training.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "988ae92c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import seaborn as sns\n",
    "import scipy.stats as stats\n",
    "from scipy.stats import boxcox, yeojohnson\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "0e9301f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "tf.config.list_physical_devices('GPU')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "df323cfb",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"ready4training.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1936b29a",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.drop('popularity', axis=1)\n",
    "y = data['popularity']\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "651a907c",
   "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>explicit</th>\n",
       "      <th>danceability</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>valence</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>key_1</th>\n",
       "      <th>key_2</th>\n",
       "      <th>key_3</th>\n",
       "      <th>key_4</th>\n",
       "      <th>key_5</th>\n",
       "      <th>key_6</th>\n",
       "      <th>key_7</th>\n",
       "      <th>key_8</th>\n",
       "      <th>key_9</th>\n",
       "      <th>key_10</th>\n",
       "      <th>key_11</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>97547</th>\n",
       "      <td>1</td>\n",
       "      <td>0.204126</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.398034</td>\n",
       "      <td>-0.205937</td>\n",
       "      <td>-0.503064</td>\n",
       "      <td>-0.298645</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27380</th>\n",
       "      <td>0</td>\n",
       "      <td>-0.991936</td>\n",
       "      <td>1</td>\n",
       "      <td>4.673269</td>\n",
       "      <td>-0.929519</td>\n",
       "      <td>0.624991</td>\n",
       "      <td>-0.962898</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96609</th>\n",
       "      <td>0</td>\n",
       "      <td>0.175096</td>\n",
       "      <td>1</td>\n",
       "      <td>1.395141</td>\n",
       "      <td>-0.473042</td>\n",
       "      <td>-0.503049</td>\n",
       "      <td>-0.959036</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86131</th>\n",
       "      <td>0</td>\n",
       "      <td>-0.469385</td>\n",
       "      <td>1</td>\n",
       "      <td>2.727288</td>\n",
       "      <td>0.188650</td>\n",
       "      <td>-0.502815</td>\n",
       "      <td>0.809730</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21611</th>\n",
       "      <td>0</td>\n",
       "      <td>-1.061610</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.054549</td>\n",
       "      <td>-0.937031</td>\n",
       "      <td>-0.046007</td>\n",
       "      <td>-0.935864</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       explicit  danceability  mode  speechiness  acousticness  \\\n",
       "97547         1      0.204126     1    -0.398034     -0.205937   \n",
       "27380         0     -0.991936     1     4.673269     -0.929519   \n",
       "96609         0      0.175096     1     1.395141     -0.473042   \n",
       "86131         0     -0.469385     1     2.727288      0.188650   \n",
       "21611         0     -1.061610     0    -0.054549     -0.937031   \n",
       "\n",
       "       instrumentalness   valence  time_signature  key_1  key_2  key_3  key_4  \\\n",
       "97547         -0.503064 -0.298645               4      0      0      0      0   \n",
       "27380          0.624991 -0.962898               4      0      1      0      0   \n",
       "96609         -0.503049 -0.959036               3      0      0      0      0   \n",
       "86131         -0.502815  0.809730               4      0      0      0      0   \n",
       "21611         -0.046007 -0.935864               3      0      0      0      1   \n",
       "\n",
       "       key_5  key_6  key_7  key_8  key_9  key_10  key_11  \n",
       "97547      1      0      0      0      0       0       0  \n",
       "27380      0      0      0      0      0       0       0  \n",
       "96609      0      0      1      0      0       0       0  \n",
       "86131      0      0      1      0      0       0       0  \n",
       "21611      0      0      0      0      0       0       0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "005514f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "from sklearn.metrics import mean_absolute_error, mean_squared_error, mean_absolute_percentage_error, r2_score\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44049a4d",
   "metadata": {},
   "source": [
    "<a id='1'>1 Обучение линейной регрессии</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6a35baf2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 488.8996982837067\n",
      "    mae -> 18.51529321190871\n",
      "    rmse -> 22.111076370988968\n",
      "    mape -> 2.152193716656925e+16\n",
      "    r2 -> 0.023323197042441457\n",
      "    \n",
      "{'alpha': 10}\n"
     ]
    }
   ],
   "source": [
    "params = {\n",
    "    \"alpha\": [1e-3, 1e-2, 0.1, 1, 10, 1e3, 1e5],\n",
    "}\n",
    "model = Ridge()\n",
    "ridge_regressor = GridSearchCV(estimator=model, param_grid=params, scoring=['neg_mean_squared_error', 'neg_mean_absolute_error'], refit='neg_mean_squared_error', n_jobs=-1)\n",
    "ridge_regressor.fit(X_train, y_train)\n",
    "preds = ridge_regressor.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")\n",
    "print(ridge_regressor.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2707f8fa",
   "metadata": {},
   "source": [
    "<a id='2'>2 Обучение SVM регрессии</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "19244b40",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.8/dist-packages/sklearn/svm/_classes.py:32: FutureWarning: The default value of `dual` will change from `True` to `'auto'` in 1.5. Set the value of `dual` explicitly to suppress the warning.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 498.8031019752922\n",
      "    mae -> 18.390421488262053\n",
      "    rmse -> 22.33390028578287\n",
      "    mape -> 2.279590079860916e+16\n",
      "    r2 -> 0.003539129492694526\n",
      "    \n",
      "{'C': 0.46415888336127775}\n"
     ]
    }
   ],
   "source": [
    "from sklearn import svm\n",
    "\n",
    "params = {\n",
    "    \"C\": np.logspace(start=-2, stop=1, num=10)\n",
    "}\n",
    "\n",
    "model = GridSearchCV(estimator=svm.LinearSVR(random_state=42), param_grid=params, scoring=['neg_mean_squared_error', 'neg_mean_absolute_error'], refit='neg_mean_squared_error', n_jobs=-1)\n",
    "model.fit(X_train, y_train)\n",
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")\n",
    "print(model.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44b233c5",
   "metadata": {},
   "source": [
    "<a id='3'>3 Обучение DecisionTree регрессии</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "781446d9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "~~~test results~~~\n",
      "\n",
      "        mse -> 448.3968790332429\n",
      "        mae -> 13.570147063528202\n",
      "        rmse -> 21.17538379895965\n",
      "        mape -> 5389375926984328.0\n",
      "        r2 -> 0.10423583445085194\n",
      "        \n",
      "~~~train results~~~\n",
      "\n",
      "        mse -> 20.176966799787934\n",
      "        mae -> 0.6945978381630007\n",
      "        rmse -> 4.491877870088181\n",
      "        mape -> 897357824417728.2\n",
      "        r2 -> 0.9597708211339882\n",
      "        \n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor\n",
    "def print_results(preds, yhat):\n",
    "    mse = mean_squared_error(yhat, preds)\n",
    "    mae = mean_absolute_error(yhat, preds)\n",
    "    r2 = r2_score(yhat, preds)\n",
    "    rmse = np.sqrt(mse)\n",
    "    mape = mean_absolute_percentage_error(yhat, preds)\n",
    "    print(\n",
    "        \"\"\"\n",
    "        mse -> {mse}\n",
    "        mae -> {mae}\n",
    "        rmse -> {rmse}\n",
    "        mape -> {mape}\n",
    "        r2 -> {r2}\n",
    "        \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    "    )\n",
    "    \n",
    "# params = {\n",
    "#     \"min_samples_split\": [2, 3],\n",
    "#     \"max_features\": ['sqrt', 'log2']\n",
    "# }\n",
    "model = DecisionTreeRegressor(random_state=42)\n",
    "model.fit(X_train, y_train)\n",
    "\n",
    "preds = model.predict(X_test)\n",
    "preds_train = model.predict(X_train)\n",
    "print(\"~\"*3 + \"test results\" + \"~\"*3)\n",
    "print_results(preds, y_test)\n",
    "print(\"~\"*3 + \"train results\" + \"~\"*3)\n",
    "print_results(preds_train, y_train)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "867d9f94",
   "metadata": {},
   "outputs": [],
   "source": [
    "path = model.cost_complexity_pruning_path(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "11014c38",
   "metadata": {},
   "outputs": [],
   "source": [
    "ccp_alphas, impurities = path.ccp_alphas, path.impurities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b9832c7d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.LineCollection at 0x7c7663dc7fd0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ax.plot(ccp_alphas[:-1], impurities[:-1], marker='o', drawstyle=\"steps-post\")\n",
    "ax.set_xlabel(\"effective alpha\")\n",
    "ax.set_ylabel(\"total impurity of leaves\")\n",
    "ax.set_title(\"Total Impurity vs effective alpha for training set\")\n",
    "ax.vlines(x=0.15, ymin=0, ymax=500)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0384cbd",
   "metadata": {},
   "source": [
    "примерно на значении 0.15 impurity перестает расти быстро. Буду пробовать значения impurity от 0 до 0.15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "eb7728b4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.001 , 0.0308, 0.0606, 0.0904, 0.1202, 0.15  ])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ccp_alphas = np.linspace(start=0.001, stop=0.15, num=6)\n",
    "ccp_alphas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "0626d129",
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = []\n",
    "for ccp_alpha in ccp_alphas:\n",
    "    model = DecisionTreeRegressor(random_state=42, ccp_alpha=ccp_alpha)\n",
    "    model.fit(X_train, y_train)\n",
    "    preds = model.predict(X_test)\n",
    "    score = mean_squared_error(preds, y_test)\n",
    "    scores.append(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "ae6f2269",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0308"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min_error = min(scores)\n",
    "index = scores.index(min_error)\n",
    "best_ccp_alpha = ccp_alphas[index]\n",
    "best_ccp_alpha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "dd35ffc2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "~~~test results~~~\n",
      "\n",
      "        mse -> 396.04322932231776\n",
      "        mae -> 15.252479508122564\n",
      "        rmse -> 19.90083489008232\n",
      "        mape -> 1.0738407937457514e+16\n",
      "        r2 -> 0.20882292133662494\n",
      "        \n",
      "~~~train results~~~\n",
      "\n",
      "        mse -> 254.28214686021\n",
      "        mae -> 12.294423490783105\n",
      "        rmse -> 15.946226727982078\n",
      "        mape -> 8156716022514243.0\n",
      "        r2 -> 0.4930079396978344\n",
      "        \n"
     ]
    }
   ],
   "source": [
    "model = DecisionTreeRegressor(random_state=42, ccp_alpha=best_ccp_alpha)\n",
    "model.fit(X_train, y_train)\n",
    "\n",
    "preds = model.predict(X_test)\n",
    "preds_train = model.predict(X_train)\n",
    "print(\"~\"*3 + \"test results\" + \"~\"*3)\n",
    "print_results(preds, y_test)\n",
    "print(\"~\"*3 + \"train results\" + \"~\"*3)\n",
    "print_results(preds_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3464cec8",
   "metadata": {},
   "source": [
    "<a id='4'>4 Обучение RandomForestRegressor</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "f0803ea4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 242.17627711414588\n",
      "    mae -> 11.22408010618234\n",
      "    rmse -> 15.562013915754795\n",
      "    mape -> 7035027430623027.0\n",
      "    r2 -> 0.5162035220836823\n",
      "    \n",
      "{'n_estimators': 250}\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "params = {\n",
    "    \"n_estimators\": [50, 100, 150, 200, 250],\n",
    "}\n",
    "model = GridSearchCV(estimator=RandomForestRegressor(random_state=42, n_jobs=-1), param_grid=params, scoring=['neg_mean_squared_error', 'neg_mean_absolute_error'], refit='neg_mean_squared_error', n_jobs=-1)\n",
    "model.fit(X_train, y_train)\n",
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")\n",
    "print(model.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d6457f93",
   "metadata": {},
   "source": [
    "<a id='5'>5 Обучение AdaBoostRegressor</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f7eaaa42",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 486.2072496945553\n",
      "    mae -> 18.543477848733136\n",
      "    rmse -> 22.05010770256135\n",
      "    mape -> 2.1509706628606256e+16\n",
      "    r2 -> 0.02870191191874727\n",
      "    \n",
      "{'learning_rate': 1.2941958414499861}\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "\n",
    "params = {\n",
    "    'learning_rate': np.logspace(start=0.001, stop=1.0, num=10)\n",
    "}\n",
    "model = GridSearchCV(estimator=AdaBoostRegressor(random_state=42), param_grid=params, scoring=['neg_mean_squared_error', 'neg_mean_absolute_error'], refit='neg_mean_squared_error', n_jobs=-1)\n",
    "model.fit(X_train, y_train)\n",
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")\n",
    "print(model.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bad878a6",
   "metadata": {},
   "source": [
    "<a id='6'>6 Обучение VotingRegressor</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "0d30af9f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 285.78311438662485\n",
      "    mae -> 12.843612664639814\n",
      "    rmse -> 16.905120951552664\n",
      "    mape -> 8886717684040270.0\n",
      "    r2 -> 0.4290899759639205\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import VotingRegressor\n",
    "\n",
    "est1 = RandomForestRegressor(random_state=42, n_estimators=250, n_jobs=-1)\n",
    "est2 = DecisionTreeRegressor(random_state=42, ccp_alpha=best_ccp_alpha)\n",
    "model = VotingRegressor([('est1', est1), ('est2', est2)])\n",
    "model.fit(X_train, y_train)\n",
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")\n",
    "del est1, est2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1daa9632",
   "metadata": {},
   "source": [
    "<a id='7'>7 Обучение НЕЙРОСЕТИ</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "19ab0679",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense (Dense)               (None, 256)               5120      \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 256)               0         \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 128)               32896     \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 64)                8256      \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 32)                2080      \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 48,385\n",
      "Trainable params: 48,385\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Dense(256, activation='relu', input_shape=(X.shape[1],)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Dense(128, activation='relu'))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Dense(64, activation='relu'))\n",
    "# model.add(Dropout(0.3))\n",
    "model.add(Dense(32, activation='relu'))\n",
    "# model.add(Dropout(0.3))\n",
    "model.add(Dense(1, activation='relu'))\n",
    "\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "671152c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow.keras.callbacks as callbacks\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint\n",
    "\n",
    "save_best = ModelCheckpoint(filepath=\"best_weights.hdf5\",\n",
    "                            monitor=\"val_root_mean_squared_error\",\n",
    "                            save_best_only=True,\n",
    "                            mode=\"auto\",\n",
    "                            verbose=1)\n",
    "\n",
    "\n",
    "plateau = callbacks.ReduceLROnPlateau(\n",
    "    monitor='val_root_mean_squared_error',    \n",
    "    factor = 0.2,\n",
    "    patience = 3,                                                                                            \n",
    "    verbose = 1,\n",
    "    min_lr = 1e-5\n",
    ") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "debce035",
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics = ['mae', 'root_mean_squared_error', 'mape', 'r2_score']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "c77fbd1c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.metrics import RootMeanSquaredError\n",
    "\n",
    "model.compile(optimizer='adam', loss='mse', metrics=['mae', RootMeanSquaredError()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "181d7cbc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 534.4277 - mae: 19.0957 - root_mean_squared_error: 23.1177\n",
      "Epoch 1: val_root_mean_squared_error improved from inf to 21.94366, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 4s 5ms/step - loss: 533.9120 - mae: 19.0893 - root_mean_squared_error: 23.1065 - val_loss: 481.5243 - val_mae: 18.2181 - val_root_mean_squared_error: 21.9437 - lr: 0.0010\n",
      "Epoch 2/50\n",
      "555/560 [============================>.] - ETA: 0s - loss: 492.4168 - mae: 18.4338 - root_mean_squared_error: 22.1905\n",
      "Epoch 2: val_root_mean_squared_error improved from 21.94366 to 21.92950, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 492.0770 - mae: 18.4286 - root_mean_squared_error: 22.1828 - val_loss: 480.9028 - val_mae: 18.2583 - val_root_mean_squared_error: 21.9295 - lr: 0.0010\n",
      "Epoch 3/50\n",
      "560/560 [==============================] - ETA: 0s - loss: 488.2264 - mae: 18.3172 - root_mean_squared_error: 22.0958\n",
      "Epoch 3: val_root_mean_squared_error improved from 21.92950 to 21.82128, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 488.2264 - mae: 18.3172 - root_mean_squared_error: 22.0958 - val_loss: 476.1684 - val_mae: 18.1484 - val_root_mean_squared_error: 21.8213 - lr: 0.0010\n",
      "Epoch 4/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 484.0132 - mae: 18.2223 - root_mean_squared_error: 22.0003\n",
      "Epoch 4: val_root_mean_squared_error did not improve from 21.82128\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 484.0359 - mae: 18.2186 - root_mean_squared_error: 22.0008 - val_loss: 476.9145 - val_mae: 18.1668 - val_root_mean_squared_error: 21.8384 - lr: 0.0010\n",
      "Epoch 5/50\n",
      "546/560 [============================>.] - ETA: 0s - loss: 483.9585 - mae: 18.2095 - root_mean_squared_error: 21.9991\n",
      "Epoch 5: val_root_mean_squared_error improved from 21.82128 to 21.74458, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 483.3726 - mae: 18.2018 - root_mean_squared_error: 21.9857 - val_loss: 472.8267 - val_mae: 18.0163 - val_root_mean_squared_error: 21.7446 - lr: 0.0010\n",
      "Epoch 6/50\n",
      "552/560 [============================>.] - ETA: 0s - loss: 480.7844 - mae: 18.1508 - root_mean_squared_error: 21.9268\n",
      "Epoch 6: val_root_mean_squared_error did not improve from 21.74458\n",
      "560/560 [==============================] - 3s 4ms/step - loss: 480.6992 - mae: 18.1518 - root_mean_squared_error: 21.9249 - val_loss: 473.2647 - val_mae: 18.0758 - val_root_mean_squared_error: 21.7546 - lr: 0.0010\n",
      "Epoch 7/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 478.8712 - mae: 18.1131 - root_mean_squared_error: 21.8831\n",
      "Epoch 7: val_root_mean_squared_error did not improve from 21.74458\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 478.8481 - mae: 18.1146 - root_mean_squared_error: 21.8826 - val_loss: 474.2388 - val_mae: 18.0768 - val_root_mean_squared_error: 21.7770 - lr: 0.0010\n",
      "Epoch 8/50\n",
      "549/560 [============================>.] - ETA: 0s - loss: 478.0509 - mae: 18.0974 - root_mean_squared_error: 21.8644\n",
      "Epoch 8: val_root_mean_squared_error improved from 21.74458 to 21.68469, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 477.7287 - mae: 18.0923 - root_mean_squared_error: 21.8570 - val_loss: 470.2259 - val_mae: 17.9688 - val_root_mean_squared_error: 21.6847 - lr: 0.0010\n",
      "Epoch 9/50\n",
      "558/560 [============================>.] - ETA: 0s - loss: 475.7781 - mae: 18.0541 - root_mean_squared_error: 21.8123\n",
      "Epoch 9: val_root_mean_squared_error improved from 21.68469 to 21.65721, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 475.8529 - mae: 18.0547 - root_mean_squared_error: 21.8141 - val_loss: 469.0348 - val_mae: 17.9029 - val_root_mean_squared_error: 21.6572 - lr: 0.0010\n",
      "Epoch 10/50\n",
      "559/560 [============================>.] - ETA: 0s - loss: 473.8624 - mae: 18.0044 - root_mean_squared_error: 21.7684\n",
      "Epoch 10: val_root_mean_squared_error improved from 21.65721 to 21.61041, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 473.9281 - mae: 18.0070 - root_mean_squared_error: 21.7699 - val_loss: 467.0099 - val_mae: 17.8690 - val_root_mean_squared_error: 21.6104 - lr: 0.0010\n",
      "Epoch 11/50\n",
      "552/560 [============================>.] - ETA: 0s - loss: 473.3816 - mae: 17.9990 - root_mean_squared_error: 21.7573\n",
      "Epoch 11: val_root_mean_squared_error did not improve from 21.61041\n",
      "560/560 [==============================] - 2s 4ms/step - loss: 473.2954 - mae: 17.9992 - root_mean_squared_error: 21.7554 - val_loss: 471.6990 - val_mae: 18.0274 - val_root_mean_squared_error: 21.7186 - lr: 0.0010\n",
      "Epoch 12/50\n",
      "556/560 [============================>.] - ETA: 0s - loss: 471.9916 - mae: 17.9698 - root_mean_squared_error: 21.7254\n",
      "Epoch 12: val_root_mean_squared_error improved from 21.61041 to 21.60835, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 472.0759 - mae: 17.9704 - root_mean_squared_error: 21.7273 - val_loss: 466.9207 - val_mae: 17.8983 - val_root_mean_squared_error: 21.6083 - lr: 0.0010\n",
      "Epoch 13/50\n",
      "550/560 [============================>.] - ETA: 0s - loss: 471.2691 - mae: 17.9628 - root_mean_squared_error: 21.7087\n",
      "Epoch 13: val_root_mean_squared_error did not improve from 21.60835\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 470.9954 - mae: 17.9556 - root_mean_squared_error: 21.7024 - val_loss: 471.2775 - val_mae: 18.0648 - val_root_mean_squared_error: 21.7089 - lr: 0.0010\n",
      "Epoch 14/50\n",
      "560/560 [==============================] - ETA: 0s - loss: 470.2653 - mae: 17.9466 - root_mean_squared_error: 21.6856\n",
      "Epoch 14: val_root_mean_squared_error did not improve from 21.60835\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 470.2653 - mae: 17.9466 - root_mean_squared_error: 21.6856 - val_loss: 471.3307 - val_mae: 18.0206 - val_root_mean_squared_error: 21.7102 - lr: 0.0010\n",
      "Epoch 15/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 469.3148 - mae: 17.9109 - root_mean_squared_error: 21.6637\n",
      "Epoch 15: val_root_mean_squared_error improved from 21.60835 to 21.57169, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 469.3274 - mae: 17.9106 - root_mean_squared_error: 21.6640 - val_loss: 465.3377 - val_mae: 17.8132 - val_root_mean_squared_error: 21.5717 - lr: 0.0010\n",
      "Epoch 16/50\n",
      "560/560 [==============================] - ETA: 0s - loss: 468.6149 - mae: 17.9023 - root_mean_squared_error: 21.6475\n",
      "Epoch 16: val_root_mean_squared_error did not improve from 21.57169\n",
      "560/560 [==============================] - 2s 4ms/step - loss: 468.6149 - mae: 17.9023 - root_mean_squared_error: 21.6475 - val_loss: 466.6220 - val_mae: 17.9362 - val_root_mean_squared_error: 21.6014 - lr: 0.0010\n",
      "Epoch 17/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 467.7665 - mae: 17.8908 - root_mean_squared_error: 21.6279\n",
      "Epoch 17: val_root_mean_squared_error did not improve from 21.57169\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 467.7248 - mae: 17.8923 - root_mean_squared_error: 21.6269 - val_loss: 465.6278 - val_mae: 17.8447 - val_root_mean_squared_error: 21.5784 - lr: 0.0010\n",
      "Epoch 18/50\n",
      "555/560 [============================>.] - ETA: 0s - loss: 466.2156 - mae: 17.8479 - root_mean_squared_error: 21.5920\n",
      "Epoch 18: val_root_mean_squared_error improved from 21.57169 to 21.56698, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 466.0617 - mae: 17.8440 - root_mean_squared_error: 21.5885 - val_loss: 465.1346 - val_mae: 17.7510 - val_root_mean_squared_error: 21.5670 - lr: 0.0010\n",
      "Epoch 19/50\n",
      "559/560 [============================>.] - ETA: 0s - loss: 466.2432 - mae: 17.8486 - root_mean_squared_error: 21.5927\n",
      "Epoch 19: val_root_mean_squared_error did not improve from 21.56698\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 466.3862 - mae: 17.8509 - root_mean_squared_error: 21.5960 - val_loss: 465.7154 - val_mae: 17.7056 - val_root_mean_squared_error: 21.5804 - lr: 0.0010\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 20/50\n",
      "557/560 [============================>.] - ETA: 0s - loss: 465.7216 - mae: 17.8321 - root_mean_squared_error: 21.5806\n",
      "Epoch 20: val_root_mean_squared_error improved from 21.56698 to 21.51177, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 465.5190 - mae: 17.8284 - root_mean_squared_error: 21.5759 - val_loss: 462.7563 - val_mae: 17.7465 - val_root_mean_squared_error: 21.5118 - lr: 0.0010\n",
      "Epoch 21/50\n",
      "555/560 [============================>.] - ETA: 0s - loss: 464.2400 - mae: 17.7965 - root_mean_squared_error: 21.5462\n",
      "Epoch 21: val_root_mean_squared_error improved from 21.51177 to 21.51137, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 2s 4ms/step - loss: 464.1308 - mae: 17.7950 - root_mean_squared_error: 21.5437 - val_loss: 462.7392 - val_mae: 17.7165 - val_root_mean_squared_error: 21.5114 - lr: 0.0010\n",
      "Epoch 22/50\n",
      "560/560 [==============================] - ETA: 0s - loss: 462.8025 - mae: 17.7558 - root_mean_squared_error: 21.5128\n",
      "Epoch 22: val_root_mean_squared_error did not improve from 21.51137\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 462.8025 - mae: 17.7558 - root_mean_squared_error: 21.5128 - val_loss: 464.3857 - val_mae: 17.8569 - val_root_mean_squared_error: 21.5496 - lr: 0.0010\n",
      "Epoch 23/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 462.8865 - mae: 17.7554 - root_mean_squared_error: 21.5148\n",
      "Epoch 23: val_root_mean_squared_error improved from 21.51137 to 21.45562, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 462.8324 - mae: 17.7550 - root_mean_squared_error: 21.5135 - val_loss: 460.3438 - val_mae: 17.7524 - val_root_mean_squared_error: 21.4556 - lr: 0.0010\n",
      "Epoch 24/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 461.1951 - mae: 17.7209 - root_mean_squared_error: 21.4755\n",
      "Epoch 24: val_root_mean_squared_error improved from 21.45562 to 21.43506, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 4ms/step - loss: 461.0845 - mae: 17.7151 - root_mean_squared_error: 21.4729 - val_loss: 459.4618 - val_mae: 17.6680 - val_root_mean_squared_error: 21.4351 - lr: 0.0010\n",
      "Epoch 25/50\n",
      "547/560 [============================>.] - ETA: 0s - loss: 459.7657 - mae: 17.6819 - root_mean_squared_error: 21.4421\n",
      "Epoch 25: val_root_mean_squared_error improved from 21.43506 to 21.43113, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 459.8412 - mae: 17.6830 - root_mean_squared_error: 21.4439 - val_loss: 459.2932 - val_mae: 17.6004 - val_root_mean_squared_error: 21.4311 - lr: 0.0010\n",
      "Epoch 26/50\n",
      "551/560 [============================>.] - ETA: 0s - loss: 458.8484 - mae: 17.6661 - root_mean_squared_error: 21.4207\n",
      "Epoch 26: val_root_mean_squared_error improved from 21.43113 to 21.40858, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 459.0597 - mae: 17.6659 - root_mean_squared_error: 21.4257 - val_loss: 458.3273 - val_mae: 17.6164 - val_root_mean_squared_error: 21.4086 - lr: 0.0010\n",
      "Epoch 27/50\n",
      "549/560 [============================>.] - ETA: 0s - loss: 457.2751 - mae: 17.6195 - root_mean_squared_error: 21.3840\n",
      "Epoch 27: val_root_mean_squared_error improved from 21.40858 to 21.37125, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 457.1917 - mae: 17.6193 - root_mean_squared_error: 21.3820 - val_loss: 456.7305 - val_mae: 17.5291 - val_root_mean_squared_error: 21.3713 - lr: 0.0010\n",
      "Epoch 28/50\n",
      "555/560 [============================>.] - ETA: 0s - loss: 455.9439 - mae: 17.5882 - root_mean_squared_error: 21.3528\n",
      "Epoch 28: val_root_mean_squared_error improved from 21.37125 to 21.34015, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 456.2275 - mae: 17.5959 - root_mean_squared_error: 21.3595 - val_loss: 455.4019 - val_mae: 17.5554 - val_root_mean_squared_error: 21.3401 - lr: 0.0010\n",
      "Epoch 29/50\n",
      "548/560 [============================>.] - ETA: 0s - loss: 455.6786 - mae: 17.5719 - root_mean_squared_error: 21.3466\n",
      "Epoch 29: val_root_mean_squared_error did not improve from 21.34015\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 455.8305 - mae: 17.5785 - root_mean_squared_error: 21.3502 - val_loss: 455.6438 - val_mae: 17.5332 - val_root_mean_squared_error: 21.3458 - lr: 0.0010\n",
      "Epoch 30/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 454.8186 - mae: 17.5423 - root_mean_squared_error: 21.3265\n",
      "Epoch 30: val_root_mean_squared_error did not improve from 21.34015\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 454.7932 - mae: 17.5438 - root_mean_squared_error: 21.3259 - val_loss: 456.1064 - val_mae: 17.5593 - val_root_mean_squared_error: 21.3566 - lr: 0.0010\n",
      "Epoch 31/50\n",
      "550/560 [============================>.] - ETA: 0s - loss: 453.1388 - mae: 17.5063 - root_mean_squared_error: 21.2871\n",
      "Epoch 31: val_root_mean_squared_error did not improve from 21.34015\n",
      "\n",
      "Epoch 31: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 453.2228 - mae: 17.5077 - root_mean_squared_error: 21.2890 - val_loss: 456.3398 - val_mae: 17.4725 - val_root_mean_squared_error: 21.3621 - lr: 0.0010\n",
      "Epoch 32/50\n",
      "548/560 [============================>.] - ETA: 0s - loss: 447.2572 - mae: 17.3523 - root_mean_squared_error: 21.1485\n",
      "Epoch 32: val_root_mean_squared_error improved from 21.34015 to 21.25226, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 2s 4ms/step - loss: 447.5686 - mae: 17.3596 - root_mean_squared_error: 21.1558 - val_loss: 451.6584 - val_mae: 17.4668 - val_root_mean_squared_error: 21.2523 - lr: 2.0000e-04\n",
      "Epoch 33/50\n",
      "559/560 [============================>.] - ETA: 0s - loss: 445.8459 - mae: 17.3289 - root_mean_squared_error: 21.1151\n",
      "Epoch 33: val_root_mean_squared_error improved from 21.25226 to 21.21735, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 445.9850 - mae: 17.3322 - root_mean_squared_error: 21.1184 - val_loss: 450.1758 - val_mae: 17.3946 - val_root_mean_squared_error: 21.2173 - lr: 2.0000e-04\n",
      "Epoch 34/50\n",
      "560/560 [==============================] - ETA: 0s - loss: 445.8347 - mae: 17.3196 - root_mean_squared_error: 21.1148\n",
      "Epoch 34: val_root_mean_squared_error improved from 21.21735 to 21.20211, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 445.8347 - mae: 17.3196 - root_mean_squared_error: 21.1148 - val_loss: 449.5295 - val_mae: 17.3668 - val_root_mean_squared_error: 21.2021 - lr: 2.0000e-04\n",
      "Epoch 35/50\n",
      "556/560 [============================>.] - ETA: 0s - loss: 444.8211 - mae: 17.3014 - root_mean_squared_error: 21.0908\n",
      "Epoch 35: val_root_mean_squared_error did not improve from 21.20211\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 444.9759 - mae: 17.3042 - root_mean_squared_error: 21.0945 - val_loss: 450.3568 - val_mae: 17.3785 - val_root_mean_squared_error: 21.2216 - lr: 2.0000e-04\n",
      "Epoch 36/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 443.7896 - mae: 17.2666 - root_mean_squared_error: 21.0663\n",
      "Epoch 36: val_root_mean_squared_error improved from 21.20211 to 21.19136, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 444.1483 - mae: 17.2753 - root_mean_squared_error: 21.0748 - val_loss: 449.0737 - val_mae: 17.3421 - val_root_mean_squared_error: 21.1914 - lr: 2.0000e-04\n",
      "Epoch 37/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 443.4491 - mae: 17.2480 - root_mean_squared_error: 21.0582\n",
      "Epoch 37: val_root_mean_squared_error did not improve from 21.19136\n",
      "560/560 [==============================] - 2s 4ms/step - loss: 443.4917 - mae: 17.2482 - root_mean_squared_error: 21.0592 - val_loss: 449.2308 - val_mae: 17.2985 - val_root_mean_squared_error: 21.1951 - lr: 2.0000e-04\n",
      "Epoch 38/50\n",
      "555/560 [============================>.] - ETA: 0s - loss: 443.8342 - mae: 17.2499 - root_mean_squared_error: 21.0674\n",
      "Epoch 38: val_root_mean_squared_error improved from 21.19136 to 21.18126, saving model to best_weights.hdf5\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "560/560 [==============================] - 3s 5ms/step - loss: 443.5633 - mae: 17.2432 - root_mean_squared_error: 21.0609 - val_loss: 448.6457 - val_mae: 17.3465 - val_root_mean_squared_error: 21.1813 - lr: 2.0000e-04\n",
      "Epoch 39/50\n",
      "557/560 [============================>.] - ETA: 0s - loss: 443.0300 - mae: 17.2414 - root_mean_squared_error: 21.0483\n",
      "Epoch 39: val_root_mean_squared_error improved from 21.18126 to 21.16651, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 4ms/step - loss: 443.0610 - mae: 17.2426 - root_mean_squared_error: 21.0490 - val_loss: 448.0213 - val_mae: 17.3213 - val_root_mean_squared_error: 21.1665 - lr: 2.0000e-04\n",
      "Epoch 40/50\n",
      "553/560 [============================>.] - ETA: 0s - loss: 443.1344 - mae: 17.2443 - root_mean_squared_error: 21.0508\n",
      "Epoch 40: val_root_mean_squared_error improved from 21.16651 to 21.15790, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 443.0339 - mae: 17.2388 - root_mean_squared_error: 21.0484 - val_loss: 447.6569 - val_mae: 17.3101 - val_root_mean_squared_error: 21.1579 - lr: 2.0000e-04\n",
      "Epoch 41/50\n",
      "551/560 [============================>.] - ETA: 0s - loss: 442.7176 - mae: 17.2254 - root_mean_squared_error: 21.0409\n",
      "Epoch 41: val_root_mean_squared_error improved from 21.15790 to 21.15602, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 442.9447 - mae: 17.2319 - root_mean_squared_error: 21.0463 - val_loss: 447.5770 - val_mae: 17.3218 - val_root_mean_squared_error: 21.1560 - lr: 2.0000e-04\n",
      "Epoch 42/50\n",
      "559/560 [============================>.] - ETA: 0s - loss: 442.6478 - mae: 17.2202 - root_mean_squared_error: 21.0392\n",
      "Epoch 42: val_root_mean_squared_error improved from 21.15602 to 21.14530, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 442.5944 - mae: 17.2186 - root_mean_squared_error: 21.0379 - val_loss: 447.1239 - val_mae: 17.2801 - val_root_mean_squared_error: 21.1453 - lr: 2.0000e-04\n",
      "Epoch 43/50\n",
      "549/560 [============================>.] - ETA: 0s - loss: 442.1453 - mae: 17.2140 - root_mean_squared_error: 21.0273\n",
      "Epoch 43: val_root_mean_squared_error did not improve from 21.14530\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 441.8412 - mae: 17.2092 - root_mean_squared_error: 21.0200 - val_loss: 447.2757 - val_mae: 17.3473 - val_root_mean_squared_error: 21.1489 - lr: 2.0000e-04\n",
      "Epoch 44/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 441.3395 - mae: 17.1987 - root_mean_squared_error: 21.0081\n",
      "Epoch 44: val_root_mean_squared_error improved from 21.14530 to 21.12741, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 4ms/step - loss: 441.2919 - mae: 17.1968 - root_mean_squared_error: 21.0069 - val_loss: 446.3676 - val_mae: 17.2914 - val_root_mean_squared_error: 21.1274 - lr: 2.0000e-04\n",
      "Epoch 45/50\n",
      "555/560 [============================>.] - ETA: 0s - loss: 441.0567 - mae: 17.1951 - root_mean_squared_error: 21.0013\n",
      "Epoch 45: val_root_mean_squared_error improved from 21.12741 to 21.12405, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 441.0352 - mae: 17.1952 - root_mean_squared_error: 21.0008 - val_loss: 446.2256 - val_mae: 17.2545 - val_root_mean_squared_error: 21.1241 - lr: 2.0000e-04\n",
      "Epoch 46/50\n",
      "552/560 [============================>.] - ETA: 0s - loss: 440.5192 - mae: 17.1691 - root_mean_squared_error: 20.9885\n",
      "Epoch 46: val_root_mean_squared_error improved from 21.12405 to 21.12392, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 440.5935 - mae: 17.1708 - root_mean_squared_error: 20.9903 - val_loss: 446.2200 - val_mae: 17.2774 - val_root_mean_squared_error: 21.1239 - lr: 2.0000e-04\n",
      "Epoch 47/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 440.9514 - mae: 17.1830 - root_mean_squared_error: 20.9988\n",
      "Epoch 47: val_root_mean_squared_error did not improve from 21.12392\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 441.1775 - mae: 17.1884 - root_mean_squared_error: 21.0042 - val_loss: 446.3133 - val_mae: 17.2781 - val_root_mean_squared_error: 21.1261 - lr: 2.0000e-04\n",
      "Epoch 48/50\n",
      "558/560 [============================>.] - ETA: 0s - loss: 441.3314 - mae: 17.1841 - root_mean_squared_error: 21.0079\n",
      "Epoch 48: val_root_mean_squared_error improved from 21.12392 to 21.12366, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 441.1317 - mae: 17.1802 - root_mean_squared_error: 21.0031 - val_loss: 446.2090 - val_mae: 17.3262 - val_root_mean_squared_error: 21.1237 - lr: 2.0000e-04\n",
      "Epoch 49/50\n",
      "554/560 [============================>.] - ETA: 0s - loss: 439.6557 - mae: 17.1571 - root_mean_squared_error: 20.9680\n",
      "Epoch 49: val_root_mean_squared_error improved from 21.12366 to 21.11277, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 439.7210 - mae: 17.1570 - root_mean_squared_error: 20.9695 - val_loss: 445.7489 - val_mae: 17.2588 - val_root_mean_squared_error: 21.1128 - lr: 2.0000e-04\n",
      "Epoch 50/50\n",
      "552/560 [============================>.] - ETA: 0s - loss: 439.7066 - mae: 17.1446 - root_mean_squared_error: 20.9692\n",
      "Epoch 50: val_root_mean_squared_error improved from 21.11277 to 21.09305, saving model to best_weights.hdf5\n",
      "560/560 [==============================] - 3s 5ms/step - loss: 439.4897 - mae: 17.1403 - root_mean_squared_error: 20.9640 - val_loss: 444.9166 - val_mae: 17.2252 - val_root_mean_squared_error: 21.0930 - lr: 2.0000e-04\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, \n",
    "                    y_train, \n",
    "                    epochs=50, \n",
    "                    validation_split=0.2, \n",
    "                    batch_size=128,\n",
    "                    verbose=1, \n",
    "                    callbacks=[save_best, plateau]\n",
    "                   )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "847a532e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['mae'], \n",
    "         label='Средняя абсолютная ошибка на обучающем наборе')\n",
    "plt.plot(history.history['val_mae'], \n",
    "         label='Средняя абсолютная ошибка на проверочном наборе')\n",
    "plt.xlabel('Эпоха обучения')\n",
    "plt.ylabel('Средняя абсолютная ошибка')\n",
    "plt.legend(loc=(1.05, 0.5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "bd9d882c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['root_mean_squared_error'], \n",
    "         label='root_mean_squared_error на обучающем наборе')\n",
    "plt.plot(history.history['val_root_mean_squared_error'], \n",
    "         label='root_mean_squared_error на проверочном наборе')\n",
    "plt.xlabel('Эпоха обучения')\n",
    "plt.ylabel('Средняя абсолютная ошибка')\n",
    "plt.legend(loc=(1.05, 0.5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "74bb2685",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'], \n",
    "         label='loss на обучающем наборе')\n",
    "plt.plot(history.history['val_loss'], \n",
    "         label='val_loss на проверочном наборе')\n",
    "plt.xlabel('Эпоха обучения')\n",
    "plt.ylabel('Средняя абсолютная ошибка')\n",
    "plt.legend(loc=(1.05, 0.5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70dda365",
   "metadata": {},
   "source": [
    "## Проверка на холд аут наборе"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "44e12a80",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 445.4984883017255\n",
      "    mae -> 17.290321515667593\n",
      "    rmse -> 21.106835108602272\n",
      "    mape -> 1.8906130316820376e+16\n",
      "    r2 -> 0.11002596073507298\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "model.load_weights('best_weights.hdf5')\n",
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c72b2a0c",
   "metadata": {},
   "source": [
    "<a id='8'>8 Обучение catboost</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "dbb061f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Name: catboost\r\n",
      "Version: 1.2.7\r\n",
      "Summary: CatBoost Python Package\r\n",
      "Home-page: https://catboost.ai\r\n",
      "Author: CatBoost Developers\r\n",
      "Author-email: \r\n",
      "License: Apache License, Version 2.0\r\n",
      "Location: /usr/local/lib/python3.8/dist-packages\r\n",
      "Requires: graphviz, matplotlib, numpy, pandas, plotly, scipy, six\r\n",
      "Required-by: \r\n"
     ]
    }
   ],
   "source": [
    "!pip show catboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "8b6c48f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.1, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "0935308d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 22.2912801\ttest: 22.3352761\tbest: 22.3352761 (0)\ttotal: 43.7ms\tremaining: 43.6s\n",
      "1:\tlearn: 22.2004646\ttest: 22.2593341\tbest: 22.2593341 (1)\ttotal: 87.1ms\tremaining: 43.5s\n",
      "2:\tlearn: 22.1127524\ttest: 22.1846365\tbest: 22.1846365 (2)\ttotal: 134ms\tremaining: 44.7s\n",
      "3:\tlearn: 22.0332782\ttest: 22.1132311\tbest: 22.1132311 (3)\ttotal: 193ms\tremaining: 48s\n",
      "4:\tlearn: 21.9582692\ttest: 22.0529818\tbest: 22.0529818 (4)\ttotal: 245ms\tremaining: 48.7s\n",
      "5:\tlearn: 21.8959086\ttest: 21.9983358\tbest: 21.9983358 (5)\ttotal: 305ms\tremaining: 50.5s\n",
      "6:\tlearn: 21.8374353\ttest: 21.9468791\tbest: 21.9468791 (6)\ttotal: 365ms\tremaining: 51.8s\n",
      "7:\tlearn: 21.7774568\ttest: 21.8982290\tbest: 21.8982290 (7)\ttotal: 420ms\tremaining: 52.1s\n",
      "8:\tlearn: 21.7250648\ttest: 21.8527350\tbest: 21.8527350 (8)\ttotal: 476ms\tremaining: 52.4s\n",
      "9:\tlearn: 21.6709969\ttest: 21.8111043\tbest: 21.8111043 (9)\ttotal: 530ms\tremaining: 52.4s\n",
      "10:\tlearn: 21.6287253\ttest: 21.7796094\tbest: 21.7796094 (10)\ttotal: 579ms\tremaining: 52.1s\n",
      "11:\tlearn: 21.5834541\ttest: 21.7418879\tbest: 21.7418879 (11)\ttotal: 628ms\tremaining: 51.7s\n",
      "12:\tlearn: 21.5475284\ttest: 21.7072985\tbest: 21.7072985 (12)\ttotal: 684ms\tremaining: 51.9s\n",
      "13:\tlearn: 21.5007653\ttest: 21.6695475\tbest: 21.6695475 (13)\ttotal: 730ms\tremaining: 51.4s\n",
      "14:\tlearn: 21.4590551\ttest: 21.6361278\tbest: 21.6361278 (14)\ttotal: 784ms\tremaining: 51.5s\n",
      "15:\tlearn: 21.4209190\ttest: 21.6073007\tbest: 21.6073007 (15)\ttotal: 841ms\tremaining: 51.7s\n",
      "16:\tlearn: 21.3917083\ttest: 21.5867977\tbest: 21.5867977 (16)\ttotal: 899ms\tremaining: 52s\n",
      "17:\tlearn: 21.3701648\ttest: 21.5701742\tbest: 21.5701742 (17)\ttotal: 958ms\tremaining: 52.3s\n",
      "18:\tlearn: 21.3356802\ttest: 21.5473547\tbest: 21.5473547 (18)\ttotal: 1.02s\tremaining: 52.7s\n",
      "19:\tlearn: 21.2986322\ttest: 21.5170872\tbest: 21.5170872 (19)\ttotal: 1.08s\tremaining: 52.9s\n",
      "20:\tlearn: 21.2745092\ttest: 21.4964341\tbest: 21.4964341 (20)\ttotal: 1.13s\tremaining: 52.7s\n",
      "21:\tlearn: 21.2423051\ttest: 21.4747190\tbest: 21.4747190 (21)\ttotal: 1.18s\tremaining: 52.3s\n",
      "22:\tlearn: 21.2143314\ttest: 21.4501885\tbest: 21.4501885 (22)\ttotal: 1.23s\tremaining: 52.2s\n",
      "23:\tlearn: 21.1908614\ttest: 21.4356105\tbest: 21.4356105 (23)\ttotal: 1.28s\tremaining: 52.2s\n",
      "24:\tlearn: 21.1566716\ttest: 21.4070846\tbest: 21.4070846 (24)\ttotal: 1.34s\tremaining: 52.2s\n",
      "25:\tlearn: 21.1179933\ttest: 21.3766211\tbest: 21.3766211 (25)\ttotal: 1.39s\tremaining: 52.1s\n",
      "26:\tlearn: 21.0920471\ttest: 21.3590936\tbest: 21.3590936 (26)\ttotal: 1.44s\tremaining: 51.9s\n",
      "27:\tlearn: 21.0621047\ttest: 21.3379844\tbest: 21.3379844 (27)\ttotal: 1.5s\tremaining: 51.9s\n",
      "28:\tlearn: 21.0358210\ttest: 21.3150176\tbest: 21.3150176 (28)\ttotal: 1.54s\tremaining: 51.6s\n",
      "29:\tlearn: 21.0108743\ttest: 21.2944117\tbest: 21.2944117 (29)\ttotal: 1.6s\tremaining: 51.7s\n",
      "30:\tlearn: 20.9839595\ttest: 21.2788874\tbest: 21.2788874 (30)\ttotal: 1.65s\tremaining: 51.7s\n",
      "31:\tlearn: 20.9613697\ttest: 21.2632804\tbest: 21.2632804 (31)\ttotal: 1.72s\tremaining: 51.9s\n",
      "32:\tlearn: 20.9378428\ttest: 21.2520981\tbest: 21.2520981 (32)\ttotal: 1.78s\tremaining: 52.1s\n",
      "33:\tlearn: 20.9035121\ttest: 21.2286800\tbest: 21.2286800 (33)\ttotal: 1.82s\tremaining: 51.9s\n",
      "34:\tlearn: 20.8839560\ttest: 21.2174193\tbest: 21.2174193 (34)\ttotal: 1.88s\tremaining: 51.8s\n",
      "35:\tlearn: 20.8725899\ttest: 21.2079646\tbest: 21.2079646 (35)\ttotal: 1.94s\tremaining: 51.9s\n",
      "36:\tlearn: 20.8450602\ttest: 21.1925817\tbest: 21.1925817 (36)\ttotal: 2s\tremaining: 52s\n",
      "37:\tlearn: 20.8258579\ttest: 21.1795155\tbest: 21.1795155 (37)\ttotal: 2.06s\tremaining: 52s\n",
      "38:\tlearn: 20.7997709\ttest: 21.1622453\tbest: 21.1622453 (38)\ttotal: 2.12s\tremaining: 52.1s\n",
      "39:\tlearn: 20.7760800\ttest: 21.1457666\tbest: 21.1457666 (39)\ttotal: 2.17s\tremaining: 52s\n",
      "40:\tlearn: 20.7603086\ttest: 21.1352347\tbest: 21.1352347 (40)\ttotal: 2.23s\tremaining: 52.1s\n",
      "41:\tlearn: 20.7347065\ttest: 21.1154001\tbest: 21.1154001 (41)\ttotal: 2.29s\tremaining: 52.2s\n",
      "42:\tlearn: 20.7094003\ttest: 21.0997873\tbest: 21.0997873 (42)\ttotal: 2.34s\tremaining: 52s\n",
      "43:\tlearn: 20.6781304\ttest: 21.0778581\tbest: 21.0778581 (43)\ttotal: 2.4s\tremaining: 52.1s\n",
      "44:\tlearn: 20.6584431\ttest: 21.0650821\tbest: 21.0650821 (44)\ttotal: 2.45s\tremaining: 52s\n",
      "45:\tlearn: 20.6439984\ttest: 21.0586123\tbest: 21.0586123 (45)\ttotal: 2.51s\tremaining: 52s\n",
      "46:\tlearn: 20.6353169\ttest: 21.0541674\tbest: 21.0541674 (46)\ttotal: 2.59s\tremaining: 52.5s\n",
      "47:\tlearn: 20.6172512\ttest: 21.0421410\tbest: 21.0421410 (47)\ttotal: 2.65s\tremaining: 52.6s\n",
      "48:\tlearn: 20.6024316\ttest: 21.0311382\tbest: 21.0311382 (48)\ttotal: 2.71s\tremaining: 52.6s\n",
      "49:\tlearn: 20.5698755\ttest: 21.0113132\tbest: 21.0113132 (49)\ttotal: 2.79s\tremaining: 52.9s\n",
      "50:\tlearn: 20.5547686\ttest: 21.0006861\tbest: 21.0006861 (50)\ttotal: 2.84s\tremaining: 52.9s\n",
      "51:\tlearn: 20.5301251\ttest: 20.9849281\tbest: 20.9849281 (51)\ttotal: 2.9s\tremaining: 52.8s\n",
      "52:\tlearn: 20.5125627\ttest: 20.9763758\tbest: 20.9763758 (52)\ttotal: 2.95s\tremaining: 52.8s\n",
      "53:\tlearn: 20.4921225\ttest: 20.9597665\tbest: 20.9597665 (53)\ttotal: 3.01s\tremaining: 52.7s\n",
      "54:\tlearn: 20.4766977\ttest: 20.9516718\tbest: 20.9516718 (54)\ttotal: 3.05s\tremaining: 52.4s\n",
      "55:\tlearn: 20.4596701\ttest: 20.9394711\tbest: 20.9394711 (55)\ttotal: 3.1s\tremaining: 52.3s\n",
      "56:\tlearn: 20.4441647\ttest: 20.9292045\tbest: 20.9292045 (56)\ttotal: 3.15s\tremaining: 52.2s\n",
      "57:\tlearn: 20.4325713\ttest: 20.9224038\tbest: 20.9224038 (57)\ttotal: 3.23s\tremaining: 52.4s\n",
      "58:\tlearn: 20.4153048\ttest: 20.9112429\tbest: 20.9112429 (58)\ttotal: 3.3s\tremaining: 52.6s\n",
      "59:\tlearn: 20.3969028\ttest: 20.9002713\tbest: 20.9002713 (59)\ttotal: 3.36s\tremaining: 52.6s\n",
      "60:\tlearn: 20.3795113\ttest: 20.8938875\tbest: 20.8938875 (60)\ttotal: 3.42s\tremaining: 52.7s\n",
      "61:\tlearn: 20.3607085\ttest: 20.8801002\tbest: 20.8801002 (61)\ttotal: 3.48s\tremaining: 52.6s\n",
      "62:\tlearn: 20.3444683\ttest: 20.8669880\tbest: 20.8669880 (62)\ttotal: 3.53s\tremaining: 52.6s\n",
      "63:\tlearn: 20.3293227\ttest: 20.8606945\tbest: 20.8606945 (63)\ttotal: 3.59s\tremaining: 52.5s\n",
      "64:\tlearn: 20.3125951\ttest: 20.8493008\tbest: 20.8493008 (64)\ttotal: 3.64s\tremaining: 52.4s\n",
      "65:\tlearn: 20.3002067\ttest: 20.8400605\tbest: 20.8400605 (65)\ttotal: 3.69s\tremaining: 52.3s\n",
      "66:\tlearn: 20.2876250\ttest: 20.8312712\tbest: 20.8312712 (66)\ttotal: 3.75s\tremaining: 52.2s\n",
      "67:\tlearn: 20.2615646\ttest: 20.8151666\tbest: 20.8151666 (67)\ttotal: 3.8s\tremaining: 52.1s\n",
      "68:\tlearn: 20.2492252\ttest: 20.8057370\tbest: 20.8057370 (68)\ttotal: 3.86s\tremaining: 52.1s\n",
      "69:\tlearn: 20.2309275\ttest: 20.7937635\tbest: 20.7937635 (69)\ttotal: 3.92s\tremaining: 52.1s\n",
      "70:\tlearn: 20.2162640\ttest: 20.7847330\tbest: 20.7847330 (70)\ttotal: 3.99s\tremaining: 52.2s\n",
      "71:\tlearn: 20.2040666\ttest: 20.7779593\tbest: 20.7779593 (71)\ttotal: 4.05s\tremaining: 52.3s\n",
      "72:\tlearn: 20.1754834\ttest: 20.7603758\tbest: 20.7603758 (72)\ttotal: 4.11s\tremaining: 52.2s\n",
      "73:\tlearn: 20.1573566\ttest: 20.7504139\tbest: 20.7504139 (73)\ttotal: 4.17s\tremaining: 52.2s\n",
      "74:\tlearn: 20.1445044\ttest: 20.7423378\tbest: 20.7423378 (74)\ttotal: 4.23s\tremaining: 52.1s\n",
      "75:\tlearn: 20.1303499\ttest: 20.7336644\tbest: 20.7336644 (75)\ttotal: 4.28s\tremaining: 52.1s\n",
      "76:\tlearn: 20.1132019\ttest: 20.7203398\tbest: 20.7203398 (76)\ttotal: 4.33s\tremaining: 51.9s\n",
      "77:\tlearn: 20.0996436\ttest: 20.7147671\tbest: 20.7147671 (77)\ttotal: 4.38s\tremaining: 51.8s\n",
      "78:\tlearn: 20.0817331\ttest: 20.7025564\tbest: 20.7025564 (78)\ttotal: 4.44s\tremaining: 51.8s\n",
      "79:\tlearn: 20.0742072\ttest: 20.6992713\tbest: 20.6992713 (79)\ttotal: 4.5s\tremaining: 51.7s\n",
      "80:\tlearn: 20.0567057\ttest: 20.6906698\tbest: 20.6906698 (80)\ttotal: 4.54s\tremaining: 51.5s\n",
      "81:\tlearn: 20.0479478\ttest: 20.6842839\tbest: 20.6842839 (81)\ttotal: 4.59s\tremaining: 51.4s\n",
      "82:\tlearn: 20.0378982\ttest: 20.6812045\tbest: 20.6812045 (82)\ttotal: 4.64s\tremaining: 51.3s\n",
      "83:\tlearn: 20.0274809\ttest: 20.6749861\tbest: 20.6749861 (83)\ttotal: 4.7s\tremaining: 51.2s\n",
      "84:\tlearn: 19.9993988\ttest: 20.6590463\tbest: 20.6590463 (84)\ttotal: 4.76s\tremaining: 51.2s\n",
      "85:\tlearn: 19.9833795\ttest: 20.6502885\tbest: 20.6502885 (85)\ttotal: 4.81s\tremaining: 51.1s\n",
      "86:\tlearn: 19.9631700\ttest: 20.6372887\tbest: 20.6372887 (86)\ttotal: 4.87s\tremaining: 51.1s\n",
      "87:\tlearn: 19.9533262\ttest: 20.6311594\tbest: 20.6311594 (87)\ttotal: 4.92s\tremaining: 51s\n",
      "88:\tlearn: 19.9441068\ttest: 20.6288965\tbest: 20.6288965 (88)\ttotal: 4.97s\tremaining: 50.9s\n",
      "89:\tlearn: 19.9289387\ttest: 20.6179303\tbest: 20.6179303 (89)\ttotal: 5.02s\tremaining: 50.8s\n",
      "90:\tlearn: 19.9063314\ttest: 20.6049632\tbest: 20.6049632 (90)\ttotal: 5.07s\tremaining: 50.6s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "91:\tlearn: 19.8937860\ttest: 20.5976608\tbest: 20.5976608 (91)\ttotal: 5.12s\tremaining: 50.6s\n",
      "92:\tlearn: 19.8806913\ttest: 20.5929463\tbest: 20.5929463 (92)\ttotal: 5.18s\tremaining: 50.5s\n",
      "93:\tlearn: 19.8592544\ttest: 20.5786896\tbest: 20.5786896 (93)\ttotal: 5.24s\tremaining: 50.5s\n",
      "94:\tlearn: 19.8417541\ttest: 20.5669827\tbest: 20.5669827 (94)\ttotal: 5.3s\tremaining: 50.5s\n",
      "95:\tlearn: 19.8307640\ttest: 20.5617950\tbest: 20.5617950 (95)\ttotal: 5.37s\tremaining: 50.5s\n",
      "96:\tlearn: 19.8089873\ttest: 20.5470972\tbest: 20.5470972 (96)\ttotal: 5.43s\tremaining: 50.5s\n",
      "97:\tlearn: 19.7876890\ttest: 20.5379554\tbest: 20.5379554 (97)\ttotal: 5.48s\tremaining: 50.5s\n",
      "98:\tlearn: 19.7807996\ttest: 20.5333419\tbest: 20.5333419 (98)\ttotal: 5.53s\tremaining: 50.3s\n",
      "99:\tlearn: 19.7644749\ttest: 20.5232801\tbest: 20.5232801 (99)\ttotal: 5.59s\tremaining: 50.3s\n",
      "100:\tlearn: 19.7493727\ttest: 20.5159449\tbest: 20.5159449 (100)\ttotal: 5.65s\tremaining: 50.3s\n",
      "101:\tlearn: 19.7332162\ttest: 20.5068519\tbest: 20.5068519 (101)\ttotal: 5.71s\tremaining: 50.3s\n",
      "102:\tlearn: 19.7207326\ttest: 20.4979783\tbest: 20.4979783 (102)\ttotal: 5.77s\tremaining: 50.3s\n",
      "103:\tlearn: 19.7080709\ttest: 20.4912700\tbest: 20.4912700 (103)\ttotal: 5.84s\tremaining: 50.3s\n",
      "104:\tlearn: 19.6941615\ttest: 20.4840695\tbest: 20.4840695 (104)\ttotal: 5.9s\tremaining: 50.3s\n",
      "105:\tlearn: 19.6857420\ttest: 20.4829172\tbest: 20.4829172 (105)\ttotal: 5.97s\tremaining: 50.3s\n",
      "106:\tlearn: 19.6766671\ttest: 20.4788246\tbest: 20.4788246 (106)\ttotal: 6.01s\tremaining: 50.2s\n",
      "107:\tlearn: 19.6564752\ttest: 20.4657289\tbest: 20.4657289 (107)\ttotal: 6.08s\tremaining: 50.2s\n",
      "108:\tlearn: 19.6381822\ttest: 20.4565821\tbest: 20.4565821 (108)\ttotal: 6.15s\tremaining: 50.3s\n",
      "109:\tlearn: 19.6377216\ttest: 20.4558471\tbest: 20.4558471 (109)\ttotal: 6.16s\tremaining: 49.8s\n",
      "110:\tlearn: 19.6274833\ttest: 20.4491750\tbest: 20.4491750 (110)\ttotal: 6.21s\tremaining: 49.8s\n",
      "111:\tlearn: 19.6137166\ttest: 20.4430249\tbest: 20.4430249 (111)\ttotal: 6.28s\tremaining: 49.8s\n",
      "112:\tlearn: 19.5996144\ttest: 20.4337271\tbest: 20.4337271 (112)\ttotal: 6.34s\tremaining: 49.7s\n",
      "113:\tlearn: 19.5781915\ttest: 20.4196548\tbest: 20.4196548 (113)\ttotal: 6.39s\tremaining: 49.7s\n",
      "114:\tlearn: 19.5678958\ttest: 20.4123216\tbest: 20.4123216 (114)\ttotal: 6.44s\tremaining: 49.6s\n",
      "115:\tlearn: 19.5509691\ttest: 20.4017700\tbest: 20.4017700 (115)\ttotal: 6.5s\tremaining: 49.5s\n",
      "116:\tlearn: 19.5380183\ttest: 20.3925495\tbest: 20.3925495 (116)\ttotal: 6.56s\tremaining: 49.5s\n",
      "117:\tlearn: 19.5286785\ttest: 20.3877461\tbest: 20.3877461 (117)\ttotal: 6.61s\tremaining: 49.4s\n",
      "118:\tlearn: 19.5112377\ttest: 20.3827215\tbest: 20.3827215 (118)\ttotal: 6.67s\tremaining: 49.4s\n",
      "119:\tlearn: 19.4860283\ttest: 20.3654568\tbest: 20.3654568 (119)\ttotal: 6.73s\tremaining: 49.4s\n",
      "120:\tlearn: 19.4756405\ttest: 20.3593143\tbest: 20.3593143 (120)\ttotal: 6.78s\tremaining: 49.3s\n",
      "121:\tlearn: 19.4594374\ttest: 20.3469451\tbest: 20.3469451 (121)\ttotal: 6.84s\tremaining: 49.2s\n",
      "122:\tlearn: 19.4436935\ttest: 20.3385437\tbest: 20.3385437 (122)\ttotal: 6.89s\tremaining: 49.1s\n",
      "123:\tlearn: 19.4226676\ttest: 20.3269672\tbest: 20.3269672 (123)\ttotal: 6.96s\tremaining: 49.2s\n",
      "124:\tlearn: 19.4105579\ttest: 20.3222002\tbest: 20.3222002 (124)\ttotal: 7.02s\tremaining: 49.1s\n",
      "125:\tlearn: 19.4009554\ttest: 20.3173784\tbest: 20.3173784 (125)\ttotal: 7.08s\tremaining: 49.1s\n",
      "126:\tlearn: 19.3836898\ttest: 20.3062718\tbest: 20.3062718 (126)\ttotal: 7.14s\tremaining: 49.1s\n",
      "127:\tlearn: 19.3718353\ttest: 20.3000820\tbest: 20.3000820 (127)\ttotal: 7.19s\tremaining: 49s\n",
      "128:\tlearn: 19.3570904\ttest: 20.2927254\tbest: 20.2927254 (128)\ttotal: 7.24s\tremaining: 48.9s\n",
      "129:\tlearn: 19.3461597\ttest: 20.2870385\tbest: 20.2870385 (129)\ttotal: 7.29s\tremaining: 48.8s\n",
      "130:\tlearn: 19.3384591\ttest: 20.2839215\tbest: 20.2839215 (130)\ttotal: 7.33s\tremaining: 48.6s\n",
      "131:\tlearn: 19.3288368\ttest: 20.2765276\tbest: 20.2765276 (131)\ttotal: 7.39s\tremaining: 48.6s\n",
      "132:\tlearn: 19.3185423\ttest: 20.2715423\tbest: 20.2715423 (132)\ttotal: 7.46s\tremaining: 48.6s\n",
      "133:\tlearn: 19.2978564\ttest: 20.2582223\tbest: 20.2582223 (133)\ttotal: 7.51s\tremaining: 48.6s\n",
      "134:\tlearn: 19.2836678\ttest: 20.2512862\tbest: 20.2512862 (134)\ttotal: 7.57s\tremaining: 48.5s\n",
      "135:\tlearn: 19.2668457\ttest: 20.2424047\tbest: 20.2424047 (135)\ttotal: 7.62s\tremaining: 48.4s\n",
      "136:\tlearn: 19.2536618\ttest: 20.2336281\tbest: 20.2336281 (136)\ttotal: 7.66s\tremaining: 48.3s\n",
      "137:\tlearn: 19.2365400\ttest: 20.2228916\tbest: 20.2228916 (137)\ttotal: 7.71s\tremaining: 48.2s\n",
      "138:\tlearn: 19.2246186\ttest: 20.2180687\tbest: 20.2180687 (138)\ttotal: 7.77s\tremaining: 48.1s\n",
      "139:\tlearn: 19.2076168\ttest: 20.2064708\tbest: 20.2064708 (139)\ttotal: 7.82s\tremaining: 48.1s\n",
      "140:\tlearn: 19.2001608\ttest: 20.2047645\tbest: 20.2047645 (140)\ttotal: 7.87s\tremaining: 48s\n",
      "141:\tlearn: 19.1877948\ttest: 20.1985476\tbest: 20.1985476 (141)\ttotal: 7.92s\tremaining: 47.9s\n",
      "142:\tlearn: 19.1739167\ttest: 20.1914578\tbest: 20.1914578 (142)\ttotal: 7.98s\tremaining: 47.8s\n",
      "143:\tlearn: 19.1621871\ttest: 20.1854026\tbest: 20.1854026 (143)\ttotal: 8.03s\tremaining: 47.7s\n",
      "144:\tlearn: 19.1547001\ttest: 20.1810215\tbest: 20.1810215 (144)\ttotal: 8.08s\tremaining: 47.7s\n",
      "145:\tlearn: 19.1462605\ttest: 20.1759004\tbest: 20.1759004 (145)\ttotal: 8.13s\tremaining: 47.6s\n",
      "146:\tlearn: 19.1305866\ttest: 20.1677574\tbest: 20.1677574 (146)\ttotal: 8.18s\tremaining: 47.5s\n",
      "147:\tlearn: 19.1135087\ttest: 20.1560055\tbest: 20.1560055 (147)\ttotal: 8.23s\tremaining: 47.4s\n",
      "148:\tlearn: 19.0980617\ttest: 20.1478634\tbest: 20.1478634 (148)\ttotal: 8.28s\tremaining: 47.3s\n",
      "149:\tlearn: 19.0927202\ttest: 20.1455262\tbest: 20.1455262 (149)\ttotal: 8.34s\tremaining: 47.2s\n",
      "150:\tlearn: 19.0811699\ttest: 20.1393281\tbest: 20.1393281 (150)\ttotal: 8.39s\tremaining: 47.2s\n",
      "151:\tlearn: 19.0724522\ttest: 20.1334173\tbest: 20.1334173 (151)\ttotal: 8.44s\tremaining: 47.1s\n",
      "152:\tlearn: 19.0609204\ttest: 20.1270535\tbest: 20.1270535 (152)\ttotal: 8.49s\tremaining: 47s\n",
      "153:\tlearn: 19.0521904\ttest: 20.1244805\tbest: 20.1244805 (153)\ttotal: 8.54s\tremaining: 46.9s\n",
      "154:\tlearn: 19.0434130\ttest: 20.1203013\tbest: 20.1203013 (154)\ttotal: 8.6s\tremaining: 46.9s\n",
      "155:\tlearn: 19.0309869\ttest: 20.1125437\tbest: 20.1125437 (155)\ttotal: 8.66s\tremaining: 46.8s\n",
      "156:\tlearn: 19.0212813\ttest: 20.1071875\tbest: 20.1071875 (156)\ttotal: 8.71s\tremaining: 46.8s\n",
      "157:\tlearn: 19.0102506\ttest: 20.1018525\tbest: 20.1018525 (157)\ttotal: 8.76s\tremaining: 46.7s\n",
      "158:\tlearn: 19.0022748\ttest: 20.0997692\tbest: 20.0997692 (158)\ttotal: 8.8s\tremaining: 46.6s\n",
      "159:\tlearn: 18.9911085\ttest: 20.0935440\tbest: 20.0935440 (159)\ttotal: 8.85s\tremaining: 46.5s\n",
      "160:\tlearn: 18.9824962\ttest: 20.0891490\tbest: 20.0891490 (160)\ttotal: 8.91s\tremaining: 46.4s\n",
      "161:\tlearn: 18.9724649\ttest: 20.0828799\tbest: 20.0828799 (161)\ttotal: 8.95s\tremaining: 46.3s\n",
      "162:\tlearn: 18.9608349\ttest: 20.0764961\tbest: 20.0764961 (162)\ttotal: 9s\tremaining: 46.2s\n",
      "163:\tlearn: 18.9489859\ttest: 20.0707247\tbest: 20.0707247 (163)\ttotal: 9.05s\tremaining: 46.1s\n",
      "164:\tlearn: 18.9353313\ttest: 20.0613041\tbest: 20.0613041 (164)\ttotal: 9.1s\tremaining: 46.1s\n",
      "165:\tlearn: 18.9250783\ttest: 20.0591936\tbest: 20.0591936 (165)\ttotal: 9.15s\tremaining: 46s\n",
      "166:\tlearn: 18.9147608\ttest: 20.0543747\tbest: 20.0543747 (166)\ttotal: 9.2s\tremaining: 45.9s\n",
      "167:\tlearn: 18.9023277\ttest: 20.0474560\tbest: 20.0474560 (167)\ttotal: 9.24s\tremaining: 45.8s\n",
      "168:\tlearn: 18.8866921\ttest: 20.0375862\tbest: 20.0375862 (168)\ttotal: 9.29s\tremaining: 45.7s\n",
      "169:\tlearn: 18.8729592\ttest: 20.0301046\tbest: 20.0301046 (169)\ttotal: 9.34s\tremaining: 45.6s\n",
      "170:\tlearn: 18.8613589\ttest: 20.0243310\tbest: 20.0243310 (170)\ttotal: 9.39s\tremaining: 45.5s\n",
      "171:\tlearn: 18.8505085\ttest: 20.0182537\tbest: 20.0182537 (171)\ttotal: 9.45s\tremaining: 45.5s\n",
      "172:\tlearn: 18.8355229\ttest: 20.0114322\tbest: 20.0114322 (172)\ttotal: 9.5s\tremaining: 45.4s\n",
      "173:\tlearn: 18.8307624\ttest: 20.0099844\tbest: 20.0099844 (173)\ttotal: 9.55s\tremaining: 45.3s\n",
      "174:\tlearn: 18.8261647\ttest: 20.0093646\tbest: 20.0093646 (174)\ttotal: 9.6s\tremaining: 45.2s\n",
      "175:\tlearn: 18.8151806\ttest: 20.0040506\tbest: 20.0040506 (175)\ttotal: 9.64s\tremaining: 45.1s\n",
      "176:\tlearn: 18.8073029\ttest: 19.9994853\tbest: 19.9994853 (176)\ttotal: 9.69s\tremaining: 45.1s\n",
      "177:\tlearn: 18.7981330\ttest: 19.9932610\tbest: 19.9932610 (177)\ttotal: 9.73s\tremaining: 44.9s\n",
      "178:\tlearn: 18.7913824\ttest: 19.9914613\tbest: 19.9914613 (178)\ttotal: 9.79s\tremaining: 44.9s\n",
      "179:\tlearn: 18.7812483\ttest: 19.9859206\tbest: 19.9859206 (179)\ttotal: 9.86s\tremaining: 44.9s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "180:\tlearn: 18.7688017\ttest: 19.9796914\tbest: 19.9796914 (180)\ttotal: 9.91s\tremaining: 44.8s\n",
      "181:\tlearn: 18.7628546\ttest: 19.9789245\tbest: 19.9789245 (181)\ttotal: 9.96s\tremaining: 44.8s\n",
      "182:\tlearn: 18.7480398\ttest: 19.9682320\tbest: 19.9682320 (182)\ttotal: 10s\tremaining: 44.7s\n",
      "183:\tlearn: 18.7400123\ttest: 19.9644725\tbest: 19.9644725 (183)\ttotal: 10.1s\tremaining: 44.6s\n",
      "184:\tlearn: 18.7361774\ttest: 19.9644090\tbest: 19.9644090 (184)\ttotal: 10.1s\tremaining: 44.6s\n",
      "185:\tlearn: 18.7279986\ttest: 19.9612709\tbest: 19.9612709 (185)\ttotal: 10.2s\tremaining: 44.5s\n",
      "186:\tlearn: 18.7204149\ttest: 19.9575725\tbest: 19.9575725 (186)\ttotal: 10.2s\tremaining: 44.5s\n",
      "187:\tlearn: 18.7103172\ttest: 19.9535310\tbest: 19.9535310 (187)\ttotal: 10.3s\tremaining: 44.4s\n",
      "188:\tlearn: 18.6991811\ttest: 19.9473771\tbest: 19.9473771 (188)\ttotal: 10.3s\tremaining: 44.3s\n",
      "189:\tlearn: 18.6882012\ttest: 19.9404139\tbest: 19.9404139 (189)\ttotal: 10.4s\tremaining: 44.2s\n",
      "190:\tlearn: 18.6749815\ttest: 19.9331782\tbest: 19.9331782 (190)\ttotal: 10.4s\tremaining: 44.1s\n",
      "191:\tlearn: 18.6654017\ttest: 19.9271358\tbest: 19.9271358 (191)\ttotal: 10.5s\tremaining: 44.1s\n",
      "192:\tlearn: 18.6527558\ttest: 19.9184028\tbest: 19.9184028 (192)\ttotal: 10.5s\tremaining: 44s\n",
      "193:\tlearn: 18.6370132\ttest: 19.9117529\tbest: 19.9117529 (193)\ttotal: 10.6s\tremaining: 43.9s\n",
      "194:\tlearn: 18.6308933\ttest: 19.9079538\tbest: 19.9079538 (194)\ttotal: 10.6s\tremaining: 43.8s\n",
      "195:\tlearn: 18.6182095\ttest: 19.9038260\tbest: 19.9038260 (195)\ttotal: 10.7s\tremaining: 43.8s\n",
      "196:\tlearn: 18.6059468\ttest: 19.8979045\tbest: 19.8979045 (196)\ttotal: 10.7s\tremaining: 43.7s\n",
      "197:\tlearn: 18.5931064\ttest: 19.8910168\tbest: 19.8910168 (197)\ttotal: 10.8s\tremaining: 43.7s\n",
      "198:\tlearn: 18.5846699\ttest: 19.8880540\tbest: 19.8880540 (198)\ttotal: 10.8s\tremaining: 43.6s\n",
      "199:\tlearn: 18.5723050\ttest: 19.8810999\tbest: 19.8810999 (199)\ttotal: 10.9s\tremaining: 43.5s\n",
      "200:\tlearn: 18.5628640\ttest: 19.8772555\tbest: 19.8772555 (200)\ttotal: 10.9s\tremaining: 43.5s\n",
      "201:\tlearn: 18.5531893\ttest: 19.8731378\tbest: 19.8731378 (201)\ttotal: 11s\tremaining: 43.4s\n",
      "202:\tlearn: 18.5344913\ttest: 19.8640283\tbest: 19.8640283 (202)\ttotal: 11s\tremaining: 43.4s\n",
      "203:\tlearn: 18.5240311\ttest: 19.8571202\tbest: 19.8571202 (203)\ttotal: 11.1s\tremaining: 43.3s\n",
      "204:\tlearn: 18.5114924\ttest: 19.8514374\tbest: 19.8514374 (204)\ttotal: 11.2s\tremaining: 43.3s\n",
      "205:\tlearn: 18.4977235\ttest: 19.8450590\tbest: 19.8450590 (205)\ttotal: 11.2s\tremaining: 43.2s\n",
      "206:\tlearn: 18.4906079\ttest: 19.8408665\tbest: 19.8408665 (206)\ttotal: 11.3s\tremaining: 43.2s\n",
      "207:\tlearn: 18.4815996\ttest: 19.8359413\tbest: 19.8359413 (207)\ttotal: 11.3s\tremaining: 43.1s\n",
      "208:\tlearn: 18.4746296\ttest: 19.8317567\tbest: 19.8317567 (208)\ttotal: 11.4s\tremaining: 43.1s\n",
      "209:\tlearn: 18.4591367\ttest: 19.8236575\tbest: 19.8236575 (209)\ttotal: 11.4s\tremaining: 43s\n",
      "210:\tlearn: 18.4473468\ttest: 19.8162334\tbest: 19.8162334 (210)\ttotal: 11.5s\tremaining: 42.9s\n",
      "211:\tlearn: 18.4367445\ttest: 19.8112580\tbest: 19.8112580 (211)\ttotal: 11.5s\tremaining: 42.8s\n",
      "212:\tlearn: 18.4201504\ttest: 19.8041223\tbest: 19.8041223 (212)\ttotal: 11.6s\tremaining: 42.8s\n",
      "213:\tlearn: 18.4128507\ttest: 19.8011121\tbest: 19.8011121 (213)\ttotal: 11.6s\tremaining: 42.7s\n",
      "214:\tlearn: 18.4015768\ttest: 19.7955417\tbest: 19.7955417 (214)\ttotal: 11.7s\tremaining: 42.7s\n",
      "215:\tlearn: 18.3927965\ttest: 19.7928164\tbest: 19.7928164 (215)\ttotal: 11.7s\tremaining: 42.6s\n",
      "216:\tlearn: 18.3803636\ttest: 19.7871574\tbest: 19.7871574 (216)\ttotal: 11.8s\tremaining: 42.6s\n",
      "217:\tlearn: 18.3731436\ttest: 19.7848850\tbest: 19.7848850 (217)\ttotal: 11.9s\tremaining: 42.5s\n",
      "218:\tlearn: 18.3651487\ttest: 19.7827265\tbest: 19.7827265 (218)\ttotal: 11.9s\tremaining: 42.5s\n",
      "219:\tlearn: 18.3539493\ttest: 19.7780361\tbest: 19.7780361 (219)\ttotal: 12s\tremaining: 42.4s\n",
      "220:\tlearn: 18.3459706\ttest: 19.7758662\tbest: 19.7758662 (220)\ttotal: 12s\tremaining: 42.3s\n",
      "221:\tlearn: 18.3351836\ttest: 19.7679372\tbest: 19.7679372 (221)\ttotal: 12.1s\tremaining: 42.3s\n",
      "222:\tlearn: 18.3235177\ttest: 19.7619770\tbest: 19.7619770 (222)\ttotal: 12.1s\tremaining: 42.2s\n",
      "223:\tlearn: 18.3136966\ttest: 19.7564722\tbest: 19.7564722 (223)\ttotal: 12.2s\tremaining: 42.1s\n",
      "224:\tlearn: 18.3061972\ttest: 19.7534750\tbest: 19.7534750 (224)\ttotal: 12.2s\tremaining: 42s\n",
      "225:\tlearn: 18.2940382\ttest: 19.7460625\tbest: 19.7460625 (225)\ttotal: 12.3s\tremaining: 42s\n",
      "226:\tlearn: 18.2839089\ttest: 19.7410069\tbest: 19.7410069 (226)\ttotal: 12.3s\tremaining: 41.9s\n",
      "227:\tlearn: 18.2762610\ttest: 19.7389261\tbest: 19.7389261 (227)\ttotal: 12.4s\tremaining: 41.9s\n",
      "228:\tlearn: 18.2643946\ttest: 19.7336778\tbest: 19.7336778 (228)\ttotal: 12.4s\tremaining: 41.8s\n",
      "229:\tlearn: 18.2580706\ttest: 19.7327869\tbest: 19.7327869 (229)\ttotal: 12.5s\tremaining: 41.8s\n",
      "230:\tlearn: 18.2468957\ttest: 19.7284569\tbest: 19.7284569 (230)\ttotal: 12.5s\tremaining: 41.7s\n",
      "231:\tlearn: 18.2356762\ttest: 19.7249036\tbest: 19.7249036 (231)\ttotal: 12.6s\tremaining: 41.7s\n",
      "232:\tlearn: 18.2296110\ttest: 19.7241663\tbest: 19.7241663 (232)\ttotal: 12.6s\tremaining: 41.6s\n",
      "233:\tlearn: 18.2190514\ttest: 19.7211092\tbest: 19.7211092 (233)\ttotal: 12.7s\tremaining: 41.5s\n",
      "234:\tlearn: 18.2057382\ttest: 19.7109121\tbest: 19.7109121 (234)\ttotal: 12.7s\tremaining: 41.4s\n",
      "235:\tlearn: 18.1963988\ttest: 19.7065478\tbest: 19.7065478 (235)\ttotal: 12.8s\tremaining: 41.4s\n",
      "236:\tlearn: 18.1835370\ttest: 19.6958084\tbest: 19.6958084 (236)\ttotal: 12.8s\tremaining: 41.3s\n",
      "237:\tlearn: 18.1715848\ttest: 19.6875936\tbest: 19.6875936 (237)\ttotal: 12.9s\tremaining: 41.3s\n",
      "238:\tlearn: 18.1607523\ttest: 19.6834247\tbest: 19.6834247 (238)\ttotal: 12.9s\tremaining: 41.2s\n",
      "239:\tlearn: 18.1490861\ttest: 19.6784734\tbest: 19.6784734 (239)\ttotal: 13s\tremaining: 41.2s\n",
      "240:\tlearn: 18.1379833\ttest: 19.6733264\tbest: 19.6733264 (240)\ttotal: 13s\tremaining: 41.1s\n",
      "241:\tlearn: 18.1260671\ttest: 19.6675982\tbest: 19.6675982 (241)\ttotal: 13.1s\tremaining: 41s\n",
      "242:\tlearn: 18.1173903\ttest: 19.6639328\tbest: 19.6639328 (242)\ttotal: 13.2s\tremaining: 41s\n",
      "243:\tlearn: 18.1063676\ttest: 19.6587498\tbest: 19.6587498 (243)\ttotal: 13.2s\tremaining: 40.9s\n",
      "244:\tlearn: 18.0991333\ttest: 19.6547839\tbest: 19.6547839 (244)\ttotal: 13.3s\tremaining: 40.9s\n",
      "245:\tlearn: 18.0838081\ttest: 19.6474647\tbest: 19.6474647 (245)\ttotal: 13.3s\tremaining: 40.8s\n",
      "246:\tlearn: 18.0771369\ttest: 19.6463268\tbest: 19.6463268 (246)\ttotal: 13.4s\tremaining: 40.7s\n",
      "247:\tlearn: 18.0700838\ttest: 19.6436987\tbest: 19.6436987 (247)\ttotal: 13.4s\tremaining: 40.7s\n",
      "248:\tlearn: 18.0579809\ttest: 19.6373024\tbest: 19.6373024 (248)\ttotal: 13.5s\tremaining: 40.6s\n",
      "249:\tlearn: 18.0466707\ttest: 19.6313574\tbest: 19.6313574 (249)\ttotal: 13.5s\tremaining: 40.6s\n",
      "250:\tlearn: 18.0348940\ttest: 19.6241660\tbest: 19.6241660 (250)\ttotal: 13.6s\tremaining: 40.5s\n",
      "251:\tlearn: 18.0298721\ttest: 19.6226593\tbest: 19.6226593 (251)\ttotal: 13.6s\tremaining: 40.5s\n",
      "252:\tlearn: 18.0227635\ttest: 19.6191202\tbest: 19.6191202 (252)\ttotal: 13.7s\tremaining: 40.4s\n",
      "253:\tlearn: 18.0157566\ttest: 19.6172999\tbest: 19.6172999 (253)\ttotal: 13.7s\tremaining: 40.4s\n",
      "254:\tlearn: 18.0053224\ttest: 19.6144403\tbest: 19.6144403 (254)\ttotal: 13.8s\tremaining: 40.3s\n",
      "255:\tlearn: 17.9996618\ttest: 19.6128360\tbest: 19.6128360 (255)\ttotal: 13.9s\tremaining: 40.3s\n",
      "256:\tlearn: 17.9885189\ttest: 19.6084482\tbest: 19.6084482 (256)\ttotal: 13.9s\tremaining: 40.2s\n",
      "257:\tlearn: 17.9785885\ttest: 19.6041504\tbest: 19.6041504 (257)\ttotal: 14s\tremaining: 40.2s\n",
      "258:\tlearn: 17.9678113\ttest: 19.6001304\tbest: 19.6001304 (258)\ttotal: 14s\tremaining: 40.1s\n",
      "259:\tlearn: 17.9632412\ttest: 19.5979564\tbest: 19.5979564 (259)\ttotal: 14.1s\tremaining: 40s\n",
      "260:\tlearn: 17.9532217\ttest: 19.5931441\tbest: 19.5931441 (260)\ttotal: 14.1s\tremaining: 40s\n",
      "261:\tlearn: 17.9453601\ttest: 19.5916858\tbest: 19.5916858 (261)\ttotal: 14.2s\tremaining: 39.9s\n",
      "262:\tlearn: 17.9357162\ttest: 19.5873793\tbest: 19.5873793 (262)\ttotal: 14.2s\tremaining: 39.8s\n",
      "263:\tlearn: 17.9302293\ttest: 19.5847478\tbest: 19.5847478 (263)\ttotal: 14.3s\tremaining: 39.7s\n",
      "264:\tlearn: 17.9225954\ttest: 19.5816535\tbest: 19.5816535 (264)\ttotal: 14.3s\tremaining: 39.6s\n",
      "265:\tlearn: 17.9164174\ttest: 19.5804925\tbest: 19.5804925 (265)\ttotal: 14.3s\tremaining: 39.6s\n",
      "266:\tlearn: 17.9073817\ttest: 19.5766923\tbest: 19.5766923 (266)\ttotal: 14.4s\tremaining: 39.5s\n",
      "267:\tlearn: 17.8925186\ttest: 19.5695640\tbest: 19.5695640 (267)\ttotal: 14.4s\tremaining: 39.5s\n",
      "268:\tlearn: 17.8817381\ttest: 19.5673504\tbest: 19.5673504 (268)\ttotal: 14.5s\tremaining: 39.4s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "269:\tlearn: 17.8760475\ttest: 19.5655871\tbest: 19.5655871 (269)\ttotal: 14.5s\tremaining: 39.3s\n",
      "270:\tlearn: 17.8660791\ttest: 19.5622267\tbest: 19.5622267 (270)\ttotal: 14.6s\tremaining: 39.3s\n",
      "271:\tlearn: 17.8544128\ttest: 19.5554543\tbest: 19.5554543 (271)\ttotal: 14.7s\tremaining: 39.2s\n",
      "272:\tlearn: 17.8478226\ttest: 19.5544436\tbest: 19.5544436 (272)\ttotal: 14.7s\tremaining: 39.2s\n",
      "273:\tlearn: 17.8367082\ttest: 19.5484945\tbest: 19.5484945 (273)\ttotal: 14.8s\tremaining: 39.1s\n",
      "274:\tlearn: 17.8311689\ttest: 19.5482691\tbest: 19.5482691 (274)\ttotal: 14.8s\tremaining: 39.1s\n",
      "275:\tlearn: 17.8238743\ttest: 19.5459187\tbest: 19.5459187 (275)\ttotal: 14.9s\tremaining: 39s\n",
      "276:\tlearn: 17.8198506\ttest: 19.5455953\tbest: 19.5455953 (276)\ttotal: 14.9s\tremaining: 39s\n",
      "277:\tlearn: 17.8141577\ttest: 19.5420902\tbest: 19.5420902 (277)\ttotal: 15s\tremaining: 38.9s\n",
      "278:\tlearn: 17.8040099\ttest: 19.5359468\tbest: 19.5359468 (278)\ttotal: 15.1s\tremaining: 38.9s\n",
      "279:\tlearn: 17.7939896\ttest: 19.5308907\tbest: 19.5308907 (279)\ttotal: 15.1s\tremaining: 38.9s\n",
      "280:\tlearn: 17.7886205\ttest: 19.5273598\tbest: 19.5273598 (280)\ttotal: 15.2s\tremaining: 38.8s\n",
      "281:\tlearn: 17.7809532\ttest: 19.5238576\tbest: 19.5238576 (281)\ttotal: 15.2s\tremaining: 38.7s\n",
      "282:\tlearn: 17.7686577\ttest: 19.5181117\tbest: 19.5181117 (282)\ttotal: 15.3s\tremaining: 38.7s\n",
      "283:\tlearn: 17.7618311\ttest: 19.5148437\tbest: 19.5148437 (283)\ttotal: 15.3s\tremaining: 38.6s\n",
      "284:\tlearn: 17.7566017\ttest: 19.5127854\tbest: 19.5127854 (284)\ttotal: 15.4s\tremaining: 38.6s\n",
      "285:\tlearn: 17.7503183\ttest: 19.5110162\tbest: 19.5110162 (285)\ttotal: 15.4s\tremaining: 38.6s\n",
      "286:\tlearn: 17.7408314\ttest: 19.5060457\tbest: 19.5060457 (286)\ttotal: 15.5s\tremaining: 38.5s\n",
      "287:\tlearn: 17.7314321\ttest: 19.5014194\tbest: 19.5014194 (287)\ttotal: 15.6s\tremaining: 38.5s\n",
      "288:\tlearn: 17.7219005\ttest: 19.4970752\tbest: 19.4970752 (288)\ttotal: 15.6s\tremaining: 38.4s\n",
      "289:\tlearn: 17.7113579\ttest: 19.4921685\tbest: 19.4921685 (289)\ttotal: 15.7s\tremaining: 38.3s\n",
      "290:\tlearn: 17.7038166\ttest: 19.4875526\tbest: 19.4875526 (290)\ttotal: 15.7s\tremaining: 38.3s\n",
      "291:\tlearn: 17.6966537\ttest: 19.4839056\tbest: 19.4839056 (291)\ttotal: 15.8s\tremaining: 38.3s\n",
      "292:\tlearn: 17.6857925\ttest: 19.4788898\tbest: 19.4788898 (292)\ttotal: 15.8s\tremaining: 38.2s\n",
      "293:\tlearn: 17.6793356\ttest: 19.4760524\tbest: 19.4760524 (293)\ttotal: 15.9s\tremaining: 38.2s\n",
      "294:\tlearn: 17.6690618\ttest: 19.4706583\tbest: 19.4706583 (294)\ttotal: 16s\tremaining: 38.1s\n",
      "295:\tlearn: 17.6606743\ttest: 19.4647104\tbest: 19.4647104 (295)\ttotal: 16s\tremaining: 38.1s\n",
      "296:\tlearn: 17.6505277\ttest: 19.4606726\tbest: 19.4606726 (296)\ttotal: 16.1s\tremaining: 38.1s\n",
      "297:\tlearn: 17.6449483\ttest: 19.4586836\tbest: 19.4586836 (297)\ttotal: 16.1s\tremaining: 38s\n",
      "298:\tlearn: 17.6378985\ttest: 19.4564994\tbest: 19.4564994 (298)\ttotal: 16.2s\tremaining: 38s\n",
      "299:\tlearn: 17.6301905\ttest: 19.4534007\tbest: 19.4534007 (299)\ttotal: 16.2s\tremaining: 37.9s\n",
      "300:\tlearn: 17.6214945\ttest: 19.4492012\tbest: 19.4492012 (300)\ttotal: 16.3s\tremaining: 37.9s\n",
      "301:\tlearn: 17.6124654\ttest: 19.4466231\tbest: 19.4466231 (301)\ttotal: 16.4s\tremaining: 37.8s\n",
      "302:\tlearn: 17.6036310\ttest: 19.4426223\tbest: 19.4426223 (302)\ttotal: 16.4s\tremaining: 37.8s\n",
      "303:\tlearn: 17.5911428\ttest: 19.4347485\tbest: 19.4347485 (303)\ttotal: 16.5s\tremaining: 37.8s\n",
      "304:\tlearn: 17.5830339\ttest: 19.4295897\tbest: 19.4295897 (304)\ttotal: 16.6s\tremaining: 37.7s\n",
      "305:\tlearn: 17.5725124\ttest: 19.4255409\tbest: 19.4255409 (305)\ttotal: 16.6s\tremaining: 37.7s\n",
      "306:\tlearn: 17.5658402\ttest: 19.4226721\tbest: 19.4226721 (306)\ttotal: 16.7s\tremaining: 37.6s\n",
      "307:\tlearn: 17.5563551\ttest: 19.4180365\tbest: 19.4180365 (307)\ttotal: 16.7s\tremaining: 37.6s\n",
      "308:\tlearn: 17.5488802\ttest: 19.4133548\tbest: 19.4133548 (308)\ttotal: 16.8s\tremaining: 37.6s\n",
      "309:\tlearn: 17.5421619\ttest: 19.4102297\tbest: 19.4102297 (309)\ttotal: 16.9s\tremaining: 37.5s\n",
      "310:\tlearn: 17.5321720\ttest: 19.4035003\tbest: 19.4035003 (310)\ttotal: 16.9s\tremaining: 37.5s\n",
      "311:\tlearn: 17.5248778\ttest: 19.4004797\tbest: 19.4004797 (311)\ttotal: 17s\tremaining: 37.4s\n",
      "312:\tlearn: 17.5166915\ttest: 19.3981141\tbest: 19.3981141 (312)\ttotal: 17s\tremaining: 37.4s\n",
      "313:\tlearn: 17.5070089\ttest: 19.3938996\tbest: 19.3938996 (313)\ttotal: 17.1s\tremaining: 37.3s\n",
      "314:\tlearn: 17.5013079\ttest: 19.3906396\tbest: 19.3906396 (314)\ttotal: 17.1s\tremaining: 37.3s\n",
      "315:\tlearn: 17.4915792\ttest: 19.3883094\tbest: 19.3883094 (315)\ttotal: 17.2s\tremaining: 37.2s\n",
      "316:\tlearn: 17.4841249\ttest: 19.3851677\tbest: 19.3851677 (316)\ttotal: 17.2s\tremaining: 37.2s\n",
      "317:\tlearn: 17.4750727\ttest: 19.3798888\tbest: 19.3798888 (317)\ttotal: 17.3s\tremaining: 37.1s\n",
      "318:\tlearn: 17.4709408\ttest: 19.3789225\tbest: 19.3789225 (318)\ttotal: 17.4s\tremaining: 37.1s\n",
      "319:\tlearn: 17.4612827\ttest: 19.3728228\tbest: 19.3728228 (319)\ttotal: 17.4s\tremaining: 37s\n",
      "320:\tlearn: 17.4539686\ttest: 19.3695665\tbest: 19.3695665 (320)\ttotal: 17.5s\tremaining: 37s\n",
      "321:\tlearn: 17.4431667\ttest: 19.3667465\tbest: 19.3667465 (321)\ttotal: 17.5s\tremaining: 36.9s\n",
      "322:\tlearn: 17.4317345\ttest: 19.3616027\tbest: 19.3616027 (322)\ttotal: 17.6s\tremaining: 36.9s\n",
      "323:\tlearn: 17.4235622\ttest: 19.3572461\tbest: 19.3572461 (323)\ttotal: 17.6s\tremaining: 36.8s\n",
      "324:\tlearn: 17.4095816\ttest: 19.3479580\tbest: 19.3479580 (324)\ttotal: 17.7s\tremaining: 36.8s\n",
      "325:\tlearn: 17.4056857\ttest: 19.3468091\tbest: 19.3468091 (325)\ttotal: 17.8s\tremaining: 36.7s\n",
      "326:\tlearn: 17.3994160\ttest: 19.3417510\tbest: 19.3417510 (326)\ttotal: 17.8s\tremaining: 36.7s\n",
      "327:\tlearn: 17.3911537\ttest: 19.3390937\tbest: 19.3390937 (327)\ttotal: 17.9s\tremaining: 36.6s\n",
      "328:\tlearn: 17.3811343\ttest: 19.3352590\tbest: 19.3352590 (328)\ttotal: 17.9s\tremaining: 36.5s\n",
      "329:\tlearn: 17.3726320\ttest: 19.3300548\tbest: 19.3300548 (329)\ttotal: 18s\tremaining: 36.5s\n",
      "330:\tlearn: 17.3655040\ttest: 19.3275906\tbest: 19.3275906 (330)\ttotal: 18s\tremaining: 36.4s\n",
      "331:\tlearn: 17.3617081\ttest: 19.3276453\tbest: 19.3275906 (330)\ttotal: 18.1s\tremaining: 36.4s\n",
      "332:\tlearn: 17.3537714\ttest: 19.3249324\tbest: 19.3249324 (332)\ttotal: 18.1s\tremaining: 36.3s\n",
      "333:\tlearn: 17.3445113\ttest: 19.3221169\tbest: 19.3221169 (333)\ttotal: 18.2s\tremaining: 36.3s\n",
      "334:\tlearn: 17.3357178\ttest: 19.3198386\tbest: 19.3198386 (334)\ttotal: 18.3s\tremaining: 36.3s\n",
      "335:\tlearn: 17.3289762\ttest: 19.3167064\tbest: 19.3167064 (335)\ttotal: 18.3s\tremaining: 36.2s\n",
      "336:\tlearn: 17.3193955\ttest: 19.3128023\tbest: 19.3128023 (336)\ttotal: 18.4s\tremaining: 36.2s\n",
      "337:\tlearn: 17.3104581\ttest: 19.3100516\tbest: 19.3100516 (337)\ttotal: 18.4s\tremaining: 36.1s\n",
      "338:\tlearn: 17.3037947\ttest: 19.3080177\tbest: 19.3080177 (338)\ttotal: 18.5s\tremaining: 36.1s\n",
      "339:\tlearn: 17.2980972\ttest: 19.3043396\tbest: 19.3043396 (339)\ttotal: 18.5s\tremaining: 36s\n",
      "340:\tlearn: 17.2897884\ttest: 19.3023849\tbest: 19.3023849 (340)\ttotal: 18.6s\tremaining: 36s\n",
      "341:\tlearn: 17.2789462\ttest: 19.2972394\tbest: 19.2972394 (341)\ttotal: 18.7s\tremaining: 35.9s\n",
      "342:\tlearn: 17.2761933\ttest: 19.2966993\tbest: 19.2966993 (342)\ttotal: 18.7s\tremaining: 35.9s\n",
      "343:\tlearn: 17.2661967\ttest: 19.2931821\tbest: 19.2931821 (343)\ttotal: 18.8s\tremaining: 35.8s\n",
      "344:\tlearn: 17.2611852\ttest: 19.2912387\tbest: 19.2912387 (344)\ttotal: 18.9s\tremaining: 35.8s\n",
      "345:\tlearn: 17.2539483\ttest: 19.2893547\tbest: 19.2893547 (345)\ttotal: 18.9s\tremaining: 35.8s\n",
      "346:\tlearn: 17.2441320\ttest: 19.2851824\tbest: 19.2851824 (346)\ttotal: 19s\tremaining: 35.7s\n",
      "347:\tlearn: 17.2364706\ttest: 19.2833339\tbest: 19.2833339 (347)\ttotal: 19s\tremaining: 35.7s\n",
      "348:\tlearn: 17.2282049\ttest: 19.2791239\tbest: 19.2791239 (348)\ttotal: 19.1s\tremaining: 35.6s\n",
      "349:\tlearn: 17.2219798\ttest: 19.2773450\tbest: 19.2773450 (349)\ttotal: 19.1s\tremaining: 35.6s\n",
      "350:\tlearn: 17.2180503\ttest: 19.2770372\tbest: 19.2770372 (350)\ttotal: 19.2s\tremaining: 35.5s\n",
      "351:\tlearn: 17.2105475\ttest: 19.2725759\tbest: 19.2725759 (351)\ttotal: 19.2s\tremaining: 35.4s\n",
      "352:\tlearn: 17.2043601\ttest: 19.2702633\tbest: 19.2702633 (352)\ttotal: 19.3s\tremaining: 35.4s\n",
      "353:\tlearn: 17.1933892\ttest: 19.2660749\tbest: 19.2660749 (353)\ttotal: 19.4s\tremaining: 35.3s\n",
      "354:\tlearn: 17.1837579\ttest: 19.2621961\tbest: 19.2621961 (354)\ttotal: 19.4s\tremaining: 35.3s\n",
      "355:\tlearn: 17.1808344\ttest: 19.2613213\tbest: 19.2613213 (355)\ttotal: 19.5s\tremaining: 35.2s\n",
      "356:\tlearn: 17.1742024\ttest: 19.2588691\tbest: 19.2588691 (356)\ttotal: 19.5s\tremaining: 35.2s\n",
      "357:\tlearn: 17.1671590\ttest: 19.2574683\tbest: 19.2574683 (357)\ttotal: 19.6s\tremaining: 35.1s\n",
      "358:\tlearn: 17.1573445\ttest: 19.2529078\tbest: 19.2529078 (358)\ttotal: 19.6s\tremaining: 35s\n",
      "359:\tlearn: 17.1494301\ttest: 19.2503213\tbest: 19.2503213 (359)\ttotal: 19.7s\tremaining: 35s\n",
      "360:\tlearn: 17.1451346\ttest: 19.2490034\tbest: 19.2490034 (360)\ttotal: 19.8s\tremaining: 35s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "361:\tlearn: 17.1370222\ttest: 19.2453363\tbest: 19.2453363 (361)\ttotal: 19.8s\tremaining: 34.9s\n",
      "362:\tlearn: 17.1308297\ttest: 19.2429167\tbest: 19.2429167 (362)\ttotal: 19.9s\tremaining: 34.9s\n",
      "363:\tlearn: 17.1193537\ttest: 19.2397886\tbest: 19.2397886 (363)\ttotal: 19.9s\tremaining: 34.8s\n",
      "364:\tlearn: 17.1133203\ttest: 19.2362375\tbest: 19.2362375 (364)\ttotal: 20s\tremaining: 34.7s\n",
      "365:\tlearn: 17.1034450\ttest: 19.2312840\tbest: 19.2312840 (365)\ttotal: 20s\tremaining: 34.7s\n",
      "366:\tlearn: 17.0943524\ttest: 19.2280632\tbest: 19.2280632 (366)\ttotal: 20.1s\tremaining: 34.6s\n",
      "367:\tlearn: 17.0856333\ttest: 19.2237453\tbest: 19.2237453 (367)\ttotal: 20.1s\tremaining: 34.6s\n",
      "368:\tlearn: 17.0793954\ttest: 19.2223595\tbest: 19.2223595 (368)\ttotal: 20.2s\tremaining: 34.5s\n",
      "369:\tlearn: 17.0716030\ttest: 19.2185200\tbest: 19.2185200 (369)\ttotal: 20.2s\tremaining: 34.5s\n",
      "370:\tlearn: 17.0600166\ttest: 19.2140791\tbest: 19.2140791 (370)\ttotal: 20.3s\tremaining: 34.4s\n",
      "371:\tlearn: 17.0527681\ttest: 19.2093273\tbest: 19.2093273 (371)\ttotal: 20.4s\tremaining: 34.4s\n",
      "372:\tlearn: 17.0453605\ttest: 19.2060034\tbest: 19.2060034 (372)\ttotal: 20.4s\tremaining: 34.3s\n",
      "373:\tlearn: 17.0401478\ttest: 19.2039651\tbest: 19.2039651 (373)\ttotal: 20.5s\tremaining: 34.2s\n",
      "374:\tlearn: 17.0319930\ttest: 19.2004591\tbest: 19.2004591 (374)\ttotal: 20.5s\tremaining: 34.2s\n",
      "375:\tlearn: 17.0227145\ttest: 19.1970456\tbest: 19.1970456 (375)\ttotal: 20.6s\tremaining: 34.2s\n",
      "376:\tlearn: 17.0164515\ttest: 19.1957953\tbest: 19.1957953 (376)\ttotal: 20.6s\tremaining: 34.1s\n",
      "377:\tlearn: 17.0059379\ttest: 19.1891640\tbest: 19.1891640 (377)\ttotal: 20.7s\tremaining: 34.1s\n",
      "378:\tlearn: 16.9997561\ttest: 19.1860384\tbest: 19.1860384 (378)\ttotal: 20.8s\tremaining: 34s\n",
      "379:\tlearn: 16.9949038\ttest: 19.1839331\tbest: 19.1839331 (379)\ttotal: 20.8s\tremaining: 34s\n",
      "380:\tlearn: 16.9868368\ttest: 19.1807208\tbest: 19.1807208 (380)\ttotal: 20.9s\tremaining: 33.9s\n",
      "381:\tlearn: 16.9816644\ttest: 19.1782204\tbest: 19.1782204 (381)\ttotal: 20.9s\tremaining: 33.9s\n",
      "382:\tlearn: 16.9741651\ttest: 19.1746040\tbest: 19.1746040 (382)\ttotal: 21s\tremaining: 33.8s\n",
      "383:\tlearn: 16.9634897\ttest: 19.1700792\tbest: 19.1700792 (383)\ttotal: 21s\tremaining: 33.7s\n",
      "384:\tlearn: 16.9567414\ttest: 19.1687158\tbest: 19.1687158 (384)\ttotal: 21.1s\tremaining: 33.7s\n",
      "385:\tlearn: 16.9474896\ttest: 19.1652026\tbest: 19.1652026 (385)\ttotal: 21.1s\tremaining: 33.6s\n",
      "386:\tlearn: 16.9410416\ttest: 19.1621953\tbest: 19.1621953 (386)\ttotal: 21.2s\tremaining: 33.6s\n",
      "387:\tlearn: 16.9301742\ttest: 19.1560043\tbest: 19.1560043 (387)\ttotal: 21.2s\tremaining: 33.5s\n",
      "388:\tlearn: 16.9252171\ttest: 19.1542602\tbest: 19.1542602 (388)\ttotal: 21.3s\tremaining: 33.4s\n",
      "389:\tlearn: 16.9201663\ttest: 19.1513025\tbest: 19.1513025 (389)\ttotal: 21.4s\tremaining: 33.4s\n",
      "390:\tlearn: 16.9145675\ttest: 19.1491131\tbest: 19.1491131 (390)\ttotal: 21.4s\tremaining: 33.4s\n",
      "391:\tlearn: 16.9069287\ttest: 19.1460426\tbest: 19.1460426 (391)\ttotal: 21.5s\tremaining: 33.3s\n",
      "392:\tlearn: 16.9008646\ttest: 19.1430192\tbest: 19.1430192 (392)\ttotal: 21.5s\tremaining: 33.3s\n",
      "393:\tlearn: 16.8951767\ttest: 19.1399606\tbest: 19.1399606 (393)\ttotal: 21.6s\tremaining: 33.2s\n",
      "394:\tlearn: 16.8883495\ttest: 19.1380876\tbest: 19.1380876 (394)\ttotal: 21.6s\tremaining: 33.2s\n",
      "395:\tlearn: 16.8812524\ttest: 19.1350516\tbest: 19.1350516 (395)\ttotal: 21.7s\tremaining: 33.1s\n",
      "396:\tlearn: 16.8751927\ttest: 19.1327789\tbest: 19.1327789 (396)\ttotal: 21.8s\tremaining: 33s\n",
      "397:\tlearn: 16.8650378\ttest: 19.1282395\tbest: 19.1282395 (397)\ttotal: 21.8s\tremaining: 33s\n",
      "398:\tlearn: 16.8547631\ttest: 19.1249350\tbest: 19.1249350 (398)\ttotal: 21.9s\tremaining: 32.9s\n",
      "399:\tlearn: 16.8471820\ttest: 19.1207645\tbest: 19.1207645 (399)\ttotal: 21.9s\tremaining: 32.9s\n",
      "400:\tlearn: 16.8394500\ttest: 19.1184360\tbest: 19.1184360 (400)\ttotal: 22s\tremaining: 32.8s\n",
      "401:\tlearn: 16.8301874\ttest: 19.1137780\tbest: 19.1137780 (401)\ttotal: 22s\tremaining: 32.8s\n",
      "402:\tlearn: 16.8257746\ttest: 19.1135896\tbest: 19.1135896 (402)\ttotal: 22.1s\tremaining: 32.7s\n",
      "403:\tlearn: 16.8164614\ttest: 19.1093815\tbest: 19.1093815 (403)\ttotal: 22.1s\tremaining: 32.6s\n",
      "404:\tlearn: 16.8068022\ttest: 19.1051612\tbest: 19.1051612 (404)\ttotal: 22.2s\tremaining: 32.6s\n",
      "405:\tlearn: 16.8016642\ttest: 19.1033555\tbest: 19.1033555 (405)\ttotal: 22.2s\tremaining: 32.5s\n",
      "406:\tlearn: 16.7911991\ttest: 19.1000519\tbest: 19.1000519 (406)\ttotal: 22.3s\tremaining: 32.5s\n",
      "407:\tlearn: 16.7850953\ttest: 19.0968457\tbest: 19.0968457 (407)\ttotal: 22.3s\tremaining: 32.4s\n",
      "408:\tlearn: 16.7767266\ttest: 19.0937121\tbest: 19.0937121 (408)\ttotal: 22.4s\tremaining: 32.3s\n",
      "409:\tlearn: 16.7699034\ttest: 19.0918606\tbest: 19.0918606 (409)\ttotal: 22.4s\tremaining: 32.3s\n",
      "410:\tlearn: 16.7645003\ttest: 19.0895700\tbest: 19.0895700 (410)\ttotal: 22.5s\tremaining: 32.2s\n",
      "411:\tlearn: 16.7587910\ttest: 19.0879732\tbest: 19.0879732 (411)\ttotal: 22.6s\tremaining: 32.2s\n",
      "412:\tlearn: 16.7518975\ttest: 19.0858036\tbest: 19.0858036 (412)\ttotal: 22.6s\tremaining: 32.1s\n",
      "413:\tlearn: 16.7451102\ttest: 19.0825022\tbest: 19.0825022 (413)\ttotal: 22.7s\tremaining: 32.1s\n",
      "414:\tlearn: 16.7363257\ttest: 19.0782991\tbest: 19.0782991 (414)\ttotal: 22.7s\tremaining: 32s\n",
      "415:\tlearn: 16.7308433\ttest: 19.0753464\tbest: 19.0753464 (415)\ttotal: 22.8s\tremaining: 31.9s\n",
      "416:\tlearn: 16.7254355\ttest: 19.0724415\tbest: 19.0724415 (416)\ttotal: 22.8s\tremaining: 31.9s\n",
      "417:\tlearn: 16.7203227\ttest: 19.0705120\tbest: 19.0705120 (417)\ttotal: 22.9s\tremaining: 31.8s\n",
      "418:\tlearn: 16.7159187\ttest: 19.0678471\tbest: 19.0678471 (418)\ttotal: 22.9s\tremaining: 31.8s\n",
      "419:\tlearn: 16.7072281\ttest: 19.0641913\tbest: 19.0641913 (419)\ttotal: 23s\tremaining: 31.7s\n",
      "420:\tlearn: 16.6999059\ttest: 19.0615659\tbest: 19.0615659 (420)\ttotal: 23s\tremaining: 31.7s\n",
      "421:\tlearn: 16.6932601\ttest: 19.0578072\tbest: 19.0578072 (421)\ttotal: 23.1s\tremaining: 31.6s\n",
      "422:\tlearn: 16.6889768\ttest: 19.0569682\tbest: 19.0569682 (422)\ttotal: 23.1s\tremaining: 31.6s\n",
      "423:\tlearn: 16.6809892\ttest: 19.0553240\tbest: 19.0553240 (423)\ttotal: 23.2s\tremaining: 31.5s\n",
      "424:\tlearn: 16.6763761\ttest: 19.0532116\tbest: 19.0532116 (424)\ttotal: 23.2s\tremaining: 31.4s\n",
      "425:\tlearn: 16.6718156\ttest: 19.0518190\tbest: 19.0518190 (425)\ttotal: 23.3s\tremaining: 31.4s\n",
      "426:\tlearn: 16.6675945\ttest: 19.0507704\tbest: 19.0507704 (426)\ttotal: 23.3s\tremaining: 31.3s\n",
      "427:\tlearn: 16.6607330\ttest: 19.0495452\tbest: 19.0495452 (427)\ttotal: 23.4s\tremaining: 31.3s\n",
      "428:\tlearn: 16.6559946\ttest: 19.0489102\tbest: 19.0489102 (428)\ttotal: 23.4s\tremaining: 31.2s\n",
      "429:\tlearn: 16.6507523\ttest: 19.0458534\tbest: 19.0458534 (429)\ttotal: 23.5s\tremaining: 31.1s\n",
      "430:\tlearn: 16.6427241\ttest: 19.0429460\tbest: 19.0429460 (430)\ttotal: 23.5s\tremaining: 31.1s\n",
      "431:\tlearn: 16.6324865\ttest: 19.0387065\tbest: 19.0387065 (431)\ttotal: 23.6s\tremaining: 31s\n",
      "432:\tlearn: 16.6282859\ttest: 19.0372269\tbest: 19.0372269 (432)\ttotal: 23.7s\tremaining: 31s\n",
      "433:\tlearn: 16.6203565\ttest: 19.0331211\tbest: 19.0331211 (433)\ttotal: 23.7s\tremaining: 30.9s\n",
      "434:\tlearn: 16.6121062\ttest: 19.0303616\tbest: 19.0303616 (434)\ttotal: 23.8s\tremaining: 30.9s\n",
      "435:\tlearn: 16.6074218\ttest: 19.0280196\tbest: 19.0280196 (435)\ttotal: 23.8s\tremaining: 30.8s\n",
      "436:\tlearn: 16.6026847\ttest: 19.0276745\tbest: 19.0276745 (436)\ttotal: 23.9s\tremaining: 30.7s\n",
      "437:\tlearn: 16.5940950\ttest: 19.0243986\tbest: 19.0243986 (437)\ttotal: 23.9s\tremaining: 30.7s\n",
      "438:\tlearn: 16.5866756\ttest: 19.0204190\tbest: 19.0204190 (438)\ttotal: 24s\tremaining: 30.6s\n",
      "439:\tlearn: 16.5792077\ttest: 19.0178362\tbest: 19.0178362 (439)\ttotal: 24s\tremaining: 30.6s\n",
      "440:\tlearn: 16.5706987\ttest: 19.0132557\tbest: 19.0132557 (440)\ttotal: 24.1s\tremaining: 30.5s\n",
      "441:\tlearn: 16.5646796\ttest: 19.0093638\tbest: 19.0093638 (441)\ttotal: 24.1s\tremaining: 30.5s\n",
      "442:\tlearn: 16.5621081\ttest: 19.0083672\tbest: 19.0083672 (442)\ttotal: 24.2s\tremaining: 30.4s\n",
      "443:\tlearn: 16.5556850\ttest: 19.0062040\tbest: 19.0062040 (443)\ttotal: 24.2s\tremaining: 30.3s\n",
      "444:\tlearn: 16.5479103\ttest: 19.0024507\tbest: 19.0024507 (444)\ttotal: 24.3s\tremaining: 30.3s\n",
      "445:\tlearn: 16.5384860\ttest: 18.9990845\tbest: 18.9990845 (445)\ttotal: 24.3s\tremaining: 30.2s\n",
      "446:\tlearn: 16.5329160\ttest: 18.9972652\tbest: 18.9972652 (446)\ttotal: 24.4s\tremaining: 30.1s\n",
      "447:\tlearn: 16.5271199\ttest: 18.9964182\tbest: 18.9964182 (447)\ttotal: 24.4s\tremaining: 30.1s\n",
      "448:\tlearn: 16.5219774\ttest: 18.9950900\tbest: 18.9950900 (448)\ttotal: 24.5s\tremaining: 30s\n",
      "449:\tlearn: 16.5155677\ttest: 18.9925560\tbest: 18.9925560 (449)\ttotal: 24.5s\tremaining: 30s\n",
      "450:\tlearn: 16.5093657\ttest: 18.9900124\tbest: 18.9900124 (450)\ttotal: 24.6s\tremaining: 29.9s\n",
      "451:\tlearn: 16.5045917\ttest: 18.9877566\tbest: 18.9877566 (451)\ttotal: 24.6s\tremaining: 29.8s\n",
      "452:\tlearn: 16.4977119\ttest: 18.9836209\tbest: 18.9836209 (452)\ttotal: 24.7s\tremaining: 29.8s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "453:\tlearn: 16.4903894\ttest: 18.9813462\tbest: 18.9813462 (453)\ttotal: 24.7s\tremaining: 29.7s\n",
      "454:\tlearn: 16.4838599\ttest: 18.9792572\tbest: 18.9792572 (454)\ttotal: 24.8s\tremaining: 29.7s\n",
      "455:\tlearn: 16.4788769\ttest: 18.9772888\tbest: 18.9772888 (455)\ttotal: 24.8s\tremaining: 29.6s\n",
      "456:\tlearn: 16.4712484\ttest: 18.9736837\tbest: 18.9736837 (456)\ttotal: 24.9s\tremaining: 29.6s\n",
      "457:\tlearn: 16.4660372\ttest: 18.9708875\tbest: 18.9708875 (457)\ttotal: 24.9s\tremaining: 29.5s\n",
      "458:\tlearn: 16.4607064\ttest: 18.9689933\tbest: 18.9689933 (458)\ttotal: 25s\tremaining: 29.4s\n",
      "459:\tlearn: 16.4529606\ttest: 18.9666909\tbest: 18.9666909 (459)\ttotal: 25s\tremaining: 29.4s\n",
      "460:\tlearn: 16.4472755\ttest: 18.9638435\tbest: 18.9638435 (460)\ttotal: 25.1s\tremaining: 29.3s\n",
      "461:\tlearn: 16.4430392\ttest: 18.9638101\tbest: 18.9638101 (461)\ttotal: 25.1s\tremaining: 29.3s\n",
      "462:\tlearn: 16.4370623\ttest: 18.9627319\tbest: 18.9627319 (462)\ttotal: 25.2s\tremaining: 29.2s\n",
      "463:\tlearn: 16.4323912\ttest: 18.9617203\tbest: 18.9617203 (463)\ttotal: 25.2s\tremaining: 29.2s\n",
      "464:\tlearn: 16.4234825\ttest: 18.9582789\tbest: 18.9582789 (464)\ttotal: 25.3s\tremaining: 29.1s\n",
      "465:\tlearn: 16.4196337\ttest: 18.9574440\tbest: 18.9574440 (465)\ttotal: 25.4s\tremaining: 29.1s\n",
      "466:\tlearn: 16.4155045\ttest: 18.9576474\tbest: 18.9574440 (465)\ttotal: 25.4s\tremaining: 29s\n",
      "467:\tlearn: 16.4117012\ttest: 18.9570504\tbest: 18.9570504 (467)\ttotal: 25.5s\tremaining: 28.9s\n",
      "468:\tlearn: 16.4052307\ttest: 18.9542922\tbest: 18.9542922 (468)\ttotal: 25.5s\tremaining: 28.9s\n",
      "469:\tlearn: 16.4017452\ttest: 18.9531507\tbest: 18.9531507 (469)\ttotal: 25.6s\tremaining: 28.8s\n",
      "470:\tlearn: 16.3981813\ttest: 18.9527800\tbest: 18.9527800 (470)\ttotal: 25.6s\tremaining: 28.8s\n",
      "471:\tlearn: 16.3926458\ttest: 18.9512981\tbest: 18.9512981 (471)\ttotal: 25.7s\tremaining: 28.7s\n",
      "472:\tlearn: 16.3852495\ttest: 18.9482098\tbest: 18.9482098 (472)\ttotal: 25.7s\tremaining: 28.7s\n",
      "473:\tlearn: 16.3823645\ttest: 18.9473082\tbest: 18.9473082 (473)\ttotal: 25.8s\tremaining: 28.6s\n",
      "474:\tlearn: 16.3763155\ttest: 18.9446373\tbest: 18.9446373 (474)\ttotal: 25.8s\tremaining: 28.5s\n",
      "475:\tlearn: 16.3705865\ttest: 18.9432517\tbest: 18.9432517 (475)\ttotal: 25.9s\tremaining: 28.5s\n",
      "476:\tlearn: 16.3632609\ttest: 18.9407734\tbest: 18.9407734 (476)\ttotal: 25.9s\tremaining: 28.4s\n",
      "477:\tlearn: 16.3574156\ttest: 18.9389862\tbest: 18.9389862 (477)\ttotal: 26s\tremaining: 28.4s\n",
      "478:\tlearn: 16.3531923\ttest: 18.9371026\tbest: 18.9371026 (478)\ttotal: 26s\tremaining: 28.3s\n",
      "479:\tlearn: 16.3467075\ttest: 18.9330482\tbest: 18.9330482 (479)\ttotal: 26.1s\tremaining: 28.2s\n",
      "480:\tlearn: 16.3402398\ttest: 18.9312383\tbest: 18.9312383 (480)\ttotal: 26.1s\tremaining: 28.2s\n",
      "481:\tlearn: 16.3318232\ttest: 18.9278879\tbest: 18.9278879 (481)\ttotal: 26.2s\tremaining: 28.1s\n",
      "482:\tlearn: 16.3286994\ttest: 18.9264104\tbest: 18.9264104 (482)\ttotal: 26.2s\tremaining: 28.1s\n",
      "483:\tlearn: 16.3236961\ttest: 18.9231317\tbest: 18.9231317 (483)\ttotal: 26.3s\tremaining: 28s\n",
      "484:\tlearn: 16.3176627\ttest: 18.9212670\tbest: 18.9212670 (484)\ttotal: 26.4s\tremaining: 28s\n",
      "485:\tlearn: 16.3120633\ttest: 18.9208322\tbest: 18.9208322 (485)\ttotal: 26.4s\tremaining: 27.9s\n",
      "486:\tlearn: 16.3045457\ttest: 18.9190509\tbest: 18.9190509 (486)\ttotal: 26.4s\tremaining: 27.9s\n",
      "487:\tlearn: 16.2971719\ttest: 18.9159235\tbest: 18.9159235 (487)\ttotal: 26.5s\tremaining: 27.8s\n",
      "488:\tlearn: 16.2929852\ttest: 18.9139675\tbest: 18.9139675 (488)\ttotal: 26.6s\tremaining: 27.8s\n",
      "489:\tlearn: 16.2886480\ttest: 18.9140559\tbest: 18.9139675 (488)\ttotal: 26.6s\tremaining: 27.7s\n",
      "490:\tlearn: 16.2835179\ttest: 18.9133436\tbest: 18.9133436 (490)\ttotal: 26.7s\tremaining: 27.6s\n",
      "491:\tlearn: 16.2769440\ttest: 18.9106991\tbest: 18.9106991 (491)\ttotal: 26.7s\tremaining: 27.6s\n",
      "492:\tlearn: 16.2700022\ttest: 18.9066432\tbest: 18.9066432 (492)\ttotal: 26.8s\tremaining: 27.5s\n",
      "493:\tlearn: 16.2635390\ttest: 18.9038326\tbest: 18.9038326 (493)\ttotal: 26.8s\tremaining: 27.5s\n",
      "494:\tlearn: 16.2583294\ttest: 18.8998326\tbest: 18.8998326 (494)\ttotal: 26.9s\tremaining: 27.4s\n",
      "495:\tlearn: 16.2555676\ttest: 18.8991071\tbest: 18.8991071 (495)\ttotal: 27s\tremaining: 27.4s\n",
      "496:\tlearn: 16.2483793\ttest: 18.8964375\tbest: 18.8964375 (496)\ttotal: 27s\tremaining: 27.3s\n",
      "497:\tlearn: 16.2389151\ttest: 18.8933737\tbest: 18.8933737 (497)\ttotal: 27.1s\tremaining: 27.3s\n",
      "498:\tlearn: 16.2332354\ttest: 18.8910812\tbest: 18.8910812 (498)\ttotal: 27.1s\tremaining: 27.2s\n",
      "499:\tlearn: 16.2276433\ttest: 18.8884086\tbest: 18.8884086 (499)\ttotal: 27.2s\tremaining: 27.2s\n",
      "500:\tlearn: 16.2232802\ttest: 18.8872963\tbest: 18.8872963 (500)\ttotal: 27.2s\tremaining: 27.1s\n",
      "501:\tlearn: 16.2188186\ttest: 18.8849707\tbest: 18.8849707 (501)\ttotal: 27.3s\tremaining: 27s\n",
      "502:\tlearn: 16.2118337\ttest: 18.8828655\tbest: 18.8828655 (502)\ttotal: 27.3s\tremaining: 27s\n",
      "503:\tlearn: 16.2039795\ttest: 18.8801037\tbest: 18.8801037 (503)\ttotal: 27.4s\tremaining: 26.9s\n",
      "504:\tlearn: 16.1961532\ttest: 18.8777122\tbest: 18.8777122 (504)\ttotal: 27.4s\tremaining: 26.9s\n",
      "505:\tlearn: 16.1917935\ttest: 18.8756498\tbest: 18.8756498 (505)\ttotal: 27.5s\tremaining: 26.8s\n",
      "506:\tlearn: 16.1866193\ttest: 18.8736052\tbest: 18.8736052 (506)\ttotal: 27.5s\tremaining: 26.7s\n",
      "507:\tlearn: 16.1832075\ttest: 18.8725377\tbest: 18.8725377 (507)\ttotal: 27.6s\tremaining: 26.7s\n",
      "508:\tlearn: 16.1795076\ttest: 18.8725537\tbest: 18.8725377 (507)\ttotal: 27.6s\tremaining: 26.6s\n",
      "509:\tlearn: 16.1753380\ttest: 18.8709571\tbest: 18.8709571 (509)\ttotal: 27.7s\tremaining: 26.6s\n",
      "510:\tlearn: 16.1661517\ttest: 18.8678081\tbest: 18.8678081 (510)\ttotal: 27.7s\tremaining: 26.5s\n",
      "511:\tlearn: 16.1591347\ttest: 18.8660189\tbest: 18.8660189 (511)\ttotal: 27.8s\tremaining: 26.5s\n",
      "512:\tlearn: 16.1528524\ttest: 18.8628151\tbest: 18.8628151 (512)\ttotal: 27.8s\tremaining: 26.4s\n",
      "513:\tlearn: 16.1468269\ttest: 18.8586864\tbest: 18.8586864 (513)\ttotal: 27.9s\tremaining: 26.4s\n",
      "514:\tlearn: 16.1414449\ttest: 18.8555053\tbest: 18.8555053 (514)\ttotal: 27.9s\tremaining: 26.3s\n",
      "515:\tlearn: 16.1339146\ttest: 18.8528459\tbest: 18.8528459 (515)\ttotal: 28s\tremaining: 26.2s\n",
      "516:\tlearn: 16.1278223\ttest: 18.8502501\tbest: 18.8502501 (516)\ttotal: 28s\tremaining: 26.2s\n",
      "517:\tlearn: 16.1217387\ttest: 18.8479599\tbest: 18.8479599 (517)\ttotal: 28.1s\tremaining: 26.1s\n",
      "518:\tlearn: 16.1168248\ttest: 18.8459846\tbest: 18.8459846 (518)\ttotal: 28.1s\tremaining: 26s\n",
      "519:\tlearn: 16.1128795\ttest: 18.8450017\tbest: 18.8450017 (519)\ttotal: 28.1s\tremaining: 26s\n",
      "520:\tlearn: 16.1042161\ttest: 18.8412089\tbest: 18.8412089 (520)\ttotal: 28.2s\tremaining: 25.9s\n",
      "521:\tlearn: 16.0987324\ttest: 18.8398538\tbest: 18.8398538 (521)\ttotal: 28.2s\tremaining: 25.9s\n",
      "522:\tlearn: 16.0947192\ttest: 18.8376465\tbest: 18.8376465 (522)\ttotal: 28.3s\tremaining: 25.8s\n",
      "523:\tlearn: 16.0914271\ttest: 18.8360309\tbest: 18.8360309 (523)\ttotal: 28.3s\tremaining: 25.7s\n",
      "524:\tlearn: 16.0840298\ttest: 18.8329359\tbest: 18.8329359 (524)\ttotal: 28.4s\tremaining: 25.7s\n",
      "525:\tlearn: 16.0776493\ttest: 18.8307811\tbest: 18.8307811 (525)\ttotal: 28.4s\tremaining: 25.6s\n",
      "526:\tlearn: 16.0711514\ttest: 18.8273480\tbest: 18.8273480 (526)\ttotal: 28.5s\tremaining: 25.6s\n",
      "527:\tlearn: 16.0651173\ttest: 18.8260556\tbest: 18.8260556 (527)\ttotal: 28.5s\tremaining: 25.5s\n",
      "528:\tlearn: 16.0601911\ttest: 18.8241946\tbest: 18.8241946 (528)\ttotal: 28.6s\tremaining: 25.5s\n",
      "529:\tlearn: 16.0539826\ttest: 18.8217667\tbest: 18.8217667 (529)\ttotal: 28.6s\tremaining: 25.4s\n",
      "530:\tlearn: 16.0494496\ttest: 18.8207926\tbest: 18.8207926 (530)\ttotal: 28.7s\tremaining: 25.3s\n",
      "531:\tlearn: 16.0431949\ttest: 18.8170560\tbest: 18.8170560 (531)\ttotal: 28.8s\tremaining: 25.3s\n",
      "532:\tlearn: 16.0362848\ttest: 18.8138064\tbest: 18.8138064 (532)\ttotal: 28.8s\tremaining: 25.2s\n",
      "533:\tlearn: 16.0298472\ttest: 18.8118016\tbest: 18.8118016 (533)\ttotal: 28.9s\tremaining: 25.2s\n",
      "534:\tlearn: 16.0216044\ttest: 18.8081726\tbest: 18.8081726 (534)\ttotal: 28.9s\tremaining: 25.1s\n",
      "535:\tlearn: 16.0154774\ttest: 18.8056525\tbest: 18.8056525 (535)\ttotal: 29s\tremaining: 25.1s\n",
      "536:\tlearn: 16.0095985\ttest: 18.8041454\tbest: 18.8041454 (536)\ttotal: 29s\tremaining: 25s\n",
      "537:\tlearn: 16.0031012\ttest: 18.8029944\tbest: 18.8029944 (537)\ttotal: 29.1s\tremaining: 25s\n",
      "538:\tlearn: 15.9984010\ttest: 18.7985926\tbest: 18.7985926 (538)\ttotal: 29.1s\tremaining: 24.9s\n",
      "539:\tlearn: 15.9941458\ttest: 18.7985360\tbest: 18.7985360 (539)\ttotal: 29.2s\tremaining: 24.9s\n",
      "540:\tlearn: 15.9891424\ttest: 18.7975662\tbest: 18.7975662 (540)\ttotal: 29.3s\tremaining: 24.8s\n",
      "541:\tlearn: 15.9832200\ttest: 18.7944217\tbest: 18.7944217 (541)\ttotal: 29.3s\tremaining: 24.8s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "542:\tlearn: 15.9778238\ttest: 18.7918362\tbest: 18.7918362 (542)\ttotal: 29.4s\tremaining: 24.7s\n",
      "543:\tlearn: 15.9725403\ttest: 18.7900665\tbest: 18.7900665 (543)\ttotal: 29.4s\tremaining: 24.7s\n",
      "544:\tlearn: 15.9687802\ttest: 18.7881512\tbest: 18.7881512 (544)\ttotal: 29.5s\tremaining: 24.6s\n",
      "545:\tlearn: 15.9634576\ttest: 18.7857440\tbest: 18.7857440 (545)\ttotal: 29.5s\tremaining: 24.6s\n",
      "546:\tlearn: 15.9587804\ttest: 18.7849450\tbest: 18.7849450 (546)\ttotal: 29.6s\tremaining: 24.5s\n",
      "547:\tlearn: 15.9532088\ttest: 18.7827051\tbest: 18.7827051 (547)\ttotal: 29.6s\tremaining: 24.4s\n",
      "548:\tlearn: 15.9485774\ttest: 18.7811903\tbest: 18.7811903 (548)\ttotal: 29.7s\tremaining: 24.4s\n",
      "549:\tlearn: 15.9408198\ttest: 18.7770051\tbest: 18.7770051 (549)\ttotal: 29.8s\tremaining: 24.4s\n",
      "550:\tlearn: 15.9354518\ttest: 18.7744129\tbest: 18.7744129 (550)\ttotal: 29.8s\tremaining: 24.3s\n",
      "551:\tlearn: 15.9290391\ttest: 18.7729675\tbest: 18.7729675 (551)\ttotal: 29.9s\tremaining: 24.2s\n",
      "552:\tlearn: 15.9222851\ttest: 18.7694295\tbest: 18.7694295 (552)\ttotal: 29.9s\tremaining: 24.2s\n",
      "553:\tlearn: 15.9197879\ttest: 18.7686844\tbest: 18.7686844 (553)\ttotal: 30s\tremaining: 24.2s\n",
      "554:\tlearn: 15.9165091\ttest: 18.7671345\tbest: 18.7671345 (554)\ttotal: 30.1s\tremaining: 24.1s\n",
      "555:\tlearn: 15.9143693\ttest: 18.7663065\tbest: 18.7663065 (555)\ttotal: 30.1s\tremaining: 24s\n",
      "556:\tlearn: 15.9086225\ttest: 18.7640723\tbest: 18.7640723 (556)\ttotal: 30.2s\tremaining: 24s\n",
      "557:\tlearn: 15.9016373\ttest: 18.7629263\tbest: 18.7629263 (557)\ttotal: 30.2s\tremaining: 23.9s\n",
      "558:\tlearn: 15.8986358\ttest: 18.7617766\tbest: 18.7617766 (558)\ttotal: 30.3s\tremaining: 23.9s\n",
      "559:\tlearn: 15.8927196\ttest: 18.7577181\tbest: 18.7577181 (559)\ttotal: 30.3s\tremaining: 23.8s\n",
      "560:\tlearn: 15.8892451\ttest: 18.7560646\tbest: 18.7560646 (560)\ttotal: 30.4s\tremaining: 23.8s\n",
      "561:\tlearn: 15.8847869\ttest: 18.7536447\tbest: 18.7536447 (561)\ttotal: 30.4s\tremaining: 23.7s\n",
      "562:\tlearn: 15.8808066\ttest: 18.7511138\tbest: 18.7511138 (562)\ttotal: 30.5s\tremaining: 23.7s\n",
      "563:\tlearn: 15.8765481\ttest: 18.7487173\tbest: 18.7487173 (563)\ttotal: 30.6s\tremaining: 23.6s\n",
      "564:\tlearn: 15.8721549\ttest: 18.7480100\tbest: 18.7480100 (564)\ttotal: 30.6s\tremaining: 23.6s\n",
      "565:\tlearn: 15.8672486\ttest: 18.7472067\tbest: 18.7472067 (565)\ttotal: 30.7s\tremaining: 23.5s\n",
      "566:\tlearn: 15.8596486\ttest: 18.7432349\tbest: 18.7432349 (566)\ttotal: 30.8s\tremaining: 23.5s\n",
      "567:\tlearn: 15.8539888\ttest: 18.7405693\tbest: 18.7405693 (567)\ttotal: 30.8s\tremaining: 23.4s\n",
      "568:\tlearn: 15.8496419\ttest: 18.7400865\tbest: 18.7400865 (568)\ttotal: 30.9s\tremaining: 23.4s\n",
      "569:\tlearn: 15.8430603\ttest: 18.7367967\tbest: 18.7367967 (569)\ttotal: 30.9s\tremaining: 23.3s\n",
      "570:\tlearn: 15.8350077\ttest: 18.7337964\tbest: 18.7337964 (570)\ttotal: 31s\tremaining: 23.3s\n",
      "571:\tlearn: 15.8279666\ttest: 18.7306696\tbest: 18.7306696 (571)\ttotal: 31s\tremaining: 23.2s\n",
      "572:\tlearn: 15.8219471\ttest: 18.7287204\tbest: 18.7287204 (572)\ttotal: 31.1s\tremaining: 23.2s\n",
      "573:\tlearn: 15.8182596\ttest: 18.7280431\tbest: 18.7280431 (573)\ttotal: 31.2s\tremaining: 23.1s\n",
      "574:\tlearn: 15.8150550\ttest: 18.7268992\tbest: 18.7268992 (574)\ttotal: 31.2s\tremaining: 23.1s\n",
      "575:\tlearn: 15.8100645\ttest: 18.7245620\tbest: 18.7245620 (575)\ttotal: 31.3s\tremaining: 23s\n",
      "576:\tlearn: 15.8057100\ttest: 18.7222326\tbest: 18.7222326 (576)\ttotal: 31.3s\tremaining: 23s\n",
      "577:\tlearn: 15.7985573\ttest: 18.7195564\tbest: 18.7195564 (577)\ttotal: 31.4s\tremaining: 22.9s\n",
      "578:\tlearn: 15.7898753\ttest: 18.7177632\tbest: 18.7177632 (578)\ttotal: 31.5s\tremaining: 22.9s\n",
      "579:\tlearn: 15.7855843\ttest: 18.7162041\tbest: 18.7162041 (579)\ttotal: 31.5s\tremaining: 22.8s\n",
      "580:\tlearn: 15.7788237\ttest: 18.7127524\tbest: 18.7127524 (580)\ttotal: 31.6s\tremaining: 22.8s\n",
      "581:\tlearn: 15.7741631\ttest: 18.7121723\tbest: 18.7121723 (581)\ttotal: 31.6s\tremaining: 22.7s\n",
      "582:\tlearn: 15.7688624\ttest: 18.7095737\tbest: 18.7095737 (582)\ttotal: 31.7s\tremaining: 22.7s\n",
      "583:\tlearn: 15.7646279\ttest: 18.7073563\tbest: 18.7073563 (583)\ttotal: 31.8s\tremaining: 22.6s\n",
      "584:\tlearn: 15.7592877\ttest: 18.7051861\tbest: 18.7051861 (584)\ttotal: 31.8s\tremaining: 22.6s\n",
      "585:\tlearn: 15.7560227\ttest: 18.7039012\tbest: 18.7039012 (585)\ttotal: 31.9s\tremaining: 22.5s\n",
      "586:\tlearn: 15.7504207\ttest: 18.7032789\tbest: 18.7032789 (586)\ttotal: 31.9s\tremaining: 22.5s\n",
      "587:\tlearn: 15.7438525\ttest: 18.7023627\tbest: 18.7023627 (587)\ttotal: 32s\tremaining: 22.4s\n",
      "588:\tlearn: 15.7385631\ttest: 18.7014214\tbest: 18.7014214 (588)\ttotal: 32.1s\tremaining: 22.4s\n",
      "589:\tlearn: 15.7334821\ttest: 18.6993507\tbest: 18.6993507 (589)\ttotal: 32.1s\tremaining: 22.3s\n",
      "590:\tlearn: 15.7307511\ttest: 18.6980711\tbest: 18.6980711 (590)\ttotal: 32.2s\tremaining: 22.3s\n",
      "591:\tlearn: 15.7253860\ttest: 18.6953128\tbest: 18.6953128 (591)\ttotal: 32.2s\tremaining: 22.2s\n",
      "592:\tlearn: 15.7211108\ttest: 18.6927974\tbest: 18.6927974 (592)\ttotal: 32.3s\tremaining: 22.2s\n",
      "593:\tlearn: 15.7147593\ttest: 18.6901749\tbest: 18.6901749 (593)\ttotal: 32.3s\tremaining: 22.1s\n",
      "594:\tlearn: 15.7070888\ttest: 18.6877562\tbest: 18.6877562 (594)\ttotal: 32.4s\tremaining: 22s\n",
      "595:\tlearn: 15.7021131\ttest: 18.6875138\tbest: 18.6875138 (595)\ttotal: 32.4s\tremaining: 22s\n",
      "596:\tlearn: 15.6960808\ttest: 18.6838783\tbest: 18.6838783 (596)\ttotal: 32.5s\tremaining: 21.9s\n",
      "597:\tlearn: 15.6909074\ttest: 18.6818927\tbest: 18.6818927 (597)\ttotal: 32.6s\tremaining: 21.9s\n",
      "598:\tlearn: 15.6856209\ttest: 18.6801199\tbest: 18.6801199 (598)\ttotal: 32.6s\tremaining: 21.8s\n",
      "599:\tlearn: 15.6800762\ttest: 18.6772037\tbest: 18.6772037 (599)\ttotal: 32.7s\tremaining: 21.8s\n",
      "600:\tlearn: 15.6761754\ttest: 18.6758285\tbest: 18.6758285 (600)\ttotal: 32.7s\tremaining: 21.7s\n",
      "601:\tlearn: 15.6698899\ttest: 18.6720146\tbest: 18.6720146 (601)\ttotal: 32.8s\tremaining: 21.7s\n",
      "602:\tlearn: 15.6642423\ttest: 18.6704572\tbest: 18.6704572 (602)\ttotal: 32.8s\tremaining: 21.6s\n",
      "603:\tlearn: 15.6585053\ttest: 18.6682954\tbest: 18.6682954 (603)\ttotal: 32.9s\tremaining: 21.6s\n",
      "604:\tlearn: 15.6559549\ttest: 18.6672214\tbest: 18.6672214 (604)\ttotal: 32.9s\tremaining: 21.5s\n",
      "605:\tlearn: 15.6483286\ttest: 18.6625907\tbest: 18.6625907 (605)\ttotal: 33s\tremaining: 21.5s\n",
      "606:\tlearn: 15.6381632\ttest: 18.6601384\tbest: 18.6601384 (606)\ttotal: 33.1s\tremaining: 21.4s\n",
      "607:\tlearn: 15.6289423\ttest: 18.6574769\tbest: 18.6574769 (607)\ttotal: 33.1s\tremaining: 21.3s\n",
      "608:\tlearn: 15.6232548\ttest: 18.6551049\tbest: 18.6551049 (608)\ttotal: 33.2s\tremaining: 21.3s\n",
      "609:\tlearn: 15.6186513\ttest: 18.6533076\tbest: 18.6533076 (609)\ttotal: 33.2s\tremaining: 21.2s\n",
      "610:\tlearn: 15.6130456\ttest: 18.6515061\tbest: 18.6515061 (610)\ttotal: 33.3s\tremaining: 21.2s\n",
      "611:\tlearn: 15.6073098\ttest: 18.6484322\tbest: 18.6484322 (611)\ttotal: 33.3s\tremaining: 21.1s\n",
      "612:\tlearn: 15.6039829\ttest: 18.6479504\tbest: 18.6479504 (612)\ttotal: 33.4s\tremaining: 21.1s\n",
      "613:\tlearn: 15.6002282\ttest: 18.6477182\tbest: 18.6477182 (613)\ttotal: 33.4s\tremaining: 21s\n",
      "614:\tlearn: 15.5964345\ttest: 18.6457482\tbest: 18.6457482 (614)\ttotal: 33.5s\tremaining: 21s\n",
      "615:\tlearn: 15.5915462\ttest: 18.6434469\tbest: 18.6434469 (615)\ttotal: 33.5s\tremaining: 20.9s\n",
      "616:\tlearn: 15.5840792\ttest: 18.6398384\tbest: 18.6398384 (616)\ttotal: 33.6s\tremaining: 20.8s\n",
      "617:\tlearn: 15.5810642\ttest: 18.6387255\tbest: 18.6387255 (617)\ttotal: 33.6s\tremaining: 20.8s\n",
      "618:\tlearn: 15.5753695\ttest: 18.6378006\tbest: 18.6378006 (618)\ttotal: 33.7s\tremaining: 20.7s\n",
      "619:\tlearn: 15.5709720\ttest: 18.6359343\tbest: 18.6359343 (619)\ttotal: 33.8s\tremaining: 20.7s\n",
      "620:\tlearn: 15.5664765\ttest: 18.6346684\tbest: 18.6346684 (620)\ttotal: 33.8s\tremaining: 20.6s\n",
      "621:\tlearn: 15.5619965\ttest: 18.6328237\tbest: 18.6328237 (621)\ttotal: 33.9s\tremaining: 20.6s\n",
      "622:\tlearn: 15.5565692\ttest: 18.6324587\tbest: 18.6324587 (622)\ttotal: 33.9s\tremaining: 20.5s\n",
      "623:\tlearn: 15.5512332\ttest: 18.6300868\tbest: 18.6300868 (623)\ttotal: 34s\tremaining: 20.5s\n",
      "624:\tlearn: 15.5434369\ttest: 18.6280630\tbest: 18.6280630 (624)\ttotal: 34.1s\tremaining: 20.4s\n",
      "625:\tlearn: 15.5368637\ttest: 18.6266121\tbest: 18.6266121 (625)\ttotal: 34.1s\tremaining: 20.4s\n",
      "626:\tlearn: 15.5319610\ttest: 18.6253913\tbest: 18.6253913 (626)\ttotal: 34.2s\tremaining: 20.3s\n",
      "627:\tlearn: 15.5277376\ttest: 18.6238515\tbest: 18.6238515 (627)\ttotal: 34.3s\tremaining: 20.3s\n",
      "628:\tlearn: 15.5219785\ttest: 18.6217597\tbest: 18.6217597 (628)\ttotal: 34.3s\tremaining: 20.3s\n",
      "629:\tlearn: 15.5169166\ttest: 18.6204059\tbest: 18.6204059 (629)\ttotal: 34.4s\tremaining: 20.2s\n",
      "630:\tlearn: 15.5106143\ttest: 18.6179679\tbest: 18.6179679 (630)\ttotal: 34.4s\tremaining: 20.1s\n",
      "631:\tlearn: 15.5062208\ttest: 18.6167576\tbest: 18.6167576 (631)\ttotal: 34.5s\tremaining: 20.1s\n",
      "632:\tlearn: 15.4999295\ttest: 18.6140765\tbest: 18.6140765 (632)\ttotal: 34.6s\tremaining: 20s\n",
      "633:\tlearn: 15.4946714\ttest: 18.6120680\tbest: 18.6120680 (633)\ttotal: 34.6s\tremaining: 20s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "634:\tlearn: 15.4908658\ttest: 18.6117930\tbest: 18.6117930 (634)\ttotal: 34.7s\tremaining: 19.9s\n",
      "635:\tlearn: 15.4859329\ttest: 18.6090849\tbest: 18.6090849 (635)\ttotal: 34.7s\tremaining: 19.9s\n",
      "636:\tlearn: 15.4765818\ttest: 18.6072163\tbest: 18.6072163 (636)\ttotal: 34.8s\tremaining: 19.8s\n",
      "637:\tlearn: 15.4718821\ttest: 18.6060182\tbest: 18.6060182 (637)\ttotal: 34.8s\tremaining: 19.8s\n",
      "638:\tlearn: 15.4652837\ttest: 18.6027335\tbest: 18.6027335 (638)\ttotal: 34.9s\tremaining: 19.7s\n",
      "639:\tlearn: 15.4619467\ttest: 18.6003539\tbest: 18.6003539 (639)\ttotal: 34.9s\tremaining: 19.7s\n",
      "640:\tlearn: 15.4581653\ttest: 18.5982817\tbest: 18.5982817 (640)\ttotal: 35s\tremaining: 19.6s\n",
      "641:\tlearn: 15.4540421\ttest: 18.5963338\tbest: 18.5963338 (641)\ttotal: 35.1s\tremaining: 19.6s\n",
      "642:\tlearn: 15.4489544\ttest: 18.5940694\tbest: 18.5940694 (642)\ttotal: 35.1s\tremaining: 19.5s\n",
      "643:\tlearn: 15.4432345\ttest: 18.5924183\tbest: 18.5924183 (643)\ttotal: 35.2s\tremaining: 19.4s\n",
      "644:\tlearn: 15.4364928\ttest: 18.5903989\tbest: 18.5903989 (644)\ttotal: 35.2s\tremaining: 19.4s\n",
      "645:\tlearn: 15.4314343\ttest: 18.5883447\tbest: 18.5883447 (645)\ttotal: 35.3s\tremaining: 19.3s\n",
      "646:\tlearn: 15.4272680\ttest: 18.5863877\tbest: 18.5863877 (646)\ttotal: 35.3s\tremaining: 19.3s\n",
      "647:\tlearn: 15.4218038\ttest: 18.5855287\tbest: 18.5855287 (647)\ttotal: 35.4s\tremaining: 19.2s\n",
      "648:\tlearn: 15.4172221\ttest: 18.5822881\tbest: 18.5822881 (648)\ttotal: 35.4s\tremaining: 19.2s\n",
      "649:\tlearn: 15.4142860\ttest: 18.5807364\tbest: 18.5807364 (649)\ttotal: 35.5s\tremaining: 19.1s\n",
      "650:\tlearn: 15.4075035\ttest: 18.5773174\tbest: 18.5773174 (650)\ttotal: 35.5s\tremaining: 19s\n",
      "651:\tlearn: 15.4043346\ttest: 18.5753480\tbest: 18.5753480 (651)\ttotal: 35.6s\tremaining: 19s\n",
      "652:\tlearn: 15.4005668\ttest: 18.5752204\tbest: 18.5752204 (652)\ttotal: 35.6s\tremaining: 18.9s\n",
      "653:\tlearn: 15.3945576\ttest: 18.5723703\tbest: 18.5723703 (653)\ttotal: 35.7s\tremaining: 18.9s\n",
      "654:\tlearn: 15.3876878\ttest: 18.5691523\tbest: 18.5691523 (654)\ttotal: 35.7s\tremaining: 18.8s\n",
      "655:\tlearn: 15.3834756\ttest: 18.5686913\tbest: 18.5686913 (655)\ttotal: 35.8s\tremaining: 18.8s\n",
      "656:\tlearn: 15.3769924\ttest: 18.5671709\tbest: 18.5671709 (656)\ttotal: 35.8s\tremaining: 18.7s\n",
      "657:\tlearn: 15.3717386\ttest: 18.5648268\tbest: 18.5648268 (657)\ttotal: 35.9s\tremaining: 18.7s\n",
      "658:\tlearn: 15.3676022\ttest: 18.5640508\tbest: 18.5640508 (658)\ttotal: 35.9s\tremaining: 18.6s\n",
      "659:\tlearn: 15.3630469\ttest: 18.5623265\tbest: 18.5623265 (659)\ttotal: 36s\tremaining: 18.5s\n",
      "660:\tlearn: 15.3556063\ttest: 18.5599417\tbest: 18.5599417 (660)\ttotal: 36s\tremaining: 18.5s\n",
      "661:\tlearn: 15.3519012\ttest: 18.5589410\tbest: 18.5589410 (661)\ttotal: 36.1s\tremaining: 18.4s\n",
      "662:\tlearn: 15.3469932\ttest: 18.5566612\tbest: 18.5566612 (662)\ttotal: 36.1s\tremaining: 18.4s\n",
      "663:\tlearn: 15.3407651\ttest: 18.5538081\tbest: 18.5538081 (663)\ttotal: 36.2s\tremaining: 18.3s\n",
      "664:\tlearn: 15.3362791\ttest: 18.5525660\tbest: 18.5525660 (664)\ttotal: 36.2s\tremaining: 18.3s\n",
      "665:\tlearn: 15.3310509\ttest: 18.5486448\tbest: 18.5486448 (665)\ttotal: 36.3s\tremaining: 18.2s\n",
      "666:\tlearn: 15.3247027\ttest: 18.5467216\tbest: 18.5467216 (666)\ttotal: 36.3s\tremaining: 18.1s\n",
      "667:\tlearn: 15.3195593\ttest: 18.5451479\tbest: 18.5451479 (667)\ttotal: 36.4s\tremaining: 18.1s\n",
      "668:\tlearn: 15.3149138\ttest: 18.5446821\tbest: 18.5446821 (668)\ttotal: 36.4s\tremaining: 18s\n",
      "669:\tlearn: 15.3088288\ttest: 18.5426434\tbest: 18.5426434 (669)\ttotal: 36.5s\tremaining: 18s\n",
      "670:\tlearn: 15.3013515\ttest: 18.5405527\tbest: 18.5405527 (670)\ttotal: 36.5s\tremaining: 17.9s\n",
      "671:\tlearn: 15.2955388\ttest: 18.5385334\tbest: 18.5385334 (671)\ttotal: 36.6s\tremaining: 17.9s\n",
      "672:\tlearn: 15.2937813\ttest: 18.5385213\tbest: 18.5385213 (672)\ttotal: 36.7s\tremaining: 17.8s\n",
      "673:\tlearn: 15.2886309\ttest: 18.5362981\tbest: 18.5362981 (673)\ttotal: 36.7s\tremaining: 17.8s\n",
      "674:\tlearn: 15.2836258\ttest: 18.5346831\tbest: 18.5346831 (674)\ttotal: 36.8s\tremaining: 17.7s\n",
      "675:\tlearn: 15.2759236\ttest: 18.5319209\tbest: 18.5319209 (675)\ttotal: 36.8s\tremaining: 17.7s\n",
      "676:\tlearn: 15.2688975\ttest: 18.5289486\tbest: 18.5289486 (676)\ttotal: 36.9s\tremaining: 17.6s\n",
      "677:\tlearn: 15.2635716\ttest: 18.5269820\tbest: 18.5269820 (677)\ttotal: 36.9s\tremaining: 17.5s\n",
      "678:\tlearn: 15.2596636\ttest: 18.5261796\tbest: 18.5261796 (678)\ttotal: 37s\tremaining: 17.5s\n",
      "679:\tlearn: 15.2545142\ttest: 18.5246202\tbest: 18.5246202 (679)\ttotal: 37s\tremaining: 17.4s\n",
      "680:\tlearn: 15.2506237\ttest: 18.5225472\tbest: 18.5225472 (680)\ttotal: 37.1s\tremaining: 17.4s\n",
      "681:\tlearn: 15.2474492\ttest: 18.5214733\tbest: 18.5214733 (681)\ttotal: 37.1s\tremaining: 17.3s\n",
      "682:\tlearn: 15.2444562\ttest: 18.5194998\tbest: 18.5194998 (682)\ttotal: 37.2s\tremaining: 17.3s\n",
      "683:\tlearn: 15.2422674\ttest: 18.5191159\tbest: 18.5191159 (683)\ttotal: 37.2s\tremaining: 17.2s\n",
      "684:\tlearn: 15.2382237\ttest: 18.5169118\tbest: 18.5169118 (684)\ttotal: 37.3s\tremaining: 17.2s\n",
      "685:\tlearn: 15.2338767\ttest: 18.5153393\tbest: 18.5153393 (685)\ttotal: 37.3s\tremaining: 17.1s\n",
      "686:\tlearn: 15.2266388\ttest: 18.5116954\tbest: 18.5116954 (686)\ttotal: 37.4s\tremaining: 17s\n",
      "687:\tlearn: 15.2217071\ttest: 18.5101826\tbest: 18.5101826 (687)\ttotal: 37.4s\tremaining: 17s\n",
      "688:\tlearn: 15.2169985\ttest: 18.5096543\tbest: 18.5096543 (688)\ttotal: 37.5s\tremaining: 16.9s\n",
      "689:\tlearn: 15.2118857\ttest: 18.5077394\tbest: 18.5077394 (689)\ttotal: 37.6s\tremaining: 16.9s\n",
      "690:\tlearn: 15.2069513\ttest: 18.5052370\tbest: 18.5052370 (690)\ttotal: 37.6s\tremaining: 16.8s\n",
      "691:\tlearn: 15.2010537\ttest: 18.5027892\tbest: 18.5027892 (691)\ttotal: 37.7s\tremaining: 16.8s\n",
      "692:\tlearn: 15.1973518\ttest: 18.5015845\tbest: 18.5015845 (692)\ttotal: 37.7s\tremaining: 16.7s\n",
      "693:\tlearn: 15.1922914\ttest: 18.5007664\tbest: 18.5007664 (693)\ttotal: 37.8s\tremaining: 16.7s\n",
      "694:\tlearn: 15.1853617\ttest: 18.4969164\tbest: 18.4969164 (694)\ttotal: 37.8s\tremaining: 16.6s\n",
      "695:\tlearn: 15.1811410\ttest: 18.4946264\tbest: 18.4946264 (695)\ttotal: 37.9s\tremaining: 16.5s\n",
      "696:\tlearn: 15.1767867\ttest: 18.4923131\tbest: 18.4923131 (696)\ttotal: 37.9s\tremaining: 16.5s\n",
      "697:\tlearn: 15.1729902\ttest: 18.4906734\tbest: 18.4906734 (697)\ttotal: 38s\tremaining: 16.4s\n",
      "698:\tlearn: 15.1665283\ttest: 18.4886961\tbest: 18.4886961 (698)\ttotal: 38.1s\tremaining: 16.4s\n",
      "699:\tlearn: 15.1617871\ttest: 18.4864429\tbest: 18.4864429 (699)\ttotal: 38.1s\tremaining: 16.3s\n",
      "700:\tlearn: 15.1562157\ttest: 18.4833479\tbest: 18.4833479 (700)\ttotal: 38.2s\tremaining: 16.3s\n",
      "701:\tlearn: 15.1527807\ttest: 18.4826708\tbest: 18.4826708 (701)\ttotal: 38.2s\tremaining: 16.2s\n",
      "702:\tlearn: 15.1457553\ttest: 18.4807521\tbest: 18.4807521 (702)\ttotal: 38.2s\tremaining: 16.2s\n",
      "703:\tlearn: 15.1416928\ttest: 18.4807352\tbest: 18.4807352 (703)\ttotal: 38.3s\tremaining: 16.1s\n",
      "704:\tlearn: 15.1352018\ttest: 18.4785557\tbest: 18.4785557 (704)\ttotal: 38.3s\tremaining: 16s\n",
      "705:\tlearn: 15.1308873\ttest: 18.4783463\tbest: 18.4783463 (705)\ttotal: 38.4s\tremaining: 16s\n",
      "706:\tlearn: 15.1244671\ttest: 18.4769818\tbest: 18.4769818 (706)\ttotal: 38.5s\tremaining: 15.9s\n",
      "707:\tlearn: 15.1183223\ttest: 18.4748147\tbest: 18.4748147 (707)\ttotal: 38.5s\tremaining: 15.9s\n",
      "708:\tlearn: 15.1129265\ttest: 18.4723186\tbest: 18.4723186 (708)\ttotal: 38.6s\tremaining: 15.8s\n",
      "709:\tlearn: 15.1071179\ttest: 18.4698788\tbest: 18.4698788 (709)\ttotal: 38.6s\tremaining: 15.8s\n",
      "710:\tlearn: 15.1000998\ttest: 18.4675745\tbest: 18.4675745 (710)\ttotal: 38.7s\tremaining: 15.7s\n",
      "711:\tlearn: 15.0960474\ttest: 18.4661270\tbest: 18.4661270 (711)\ttotal: 38.7s\tremaining: 15.7s\n",
      "712:\tlearn: 15.0906938\ttest: 18.4646556\tbest: 18.4646556 (712)\ttotal: 38.8s\tremaining: 15.6s\n",
      "713:\tlearn: 15.0863645\ttest: 18.4630068\tbest: 18.4630068 (713)\ttotal: 38.8s\tremaining: 15.6s\n",
      "714:\tlearn: 15.0797192\ttest: 18.4598531\tbest: 18.4598531 (714)\ttotal: 38.9s\tremaining: 15.5s\n",
      "715:\tlearn: 15.0741708\ttest: 18.4579340\tbest: 18.4579340 (715)\ttotal: 38.9s\tremaining: 15.4s\n",
      "716:\tlearn: 15.0692131\ttest: 18.4580309\tbest: 18.4579340 (715)\ttotal: 39s\tremaining: 15.4s\n",
      "717:\tlearn: 15.0668591\ttest: 18.4573694\tbest: 18.4573694 (717)\ttotal: 39s\tremaining: 15.3s\n",
      "718:\tlearn: 15.0628377\ttest: 18.4561639\tbest: 18.4561639 (718)\ttotal: 39.1s\tremaining: 15.3s\n",
      "719:\tlearn: 15.0586680\ttest: 18.4549707\tbest: 18.4549707 (719)\ttotal: 39.1s\tremaining: 15.2s\n",
      "720:\tlearn: 15.0533746\ttest: 18.4522448\tbest: 18.4522448 (720)\ttotal: 39.2s\tremaining: 15.2s\n",
      "721:\tlearn: 15.0510300\ttest: 18.4519294\tbest: 18.4519294 (721)\ttotal: 39.2s\tremaining: 15.1s\n",
      "722:\tlearn: 15.0447438\ttest: 18.4499188\tbest: 18.4499188 (722)\ttotal: 39.3s\tremaining: 15.1s\n",
      "723:\tlearn: 15.0409923\ttest: 18.4486036\tbest: 18.4486036 (723)\ttotal: 39.4s\tremaining: 15s\n",
      "724:\tlearn: 15.0363012\ttest: 18.4475650\tbest: 18.4475650 (724)\ttotal: 39.4s\tremaining: 14.9s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "725:\tlearn: 15.0332066\ttest: 18.4465627\tbest: 18.4465627 (725)\ttotal: 39.5s\tremaining: 14.9s\n",
      "726:\tlearn: 15.0257149\ttest: 18.4434548\tbest: 18.4434548 (726)\ttotal: 39.5s\tremaining: 14.8s\n",
      "727:\tlearn: 15.0208822\ttest: 18.4415940\tbest: 18.4415940 (727)\ttotal: 39.6s\tremaining: 14.8s\n",
      "728:\tlearn: 15.0161620\ttest: 18.4389450\tbest: 18.4389450 (728)\ttotal: 39.6s\tremaining: 14.7s\n",
      "729:\tlearn: 15.0119501\ttest: 18.4376831\tbest: 18.4376831 (729)\ttotal: 39.7s\tremaining: 14.7s\n",
      "730:\tlearn: 15.0048478\ttest: 18.4353039\tbest: 18.4353039 (730)\ttotal: 39.7s\tremaining: 14.6s\n",
      "731:\tlearn: 15.0020240\ttest: 18.4338810\tbest: 18.4338810 (731)\ttotal: 39.8s\tremaining: 14.6s\n",
      "732:\tlearn: 14.9992455\ttest: 18.4334341\tbest: 18.4334341 (732)\ttotal: 39.8s\tremaining: 14.5s\n",
      "733:\tlearn: 14.9948034\ttest: 18.4310375\tbest: 18.4310375 (733)\ttotal: 39.9s\tremaining: 14.5s\n",
      "734:\tlearn: 14.9878489\ttest: 18.4284164\tbest: 18.4284164 (734)\ttotal: 39.9s\tremaining: 14.4s\n",
      "735:\tlearn: 14.9819997\ttest: 18.4259899\tbest: 18.4259899 (735)\ttotal: 40s\tremaining: 14.3s\n",
      "736:\tlearn: 14.9769297\ttest: 18.4222992\tbest: 18.4222992 (736)\ttotal: 40.1s\tremaining: 14.3s\n",
      "737:\tlearn: 14.9721754\ttest: 18.4215049\tbest: 18.4215049 (737)\ttotal: 40.1s\tremaining: 14.2s\n",
      "738:\tlearn: 14.9676389\ttest: 18.4199751\tbest: 18.4199751 (738)\ttotal: 40.2s\tremaining: 14.2s\n",
      "739:\tlearn: 14.9624079\ttest: 18.4178554\tbest: 18.4178554 (739)\ttotal: 40.2s\tremaining: 14.1s\n",
      "740:\tlearn: 14.9579374\ttest: 18.4172623\tbest: 18.4172623 (740)\ttotal: 40.3s\tremaining: 14.1s\n",
      "741:\tlearn: 14.9548440\ttest: 18.4164329\tbest: 18.4164329 (741)\ttotal: 40.3s\tremaining: 14s\n",
      "742:\tlearn: 14.9488620\ttest: 18.4149648\tbest: 18.4149648 (742)\ttotal: 40.4s\tremaining: 14s\n",
      "743:\tlearn: 14.9435067\ttest: 18.4134909\tbest: 18.4134909 (743)\ttotal: 40.4s\tremaining: 13.9s\n",
      "744:\tlearn: 14.9399912\ttest: 18.4124296\tbest: 18.4124296 (744)\ttotal: 40.5s\tremaining: 13.9s\n",
      "745:\tlearn: 14.9332992\ttest: 18.4107623\tbest: 18.4107623 (745)\ttotal: 40.5s\tremaining: 13.8s\n",
      "746:\tlearn: 14.9312300\ttest: 18.4098933\tbest: 18.4098933 (746)\ttotal: 40.6s\tremaining: 13.7s\n",
      "747:\tlearn: 14.9269193\ttest: 18.4075711\tbest: 18.4075711 (747)\ttotal: 40.6s\tremaining: 13.7s\n",
      "748:\tlearn: 14.9199391\ttest: 18.4067662\tbest: 18.4067662 (748)\ttotal: 40.7s\tremaining: 13.6s\n",
      "749:\tlearn: 14.9149038\ttest: 18.4060862\tbest: 18.4060862 (749)\ttotal: 40.7s\tremaining: 13.6s\n",
      "750:\tlearn: 14.9125274\ttest: 18.4057345\tbest: 18.4057345 (750)\ttotal: 40.8s\tremaining: 13.5s\n",
      "751:\tlearn: 14.9086628\ttest: 18.4043150\tbest: 18.4043150 (751)\ttotal: 40.9s\tremaining: 13.5s\n",
      "752:\tlearn: 14.9035034\ttest: 18.4025476\tbest: 18.4025476 (752)\ttotal: 40.9s\tremaining: 13.4s\n",
      "753:\tlearn: 14.8980735\ttest: 18.4017420\tbest: 18.4017420 (753)\ttotal: 41s\tremaining: 13.4s\n",
      "754:\tlearn: 14.8928616\ttest: 18.3978883\tbest: 18.3978883 (754)\ttotal: 41s\tremaining: 13.3s\n",
      "755:\tlearn: 14.8909222\ttest: 18.3976783\tbest: 18.3976783 (755)\ttotal: 41s\tremaining: 13.2s\n",
      "756:\tlearn: 14.8862492\ttest: 18.3963315\tbest: 18.3963315 (756)\ttotal: 41.1s\tremaining: 13.2s\n",
      "757:\tlearn: 14.8802563\ttest: 18.3937372\tbest: 18.3937372 (757)\ttotal: 41.1s\tremaining: 13.1s\n",
      "758:\tlearn: 14.8754564\ttest: 18.3927337\tbest: 18.3927337 (758)\ttotal: 41.2s\tremaining: 13.1s\n",
      "759:\tlearn: 14.8685384\ttest: 18.3904994\tbest: 18.3904994 (759)\ttotal: 41.2s\tremaining: 13s\n",
      "760:\tlearn: 14.8643792\ttest: 18.3879549\tbest: 18.3879549 (760)\ttotal: 41.3s\tremaining: 13s\n",
      "761:\tlearn: 14.8599365\ttest: 18.3852491\tbest: 18.3852491 (761)\ttotal: 41.3s\tremaining: 12.9s\n",
      "762:\tlearn: 14.8548069\ttest: 18.3832021\tbest: 18.3832021 (762)\ttotal: 41.4s\tremaining: 12.8s\n",
      "763:\tlearn: 14.8485695\ttest: 18.3814022\tbest: 18.3814022 (763)\ttotal: 41.4s\tremaining: 12.8s\n",
      "764:\tlearn: 14.8453421\ttest: 18.3791732\tbest: 18.3791732 (764)\ttotal: 41.5s\tremaining: 12.7s\n",
      "765:\tlearn: 14.8420553\ttest: 18.3775248\tbest: 18.3775248 (765)\ttotal: 41.5s\tremaining: 12.7s\n",
      "766:\tlearn: 14.8382576\ttest: 18.3757915\tbest: 18.3757915 (766)\ttotal: 41.6s\tremaining: 12.6s\n",
      "767:\tlearn: 14.8342107\ttest: 18.3739496\tbest: 18.3739496 (767)\ttotal: 41.6s\tremaining: 12.6s\n",
      "768:\tlearn: 14.8287489\ttest: 18.3719629\tbest: 18.3719629 (768)\ttotal: 41.7s\tremaining: 12.5s\n",
      "769:\tlearn: 14.8251850\ttest: 18.3701893\tbest: 18.3701893 (769)\ttotal: 41.7s\tremaining: 12.5s\n",
      "770:\tlearn: 14.8191780\ttest: 18.3686227\tbest: 18.3686227 (770)\ttotal: 41.8s\tremaining: 12.4s\n",
      "771:\tlearn: 14.8154758\ttest: 18.3682168\tbest: 18.3682168 (771)\ttotal: 41.8s\tremaining: 12.4s\n",
      "772:\tlearn: 14.8107033\ttest: 18.3671103\tbest: 18.3671103 (772)\ttotal: 41.9s\tremaining: 12.3s\n",
      "773:\tlearn: 14.8043259\ttest: 18.3638433\tbest: 18.3638433 (773)\ttotal: 41.9s\tremaining: 12.2s\n",
      "774:\tlearn: 14.7980187\ttest: 18.3629189\tbest: 18.3629189 (774)\ttotal: 42s\tremaining: 12.2s\n",
      "775:\tlearn: 14.7913822\ttest: 18.3598615\tbest: 18.3598615 (775)\ttotal: 42s\tremaining: 12.1s\n",
      "776:\tlearn: 14.7888817\ttest: 18.3590705\tbest: 18.3590705 (776)\ttotal: 42.1s\tremaining: 12.1s\n",
      "777:\tlearn: 14.7860683\ttest: 18.3595062\tbest: 18.3590705 (776)\ttotal: 42.1s\tremaining: 12s\n",
      "778:\tlearn: 14.7807094\ttest: 18.3580295\tbest: 18.3580295 (778)\ttotal: 42.2s\tremaining: 12s\n",
      "779:\tlearn: 14.7767929\ttest: 18.3564984\tbest: 18.3564984 (779)\ttotal: 42.2s\tremaining: 11.9s\n",
      "780:\tlearn: 14.7719793\ttest: 18.3537612\tbest: 18.3537612 (780)\ttotal: 42.3s\tremaining: 11.9s\n",
      "781:\tlearn: 14.7667121\ttest: 18.3520431\tbest: 18.3520431 (781)\ttotal: 42.4s\tremaining: 11.8s\n",
      "782:\tlearn: 14.7633607\ttest: 18.3505483\tbest: 18.3505483 (782)\ttotal: 42.4s\tremaining: 11.8s\n",
      "783:\tlearn: 14.7591673\ttest: 18.3509752\tbest: 18.3505483 (782)\ttotal: 42.5s\tremaining: 11.7s\n",
      "784:\tlearn: 14.7534225\ttest: 18.3487097\tbest: 18.3487097 (784)\ttotal: 42.5s\tremaining: 11.6s\n",
      "785:\tlearn: 14.7491118\ttest: 18.3480232\tbest: 18.3480232 (785)\ttotal: 42.6s\tremaining: 11.6s\n",
      "786:\tlearn: 14.7474239\ttest: 18.3481811\tbest: 18.3480232 (785)\ttotal: 42.6s\tremaining: 11.5s\n",
      "787:\tlearn: 14.7436320\ttest: 18.3472859\tbest: 18.3472859 (787)\ttotal: 42.7s\tremaining: 11.5s\n",
      "788:\tlearn: 14.7419854\ttest: 18.3473648\tbest: 18.3472859 (787)\ttotal: 42.7s\tremaining: 11.4s\n",
      "789:\tlearn: 14.7361637\ttest: 18.3464036\tbest: 18.3464036 (789)\ttotal: 42.8s\tremaining: 11.4s\n",
      "790:\tlearn: 14.7296024\ttest: 18.3445185\tbest: 18.3445185 (790)\ttotal: 42.8s\tremaining: 11.3s\n",
      "791:\tlearn: 14.7272535\ttest: 18.3444603\tbest: 18.3444603 (791)\ttotal: 42.9s\tremaining: 11.3s\n",
      "792:\tlearn: 14.7243375\ttest: 18.3439526\tbest: 18.3439526 (792)\ttotal: 42.9s\tremaining: 11.2s\n",
      "793:\tlearn: 14.7197277\ttest: 18.3423761\tbest: 18.3423761 (793)\ttotal: 43s\tremaining: 11.2s\n",
      "794:\tlearn: 14.7151669\ttest: 18.3403031\tbest: 18.3403031 (794)\ttotal: 43s\tremaining: 11.1s\n",
      "795:\tlearn: 14.7110691\ttest: 18.3394291\tbest: 18.3394291 (795)\ttotal: 43.1s\tremaining: 11s\n",
      "796:\tlearn: 14.7051930\ttest: 18.3382502\tbest: 18.3382502 (796)\ttotal: 43.1s\tremaining: 11s\n",
      "797:\tlearn: 14.7030409\ttest: 18.3379011\tbest: 18.3379011 (797)\ttotal: 43.2s\tremaining: 10.9s\n",
      "798:\tlearn: 14.6980542\ttest: 18.3355367\tbest: 18.3355367 (798)\ttotal: 43.2s\tremaining: 10.9s\n",
      "799:\tlearn: 14.6956004\ttest: 18.3350789\tbest: 18.3350789 (799)\ttotal: 43.3s\tremaining: 10.8s\n",
      "800:\tlearn: 14.6907744\ttest: 18.3338478\tbest: 18.3338478 (800)\ttotal: 43.4s\tremaining: 10.8s\n",
      "801:\tlearn: 14.6829462\ttest: 18.3309620\tbest: 18.3309620 (801)\ttotal: 43.4s\tremaining: 10.7s\n",
      "802:\tlearn: 14.6773749\ttest: 18.3293643\tbest: 18.3293643 (802)\ttotal: 43.5s\tremaining: 10.7s\n",
      "803:\tlearn: 14.6720113\ttest: 18.3272880\tbest: 18.3272880 (803)\ttotal: 43.5s\tremaining: 10.6s\n",
      "804:\tlearn: 14.6658307\ttest: 18.3257044\tbest: 18.3257044 (804)\ttotal: 43.6s\tremaining: 10.6s\n",
      "805:\tlearn: 14.6621186\ttest: 18.3245142\tbest: 18.3245142 (805)\ttotal: 43.7s\tremaining: 10.5s\n",
      "806:\tlearn: 14.6602440\ttest: 18.3237821\tbest: 18.3237821 (806)\ttotal: 43.7s\tremaining: 10.5s\n",
      "807:\tlearn: 14.6549711\ttest: 18.3222495\tbest: 18.3222495 (807)\ttotal: 43.8s\tremaining: 10.4s\n",
      "808:\tlearn: 14.6500759\ttest: 18.3209475\tbest: 18.3209475 (808)\ttotal: 43.8s\tremaining: 10.3s\n",
      "809:\tlearn: 14.6483383\ttest: 18.3205072\tbest: 18.3205072 (809)\ttotal: 43.9s\tremaining: 10.3s\n",
      "810:\tlearn: 14.6439478\ttest: 18.3199023\tbest: 18.3199023 (810)\ttotal: 43.9s\tremaining: 10.2s\n",
      "811:\tlearn: 14.6403800\ttest: 18.3184634\tbest: 18.3184634 (811)\ttotal: 44s\tremaining: 10.2s\n",
      "812:\tlearn: 14.6357683\ttest: 18.3165549\tbest: 18.3165549 (812)\ttotal: 44.1s\tremaining: 10.1s\n",
      "813:\tlearn: 14.6299965\ttest: 18.3164870\tbest: 18.3164870 (813)\ttotal: 44.1s\tremaining: 10.1s\n",
      "814:\tlearn: 14.6268814\ttest: 18.3153191\tbest: 18.3153191 (814)\ttotal: 44.2s\tremaining: 10s\n",
      "815:\tlearn: 14.6220875\ttest: 18.3123417\tbest: 18.3123417 (815)\ttotal: 44.2s\tremaining: 9.98s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "816:\tlearn: 14.6164181\ttest: 18.3097150\tbest: 18.3097150 (816)\ttotal: 44.3s\tremaining: 9.92s\n",
      "817:\tlearn: 14.6119080\ttest: 18.3071577\tbest: 18.3071577 (817)\ttotal: 44.4s\tremaining: 9.87s\n",
      "818:\tlearn: 14.6060589\ttest: 18.3054611\tbest: 18.3054611 (818)\ttotal: 44.4s\tremaining: 9.82s\n",
      "819:\tlearn: 14.6018087\ttest: 18.3035143\tbest: 18.3035143 (819)\ttotal: 44.5s\tremaining: 9.77s\n",
      "820:\tlearn: 14.5991342\ttest: 18.3020617\tbest: 18.3020617 (820)\ttotal: 44.6s\tremaining: 9.71s\n",
      "821:\tlearn: 14.5970497\ttest: 18.3016062\tbest: 18.3016062 (821)\ttotal: 44.6s\tremaining: 9.66s\n",
      "822:\tlearn: 14.5930109\ttest: 18.3008771\tbest: 18.3008771 (822)\ttotal: 44.7s\tremaining: 9.61s\n",
      "823:\tlearn: 14.5892025\ttest: 18.3002382\tbest: 18.3002382 (823)\ttotal: 44.7s\tremaining: 9.55s\n",
      "824:\tlearn: 14.5815775\ttest: 18.2981540\tbest: 18.2981540 (824)\ttotal: 44.8s\tremaining: 9.5s\n",
      "825:\tlearn: 14.5755744\ttest: 18.2953098\tbest: 18.2953098 (825)\ttotal: 44.9s\tremaining: 9.45s\n",
      "826:\tlearn: 14.5693845\ttest: 18.2928318\tbest: 18.2928318 (826)\ttotal: 44.9s\tremaining: 9.4s\n",
      "827:\tlearn: 14.5656105\ttest: 18.2921429\tbest: 18.2921429 (827)\ttotal: 45s\tremaining: 9.34s\n",
      "828:\tlearn: 14.5595627\ttest: 18.2895245\tbest: 18.2895245 (828)\ttotal: 45s\tremaining: 9.29s\n",
      "829:\tlearn: 14.5540980\ttest: 18.2880369\tbest: 18.2880369 (829)\ttotal: 45.1s\tremaining: 9.23s\n",
      "830:\tlearn: 14.5497238\ttest: 18.2863024\tbest: 18.2863024 (830)\ttotal: 45.1s\tremaining: 9.18s\n",
      "831:\tlearn: 14.5442365\ttest: 18.2845987\tbest: 18.2845987 (831)\ttotal: 45.2s\tremaining: 9.12s\n",
      "832:\tlearn: 14.5386646\ttest: 18.2830932\tbest: 18.2830932 (832)\ttotal: 45.2s\tremaining: 9.07s\n",
      "833:\tlearn: 14.5367031\ttest: 18.2819527\tbest: 18.2819527 (833)\ttotal: 45.3s\tremaining: 9.01s\n",
      "834:\tlearn: 14.5338126\ttest: 18.2811645\tbest: 18.2811645 (834)\ttotal: 45.4s\tremaining: 8.96s\n",
      "835:\tlearn: 14.5273458\ttest: 18.2795228\tbest: 18.2795228 (835)\ttotal: 45.4s\tremaining: 8.91s\n",
      "836:\tlearn: 14.5242365\ttest: 18.2794464\tbest: 18.2794464 (836)\ttotal: 45.5s\tremaining: 8.85s\n",
      "837:\tlearn: 14.5193660\ttest: 18.2788815\tbest: 18.2788815 (837)\ttotal: 45.5s\tremaining: 8.8s\n",
      "838:\tlearn: 14.5144094\ttest: 18.2775106\tbest: 18.2775106 (838)\ttotal: 45.6s\tremaining: 8.75s\n",
      "839:\tlearn: 14.5093610\ttest: 18.2760016\tbest: 18.2760016 (839)\ttotal: 45.6s\tremaining: 8.69s\n",
      "840:\tlearn: 14.5052631\ttest: 18.2751246\tbest: 18.2751246 (840)\ttotal: 45.7s\tremaining: 8.64s\n",
      "841:\tlearn: 14.4998962\ttest: 18.2727920\tbest: 18.2727920 (841)\ttotal: 45.8s\tremaining: 8.59s\n",
      "842:\tlearn: 14.4954758\ttest: 18.2709278\tbest: 18.2709278 (842)\ttotal: 45.8s\tremaining: 8.53s\n",
      "843:\tlearn: 14.4912715\ttest: 18.2687096\tbest: 18.2687096 (843)\ttotal: 45.9s\tremaining: 8.48s\n",
      "844:\tlearn: 14.4883934\ttest: 18.2683607\tbest: 18.2683607 (844)\ttotal: 45.9s\tremaining: 8.42s\n",
      "845:\tlearn: 14.4850871\ttest: 18.2668004\tbest: 18.2668004 (845)\ttotal: 46s\tremaining: 8.37s\n",
      "846:\tlearn: 14.4819624\ttest: 18.2646821\tbest: 18.2646821 (846)\ttotal: 46s\tremaining: 8.31s\n",
      "847:\tlearn: 14.4785066\ttest: 18.2624061\tbest: 18.2624061 (847)\ttotal: 46.1s\tremaining: 8.26s\n",
      "848:\tlearn: 14.4746102\ttest: 18.2605729\tbest: 18.2605729 (848)\ttotal: 46.1s\tremaining: 8.21s\n",
      "849:\tlearn: 14.4709229\ttest: 18.2604896\tbest: 18.2604896 (849)\ttotal: 46.2s\tremaining: 8.15s\n",
      "850:\tlearn: 14.4669280\ttest: 18.2581772\tbest: 18.2581772 (850)\ttotal: 46.2s\tremaining: 8.1s\n",
      "851:\tlearn: 14.4618825\ttest: 18.2571197\tbest: 18.2571197 (851)\ttotal: 46.3s\tremaining: 8.04s\n",
      "852:\tlearn: 14.4579355\ttest: 18.2550541\tbest: 18.2550541 (852)\ttotal: 46.4s\tremaining: 7.99s\n",
      "853:\tlearn: 14.4531610\ttest: 18.2530194\tbest: 18.2530194 (853)\ttotal: 46.4s\tremaining: 7.94s\n",
      "854:\tlearn: 14.4481376\ttest: 18.2515276\tbest: 18.2515276 (854)\ttotal: 46.5s\tremaining: 7.88s\n",
      "855:\tlearn: 14.4435551\ttest: 18.2497739\tbest: 18.2497739 (855)\ttotal: 46.5s\tremaining: 7.83s\n",
      "856:\tlearn: 14.4382009\ttest: 18.2472351\tbest: 18.2472351 (856)\ttotal: 46.6s\tremaining: 7.78s\n",
      "857:\tlearn: 14.4354566\ttest: 18.2457597\tbest: 18.2457597 (857)\ttotal: 46.6s\tremaining: 7.72s\n",
      "858:\tlearn: 14.4322478\ttest: 18.2451559\tbest: 18.2451559 (858)\ttotal: 46.7s\tremaining: 7.67s\n",
      "859:\tlearn: 14.4280800\ttest: 18.2430771\tbest: 18.2430771 (859)\ttotal: 46.8s\tremaining: 7.61s\n",
      "860:\tlearn: 14.4241794\ttest: 18.2415380\tbest: 18.2415380 (860)\ttotal: 46.8s\tremaining: 7.55s\n",
      "861:\tlearn: 14.4201482\ttest: 18.2409885\tbest: 18.2409885 (861)\ttotal: 46.9s\tremaining: 7.5s\n",
      "862:\tlearn: 14.4171856\ttest: 18.2405912\tbest: 18.2405912 (862)\ttotal: 46.9s\tremaining: 7.45s\n",
      "863:\tlearn: 14.4118759\ttest: 18.2387757\tbest: 18.2387757 (863)\ttotal: 47s\tremaining: 7.39s\n",
      "864:\tlearn: 14.4072251\ttest: 18.2364207\tbest: 18.2364207 (864)\ttotal: 47s\tremaining: 7.34s\n",
      "865:\tlearn: 14.4013412\ttest: 18.2331345\tbest: 18.2331345 (865)\ttotal: 47.1s\tremaining: 7.29s\n",
      "866:\tlearn: 14.3975079\ttest: 18.2321406\tbest: 18.2321406 (866)\ttotal: 47.1s\tremaining: 7.23s\n",
      "867:\tlearn: 14.3923964\ttest: 18.2297642\tbest: 18.2297642 (867)\ttotal: 47.2s\tremaining: 7.18s\n",
      "868:\tlearn: 14.3850860\ttest: 18.2273363\tbest: 18.2273363 (868)\ttotal: 47.3s\tremaining: 7.12s\n",
      "869:\tlearn: 14.3823749\ttest: 18.2270102\tbest: 18.2270102 (869)\ttotal: 47.3s\tremaining: 7.07s\n",
      "870:\tlearn: 14.3792961\ttest: 18.2262523\tbest: 18.2262523 (870)\ttotal: 47.4s\tremaining: 7.01s\n",
      "871:\tlearn: 14.3747673\ttest: 18.2250753\tbest: 18.2250753 (871)\ttotal: 47.4s\tremaining: 6.96s\n",
      "872:\tlearn: 14.3704039\ttest: 18.2231000\tbest: 18.2231000 (872)\ttotal: 47.5s\tremaining: 6.9s\n",
      "873:\tlearn: 14.3668748\ttest: 18.2220514\tbest: 18.2220514 (873)\ttotal: 47.5s\tremaining: 6.85s\n",
      "874:\tlearn: 14.3646076\ttest: 18.2219614\tbest: 18.2219614 (874)\ttotal: 47.6s\tremaining: 6.8s\n",
      "875:\tlearn: 14.3622064\ttest: 18.2216497\tbest: 18.2216497 (875)\ttotal: 47.7s\tremaining: 6.75s\n",
      "876:\tlearn: 14.3572861\ttest: 18.2192597\tbest: 18.2192597 (876)\ttotal: 47.7s\tremaining: 6.69s\n",
      "877:\tlearn: 14.3527446\ttest: 18.2176403\tbest: 18.2176403 (877)\ttotal: 47.8s\tremaining: 6.64s\n",
      "878:\tlearn: 14.3481771\ttest: 18.2164383\tbest: 18.2164383 (878)\ttotal: 47.8s\tremaining: 6.58s\n",
      "879:\tlearn: 14.3442304\ttest: 18.2141463\tbest: 18.2141463 (879)\ttotal: 47.9s\tremaining: 6.53s\n",
      "880:\tlearn: 14.3387412\ttest: 18.2120469\tbest: 18.2120469 (880)\ttotal: 48s\tremaining: 6.48s\n",
      "881:\tlearn: 14.3362359\ttest: 18.2106978\tbest: 18.2106978 (881)\ttotal: 48s\tremaining: 6.42s\n",
      "882:\tlearn: 14.3336903\ttest: 18.2102969\tbest: 18.2102969 (882)\ttotal: 48.1s\tremaining: 6.37s\n",
      "883:\tlearn: 14.3282393\ttest: 18.2088567\tbest: 18.2088567 (883)\ttotal: 48.1s\tremaining: 6.32s\n",
      "884:\tlearn: 14.3248151\ttest: 18.2069469\tbest: 18.2069469 (884)\ttotal: 48.2s\tremaining: 6.26s\n",
      "885:\tlearn: 14.3220715\ttest: 18.2058995\tbest: 18.2058995 (885)\ttotal: 48.3s\tremaining: 6.21s\n",
      "886:\tlearn: 14.3201664\ttest: 18.2052403\tbest: 18.2052403 (886)\ttotal: 48.3s\tremaining: 6.15s\n",
      "887:\tlearn: 14.3159795\ttest: 18.2042384\tbest: 18.2042384 (887)\ttotal: 48.4s\tremaining: 6.1s\n",
      "888:\tlearn: 14.3126885\ttest: 18.2040501\tbest: 18.2040501 (888)\ttotal: 48.4s\tremaining: 6.04s\n",
      "889:\tlearn: 14.3091111\ttest: 18.2038021\tbest: 18.2038021 (889)\ttotal: 48.5s\tremaining: 5.99s\n",
      "890:\tlearn: 14.3052223\ttest: 18.2038912\tbest: 18.2038021 (889)\ttotal: 48.5s\tremaining: 5.93s\n",
      "891:\tlearn: 14.2997243\ttest: 18.2001614\tbest: 18.2001614 (891)\ttotal: 48.6s\tremaining: 5.88s\n",
      "892:\tlearn: 14.2961998\ttest: 18.1996661\tbest: 18.1996661 (892)\ttotal: 48.6s\tremaining: 5.83s\n",
      "893:\tlearn: 14.2909614\ttest: 18.1983587\tbest: 18.1983587 (893)\ttotal: 48.7s\tremaining: 5.77s\n",
      "894:\tlearn: 14.2867119\ttest: 18.1980891\tbest: 18.1980891 (894)\ttotal: 48.7s\tremaining: 5.72s\n",
      "895:\tlearn: 14.2835974\ttest: 18.1966005\tbest: 18.1966005 (895)\ttotal: 48.8s\tremaining: 5.66s\n",
      "896:\tlearn: 14.2786215\ttest: 18.1951766\tbest: 18.1951766 (896)\ttotal: 48.8s\tremaining: 5.61s\n",
      "897:\tlearn: 14.2745516\ttest: 18.1941788\tbest: 18.1941788 (897)\ttotal: 48.9s\tremaining: 5.55s\n",
      "898:\tlearn: 14.2719659\ttest: 18.1937897\tbest: 18.1937897 (898)\ttotal: 48.9s\tremaining: 5.5s\n",
      "899:\tlearn: 14.2689756\ttest: 18.1930655\tbest: 18.1930655 (899)\ttotal: 49s\tremaining: 5.44s\n",
      "900:\tlearn: 14.2637002\ttest: 18.1910112\tbest: 18.1910112 (900)\ttotal: 49s\tremaining: 5.39s\n",
      "901:\tlearn: 14.2583124\ttest: 18.1892723\tbest: 18.1892723 (901)\ttotal: 49.1s\tremaining: 5.33s\n",
      "902:\tlearn: 14.2549660\ttest: 18.1872912\tbest: 18.1872912 (902)\ttotal: 49.1s\tremaining: 5.28s\n",
      "903:\tlearn: 14.2492271\ttest: 18.1855158\tbest: 18.1855158 (903)\ttotal: 49.2s\tremaining: 5.22s\n",
      "904:\tlearn: 14.2437248\ttest: 18.1849858\tbest: 18.1849858 (904)\ttotal: 49.2s\tremaining: 5.17s\n",
      "905:\tlearn: 14.2393849\ttest: 18.1842564\tbest: 18.1842564 (905)\ttotal: 49.3s\tremaining: 5.12s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "906:\tlearn: 14.2354112\ttest: 18.1827744\tbest: 18.1827744 (906)\ttotal: 49.4s\tremaining: 5.06s\n",
      "907:\tlearn: 14.2311559\ttest: 18.1814571\tbest: 18.1814571 (907)\ttotal: 49.4s\tremaining: 5.01s\n",
      "908:\tlearn: 14.2274990\ttest: 18.1808015\tbest: 18.1808015 (908)\ttotal: 49.5s\tremaining: 4.95s\n",
      "909:\tlearn: 14.2247651\ttest: 18.1793498\tbest: 18.1793498 (909)\ttotal: 49.5s\tremaining: 4.9s\n",
      "910:\tlearn: 14.2197113\ttest: 18.1785173\tbest: 18.1785173 (910)\ttotal: 49.6s\tremaining: 4.84s\n",
      "911:\tlearn: 14.2160307\ttest: 18.1767821\tbest: 18.1767821 (911)\ttotal: 49.7s\tremaining: 4.79s\n",
      "912:\tlearn: 14.2113071\ttest: 18.1755055\tbest: 18.1755055 (912)\ttotal: 49.7s\tremaining: 4.74s\n",
      "913:\tlearn: 14.2080172\ttest: 18.1740485\tbest: 18.1740485 (913)\ttotal: 49.8s\tremaining: 4.68s\n",
      "914:\tlearn: 14.2025107\ttest: 18.1732528\tbest: 18.1732528 (914)\ttotal: 49.8s\tremaining: 4.63s\n",
      "915:\tlearn: 14.1963832\ttest: 18.1707111\tbest: 18.1707111 (915)\ttotal: 49.9s\tremaining: 4.57s\n",
      "916:\tlearn: 14.1922432\ttest: 18.1679822\tbest: 18.1679822 (916)\ttotal: 49.9s\tremaining: 4.52s\n",
      "917:\tlearn: 14.1882614\ttest: 18.1671057\tbest: 18.1671057 (917)\ttotal: 50s\tremaining: 4.46s\n",
      "918:\tlearn: 14.1840836\ttest: 18.1662003\tbest: 18.1662003 (918)\ttotal: 50s\tremaining: 4.41s\n",
      "919:\tlearn: 14.1800578\ttest: 18.1641313\tbest: 18.1641313 (919)\ttotal: 50.1s\tremaining: 4.35s\n",
      "920:\tlearn: 14.1769313\ttest: 18.1638365\tbest: 18.1638365 (920)\ttotal: 50.1s\tremaining: 4.3s\n",
      "921:\tlearn: 14.1724553\ttest: 18.1629377\tbest: 18.1629377 (921)\ttotal: 50.1s\tremaining: 4.24s\n",
      "922:\tlearn: 14.1702957\ttest: 18.1624270\tbest: 18.1624270 (922)\ttotal: 50.2s\tremaining: 4.19s\n",
      "923:\tlearn: 14.1651514\ttest: 18.1600413\tbest: 18.1600413 (923)\ttotal: 50.2s\tremaining: 4.13s\n",
      "924:\tlearn: 14.1605247\ttest: 18.1584832\tbest: 18.1584832 (924)\ttotal: 50.3s\tremaining: 4.08s\n",
      "925:\tlearn: 14.1557768\ttest: 18.1574951\tbest: 18.1574951 (925)\ttotal: 50.3s\tremaining: 4.02s\n",
      "926:\tlearn: 14.1528434\ttest: 18.1565733\tbest: 18.1565733 (926)\ttotal: 50.4s\tremaining: 3.97s\n",
      "927:\tlearn: 14.1481818\ttest: 18.1552399\tbest: 18.1552399 (927)\ttotal: 50.5s\tremaining: 3.91s\n",
      "928:\tlearn: 14.1445418\ttest: 18.1529090\tbest: 18.1529090 (928)\ttotal: 50.5s\tremaining: 3.86s\n",
      "929:\tlearn: 14.1395139\ttest: 18.1514507\tbest: 18.1514507 (929)\ttotal: 50.6s\tremaining: 3.81s\n",
      "930:\tlearn: 14.1363626\ttest: 18.1501442\tbest: 18.1501442 (930)\ttotal: 50.6s\tremaining: 3.75s\n",
      "931:\tlearn: 14.1329952\ttest: 18.1499655\tbest: 18.1499655 (931)\ttotal: 50.7s\tremaining: 3.7s\n",
      "932:\tlearn: 14.1294542\ttest: 18.1500651\tbest: 18.1499655 (931)\ttotal: 50.7s\tremaining: 3.64s\n",
      "933:\tlearn: 14.1248846\ttest: 18.1491252\tbest: 18.1491252 (933)\ttotal: 50.8s\tremaining: 3.59s\n",
      "934:\tlearn: 14.1216095\ttest: 18.1478066\tbest: 18.1478066 (934)\ttotal: 50.8s\tremaining: 3.53s\n",
      "935:\tlearn: 14.1170803\ttest: 18.1460149\tbest: 18.1460149 (935)\ttotal: 50.9s\tremaining: 3.48s\n",
      "936:\tlearn: 14.1128582\ttest: 18.1441649\tbest: 18.1441649 (936)\ttotal: 50.9s\tremaining: 3.42s\n",
      "937:\tlearn: 14.1097026\ttest: 18.1428911\tbest: 18.1428911 (937)\ttotal: 51s\tremaining: 3.37s\n",
      "938:\tlearn: 14.1084000\ttest: 18.1422895\tbest: 18.1422895 (938)\ttotal: 51s\tremaining: 3.31s\n",
      "939:\tlearn: 14.1038716\ttest: 18.1392544\tbest: 18.1392544 (939)\ttotal: 51.1s\tremaining: 3.26s\n",
      "940:\tlearn: 14.0996777\ttest: 18.1383281\tbest: 18.1383281 (940)\ttotal: 51.1s\tremaining: 3.2s\n",
      "941:\tlearn: 14.0971370\ttest: 18.1387561\tbest: 18.1383281 (940)\ttotal: 51.2s\tremaining: 3.15s\n",
      "942:\tlearn: 14.0914538\ttest: 18.1373948\tbest: 18.1373948 (942)\ttotal: 51.2s\tremaining: 3.1s\n",
      "943:\tlearn: 14.0864653\ttest: 18.1362305\tbest: 18.1362305 (943)\ttotal: 51.3s\tremaining: 3.04s\n",
      "944:\tlearn: 14.0827559\ttest: 18.1348700\tbest: 18.1348700 (944)\ttotal: 51.3s\tremaining: 2.99s\n",
      "945:\tlearn: 14.0788729\ttest: 18.1330909\tbest: 18.1330909 (945)\ttotal: 51.4s\tremaining: 2.93s\n",
      "946:\tlearn: 14.0755123\ttest: 18.1319437\tbest: 18.1319437 (946)\ttotal: 51.4s\tremaining: 2.88s\n",
      "947:\tlearn: 14.0720840\ttest: 18.1306312\tbest: 18.1306312 (947)\ttotal: 51.5s\tremaining: 2.82s\n",
      "948:\tlearn: 14.0681159\ttest: 18.1288261\tbest: 18.1288261 (948)\ttotal: 51.5s\tremaining: 2.77s\n",
      "949:\tlearn: 14.0638466\ttest: 18.1290015\tbest: 18.1288261 (948)\ttotal: 51.6s\tremaining: 2.71s\n",
      "950:\tlearn: 14.0605750\ttest: 18.1279163\tbest: 18.1279163 (950)\ttotal: 51.6s\tremaining: 2.66s\n",
      "951:\tlearn: 14.0551422\ttest: 18.1266726\tbest: 18.1266726 (951)\ttotal: 51.7s\tremaining: 2.61s\n",
      "952:\tlearn: 14.0503426\ttest: 18.1254481\tbest: 18.1254481 (952)\ttotal: 51.8s\tremaining: 2.55s\n",
      "953:\tlearn: 14.0464770\ttest: 18.1237769\tbest: 18.1237769 (953)\ttotal: 51.8s\tremaining: 2.5s\n",
      "954:\tlearn: 14.0433645\ttest: 18.1228175\tbest: 18.1228175 (954)\ttotal: 51.8s\tremaining: 2.44s\n",
      "955:\tlearn: 14.0395265\ttest: 18.1219675\tbest: 18.1219675 (955)\ttotal: 51.9s\tremaining: 2.39s\n",
      "956:\tlearn: 14.0357226\ttest: 18.1192934\tbest: 18.1192934 (956)\ttotal: 51.9s\tremaining: 2.33s\n",
      "957:\tlearn: 14.0304080\ttest: 18.1177875\tbest: 18.1177875 (957)\ttotal: 52s\tremaining: 2.28s\n",
      "958:\tlearn: 14.0282683\ttest: 18.1174634\tbest: 18.1174634 (958)\ttotal: 52.1s\tremaining: 2.23s\n",
      "959:\tlearn: 14.0245696\ttest: 18.1155378\tbest: 18.1155378 (959)\ttotal: 52.1s\tremaining: 2.17s\n",
      "960:\tlearn: 14.0200615\ttest: 18.1143127\tbest: 18.1143127 (960)\ttotal: 52.2s\tremaining: 2.12s\n",
      "961:\tlearn: 14.0160576\ttest: 18.1133875\tbest: 18.1133875 (961)\ttotal: 52.2s\tremaining: 2.06s\n",
      "962:\tlearn: 14.0106856\ttest: 18.1128256\tbest: 18.1128256 (962)\ttotal: 52.3s\tremaining: 2.01s\n",
      "963:\tlearn: 14.0080384\ttest: 18.1128357\tbest: 18.1128256 (962)\ttotal: 52.3s\tremaining: 1.95s\n",
      "964:\tlearn: 14.0031975\ttest: 18.1119645\tbest: 18.1119645 (964)\ttotal: 52.3s\tremaining: 1.9s\n",
      "965:\tlearn: 13.9993206\ttest: 18.1103229\tbest: 18.1103229 (965)\ttotal: 52.4s\tremaining: 1.84s\n",
      "966:\tlearn: 13.9950910\ttest: 18.1096958\tbest: 18.1096958 (966)\ttotal: 52.4s\tremaining: 1.79s\n",
      "967:\tlearn: 13.9905270\ttest: 18.1077235\tbest: 18.1077235 (967)\ttotal: 52.5s\tremaining: 1.74s\n",
      "968:\tlearn: 13.9876586\ttest: 18.1064396\tbest: 18.1064396 (968)\ttotal: 52.5s\tremaining: 1.68s\n",
      "969:\tlearn: 13.9824300\ttest: 18.1046716\tbest: 18.1046716 (969)\ttotal: 52.6s\tremaining: 1.63s\n",
      "970:\tlearn: 13.9785209\ttest: 18.1034895\tbest: 18.1034895 (970)\ttotal: 52.7s\tremaining: 1.57s\n",
      "971:\tlearn: 13.9748781\ttest: 18.1029638\tbest: 18.1029638 (971)\ttotal: 52.7s\tremaining: 1.52s\n",
      "972:\tlearn: 13.9706441\ttest: 18.1016058\tbest: 18.1016058 (972)\ttotal: 52.8s\tremaining: 1.46s\n",
      "973:\tlearn: 13.9673376\ttest: 18.1008364\tbest: 18.1008364 (973)\ttotal: 52.8s\tremaining: 1.41s\n",
      "974:\tlearn: 13.9638206\ttest: 18.0997578\tbest: 18.0997578 (974)\ttotal: 52.9s\tremaining: 1.35s\n",
      "975:\tlearn: 13.9594115\ttest: 18.0976878\tbest: 18.0976878 (975)\ttotal: 52.9s\tremaining: 1.3s\n",
      "976:\tlearn: 13.9540039\ttest: 18.0955302\tbest: 18.0955302 (976)\ttotal: 53s\tremaining: 1.25s\n",
      "977:\tlearn: 13.9495503\ttest: 18.0943034\tbest: 18.0943034 (977)\ttotal: 53s\tremaining: 1.19s\n",
      "978:\tlearn: 13.9465029\ttest: 18.0932871\tbest: 18.0932871 (978)\ttotal: 53.1s\tremaining: 1.14s\n",
      "979:\tlearn: 13.9405638\ttest: 18.0911510\tbest: 18.0911510 (979)\ttotal: 53.1s\tremaining: 1.08s\n",
      "980:\tlearn: 13.9356889\ttest: 18.0906093\tbest: 18.0906093 (980)\ttotal: 53.1s\tremaining: 1.03s\n",
      "981:\tlearn: 13.9313328\ttest: 18.0895294\tbest: 18.0895294 (981)\ttotal: 53.2s\tremaining: 975ms\n",
      "982:\tlearn: 13.9277851\ttest: 18.0884627\tbest: 18.0884627 (982)\ttotal: 53.2s\tremaining: 921ms\n",
      "983:\tlearn: 13.9233700\ttest: 18.0872777\tbest: 18.0872777 (983)\ttotal: 53.3s\tremaining: 867ms\n",
      "984:\tlearn: 13.9192154\ttest: 18.0860883\tbest: 18.0860883 (984)\ttotal: 53.4s\tremaining: 813ms\n",
      "985:\tlearn: 13.9153087\ttest: 18.0851747\tbest: 18.0851747 (985)\ttotal: 53.4s\tremaining: 758ms\n",
      "986:\tlearn: 13.9118961\ttest: 18.0847360\tbest: 18.0847360 (986)\ttotal: 53.5s\tremaining: 704ms\n",
      "987:\tlearn: 13.9089409\ttest: 18.0824425\tbest: 18.0824425 (987)\ttotal: 53.5s\tremaining: 650ms\n",
      "988:\tlearn: 13.9040834\ttest: 18.0818263\tbest: 18.0818263 (988)\ttotal: 53.6s\tremaining: 596ms\n",
      "989:\tlearn: 13.8986598\ttest: 18.0792189\tbest: 18.0792189 (989)\ttotal: 53.6s\tremaining: 542ms\n",
      "990:\tlearn: 13.8936912\ttest: 18.0774375\tbest: 18.0774375 (990)\ttotal: 53.7s\tremaining: 488ms\n",
      "991:\tlearn: 13.8899476\ttest: 18.0766256\tbest: 18.0766256 (991)\ttotal: 53.7s\tremaining: 433ms\n",
      "992:\tlearn: 13.8872230\ttest: 18.0756314\tbest: 18.0756314 (992)\ttotal: 53.8s\tremaining: 379ms\n",
      "993:\tlearn: 13.8833543\ttest: 18.0736533\tbest: 18.0736533 (993)\ttotal: 53.8s\tremaining: 325ms\n",
      "994:\tlearn: 13.8799269\ttest: 18.0727305\tbest: 18.0727305 (994)\ttotal: 53.9s\tremaining: 271ms\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "995:\tlearn: 13.8775098\ttest: 18.0718681\tbest: 18.0718681 (995)\ttotal: 54s\tremaining: 217ms\n",
      "996:\tlearn: 13.8737299\ttest: 18.0701081\tbest: 18.0701081 (996)\ttotal: 54s\tremaining: 162ms\n",
      "997:\tlearn: 13.8706310\ttest: 18.0693682\tbest: 18.0693682 (997)\ttotal: 54s\tremaining: 108ms\n",
      "998:\tlearn: 13.8670653\ttest: 18.0685451\tbest: 18.0685451 (998)\ttotal: 54.1s\tremaining: 54.1ms\n",
      "999:\tlearn: 13.8640947\ttest: 18.0672249\tbest: 18.0672249 (999)\ttotal: 54.1s\tremaining: 0us\n",
      "\n",
      "bestTest = 18.06722495\n",
      "bestIteration = 999\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostRegressor at 0x7c764ae27e50>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from catboost import CatBoostRegressor\n",
    "\n",
    "model = CatBoostRegressor(iterations=1000,\n",
    "                             learning_rate=0.07,\n",
    "                             depth=12,\n",
    "                             eval_metric='RMSE',\n",
    "                             random_seed = 42,\n",
    "                             bagging_temperature = 0.3)\n",
    "model.fit(X_train, y_train,\n",
    "              eval_set=(X_val, y_val),\n",
    "              use_best_model=True,\n",
    "              verbose=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "c1339fc5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    mse -> 328.0699356325211\n",
      "    mae -> 14.363748889907411\n",
      "    rmse -> 18.112700948023214\n",
      "    mape -> 1.2354343837832418e+16\n",
      "    r2 -> 0.34461343092479235\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ab4941b",
   "metadata": {},
   "source": [
    "<a id='9'>9 Обучение VotingRegressor catboost + randomforest</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "4e89150b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 22.2912801\ttotal: 64.1ms\tremaining: 1m 4s\n",
      "1:\tlearn: 22.2004646\ttotal: 116ms\tremaining: 57.8s\n",
      "2:\tlearn: 22.1127524\ttotal: 176ms\tremaining: 58.6s\n",
      "3:\tlearn: 22.0332782\ttotal: 240ms\tremaining: 59.7s\n",
      "4:\tlearn: 21.9582692\ttotal: 295ms\tremaining: 58.7s\n",
      "5:\tlearn: 21.8959086\ttotal: 351ms\tremaining: 58.1s\n",
      "6:\tlearn: 21.8374353\ttotal: 410ms\tremaining: 58.1s\n",
      "7:\tlearn: 21.7774568\ttotal: 465ms\tremaining: 57.7s\n",
      "8:\tlearn: 21.7250648\ttotal: 525ms\tremaining: 57.8s\n",
      "9:\tlearn: 21.6709969\ttotal: 586ms\tremaining: 58s\n",
      "10:\tlearn: 21.6287253\ttotal: 635ms\tremaining: 57.1s\n",
      "11:\tlearn: 21.5834541\ttotal: 684ms\tremaining: 56.3s\n",
      "12:\tlearn: 21.5475284\ttotal: 735ms\tremaining: 55.8s\n",
      "13:\tlearn: 21.5007653\ttotal: 789ms\tremaining: 55.6s\n",
      "14:\tlearn: 21.4590551\ttotal: 839ms\tremaining: 55.1s\n",
      "15:\tlearn: 21.4209190\ttotal: 894ms\tremaining: 55s\n",
      "16:\tlearn: 21.3917083\ttotal: 951ms\tremaining: 55s\n",
      "17:\tlearn: 21.3701648\ttotal: 1.01s\tremaining: 55s\n",
      "18:\tlearn: 21.3356802\ttotal: 1.06s\tremaining: 55s\n",
      "19:\tlearn: 21.2986322\ttotal: 1.13s\tremaining: 55.4s\n",
      "20:\tlearn: 21.2745092\ttotal: 1.2s\tremaining: 55.7s\n",
      "21:\tlearn: 21.2423051\ttotal: 1.25s\tremaining: 55.6s\n",
      "22:\tlearn: 21.2143314\ttotal: 1.3s\tremaining: 55.2s\n",
      "23:\tlearn: 21.1908614\ttotal: 1.35s\tremaining: 55.1s\n",
      "24:\tlearn: 21.1566716\ttotal: 1.41s\tremaining: 55.1s\n",
      "25:\tlearn: 21.1179933\ttotal: 1.47s\tremaining: 54.9s\n",
      "26:\tlearn: 21.0920471\ttotal: 1.51s\tremaining: 54.6s\n",
      "27:\tlearn: 21.0621047\ttotal: 1.56s\tremaining: 54.3s\n",
      "28:\tlearn: 21.0358210\ttotal: 1.61s\tremaining: 54s\n",
      "29:\tlearn: 21.0108743\ttotal: 1.66s\tremaining: 53.8s\n",
      "30:\tlearn: 20.9839595\ttotal: 1.72s\tremaining: 53.9s\n",
      "31:\tlearn: 20.9613697\ttotal: 1.78s\tremaining: 54s\n",
      "32:\tlearn: 20.9378428\ttotal: 1.84s\tremaining: 54s\n",
      "33:\tlearn: 20.9035121\ttotal: 1.89s\tremaining: 53.8s\n",
      "34:\tlearn: 20.8839560\ttotal: 1.94s\tremaining: 53.5s\n",
      "35:\tlearn: 20.8725899\ttotal: 1.99s\tremaining: 53.2s\n",
      "36:\tlearn: 20.8450602\ttotal: 2.04s\tremaining: 53s\n",
      "37:\tlearn: 20.8258579\ttotal: 2.1s\tremaining: 53.2s\n",
      "38:\tlearn: 20.7997709\ttotal: 2.16s\tremaining: 53.1s\n",
      "39:\tlearn: 20.7760800\ttotal: 2.21s\tremaining: 53.1s\n",
      "40:\tlearn: 20.7603086\ttotal: 2.26s\tremaining: 52.8s\n",
      "41:\tlearn: 20.7347065\ttotal: 2.31s\tremaining: 52.6s\n",
      "42:\tlearn: 20.7094003\ttotal: 2.35s\tremaining: 52.3s\n",
      "43:\tlearn: 20.6781304\ttotal: 2.41s\tremaining: 52.3s\n",
      "44:\tlearn: 20.6584431\ttotal: 2.47s\tremaining: 52.5s\n",
      "45:\tlearn: 20.6439984\ttotal: 2.52s\tremaining: 52.2s\n",
      "46:\tlearn: 20.6353169\ttotal: 2.56s\tremaining: 52s\n",
      "47:\tlearn: 20.6172512\ttotal: 2.61s\tremaining: 51.8s\n",
      "48:\tlearn: 20.6024316\ttotal: 2.66s\tremaining: 51.6s\n",
      "49:\tlearn: 20.5698755\ttotal: 2.72s\tremaining: 51.6s\n",
      "50:\tlearn: 20.5547686\ttotal: 2.76s\tremaining: 51.4s\n",
      "51:\tlearn: 20.5301251\ttotal: 2.81s\tremaining: 51.2s\n",
      "52:\tlearn: 20.5125627\ttotal: 2.86s\tremaining: 51.1s\n",
      "53:\tlearn: 20.4921225\ttotal: 2.93s\tremaining: 51.3s\n",
      "54:\tlearn: 20.4766977\ttotal: 2.98s\tremaining: 51.2s\n",
      "55:\tlearn: 20.4596701\ttotal: 3.04s\tremaining: 51.3s\n",
      "56:\tlearn: 20.4441647\ttotal: 3.1s\tremaining: 51.3s\n",
      "57:\tlearn: 20.4325713\ttotal: 3.16s\tremaining: 51.4s\n",
      "58:\tlearn: 20.4153048\ttotal: 3.22s\tremaining: 51.4s\n",
      "59:\tlearn: 20.3969028\ttotal: 3.28s\tremaining: 51.4s\n",
      "60:\tlearn: 20.3795113\ttotal: 3.34s\tremaining: 51.4s\n",
      "61:\tlearn: 20.3607085\ttotal: 3.39s\tremaining: 51.3s\n",
      "62:\tlearn: 20.3444683\ttotal: 3.45s\tremaining: 51.3s\n",
      "63:\tlearn: 20.3293227\ttotal: 3.52s\tremaining: 51.4s\n",
      "64:\tlearn: 20.3125951\ttotal: 3.57s\tremaining: 51.4s\n",
      "65:\tlearn: 20.3002067\ttotal: 3.62s\tremaining: 51.2s\n",
      "66:\tlearn: 20.2876250\ttotal: 3.68s\tremaining: 51.2s\n",
      "67:\tlearn: 20.2615646\ttotal: 3.73s\tremaining: 51.1s\n",
      "68:\tlearn: 20.2492252\ttotal: 3.79s\tremaining: 51.1s\n",
      "69:\tlearn: 20.2309275\ttotal: 3.85s\tremaining: 51.1s\n",
      "70:\tlearn: 20.2162640\ttotal: 3.9s\tremaining: 51s\n",
      "71:\tlearn: 20.2040666\ttotal: 3.94s\tremaining: 50.9s\n",
      "72:\tlearn: 20.1754834\ttotal: 4s\tremaining: 50.8s\n",
      "73:\tlearn: 20.1573566\ttotal: 4.06s\tremaining: 50.8s\n",
      "74:\tlearn: 20.1445044\ttotal: 4.11s\tremaining: 50.7s\n",
      "75:\tlearn: 20.1303499\ttotal: 4.17s\tremaining: 50.7s\n",
      "76:\tlearn: 20.1132019\ttotal: 4.22s\tremaining: 50.6s\n",
      "77:\tlearn: 20.0996436\ttotal: 4.28s\tremaining: 50.6s\n",
      "78:\tlearn: 20.0817331\ttotal: 4.33s\tremaining: 50.5s\n",
      "79:\tlearn: 20.0742072\ttotal: 4.39s\tremaining: 50.5s\n",
      "80:\tlearn: 20.0567057\ttotal: 4.45s\tremaining: 50.5s\n",
      "81:\tlearn: 20.0479478\ttotal: 4.51s\tremaining: 50.6s\n",
      "82:\tlearn: 20.0378982\ttotal: 4.56s\tremaining: 50.4s\n",
      "83:\tlearn: 20.0274809\ttotal: 4.61s\tremaining: 50.3s\n",
      "84:\tlearn: 19.9993988\ttotal: 4.65s\tremaining: 50.1s\n",
      "85:\tlearn: 19.9833795\ttotal: 4.71s\tremaining: 50s\n",
      "86:\tlearn: 19.9631700\ttotal: 4.77s\tremaining: 50.1s\n",
      "87:\tlearn: 19.9533262\ttotal: 4.84s\tremaining: 50.2s\n",
      "88:\tlearn: 19.9441068\ttotal: 4.9s\tremaining: 50.2s\n",
      "89:\tlearn: 19.9289387\ttotal: 4.95s\tremaining: 50.1s\n",
      "90:\tlearn: 19.9063314\ttotal: 5.01s\tremaining: 50s\n",
      "91:\tlearn: 19.8937860\ttotal: 5.05s\tremaining: 49.9s\n",
      "92:\tlearn: 19.8806913\ttotal: 5.09s\tremaining: 49.7s\n",
      "93:\tlearn: 19.8592544\ttotal: 5.14s\tremaining: 49.5s\n",
      "94:\tlearn: 19.8417541\ttotal: 5.19s\tremaining: 49.5s\n",
      "95:\tlearn: 19.8307640\ttotal: 5.25s\tremaining: 49.5s\n",
      "96:\tlearn: 19.8089873\ttotal: 5.31s\tremaining: 49.5s\n",
      "97:\tlearn: 19.7876890\ttotal: 5.38s\tremaining: 49.5s\n",
      "98:\tlearn: 19.7807996\ttotal: 5.44s\tremaining: 49.5s\n",
      "99:\tlearn: 19.7644749\ttotal: 5.5s\tremaining: 49.6s\n",
      "100:\tlearn: 19.7493727\ttotal: 5.56s\tremaining: 49.5s\n",
      "101:\tlearn: 19.7332162\ttotal: 5.62s\tremaining: 49.5s\n",
      "102:\tlearn: 19.7207326\ttotal: 5.68s\tremaining: 49.5s\n",
      "103:\tlearn: 19.7080709\ttotal: 5.74s\tremaining: 49.4s\n",
      "104:\tlearn: 19.6941615\ttotal: 5.79s\tremaining: 49.4s\n",
      "105:\tlearn: 19.6857420\ttotal: 5.85s\tremaining: 49.3s\n",
      "106:\tlearn: 19.6766671\ttotal: 5.9s\tremaining: 49.2s\n",
      "107:\tlearn: 19.6564752\ttotal: 5.95s\tremaining: 49.1s\n",
      "108:\tlearn: 19.6381822\ttotal: 5.99s\tremaining: 49s\n",
      "109:\tlearn: 19.6377216\ttotal: 6s\tremaining: 48.5s\n",
      "110:\tlearn: 19.6274833\ttotal: 6.04s\tremaining: 48.4s\n",
      "111:\tlearn: 19.6137166\ttotal: 6.1s\tremaining: 48.4s\n",
      "112:\tlearn: 19.5996144\ttotal: 6.15s\tremaining: 48.3s\n",
      "113:\tlearn: 19.5781915\ttotal: 6.21s\tremaining: 48.2s\n",
      "114:\tlearn: 19.5678958\ttotal: 6.25s\tremaining: 48.1s\n",
      "115:\tlearn: 19.5509691\ttotal: 6.3s\tremaining: 48s\n",
      "116:\tlearn: 19.5380183\ttotal: 6.35s\tremaining: 47.9s\n",
      "117:\tlearn: 19.5286785\ttotal: 6.41s\tremaining: 47.9s\n",
      "118:\tlearn: 19.5112377\ttotal: 6.47s\tremaining: 47.9s\n",
      "119:\tlearn: 19.4860283\ttotal: 6.55s\tremaining: 48s\n",
      "120:\tlearn: 19.4756405\ttotal: 6.61s\tremaining: 48s\n",
      "121:\tlearn: 19.4594374\ttotal: 6.67s\tremaining: 48s\n",
      "122:\tlearn: 19.4436935\ttotal: 6.72s\tremaining: 47.9s\n",
      "123:\tlearn: 19.4226676\ttotal: 6.79s\tremaining: 48s\n",
      "124:\tlearn: 19.4105579\ttotal: 6.83s\tremaining: 47.8s\n",
      "125:\tlearn: 19.4009554\ttotal: 6.89s\tremaining: 47.8s\n",
      "126:\tlearn: 19.3836898\ttotal: 6.95s\tremaining: 47.8s\n",
      "127:\tlearn: 19.3718353\ttotal: 7.01s\tremaining: 47.7s\n",
      "128:\tlearn: 19.3570904\ttotal: 7.05s\tremaining: 47.6s\n",
      "129:\tlearn: 19.3461597\ttotal: 7.1s\tremaining: 47.5s\n",
      "130:\tlearn: 19.3384591\ttotal: 7.16s\tremaining: 47.5s\n",
      "131:\tlearn: 19.3288368\ttotal: 7.21s\tremaining: 47.4s\n",
      "132:\tlearn: 19.3185423\ttotal: 7.26s\tremaining: 47.4s\n",
      "133:\tlearn: 19.2978564\ttotal: 7.32s\tremaining: 47.3s\n",
      "134:\tlearn: 19.2836678\ttotal: 7.37s\tremaining: 47.2s\n",
      "135:\tlearn: 19.2668457\ttotal: 7.43s\tremaining: 47.2s\n",
      "136:\tlearn: 19.2536618\ttotal: 7.48s\tremaining: 47.1s\n",
      "137:\tlearn: 19.2365400\ttotal: 7.53s\tremaining: 47.1s\n",
      "138:\tlearn: 19.2246186\ttotal: 7.59s\tremaining: 47s\n",
      "139:\tlearn: 19.2076168\ttotal: 7.64s\tremaining: 46.9s\n",
      "140:\tlearn: 19.2001608\ttotal: 7.71s\tremaining: 46.9s\n",
      "141:\tlearn: 19.1877948\ttotal: 7.76s\tremaining: 46.9s\n",
      "142:\tlearn: 19.1739167\ttotal: 7.82s\tremaining: 46.9s\n",
      "143:\tlearn: 19.1621871\ttotal: 7.88s\tremaining: 46.8s\n",
      "144:\tlearn: 19.1547001\ttotal: 7.93s\tremaining: 46.8s\n",
      "145:\tlearn: 19.1462605\ttotal: 7.99s\tremaining: 46.7s\n",
      "146:\tlearn: 19.1305866\ttotal: 8.05s\tremaining: 46.7s\n",
      "147:\tlearn: 19.1135087\ttotal: 8.11s\tremaining: 46.7s\n",
      "148:\tlearn: 19.0980617\ttotal: 8.17s\tremaining: 46.7s\n",
      "149:\tlearn: 19.0927202\ttotal: 8.22s\tremaining: 46.6s\n",
      "150:\tlearn: 19.0811699\ttotal: 8.27s\tremaining: 46.5s\n",
      "151:\tlearn: 19.0724522\ttotal: 8.33s\tremaining: 46.5s\n",
      "152:\tlearn: 19.0609204\ttotal: 8.38s\tremaining: 46.4s\n",
      "153:\tlearn: 19.0521904\ttotal: 8.44s\tremaining: 46.4s\n",
      "154:\tlearn: 19.0434130\ttotal: 8.49s\tremaining: 46.3s\n",
      "155:\tlearn: 19.0309869\ttotal: 8.54s\tremaining: 46.2s\n",
      "156:\tlearn: 19.0212813\ttotal: 8.58s\tremaining: 46.1s\n",
      "157:\tlearn: 19.0102506\ttotal: 8.63s\tremaining: 46s\n",
      "158:\tlearn: 19.0022748\ttotal: 8.68s\tremaining: 45.9s\n",
      "159:\tlearn: 18.9911085\ttotal: 8.75s\tremaining: 45.9s\n",
      "160:\tlearn: 18.9824962\ttotal: 8.81s\tremaining: 45.9s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "161:\tlearn: 18.9724649\ttotal: 8.88s\tremaining: 45.9s\n",
      "162:\tlearn: 18.9608349\ttotal: 8.94s\tremaining: 45.9s\n",
      "163:\tlearn: 18.9489859\ttotal: 9s\tremaining: 45.9s\n",
      "164:\tlearn: 18.9353313\ttotal: 9.06s\tremaining: 45.9s\n",
      "165:\tlearn: 18.9250783\ttotal: 9.11s\tremaining: 45.8s\n",
      "166:\tlearn: 18.9147608\ttotal: 9.18s\tremaining: 45.8s\n",
      "167:\tlearn: 18.9023277\ttotal: 9.24s\tremaining: 45.7s\n",
      "168:\tlearn: 18.8866921\ttotal: 9.29s\tremaining: 45.7s\n",
      "169:\tlearn: 18.8729592\ttotal: 9.35s\tremaining: 45.6s\n",
      "170:\tlearn: 18.8613589\ttotal: 9.41s\tremaining: 45.6s\n",
      "171:\tlearn: 18.8505085\ttotal: 9.45s\tremaining: 45.5s\n",
      "172:\tlearn: 18.8355229\ttotal: 9.5s\tremaining: 45.4s\n",
      "173:\tlearn: 18.8307624\ttotal: 9.55s\tremaining: 45.3s\n",
      "174:\tlearn: 18.8261647\ttotal: 9.61s\tremaining: 45.3s\n",
      "175:\tlearn: 18.8151806\ttotal: 9.66s\tremaining: 45.2s\n",
      "176:\tlearn: 18.8073029\ttotal: 9.72s\tremaining: 45.2s\n",
      "177:\tlearn: 18.7981330\ttotal: 9.79s\tremaining: 45.2s\n",
      "178:\tlearn: 18.7913824\ttotal: 9.85s\tremaining: 45.2s\n",
      "179:\tlearn: 18.7812483\ttotal: 9.9s\tremaining: 45.1s\n",
      "180:\tlearn: 18.7688017\ttotal: 9.95s\tremaining: 45s\n",
      "181:\tlearn: 18.7628546\ttotal: 10s\tremaining: 44.9s\n",
      "182:\tlearn: 18.7480398\ttotal: 10s\tremaining: 44.9s\n",
      "183:\tlearn: 18.7400123\ttotal: 10.1s\tremaining: 44.8s\n",
      "184:\tlearn: 18.7361774\ttotal: 10.2s\tremaining: 44.8s\n",
      "185:\tlearn: 18.7279986\ttotal: 10.2s\tremaining: 44.7s\n",
      "186:\tlearn: 18.7204149\ttotal: 10.3s\tremaining: 44.6s\n",
      "187:\tlearn: 18.7103172\ttotal: 10.3s\tremaining: 44.5s\n",
      "188:\tlearn: 18.6991811\ttotal: 10.4s\tremaining: 44.5s\n",
      "189:\tlearn: 18.6882012\ttotal: 10.4s\tremaining: 44.4s\n",
      "190:\tlearn: 18.6749815\ttotal: 10.5s\tremaining: 44.3s\n",
      "191:\tlearn: 18.6654017\ttotal: 10.5s\tremaining: 44.2s\n",
      "192:\tlearn: 18.6527558\ttotal: 10.6s\tremaining: 44.2s\n",
      "193:\tlearn: 18.6370132\ttotal: 10.6s\tremaining: 44.2s\n",
      "194:\tlearn: 18.6308933\ttotal: 10.7s\tremaining: 44.1s\n",
      "195:\tlearn: 18.6182095\ttotal: 10.7s\tremaining: 44s\n",
      "196:\tlearn: 18.6059468\ttotal: 10.8s\tremaining: 43.9s\n",
      "197:\tlearn: 18.5931064\ttotal: 10.8s\tremaining: 43.9s\n",
      "198:\tlearn: 18.5846699\ttotal: 10.9s\tremaining: 43.8s\n",
      "199:\tlearn: 18.5723050\ttotal: 10.9s\tremaining: 43.7s\n",
      "200:\tlearn: 18.5628640\ttotal: 11s\tremaining: 43.7s\n",
      "201:\tlearn: 18.5531893\ttotal: 11s\tremaining: 43.6s\n",
      "202:\tlearn: 18.5344913\ttotal: 11.1s\tremaining: 43.5s\n",
      "203:\tlearn: 18.5240311\ttotal: 11.1s\tremaining: 43.4s\n",
      "204:\tlearn: 18.5114924\ttotal: 11.2s\tremaining: 43.3s\n",
      "205:\tlearn: 18.4977235\ttotal: 11.2s\tremaining: 43.3s\n",
      "206:\tlearn: 18.4906079\ttotal: 11.3s\tremaining: 43.3s\n",
      "207:\tlearn: 18.4815996\ttotal: 11.3s\tremaining: 43.2s\n",
      "208:\tlearn: 18.4746296\ttotal: 11.4s\tremaining: 43.1s\n",
      "209:\tlearn: 18.4591367\ttotal: 11.5s\tremaining: 43.1s\n",
      "210:\tlearn: 18.4473468\ttotal: 11.5s\tremaining: 43.1s\n",
      "211:\tlearn: 18.4367445\ttotal: 11.6s\tremaining: 43s\n",
      "212:\tlearn: 18.4201504\ttotal: 11.6s\tremaining: 43s\n",
      "213:\tlearn: 18.4128507\ttotal: 11.7s\tremaining: 43s\n",
      "214:\tlearn: 18.4015768\ttotal: 11.7s\tremaining: 42.9s\n",
      "215:\tlearn: 18.3927965\ttotal: 11.8s\tremaining: 42.8s\n",
      "216:\tlearn: 18.3803636\ttotal: 11.8s\tremaining: 42.7s\n",
      "217:\tlearn: 18.3731436\ttotal: 11.9s\tremaining: 42.7s\n",
      "218:\tlearn: 18.3651487\ttotal: 11.9s\tremaining: 42.6s\n",
      "219:\tlearn: 18.3539493\ttotal: 12s\tremaining: 42.5s\n",
      "220:\tlearn: 18.3459706\ttotal: 12s\tremaining: 42.4s\n",
      "221:\tlearn: 18.3351836\ttotal: 12.1s\tremaining: 42.4s\n",
      "222:\tlearn: 18.3235177\ttotal: 12.1s\tremaining: 42.3s\n",
      "223:\tlearn: 18.3136966\ttotal: 12.2s\tremaining: 42.3s\n",
      "224:\tlearn: 18.3061972\ttotal: 12.3s\tremaining: 42.2s\n",
      "225:\tlearn: 18.2940382\ttotal: 12.3s\tremaining: 42.2s\n",
      "226:\tlearn: 18.2839089\ttotal: 12.4s\tremaining: 42.1s\n",
      "227:\tlearn: 18.2762610\ttotal: 12.4s\tremaining: 42.1s\n",
      "228:\tlearn: 18.2643946\ttotal: 12.5s\tremaining: 42s\n",
      "229:\tlearn: 18.2580706\ttotal: 12.5s\tremaining: 41.9s\n",
      "230:\tlearn: 18.2468957\ttotal: 12.6s\tremaining: 41.9s\n",
      "231:\tlearn: 18.2356762\ttotal: 12.6s\tremaining: 41.9s\n",
      "232:\tlearn: 18.2296110\ttotal: 12.7s\tremaining: 41.8s\n",
      "233:\tlearn: 18.2190514\ttotal: 12.8s\tremaining: 41.8s\n",
      "234:\tlearn: 18.2057382\ttotal: 12.8s\tremaining: 41.7s\n",
      "235:\tlearn: 18.1963988\ttotal: 12.9s\tremaining: 41.7s\n",
      "236:\tlearn: 18.1835370\ttotal: 12.9s\tremaining: 41.6s\n",
      "237:\tlearn: 18.1715848\ttotal: 13s\tremaining: 41.6s\n",
      "238:\tlearn: 18.1607523\ttotal: 13s\tremaining: 41.5s\n",
      "239:\tlearn: 18.1490861\ttotal: 13.1s\tremaining: 41.5s\n",
      "240:\tlearn: 18.1379833\ttotal: 13.2s\tremaining: 41.4s\n",
      "241:\tlearn: 18.1260671\ttotal: 13.2s\tremaining: 41.4s\n",
      "242:\tlearn: 18.1173903\ttotal: 13.3s\tremaining: 41.4s\n",
      "243:\tlearn: 18.1063676\ttotal: 13.3s\tremaining: 41.3s\n",
      "244:\tlearn: 18.0991333\ttotal: 13.4s\tremaining: 41.3s\n",
      "245:\tlearn: 18.0838081\ttotal: 13.5s\tremaining: 41.2s\n",
      "246:\tlearn: 18.0771369\ttotal: 13.5s\tremaining: 41.2s\n",
      "247:\tlearn: 18.0700838\ttotal: 13.6s\tremaining: 41.2s\n",
      "248:\tlearn: 18.0579809\ttotal: 13.6s\tremaining: 41.1s\n",
      "249:\tlearn: 18.0466707\ttotal: 13.7s\tremaining: 41.1s\n",
      "250:\tlearn: 18.0348940\ttotal: 13.8s\tremaining: 41.1s\n",
      "251:\tlearn: 18.0298721\ttotal: 13.8s\tremaining: 41s\n",
      "252:\tlearn: 18.0227635\ttotal: 13.9s\tremaining: 40.9s\n",
      "253:\tlearn: 18.0157566\ttotal: 13.9s\tremaining: 40.9s\n",
      "254:\tlearn: 18.0053224\ttotal: 14s\tremaining: 40.8s\n",
      "255:\tlearn: 17.9996618\ttotal: 14s\tremaining: 40.8s\n",
      "256:\tlearn: 17.9885189\ttotal: 14.1s\tremaining: 40.7s\n",
      "257:\tlearn: 17.9785885\ttotal: 14.1s\tremaining: 40.7s\n",
      "258:\tlearn: 17.9678113\ttotal: 14.2s\tremaining: 40.6s\n",
      "259:\tlearn: 17.9632412\ttotal: 14.3s\tremaining: 40.6s\n",
      "260:\tlearn: 17.9532217\ttotal: 14.3s\tremaining: 40.5s\n",
      "261:\tlearn: 17.9453601\ttotal: 14.3s\tremaining: 40.4s\n",
      "262:\tlearn: 17.9357162\ttotal: 14.4s\tremaining: 40.4s\n",
      "263:\tlearn: 17.9302293\ttotal: 14.4s\tremaining: 40.3s\n",
      "264:\tlearn: 17.9225954\ttotal: 14.5s\tremaining: 40.2s\n",
      "265:\tlearn: 17.9164174\ttotal: 14.5s\tremaining: 40.1s\n",
      "266:\tlearn: 17.9073817\ttotal: 14.6s\tremaining: 40s\n",
      "267:\tlearn: 17.8925186\ttotal: 14.6s\tremaining: 39.9s\n",
      "268:\tlearn: 17.8817381\ttotal: 14.6s\tremaining: 39.8s\n",
      "269:\tlearn: 17.8760475\ttotal: 14.7s\tremaining: 39.7s\n",
      "270:\tlearn: 17.8660791\ttotal: 14.7s\tremaining: 39.6s\n",
      "271:\tlearn: 17.8544128\ttotal: 14.8s\tremaining: 39.6s\n",
      "272:\tlearn: 17.8478226\ttotal: 14.8s\tremaining: 39.5s\n",
      "273:\tlearn: 17.8367082\ttotal: 14.9s\tremaining: 39.4s\n",
      "274:\tlearn: 17.8311689\ttotal: 14.9s\tremaining: 39.3s\n",
      "275:\tlearn: 17.8238743\ttotal: 15s\tremaining: 39.2s\n",
      "276:\tlearn: 17.8198506\ttotal: 15s\tremaining: 39.2s\n",
      "277:\tlearn: 17.8141577\ttotal: 15.1s\tremaining: 39.1s\n",
      "278:\tlearn: 17.8040099\ttotal: 15.1s\tremaining: 39.1s\n",
      "279:\tlearn: 17.7939896\ttotal: 15.2s\tremaining: 39s\n",
      "280:\tlearn: 17.7886205\ttotal: 15.2s\tremaining: 39s\n",
      "281:\tlearn: 17.7809532\ttotal: 15.3s\tremaining: 38.9s\n",
      "282:\tlearn: 17.7686577\ttotal: 15.3s\tremaining: 38.8s\n",
      "283:\tlearn: 17.7618311\ttotal: 15.4s\tremaining: 38.8s\n",
      "284:\tlearn: 17.7566017\ttotal: 15.4s\tremaining: 38.7s\n",
      "285:\tlearn: 17.7503183\ttotal: 15.5s\tremaining: 38.6s\n",
      "286:\tlearn: 17.7408314\ttotal: 15.5s\tremaining: 38.6s\n",
      "287:\tlearn: 17.7314321\ttotal: 15.6s\tremaining: 38.5s\n",
      "288:\tlearn: 17.7219005\ttotal: 15.7s\tremaining: 38.5s\n",
      "289:\tlearn: 17.7113579\ttotal: 15.7s\tremaining: 38.5s\n",
      "290:\tlearn: 17.7038166\ttotal: 15.8s\tremaining: 38.4s\n",
      "291:\tlearn: 17.6966537\ttotal: 15.8s\tremaining: 38.3s\n",
      "292:\tlearn: 17.6857925\ttotal: 15.9s\tremaining: 38.3s\n",
      "293:\tlearn: 17.6793356\ttotal: 15.9s\tremaining: 38.3s\n",
      "294:\tlearn: 17.6690618\ttotal: 16s\tremaining: 38.2s\n",
      "295:\tlearn: 17.6606743\ttotal: 16s\tremaining: 38.1s\n",
      "296:\tlearn: 17.6505277\ttotal: 16.1s\tremaining: 38s\n",
      "297:\tlearn: 17.6449483\ttotal: 16.1s\tremaining: 38s\n",
      "298:\tlearn: 17.6378985\ttotal: 16.2s\tremaining: 37.9s\n",
      "299:\tlearn: 17.6301905\ttotal: 16.2s\tremaining: 37.9s\n",
      "300:\tlearn: 17.6214945\ttotal: 16.3s\tremaining: 37.8s\n",
      "301:\tlearn: 17.6124654\ttotal: 16.4s\tremaining: 37.8s\n",
      "302:\tlearn: 17.6036310\ttotal: 16.4s\tremaining: 37.8s\n",
      "303:\tlearn: 17.5911428\ttotal: 16.5s\tremaining: 37.7s\n",
      "304:\tlearn: 17.5830339\ttotal: 16.5s\tremaining: 37.7s\n",
      "305:\tlearn: 17.5725124\ttotal: 16.6s\tremaining: 37.6s\n",
      "306:\tlearn: 17.5658402\ttotal: 16.6s\tremaining: 37.5s\n",
      "307:\tlearn: 17.5563551\ttotal: 16.7s\tremaining: 37.4s\n",
      "308:\tlearn: 17.5488802\ttotal: 16.7s\tremaining: 37.4s\n",
      "309:\tlearn: 17.5421619\ttotal: 16.8s\tremaining: 37.3s\n",
      "310:\tlearn: 17.5321720\ttotal: 16.8s\tremaining: 37.2s\n",
      "311:\tlearn: 17.5248778\ttotal: 16.9s\tremaining: 37.2s\n",
      "312:\tlearn: 17.5166915\ttotal: 16.9s\tremaining: 37.1s\n",
      "313:\tlearn: 17.5070089\ttotal: 17s\tremaining: 37s\n",
      "314:\tlearn: 17.5013079\ttotal: 17s\tremaining: 37s\n",
      "315:\tlearn: 17.4915792\ttotal: 17.1s\tremaining: 36.9s\n",
      "316:\tlearn: 17.4841249\ttotal: 17.1s\tremaining: 36.9s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "317:\tlearn: 17.4750727\ttotal: 17.2s\tremaining: 36.8s\n",
      "318:\tlearn: 17.4709408\ttotal: 17.2s\tremaining: 36.7s\n",
      "319:\tlearn: 17.4612827\ttotal: 17.3s\tremaining: 36.7s\n",
      "320:\tlearn: 17.4539686\ttotal: 17.3s\tremaining: 36.6s\n",
      "321:\tlearn: 17.4431667\ttotal: 17.3s\tremaining: 36.5s\n",
      "322:\tlearn: 17.4317345\ttotal: 17.4s\tremaining: 36.5s\n",
      "323:\tlearn: 17.4235622\ttotal: 17.4s\tremaining: 36.4s\n",
      "324:\tlearn: 17.4095816\ttotal: 17.5s\tremaining: 36.3s\n",
      "325:\tlearn: 17.4056857\ttotal: 17.5s\tremaining: 36.3s\n",
      "326:\tlearn: 17.3994160\ttotal: 17.6s\tremaining: 36.2s\n",
      "327:\tlearn: 17.3911537\ttotal: 17.6s\tremaining: 36.1s\n",
      "328:\tlearn: 17.3811343\ttotal: 17.7s\tremaining: 36s\n",
      "329:\tlearn: 17.3726320\ttotal: 17.7s\tremaining: 36s\n",
      "330:\tlearn: 17.3655040\ttotal: 17.8s\tremaining: 35.9s\n",
      "331:\tlearn: 17.3617081\ttotal: 17.8s\tremaining: 35.8s\n",
      "332:\tlearn: 17.3537714\ttotal: 17.9s\tremaining: 35.8s\n",
      "333:\tlearn: 17.3445113\ttotal: 17.9s\tremaining: 35.7s\n",
      "334:\tlearn: 17.3357178\ttotal: 18s\tremaining: 35.7s\n",
      "335:\tlearn: 17.3289762\ttotal: 18s\tremaining: 35.6s\n",
      "336:\tlearn: 17.3193955\ttotal: 18.1s\tremaining: 35.5s\n",
      "337:\tlearn: 17.3104581\ttotal: 18.1s\tremaining: 35.5s\n",
      "338:\tlearn: 17.3037947\ttotal: 18.2s\tremaining: 35.4s\n",
      "339:\tlearn: 17.2980972\ttotal: 18.2s\tremaining: 35.4s\n",
      "340:\tlearn: 17.2897884\ttotal: 18.3s\tremaining: 35.3s\n",
      "341:\tlearn: 17.2789462\ttotal: 18.3s\tremaining: 35.2s\n",
      "342:\tlearn: 17.2761933\ttotal: 18.4s\tremaining: 35.2s\n",
      "343:\tlearn: 17.2661967\ttotal: 18.4s\tremaining: 35.1s\n",
      "344:\tlearn: 17.2611852\ttotal: 18.5s\tremaining: 35.1s\n",
      "345:\tlearn: 17.2539483\ttotal: 18.5s\tremaining: 35s\n",
      "346:\tlearn: 17.2441320\ttotal: 18.6s\tremaining: 34.9s\n",
      "347:\tlearn: 17.2364706\ttotal: 18.6s\tremaining: 34.9s\n",
      "348:\tlearn: 17.2282049\ttotal: 18.7s\tremaining: 34.8s\n",
      "349:\tlearn: 17.2219798\ttotal: 18.7s\tremaining: 34.7s\n",
      "350:\tlearn: 17.2180503\ttotal: 18.7s\tremaining: 34.7s\n",
      "351:\tlearn: 17.2105475\ttotal: 18.8s\tremaining: 34.6s\n",
      "352:\tlearn: 17.2043601\ttotal: 18.9s\tremaining: 34.5s\n",
      "353:\tlearn: 17.1933892\ttotal: 18.9s\tremaining: 34.5s\n",
      "354:\tlearn: 17.1837579\ttotal: 19s\tremaining: 34.4s\n",
      "355:\tlearn: 17.1808344\ttotal: 19s\tremaining: 34.4s\n",
      "356:\tlearn: 17.1742024\ttotal: 19.1s\tremaining: 34.3s\n",
      "357:\tlearn: 17.1671590\ttotal: 19.1s\tremaining: 34.3s\n",
      "358:\tlearn: 17.1573445\ttotal: 19.2s\tremaining: 34.3s\n",
      "359:\tlearn: 17.1494301\ttotal: 19.2s\tremaining: 34.2s\n",
      "360:\tlearn: 17.1451346\ttotal: 19.3s\tremaining: 34.1s\n",
      "361:\tlearn: 17.1370222\ttotal: 19.3s\tremaining: 34.1s\n",
      "362:\tlearn: 17.1308297\ttotal: 19.4s\tremaining: 34s\n",
      "363:\tlearn: 17.1193537\ttotal: 19.4s\tremaining: 34s\n",
      "364:\tlearn: 17.1133203\ttotal: 19.5s\tremaining: 33.9s\n",
      "365:\tlearn: 17.1034450\ttotal: 19.5s\tremaining: 33.9s\n",
      "366:\tlearn: 17.0943524\ttotal: 19.6s\tremaining: 33.8s\n",
      "367:\tlearn: 17.0856333\ttotal: 19.6s\tremaining: 33.7s\n",
      "368:\tlearn: 17.0793954\ttotal: 19.7s\tremaining: 33.7s\n",
      "369:\tlearn: 17.0716030\ttotal: 19.8s\tremaining: 33.6s\n",
      "370:\tlearn: 17.0600166\ttotal: 19.8s\tremaining: 33.6s\n",
      "371:\tlearn: 17.0527681\ttotal: 19.9s\tremaining: 33.6s\n",
      "372:\tlearn: 17.0453605\ttotal: 19.9s\tremaining: 33.5s\n",
      "373:\tlearn: 17.0401478\ttotal: 20s\tremaining: 33.5s\n",
      "374:\tlearn: 17.0319930\ttotal: 20s\tremaining: 33.4s\n",
      "375:\tlearn: 17.0227145\ttotal: 20.1s\tremaining: 33.3s\n",
      "376:\tlearn: 17.0164515\ttotal: 20.1s\tremaining: 33.3s\n",
      "377:\tlearn: 17.0059379\ttotal: 20.2s\tremaining: 33.2s\n",
      "378:\tlearn: 16.9997561\ttotal: 20.3s\tremaining: 33.2s\n",
      "379:\tlearn: 16.9949038\ttotal: 20.3s\tremaining: 33.1s\n",
      "380:\tlearn: 16.9868368\ttotal: 20.4s\tremaining: 33.1s\n",
      "381:\tlearn: 16.9816644\ttotal: 20.4s\tremaining: 33s\n",
      "382:\tlearn: 16.9741651\ttotal: 20.5s\tremaining: 33s\n",
      "383:\tlearn: 16.9634897\ttotal: 20.5s\tremaining: 32.9s\n",
      "384:\tlearn: 16.9567414\ttotal: 20.6s\tremaining: 32.9s\n",
      "385:\tlearn: 16.9474896\ttotal: 20.6s\tremaining: 32.8s\n",
      "386:\tlearn: 16.9410416\ttotal: 20.7s\tremaining: 32.8s\n",
      "387:\tlearn: 16.9301742\ttotal: 20.7s\tremaining: 32.7s\n",
      "388:\tlearn: 16.9252171\ttotal: 20.8s\tremaining: 32.6s\n",
      "389:\tlearn: 16.9201663\ttotal: 20.8s\tremaining: 32.5s\n",
      "390:\tlearn: 16.9145675\ttotal: 20.8s\tremaining: 32.5s\n",
      "391:\tlearn: 16.9069287\ttotal: 20.9s\tremaining: 32.4s\n",
      "392:\tlearn: 16.9008646\ttotal: 20.9s\tremaining: 32.3s\n",
      "393:\tlearn: 16.8951767\ttotal: 21s\tremaining: 32.3s\n",
      "394:\tlearn: 16.8883495\ttotal: 21s\tremaining: 32.2s\n",
      "395:\tlearn: 16.8812524\ttotal: 21.1s\tremaining: 32.2s\n",
      "396:\tlearn: 16.8751927\ttotal: 21.1s\tremaining: 32.1s\n",
      "397:\tlearn: 16.8650378\ttotal: 21.2s\tremaining: 32s\n",
      "398:\tlearn: 16.8547631\ttotal: 21.2s\tremaining: 32s\n",
      "399:\tlearn: 16.8471820\ttotal: 21.3s\tremaining: 31.9s\n",
      "400:\tlearn: 16.8394500\ttotal: 21.3s\tremaining: 31.8s\n",
      "401:\tlearn: 16.8301874\ttotal: 21.4s\tremaining: 31.8s\n",
      "402:\tlearn: 16.8257746\ttotal: 21.4s\tremaining: 31.7s\n",
      "403:\tlearn: 16.8164614\ttotal: 21.4s\tremaining: 31.6s\n",
      "404:\tlearn: 16.8068022\ttotal: 21.5s\tremaining: 31.6s\n",
      "405:\tlearn: 16.8016642\ttotal: 21.5s\tremaining: 31.5s\n",
      "406:\tlearn: 16.7911991\ttotal: 21.6s\tremaining: 31.5s\n",
      "407:\tlearn: 16.7850953\ttotal: 21.7s\tremaining: 31.4s\n",
      "408:\tlearn: 16.7767266\ttotal: 21.7s\tremaining: 31.4s\n",
      "409:\tlearn: 16.7699034\ttotal: 21.8s\tremaining: 31.3s\n",
      "410:\tlearn: 16.7645003\ttotal: 21.8s\tremaining: 31.3s\n",
      "411:\tlearn: 16.7587910\ttotal: 21.8s\tremaining: 31.2s\n",
      "412:\tlearn: 16.7518975\ttotal: 21.9s\tremaining: 31.1s\n",
      "413:\tlearn: 16.7451102\ttotal: 21.9s\tremaining: 31s\n",
      "414:\tlearn: 16.7363257\ttotal: 22s\tremaining: 31s\n",
      "415:\tlearn: 16.7308433\ttotal: 22s\tremaining: 30.9s\n",
      "416:\tlearn: 16.7254355\ttotal: 22.1s\tremaining: 30.9s\n",
      "417:\tlearn: 16.7203227\ttotal: 22.1s\tremaining: 30.8s\n",
      "418:\tlearn: 16.7159187\ttotal: 22.1s\tremaining: 30.7s\n",
      "419:\tlearn: 16.7072281\ttotal: 22.2s\tremaining: 30.7s\n",
      "420:\tlearn: 16.6999059\ttotal: 22.2s\tremaining: 30.6s\n",
      "421:\tlearn: 16.6932601\ttotal: 22.3s\tremaining: 30.5s\n",
      "422:\tlearn: 16.6889768\ttotal: 22.3s\tremaining: 30.5s\n",
      "423:\tlearn: 16.6809892\ttotal: 22.4s\tremaining: 30.4s\n",
      "424:\tlearn: 16.6763761\ttotal: 22.4s\tremaining: 30.3s\n",
      "425:\tlearn: 16.6718156\ttotal: 22.5s\tremaining: 30.3s\n",
      "426:\tlearn: 16.6675945\ttotal: 22.5s\tremaining: 30.2s\n",
      "427:\tlearn: 16.6607330\ttotal: 22.6s\tremaining: 30.1s\n",
      "428:\tlearn: 16.6559946\ttotal: 22.6s\tremaining: 30.1s\n",
      "429:\tlearn: 16.6507523\ttotal: 22.6s\tremaining: 30s\n",
      "430:\tlearn: 16.6427241\ttotal: 22.7s\tremaining: 30s\n",
      "431:\tlearn: 16.6324865\ttotal: 22.8s\tremaining: 29.9s\n",
      "432:\tlearn: 16.6282859\ttotal: 22.8s\tremaining: 29.9s\n",
      "433:\tlearn: 16.6203565\ttotal: 22.8s\tremaining: 29.8s\n",
      "434:\tlearn: 16.6121062\ttotal: 22.9s\tremaining: 29.7s\n",
      "435:\tlearn: 16.6074218\ttotal: 22.9s\tremaining: 29.7s\n",
      "436:\tlearn: 16.6026847\ttotal: 23s\tremaining: 29.6s\n",
      "437:\tlearn: 16.5940950\ttotal: 23s\tremaining: 29.5s\n",
      "438:\tlearn: 16.5866756\ttotal: 23.1s\tremaining: 29.5s\n",
      "439:\tlearn: 16.5792077\ttotal: 23.1s\tremaining: 29.4s\n",
      "440:\tlearn: 16.5706987\ttotal: 23.2s\tremaining: 29.4s\n",
      "441:\tlearn: 16.5646796\ttotal: 23.2s\tremaining: 29.3s\n",
      "442:\tlearn: 16.5621081\ttotal: 23.3s\tremaining: 29.2s\n",
      "443:\tlearn: 16.5556850\ttotal: 23.3s\tremaining: 29.2s\n",
      "444:\tlearn: 16.5479103\ttotal: 23.3s\tremaining: 29.1s\n",
      "445:\tlearn: 16.5384860\ttotal: 23.4s\tremaining: 29.1s\n",
      "446:\tlearn: 16.5329160\ttotal: 23.4s\tremaining: 29s\n",
      "447:\tlearn: 16.5271199\ttotal: 23.5s\tremaining: 28.9s\n",
      "448:\tlearn: 16.5219774\ttotal: 23.5s\tremaining: 28.9s\n",
      "449:\tlearn: 16.5155677\ttotal: 23.6s\tremaining: 28.8s\n",
      "450:\tlearn: 16.5093657\ttotal: 23.6s\tremaining: 28.8s\n",
      "451:\tlearn: 16.5045917\ttotal: 23.7s\tremaining: 28.7s\n",
      "452:\tlearn: 16.4977119\ttotal: 23.7s\tremaining: 28.7s\n",
      "453:\tlearn: 16.4903894\ttotal: 23.8s\tremaining: 28.6s\n",
      "454:\tlearn: 16.4838599\ttotal: 23.8s\tremaining: 28.5s\n",
      "455:\tlearn: 16.4788769\ttotal: 23.9s\tremaining: 28.5s\n",
      "456:\tlearn: 16.4712484\ttotal: 23.9s\tremaining: 28.4s\n",
      "457:\tlearn: 16.4660372\ttotal: 24s\tremaining: 28.4s\n",
      "458:\tlearn: 16.4607064\ttotal: 24s\tremaining: 28.3s\n",
      "459:\tlearn: 16.4529606\ttotal: 24.1s\tremaining: 28.2s\n",
      "460:\tlearn: 16.4472755\ttotal: 24.1s\tremaining: 28.2s\n",
      "461:\tlearn: 16.4430392\ttotal: 24.2s\tremaining: 28.1s\n",
      "462:\tlearn: 16.4370623\ttotal: 24.2s\tremaining: 28.1s\n",
      "463:\tlearn: 16.4323912\ttotal: 24.3s\tremaining: 28s\n",
      "464:\tlearn: 16.4234825\ttotal: 24.3s\tremaining: 28s\n",
      "465:\tlearn: 16.4196337\ttotal: 24.4s\tremaining: 27.9s\n",
      "466:\tlearn: 16.4155045\ttotal: 24.4s\tremaining: 27.9s\n",
      "467:\tlearn: 16.4117012\ttotal: 24.5s\tremaining: 27.8s\n",
      "468:\tlearn: 16.4052307\ttotal: 24.5s\tremaining: 27.8s\n",
      "469:\tlearn: 16.4017452\ttotal: 24.6s\tremaining: 27.7s\n",
      "470:\tlearn: 16.3981813\ttotal: 24.6s\tremaining: 27.7s\n",
      "471:\tlearn: 16.3926458\ttotal: 24.7s\tremaining: 27.6s\n",
      "472:\tlearn: 16.3852495\ttotal: 24.7s\tremaining: 27.6s\n",
      "473:\tlearn: 16.3823645\ttotal: 24.8s\tremaining: 27.5s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "474:\tlearn: 16.3763155\ttotal: 24.9s\tremaining: 27.5s\n",
      "475:\tlearn: 16.3705865\ttotal: 24.9s\tremaining: 27.4s\n",
      "476:\tlearn: 16.3632609\ttotal: 25s\tremaining: 27.4s\n",
      "477:\tlearn: 16.3574156\ttotal: 25s\tremaining: 27.3s\n",
      "478:\tlearn: 16.3531923\ttotal: 25.1s\tremaining: 27.3s\n",
      "479:\tlearn: 16.3467075\ttotal: 25.1s\tremaining: 27.2s\n",
      "480:\tlearn: 16.3402398\ttotal: 25.2s\tremaining: 27.2s\n",
      "481:\tlearn: 16.3318232\ttotal: 25.2s\tremaining: 27.1s\n",
      "482:\tlearn: 16.3286994\ttotal: 25.2s\tremaining: 27s\n",
      "483:\tlearn: 16.3236961\ttotal: 25.3s\tremaining: 27s\n",
      "484:\tlearn: 16.3176627\ttotal: 25.3s\tremaining: 26.9s\n",
      "485:\tlearn: 16.3120633\ttotal: 25.4s\tremaining: 26.8s\n",
      "486:\tlearn: 16.3045457\ttotal: 25.4s\tremaining: 26.8s\n",
      "487:\tlearn: 16.2971719\ttotal: 25.5s\tremaining: 26.7s\n",
      "488:\tlearn: 16.2929852\ttotal: 25.5s\tremaining: 26.7s\n",
      "489:\tlearn: 16.2886480\ttotal: 25.6s\tremaining: 26.6s\n",
      "490:\tlearn: 16.2835179\ttotal: 25.6s\tremaining: 26.6s\n",
      "491:\tlearn: 16.2769440\ttotal: 25.7s\tremaining: 26.5s\n",
      "492:\tlearn: 16.2700022\ttotal: 25.7s\tremaining: 26.5s\n",
      "493:\tlearn: 16.2635390\ttotal: 25.8s\tremaining: 26.4s\n",
      "494:\tlearn: 16.2583294\ttotal: 25.8s\tremaining: 26.4s\n",
      "495:\tlearn: 16.2555676\ttotal: 25.9s\tremaining: 26.3s\n",
      "496:\tlearn: 16.2483793\ttotal: 25.9s\tremaining: 26.3s\n",
      "497:\tlearn: 16.2389151\ttotal: 26s\tremaining: 26.2s\n",
      "498:\tlearn: 16.2332354\ttotal: 26s\tremaining: 26.2s\n",
      "499:\tlearn: 16.2276433\ttotal: 26.1s\tremaining: 26.1s\n",
      "500:\tlearn: 16.2232802\ttotal: 26.1s\tremaining: 26s\n",
      "501:\tlearn: 16.2188186\ttotal: 26.2s\tremaining: 26s\n",
      "502:\tlearn: 16.2118337\ttotal: 26.2s\tremaining: 25.9s\n",
      "503:\tlearn: 16.2039795\ttotal: 26.3s\tremaining: 25.9s\n",
      "504:\tlearn: 16.1961532\ttotal: 26.4s\tremaining: 25.8s\n",
      "505:\tlearn: 16.1917935\ttotal: 26.4s\tremaining: 25.8s\n",
      "506:\tlearn: 16.1866193\ttotal: 26.4s\tremaining: 25.7s\n",
      "507:\tlearn: 16.1832075\ttotal: 26.5s\tremaining: 25.6s\n",
      "508:\tlearn: 16.1795076\ttotal: 26.5s\tremaining: 25.6s\n",
      "509:\tlearn: 16.1753380\ttotal: 26.6s\tremaining: 25.5s\n",
      "510:\tlearn: 16.1661517\ttotal: 26.6s\tremaining: 25.5s\n",
      "511:\tlearn: 16.1591347\ttotal: 26.7s\tremaining: 25.4s\n",
      "512:\tlearn: 16.1528524\ttotal: 26.8s\tremaining: 25.4s\n",
      "513:\tlearn: 16.1468269\ttotal: 26.8s\tremaining: 25.4s\n",
      "514:\tlearn: 16.1414449\ttotal: 26.9s\tremaining: 25.3s\n",
      "515:\tlearn: 16.1339146\ttotal: 26.9s\tremaining: 25.3s\n",
      "516:\tlearn: 16.1278223\ttotal: 27s\tremaining: 25.2s\n",
      "517:\tlearn: 16.1217387\ttotal: 27s\tremaining: 25.1s\n",
      "518:\tlearn: 16.1168248\ttotal: 27.1s\tremaining: 25.1s\n",
      "519:\tlearn: 16.1128795\ttotal: 27.1s\tremaining: 25s\n",
      "520:\tlearn: 16.1042161\ttotal: 27.2s\tremaining: 25s\n",
      "521:\tlearn: 16.0987324\ttotal: 27.2s\tremaining: 24.9s\n",
      "522:\tlearn: 16.0947192\ttotal: 27.3s\tremaining: 24.9s\n",
      "523:\tlearn: 16.0914271\ttotal: 27.3s\tremaining: 24.8s\n",
      "524:\tlearn: 16.0840298\ttotal: 27.4s\tremaining: 24.8s\n",
      "525:\tlearn: 16.0776493\ttotal: 27.4s\tremaining: 24.7s\n",
      "526:\tlearn: 16.0711514\ttotal: 27.5s\tremaining: 24.7s\n",
      "527:\tlearn: 16.0651173\ttotal: 27.5s\tremaining: 24.6s\n",
      "528:\tlearn: 16.0601911\ttotal: 27.6s\tremaining: 24.6s\n",
      "529:\tlearn: 16.0539826\ttotal: 27.6s\tremaining: 24.5s\n",
      "530:\tlearn: 16.0494496\ttotal: 27.7s\tremaining: 24.4s\n",
      "531:\tlearn: 16.0431949\ttotal: 27.7s\tremaining: 24.4s\n",
      "532:\tlearn: 16.0362848\ttotal: 27.7s\tremaining: 24.3s\n",
      "533:\tlearn: 16.0298472\ttotal: 27.8s\tremaining: 24.3s\n",
      "534:\tlearn: 16.0216044\ttotal: 27.8s\tremaining: 24.2s\n",
      "535:\tlearn: 16.0154774\ttotal: 27.9s\tremaining: 24.2s\n",
      "536:\tlearn: 16.0095985\ttotal: 27.9s\tremaining: 24.1s\n",
      "537:\tlearn: 16.0031012\ttotal: 28s\tremaining: 24s\n",
      "538:\tlearn: 15.9984010\ttotal: 28s\tremaining: 24s\n",
      "539:\tlearn: 15.9941458\ttotal: 28.1s\tremaining: 23.9s\n",
      "540:\tlearn: 15.9891424\ttotal: 28.2s\tremaining: 23.9s\n",
      "541:\tlearn: 15.9832200\ttotal: 28.2s\tremaining: 23.8s\n",
      "542:\tlearn: 15.9778238\ttotal: 28.3s\tremaining: 23.8s\n",
      "543:\tlearn: 15.9725403\ttotal: 28.3s\tremaining: 23.7s\n",
      "544:\tlearn: 15.9687802\ttotal: 28.4s\tremaining: 23.7s\n",
      "545:\tlearn: 15.9634576\ttotal: 28.4s\tremaining: 23.6s\n",
      "546:\tlearn: 15.9587804\ttotal: 28.4s\tremaining: 23.6s\n",
      "547:\tlearn: 15.9532088\ttotal: 28.5s\tremaining: 23.5s\n",
      "548:\tlearn: 15.9485774\ttotal: 28.5s\tremaining: 23.4s\n",
      "549:\tlearn: 15.9408198\ttotal: 28.6s\tremaining: 23.4s\n",
      "550:\tlearn: 15.9354518\ttotal: 28.6s\tremaining: 23.3s\n",
      "551:\tlearn: 15.9290391\ttotal: 28.7s\tremaining: 23.3s\n",
      "552:\tlearn: 15.9222851\ttotal: 28.7s\tremaining: 23.2s\n",
      "553:\tlearn: 15.9197879\ttotal: 28.8s\tremaining: 23.2s\n",
      "554:\tlearn: 15.9165091\ttotal: 28.8s\tremaining: 23.1s\n",
      "555:\tlearn: 15.9143693\ttotal: 28.9s\tremaining: 23.1s\n",
      "556:\tlearn: 15.9086225\ttotal: 28.9s\tremaining: 23s\n",
      "557:\tlearn: 15.9016373\ttotal: 29s\tremaining: 23s\n",
      "558:\tlearn: 15.8986358\ttotal: 29s\tremaining: 22.9s\n",
      "559:\tlearn: 15.8927196\ttotal: 29.1s\tremaining: 22.9s\n",
      "560:\tlearn: 15.8892451\ttotal: 29.1s\tremaining: 22.8s\n",
      "561:\tlearn: 15.8847869\ttotal: 29.2s\tremaining: 22.8s\n",
      "562:\tlearn: 15.8808066\ttotal: 29.3s\tremaining: 22.7s\n",
      "563:\tlearn: 15.8765481\ttotal: 29.3s\tremaining: 22.7s\n",
      "564:\tlearn: 15.8721549\ttotal: 29.4s\tremaining: 22.6s\n",
      "565:\tlearn: 15.8672486\ttotal: 29.4s\tremaining: 22.6s\n",
      "566:\tlearn: 15.8596486\ttotal: 29.5s\tremaining: 22.5s\n",
      "567:\tlearn: 15.8539888\ttotal: 29.5s\tremaining: 22.4s\n",
      "568:\tlearn: 15.8496419\ttotal: 29.6s\tremaining: 22.4s\n",
      "569:\tlearn: 15.8430603\ttotal: 29.6s\tremaining: 22.4s\n",
      "570:\tlearn: 15.8350077\ttotal: 29.7s\tremaining: 22.3s\n",
      "571:\tlearn: 15.8279666\ttotal: 29.8s\tremaining: 22.3s\n",
      "572:\tlearn: 15.8219471\ttotal: 29.8s\tremaining: 22.2s\n",
      "573:\tlearn: 15.8182596\ttotal: 29.9s\tremaining: 22.2s\n",
      "574:\tlearn: 15.8150550\ttotal: 29.9s\tremaining: 22.1s\n",
      "575:\tlearn: 15.8100645\ttotal: 30s\tremaining: 22.1s\n",
      "576:\tlearn: 15.8057100\ttotal: 30s\tremaining: 22s\n",
      "577:\tlearn: 15.7985573\ttotal: 30.1s\tremaining: 22s\n",
      "578:\tlearn: 15.7898753\ttotal: 30.1s\tremaining: 21.9s\n",
      "579:\tlearn: 15.7855843\ttotal: 30.2s\tremaining: 21.9s\n",
      "580:\tlearn: 15.7788237\ttotal: 30.2s\tremaining: 21.8s\n",
      "581:\tlearn: 15.7741631\ttotal: 30.3s\tremaining: 21.8s\n",
      "582:\tlearn: 15.7688624\ttotal: 30.3s\tremaining: 21.7s\n",
      "583:\tlearn: 15.7646279\ttotal: 30.4s\tremaining: 21.6s\n",
      "584:\tlearn: 15.7592877\ttotal: 30.4s\tremaining: 21.6s\n",
      "585:\tlearn: 15.7560227\ttotal: 30.5s\tremaining: 21.5s\n",
      "586:\tlearn: 15.7504207\ttotal: 30.5s\tremaining: 21.5s\n",
      "587:\tlearn: 15.7438525\ttotal: 30.6s\tremaining: 21.4s\n",
      "588:\tlearn: 15.7385631\ttotal: 30.6s\tremaining: 21.4s\n",
      "589:\tlearn: 15.7334821\ttotal: 30.7s\tremaining: 21.3s\n",
      "590:\tlearn: 15.7307511\ttotal: 30.7s\tremaining: 21.3s\n",
      "591:\tlearn: 15.7253860\ttotal: 30.8s\tremaining: 21.2s\n",
      "592:\tlearn: 15.7211108\ttotal: 30.8s\tremaining: 21.1s\n",
      "593:\tlearn: 15.7147593\ttotal: 30.8s\tremaining: 21.1s\n",
      "594:\tlearn: 15.7070888\ttotal: 30.9s\tremaining: 21s\n",
      "595:\tlearn: 15.7021131\ttotal: 30.9s\tremaining: 21s\n",
      "596:\tlearn: 15.6960808\ttotal: 31s\tremaining: 20.9s\n",
      "597:\tlearn: 15.6909074\ttotal: 31s\tremaining: 20.9s\n",
      "598:\tlearn: 15.6856209\ttotal: 31.1s\tremaining: 20.8s\n",
      "599:\tlearn: 15.6800762\ttotal: 31.1s\tremaining: 20.8s\n",
      "600:\tlearn: 15.6761754\ttotal: 31.2s\tremaining: 20.7s\n",
      "601:\tlearn: 15.6698899\ttotal: 31.2s\tremaining: 20.6s\n",
      "602:\tlearn: 15.6642423\ttotal: 31.3s\tremaining: 20.6s\n",
      "603:\tlearn: 15.6585053\ttotal: 31.3s\tremaining: 20.6s\n",
      "604:\tlearn: 15.6559549\ttotal: 31.4s\tremaining: 20.5s\n",
      "605:\tlearn: 15.6483286\ttotal: 31.5s\tremaining: 20.5s\n",
      "606:\tlearn: 15.6381632\ttotal: 31.5s\tremaining: 20.4s\n",
      "607:\tlearn: 15.6289423\ttotal: 31.6s\tremaining: 20.4s\n",
      "608:\tlearn: 15.6232548\ttotal: 31.6s\tremaining: 20.3s\n",
      "609:\tlearn: 15.6186513\ttotal: 31.6s\tremaining: 20.2s\n",
      "610:\tlearn: 15.6130456\ttotal: 31.7s\tremaining: 20.2s\n",
      "611:\tlearn: 15.6073098\ttotal: 31.7s\tremaining: 20.1s\n",
      "612:\tlearn: 15.6039829\ttotal: 31.8s\tremaining: 20.1s\n",
      "613:\tlearn: 15.6002282\ttotal: 31.8s\tremaining: 20s\n",
      "614:\tlearn: 15.5964345\ttotal: 31.9s\tremaining: 20s\n",
      "615:\tlearn: 15.5915462\ttotal: 31.9s\tremaining: 19.9s\n",
      "616:\tlearn: 15.5840792\ttotal: 32s\tremaining: 19.9s\n",
      "617:\tlearn: 15.5810642\ttotal: 32s\tremaining: 19.8s\n",
      "618:\tlearn: 15.5753695\ttotal: 32.1s\tremaining: 19.7s\n",
      "619:\tlearn: 15.5709720\ttotal: 32.1s\tremaining: 19.7s\n",
      "620:\tlearn: 15.5664765\ttotal: 32.2s\tremaining: 19.7s\n",
      "621:\tlearn: 15.5619965\ttotal: 32.3s\tremaining: 19.6s\n",
      "622:\tlearn: 15.5565692\ttotal: 32.3s\tremaining: 19.6s\n",
      "623:\tlearn: 15.5512332\ttotal: 32.4s\tremaining: 19.5s\n",
      "624:\tlearn: 15.5434369\ttotal: 32.4s\tremaining: 19.4s\n",
      "625:\tlearn: 15.5368637\ttotal: 32.5s\tremaining: 19.4s\n",
      "626:\tlearn: 15.5319610\ttotal: 32.5s\tremaining: 19.4s\n",
      "627:\tlearn: 15.5277376\ttotal: 32.6s\tremaining: 19.3s\n",
      "628:\tlearn: 15.5219785\ttotal: 32.6s\tremaining: 19.3s\n",
      "629:\tlearn: 15.5169166\ttotal: 32.7s\tremaining: 19.2s\n",
      "630:\tlearn: 15.5106143\ttotal: 32.8s\tremaining: 19.2s\n",
      "631:\tlearn: 15.5062208\ttotal: 32.8s\tremaining: 19.1s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "632:\tlearn: 15.4999295\ttotal: 32.8s\tremaining: 19s\n",
      "633:\tlearn: 15.4946714\ttotal: 32.9s\tremaining: 19s\n",
      "634:\tlearn: 15.4908658\ttotal: 32.9s\tremaining: 18.9s\n",
      "635:\tlearn: 15.4859329\ttotal: 33s\tremaining: 18.9s\n",
      "636:\tlearn: 15.4765818\ttotal: 33s\tremaining: 18.8s\n",
      "637:\tlearn: 15.4718821\ttotal: 33.1s\tremaining: 18.8s\n",
      "638:\tlearn: 15.4652837\ttotal: 33.1s\tremaining: 18.7s\n",
      "639:\tlearn: 15.4619467\ttotal: 33.2s\tremaining: 18.7s\n",
      "640:\tlearn: 15.4581653\ttotal: 33.2s\tremaining: 18.6s\n",
      "641:\tlearn: 15.4540421\ttotal: 33.3s\tremaining: 18.6s\n",
      "642:\tlearn: 15.4489544\ttotal: 33.3s\tremaining: 18.5s\n",
      "643:\tlearn: 15.4432345\ttotal: 33.4s\tremaining: 18.5s\n",
      "644:\tlearn: 15.4364928\ttotal: 33.5s\tremaining: 18.4s\n",
      "645:\tlearn: 15.4314343\ttotal: 33.5s\tremaining: 18.4s\n",
      "646:\tlearn: 15.4272680\ttotal: 33.6s\tremaining: 18.3s\n",
      "647:\tlearn: 15.4218038\ttotal: 33.6s\tremaining: 18.3s\n",
      "648:\tlearn: 15.4172221\ttotal: 33.7s\tremaining: 18.2s\n",
      "649:\tlearn: 15.4142860\ttotal: 33.7s\tremaining: 18.1s\n",
      "650:\tlearn: 15.4075035\ttotal: 33.8s\tremaining: 18.1s\n",
      "651:\tlearn: 15.4043346\ttotal: 33.8s\tremaining: 18.1s\n",
      "652:\tlearn: 15.4005668\ttotal: 33.9s\tremaining: 18s\n",
      "653:\tlearn: 15.3945576\ttotal: 33.9s\tremaining: 17.9s\n",
      "654:\tlearn: 15.3876878\ttotal: 34s\tremaining: 17.9s\n",
      "655:\tlearn: 15.3834756\ttotal: 34s\tremaining: 17.8s\n",
      "656:\tlearn: 15.3769924\ttotal: 34.1s\tremaining: 17.8s\n",
      "657:\tlearn: 15.3717386\ttotal: 34.1s\tremaining: 17.7s\n",
      "658:\tlearn: 15.3676022\ttotal: 34.2s\tremaining: 17.7s\n",
      "659:\tlearn: 15.3630469\ttotal: 34.2s\tremaining: 17.6s\n",
      "660:\tlearn: 15.3556063\ttotal: 34.3s\tremaining: 17.6s\n",
      "661:\tlearn: 15.3519012\ttotal: 34.4s\tremaining: 17.5s\n",
      "662:\tlearn: 15.3469932\ttotal: 34.4s\tremaining: 17.5s\n",
      "663:\tlearn: 15.3407651\ttotal: 34.5s\tremaining: 17.4s\n",
      "664:\tlearn: 15.3362791\ttotal: 34.5s\tremaining: 17.4s\n",
      "665:\tlearn: 15.3310509\ttotal: 34.6s\tremaining: 17.3s\n",
      "666:\tlearn: 15.3247027\ttotal: 34.6s\tremaining: 17.3s\n",
      "667:\tlearn: 15.3195593\ttotal: 34.7s\tremaining: 17.2s\n",
      "668:\tlearn: 15.3149138\ttotal: 34.7s\tremaining: 17.2s\n",
      "669:\tlearn: 15.3088288\ttotal: 34.8s\tremaining: 17.1s\n",
      "670:\tlearn: 15.3013515\ttotal: 34.9s\tremaining: 17.1s\n",
      "671:\tlearn: 15.2955388\ttotal: 34.9s\tremaining: 17s\n",
      "672:\tlearn: 15.2937813\ttotal: 35s\tremaining: 17s\n",
      "673:\tlearn: 15.2886309\ttotal: 35s\tremaining: 17s\n",
      "674:\tlearn: 15.2836258\ttotal: 35.1s\tremaining: 16.9s\n",
      "675:\tlearn: 15.2759236\ttotal: 35.2s\tremaining: 16.9s\n",
      "676:\tlearn: 15.2688975\ttotal: 35.2s\tremaining: 16.8s\n",
      "677:\tlearn: 15.2635716\ttotal: 35.3s\tremaining: 16.8s\n",
      "678:\tlearn: 15.2596636\ttotal: 35.3s\tremaining: 16.7s\n",
      "679:\tlearn: 15.2545142\ttotal: 35.4s\tremaining: 16.7s\n",
      "680:\tlearn: 15.2506237\ttotal: 35.4s\tremaining: 16.6s\n",
      "681:\tlearn: 15.2474492\ttotal: 35.5s\tremaining: 16.6s\n",
      "682:\tlearn: 15.2444562\ttotal: 35.6s\tremaining: 16.5s\n",
      "683:\tlearn: 15.2422674\ttotal: 35.6s\tremaining: 16.5s\n",
      "684:\tlearn: 15.2382237\ttotal: 35.7s\tremaining: 16.4s\n",
      "685:\tlearn: 15.2338767\ttotal: 35.7s\tremaining: 16.4s\n",
      "686:\tlearn: 15.2266388\ttotal: 35.8s\tremaining: 16.3s\n",
      "687:\tlearn: 15.2217071\ttotal: 35.8s\tremaining: 16.3s\n",
      "688:\tlearn: 15.2169985\ttotal: 35.9s\tremaining: 16.2s\n",
      "689:\tlearn: 15.2118857\ttotal: 35.9s\tremaining: 16.1s\n",
      "690:\tlearn: 15.2069513\ttotal: 36s\tremaining: 16.1s\n",
      "691:\tlearn: 15.2010537\ttotal: 36s\tremaining: 16s\n",
      "692:\tlearn: 15.1973518\ttotal: 36.1s\tremaining: 16s\n",
      "693:\tlearn: 15.1922914\ttotal: 36.1s\tremaining: 15.9s\n",
      "694:\tlearn: 15.1853617\ttotal: 36.2s\tremaining: 15.9s\n",
      "695:\tlearn: 15.1811410\ttotal: 36.3s\tremaining: 15.8s\n",
      "696:\tlearn: 15.1767867\ttotal: 36.3s\tremaining: 15.8s\n",
      "697:\tlearn: 15.1729902\ttotal: 36.4s\tremaining: 15.7s\n",
      "698:\tlearn: 15.1665283\ttotal: 36.4s\tremaining: 15.7s\n",
      "699:\tlearn: 15.1617871\ttotal: 36.5s\tremaining: 15.6s\n",
      "700:\tlearn: 15.1562157\ttotal: 36.6s\tremaining: 15.6s\n",
      "701:\tlearn: 15.1527807\ttotal: 36.6s\tremaining: 15.6s\n",
      "702:\tlearn: 15.1457553\ttotal: 36.7s\tremaining: 15.5s\n",
      "703:\tlearn: 15.1416928\ttotal: 36.7s\tremaining: 15.4s\n",
      "704:\tlearn: 15.1352018\ttotal: 36.8s\tremaining: 15.4s\n",
      "705:\tlearn: 15.1308873\ttotal: 36.9s\tremaining: 15.4s\n",
      "706:\tlearn: 15.1244671\ttotal: 36.9s\tremaining: 15.3s\n",
      "707:\tlearn: 15.1183223\ttotal: 37s\tremaining: 15.3s\n",
      "708:\tlearn: 15.1129265\ttotal: 37s\tremaining: 15.2s\n",
      "709:\tlearn: 15.1071179\ttotal: 37.1s\tremaining: 15.2s\n",
      "710:\tlearn: 15.1000998\ttotal: 37.2s\tremaining: 15.1s\n",
      "711:\tlearn: 15.0960474\ttotal: 37.2s\tremaining: 15.1s\n",
      "712:\tlearn: 15.0906938\ttotal: 37.3s\tremaining: 15s\n",
      "713:\tlearn: 15.0863645\ttotal: 37.3s\tremaining: 15s\n",
      "714:\tlearn: 15.0797192\ttotal: 37.4s\tremaining: 14.9s\n",
      "715:\tlearn: 15.0741708\ttotal: 37.5s\tremaining: 14.9s\n",
      "716:\tlearn: 15.0692131\ttotal: 37.5s\tremaining: 14.8s\n",
      "717:\tlearn: 15.0668591\ttotal: 37.6s\tremaining: 14.8s\n",
      "718:\tlearn: 15.0628377\ttotal: 37.6s\tremaining: 14.7s\n",
      "719:\tlearn: 15.0586680\ttotal: 37.7s\tremaining: 14.7s\n",
      "720:\tlearn: 15.0533746\ttotal: 37.7s\tremaining: 14.6s\n",
      "721:\tlearn: 15.0510300\ttotal: 37.8s\tremaining: 14.6s\n",
      "722:\tlearn: 15.0447438\ttotal: 37.8s\tremaining: 14.5s\n",
      "723:\tlearn: 15.0409923\ttotal: 37.9s\tremaining: 14.4s\n",
      "724:\tlearn: 15.0363012\ttotal: 38s\tremaining: 14.4s\n",
      "725:\tlearn: 15.0332066\ttotal: 38s\tremaining: 14.3s\n",
      "726:\tlearn: 15.0257149\ttotal: 38.1s\tremaining: 14.3s\n",
      "727:\tlearn: 15.0208822\ttotal: 38.1s\tremaining: 14.2s\n",
      "728:\tlearn: 15.0161620\ttotal: 38.2s\tremaining: 14.2s\n",
      "729:\tlearn: 15.0119501\ttotal: 38.2s\tremaining: 14.1s\n",
      "730:\tlearn: 15.0048478\ttotal: 38.3s\tremaining: 14.1s\n",
      "731:\tlearn: 15.0020240\ttotal: 38.4s\tremaining: 14.1s\n",
      "732:\tlearn: 14.9992455\ttotal: 38.4s\tremaining: 14s\n",
      "733:\tlearn: 14.9948034\ttotal: 38.5s\tremaining: 13.9s\n",
      "734:\tlearn: 14.9878489\ttotal: 38.6s\tremaining: 13.9s\n",
      "735:\tlearn: 14.9819997\ttotal: 38.6s\tremaining: 13.9s\n",
      "736:\tlearn: 14.9769297\ttotal: 38.7s\tremaining: 13.8s\n",
      "737:\tlearn: 14.9721754\ttotal: 38.7s\tremaining: 13.8s\n",
      "738:\tlearn: 14.9676389\ttotal: 38.8s\tremaining: 13.7s\n",
      "739:\tlearn: 14.9624079\ttotal: 38.8s\tremaining: 13.6s\n",
      "740:\tlearn: 14.9579374\ttotal: 38.9s\tremaining: 13.6s\n",
      "741:\tlearn: 14.9548440\ttotal: 38.9s\tremaining: 13.5s\n",
      "742:\tlearn: 14.9488620\ttotal: 39s\tremaining: 13.5s\n",
      "743:\tlearn: 14.9435067\ttotal: 39s\tremaining: 13.4s\n",
      "744:\tlearn: 14.9399912\ttotal: 39.1s\tremaining: 13.4s\n",
      "745:\tlearn: 14.9332992\ttotal: 39.1s\tremaining: 13.3s\n",
      "746:\tlearn: 14.9312300\ttotal: 39.2s\tremaining: 13.3s\n",
      "747:\tlearn: 14.9269193\ttotal: 39.2s\tremaining: 13.2s\n",
      "748:\tlearn: 14.9199391\ttotal: 39.3s\tremaining: 13.2s\n",
      "749:\tlearn: 14.9149038\ttotal: 39.3s\tremaining: 13.1s\n",
      "750:\tlearn: 14.9125274\ttotal: 39.4s\tremaining: 13.1s\n",
      "751:\tlearn: 14.9086628\ttotal: 39.5s\tremaining: 13s\n",
      "752:\tlearn: 14.9035034\ttotal: 39.5s\tremaining: 13s\n",
      "753:\tlearn: 14.8980735\ttotal: 39.6s\tremaining: 12.9s\n",
      "754:\tlearn: 14.8928616\ttotal: 39.6s\tremaining: 12.9s\n",
      "755:\tlearn: 14.8909222\ttotal: 39.7s\tremaining: 12.8s\n",
      "756:\tlearn: 14.8862492\ttotal: 39.7s\tremaining: 12.8s\n",
      "757:\tlearn: 14.8802563\ttotal: 39.8s\tremaining: 12.7s\n",
      "758:\tlearn: 14.8754564\ttotal: 39.8s\tremaining: 12.6s\n",
      "759:\tlearn: 14.8685384\ttotal: 39.9s\tremaining: 12.6s\n",
      "760:\tlearn: 14.8643792\ttotal: 39.9s\tremaining: 12.5s\n",
      "761:\tlearn: 14.8599365\ttotal: 40s\tremaining: 12.5s\n",
      "762:\tlearn: 14.8548069\ttotal: 40.1s\tremaining: 12.4s\n",
      "763:\tlearn: 14.8485695\ttotal: 40.1s\tremaining: 12.4s\n",
      "764:\tlearn: 14.8453421\ttotal: 40.2s\tremaining: 12.3s\n",
      "765:\tlearn: 14.8420553\ttotal: 40.2s\tremaining: 12.3s\n",
      "766:\tlearn: 14.8382576\ttotal: 40.3s\tremaining: 12.2s\n",
      "767:\tlearn: 14.8342107\ttotal: 40.4s\tremaining: 12.2s\n",
      "768:\tlearn: 14.8287489\ttotal: 40.4s\tremaining: 12.1s\n",
      "769:\tlearn: 14.8251850\ttotal: 40.5s\tremaining: 12.1s\n",
      "770:\tlearn: 14.8191780\ttotal: 40.5s\tremaining: 12s\n",
      "771:\tlearn: 14.8154758\ttotal: 40.6s\tremaining: 12s\n",
      "772:\tlearn: 14.8107033\ttotal: 40.7s\tremaining: 11.9s\n",
      "773:\tlearn: 14.8043259\ttotal: 40.7s\tremaining: 11.9s\n",
      "774:\tlearn: 14.7980187\ttotal: 40.8s\tremaining: 11.8s\n",
      "775:\tlearn: 14.7913822\ttotal: 40.8s\tremaining: 11.8s\n",
      "776:\tlearn: 14.7888817\ttotal: 40.9s\tremaining: 11.7s\n",
      "777:\tlearn: 14.7860683\ttotal: 41s\tremaining: 11.7s\n",
      "778:\tlearn: 14.7807094\ttotal: 41s\tremaining: 11.6s\n",
      "779:\tlearn: 14.7767929\ttotal: 41.1s\tremaining: 11.6s\n",
      "780:\tlearn: 14.7719793\ttotal: 41.1s\tremaining: 11.5s\n",
      "781:\tlearn: 14.7667121\ttotal: 41.2s\tremaining: 11.5s\n",
      "782:\tlearn: 14.7633607\ttotal: 41.2s\tremaining: 11.4s\n",
      "783:\tlearn: 14.7591673\ttotal: 41.3s\tremaining: 11.4s\n",
      "784:\tlearn: 14.7534225\ttotal: 41.4s\tremaining: 11.3s\n",
      "785:\tlearn: 14.7491118\ttotal: 41.4s\tremaining: 11.3s\n",
      "786:\tlearn: 14.7474239\ttotal: 41.5s\tremaining: 11.2s\n",
      "787:\tlearn: 14.7436320\ttotal: 41.5s\tremaining: 11.2s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "788:\tlearn: 14.7419854\ttotal: 41.6s\tremaining: 11.1s\n",
      "789:\tlearn: 14.7361637\ttotal: 41.6s\tremaining: 11.1s\n",
      "790:\tlearn: 14.7296024\ttotal: 41.7s\tremaining: 11s\n",
      "791:\tlearn: 14.7272535\ttotal: 41.8s\tremaining: 11s\n",
      "792:\tlearn: 14.7243375\ttotal: 41.8s\tremaining: 10.9s\n",
      "793:\tlearn: 14.7197277\ttotal: 41.9s\tremaining: 10.9s\n",
      "794:\tlearn: 14.7151669\ttotal: 41.9s\tremaining: 10.8s\n",
      "795:\tlearn: 14.7110691\ttotal: 42s\tremaining: 10.8s\n",
      "796:\tlearn: 14.7051930\ttotal: 42s\tremaining: 10.7s\n",
      "797:\tlearn: 14.7030409\ttotal: 42s\tremaining: 10.6s\n",
      "798:\tlearn: 14.6980542\ttotal: 42.1s\tremaining: 10.6s\n",
      "799:\tlearn: 14.6956004\ttotal: 42.1s\tremaining: 10.5s\n",
      "800:\tlearn: 14.6907744\ttotal: 42.2s\tremaining: 10.5s\n",
      "801:\tlearn: 14.6829462\ttotal: 42.2s\tremaining: 10.4s\n",
      "802:\tlearn: 14.6773749\ttotal: 42.3s\tremaining: 10.4s\n",
      "803:\tlearn: 14.6720113\ttotal: 42.3s\tremaining: 10.3s\n",
      "804:\tlearn: 14.6658307\ttotal: 42.4s\tremaining: 10.3s\n",
      "805:\tlearn: 14.6621186\ttotal: 42.5s\tremaining: 10.2s\n",
      "806:\tlearn: 14.6602440\ttotal: 42.5s\tremaining: 10.2s\n",
      "807:\tlearn: 14.6549711\ttotal: 42.6s\tremaining: 10.1s\n",
      "808:\tlearn: 14.6500759\ttotal: 42.6s\tremaining: 10.1s\n",
      "809:\tlearn: 14.6483383\ttotal: 42.7s\tremaining: 10s\n",
      "810:\tlearn: 14.6439478\ttotal: 42.8s\tremaining: 9.96s\n",
      "811:\tlearn: 14.6403800\ttotal: 42.8s\tremaining: 9.91s\n",
      "812:\tlearn: 14.6357683\ttotal: 42.8s\tremaining: 9.85s\n",
      "813:\tlearn: 14.6299965\ttotal: 42.9s\tremaining: 9.8s\n",
      "814:\tlearn: 14.6268814\ttotal: 42.9s\tremaining: 9.74s\n",
      "815:\tlearn: 14.6220875\ttotal: 43s\tremaining: 9.69s\n",
      "816:\tlearn: 14.6164181\ttotal: 43s\tremaining: 9.64s\n",
      "817:\tlearn: 14.6119080\ttotal: 43.1s\tremaining: 9.59s\n",
      "818:\tlearn: 14.6060589\ttotal: 43.1s\tremaining: 9.53s\n",
      "819:\tlearn: 14.6018087\ttotal: 43.2s\tremaining: 9.48s\n",
      "820:\tlearn: 14.5991342\ttotal: 43.2s\tremaining: 9.43s\n",
      "821:\tlearn: 14.5970497\ttotal: 43.3s\tremaining: 9.37s\n",
      "822:\tlearn: 14.5930109\ttotal: 43.3s\tremaining: 9.32s\n",
      "823:\tlearn: 14.5892025\ttotal: 43.4s\tremaining: 9.26s\n",
      "824:\tlearn: 14.5815775\ttotal: 43.4s\tremaining: 9.21s\n",
      "825:\tlearn: 14.5755744\ttotal: 43.5s\tremaining: 9.16s\n",
      "826:\tlearn: 14.5693845\ttotal: 43.5s\tremaining: 9.11s\n",
      "827:\tlearn: 14.5656105\ttotal: 43.6s\tremaining: 9.05s\n",
      "828:\tlearn: 14.5595627\ttotal: 43.6s\tremaining: 9s\n",
      "829:\tlearn: 14.5540980\ttotal: 43.7s\tremaining: 8.95s\n",
      "830:\tlearn: 14.5497238\ttotal: 43.7s\tremaining: 8.89s\n",
      "831:\tlearn: 14.5442365\ttotal: 43.8s\tremaining: 8.84s\n",
      "832:\tlearn: 14.5386646\ttotal: 43.8s\tremaining: 8.78s\n",
      "833:\tlearn: 14.5367031\ttotal: 43.8s\tremaining: 8.73s\n",
      "834:\tlearn: 14.5338126\ttotal: 43.9s\tremaining: 8.67s\n",
      "835:\tlearn: 14.5273458\ttotal: 43.9s\tremaining: 8.62s\n",
      "836:\tlearn: 14.5242365\ttotal: 44s\tremaining: 8.57s\n",
      "837:\tlearn: 14.5193660\ttotal: 44s\tremaining: 8.51s\n",
      "838:\tlearn: 14.5144094\ttotal: 44.1s\tremaining: 8.46s\n",
      "839:\tlearn: 14.5093610\ttotal: 44.1s\tremaining: 8.41s\n",
      "840:\tlearn: 14.5052631\ttotal: 44.2s\tremaining: 8.36s\n",
      "841:\tlearn: 14.4998962\ttotal: 44.3s\tremaining: 8.31s\n",
      "842:\tlearn: 14.4954758\ttotal: 44.3s\tremaining: 8.25s\n",
      "843:\tlearn: 14.4912715\ttotal: 44.4s\tremaining: 8.2s\n",
      "844:\tlearn: 14.4883934\ttotal: 44.4s\tremaining: 8.15s\n",
      "845:\tlearn: 14.4850871\ttotal: 44.5s\tremaining: 8.1s\n",
      "846:\tlearn: 14.4819624\ttotal: 44.5s\tremaining: 8.04s\n",
      "847:\tlearn: 14.4785066\ttotal: 44.6s\tremaining: 7.99s\n",
      "848:\tlearn: 14.4746102\ttotal: 44.6s\tremaining: 7.94s\n",
      "849:\tlearn: 14.4709229\ttotal: 44.7s\tremaining: 7.88s\n",
      "850:\tlearn: 14.4669280\ttotal: 44.7s\tremaining: 7.83s\n",
      "851:\tlearn: 14.4618825\ttotal: 44.8s\tremaining: 7.78s\n",
      "852:\tlearn: 14.4579355\ttotal: 44.8s\tremaining: 7.72s\n",
      "853:\tlearn: 14.4531610\ttotal: 44.9s\tremaining: 7.67s\n",
      "854:\tlearn: 14.4481376\ttotal: 44.9s\tremaining: 7.62s\n",
      "855:\tlearn: 14.4435551\ttotal: 45s\tremaining: 7.57s\n",
      "856:\tlearn: 14.4382009\ttotal: 45s\tremaining: 7.51s\n",
      "857:\tlearn: 14.4354566\ttotal: 45.1s\tremaining: 7.46s\n",
      "858:\tlearn: 14.4322478\ttotal: 45.2s\tremaining: 7.41s\n",
      "859:\tlearn: 14.4280800\ttotal: 45.2s\tremaining: 7.36s\n",
      "860:\tlearn: 14.4241794\ttotal: 45.2s\tremaining: 7.3s\n",
      "861:\tlearn: 14.4201482\ttotal: 45.3s\tremaining: 7.25s\n",
      "862:\tlearn: 14.4171856\ttotal: 45.3s\tremaining: 7.19s\n",
      "863:\tlearn: 14.4118759\ttotal: 45.4s\tremaining: 7.14s\n",
      "864:\tlearn: 14.4072251\ttotal: 45.4s\tremaining: 7.09s\n",
      "865:\tlearn: 14.4013412\ttotal: 45.4s\tremaining: 7.03s\n",
      "866:\tlearn: 14.3975079\ttotal: 45.5s\tremaining: 6.98s\n",
      "867:\tlearn: 14.3923964\ttotal: 45.6s\tremaining: 6.93s\n",
      "868:\tlearn: 14.3850860\ttotal: 45.6s\tremaining: 6.88s\n",
      "869:\tlearn: 14.3823749\ttotal: 45.7s\tremaining: 6.82s\n",
      "870:\tlearn: 14.3792961\ttotal: 45.7s\tremaining: 6.77s\n",
      "871:\tlearn: 14.3747673\ttotal: 45.8s\tremaining: 6.72s\n",
      "872:\tlearn: 14.3704039\ttotal: 45.8s\tremaining: 6.67s\n",
      "873:\tlearn: 14.3668748\ttotal: 45.9s\tremaining: 6.62s\n",
      "874:\tlearn: 14.3646076\ttotal: 45.9s\tremaining: 6.56s\n",
      "875:\tlearn: 14.3622064\ttotal: 46s\tremaining: 6.51s\n",
      "876:\tlearn: 14.3572861\ttotal: 46.1s\tremaining: 6.46s\n",
      "877:\tlearn: 14.3527446\ttotal: 46.1s\tremaining: 6.41s\n",
      "878:\tlearn: 14.3481771\ttotal: 46.2s\tremaining: 6.36s\n",
      "879:\tlearn: 14.3442304\ttotal: 46.2s\tremaining: 6.3s\n",
      "880:\tlearn: 14.3387412\ttotal: 46.3s\tremaining: 6.25s\n",
      "881:\tlearn: 14.3362359\ttotal: 46.4s\tremaining: 6.2s\n",
      "882:\tlearn: 14.3336903\ttotal: 46.4s\tremaining: 6.15s\n",
      "883:\tlearn: 14.3282393\ttotal: 46.5s\tremaining: 6.1s\n",
      "884:\tlearn: 14.3248151\ttotal: 46.5s\tremaining: 6.04s\n",
      "885:\tlearn: 14.3220715\ttotal: 46.6s\tremaining: 5.99s\n",
      "886:\tlearn: 14.3201664\ttotal: 46.6s\tremaining: 5.94s\n",
      "887:\tlearn: 14.3159795\ttotal: 46.7s\tremaining: 5.89s\n",
      "888:\tlearn: 14.3126885\ttotal: 46.7s\tremaining: 5.83s\n",
      "889:\tlearn: 14.3091111\ttotal: 46.8s\tremaining: 5.78s\n",
      "890:\tlearn: 14.3052223\ttotal: 46.8s\tremaining: 5.73s\n",
      "891:\tlearn: 14.2997243\ttotal: 46.9s\tremaining: 5.68s\n",
      "892:\tlearn: 14.2961998\ttotal: 47s\tremaining: 5.63s\n",
      "893:\tlearn: 14.2909614\ttotal: 47s\tremaining: 5.58s\n",
      "894:\tlearn: 14.2867119\ttotal: 47.1s\tremaining: 5.52s\n",
      "895:\tlearn: 14.2835974\ttotal: 47.1s\tremaining: 5.47s\n",
      "896:\tlearn: 14.2786215\ttotal: 47.2s\tremaining: 5.42s\n",
      "897:\tlearn: 14.2745516\ttotal: 47.3s\tremaining: 5.37s\n",
      "898:\tlearn: 14.2719659\ttotal: 47.3s\tremaining: 5.31s\n",
      "899:\tlearn: 14.2689756\ttotal: 47.4s\tremaining: 5.26s\n",
      "900:\tlearn: 14.2637002\ttotal: 47.4s\tremaining: 5.21s\n",
      "901:\tlearn: 14.2583124\ttotal: 47.5s\tremaining: 5.16s\n",
      "902:\tlearn: 14.2549660\ttotal: 47.5s\tremaining: 5.11s\n",
      "903:\tlearn: 14.2492271\ttotal: 47.6s\tremaining: 5.05s\n",
      "904:\tlearn: 14.2437248\ttotal: 47.6s\tremaining: 5s\n",
      "905:\tlearn: 14.2393849\ttotal: 47.7s\tremaining: 4.95s\n",
      "906:\tlearn: 14.2354112\ttotal: 47.8s\tremaining: 4.9s\n",
      "907:\tlearn: 14.2311559\ttotal: 47.8s\tremaining: 4.84s\n",
      "908:\tlearn: 14.2274990\ttotal: 47.9s\tremaining: 4.79s\n",
      "909:\tlearn: 14.2247651\ttotal: 47.9s\tremaining: 4.74s\n",
      "910:\tlearn: 14.2197113\ttotal: 48s\tremaining: 4.69s\n",
      "911:\tlearn: 14.2160307\ttotal: 48.1s\tremaining: 4.64s\n",
      "912:\tlearn: 14.2113071\ttotal: 48.1s\tremaining: 4.58s\n",
      "913:\tlearn: 14.2080172\ttotal: 48.2s\tremaining: 4.53s\n",
      "914:\tlearn: 14.2025107\ttotal: 48.2s\tremaining: 4.48s\n",
      "915:\tlearn: 14.1963832\ttotal: 48.3s\tremaining: 4.43s\n",
      "916:\tlearn: 14.1922432\ttotal: 48.3s\tremaining: 4.38s\n",
      "917:\tlearn: 14.1882614\ttotal: 48.4s\tremaining: 4.32s\n",
      "918:\tlearn: 14.1840836\ttotal: 48.4s\tremaining: 4.27s\n",
      "919:\tlearn: 14.1800578\ttotal: 48.5s\tremaining: 4.21s\n",
      "920:\tlearn: 14.1769313\ttotal: 48.5s\tremaining: 4.16s\n",
      "921:\tlearn: 14.1724553\ttotal: 48.6s\tremaining: 4.11s\n",
      "922:\tlearn: 14.1702957\ttotal: 48.6s\tremaining: 4.05s\n",
      "923:\tlearn: 14.1651514\ttotal: 48.7s\tremaining: 4s\n",
      "924:\tlearn: 14.1605247\ttotal: 48.7s\tremaining: 3.95s\n",
      "925:\tlearn: 14.1557768\ttotal: 48.7s\tremaining: 3.9s\n",
      "926:\tlearn: 14.1528434\ttotal: 48.8s\tremaining: 3.84s\n",
      "927:\tlearn: 14.1481818\ttotal: 48.8s\tremaining: 3.79s\n",
      "928:\tlearn: 14.1445418\ttotal: 48.9s\tremaining: 3.74s\n",
      "929:\tlearn: 14.1395139\ttotal: 49s\tremaining: 3.69s\n",
      "930:\tlearn: 14.1363626\ttotal: 49s\tremaining: 3.63s\n",
      "931:\tlearn: 14.1329952\ttotal: 49.1s\tremaining: 3.58s\n",
      "932:\tlearn: 14.1294542\ttotal: 49.1s\tremaining: 3.53s\n",
      "933:\tlearn: 14.1248846\ttotal: 49.2s\tremaining: 3.48s\n",
      "934:\tlearn: 14.1216095\ttotal: 49.3s\tremaining: 3.42s\n",
      "935:\tlearn: 14.1170803\ttotal: 49.3s\tremaining: 3.37s\n",
      "936:\tlearn: 14.1128582\ttotal: 49.4s\tremaining: 3.32s\n",
      "937:\tlearn: 14.1097026\ttotal: 49.5s\tremaining: 3.27s\n",
      "938:\tlearn: 14.1084000\ttotal: 49.5s\tremaining: 3.22s\n",
      "939:\tlearn: 14.1038716\ttotal: 49.6s\tremaining: 3.16s\n",
      "940:\tlearn: 14.0996777\ttotal: 49.6s\tremaining: 3.11s\n",
      "941:\tlearn: 14.0971370\ttotal: 49.7s\tremaining: 3.06s\n",
      "942:\tlearn: 14.0914538\ttotal: 49.7s\tremaining: 3.01s\n",
      "943:\tlearn: 14.0864653\ttotal: 49.8s\tremaining: 2.95s\n",
      "944:\tlearn: 14.0827559\ttotal: 49.9s\tremaining: 2.9s\n",
      "945:\tlearn: 14.0788729\ttotal: 49.9s\tremaining: 2.85s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "946:\tlearn: 14.0755123\ttotal: 50s\tremaining: 2.8s\n",
      "947:\tlearn: 14.0720840\ttotal: 50s\tremaining: 2.74s\n",
      "948:\tlearn: 14.0681159\ttotal: 50.1s\tremaining: 2.69s\n",
      "949:\tlearn: 14.0638466\ttotal: 50.1s\tremaining: 2.64s\n",
      "950:\tlearn: 14.0605750\ttotal: 50.1s\tremaining: 2.58s\n",
      "951:\tlearn: 14.0551422\ttotal: 50.2s\tremaining: 2.53s\n",
      "952:\tlearn: 14.0503426\ttotal: 50.2s\tremaining: 2.48s\n",
      "953:\tlearn: 14.0464770\ttotal: 50.3s\tremaining: 2.42s\n",
      "954:\tlearn: 14.0433645\ttotal: 50.3s\tremaining: 2.37s\n",
      "955:\tlearn: 14.0395265\ttotal: 50.4s\tremaining: 2.32s\n",
      "956:\tlearn: 14.0357226\ttotal: 50.4s\tremaining: 2.27s\n",
      "957:\tlearn: 14.0304080\ttotal: 50.5s\tremaining: 2.21s\n",
      "958:\tlearn: 14.0282683\ttotal: 50.6s\tremaining: 2.16s\n",
      "959:\tlearn: 14.0245696\ttotal: 50.6s\tremaining: 2.11s\n",
      "960:\tlearn: 14.0200615\ttotal: 50.7s\tremaining: 2.06s\n",
      "961:\tlearn: 14.0160576\ttotal: 50.7s\tremaining: 2s\n",
      "962:\tlearn: 14.0106856\ttotal: 50.8s\tremaining: 1.95s\n",
      "963:\tlearn: 14.0080384\ttotal: 50.8s\tremaining: 1.9s\n",
      "964:\tlearn: 14.0031975\ttotal: 50.9s\tremaining: 1.84s\n",
      "965:\tlearn: 13.9993206\ttotal: 50.9s\tremaining: 1.79s\n",
      "966:\tlearn: 13.9950910\ttotal: 51s\tremaining: 1.74s\n",
      "967:\tlearn: 13.9905270\ttotal: 51.1s\tremaining: 1.69s\n",
      "968:\tlearn: 13.9876586\ttotal: 51.1s\tremaining: 1.64s\n",
      "969:\tlearn: 13.9824300\ttotal: 51.2s\tremaining: 1.58s\n",
      "970:\tlearn: 13.9785209\ttotal: 51.2s\tremaining: 1.53s\n",
      "971:\tlearn: 13.9748781\ttotal: 51.3s\tremaining: 1.48s\n",
      "972:\tlearn: 13.9706441\ttotal: 51.4s\tremaining: 1.42s\n",
      "973:\tlearn: 13.9673376\ttotal: 51.4s\tremaining: 1.37s\n",
      "974:\tlearn: 13.9638206\ttotal: 51.5s\tremaining: 1.32s\n",
      "975:\tlearn: 13.9594115\ttotal: 51.5s\tremaining: 1.27s\n",
      "976:\tlearn: 13.9540039\ttotal: 51.6s\tremaining: 1.21s\n",
      "977:\tlearn: 13.9495503\ttotal: 51.7s\tremaining: 1.16s\n",
      "978:\tlearn: 13.9465029\ttotal: 51.7s\tremaining: 1.11s\n",
      "979:\tlearn: 13.9405638\ttotal: 51.8s\tremaining: 1.06s\n",
      "980:\tlearn: 13.9356889\ttotal: 51.8s\tremaining: 1s\n",
      "981:\tlearn: 13.9313328\ttotal: 51.9s\tremaining: 951ms\n",
      "982:\tlearn: 13.9277851\ttotal: 51.9s\tremaining: 898ms\n",
      "983:\tlearn: 13.9233700\ttotal: 52s\tremaining: 845ms\n",
      "984:\tlearn: 13.9192154\ttotal: 52s\tremaining: 793ms\n",
      "985:\tlearn: 13.9153087\ttotal: 52.1s\tremaining: 740ms\n",
      "986:\tlearn: 13.9118961\ttotal: 52.2s\tremaining: 687ms\n",
      "987:\tlearn: 13.9089409\ttotal: 52.2s\tremaining: 634ms\n",
      "988:\tlearn: 13.9040834\ttotal: 52.3s\tremaining: 581ms\n",
      "989:\tlearn: 13.8986598\ttotal: 52.3s\tremaining: 529ms\n",
      "990:\tlearn: 13.8936912\ttotal: 52.4s\tremaining: 476ms\n",
      "991:\tlearn: 13.8899476\ttotal: 52.5s\tremaining: 423ms\n",
      "992:\tlearn: 13.8872230\ttotal: 52.5s\tremaining: 370ms\n",
      "993:\tlearn: 13.8833543\ttotal: 52.6s\tremaining: 317ms\n",
      "994:\tlearn: 13.8799269\ttotal: 52.6s\tremaining: 264ms\n",
      "995:\tlearn: 13.8775098\ttotal: 52.7s\tremaining: 212ms\n",
      "996:\tlearn: 13.8737299\ttotal: 52.7s\tremaining: 159ms\n",
      "997:\tlearn: 13.8706310\ttotal: 52.8s\tremaining: 106ms\n",
      "998:\tlearn: 13.8670653\ttotal: 52.9s\tremaining: 52.9ms\n",
      "999:\tlearn: 13.8640947\ttotal: 52.9s\tremaining: 0us\n",
      "\n",
      "    mse -> 274.4539677851845\n",
      "    mae -> 12.809578681478163\n",
      "    rmse -> 16.566652280566057\n",
      "    mape -> 9933608988251572.0\n",
      "    r2 -> 0.4517222555946421\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "est1 = RandomForestRegressor(random_state=42, n_estimators=250, n_jobs=-1)\n",
    "est2 = CatBoostRegressor(iterations=1000,\n",
    "                             learning_rate=0.07,\n",
    "                             depth=12,\n",
    "                             eval_metric='RMSE',\n",
    "                             random_seed = 42,\n",
    "                             bagging_temperature = 0.3)\n",
    "\n",
    "model = VotingRegressor([('est1', est1), ('est2', est2)])\n",
    "model.fit(X_train, y_train)\n",
    "preds = model.predict(X_test)\n",
    "mse = mean_squared_error(y_test, preds)\n",
    "mae = mean_absolute_error(y_test, preds)\n",
    "r2 = r2_score(y_test, preds)\n",
    "rmse = np.sqrt(mse)\n",
    "mape = mean_absolute_percentage_error(y_test, preds)\n",
    "print(\n",
    "    \"\"\"\n",
    "    mse -> {mse}\n",
    "    mae -> {mae}\n",
    "    rmse -> {rmse}\n",
    "    mape -> {mape}\n",
    "    r2 -> {r2}\n",
    "    \"\"\".format(mse=mse, mae=mae, rmse=rmse, mape=mape, r2=r2)\n",
    ")\n",
    "del est1, est2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10781cba",
   "metadata": {},
   "source": [
    "<a id='10'>10 кросс валидация лучшего варианта (random forest)</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "517ff03f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "scoring = {\n",
    "    'mse': 'neg_mean_squared_error',\n",
    "    'mae': 'neg_mean_absolute_error',\n",
    "    'r2': 'r2'\n",
    "}\n",
    "\n",
    "\n",
    "model = RandomForestRegressor(random_state=42, n_estimators=250, n_jobs=-1)\n",
    "scores = cross_validate(model, X, y, cv=5, scoring=scoring)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c214dbc0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Средняя rmse: 18.073452772078983\n",
      "Средняя mse: 326.6496951045695\n",
      "Средняя mae: 13.458432433253924\n",
      "Средняя r2: 0.33985006130141515\n"
     ]
    }
   ],
   "source": [
    "scores\n",
    "print(f\"Средняя rmse: {np.sqrt(abs(scores['test_mse']).mean())}\")\n",
    "print(f\"Средняя mse: {abs(scores['test_mse']).mean()}\")\n",
    "print(f\"Средняя mae: {abs(scores['test_mae']).mean()}\")\n",
    "print(f\"Средняя r2: {(abs(scores['test_r2']).mean())}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08a9dd9f",
   "metadata": {},
   "source": [
    "<a id='11'>11 Вывод</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f058594d",
   "metadata": {},
   "source": [
    "Независимо от подхода, не удалось получить хорошие прогнозы популярности трека. Это свидетельствует о том, что данные вполне близки к положению дел в реальности. В реальной жизни очень трудно предсказать популярность трека, чаще всего она зависит от совершеннго случайных событий, которые не имеют каких либо шаблонов, по которым можно обучить модель машинного обучения"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ca778df",
   "metadata": {},
   "source": [
    "Лучшей моделью на данных данных является случайный лес. "
   ]
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}