15m 36s\n",
      "619:\tlearn: 0.1774395\ttotal: 25m 24s\tremaining: 15m 34s\n",
      "620:\tlearn: 0.1774224\ttotal: 25m 27s\tremaining: 15m 31s\n",
      "621:\tlearn: 0.1773927\ttotal: 25m 29s\tremaining: 15m 29s\n",
      "622:\tlearn: 0.1773528\ttotal: 25m 31s\tremaining: 15m 26s\n",
      "623:\tlearn: 0.1773398\ttotal: 25m 34s\tremaining: 15m 24s\n",
      "624:\tlearn: 0.1773074\ttotal: 25m 37s\tremaining: 15m 22s\n",
      "625:\tlearn: 0.1773030\ttotal: 25m 39s\tremaining: 15m 19s\n",
      "626:\tlearn: 0.1772859\ttotal: 25m 42s\tremaining: 15m 17s\n",
      "627:\tlearn: 0.1772769\ttotal: 25m 44s\tremaining: 15m 14s\n",
      "628:\tlearn: 0.1772528\ttotal: 25m 47s\tremaining: 15m 12s\n",
      "629:\tlearn: 0.1772322\ttotal: 25m 49s\tremaining: 15m 9s\n",
      "630:\tlearn: 0.1771815\ttotal: 25m 52s\tremaining: 15m 7s\n",
      "631:\tlearn: 0.1771587\ttotal: 25m 54s\tremaining: 15m 5s\n",
      "632:\tlearn: 0.1771335\ttotal: 25m 56s\tremaining: 15m 2s\n",
      "633:\tlearn: 0.1771228\ttotal: 25m 58s\tremaining: 14m 59s\n",
      "634:\tlearn: 0.1771173\ttotal: 26m 1s\tremaining: 14m 57s\n",
      "635:\tlearn: 0.1771129\ttotal: 26m 3s\tremaining: 14m 54s\n",
      "636:\tlearn: 0.1770967\ttotal: 26m 6s\tremaining: 14m 52s\n",
      "637:\tlearn: 0.1770853\ttotal: 26m 8s\tremaining: 14m 50s\n",
      "638:\tlearn: 0.1770755\ttotal: 26m 11s\tremaining: 14m 47s\n",
      "639:\tlearn: 0.1770622\ttotal: 26m 13s\tremaining: 14m 45s\n",
      "640:\tlearn: 0.1770589\ttotal: 26m 16s\tremaining: 14m 42s\n",
      "641:\tlearn: 0.1770469\ttotal: 26m 18s\tremaining: 14m 40s\n",
      "642:\tlearn: 0.1770248\ttotal: 26m 21s\tremaining: 14m 37s\n",
      "643:\tlearn: 0.1770010\ttotal: 26m 23s\tremaining: 14m 35s\n",
      "644:\tlearn: 0.1769974\ttotal: 26m 25s\tremaining: 14m 32s\n",
      "645:\tlearn: 0.1769579\ttotal: 26m 28s\tremaining: 14m 30s\n",
      "646:\tlearn: 0.1769537\ttotal: 26m 30s\tremaining: 14m 27s\n",
      "647:\tlearn: 0.1769378\ttotal: 26m 33s\tremaining: 14m 25s\n",
      "648:\tlearn: 0.1768936\ttotal: 26m 35s\tremaining: 14m 23s\n",
      "649:\tlearn: 0.1768890\ttotal: 26m 38s\tremaining: 14m 20s\n",
      "650:\tlearn: 0.1768855\ttotal: 26m 40s\tremaining: 14m 18s\n",
      "651:\tlearn: 0.1768575\ttotal: 26m 43s\tremaining: 14m 15s\n",
      "652:\tlearn: 0.1768176\ttotal: 26m 45s\tremaining: 14m 13s\n",
      "653:\tlearn: 0.1767592\ttotal: 26m 48s\tremaining: 14m 10s\n",
      "654:\tlearn: 0.1767366\ttotal: 26m 50s\tremaining: 14m 8s\n",
      "655:\tlearn: 0.1767314\ttotal: 26m 52s\tremaining: 14m 5s\n",
      "656:\tlearn: 0.1767191\ttotal: 26m 55s\tremaining: 14m 3s\n",
      "657:\tlearn: 0.1767169\ttotal: 26m 57s\tremaining: 14m\n",
      "658:\tlearn: 0.1766786\ttotal: 27m\tremaining: 13m 58s\n",
      "659:\tlearn: 0.1766431\ttotal: 27m 2s\tremaining: 13m 55s\n",
      "660:\tlearn: 0.1765857\ttotal: 27m 5s\tremaining: 13m 53s\n",
      "661:\tlearn: 0.1765634\ttotal: 27m 7s\tremaining: 13m 51s\n",
      "662:\tlearn: 0.1765286\ttotal: 27m 10s\tremaining: 13m 48s\n",
      "663:\tlearn: 0.1765234\ttotal: 27m 12s\tremaining: 13m 46s\n",
      "664:\tlearn: 0.1764730\ttotal: 27m 15s\tremaining: 13m 43s\n",
      "665:\tlearn: 0.1764522\ttotal: 27m 17s\tremaining: 13m 41s\n",
      "666:\tlearn: 0.1764335\ttotal: 27m 19s\tremaining: 13m 38s\n",
      "667:\tlearn: 0.1764287\ttotal: 27m 22s\tremaining: 13m 36s\n",
      "668:\tlearn: 0.1763923\ttotal: 27m 24s\tremaining: 13m 33s\n",
      "669:\tlearn: 0.1763514\ttotal: 27m 27s\tremaining: 13m 31s\n",
      "670:\tlearn: 0.1763147\ttotal: 27m 29s\tremaining: 13m 28s\n",
      "671:\tlearn: 0.1762886\ttotal: 27m 32s\tremaining: 13m 26s\n",
      "672:\tlearn: 0.1762459\ttotal: 27m 34s\tremaining: 13m 23s\n",
      "673:\tlearn: 0.1762235\ttotal: 27m 37s\tremaining: 13m 21s\n",
      "674:\tlearn: 0.1761946\ttotal: 27m 39s\tremaining: 13m 18s\n",
      "675:\tlearn: 0.1761508\ttotal: 27m 42s\tremaining: 13m 16s\n",
      "676:\tlearn: 0.1761456\ttotal: 27m 44s\tremaining: 13m 13s\n",
      "677:\tlearn: 0.1761144\ttotal: 27m 46s\tremaining: 13m 11s\n",
      "678:\tlearn: 0.1760605\ttotal: 27m 49s\tremaining: 13m 9s\n",
      "679:\tlearn: 0.1760571\ttotal: 27m 51s\tremaining: 13m 6s\n",
      "680:\tlearn: 0.1760241\ttotal: 27m 54s\tremaining: 13m 4s\n",
      "681:\tlearn: 0.1760014\ttotal: 27m 56s\tremaining: 13m 1s\n",
      "682:\tlearn: 0.1759775\ttotal: 27m 58s\tremaining: 12m 59s\n",
      "683:\tlearn: 0.1759408\ttotal: 28m 1s\tremaining: 12m 56s\n",
      "684:\tlearn: 0.1759306\ttotal: 28m 4s\tremaining: 12m 54s\n",
      "685:\tlearn: 0.1758791\ttotal: 28m 6s\tremaining: 12m 52s\n",
      "686:\tlearn: 0.1758643\ttotal: 28m 9s\tremaining: 12m 49s\n",
      "687:\tlearn: 0.1758212\ttotal: 28m 11s\tremaining: 12m 47s\n",
      "688:\tlearn: 0.1757914\ttotal: 28m 14s\tremaining: 12m 44s\n",
      "689:\tlearn: 0.1757788\ttotal: 28m 16s\tremaining: 12m 42s\n",
      "690:\tlearn: 0.1757451\ttotal: 28m 18s\tremaining: 12m 39s\n",
      "691:\tlearn: 0.1757163\ttotal: 28m 21s\tremaining: 12m 37s\n",
      "692:\tlearn: 0.1756915\ttotal: 28m 23s\tremaining: 12m 34s\n",
      "693:\tlearn: 0.1756873\ttotal: 28m 26s\tremaining: 12m 32s\n",
      "694:\tlearn: 0.1756485\ttotal: 28m 28s\tremaining: 12m 29s\n",
      "695:\tlearn: 0.1756348\ttotal: 28m 31s\tremaining: 12m 27s\n",
      "696:\tlearn: 0.1756212\ttotal: 28m 33s\tremaining: 12m 24s\n",
      "697:\tlearn: 0.1755864\ttotal: 28m 36s\tremaining: 12m 22s\n",
      "698:\tlearn: 0.1755770\ttotal: 28m 38s\tremaining: 12m 20s\n",
      "699:\tlearn: 0.1755485\ttotal: 28m 41s\tremaining: 12m 17s\n",
      "700:\tlearn: 0.1755252\ttotal: 28m 43s\tremaining: 12m 15s\n",
      "701:\tlearn: 0.1754980\ttotal: 28m 46s\tremaining: 12m 12s\n",
      "702:\tlearn: 0.1754676\ttotal: 28m 48s\tremaining: 12m 10s\n",
      "703:\tlearn: 0.1754535\ttotal: 28m 51s\tremaining: 12m 7s\n",
      "704:\tlearn: 0.1754339\ttotal: 28m 53s\tremaining: 12m 5s\n",
      "705:\tlearn: 0.1754095\ttotal: 28m 55s\tremaining: 12m 2s\n",
      "706:\tlearn: 0.1753885\ttotal: 28m 57s\tremaining: 12m\n",
      "707:\tlearn: 0.1753777\ttotal: 29m\tremaining: 11m 57s\n",
      "708:\tlearn: 0.1753476\ttotal: 29m 2s\tremaining: 11m 55s\n",
      "709:\tlearn: 0.1753252\ttotal: 29m 5s\tremaining: 11m 52s\n",
      "710:\tlearn: 0.1753200\ttotal: 29m 8s\tremaining: 11m 50s\n",
      "711:\tlearn: 0.1753031\ttotal: 29m 10s\tremaining: 11m 48s\n",
      "712:\tlearn: 0.1752991\ttotal: 29m 12s\tremaining: 11m 45s\n",
      "713:\tlearn: 0.1752854\ttotal: 29m 14s\tremaining: 11m 42s\n",
      "714:\tlearn: 0.1752580\ttotal: 29m 17s\tremaining: 11m 40s\n",
      "715:\tlearn: 0.1752504\ttotal: 29m 19s\tremaining: 11m 38s\n",
      "716:\tlearn: 0.1752219\ttotal: 29m 22s\tremaining: 11m 35s\n",
      "717:\tlearn: 0.1752174\ttotal: 29m 24s\tremaining: 11m 33s\n",
      "718:\tlearn: 0.1752127\ttotal: 29m 27s\tremaining: 11m 30s\n",
      "719:\tlearn: 0.1751643\ttotal: 29m 29s\tremaining: 11m 28s\n",
      "720:\tlearn: 0.1751526\ttotal: 29m 32s\tremaining: 11m 25s\n",
      "721:\tlearn: 0.1751366\ttotal: 29m 34s\tremaining: 11m 23s\n",
      "722:\tlearn: 0.1751185\ttotal: 29m 37s\tremaining: 11m 20s\n",
      "723:\tlearn: 0.1751136\ttotal: 29m 39s\tremaining: 11m 18s\n",
      "724:\tlearn: 0.1750737\ttotal: 29m 42s\tremaining: 11m 16s\n",
      "725:\tlearn: 0.1750595\ttotal: 29m 44s\tremaining: 11m 13s\n",
      "726:\tlearn: 0.1750174\ttotal: 29m 47s\tremaining: 11m 11s\n",
      "727:\tlearn: 0.1749805\ttotal: 29m 49s\tremaining: 11m 8s\n",
      "728:\tlearn: 0.1749349\ttotal: 29m 52s\tremaining: 11m 6s\n",
      "729:\tlearn: 0.1749335\ttotal: 29m 54s\tremaining: 11m 3s\n",
      "730:\tlearn: 0.1749025\ttotal: 29m 57s\tremaining: 11m 1s\n",
      "731:\tlearn: 0.1748899\ttotal: 29m 59s\tremaining: 10m 58s\n",
      "732:\tlearn: 0.1748722\ttotal: 30m 1s\tremaining: 10m 56s\n",
      "733:\tlearn: 0.1748658\ttotal: 30m 4s\tremaining: 10m 53s\n",
      "734:\tlearn: 0.1748305\ttotal: 30m 7s\tremaining: 10m 51s\n",
      "735:\tlearn: 0.1748120\ttotal: 30m 9s\tremaining: 10m 49s\n",
      "736:\tlearn: 0.1747669\ttotal: 30m 12s\tremaining: 10m 46s\n",
      "737:\tlearn: 0.1747394\ttotal: 30m 14s\tremaining: 10m 44s\n",
      "738:\tlearn: 0.1747351\ttotal: 30m 16s\tremaining: 10m 41s\n",
      "739:\tlearn: 0.1747221\ttotal: 30m 19s\tremaining: 10m 39s\n",
      "740:\tlearn: 0.1746764\ttotal: 30m 21s\tremaining: 10m 36s\n",
      "741:\tlearn: 0.1746311\ttotal: 30m 24s\tremaining: 10m 34s\n",
      "742:\tlearn: 0.1745863\ttotal: 30m 26s\tremaining: 10m 31s\n",
      "743:\tlearn: 0.1745550\ttotal: 30m 28s\tremaining: 10m 29s\n",
      "744:\tlearn: 0.1745410\ttotal: 30m 31s\tremaining: 10m 26s\n",
      "745:\tlearn: 0.1745233\ttotal: 30m 34s\tremaining: 10m 24s\n",
      "746:\tlearn: 0.1744876\ttotal: 30m 36s\tremaining: 10m 21s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "747:\tlearn: 0.1744705\ttotal: 30m 39s\tremaining: 10m 19s\n",
      "748:\tlearn: 0.1744398\ttotal: 30m 41s\tremaining: 10m 17s\n",
      "749:\tlearn: 0.1743737\ttotal: 30m 43s\tremaining: 10m 14s\n",
      "750:\tlearn: 0.1743699\ttotal: 30m 46s\tremaining: 10m 12s\n",
      "751:\tlearn: 0.1743543\ttotal: 30m 48s\tremaining: 10m 9s\n",
      "752:\tlearn: 0.1743334\ttotal: 30m 51s\tremaining: 10m 7s\n",
      "753:\tlearn: 0.1743214\ttotal: 30m 53s\tremaining: 10m 4s\n",
      "754:\tlearn: 0.1743147\ttotal: 30m 56s\tremaining: 10m 2s\n",
      "755:\tlearn: 0.1742911\ttotal: 30m 58s\tremaining: 9m 59s\n",
      "756:\tlearn: 0.1742757\ttotal: 31m\tremaining: 9m 57s\n",
      "757:\tlearn: 0.1742404\ttotal: 31m 3s\tremaining: 9m 54s\n",
      "758:\tlearn: 0.1741884\ttotal: 31m 5s\tremaining: 9m 52s\n",
      "759:\tlearn: 0.1741847\ttotal: 31m 8s\tremaining: 9m 49s\n",
      "760:\tlearn: 0.1741676\ttotal: 31m 10s\tremaining: 9m 47s\n",
      "761:\tlearn: 0.1741241\ttotal: 31m 13s\tremaining: 9m 45s\n",
      "762:\tlearn: 0.1741218\ttotal: 31m 16s\tremaining: 9m 42s\n",
      "763:\tlearn: 0.1741122\ttotal: 31m 18s\tremaining: 9m 40s\n",
      "764:\tlearn: 0.1740895\ttotal: 31m 20s\tremaining: 9m 37s\n",
      "765:\tlearn: 0.1740800\ttotal: 31m 22s\tremaining: 9m 35s\n",
      "766:\tlearn: 0.1740602\ttotal: 31m 25s\tremaining: 9m 32s\n",
      "767:\tlearn: 0.1740195\ttotal: 31m 27s\tremaining: 9m 30s\n",
      "768:\tlearn: 0.1739844\ttotal: 31m 30s\tremaining: 9m 27s\n",
      "769:\tlearn: 0.1739794\ttotal: 31m 32s\tremaining: 9m 25s\n",
      "770:\tlearn: 0.1739474\ttotal: 31m 35s\tremaining: 9m 22s\n",
      "771:\tlearn: 0.1739051\ttotal: 31m 37s\tremaining: 9m 20s\n",
      "772:\tlearn: 0.1739012\ttotal: 31m 40s\tremaining: 9m 18s\n",
      "773:\tlearn: 0.1738746\ttotal: 31m 42s\tremaining: 9m 15s\n",
      "774:\tlearn: 0.1738702\ttotal: 31m 45s\tremaining: 9m 13s\n",
      "775:\tlearn: 0.1738691\ttotal: 31m 47s\tremaining: 9m 10s\n",
      "776:\tlearn: 0.1738035\ttotal: 31m 49s\tremaining: 9m 8s\n",
      "777:\tlearn: 0.1738001\ttotal: 31m 52s\tremaining: 9m 5s\n",
      "778:\tlearn: 0.1737608\ttotal: 31m 54s\tremaining: 9m 3s\n",
      "779:\tlearn: 0.1737567\ttotal: 31m 56s\tremaining: 9m\n",
      "780:\tlearn: 0.1737367\ttotal: 31m 59s\tremaining: 8m 58s\n",
      "781:\tlearn: 0.1737317\ttotal: 32m 1s\tremaining: 8m 55s\n",
      "782:\tlearn: 0.1737156\ttotal: 32m 3s\tremaining: 8m 53s\n",
      "783:\tlearn: 0.1737113\ttotal: 32m 6s\tremaining: 8m 50s\n",
      "784:\tlearn: 0.1737060\ttotal: 32m 8s\tremaining: 8m 48s\n",
      "785:\tlearn: 0.1737019\ttotal: 32m 11s\tremaining: 8m 45s\n",
      "786:\tlearn: 0.1736878\ttotal: 32m 13s\tremaining: 8m 43s\n",
      "787:\tlearn: 0.1736699\ttotal: 32m 16s\tremaining: 8m 41s\n",
      "788:\tlearn: 0.1736644\ttotal: 32m 19s\tremaining: 8m 38s\n",
      "789:\tlearn: 0.1736026\ttotal: 32m 21s\tremaining: 8m 36s\n",
      "790:\tlearn: 0.1735845\ttotal: 32m 23s\tremaining: 8m 33s\n",
      "791:\tlearn: 0.1735476\ttotal: 32m 26s\tremaining: 8m 31s\n",
      "792:\tlearn: 0.1735289\ttotal: 32m 28s\tremaining: 8m 28s\n",
      "793:\tlearn: 0.1735031\ttotal: 32m 31s\tremaining: 8m 26s\n",
      "794:\tlearn: 0.1734567\ttotal: 32m 33s\tremaining: 8m 23s\n",
      "795:\tlearn: 0.1734513\ttotal: 32m 36s\tremaining: 8m 21s\n",
      "796:\tlearn: 0.1734273\ttotal: 32m 38s\tremaining: 8m 18s\n",
      "797:\tlearn: 0.1733720\ttotal: 32m 40s\tremaining: 8m 16s\n",
      "798:\tlearn: 0.1733675\ttotal: 32m 43s\tremaining: 8m 14s\n",
      "799:\tlearn: 0.1733530\ttotal: 32m 45s\tremaining: 8m 11s\n",
      "800:\tlearn: 0.1733105\ttotal: 32m 48s\tremaining: 8m 9s\n",
      "801:\tlearn: 0.1732672\ttotal: 32m 50s\tremaining: 8m 6s\n",
      "802:\tlearn: 0.1732245\ttotal: 32m 53s\tremaining: 8m 4s\n",
      "803:\tlearn: 0.1732144\ttotal: 32m 56s\tremaining: 8m 1s\n",
      "804:\tlearn: 0.1732113\ttotal: 32m 58s\tremaining: 7m 59s\n",
      "805:\tlearn: 0.1731924\ttotal: 33m\tremaining: 7m 56s\n",
      "806:\tlearn: 0.1731748\ttotal: 33m 3s\tremaining: 7m 54s\n",
      "807:\tlearn: 0.1731257\ttotal: 33m 6s\tremaining: 7m 51s\n",
      "808:\tlearn: 0.1730817\ttotal: 33m 8s\tremaining: 7m 49s\n",
      "809:\tlearn: 0.1730767\ttotal: 33m 11s\tremaining: 7m 47s\n",
      "810:\tlearn: 0.1730722\ttotal: 33m 13s\tremaining: 7m 44s\n",
      "811:\tlearn: 0.1730521\ttotal: 33m 16s\tremaining: 7m 42s\n",
      "812:\tlearn: 0.1730392\ttotal: 33m 18s\tremaining: 7m 39s\n",
      "813:\tlearn: 0.1730108\ttotal: 33m 21s\tremaining: 7m 37s\n",
      "814:\tlearn: 0.1729698\ttotal: 33m 23s\tremaining: 7m 34s\n",
      "815:\tlearn: 0.1729601\ttotal: 33m 26s\tremaining: 7m 32s\n",
      "816:\tlearn: 0.1729318\ttotal: 33m 28s\tremaining: 7m 29s\n",
      "817:\tlearn: 0.1729272\ttotal: 33m 30s\tremaining: 7m 27s\n",
      "818:\tlearn: 0.1729039\ttotal: 33m 33s\tremaining: 7m 25s\n",
      "819:\tlearn: 0.1728955\ttotal: 33m 36s\tremaining: 7m 22s\n",
      "820:\tlearn: 0.1728911\ttotal: 33m 38s\tremaining: 7m 20s\n",
      "821:\tlearn: 0.1728837\ttotal: 33m 41s\tremaining: 7m 17s\n",
      "822:\tlearn: 0.1728296\ttotal: 33m 43s\tremaining: 7m 15s\n",
      "823:\tlearn: 0.1728151\ttotal: 33m 46s\tremaining: 7m 12s\n",
      "824:\tlearn: 0.1728087\ttotal: 33m 48s\tremaining: 7m 10s\n",
      "825:\tlearn: 0.1727690\ttotal: 33m 50s\tremaining: 7m 7s\n",
      "826:\tlearn: 0.1727613\ttotal: 33m 53s\tremaining: 7m 5s\n",
      "827:\tlearn: 0.1727318\ttotal: 33m 55s\tremaining: 7m 2s\n",
      "828:\tlearn: 0.1727128\ttotal: 33m 58s\tremaining: 7m\n",
      "829:\tlearn: 0.1726753\ttotal: 34m\tremaining: 6m 58s\n",
      "830:\tlearn: 0.1726234\ttotal: 34m 3s\tremaining: 6m 55s\n",
      "831:\tlearn: 0.1725961\ttotal: 34m 5s\tremaining: 6m 53s\n",
      "832:\tlearn: 0.1725915\ttotal: 34m 8s\tremaining: 6m 50s\n",
      "833:\tlearn: 0.1725717\ttotal: 34m 10s\tremaining: 6m 48s\n",
      "834:\tlearn: 0.1725525\ttotal: 34m 13s\tremaining: 6m 45s\n",
      "835:\tlearn: 0.1725107\ttotal: 34m 15s\tremaining: 6m 43s\n",
      "836:\tlearn: 0.1724955\ttotal: 34m 17s\tremaining: 6m 40s\n",
      "837:\tlearn: 0.1724686\ttotal: 34m 19s\tremaining: 6m 38s\n",
      "838:\tlearn: 0.1724653\ttotal: 34m 22s\tremaining: 6m 35s\n",
      "839:\tlearn: 0.1724321\ttotal: 34m 24s\tremaining: 6m 33s\n",
      "840:\tlearn: 0.1724075\ttotal: 34m 27s\tremaining: 6m 30s\n",
      "841:\tlearn: 0.1723690\ttotal: 34m 29s\tremaining: 6m 28s\n",
      "842:\tlearn: 0.1723554\ttotal: 34m 32s\tremaining: 6m 25s\n",
      "843:\tlearn: 0.1723519\ttotal: 34m 34s\tremaining: 6m 23s\n",
      "844:\tlearn: 0.1723471\ttotal: 34m 37s\tremaining: 6m 21s\n",
      "845:\tlearn: 0.1723418\ttotal: 34m 39s\tremaining: 6m 18s\n",
      "846:\tlearn: 0.1723351\ttotal: 34m 42s\tremaining: 6m 16s\n",
      "847:\tlearn: 0.1723320\ttotal: 34m 44s\tremaining: 6m 13s\n",
      "848:\tlearn: 0.1723009\ttotal: 34m 47s\tremaining: 6m 11s\n",
      "849:\tlearn: 0.1722851\ttotal: 34m 49s\tremaining: 6m 8s\n",
      "850:\tlearn: 0.1722640\ttotal: 34m 51s\tremaining: 6m 6s\n",
      "851:\tlearn: 0.1722563\ttotal: 34m 54s\tremaining: 6m 3s\n",
      "852:\tlearn: 0.1722531\ttotal: 34m 56s\tremaining: 6m 1s\n",
      "853:\tlearn: 0.1722223\ttotal: 34m 59s\tremaining: 5m 58s\n",
      "854:\tlearn: 0.1722108\ttotal: 35m 1s\tremaining: 5m 56s\n",
      "855:\tlearn: 0.1721986\ttotal: 35m 4s\tremaining: 5m 54s\n",
      "856:\tlearn: 0.1721892\ttotal: 35m 6s\tremaining: 5m 51s\n",
      "857:\tlearn: 0.1721858\ttotal: 35m 9s\tremaining: 5m 49s\n",
      "858:\tlearn: 0.1721656\ttotal: 35m 12s\tremaining: 5m 46s\n",
      "859:\tlearn: 0.1721264\ttotal: 35m 14s\tremaining: 5m 44s\n",
      "860:\tlearn: 0.1721218\ttotal: 35m 16s\tremaining: 5m 41s\n",
      "861:\tlearn: 0.1721006\ttotal: 35m 19s\tremaining: 5m 39s\n",
      "862:\tlearn: 0.1720910\ttotal: 35m 21s\tremaining: 5m 36s\n",
      "863:\tlearn: 0.1720860\ttotal: 35m 24s\tremaining: 5m 34s\n",
      "864:\tlearn: 0.1720831\ttotal: 35m 26s\tremaining: 5m 31s\n",
      "865:\tlearn: 0.1720797\ttotal: 35m 29s\tremaining: 5m 29s\n",
      "866:\tlearn: 0.1720657\ttotal: 35m 31s\tremaining: 5m 27s\n",
      "867:\tlearn: 0.1720537\ttotal: 35m 33s\tremaining: 5m 24s\n",
      "868:\tlearn: 0.1720266\ttotal: 35m 36s\tremaining: 5m 22s\n",
      "869:\tlearn: 0.1720130\ttotal: 35m 39s\tremaining: 5m 19s\n",
      "870:\tlearn: 0.1720098\ttotal: 35m 41s\tremaining: 5m 17s\n",
      "871:\tlearn: 0.1719914\ttotal: 35m 44s\tremaining: 5m 14s\n",
      "872:\tlearn: 0.1719835\ttotal: 35m 46s\tremaining: 5m 12s\n",
      "873:\tlearn: 0.1719538\ttotal: 35m 48s\tremaining: 5m 9s\n",
      "874:\tlearn: 0.1719177\ttotal: 35m 51s\tremaining: 5m 7s\n",
      "875:\tlearn: 0.1718880\ttotal: 35m 53s\tremaining: 5m 4s\n",
      "876:\tlearn: 0.1718453\ttotal: 35m 56s\tremaining: 5m 2s\n",
      "877:\tlearn: 0.1717994\ttotal: 35m 58s\tremaining: 4m 59s\n",
      "878:\tlearn: 0.1717911\ttotal: 36m\tremaining: 4m 57s\n",
      "879:\tlearn: 0.1717795\ttotal: 36m 3s\tremaining: 4m 55s\n",
      "880:\tlearn: 0.1717649\ttotal: 36m 5s\tremaining: 4m 52s\n",
      "881:\tlearn: 0.1717374\ttotal: 36m 8s\tremaining: 4m 50s\n",
      "882:\tlearn: 0.1717095\ttotal: 36m 11s\tremaining: 4m 47s\n",
      "883:\tlearn: 0.1716991\ttotal: 36m 13s\tremaining: 4m 45s\n",
      "884:\tlearn: 0.1716812\ttotal: 36m 15s\tremaining: 4m 42s\n",
      "885:\tlearn: 0.1716492\ttotal: 36m 17s\tremaining: 4m 40s\n",
      "886:\tlearn: 0.1716246\ttotal: 36m 20s\tremaining: 4m 37s\n",
      "887:\tlearn: 0.1716223\ttotal: 36m 23s\tremaining: 4m 35s\n",
      "888:\tlearn: 0.1716198\ttotal: 36m 25s\tremaining: 4m 32s\n",
      "889:\tlearn: 0.1716107\ttotal: 36m 27s\tremaining: 4m 30s\n",
      "890:\tlearn: 0.1715901\ttotal: 36m 29s\tremaining: 4m 27s\n",
      "891:\tlearn: 0.1715698\ttotal: 36m 32s\tremaining: 4m 25s\n",
      "892:\tlearn: 0.1715665\ttotal: 36m 34s\tremaining: 4m 22s\n",
      "893:\tlearn: 0.1715384\ttotal: 36m 37s\tremaining: 4m 20s\n",
      "894:\tlearn: 0.1715338\ttotal: 36m 40s\tremaining: 4m 18s\n",
      "895:\tlearn: 0.1714851\ttotal: 36m 42s\tremaining: 4m 15s\n",
      "896:\tlearn: 0.1714715\ttotal: 36m 45s\tremaining: 4m 13s\n",
      "897:\tlearn: 0.1714660\ttotal: 36m 47s\tremaining: 4m 10s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "898:\tlearn: 0.1714532\ttotal: 36m 50s\tremaining: 4m 8s\n",
      "899:\tlearn: 0.1714444\ttotal: 36m 52s\tremaining: 4m 5s\n",
      "900:\tlearn: 0.1714361\ttotal: 36m 54s\tremaining: 4m 3s\n",
      "901:\tlearn: 0.1714061\ttotal: 36m 56s\tremaining: 4m\n",
      "902:\tlearn: 0.1713836\ttotal: 36m 59s\tremaining: 3m 58s\n",
      "903:\tlearn: 0.1713671\ttotal: 37m 2s\tremaining: 3m 55s\n",
      "904:\tlearn: 0.1713425\ttotal: 37m 4s\tremaining: 3m 53s\n",
      "905:\tlearn: 0.1713129\ttotal: 37m 7s\tremaining: 3m 51s\n",
      "906:\tlearn: 0.1713058\ttotal: 37m 9s\tremaining: 3m 48s\n",
      "907:\tlearn: 0.1713029\ttotal: 37m 12s\tremaining: 3m 46s\n",
      "908:\tlearn: 0.1712893\ttotal: 37m 14s\tremaining: 3m 43s\n",
      "909:\tlearn: 0.1712700\ttotal: 37m 17s\tremaining: 3m 41s\n",
      "910:\tlearn: 0.1712570\ttotal: 37m 19s\tremaining: 3m 38s\n",
      "911:\tlearn: 0.1712469\ttotal: 37m 21s\tremaining: 3m 36s\n",
      "912:\tlearn: 0.1712342\ttotal: 37m 24s\tremaining: 3m 33s\n",
      "913:\tlearn: 0.1712204\ttotal: 37m 26s\tremaining: 3m 31s\n",
      "914:\tlearn: 0.1712114\ttotal: 37m 29s\tremaining: 3m 28s\n",
      "915:\tlearn: 0.1711922\ttotal: 37m 31s\tremaining: 3m 26s\n",
      "916:\tlearn: 0.1711872\ttotal: 37m 34s\tremaining: 3m 24s\n",
      "917:\tlearn: 0.1711494\ttotal: 37m 36s\tremaining: 3m 21s\n",
      "918:\tlearn: 0.1711397\ttotal: 37m 39s\tremaining: 3m 19s\n",
      "919:\tlearn: 0.1711342\ttotal: 37m 41s\tremaining: 3m 16s\n",
      "920:\tlearn: 0.1711286\ttotal: 37m 44s\tremaining: 3m 14s\n",
      "921:\tlearn: 0.1711233\ttotal: 37m 46s\tremaining: 3m 11s\n",
      "922:\tlearn: 0.1710868\ttotal: 37m 49s\tremaining: 3m 9s\n",
      "923:\tlearn: 0.1710061\ttotal: 37m 51s\tremaining: 3m 6s\n",
      "924:\tlearn: 0.1710007\ttotal: 37m 53s\tremaining: 3m 4s\n",
      "925:\tlearn: 0.1709904\ttotal: 37m 56s\tremaining: 3m 1s\n",
      "926:\tlearn: 0.1709656\ttotal: 37m 58s\tremaining: 2m 59s\n",
      "927:\tlearn: 0.1709508\ttotal: 38m 1s\tremaining: 2m 57s\n",
      "928:\tlearn: 0.1709433\ttotal: 38m 3s\tremaining: 2m 54s\n",
      "929:\tlearn: 0.1709296\ttotal: 38m 6s\tremaining: 2m 52s\n",
      "930:\tlearn: 0.1709042\ttotal: 38m 8s\tremaining: 2m 49s\n",
      "931:\tlearn: 0.1708843\ttotal: 38m 11s\tremaining: 2m 47s\n",
      "932:\tlearn: 0.1708671\ttotal: 38m 13s\tremaining: 2m 44s\n",
      "933:\tlearn: 0.1708612\ttotal: 38m 15s\tremaining: 2m 42s\n",
      "934:\tlearn: 0.1708445\ttotal: 38m 18s\tremaining: 2m 39s\n",
      "935:\tlearn: 0.1708315\ttotal: 38m 20s\tremaining: 2m 37s\n",
      "936:\tlearn: 0.1707911\ttotal: 38m 23s\tremaining: 2m 34s\n",
      "937:\tlearn: 0.1707837\ttotal: 38m 26s\tremaining: 2m 32s\n",
      "938:\tlearn: 0.1707684\ttotal: 38m 28s\tremaining: 2m 29s\n",
      "939:\tlearn: 0.1707401\ttotal: 38m 30s\tremaining: 2m 27s\n",
      "940:\tlearn: 0.1706954\ttotal: 38m 33s\tremaining: 2m 25s\n",
      "941:\tlearn: 0.1706817\ttotal: 38m 35s\tremaining: 2m 22s\n",
      "942:\tlearn: 0.1706516\ttotal: 38m 38s\tremaining: 2m 20s\n",
      "943:\tlearn: 0.1706334\ttotal: 38m 40s\tremaining: 2m 17s\n",
      "944:\tlearn: 0.1705950\ttotal: 38m 43s\tremaining: 2m 15s\n",
      "945:\tlearn: 0.1705701\ttotal: 38m 45s\tremaining: 2m 12s\n",
      "946:\tlearn: 0.1705450\ttotal: 38m 48s\tremaining: 2m 10s\n",
      "947:\tlearn: 0.1705414\ttotal: 38m 50s\tremaining: 2m 7s\n",
      "948:\tlearn: 0.1705383\ttotal: 38m 52s\tremaining: 2m 5s\n",
      "949:\tlearn: 0.1705340\ttotal: 38m 55s\tremaining: 2m 2s\n",
      "950:\tlearn: 0.1705313\ttotal: 38m 57s\tremaining: 2m\n",
      "951:\tlearn: 0.1705179\ttotal: 39m\tremaining: 1m 58s\n",
      "952:\tlearn: 0.1705123\ttotal: 39m 2s\tremaining: 1m 55s\n",
      "953:\tlearn: 0.1705088\ttotal: 39m 5s\tremaining: 1m 53s\n",
      "954:\tlearn: 0.1704888\ttotal: 39m 7s\tremaining: 1m 50s\n",
      "955:\tlearn: 0.1704756\ttotal: 39m 10s\tremaining: 1m 48s\n",
      "956:\tlearn: 0.1704651\ttotal: 39m 12s\tremaining: 1m 45s\n",
      "957:\tlearn: 0.1704442\ttotal: 39m 14s\tremaining: 1m 43s\n",
      "958:\tlearn: 0.1704183\ttotal: 39m 17s\tremaining: 1m 40s\n",
      "959:\tlearn: 0.1704158\ttotal: 39m 19s\tremaining: 1m 38s\n",
      "960:\tlearn: 0.1704082\ttotal: 39m 22s\tremaining: 1m 35s\n",
      "961:\tlearn: 0.1703909\ttotal: 39m 24s\tremaining: 1m 33s\n",
      "962:\tlearn: 0.1703737\ttotal: 39m 27s\tremaining: 1m 30s\n",
      "963:\tlearn: 0.1703370\ttotal: 39m 29s\tremaining: 1m 28s\n",
      "964:\tlearn: 0.1703338\ttotal: 39m 32s\tremaining: 1m 26s\n",
      "965:\tlearn: 0.1703284\ttotal: 39m 34s\tremaining: 1m 23s\n",
      "966:\tlearn: 0.1702933\ttotal: 39m 37s\tremaining: 1m 21s\n",
      "967:\tlearn: 0.1702753\ttotal: 39m 39s\tremaining: 1m 18s\n",
      "968:\tlearn: 0.1702440\ttotal: 39m 42s\tremaining: 1m 16s\n",
      "969:\tlearn: 0.1702059\ttotal: 39m 44s\tremaining: 1m 13s\n",
      "970:\tlearn: 0.1702013\ttotal: 39m 47s\tremaining: 1m 11s\n",
      "971:\tlearn: 0.1701981\ttotal: 39m 49s\tremaining: 1m 8s\n",
      "972:\tlearn: 0.1701851\ttotal: 39m 51s\tremaining: 1m 6s\n",
      "973:\tlearn: 0.1701620\ttotal: 39m 54s\tremaining: 1m 3s\n",
      "974:\tlearn: 0.1701590\ttotal: 39m 56s\tremaining: 1m 1s\n",
      "975:\tlearn: 0.1701252\ttotal: 39m 59s\tremaining: 59s\n",
      "976:\tlearn: 0.1700938\ttotal: 40m 1s\tremaining: 56.5s\n",
      "977:\tlearn: 0.1700905\ttotal: 40m 4s\tremaining: 54.1s\n",
      "978:\tlearn: 0.1700863\ttotal: 40m 6s\tremaining: 51.6s\n",
      "979:\tlearn: 0.1700685\ttotal: 40m 9s\tremaining: 49.2s\n",
      "980:\tlearn: 0.1700290\ttotal: 40m 11s\tremaining: 46.7s\n",
      "981:\tlearn: 0.1700043\ttotal: 40m 14s\tremaining: 44.3s\n",
      "982:\tlearn: 0.1699759\ttotal: 40m 16s\tremaining: 41.8s\n",
      "983:\tlearn: 0.1699698\ttotal: 40m 19s\tremaining: 39.3s\n",
      "984:\tlearn: 0.1699638\ttotal: 40m 21s\tremaining: 36.9s\n",
      "985:\tlearn: 0.1699552\ttotal: 40m 24s\tremaining: 34.4s\n",
      "986:\tlearn: 0.1699347\ttotal: 40m 26s\tremaining: 32s\n",
      "987:\tlearn: 0.1699048\ttotal: 40m 29s\tremaining: 29.5s\n",
      "988:\tlearn: 0.1698914\ttotal: 40m 31s\tremaining: 27s\n",
      "989:\tlearn: 0.1698886\ttotal: 40m 33s\tremaining: 24.6s\n",
      "990:\tlearn: 0.1698835\ttotal: 40m 36s\tremaining: 22.1s\n",
      "991:\tlearn: 0.1698755\ttotal: 40m 38s\tremaining: 19.7s\n",
      "992:\tlearn: 0.1698369\ttotal: 40m 41s\tremaining: 17.2s\n",
      "993:\tlearn: 0.1698198\ttotal: 40m 43s\tremaining: 14.8s\n",
      "994:\tlearn: 0.1697923\ttotal: 40m 46s\tremaining: 12.3s\n",
      "995:\tlearn: 0.1697860\ttotal: 40m 48s\tremaining: 9.84s\n",
      "996:\tlearn: 0.1697755\ttotal: 40m 51s\tremaining: 7.38s\n",
      "997:\tlearn: 0.1697734\ttotal: 40m 53s\tremaining: 4.92s\n",
      "998:\tlearn: 0.1697286\ttotal: 40m 56s\tremaining: 2.46s\n",
      "999:\tlearn: 0.1697253\ttotal: 40m 58s\tremaining: 0us\n",
      "Training Data metrics\n",
      "accuracy_score=0.9470761542909306 and f1_score=0.1637266253826429\n",
      "Confusion Matrix\n",
      "[[6349786    1866]\n",
      " [ 354920   34926]]\n",
      "\n",
      "\n",
      "classification report\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.95      1.00      0.97   6351652\n",
      "          1       0.95      0.09      0.16    389846\n",
      "\n",
      "avg / total       0.95      0.95      0.93   6741498\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "classifier_cb=prediction(CB,X,y,X_test,submission)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x25b8f77e4a8>"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "importance_cb=classifier_cb.feature_importances_\n",
    "imp_cb=pd.DataFrame(importance_cb,index=X.columns)\n",
    "imp.plot(kind='bar')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}