HeartDiseasePrediction_Zindi_Hackathon_DSN_2024.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mzimpSm8i3I1",
        "outputId": "585d2c24-92c9-43b8-bb7c-ad84f255e7a5"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "      Id  Age  Sex  cp  trestbps  chol  fbs  restecg  thalach  exang  oldpeak  \\\n",
            "0  16167   33    0   1       158   205    1        0      154      0      1.5   \n",
            "1  11275   53    1   2       198   154    0        1      104      0      0.8   \n",
            "2  13251   37    1   2       101   202    1        0      155      0      2.1   \n",
            "3  19921   75    0   0       113   306    1        2       88      1      4.9   \n",
            "4  11293   35    1   2       139   419    1        1      166      1      0.9   \n",
            "\n",
            "   slope  ca  thal  target  \n",
            "0      1   4     1       1  \n",
            "1      2   1     0       0  \n",
            "2      1   3     1       1  \n",
            "3      0   2     2       1  \n",
            "4      2   4     0       1  \n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 7303 entries, 0 to 7302\n",
            "Data columns (total 15 columns):\n",
            " #   Column    Non-Null Count  Dtype  \n",
            "---  ------    --------------  -----  \n",
            " 0   Id        7303 non-null   int64  \n",
            " 1   Age       7303 non-null   int64  \n",
            " 2   Sex       7303 non-null   int64  \n",
            " 3   cp        7303 non-null   int64  \n",
            " 4   trestbps  7303 non-null   int64  \n",
            " 5   chol      7303 non-null   int64  \n",
            " 6   fbs       7303 non-null   int64  \n",
            " 7   restecg   7303 non-null   int64  \n",
            " 8   thalach   7303 non-null   int64  \n",
            " 9   exang     7303 non-null   int64  \n",
            " 10  oldpeak   7303 non-null   float64\n",
            " 11  slope     7303 non-null   int64  \n",
            " 12  ca        7303 non-null   int64  \n",
            " 13  thal      7303 non-null   int64  \n",
            " 14  target    7303 non-null   int64  \n",
            "dtypes: float64(1), int64(14)\n",
            "memory usage: 855.9 KB\n",
            "None\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "# Load the training data\n",
        "train_data = pd.read_csv('/content/Train Dataset .csv')\n",
        "\n",
        "# Display the first few rows and basic information about the dataset\n",
        "print(train_data.head())\n",
        "print(train_data.info())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "PHYVvqP5DHvG",
        "outputId": "b5ba4909-587b-4308-949e-5fdf6646fccb"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                 Id          Age          Sex           cp     trestbps  \\\n",
            "count   7303.000000  7303.000000  7303.000000  7303.000000  7303.000000   \n",
            "mean   15021.535396    53.172669     0.499658     1.502533   147.447487   \n",
            "std     2886.026080    14.185970     0.500034     1.115594    31.099538   \n",
            "min    10001.000000    29.000000     0.000000     0.000000    94.000000   \n",
            "25%    12521.500000    41.000000     0.000000     1.000000   120.000000   \n",
            "50%    15054.000000    53.000000     0.000000     1.000000   148.000000   \n",
            "75%    17513.500000    65.000000     1.000000     3.000000   174.000000   \n",
            "max    19998.000000    77.000000     1.000000     3.000000   200.000000   \n",
            "\n",
            "              chol          fbs      restecg      thalach        exang  \\\n",
            "count  7303.000000  7303.000000  7303.000000  7303.000000  7303.000000   \n",
            "mean    342.805970     0.493085     1.013008   136.506093     0.503218   \n",
            "std     127.291998     0.499986     0.815806    38.141966     0.500024   \n",
            "min     126.000000     0.000000     0.000000    71.000000     0.000000   \n",
            "25%     231.000000     0.000000     0.000000   104.000000     0.000000   \n",
            "50%     341.000000     0.000000     1.000000   137.000000     1.000000   \n",
            "75%     450.000000     1.000000     2.000000   170.000000     1.000000   \n",
            "max     564.000000     1.000000     2.000000   202.000000     1.000000   \n",
            "\n",
            "           oldpeak        slope           ca         thal       target  \n",
            "count  7303.000000  7303.000000  7303.000000  7303.000000  7303.000000  \n",
            "mean      3.129851     0.991510     2.019033     1.502259     0.813501  \n",
            "std       1.791160     0.817291     1.410546     1.113137     0.389535  \n",
            "min       0.000000     0.000000     0.000000     0.000000     0.000000  \n",
            "25%       1.600000     0.000000     1.000000     1.000000     1.000000  \n",
            "50%       3.100000     1.000000     2.000000     1.000000     1.000000  \n",
            "75%       4.700000     2.000000     3.000000     2.000000     1.000000  \n",
            "max       6.200000     2.000000     4.000000     3.000000     1.000000  \n",
            "Id          0\n",
            "Age         0\n",
            "Sex         0\n",
            "cp          0\n",
            "trestbps    0\n",
            "chol        0\n",
            "fbs         0\n",
            "restecg     0\n",
            "thalach     0\n",
            "exang       0\n",
            "oldpeak     0\n",
            "slope       0\n",
            "ca          0\n",
            "thal        0\n",
            "target      0\n",
            "dtype: int64\n"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1200x1000 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# Basic statistical summary of the numerical columns\n",
        "print(train_data.describe())\n",
        "\n",
        "# Check for missing values\n",
        "print(train_data.isnull().sum())\n",
        "\n",
        "# Distribution of the target variable\n",
        "plt.figure(figsize=(8, 6))\n",
        "train_data['target'].value_counts().plot(kind='bar')\n",
        "plt.title('Distribution of Heart Disease')\n",
        "plt.xlabel('Heart Disease')\n",
        "plt.ylabel('Count')\n",
        "plt.xticks([0, 1], ['No Heart Disease', 'Heart Disease'])\n",
        "plt.show()\n",
        "\n",
        "# Correlation heatmap\n",
        "plt.figure(figsize=(12, 10))\n",
        "sns.heatmap(train_data.corr(), annot=True, cmap='coolwarm', linewidths=0.5)\n",
        "plt.title('Correlation Heatmap of Features')\n",
        "plt.show()\n",
        "\n",
        "# Age distribution\n",
        "plt.figure(figsize=(10, 6))\n",
        "sns.histplot(data=train_data, x='Age', hue='target', kde=True, multiple='stack')\n",
        "plt.title('Age Distribution by Heart Disease')\n",
        "plt.show()\n",
        "\n",
        "# Boxplot of cholesterol levels by heart disease\n",
        "plt.figure(figsize=(10, 6))\n",
        "sns.boxplot(x='target', y='chol', data=train_data)\n",
        "plt.title('Cholesterol Levels by Heart Disease')\n",
        "plt.xlabel('Heart Disease')\n",
        "plt.ylabel('Cholesterol')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nFOoRTEXcPUq"
      },
      "source": [
        "**Initial Data Exploration and Insights**\n",
        "\n",
        "The preliminary analysis of the dataset for the Zindi hackathon has yielded several important insights. The dataset presents an imbalance in the distribution of heart disease cases, with a higher number of positive cases compared to negative ones. This imbalance will require careful consideration during the modeling phase to ensure fair and accurate predictions.\n",
        "\n",
        "Fortunately, the dataset is complete, with no missing values, which simplifies the preprocessing steps. The age range of participants spans from 29 to 77 years, suggesting the need for age-based analysis or grouping in our model. Cholesterol levels show significant variation (126 to 564), indicating that this could be a crucial factor in predicting heart disease.\n",
        "\n",
        "Further analysis revealed that exercise-induced angina ('exang') has the strongest correlation with the target variable, while chest pain type ('cp') shows a negative correlation. Most other features exhibit weak correlations with the target, and there's a lack of strong correlations between features, which is favorable for avoiding multicollinearity in the model.\n",
        "\n",
        "Age appears to play a significant role, with heart disease cases becoming more prevalent in older age groups, particularly after 50. Interestingly, cholesterol levels alone do not show a marked difference between those with and without heart disease, suggesting it may not be a strong predictor in isolation.\n",
        "\n",
        "**Next Steps and Action Plan**\n",
        "\n",
        "Based on these observations, the next phase of the project will focus on data preprocessing and feature engineering. The following steps will be undertaken:\n",
        "\n",
        "1. Addressing the class imbalance through techniques such as oversampling, undersampling, or applying class weights.\n",
        "\n",
        "2. Encoding categorical variables, ensuring proper representation of features like chest pain type and thalassemia.\n",
        "\n",
        "3. Scaling numerical features to ensure all variables are on a comparable scale for modeling.\n",
        "\n",
        "4. Creating age group categories to better capture the relationship between age and heart disease prevalence.\n",
        "\n",
        "5. Splitting the data into training and testing sets for model development and evaluation.\n",
        "\n",
        "In terms of modeling strategy, the focus will be on features that show stronger correlations with the target, such as exercise-induced angina and chest pain type. Given the potential for non-linear relationships, models capable of capturing such complexities (e.g., Random Forest, Gradient Boosting) will be considered. Feature selection techniques may also be employed to identify the most predictive variables.\n",
        "\n",
        "By following this structured approach, the aim is to develop a robust and accurate model for predicting heart disease, taking into account the nuances and challenges presented by the dataset. Regular evaluation and iteration will be key to improving the model's performance and achieving success in the Zindi hackathon."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pmweBPTMwewJ",
        "outputId": "ae138a44-2fc3-415a-aa22-97db75adefb6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Shape of training data before SMOTE: (5842, 26)\n",
            "Shape of training data after SMOTE: (9504, 26)\n",
            "Distribution of target variable before SMOTE: [1090 4752]\n",
            "Distribution of target variable after SMOTE: [4752 4752]\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE\n",
        "\n",
        "# Load the data\n",
        "train_data = pd.read_csv('/content/Train Dataset .csv')\n",
        "\n",
        "# Separate features and target\n",
        "X = train_data.drop(['Id', 'target'], axis=1)\n",
        "y = train_data['target']\n",
        "\n",
        "# Define categorical and numerical columns\n",
        "categorical_features = ['Sex', 'cp', 'fbs', 'restecg', 'exang', 'slope', 'ca', 'thal']\n",
        "numerical_features = ['Age', 'trestbps', 'chol', 'thalach', 'oldpeak']\n",
        "\n",
        "# Create age groups\n",
        "X['AgeGroup'] = pd.cut(X['Age'], bins=[0, 40, 50, 60, 70, 100], labels=['0-40', '41-50', '51-60', '61-70', '70+'])\n",
        "categorical_features.append('AgeGroup')\n",
        "\n",
        "# Split the data\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)\n",
        "\n",
        "# Create preprocessing steps\n",
        "categorical_transformer = OneHotEncoder(drop='first', sparse_output=False)\n",
        "numerical_transformer = StandardScaler()\n",
        "\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numerical_transformer, numerical_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ])\n",
        "\n",
        "# Fit the preprocessor on the training data\n",
        "X_train_preprocessed = preprocessor.fit_transform(X_train)\n",
        "X_test_preprocessed = preprocessor.transform(X_test)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_balanced, y_train_balanced = smote.fit_resample(X_train_preprocessed, y_train)\n",
        "\n",
        "print(\"Shape of training data before SMOTE:\", X_train_preprocessed.shape)\n",
        "print(\"Shape of training data after SMOTE:\", X_train_balanced.shape)\n",
        "print(\"Distribution of target variable before SMOTE:\", np.bincount(y_train))\n",
        "print(\"Distribution of target variable after SMOTE:\", np.bincount(y_train_balanced))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RFsUEkH3IP5o"
      },
      "source": [
        "These results show that our preprocessing steps have successfully addressed the class imbalance issue. We now have a balanced dataset to work with, which should help our models learn to predict both classes more effectively.\n",
        "Now that we have our preprocessed and balanced data, we're ready to move on to the next step: **model selection, training, and evaluation** . We'll try a few different algorithms and compare their performance."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0zN49LDkNabt",
        "outputId": "154ffd92-2893-4090-f25e-26c7446475ba"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Shape of training data before SMOTE: (5842, 26)\n",
            "Shape of training data after SMOTE: (9504, 26)\n",
            "Distribution of target variable before SMOTE: [1090 4752]\n",
            "Distribution of target variable after SMOTE: [4752 4752]\n",
            "\n",
            "Logistic Regression Results:\n",
            "Accuracy: 0.8158795345653662\n",
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.50      1.00      0.67       272\n",
            "           1       1.00      0.77      0.87      1189\n",
            "\n",
            "    accuracy                           0.82      1461\n",
            "   macro avg       0.75      0.89      0.77      1461\n",
            "weighted avg       0.91      0.82      0.83      1461\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[272   0]\n",
            " [269 920]]\n",
            "\n",
            "Random Forest Results:\n",
            "Accuracy: 0.813141683778234\n",
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.50      0.54      0.52       272\n",
            "           1       0.89      0.87      0.88      1189\n",
            "\n",
            "    accuracy                           0.81      1461\n",
            "   macro avg       0.70      0.71      0.70      1461\n",
            "weighted avg       0.82      0.81      0.82      1461\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[ 148  124]\n",
            " [ 149 1040]]\n",
            "\n",
            "XGBoost Results:\n",
            "Accuracy: 0.8199863107460643\n",
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.52      0.52      0.52       272\n",
            "           1       0.89      0.89      0.89      1189\n",
            "\n",
            "    accuracy                           0.82      1461\n",
            "   macro avg       0.70      0.70      0.70      1461\n",
            "weighted avg       0.82      0.82      0.82      1461\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[ 141  131]\n",
            " [ 132 1057]]\n"
          ]
        }
      ],
      "source": [
        "   import pandas as pd\n",
        "   import numpy as np\n",
        "   from sklearn.model_selection import train_test_split\n",
        "   from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "   from sklearn.compose import ColumnTransformer\n",
        "   from sklearn.pipeline import Pipeline\n",
        "   from imblearn.over_sampling import SMOTE\n",
        "   from sklearn.linear_model import LogisticRegression\n",
        "   from sklearn.ensemble import RandomForestClassifier\n",
        "   from xgboost import XGBClassifier\n",
        "   from sklearn.metrics import accuracy_score, classification_report, confusion_matrix\n",
        "\n",
        "   # Load the data\n",
        "   train_data = pd.read_csv('/content/Train Dataset .csv')\n",
        "\n",
        "   # Separate features and target\n",
        "   X = train_data.drop(['Id', 'target'], axis=1)\n",
        "   y = train_data['target']\n",
        "\n",
        "   # Define categorical and numerical columns\n",
        "   categorical_features = ['Sex', 'cp', 'fbs', 'restecg', 'exang', 'slope', 'ca', 'thal']\n",
        "   numerical_features = ['Age', 'trestbps', 'chol', 'thalach', 'oldpeak']\n",
        "\n",
        "   # Create age groups\n",
        "   X['AgeGroup'] = pd.cut(X['Age'], bins=[0, 40, 50, 60, 70, 100], labels=['0-40', '41-50', '51-60', '61-70', '70+'])\n",
        "   categorical_features.append('AgeGroup')\n",
        "\n",
        "   # Split the data\n",
        "   X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)\n",
        "\n",
        "# Create preprocessing steps\n",
        "categorical_transformer = OneHotEncoder(drop='first', sparse_output=False)\n",
        "numerical_transformer = StandardScaler()\n",
        "\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numerical_transformer, numerical_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ])\n",
        "\n",
        "# Fit the preprocessor on the training data\n",
        "X_train_preprocessed = preprocessor.fit_transform(X_train)\n",
        "X_test_preprocessed = preprocessor.transform(X_test)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_balanced, y_train_balanced = smote.fit_resample(X_train_preprocessed, y_train)\n",
        "\n",
        "print(\"Shape of training data before SMOTE:\", X_train_preprocessed.shape)\n",
        "print(\"Shape of training data after SMOTE:\", X_train_balanced.shape)\n",
        "print(\"Distribution of target variable before SMOTE:\", np.bincount(y_train))\n",
        "print(\"Distribution of target variable after SMOTE:\", np.bincount(y_train_balanced))\n",
        "\n",
        "# Logistic Regression\n",
        "lr_model = LogisticRegression(random_state=42)\n",
        "lr_model.fit(X_train_balanced, y_train_balanced)\n",
        "lr_pred = lr_model.predict(X_test_preprocessed)\n",
        "\n",
        "# Random Forest\n",
        "rf_model = RandomForestClassifier(n_estimators=100, random_state=42)\n",
        "rf_model.fit(X_train_balanced, y_train_balanced)\n",
        "rf_pred = rf_model.predict(X_test_preprocessed)\n",
        "\n",
        "# XGBoost\n",
        "xgb_model = XGBClassifier(eval_metric='logloss', random_state=42)\n",
        "xgb_model.fit(X_train_balanced, y_train_balanced)\n",
        "xgb_pred = xgb_model.predict(X_test_preprocessed)\n",
        "\n",
        "# Evaluate models\n",
        "models = [('Logistic Regression', lr_pred),\n",
        "          ('Random Forest', rf_pred),\n",
        "          ('XGBoost', xgb_pred)]\n",
        "\n",
        "for name, predictions in models:\n",
        "    print(f\"\\n{name} Results:\")\n",
        "    print(\"Accuracy:\", accuracy_score(y_test, predictions))\n",
        "    print(\"\\nClassification Report:\")\n",
        "    print(classification_report(y_test, predictions))\n",
        "    print(\"\\nConfusion Matrix:\")\n",
        "    print(confusion_matrix(y_test, predictions))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x2YU9TfbT9E2"
      },
      "source": [
        " Three different machine learning algorithms were implemented and evaluated: Logistic Regression, Random Forest, and XGBoost. Each model demonstrated unique strengths and weaknesses in predicting heart disease cases.\n",
        "\n",
        "The Logistic Regression model achieved an accuracy of 81.59%. It showed perfect recall (1.00) for predicting the absence of heart disease (class 0) but with low precision (0.50). Conversely, it demonstrated high precision (1.00) for predicting the presence of heart disease (class 1) but with lower recall (0.77). This model tends to overpredict the absence of heart disease, which could be problematic in a medical care where false negatives might lead to missed diagnoses.\n",
        "\n",
        "The Random Forest model performed slightly differently, with an accuracy of 81.31%. It showed a more balanced performance between classes, slightly favoring the prediction of heart disease cases (class 1) over non-heart disease cases. This balanced approach could be more suitable for medical applications where both false positives and false negatives carry significant consequences.\n",
        "\n",
        "The XGBoost model emerged as the top performer with an accuracy of 82.00%. It demonstrated the most balanced performance across both classes and achieved the highest overall accuracy among the three models. This balanced and accurate performance makes XGBoost a promising candidate for further development and refinement.\n",
        "\n",
        "Based on these results, the next phase of this project will focus on several key areas to enhance model performance and reliability:\n",
        "\n",
        "1. **Hyperparameter tuning**: Efforts will be made to optimize the hyperparameters of the Random Forest and XGBoost models, which have shown potential for improvement.\n",
        "\n",
        "2. **Feature importance analysis**: An in-depth examination of which features contribute most significantly to the predictions will be conducted. This analysis could provide valuable insights into the factors most strongly associated with heart disease risk.\n",
        "\n",
        "3. **Cross-validation implementation**: To obtain a more robust estimate of model performance and reduce the risk of overfitting, k-fold cross-validation will be implemented.\n",
        "\n",
        "4. **Exploration of ensemble methods**: Combining predictions from multiple models through ensemble techniques will be investigated as a potential strategy to further improve overall performance.\n",
        "\n",
        "These steps aim to refine the models, increase their accuracy and reliability, and potentially uncover additional insights into heart disease prediction. The focus will be on maintaining a balance between predictive power and interpretability, ensuring that the final model not only performs well statistically but also provides meaningful and actionable insights medically."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yT1YquofVaEC",
        "outputId": "93d981c3-3867-412b-839b-6e51b4bfbca8"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Best parameters: {'learning_rate': 0.01, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 200}\n",
            "Best cross-validation score: 0.8826811927240511\n",
            "\n",
            "Best XGBoost Model Results:\n",
            "Accuracy: 0.8158795345653662\n",
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.50      1.00      0.67       272\n",
            "           1       1.00      0.77      0.87      1189\n",
            "\n",
            "    accuracy                           0.82      1461\n",
            "   macro avg       0.75      0.89      0.77      1461\n",
            "weighted avg       0.91      0.82      0.83      1461\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[272   0]\n",
            " [269 920]]\n"
          ]
        }
      ],
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "param_grid = {\n",
        "    'max_depth': [3, 4, 5],\n",
        "    'learning_rate': [0.01, 0.1, 0.3],\n",
        "    'n_estimators': [100, 200],\n",
        "    'min_child_weight': [1, 3, 5]\n",
        "}\n",
        "\n",
        "xgb_model = XGBClassifier(eval_metric='logloss', random_state=42)\n",
        "grid_search = GridSearchCV(estimator=xgb_model, param_grid=param_grid, cv=5, scoring='accuracy', n_jobs=-1)\n",
        "grid_search.fit(X_train_balanced, y_train_balanced)\n",
        "\n",
        "print(\"Best parameters:\", grid_search.best_params_)\n",
        "print(\"Best cross-validation score:\", grid_search.best_score_)\n",
        "\n",
        "best_xgb_model = grid_search.best_estimator_\n",
        "best_xgb_pred = best_xgb_model.predict(X_test_preprocessed)\n",
        "\n",
        "print(\"\\nBest XGBoost Model Results:\")\n",
        "print(\"Accuracy:\", accuracy_score(y_test, best_xgb_pred))\n",
        "print(\"\\nClassification Report:\")\n",
        "print(classification_report(y_test, best_xgb_pred))\n",
        "print(\"\\nConfusion Matrix:\")\n",
        "print(confusion_matrix(y_test, best_xgb_pred))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qS_tMcqU5NJp"
      },
      "source": [
        "The hyperparameter tuning process for XGBoost has yielded interesting results. The best parameters identified include a low learning rate of 0.01, a max depth of 3, a min child weight of 1, and 200 estimators. This configuration suggests a relatively conservative model that prioritizes preventing overfitting, which is crucial in medical applications where generalizability is key.\n",
        "\n",
        "The best cross-validation score achieved was 88.27%, which is encouraging and represents an improvement over our previous models. However, the performance on the test set (81.59% accuracy) is similar to our earlier models, indicating a potential discrepancy between cross-validation and test set performance. This could suggest some degree of overfitting or variability in the data that we need to address.\n",
        "\n",
        "Interestingly, the performance characteristics of this tuned XGBoost model on the test set mirror those of our earlier Logistic Regression model, with perfect recall but low precision for non-heart disease cases, and high precision but lower recall for heart disease cases. This similarity across different model types suggests that these performance patterns might be intrinsic to our dataset or problem structure.\n",
        "\n",
        "Given these observations, our next steps will focus on further refining our approach and gaining deeper insights into the problem:\n",
        "\n",
        "1. **Feature Importance Analysis**: We will examine which features are most influential in our best XGBoost model. This analysis will provide valuable insights into the factors most predictive of heart disease, potentially guiding future data collection efforts or informing medical professionals about key risk factors.\n",
        "\n",
        "2. **Ensemble Method Exploration**: Since different models are performing similarly but with varying strengths, we will implement an ensemble method to combine their predictions. This approach could leverage the strengths of each model to produce more robust and accurate predictions overall.\n",
        "\n",
        "3. **Addressing Class Imbalance**: Given the model's continued struggle with the minority class (non-heart disease cases), we will revisit our approach to handling class imbalance. This may involve adjusting class weights, exploring different resampling techniques, or implementing advanced methods like SMOTE (Synthetic Minority Over-sampling Technique) to create a more balanced dataset.\n",
        "\n",
        "4. **Cross-Validation Strategy Review**: The discrepancy between cross-validation and test set performance suggests we might need to reassess our cross-validation strategy. We will consider techniques like stratified k-fold cross-validation to ensure our validation process better represents the test set distribution.\n",
        "\n",
        "5. **Additional Feature Engineering**: Based on the feature importance analysis, we may identify opportunities for creating new features or transforming existing ones to capture more predictive information.\n",
        "\n",
        "6. **Model Interpretability**: Given the medical nature of this prediction task, we will also focus on making our model more interpretable. This could involve using techniques like SHAP (SHapley Additive exPlanations) values to understand how each feature contributes to individual predictions.\n",
        "\n",
        "By implementing these steps, we aim to not only improve the predictive performance of our model but also to derive meaningful insights that could be valuable in a clinical setting. Our goal is to develop a model that is not only accurate but also provides interpretable and actionable information for heart disease risk assessment.\n",
        "\n",
        "**Note:** In tree-based models like XGBoost, a \"child\" refers to a node resulting from a split of a parent node.\n",
        "Imagine a decision tree - each node represents a decision point based on a feature. When a node is split, it creates two or more \"children,\" which are subsequent nodes in the tree structure.\n",
        "The `min_child_weight` parameter determines how much weight (or how many samples) is required in a child node to allow further splitting. This helps prevent the model from creating very specific rules that might only apply to a few samples, leading to overfitting."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 827
        },
        "id": "Rm004gs-6PCT",
        "outputId": "a27bf42c-b5f7-4e30-be9f-89e05d519194"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Feature Importance:\n",
            "cat__exang_1: 0.6651\n",
            "cat__Sex_1: 0.0823\n",
            "cat__cp_2: 0.0647\n",
            "cat__cp_1: 0.0640\n",
            "cat__cp_3: 0.0557\n",
            "cat__thal_1: 0.0356\n",
            "cat__restecg_2: 0.0204\n",
            "cat__ca_3: 0.0121\n",
            "num__trestbps: 0.0000\n",
            "num__thalach: 0.0000\n"
          ]
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Get feature importance\n",
        "importance = best_xgb_model.feature_importances_\n",
        "feature_names = preprocessor.get_feature_names_out()\n",
        "\n",
        "# Sort features by importance\n",
        "feature_importance = sorted(zip(importance, feature_names), reverse=True)\n",
        "\n",
        "# Plot feature importance\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.bar([x[1] for x in feature_importance[:10]], [x[0] for x in feature_importance[:10]])\n",
        "plt.title('Top 10 Feature Importance')\n",
        "plt.xlabel('Features')\n",
        "plt.ylabel('Importance')\n",
        "plt.xticks(rotation=45, ha='right')\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "\n",
        "# Print feature importance\n",
        "print(\"Feature Importance:\")\n",
        "for importance, name in feature_importance[:10]:\n",
        "    print(f\"{name}: {importance:.4f}\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zw8SqQ4IZlnH"
      },
      "source": [
        "The feature importance analysis reveals that our model's predictions are heavily influenced by a small subset of features:\n",
        "\n",
        "1. Exercise-induced angina (exang) is by far the most crucial predictor, accounting for about 66.5% of the model's predictive power. This aligns with medical knowledge that angina during physical activity is a strong indicator of heart disease.\n",
        "\n",
        "2. The patient's sex is the second most important feature, contributing 8.2% to the predictions. This reflects the known differences in heart disease risk between males and females.\n",
        "\n",
        "3. Different categories of chest pain (cp) collectively account for about 18.4% of the model's predictive power, underscoring the importance of this symptom in heart disease diagnosis.\n",
        "\n",
        "4. Other features like thalassemia, resting electrocardiographic results, and coronary artery status play smaller but still significant roles in the predictions.\n",
        "\n",
        "5. Interestingly, some numerical features like resting blood pressure and maximum heart rate appear to have no importance in the model's predictions, which is unexpected given their clinical relevance.\n",
        "\n",
        "These findings provide crucial insights but also raise some concerns about potential overfitting and the underutilization of certain features. To address these issues and further improve our model, we will proceed with the following steps:\n",
        "\n",
        "1. **Ensemble Method Implementation**: As suggested earlier, we will create a voting classifier that combines predictions from our best Logistic Regression, Random Forest, and XGBoost models. This ensemble approach should help balance out the strengths and weaknesses of individual models, potentially leading to more robust and accurate predictions.\n",
        "\n",
        "2. **Feature Engineering**: We will create interaction features, especially between the top predictors like exercise-induced angina, sex, and chest pain categories. This could capture more complex relationships between these important factors.\n",
        "\n",
        "3. **Regularization Techniques**: To address the potential overfitting indicated by the heavy reliance on a single feature, we will experiment with different regularization techniques in our models.\n",
        "\n",
        "4. **Reinvestigation of Numerical Features**: We'll revisit our preprocessing of numerical features, especially resting blood pressure and maximum heart rate, to understand why they're not contributing to the predictions. This may involve trying different scaling methods or creating derived features.\n",
        "\n",
        "5. **Domain Expert Consultation**:I also suggest We seek input from medical experts to validate whether our feature importances align with clinical knowledge about heart disease risk factors. This could help identify any discrepancies between our model's learning and established medical understanding. But I don't think we have that time 😊\n",
        "\n",
        "6. **Cross-Model Comparison**: We'll compare feature importances across different models (e.g., Random Forest) to identify consistencies and discrepancies in feature relevance.\n",
        "\n",
        "7. **Advanced Feature Selection**: We'll employ more sophisticated feature selection techniques to ensure we're using the most relevant set of predictors while avoiding overfitting.\n",
        "\n",
        "By implementing these steps, we aim to develop a more robust and clinically relevant model for heart disease prediction. The ensemble method, in particular, should help us leverage the strengths of different modeling approaches while mitigating their individual weaknesses.\n",
        "\n",
        "Our next immediate step will be to implement the voting classifier ensemble. We'll combine our best Logistic Regression, Random Forest, and XGBoost models, and evaluate the performance of this ensemble against our individual models. This approach should provide a more balanced and potentially more accurate prediction of heart disease risk, which is crucial for the success of the project."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kWX9M7zUMPd8"
      },
      "source": [
        "# **Ensemble submission**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_35Ce0A6RnMp",
        "outputId": "0a555fb3-8a5a-43f5-c3ad-b9535c51f155"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Cross-validation scores: [0.88396624 0.87552743 0.88607595 0.87869198 0.88765823]\n",
            "Mean CV score: 0.8824 (+/- 0.0091)\n",
            "Ensemble Model Accuracy on validation set: 0.8159\n",
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.49      1.00      0.66       260\n",
            "           1       1.00      0.78      0.87      1201\n",
            "\n",
            "    accuracy                           0.82      1461\n",
            "   macro avg       0.75      0.89      0.77      1461\n",
            "weighted avg       0.91      0.82      0.84      1461\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[260   0]\n",
            " [269 932]]\n",
            "Submission file 'ensemble_submission.csv' has been created.\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from xgboost import XGBClassifier\n",
        "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.ensemble import VotingClassifier\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Load the datasets\n",
        "train_df = pd.read_csv('/content/Train Dataset .csv')\n",
        "test_df = pd.read_csv('/content/Test Dataset.csv')\n",
        "\n",
        "# Separate features and target in the training data\n",
        "X = train_df.drop('target', axis=1)\n",
        "y = train_df['target']\n",
        "\n",
        "# Split the data into training and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# Identify numeric and categorical columns\n",
        "numeric_features = X.select_dtypes(include=[np.number]).columns\n",
        "categorical_features = X.select_dtypes(exclude=[np.number]).columns\n",
        "\n",
        "# Create preprocessing pipelines\n",
        "numeric_transformer = Pipeline(steps=[\n",
        "    ('scaler', StandardScaler())\n",
        "])\n",
        "\n",
        "categorical_transformer = Pipeline(steps=[\n",
        "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
        "])\n",
        "\n",
        "# Combine preprocessing steps\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numeric_transformer, numeric_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ])\n",
        "\n",
        "\n",
        "\n",
        "# Create the full preprocessing pipeline\n",
        "full_pipeline = Pipeline([\n",
        "    ('preprocessor', preprocessor)\n",
        "])\n",
        "\n",
        "# Fit the pipeline and transform the data\n",
        "X_train_processed = full_pipeline.fit_transform(X_train, y_train)\n",
        "X_val_processed = full_pipeline.transform(X_val)\n",
        "\n",
        "# Change the column names in test_df to match the training data\n",
        "test_df.rename(columns={'id': 'Id', 'age': 'Age', 'sex': 'Sex'}, inplace=True)\n",
        "\n",
        "# Transform the test data\n",
        "X_test_processed = full_pipeline.transform(test_df)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_processed, y_train_balanced = smote.fit_resample(X_train_processed, y_train)\n",
        "\n",
        "# Create base models\n",
        "best_lr_model = LogisticRegression(C=1.0, penalty='l2', random_state=42)\n",
        "best_rf_model = RandomForestClassifier(n_estimators=100, max_depth=5, random_state=42)\n",
        "best_xgb_model = XGBClassifier(learning_rate=0.01, max_depth=3, min_child_weight=1, n_estimators=200, random_state=42)\n",
        "\n",
        "# Create the voting classifier\n",
        "voting_clf = VotingClassifier(\n",
        "    estimators=[\n",
        "        ('lr', best_lr_model),\n",
        "        ('rf', best_rf_model),\n",
        "        ('xgb', best_xgb_model)\n",
        "    ],\n",
        "    voting='soft'\n",
        ")\n",
        "\n",
        "# Fit the voting classifier\n",
        "voting_clf.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Perform cross-validation\n",
        "cv_scores = cross_val_score(voting_clf, X_train_processed, y_train_balanced, cv=5)\n",
        "print(f\"Cross-validation scores: {cv_scores}\")\n",
        "print(f\"Mean CV score: {cv_scores.mean():.4f} (+/- {cv_scores.std() * 2:.4f})\")\n",
        "\n",
        "# Make predictions on the validation set\n",
        "y_val_pred = voting_clf.predict(X_val_processed)\n",
        "\n",
        "# Calculate accuracy\n",
        "accuracy = accuracy_score(y_val, y_val_pred)\n",
        "print(f\"Ensemble Model Accuracy on validation set: {accuracy:.4f}\")\n",
        "\n",
        "# Print classification report\n",
        "print(\"\\nClassification Report:\")\n",
        "print(classification_report(y_val, y_val_pred))\n",
        "\n",
        "# Print confusion matrix\n",
        "print(\"\\nConfusion Matrix:\")\n",
        "print(confusion_matrix(y_val, y_val_pred))\n",
        "\n",
        "# Make predictions on the test set\n",
        "test_predictions = voting_clf.predict(X_test_processed)\n",
        "\n",
        "# Create a submission DataFrame\n",
        "submission = pd.DataFrame({\n",
        "    'ID': test_df['Id'],\n",
        "    'target': test_predictions\n",
        "})\n",
        "\n",
        "# Save the submission file\n",
        "submission.to_csv('ensemble_submission.csv', index=False)\n",
        "print(\"Submission file 'ensemble_submission.csv' has been created.\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FXax62ps8Pow",
        "outputId": "115d070c-9133-4692-90ff-0c67ab25b60a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting catboost\n",
            "  Downloading catboost-1.2.7-cp310-cp310-manylinux2014_x86_64.whl.metadata (1.2 kB)\n",
            "Requirement already satisfied: graphviz in /usr/local/lib/python3.10/dist-packages (from catboost) (0.20.3)\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.10/dist-packages (from catboost) (3.7.1)\n",
            "Requirement already satisfied: numpy<2.0,>=1.16.0 in /usr/local/lib/python3.10/dist-packages (from catboost) (1.26.4)\n",
            "Requirement already satisfied: pandas>=0.24 in /usr/local/lib/python3.10/dist-packages (from catboost) (2.1.4)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from catboost) (1.13.1)\n",
            "Requirement already satisfied: plotly in /usr/local/lib/python3.10/dist-packages (from catboost) (5.24.1)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.10/dist-packages (from catboost) (1.16.0)\n",
            "Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.10/dist-packages (from pandas>=0.24->catboost) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=0.24->catboost) (2024.2)\n",
            "Requirement already satisfied: tzdata>=2022.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=0.24->catboost) (2024.1)\n",
            "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (1.3.0)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (0.12.1)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (4.53.1)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (1.4.7)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (24.1)\n",
            "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (10.4.0)\n",
            "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->catboost) (3.1.4)\n",
            "Requirement already satisfied: tenacity>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from plotly->catboost) (9.0.0)\n",
            "Downloading catboost-1.2.7-cp310-cp310-manylinux2014_x86_64.whl (98.7 MB)\n",
            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m98.7/98.7 MB\u001b[0m \u001b[31m7.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hInstalling collected packages: catboost\n",
            "Successfully installed catboost-1.2.7\n"
          ]
        }
      ],
      "source": [
        "!pip install catboost"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BJ6d9Qf2IB3c"
      },
      "source": [
        "# **Only the bestxgboost with catboost**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wE5pgzytIKMh",
        "outputId": "360cde99-8c69-4180-a933-ce45c2f090c8"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0:\tlearn: 0.8823840\ttest: 0.8158795\tbest: 0.8158795 (0)\ttotal: 54.7ms\tremaining: 54.7s\n",
            "100:\tlearn: 0.8823840\ttest: 0.8158795\tbest: 0.8158795 (0)\ttotal: 665ms\tremaining: 5.92s\n",
            "200:\tlearn: 0.8823840\ttest: 0.8158795\tbest: 0.8158795 (0)\ttotal: 1.28s\tremaining: 5.08s\n",
            "300:\tlearn: 0.8848101\ttest: 0.8151951\tbest: 0.8158795 (0)\ttotal: 1.91s\tremaining: 4.43s\n",
            "400:\tlearn: 0.8888186\ttest: 0.8158795\tbest: 0.8158795 (0)\ttotal: 2.53s\tremaining: 3.78s\n",
            "500:\tlearn: 0.8962025\ttest: 0.8158795\tbest: 0.8172485 (414)\ttotal: 3.15s\tremaining: 3.14s\n",
            "600:\tlearn: 0.9032700\ttest: 0.8165640\tbest: 0.8186174 (583)\ttotal: 3.79s\tremaining: 2.52s\n",
            "700:\tlearn: 0.9102321\ttest: 0.8151951\tbest: 0.8186174 (583)\ttotal: 4.43s\tremaining: 1.89s\n",
            "800:\tlearn: 0.9186709\ttest: 0.8104038\tbest: 0.8186174 (583)\ttotal: 5.06s\tremaining: 1.26s\n",
            "900:\tlearn: 0.9258439\ttest: 0.8151951\tbest: 0.8186174 (583)\ttotal: 5.89s\tremaining: 647ms\n",
            "999:\tlearn: 0.9328059\ttest: 0.8179329\tbest: 0.8186174 (583)\ttotal: 7.22s\tremaining: 0us\n",
            "\n",
            "bestTest = 0.8186173854\n",
            "bestIteration = 583\n",
            "\n",
            "Shrink model to first 584 iterations.\n",
            "0:\tlearn: 0.8819884\ttotal: 16.2ms\tremaining: 16.2s\n",
            "100:\tlearn: 0.8821203\ttotal: 1.16s\tremaining: 10.3s\n",
            "200:\tlearn: 0.8898998\ttotal: 1.99s\tremaining: 7.89s\n",
            "300:\tlearn: 0.9106013\ttotal: 2.56s\tremaining: 5.95s\n",
            "400:\tlearn: 0.9365770\ttotal: 3.14s\tremaining: 4.68s\n",
            "500:\tlearn: 0.9570148\ttotal: 3.72s\tremaining: 3.7s\n",
            "600:\tlearn: 0.9687500\ttotal: 4.28s\tremaining: 2.84s\n",
            "700:\tlearn: 0.9742880\ttotal: 4.86s\tremaining: 2.07s\n",
            "800:\tlearn: 0.9783755\ttotal: 5.43s\tremaining: 1.35s\n",
            "900:\tlearn: 0.9804852\ttotal: 6.03s\tremaining: 662ms\n",
            "999:\tlearn: 0.9825949\ttotal: 6.6s\tremaining: 0us\n",
            "0:\tlearn: 0.8840981\ttotal: 5.78ms\tremaining: 5.78s\n",
            "100:\tlearn: 0.8840981\ttotal: 532ms\tremaining: 4.73s\n",
            "200:\tlearn: 0.8840981\ttotal: 1.08s\tremaining: 4.3s\n",
            "300:\tlearn: 0.8856804\ttotal: 1.64s\tremaining: 3.81s\n",
            "400:\tlearn: 0.8906909\ttotal: 2.21s\tremaining: 3.31s\n",
            "500:\tlearn: 0.8972838\ttotal: 2.77s\tremaining: 2.76s\n",
            "600:\tlearn: 0.9026899\ttotal: 3.35s\tremaining: 2.22s\n",
            "700:\tlearn: 0.9096783\ttotal: 3.92s\tremaining: 1.67s\n",
            "800:\tlearn: 0.9167985\ttotal: 4.49s\tremaining: 1.11s\n",
            "900:\tlearn: 0.9239188\ttotal: 5.39s\tremaining: 592ms\n",
            "999:\tlearn: 0.9309072\ttotal: 6.74s\tremaining: 0us\n",
            "0:\tlearn: 0.8814610\ttotal: 8.8ms\tremaining: 8.8s\n",
            "100:\tlearn: 0.8814610\ttotal: 1.17s\tremaining: 10.4s\n",
            "200:\tlearn: 0.8814610\ttotal: 1.71s\tremaining: 6.79s\n",
            "300:\tlearn: 0.8825158\ttotal: 2.29s\tremaining: 5.32s\n",
            "400:\tlearn: 0.8850211\ttotal: 2.85s\tremaining: 4.25s\n",
            "500:\tlearn: 0.8904272\ttotal: 3.43s\tremaining: 3.42s\n",
            "600:\tlearn: 0.8962289\ttotal: 3.99s\tremaining: 2.65s\n",
            "700:\tlearn: 0.9017669\ttotal: 4.58s\tremaining: 1.95s\n",
            "800:\tlearn: 0.9083597\ttotal: 5.16s\tremaining: 1.28s\n",
            "900:\tlearn: 0.9165348\ttotal: 5.71s\tremaining: 628ms\n",
            "999:\tlearn: 0.9235232\ttotal: 6.29s\tremaining: 0us\n",
            "0:\tlearn: 0.8833070\ttotal: 6.76ms\tremaining: 6.76s\n",
            "100:\tlearn: 0.8833070\ttotal: 531ms\tremaining: 4.73s\n",
            "200:\tlearn: 0.8833070\ttotal: 1.11s\tremaining: 4.43s\n",
            "300:\tlearn: 0.8838344\ttotal: 1.68s\tremaining: 3.9s\n",
            "400:\tlearn: 0.8856804\ttotal: 2.3s\tremaining: 3.44s\n",
            "500:\tlearn: 0.8914821\ttotal: 2.89s\tremaining: 2.88s\n",
            "600:\tlearn: 0.8957015\ttotal: 3.46s\tremaining: 2.29s\n",
            "700:\tlearn: 0.9032173\ttotal: 4.04s\tremaining: 1.72s\n",
            "800:\tlearn: 0.9111287\ttotal: 4.64s\tremaining: 1.15s\n",
            "900:\tlearn: 0.9175897\ttotal: 5.78s\tremaining: 635ms\n",
            "999:\tlearn: 0.9248418\ttotal: 7.12s\tremaining: 0us\n",
            "0:\tlearn: 0.8810654\ttotal: 9.23ms\tremaining: 9.22s\n",
            "100:\tlearn: 0.8810654\ttotal: 774ms\tremaining: 6.88s\n",
            "200:\tlearn: 0.8810654\ttotal: 1.32s\tremaining: 5.26s\n",
            "300:\tlearn: 0.8822521\ttotal: 1.91s\tremaining: 4.42s\n",
            "400:\tlearn: 0.8850211\ttotal: 2.47s\tremaining: 3.69s\n",
            "500:\tlearn: 0.8901635\ttotal: 3.05s\tremaining: 3.04s\n",
            "600:\tlearn: 0.8960970\ttotal: 3.59s\tremaining: 2.38s\n",
            "700:\tlearn: 0.9030854\ttotal: 4.17s\tremaining: 1.78s\n",
            "800:\tlearn: 0.9104694\ttotal: 4.75s\tremaining: 1.18s\n",
            "900:\tlearn: 0.9153481\ttotal: 5.32s\tremaining: 584ms\n",
            "999:\tlearn: 0.9223365\ttotal: 5.91s\tremaining: 0us\n",
            "Cross-validation scores: [0.57489451 0.88291139 0.89609705 0.88765823 0.89662447]\n",
            "Mean CV score: 0.8276 (+/- 0.2530)\n",
            "CatBoost Model Accuracy on validation set: 0.8186\n",
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.50      0.96      0.65       260\n",
            "           1       0.99      0.79      0.88      1201\n",
            "\n",
            "    accuracy                           0.82      1461\n",
            "   macro avg       0.74      0.87      0.76      1461\n",
            "weighted avg       0.90      0.82      0.84      1461\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[249  11]\n",
            " [254 947]]\n",
            "Submission file 'catboost_submission.csv' has been created.\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE\n",
        "from catboost import CatBoostClassifier\n",
        "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix\n",
        "from sklearn.model_selection import cross_val_score\n",
        "\n",
        "# Load the datasets\n",
        "train_df = pd.read_csv('/content/Train Dataset.csv')\n",
        "test_df = pd.read_csv('/content/Test Dataset.csv')\n",
        "\n",
        "# Separate features and target in the training data\n",
        "X = train_df.drop('target', axis=1)\n",
        "y = train_df['target']\n",
        "\n",
        "# Split the data into training and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# Identify numeric and categorical columns\n",
        "numeric_features = X.select_dtypes(include=[np.number]).columns\n",
        "categorical_features = X.select_dtypes(exclude=[np.number]).columns\n",
        "\n",
        "# Create preprocessing pipelines\n",
        "numeric_transformer = Pipeline(steps=[\n",
        "    ('scaler', StandardScaler())\n",
        "])\n",
        "\n",
        "categorical_transformer = Pipeline(steps=[\n",
        "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
        "])\n",
        "\n",
        "# Combine preprocessing steps\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numeric_transformer, numeric_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ]\n",
        ")\n",
        "\n",
        "# Create the full preprocessing pipeline\n",
        "full_pipeline = Pipeline([\n",
        "    ('preprocessor', preprocessor)\n",
        "])\n",
        "\n",
        "# Fit the pipeline and transform the data\n",
        "X_train_processed = full_pipeline.fit_transform(X_train, y_train)\n",
        "X_val_processed = full_pipeline.transform(X_val)\n",
        "\n",
        "# Change the column names in test_df to match the training data\n",
        "test_df.rename(columns={'id': 'Id', 'age': 'Age', 'sex': 'Sex'}, inplace=True)\n",
        "\n",
        "# Transform the test data\n",
        "X_test_processed = full_pipeline.transform(test_df)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_processed, y_train_balanced = smote.fit_resample(X_train_processed, y_train)\n",
        "\n",
        "# Create the CatBoost model\n",
        "catboost_model = CatBoostClassifier(\n",
        "    iterations=1000,\n",
        "    learning_rate=0.01,\n",
        "    depth=6,\n",
        "    loss_function='Logloss',\n",
        "    eval_metric='Accuracy',\n",
        "    verbose=100\n",
        ")\n",
        "\n",
        "# Fit the CatBoost model\n",
        "catboost_model.fit(X_train_processed, y_train_balanced, eval_set=(X_val_processed, y_val))\n",
        "\n",
        "# Perform cross-validation\n",
        "cv_scores = cross_val_score(catboost_model, X_train_processed, y_train_balanced, cv=5)\n",
        "print(f\"Cross-validation scores: {cv_scores}\")\n",
        "print(f\"Mean CV score: {cv_scores.mean():.4f} (+/- {cv_scores.std() * 2:.4f})\")\n",
        "\n",
        "# Make predictions on the validation set\n",
        "y_val_pred = catboost_model.predict(X_val_processed)\n",
        "\n",
        "# Calculate accuracy\n",
        "accuracy = accuracy_score(y_val, y_val_pred)\n",
        "print(f\"CatBoost Model Accuracy on validation set: {accuracy:.4f}\")\n",
        "\n",
        "# Print classification report\n",
        "print(\"\\nClassification Report:\")\n",
        "print(classification_report(y_val, y_val_pred))\n",
        "\n",
        "# Print confusion matrix\n",
        "print(\"\\nConfusion Matrix:\")\n",
        "print(confusion_matrix(y_val, y_val_pred))\n",
        "\n",
        "# Make predictions on the test set\n",
        "test_predictions = catboost_model.predict(X_test_processed)\n",
        "\n",
        "# Create a submission DataFrame\n",
        "submission = pd.DataFrame({\n",
        "    'ID': test_df['Id'],\n",
        "    'target': test_predictions\n",
        "})\n",
        "\n",
        "# Save the submission file\n",
        "submission.to_csv('catboost_submission.csv', index=False)\n",
        "print(\"Submission file 'catboost_submission.csv' has been created.\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4w501_wGLN4Z"
      },
      "source": [
        "# **CatBoost Model with Hyperparameter Tuning and Evaluation**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VmD11eKRLOf4",
        "outputId": "60034d67-0cc0-457e-a0fb-020896dec065"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0:\tlearn: 0.6768395\ttotal: 3.62ms\tremaining: 1.8s\n",
            "100:\tlearn: 0.2716152\ttotal: 320ms\tremaining: 1.26s\n",
            "200:\tlearn: 0.2034932\ttotal: 664ms\tremaining: 987ms\n",
            "300:\tlearn: 0.1713874\ttotal: 1.32s\tremaining: 876ms\n",
            "400:\tlearn: 0.1538565\ttotal: 2.11s\tremaining: 522ms\n",
            "499:\tlearn: 0.1404599\ttotal: 2.44s\tremaining: 0us\n",
            "0:\tlearn: 0.6767305\ttotal: 3.47ms\tremaining: 1.73s\n",
            "100:\tlearn: 0.2930501\ttotal: 352ms\tremaining: 1.39s\n",
            "200:\tlearn: 0.2471801\ttotal: 682ms\tremaining: 1.01s\n",
            "300:\tlearn: 0.2252278\ttotal: 1.03s\tremaining: 681ms\n",
            "400:\tlearn: 0.2149575\ttotal: 1.38s\tremaining: 342ms\n",
            "499:\tlearn: 0.2068477\ttotal: 1.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6769112\ttotal: 3.39ms\tremaining: 1.69s\n",
            "100:\tlearn: 0.2994007\ttotal: 323ms\tremaining: 1.27s\n",
            "200:\tlearn: 0.2527056\ttotal: 673ms\tremaining: 1s\n",
            "300:\tlearn: 0.2345149\ttotal: 1s\tremaining: 663ms\n",
            "400:\tlearn: 0.2236995\ttotal: 1.36s\tremaining: 335ms\n",
            "499:\tlearn: 0.2161691\ttotal: 1.68s\tremaining: 0us\n",
            "0:\tlearn: 0.6767756\ttotal: 3.45ms\tremaining: 1.72s\n",
            "100:\tlearn: 0.3022649\ttotal: 329ms\tremaining: 1.3s\n",
            "200:\tlearn: 0.2553940\ttotal: 664ms\tremaining: 987ms\n",
            "300:\tlearn: 0.2339026\ttotal: 993ms\tremaining: 657ms\n",
            "400:\tlearn: 0.2229097\ttotal: 1.33s\tremaining: 329ms\n",
            "499:\tlearn: 0.2152515\ttotal: 1.68s\tremaining: 0us\n",
            "0:\tlearn: 0.6769307\ttotal: 3.41ms\tremaining: 1.7s\n",
            "100:\tlearn: 0.3042785\ttotal: 324ms\tremaining: 1.28s\n",
            "200:\tlearn: 0.2582687\ttotal: 660ms\tremaining: 982ms\n",
            "300:\tlearn: 0.2368005\ttotal: 1.31s\tremaining: 870ms\n",
            "400:\tlearn: 0.2258895\ttotal: 1.85s\tremaining: 457ms\n",
            "499:\tlearn: 0.2176265\ttotal: 2.51s\tremaining: 0us\n",
            "0:\tlearn: 0.6168264\ttotal: 8.91ms\tremaining: 4.45s\n",
            "100:\tlearn: 0.1467576\ttotal: 844ms\tremaining: 3.33s\n",
            "200:\tlearn: 0.1022322\ttotal: 1.54s\tremaining: 2.29s\n",
            "300:\tlearn: 0.0823024\ttotal: 1.9s\tremaining: 1.25s\n",
            "400:\tlearn: 0.0701954\ttotal: 2.22s\tremaining: 548ms\n",
            "499:\tlearn: 0.0611351\ttotal: 2.57s\tremaining: 0us\n",
            "0:\tlearn: 0.6163064\ttotal: 3.43ms\tremaining: 1.71s\n",
            "100:\tlearn: 0.2055189\ttotal: 337ms\tremaining: 1.33s\n",
            "200:\tlearn: 0.1790796\ttotal: 687ms\tremaining: 1.02s\n",
            "300:\tlearn: 0.1596897\ttotal: 1.01s\tremaining: 671ms\n",
            "400:\tlearn: 0.1454243\ttotal: 1.35s\tremaining: 333ms\n",
            "499:\tlearn: 0.1350544\ttotal: 1.69s\tremaining: 0us\n",
            "0:\tlearn: 0.6171422\ttotal: 3.36ms\tremaining: 1.67s\n",
            "100:\tlearn: 0.2152861\ttotal: 342ms\tremaining: 1.35s\n",
            "200:\tlearn: 0.1922527\ttotal: 676ms\tremaining: 1s\n",
            "300:\tlearn: 0.1734693\ttotal: 1.02s\tremaining: 673ms\n",
            "400:\tlearn: 0.1601406\ttotal: 1.36s\tremaining: 335ms\n",
            "499:\tlearn: 0.1485858\ttotal: 1.68s\tremaining: 0us\n",
            "0:\tlearn: 0.6165125\ttotal: 3.42ms\tremaining: 1.7s\n",
            "100:\tlearn: 0.2166778\ttotal: 352ms\tremaining: 1.39s\n",
            "200:\tlearn: 0.1918982\ttotal: 678ms\tremaining: 1.01s\n",
            "300:\tlearn: 0.1718202\ttotal: 1.02s\tremaining: 672ms\n",
            "400:\tlearn: 0.1574752\ttotal: 1.36s\tremaining: 337ms\n",
            "499:\tlearn: 0.1456568\ttotal: 1.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6172565\ttotal: 3.45ms\tremaining: 1.72s\n",
            "100:\tlearn: 0.2156989\ttotal: 331ms\tremaining: 1.31s\n",
            "200:\tlearn: 0.1930081\ttotal: 690ms\tremaining: 1.03s\n",
            "300:\tlearn: 0.1731520\ttotal: 1.03s\tremaining: 679ms\n",
            "400:\tlearn: 0.1584711\ttotal: 1.36s\tremaining: 335ms\n",
            "499:\tlearn: 0.1474459\ttotal: 1.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6785358\ttotal: 3.61ms\tremaining: 1.8s\n",
            "100:\tlearn: 0.2850374\ttotal: 332ms\tremaining: 1.31s\n",
            "200:\tlearn: 0.2125648\ttotal: 680ms\tremaining: 1.01s\n",
            "300:\tlearn: 0.1800752\ttotal: 1.01s\tremaining: 671ms\n",
            "400:\tlearn: 0.1610291\ttotal: 1.34s\tremaining: 331ms\n",
            "499:\tlearn: 0.1474700\ttotal: 1.88s\tremaining: 0us\n",
            "0:\tlearn: 0.6784351\ttotal: 4.34ms\tremaining: 2.16s\n",
            "100:\tlearn: 0.3023584\ttotal: 850ms\tremaining: 3.36s\n",
            "200:\tlearn: 0.2512691\ttotal: 1.45s\tremaining: 2.16s\n",
            "300:\tlearn: 0.2299598\ttotal: 2.23s\tremaining: 1.47s\n",
            "400:\tlearn: 0.2184524\ttotal: 2.81s\tremaining: 694ms\n",
            "499:\tlearn: 0.2093141\ttotal: 3.14s\tremaining: 0us\n",
            "0:\tlearn: 0.6786057\ttotal: 3.9ms\tremaining: 1.94s\n",
            "100:\tlearn: 0.3079947\ttotal: 329ms\tremaining: 1.3s\n",
            "200:\tlearn: 0.2605032\ttotal: 682ms\tremaining: 1.01s\n",
            "300:\tlearn: 0.2385752\ttotal: 1.01s\tremaining: 671ms\n",
            "400:\tlearn: 0.2282977\ttotal: 1.34s\tremaining: 330ms\n",
            "499:\tlearn: 0.2207810\ttotal: 1.69s\tremaining: 0us\n",
            "0:\tlearn: 0.6784781\ttotal: 3.49ms\tremaining: 1.74s\n",
            "100:\tlearn: 0.3053879\ttotal: 322ms\tremaining: 1.27s\n",
            "200:\tlearn: 0.2590176\ttotal: 666ms\tremaining: 991ms\n",
            "300:\tlearn: 0.2381173\ttotal: 1.02s\tremaining: 677ms\n",
            "400:\tlearn: 0.2267030\ttotal: 1.34s\tremaining: 331ms\n",
            "499:\tlearn: 0.2196280\ttotal: 1.67s\tremaining: 0us\n",
            "0:\tlearn: 0.6786280\ttotal: 3.35ms\tremaining: 1.67s\n",
            "100:\tlearn: 0.3097787\ttotal: 320ms\tremaining: 1.26s\n",
            "200:\tlearn: 0.2612665\ttotal: 650ms\tremaining: 967ms\n",
            "300:\tlearn: 0.2426396\ttotal: 993ms\tremaining: 656ms\n",
            "400:\tlearn: 0.2304073\ttotal: 1.32s\tremaining: 326ms\n",
            "499:\tlearn: 0.2221995\ttotal: 1.65s\tremaining: 0us\n",
            "0:\tlearn: 0.6241369\ttotal: 3.57ms\tremaining: 1.78s\n",
            "100:\tlearn: 0.1464366\ttotal: 367ms\tremaining: 1.45s\n",
            "200:\tlearn: 0.1055696\ttotal: 708ms\tremaining: 1.05s\n",
            "300:\tlearn: 0.0854655\ttotal: 1.04s\tremaining: 691ms\n",
            "400:\tlearn: 0.0735692\ttotal: 1.4s\tremaining: 345ms\n",
            "499:\tlearn: 0.0652021\ttotal: 1.74s\tremaining: 0us\n",
            "0:\tlearn: 0.6236538\ttotal: 3.4ms\tremaining: 1.7s\n",
            "100:\tlearn: 0.2091302\ttotal: 337ms\tremaining: 1.33s\n",
            "200:\tlearn: 0.1842643\ttotal: 688ms\tremaining: 1.02s\n",
            "300:\tlearn: 0.1650527\ttotal: 1.02s\tremaining: 675ms\n",
            "400:\tlearn: 0.1506773\ttotal: 1.36s\tremaining: 335ms\n",
            "499:\tlearn: 0.1394139\ttotal: 1.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6244458\ttotal: 3.5ms\tremaining: 1.74s\n",
            "100:\tlearn: 0.2199537\ttotal: 407ms\tremaining: 1.61s\n",
            "200:\tlearn: 0.1954331\ttotal: 1.15s\tremaining: 1.71s\n",
            "300:\tlearn: 0.1761642\ttotal: 1.81s\tremaining: 1.2s\n",
            "400:\tlearn: 0.1632704\ttotal: 2.52s\tremaining: 622ms\n",
            "499:\tlearn: 0.1525670\ttotal: 3.18s\tremaining: 0us\n",
            "0:\tlearn: 0.6238464\ttotal: 12.9ms\tremaining: 6.41s\n",
            "100:\tlearn: 0.2205674\ttotal: 517ms\tremaining: 2.04s\n",
            "200:\tlearn: 0.1939861\ttotal: 859ms\tremaining: 1.28s\n",
            "300:\tlearn: 0.1743951\ttotal: 1.2s\tremaining: 790ms\n",
            "400:\tlearn: 0.1612890\ttotal: 1.54s\tremaining: 382ms\n",
            "499:\tlearn: 0.1502844\ttotal: 1.88s\tremaining: 0us\n",
            "0:\tlearn: 0.6245497\ttotal: 3.56ms\tremaining: 1.78s\n",
            "100:\tlearn: 0.2193424\ttotal: 332ms\tremaining: 1.31s\n",
            "200:\tlearn: 0.1964652\ttotal: 682ms\tremaining: 1.01s\n",
            "300:\tlearn: 0.1774236\ttotal: 1.01s\tremaining: 672ms\n",
            "400:\tlearn: 0.1636681\ttotal: 1.35s\tremaining: 333ms\n",
            "499:\tlearn: 0.1523233\ttotal: 1.69s\tremaining: 0us\n",
            "0:\tlearn: 0.6768395\ttotal: 3.51ms\tremaining: 3.51s\n",
            "100:\tlearn: 0.2716152\ttotal: 355ms\tremaining: 3.16s\n",
            "200:\tlearn: 0.2034932\ttotal: 704ms\tremaining: 2.8s\n",
            "300:\tlearn: 0.1713874\ttotal: 1.04s\tremaining: 2.41s\n",
            "400:\tlearn: 0.1538565\ttotal: 1.38s\tremaining: 2.06s\n",
            "500:\tlearn: 0.1403932\ttotal: 1.74s\tremaining: 1.73s\n",
            "600:\tlearn: 0.1311400\ttotal: 2.08s\tremaining: 1.38s\n",
            "700:\tlearn: 0.1238436\ttotal: 2.41s\tremaining: 1.03s\n",
            "800:\tlearn: 0.1175345\ttotal: 2.76s\tremaining: 686ms\n",
            "900:\tlearn: 0.1110811\ttotal: 3.09s\tremaining: 339ms\n",
            "999:\tlearn: 0.1039806\ttotal: 3.42s\tremaining: 0us\n",
            "0:\tlearn: 0.6767305\ttotal: 4.66ms\tremaining: 4.65s\n",
            "100:\tlearn: 0.2930501\ttotal: 344ms\tremaining: 3.06s\n",
            "200:\tlearn: 0.2471801\ttotal: 680ms\tremaining: 2.7s\n",
            "300:\tlearn: 0.2252278\ttotal: 1.01s\tremaining: 2.36s\n",
            "400:\tlearn: 0.2149575\ttotal: 1.38s\tremaining: 2.07s\n",
            "500:\tlearn: 0.2068217\ttotal: 1.72s\tremaining: 1.72s\n",
            "600:\tlearn: 0.1997505\ttotal: 2.06s\tremaining: 1.37s\n",
            "700:\tlearn: 0.1952194\ttotal: 2.4s\tremaining: 1.02s\n",
            "800:\tlearn: 0.1905585\ttotal: 2.73s\tremaining: 679ms\n",
            "900:\tlearn: 0.1851906\ttotal: 3.34s\tremaining: 367ms\n",
            "999:\tlearn: 0.1801336\ttotal: 4.04s\tremaining: 0us\n",
            "0:\tlearn: 0.6769112\ttotal: 7.75ms\tremaining: 7.74s\n",
            "100:\tlearn: 0.2994007\ttotal: 734ms\tremaining: 6.53s\n",
            "200:\tlearn: 0.2527056\ttotal: 1.59s\tremaining: 6.34s\n",
            "300:\tlearn: 0.2345149\ttotal: 2.13s\tremaining: 4.94s\n",
            "400:\tlearn: 0.2236995\ttotal: 2.47s\tremaining: 3.69s\n",
            "500:\tlearn: 0.2161252\ttotal: 2.79s\tremaining: 2.78s\n",
            "600:\tlearn: 0.2112796\ttotal: 3.13s\tremaining: 2.08s\n",
            "700:\tlearn: 0.2062114\ttotal: 3.47s\tremaining: 1.48s\n",
            "800:\tlearn: 0.2017202\ttotal: 3.8s\tremaining: 943ms\n",
            "900:\tlearn: 0.1970194\ttotal: 4.12s\tremaining: 453ms\n",
            "999:\tlearn: 0.1925832\ttotal: 4.47s\tremaining: 0us\n",
            "0:\tlearn: 0.6767756\ttotal: 3.22ms\tremaining: 3.22s\n",
            "100:\tlearn: 0.3022649\ttotal: 324ms\tremaining: 2.88s\n",
            "200:\tlearn: 0.2553940\ttotal: 662ms\tremaining: 2.63s\n",
            "300:\tlearn: 0.2339026\ttotal: 993ms\tremaining: 2.31s\n",
            "400:\tlearn: 0.2229097\ttotal: 1.31s\tremaining: 1.97s\n",
            "500:\tlearn: 0.2152154\ttotal: 1.66s\tremaining: 1.65s\n",
            "600:\tlearn: 0.2095077\ttotal: 1.99s\tremaining: 1.32s\n",
            "700:\tlearn: 0.2044608\ttotal: 2.32s\tremaining: 991ms\n",
            "800:\tlearn: 0.2000496\ttotal: 2.66s\tremaining: 661ms\n",
            "900:\tlearn: 0.1957846\ttotal: 2.99s\tremaining: 329ms\n",
            "999:\tlearn: 0.1904941\ttotal: 3.32s\tremaining: 0us\n",
            "0:\tlearn: 0.6769307\ttotal: 3.4ms\tremaining: 3.39s\n",
            "100:\tlearn: 0.3042785\ttotal: 341ms\tremaining: 3.04s\n",
            "200:\tlearn: 0.2582687\ttotal: 668ms\tremaining: 2.65s\n",
            "300:\tlearn: 0.2368005\ttotal: 1s\tremaining: 2.33s\n",
            "400:\tlearn: 0.2258895\ttotal: 1.36s\tremaining: 2.03s\n",
            "500:\tlearn: 0.2175847\ttotal: 1.68s\tremaining: 1.68s\n",
            "600:\tlearn: 0.2118723\ttotal: 2.02s\tremaining: 1.34s\n",
            "700:\tlearn: 0.2071846\ttotal: 2.38s\tremaining: 1.02s\n",
            "800:\tlearn: 0.2028292\ttotal: 2.71s\tremaining: 674ms\n",
            "900:\tlearn: 0.1976409\ttotal: 3.04s\tremaining: 334ms\n",
            "999:\tlearn: 0.1926216\ttotal: 3.39s\tremaining: 0us\n",
            "0:\tlearn: 0.6168264\ttotal: 3.24ms\tremaining: 3.24s\n",
            "100:\tlearn: 0.1467576\ttotal: 384ms\tremaining: 3.41s\n",
            "200:\tlearn: 0.1022322\ttotal: 988ms\tremaining: 3.93s\n",
            "300:\tlearn: 0.0823024\ttotal: 1.77s\tremaining: 4.12s\n",
            "400:\tlearn: 0.0701954\ttotal: 2.42s\tremaining: 3.62s\n",
            "500:\tlearn: 0.0610556\ttotal: 3.12s\tremaining: 3.1s\n",
            "600:\tlearn: 0.0542644\ttotal: 3.65s\tremaining: 2.42s\n",
            "700:\tlearn: 0.0476720\ttotal: 3.99s\tremaining: 1.7s\n",
            "800:\tlearn: 0.0424358\ttotal: 4.32s\tremaining: 1.07s\n",
            "900:\tlearn: 0.0377837\ttotal: 4.65s\tremaining: 511ms\n",
            "999:\tlearn: 0.0338997\ttotal: 5s\tremaining: 0us\n",
            "0:\tlearn: 0.6163064\ttotal: 4.24ms\tremaining: 4.23s\n",
            "100:\tlearn: 0.2055189\ttotal: 335ms\tremaining: 2.98s\n",
            "200:\tlearn: 0.1790796\ttotal: 666ms\tremaining: 2.65s\n",
            "300:\tlearn: 0.1596897\ttotal: 1.01s\tremaining: 2.35s\n",
            "400:\tlearn: 0.1454243\ttotal: 1.33s\tremaining: 1.99s\n",
            "500:\tlearn: 0.1349360\ttotal: 1.67s\tremaining: 1.66s\n",
            "600:\tlearn: 0.1245749\ttotal: 2.02s\tremaining: 1.34s\n",
            "700:\tlearn: 0.1156814\ttotal: 2.34s\tremaining: 1000ms\n",
            "800:\tlearn: 0.1076715\ttotal: 2.67s\tremaining: 664ms\n",
            "900:\tlearn: 0.1009637\ttotal: 3.02s\tremaining: 331ms\n",
            "999:\tlearn: 0.0946598\ttotal: 3.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6171422\ttotal: 3.33ms\tremaining: 3.33s\n",
            "100:\tlearn: 0.2152861\ttotal: 910ms\tremaining: 8.1s\n",
            "200:\tlearn: 0.1922527\ttotal: 2.29s\tremaining: 9.1s\n",
            "300:\tlearn: 0.1734693\ttotal: 2.64s\tremaining: 6.13s\n",
            "400:\tlearn: 0.1601406\ttotal: 2.96s\tremaining: 4.43s\n",
            "500:\tlearn: 0.1484847\ttotal: 3.31s\tremaining: 3.29s\n",
            "600:\tlearn: 0.1385020\ttotal: 3.96s\tremaining: 2.63s\n",
            "700:\tlearn: 0.1292086\ttotal: 4.34s\tremaining: 1.85s\n",
            "800:\tlearn: 0.1207431\ttotal: 4.76s\tremaining: 1.18s\n",
            "900:\tlearn: 0.1132883\ttotal: 5.55s\tremaining: 610ms\n",
            "999:\tlearn: 0.1066286\ttotal: 6.55s\tremaining: 0us\n",
            "0:\tlearn: 0.6165125\ttotal: 6.79ms\tremaining: 6.78s\n",
            "100:\tlearn: 0.2166778\ttotal: 991ms\tremaining: 8.82s\n",
            "200:\tlearn: 0.1918982\ttotal: 2.07s\tremaining: 8.21s\n",
            "300:\tlearn: 0.1718202\ttotal: 3.04s\tremaining: 7.05s\n",
            "400:\tlearn: 0.1574752\ttotal: 5.32s\tremaining: 7.94s\n",
            "500:\tlearn: 0.1455309\ttotal: 5.68s\tremaining: 5.66s\n",
            "600:\tlearn: 0.1353873\ttotal: 6.01s\tremaining: 3.99s\n",
            "700:\tlearn: 0.1259901\ttotal: 6.34s\tremaining: 2.7s\n",
            "800:\tlearn: 0.1174869\ttotal: 6.68s\tremaining: 1.66s\n",
            "900:\tlearn: 0.1096685\ttotal: 7.01s\tremaining: 771ms\n",
            "999:\tlearn: 0.1027028\ttotal: 7.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6172565\ttotal: 3.38ms\tremaining: 3.38s\n",
            "100:\tlearn: 0.2156989\ttotal: 353ms\tremaining: 3.14s\n",
            "200:\tlearn: 0.1930081\ttotal: 686ms\tremaining: 2.73s\n",
            "300:\tlearn: 0.1731520\ttotal: 1.02s\tremaining: 2.37s\n",
            "400:\tlearn: 0.1584711\ttotal: 1.36s\tremaining: 2.04s\n",
            "500:\tlearn: 0.1473446\ttotal: 1.69s\tremaining: 1.69s\n",
            "600:\tlearn: 0.1374751\ttotal: 2.03s\tremaining: 1.35s\n",
            "700:\tlearn: 0.1284635\ttotal: 2.37s\tremaining: 1.01s\n",
            "800:\tlearn: 0.1199039\ttotal: 2.7s\tremaining: 672ms\n",
            "900:\tlearn: 0.1125909\ttotal: 3.04s\tremaining: 334ms\n",
            "999:\tlearn: 0.1057958\ttotal: 3.38s\tremaining: 0us\n",
            "0:\tlearn: 0.6785358\ttotal: 3.44ms\tremaining: 3.43s\n",
            "100:\tlearn: 0.2850374\ttotal: 324ms\tremaining: 2.88s\n",
            "200:\tlearn: 0.2125648\ttotal: 650ms\tremaining: 2.58s\n",
            "300:\tlearn: 0.1800752\ttotal: 990ms\tremaining: 2.3s\n",
            "400:\tlearn: 0.1610291\ttotal: 1.32s\tremaining: 1.97s\n",
            "500:\tlearn: 0.1474350\ttotal: 1.65s\tremaining: 1.65s\n",
            "600:\tlearn: 0.1381043\ttotal: 2s\tremaining: 1.33s\n",
            "700:\tlearn: 0.1297824\ttotal: 2.34s\tremaining: 998ms\n",
            "800:\tlearn: 0.1218383\ttotal: 2.7s\tremaining: 670ms\n",
            "900:\tlearn: 0.1145736\ttotal: 3.03s\tremaining: 332ms\n",
            "999:\tlearn: 0.1075639\ttotal: 3.36s\tremaining: 0us\n",
            "0:\tlearn: 0.6784351\ttotal: 6.26ms\tremaining: 6.26s\n",
            "100:\tlearn: 0.3023584\ttotal: 784ms\tremaining: 6.98s\n",
            "200:\tlearn: 0.2512691\ttotal: 1.54s\tremaining: 6.12s\n",
            "300:\tlearn: 0.2299598\ttotal: 2.32s\tremaining: 5.39s\n",
            "400:\tlearn: 0.2184524\ttotal: 3.15s\tremaining: 4.7s\n",
            "500:\tlearn: 0.2091627\ttotal: 3.63s\tremaining: 3.61s\n",
            "600:\tlearn: 0.2031477\ttotal: 3.96s\tremaining: 2.63s\n",
            "700:\tlearn: 0.1976611\ttotal: 4.29s\tremaining: 1.83s\n",
            "800:\tlearn: 0.1927825\ttotal: 4.64s\tremaining: 1.15s\n",
            "900:\tlearn: 0.1881974\ttotal: 4.98s\tremaining: 547ms\n",
            "999:\tlearn: 0.1833175\ttotal: 5.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6786057\ttotal: 4.34ms\tremaining: 4.34s\n",
            "100:\tlearn: 0.3079947\ttotal: 323ms\tremaining: 2.87s\n",
            "200:\tlearn: 0.2605032\ttotal: 652ms\tremaining: 2.59s\n",
            "300:\tlearn: 0.2385752\ttotal: 995ms\tremaining: 2.31s\n",
            "400:\tlearn: 0.2282977\ttotal: 1.32s\tremaining: 1.98s\n",
            "500:\tlearn: 0.2206457\ttotal: 1.65s\tremaining: 1.64s\n",
            "600:\tlearn: 0.2140130\ttotal: 1.99s\tremaining: 1.32s\n",
            "700:\tlearn: 0.2092533\ttotal: 2.32s\tremaining: 989ms\n",
            "800:\tlearn: 0.2046128\ttotal: 2.65s\tremaining: 658ms\n",
            "900:\tlearn: 0.2001008\ttotal: 2.99s\tremaining: 328ms\n",
            "999:\tlearn: 0.1953793\ttotal: 3.31s\tremaining: 0us\n",
            "0:\tlearn: 0.6784781\ttotal: 3.4ms\tremaining: 3.4s\n",
            "100:\tlearn: 0.3053879\ttotal: 326ms\tremaining: 2.9s\n",
            "200:\tlearn: 0.2590176\ttotal: 673ms\tremaining: 2.67s\n",
            "300:\tlearn: 0.2381173\ttotal: 1s\tremaining: 2.33s\n",
            "400:\tlearn: 0.2267030\ttotal: 1.33s\tremaining: 1.99s\n",
            "500:\tlearn: 0.2195111\ttotal: 1.72s\tremaining: 1.72s\n",
            "600:\tlearn: 0.2137475\ttotal: 2.06s\tremaining: 1.37s\n",
            "700:\tlearn: 0.2083424\ttotal: 2.38s\tremaining: 1.02s\n",
            "800:\tlearn: 0.2034713\ttotal: 2.73s\tremaining: 679ms\n",
            "900:\tlearn: 0.1988089\ttotal: 3.06s\tremaining: 336ms\n",
            "999:\tlearn: 0.1940208\ttotal: 3.38s\tremaining: 0us\n",
            "0:\tlearn: 0.6786280\ttotal: 3.36ms\tremaining: 3.36s\n",
            "100:\tlearn: 0.3097787\ttotal: 348ms\tremaining: 3.1s\n",
            "200:\tlearn: 0.2612665\ttotal: 692ms\tremaining: 2.75s\n",
            "300:\tlearn: 0.2426396\ttotal: 1.13s\tremaining: 2.62s\n",
            "400:\tlearn: 0.2304073\ttotal: 1.93s\tremaining: 2.88s\n",
            "500:\tlearn: 0.2221605\ttotal: 2.67s\tremaining: 2.66s\n",
            "600:\tlearn: 0.2161092\ttotal: 3.42s\tremaining: 2.27s\n",
            "700:\tlearn: 0.2106787\ttotal: 4.24s\tremaining: 1.81s\n",
            "800:\tlearn: 0.2063875\ttotal: 4.7s\tremaining: 1.17s\n",
            "900:\tlearn: 0.2008299\ttotal: 5.03s\tremaining: 552ms\n",
            "999:\tlearn: 0.1954694\ttotal: 5.39s\tremaining: 0us\n",
            "0:\tlearn: 0.6241369\ttotal: 3.45ms\tremaining: 3.45s\n",
            "100:\tlearn: 0.1464366\ttotal: 338ms\tremaining: 3.01s\n",
            "200:\tlearn: 0.1055696\ttotal: 675ms\tremaining: 2.68s\n",
            "300:\tlearn: 0.0854655\ttotal: 1.02s\tremaining: 2.38s\n",
            "400:\tlearn: 0.0735692\ttotal: 1.35s\tremaining: 2.02s\n",
            "500:\tlearn: 0.0651612\ttotal: 1.68s\tremaining: 1.68s\n",
            "600:\tlearn: 0.0580071\ttotal: 2.04s\tremaining: 1.35s\n",
            "700:\tlearn: 0.0515631\ttotal: 2.36s\tremaining: 1.01s\n",
            "800:\tlearn: 0.0462228\ttotal: 2.71s\tremaining: 674ms\n",
            "900:\tlearn: 0.0419313\ttotal: 3.06s\tremaining: 336ms\n",
            "999:\tlearn: 0.0383002\ttotal: 3.38s\tremaining: 0us\n",
            "0:\tlearn: 0.6236538\ttotal: 3.77ms\tremaining: 3.76s\n",
            "100:\tlearn: 0.2091302\ttotal: 335ms\tremaining: 2.98s\n",
            "200:\tlearn: 0.1842643\ttotal: 671ms\tremaining: 2.67s\n",
            "300:\tlearn: 0.1650527\ttotal: 998ms\tremaining: 2.32s\n",
            "400:\tlearn: 0.1506773\ttotal: 1.33s\tremaining: 1.99s\n",
            "500:\tlearn: 0.1393380\ttotal: 1.68s\tremaining: 1.67s\n",
            "600:\tlearn: 0.1295097\ttotal: 2s\tremaining: 1.33s\n",
            "700:\tlearn: 0.1213831\ttotal: 2.34s\tremaining: 998ms\n",
            "800:\tlearn: 0.1139029\ttotal: 2.69s\tremaining: 668ms\n",
            "900:\tlearn: 0.1071460\ttotal: 3.01s\tremaining: 331ms\n",
            "999:\tlearn: 0.1009566\ttotal: 3.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6244458\ttotal: 3.33ms\tremaining: 3.33s\n",
            "100:\tlearn: 0.2199537\ttotal: 329ms\tremaining: 2.93s\n",
            "200:\tlearn: 0.1954331\ttotal: 657ms\tremaining: 2.61s\n",
            "300:\tlearn: 0.1761642\ttotal: 1s\tremaining: 2.33s\n",
            "400:\tlearn: 0.1632704\ttotal: 1.32s\tremaining: 1.98s\n",
            "500:\tlearn: 0.1523811\ttotal: 1.66s\tremaining: 1.65s\n",
            "600:\tlearn: 0.1425696\ttotal: 2.23s\tremaining: 1.48s\n",
            "700:\tlearn: 0.1335605\ttotal: 2.92s\tremaining: 1.25s\n",
            "800:\tlearn: 0.1255692\ttotal: 3.71s\tremaining: 922ms\n",
            "900:\tlearn: 0.1182623\ttotal: 4.51s\tremaining: 495ms\n",
            "999:\tlearn: 0.1118726\ttotal: 5.13s\tremaining: 0us\n",
            "0:\tlearn: 0.6238464\ttotal: 3.25ms\tremaining: 3.25s\n",
            "100:\tlearn: 0.2205674\ttotal: 339ms\tremaining: 3.02s\n",
            "200:\tlearn: 0.1939861\ttotal: 674ms\tremaining: 2.68s\n",
            "300:\tlearn: 0.1743951\ttotal: 1.03s\tremaining: 2.4s\n",
            "400:\tlearn: 0.1612890\ttotal: 1.38s\tremaining: 2.06s\n",
            "500:\tlearn: 0.1501797\ttotal: 1.71s\tremaining: 1.7s\n",
            "600:\tlearn: 0.1406122\ttotal: 2.06s\tremaining: 1.36s\n",
            "700:\tlearn: 0.1313474\ttotal: 2.4s\tremaining: 1.02s\n",
            "800:\tlearn: 0.1236900\ttotal: 2.73s\tremaining: 678ms\n",
            "900:\tlearn: 0.1160800\ttotal: 3.08s\tremaining: 338ms\n",
            "999:\tlearn: 0.1089241\ttotal: 3.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6245497\ttotal: 3.35ms\tremaining: 3.34s\n",
            "100:\tlearn: 0.2193424\ttotal: 333ms\tremaining: 2.96s\n",
            "200:\tlearn: 0.1964652\ttotal: 687ms\tremaining: 2.73s\n",
            "300:\tlearn: 0.1774236\ttotal: 1.01s\tremaining: 2.35s\n",
            "400:\tlearn: 0.1636681\ttotal: 1.34s\tremaining: 2.01s\n",
            "500:\tlearn: 0.1520808\ttotal: 1.69s\tremaining: 1.68s\n",
            "600:\tlearn: 0.1422436\ttotal: 2.02s\tremaining: 1.34s\n",
            "700:\tlearn: 0.1332921\ttotal: 2.35s\tremaining: 1s\n",
            "800:\tlearn: 0.1252974\ttotal: 2.7s\tremaining: 671ms\n",
            "900:\tlearn: 0.1180055\ttotal: 3.03s\tremaining: 333ms\n",
            "999:\tlearn: 0.1111530\ttotal: 3.36s\tremaining: 0us\n",
            "0:\tlearn: 0.6786722\ttotal: 5.5ms\tremaining: 2.74s\n",
            "100:\tlearn: 0.2537425\ttotal: 534ms\tremaining: 2.11s\n",
            "200:\tlearn: 0.1817053\ttotal: 1.07s\tremaining: 1.59s\n",
            "300:\tlearn: 0.1529348\ttotal: 1.62s\tremaining: 1.07s\n",
            "400:\tlearn: 0.1359096\ttotal: 2.19s\tremaining: 542ms\n",
            "499:\tlearn: 0.1238373\ttotal: 2.87s\tremaining: 0us\n",
            "0:\tlearn: 0.6784807\ttotal: 9.2ms\tremaining: 4.59s\n",
            "100:\tlearn: 0.2852376\ttotal: 1.15s\tremaining: 4.54s\n",
            "200:\tlearn: 0.2333216\ttotal: 2.14s\tremaining: 3.19s\n",
            "300:\tlearn: 0.2127075\ttotal: 3.37s\tremaining: 2.23s\n",
            "400:\tlearn: 0.1987462\ttotal: 3.9s\tremaining: 962ms\n",
            "499:\tlearn: 0.1876897\ttotal: 4.44s\tremaining: 0us\n",
            "0:\tlearn: 0.6787581\ttotal: 5.59ms\tremaining: 2.79s\n",
            "100:\tlearn: 0.2893615\ttotal: 515ms\tremaining: 2.03s\n",
            "200:\tlearn: 0.2429830\ttotal: 1.06s\tremaining: 1.57s\n",
            "300:\tlearn: 0.2217212\ttotal: 1.6s\tremaining: 1.06s\n",
            "400:\tlearn: 0.2099069\ttotal: 2.13s\tremaining: 526ms\n",
            "499:\tlearn: 0.1996944\ttotal: 2.67s\tremaining: 0us\n",
            "0:\tlearn: 0.6785359\ttotal: 5.55ms\tremaining: 2.77s\n",
            "100:\tlearn: 0.2882058\ttotal: 508ms\tremaining: 2.01s\n",
            "200:\tlearn: 0.2437079\ttotal: 1.06s\tremaining: 1.57s\n",
            "300:\tlearn: 0.2221950\ttotal: 1.58s\tremaining: 1.04s\n",
            "400:\tlearn: 0.2084851\ttotal: 2.13s\tremaining: 526ms\n",
            "499:\tlearn: 0.1981636\ttotal: 2.65s\tremaining: 0us\n",
            "0:\tlearn: 0.6787666\ttotal: 5.53ms\tremaining: 2.76s\n",
            "100:\tlearn: 0.2911528\ttotal: 522ms\tremaining: 2.06s\n",
            "200:\tlearn: 0.2446408\ttotal: 1.04s\tremaining: 1.55s\n",
            "300:\tlearn: 0.2235262\ttotal: 1.59s\tremaining: 1.05s\n",
            "400:\tlearn: 0.2100603\ttotal: 2.12s\tremaining: 524ms\n",
            "499:\tlearn: 0.2002925\ttotal: 2.66s\tremaining: 0us\n",
            "0:\tlearn: 0.6247335\ttotal: 5.68ms\tremaining: 2.83s\n",
            "100:\tlearn: 0.1224462\ttotal: 877ms\tremaining: 3.46s\n",
            "200:\tlearn: 0.0821964\ttotal: 2.18s\tremaining: 3.24s\n",
            "300:\tlearn: 0.0590708\ttotal: 3.47s\tremaining: 2.29s\n",
            "400:\tlearn: 0.0437326\ttotal: 4.22s\tremaining: 1.04s\n",
            "499:\tlearn: 0.0338875\ttotal: 4.76s\tremaining: 0us\n",
            "0:\tlearn: 0.6238663\ttotal: 5.47ms\tremaining: 2.73s\n",
            "100:\tlearn: 0.1872826\ttotal: 545ms\tremaining: 2.15s\n",
            "200:\tlearn: 0.1469977\ttotal: 1.1s\tremaining: 1.63s\n",
            "300:\tlearn: 0.1183425\ttotal: 1.63s\tremaining: 1.07s\n",
            "400:\tlearn: 0.0969070\ttotal: 2.19s\tremaining: 541ms\n",
            "499:\tlearn: 0.0815185\ttotal: 2.74s\tremaining: 0us\n",
            "0:\tlearn: 0.6251516\ttotal: 5.57ms\tremaining: 2.78s\n",
            "100:\tlearn: 0.1986659\ttotal: 548ms\tremaining: 2.16s\n",
            "200:\tlearn: 0.1590310\ttotal: 1.1s\tremaining: 1.64s\n",
            "300:\tlearn: 0.1299918\ttotal: 1.64s\tremaining: 1.08s\n",
            "400:\tlearn: 0.1084451\ttotal: 2.17s\tremaining: 537ms\n",
            "499:\tlearn: 0.0914640\ttotal: 2.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6241247\ttotal: 5.5ms\tremaining: 2.74s\n",
            "100:\tlearn: 0.1978606\ttotal: 552ms\tremaining: 2.18s\n",
            "200:\tlearn: 0.1569792\ttotal: 1.08s\tremaining: 1.6s\n",
            "300:\tlearn: 0.1270850\ttotal: 1.64s\tremaining: 1.08s\n",
            "400:\tlearn: 0.1055732\ttotal: 2.17s\tremaining: 535ms\n",
            "499:\tlearn: 0.0881273\ttotal: 2.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6251605\ttotal: 5.21ms\tremaining: 2.6s\n",
            "100:\tlearn: 0.1999188\ttotal: 679ms\tremaining: 2.68s\n",
            "200:\tlearn: 0.1604473\ttotal: 1.96s\tremaining: 2.92s\n",
            "300:\tlearn: 0.1295998\ttotal: 3.26s\tremaining: 2.15s\n",
            "400:\tlearn: 0.1082992\ttotal: 4.18s\tremaining: 1.03s\n",
            "499:\tlearn: 0.0909704\ttotal: 4.73s\tremaining: 0us\n",
            "0:\tlearn: 0.6804556\ttotal: 5.77ms\tremaining: 2.88s\n",
            "100:\tlearn: 0.2644003\ttotal: 529ms\tremaining: 2.09s\n",
            "200:\tlearn: 0.1892531\ttotal: 1.1s\tremaining: 1.64s\n",
            "300:\tlearn: 0.1605806\ttotal: 1.63s\tremaining: 1.08s\n",
            "400:\tlearn: 0.1420032\ttotal: 2.19s\tremaining: 541ms\n",
            "499:\tlearn: 0.1301085\ttotal: 2.76s\tremaining: 0us\n",
            "0:\tlearn: 0.6802679\ttotal: 5.7ms\tremaining: 2.85s\n",
            "100:\tlearn: 0.2922426\ttotal: 519ms\tremaining: 2.05s\n",
            "200:\tlearn: 0.2404537\ttotal: 1.07s\tremaining: 1.59s\n",
            "300:\tlearn: 0.2183429\ttotal: 1.6s\tremaining: 1.06s\n",
            "400:\tlearn: 0.2052006\ttotal: 2.16s\tremaining: 533ms\n",
            "499:\tlearn: 0.1951121\ttotal: 2.68s\tremaining: 0us\n",
            "0:\tlearn: 0.6805379\ttotal: 5.28ms\tremaining: 2.63s\n",
            "100:\tlearn: 0.2961043\ttotal: 517ms\tremaining: 2.04s\n",
            "200:\tlearn: 0.2483248\ttotal: 1.04s\tremaining: 1.55s\n",
            "300:\tlearn: 0.2284318\ttotal: 1.61s\tremaining: 1.06s\n",
            "400:\tlearn: 0.2152916\ttotal: 2.15s\tremaining: 530ms\n",
            "499:\tlearn: 0.2054652\ttotal: 2.67s\tremaining: 0us\n",
            "0:\tlearn: 0.6803194\ttotal: 5.56ms\tremaining: 2.77s\n",
            "100:\tlearn: 0.2977526\ttotal: 643ms\tremaining: 2.54s\n",
            "200:\tlearn: 0.2486288\ttotal: 1.95s\tremaining: 2.9s\n",
            "300:\tlearn: 0.2278670\ttotal: 3.17s\tremaining: 2.1s\n",
            "400:\tlearn: 0.2146822\ttotal: 4.16s\tremaining: 1.03s\n",
            "499:\tlearn: 0.2046170\ttotal: 4.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6805445\ttotal: 5.73ms\tremaining: 2.86s\n",
            "100:\tlearn: 0.2996263\ttotal: 519ms\tremaining: 2.05s\n",
            "200:\tlearn: 0.2503341\ttotal: 1.91s\tremaining: 2.84s\n",
            "300:\tlearn: 0.2306920\ttotal: 3.25s\tremaining: 2.15s\n",
            "400:\tlearn: 0.2170860\ttotal: 3.79s\tremaining: 936ms\n",
            "499:\tlearn: 0.2066214\ttotal: 4.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6326151\ttotal: 5.73ms\tremaining: 2.86s\n",
            "100:\tlearn: 0.1305059\ttotal: 565ms\tremaining: 2.23s\n",
            "200:\tlearn: 0.0883408\ttotal: 1.11s\tremaining: 1.65s\n",
            "300:\tlearn: 0.0644129\ttotal: 1.68s\tremaining: 1.11s\n",
            "400:\tlearn: 0.0495191\ttotal: 2.21s\tremaining: 547ms\n",
            "499:\tlearn: 0.0397330\ttotal: 2.77s\tremaining: 0us\n",
            "0:\tlearn: 0.6317614\ttotal: 5.49ms\tremaining: 2.74s\n",
            "100:\tlearn: 0.1952039\ttotal: 564ms\tremaining: 2.23s\n",
            "200:\tlearn: 0.1574880\ttotal: 1.1s\tremaining: 1.64s\n",
            "300:\tlearn: 0.1298218\ttotal: 1.72s\tremaining: 1.14s\n",
            "400:\tlearn: 0.1093238\ttotal: 2.85s\tremaining: 702ms\n",
            "499:\tlearn: 0.0927001\ttotal: 4s\tremaining: 0us\n",
            "0:\tlearn: 0.6330237\ttotal: 14.9ms\tremaining: 7.41s\n",
            "100:\tlearn: 0.2057656\ttotal: 1.14s\tremaining: 4.51s\n",
            "200:\tlearn: 0.1687522\ttotal: 1.68s\tremaining: 2.5s\n",
            "300:\tlearn: 0.1399146\ttotal: 2.21s\tremaining: 1.46s\n",
            "400:\tlearn: 0.1189365\ttotal: 2.76s\tremaining: 682ms\n",
            "499:\tlearn: 0.1014685\ttotal: 3.29s\tremaining: 0us\n",
            "0:\tlearn: 0.6320028\ttotal: 5.66ms\tremaining: 2.82s\n",
            "100:\tlearn: 0.2053565\ttotal: 607ms\tremaining: 2.4s\n",
            "200:\tlearn: 0.1667900\ttotal: 1.15s\tremaining: 1.71s\n",
            "300:\tlearn: 0.1375627\ttotal: 1.71s\tremaining: 1.13s\n",
            "400:\tlearn: 0.1170982\ttotal: 2.28s\tremaining: 563ms\n",
            "499:\tlearn: 0.1003344\ttotal: 2.81s\tremaining: 0us\n",
            "0:\tlearn: 0.6330252\ttotal: 5.48ms\tremaining: 2.74s\n",
            "100:\tlearn: 0.2079586\ttotal: 561ms\tremaining: 2.21s\n",
            "200:\tlearn: 0.1705688\ttotal: 1.09s\tremaining: 1.62s\n",
            "300:\tlearn: 0.1410444\ttotal: 1.65s\tremaining: 1.09s\n",
            "400:\tlearn: 0.1179807\ttotal: 2.18s\tremaining: 538ms\n",
            "499:\tlearn: 0.1006520\ttotal: 2.73s\tremaining: 0us\n",
            "0:\tlearn: 0.6786722\ttotal: 5.79ms\tremaining: 5.79s\n",
            "100:\tlearn: 0.2537425\ttotal: 551ms\tremaining: 4.9s\n",
            "200:\tlearn: 0.1817053\ttotal: 1.08s\tremaining: 4.3s\n",
            "300:\tlearn: 0.1529348\ttotal: 1.66s\tremaining: 3.85s\n",
            "400:\tlearn: 0.1359096\ttotal: 2.85s\tremaining: 4.25s\n",
            "500:\tlearn: 0.1237691\ttotal: 4.08s\tremaining: 4.06s\n",
            "600:\tlearn: 0.1135050\ttotal: 5.11s\tremaining: 3.39s\n",
            "700:\tlearn: 0.1046083\ttotal: 5.82s\tremaining: 2.48s\n",
            "800:\tlearn: 0.0964807\ttotal: 6.35s\tremaining: 1.58s\n",
            "900:\tlearn: 0.0894195\ttotal: 6.89s\tremaining: 757ms\n",
            "999:\tlearn: 0.0821917\ttotal: 7.42s\tremaining: 0us\n",
            "0:\tlearn: 0.6784807\ttotal: 5.25ms\tremaining: 5.24s\n",
            "100:\tlearn: 0.2852376\ttotal: 513ms\tremaining: 4.57s\n",
            "200:\tlearn: 0.2333216\ttotal: 1.03s\tremaining: 4.1s\n",
            "300:\tlearn: 0.2127075\ttotal: 1.56s\tremaining: 3.64s\n",
            "400:\tlearn: 0.1987462\ttotal: 2.1s\tremaining: 3.14s\n",
            "500:\tlearn: 0.1875774\ttotal: 2.66s\tremaining: 2.65s\n",
            "600:\tlearn: 0.1785658\ttotal: 3.2s\tremaining: 2.12s\n",
            "700:\tlearn: 0.1703253\ttotal: 3.74s\tremaining: 1.59s\n",
            "800:\tlearn: 0.1624888\ttotal: 4.29s\tremaining: 1.07s\n",
            "900:\tlearn: 0.1551775\ttotal: 4.84s\tremaining: 532ms\n",
            "999:\tlearn: 0.1479056\ttotal: 5.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6787581\ttotal: 5.62ms\tremaining: 5.61s\n",
            "100:\tlearn: 0.2893615\ttotal: 527ms\tremaining: 4.69s\n",
            "200:\tlearn: 0.2429830\ttotal: 1.08s\tremaining: 4.29s\n",
            "300:\tlearn: 0.2217212\ttotal: 1.62s\tremaining: 3.77s\n",
            "400:\tlearn: 0.2099069\ttotal: 2.25s\tremaining: 3.36s\n",
            "500:\tlearn: 0.1995690\ttotal: 3.55s\tremaining: 3.54s\n",
            "600:\tlearn: 0.1908860\ttotal: 4.65s\tremaining: 3.09s\n",
            "700:\tlearn: 0.1824731\ttotal: 5.67s\tremaining: 2.42s\n",
            "800:\tlearn: 0.1748331\ttotal: 6.22s\tremaining: 1.54s\n",
            "900:\tlearn: 0.1669596\ttotal: 6.75s\tremaining: 742ms\n",
            "999:\tlearn: 0.1596641\ttotal: 7.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6785359\ttotal: 5.49ms\tremaining: 5.49s\n",
            "100:\tlearn: 0.2882058\ttotal: 517ms\tremaining: 4.6s\n",
            "200:\tlearn: 0.2437079\ttotal: 1.05s\tremaining: 4.16s\n",
            "300:\tlearn: 0.2221950\ttotal: 1.59s\tremaining: 3.69s\n",
            "400:\tlearn: 0.2084851\ttotal: 2.13s\tremaining: 3.19s\n",
            "500:\tlearn: 0.1980283\ttotal: 2.68s\tremaining: 2.67s\n",
            "600:\tlearn: 0.1891770\ttotal: 3.22s\tremaining: 2.14s\n",
            "700:\tlearn: 0.1812339\ttotal: 3.82s\tremaining: 1.63s\n",
            "800:\tlearn: 0.1736283\ttotal: 4.36s\tremaining: 1.08s\n",
            "900:\tlearn: 0.1657437\ttotal: 4.91s\tremaining: 540ms\n",
            "999:\tlearn: 0.1587883\ttotal: 5.44s\tremaining: 0us\n",
            "0:\tlearn: 0.6787666\ttotal: 7.74ms\tremaining: 7.73s\n",
            "100:\tlearn: 0.2911528\ttotal: 540ms\tremaining: 4.81s\n",
            "200:\tlearn: 0.2446408\ttotal: 1.08s\tremaining: 4.29s\n",
            "300:\tlearn: 0.2235262\ttotal: 1.61s\tremaining: 3.75s\n",
            "400:\tlearn: 0.2100603\ttotal: 2.16s\tremaining: 3.23s\n",
            "500:\tlearn: 0.2001963\ttotal: 3.29s\tremaining: 3.28s\n",
            "600:\tlearn: 0.1907666\ttotal: 4.63s\tremaining: 3.07s\n",
            "700:\tlearn: 0.1828722\ttotal: 5.84s\tremaining: 2.49s\n",
            "800:\tlearn: 0.1748580\ttotal: 6.4s\tremaining: 1.59s\n",
            "900:\tlearn: 0.1672650\ttotal: 6.93s\tremaining: 762ms\n",
            "999:\tlearn: 0.1593780\ttotal: 7.47s\tremaining: 0us\n",
            "0:\tlearn: 0.6247335\ttotal: 5.72ms\tremaining: 5.71s\n",
            "100:\tlearn: 0.1224462\ttotal: 548ms\tremaining: 4.88s\n",
            "200:\tlearn: 0.0821964\ttotal: 1.1s\tremaining: 4.38s\n",
            "300:\tlearn: 0.0590708\ttotal: 1.65s\tremaining: 3.83s\n",
            "400:\tlearn: 0.0437326\ttotal: 2.2s\tremaining: 3.29s\n",
            "500:\tlearn: 0.0337963\ttotal: 2.76s\tremaining: 2.75s\n",
            "600:\tlearn: 0.0261848\ttotal: 3.31s\tremaining: 2.19s\n",
            "700:\tlearn: 0.0205883\ttotal: 3.87s\tremaining: 1.65s\n",
            "800:\tlearn: 0.0163323\ttotal: 4.42s\tremaining: 1.1s\n",
            "900:\tlearn: 0.0133739\ttotal: 4.99s\tremaining: 548ms\n",
            "999:\tlearn: 0.0109513\ttotal: 5.52s\tremaining: 0us\n",
            "0:\tlearn: 0.6238663\ttotal: 5.49ms\tremaining: 5.48s\n",
            "100:\tlearn: 0.1872826\ttotal: 551ms\tremaining: 4.9s\n",
            "200:\tlearn: 0.1469977\ttotal: 1.08s\tremaining: 4.3s\n",
            "300:\tlearn: 0.1183425\ttotal: 1.63s\tremaining: 3.78s\n",
            "400:\tlearn: 0.0969070\ttotal: 2.17s\tremaining: 3.24s\n",
            "500:\tlearn: 0.0813828\ttotal: 3.11s\tremaining: 3.1s\n",
            "600:\tlearn: 0.0670880\ttotal: 4.37s\tremaining: 2.9s\n",
            "700:\tlearn: 0.0563242\ttotal: 5.68s\tremaining: 2.42s\n",
            "800:\tlearn: 0.0481329\ttotal: 6.34s\tremaining: 1.57s\n",
            "900:\tlearn: 0.0410944\ttotal: 6.87s\tremaining: 755ms\n",
            "999:\tlearn: 0.0355313\ttotal: 7.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6251516\ttotal: 5.42ms\tremaining: 5.41s\n",
            "100:\tlearn: 0.1986659\ttotal: 539ms\tremaining: 4.79s\n",
            "200:\tlearn: 0.1590310\ttotal: 1.09s\tremaining: 4.35s\n",
            "300:\tlearn: 0.1299918\ttotal: 1.63s\tremaining: 3.78s\n",
            "400:\tlearn: 0.1084451\ttotal: 2.18s\tremaining: 3.26s\n",
            "500:\tlearn: 0.0913753\ttotal: 2.73s\tremaining: 2.72s\n",
            "600:\tlearn: 0.0773058\ttotal: 3.28s\tremaining: 2.18s\n",
            "700:\tlearn: 0.0648801\ttotal: 3.81s\tremaining: 1.63s\n",
            "800:\tlearn: 0.0551425\ttotal: 4.36s\tremaining: 1.08s\n",
            "900:\tlearn: 0.0475468\ttotal: 4.91s\tremaining: 540ms\n",
            "999:\tlearn: 0.0411743\ttotal: 5.44s\tremaining: 0us\n",
            "0:\tlearn: 0.6241247\ttotal: 5.48ms\tremaining: 5.47s\n",
            "100:\tlearn: 0.1978606\ttotal: 543ms\tremaining: 4.83s\n",
            "200:\tlearn: 0.1569792\ttotal: 1.07s\tremaining: 4.27s\n",
            "300:\tlearn: 0.1270850\ttotal: 1.67s\tremaining: 3.89s\n",
            "400:\tlearn: 0.1055732\ttotal: 2.21s\tremaining: 3.31s\n",
            "500:\tlearn: 0.0880240\ttotal: 3.02s\tremaining: 3.01s\n",
            "600:\tlearn: 0.0744480\ttotal: 4.3s\tremaining: 2.85s\n",
            "700:\tlearn: 0.0634380\ttotal: 5.49s\tremaining: 2.34s\n",
            "800:\tlearn: 0.0546343\ttotal: 6.24s\tremaining: 1.55s\n",
            "900:\tlearn: 0.0470153\ttotal: 6.79s\tremaining: 746ms\n",
            "999:\tlearn: 0.0402144\ttotal: 7.32s\tremaining: 0us\n",
            "0:\tlearn: 0.6251605\ttotal: 5.52ms\tremaining: 5.52s\n",
            "100:\tlearn: 0.1999188\ttotal: 562ms\tremaining: 5s\n",
            "200:\tlearn: 0.1604473\ttotal: 1.09s\tremaining: 4.33s\n",
            "300:\tlearn: 0.1295998\ttotal: 1.63s\tremaining: 3.79s\n",
            "400:\tlearn: 0.1082992\ttotal: 2.16s\tremaining: 3.23s\n",
            "500:\tlearn: 0.0908491\ttotal: 2.7s\tremaining: 2.69s\n",
            "600:\tlearn: 0.0762866\ttotal: 3.24s\tremaining: 2.15s\n",
            "700:\tlearn: 0.0648950\ttotal: 3.78s\tremaining: 1.61s\n",
            "800:\tlearn: 0.0547766\ttotal: 4.33s\tremaining: 1.07s\n",
            "900:\tlearn: 0.0468623\ttotal: 4.87s\tremaining: 535ms\n",
            "999:\tlearn: 0.0402823\ttotal: 5.42s\tremaining: 0us\n",
            "0:\tlearn: 0.6804556\ttotal: 5.56ms\tremaining: 5.55s\n",
            "100:\tlearn: 0.2644003\ttotal: 529ms\tremaining: 4.71s\n",
            "200:\tlearn: 0.1892531\ttotal: 1.08s\tremaining: 4.28s\n",
            "300:\tlearn: 0.1605806\ttotal: 1.61s\tremaining: 3.73s\n",
            "400:\tlearn: 0.1420032\ttotal: 2.16s\tremaining: 3.23s\n",
            "500:\tlearn: 0.1300274\ttotal: 2.7s\tremaining: 2.69s\n",
            "600:\tlearn: 0.1197858\ttotal: 3.79s\tremaining: 2.51s\n",
            "700:\tlearn: 0.1111824\ttotal: 5.1s\tremaining: 2.17s\n",
            "800:\tlearn: 0.1030633\ttotal: 6.32s\tremaining: 1.57s\n",
            "900:\tlearn: 0.0949347\ttotal: 6.86s\tremaining: 753ms\n",
            "999:\tlearn: 0.0877235\ttotal: 7.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6802679\ttotal: 5.56ms\tremaining: 5.55s\n",
            "100:\tlearn: 0.2922426\ttotal: 538ms\tremaining: 4.79s\n",
            "200:\tlearn: 0.2404537\ttotal: 1.07s\tremaining: 4.25s\n",
            "300:\tlearn: 0.2183429\ttotal: 1.62s\tremaining: 3.77s\n",
            "400:\tlearn: 0.2052006\ttotal: 2.16s\tremaining: 3.23s\n",
            "500:\tlearn: 0.1950448\ttotal: 2.71s\tremaining: 2.7s\n",
            "600:\tlearn: 0.1858464\ttotal: 3.25s\tremaining: 2.16s\n",
            "700:\tlearn: 0.1780320\ttotal: 3.79s\tremaining: 1.61s\n",
            "800:\tlearn: 0.1705372\ttotal: 4.32s\tremaining: 1.07s\n",
            "900:\tlearn: 0.1636255\ttotal: 4.87s\tremaining: 535ms\n",
            "999:\tlearn: 0.1568133\ttotal: 5.39s\tremaining: 0us\n",
            "0:\tlearn: 0.6805379\ttotal: 5.17ms\tremaining: 5.17s\n",
            "100:\tlearn: 0.2961043\ttotal: 506ms\tremaining: 4.5s\n",
            "200:\tlearn: 0.2483248\ttotal: 1.05s\tremaining: 4.19s\n",
            "300:\tlearn: 0.2284318\ttotal: 1.57s\tremaining: 3.65s\n",
            "400:\tlearn: 0.2152916\ttotal: 2.11s\tremaining: 3.15s\n",
            "500:\tlearn: 0.2054001\ttotal: 2.64s\tremaining: 2.63s\n",
            "600:\tlearn: 0.1970758\ttotal: 3.36s\tremaining: 2.23s\n",
            "700:\tlearn: 0.1896544\ttotal: 4.61s\tremaining: 1.96s\n",
            "800:\tlearn: 0.1822585\ttotal: 5.84s\tremaining: 1.45s\n",
            "900:\tlearn: 0.1750909\ttotal: 6.78s\tremaining: 746ms\n",
            "999:\tlearn: 0.1678348\ttotal: 7.33s\tremaining: 0us\n",
            "0:\tlearn: 0.6803194\ttotal: 5.55ms\tremaining: 5.55s\n",
            "100:\tlearn: 0.2977526\ttotal: 510ms\tremaining: 4.54s\n",
            "200:\tlearn: 0.2486288\ttotal: 1.07s\tremaining: 4.24s\n",
            "300:\tlearn: 0.2278670\ttotal: 1.62s\tremaining: 3.77s\n",
            "400:\tlearn: 0.2146822\ttotal: 2.18s\tremaining: 3.26s\n",
            "500:\tlearn: 0.2045659\ttotal: 2.73s\tremaining: 2.71s\n",
            "600:\tlearn: 0.1958216\ttotal: 3.26s\tremaining: 2.16s\n",
            "700:\tlearn: 0.1881869\ttotal: 3.81s\tremaining: 1.62s\n",
            "800:\tlearn: 0.1805263\ttotal: 4.35s\tremaining: 1.08s\n",
            "900:\tlearn: 0.1736213\ttotal: 4.89s\tremaining: 537ms\n",
            "999:\tlearn: 0.1666621\ttotal: 5.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6805445\ttotal: 5.5ms\tremaining: 5.5s\n",
            "100:\tlearn: 0.2996263\ttotal: 542ms\tremaining: 4.83s\n",
            "200:\tlearn: 0.2503341\ttotal: 1.06s\tremaining: 4.21s\n",
            "300:\tlearn: 0.2306920\ttotal: 1.61s\tremaining: 3.74s\n",
            "400:\tlearn: 0.2170860\ttotal: 2.15s\tremaining: 3.21s\n",
            "500:\tlearn: 0.2064008\ttotal: 2.7s\tremaining: 2.69s\n",
            "600:\tlearn: 0.1977311\ttotal: 3.24s\tremaining: 2.15s\n",
            "700:\tlearn: 0.1896346\ttotal: 4.33s\tremaining: 1.85s\n",
            "800:\tlearn: 0.1823272\ttotal: 5.57s\tremaining: 1.38s\n",
            "900:\tlearn: 0.1750823\ttotal: 6.71s\tremaining: 738ms\n",
            "999:\tlearn: 0.1678149\ttotal: 7.25s\tremaining: 0us\n",
            "0:\tlearn: 0.6326151\ttotal: 5.65ms\tremaining: 5.65s\n",
            "100:\tlearn: 0.1305059\ttotal: 560ms\tremaining: 4.98s\n",
            "200:\tlearn: 0.0883408\ttotal: 1.12s\tremaining: 4.46s\n",
            "300:\tlearn: 0.0644129\ttotal: 1.66s\tremaining: 3.85s\n",
            "400:\tlearn: 0.0495191\ttotal: 2.21s\tremaining: 3.3s\n",
            "500:\tlearn: 0.0396159\ttotal: 2.77s\tremaining: 2.76s\n",
            "600:\tlearn: 0.0325754\ttotal: 3.34s\tremaining: 2.22s\n",
            "700:\tlearn: 0.0269145\ttotal: 3.89s\tremaining: 1.66s\n",
            "800:\tlearn: 0.0227888\ttotal: 4.46s\tremaining: 1.11s\n",
            "900:\tlearn: 0.0191283\ttotal: 5.01s\tremaining: 550ms\n",
            "999:\tlearn: 0.0162526\ttotal: 5.56s\tremaining: 0us\n",
            "0:\tlearn: 0.6317614\ttotal: 5.78ms\tremaining: 5.77s\n",
            "100:\tlearn: 0.1952039\ttotal: 542ms\tremaining: 4.82s\n",
            "200:\tlearn: 0.1574880\ttotal: 1.1s\tremaining: 4.36s\n",
            "300:\tlearn: 0.1298218\ttotal: 1.65s\tremaining: 3.83s\n",
            "400:\tlearn: 0.1093238\ttotal: 2.19s\tremaining: 3.28s\n",
            "500:\tlearn: 0.0925377\ttotal: 2.74s\tremaining: 2.73s\n",
            "600:\tlearn: 0.0798585\ttotal: 3.27s\tremaining: 2.17s\n",
            "700:\tlearn: 0.0689698\ttotal: 4.17s\tremaining: 1.78s\n",
            "800:\tlearn: 0.0603222\ttotal: 5.33s\tremaining: 1.32s\n",
            "900:\tlearn: 0.0528103\ttotal: 6.68s\tremaining: 735ms\n",
            "999:\tlearn: 0.0463955\ttotal: 7.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6330237\ttotal: 5.61ms\tremaining: 5.61s\n",
            "100:\tlearn: 0.2057656\ttotal: 569ms\tremaining: 5.06s\n",
            "200:\tlearn: 0.1687522\ttotal: 1.11s\tremaining: 4.4s\n",
            "300:\tlearn: 0.1399146\ttotal: 1.7s\tremaining: 3.96s\n",
            "400:\tlearn: 0.1189365\ttotal: 2.26s\tremaining: 3.37s\n",
            "500:\tlearn: 0.1013225\ttotal: 2.79s\tremaining: 2.78s\n",
            "600:\tlearn: 0.0875004\ttotal: 3.34s\tremaining: 2.21s\n",
            "700:\tlearn: 0.0764018\ttotal: 3.88s\tremaining: 1.66s\n",
            "800:\tlearn: 0.0670599\ttotal: 4.43s\tremaining: 1.1s\n",
            "900:\tlearn: 0.0588593\ttotal: 4.97s\tremaining: 546ms\n",
            "999:\tlearn: 0.0516314\ttotal: 5.52s\tremaining: 0us\n",
            "0:\tlearn: 0.6320028\ttotal: 5.21ms\tremaining: 5.2s\n",
            "100:\tlearn: 0.2053565\ttotal: 526ms\tremaining: 4.68s\n",
            "200:\tlearn: 0.1667900\ttotal: 1.08s\tremaining: 4.28s\n",
            "300:\tlearn: 0.1375627\ttotal: 1.61s\tremaining: 3.73s\n",
            "400:\tlearn: 0.1170982\ttotal: 2.17s\tremaining: 3.24s\n",
            "500:\tlearn: 0.1001822\ttotal: 2.73s\tremaining: 2.72s\n",
            "600:\tlearn: 0.0859187\ttotal: 3.27s\tremaining: 2.17s\n",
            "700:\tlearn: 0.0747282\ttotal: 3.94s\tremaining: 1.68s\n",
            "800:\tlearn: 0.0651034\ttotal: 5.16s\tremaining: 1.28s\n",
            "900:\tlearn: 0.0572397\ttotal: 6.44s\tremaining: 708ms\n",
            "999:\tlearn: 0.0508443\ttotal: 7.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6330252\ttotal: 5.46ms\tremaining: 5.46s\n",
            "100:\tlearn: 0.2079586\ttotal: 556ms\tremaining: 4.95s\n",
            "200:\tlearn: 0.1705688\ttotal: 1.08s\tremaining: 4.3s\n",
            "300:\tlearn: 0.1410444\ttotal: 1.65s\tremaining: 3.82s\n",
            "400:\tlearn: 0.1179807\ttotal: 2.17s\tremaining: 3.24s\n",
            "500:\tlearn: 0.1005336\ttotal: 2.71s\tremaining: 2.7s\n",
            "600:\tlearn: 0.0865084\ttotal: 3.24s\tremaining: 2.15s\n",
            "700:\tlearn: 0.0751082\ttotal: 3.79s\tremaining: 1.62s\n",
            "800:\tlearn: 0.0659065\ttotal: 4.33s\tremaining: 1.07s\n",
            "900:\tlearn: 0.0576182\ttotal: 4.88s\tremaining: 536ms\n",
            "999:\tlearn: 0.0508106\ttotal: 5.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6781719\ttotal: 4.04ms\tremaining: 2.02s\n",
            "100:\tlearn: 0.3021006\ttotal: 370ms\tremaining: 1.46s\n",
            "200:\tlearn: 0.2506639\ttotal: 745ms\tremaining: 1.11s\n",
            "300:\tlearn: 0.2267491\ttotal: 1.15s\tremaining: 760ms\n",
            "400:\tlearn: 0.2131215\ttotal: 1.52s\tremaining: 375ms\n",
            "499:\tlearn: 0.2052175\ttotal: 1.89s\tremaining: 0us\n",
            "Validation Set Metrics:\n",
            "Accuracy: 0.8151950718685832\n",
            "Precision: 0.9978609625668449\n",
            "Recall: 0.7768526228143214\n",
            "F1-score: 0.8735955056179775\n",
            "AUC-ROC: 0.8845801575610068\n",
            "Submission file 'improved_catboost_submission.csv' has been created.\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split, GridSearchCV\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE\n",
        "from catboost import CatBoostClassifier\n",
        "from sklearn.metrics import accuracy_score , precision_score, recall_score, f1_score, roc_auc_score, classification_report, confusion_matrix\n",
        "from sklearn.model_selection import cross_val_score\n",
        "\n",
        "\n",
        "# Load the datasets\n",
        "train_df = pd.read_csv('/content/Train Dataset.csv')\n",
        "test_df = pd.read_csv('/content/Test Dataset.csv')\n",
        "\n",
        "# Separate features and target in the training data\n",
        "X = train_df.drop('target', axis=1)\n",
        "y = train_df['target']\n",
        "\n",
        "# Split the data into training and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# Identify numeric and categorical columns\n",
        "numeric_features = X.select_dtypes(include=[np.number]).columns\n",
        "categorical_features = X.select_dtypes(exclude=[np.number]).columns\n",
        "\n",
        "# Create preprocessing pipelines\n",
        "numeric_transformer = Pipeline(steps=[\n",
        "    ('scaler', StandardScaler())\n",
        "])\n",
        "\n",
        "categorical_transformer = Pipeline(steps=[\n",
        "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
        "])\n",
        "\n",
        "# Combine preprocessing steps\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numeric_transformer, numeric_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ]\n",
        ")\n",
        "\n",
        "# Create the full preprocessing pipeline\n",
        "full_pipeline = Pipeline([\n",
        "    ('preprocessor', preprocessor)\n",
        "])\n",
        "\n",
        "# Fit the pipeline and transform the data\n",
        "X_train_processed = full_pipeline.fit_transform(X_train, y_train)\n",
        "X_val_processed = full_pipeline.transform(X_val)\n",
        "\n",
        "# Change the column names in test_df to match the training data\n",
        "test_df.rename(columns={'id': 'Id', 'age': 'Age', 'sex': 'Sex'}, inplace=True)\n",
        "\n",
        "# Transform the test data\n",
        "X_test_processed = full_pipeline.transform(test_df)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_processed, y_train_balanced = smote.fit_resample(X_train_processed, y_train)\n",
        "\n",
        "# Define the CatBoost model and hyperparameter grid\n",
        "catboost_model = CatBoostClassifier(verbose=100)\n",
        "param_grid = {\n",
        "    'iterations': [500, 1000],\n",
        "    'learning_rate': [0.01, 0.05],\n",
        "    'depth': [4, 6],\n",
        "    'l2_leaf_reg': [1, 3]\n",
        "}\n",
        "\n",
        "# Perform grid search hyperparameter tuning\n",
        "grid_search = GridSearchCV(catboost_model, param_grid, cv=5, scoring='f1_macro')\n",
        "grid_search.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Get the best CatBoost model\n",
        "best_catboost_model = grid_search.best_estimator_\n",
        "\n",
        "# Evaluate the best CatBoost model on the validation set\n",
        "y_val_pred = best_catboost_model.predict(X_val_processed)\n",
        "\n",
        "# Calculate accuracy and other metrics\n",
        "accuracy = accuracy_score(y_val, y_val_pred)\n",
        "precision = precision_score(y_val, y_val_pred)\n",
        "recall = recall_score(y_val, y_val_pred)\n",
        "f1_score = f1_score(y_val, y_val_pred)\n",
        "auc_roc = roc_auc_score(y_val, y_val_pred)\n",
        "\n",
        "print(\"Validation Set Metrics:\")\n",
        "print(\"Accuracy:\", accuracy)\n",
        "print(\"Precision:\", precision)\n",
        "print(\"Recall:\", recall)\n",
        "print(\"F1-score:\", f1_score)\n",
        "print(\"AUC-ROC:\", auc_roc)\n",
        "\n",
        "# Make predictions on the test set\n",
        "test_predictions = best_catboost_model.predict(X_test_processed)\n",
        "\n",
        "# Create a submission DataFrame\n",
        "submission = pd.DataFrame({\n",
        "    'ID': test_df['Id'],\n",
        "    'target': test_predictions\n",
        "})\n",
        "\n",
        "# Save the submission file\n",
        "submission.to_csv('improved_catboost_submission.csv', index=False)\n",
        "print(\"Submission file 'improved_catboost_submission.csv' has been created.\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YS3OqdKoTS3z"
      },
      "source": [
        "# **CatBoost Model with Enhanced Hyperparameter Tuning and Evaluation**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v4nWE-0DTTyB",
        "outputId": "dca746ed-bb81-4d4a-f882-eb1a64cc0c46"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0:\tlearn: 0.6551995\ttotal: 23.4ms\tremaining: 23.3s\n",
            "100:\tlearn: 0.1662994\ttotal: 1.25s\tremaining: 11.2s\n",
            "200:\tlearn: 0.1239749\ttotal: 1.79s\tremaining: 7.13s\n",
            "300:\tlearn: 0.0993934\ttotal: 2.37s\tremaining: 5.49s\n",
            "400:\tlearn: 0.0794829\ttotal: 2.98s\tremaining: 4.45s\n",
            "500:\tlearn: 0.0679797\ttotal: 3.55s\tremaining: 3.53s\n",
            "600:\tlearn: 0.0594387\ttotal: 4.11s\tremaining: 2.73s\n",
            "700:\tlearn: 0.0522209\ttotal: 4.67s\tremaining: 1.99s\n",
            "800:\tlearn: 0.0464026\ttotal: 5.25s\tremaining: 1.3s\n",
            "900:\tlearn: 0.0415985\ttotal: 5.78s\tremaining: 635ms\n",
            "999:\tlearn: 0.0374992\ttotal: 6.32s\tremaining: 0us\n",
            "0:\tlearn: 0.6565370\ttotal: 6.74ms\tremaining: 6.73s\n",
            "100:\tlearn: 0.2222703\ttotal: 543ms\tremaining: 4.84s\n",
            "200:\tlearn: 0.1908360\ttotal: 1.08s\tremaining: 4.29s\n",
            "300:\tlearn: 0.1690689\ttotal: 1.63s\tremaining: 3.79s\n",
            "400:\tlearn: 0.1488195\ttotal: 2.18s\tremaining: 3.25s\n",
            "500:\tlearn: 0.1329419\ttotal: 2.75s\tremaining: 2.74s\n",
            "600:\tlearn: 0.1202657\ttotal: 3.29s\tremaining: 2.18s\n",
            "700:\tlearn: 0.1098313\ttotal: 3.85s\tremaining: 1.64s\n",
            "800:\tlearn: 0.1003899\ttotal: 4.8s\tremaining: 1.19s\n",
            "900:\tlearn: 0.0919323\ttotal: 6.15s\tremaining: 676ms\n",
            "999:\tlearn: 0.0845866\ttotal: 7.46s\tremaining: 0us\n",
            "0:\tlearn: 0.6569513\ttotal: 5.48ms\tremaining: 5.47s\n",
            "100:\tlearn: 0.2325094\ttotal: 539ms\tremaining: 4.79s\n",
            "200:\tlearn: 0.2001111\ttotal: 1.09s\tremaining: 4.32s\n",
            "300:\tlearn: 0.1795530\ttotal: 1.63s\tremaining: 3.79s\n",
            "400:\tlearn: 0.1598434\ttotal: 2.19s\tremaining: 3.27s\n",
            "500:\tlearn: 0.1454546\ttotal: 2.72s\tremaining: 2.71s\n",
            "600:\tlearn: 0.1323800\ttotal: 3.29s\tremaining: 2.19s\n",
            "700:\tlearn: 0.1217579\ttotal: 3.83s\tremaining: 1.64s\n",
            "800:\tlearn: 0.1116992\ttotal: 4.4s\tremaining: 1.09s\n",
            "900:\tlearn: 0.1029881\ttotal: 4.95s\tremaining: 544ms\n",
            "999:\tlearn: 0.0951663\ttotal: 5.52s\tremaining: 0us\n",
            "0:\tlearn: 0.6566074\ttotal: 5.58ms\tremaining: 5.58s\n",
            "100:\tlearn: 0.2318450\ttotal: 562ms\tremaining: 5s\n",
            "200:\tlearn: 0.1997540\ttotal: 1.1s\tremaining: 4.37s\n",
            "300:\tlearn: 0.1776866\ttotal: 1.67s\tremaining: 3.87s\n",
            "400:\tlearn: 0.1579653\ttotal: 2.2s\tremaining: 3.28s\n",
            "500:\tlearn: 0.1440431\ttotal: 2.78s\tremaining: 2.77s\n",
            "600:\tlearn: 0.1319674\ttotal: 3.31s\tremaining: 2.2s\n",
            "700:\tlearn: 0.1210091\ttotal: 3.86s\tremaining: 1.65s\n",
            "800:\tlearn: 0.1110523\ttotal: 4.92s\tremaining: 1.22s\n",
            "900:\tlearn: 0.1021229\ttotal: 6.13s\tremaining: 673ms\n",
            "999:\tlearn: 0.0942746\ttotal: 7.41s\tremaining: 0us\n",
            "0:\tlearn: 0.6569373\ttotal: 5.62ms\tremaining: 5.61s\n",
            "100:\tlearn: 0.2359418\ttotal: 523ms\tremaining: 4.65s\n",
            "200:\tlearn: 0.2040217\ttotal: 1.06s\tremaining: 4.2s\n",
            "300:\tlearn: 0.1823628\ttotal: 1.62s\tremaining: 3.76s\n",
            "400:\tlearn: 0.1613403\ttotal: 2.15s\tremaining: 3.21s\n",
            "500:\tlearn: 0.1464852\ttotal: 2.71s\tremaining: 2.7s\n",
            "600:\tlearn: 0.1340165\ttotal: 3.25s\tremaining: 2.16s\n",
            "700:\tlearn: 0.1234545\ttotal: 3.81s\tremaining: 1.62s\n",
            "800:\tlearn: 0.1135049\ttotal: 4.35s\tremaining: 1.08s\n",
            "900:\tlearn: 0.1045755\ttotal: 4.89s\tremaining: 537ms\n",
            "999:\tlearn: 0.0967219\ttotal: 5.49s\tremaining: 0us\n",
            "0:\tlearn: 0.6505014\ttotal: 3.5ms\tremaining: 1.75s\n",
            "100:\tlearn: 0.1823633\ttotal: 341ms\tremaining: 1.35s\n",
            "200:\tlearn: 0.1398175\ttotal: 695ms\tremaining: 1.03s\n",
            "300:\tlearn: 0.1152339\ttotal: 1.02s\tremaining: 677ms\n",
            "400:\tlearn: 0.0965334\ttotal: 1.36s\tremaining: 335ms\n",
            "499:\tlearn: 0.0857849\ttotal: 1.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6501542\ttotal: 3.63ms\tremaining: 1.81s\n",
            "100:\tlearn: 0.2315398\ttotal: 336ms\tremaining: 1.33s\n",
            "200:\tlearn: 0.2035669\ttotal: 660ms\tremaining: 982ms\n",
            "300:\tlearn: 0.1893426\ttotal: 1.01s\tremaining: 671ms\n",
            "400:\tlearn: 0.1739890\ttotal: 1.35s\tremaining: 332ms\n",
            "499:\tlearn: 0.1638541\ttotal: 1.67s\tremaining: 0us\n",
            "0:\tlearn: 0.6506329\ttotal: 3.4ms\tremaining: 1.7s\n",
            "100:\tlearn: 0.2390796\ttotal: 344ms\tremaining: 1.36s\n",
            "200:\tlearn: 0.2141529\ttotal: 770ms\tremaining: 1.15s\n",
            "300:\tlearn: 0.2005501\ttotal: 1.46s\tremaining: 966ms\n",
            "400:\tlearn: 0.1867207\ttotal: 2.25s\tremaining: 557ms\n",
            "499:\tlearn: 0.1764557\ttotal: 2.86s\tremaining: 0us\n",
            "0:\tlearn: 0.6503168\ttotal: 7.6ms\tremaining: 3.79s\n",
            "100:\tlearn: 0.2396773\ttotal: 811ms\tremaining: 3.2s\n",
            "200:\tlearn: 0.2126715\ttotal: 1.32s\tremaining: 1.97s\n",
            "300:\tlearn: 0.1987715\ttotal: 1.66s\tremaining: 1.1s\n",
            "400:\tlearn: 0.1845752\ttotal: 1.99s\tremaining: 491ms\n",
            "499:\tlearn: 0.1750543\ttotal: 2.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6507205\ttotal: 3.54ms\tremaining: 1.76s\n",
            "100:\tlearn: 0.2406717\ttotal: 331ms\tremaining: 1.31s\n",
            "200:\tlearn: 0.2141761\ttotal: 664ms\tremaining: 987ms\n",
            "300:\tlearn: 0.2014697\ttotal: 1.02s\tremaining: 672ms\n",
            "400:\tlearn: 0.1868097\ttotal: 1.34s\tremaining: 332ms\n",
            "499:\tlearn: 0.1768014\ttotal: 1.67s\tremaining: 0us\n",
            "0:\tlearn: 0.6498468\ttotal: 4ms\tremaining: 7.99s\n",
            "100:\tlearn: 0.1842176\ttotal: 355ms\tremaining: 6.68s\n",
            "200:\tlearn: 0.1382964\ttotal: 692ms\tremaining: 6.19s\n",
            "300:\tlearn: 0.1166029\ttotal: 1.02s\tremaining: 5.74s\n",
            "400:\tlearn: 0.0954979\ttotal: 1.38s\tremaining: 5.52s\n",
            "500:\tlearn: 0.0851314\ttotal: 1.71s\tremaining: 5.12s\n",
            "600:\tlearn: 0.0782310\ttotal: 2.05s\tremaining: 4.76s\n",
            "700:\tlearn: 0.0721868\ttotal: 2.39s\tremaining: 4.44s\n",
            "800:\tlearn: 0.0667088\ttotal: 2.72s\tremaining: 4.07s\n",
            "900:\tlearn: 0.0620673\ttotal: 3.05s\tremaining: 3.71s\n",
            "1000:\tlearn: 0.0579580\ttotal: 3.4s\tremaining: 3.39s\n",
            "1100:\tlearn: 0.0545328\ttotal: 3.73s\tremaining: 3.04s\n",
            "1200:\tlearn: 0.0511322\ttotal: 4.06s\tremaining: 2.7s\n",
            "1300:\tlearn: 0.0481097\ttotal: 4.41s\tremaining: 2.37s\n",
            "1400:\tlearn: 0.0452197\ttotal: 4.75s\tremaining: 2.03s\n",
            "1500:\tlearn: 0.0425064\ttotal: 5.08s\tremaining: 1.69s\n",
            "1600:\tlearn: 0.0400499\ttotal: 5.43s\tremaining: 1.35s\n",
            "1700:\tlearn: 0.0377542\ttotal: 5.75s\tremaining: 1.01s\n",
            "1800:\tlearn: 0.0358186\ttotal: 6.08s\tremaining: 672ms\n",
            "1900:\tlearn: 0.0339424\ttotal: 6.43s\tremaining: 335ms\n",
            "1999:\tlearn: 0.0321414\ttotal: 6.81s\tremaining: 0us\n",
            "0:\tlearn: 0.6494702\ttotal: 5.17ms\tremaining: 10.3s\n",
            "100:\tlearn: 0.2323547\ttotal: 778ms\tremaining: 14.6s\n",
            "200:\tlearn: 0.2044304\ttotal: 1.49s\tremaining: 13.3s\n",
            "300:\tlearn: 0.1892481\ttotal: 2.12s\tremaining: 11.9s\n",
            "400:\tlearn: 0.1736828\ttotal: 2.93s\tremaining: 11.7s\n",
            "500:\tlearn: 0.1633174\ttotal: 3.44s\tremaining: 10.3s\n",
            "600:\tlearn: 0.1550564\ttotal: 3.78s\tremaining: 8.8s\n",
            "700:\tlearn: 0.1476343\ttotal: 4.11s\tremaining: 7.62s\n",
            "800:\tlearn: 0.1407810\ttotal: 4.46s\tremaining: 6.67s\n",
            "900:\tlearn: 0.1347446\ttotal: 4.79s\tremaining: 5.84s\n",
            "1000:\tlearn: 0.1292145\ttotal: 5.13s\tremaining: 5.12s\n",
            "1100:\tlearn: 0.1240333\ttotal: 5.48s\tremaining: 4.47s\n",
            "1200:\tlearn: 0.1192733\ttotal: 5.82s\tremaining: 3.87s\n",
            "1300:\tlearn: 0.1148784\ttotal: 6.17s\tremaining: 3.31s\n",
            "1400:\tlearn: 0.1105475\ttotal: 6.51s\tremaining: 2.78s\n",
            "1500:\tlearn: 0.1065834\ttotal: 6.85s\tremaining: 2.28s\n",
            "1600:\tlearn: 0.1026910\ttotal: 7.18s\tremaining: 1.79s\n",
            "1700:\tlearn: 0.0990951\ttotal: 7.53s\tremaining: 1.32s\n",
            "1800:\tlearn: 0.0955643\ttotal: 7.86s\tremaining: 869ms\n",
            "1900:\tlearn: 0.0922242\ttotal: 8.19s\tremaining: 427ms\n",
            "1999:\tlearn: 0.0890826\ttotal: 8.53s\tremaining: 0us\n",
            "0:\tlearn: 0.6499418\ttotal: 3.36ms\tremaining: 6.72s\n",
            "100:\tlearn: 0.2396340\ttotal: 335ms\tremaining: 6.3s\n",
            "200:\tlearn: 0.2137983\ttotal: 680ms\tremaining: 6.09s\n",
            "300:\tlearn: 0.2013967\ttotal: 1.01s\tremaining: 5.68s\n",
            "400:\tlearn: 0.1870972\ttotal: 1.34s\tremaining: 5.35s\n",
            "500:\tlearn: 0.1770104\ttotal: 1.68s\tremaining: 5.03s\n",
            "600:\tlearn: 0.1688995\ttotal: 2.02s\tremaining: 4.7s\n",
            "700:\tlearn: 0.1615293\ttotal: 2.36s\tremaining: 4.37s\n",
            "800:\tlearn: 0.1549683\ttotal: 2.71s\tremaining: 4.06s\n",
            "900:\tlearn: 0.1489369\ttotal: 3.07s\tremaining: 3.75s\n",
            "1000:\tlearn: 0.1434338\ttotal: 3.41s\tremaining: 3.4s\n",
            "1100:\tlearn: 0.1380912\ttotal: 3.76s\tremaining: 3.07s\n",
            "1200:\tlearn: 0.1329200\ttotal: 4.11s\tremaining: 2.73s\n",
            "1300:\tlearn: 0.1283367\ttotal: 4.55s\tremaining: 2.44s\n",
            "1400:\tlearn: 0.1238978\ttotal: 5.3s\tremaining: 2.27s\n",
            "1500:\tlearn: 0.1193266\ttotal: 5.97s\tremaining: 1.99s\n",
            "1600:\tlearn: 0.1152374\ttotal: 6.66s\tremaining: 1.66s\n",
            "1700:\tlearn: 0.1112428\ttotal: 7.44s\tremaining: 1.31s\n",
            "1800:\tlearn: 0.1076384\ttotal: 7.95s\tremaining: 878ms\n",
            "1900:\tlearn: 0.1040639\ttotal: 8.28s\tremaining: 431ms\n",
            "1999:\tlearn: 0.1006111\ttotal: 8.6s\tremaining: 0us\n",
            "0:\tlearn: 0.6495997\ttotal: 3.28ms\tremaining: 6.56s\n",
            "100:\tlearn: 0.2402797\ttotal: 360ms\tremaining: 6.76s\n",
            "200:\tlearn: 0.2123485\ttotal: 696ms\tremaining: 6.23s\n",
            "300:\tlearn: 0.1986391\ttotal: 1.03s\tremaining: 5.8s\n",
            "400:\tlearn: 0.1842637\ttotal: 1.39s\tremaining: 5.53s\n",
            "500:\tlearn: 0.1749444\ttotal: 1.71s\tremaining: 5.13s\n",
            "600:\tlearn: 0.1663244\ttotal: 2.05s\tremaining: 4.78s\n",
            "700:\tlearn: 0.1588834\ttotal: 2.41s\tremaining: 4.47s\n",
            "800:\tlearn: 0.1518542\ttotal: 2.74s\tremaining: 4.11s\n",
            "900:\tlearn: 0.1458144\ttotal: 3.08s\tremaining: 3.76s\n",
            "1000:\tlearn: 0.1403644\ttotal: 3.43s\tremaining: 3.42s\n",
            "1100:\tlearn: 0.1346535\ttotal: 3.77s\tremaining: 3.08s\n",
            "1200:\tlearn: 0.1296532\ttotal: 4.12s\tremaining: 2.74s\n",
            "1300:\tlearn: 0.1249142\ttotal: 4.46s\tremaining: 2.4s\n",
            "1400:\tlearn: 0.1201626\ttotal: 4.79s\tremaining: 2.05s\n",
            "1500:\tlearn: 0.1159609\ttotal: 5.14s\tremaining: 1.71s\n",
            "1600:\tlearn: 0.1118371\ttotal: 5.47s\tremaining: 1.36s\n",
            "1700:\tlearn: 0.1079389\ttotal: 5.8s\tremaining: 1.02s\n",
            "1800:\tlearn: 0.1043561\ttotal: 6.15s\tremaining: 680ms\n",
            "1900:\tlearn: 0.1007298\ttotal: 6.48s\tremaining: 338ms\n",
            "1999:\tlearn: 0.0974076\ttotal: 6.81s\tremaining: 0us\n",
            "0:\tlearn: 0.6500148\ttotal: 3.77ms\tremaining: 7.54s\n",
            "100:\tlearn: 0.2405399\ttotal: 354ms\tremaining: 6.66s\n",
            "200:\tlearn: 0.2148661\ttotal: 691ms\tremaining: 6.19s\n",
            "300:\tlearn: 0.2004341\ttotal: 1.03s\tremaining: 5.8s\n",
            "400:\tlearn: 0.1859958\ttotal: 1.38s\tremaining: 5.51s\n",
            "500:\tlearn: 0.1764501\ttotal: 1.71s\tremaining: 5.13s\n",
            "600:\tlearn: 0.1679475\ttotal: 2.29s\tremaining: 5.32s\n",
            "700:\tlearn: 0.1607292\ttotal: 3s\tremaining: 5.55s\n",
            "800:\tlearn: 0.1538183\ttotal: 3.79s\tremaining: 5.67s\n",
            "900:\tlearn: 0.1478962\ttotal: 4.53s\tremaining: 5.53s\n",
            "1000:\tlearn: 0.1420204\ttotal: 5.42s\tremaining: 5.41s\n",
            "1100:\tlearn: 0.1368710\ttotal: 5.79s\tremaining: 4.73s\n",
            "1200:\tlearn: 0.1320139\ttotal: 6.13s\tremaining: 4.08s\n",
            "1300:\tlearn: 0.1274155\ttotal: 6.48s\tremaining: 3.48s\n",
            "1400:\tlearn: 0.1228205\ttotal: 6.83s\tremaining: 2.92s\n",
            "1500:\tlearn: 0.1187674\ttotal: 7.17s\tremaining: 2.38s\n",
            "1600:\tlearn: 0.1144714\ttotal: 7.51s\tremaining: 1.87s\n",
            "1700:\tlearn: 0.1106662\ttotal: 7.84s\tremaining: 1.38s\n",
            "1800:\tlearn: 0.1071594\ttotal: 8.17s\tremaining: 903ms\n",
            "1900:\tlearn: 0.1035813\ttotal: 8.53s\tremaining: 444ms\n",
            "1999:\tlearn: 0.1000906\ttotal: 8.86s\tremaining: 0us\n",
            "0:\tlearn: 0.6345013\ttotal: 13.9ms\tremaining: 27.8s\n",
            "100:\tlearn: 0.1122574\ttotal: 1.25s\tremaining: 23.5s\n",
            "200:\tlearn: 0.0645597\ttotal: 2.51s\tremaining: 22.5s\n",
            "300:\tlearn: 0.0404946\ttotal: 3.78s\tremaining: 21.3s\n",
            "400:\tlearn: 0.0282897\ttotal: 5.04s\tremaining: 20.1s\n",
            "500:\tlearn: 0.0203990\ttotal: 6.31s\tremaining: 18.9s\n",
            "600:\tlearn: 0.0152961\ttotal: 10.1s\tremaining: 23.6s\n",
            "700:\tlearn: 0.0118608\ttotal: 11.4s\tremaining: 21.1s\n",
            "800:\tlearn: 0.0093658\ttotal: 12.6s\tremaining: 18.9s\n",
            "900:\tlearn: 0.0076631\ttotal: 13.9s\tremaining: 16.9s\n",
            "1000:\tlearn: 0.0063673\ttotal: 15.1s\tremaining: 15.1s\n",
            "1100:\tlearn: 0.0053682\ttotal: 16.4s\tremaining: 13.4s\n",
            "1200:\tlearn: 0.0047305\ttotal: 17.7s\tremaining: 11.8s\n",
            "1300:\tlearn: 0.0041953\ttotal: 19s\tremaining: 10.2s\n",
            "1400:\tlearn: 0.0039508\ttotal: 20.7s\tremaining: 8.85s\n",
            "1500:\tlearn: 0.0037337\ttotal: 23.3s\tremaining: 7.74s\n",
            "1600:\tlearn: 0.0036278\ttotal: 24.6s\tremaining: 6.13s\n",
            "1700:\tlearn: 0.0035919\ttotal: 25.8s\tremaining: 4.54s\n",
            "1800:\tlearn: 0.0035601\ttotal: 27.1s\tremaining: 2.99s\n",
            "1900:\tlearn: 0.0034787\ttotal: 28.3s\tremaining: 1.47s\n",
            "1999:\tlearn: 0.0033877\ttotal: 29.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6364892\ttotal: 14.3ms\tremaining: 28.6s\n",
            "100:\tlearn: 0.1704762\ttotal: 1.22s\tremaining: 22.9s\n",
            "200:\tlearn: 0.1165878\ttotal: 2.49s\tremaining: 22.3s\n",
            "300:\tlearn: 0.0828537\ttotal: 3.92s\tremaining: 22.1s\n",
            "400:\tlearn: 0.0615333\ttotal: 6.58s\tremaining: 26.2s\n",
            "500:\tlearn: 0.0468709\ttotal: 8.09s\tremaining: 24.2s\n",
            "600:\tlearn: 0.0358557\ttotal: 9.36s\tremaining: 21.8s\n",
            "700:\tlearn: 0.0281651\ttotal: 10.6s\tremaining: 19.7s\n",
            "800:\tlearn: 0.0224921\ttotal: 11.9s\tremaining: 17.9s\n",
            "900:\tlearn: 0.0185705\ttotal: 13.2s\tremaining: 16.1s\n",
            "1000:\tlearn: 0.0155797\ttotal: 14.5s\tremaining: 14.4s\n",
            "1100:\tlearn: 0.0132587\ttotal: 15.7s\tremaining: 12.8s\n",
            "1200:\tlearn: 0.0114191\ttotal: 17s\tremaining: 11.3s\n",
            "1300:\tlearn: 0.0098538\ttotal: 19.7s\tremaining: 10.6s\n",
            "1400:\tlearn: 0.0087787\ttotal: 21.3s\tremaining: 9.12s\n",
            "1500:\tlearn: 0.0079489\ttotal: 22.6s\tremaining: 7.51s\n",
            "1600:\tlearn: 0.0071239\ttotal: 23.9s\tremaining: 5.95s\n",
            "1700:\tlearn: 0.0066771\ttotal: 25.1s\tremaining: 4.41s\n",
            "1800:\tlearn: 0.0063204\ttotal: 26.3s\tremaining: 2.9s\n",
            "1900:\tlearn: 0.0058406\ttotal: 27.5s\tremaining: 1.43s\n",
            "1999:\tlearn: 0.0055108\ttotal: 28.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6376526\ttotal: 12.8ms\tremaining: 25.5s\n",
            "100:\tlearn: 0.1786079\ttotal: 1.17s\tremaining: 22s\n",
            "200:\tlearn: 0.1210703\ttotal: 3.52s\tremaining: 31.5s\n",
            "300:\tlearn: 0.0875128\ttotal: 5.45s\tremaining: 30.8s\n",
            "400:\tlearn: 0.0652470\ttotal: 6.73s\tremaining: 26.8s\n",
            "500:\tlearn: 0.0498762\ttotal: 7.98s\tremaining: 23.9s\n",
            "600:\tlearn: 0.0393184\ttotal: 9.28s\tremaining: 21.6s\n",
            "700:\tlearn: 0.0315133\ttotal: 10.5s\tremaining: 19.5s\n",
            "800:\tlearn: 0.0258843\ttotal: 11.8s\tremaining: 17.7s\n",
            "900:\tlearn: 0.0215895\ttotal: 13.1s\tremaining: 16s\n",
            "1000:\tlearn: 0.0183673\ttotal: 14.4s\tremaining: 14.3s\n",
            "1100:\tlearn: 0.0155112\ttotal: 16.6s\tremaining: 13.6s\n",
            "1200:\tlearn: 0.0133471\ttotal: 18.7s\tremaining: 12.4s\n",
            "1300:\tlearn: 0.0117124\ttotal: 20s\tremaining: 10.7s\n",
            "1400:\tlearn: 0.0103755\ttotal: 21.2s\tremaining: 9.08s\n",
            "1500:\tlearn: 0.0092762\ttotal: 22.5s\tremaining: 7.48s\n",
            "1600:\tlearn: 0.0082900\ttotal: 23.8s\tremaining: 5.92s\n",
            "1700:\tlearn: 0.0075096\ttotal: 25s\tremaining: 4.4s\n",
            "1800:\tlearn: 0.0068343\ttotal: 26.3s\tremaining: 2.9s\n",
            "1900:\tlearn: 0.0063378\ttotal: 27.5s\tremaining: 1.43s\n",
            "1999:\tlearn: 0.0058855\ttotal: 29.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6342298\ttotal: 20.1ms\tremaining: 40.1s\n",
            "100:\tlearn: 0.1791848\ttotal: 2.08s\tremaining: 39s\n",
            "200:\tlearn: 0.1204113\ttotal: 3.34s\tremaining: 29.9s\n",
            "300:\tlearn: 0.0873456\ttotal: 4.62s\tremaining: 26.1s\n",
            "400:\tlearn: 0.0655726\ttotal: 5.9s\tremaining: 23.5s\n",
            "500:\tlearn: 0.0503180\ttotal: 7.2s\tremaining: 21.5s\n",
            "600:\tlearn: 0.0390869\ttotal: 8.47s\tremaining: 19.7s\n",
            "700:\tlearn: 0.0309112\ttotal: 9.8s\tremaining: 18.2s\n",
            "800:\tlearn: 0.0252472\ttotal: 11.1s\tremaining: 16.6s\n",
            "900:\tlearn: 0.0211297\ttotal: 13.2s\tremaining: 16.1s\n",
            "1000:\tlearn: 0.0177437\ttotal: 15.4s\tremaining: 15.4s\n",
            "1100:\tlearn: 0.0150317\ttotal: 16.7s\tremaining: 13.6s\n",
            "1200:\tlearn: 0.0129236\ttotal: 19.5s\tremaining: 13s\n",
            "1300:\tlearn: 0.0112231\ttotal: 20.8s\tremaining: 11.2s\n",
            "1400:\tlearn: 0.0099361\ttotal: 22.1s\tremaining: 9.44s\n",
            "1500:\tlearn: 0.0088423\ttotal: 23.4s\tremaining: 7.78s\n",
            "1600:\tlearn: 0.0080334\ttotal: 24.7s\tremaining: 6.14s\n",
            "1700:\tlearn: 0.0075281\ttotal: 26.9s\tremaining: 4.72s\n",
            "1800:\tlearn: 0.0070534\ttotal: 29s\tremaining: 3.2s\n",
            "1900:\tlearn: 0.0065181\ttotal: 30.3s\tremaining: 1.58s\n",
            "1999:\tlearn: 0.0061828\ttotal: 31.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6370645\ttotal: 11.9ms\tremaining: 23.9s\n",
            "100:\tlearn: 0.1835301\ttotal: 1.22s\tremaining: 22.9s\n",
            "200:\tlearn: 0.1243510\ttotal: 2.5s\tremaining: 22.4s\n",
            "300:\tlearn: 0.0886331\ttotal: 3.79s\tremaining: 21.4s\n",
            "400:\tlearn: 0.0659519\ttotal: 5.09s\tremaining: 20.3s\n",
            "500:\tlearn: 0.0503190\ttotal: 6.37s\tremaining: 19s\n",
            "600:\tlearn: 0.0389733\ttotal: 8.47s\tremaining: 19.7s\n",
            "700:\tlearn: 0.0307467\ttotal: 10.8s\tremaining: 20s\n",
            "800:\tlearn: 0.0247846\ttotal: 12.1s\tremaining: 18.1s\n",
            "900:\tlearn: 0.0205997\ttotal: 13.4s\tremaining: 16.3s\n",
            "1000:\tlearn: 0.0173960\ttotal: 14.7s\tremaining: 14.6s\n",
            "1100:\tlearn: 0.0148142\ttotal: 16s\tremaining: 13.1s\n",
            "1200:\tlearn: 0.0127064\ttotal: 17.3s\tremaining: 11.5s\n",
            "1300:\tlearn: 0.0111239\ttotal: 18.6s\tremaining: 9.97s\n",
            "1400:\tlearn: 0.0100436\ttotal: 19.9s\tremaining: 8.49s\n",
            "1500:\tlearn: 0.0089159\ttotal: 22s\tremaining: 7.3s\n",
            "1600:\tlearn: 0.0079733\ttotal: 24.3s\tremaining: 6.05s\n",
            "1700:\tlearn: 0.0072660\ttotal: 25.6s\tremaining: 4.49s\n",
            "1800:\tlearn: 0.0066707\ttotal: 26.8s\tremaining: 2.96s\n",
            "1900:\tlearn: 0.0063162\ttotal: 28.1s\tremaining: 1.46s\n",
            "1999:\tlearn: 0.0059834\ttotal: 29.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6790621\ttotal: 15.5ms\tremaining: 7.72s\n",
            "100:\tlearn: 0.2414486\ttotal: 1.25s\tremaining: 4.94s\n",
            "200:\tlearn: 0.1663736\ttotal: 2.51s\tremaining: 3.73s\n",
            "300:\tlearn: 0.1345052\ttotal: 3.81s\tremaining: 2.52s\n",
            "400:\tlearn: 0.1163404\ttotal: 5.91s\tremaining: 1.46s\n",
            "499:\tlearn: 0.1005855\ttotal: 8.26s\tremaining: 0us\n",
            "0:\tlearn: 0.6794683\ttotal: 12.3ms\tremaining: 6.14s\n",
            "100:\tlearn: 0.2747062\ttotal: 1.16s\tremaining: 4.59s\n",
            "200:\tlearn: 0.2225374\ttotal: 2.43s\tremaining: 3.61s\n",
            "300:\tlearn: 0.1953368\ttotal: 3.7s\tremaining: 2.44s\n",
            "400:\tlearn: 0.1753159\ttotal: 4.97s\tremaining: 1.23s\n",
            "499:\tlearn: 0.1578236\ttotal: 6.25s\tremaining: 0us\n",
            "0:\tlearn: 0.6797318\ttotal: 12.7ms\tremaining: 6.34s\n",
            "100:\tlearn: 0.2857417\ttotal: 1.19s\tremaining: 4.7s\n",
            "200:\tlearn: 0.2294337\ttotal: 2.46s\tremaining: 3.66s\n",
            "300:\tlearn: 0.2019028\ttotal: 4.36s\tremaining: 2.88s\n",
            "400:\tlearn: 0.1817297\ttotal: 6.72s\tremaining: 1.66s\n",
            "499:\tlearn: 0.1651237\ttotal: 7.97s\tremaining: 0us\n",
            "0:\tlearn: 0.6795367\ttotal: 12.2ms\tremaining: 6.09s\n",
            "100:\tlearn: 0.2847877\ttotal: 1.17s\tremaining: 4.62s\n",
            "200:\tlearn: 0.2307233\ttotal: 2.42s\tremaining: 3.6s\n",
            "300:\tlearn: 0.2036641\ttotal: 3.71s\tremaining: 2.45s\n",
            "400:\tlearn: 0.1825551\ttotal: 4.97s\tremaining: 1.23s\n",
            "499:\tlearn: 0.1648799\ttotal: 6.24s\tremaining: 0us\n",
            "0:\tlearn: 0.6796077\ttotal: 12ms\tremaining: 5.96s\n",
            "100:\tlearn: 0.2831980\ttotal: 1.16s\tremaining: 4.58s\n",
            "200:\tlearn: 0.2340969\ttotal: 2.94s\tremaining: 4.37s\n",
            "300:\tlearn: 0.2052637\ttotal: 5.33s\tremaining: 3.52s\n",
            "400:\tlearn: 0.1832367\ttotal: 6.62s\tremaining: 1.63s\n",
            "499:\tlearn: 0.1650272\ttotal: 7.92s\tremaining: 0us\n",
            "0:\tlearn: 0.6782940\ttotal: 3.43ms\tremaining: 6.86s\n",
            "100:\tlearn: 0.2869110\ttotal: 385ms\tremaining: 7.25s\n",
            "200:\tlearn: 0.2133993\ttotal: 718ms\tremaining: 6.43s\n",
            "300:\tlearn: 0.1803640\ttotal: 1.08s\tremaining: 6.08s\n",
            "400:\tlearn: 0.1627195\ttotal: 1.41s\tremaining: 5.62s\n",
            "500:\tlearn: 0.1491169\ttotal: 1.74s\tremaining: 5.21s\n",
            "600:\tlearn: 0.1378641\ttotal: 2.09s\tremaining: 4.88s\n",
            "700:\tlearn: 0.1297999\ttotal: 2.44s\tremaining: 4.52s\n",
            "800:\tlearn: 0.1226066\ttotal: 2.77s\tremaining: 4.15s\n",
            "900:\tlearn: 0.1148922\ttotal: 3.12s\tremaining: 3.8s\n",
            "1000:\tlearn: 0.1075298\ttotal: 3.45s\tremaining: 3.44s\n",
            "1100:\tlearn: 0.1013133\ttotal: 3.79s\tremaining: 3.09s\n",
            "1200:\tlearn: 0.0964610\ttotal: 4.13s\tremaining: 2.75s\n",
            "1300:\tlearn: 0.0924981\ttotal: 4.46s\tremaining: 2.4s\n",
            "1400:\tlearn: 0.0886436\ttotal: 4.81s\tremaining: 2.05s\n",
            "1500:\tlearn: 0.0857458\ttotal: 5.17s\tremaining: 1.72s\n",
            "1600:\tlearn: 0.0829205\ttotal: 5.5s\tremaining: 1.37s\n",
            "1700:\tlearn: 0.0804534\ttotal: 5.83s\tremaining: 1.02s\n",
            "1800:\tlearn: 0.0783304\ttotal: 6.18s\tremaining: 683ms\n",
            "1900:\tlearn: 0.0764294\ttotal: 6.52s\tremaining: 340ms\n",
            "1999:\tlearn: 0.0744821\ttotal: 6.85s\tremaining: 0us\n",
            "0:\tlearn: 0.6781636\ttotal: 5.62ms\tremaining: 11.2s\n",
            "100:\tlearn: 0.2998076\ttotal: 742ms\tremaining: 14s\n",
            "200:\tlearn: 0.2528935\ttotal: 1.47s\tremaining: 13.2s\n",
            "300:\tlearn: 0.2315427\ttotal: 2.19s\tremaining: 12.4s\n",
            "400:\tlearn: 0.2190662\ttotal: 3.06s\tremaining: 12.2s\n",
            "500:\tlearn: 0.2104895\ttotal: 3.61s\tremaining: 10.8s\n",
            "600:\tlearn: 0.2045114\ttotal: 3.94s\tremaining: 9.17s\n",
            "700:\tlearn: 0.1993296\ttotal: 4.29s\tremaining: 7.96s\n",
            "800:\tlearn: 0.1943087\ttotal: 4.63s\tremaining: 6.93s\n",
            "900:\tlearn: 0.1890008\ttotal: 4.98s\tremaining: 6.08s\n",
            "1000:\tlearn: 0.1839745\ttotal: 5.33s\tremaining: 5.32s\n",
            "1100:\tlearn: 0.1786739\ttotal: 5.68s\tremaining: 4.63s\n",
            "1200:\tlearn: 0.1741502\ttotal: 6.02s\tremaining: 4s\n",
            "1300:\tlearn: 0.1703594\ttotal: 6.37s\tremaining: 3.42s\n",
            "1400:\tlearn: 0.1670629\ttotal: 6.72s\tremaining: 2.87s\n",
            "1500:\tlearn: 0.1641909\ttotal: 7.06s\tremaining: 2.35s\n",
            "1600:\tlearn: 0.1612430\ttotal: 7.41s\tremaining: 1.85s\n",
            "1700:\tlearn: 0.1585891\ttotal: 7.74s\tremaining: 1.36s\n",
            "1800:\tlearn: 0.1559317\ttotal: 8.08s\tremaining: 892ms\n",
            "1900:\tlearn: 0.1534312\ttotal: 8.43s\tremaining: 439ms\n",
            "1999:\tlearn: 0.1510617\ttotal: 8.75s\tremaining: 0us\n",
            "0:\tlearn: 0.6783219\ttotal: 4.02ms\tremaining: 8.03s\n",
            "100:\tlearn: 0.3066887\ttotal: 353ms\tremaining: 6.64s\n",
            "200:\tlearn: 0.2616208\ttotal: 682ms\tremaining: 6.11s\n",
            "300:\tlearn: 0.2407598\ttotal: 1.02s\tremaining: 5.75s\n",
            "400:\tlearn: 0.2287328\ttotal: 1.37s\tremaining: 5.45s\n",
            "500:\tlearn: 0.2208694\ttotal: 1.7s\tremaining: 5.08s\n",
            "600:\tlearn: 0.2150090\ttotal: 2.02s\tremaining: 4.71s\n",
            "700:\tlearn: 0.2103851\ttotal: 2.37s\tremaining: 4.4s\n",
            "800:\tlearn: 0.2051921\ttotal: 2.71s\tremaining: 4.05s\n",
            "900:\tlearn: 0.2008327\ttotal: 3.04s\tremaining: 3.71s\n",
            "1000:\tlearn: 0.1962205\ttotal: 3.38s\tremaining: 3.38s\n",
            "1100:\tlearn: 0.1911279\ttotal: 3.71s\tremaining: 3.03s\n",
            "1200:\tlearn: 0.1865697\ttotal: 4.04s\tremaining: 2.69s\n",
            "1300:\tlearn: 0.1832577\ttotal: 4.49s\tremaining: 2.41s\n",
            "1400:\tlearn: 0.1800487\ttotal: 5.26s\tremaining: 2.25s\n",
            "1500:\tlearn: 0.1770420\ttotal: 6.13s\tremaining: 2.04s\n",
            "1600:\tlearn: 0.1741753\ttotal: 6.87s\tremaining: 1.71s\n",
            "1700:\tlearn: 0.1713314\ttotal: 7.61s\tremaining: 1.34s\n",
            "1800:\tlearn: 0.1687969\ttotal: 7.99s\tremaining: 883ms\n",
            "1900:\tlearn: 0.1661854\ttotal: 8.33s\tremaining: 434ms\n",
            "1999:\tlearn: 0.1638919\ttotal: 8.69s\tremaining: 0us\n",
            "0:\tlearn: 0.6782090\ttotal: 3.44ms\tremaining: 6.87s\n",
            "100:\tlearn: 0.3087542\ttotal: 333ms\tremaining: 6.26s\n",
            "200:\tlearn: 0.2607212\ttotal: 661ms\tremaining: 5.92s\n",
            "300:\tlearn: 0.2399757\ttotal: 1.01s\tremaining: 5.7s\n",
            "400:\tlearn: 0.2272236\ttotal: 1.34s\tremaining: 5.34s\n",
            "500:\tlearn: 0.2192555\ttotal: 1.68s\tremaining: 5.03s\n",
            "600:\tlearn: 0.2129123\ttotal: 2.03s\tremaining: 4.73s\n",
            "700:\tlearn: 0.2081180\ttotal: 2.36s\tremaining: 4.38s\n",
            "800:\tlearn: 0.2033223\ttotal: 2.7s\tremaining: 4.04s\n",
            "900:\tlearn: 0.1986462\ttotal: 3.05s\tremaining: 3.72s\n",
            "1000:\tlearn: 0.1935920\ttotal: 3.39s\tremaining: 3.39s\n",
            "1100:\tlearn: 0.1889637\ttotal: 3.75s\tremaining: 3.06s\n",
            "1200:\tlearn: 0.1847304\ttotal: 4.09s\tremaining: 2.72s\n",
            "1300:\tlearn: 0.1811227\ttotal: 4.42s\tremaining: 2.37s\n",
            "1400:\tlearn: 0.1780063\ttotal: 4.77s\tremaining: 2.04s\n",
            "1500:\tlearn: 0.1749599\ttotal: 5.15s\tremaining: 1.71s\n",
            "1600:\tlearn: 0.1718710\ttotal: 5.48s\tremaining: 1.37s\n",
            "1700:\tlearn: 0.1691524\ttotal: 5.84s\tremaining: 1.03s\n",
            "1800:\tlearn: 0.1664792\ttotal: 6.18s\tremaining: 683ms\n",
            "1900:\tlearn: 0.1637392\ttotal: 6.51s\tremaining: 339ms\n",
            "1999:\tlearn: 0.1612501\ttotal: 6.85s\tremaining: 0us\n",
            "0:\tlearn: 0.6783600\ttotal: 3.52ms\tremaining: 7.04s\n",
            "100:\tlearn: 0.3090280\ttotal: 327ms\tremaining: 6.15s\n",
            "200:\tlearn: 0.2631914\ttotal: 659ms\tremaining: 5.9s\n",
            "300:\tlearn: 0.2421571\ttotal: 1.02s\tremaining: 5.74s\n",
            "400:\tlearn: 0.2305807\ttotal: 1.35s\tremaining: 5.38s\n",
            "500:\tlearn: 0.2224044\ttotal: 1.69s\tremaining: 5.04s\n",
            "600:\tlearn: 0.2154860\ttotal: 2.42s\tremaining: 5.63s\n",
            "700:\tlearn: 0.2101089\ttotal: 3s\tremaining: 5.57s\n",
            "800:\tlearn: 0.2053512\ttotal: 3.82s\tremaining: 5.71s\n",
            "900:\tlearn: 0.2009332\ttotal: 4.64s\tremaining: 5.66s\n",
            "1000:\tlearn: 0.1959479\ttotal: 5.38s\tremaining: 5.37s\n",
            "1100:\tlearn: 0.1903124\ttotal: 5.71s\tremaining: 4.66s\n",
            "1200:\tlearn: 0.1861274\ttotal: 6.06s\tremaining: 4.03s\n",
            "1300:\tlearn: 0.1826983\ttotal: 6.4s\tremaining: 3.44s\n",
            "1400:\tlearn: 0.1794346\ttotal: 6.73s\tremaining: 2.88s\n",
            "1500:\tlearn: 0.1765991\ttotal: 7.09s\tremaining: 2.36s\n",
            "1600:\tlearn: 0.1735110\ttotal: 7.43s\tremaining: 1.85s\n",
            "1700:\tlearn: 0.1708619\ttotal: 7.76s\tremaining: 1.36s\n",
            "1800:\tlearn: 0.1682820\ttotal: 8.11s\tremaining: 896ms\n",
            "1900:\tlearn: 0.1657899\ttotal: 8.44s\tremaining: 440ms\n",
            "1999:\tlearn: 0.1634803\ttotal: 8.78s\tremaining: 0us\n",
            "0:\tlearn: 0.6785293\ttotal: 3.49ms\tremaining: 6.97s\n",
            "100:\tlearn: 0.2856746\ttotal: 348ms\tremaining: 6.55s\n",
            "200:\tlearn: 0.2145588\ttotal: 677ms\tremaining: 6.05s\n",
            "300:\tlearn: 0.1812029\ttotal: 1.02s\tremaining: 5.76s\n",
            "400:\tlearn: 0.1608276\ttotal: 1.36s\tremaining: 5.44s\n",
            "500:\tlearn: 0.1483873\ttotal: 1.7s\tremaining: 5.08s\n",
            "600:\tlearn: 0.1389332\ttotal: 2.04s\tremaining: 4.74s\n",
            "700:\tlearn: 0.1316640\ttotal: 2.4s\tremaining: 4.45s\n",
            "800:\tlearn: 0.1231084\ttotal: 2.73s\tremaining: 4.08s\n",
            "900:\tlearn: 0.1157335\ttotal: 3.07s\tremaining: 3.75s\n",
            "1000:\tlearn: 0.1080323\ttotal: 3.44s\tremaining: 3.43s\n",
            "1100:\tlearn: 0.1004004\ttotal: 3.77s\tremaining: 3.08s\n",
            "1200:\tlearn: 0.0961105\ttotal: 4.14s\tremaining: 2.75s\n",
            "1300:\tlearn: 0.0922029\ttotal: 4.48s\tremaining: 2.41s\n",
            "1400:\tlearn: 0.0887462\ttotal: 4.82s\tremaining: 2.06s\n",
            "1500:\tlearn: 0.0859452\ttotal: 5.17s\tremaining: 1.72s\n",
            "1600:\tlearn: 0.0832971\ttotal: 5.5s\tremaining: 1.37s\n",
            "1700:\tlearn: 0.0806839\ttotal: 5.84s\tremaining: 1.03s\n",
            "1800:\tlearn: 0.0784845\ttotal: 6.24s\tremaining: 689ms\n",
            "1900:\tlearn: 0.0764603\ttotal: 7.03s\tremaining: 366ms\n",
            "1999:\tlearn: 0.0744071\ttotal: 7.84s\tremaining: 0us\n",
            "0:\tlearn: 0.6784109\ttotal: 4.54ms\tremaining: 9.07s\n",
            "100:\tlearn: 0.3005926\ttotal: 657ms\tremaining: 12.4s\n",
            "200:\tlearn: 0.2508312\ttotal: 1.47s\tremaining: 13.1s\n",
            "300:\tlearn: 0.2294717\ttotal: 1.91s\tremaining: 10.8s\n",
            "400:\tlearn: 0.2179180\ttotal: 2.26s\tremaining: 9.02s\n",
            "500:\tlearn: 0.2102119\ttotal: 2.61s\tremaining: 7.82s\n",
            "600:\tlearn: 0.2033372\ttotal: 2.94s\tremaining: 6.85s\n",
            "700:\tlearn: 0.1974574\ttotal: 3.29s\tremaining: 6.11s\n",
            "800:\tlearn: 0.1926594\ttotal: 3.63s\tremaining: 5.44s\n",
            "900:\tlearn: 0.1883863\ttotal: 3.97s\tremaining: 4.84s\n",
            "1000:\tlearn: 0.1832510\ttotal: 4.32s\tremaining: 4.31s\n",
            "1100:\tlearn: 0.1784695\ttotal: 4.64s\tremaining: 3.79s\n",
            "1200:\tlearn: 0.1745176\ttotal: 4.98s\tremaining: 3.31s\n",
            "1300:\tlearn: 0.1707742\ttotal: 5.34s\tremaining: 2.87s\n",
            "1400:\tlearn: 0.1672961\ttotal: 5.68s\tremaining: 2.43s\n",
            "1500:\tlearn: 0.1640750\ttotal: 6.01s\tremaining: 2s\n",
            "1600:\tlearn: 0.1610999\ttotal: 6.38s\tremaining: 1.59s\n",
            "1700:\tlearn: 0.1583870\ttotal: 6.72s\tremaining: 1.18s\n",
            "1800:\tlearn: 0.1556950\ttotal: 7.05s\tremaining: 779ms\n",
            "1900:\tlearn: 0.1531512\ttotal: 7.41s\tremaining: 386ms\n",
            "1999:\tlearn: 0.1506585\ttotal: 7.74s\tremaining: 0us\n",
            "0:\tlearn: 0.6785715\ttotal: 3.42ms\tremaining: 6.83s\n",
            "100:\tlearn: 0.3053496\ttotal: 334ms\tremaining: 6.28s\n",
            "200:\tlearn: 0.2597632\ttotal: 689ms\tremaining: 6.17s\n",
            "300:\tlearn: 0.2409461\ttotal: 1.02s\tremaining: 5.77s\n",
            "400:\tlearn: 0.2298425\ttotal: 1.36s\tremaining: 5.43s\n",
            "500:\tlearn: 0.2212734\ttotal: 1.72s\tremaining: 5.13s\n",
            "600:\tlearn: 0.2161759\ttotal: 2.05s\tremaining: 4.76s\n",
            "700:\tlearn: 0.2110230\ttotal: 2.38s\tremaining: 4.42s\n",
            "800:\tlearn: 0.2063341\ttotal: 2.79s\tremaining: 4.17s\n",
            "900:\tlearn: 0.2014823\ttotal: 3.12s\tremaining: 3.81s\n",
            "1000:\tlearn: 0.1959823\ttotal: 3.47s\tremaining: 3.46s\n",
            "1100:\tlearn: 0.1911606\ttotal: 4.18s\tremaining: 3.41s\n",
            "1200:\tlearn: 0.1866483\ttotal: 4.9s\tremaining: 3.26s\n",
            "1300:\tlearn: 0.1832892\ttotal: 5.69s\tremaining: 3.06s\n",
            "1400:\tlearn: 0.1800059\ttotal: 6.54s\tremaining: 2.8s\n",
            "1500:\tlearn: 0.1769837\ttotal: 7.1s\tremaining: 2.36s\n",
            "1600:\tlearn: 0.1740141\ttotal: 7.43s\tremaining: 1.85s\n",
            "1700:\tlearn: 0.1713364\ttotal: 7.79s\tremaining: 1.37s\n",
            "1800:\tlearn: 0.1686931\ttotal: 8.12s\tremaining: 897ms\n",
            "1900:\tlearn: 0.1662674\ttotal: 8.46s\tremaining: 440ms\n",
            "1999:\tlearn: 0.1637540\ttotal: 8.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6784645\ttotal: 3.4ms\tremaining: 6.79s\n",
            "100:\tlearn: 0.3076546\ttotal: 341ms\tremaining: 6.42s\n",
            "200:\tlearn: 0.2608560\ttotal: 696ms\tremaining: 6.23s\n",
            "300:\tlearn: 0.2393815\ttotal: 1.04s\tremaining: 5.85s\n",
            "400:\tlearn: 0.2275606\ttotal: 1.38s\tremaining: 5.49s\n",
            "500:\tlearn: 0.2198455\ttotal: 1.73s\tremaining: 5.18s\n",
            "600:\tlearn: 0.2139582\ttotal: 2.06s\tremaining: 4.79s\n",
            "700:\tlearn: 0.2083662\ttotal: 2.4s\tremaining: 4.44s\n",
            "800:\tlearn: 0.2035131\ttotal: 2.75s\tremaining: 4.11s\n",
            "900:\tlearn: 0.1987360\ttotal: 3.07s\tremaining: 3.74s\n",
            "1000:\tlearn: 0.1937729\ttotal: 3.41s\tremaining: 3.4s\n",
            "1100:\tlearn: 0.1896500\ttotal: 3.76s\tremaining: 3.07s\n",
            "1200:\tlearn: 0.1848477\ttotal: 4.09s\tremaining: 2.72s\n",
            "1300:\tlearn: 0.1814847\ttotal: 4.43s\tremaining: 2.38s\n",
            "1400:\tlearn: 0.1783528\ttotal: 4.77s\tremaining: 2.04s\n",
            "1500:\tlearn: 0.1753518\ttotal: 5.1s\tremaining: 1.7s\n",
            "1600:\tlearn: 0.1722669\ttotal: 5.44s\tremaining: 1.36s\n",
            "1700:\tlearn: 0.1694671\ttotal: 5.79s\tremaining: 1.02s\n",
            "1800:\tlearn: 0.1667210\ttotal: 6.12s\tremaining: 677ms\n",
            "1900:\tlearn: 0.1641957\ttotal: 6.46s\tremaining: 337ms\n",
            "1999:\tlearn: 0.1618738\ttotal: 6.81s\tremaining: 0us\n",
            "0:\tlearn: 0.6785979\ttotal: 3.4ms\tremaining: 6.79s\n",
            "100:\tlearn: 0.3076853\ttotal: 339ms\tremaining: 6.38s\n",
            "200:\tlearn: 0.2614093\ttotal: 664ms\tremaining: 5.94s\n",
            "300:\tlearn: 0.2435798\ttotal: 1.25s\tremaining: 7.05s\n",
            "400:\tlearn: 0.2317546\ttotal: 2.01s\tremaining: 8.01s\n",
            "500:\tlearn: 0.2233695\ttotal: 2.7s\tremaining: 8.07s\n",
            "600:\tlearn: 0.2169005\ttotal: 3.43s\tremaining: 7.99s\n",
            "700:\tlearn: 0.2113338\ttotal: 4.06s\tremaining: 7.53s\n",
            "800:\tlearn: 0.2066698\ttotal: 4.51s\tremaining: 6.75s\n",
            "900:\tlearn: 0.2013363\ttotal: 4.84s\tremaining: 5.91s\n",
            "1000:\tlearn: 0.1962535\ttotal: 5.19s\tremaining: 5.18s\n",
            "1100:\tlearn: 0.1909568\ttotal: 5.53s\tremaining: 4.52s\n",
            "1200:\tlearn: 0.1866246\ttotal: 5.86s\tremaining: 3.9s\n",
            "1300:\tlearn: 0.1832001\ttotal: 6.21s\tremaining: 3.34s\n",
            "1400:\tlearn: 0.1801995\ttotal: 6.55s\tremaining: 2.8s\n",
            "1500:\tlearn: 0.1771538\ttotal: 6.88s\tremaining: 2.29s\n",
            "1600:\tlearn: 0.1743117\ttotal: 7.24s\tremaining: 1.8s\n",
            "1700:\tlearn: 0.1714935\ttotal: 7.56s\tremaining: 1.33s\n",
            "1800:\tlearn: 0.1688664\ttotal: 7.89s\tremaining: 872ms\n",
            "1900:\tlearn: 0.1663796\ttotal: 8.25s\tremaining: 430ms\n",
            "1999:\tlearn: 0.1639018\ttotal: 8.58s\tremaining: 0us\n",
            "0:\tlearn: 0.6768335\ttotal: 3.38ms\tremaining: 3.37s\n",
            "100:\tlearn: 0.2781192\ttotal: 340ms\tremaining: 3.03s\n",
            "200:\tlearn: 0.2014827\ttotal: 673ms\tremaining: 2.67s\n",
            "300:\tlearn: 0.1699561\ttotal: 1.01s\tremaining: 2.34s\n",
            "400:\tlearn: 0.1540172\ttotal: 1.36s\tremaining: 2.03s\n",
            "500:\tlearn: 0.1424334\ttotal: 1.68s\tremaining: 1.68s\n",
            "600:\tlearn: 0.1327079\ttotal: 2.02s\tremaining: 1.34s\n",
            "700:\tlearn: 0.1254979\ttotal: 2.37s\tremaining: 1.01s\n",
            "800:\tlearn: 0.1187946\ttotal: 2.71s\tremaining: 673ms\n",
            "900:\tlearn: 0.1117357\ttotal: 3.04s\tremaining: 334ms\n",
            "999:\tlearn: 0.1043698\ttotal: 3.39s\tremaining: 0us\n",
            "0:\tlearn: 0.6767047\ttotal: 4.71ms\tremaining: 4.71s\n",
            "100:\tlearn: 0.2948316\ttotal: 327ms\tremaining: 2.91s\n",
            "200:\tlearn: 0.2447038\ttotal: 654ms\tremaining: 2.6s\n",
            "300:\tlearn: 0.2253484\ttotal: 1s\tremaining: 2.34s\n",
            "400:\tlearn: 0.2146856\ttotal: 1.33s\tremaining: 1.99s\n",
            "500:\tlearn: 0.2068638\ttotal: 1.66s\tremaining: 1.65s\n",
            "600:\tlearn: 0.2004448\ttotal: 2.23s\tremaining: 1.48s\n",
            "700:\tlearn: 0.1947712\ttotal: 2.95s\tremaining: 1.26s\n",
            "800:\tlearn: 0.1906099\ttotal: 3.65s\tremaining: 907ms\n",
            "900:\tlearn: 0.1859310\ttotal: 4.28s\tremaining: 470ms\n",
            "999:\tlearn: 0.1796832\ttotal: 5.01s\tremaining: 0us\n",
            "0:\tlearn: 0.6768778\ttotal: 13.5ms\tremaining: 13.5s\n",
            "100:\tlearn: 0.2989495\ttotal: 411ms\tremaining: 3.66s\n",
            "200:\tlearn: 0.2580170\ttotal: 749ms\tremaining: 2.98s\n",
            "300:\tlearn: 0.2359303\ttotal: 1.09s\tremaining: 2.53s\n",
            "400:\tlearn: 0.2245248\ttotal: 1.42s\tremaining: 2.12s\n",
            "500:\tlearn: 0.2177173\ttotal: 1.75s\tremaining: 1.75s\n",
            "600:\tlearn: 0.2114408\ttotal: 2.09s\tremaining: 1.39s\n",
            "700:\tlearn: 0.2066680\ttotal: 2.41s\tremaining: 1.03s\n",
            "800:\tlearn: 0.2023595\ttotal: 2.76s\tremaining: 685ms\n",
            "900:\tlearn: 0.1980624\ttotal: 3.1s\tremaining: 340ms\n",
            "999:\tlearn: 0.1928056\ttotal: 3.42s\tremaining: 0us\n",
            "0:\tlearn: 0.6767637\ttotal: 4.4ms\tremaining: 4.39s\n",
            "100:\tlearn: 0.3013773\ttotal: 352ms\tremaining: 3.13s\n",
            "200:\tlearn: 0.2539953\ttotal: 677ms\tremaining: 2.69s\n",
            "300:\tlearn: 0.2346709\ttotal: 1.01s\tremaining: 2.35s\n",
            "400:\tlearn: 0.2229896\ttotal: 1.36s\tremaining: 2.03s\n",
            "500:\tlearn: 0.2147378\ttotal: 1.68s\tremaining: 1.67s\n",
            "600:\tlearn: 0.2093317\ttotal: 2.01s\tremaining: 1.33s\n",
            "700:\tlearn: 0.2045684\ttotal: 2.36s\tremaining: 1s\n",
            "800:\tlearn: 0.1997414\ttotal: 2.68s\tremaining: 665ms\n",
            "900:\tlearn: 0.1953276\ttotal: 3s\tremaining: 330ms\n",
            "999:\tlearn: 0.1903402\ttotal: 3.35s\tremaining: 0us\n",
            "0:\tlearn: 0.6769071\ttotal: 3.41ms\tremaining: 3.4s\n",
            "100:\tlearn: 0.3021804\ttotal: 324ms\tremaining: 2.88s\n",
            "200:\tlearn: 0.2577904\ttotal: 671ms\tremaining: 2.67s\n",
            "300:\tlearn: 0.2372846\ttotal: 1.02s\tremaining: 2.37s\n",
            "400:\tlearn: 0.2258501\ttotal: 1.35s\tremaining: 2.01s\n",
            "500:\tlearn: 0.2180781\ttotal: 1.68s\tremaining: 1.67s\n",
            "600:\tlearn: 0.2118302\ttotal: 2.01s\tremaining: 1.34s\n",
            "700:\tlearn: 0.2073837\ttotal: 2.34s\tremaining: 999ms\n",
            "800:\tlearn: 0.2023704\ttotal: 2.67s\tremaining: 663ms\n",
            "900:\tlearn: 0.1974317\ttotal: 3.08s\tremaining: 339ms\n",
            "999:\tlearn: 0.1923426\ttotal: 3.96s\tremaining: 0us\n",
            "0:\tlearn: 0.6201221\ttotal: 11.9ms\tremaining: 23.8s\n",
            "100:\tlearn: 0.1221653\ttotal: 1.21s\tremaining: 22.7s\n",
            "200:\tlearn: 0.0797993\ttotal: 2.43s\tremaining: 21.8s\n",
            "300:\tlearn: 0.0569831\ttotal: 2.99s\tremaining: 16.9s\n",
            "400:\tlearn: 0.0429692\ttotal: 3.54s\tremaining: 14.1s\n",
            "500:\tlearn: 0.0329577\ttotal: 4.1s\tremaining: 12.3s\n",
            "600:\tlearn: 0.0258211\ttotal: 4.65s\tremaining: 10.8s\n",
            "700:\tlearn: 0.0203731\ttotal: 5.21s\tremaining: 9.66s\n",
            "800:\tlearn: 0.0160838\ttotal: 5.75s\tremaining: 8.61s\n",
            "900:\tlearn: 0.0130557\ttotal: 6.32s\tremaining: 7.71s\n",
            "1000:\tlearn: 0.0106159\ttotal: 6.85s\tremaining: 6.83s\n",
            "1100:\tlearn: 0.0087721\ttotal: 7.42s\tremaining: 6.06s\n",
            "1200:\tlearn: 0.0073157\ttotal: 7.98s\tremaining: 5.31s\n",
            "1300:\tlearn: 0.0062841\ttotal: 8.52s\tremaining: 4.58s\n",
            "1400:\tlearn: 0.0054247\ttotal: 9.08s\tremaining: 3.88s\n",
            "1500:\tlearn: 0.0048860\ttotal: 9.61s\tremaining: 3.19s\n",
            "1600:\tlearn: 0.0046357\ttotal: 10.1s\tremaining: 2.53s\n",
            "1700:\tlearn: 0.0043491\ttotal: 10.7s\tremaining: 1.87s\n",
            "1800:\tlearn: 0.0042348\ttotal: 11.2s\tremaining: 1.24s\n",
            "1900:\tlearn: 0.0041792\ttotal: 11.7s\tremaining: 608ms\n",
            "1999:\tlearn: 0.0039883\ttotal: 12.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6219955\ttotal: 16.1ms\tremaining: 32.1s\n",
            "100:\tlearn: 0.1855856\ttotal: 1.35s\tremaining: 25.4s\n",
            "200:\tlearn: 0.1478639\ttotal: 2.74s\tremaining: 24.5s\n",
            "300:\tlearn: 0.1178822\ttotal: 3.28s\tremaining: 18.5s\n",
            "400:\tlearn: 0.0967154\ttotal: 3.85s\tremaining: 15.3s\n",
            "500:\tlearn: 0.0794809\ttotal: 4.38s\tremaining: 13.1s\n",
            "600:\tlearn: 0.0659896\ttotal: 4.99s\tremaining: 11.6s\n",
            "700:\tlearn: 0.0557769\ttotal: 5.54s\tremaining: 10.3s\n",
            "800:\tlearn: 0.0475205\ttotal: 6.08s\tremaining: 9.1s\n",
            "900:\tlearn: 0.0403851\ttotal: 6.64s\tremaining: 8.1s\n",
            "1000:\tlearn: 0.0349606\ttotal: 7.18s\tremaining: 7.17s\n",
            "1100:\tlearn: 0.0302566\ttotal: 7.74s\tremaining: 6.32s\n",
            "1200:\tlearn: 0.0259435\ttotal: 8.28s\tremaining: 5.51s\n",
            "1300:\tlearn: 0.0225711\ttotal: 8.85s\tremaining: 4.76s\n",
            "1400:\tlearn: 0.0197029\ttotal: 9.4s\tremaining: 4.02s\n",
            "1500:\tlearn: 0.0173748\ttotal: 9.97s\tremaining: 3.31s\n",
            "1600:\tlearn: 0.0152238\ttotal: 10.5s\tremaining: 2.62s\n",
            "1700:\tlearn: 0.0135092\ttotal: 11.1s\tremaining: 1.95s\n",
            "1800:\tlearn: 0.0120849\ttotal: 11.6s\tremaining: 1.29s\n",
            "1900:\tlearn: 0.0108323\ttotal: 12.2s\tremaining: 635ms\n",
            "1999:\tlearn: 0.0097229\ttotal: 13s\tremaining: 0us\n",
            "0:\tlearn: 0.6227508\ttotal: 13.8ms\tremaining: 27.6s\n",
            "100:\tlearn: 0.2004496\ttotal: 1.25s\tremaining: 23.4s\n",
            "200:\tlearn: 0.1602804\ttotal: 2.63s\tremaining: 23.6s\n",
            "300:\tlearn: 0.1305141\ttotal: 3.3s\tremaining: 18.6s\n",
            "400:\tlearn: 0.1087834\ttotal: 3.85s\tremaining: 15.3s\n",
            "500:\tlearn: 0.0917001\ttotal: 4.38s\tremaining: 13.1s\n",
            "600:\tlearn: 0.0773344\ttotal: 4.94s\tremaining: 11.5s\n",
            "700:\tlearn: 0.0657086\ttotal: 5.48s\tremaining: 10.2s\n",
            "800:\tlearn: 0.0559737\ttotal: 6.04s\tremaining: 9.05s\n",
            "900:\tlearn: 0.0478426\ttotal: 6.58s\tremaining: 8.02s\n",
            "1000:\tlearn: 0.0412026\ttotal: 7.15s\tremaining: 7.14s\n",
            "1100:\tlearn: 0.0356101\ttotal: 7.7s\tremaining: 6.29s\n",
            "1200:\tlearn: 0.0308573\ttotal: 8.24s\tremaining: 5.48s\n",
            "1300:\tlearn: 0.0269250\ttotal: 8.79s\tremaining: 4.72s\n",
            "1400:\tlearn: 0.0235232\ttotal: 9.34s\tremaining: 3.99s\n",
            "1500:\tlearn: 0.0207352\ttotal: 9.9s\tremaining: 3.29s\n",
            "1600:\tlearn: 0.0183878\ttotal: 10.4s\tremaining: 2.6s\n",
            "1700:\tlearn: 0.0162213\ttotal: 11s\tremaining: 1.93s\n",
            "1800:\tlearn: 0.0144374\ttotal: 11.5s\tremaining: 1.27s\n",
            "1900:\tlearn: 0.0129370\ttotal: 12s\tremaining: 627ms\n",
            "1999:\tlearn: 0.0116510\ttotal: 12.6s\tremaining: 0us\n",
            "0:\tlearn: 0.6221085\ttotal: 10.4ms\tremaining: 20.8s\n",
            "100:\tlearn: 0.1972229\ttotal: 1.24s\tremaining: 23.2s\n",
            "200:\tlearn: 0.1605018\ttotal: 2.49s\tremaining: 22.3s\n",
            "300:\tlearn: 0.1307175\ttotal: 3.58s\tremaining: 20.2s\n",
            "400:\tlearn: 0.1076946\ttotal: 4.1s\tremaining: 16.4s\n",
            "500:\tlearn: 0.0904924\ttotal: 4.64s\tremaining: 13.9s\n",
            "600:\tlearn: 0.0757838\ttotal: 5.21s\tremaining: 12.1s\n",
            "700:\tlearn: 0.0635417\ttotal: 5.75s\tremaining: 10.6s\n",
            "800:\tlearn: 0.0543541\ttotal: 6.32s\tremaining: 9.46s\n",
            "900:\tlearn: 0.0459525\ttotal: 6.86s\tremaining: 8.36s\n",
            "1000:\tlearn: 0.0396381\ttotal: 7.43s\tremaining: 7.41s\n",
            "1100:\tlearn: 0.0341377\ttotal: 7.98s\tremaining: 6.52s\n",
            "1200:\tlearn: 0.0298644\ttotal: 8.57s\tremaining: 5.7s\n",
            "1300:\tlearn: 0.0261083\ttotal: 9.11s\tremaining: 4.89s\n",
            "1400:\tlearn: 0.0229340\ttotal: 9.67s\tremaining: 4.13s\n",
            "1500:\tlearn: 0.0201735\ttotal: 10.2s\tremaining: 3.4s\n",
            "1600:\tlearn: 0.0178867\ttotal: 10.8s\tremaining: 2.68s\n",
            "1700:\tlearn: 0.0159420\ttotal: 11.3s\tremaining: 1.99s\n",
            "1800:\tlearn: 0.0142676\ttotal: 11.9s\tremaining: 1.31s\n",
            "1900:\tlearn: 0.0127571\ttotal: 12.4s\tremaining: 646ms\n",
            "1999:\tlearn: 0.0114848\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6227409\ttotal: 5.96ms\tremaining: 11.9s\n",
            "100:\tlearn: 0.2015869\ttotal: 1.03s\tremaining: 19.5s\n",
            "200:\tlearn: 0.1626762\ttotal: 2.22s\tremaining: 19.9s\n",
            "300:\tlearn: 0.1322656\ttotal: 3.34s\tremaining: 18.9s\n",
            "400:\tlearn: 0.1091387\ttotal: 4.08s\tremaining: 16.3s\n",
            "500:\tlearn: 0.0904441\ttotal: 4.62s\tremaining: 13.8s\n",
            "600:\tlearn: 0.0761777\ttotal: 5.17s\tremaining: 12s\n",
            "700:\tlearn: 0.0645932\ttotal: 5.72s\tremaining: 10.6s\n",
            "800:\tlearn: 0.0551144\ttotal: 6.26s\tremaining: 9.37s\n",
            "900:\tlearn: 0.0472130\ttotal: 6.82s\tremaining: 8.32s\n",
            "1000:\tlearn: 0.0406474\ttotal: 7.39s\tremaining: 7.38s\n",
            "1100:\tlearn: 0.0354328\ttotal: 7.93s\tremaining: 6.47s\n",
            "1200:\tlearn: 0.0307194\ttotal: 8.48s\tremaining: 5.64s\n",
            "1300:\tlearn: 0.0268233\ttotal: 9.03s\tremaining: 4.85s\n",
            "1400:\tlearn: 0.0236837\ttotal: 9.6s\tremaining: 4.11s\n",
            "1500:\tlearn: 0.0207744\ttotal: 10.2s\tremaining: 3.37s\n",
            "1600:\tlearn: 0.0183841\ttotal: 10.7s\tremaining: 2.67s\n",
            "1700:\tlearn: 0.0162936\ttotal: 11.2s\tremaining: 1.97s\n",
            "1800:\tlearn: 0.0146446\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0131562\ttotal: 12.3s\tremaining: 641ms\n",
            "1999:\tlearn: 0.0117910\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6772010\ttotal: 2.92ms\tremaining: 2.92s\n",
            "100:\tlearn: 0.2665558\ttotal: 605ms\tremaining: 5.38s\n",
            "200:\tlearn: 0.1924888\ttotal: 1.86s\tremaining: 7.41s\n",
            "300:\tlearn: 0.1645454\ttotal: 3.02s\tremaining: 7.02s\n",
            "400:\tlearn: 0.1450697\ttotal: 4.15s\tremaining: 6.2s\n",
            "500:\tlearn: 0.1321570\ttotal: 4.71s\tremaining: 4.69s\n",
            "600:\tlearn: 0.1227522\ttotal: 5.25s\tremaining: 3.49s\n",
            "700:\tlearn: 0.1142718\ttotal: 5.84s\tremaining: 2.49s\n",
            "800:\tlearn: 0.1068095\ttotal: 6.38s\tremaining: 1.58s\n",
            "900:\tlearn: 0.0994714\ttotal: 6.94s\tremaining: 763ms\n",
            "999:\tlearn: 0.0919176\ttotal: 7.48s\tremaining: 0us\n",
            "0:\tlearn: 0.6770685\ttotal: 2.79ms\tremaining: 2.78s\n",
            "100:\tlearn: 0.2944577\ttotal: 546ms\tremaining: 4.86s\n",
            "200:\tlearn: 0.2428767\ttotal: 1.1s\tremaining: 4.37s\n",
            "300:\tlearn: 0.2217330\ttotal: 1.63s\tremaining: 3.79s\n",
            "400:\tlearn: 0.2081390\ttotal: 2.19s\tremaining: 3.28s\n",
            "500:\tlearn: 0.1987292\ttotal: 2.73s\tremaining: 2.71s\n",
            "600:\tlearn: 0.1901609\ttotal: 3.29s\tremaining: 2.18s\n",
            "700:\tlearn: 0.1823484\ttotal: 3.83s\tremaining: 1.64s\n",
            "800:\tlearn: 0.1749717\ttotal: 4.44s\tremaining: 1.1s\n",
            "900:\tlearn: 0.1681721\ttotal: 4.98s\tremaining: 548ms\n",
            "999:\tlearn: 0.1616028\ttotal: 5.52s\tremaining: 0us\n",
            "0:\tlearn: 0.6772538\ttotal: 2.93ms\tremaining: 2.93s\n",
            "100:\tlearn: 0.3043448\ttotal: 535ms\tremaining: 4.76s\n",
            "200:\tlearn: 0.2546110\ttotal: 1.72s\tremaining: 6.82s\n",
            "300:\tlearn: 0.2327713\ttotal: 2.81s\tremaining: 6.53s\n",
            "400:\tlearn: 0.2194305\ttotal: 3.89s\tremaining: 5.81s\n",
            "500:\tlearn: 0.2101142\ttotal: 4.41s\tremaining: 4.39s\n",
            "600:\tlearn: 0.2018534\ttotal: 4.97s\tremaining: 3.3s\n",
            "700:\tlearn: 0.1943547\ttotal: 5.51s\tremaining: 2.35s\n",
            "800:\tlearn: 0.1872298\ttotal: 6.07s\tremaining: 1.51s\n",
            "900:\tlearn: 0.1800332\ttotal: 6.63s\tremaining: 728ms\n",
            "999:\tlearn: 0.1733201\ttotal: 7.15s\tremaining: 0us\n",
            "0:\tlearn: 0.6771467\ttotal: 2.94ms\tremaining: 2.93s\n",
            "100:\tlearn: 0.3007175\ttotal: 531ms\tremaining: 4.72s\n",
            "200:\tlearn: 0.2511896\ttotal: 1.05s\tremaining: 4.19s\n",
            "300:\tlearn: 0.2310720\ttotal: 1.6s\tremaining: 3.71s\n",
            "400:\tlearn: 0.2180939\ttotal: 2.13s\tremaining: 3.18s\n",
            "500:\tlearn: 0.2079355\ttotal: 2.68s\tremaining: 2.67s\n",
            "600:\tlearn: 0.1989609\ttotal: 3.21s\tremaining: 2.13s\n",
            "700:\tlearn: 0.1912465\ttotal: 3.77s\tremaining: 1.61s\n",
            "800:\tlearn: 0.1844722\ttotal: 4.3s\tremaining: 1.07s\n",
            "900:\tlearn: 0.1778174\ttotal: 4.86s\tremaining: 534ms\n",
            "999:\tlearn: 0.1710150\ttotal: 5.38s\tremaining: 0us\n",
            "0:\tlearn: 0.6772945\ttotal: 3.07ms\tremaining: 3.06s\n",
            "100:\tlearn: 0.3029536\ttotal: 511ms\tremaining: 4.55s\n",
            "200:\tlearn: 0.2559581\ttotal: 1.23s\tremaining: 4.89s\n",
            "300:\tlearn: 0.2345299\ttotal: 2.27s\tremaining: 5.28s\n",
            "400:\tlearn: 0.2208672\ttotal: 3.65s\tremaining: 5.45s\n",
            "500:\tlearn: 0.2103393\ttotal: 4.62s\tremaining: 4.6s\n",
            "600:\tlearn: 0.2017056\ttotal: 5.19s\tremaining: 3.44s\n",
            "700:\tlearn: 0.1941837\ttotal: 5.74s\tremaining: 2.45s\n",
            "800:\tlearn: 0.1869395\ttotal: 6.3s\tremaining: 1.56s\n",
            "900:\tlearn: 0.1799791\ttotal: 6.84s\tremaining: 752ms\n",
            "999:\tlearn: 0.1729827\ttotal: 7.39s\tremaining: 0us\n",
            "0:\tlearn: 0.6792296\ttotal: 13.6ms\tremaining: 6.78s\n",
            "100:\tlearn: 0.2712141\ttotal: 1.27s\tremaining: 5.02s\n",
            "200:\tlearn: 0.2172977\ttotal: 2.59s\tremaining: 3.85s\n",
            "300:\tlearn: 0.1925217\ttotal: 3.98s\tremaining: 2.63s\n",
            "400:\tlearn: 0.1750928\ttotal: 5.32s\tremaining: 1.31s\n",
            "499:\tlearn: 0.1590440\ttotal: 6.66s\tremaining: 0us\n",
            "Validation Set Metrics:\n",
            "Accuracy: 0.8193018480492813\n",
            "Precision: 0.9795291709314228\n",
            "Recall: 0.7968359700249792\n",
            "F1-score: 0.8787878787878788\n",
            "AUC-ROC: 0.8599564465509512\n",
            "Submission file '2nd_improvedcatboost_submission.csv' has been created.\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split, RandomizedSearchCV\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE\n",
        "from catboost import CatBoostClassifier\n",
        "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score\n",
        "\n",
        "# Load the datasets\n",
        "train_df = pd.read_csv('/content/Train Dataset.csv')\n",
        "test_df = pd.read_csv('/content/Test Dataset.csv')\n",
        "\n",
        "# Separate features and target in the training data\n",
        "X = train_df.drop('target', axis=1)\n",
        "y = train_df['target']\n",
        "\n",
        "# Split the data into training and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# Identify numeric and categorical columns\n",
        "numeric_features = X.select_dtypes(include=[np.number]).columns\n",
        "categorical_features = X.select_dtypes(exclude=[np.number]).columns\n",
        "\n",
        "# Create preprocessing pipelines\n",
        "numeric_transformer = Pipeline(steps=[\n",
        "    ('scaler', StandardScaler())\n",
        "])\n",
        "\n",
        "categorical_transformer = Pipeline(steps=[\n",
        "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
        "])\n",
        "\n",
        "# Combine preprocessing steps\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numeric_transformer, numeric_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ]\n",
        ")\n",
        "\n",
        "# Create the full preprocessing pipeline\n",
        "full_pipeline = Pipeline([\n",
        "    ('preprocessor', preprocessor)\n",
        "])\n",
        "\n",
        "# Fit the pipeline and transform the data\n",
        "X_train_processed = full_pipeline.fit_transform(X_train, y_train)\n",
        "X_val_processed = full_pipeline.transform(X_val)\n",
        "\n",
        "# Change the column names in test_df to match the training data\n",
        "test_df.rename(columns={'id': 'Id', 'age': 'Age', 'sex': 'Sex'}, inplace=True)\n",
        "\n",
        "# Transform the test data\n",
        "X_test_processed = full_pipeline.transform(test_df)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_processed, y_train_balanced = smote.fit_resample(X_train_processed, y_train)\n",
        "\n",
        "# Define the CatBoost model and hyperparameter grid\n",
        "catboost_model = CatBoostClassifier(verbose=100)\n",
        "param_grid = {\n",
        "    'iterations': [500, 1000, 2000],\n",
        "    'learning_rate': [0.01, 0.03, 0.05],\n",
        "    'depth': [4, 6, 8],\n",
        "    'l2_leaf_reg': [1, 3, 5],\n",
        "    'random_seed': [42, 123, 987]\n",
        "}\n",
        "\n",
        "# Perform randomized search hyperparameter tuning\n",
        "random_search = RandomizedSearchCV(catboost_model, param_distributions=param_grid, n_iter=10, cv=5, scoring='f1_macro')\n",
        "random_search.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Get the best CatBoost model\n",
        "best_catboost_model = random_search.best_estimator_\n",
        "\n",
        "# Evaluate the best CatBoost model on the validation set\n",
        "y_val_pred = best_catboost_model.predict(X_val_processed)\n",
        "\n",
        "# Calculate accuracy and other metrics\n",
        "accuracy = accuracy_score(y_val, y_val_pred)\n",
        "precision = precision_score(y_val, y_val_pred)\n",
        "recall = recall_score(y_val, y_val_pred)\n",
        "f1_score = f1_score(y_val, y_val_pred)\n",
        "auc_roc = roc_auc_score(y_val, y_val_pred)\n",
        "\n",
        "print(\"Validation Set Metrics:\")\n",
        "print(\"Accuracy:\", accuracy)\n",
        "print(\"Precision:\", precision)\n",
        "print(\"Recall:\", recall)\n",
        "print(\"F1-score:\", f1_score)\n",
        "print(\"AUC-ROC:\", auc_roc)\n",
        "\n",
        "# Make predictions on the test set\n",
        "test_predictions = best_catboost_model.predict(X_test_processed)\n",
        "\n",
        "# Create a submission DataFrame\n",
        "submission = pd.DataFrame({\n",
        "    'ID': test_df['Id'],\n",
        "    'target': test_predictions\n",
        "})\n",
        "\n",
        "# Save the submission file\n",
        "submission.to_csv('2nd_improvedcatboost_submission.csv', index=False)\n",
        "print(\"Submission file '2nd_improvedcatboost_submission.csv' has been created.\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k7CZzeYjZncW"
      },
      "source": [
        "# **Ensemble Model with CatBoost, SVM, and Random Forest**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JoRAJ23IZoJG",
        "outputId": "36572ab0-cb13-4dae-ef7e-9bd749cd6c48"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0:\tlearn: 0.6151653\ttotal: 5.91ms\tremaining: 5.9s\n",
            "100:\tlearn: 0.1123497\ttotal: 2.45s\tremaining: 21.8s\n",
            "200:\tlearn: 0.0664614\ttotal: 4.1s\tremaining: 16.3s\n",
            "300:\tlearn: 0.0418213\ttotal: 5.38s\tremaining: 12.5s\n",
            "400:\tlearn: 0.0288066\ttotal: 6.69s\tremaining: 10s\n",
            "500:\tlearn: 0.0208242\ttotal: 7.99s\tremaining: 7.96s\n",
            "600:\tlearn: 0.0157247\ttotal: 9.27s\tremaining: 6.15s\n",
            "700:\tlearn: 0.0120834\ttotal: 10.6s\tremaining: 4.51s\n",
            "800:\tlearn: 0.0095717\ttotal: 11.8s\tremaining: 2.94s\n",
            "900:\tlearn: 0.0077468\ttotal: 13.1s\tremaining: 1.44s\n",
            "999:\tlearn: 0.0063959\ttotal: 15.6s\tremaining: 0us\n",
            "0:\tlearn: 0.6145342\ttotal: 3.43ms\tremaining: 3.43s\n",
            "100:\tlearn: 0.1716079\ttotal: 1.4s\tremaining: 12.5s\n",
            "200:\tlearn: 0.1152128\ttotal: 2.66s\tremaining: 10.6s\n",
            "300:\tlearn: 0.0815506\ttotal: 3.94s\tremaining: 9.14s\n",
            "400:\tlearn: 0.0603202\ttotal: 5.21s\tremaining: 7.79s\n",
            "500:\tlearn: 0.0459861\ttotal: 6.49s\tremaining: 6.46s\n",
            "600:\tlearn: 0.0355466\ttotal: 7.77s\tremaining: 5.16s\n",
            "700:\tlearn: 0.0282270\ttotal: 9.05s\tremaining: 3.86s\n",
            "800:\tlearn: 0.0227391\ttotal: 10.3s\tremaining: 2.56s\n",
            "900:\tlearn: 0.0187599\ttotal: 13s\tremaining: 1.42s\n",
            "999:\tlearn: 0.0157183\ttotal: 14.6s\tremaining: 0us\n",
            "0:\tlearn: 0.6153588\ttotal: 2.93ms\tremaining: 2.93s\n",
            "100:\tlearn: 0.1808889\ttotal: 1.23s\tremaining: 11s\n",
            "200:\tlearn: 0.1225503\ttotal: 2.51s\tremaining: 9.98s\n",
            "300:\tlearn: 0.0872594\ttotal: 3.78s\tremaining: 8.78s\n",
            "400:\tlearn: 0.0656563\ttotal: 5.08s\tremaining: 7.58s\n",
            "500:\tlearn: 0.0499519\ttotal: 6.33s\tremaining: 6.31s\n",
            "600:\tlearn: 0.0393531\ttotal: 7.61s\tremaining: 5.05s\n",
            "700:\tlearn: 0.0315056\ttotal: 8.89s\tremaining: 3.79s\n",
            "800:\tlearn: 0.0253459\ttotal: 11.3s\tremaining: 2.82s\n",
            "900:\tlearn: 0.0208723\ttotal: 13.1s\tremaining: 1.44s\n",
            "999:\tlearn: 0.0173924\ttotal: 14.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6147722\ttotal: 2.97ms\tremaining: 2.97s\n",
            "100:\tlearn: 0.1756435\ttotal: 1.29s\tremaining: 11.5s\n",
            "200:\tlearn: 0.1219027\ttotal: 2.57s\tremaining: 10.2s\n",
            "300:\tlearn: 0.0876868\ttotal: 3.81s\tremaining: 8.86s\n",
            "400:\tlearn: 0.0657067\ttotal: 5.07s\tremaining: 7.57s\n",
            "500:\tlearn: 0.0502529\ttotal: 6.33s\tremaining: 6.3s\n",
            "600:\tlearn: 0.0393321\ttotal: 7.6s\tremaining: 5.05s\n",
            "700:\tlearn: 0.0311794\ttotal: 10s\tremaining: 4.28s\n",
            "800:\tlearn: 0.0251796\ttotal: 11.9s\tremaining: 2.96s\n",
            "900:\tlearn: 0.0211944\ttotal: 13.2s\tremaining: 1.45s\n",
            "999:\tlearn: 0.0179902\ttotal: 14.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6154900\ttotal: 2.89ms\tremaining: 2.89s\n",
            "100:\tlearn: 0.1794109\ttotal: 1.23s\tremaining: 11s\n",
            "200:\tlearn: 0.1221254\ttotal: 2.48s\tremaining: 9.87s\n",
            "300:\tlearn: 0.0883088\ttotal: 3.76s\tremaining: 8.74s\n",
            "400:\tlearn: 0.0659128\ttotal: 5.04s\tremaining: 7.53s\n",
            "500:\tlearn: 0.0498539\ttotal: 6.33s\tremaining: 6.3s\n",
            "600:\tlearn: 0.0388612\ttotal: 8.85s\tremaining: 5.87s\n",
            "700:\tlearn: 0.0309835\ttotal: 10.7s\tremaining: 4.56s\n",
            "800:\tlearn: 0.0252468\ttotal: 11.9s\tremaining: 2.97s\n",
            "900:\tlearn: 0.0208441\ttotal: 13.2s\tremaining: 1.45s\n",
            "999:\tlearn: 0.0175449\ttotal: 14.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6151653\ttotal: 2.93ms\tremaining: 1.46s\n",
            "100:\tlearn: 0.1123497\ttotal: 1.31s\tremaining: 5.16s\n",
            "200:\tlearn: 0.0664614\ttotal: 2.6s\tremaining: 3.87s\n",
            "300:\tlearn: 0.0418213\ttotal: 3.87s\tremaining: 2.56s\n",
            "400:\tlearn: 0.0288066\ttotal: 5.18s\tremaining: 1.28s\n",
            "499:\tlearn: 0.0208594\ttotal: 8.84s\tremaining: 0us\n",
            "0:\tlearn: 0.6145342\ttotal: 2.71ms\tremaining: 1.35s\n",
            "100:\tlearn: 0.1716079\ttotal: 1.22s\tremaining: 4.82s\n",
            "200:\tlearn: 0.1152128\ttotal: 2.52s\tremaining: 3.75s\n",
            "300:\tlearn: 0.0815506\ttotal: 3.79s\tremaining: 2.51s\n",
            "400:\tlearn: 0.0603202\ttotal: 5.06s\tremaining: 1.25s\n",
            "499:\tlearn: 0.0461105\ttotal: 6.32s\tremaining: 0us\n",
            "0:\tlearn: 0.6153588\ttotal: 2.87ms\tremaining: 1.43s\n",
            "100:\tlearn: 0.1808889\ttotal: 1.25s\tremaining: 4.93s\n",
            "200:\tlearn: 0.1225503\ttotal: 2.51s\tremaining: 3.73s\n",
            "300:\tlearn: 0.0872594\ttotal: 4.16s\tremaining: 2.75s\n",
            "400:\tlearn: 0.0656563\ttotal: 7.57s\tremaining: 1.87s\n",
            "499:\tlearn: 0.0500991\ttotal: 8.85s\tremaining: 0us\n",
            "0:\tlearn: 0.6147722\ttotal: 2.91ms\tremaining: 1.45s\n",
            "100:\tlearn: 0.1756435\ttotal: 1.26s\tremaining: 4.99s\n",
            "200:\tlearn: 0.1219027\ttotal: 2.54s\tremaining: 3.78s\n",
            "300:\tlearn: 0.0876868\ttotal: 3.82s\tremaining: 2.52s\n",
            "400:\tlearn: 0.0657067\ttotal: 5.09s\tremaining: 1.26s\n",
            "499:\tlearn: 0.0504198\ttotal: 6.44s\tremaining: 0us\n",
            "0:\tlearn: 0.6154900\ttotal: 3.55ms\tremaining: 1.77s\n",
            "100:\tlearn: 0.1794109\ttotal: 1.63s\tremaining: 6.43s\n",
            "200:\tlearn: 0.1221254\ttotal: 4.67s\tremaining: 6.95s\n",
            "300:\tlearn: 0.0883088\ttotal: 7.27s\tremaining: 4.81s\n",
            "400:\tlearn: 0.0659128\ttotal: 8.93s\tremaining: 2.2s\n",
            "499:\tlearn: 0.0499787\ttotal: 10.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6793506\ttotal: 5.58ms\tremaining: 5.58s\n",
            "100:\tlearn: 0.2610784\ttotal: 555ms\tremaining: 4.94s\n",
            "200:\tlearn: 0.1927226\ttotal: 1.11s\tremaining: 4.41s\n",
            "300:\tlearn: 0.1617228\ttotal: 1.65s\tremaining: 3.84s\n",
            "400:\tlearn: 0.1429668\ttotal: 2.21s\tremaining: 3.3s\n",
            "500:\tlearn: 0.1307759\ttotal: 2.75s\tremaining: 2.73s\n",
            "600:\tlearn: 0.1199879\ttotal: 3.29s\tremaining: 2.19s\n",
            "700:\tlearn: 0.1113058\ttotal: 3.86s\tremaining: 1.65s\n",
            "800:\tlearn: 0.1032290\ttotal: 4.42s\tremaining: 1.1s\n",
            "900:\tlearn: 0.0951055\ttotal: 4.97s\tremaining: 546ms\n",
            "999:\tlearn: 0.0877303\ttotal: 5.61s\tremaining: 0us\n",
            "0:\tlearn: 0.6798133\ttotal: 11.7ms\tremaining: 11.7s\n",
            "100:\tlearn: 0.2889438\ttotal: 1.27s\tremaining: 11.3s\n",
            "200:\tlearn: 0.2395739\ttotal: 2.52s\tremaining: 10s\n",
            "300:\tlearn: 0.2188545\ttotal: 3.29s\tremaining: 7.64s\n",
            "400:\tlearn: 0.2050619\ttotal: 3.83s\tremaining: 5.72s\n",
            "500:\tlearn: 0.1956103\ttotal: 4.37s\tremaining: 4.35s\n",
            "600:\tlearn: 0.1861840\ttotal: 4.93s\tremaining: 3.28s\n",
            "700:\tlearn: 0.1779782\ttotal: 5.47s\tremaining: 2.33s\n",
            "800:\tlearn: 0.1701770\ttotal: 6.02s\tremaining: 1.5s\n",
            "900:\tlearn: 0.1629168\ttotal: 6.57s\tremaining: 722ms\n",
            "999:\tlearn: 0.1564936\ttotal: 7.09s\tremaining: 0us\n",
            "0:\tlearn: 0.6799623\ttotal: 5.41ms\tremaining: 5.4s\n",
            "100:\tlearn: 0.2976966\ttotal: 523ms\tremaining: 4.65s\n",
            "200:\tlearn: 0.2498509\ttotal: 1.03s\tremaining: 4.12s\n",
            "300:\tlearn: 0.2291381\ttotal: 1.58s\tremaining: 3.68s\n",
            "400:\tlearn: 0.2157448\ttotal: 2.12s\tremaining: 3.17s\n",
            "500:\tlearn: 0.2055916\ttotal: 2.67s\tremaining: 2.66s\n",
            "600:\tlearn: 0.1971933\ttotal: 3.2s\tremaining: 2.13s\n",
            "700:\tlearn: 0.1892561\ttotal: 3.75s\tremaining: 1.6s\n",
            "800:\tlearn: 0.1820051\ttotal: 4.29s\tremaining: 1.06s\n",
            "900:\tlearn: 0.1743188\ttotal: 4.84s\tremaining: 532ms\n",
            "999:\tlearn: 0.1673061\ttotal: 5.39s\tremaining: 0us\n",
            "0:\tlearn: 0.6798407\ttotal: 5.44ms\tremaining: 5.43s\n",
            "100:\tlearn: 0.2976196\ttotal: 1.11s\tremaining: 9.91s\n",
            "200:\tlearn: 0.2489349\ttotal: 2.28s\tremaining: 9.08s\n",
            "300:\tlearn: 0.2271681\ttotal: 3.64s\tremaining: 8.46s\n",
            "400:\tlearn: 0.2145400\ttotal: 4.21s\tremaining: 6.29s\n",
            "500:\tlearn: 0.2042814\ttotal: 4.74s\tremaining: 4.72s\n",
            "600:\tlearn: 0.1955177\ttotal: 5.3s\tremaining: 3.52s\n",
            "700:\tlearn: 0.1879364\ttotal: 5.84s\tremaining: 2.49s\n",
            "800:\tlearn: 0.1807753\ttotal: 6.4s\tremaining: 1.59s\n",
            "900:\tlearn: 0.1734264\ttotal: 6.94s\tremaining: 762ms\n",
            "999:\tlearn: 0.1662376\ttotal: 7.49s\tremaining: 0us\n",
            "0:\tlearn: 0.6799523\ttotal: 5.22ms\tremaining: 5.21s\n",
            "100:\tlearn: 0.2980245\ttotal: 507ms\tremaining: 4.51s\n",
            "200:\tlearn: 0.2504249\ttotal: 1.02s\tremaining: 4.07s\n",
            "300:\tlearn: 0.2307104\ttotal: 1.57s\tremaining: 3.65s\n",
            "400:\tlearn: 0.2169661\ttotal: 2.1s\tremaining: 3.15s\n",
            "500:\tlearn: 0.2066040\ttotal: 2.65s\tremaining: 2.65s\n",
            "600:\tlearn: 0.1977188\ttotal: 3.18s\tremaining: 2.11s\n",
            "700:\tlearn: 0.1903455\ttotal: 3.74s\tremaining: 1.6s\n",
            "800:\tlearn: 0.1831098\ttotal: 4.28s\tremaining: 1.06s\n",
            "900:\tlearn: 0.1755439\ttotal: 4.84s\tremaining: 532ms\n",
            "999:\tlearn: 0.1682493\ttotal: 5.38s\tremaining: 0us\n",
            "0:\tlearn: 0.6498468\ttotal: 3.35ms\tremaining: 1.67s\n",
            "100:\tlearn: 0.1842176\ttotal: 344ms\tremaining: 1.36s\n",
            "200:\tlearn: 0.1382964\ttotal: 928ms\tremaining: 1.38s\n",
            "300:\tlearn: 0.1166029\ttotal: 1.63s\tremaining: 1.08s\n",
            "400:\tlearn: 0.0954979\ttotal: 2.36s\tremaining: 584ms\n",
            "499:\tlearn: 0.0852230\ttotal: 2.94s\tremaining: 0us\n",
            "0:\tlearn: 0.6494702\ttotal: 4.17ms\tremaining: 2.08s\n",
            "100:\tlearn: 0.2323547\ttotal: 586ms\tremaining: 2.31s\n",
            "200:\tlearn: 0.2044304\ttotal: 914ms\tremaining: 1.36s\n",
            "300:\tlearn: 0.1892481\ttotal: 1.26s\tremaining: 832ms\n",
            "400:\tlearn: 0.1736828\ttotal: 1.59s\tremaining: 392ms\n",
            "499:\tlearn: 0.1633796\ttotal: 1.92s\tremaining: 0us\n",
            "0:\tlearn: 0.6499418\ttotal: 3.87ms\tremaining: 1.93s\n",
            "100:\tlearn: 0.2396340\ttotal: 346ms\tremaining: 1.37s\n",
            "200:\tlearn: 0.2137983\ttotal: 681ms\tremaining: 1.01s\n",
            "300:\tlearn: 0.2013967\ttotal: 1.02s\tremaining: 673ms\n",
            "400:\tlearn: 0.1870972\ttotal: 1.38s\tremaining: 340ms\n",
            "499:\tlearn: 0.1771006\ttotal: 1.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6495997\ttotal: 3.29ms\tremaining: 1.64s\n",
            "100:\tlearn: 0.2402797\ttotal: 331ms\tremaining: 1.31s\n",
            "200:\tlearn: 0.2123485\ttotal: 691ms\tremaining: 1.03s\n",
            "300:\tlearn: 0.1986391\ttotal: 1.02s\tremaining: 675ms\n",
            "400:\tlearn: 0.1842637\ttotal: 1.37s\tremaining: 338ms\n",
            "499:\tlearn: 0.1750319\ttotal: 1.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6500148\ttotal: 3.38ms\tremaining: 1.69s\n",
            "100:\tlearn: 0.2405399\ttotal: 335ms\tremaining: 1.32s\n",
            "200:\tlearn: 0.2148661\ttotal: 688ms\tremaining: 1.02s\n",
            "300:\tlearn: 0.2004341\ttotal: 1.02s\tremaining: 678ms\n",
            "400:\tlearn: 0.1859958\ttotal: 1.36s\tremaining: 336ms\n",
            "499:\tlearn: 0.1765408\ttotal: 1.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6764286\ttotal: 2.94ms\tremaining: 1.47s\n",
            "100:\tlearn: 0.2508646\ttotal: 1.17s\tremaining: 4.63s\n",
            "200:\tlearn: 0.1760537\ttotal: 2.47s\tremaining: 3.67s\n",
            "300:\tlearn: 0.1458719\ttotal: 4.58s\tremaining: 3.03s\n",
            "400:\tlearn: 0.1261938\ttotal: 6.84s\tremaining: 1.69s\n",
            "499:\tlearn: 0.1115623\ttotal: 8.12s\tremaining: 0us\n",
            "0:\tlearn: 0.6762966\ttotal: 2.9ms\tremaining: 1.45s\n",
            "100:\tlearn: 0.2843405\ttotal: 1.1s\tremaining: 4.33s\n",
            "200:\tlearn: 0.2320997\ttotal: 2.3s\tremaining: 3.43s\n",
            "300:\tlearn: 0.2063673\ttotal: 3.56s\tremaining: 2.35s\n",
            "400:\tlearn: 0.1867159\ttotal: 4.8s\tremaining: 1.18s\n",
            "499:\tlearn: 0.1706659\ttotal: 6.08s\tremaining: 0us\n",
            "0:\tlearn: 0.6764830\ttotal: 2.88ms\tremaining: 1.44s\n",
            "100:\tlearn: 0.2937408\ttotal: 1.03s\tremaining: 4.08s\n",
            "200:\tlearn: 0.2419446\ttotal: 2.53s\tremaining: 3.77s\n",
            "300:\tlearn: 0.2146361\ttotal: 5.12s\tremaining: 3.38s\n",
            "400:\tlearn: 0.1956190\ttotal: 6.59s\tremaining: 1.63s\n",
            "499:\tlearn: 0.1801269\ttotal: 7.84s\tremaining: 0us\n",
            "0:\tlearn: 0.6763732\ttotal: 2.98ms\tremaining: 1.49s\n",
            "100:\tlearn: 0.2913394\ttotal: 1.05s\tremaining: 4.14s\n",
            "200:\tlearn: 0.2411850\ttotal: 2.23s\tremaining: 3.32s\n",
            "300:\tlearn: 0.2139744\ttotal: 3.5s\tremaining: 2.31s\n",
            "400:\tlearn: 0.1957094\ttotal: 4.75s\tremaining: 1.17s\n",
            "499:\tlearn: 0.1798894\ttotal: 6.82s\tremaining: 0us\n",
            "0:\tlearn: 0.6765251\ttotal: 3.15ms\tremaining: 1.57s\n",
            "100:\tlearn: 0.2947073\ttotal: 1.62s\tremaining: 6.4s\n",
            "200:\tlearn: 0.2448800\ttotal: 4.09s\tremaining: 6.09s\n",
            "300:\tlearn: 0.2179275\ttotal: 5.35s\tremaining: 3.54s\n",
            "400:\tlearn: 0.1989630\ttotal: 6.62s\tremaining: 1.63s\n",
            "499:\tlearn: 0.1831672\ttotal: 7.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6276837\ttotal: 3.37ms\tremaining: 1.68s\n",
            "100:\tlearn: 0.1502910\ttotal: 617ms\tremaining: 2.44s\n",
            "200:\tlearn: 0.1077268\ttotal: 1.34s\tremaining: 1.99s\n",
            "300:\tlearn: 0.0874599\ttotal: 1.68s\tremaining: 1.11s\n",
            "400:\tlearn: 0.0764990\ttotal: 2.01s\tremaining: 496ms\n",
            "499:\tlearn: 0.0677052\ttotal: 2.35s\tremaining: 0us\n",
            "0:\tlearn: 0.6271478\ttotal: 3.47ms\tremaining: 1.73s\n",
            "100:\tlearn: 0.2106231\ttotal: 362ms\tremaining: 1.43s\n",
            "200:\tlearn: 0.1845374\ttotal: 699ms\tremaining: 1.04s\n",
            "300:\tlearn: 0.1655504\ttotal: 1.03s\tremaining: 683ms\n",
            "400:\tlearn: 0.1525224\ttotal: 1.38s\tremaining: 341ms\n",
            "499:\tlearn: 0.1413106\ttotal: 1.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6278940\ttotal: 3.42ms\tremaining: 1.71s\n",
            "100:\tlearn: 0.2212609\ttotal: 331ms\tremaining: 1.31s\n",
            "200:\tlearn: 0.1963040\ttotal: 703ms\tremaining: 1.05s\n",
            "300:\tlearn: 0.1786086\ttotal: 1.77s\tremaining: 1.17s\n",
            "400:\tlearn: 0.1657535\ttotal: 3.49s\tremaining: 861ms\n",
            "499:\tlearn: 0.1548192\ttotal: 4.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6274106\ttotal: 3.67ms\tremaining: 1.83s\n",
            "100:\tlearn: 0.2208258\ttotal: 567ms\tremaining: 2.24s\n",
            "200:\tlearn: 0.1949890\ttotal: 983ms\tremaining: 1.46s\n",
            "300:\tlearn: 0.1769808\ttotal: 1.33s\tremaining: 878ms\n",
            "400:\tlearn: 0.1637753\ttotal: 1.66s\tremaining: 411ms\n",
            "499:\tlearn: 0.1524461\ttotal: 1.99s\tremaining: 0us\n",
            "0:\tlearn: 0.6280315\ttotal: 7.18ms\tremaining: 3.58s\n",
            "100:\tlearn: 0.2221093\ttotal: 345ms\tremaining: 1.36s\n",
            "200:\tlearn: 0.1970345\ttotal: 687ms\tremaining: 1.02s\n",
            "300:\tlearn: 0.1783569\ttotal: 1.03s\tremaining: 680ms\n",
            "400:\tlearn: 0.1650545\ttotal: 1.37s\tremaining: 337ms\n",
            "499:\tlearn: 0.1546492\ttotal: 1.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6446717\ttotal: 2.9ms\tremaining: 5.8s\n",
            "100:\tlearn: 0.1837057\ttotal: 352ms\tremaining: 6.61s\n",
            "200:\tlearn: 0.1374914\ttotal: 688ms\tremaining: 6.15s\n",
            "300:\tlearn: 0.1160950\ttotal: 1.02s\tremaining: 5.78s\n",
            "400:\tlearn: 0.0969178\ttotal: 1.37s\tremaining: 5.46s\n",
            "500:\tlearn: 0.0866321\ttotal: 1.7s\tremaining: 5.08s\n",
            "600:\tlearn: 0.0786777\ttotal: 2.04s\tremaining: 4.75s\n",
            "700:\tlearn: 0.0728135\ttotal: 2.38s\tremaining: 4.41s\n",
            "800:\tlearn: 0.0678201\ttotal: 2.71s\tremaining: 4.06s\n",
            "900:\tlearn: 0.0632640\ttotal: 3.04s\tremaining: 3.71s\n",
            "1000:\tlearn: 0.0588100\ttotal: 3.38s\tremaining: 3.38s\n",
            "1100:\tlearn: 0.0550746\ttotal: 3.72s\tremaining: 3.04s\n",
            "1200:\tlearn: 0.0520210\ttotal: 4.05s\tremaining: 2.7s\n",
            "1300:\tlearn: 0.0486772\ttotal: 4.4s\tremaining: 2.36s\n",
            "1400:\tlearn: 0.0459686\ttotal: 4.75s\tremaining: 2.03s\n",
            "1500:\tlearn: 0.0431105\ttotal: 5.08s\tremaining: 1.69s\n",
            "1600:\tlearn: 0.0407112\ttotal: 5.43s\tremaining: 1.35s\n",
            "1700:\tlearn: 0.0385332\ttotal: 5.76s\tremaining: 1.01s\n",
            "1800:\tlearn: 0.0364899\ttotal: 6.09s\tremaining: 673ms\n",
            "1900:\tlearn: 0.0345887\ttotal: 6.44s\tremaining: 335ms\n",
            "1999:\tlearn: 0.0326740\ttotal: 6.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6442642\ttotal: 5.48ms\tremaining: 11s\n",
            "100:\tlearn: 0.2314324\ttotal: 696ms\tremaining: 13.1s\n",
            "200:\tlearn: 0.2032641\ttotal: 1.22s\tremaining: 10.9s\n",
            "300:\tlearn: 0.1893548\ttotal: 1.95s\tremaining: 11s\n",
            "400:\tlearn: 0.1745479\ttotal: 2.59s\tremaining: 10.3s\n",
            "500:\tlearn: 0.1641597\ttotal: 3.06s\tremaining: 9.17s\n",
            "600:\tlearn: 0.1558088\ttotal: 3.42s\tremaining: 7.96s\n",
            "700:\tlearn: 0.1483165\ttotal: 3.75s\tremaining: 6.95s\n",
            "800:\tlearn: 0.1415247\ttotal: 4.08s\tremaining: 6.11s\n",
            "900:\tlearn: 0.1353262\ttotal: 4.43s\tremaining: 5.4s\n",
            "1000:\tlearn: 0.1296853\ttotal: 4.76s\tremaining: 4.75s\n",
            "1100:\tlearn: 0.1245368\ttotal: 5.09s\tremaining: 4.15s\n",
            "1200:\tlearn: 0.1195733\ttotal: 5.44s\tremaining: 3.62s\n",
            "1300:\tlearn: 0.1149818\ttotal: 5.77s\tremaining: 3.1s\n",
            "1400:\tlearn: 0.1108684\ttotal: 6.1s\tremaining: 2.61s\n",
            "1500:\tlearn: 0.1068645\ttotal: 6.46s\tremaining: 2.15s\n",
            "1600:\tlearn: 0.1029232\ttotal: 6.79s\tremaining: 1.69s\n",
            "1700:\tlearn: 0.0994358\ttotal: 7.12s\tremaining: 1.25s\n",
            "1800:\tlearn: 0.0961811\ttotal: 7.47s\tremaining: 825ms\n",
            "1900:\tlearn: 0.0929409\ttotal: 7.8s\tremaining: 406ms\n",
            "1999:\tlearn: 0.0897757\ttotal: 8.12s\tremaining: 0us\n",
            "0:\tlearn: 0.6448087\ttotal: 2.86ms\tremaining: 5.72s\n",
            "100:\tlearn: 0.2385095\ttotal: 350ms\tremaining: 6.58s\n",
            "200:\tlearn: 0.2146081\ttotal: 683ms\tremaining: 6.11s\n",
            "300:\tlearn: 0.2004198\ttotal: 1.04s\tremaining: 5.87s\n",
            "400:\tlearn: 0.1867755\ttotal: 1.39s\tremaining: 5.52s\n",
            "500:\tlearn: 0.1769974\ttotal: 1.71s\tremaining: 5.12s\n",
            "600:\tlearn: 0.1687894\ttotal: 2.04s\tremaining: 4.75s\n",
            "700:\tlearn: 0.1612434\ttotal: 2.39s\tremaining: 4.43s\n",
            "800:\tlearn: 0.1546413\ttotal: 2.72s\tremaining: 4.08s\n",
            "900:\tlearn: 0.1487425\ttotal: 3.06s\tremaining: 3.73s\n",
            "1000:\tlearn: 0.1429980\ttotal: 3.41s\tremaining: 3.4s\n",
            "1100:\tlearn: 0.1376025\ttotal: 3.75s\tremaining: 3.06s\n",
            "1200:\tlearn: 0.1323893\ttotal: 4.08s\tremaining: 2.71s\n",
            "1300:\tlearn: 0.1277687\ttotal: 4.46s\tremaining: 2.4s\n",
            "1400:\tlearn: 0.1234493\ttotal: 5.17s\tremaining: 2.21s\n",
            "1500:\tlearn: 0.1189398\ttotal: 5.75s\tremaining: 1.91s\n",
            "1600:\tlearn: 0.1148916\ttotal: 6.48s\tremaining: 1.61s\n",
            "1700:\tlearn: 0.1111446\ttotal: 7.29s\tremaining: 1.28s\n",
            "1800:\tlearn: 0.1074523\ttotal: 7.82s\tremaining: 864ms\n",
            "1900:\tlearn: 0.1039902\ttotal: 8.14s\tremaining: 424ms\n",
            "1999:\tlearn: 0.1006575\ttotal: 8.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6444333\ttotal: 2.91ms\tremaining: 5.82s\n",
            "100:\tlearn: 0.2406091\ttotal: 335ms\tremaining: 6.3s\n",
            "200:\tlearn: 0.2145620\ttotal: 675ms\tremaining: 6.04s\n",
            "300:\tlearn: 0.1990483\ttotal: 1.03s\tremaining: 5.84s\n",
            "400:\tlearn: 0.1840673\ttotal: 1.37s\tremaining: 5.46s\n",
            "500:\tlearn: 0.1746477\ttotal: 1.71s\tremaining: 5.13s\n",
            "600:\tlearn: 0.1663897\ttotal: 2.07s\tremaining: 4.82s\n",
            "700:\tlearn: 0.1585023\ttotal: 2.4s\tremaining: 4.46s\n",
            "800:\tlearn: 0.1518272\ttotal: 2.73s\tremaining: 4.09s\n",
            "900:\tlearn: 0.1456163\ttotal: 3.09s\tremaining: 3.76s\n",
            "1000:\tlearn: 0.1402458\ttotal: 3.41s\tremaining: 3.41s\n",
            "1100:\tlearn: 0.1343346\ttotal: 3.74s\tremaining: 3.06s\n",
            "1200:\tlearn: 0.1294246\ttotal: 4.09s\tremaining: 2.72s\n",
            "1300:\tlearn: 0.1246320\ttotal: 4.43s\tremaining: 2.38s\n",
            "1400:\tlearn: 0.1200455\ttotal: 4.75s\tremaining: 2.03s\n",
            "1500:\tlearn: 0.1158496\ttotal: 5.1s\tremaining: 1.7s\n",
            "1600:\tlearn: 0.1117373\ttotal: 5.43s\tremaining: 1.35s\n",
            "1700:\tlearn: 0.1080329\ttotal: 5.77s\tremaining: 1.01s\n",
            "1800:\tlearn: 0.1039516\ttotal: 6.12s\tremaining: 676ms\n",
            "1900:\tlearn: 0.1002890\ttotal: 6.46s\tremaining: 336ms\n",
            "1999:\tlearn: 0.0971149\ttotal: 6.78s\tremaining: 0us\n",
            "0:\tlearn: 0.6448733\ttotal: 2.89ms\tremaining: 5.79s\n",
            "100:\tlearn: 0.2417710\ttotal: 335ms\tremaining: 6.29s\n",
            "200:\tlearn: 0.2168660\ttotal: 669ms\tremaining: 5.99s\n",
            "300:\tlearn: 0.2018158\ttotal: 1.01s\tremaining: 5.71s\n",
            "400:\tlearn: 0.1858587\ttotal: 1.35s\tremaining: 5.4s\n",
            "500:\tlearn: 0.1763913\ttotal: 1.69s\tremaining: 5.05s\n",
            "600:\tlearn: 0.1679184\ttotal: 2.07s\tremaining: 4.83s\n",
            "700:\tlearn: 0.1605837\ttotal: 2.78s\tremaining: 5.15s\n",
            "800:\tlearn: 0.1540247\ttotal: 3.46s\tremaining: 5.17s\n",
            "900:\tlearn: 0.1478365\ttotal: 4.1s\tremaining: 5s\n",
            "1000:\tlearn: 0.1422727\ttotal: 4.88s\tremaining: 4.87s\n",
            "1100:\tlearn: 0.1368563\ttotal: 5.63s\tremaining: 4.6s\n",
            "1200:\tlearn: 0.1319655\ttotal: 5.97s\tremaining: 3.97s\n",
            "1300:\tlearn: 0.1271257\ttotal: 6.32s\tremaining: 3.39s\n",
            "1400:\tlearn: 0.1225418\ttotal: 6.65s\tremaining: 2.84s\n",
            "1500:\tlearn: 0.1182077\ttotal: 6.98s\tremaining: 2.32s\n",
            "1600:\tlearn: 0.1141852\ttotal: 7.32s\tremaining: 1.82s\n",
            "1700:\tlearn: 0.1103683\ttotal: 7.66s\tremaining: 1.34s\n",
            "1800:\tlearn: 0.1066581\ttotal: 7.98s\tremaining: 882ms\n",
            "1900:\tlearn: 0.1031007\ttotal: 8.34s\tremaining: 434ms\n",
            "1999:\tlearn: 0.0995880\ttotal: 8.67s\tremaining: 0us\n",
            "0:\tlearn: 0.6201221\ttotal: 5.63ms\tremaining: 11.3s\n",
            "100:\tlearn: 0.1221653\ttotal: 564ms\tremaining: 10.6s\n",
            "200:\tlearn: 0.0797993\ttotal: 1.1s\tremaining: 9.87s\n",
            "300:\tlearn: 0.0569831\ttotal: 1.67s\tremaining: 9.45s\n",
            "400:\tlearn: 0.0429692\ttotal: 2.22s\tremaining: 8.85s\n",
            "500:\tlearn: 0.0329577\ttotal: 2.78s\tremaining: 8.33s\n",
            "600:\tlearn: 0.0258211\ttotal: 3.34s\tremaining: 7.78s\n",
            "700:\tlearn: 0.0203731\ttotal: 3.88s\tremaining: 7.2s\n",
            "800:\tlearn: 0.0160838\ttotal: 4.45s\tremaining: 6.67s\n",
            "900:\tlearn: 0.0130557\ttotal: 5s\tremaining: 6.09s\n",
            "1000:\tlearn: 0.0106159\ttotal: 5.57s\tremaining: 5.55s\n",
            "1100:\tlearn: 0.0087721\ttotal: 6.14s\tremaining: 5.02s\n",
            "1200:\tlearn: 0.0073157\ttotal: 6.86s\tremaining: 4.56s\n",
            "1300:\tlearn: 0.0062841\ttotal: 8.16s\tremaining: 4.38s\n",
            "1400:\tlearn: 0.0054247\ttotal: 9.46s\tremaining: 4.05s\n",
            "1500:\tlearn: 0.0048860\ttotal: 10.3s\tremaining: 3.44s\n",
            "1600:\tlearn: 0.0046357\ttotal: 10.9s\tremaining: 2.71s\n",
            "1700:\tlearn: 0.0043491\ttotal: 11.4s\tremaining: 2s\n",
            "1800:\tlearn: 0.0042348\ttotal: 11.9s\tremaining: 1.32s\n",
            "1900:\tlearn: 0.0041792\ttotal: 12.4s\tremaining: 646ms\n",
            "1999:\tlearn: 0.0039883\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6219955\ttotal: 5.51ms\tremaining: 11s\n",
            "100:\tlearn: 0.1855856\ttotal: 567ms\tremaining: 10.7s\n",
            "200:\tlearn: 0.1478639\ttotal: 1.1s\tremaining: 9.83s\n",
            "300:\tlearn: 0.1178822\ttotal: 1.65s\tremaining: 9.32s\n",
            "400:\tlearn: 0.0967154\ttotal: 2.18s\tremaining: 8.71s\n",
            "500:\tlearn: 0.0794809\ttotal: 2.75s\tremaining: 8.22s\n",
            "600:\tlearn: 0.0659896\ttotal: 3.28s\tremaining: 7.64s\n",
            "700:\tlearn: 0.0557769\ttotal: 3.83s\tremaining: 7.09s\n",
            "800:\tlearn: 0.0475205\ttotal: 4.36s\tremaining: 6.53s\n",
            "900:\tlearn: 0.0403851\ttotal: 4.91s\tremaining: 5.99s\n",
            "1000:\tlearn: 0.0349606\ttotal: 5.45s\tremaining: 5.44s\n",
            "1100:\tlearn: 0.0302566\ttotal: 6s\tremaining: 4.9s\n",
            "1200:\tlearn: 0.0259435\ttotal: 6.56s\tremaining: 4.36s\n",
            "1300:\tlearn: 0.0225711\ttotal: 7.46s\tremaining: 4s\n",
            "1400:\tlearn: 0.0197029\ttotal: 8.71s\tremaining: 3.72s\n",
            "1500:\tlearn: 0.0173748\ttotal: 9.98s\tremaining: 3.32s\n",
            "1600:\tlearn: 0.0152238\ttotal: 10.6s\tremaining: 2.65s\n",
            "1700:\tlearn: 0.0135092\ttotal: 11.2s\tremaining: 1.97s\n",
            "1800:\tlearn: 0.0120849\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0108323\ttotal: 12.3s\tremaining: 642ms\n",
            "1999:\tlearn: 0.0097229\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6227508\ttotal: 5.37ms\tremaining: 10.7s\n",
            "100:\tlearn: 0.2004496\ttotal: 524ms\tremaining: 9.85s\n",
            "200:\tlearn: 0.1602804\ttotal: 1.09s\tremaining: 9.74s\n",
            "300:\tlearn: 0.1305141\ttotal: 1.63s\tremaining: 9.19s\n",
            "400:\tlearn: 0.1087834\ttotal: 2.2s\tremaining: 8.77s\n",
            "500:\tlearn: 0.0917001\ttotal: 2.73s\tremaining: 8.18s\n",
            "600:\tlearn: 0.0773344\ttotal: 3.28s\tremaining: 7.64s\n",
            "700:\tlearn: 0.0657086\ttotal: 3.84s\tremaining: 7.11s\n",
            "800:\tlearn: 0.0559737\ttotal: 4.36s\tremaining: 6.53s\n",
            "900:\tlearn: 0.0478426\ttotal: 4.9s\tremaining: 5.98s\n",
            "1000:\tlearn: 0.0412026\ttotal: 5.43s\tremaining: 5.42s\n",
            "1100:\tlearn: 0.0356101\ttotal: 5.97s\tremaining: 4.88s\n",
            "1200:\tlearn: 0.0308573\ttotal: 6.51s\tremaining: 4.33s\n",
            "1300:\tlearn: 0.0269250\ttotal: 7.14s\tremaining: 3.84s\n",
            "1400:\tlearn: 0.0235232\ttotal: 8.39s\tremaining: 3.59s\n",
            "1500:\tlearn: 0.0207352\ttotal: 9.73s\tremaining: 3.23s\n",
            "1600:\tlearn: 0.0183878\ttotal: 10.5s\tremaining: 2.62s\n",
            "1700:\tlearn: 0.0162213\ttotal: 11.1s\tremaining: 1.95s\n",
            "1800:\tlearn: 0.0144374\ttotal: 11.6s\tremaining: 1.28s\n",
            "1900:\tlearn: 0.0129370\ttotal: 12.2s\tremaining: 633ms\n",
            "1999:\tlearn: 0.0116510\ttotal: 12.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6221085\ttotal: 5.57ms\tremaining: 11.1s\n",
            "100:\tlearn: 0.1972229\ttotal: 546ms\tremaining: 10.3s\n",
            "200:\tlearn: 0.1605018\ttotal: 1.08s\tremaining: 9.65s\n",
            "300:\tlearn: 0.1307175\ttotal: 1.62s\tremaining: 9.16s\n",
            "400:\tlearn: 0.1076946\ttotal: 2.15s\tremaining: 8.59s\n",
            "500:\tlearn: 0.0904924\ttotal: 2.69s\tremaining: 8.06s\n",
            "600:\tlearn: 0.0757838\ttotal: 3.25s\tremaining: 7.56s\n",
            "700:\tlearn: 0.0635417\ttotal: 3.78s\tremaining: 7.01s\n",
            "800:\tlearn: 0.0543541\ttotal: 4.34s\tremaining: 6.5s\n",
            "900:\tlearn: 0.0459525\ttotal: 4.87s\tremaining: 5.94s\n",
            "1000:\tlearn: 0.0396381\ttotal: 5.46s\tremaining: 5.44s\n",
            "1100:\tlearn: 0.0341377\ttotal: 6s\tremaining: 4.9s\n",
            "1200:\tlearn: 0.0298644\ttotal: 6.56s\tremaining: 4.36s\n",
            "1300:\tlearn: 0.0261083\ttotal: 7.09s\tremaining: 3.81s\n",
            "1400:\tlearn: 0.0229340\ttotal: 8.17s\tremaining: 3.49s\n",
            "1500:\tlearn: 0.0201735\ttotal: 9.47s\tremaining: 3.15s\n",
            "1600:\tlearn: 0.0178867\ttotal: 10.6s\tremaining: 2.63s\n",
            "1700:\tlearn: 0.0159420\ttotal: 11.1s\tremaining: 1.95s\n",
            "1800:\tlearn: 0.0142676\ttotal: 11.7s\tremaining: 1.29s\n",
            "1900:\tlearn: 0.0127571\ttotal: 12.2s\tremaining: 635ms\n",
            "1999:\tlearn: 0.0114848\ttotal: 12.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6227409\ttotal: 5.43ms\tremaining: 10.9s\n",
            "100:\tlearn: 0.2015869\ttotal: 562ms\tremaining: 10.6s\n",
            "200:\tlearn: 0.1626762\ttotal: 1.1s\tremaining: 9.82s\n",
            "300:\tlearn: 0.1322656\ttotal: 1.64s\tremaining: 9.26s\n",
            "400:\tlearn: 0.1091387\ttotal: 2.19s\tremaining: 8.72s\n",
            "500:\tlearn: 0.0904441\ttotal: 2.74s\tremaining: 8.2s\n",
            "600:\tlearn: 0.0761777\ttotal: 3.29s\tremaining: 7.65s\n",
            "700:\tlearn: 0.0645932\ttotal: 3.84s\tremaining: 7.12s\n",
            "800:\tlearn: 0.0551144\ttotal: 4.38s\tremaining: 6.56s\n",
            "900:\tlearn: 0.0472130\ttotal: 4.93s\tremaining: 6.02s\n",
            "1000:\tlearn: 0.0406474\ttotal: 5.47s\tremaining: 5.46s\n",
            "1100:\tlearn: 0.0354328\ttotal: 6.05s\tremaining: 4.94s\n",
            "1200:\tlearn: 0.0307194\ttotal: 6.59s\tremaining: 4.38s\n",
            "1300:\tlearn: 0.0268233\ttotal: 7.15s\tremaining: 3.84s\n",
            "1400:\tlearn: 0.0236837\ttotal: 8.04s\tremaining: 3.44s\n",
            "1500:\tlearn: 0.0207744\ttotal: 9.28s\tremaining: 3.08s\n",
            "1600:\tlearn: 0.0183841\ttotal: 10.5s\tremaining: 2.63s\n",
            "1700:\tlearn: 0.0162936\ttotal: 11.1s\tremaining: 1.95s\n",
            "1800:\tlearn: 0.0146446\ttotal: 11.6s\tremaining: 1.28s\n",
            "1900:\tlearn: 0.0131562\ttotal: 12.2s\tremaining: 635ms\n",
            "1999:\tlearn: 0.0117910\ttotal: 12.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6367389\ttotal: 5.97ms\tremaining: 11.9s\n",
            "100:\tlearn: 0.1321354\ttotal: 563ms\tremaining: 10.6s\n",
            "200:\tlearn: 0.0907131\ttotal: 1.15s\tremaining: 10.3s\n",
            "300:\tlearn: 0.0676683\ttotal: 1.69s\tremaining: 9.51s\n",
            "400:\tlearn: 0.0538446\ttotal: 2.24s\tremaining: 8.94s\n",
            "500:\tlearn: 0.0447386\ttotal: 2.77s\tremaining: 8.3s\n",
            "600:\tlearn: 0.0369655\ttotal: 3.35s\tremaining: 7.8s\n",
            "700:\tlearn: 0.0311294\ttotal: 3.89s\tremaining: 7.21s\n",
            "800:\tlearn: 0.0265815\ttotal: 4.45s\tremaining: 6.66s\n",
            "900:\tlearn: 0.0226380\ttotal: 5.01s\tremaining: 6.11s\n",
            "1000:\tlearn: 0.0192693\ttotal: 5.55s\tremaining: 5.54s\n",
            "1100:\tlearn: 0.0167697\ttotal: 6.12s\tremaining: 4.99s\n",
            "1200:\tlearn: 0.0150063\ttotal: 6.65s\tremaining: 4.42s\n",
            "1300:\tlearn: 0.0133147\ttotal: 7.21s\tremaining: 3.88s\n",
            "1400:\tlearn: 0.0118297\ttotal: 8.04s\tremaining: 3.44s\n",
            "1500:\tlearn: 0.0105826\ttotal: 9.16s\tremaining: 3.05s\n",
            "1600:\tlearn: 0.0097917\ttotal: 10.5s\tremaining: 2.61s\n",
            "1700:\tlearn: 0.0091348\ttotal: 11.3s\tremaining: 1.99s\n",
            "1800:\tlearn: 0.0083237\ttotal: 11.8s\tremaining: 1.31s\n",
            "1900:\tlearn: 0.0077114\ttotal: 12.4s\tremaining: 646ms\n",
            "1999:\tlearn: 0.0071388\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6362568\ttotal: 5.42ms\tremaining: 10.8s\n",
            "100:\tlearn: 0.1983941\ttotal: 571ms\tremaining: 10.7s\n",
            "200:\tlearn: 0.1604347\ttotal: 1.1s\tremaining: 9.85s\n",
            "300:\tlearn: 0.1314823\ttotal: 1.65s\tremaining: 9.33s\n",
            "400:\tlearn: 0.1123390\ttotal: 2.2s\tremaining: 8.77s\n",
            "500:\tlearn: 0.0972839\ttotal: 2.73s\tremaining: 8.16s\n",
            "600:\tlearn: 0.0845020\ttotal: 3.28s\tremaining: 7.63s\n",
            "700:\tlearn: 0.0743540\ttotal: 3.82s\tremaining: 7.08s\n",
            "800:\tlearn: 0.0650296\ttotal: 4.38s\tremaining: 6.55s\n",
            "900:\tlearn: 0.0579049\ttotal: 4.91s\tremaining: 5.99s\n",
            "1000:\tlearn: 0.0511330\ttotal: 5.48s\tremaining: 5.46s\n",
            "1100:\tlearn: 0.0456899\ttotal: 6.01s\tremaining: 4.91s\n",
            "1200:\tlearn: 0.0408933\ttotal: 6.55s\tremaining: 4.36s\n",
            "1300:\tlearn: 0.0370312\ttotal: 7.1s\tremaining: 3.82s\n",
            "1400:\tlearn: 0.0334155\ttotal: 7.65s\tremaining: 3.27s\n",
            "1500:\tlearn: 0.0304426\ttotal: 8.74s\tremaining: 2.91s\n",
            "1600:\tlearn: 0.0277736\ttotal: 9.96s\tremaining: 2.48s\n",
            "1700:\tlearn: 0.0253813\ttotal: 11.2s\tremaining: 1.97s\n",
            "1800:\tlearn: 0.0232809\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0213509\ttotal: 12.3s\tremaining: 640ms\n",
            "1999:\tlearn: 0.0196637\ttotal: 12.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6368878\ttotal: 5.48ms\tremaining: 11s\n",
            "100:\tlearn: 0.2088517\ttotal: 546ms\tremaining: 10.3s\n",
            "200:\tlearn: 0.1745870\ttotal: 1.07s\tremaining: 9.61s\n",
            "300:\tlearn: 0.1453802\ttotal: 1.62s\tremaining: 9.15s\n",
            "400:\tlearn: 0.1245890\ttotal: 2.15s\tremaining: 8.59s\n",
            "500:\tlearn: 0.1080243\ttotal: 2.7s\tremaining: 8.07s\n",
            "600:\tlearn: 0.0934156\ttotal: 3.24s\tremaining: 7.55s\n",
            "700:\tlearn: 0.0825661\ttotal: 3.8s\tremaining: 7.04s\n",
            "800:\tlearn: 0.0727361\ttotal: 4.37s\tremaining: 6.54s\n",
            "900:\tlearn: 0.0645731\ttotal: 4.92s\tremaining: 6s\n",
            "1000:\tlearn: 0.0576882\ttotal: 5.47s\tremaining: 5.46s\n",
            "1100:\tlearn: 0.0521867\ttotal: 6s\tremaining: 4.9s\n",
            "1200:\tlearn: 0.0468730\ttotal: 6.55s\tremaining: 4.36s\n",
            "1300:\tlearn: 0.0422171\ttotal: 7.08s\tremaining: 3.81s\n",
            "1400:\tlearn: 0.0383843\ttotal: 7.64s\tremaining: 3.27s\n",
            "1500:\tlearn: 0.0348579\ttotal: 8.44s\tremaining: 2.81s\n",
            "1600:\tlearn: 0.0319575\ttotal: 9.82s\tremaining: 2.45s\n",
            "1700:\tlearn: 0.0293848\ttotal: 11s\tremaining: 1.94s\n",
            "1800:\tlearn: 0.0270633\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0250557\ttotal: 12.3s\tremaining: 641ms\n",
            "1999:\tlearn: 0.0231755\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6366657\ttotal: 5.79ms\tremaining: 11.6s\n",
            "100:\tlearn: 0.2067614\ttotal: 546ms\tremaining: 10.3s\n",
            "200:\tlearn: 0.1703524\ttotal: 1.11s\tremaining: 9.93s\n",
            "300:\tlearn: 0.1428716\ttotal: 1.64s\tremaining: 9.26s\n",
            "400:\tlearn: 0.1223835\ttotal: 2.18s\tremaining: 8.7s\n",
            "500:\tlearn: 0.1072099\ttotal: 2.71s\tremaining: 8.12s\n",
            "600:\tlearn: 0.0938485\ttotal: 3.26s\tremaining: 7.58s\n",
            "700:\tlearn: 0.0824903\ttotal: 3.82s\tremaining: 7.07s\n",
            "800:\tlearn: 0.0729738\ttotal: 4.36s\tremaining: 6.52s\n",
            "900:\tlearn: 0.0649785\ttotal: 4.91s\tremaining: 5.99s\n",
            "1000:\tlearn: 0.0579589\ttotal: 5.45s\tremaining: 5.43s\n",
            "1100:\tlearn: 0.0519097\ttotal: 6.02s\tremaining: 4.92s\n",
            "1200:\tlearn: 0.0465538\ttotal: 6.56s\tremaining: 4.36s\n",
            "1300:\tlearn: 0.0419236\ttotal: 7.11s\tremaining: 3.82s\n",
            "1400:\tlearn: 0.0380271\ttotal: 7.64s\tremaining: 3.27s\n",
            "1500:\tlearn: 0.0347980\ttotal: 8.2s\tremaining: 2.73s\n",
            "1600:\tlearn: 0.0317407\ttotal: 9.27s\tremaining: 2.31s\n",
            "1700:\tlearn: 0.0292879\ttotal: 10.5s\tremaining: 1.84s\n",
            "1800:\tlearn: 0.0269535\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0248619\ttotal: 12.4s\tremaining: 643ms\n",
            "1999:\tlearn: 0.0229455\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6370995\ttotal: 5.96ms\tremaining: 11.9s\n",
            "100:\tlearn: 0.2128846\ttotal: 547ms\tremaining: 10.3s\n",
            "200:\tlearn: 0.1737828\ttotal: 1.1s\tremaining: 9.83s\n",
            "300:\tlearn: 0.1454526\ttotal: 1.63s\tremaining: 9.23s\n",
            "400:\tlearn: 0.1243457\ttotal: 2.19s\tremaining: 8.73s\n",
            "500:\tlearn: 0.1081336\ttotal: 2.73s\tremaining: 8.17s\n",
            "600:\tlearn: 0.0937888\ttotal: 3.29s\tremaining: 7.65s\n",
            "700:\tlearn: 0.0818255\ttotal: 3.82s\tremaining: 7.09s\n",
            "800:\tlearn: 0.0719285\ttotal: 4.38s\tremaining: 6.56s\n",
            "900:\tlearn: 0.0635387\ttotal: 4.91s\tremaining: 5.99s\n",
            "1000:\tlearn: 0.0562134\ttotal: 5.46s\tremaining: 5.45s\n",
            "1100:\tlearn: 0.0502387\ttotal: 5.99s\tremaining: 4.89s\n",
            "1200:\tlearn: 0.0451936\ttotal: 6.54s\tremaining: 4.35s\n",
            "1300:\tlearn: 0.0407990\ttotal: 7.12s\tremaining: 3.82s\n",
            "1400:\tlearn: 0.0369245\ttotal: 7.66s\tremaining: 3.27s\n",
            "1500:\tlearn: 0.0336316\ttotal: 8.26s\tremaining: 2.75s\n",
            "1600:\tlearn: 0.0306901\ttotal: 9.06s\tremaining: 2.26s\n",
            "1700:\tlearn: 0.0283501\ttotal: 10.5s\tremaining: 1.84s\n",
            "1800:\tlearn: 0.0259690\ttotal: 11.7s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0239637\ttotal: 12.3s\tremaining: 639ms\n",
            "1999:\tlearn: 0.0221769\ttotal: 12.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6772010\ttotal: 2.96ms\tremaining: 5.91s\n",
            "100:\tlearn: 0.2595228\ttotal: 1.15s\tremaining: 21.7s\n",
            "200:\tlearn: 0.1819896\ttotal: 2.46s\tremaining: 22.1s\n",
            "300:\tlearn: 0.1512738\ttotal: 3.69s\tremaining: 20.9s\n",
            "400:\tlearn: 0.1316534\ttotal: 4.98s\tremaining: 19.8s\n",
            "500:\tlearn: 0.1176172\ttotal: 6.26s\tremaining: 18.7s\n",
            "600:\tlearn: 0.1060463\ttotal: 7.56s\tremaining: 17.6s\n",
            "700:\tlearn: 0.0956697\ttotal: 8.99s\tremaining: 16.7s\n",
            "800:\tlearn: 0.0868511\ttotal: 11.5s\tremaining: 17.3s\n",
            "900:\tlearn: 0.0790283\ttotal: 13.2s\tremaining: 16.1s\n",
            "1000:\tlearn: 0.0708060\ttotal: 14.5s\tremaining: 14.5s\n",
            "1100:\tlearn: 0.0636852\ttotal: 15.8s\tremaining: 12.9s\n",
            "1200:\tlearn: 0.0579909\ttotal: 17.1s\tremaining: 11.4s\n",
            "1300:\tlearn: 0.0534688\ttotal: 18.4s\tremaining: 9.91s\n",
            "1400:\tlearn: 0.0498310\ttotal: 19.8s\tremaining: 8.45s\n",
            "1500:\tlearn: 0.0465047\ttotal: 21s\tremaining: 7s\n",
            "1600:\tlearn: 0.0433581\ttotal: 22.6s\tremaining: 5.63s\n",
            "1700:\tlearn: 0.0405743\ttotal: 25.3s\tremaining: 4.44s\n",
            "1800:\tlearn: 0.0379931\ttotal: 26.7s\tremaining: 2.95s\n",
            "1900:\tlearn: 0.0355945\ttotal: 28s\tremaining: 1.46s\n",
            "1999:\tlearn: 0.0333772\ttotal: 29.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6770685\ttotal: 2.92ms\tremaining: 5.83s\n",
            "100:\tlearn: 0.2926565\ttotal: 1.07s\tremaining: 20.1s\n",
            "200:\tlearn: 0.2362170\ttotal: 2.32s\tremaining: 20.7s\n",
            "300:\tlearn: 0.2117956\ttotal: 3.6s\tremaining: 20.3s\n",
            "400:\tlearn: 0.1934406\ttotal: 4.91s\tremaining: 19.6s\n",
            "500:\tlearn: 0.1786773\ttotal: 6.32s\tremaining: 18.9s\n",
            "600:\tlearn: 0.1652647\ttotal: 8.89s\tremaining: 20.7s\n",
            "700:\tlearn: 0.1542063\ttotal: 10.5s\tremaining: 19.4s\n",
            "800:\tlearn: 0.1443765\ttotal: 11.7s\tremaining: 17.5s\n",
            "900:\tlearn: 0.1346257\ttotal: 13s\tremaining: 15.9s\n",
            "1000:\tlearn: 0.1259169\ttotal: 14.3s\tremaining: 14.3s\n",
            "1100:\tlearn: 0.1177983\ttotal: 15.6s\tremaining: 12.7s\n",
            "1200:\tlearn: 0.1104710\ttotal: 16.8s\tremaining: 11.2s\n",
            "1300:\tlearn: 0.1037329\ttotal: 18.1s\tremaining: 9.72s\n",
            "1400:\tlearn: 0.0977011\ttotal: 19.4s\tremaining: 8.28s\n",
            "1500:\tlearn: 0.0921507\ttotal: 21.7s\tremaining: 7.23s\n",
            "1600:\tlearn: 0.0870422\ttotal: 23.7s\tremaining: 5.9s\n",
            "1700:\tlearn: 0.0824677\ttotal: 25s\tremaining: 4.39s\n",
            "1800:\tlearn: 0.0781738\ttotal: 26.2s\tremaining: 2.89s\n",
            "1900:\tlearn: 0.0737521\ttotal: 27.4s\tremaining: 1.43s\n",
            "1999:\tlearn: 0.0697408\ttotal: 28.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6772538\ttotal: 2.96ms\tremaining: 5.92s\n",
            "100:\tlearn: 0.2990203\ttotal: 1.05s\tremaining: 19.8s\n",
            "200:\tlearn: 0.2491030\ttotal: 2.25s\tremaining: 20.2s\n",
            "300:\tlearn: 0.2227953\ttotal: 3.5s\tremaining: 19.7s\n",
            "400:\tlearn: 0.2044149\ttotal: 5.59s\tremaining: 22.3s\n",
            "500:\tlearn: 0.1891435\ttotal: 7.6s\tremaining: 22.7s\n",
            "600:\tlearn: 0.1759844\ttotal: 8.87s\tremaining: 20.6s\n",
            "700:\tlearn: 0.1642765\ttotal: 10.1s\tremaining: 18.8s\n",
            "800:\tlearn: 0.1530450\ttotal: 11.4s\tremaining: 17.1s\n",
            "900:\tlearn: 0.1431734\ttotal: 12.7s\tremaining: 15.5s\n",
            "1000:\tlearn: 0.1339446\ttotal: 13.9s\tremaining: 13.9s\n",
            "1100:\tlearn: 0.1260052\ttotal: 15.2s\tremaining: 12.4s\n",
            "1200:\tlearn: 0.1186500\ttotal: 16.5s\tremaining: 11s\n",
            "1300:\tlearn: 0.1121656\ttotal: 18.6s\tremaining: 10s\n",
            "1400:\tlearn: 0.1063036\ttotal: 20.6s\tremaining: 8.81s\n",
            "1500:\tlearn: 0.1007855\ttotal: 21.9s\tremaining: 7.27s\n",
            "1600:\tlearn: 0.0959298\ttotal: 23.1s\tremaining: 5.77s\n",
            "1700:\tlearn: 0.0915254\ttotal: 24.4s\tremaining: 4.29s\n",
            "1800:\tlearn: 0.0870779\ttotal: 25.7s\tremaining: 2.83s\n",
            "1900:\tlearn: 0.0826419\ttotal: 26.9s\tremaining: 1.4s\n",
            "1999:\tlearn: 0.0784255\ttotal: 28.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6771467\ttotal: 3.07ms\tremaining: 6.14s\n",
            "100:\tlearn: 0.2967799\ttotal: 1.05s\tremaining: 19.7s\n",
            "200:\tlearn: 0.2457343\ttotal: 3.04s\tremaining: 27.2s\n",
            "300:\tlearn: 0.2206136\ttotal: 5.36s\tremaining: 30.3s\n",
            "400:\tlearn: 0.2021641\ttotal: 6.65s\tremaining: 26.5s\n",
            "500:\tlearn: 0.1859581\ttotal: 7.94s\tremaining: 23.8s\n",
            "600:\tlearn: 0.1723875\ttotal: 9.18s\tremaining: 21.4s\n",
            "700:\tlearn: 0.1605191\ttotal: 10.4s\tremaining: 19.3s\n",
            "800:\tlearn: 0.1498586\ttotal: 11.7s\tremaining: 17.6s\n",
            "900:\tlearn: 0.1404446\ttotal: 12.9s\tremaining: 15.8s\n",
            "1000:\tlearn: 0.1317549\ttotal: 14.2s\tremaining: 14.2s\n",
            "1100:\tlearn: 0.1236730\ttotal: 16.1s\tremaining: 13.1s\n",
            "1200:\tlearn: 0.1161290\ttotal: 18.6s\tremaining: 12.4s\n",
            "1300:\tlearn: 0.1094551\ttotal: 19.9s\tremaining: 10.7s\n",
            "1400:\tlearn: 0.1037710\ttotal: 21.1s\tremaining: 9.03s\n",
            "1500:\tlearn: 0.0984868\ttotal: 22.4s\tremaining: 7.44s\n",
            "1600:\tlearn: 0.0938932\ttotal: 23.7s\tremaining: 5.9s\n",
            "1700:\tlearn: 0.0890597\ttotal: 24.9s\tremaining: 4.38s\n",
            "1800:\tlearn: 0.0849017\ttotal: 26.2s\tremaining: 2.89s\n",
            "1900:\tlearn: 0.0808040\ttotal: 27.5s\tremaining: 1.43s\n",
            "1999:\tlearn: 0.0770119\ttotal: 29.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6772945\ttotal: 3.52ms\tremaining: 7.03s\n",
            "100:\tlearn: 0.2999643\ttotal: 2.15s\tremaining: 40.4s\n",
            "200:\tlearn: 0.2491939\ttotal: 3.38s\tremaining: 30.3s\n",
            "300:\tlearn: 0.2230225\ttotal: 4.63s\tremaining: 26.1s\n",
            "400:\tlearn: 0.2039077\ttotal: 5.9s\tremaining: 23.5s\n",
            "500:\tlearn: 0.1884974\ttotal: 7.19s\tremaining: 21.5s\n",
            "600:\tlearn: 0.1754852\ttotal: 8.47s\tremaining: 19.7s\n",
            "700:\tlearn: 0.1636070\ttotal: 9.76s\tremaining: 18.1s\n",
            "800:\tlearn: 0.1530573\ttotal: 11s\tremaining: 16.5s\n",
            "900:\tlearn: 0.1431524\ttotal: 12.4s\tremaining: 15.2s\n",
            "1000:\tlearn: 0.1339094\ttotal: 14.8s\tremaining: 14.8s\n",
            "1100:\tlearn: 0.1257609\ttotal: 16.6s\tremaining: 13.5s\n",
            "1200:\tlearn: 0.1182149\ttotal: 17.8s\tremaining: 11.9s\n",
            "1300:\tlearn: 0.1115127\ttotal: 19.1s\tremaining: 10.3s\n",
            "1400:\tlearn: 0.1052712\ttotal: 20.4s\tremaining: 8.73s\n",
            "1500:\tlearn: 0.0995623\ttotal: 21.7s\tremaining: 7.22s\n",
            "1600:\tlearn: 0.0942141\ttotal: 23s\tremaining: 5.73s\n",
            "1700:\tlearn: 0.0892088\ttotal: 24.3s\tremaining: 4.26s\n",
            "1800:\tlearn: 0.0845783\ttotal: 25.5s\tremaining: 2.82s\n",
            "1900:\tlearn: 0.0803128\ttotal: 27.9s\tremaining: 1.45s\n",
            "1999:\tlearn: 0.0763854\ttotal: 29.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6760899\ttotal: 3.79ms\tremaining: 1.89s\n",
            "100:\tlearn: 0.2802690\ttotal: 1.21s\tremaining: 4.78s\n",
            "200:\tlearn: 0.2248432\ttotal: 2.57s\tremaining: 3.82s\n",
            "300:\tlearn: 0.2000283\ttotal: 3.96s\tremaining: 2.62s\n",
            "400:\tlearn: 0.1832215\ttotal: 5.35s\tremaining: 1.32s\n",
            "499:\tlearn: 0.1690160\ttotal: 6.71s\tremaining: 0us\n",
            "0:\tlearn: 0.6760899\ttotal: 3.69ms\tremaining: 1.84s\n",
            "100:\tlearn: 0.2802690\ttotal: 1.2s\tremaining: 4.72s\n",
            "200:\tlearn: 0.2248432\ttotal: 2.54s\tremaining: 3.78s\n",
            "300:\tlearn: 0.2000283\ttotal: 3.91s\tremaining: 2.59s\n",
            "400:\tlearn: 0.1832215\ttotal: 5.42s\tremaining: 1.34s\n",
            "499:\tlearn: 0.1690160\ttotal: 8.01s\tremaining: 0us\n",
            "Validation Set Metrics:\n",
            "Accuracy: 0.8179329226557153\n",
            "Precision: 0.9097283085013146\n",
            "Recall: 0.8642797668609492\n",
            "F1-score: 0.8864218616567037\n",
            "AUC-ROC: 0.7340629603535516\n",
            "Submission file 'Cat-svm-rf_ensemble_submission.csv' has been created.\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split, RandomizedSearchCV\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE\n",
        "from catboost import CatBoostClassifier\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score\n",
        "from sklearn.ensemble import VotingClassifier\n",
        "\n",
        "# Load the datasets\n",
        "train_df = pd.read_csv('/content/Train Dataset.csv')\n",
        "test_df = pd.read_csv('/content/Test Dataset.csv')\n",
        "\n",
        "# Separate features and target in the training data\n",
        "X = train_df.drop('target', axis=1)\n",
        "y = train_df['target']\n",
        "\n",
        "# Split the data into training and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# Identify numeric and categorical columns\n",
        "numeric_features = X.select_dtypes(include=[np.number]).columns\n",
        "categorical_features = X.select_dtypes(exclude=[np.number]).columns\n",
        "\n",
        "# Create preprocessing pipelines\n",
        "numeric_transformer = Pipeline(steps=[\n",
        "    ('scaler', StandardScaler())\n",
        "])\n",
        "\n",
        "categorical_transformer = Pipeline(steps=[\n",
        "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
        "])\n",
        "\n",
        "# Combine preprocessing steps\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numeric_transformer, numeric_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "    ]\n",
        ")\n",
        "\n",
        "# Create the full preprocessing pipeline\n",
        "full_pipeline = Pipeline([\n",
        "    ('preprocessor', preprocessor)\n",
        "])\n",
        "\n",
        "# Fit the pipeline and transform the data\n",
        "X_train_processed = full_pipeline.fit_transform(X_train, y_train)\n",
        "X_val_processed = full_pipeline.transform(X_val)\n",
        "\n",
        "# Change the column names in test_df to match the training data\n",
        "test_df.rename(columns={'id': 'Id', 'age': 'Age', 'sex': 'Sex'}, inplace=True)\n",
        "\n",
        "# Transform the test data\n",
        "X_test_processed = full_pipeline.transform(test_df)\n",
        "\n",
        "# Apply SMOTE to handle class imbalance\n",
        "smote = SMOTE(random_state=42)\n",
        "X_train_processed, y_train_balanced = smote.fit_resample(X_train_processed, y_train)\n",
        "\n",
        "# Define individual models\n",
        "catboost_model = CatBoostClassifier(verbose=100)\n",
        "svm_model = SVC(probability=True)\n",
        "rf_model = RandomForestClassifier(random_state=42)\n",
        "\n",
        "# Define hyperparameter grids for each model (adjust as needed)\n",
        "catboost_param_grid = {\n",
        "    'iterations': [500, 1000, 2000],\n",
        "    'learning_rate': [0.01, 0.03, 0.05],\n",
        "    'depth': [4, 6, 8],\n",
        "    'l2_leaf_reg': [1, 3, 5],\n",
        "    'random_seed': [42, 123, 987]\n",
        "}\n",
        "\n",
        "svm_param_grid = {\n",
        "    'C': [0.1, 1, 10],\n",
        "    'kernel': ['linear', 'rbf', 'poly'],\n",
        "    'gamma': ['scale', 'auto']\n",
        "}\n",
        "\n",
        "rf_param_grid = {\n",
        "    'n_estimators': [100, 200, 300],\n",
        "    'max_depth': [None, 5, 10],\n",
        "    'min_samples_split': [2, 5, 10],\n",
        "    'min_samples_leaf': [1, 2, 4]\n",
        "}\n",
        "\n",
        "# Perform randomized search hyperparameter tuning for each model\n",
        "catboost_random_search = RandomizedSearchCV(catboost_model, param_distributions=catboost_param_grid, n_iter=10, cv=5, scoring='f1_macro')\n",
        "svm_random_search = RandomizedSearchCV(svm_model, param_distributions=svm_param_grid, n_iter=10, cv=5, scoring='f1_macro')\n",
        "rf_random_search = RandomizedSearchCV(rf_model, param_distributions=rf_param_grid, n_iter=10, cv=5, scoring='f1_macro')\n",
        "\n",
        "# Fit the random searches to the training data\n",
        "catboost_random_search.fit(X_train_processed, y_train_balanced)\n",
        "svm_random_search.fit(X_train_processed, y_train_balanced)\n",
        "rf_random_search.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Get the best models from each random search\n",
        "best_catboost_model = catboost_random_search.best_estimator_\n",
        "best_svm_model = svm_random_search.best_estimator_\n",
        "best_rf_model = rf_random_search.best_estimator_\n",
        "\n",
        "# Create the ensemble model using VotingClassifier\n",
        "ensemble_model = VotingClassifier(\n",
        "    estimators=[\n",
        "        ('catboost', best_catboost_model),\n",
        "        ('svm', best_svm_model),\n",
        "        ('rf', best_rf_model)\n",
        "    ],\n",
        "    voting='soft'\n",
        ")\n",
        "\n",
        "# Fit the ensemble model to the training data\n",
        "ensemble_model.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Evaluate the ensemble model on the validation set\n",
        "y_val_pred = ensemble_model.predict(X_val_processed)\n",
        "\n",
        "# Calculate accuracy and other metrics\n",
        "accuracy = accuracy_score(y_val, y_val_pred)\n",
        "precision = precision_score(y_val, y_val_pred)\n",
        "recall = recall_score(y_val, y_val_pred)\n",
        "f1_score = f1_score(y_val, y_val_pred)\n",
        "auc_roc = roc_auc_score(y_val, y_val_pred)\n",
        "\n",
        "print(\"Validation Set Metrics:\")\n",
        "print(\"Accuracy:\", accuracy)\n",
        "print(\"Precision:\", precision)\n",
        "print(\"Recall:\", recall)\n",
        "print(\"F1-score:\", f1_score)\n",
        "print(\"AUC-ROC:\", auc_roc)\n",
        "\n",
        "# Make predictions on the test set\n",
        "test_predictions = ensemble_model.predict(X_test_processed)\n",
        "\n",
        "# Create a submission DataFrame\n",
        "submission = pd.DataFrame({\n",
        "    'ID': test_df['Id'],\n",
        "    'target': test_predictions\n",
        "})\n",
        "\n",
        "# Save the submission file\n",
        "submission.to_csv('Cat-svm-rf_ensemble_submission.csv', index=False)\n",
        "print(\"Submission file 'Cat-svm-rf_ensemble_submission.csv' has been created.\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WpCqlXGDVByD"
      },
      "source": [
        "# **Enhanced Ensemble Model with CatBoost, SVM, and Random Forest**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yu2uRsbhBTGQ",
        "outputId": "cd165f18-8e85-4baf-e420-8bb7e1c3b33f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0:\tlearn: 0.6849289\ttotal: 50.1ms\tremaining: 2m 30s\n",
            "100:\tlearn: 0.3702167\ttotal: 552ms\tremaining: 15.8s\n",
            "200:\tlearn: 0.2938486\ttotal: 1.09s\tremaining: 15.2s\n",
            "300:\tlearn: 0.2592957\ttotal: 1.62s\tremaining: 14.5s\n",
            "400:\tlearn: 0.2385188\ttotal: 2.17s\tremaining: 14.1s\n",
            "500:\tlearn: 0.2269051\ttotal: 2.73s\tremaining: 13.6s\n",
            "600:\tlearn: 0.2168982\ttotal: 3.27s\tremaining: 13.1s\n",
            "700:\tlearn: 0.2089873\ttotal: 3.84s\tremaining: 12.6s\n",
            "800:\tlearn: 0.2021213\ttotal: 4.37s\tremaining: 12s\n",
            "900:\tlearn: 0.1963590\ttotal: 4.92s\tremaining: 11.5s\n",
            "1000:\tlearn: 0.1912010\ttotal: 5.45s\tremaining: 10.9s\n",
            "1100:\tlearn: 0.1867668\ttotal: 6.01s\tremaining: 10.4s\n",
            "1200:\tlearn: 0.1825468\ttotal: 6.77s\tremaining: 10.1s\n",
            "1300:\tlearn: 0.1784083\ttotal: 8.05s\tremaining: 10.5s\n",
            "1400:\tlearn: 0.1742846\ttotal: 9.23s\tremaining: 10.5s\n",
            "1500:\tlearn: 0.1706369\ttotal: 10.1s\tremaining: 10.1s\n",
            "1600:\tlearn: 0.1671478\ttotal: 10.6s\tremaining: 9.28s\n",
            "1700:\tlearn: 0.1636207\ttotal: 11.2s\tremaining: 8.54s\n",
            "1800:\tlearn: 0.1599342\ttotal: 11.7s\tremaining: 7.81s\n",
            "1900:\tlearn: 0.1564091\ttotal: 12.3s\tremaining: 7.1s\n",
            "2000:\tlearn: 0.1531296\ttotal: 12.8s\tremaining: 6.4s\n",
            "2100:\tlearn: 0.1499079\ttotal: 13.4s\tremaining: 5.73s\n",
            "2200:\tlearn: 0.1468506\ttotal: 13.9s\tremaining: 5.06s\n",
            "2300:\tlearn: 0.1437968\ttotal: 14.5s\tremaining: 4.4s\n",
            "2400:\tlearn: 0.1410485\ttotal: 15s\tremaining: 3.75s\n",
            "2500:\tlearn: 0.1385005\ttotal: 15.6s\tremaining: 3.11s\n",
            "2600:\tlearn: 0.1360120\ttotal: 16.1s\tremaining: 2.48s\n",
            "2700:\tlearn: 0.1333488\ttotal: 16.7s\tremaining: 1.85s\n",
            "2800:\tlearn: 0.1310438\ttotal: 17.2s\tremaining: 1.23s\n",
            "2900:\tlearn: 0.1286518\ttotal: 17.8s\tremaining: 607ms\n",
            "2999:\tlearn: 0.1264375\ttotal: 18.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6850157\ttotal: 3.15ms\tremaining: 9.45s\n",
            "100:\tlearn: 0.3728875\ttotal: 538ms\tremaining: 15.4s\n",
            "200:\tlearn: 0.2988398\ttotal: 1.04s\tremaining: 14.5s\n",
            "300:\tlearn: 0.2658620\ttotal: 1.96s\tremaining: 17.6s\n",
            "400:\tlearn: 0.2454030\ttotal: 3.1s\tremaining: 20.1s\n",
            "500:\tlearn: 0.2318228\ttotal: 4.35s\tremaining: 21.7s\n",
            "600:\tlearn: 0.2215363\ttotal: 5.12s\tremaining: 20.4s\n",
            "700:\tlearn: 0.2134932\ttotal: 5.67s\tremaining: 18.6s\n",
            "800:\tlearn: 0.2071238\ttotal: 6.22s\tremaining: 17.1s\n",
            "900:\tlearn: 0.2013448\ttotal: 6.79s\tremaining: 15.8s\n",
            "1000:\tlearn: 0.1961610\ttotal: 7.34s\tremaining: 14.7s\n",
            "1100:\tlearn: 0.1916964\ttotal: 7.9s\tremaining: 13.6s\n",
            "1200:\tlearn: 0.1871023\ttotal: 8.44s\tremaining: 12.6s\n",
            "1300:\tlearn: 0.1829675\ttotal: 9.01s\tremaining: 11.8s\n",
            "1400:\tlearn: 0.1789736\ttotal: 9.56s\tremaining: 10.9s\n",
            "1500:\tlearn: 0.1755978\ttotal: 10.1s\tremaining: 10.1s\n",
            "1600:\tlearn: 0.1719887\ttotal: 10.7s\tremaining: 9.36s\n",
            "1700:\tlearn: 0.1684407\ttotal: 11.3s\tremaining: 8.59s\n",
            "1800:\tlearn: 0.1650484\ttotal: 11.8s\tremaining: 7.87s\n",
            "1900:\tlearn: 0.1615049\ttotal: 12.4s\tremaining: 7.15s\n",
            "2000:\tlearn: 0.1579855\ttotal: 12.9s\tremaining: 6.46s\n",
            "2100:\tlearn: 0.1548995\ttotal: 13.5s\tremaining: 5.77s\n",
            "2200:\tlearn: 0.1515964\ttotal: 14s\tremaining: 5.1s\n",
            "2300:\tlearn: 0.1486746\ttotal: 14.7s\tremaining: 4.47s\n",
            "2400:\tlearn: 0.1460480\ttotal: 15.8s\tremaining: 3.95s\n",
            "2500:\tlearn: 0.1433867\ttotal: 17s\tremaining: 3.4s\n",
            "2600:\tlearn: 0.1409985\ttotal: 18.2s\tremaining: 2.79s\n",
            "2700:\tlearn: 0.1386078\ttotal: 18.7s\tremaining: 2.07s\n",
            "2800:\tlearn: 0.1363035\ttotal: 19.3s\tremaining: 1.37s\n",
            "2900:\tlearn: 0.1341308\ttotal: 19.8s\tremaining: 676ms\n",
            "2999:\tlearn: 0.1320330\ttotal: 20.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6849502\ttotal: 2.92ms\tremaining: 8.77s\n",
            "100:\tlearn: 0.3699299\ttotal: 532ms\tremaining: 15.3s\n",
            "200:\tlearn: 0.2965105\ttotal: 1.03s\tremaining: 14.4s\n",
            "300:\tlearn: 0.2641120\ttotal: 1.59s\tremaining: 14.3s\n",
            "400:\tlearn: 0.2432702\ttotal: 2.14s\tremaining: 13.9s\n",
            "500:\tlearn: 0.2296673\ttotal: 2.71s\tremaining: 13.5s\n",
            "600:\tlearn: 0.2190357\ttotal: 3.25s\tremaining: 13s\n",
            "700:\tlearn: 0.2108527\ttotal: 3.81s\tremaining: 12.5s\n",
            "800:\tlearn: 0.2044811\ttotal: 4.36s\tremaining: 12s\n",
            "900:\tlearn: 0.1989913\ttotal: 4.93s\tremaining: 11.5s\n",
            "1000:\tlearn: 0.1938876\ttotal: 5.48s\tremaining: 10.9s\n",
            "1100:\tlearn: 0.1893927\ttotal: 6.04s\tremaining: 10.4s\n",
            "1200:\tlearn: 0.1849903\ttotal: 6.61s\tremaining: 9.89s\n",
            "1300:\tlearn: 0.1811827\ttotal: 7.14s\tremaining: 9.33s\n",
            "1400:\tlearn: 0.1771633\ttotal: 8.02s\tremaining: 9.15s\n",
            "1500:\tlearn: 0.1734895\ttotal: 9.19s\tremaining: 9.18s\n",
            "1600:\tlearn: 0.1696803\ttotal: 10.4s\tremaining: 9.13s\n",
            "1700:\tlearn: 0.1660988\ttotal: 11.3s\tremaining: 8.64s\n",
            "1800:\tlearn: 0.1625727\ttotal: 11.9s\tremaining: 7.9s\n",
            "1900:\tlearn: 0.1591002\ttotal: 12.4s\tremaining: 7.17s\n",
            "2000:\tlearn: 0.1558091\ttotal: 13s\tremaining: 6.47s\n",
            "2100:\tlearn: 0.1528556\ttotal: 13.5s\tremaining: 5.77s\n",
            "2200:\tlearn: 0.1499566\ttotal: 14.1s\tremaining: 5.1s\n",
            "2300:\tlearn: 0.1469699\ttotal: 14.6s\tremaining: 4.43s\n",
            "2400:\tlearn: 0.1442154\ttotal: 15.1s\tremaining: 3.77s\n",
            "2500:\tlearn: 0.1416626\ttotal: 15.7s\tremaining: 3.13s\n",
            "2600:\tlearn: 0.1391640\ttotal: 16.2s\tremaining: 2.49s\n",
            "2700:\tlearn: 0.1367652\ttotal: 16.8s\tremaining: 1.86s\n",
            "2800:\tlearn: 0.1342274\ttotal: 17.3s\tremaining: 1.23s\n",
            "2900:\tlearn: 0.1319918\ttotal: 17.9s\tremaining: 610ms\n",
            "2999:\tlearn: 0.1298099\ttotal: 18.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6848234\ttotal: 2.73ms\tremaining: 8.18s\n",
            "100:\tlearn: 0.3655791\ttotal: 538ms\tremaining: 15.4s\n",
            "200:\tlearn: 0.2902426\ttotal: 1.09s\tremaining: 15.1s\n",
            "300:\tlearn: 0.2581238\ttotal: 1.62s\tremaining: 14.6s\n",
            "400:\tlearn: 0.2367134\ttotal: 2.16s\tremaining: 14s\n",
            "500:\tlearn: 0.2234769\ttotal: 3.33s\tremaining: 16.6s\n",
            "600:\tlearn: 0.2141053\ttotal: 4.34s\tremaining: 17.3s\n",
            "700:\tlearn: 0.2063986\ttotal: 5.64s\tremaining: 18.5s\n",
            "800:\tlearn: 0.2005402\ttotal: 6.26s\tremaining: 17.2s\n",
            "900:\tlearn: 0.1949591\ttotal: 6.82s\tremaining: 15.9s\n",
            "1000:\tlearn: 0.1899110\ttotal: 7.36s\tremaining: 14.7s\n",
            "1100:\tlearn: 0.1854461\ttotal: 7.93s\tremaining: 13.7s\n",
            "1200:\tlearn: 0.1811289\ttotal: 8.48s\tremaining: 12.7s\n",
            "1300:\tlearn: 0.1772359\ttotal: 9.03s\tremaining: 11.8s\n",
            "1400:\tlearn: 0.1731762\ttotal: 9.58s\tremaining: 10.9s\n",
            "1500:\tlearn: 0.1695573\ttotal: 10.1s\tremaining: 10.1s\n",
            "1600:\tlearn: 0.1660645\ttotal: 10.7s\tremaining: 9.34s\n",
            "1700:\tlearn: 0.1626182\ttotal: 11.2s\tremaining: 8.58s\n",
            "1800:\tlearn: 0.1591716\ttotal: 11.8s\tremaining: 7.85s\n",
            "1900:\tlearn: 0.1557056\ttotal: 12.3s\tremaining: 7.13s\n",
            "2000:\tlearn: 0.1520051\ttotal: 12.9s\tremaining: 6.44s\n",
            "2100:\tlearn: 0.1488027\ttotal: 13.5s\tremaining: 5.76s\n",
            "2200:\tlearn: 0.1452842\ttotal: 14s\tremaining: 5.09s\n",
            "2300:\tlearn: 0.1422240\ttotal: 14.6s\tremaining: 4.43s\n",
            "2400:\tlearn: 0.1392928\ttotal: 16.1s\tremaining: 4.03s\n",
            "2500:\tlearn: 0.1365924\ttotal: 17.4s\tremaining: 3.47s\n",
            "2600:\tlearn: 0.1338918\ttotal: 18.8s\tremaining: 2.88s\n",
            "2700:\tlearn: 0.1314732\ttotal: 19.5s\tremaining: 2.15s\n",
            "2800:\tlearn: 0.1290082\ttotal: 20s\tremaining: 1.42s\n",
            "2900:\tlearn: 0.1266574\ttotal: 20.6s\tremaining: 702ms\n",
            "2999:\tlearn: 0.1242777\ttotal: 21.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6848996\ttotal: 3.72ms\tremaining: 11.1s\n",
            "100:\tlearn: 0.3680553\ttotal: 529ms\tremaining: 15.2s\n",
            "200:\tlearn: 0.2941121\ttotal: 1.05s\tremaining: 14.7s\n",
            "300:\tlearn: 0.2613839\ttotal: 1.61s\tremaining: 14.5s\n",
            "400:\tlearn: 0.2405673\ttotal: 2.17s\tremaining: 14.1s\n",
            "500:\tlearn: 0.2281903\ttotal: 2.75s\tremaining: 13.7s\n",
            "600:\tlearn: 0.2184132\ttotal: 3.3s\tremaining: 13.2s\n",
            "700:\tlearn: 0.2100810\ttotal: 3.9s\tremaining: 12.8s\n",
            "800:\tlearn: 0.2036710\ttotal: 4.45s\tremaining: 12.2s\n",
            "900:\tlearn: 0.1979447\ttotal: 5.02s\tremaining: 11.7s\n",
            "1000:\tlearn: 0.1926032\ttotal: 5.6s\tremaining: 11.2s\n",
            "1100:\tlearn: 0.1878643\ttotal: 6.15s\tremaining: 10.6s\n",
            "1200:\tlearn: 0.1830854\ttotal: 6.72s\tremaining: 10.1s\n",
            "1300:\tlearn: 0.1792115\ttotal: 7.27s\tremaining: 9.49s\n",
            "1400:\tlearn: 0.1750275\ttotal: 8.05s\tremaining: 9.18s\n",
            "1500:\tlearn: 0.1713379\ttotal: 9.29s\tremaining: 9.28s\n",
            "1600:\tlearn: 0.1679463\ttotal: 10.6s\tremaining: 9.24s\n",
            "1700:\tlearn: 0.1645556\ttotal: 11.5s\tremaining: 8.76s\n",
            "1800:\tlearn: 0.1610316\ttotal: 12s\tremaining: 8.02s\n",
            "1900:\tlearn: 0.1576126\ttotal: 12.6s\tremaining: 7.28s\n",
            "2000:\tlearn: 0.1540751\ttotal: 13.2s\tremaining: 6.58s\n",
            "2100:\tlearn: 0.1509898\ttotal: 13.7s\tremaining: 5.88s\n",
            "2200:\tlearn: 0.1482915\ttotal: 14.3s\tremaining: 5.18s\n",
            "2300:\tlearn: 0.1454052\ttotal: 14.8s\tremaining: 4.51s\n",
            "2400:\tlearn: 0.1424040\ttotal: 15.4s\tremaining: 3.84s\n",
            "2500:\tlearn: 0.1397600\ttotal: 15.9s\tremaining: 3.18s\n",
            "2600:\tlearn: 0.1372237\ttotal: 16.5s\tremaining: 2.53s\n",
            "2700:\tlearn: 0.1347368\ttotal: 17.1s\tremaining: 1.89s\n",
            "2800:\tlearn: 0.1325286\ttotal: 17.6s\tremaining: 1.25s\n",
            "2900:\tlearn: 0.1301060\ttotal: 18.2s\tremaining: 621ms\n",
            "2999:\tlearn: 0.1278615\ttotal: 18.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6844075\ttotal: 2.95ms\tremaining: 5.9s\n",
            "100:\tlearn: 0.3590925\ttotal: 524ms\tremaining: 9.85s\n",
            "200:\tlearn: 0.2863861\ttotal: 1.06s\tremaining: 9.46s\n",
            "300:\tlearn: 0.2552085\ttotal: 1.6s\tremaining: 9.04s\n",
            "400:\tlearn: 0.2335641\ttotal: 2.29s\tremaining: 9.15s\n",
            "500:\tlearn: 0.2207115\ttotal: 3.44s\tremaining: 10.3s\n",
            "600:\tlearn: 0.2113543\ttotal: 4.89s\tremaining: 11.4s\n",
            "700:\tlearn: 0.2038072\ttotal: 5.89s\tremaining: 10.9s\n",
            "800:\tlearn: 0.1974884\ttotal: 6.47s\tremaining: 9.68s\n",
            "900:\tlearn: 0.1914106\ttotal: 7.01s\tremaining: 8.55s\n",
            "1000:\tlearn: 0.1862874\ttotal: 7.58s\tremaining: 7.57s\n",
            "1100:\tlearn: 0.1817207\ttotal: 8.15s\tremaining: 6.66s\n",
            "1200:\tlearn: 0.1772067\ttotal: 8.71s\tremaining: 5.79s\n",
            "1300:\tlearn: 0.1729240\ttotal: 9.27s\tremaining: 4.98s\n",
            "1400:\tlearn: 0.1687002\ttotal: 9.82s\tremaining: 4.2s\n",
            "1500:\tlearn: 0.1650208\ttotal: 10.4s\tremaining: 3.45s\n",
            "1600:\tlearn: 0.1612969\ttotal: 10.9s\tremaining: 2.73s\n",
            "1700:\tlearn: 0.1575750\ttotal: 11.5s\tremaining: 2.02s\n",
            "1800:\tlearn: 0.1540697\ttotal: 12s\tremaining: 1.33s\n",
            "1900:\tlearn: 0.1503087\ttotal: 12.6s\tremaining: 657ms\n",
            "1999:\tlearn: 0.1466661\ttotal: 13.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6844929\ttotal: 3.02ms\tremaining: 6.04s\n",
            "100:\tlearn: 0.3624252\ttotal: 519ms\tremaining: 9.75s\n",
            "200:\tlearn: 0.2907750\ttotal: 1.08s\tremaining: 9.64s\n",
            "300:\tlearn: 0.2581778\ttotal: 1.62s\tremaining: 9.13s\n",
            "400:\tlearn: 0.2384507\ttotal: 2.24s\tremaining: 8.92s\n",
            "500:\tlearn: 0.2262720\ttotal: 3.61s\tremaining: 10.8s\n",
            "600:\tlearn: 0.2166521\ttotal: 4.89s\tremaining: 11.4s\n",
            "700:\tlearn: 0.2087555\ttotal: 5.83s\tremaining: 10.8s\n",
            "800:\tlearn: 0.2027198\ttotal: 6.41s\tremaining: 9.59s\n",
            "900:\tlearn: 0.1967584\ttotal: 7.02s\tremaining: 8.56s\n",
            "1000:\tlearn: 0.1914429\ttotal: 7.58s\tremaining: 7.56s\n",
            "1100:\tlearn: 0.1866374\ttotal: 8.15s\tremaining: 6.65s\n",
            "1200:\tlearn: 0.1820353\ttotal: 8.72s\tremaining: 5.8s\n",
            "1300:\tlearn: 0.1778491\ttotal: 9.29s\tremaining: 4.99s\n",
            "1400:\tlearn: 0.1737500\ttotal: 9.85s\tremaining: 4.21s\n",
            "1500:\tlearn: 0.1699622\ttotal: 10.4s\tremaining: 3.46s\n",
            "1600:\tlearn: 0.1662711\ttotal: 11s\tremaining: 2.74s\n",
            "1700:\tlearn: 0.1626741\ttotal: 11.5s\tremaining: 2.03s\n",
            "1800:\tlearn: 0.1590811\ttotal: 12.1s\tremaining: 1.34s\n",
            "1900:\tlearn: 0.1556431\ttotal: 12.7s\tremaining: 659ms\n",
            "1999:\tlearn: 0.1519588\ttotal: 13.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6844298\ttotal: 2.88ms\tremaining: 5.76s\n",
            "100:\tlearn: 0.3593491\ttotal: 531ms\tremaining: 9.98s\n",
            "200:\tlearn: 0.2880562\ttotal: 1.09s\tremaining: 9.73s\n",
            "300:\tlearn: 0.2570938\ttotal: 1.63s\tremaining: 9.2s\n",
            "400:\tlearn: 0.2362208\ttotal: 2.4s\tremaining: 9.58s\n",
            "500:\tlearn: 0.2238402\ttotal: 3.74s\tremaining: 11.2s\n",
            "600:\tlearn: 0.2136095\ttotal: 4.98s\tremaining: 11.6s\n",
            "700:\tlearn: 0.2060237\ttotal: 5.84s\tremaining: 10.8s\n",
            "800:\tlearn: 0.1999429\ttotal: 6.41s\tremaining: 9.59s\n",
            "900:\tlearn: 0.1945486\ttotal: 6.96s\tremaining: 8.49s\n",
            "1000:\tlearn: 0.1893539\ttotal: 7.52s\tremaining: 7.5s\n",
            "1100:\tlearn: 0.1846857\ttotal: 8.09s\tremaining: 6.6s\n",
            "1200:\tlearn: 0.1802026\ttotal: 8.63s\tremaining: 5.74s\n",
            "1300:\tlearn: 0.1760982\ttotal: 9.19s\tremaining: 4.93s\n",
            "1400:\tlearn: 0.1720056\ttotal: 9.77s\tremaining: 4.18s\n",
            "1500:\tlearn: 0.1682622\ttotal: 10.3s\tremaining: 3.44s\n",
            "1600:\tlearn: 0.1644831\ttotal: 10.9s\tremaining: 2.71s\n",
            "1700:\tlearn: 0.1607538\ttotal: 11.4s\tremaining: 2.01s\n",
            "1800:\tlearn: 0.1571132\ttotal: 12s\tremaining: 1.32s\n",
            "1900:\tlearn: 0.1534737\ttotal: 12.5s\tremaining: 652ms\n",
            "1999:\tlearn: 0.1499757\ttotal: 13.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6842976\ttotal: 2.74ms\tremaining: 5.48s\n",
            "100:\tlearn: 0.3529742\ttotal: 511ms\tremaining: 9.61s\n",
            "200:\tlearn: 0.2836692\ttotal: 1.05s\tremaining: 9.41s\n",
            "300:\tlearn: 0.2531307\ttotal: 1.57s\tremaining: 8.87s\n",
            "400:\tlearn: 0.2313319\ttotal: 2.17s\tremaining: 8.64s\n",
            "500:\tlearn: 0.2184175\ttotal: 3.53s\tremaining: 10.6s\n",
            "600:\tlearn: 0.2092054\ttotal: 4.77s\tremaining: 11.1s\n",
            "700:\tlearn: 0.2019148\ttotal: 5.83s\tremaining: 10.8s\n",
            "800:\tlearn: 0.1962331\ttotal: 6.38s\tremaining: 9.55s\n",
            "900:\tlearn: 0.1905397\ttotal: 6.92s\tremaining: 8.44s\n",
            "1000:\tlearn: 0.1855159\ttotal: 7.47s\tremaining: 7.46s\n",
            "1100:\tlearn: 0.1808641\ttotal: 8.03s\tremaining: 6.56s\n",
            "1200:\tlearn: 0.1765032\ttotal: 8.58s\tremaining: 5.71s\n",
            "1300:\tlearn: 0.1725044\ttotal: 9.13s\tremaining: 4.91s\n",
            "1400:\tlearn: 0.1685046\ttotal: 9.71s\tremaining: 4.15s\n",
            "1500:\tlearn: 0.1648284\ttotal: 10.3s\tremaining: 3.42s\n",
            "1600:\tlearn: 0.1611087\ttotal: 10.8s\tremaining: 2.7s\n",
            "1700:\tlearn: 0.1575095\ttotal: 11.4s\tremaining: 2s\n",
            "1800:\tlearn: 0.1538747\ttotal: 11.9s\tremaining: 1.32s\n",
            "1900:\tlearn: 0.1503461\ttotal: 12.5s\tremaining: 651ms\n",
            "1999:\tlearn: 0.1467289\ttotal: 13s\tremaining: 0us\n",
            "0:\tlearn: 0.6843736\ttotal: 2.73ms\tremaining: 5.46s\n",
            "100:\tlearn: 0.3568502\ttotal: 510ms\tremaining: 9.6s\n",
            "200:\tlearn: 0.2862356\ttotal: 1.05s\tremaining: 9.42s\n",
            "300:\tlearn: 0.2547387\ttotal: 1.59s\tremaining: 8.99s\n",
            "400:\tlearn: 0.2362175\ttotal: 2.15s\tremaining: 8.58s\n",
            "500:\tlearn: 0.2249817\ttotal: 3.33s\tremaining: 9.98s\n",
            "600:\tlearn: 0.2150017\ttotal: 4.67s\tremaining: 10.9s\n",
            "700:\tlearn: 0.2062723\ttotal: 5.58s\tremaining: 10.3s\n",
            "800:\tlearn: 0.1992246\ttotal: 6.16s\tremaining: 9.22s\n",
            "900:\tlearn: 0.1932025\ttotal: 6.73s\tremaining: 8.21s\n",
            "1000:\tlearn: 0.1879736\ttotal: 7.29s\tremaining: 7.28s\n",
            "1100:\tlearn: 0.1830240\ttotal: 7.85s\tremaining: 6.41s\n",
            "1200:\tlearn: 0.1785086\ttotal: 8.42s\tremaining: 5.6s\n",
            "1300:\tlearn: 0.1743498\ttotal: 8.97s\tremaining: 4.82s\n",
            "1400:\tlearn: 0.1704070\ttotal: 9.53s\tremaining: 4.08s\n",
            "1500:\tlearn: 0.1664806\ttotal: 10.1s\tremaining: 3.35s\n",
            "1600:\tlearn: 0.1628772\ttotal: 10.7s\tremaining: 2.66s\n",
            "1700:\tlearn: 0.1592271\ttotal: 11.9s\tremaining: 2.09s\n",
            "1800:\tlearn: 0.1555461\ttotal: 13.2s\tremaining: 1.45s\n",
            "1900:\tlearn: 0.1519595\ttotal: 13.7s\tremaining: 715ms\n",
            "1999:\tlearn: 0.1484205\ttotal: 14.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6590140\ttotal: 3.06ms\tremaining: 3.05s\n",
            "100:\tlearn: 0.2086665\ttotal: 4.87s\tremaining: 43.3s\n",
            "200:\tlearn: 0.1416141\ttotal: 8.5s\tremaining: 33.8s\n",
            "300:\tlearn: 0.1013313\ttotal: 12.1s\tremaining: 28.2s\n",
            "400:\tlearn: 0.0744493\ttotal: 17s\tremaining: 25.4s\n",
            "500:\tlearn: 0.0575410\ttotal: 20.7s\tremaining: 20.7s\n",
            "600:\tlearn: 0.0458989\ttotal: 24.3s\tremaining: 16.2s\n",
            "700:\tlearn: 0.0373236\ttotal: 28.5s\tremaining: 12.1s\n",
            "800:\tlearn: 0.0309314\ttotal: 33.1s\tremaining: 8.23s\n",
            "900:\tlearn: 0.0260076\ttotal: 36.7s\tremaining: 4.03s\n",
            "999:\tlearn: 0.0223259\ttotal: 40.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6594064\ttotal: 2.87ms\tremaining: 2.87s\n",
            "100:\tlearn: 0.2131869\ttotal: 4.87s\tremaining: 43.4s\n",
            "200:\tlearn: 0.1464729\ttotal: 8.49s\tremaining: 33.8s\n",
            "300:\tlearn: 0.1049121\ttotal: 12.1s\tremaining: 28.1s\n",
            "400:\tlearn: 0.0777616\ttotal: 17.3s\tremaining: 25.8s\n",
            "500:\tlearn: 0.0602667\ttotal: 20.9s\tremaining: 20.8s\n",
            "600:\tlearn: 0.0486871\ttotal: 24.6s\tremaining: 16.3s\n",
            "700:\tlearn: 0.0396294\ttotal: 30.3s\tremaining: 12.9s\n",
            "800:\tlearn: 0.0326993\ttotal: 35.9s\tremaining: 8.93s\n",
            "900:\tlearn: 0.0273855\ttotal: 39.7s\tremaining: 4.36s\n",
            "999:\tlearn: 0.0235225\ttotal: 44.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6591319\ttotal: 2.93ms\tremaining: 2.93s\n",
            "100:\tlearn: 0.2109581\ttotal: 3.21s\tremaining: 28.6s\n",
            "200:\tlearn: 0.1468160\ttotal: 6.75s\tremaining: 26.8s\n",
            "300:\tlearn: 0.1044023\ttotal: 10.7s\tremaining: 24.8s\n",
            "400:\tlearn: 0.0774795\ttotal: 15.5s\tremaining: 23.1s\n",
            "500:\tlearn: 0.0607115\ttotal: 19.2s\tremaining: 19.1s\n",
            "600:\tlearn: 0.0487912\ttotal: 22.8s\tremaining: 15.1s\n",
            "700:\tlearn: 0.0400200\ttotal: 27.9s\tremaining: 11.9s\n",
            "800:\tlearn: 0.0330774\ttotal: 31.5s\tremaining: 7.82s\n",
            "900:\tlearn: 0.0276082\ttotal: 35.1s\tremaining: 3.86s\n",
            "999:\tlearn: 0.0235308\ttotal: 40.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6586309\ttotal: 2.9ms\tremaining: 2.9s\n",
            "100:\tlearn: 0.2073298\ttotal: 3.27s\tremaining: 29.1s\n",
            "200:\tlearn: 0.1406235\ttotal: 6.91s\tremaining: 27.5s\n",
            "300:\tlearn: 0.1006856\ttotal: 11.7s\tremaining: 27.3s\n",
            "400:\tlearn: 0.0729240\ttotal: 15.6s\tremaining: 23.3s\n",
            "500:\tlearn: 0.0560852\ttotal: 19.3s\tremaining: 19.2s\n",
            "600:\tlearn: 0.0447404\ttotal: 23.5s\tremaining: 15.6s\n",
            "700:\tlearn: 0.0359915\ttotal: 28.1s\tremaining: 12s\n",
            "800:\tlearn: 0.0297454\ttotal: 31.8s\tremaining: 7.9s\n",
            "900:\tlearn: 0.0246574\ttotal: 35.5s\tremaining: 3.9s\n",
            "999:\tlearn: 0.0204828\ttotal: 40.6s\tremaining: 0us\n",
            "0:\tlearn: 0.6589918\ttotal: 2.96ms\tremaining: 2.96s\n",
            "100:\tlearn: 0.2097053\ttotal: 3.21s\tremaining: 28.6s\n",
            "200:\tlearn: 0.1441171\ttotal: 6.88s\tremaining: 27.3s\n",
            "300:\tlearn: 0.1037385\ttotal: 11.9s\tremaining: 27.7s\n",
            "400:\tlearn: 0.0768948\ttotal: 15.5s\tremaining: 23.2s\n",
            "500:\tlearn: 0.0597963\ttotal: 19.2s\tremaining: 19.1s\n",
            "600:\tlearn: 0.0479334\ttotal: 24.3s\tremaining: 16.1s\n",
            "700:\tlearn: 0.0392803\ttotal: 28.1s\tremaining: 12s\n",
            "800:\tlearn: 0.0327087\ttotal: 31.7s\tremaining: 7.88s\n",
            "900:\tlearn: 0.0276622\ttotal: 36.3s\tremaining: 3.99s\n",
            "999:\tlearn: 0.0235799\ttotal: 40.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6852436\ttotal: 2.81ms\tremaining: 2.8s\n",
            "100:\tlearn: 0.3780766\ttotal: 985ms\tremaining: 8.77s\n",
            "200:\tlearn: 0.2935879\ttotal: 2.12s\tremaining: 8.41s\n",
            "300:\tlearn: 0.2586953\ttotal: 3.3s\tremaining: 7.67s\n",
            "400:\tlearn: 0.2370745\ttotal: 4.54s\tremaining: 6.79s\n",
            "500:\tlearn: 0.2217851\ttotal: 5.79s\tremaining: 5.77s\n",
            "600:\tlearn: 0.2097103\ttotal: 7.58s\tremaining: 5.03s\n",
            "700:\tlearn: 0.1998340\ttotal: 10s\tremaining: 4.28s\n",
            "800:\tlearn: 0.1907364\ttotal: 11.3s\tremaining: 2.81s\n",
            "900:\tlearn: 0.1829854\ttotal: 12.5s\tremaining: 1.38s\n",
            "999:\tlearn: 0.1754610\ttotal: 13.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6853284\ttotal: 2.97ms\tremaining: 2.97s\n",
            "100:\tlearn: 0.3812160\ttotal: 1.08s\tremaining: 9.66s\n",
            "200:\tlearn: 0.2977474\ttotal: 2.23s\tremaining: 8.84s\n",
            "300:\tlearn: 0.2626837\ttotal: 3.43s\tremaining: 7.96s\n",
            "400:\tlearn: 0.2404415\ttotal: 4.67s\tremaining: 6.98s\n",
            "500:\tlearn: 0.2254268\ttotal: 6.03s\tremaining: 6.01s\n",
            "600:\tlearn: 0.2138953\ttotal: 8.72s\tremaining: 5.79s\n",
            "700:\tlearn: 0.2043874\ttotal: 10.1s\tremaining: 4.32s\n",
            "800:\tlearn: 0.1958554\ttotal: 11.4s\tremaining: 2.83s\n",
            "900:\tlearn: 0.1878630\ttotal: 12.6s\tremaining: 1.39s\n",
            "999:\tlearn: 0.1806784\ttotal: 13.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6852649\ttotal: 2.99ms\tremaining: 2.98s\n",
            "100:\tlearn: 0.3780012\ttotal: 1.02s\tremaining: 9.09s\n",
            "200:\tlearn: 0.2948869\ttotal: 2.13s\tremaining: 8.49s\n",
            "300:\tlearn: 0.2587000\ttotal: 3.33s\tremaining: 7.74s\n",
            "400:\tlearn: 0.2388561\ttotal: 4.56s\tremaining: 6.82s\n",
            "500:\tlearn: 0.2236818\ttotal: 6.76s\tremaining: 6.73s\n",
            "600:\tlearn: 0.2120424\ttotal: 8.73s\tremaining: 5.8s\n",
            "700:\tlearn: 0.2022223\ttotal: 9.97s\tremaining: 4.25s\n",
            "800:\tlearn: 0.1934170\ttotal: 11.2s\tremaining: 2.79s\n",
            "900:\tlearn: 0.1854887\ttotal: 12.5s\tremaining: 1.37s\n",
            "999:\tlearn: 0.1784584\ttotal: 13.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6851405\ttotal: 3.12ms\tremaining: 3.12s\n",
            "100:\tlearn: 0.3749232\ttotal: 1.01s\tremaining: 9s\n",
            "200:\tlearn: 0.2907369\ttotal: 2.12s\tremaining: 8.44s\n",
            "300:\tlearn: 0.2561812\ttotal: 3.34s\tremaining: 7.76s\n",
            "400:\tlearn: 0.2355944\ttotal: 5s\tremaining: 7.46s\n",
            "500:\tlearn: 0.2205425\ttotal: 7.47s\tremaining: 7.44s\n",
            "600:\tlearn: 0.2087647\ttotal: 8.71s\tremaining: 5.78s\n",
            "700:\tlearn: 0.1985515\ttotal: 9.98s\tremaining: 4.26s\n",
            "800:\tlearn: 0.1901201\ttotal: 11.3s\tremaining: 2.8s\n",
            "900:\tlearn: 0.1826467\ttotal: 12.5s\tremaining: 1.38s\n",
            "999:\tlearn: 0.1752740\ttotal: 13.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6852176\ttotal: 2.87ms\tremaining: 2.87s\n",
            "100:\tlearn: 0.3778506\ttotal: 1.02s\tremaining: 9.04s\n",
            "200:\tlearn: 0.2942347\ttotal: 2.16s\tremaining: 8.6s\n",
            "300:\tlearn: 0.2595617\ttotal: 3.36s\tremaining: 7.81s\n",
            "400:\tlearn: 0.2375882\ttotal: 5.97s\tremaining: 8.92s\n",
            "500:\tlearn: 0.2227781\ttotal: 7.58s\tremaining: 7.55s\n",
            "600:\tlearn: 0.2119575\ttotal: 8.86s\tremaining: 5.88s\n",
            "700:\tlearn: 0.2020111\ttotal: 10.1s\tremaining: 4.32s\n",
            "800:\tlearn: 0.1929771\ttotal: 11.4s\tremaining: 2.83s\n",
            "900:\tlearn: 0.1853320\ttotal: 12.7s\tremaining: 1.39s\n",
            "999:\tlearn: 0.1781300\ttotal: 13.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6870521\ttotal: 11.9ms\tremaining: 35.8s\n",
            "100:\tlearn: 0.3718630\ttotal: 1.09s\tremaining: 31.2s\n",
            "200:\tlearn: 0.2889148\ttotal: 2.23s\tremaining: 31.1s\n",
            "300:\tlearn: 0.2548081\ttotal: 4.35s\tremaining: 39s\n",
            "400:\tlearn: 0.2337375\ttotal: 6.44s\tremaining: 41.8s\n",
            "500:\tlearn: 0.2177216\ttotal: 7.68s\tremaining: 38.3s\n",
            "600:\tlearn: 0.2061095\ttotal: 8.94s\tremaining: 35.7s\n",
            "700:\tlearn: 0.1951439\ttotal: 10.2s\tremaining: 33.4s\n",
            "800:\tlearn: 0.1859570\ttotal: 11.4s\tremaining: 31.3s\n",
            "900:\tlearn: 0.1773362\ttotal: 12.6s\tremaining: 29.4s\n",
            "1000:\tlearn: 0.1696604\ttotal: 13.9s\tremaining: 27.7s\n",
            "1100:\tlearn: 0.1625505\ttotal: 15.1s\tremaining: 26.1s\n",
            "1200:\tlearn: 0.1559986\ttotal: 16.9s\tremaining: 25.4s\n",
            "1300:\tlearn: 0.1498398\ttotal: 19.1s\tremaining: 25s\n",
            "1400:\tlearn: 0.1442031\ttotal: 20.4s\tremaining: 23.3s\n",
            "1500:\tlearn: 0.1386029\ttotal: 21.6s\tremaining: 21.6s\n",
            "1600:\tlearn: 0.1336207\ttotal: 22.9s\tremaining: 20s\n",
            "1700:\tlearn: 0.1285794\ttotal: 24.1s\tremaining: 18.4s\n",
            "1800:\tlearn: 0.1239130\ttotal: 25.3s\tremaining: 16.9s\n",
            "1900:\tlearn: 0.1192380\ttotal: 26.6s\tremaining: 15.4s\n",
            "2000:\tlearn: 0.1148347\ttotal: 27.8s\tremaining: 13.9s\n",
            "2100:\tlearn: 0.1106871\ttotal: 29.4s\tremaining: 12.6s\n",
            "2200:\tlearn: 0.1068864\ttotal: 32s\tremaining: 11.6s\n",
            "2300:\tlearn: 0.1032362\ttotal: 33.2s\tremaining: 10.1s\n",
            "2400:\tlearn: 0.0997916\ttotal: 34.4s\tremaining: 8.59s\n",
            "2500:\tlearn: 0.0965328\ttotal: 35.7s\tremaining: 7.13s\n",
            "2600:\tlearn: 0.0934634\ttotal: 37s\tremaining: 5.67s\n",
            "2700:\tlearn: 0.0905805\ttotal: 38.2s\tremaining: 4.23s\n",
            "2800:\tlearn: 0.0877233\ttotal: 39.4s\tremaining: 2.8s\n",
            "2900:\tlearn: 0.0851705\ttotal: 40.7s\tremaining: 1.39s\n",
            "2999:\tlearn: 0.0828746\ttotal: 42s\tremaining: 0us\n",
            "0:\tlearn: 0.6871162\ttotal: 16.6ms\tremaining: 49.6s\n",
            "100:\tlearn: 0.3762849\ttotal: 2.37s\tremaining: 1m 7s\n",
            "200:\tlearn: 0.2921564\ttotal: 3.75s\tremaining: 52.3s\n",
            "300:\tlearn: 0.2556317\ttotal: 4.95s\tremaining: 44.4s\n",
            "400:\tlearn: 0.2346940\ttotal: 6.2s\tremaining: 40.2s\n",
            "500:\tlearn: 0.2196354\ttotal: 7.46s\tremaining: 37.2s\n",
            "600:\tlearn: 0.2071306\ttotal: 8.73s\tremaining: 34.9s\n",
            "700:\tlearn: 0.1971736\ttotal: 10s\tremaining: 32.8s\n",
            "800:\tlearn: 0.1879399\ttotal: 11.2s\tremaining: 30.9s\n",
            "900:\tlearn: 0.1801303\ttotal: 12.5s\tremaining: 29.1s\n",
            "1000:\tlearn: 0.1731042\ttotal: 14.9s\tremaining: 29.8s\n",
            "1100:\tlearn: 0.1661613\ttotal: 16.8s\tremaining: 28.9s\n",
            "1200:\tlearn: 0.1594857\ttotal: 18.1s\tremaining: 27.1s\n",
            "1300:\tlearn: 0.1534432\ttotal: 19.4s\tremaining: 25.3s\n",
            "1400:\tlearn: 0.1477283\ttotal: 20.7s\tremaining: 23.6s\n",
            "1500:\tlearn: 0.1422667\ttotal: 22.1s\tremaining: 22s\n",
            "1600:\tlearn: 0.1371533\ttotal: 23.5s\tremaining: 20.6s\n",
            "1700:\tlearn: 0.1320757\ttotal: 24.9s\tremaining: 19s\n",
            "1800:\tlearn: 0.1275730\ttotal: 26.7s\tremaining: 17.7s\n",
            "1900:\tlearn: 0.1232790\ttotal: 29.4s\tremaining: 17s\n",
            "2000:\tlearn: 0.1191417\ttotal: 30.7s\tremaining: 15.3s\n",
            "2100:\tlearn: 0.1153640\ttotal: 32s\tremaining: 13.7s\n",
            "2200:\tlearn: 0.1115415\ttotal: 33.4s\tremaining: 12.1s\n",
            "2300:\tlearn: 0.1079608\ttotal: 34.6s\tremaining: 10.5s\n",
            "2400:\tlearn: 0.1048192\ttotal: 35.9s\tremaining: 8.96s\n",
            "2500:\tlearn: 0.1016948\ttotal: 37.2s\tremaining: 7.42s\n",
            "2600:\tlearn: 0.0987396\ttotal: 38.5s\tremaining: 5.9s\n",
            "2700:\tlearn: 0.0959533\ttotal: 40.2s\tremaining: 4.45s\n",
            "2800:\tlearn: 0.0933771\ttotal: 42.7s\tremaining: 3.03s\n",
            "2900:\tlearn: 0.0907473\ttotal: 44s\tremaining: 1.5s\n",
            "2999:\tlearn: 0.0883684\ttotal: 45.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6867492\ttotal: 21.6ms\tremaining: 1m 4s\n",
            "100:\tlearn: 0.3727338\ttotal: 1.14s\tremaining: 32.6s\n",
            "200:\tlearn: 0.2914114\ttotal: 2.29s\tremaining: 32s\n",
            "300:\tlearn: 0.2543703\ttotal: 3.57s\tremaining: 32s\n",
            "400:\tlearn: 0.2329630\ttotal: 4.84s\tremaining: 31.4s\n",
            "500:\tlearn: 0.2179842\ttotal: 6.1s\tremaining: 30.4s\n",
            "600:\tlearn: 0.2057205\ttotal: 7.82s\tremaining: 31.2s\n",
            "700:\tlearn: 0.1951377\ttotal: 10.3s\tremaining: 33.9s\n",
            "800:\tlearn: 0.1866081\ttotal: 11.6s\tremaining: 31.8s\n",
            "900:\tlearn: 0.1784528\ttotal: 12.9s\tremaining: 30s\n",
            "1000:\tlearn: 0.1710893\ttotal: 14.1s\tremaining: 28.3s\n",
            "1100:\tlearn: 0.1640140\ttotal: 15.4s\tremaining: 26.6s\n",
            "1200:\tlearn: 0.1574815\ttotal: 16.7s\tremaining: 25.1s\n",
            "1300:\tlearn: 0.1513703\ttotal: 18s\tremaining: 23.5s\n",
            "1400:\tlearn: 0.1453327\ttotal: 19.3s\tremaining: 22s\n",
            "1500:\tlearn: 0.1395798\ttotal: 21.1s\tremaining: 21.1s\n",
            "1600:\tlearn: 0.1346524\ttotal: 23.6s\tremaining: 20.7s\n",
            "1700:\tlearn: 0.1296564\ttotal: 24.9s\tremaining: 19s\n",
            "1800:\tlearn: 0.1250879\ttotal: 26.2s\tremaining: 17.5s\n",
            "1900:\tlearn: 0.1211504\ttotal: 27.5s\tremaining: 15.9s\n",
            "2000:\tlearn: 0.1172380\ttotal: 28.8s\tremaining: 14.4s\n",
            "2100:\tlearn: 0.1136063\ttotal: 30.1s\tremaining: 12.9s\n",
            "2200:\tlearn: 0.1100875\ttotal: 31.4s\tremaining: 11.4s\n",
            "2300:\tlearn: 0.1065625\ttotal: 32.7s\tremaining: 9.93s\n",
            "2400:\tlearn: 0.1031574\ttotal: 34.6s\tremaining: 8.64s\n",
            "2500:\tlearn: 0.0998258\ttotal: 37s\tremaining: 7.39s\n",
            "2600:\tlearn: 0.0968334\ttotal: 38.3s\tremaining: 5.88s\n",
            "2700:\tlearn: 0.0940891\ttotal: 39.6s\tremaining: 4.38s\n",
            "2800:\tlearn: 0.0914162\ttotal: 40.9s\tremaining: 2.9s\n",
            "2900:\tlearn: 0.0887512\ttotal: 42.2s\tremaining: 1.44s\n",
            "2999:\tlearn: 0.0861317\ttotal: 43.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6869888\ttotal: 12.6ms\tremaining: 37.8s\n",
            "100:\tlearn: 0.3671916\ttotal: 1.11s\tremaining: 31.8s\n",
            "200:\tlearn: 0.2866126\ttotal: 2.23s\tremaining: 31.1s\n",
            "300:\tlearn: 0.2531556\ttotal: 3.75s\tremaining: 33.6s\n",
            "400:\tlearn: 0.2315881\ttotal: 6.42s\tremaining: 41.6s\n",
            "500:\tlearn: 0.2156324\ttotal: 7.71s\tremaining: 38.4s\n",
            "600:\tlearn: 0.2039811\ttotal: 8.97s\tremaining: 35.8s\n",
            "700:\tlearn: 0.1935114\ttotal: 10.3s\tremaining: 33.7s\n",
            "800:\tlearn: 0.1849604\ttotal: 11.5s\tremaining: 31.7s\n",
            "900:\tlearn: 0.1765556\ttotal: 12.8s\tremaining: 29.9s\n",
            "1000:\tlearn: 0.1692553\ttotal: 14.1s\tremaining: 28.1s\n",
            "1100:\tlearn: 0.1622111\ttotal: 15.4s\tremaining: 26.5s\n",
            "1200:\tlearn: 0.1556426\ttotal: 17s\tremaining: 25.4s\n",
            "1300:\tlearn: 0.1497275\ttotal: 19.7s\tremaining: 25.7s\n",
            "1400:\tlearn: 0.1439305\ttotal: 20.9s\tremaining: 23.9s\n",
            "1500:\tlearn: 0.1379786\ttotal: 22.2s\tremaining: 22.2s\n",
            "1600:\tlearn: 0.1324985\ttotal: 23.5s\tremaining: 20.5s\n",
            "1700:\tlearn: 0.1274488\ttotal: 24.8s\tremaining: 18.9s\n",
            "1800:\tlearn: 0.1225240\ttotal: 26.1s\tremaining: 17.4s\n",
            "1900:\tlearn: 0.1180232\ttotal: 27.3s\tremaining: 15.8s\n",
            "2000:\tlearn: 0.1137975\ttotal: 28.6s\tremaining: 14.3s\n",
            "2100:\tlearn: 0.1098038\ttotal: 30.3s\tremaining: 13s\n",
            "2200:\tlearn: 0.1060497\ttotal: 32.9s\tremaining: 11.9s\n",
            "2300:\tlearn: 0.1025874\ttotal: 34.2s\tremaining: 10.4s\n",
            "2400:\tlearn: 0.0992637\ttotal: 35.5s\tremaining: 8.85s\n",
            "2500:\tlearn: 0.0960746\ttotal: 36.7s\tremaining: 7.33s\n",
            "2600:\tlearn: 0.0928890\ttotal: 38s\tremaining: 5.83s\n",
            "2700:\tlearn: 0.0900475\ttotal: 39.3s\tremaining: 4.35s\n",
            "2800:\tlearn: 0.0873933\ttotal: 40.6s\tremaining: 2.88s\n",
            "2900:\tlearn: 0.0849383\ttotal: 41.9s\tremaining: 1.43s\n",
            "2999:\tlearn: 0.0824075\ttotal: 43.6s\tremaining: 0us\n",
            "0:\tlearn: 0.6867347\ttotal: 29.9ms\tremaining: 1m 29s\n",
            "100:\tlearn: 0.3700745\ttotal: 2.24s\tremaining: 1m 4s\n",
            "200:\tlearn: 0.2881306\ttotal: 3.44s\tremaining: 47.9s\n",
            "300:\tlearn: 0.2548475\ttotal: 4.77s\tremaining: 42.8s\n",
            "400:\tlearn: 0.2330775\ttotal: 6.08s\tremaining: 39.4s\n",
            "500:\tlearn: 0.2185586\ttotal: 7.36s\tremaining: 36.7s\n",
            "600:\tlearn: 0.2056791\ttotal: 8.69s\tremaining: 34.7s\n",
            "700:\tlearn: 0.1953076\ttotal: 9.98s\tremaining: 32.7s\n",
            "800:\tlearn: 0.1865724\ttotal: 11.3s\tremaining: 30.9s\n",
            "900:\tlearn: 0.1785620\ttotal: 13.1s\tremaining: 30.6s\n",
            "1000:\tlearn: 0.1711392\ttotal: 15.5s\tremaining: 30.9s\n",
            "1100:\tlearn: 0.1639162\ttotal: 16.8s\tremaining: 28.9s\n",
            "1200:\tlearn: 0.1571145\ttotal: 18.1s\tremaining: 27.1s\n",
            "1300:\tlearn: 0.1511049\ttotal: 19.4s\tremaining: 25.3s\n",
            "1400:\tlearn: 0.1452282\ttotal: 20.7s\tremaining: 23.6s\n",
            "1500:\tlearn: 0.1397573\ttotal: 22s\tremaining: 22s\n",
            "1600:\tlearn: 0.1344870\ttotal: 23.3s\tremaining: 20.4s\n",
            "1700:\tlearn: 0.1293630\ttotal: 24.6s\tremaining: 18.8s\n",
            "1800:\tlearn: 0.1247107\ttotal: 26.7s\tremaining: 17.8s\n",
            "1900:\tlearn: 0.1203493\ttotal: 28.9s\tremaining: 16.7s\n",
            "2000:\tlearn: 0.1163409\ttotal: 30.2s\tremaining: 15.1s\n",
            "2100:\tlearn: 0.1127306\ttotal: 31.5s\tremaining: 13.5s\n",
            "2200:\tlearn: 0.1092346\ttotal: 32.8s\tremaining: 11.9s\n",
            "2300:\tlearn: 0.1058206\ttotal: 34.1s\tremaining: 10.3s\n",
            "2400:\tlearn: 0.1024726\ttotal: 35.3s\tremaining: 8.82s\n",
            "2500:\tlearn: 0.0991897\ttotal: 36.6s\tremaining: 7.3s\n",
            "2600:\tlearn: 0.0960019\ttotal: 37.9s\tremaining: 5.81s\n",
            "2700:\tlearn: 0.0931751\ttotal: 40.1s\tremaining: 4.44s\n",
            "2800:\tlearn: 0.0904738\ttotal: 42.2s\tremaining: 3s\n",
            "2900:\tlearn: 0.0878508\ttotal: 43.5s\tremaining: 1.48s\n",
            "2999:\tlearn: 0.0853380\ttotal: 44.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6774414\ttotal: 2.92ms\tremaining: 5.84s\n",
            "100:\tlearn: 0.2956356\ttotal: 526ms\tremaining: 9.88s\n",
            "200:\tlearn: 0.2455957\ttotal: 1.09s\tremaining: 9.8s\n",
            "300:\tlearn: 0.2201280\ttotal: 1.65s\tremaining: 9.33s\n",
            "400:\tlearn: 0.2057719\ttotal: 2.22s\tremaining: 8.87s\n",
            "500:\tlearn: 0.1948204\ttotal: 2.79s\tremaining: 8.34s\n",
            "600:\tlearn: 0.1860104\ttotal: 3.35s\tremaining: 7.81s\n",
            "700:\tlearn: 0.1781129\ttotal: 3.93s\tremaining: 7.29s\n",
            "800:\tlearn: 0.1705578\ttotal: 4.5s\tremaining: 6.73s\n",
            "900:\tlearn: 0.1636794\ttotal: 5.07s\tremaining: 6.19s\n",
            "1000:\tlearn: 0.1565095\ttotal: 5.63s\tremaining: 5.61s\n",
            "1100:\tlearn: 0.1501771\ttotal: 6.2s\tremaining: 5.06s\n",
            "1200:\tlearn: 0.1445379\ttotal: 6.96s\tremaining: 4.63s\n",
            "1300:\tlearn: 0.1394933\ttotal: 8.29s\tremaining: 4.46s\n",
            "1400:\tlearn: 0.1346065\ttotal: 9.65s\tremaining: 4.13s\n",
            "1500:\tlearn: 0.1301363\ttotal: 10.3s\tremaining: 3.42s\n",
            "1600:\tlearn: 0.1259089\ttotal: 10.8s\tremaining: 2.7s\n",
            "1700:\tlearn: 0.1219842\ttotal: 11.4s\tremaining: 2s\n",
            "1800:\tlearn: 0.1180988\ttotal: 12s\tremaining: 1.32s\n",
            "1900:\tlearn: 0.1143633\ttotal: 12.5s\tremaining: 653ms\n",
            "1999:\tlearn: 0.1107051\ttotal: 13.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6776294\ttotal: 2.83ms\tremaining: 5.66s\n",
            "100:\tlearn: 0.2982489\ttotal: 532ms\tremaining: 10s\n",
            "200:\tlearn: 0.2453494\ttotal: 1.11s\tremaining: 9.9s\n",
            "300:\tlearn: 0.2243899\ttotal: 1.67s\tremaining: 9.4s\n",
            "400:\tlearn: 0.2102616\ttotal: 2.24s\tremaining: 8.92s\n",
            "500:\tlearn: 0.2000767\ttotal: 2.8s\tremaining: 8.38s\n",
            "600:\tlearn: 0.1910447\ttotal: 3.38s\tremaining: 7.87s\n",
            "700:\tlearn: 0.1830339\ttotal: 3.97s\tremaining: 7.35s\n",
            "800:\tlearn: 0.1755274\ttotal: 4.54s\tremaining: 6.79s\n",
            "900:\tlearn: 0.1686572\ttotal: 5.12s\tremaining: 6.25s\n",
            "1000:\tlearn: 0.1620283\ttotal: 5.66s\tremaining: 5.65s\n",
            "1100:\tlearn: 0.1559299\ttotal: 6.24s\tremaining: 5.1s\n",
            "1200:\tlearn: 0.1503054\ttotal: 7.44s\tremaining: 4.95s\n",
            "1300:\tlearn: 0.1455101\ttotal: 8.65s\tremaining: 4.65s\n",
            "1400:\tlearn: 0.1408729\ttotal: 9.78s\tremaining: 4.18s\n",
            "1500:\tlearn: 0.1364785\ttotal: 10.4s\tremaining: 3.45s\n",
            "1600:\tlearn: 0.1323674\ttotal: 10.9s\tremaining: 2.73s\n",
            "1700:\tlearn: 0.1283831\ttotal: 11.5s\tremaining: 2.02s\n",
            "1800:\tlearn: 0.1246948\ttotal: 12.1s\tremaining: 1.34s\n",
            "1900:\tlearn: 0.1213072\ttotal: 12.7s\tremaining: 659ms\n",
            "1999:\tlearn: 0.1177895\ttotal: 13.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6774793\ttotal: 2.92ms\tremaining: 5.84s\n",
            "100:\tlearn: 0.2983173\ttotal: 532ms\tremaining: 10s\n",
            "200:\tlearn: 0.2435549\ttotal: 1.09s\tremaining: 9.8s\n",
            "300:\tlearn: 0.2201806\ttotal: 1.65s\tremaining: 9.3s\n",
            "400:\tlearn: 0.2062667\ttotal: 2.22s\tremaining: 8.84s\n",
            "500:\tlearn: 0.1956077\ttotal: 2.78s\tremaining: 8.33s\n",
            "600:\tlearn: 0.1871753\ttotal: 3.34s\tremaining: 7.78s\n",
            "700:\tlearn: 0.1794293\ttotal: 3.91s\tremaining: 7.24s\n",
            "800:\tlearn: 0.1723876\ttotal: 4.46s\tremaining: 6.68s\n",
            "900:\tlearn: 0.1657720\ttotal: 5.04s\tremaining: 6.14s\n",
            "1000:\tlearn: 0.1592387\ttotal: 5.6s\tremaining: 5.58s\n",
            "1100:\tlearn: 0.1532159\ttotal: 6.25s\tremaining: 5.11s\n",
            "1200:\tlearn: 0.1477421\ttotal: 7.45s\tremaining: 4.96s\n",
            "1300:\tlearn: 0.1428988\ttotal: 8.78s\tremaining: 4.72s\n",
            "1400:\tlearn: 0.1381922\ttotal: 9.57s\tremaining: 4.09s\n",
            "1500:\tlearn: 0.1339415\ttotal: 10.1s\tremaining: 3.37s\n",
            "1600:\tlearn: 0.1295895\ttotal: 10.7s\tremaining: 2.66s\n",
            "1700:\tlearn: 0.1255590\ttotal: 11.3s\tremaining: 1.98s\n",
            "1800:\tlearn: 0.1215742\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.1181401\ttotal: 12.4s\tremaining: 644ms\n",
            "1999:\tlearn: 0.1146225\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6772583\ttotal: 2.79ms\tremaining: 5.58s\n",
            "100:\tlearn: 0.2961824\ttotal: 532ms\tremaining: 10s\n",
            "200:\tlearn: 0.2405045\ttotal: 1.1s\tremaining: 9.86s\n",
            "300:\tlearn: 0.2179204\ttotal: 1.66s\tremaining: 9.35s\n",
            "400:\tlearn: 0.2025744\ttotal: 2.23s\tremaining: 8.88s\n",
            "500:\tlearn: 0.1925333\ttotal: 2.77s\tremaining: 8.3s\n",
            "600:\tlearn: 0.1838317\ttotal: 3.35s\tremaining: 7.8s\n",
            "700:\tlearn: 0.1764590\ttotal: 3.91s\tremaining: 7.25s\n",
            "800:\tlearn: 0.1693147\ttotal: 4.49s\tremaining: 6.72s\n",
            "900:\tlearn: 0.1623722\ttotal: 5.06s\tremaining: 6.17s\n",
            "1000:\tlearn: 0.1558198\ttotal: 5.61s\tremaining: 5.6s\n",
            "1100:\tlearn: 0.1494660\ttotal: 6.38s\tremaining: 5.21s\n",
            "1200:\tlearn: 0.1433162\ttotal: 7.59s\tremaining: 5.05s\n",
            "1300:\tlearn: 0.1381321\ttotal: 8.87s\tremaining: 4.77s\n",
            "1400:\tlearn: 0.1332840\ttotal: 9.64s\tremaining: 4.12s\n",
            "1500:\tlearn: 0.1288158\ttotal: 10.2s\tremaining: 3.39s\n",
            "1600:\tlearn: 0.1249066\ttotal: 10.8s\tremaining: 2.68s\n",
            "1700:\tlearn: 0.1210796\ttotal: 11.3s\tremaining: 1.99s\n",
            "1800:\tlearn: 0.1171195\ttotal: 11.9s\tremaining: 1.31s\n",
            "1900:\tlearn: 0.1134254\ttotal: 12.5s\tremaining: 649ms\n",
            "1999:\tlearn: 0.1101960\ttotal: 13s\tremaining: 0us\n",
            "0:\tlearn: 0.6774331\ttotal: 2.88ms\tremaining: 5.75s\n",
            "100:\tlearn: 0.2968571\ttotal: 528ms\tremaining: 9.93s\n",
            "200:\tlearn: 0.2441170\ttotal: 1.1s\tremaining: 9.86s\n",
            "300:\tlearn: 0.2220740\ttotal: 1.67s\tremaining: 9.42s\n",
            "400:\tlearn: 0.2061188\ttotal: 2.24s\tremaining: 8.95s\n",
            "500:\tlearn: 0.1959338\ttotal: 2.79s\tremaining: 8.36s\n",
            "600:\tlearn: 0.1866170\ttotal: 3.36s\tremaining: 7.82s\n",
            "700:\tlearn: 0.1789465\ttotal: 3.91s\tremaining: 7.25s\n",
            "800:\tlearn: 0.1716928\ttotal: 4.48s\tremaining: 6.7s\n",
            "900:\tlearn: 0.1642376\ttotal: 5.05s\tremaining: 6.16s\n",
            "1000:\tlearn: 0.1578734\ttotal: 5.6s\tremaining: 5.59s\n",
            "1100:\tlearn: 0.1522010\ttotal: 6.52s\tremaining: 5.33s\n",
            "1200:\tlearn: 0.1458519\ttotal: 7.65s\tremaining: 5.09s\n",
            "1300:\tlearn: 0.1405529\ttotal: 9.02s\tremaining: 4.85s\n",
            "1400:\tlearn: 0.1361902\ttotal: 9.69s\tremaining: 4.14s\n",
            "1500:\tlearn: 0.1318194\ttotal: 10.3s\tremaining: 3.41s\n",
            "1600:\tlearn: 0.1277253\ttotal: 10.8s\tremaining: 2.7s\n",
            "1700:\tlearn: 0.1240444\ttotal: 11.4s\tremaining: 2s\n",
            "1800:\tlearn: 0.1203116\ttotal: 11.9s\tremaining: 1.32s\n",
            "1900:\tlearn: 0.1166570\ttotal: 12.5s\tremaining: 652ms\n",
            "1999:\tlearn: 0.1132056\ttotal: 13.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6691456\ttotal: 5.31ms\tremaining: 10.6s\n",
            "100:\tlearn: 0.2452537\ttotal: 556ms\tremaining: 10.5s\n",
            "200:\tlearn: 0.2059206\ttotal: 1.12s\tremaining: 10s\n",
            "300:\tlearn: 0.1859404\ttotal: 1.68s\tremaining: 9.47s\n",
            "400:\tlearn: 0.1709514\ttotal: 2.25s\tremaining: 8.97s\n",
            "500:\tlearn: 0.1570436\ttotal: 2.8s\tremaining: 8.38s\n",
            "600:\tlearn: 0.1442946\ttotal: 3.37s\tremaining: 7.84s\n",
            "700:\tlearn: 0.1349082\ttotal: 3.91s\tremaining: 7.25s\n",
            "800:\tlearn: 0.1260837\ttotal: 4.49s\tremaining: 6.71s\n",
            "900:\tlearn: 0.1184880\ttotal: 5.04s\tremaining: 6.14s\n",
            "1000:\tlearn: 0.1114940\ttotal: 5.61s\tremaining: 5.59s\n",
            "1100:\tlearn: 0.1052486\ttotal: 6.54s\tremaining: 5.34s\n",
            "1200:\tlearn: 0.0990911\ttotal: 7.8s\tremaining: 5.19s\n",
            "1300:\tlearn: 0.0938863\ttotal: 9.07s\tremaining: 4.88s\n",
            "1400:\tlearn: 0.0886912\ttotal: 9.65s\tremaining: 4.13s\n",
            "1500:\tlearn: 0.0839687\ttotal: 10.2s\tremaining: 3.4s\n",
            "1600:\tlearn: 0.0795733\ttotal: 10.8s\tremaining: 2.69s\n",
            "1700:\tlearn: 0.0754701\ttotal: 11.4s\tremaining: 2s\n",
            "1800:\tlearn: 0.0717224\ttotal: 11.9s\tremaining: 1.32s\n",
            "1900:\tlearn: 0.0683687\ttotal: 12.5s\tremaining: 651ms\n",
            "1999:\tlearn: 0.0654150\ttotal: 13s\tremaining: 0us\n",
            "0:\tlearn: 0.6694856\ttotal: 5.47ms\tremaining: 10.9s\n",
            "100:\tlearn: 0.2470364\ttotal: 547ms\tremaining: 10.3s\n",
            "200:\tlearn: 0.2111098\ttotal: 1.11s\tremaining: 9.98s\n",
            "300:\tlearn: 0.1925332\ttotal: 1.68s\tremaining: 9.5s\n",
            "400:\tlearn: 0.1769528\ttotal: 2.27s\tremaining: 9.04s\n",
            "500:\tlearn: 0.1626640\ttotal: 2.84s\tremaining: 8.5s\n",
            "600:\tlearn: 0.1507968\ttotal: 3.47s\tremaining: 8.07s\n",
            "700:\tlearn: 0.1414519\ttotal: 4.04s\tremaining: 7.49s\n",
            "800:\tlearn: 0.1328229\ttotal: 4.63s\tremaining: 6.93s\n",
            "900:\tlearn: 0.1250681\ttotal: 5.22s\tremaining: 6.37s\n",
            "1000:\tlearn: 0.1182184\ttotal: 6s\tremaining: 5.99s\n",
            "1100:\tlearn: 0.1118747\ttotal: 7.33s\tremaining: 5.98s\n",
            "1200:\tlearn: 0.1057567\ttotal: 8.75s\tremaining: 5.82s\n",
            "1300:\tlearn: 0.1003958\ttotal: 9.34s\tremaining: 5.01s\n",
            "1400:\tlearn: 0.0954214\ttotal: 9.89s\tremaining: 4.23s\n",
            "1500:\tlearn: 0.0909185\ttotal: 10.5s\tremaining: 3.48s\n",
            "1600:\tlearn: 0.0867408\ttotal: 11s\tremaining: 2.75s\n",
            "1700:\tlearn: 0.0826839\ttotal: 11.6s\tremaining: 2.04s\n",
            "1800:\tlearn: 0.0789176\ttotal: 12.2s\tremaining: 1.35s\n",
            "1900:\tlearn: 0.0754043\ttotal: 12.8s\tremaining: 665ms\n",
            "1999:\tlearn: 0.0721567\ttotal: 13.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6692877\ttotal: 5.64ms\tremaining: 11.3s\n",
            "100:\tlearn: 0.2463141\ttotal: 548ms\tremaining: 10.3s\n",
            "200:\tlearn: 0.2073298\ttotal: 1.11s\tremaining: 9.98s\n",
            "300:\tlearn: 0.1885192\ttotal: 1.69s\tremaining: 9.52s\n",
            "400:\tlearn: 0.1731951\ttotal: 2.26s\tremaining: 9s\n",
            "500:\tlearn: 0.1599140\ttotal: 2.82s\tremaining: 8.43s\n",
            "600:\tlearn: 0.1486885\ttotal: 3.39s\tremaining: 7.9s\n",
            "700:\tlearn: 0.1390470\ttotal: 3.97s\tremaining: 7.36s\n",
            "800:\tlearn: 0.1304383\ttotal: 4.53s\tremaining: 6.79s\n",
            "900:\tlearn: 0.1225101\ttotal: 5.17s\tremaining: 6.31s\n",
            "1000:\tlearn: 0.1154907\ttotal: 6.46s\tremaining: 6.45s\n",
            "1100:\tlearn: 0.1090102\ttotal: 7.81s\tremaining: 6.37s\n",
            "1200:\tlearn: 0.1029350\ttotal: 8.64s\tremaining: 5.75s\n",
            "1300:\tlearn: 0.0973157\ttotal: 9.21s\tremaining: 4.95s\n",
            "1400:\tlearn: 0.0923378\ttotal: 9.77s\tremaining: 4.17s\n",
            "1500:\tlearn: 0.0872774\ttotal: 10.3s\tremaining: 3.44s\n",
            "1600:\tlearn: 0.0829669\ttotal: 10.9s\tremaining: 2.72s\n",
            "1700:\tlearn: 0.0790655\ttotal: 11.5s\tremaining: 2.02s\n",
            "1800:\tlearn: 0.0753541\ttotal: 12s\tremaining: 1.33s\n",
            "1900:\tlearn: 0.0716666\ttotal: 12.6s\tremaining: 657ms\n",
            "1999:\tlearn: 0.0681354\ttotal: 13.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6688916\ttotal: 7.63ms\tremaining: 15.2s\n",
            "100:\tlearn: 0.2425630\ttotal: 535ms\tremaining: 10.1s\n",
            "200:\tlearn: 0.2038320\ttotal: 1.1s\tremaining: 9.86s\n",
            "300:\tlearn: 0.1849371\ttotal: 1.66s\tremaining: 9.38s\n",
            "400:\tlearn: 0.1700989\ttotal: 2.23s\tremaining: 8.89s\n",
            "500:\tlearn: 0.1562137\ttotal: 2.77s\tremaining: 8.3s\n",
            "600:\tlearn: 0.1438230\ttotal: 3.35s\tremaining: 7.8s\n",
            "700:\tlearn: 0.1339653\ttotal: 3.92s\tremaining: 7.26s\n",
            "800:\tlearn: 0.1250836\ttotal: 4.46s\tremaining: 6.68s\n",
            "900:\tlearn: 0.1169462\ttotal: 5.35s\tremaining: 6.53s\n",
            "1000:\tlearn: 0.1097214\ttotal: 6.61s\tremaining: 6.6s\n",
            "1100:\tlearn: 0.1031945\ttotal: 7.88s\tremaining: 6.44s\n",
            "1200:\tlearn: 0.0974513\ttotal: 8.45s\tremaining: 5.62s\n",
            "1300:\tlearn: 0.0923059\ttotal: 9.02s\tremaining: 4.84s\n",
            "1400:\tlearn: 0.0874043\ttotal: 9.58s\tremaining: 4.09s\n",
            "1500:\tlearn: 0.0825770\ttotal: 10.2s\tremaining: 3.38s\n",
            "1600:\tlearn: 0.0785289\ttotal: 10.7s\tremaining: 2.67s\n",
            "1700:\tlearn: 0.0748145\ttotal: 11.3s\tremaining: 1.98s\n",
            "1800:\tlearn: 0.0714557\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.0682044\ttotal: 12.4s\tremaining: 644ms\n",
            "1999:\tlearn: 0.0649905\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6680118\ttotal: 5.43ms\tremaining: 10.9s\n",
            "100:\tlearn: 0.2477588\ttotal: 565ms\tremaining: 10.6s\n",
            "200:\tlearn: 0.2069525\ttotal: 1.13s\tremaining: 10.2s\n",
            "300:\tlearn: 0.1862553\ttotal: 1.7s\tremaining: 9.58s\n",
            "400:\tlearn: 0.1712188\ttotal: 2.27s\tremaining: 9.06s\n",
            "500:\tlearn: 0.1580399\ttotal: 2.83s\tremaining: 8.48s\n",
            "600:\tlearn: 0.1465379\ttotal: 3.4s\tremaining: 7.92s\n",
            "700:\tlearn: 0.1368077\ttotal: 3.96s\tremaining: 7.33s\n",
            "800:\tlearn: 0.1281895\ttotal: 4.54s\tremaining: 6.79s\n",
            "900:\tlearn: 0.1207233\ttotal: 5.63s\tremaining: 6.87s\n",
            "1000:\tlearn: 0.1133443\ttotal: 6.99s\tremaining: 6.98s\n",
            "1100:\tlearn: 0.1069845\ttotal: 8.11s\tremaining: 6.62s\n",
            "1200:\tlearn: 0.1010519\ttotal: 8.68s\tremaining: 5.78s\n",
            "1300:\tlearn: 0.0956058\ttotal: 9.26s\tremaining: 4.97s\n",
            "1400:\tlearn: 0.0905919\ttotal: 9.82s\tremaining: 4.2s\n",
            "1500:\tlearn: 0.0861545\ttotal: 10.4s\tremaining: 3.46s\n",
            "1600:\tlearn: 0.0819681\ttotal: 11s\tremaining: 2.73s\n",
            "1700:\tlearn: 0.0780482\ttotal: 11.5s\tremaining: 2.03s\n",
            "1800:\tlearn: 0.0744250\ttotal: 12.1s\tremaining: 1.33s\n",
            "1900:\tlearn: 0.0708932\ttotal: 12.7s\tremaining: 660ms\n",
            "1999:\tlearn: 0.0677277\ttotal: 13.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6709367\ttotal: 12.7ms\tremaining: 38s\n",
            "100:\tlearn: 0.2285444\ttotal: 1.19s\tremaining: 34s\n",
            "200:\tlearn: 0.1844295\ttotal: 2.45s\tremaining: 34.1s\n",
            "300:\tlearn: 0.1551811\ttotal: 3.73s\tremaining: 33.5s\n",
            "400:\tlearn: 0.1322041\ttotal: 5.68s\tremaining: 36.8s\n",
            "500:\tlearn: 0.1140369\ttotal: 7.88s\tremaining: 39.3s\n",
            "600:\tlearn: 0.0997993\ttotal: 9.17s\tremaining: 36.6s\n",
            "700:\tlearn: 0.0873982\ttotal: 10.5s\tremaining: 34.4s\n",
            "800:\tlearn: 0.0773603\ttotal: 11.7s\tremaining: 32.2s\n",
            "900:\tlearn: 0.0689072\ttotal: 13s\tremaining: 30.3s\n",
            "1000:\tlearn: 0.0614621\ttotal: 14.3s\tremaining: 28.5s\n",
            "1100:\tlearn: 0.0551293\ttotal: 15.5s\tremaining: 26.8s\n",
            "1200:\tlearn: 0.0497164\ttotal: 16.8s\tremaining: 25.2s\n",
            "1300:\tlearn: 0.0448874\ttotal: 18.8s\tremaining: 24.5s\n",
            "1400:\tlearn: 0.0408651\ttotal: 21s\tremaining: 24s\n",
            "1500:\tlearn: 0.0372746\ttotal: 22.3s\tremaining: 22.2s\n",
            "1600:\tlearn: 0.0341220\ttotal: 23.5s\tremaining: 20.5s\n",
            "1700:\tlearn: 0.0314913\ttotal: 24.7s\tremaining: 18.9s\n",
            "1800:\tlearn: 0.0291282\ttotal: 26s\tremaining: 17.3s\n",
            "1900:\tlearn: 0.0269516\ttotal: 27.3s\tremaining: 15.8s\n",
            "2000:\tlearn: 0.0249529\ttotal: 28.5s\tremaining: 14.2s\n",
            "2100:\tlearn: 0.0231305\ttotal: 29.8s\tremaining: 12.8s\n",
            "2200:\tlearn: 0.0215309\ttotal: 31.7s\tremaining: 11.5s\n",
            "2300:\tlearn: 0.0201014\ttotal: 34.1s\tremaining: 10.3s\n",
            "2400:\tlearn: 0.0188826\ttotal: 35.3s\tremaining: 8.81s\n",
            "2500:\tlearn: 0.0177882\ttotal: 36.5s\tremaining: 7.29s\n",
            "2600:\tlearn: 0.0167463\ttotal: 37.8s\tremaining: 5.8s\n",
            "2700:\tlearn: 0.0157381\ttotal: 39.1s\tremaining: 4.33s\n",
            "2800:\tlearn: 0.0148155\ttotal: 40.4s\tremaining: 2.87s\n",
            "2900:\tlearn: 0.0139437\ttotal: 41.7s\tremaining: 1.42s\n",
            "2999:\tlearn: 0.0131817\ttotal: 43.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6712813\ttotal: 27.6ms\tremaining: 1m 22s\n",
            "100:\tlearn: 0.2361646\ttotal: 2.32s\tremaining: 1m 6s\n",
            "200:\tlearn: 0.1873042\ttotal: 5.41s\tremaining: 1m 15s\n",
            "300:\tlearn: 0.1577279\ttotal: 7.44s\tremaining: 1m 6s\n",
            "400:\tlearn: 0.1351405\ttotal: 8.99s\tremaining: 58.3s\n",
            "500:\tlearn: 0.1181406\ttotal: 10.3s\tremaining: 51.1s\n",
            "600:\tlearn: 0.1034451\ttotal: 11.5s\tremaining: 46.1s\n",
            "700:\tlearn: 0.0927525\ttotal: 12.8s\tremaining: 42s\n",
            "800:\tlearn: 0.0832835\ttotal: 14.1s\tremaining: 38.6s\n",
            "900:\tlearn: 0.0747854\ttotal: 15.9s\tremaining: 37.1s\n",
            "1000:\tlearn: 0.0673413\ttotal: 18.3s\tremaining: 36.5s\n",
            "1100:\tlearn: 0.0608083\ttotal: 19.6s\tremaining: 33.7s\n",
            "1200:\tlearn: 0.0551713\ttotal: 20.8s\tremaining: 31.2s\n",
            "1300:\tlearn: 0.0499398\ttotal: 22.1s\tremaining: 28.9s\n",
            "1400:\tlearn: 0.0457000\ttotal: 23.4s\tremaining: 26.7s\n",
            "1500:\tlearn: 0.0414949\ttotal: 24.6s\tremaining: 24.6s\n",
            "1600:\tlearn: 0.0383115\ttotal: 26s\tremaining: 22.7s\n",
            "1700:\tlearn: 0.0352197\ttotal: 27.3s\tremaining: 20.8s\n",
            "1800:\tlearn: 0.0325710\ttotal: 29.2s\tremaining: 19.4s\n",
            "1900:\tlearn: 0.0300642\ttotal: 31.4s\tremaining: 18.1s\n",
            "2000:\tlearn: 0.0279047\ttotal: 32.6s\tremaining: 16.3s\n",
            "2100:\tlearn: 0.0258648\ttotal: 33.9s\tremaining: 14.5s\n",
            "2200:\tlearn: 0.0242338\ttotal: 35.2s\tremaining: 12.8s\n",
            "2300:\tlearn: 0.0226502\ttotal: 36.5s\tremaining: 11.1s\n",
            "2400:\tlearn: 0.0211510\ttotal: 37.8s\tremaining: 9.43s\n",
            "2500:\tlearn: 0.0197838\ttotal: 39.1s\tremaining: 7.79s\n",
            "2600:\tlearn: 0.0186337\ttotal: 40.4s\tremaining: 6.19s\n",
            "2700:\tlearn: 0.0175671\ttotal: 42.3s\tremaining: 4.68s\n",
            "2800:\tlearn: 0.0165643\ttotal: 44.5s\tremaining: 3.16s\n",
            "2900:\tlearn: 0.0156722\ttotal: 45.8s\tremaining: 1.56s\n",
            "2999:\tlearn: 0.0147788\ttotal: 47.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6699190\ttotal: 13.2ms\tremaining: 39.5s\n",
            "100:\tlearn: 0.2324198\ttotal: 1.23s\tremaining: 35.3s\n",
            "200:\tlearn: 0.1855523\ttotal: 2.5s\tremaining: 34.9s\n",
            "300:\tlearn: 0.1576469\ttotal: 3.77s\tremaining: 33.8s\n",
            "400:\tlearn: 0.1354314\ttotal: 5.04s\tremaining: 32.7s\n",
            "500:\tlearn: 0.1179388\ttotal: 6.32s\tremaining: 31.5s\n",
            "600:\tlearn: 0.1038258\ttotal: 8.35s\tremaining: 33.3s\n",
            "700:\tlearn: 0.0917535\ttotal: 10.5s\tremaining: 34.6s\n",
            "800:\tlearn: 0.0819360\ttotal: 11.8s\tremaining: 32.4s\n",
            "900:\tlearn: 0.0732416\ttotal: 13s\tremaining: 30.4s\n",
            "1000:\tlearn: 0.0659266\ttotal: 14.3s\tremaining: 28.6s\n",
            "1100:\tlearn: 0.0591345\ttotal: 15.6s\tremaining: 26.9s\n",
            "1200:\tlearn: 0.0534576\ttotal: 16.9s\tremaining: 25.3s\n",
            "1300:\tlearn: 0.0487220\ttotal: 18.2s\tremaining: 23.7s\n",
            "1400:\tlearn: 0.0441299\ttotal: 19.4s\tremaining: 22.2s\n",
            "1500:\tlearn: 0.0403531\ttotal: 21.4s\tremaining: 21.4s\n",
            "1600:\tlearn: 0.0370369\ttotal: 23.7s\tremaining: 20.7s\n",
            "1700:\tlearn: 0.0341067\ttotal: 24.9s\tremaining: 19s\n",
            "1800:\tlearn: 0.0314169\ttotal: 26.2s\tremaining: 17.4s\n",
            "1900:\tlearn: 0.0290098\ttotal: 27.5s\tremaining: 15.9s\n",
            "2000:\tlearn: 0.0265921\ttotal: 28.8s\tremaining: 14.4s\n",
            "2100:\tlearn: 0.0247301\ttotal: 30.1s\tremaining: 12.9s\n",
            "2200:\tlearn: 0.0229391\ttotal: 31.3s\tremaining: 11.4s\n",
            "2300:\tlearn: 0.0213779\ttotal: 32.6s\tremaining: 9.9s\n",
            "2400:\tlearn: 0.0199806\ttotal: 34.6s\tremaining: 8.63s\n",
            "2500:\tlearn: 0.0187249\ttotal: 36.7s\tremaining: 7.33s\n",
            "2600:\tlearn: 0.0175401\ttotal: 38s\tremaining: 5.83s\n",
            "2700:\tlearn: 0.0165002\ttotal: 39.3s\tremaining: 4.35s\n",
            "2800:\tlearn: 0.0155645\ttotal: 40.6s\tremaining: 2.88s\n",
            "2900:\tlearn: 0.0147245\ttotal: 41.9s\tremaining: 1.43s\n",
            "2999:\tlearn: 0.0139414\ttotal: 43.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6695228\ttotal: 11.9ms\tremaining: 35.7s\n",
            "100:\tlearn: 0.2295811\ttotal: 1.23s\tremaining: 35.3s\n",
            "200:\tlearn: 0.1850691\ttotal: 2.5s\tremaining: 34.8s\n",
            "300:\tlearn: 0.1552584\ttotal: 4.62s\tremaining: 41.4s\n",
            "400:\tlearn: 0.1318137\ttotal: 6.73s\tremaining: 43.6s\n",
            "500:\tlearn: 0.1132058\ttotal: 8s\tremaining: 39.9s\n",
            "600:\tlearn: 0.0985245\ttotal: 9.29s\tremaining: 37.1s\n",
            "700:\tlearn: 0.0862199\ttotal: 10.5s\tremaining: 34.6s\n",
            "800:\tlearn: 0.0768162\ttotal: 11.8s\tremaining: 32.4s\n",
            "900:\tlearn: 0.0685681\ttotal: 13.1s\tremaining: 30.4s\n",
            "1000:\tlearn: 0.0618078\ttotal: 14.3s\tremaining: 28.6s\n",
            "1100:\tlearn: 0.0557198\ttotal: 15.6s\tremaining: 26.9s\n",
            "1200:\tlearn: 0.0502663\ttotal: 17.6s\tremaining: 26.4s\n",
            "1300:\tlearn: 0.0453925\ttotal: 19.8s\tremaining: 25.9s\n",
            "1400:\tlearn: 0.0412058\ttotal: 21.1s\tremaining: 24.1s\n",
            "1500:\tlearn: 0.0378140\ttotal: 22.4s\tremaining: 22.3s\n",
            "1600:\tlearn: 0.0347712\ttotal: 23.7s\tremaining: 20.7s\n",
            "1700:\tlearn: 0.0319569\ttotal: 25s\tremaining: 19.1s\n",
            "1800:\tlearn: 0.0294213\ttotal: 26.2s\tremaining: 17.5s\n",
            "1900:\tlearn: 0.0270896\ttotal: 27.5s\tremaining: 15.9s\n",
            "2000:\tlearn: 0.0252925\ttotal: 28.8s\tremaining: 14.4s\n",
            "2100:\tlearn: 0.0234886\ttotal: 30.9s\tremaining: 13.2s\n",
            "2200:\tlearn: 0.0218954\ttotal: 33s\tremaining: 12s\n",
            "2300:\tlearn: 0.0205434\ttotal: 34.3s\tremaining: 10.4s\n",
            "2400:\tlearn: 0.0192140\ttotal: 35.5s\tremaining: 8.86s\n",
            "2500:\tlearn: 0.0179607\ttotal: 36.8s\tremaining: 7.34s\n",
            "2600:\tlearn: 0.0168102\ttotal: 38.1s\tremaining: 5.84s\n",
            "2700:\tlearn: 0.0158225\ttotal: 39.3s\tremaining: 4.35s\n",
            "2800:\tlearn: 0.0149381\ttotal: 40.6s\tremaining: 2.88s\n",
            "2900:\tlearn: 0.0141427\ttotal: 41.9s\tremaining: 1.43s\n",
            "2999:\tlearn: 0.0133640\ttotal: 43.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6699130\ttotal: 38ms\tremaining: 1m 53s\n",
            "100:\tlearn: 0.2341581\ttotal: 1.97s\tremaining: 56.4s\n",
            "200:\tlearn: 0.1854400\ttotal: 3.25s\tremaining: 45.3s\n",
            "300:\tlearn: 0.1558955\ttotal: 4.54s\tremaining: 40.7s\n",
            "400:\tlearn: 0.1330250\ttotal: 5.81s\tremaining: 37.7s\n",
            "500:\tlearn: 0.1154174\ttotal: 7.08s\tremaining: 35.3s\n",
            "600:\tlearn: 0.1013568\ttotal: 8.37s\tremaining: 33.4s\n",
            "700:\tlearn: 0.0893892\ttotal: 9.64s\tremaining: 31.6s\n",
            "800:\tlearn: 0.0797198\ttotal: 10.9s\tremaining: 30.1s\n",
            "900:\tlearn: 0.0710238\ttotal: 13.2s\tremaining: 30.8s\n",
            "1000:\tlearn: 0.0638186\ttotal: 15.2s\tremaining: 30.4s\n",
            "1100:\tlearn: 0.0579537\ttotal: 16.5s\tremaining: 28.4s\n",
            "1200:\tlearn: 0.0523631\ttotal: 17.8s\tremaining: 26.6s\n",
            "1300:\tlearn: 0.0476517\ttotal: 19s\tremaining: 24.9s\n",
            "1400:\tlearn: 0.0437503\ttotal: 20.3s\tremaining: 23.2s\n",
            "1500:\tlearn: 0.0400072\ttotal: 21.6s\tremaining: 21.6s\n",
            "1600:\tlearn: 0.0365423\ttotal: 22.9s\tremaining: 20s\n",
            "1700:\tlearn: 0.0336720\ttotal: 24.1s\tremaining: 18.4s\n",
            "1800:\tlearn: 0.0311372\ttotal: 26.3s\tremaining: 17.5s\n",
            "1900:\tlearn: 0.0288399\ttotal: 28.2s\tremaining: 16.3s\n",
            "2000:\tlearn: 0.0267164\ttotal: 29.5s\tremaining: 14.7s\n",
            "2100:\tlearn: 0.0248874\ttotal: 30.8s\tremaining: 13.2s\n",
            "2200:\tlearn: 0.0231517\ttotal: 32s\tremaining: 11.6s\n",
            "2300:\tlearn: 0.0216044\ttotal: 33.3s\tremaining: 10.1s\n",
            "2400:\tlearn: 0.0202742\ttotal: 34.6s\tremaining: 8.63s\n",
            "2500:\tlearn: 0.0189984\ttotal: 35.8s\tremaining: 7.15s\n",
            "2600:\tlearn: 0.0178570\ttotal: 37.1s\tremaining: 5.69s\n",
            "2700:\tlearn: 0.0168994\ttotal: 39.3s\tremaining: 4.35s\n",
            "2800:\tlearn: 0.0159334\ttotal: 41.4s\tremaining: 2.94s\n",
            "2900:\tlearn: 0.0149844\ttotal: 42.6s\tremaining: 1.46s\n",
            "2999:\tlearn: 0.0141996\ttotal: 43.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6852436\ttotal: 2.82ms\tremaining: 2.82s\n",
            "100:\tlearn: 0.3765501\ttotal: 528ms\tremaining: 4.7s\n",
            "200:\tlearn: 0.2977429\ttotal: 1.05s\tremaining: 4.18s\n",
            "300:\tlearn: 0.2621201\ttotal: 1.6s\tremaining: 3.73s\n",
            "400:\tlearn: 0.2426487\ttotal: 2.14s\tremaining: 3.2s\n",
            "500:\tlearn: 0.2300174\ttotal: 2.7s\tremaining: 2.69s\n",
            "600:\tlearn: 0.2198556\ttotal: 3.25s\tremaining: 2.16s\n",
            "700:\tlearn: 0.2114852\ttotal: 3.81s\tremaining: 1.62s\n",
            "800:\tlearn: 0.2049745\ttotal: 4.36s\tremaining: 1.08s\n",
            "900:\tlearn: 0.1993183\ttotal: 4.93s\tremaining: 541ms\n",
            "999:\tlearn: 0.1942966\ttotal: 5.48s\tremaining: 0us\n",
            "0:\tlearn: 0.6853284\ttotal: 2.77ms\tremaining: 2.77s\n",
            "100:\tlearn: 0.3792800\ttotal: 512ms\tremaining: 4.56s\n",
            "200:\tlearn: 0.3026723\ttotal: 1.12s\tremaining: 4.45s\n",
            "300:\tlearn: 0.2677206\ttotal: 2.45s\tremaining: 5.68s\n",
            "400:\tlearn: 0.2474042\ttotal: 3.69s\tremaining: 5.52s\n",
            "500:\tlearn: 0.2348176\ttotal: 4.54s\tremaining: 4.52s\n",
            "600:\tlearn: 0.2248649\ttotal: 5.11s\tremaining: 3.39s\n",
            "700:\tlearn: 0.2173233\ttotal: 5.66s\tremaining: 2.41s\n",
            "800:\tlearn: 0.2104950\ttotal: 6.23s\tremaining: 1.55s\n",
            "900:\tlearn: 0.2046161\ttotal: 6.78s\tremaining: 745ms\n",
            "999:\tlearn: 0.1997916\ttotal: 7.35s\tremaining: 0us\n",
            "0:\tlearn: 0.6852649\ttotal: 2.81ms\tremaining: 2.81s\n",
            "100:\tlearn: 0.3766394\ttotal: 520ms\tremaining: 4.62s\n",
            "200:\tlearn: 0.3006610\ttotal: 1.05s\tremaining: 4.18s\n",
            "300:\tlearn: 0.2650506\ttotal: 1.6s\tremaining: 3.73s\n",
            "400:\tlearn: 0.2451767\ttotal: 2.15s\tremaining: 3.21s\n",
            "500:\tlearn: 0.2314600\ttotal: 2.71s\tremaining: 2.7s\n",
            "600:\tlearn: 0.2213122\ttotal: 3.26s\tremaining: 2.16s\n",
            "700:\tlearn: 0.2134420\ttotal: 3.82s\tremaining: 1.63s\n",
            "800:\tlearn: 0.2070866\ttotal: 4.38s\tremaining: 1.09s\n",
            "900:\tlearn: 0.2017489\ttotal: 4.95s\tremaining: 544ms\n",
            "999:\tlearn: 0.1965782\ttotal: 5.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6851405\ttotal: 4.38ms\tremaining: 4.37s\n",
            "100:\tlearn: 0.3721932\ttotal: 550ms\tremaining: 4.9s\n",
            "200:\tlearn: 0.2950510\ttotal: 1.3s\tremaining: 5.17s\n",
            "300:\tlearn: 0.2619596\ttotal: 2.56s\tremaining: 5.94s\n",
            "400:\tlearn: 0.2404144\ttotal: 3.82s\tremaining: 5.7s\n",
            "500:\tlearn: 0.2272253\ttotal: 4.53s\tremaining: 4.51s\n",
            "600:\tlearn: 0.2171139\ttotal: 5.08s\tremaining: 3.38s\n",
            "700:\tlearn: 0.2096092\ttotal: 5.65s\tremaining: 2.41s\n",
            "800:\tlearn: 0.2035697\ttotal: 6.21s\tremaining: 1.54s\n",
            "900:\tlearn: 0.1981186\ttotal: 6.77s\tremaining: 743ms\n",
            "999:\tlearn: 0.1931562\ttotal: 7.34s\tremaining: 0us\n",
            "0:\tlearn: 0.6852176\ttotal: 2.9ms\tremaining: 2.9s\n",
            "100:\tlearn: 0.3743535\ttotal: 524ms\tremaining: 4.66s\n",
            "200:\tlearn: 0.2987376\ttotal: 1.07s\tremaining: 4.26s\n",
            "300:\tlearn: 0.2659870\ttotal: 1.62s\tremaining: 3.77s\n",
            "400:\tlearn: 0.2456408\ttotal: 2.19s\tremaining: 3.28s\n",
            "500:\tlearn: 0.2316350\ttotal: 2.74s\tremaining: 2.73s\n",
            "600:\tlearn: 0.2215944\ttotal: 3.31s\tremaining: 2.2s\n",
            "700:\tlearn: 0.2129443\ttotal: 3.89s\tremaining: 1.66s\n",
            "800:\tlearn: 0.2065266\ttotal: 4.45s\tremaining: 1.1s\n",
            "900:\tlearn: 0.2007859\ttotal: 5.02s\tremaining: 551ms\n",
            "999:\tlearn: 0.1957585\ttotal: 5.57s\tremaining: 0us\n",
            "0:\tlearn: 0.6874570\ttotal: 15.7ms\tremaining: 47.1s\n",
            "100:\tlearn: 0.3783524\ttotal: 1.39s\tremaining: 39.8s\n",
            "200:\tlearn: 0.2959401\ttotal: 3.68s\tremaining: 51.3s\n",
            "300:\tlearn: 0.2595547\ttotal: 4.97s\tremaining: 44.6s\n",
            "400:\tlearn: 0.2375487\ttotal: 6.19s\tremaining: 40.1s\n",
            "500:\tlearn: 0.2217394\ttotal: 7.46s\tremaining: 37.2s\n",
            "600:\tlearn: 0.2102662\ttotal: 8.69s\tremaining: 34.7s\n",
            "700:\tlearn: 0.2002991\ttotal: 9.93s\tremaining: 32.6s\n",
            "800:\tlearn: 0.1913001\ttotal: 11.2s\tremaining: 30.7s\n",
            "900:\tlearn: 0.1834046\ttotal: 12.4s\tremaining: 29s\n",
            "1000:\tlearn: 0.1760501\ttotal: 13.7s\tremaining: 27.3s\n",
            "1100:\tlearn: 0.1690036\ttotal: 16.3s\tremaining: 28s\n",
            "1200:\tlearn: 0.1626698\ttotal: 17.9s\tremaining: 26.8s\n",
            "1300:\tlearn: 0.1567028\ttotal: 19.2s\tremaining: 25s\n",
            "1400:\tlearn: 0.1510527\ttotal: 20.4s\tremaining: 23.3s\n",
            "1500:\tlearn: 0.1458242\ttotal: 21.6s\tremaining: 21.6s\n",
            "1600:\tlearn: 0.1409535\ttotal: 22.9s\tremaining: 20s\n",
            "1700:\tlearn: 0.1362789\ttotal: 24.2s\tremaining: 18.5s\n",
            "1800:\tlearn: 0.1317220\ttotal: 25.4s\tremaining: 16.9s\n",
            "1900:\tlearn: 0.1272557\ttotal: 26.7s\tremaining: 15.4s\n",
            "2000:\tlearn: 0.1230016\ttotal: 28.9s\tremaining: 14.4s\n",
            "2100:\tlearn: 0.1188620\ttotal: 30.9s\tremaining: 13.2s\n",
            "2200:\tlearn: 0.1149822\ttotal: 32.1s\tremaining: 11.7s\n",
            "2300:\tlearn: 0.1114443\ttotal: 33.4s\tremaining: 10.1s\n",
            "2400:\tlearn: 0.1079260\ttotal: 34.6s\tremaining: 8.64s\n",
            "2500:\tlearn: 0.1045826\ttotal: 35.9s\tremaining: 7.16s\n",
            "2600:\tlearn: 0.1014739\ttotal: 37.2s\tremaining: 5.7s\n",
            "2700:\tlearn: 0.0985198\ttotal: 38.4s\tremaining: 4.25s\n",
            "2800:\tlearn: 0.0957556\ttotal: 39.7s\tremaining: 2.82s\n",
            "2900:\tlearn: 0.0931893\ttotal: 41.8s\tremaining: 1.43s\n",
            "2999:\tlearn: 0.0907240\ttotal: 43.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6875202\ttotal: 12.7ms\tremaining: 38.2s\n",
            "100:\tlearn: 0.3835562\ttotal: 1.08s\tremaining: 31.1s\n",
            "200:\tlearn: 0.2974245\ttotal: 2.26s\tremaining: 31.4s\n",
            "300:\tlearn: 0.2630902\ttotal: 3.49s\tremaining: 31.3s\n",
            "400:\tlearn: 0.2400616\ttotal: 4.77s\tremaining: 30.9s\n",
            "500:\tlearn: 0.2254507\ttotal: 6.03s\tremaining: 30.1s\n",
            "600:\tlearn: 0.2137133\ttotal: 7.29s\tremaining: 29.1s\n",
            "700:\tlearn: 0.2041135\ttotal: 8.55s\tremaining: 28s\n",
            "800:\tlearn: 0.1955167\ttotal: 10.4s\tremaining: 28.6s\n",
            "900:\tlearn: 0.1877055\ttotal: 12.8s\tremaining: 29.8s\n",
            "1000:\tlearn: 0.1804787\ttotal: 14.1s\tremaining: 28.1s\n",
            "1100:\tlearn: 0.1734512\ttotal: 15.4s\tremaining: 26.5s\n",
            "1200:\tlearn: 0.1667860\ttotal: 16.6s\tremaining: 24.9s\n",
            "1300:\tlearn: 0.1607728\ttotal: 17.9s\tremaining: 23.4s\n",
            "1400:\tlearn: 0.1551520\ttotal: 19.2s\tremaining: 21.9s\n",
            "1500:\tlearn: 0.1496612\ttotal: 20.5s\tremaining: 20.4s\n",
            "1600:\tlearn: 0.1446156\ttotal: 21.7s\tremaining: 19s\n",
            "1700:\tlearn: 0.1396436\ttotal: 23.7s\tremaining: 18.1s\n",
            "1800:\tlearn: 0.1350259\ttotal: 25.9s\tremaining: 17.3s\n",
            "1900:\tlearn: 0.1307490\ttotal: 27.2s\tremaining: 15.7s\n",
            "2000:\tlearn: 0.1268356\ttotal: 28.5s\tremaining: 14.2s\n",
            "2100:\tlearn: 0.1232047\ttotal: 29.7s\tremaining: 12.7s\n",
            "2200:\tlearn: 0.1195506\ttotal: 31s\tremaining: 11.3s\n",
            "2300:\tlearn: 0.1161377\ttotal: 32.3s\tremaining: 9.8s\n",
            "2400:\tlearn: 0.1130624\ttotal: 33.5s\tremaining: 8.37s\n",
            "2500:\tlearn: 0.1099174\ttotal: 34.8s\tremaining: 6.94s\n",
            "2600:\tlearn: 0.1069904\ttotal: 36.7s\tremaining: 5.63s\n",
            "2700:\tlearn: 0.1041142\ttotal: 39s\tremaining: 4.32s\n",
            "2800:\tlearn: 0.1014501\ttotal: 40.3s\tremaining: 2.86s\n",
            "2900:\tlearn: 0.0990455\ttotal: 41.5s\tremaining: 1.42s\n",
            "2999:\tlearn: 0.0967413\ttotal: 42.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6871499\ttotal: 12.5ms\tremaining: 37.4s\n",
            "100:\tlearn: 0.3801003\ttotal: 1.08s\tremaining: 31s\n",
            "200:\tlearn: 0.2949619\ttotal: 2.22s\tremaining: 30.9s\n",
            "300:\tlearn: 0.2593275\ttotal: 3.43s\tremaining: 30.8s\n",
            "400:\tlearn: 0.2382494\ttotal: 4.68s\tremaining: 30.3s\n",
            "500:\tlearn: 0.2226098\ttotal: 6.23s\tremaining: 31.1s\n",
            "600:\tlearn: 0.2110134\ttotal: 8.87s\tremaining: 35.4s\n",
            "700:\tlearn: 0.2013670\ttotal: 10.1s\tremaining: 33.3s\n",
            "800:\tlearn: 0.1931965\ttotal: 11.4s\tremaining: 31.3s\n",
            "900:\tlearn: 0.1850530\ttotal: 12.6s\tremaining: 29.4s\n",
            "1000:\tlearn: 0.1779543\ttotal: 13.9s\tremaining: 27.8s\n",
            "1100:\tlearn: 0.1710220\ttotal: 15.1s\tremaining: 26.1s\n",
            "1200:\tlearn: 0.1650891\ttotal: 16.4s\tremaining: 24.6s\n",
            "1300:\tlearn: 0.1591477\ttotal: 17.7s\tremaining: 23.1s\n",
            "1400:\tlearn: 0.1534933\ttotal: 19.1s\tremaining: 21.8s\n",
            "1500:\tlearn: 0.1480679\ttotal: 21.7s\tremaining: 21.7s\n",
            "1600:\tlearn: 0.1430151\ttotal: 23.1s\tremaining: 20.2s\n",
            "1700:\tlearn: 0.1381356\ttotal: 24.4s\tremaining: 18.6s\n",
            "1800:\tlearn: 0.1336630\ttotal: 25.6s\tremaining: 17.1s\n",
            "1900:\tlearn: 0.1297932\ttotal: 26.9s\tremaining: 15.6s\n",
            "2000:\tlearn: 0.1259173\ttotal: 28.2s\tremaining: 14.1s\n",
            "2100:\tlearn: 0.1222335\ttotal: 29.4s\tremaining: 12.6s\n",
            "2200:\tlearn: 0.1185331\ttotal: 30.7s\tremaining: 11.1s\n",
            "2300:\tlearn: 0.1150232\ttotal: 32s\tremaining: 9.72s\n",
            "2400:\tlearn: 0.1116353\ttotal: 34.6s\tremaining: 8.63s\n",
            "2500:\tlearn: 0.1084537\ttotal: 36.1s\tremaining: 7.21s\n",
            "2600:\tlearn: 0.1056803\ttotal: 37.4s\tremaining: 5.74s\n",
            "2700:\tlearn: 0.1028935\ttotal: 38.7s\tremaining: 4.28s\n",
            "2800:\tlearn: 0.1001854\ttotal: 39.9s\tremaining: 2.83s\n",
            "2900:\tlearn: 0.0975970\ttotal: 41.2s\tremaining: 1.4s\n",
            "2999:\tlearn: 0.0951636\ttotal: 42.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6873965\ttotal: 12.3ms\tremaining: 36.8s\n",
            "100:\tlearn: 0.3779673\ttotal: 1.03s\tremaining: 29.5s\n",
            "200:\tlearn: 0.2932554\ttotal: 2.18s\tremaining: 30.4s\n",
            "300:\tlearn: 0.2551905\ttotal: 4.23s\tremaining: 37.9s\n",
            "400:\tlearn: 0.2341523\ttotal: 6.31s\tremaining: 40.9s\n",
            "500:\tlearn: 0.2195482\ttotal: 7.56s\tremaining: 37.7s\n",
            "600:\tlearn: 0.2076671\ttotal: 8.83s\tremaining: 35.3s\n",
            "700:\tlearn: 0.1978580\ttotal: 10.1s\tremaining: 33s\n",
            "800:\tlearn: 0.1896915\ttotal: 11.3s\tremaining: 31.1s\n",
            "900:\tlearn: 0.1819833\ttotal: 12.6s\tremaining: 29.3s\n",
            "1000:\tlearn: 0.1744667\ttotal: 13.9s\tremaining: 27.7s\n",
            "1100:\tlearn: 0.1678559\ttotal: 15.1s\tremaining: 26.1s\n",
            "1200:\tlearn: 0.1616736\ttotal: 17.2s\tremaining: 25.7s\n",
            "1300:\tlearn: 0.1559295\ttotal: 19.3s\tremaining: 25.2s\n",
            "1400:\tlearn: 0.1504961\ttotal: 20.6s\tremaining: 23.5s\n",
            "1500:\tlearn: 0.1451654\ttotal: 21.8s\tremaining: 21.8s\n",
            "1600:\tlearn: 0.1398595\ttotal: 23.1s\tremaining: 20.2s\n",
            "1700:\tlearn: 0.1349840\ttotal: 24.3s\tremaining: 18.6s\n",
            "1800:\tlearn: 0.1301127\ttotal: 25.6s\tremaining: 17s\n",
            "1900:\tlearn: 0.1256482\ttotal: 26.8s\tremaining: 15.5s\n",
            "2000:\tlearn: 0.1214712\ttotal: 28.1s\tremaining: 14s\n",
            "2100:\tlearn: 0.1176153\ttotal: 30s\tremaining: 12.8s\n",
            "2200:\tlearn: 0.1138459\ttotal: 32.3s\tremaining: 11.7s\n",
            "2300:\tlearn: 0.1102236\ttotal: 33.5s\tremaining: 10.2s\n",
            "2400:\tlearn: 0.1068554\ttotal: 34.8s\tremaining: 8.68s\n",
            "2500:\tlearn: 0.1036457\ttotal: 36s\tremaining: 7.19s\n",
            "2600:\tlearn: 0.1006206\ttotal: 37.3s\tremaining: 5.72s\n",
            "2700:\tlearn: 0.0977117\ttotal: 38.6s\tremaining: 4.27s\n",
            "2800:\tlearn: 0.0950132\ttotal: 39.8s\tremaining: 2.83s\n",
            "2900:\tlearn: 0.0922922\ttotal: 41.1s\tremaining: 1.4s\n",
            "2999:\tlearn: 0.0898121\ttotal: 42.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6871332\ttotal: 23ms\tremaining: 1m 8s\n",
            "100:\tlearn: 0.3787825\ttotal: 2.15s\tremaining: 1m 1s\n",
            "200:\tlearn: 0.2963930\ttotal: 3.27s\tremaining: 45.5s\n",
            "300:\tlearn: 0.2584079\ttotal: 4.51s\tremaining: 40.4s\n",
            "400:\tlearn: 0.2370151\ttotal: 5.75s\tremaining: 37.3s\n",
            "500:\tlearn: 0.2211256\ttotal: 7.01s\tremaining: 35s\n",
            "600:\tlearn: 0.2088466\ttotal: 8.27s\tremaining: 33s\n",
            "700:\tlearn: 0.1991674\ttotal: 9.53s\tremaining: 31.2s\n",
            "800:\tlearn: 0.1906995\ttotal: 10.8s\tremaining: 29.6s\n",
            "900:\tlearn: 0.1830827\ttotal: 12.1s\tremaining: 28.1s\n",
            "1000:\tlearn: 0.1757911\ttotal: 14.6s\tremaining: 29.2s\n",
            "1100:\tlearn: 0.1692189\ttotal: 16.2s\tremaining: 27.9s\n",
            "1200:\tlearn: 0.1630473\ttotal: 17.4s\tremaining: 26.1s\n",
            "1300:\tlearn: 0.1573459\ttotal: 18.7s\tremaining: 24.4s\n",
            "1400:\tlearn: 0.1516960\ttotal: 19.9s\tremaining: 22.7s\n",
            "1500:\tlearn: 0.1465727\ttotal: 21.2s\tremaining: 21.2s\n",
            "1600:\tlearn: 0.1414630\ttotal: 22.5s\tremaining: 19.6s\n",
            "1700:\tlearn: 0.1367037\ttotal: 23.7s\tremaining: 18.1s\n",
            "1800:\tlearn: 0.1323205\ttotal: 24.9s\tremaining: 16.6s\n",
            "1900:\tlearn: 0.1280598\ttotal: 27.4s\tremaining: 15.8s\n",
            "2000:\tlearn: 0.1243207\ttotal: 29s\tremaining: 14.5s\n",
            "2100:\tlearn: 0.1206896\ttotal: 30.2s\tremaining: 12.9s\n",
            "2200:\tlearn: 0.1171215\ttotal: 31.5s\tremaining: 11.4s\n",
            "2300:\tlearn: 0.1138161\ttotal: 32.7s\tremaining: 9.95s\n",
            "2400:\tlearn: 0.1106258\ttotal: 34s\tremaining: 8.48s\n",
            "2500:\tlearn: 0.1075658\ttotal: 35.2s\tremaining: 7.03s\n",
            "2600:\tlearn: 0.1045390\ttotal: 36.5s\tremaining: 5.6s\n",
            "2700:\tlearn: 0.1015785\ttotal: 37.7s\tremaining: 4.18s\n",
            "2800:\tlearn: 0.0988827\ttotal: 40s\tremaining: 2.84s\n",
            "2900:\tlearn: 0.0960397\ttotal: 41.9s\tremaining: 1.43s\n",
            "2999:\tlearn: 0.0934214\ttotal: 43.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6706999\ttotal: 12.1ms\tremaining: 12.1s\n",
            "100:\tlearn: 0.2352128\ttotal: 1.17s\tremaining: 10.4s\n",
            "200:\tlearn: 0.1903086\ttotal: 2.4s\tremaining: 9.54s\n",
            "300:\tlearn: 0.1611033\ttotal: 3.66s\tremaining: 8.51s\n",
            "400:\tlearn: 0.1400993\ttotal: 4.91s\tremaining: 7.34s\n",
            "500:\tlearn: 0.1219636\ttotal: 6.17s\tremaining: 6.14s\n",
            "600:\tlearn: 0.1076382\ttotal: 7.43s\tremaining: 4.93s\n",
            "700:\tlearn: 0.0958738\ttotal: 9.7s\tremaining: 4.14s\n",
            "800:\tlearn: 0.0862360\ttotal: 11.6s\tremaining: 2.88s\n",
            "900:\tlearn: 0.0777726\ttotal: 12.8s\tremaining: 1.41s\n",
            "999:\tlearn: 0.0704683\ttotal: 14.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6709562\ttotal: 12.3ms\tremaining: 12.3s\n",
            "100:\tlearn: 0.2373797\ttotal: 1.19s\tremaining: 10.6s\n",
            "200:\tlearn: 0.1946068\ttotal: 2.46s\tremaining: 9.8s\n",
            "300:\tlearn: 0.1667993\ttotal: 3.7s\tremaining: 8.6s\n",
            "400:\tlearn: 0.1454547\ttotal: 4.97s\tremaining: 7.43s\n",
            "500:\tlearn: 0.1276833\ttotal: 6.23s\tremaining: 6.21s\n",
            "600:\tlearn: 0.1134806\ttotal: 8.38s\tremaining: 5.56s\n",
            "700:\tlearn: 0.1019368\ttotal: 10.4s\tremaining: 4.44s\n",
            "800:\tlearn: 0.0922288\ttotal: 11.7s\tremaining: 2.9s\n",
            "900:\tlearn: 0.0836051\ttotal: 12.9s\tremaining: 1.42s\n",
            "999:\tlearn: 0.0760023\ttotal: 14.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6695213\ttotal: 12.2ms\tremaining: 12.2s\n",
            "100:\tlearn: 0.2382536\ttotal: 1.14s\tremaining: 10.2s\n",
            "200:\tlearn: 0.1929849\ttotal: 2.39s\tremaining: 9.49s\n",
            "300:\tlearn: 0.1647432\ttotal: 3.64s\tremaining: 8.45s\n",
            "400:\tlearn: 0.1441991\ttotal: 4.89s\tremaining: 7.31s\n",
            "500:\tlearn: 0.1266711\ttotal: 6.86s\tremaining: 6.83s\n",
            "600:\tlearn: 0.1134213\ttotal: 9.06s\tremaining: 6.02s\n",
            "700:\tlearn: 0.1019476\ttotal: 10.3s\tremaining: 4.41s\n",
            "800:\tlearn: 0.0910762\ttotal: 11.6s\tremaining: 2.88s\n",
            "900:\tlearn: 0.0826027\ttotal: 12.8s\tremaining: 1.41s\n",
            "999:\tlearn: 0.0746350\ttotal: 14.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6704750\ttotal: 23.9ms\tremaining: 23.9s\n",
            "100:\tlearn: 0.2319373\ttotal: 1.16s\tremaining: 10.4s\n",
            "200:\tlearn: 0.1891075\ttotal: 2.4s\tremaining: 9.56s\n",
            "300:\tlearn: 0.1609725\ttotal: 3.69s\tremaining: 8.56s\n",
            "400:\tlearn: 0.1400529\ttotal: 5.48s\tremaining: 8.19s\n",
            "500:\tlearn: 0.1217219\ttotal: 7.85s\tremaining: 7.82s\n",
            "600:\tlearn: 0.1069202\ttotal: 9.09s\tremaining: 6.03s\n",
            "700:\tlearn: 0.0958382\ttotal: 10.3s\tremaining: 4.41s\n",
            "800:\tlearn: 0.0858644\ttotal: 11.6s\tremaining: 2.88s\n",
            "900:\tlearn: 0.0771302\ttotal: 12.8s\tremaining: 1.41s\n",
            "999:\tlearn: 0.0695179\ttotal: 14.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6694550\ttotal: 12.2ms\tremaining: 12.2s\n",
            "100:\tlearn: 0.2373992\ttotal: 1.17s\tremaining: 10.4s\n",
            "200:\tlearn: 0.1905675\ttotal: 2.42s\tremaining: 9.63s\n",
            "300:\tlearn: 0.1625193\ttotal: 3.95s\tremaining: 9.17s\n",
            "400:\tlearn: 0.1413483\ttotal: 6.56s\tremaining: 9.8s\n",
            "500:\tlearn: 0.1242272\ttotal: 7.82s\tremaining: 7.79s\n",
            "600:\tlearn: 0.1099669\ttotal: 9.07s\tremaining: 6.02s\n",
            "700:\tlearn: 0.0981539\ttotal: 10.3s\tremaining: 4.4s\n",
            "800:\tlearn: 0.0884142\ttotal: 11.6s\tremaining: 2.88s\n",
            "900:\tlearn: 0.0801216\ttotal: 12.9s\tremaining: 1.41s\n",
            "999:\tlearn: 0.0729732\ttotal: 14.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6849289\ttotal: 2.81ms\tremaining: 5.61s\n",
            "100:\tlearn: 0.3702167\ttotal: 544ms\tremaining: 10.2s\n",
            "200:\tlearn: 0.2938486\ttotal: 1.07s\tremaining: 9.6s\n",
            "300:\tlearn: 0.2592957\ttotal: 1.62s\tremaining: 9.16s\n",
            "400:\tlearn: 0.2385188\ttotal: 2.18s\tremaining: 8.69s\n",
            "500:\tlearn: 0.2269051\ttotal: 3.37s\tremaining: 10.1s\n",
            "600:\tlearn: 0.2168982\ttotal: 4.7s\tremaining: 10.9s\n",
            "700:\tlearn: 0.2089873\ttotal: 5.52s\tremaining: 10.2s\n",
            "800:\tlearn: 0.2021213\ttotal: 6.08s\tremaining: 9.1s\n",
            "900:\tlearn: 0.1963590\ttotal: 6.62s\tremaining: 8.07s\n",
            "1000:\tlearn: 0.1912010\ttotal: 7.16s\tremaining: 7.14s\n",
            "1100:\tlearn: 0.1867668\ttotal: 7.71s\tremaining: 6.3s\n",
            "1200:\tlearn: 0.1825468\ttotal: 8.26s\tremaining: 5.49s\n",
            "1300:\tlearn: 0.1784083\ttotal: 8.81s\tremaining: 4.73s\n",
            "1400:\tlearn: 0.1742846\ttotal: 9.34s\tremaining: 3.99s\n",
            "1500:\tlearn: 0.1706369\ttotal: 9.91s\tremaining: 3.29s\n",
            "1600:\tlearn: 0.1671478\ttotal: 10.5s\tremaining: 2.6s\n",
            "1700:\tlearn: 0.1636207\ttotal: 11s\tremaining: 1.94s\n",
            "1800:\tlearn: 0.1599342\ttotal: 11.5s\tremaining: 1.27s\n",
            "1900:\tlearn: 0.1564091\ttotal: 12.1s\tremaining: 631ms\n",
            "1999:\tlearn: 0.1531548\ttotal: 12.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6850157\ttotal: 2.86ms\tremaining: 5.72s\n",
            "100:\tlearn: 0.3728875\ttotal: 523ms\tremaining: 9.82s\n",
            "200:\tlearn: 0.2988398\ttotal: 1.04s\tremaining: 9.35s\n",
            "300:\tlearn: 0.2658620\ttotal: 1.58s\tremaining: 8.93s\n",
            "400:\tlearn: 0.2454030\ttotal: 2.14s\tremaining: 8.54s\n",
            "500:\tlearn: 0.2318228\ttotal: 3.41s\tremaining: 10.2s\n",
            "600:\tlearn: 0.2215363\ttotal: 4.57s\tremaining: 10.6s\n",
            "700:\tlearn: 0.2134932\ttotal: 5.59s\tremaining: 10.4s\n",
            "800:\tlearn: 0.2071238\ttotal: 6.14s\tremaining: 9.19s\n",
            "900:\tlearn: 0.2013448\ttotal: 6.68s\tremaining: 8.15s\n",
            "1000:\tlearn: 0.1961610\ttotal: 7.24s\tremaining: 7.23s\n",
            "1100:\tlearn: 0.1916964\ttotal: 7.79s\tremaining: 6.36s\n",
            "1200:\tlearn: 0.1871023\ttotal: 8.35s\tremaining: 5.55s\n",
            "1300:\tlearn: 0.1829675\ttotal: 8.89s\tremaining: 4.78s\n",
            "1400:\tlearn: 0.1789736\ttotal: 9.46s\tremaining: 4.04s\n",
            "1500:\tlearn: 0.1755978\ttotal: 10s\tremaining: 3.33s\n",
            "1600:\tlearn: 0.1719887\ttotal: 10.6s\tremaining: 2.64s\n",
            "1700:\tlearn: 0.1684407\ttotal: 11.1s\tremaining: 1.96s\n",
            "1800:\tlearn: 0.1650484\ttotal: 11.7s\tremaining: 1.29s\n",
            "1900:\tlearn: 0.1615049\ttotal: 12.2s\tremaining: 638ms\n",
            "1999:\tlearn: 0.1580308\ttotal: 12.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6849502\ttotal: 2.88ms\tremaining: 5.75s\n",
            "100:\tlearn: 0.3699299\ttotal: 528ms\tremaining: 9.94s\n",
            "200:\tlearn: 0.2965105\ttotal: 1.04s\tremaining: 9.3s\n",
            "300:\tlearn: 0.2641120\ttotal: 1.6s\tremaining: 9.05s\n",
            "400:\tlearn: 0.2432702\ttotal: 2.15s\tremaining: 8.59s\n",
            "500:\tlearn: 0.2296673\ttotal: 3.21s\tremaining: 9.62s\n",
            "600:\tlearn: 0.2190357\ttotal: 4.5s\tremaining: 10.5s\n",
            "700:\tlearn: 0.2108527\ttotal: 5.59s\tremaining: 10.4s\n",
            "800:\tlearn: 0.2044811\ttotal: 6.13s\tremaining: 9.18s\n",
            "900:\tlearn: 0.1989913\ttotal: 6.69s\tremaining: 8.16s\n",
            "1000:\tlearn: 0.1938876\ttotal: 7.25s\tremaining: 7.23s\n",
            "1100:\tlearn: 0.1893927\ttotal: 7.81s\tremaining: 6.38s\n",
            "1200:\tlearn: 0.1849903\ttotal: 8.38s\tremaining: 5.57s\n",
            "1300:\tlearn: 0.1811827\ttotal: 8.96s\tremaining: 4.81s\n",
            "1400:\tlearn: 0.1771633\ttotal: 9.53s\tremaining: 4.07s\n",
            "1500:\tlearn: 0.1734895\ttotal: 10.1s\tremaining: 3.35s\n",
            "1600:\tlearn: 0.1696803\ttotal: 10.6s\tremaining: 2.65s\n",
            "1700:\tlearn: 0.1660988\ttotal: 11.2s\tremaining: 1.96s\n",
            "1800:\tlearn: 0.1625727\ttotal: 11.7s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.1591002\ttotal: 12.3s\tremaining: 640ms\n",
            "1999:\tlearn: 0.1558495\ttotal: 12.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6848234\ttotal: 3.06ms\tremaining: 6.11s\n",
            "100:\tlearn: 0.3655791\ttotal: 536ms\tremaining: 10.1s\n",
            "200:\tlearn: 0.2902426\ttotal: 1.06s\tremaining: 9.51s\n",
            "300:\tlearn: 0.2581238\ttotal: 1.61s\tremaining: 9.07s\n",
            "400:\tlearn: 0.2367134\ttotal: 2.15s\tremaining: 8.58s\n",
            "500:\tlearn: 0.2234769\ttotal: 3.14s\tremaining: 9.38s\n",
            "600:\tlearn: 0.2141053\ttotal: 4.29s\tremaining: 9.99s\n",
            "700:\tlearn: 0.2063986\ttotal: 5.6s\tremaining: 10.4s\n",
            "800:\tlearn: 0.2005402\ttotal: 6.16s\tremaining: 9.21s\n",
            "900:\tlearn: 0.1949591\ttotal: 6.71s\tremaining: 8.19s\n",
            "1000:\tlearn: 0.1899110\ttotal: 7.26s\tremaining: 7.25s\n",
            "1100:\tlearn: 0.1854461\ttotal: 7.83s\tremaining: 6.39s\n",
            "1200:\tlearn: 0.1811289\ttotal: 8.37s\tremaining: 5.57s\n",
            "1300:\tlearn: 0.1772359\ttotal: 8.93s\tremaining: 4.8s\n",
            "1400:\tlearn: 0.1731762\ttotal: 9.47s\tremaining: 4.05s\n",
            "1500:\tlearn: 0.1695573\ttotal: 10s\tremaining: 3.34s\n",
            "1600:\tlearn: 0.1660645\ttotal: 10.6s\tremaining: 2.64s\n",
            "1700:\tlearn: 0.1626182\ttotal: 11.2s\tremaining: 1.96s\n",
            "1800:\tlearn: 0.1591716\ttotal: 11.7s\tremaining: 1.29s\n",
            "1900:\tlearn: 0.1557056\ttotal: 12.3s\tremaining: 638ms\n",
            "1999:\tlearn: 0.1520350\ttotal: 12.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6848996\ttotal: 2.87ms\tremaining: 5.74s\n",
            "100:\tlearn: 0.3680553\ttotal: 512ms\tremaining: 9.62s\n",
            "200:\tlearn: 0.2941121\ttotal: 1.08s\tremaining: 9.64s\n",
            "300:\tlearn: 0.2613839\ttotal: 1.61s\tremaining: 9.1s\n",
            "400:\tlearn: 0.2405673\ttotal: 2.17s\tremaining: 8.67s\n",
            "500:\tlearn: 0.2281903\ttotal: 3.1s\tremaining: 9.27s\n",
            "600:\tlearn: 0.2184132\ttotal: 4.36s\tremaining: 10.2s\n",
            "700:\tlearn: 0.2100810\ttotal: 5.67s\tremaining: 10.5s\n",
            "800:\tlearn: 0.2036710\ttotal: 6.24s\tremaining: 9.34s\n",
            "900:\tlearn: 0.1979447\ttotal: 6.78s\tremaining: 8.27s\n",
            "1000:\tlearn: 0.1926032\ttotal: 7.34s\tremaining: 7.32s\n",
            "1100:\tlearn: 0.1878643\ttotal: 7.88s\tremaining: 6.43s\n",
            "1200:\tlearn: 0.1830854\ttotal: 8.44s\tremaining: 5.61s\n",
            "1300:\tlearn: 0.1792115\ttotal: 9s\tremaining: 4.84s\n",
            "1400:\tlearn: 0.1750275\ttotal: 9.55s\tremaining: 4.08s\n",
            "1500:\tlearn: 0.1713379\ttotal: 10.1s\tremaining: 3.37s\n",
            "1600:\tlearn: 0.1679463\ttotal: 10.7s\tremaining: 2.66s\n",
            "1700:\tlearn: 0.1645556\ttotal: 11.2s\tremaining: 1.98s\n",
            "1800:\tlearn: 0.1610316\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.1576126\ttotal: 12.4s\tremaining: 644ms\n",
            "1999:\tlearn: 0.1541038\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6878794\ttotal: 35.2ms\tremaining: 1m 10s\n",
            "100:\tlearn: 0.3890668\ttotal: 3.25s\tremaining: 1m 1s\n",
            "200:\tlearn: 0.2977377\ttotal: 7.28s\tremaining: 1m 5s\n",
            "300:\tlearn: 0.2550239\ttotal: 10.6s\tremaining: 59.9s\n",
            "400:\tlearn: 0.2285433\ttotal: 14.1s\tremaining: 56.1s\n",
            "500:\tlearn: 0.2103904\ttotal: 19s\tremaining: 56.9s\n",
            "600:\tlearn: 0.1949840\ttotal: 22.6s\tremaining: 52.6s\n",
            "700:\tlearn: 0.1818261\ttotal: 26.2s\tremaining: 48.6s\n",
            "800:\tlearn: 0.1698758\ttotal: 31s\tremaining: 46.3s\n",
            "900:\tlearn: 0.1589180\ttotal: 34.9s\tremaining: 42.5s\n",
            "1000:\tlearn: 0.1487939\ttotal: 38.5s\tremaining: 38.4s\n",
            "1100:\tlearn: 0.1403625\ttotal: 42.5s\tremaining: 34.7s\n",
            "1200:\tlearn: 0.1319027\ttotal: 47.1s\tremaining: 31.3s\n",
            "1300:\tlearn: 0.1243531\ttotal: 50.7s\tremaining: 27.3s\n",
            "1400:\tlearn: 0.1175627\ttotal: 54.4s\tremaining: 23.2s\n",
            "1500:\tlearn: 0.1109245\ttotal: 59.4s\tremaining: 19.7s\n",
            "1600:\tlearn: 0.1050425\ttotal: 1m 3s\tremaining: 15.7s\n",
            "1700:\tlearn: 0.0996763\ttotal: 1m 6s\tremaining: 11.7s\n",
            "1800:\tlearn: 0.0947993\ttotal: 1m 11s\tremaining: 7.92s\n",
            "1900:\tlearn: 0.0903407\ttotal: 1m 15s\tremaining: 3.92s\n",
            "1999:\tlearn: 0.0862551\ttotal: 1m 18s\tremaining: 0us\n",
            "0:\tlearn: 0.6885432\ttotal: 36.2ms\tremaining: 1m 12s\n",
            "100:\tlearn: 0.3867740\ttotal: 3.37s\tremaining: 1m 3s\n",
            "200:\tlearn: 0.2989810\ttotal: 7s\tremaining: 1m 2s\n",
            "300:\tlearn: 0.2586152\ttotal: 10.4s\tremaining: 58.8s\n",
            "400:\tlearn: 0.2342127\ttotal: 14s\tremaining: 55.9s\n",
            "500:\tlearn: 0.2158961\ttotal: 19s\tremaining: 56.9s\n",
            "600:\tlearn: 0.2007999\ttotal: 22.7s\tremaining: 52.8s\n",
            "700:\tlearn: 0.1872122\ttotal: 26.4s\tremaining: 48.9s\n",
            "800:\tlearn: 0.1749156\ttotal: 31.5s\tremaining: 47.1s\n",
            "900:\tlearn: 0.1639075\ttotal: 35.1s\tremaining: 42.8s\n",
            "1000:\tlearn: 0.1539572\ttotal: 38.8s\tremaining: 38.7s\n",
            "1100:\tlearn: 0.1449590\ttotal: 43.7s\tremaining: 35.7s\n",
            "1200:\tlearn: 0.1362142\ttotal: 47.6s\tremaining: 31.7s\n",
            "1300:\tlearn: 0.1284773\ttotal: 51.4s\tremaining: 27.6s\n",
            "1400:\tlearn: 0.1211422\ttotal: 55.6s\tremaining: 23.8s\n",
            "1500:\tlearn: 0.1144855\ttotal: 1m\tremaining: 20s\n",
            "1600:\tlearn: 0.1086725\ttotal: 1m 3s\tremaining: 15.9s\n",
            "1700:\tlearn: 0.1032766\ttotal: 1m 7s\tremaining: 11.9s\n",
            "1800:\tlearn: 0.0982651\ttotal: 1m 12s\tremaining: 8.04s\n",
            "1900:\tlearn: 0.0937399\ttotal: 1m 16s\tremaining: 3.98s\n",
            "1999:\tlearn: 0.0895823\ttotal: 1m 20s\tremaining: 0us\n",
            "0:\tlearn: 0.6875591\ttotal: 37.4ms\tremaining: 1m 14s\n",
            "100:\tlearn: 0.3914548\ttotal: 4.39s\tremaining: 1m 22s\n",
            "200:\tlearn: 0.2971096\ttotal: 7.45s\tremaining: 1m 6s\n",
            "300:\tlearn: 0.2560457\ttotal: 10.9s\tremaining: 1m 1s\n",
            "400:\tlearn: 0.2311334\ttotal: 15.7s\tremaining: 1m 2s\n",
            "500:\tlearn: 0.2124624\ttotal: 19.7s\tremaining: 58.9s\n",
            "600:\tlearn: 0.1977720\ttotal: 23.4s\tremaining: 54.4s\n",
            "700:\tlearn: 0.1848111\ttotal: 27.7s\tremaining: 51.4s\n",
            "800:\tlearn: 0.1729368\ttotal: 32s\tremaining: 47.9s\n",
            "900:\tlearn: 0.1619727\ttotal: 35.6s\tremaining: 43.4s\n",
            "1000:\tlearn: 0.1522646\ttotal: 39.2s\tremaining: 39.1s\n",
            "1100:\tlearn: 0.1429716\ttotal: 44.2s\tremaining: 36.1s\n",
            "1200:\tlearn: 0.1346852\ttotal: 47.9s\tremaining: 31.8s\n",
            "1300:\tlearn: 0.1267069\ttotal: 51.6s\tremaining: 27.7s\n",
            "1400:\tlearn: 0.1193382\ttotal: 56.7s\tremaining: 24.3s\n",
            "1500:\tlearn: 0.1130877\ttotal: 1m\tremaining: 20.1s\n",
            "1600:\tlearn: 0.1071453\ttotal: 1m 4s\tremaining: 16s\n",
            "1700:\tlearn: 0.1021147\ttotal: 1m 9s\tremaining: 12.2s\n",
            "1800:\tlearn: 0.0973540\ttotal: 1m 12s\tremaining: 8.05s\n",
            "1900:\tlearn: 0.0929661\ttotal: 1m 16s\tremaining: 3.98s\n",
            "1999:\tlearn: 0.0889589\ttotal: 1m 21s\tremaining: 0us\n",
            "0:\tlearn: 0.6879336\ttotal: 36.2ms\tremaining: 1m 12s\n",
            "100:\tlearn: 0.3816595\ttotal: 2.72s\tremaining: 51.1s\n",
            "200:\tlearn: 0.2929880\ttotal: 5.8s\tremaining: 51.9s\n",
            "300:\tlearn: 0.2519855\ttotal: 9.11s\tremaining: 51.4s\n",
            "400:\tlearn: 0.2277039\ttotal: 14.3s\tremaining: 57s\n",
            "500:\tlearn: 0.2086755\ttotal: 18.1s\tremaining: 54.3s\n",
            "600:\tlearn: 0.1934992\ttotal: 21.9s\tremaining: 51s\n",
            "700:\tlearn: 0.1798179\ttotal: 27.3s\tremaining: 50.5s\n",
            "800:\tlearn: 0.1682417\ttotal: 31s\tremaining: 46.4s\n",
            "900:\tlearn: 0.1570994\ttotal: 34.7s\tremaining: 42.3s\n",
            "1000:\tlearn: 0.1474516\ttotal: 40s\tremaining: 39.9s\n",
            "1100:\tlearn: 0.1383073\ttotal: 43.7s\tremaining: 35.7s\n",
            "1200:\tlearn: 0.1300619\ttotal: 47.4s\tremaining: 31.5s\n",
            "1300:\tlearn: 0.1224685\ttotal: 52.2s\tremaining: 28s\n",
            "1400:\tlearn: 0.1156591\ttotal: 56.3s\tremaining: 24.1s\n",
            "1500:\tlearn: 0.1094515\ttotal: 1m\tremaining: 20s\n",
            "1600:\tlearn: 0.1036382\ttotal: 1m 4s\tremaining: 16s\n",
            "1700:\tlearn: 0.0982150\ttotal: 1m 8s\tremaining: 12.1s\n",
            "1800:\tlearn: 0.0934382\ttotal: 1m 12s\tremaining: 8.02s\n",
            "1900:\tlearn: 0.0890034\ttotal: 1m 16s\tremaining: 3.99s\n",
            "1999:\tlearn: 0.0849232\ttotal: 1m 21s\tremaining: 0us\n",
            "0:\tlearn: 0.6873200\ttotal: 36.6ms\tremaining: 1m 13s\n",
            "100:\tlearn: 0.3861556\ttotal: 2.81s\tremaining: 52.8s\n",
            "200:\tlearn: 0.2962213\ttotal: 6.13s\tremaining: 54.9s\n",
            "300:\tlearn: 0.2548701\ttotal: 11.1s\tremaining: 1m 2s\n",
            "400:\tlearn: 0.2313588\ttotal: 14.7s\tremaining: 58.5s\n",
            "500:\tlearn: 0.2134847\ttotal: 18.4s\tremaining: 55s\n",
            "600:\tlearn: 0.1981134\ttotal: 23.4s\tremaining: 54.5s\n",
            "700:\tlearn: 0.1842945\ttotal: 27.2s\tremaining: 50.3s\n",
            "800:\tlearn: 0.1719717\ttotal: 30.8s\tremaining: 46.1s\n",
            "900:\tlearn: 0.1612883\ttotal: 35.4s\tremaining: 43.1s\n",
            "1000:\tlearn: 0.1511487\ttotal: 39.6s\tremaining: 39.5s\n",
            "1100:\tlearn: 0.1426445\ttotal: 43.4s\tremaining: 35.4s\n",
            "1200:\tlearn: 0.1342949\ttotal: 47.7s\tremaining: 31.7s\n",
            "1300:\tlearn: 0.1263694\ttotal: 52.4s\tremaining: 28.1s\n",
            "1400:\tlearn: 0.1194790\ttotal: 56.1s\tremaining: 24s\n",
            "1500:\tlearn: 0.1132208\ttotal: 59.9s\tremaining: 19.9s\n",
            "1600:\tlearn: 0.1074794\ttotal: 1m 4s\tremaining: 16.2s\n",
            "1700:\tlearn: 0.1019938\ttotal: 1m 8s\tremaining: 12.1s\n",
            "1800:\tlearn: 0.0971448\ttotal: 1m 12s\tremaining: 8.01s\n",
            "1900:\tlearn: 0.0928321\ttotal: 1m 17s\tremaining: 4.05s\n",
            "1999:\tlearn: 0.0888168\ttotal: 1m 21s\tremaining: 0us\n",
            "0:\tlearn: 0.6878794\ttotal: 35.8ms\tremaining: 1m 47s\n",
            "100:\tlearn: 0.3890668\ttotal: 2.94s\tremaining: 1m 24s\n",
            "200:\tlearn: 0.2977377\ttotal: 7.51s\tremaining: 1m 44s\n",
            "300:\tlearn: 0.2550239\ttotal: 10.9s\tremaining: 1m 37s\n",
            "400:\tlearn: 0.2285433\ttotal: 14.5s\tremaining: 1m 33s\n",
            "500:\tlearn: 0.2103904\ttotal: 18.6s\tremaining: 1m 32s\n",
            "600:\tlearn: 0.1949840\ttotal: 23.2s\tremaining: 1m 32s\n",
            "700:\tlearn: 0.1818261\ttotal: 26.9s\tremaining: 1m 28s\n",
            "800:\tlearn: 0.1698758\ttotal: 30.6s\tremaining: 1m 24s\n",
            "900:\tlearn: 0.1589180\ttotal: 35.7s\tremaining: 1m 23s\n",
            "1000:\tlearn: 0.1487939\ttotal: 39.4s\tremaining: 1m 18s\n",
            "1100:\tlearn: 0.1403625\ttotal: 43.1s\tremaining: 1m 14s\n",
            "1200:\tlearn: 0.1319027\ttotal: 48.2s\tremaining: 1m 12s\n",
            "1300:\tlearn: 0.1243531\ttotal: 52s\tremaining: 1m 7s\n",
            "1400:\tlearn: 0.1175627\ttotal: 55.7s\tremaining: 1m 3s\n",
            "1500:\tlearn: 0.1109245\ttotal: 1m\tremaining: 1m\n",
            "1600:\tlearn: 0.1050425\ttotal: 1m 4s\tremaining: 56.3s\n",
            "1700:\tlearn: 0.0996763\ttotal: 1m 8s\tremaining: 52s\n",
            "1800:\tlearn: 0.0947993\ttotal: 1m 13s\tremaining: 48.7s\n",
            "1900:\tlearn: 0.0903407\ttotal: 1m 16s\tremaining: 44.4s\n",
            "2000:\tlearn: 0.0862148\ttotal: 1m 20s\tremaining: 40.3s\n",
            "2100:\tlearn: 0.0822814\ttotal: 1m 25s\tremaining: 36.6s\n",
            "2200:\tlearn: 0.0787032\ttotal: 1m 29s\tremaining: 32.5s\n",
            "2300:\tlearn: 0.0754192\ttotal: 1m 33s\tremaining: 28.4s\n",
            "2400:\tlearn: 0.0725374\ttotal: 1m 37s\tremaining: 24.4s\n",
            "2500:\tlearn: 0.0697554\ttotal: 1m 42s\tremaining: 20.4s\n",
            "2600:\tlearn: 0.0672138\ttotal: 1m 46s\tremaining: 16.3s\n",
            "2700:\tlearn: 0.0646929\ttotal: 1m 50s\tremaining: 12.2s\n",
            "2800:\tlearn: 0.0622309\ttotal: 1m 55s\tremaining: 8.17s\n",
            "2900:\tlearn: 0.0598831\ttotal: 1m 58s\tremaining: 4.05s\n",
            "2999:\tlearn: 0.0576239\ttotal: 2m 2s\tremaining: 0us\n",
            "0:\tlearn: 0.6885432\ttotal: 59.1ms\tremaining: 2m 57s\n",
            "100:\tlearn: 0.3867740\ttotal: 3.95s\tremaining: 1m 53s\n",
            "200:\tlearn: 0.2989810\ttotal: 6.93s\tremaining: 1m 36s\n",
            "300:\tlearn: 0.2586152\ttotal: 10.4s\tremaining: 1m 32s\n",
            "400:\tlearn: 0.2342127\ttotal: 15s\tremaining: 1m 37s\n",
            "500:\tlearn: 0.2158961\ttotal: 18.9s\tremaining: 1m 34s\n",
            "600:\tlearn: 0.2007999\ttotal: 22.6s\tremaining: 1m 30s\n",
            "700:\tlearn: 0.1872122\ttotal: 26.8s\tremaining: 1m 27s\n",
            "800:\tlearn: 0.1749156\ttotal: 31.4s\tremaining: 1m 26s\n",
            "900:\tlearn: 0.1639075\ttotal: 35.1s\tremaining: 1m 21s\n",
            "1000:\tlearn: 0.1539572\ttotal: 38.8s\tremaining: 1m 17s\n",
            "1100:\tlearn: 0.1449590\ttotal: 44s\tremaining: 1m 15s\n",
            "1200:\tlearn: 0.1362142\ttotal: 47.7s\tremaining: 1m 11s\n",
            "1300:\tlearn: 0.1284773\ttotal: 51.5s\tremaining: 1m 7s\n",
            "1400:\tlearn: 0.1211422\ttotal: 56.6s\tremaining: 1m 4s\n",
            "1500:\tlearn: 0.1144855\ttotal: 1m\tremaining: 1m\n",
            "1600:\tlearn: 0.1086725\ttotal: 1m 3s\tremaining: 55.9s\n",
            "1700:\tlearn: 0.1032766\ttotal: 1m 9s\tremaining: 52.8s\n",
            "1800:\tlearn: 0.0982651\ttotal: 1m 12s\tremaining: 48.5s\n",
            "1900:\tlearn: 0.0937399\ttotal: 1m 16s\tremaining: 44.3s\n",
            "2000:\tlearn: 0.0895561\ttotal: 1m 21s\tremaining: 40.8s\n",
            "2100:\tlearn: 0.0856893\ttotal: 1m 25s\tremaining: 36.6s\n",
            "2200:\tlearn: 0.0822665\ttotal: 1m 29s\tremaining: 32.4s\n",
            "2300:\tlearn: 0.0788971\ttotal: 1m 34s\tremaining: 28.6s\n",
            "2400:\tlearn: 0.0755965\ttotal: 1m 38s\tremaining: 24.5s\n",
            "2500:\tlearn: 0.0726144\ttotal: 1m 41s\tremaining: 20.3s\n",
            "2600:\tlearn: 0.0697558\ttotal: 1m 46s\tremaining: 16.4s\n",
            "2700:\tlearn: 0.0671422\ttotal: 1m 50s\tremaining: 12.3s\n",
            "2800:\tlearn: 0.0647666\ttotal: 1m 54s\tremaining: 8.15s\n",
            "2900:\tlearn: 0.0624002\ttotal: 1m 59s\tremaining: 4.07s\n",
            "2999:\tlearn: 0.0602760\ttotal: 2m 3s\tremaining: 0us\n",
            "0:\tlearn: 0.6875591\ttotal: 36.3ms\tremaining: 1m 48s\n",
            "100:\tlearn: 0.3914548\ttotal: 2.89s\tremaining: 1m 22s\n",
            "200:\tlearn: 0.2971096\ttotal: 5.96s\tremaining: 1m 22s\n",
            "300:\tlearn: 0.2560457\ttotal: 10.9s\tremaining: 1m 37s\n",
            "400:\tlearn: 0.2311334\ttotal: 14.5s\tremaining: 1m 34s\n",
            "500:\tlearn: 0.2124624\ttotal: 18.3s\tremaining: 1m 31s\n",
            "600:\tlearn: 0.1977720\ttotal: 23.3s\tremaining: 1m 32s\n",
            "700:\tlearn: 0.1848111\ttotal: 26.9s\tremaining: 1m 28s\n",
            "800:\tlearn: 0.1729368\ttotal: 30.6s\tremaining: 1m 23s\n",
            "900:\tlearn: 0.1619727\ttotal: 35.4s\tremaining: 1m 22s\n",
            "1000:\tlearn: 0.1522646\ttotal: 39.3s\tremaining: 1m 18s\n",
            "1100:\tlearn: 0.1429716\ttotal: 42.9s\tremaining: 1m 14s\n",
            "1200:\tlearn: 0.1346852\ttotal: 47.5s\tremaining: 1m 11s\n",
            "1300:\tlearn: 0.1267069\ttotal: 51.8s\tremaining: 1m 7s\n",
            "1400:\tlearn: 0.1193382\ttotal: 55.6s\tremaining: 1m 3s\n",
            "1500:\tlearn: 0.1130877\ttotal: 59.7s\tremaining: 59.6s\n",
            "1600:\tlearn: 0.1071453\ttotal: 1m 4s\tremaining: 56.4s\n",
            "1700:\tlearn: 0.1021147\ttotal: 1m 8s\tremaining: 52.1s\n",
            "1800:\tlearn: 0.0973540\ttotal: 1m 12s\tremaining: 48s\n",
            "1900:\tlearn: 0.0929661\ttotal: 1m 17s\tremaining: 44.7s\n",
            "2000:\tlearn: 0.0889198\ttotal: 1m 21s\tremaining: 40.5s\n",
            "2100:\tlearn: 0.0849214\ttotal: 1m 24s\tremaining: 36.3s\n",
            "2200:\tlearn: 0.0815273\ttotal: 1m 30s\tremaining: 32.7s\n",
            "2300:\tlearn: 0.0782887\ttotal: 1m 33s\tremaining: 28.5s\n",
            "2400:\tlearn: 0.0751770\ttotal: 1m 37s\tremaining: 24.3s\n",
            "2500:\tlearn: 0.0722002\ttotal: 1m 42s\tremaining: 20.5s\n",
            "2600:\tlearn: 0.0692336\ttotal: 1m 46s\tremaining: 16.4s\n",
            "2700:\tlearn: 0.0665802\ttotal: 1m 50s\tremaining: 12.2s\n",
            "2800:\tlearn: 0.0641347\ttotal: 1m 55s\tremaining: 8.21s\n",
            "2900:\tlearn: 0.0618853\ttotal: 1m 59s\tremaining: 4.07s\n",
            "2999:\tlearn: 0.0597579\ttotal: 2m 3s\tremaining: 0us\n",
            "0:\tlearn: 0.6879336\ttotal: 35.9ms\tremaining: 1m 47s\n",
            "100:\tlearn: 0.3816595\ttotal: 3.72s\tremaining: 1m 46s\n",
            "200:\tlearn: 0.2929880\ttotal: 7.18s\tremaining: 1m 40s\n",
            "300:\tlearn: 0.2519855\ttotal: 10.5s\tremaining: 1m 33s\n",
            "400:\tlearn: 0.2277039\ttotal: 14s\tremaining: 1m 30s\n",
            "500:\tlearn: 0.2086755\ttotal: 19.2s\tremaining: 1m 35s\n",
            "600:\tlearn: 0.1934992\ttotal: 22.9s\tremaining: 1m 31s\n",
            "700:\tlearn: 0.1798179\ttotal: 26.5s\tremaining: 1m 26s\n",
            "800:\tlearn: 0.1682417\ttotal: 31.7s\tremaining: 1m 27s\n",
            "900:\tlearn: 0.1570994\ttotal: 35.4s\tremaining: 1m 22s\n",
            "1000:\tlearn: 0.1474516\ttotal: 39.3s\tremaining: 1m 18s\n",
            "1100:\tlearn: 0.1383073\ttotal: 44.4s\tremaining: 1m 16s\n",
            "1200:\tlearn: 0.1300619\ttotal: 48.1s\tremaining: 1m 12s\n",
            "1300:\tlearn: 0.1224685\ttotal: 51.8s\tremaining: 1m 7s\n",
            "1400:\tlearn: 0.1156591\ttotal: 56.6s\tremaining: 1m 4s\n",
            "1500:\tlearn: 0.1094515\ttotal: 1m\tremaining: 1m\n",
            "1600:\tlearn: 0.1036382\ttotal: 1m 4s\tremaining: 56s\n",
            "1700:\tlearn: 0.0982150\ttotal: 1m 8s\tremaining: 52.4s\n",
            "1800:\tlearn: 0.0934382\ttotal: 1m 12s\tremaining: 48.6s\n",
            "1900:\tlearn: 0.0890034\ttotal: 1m 16s\tremaining: 44.3s\n",
            "2000:\tlearn: 0.0848859\ttotal: 1m 20s\tremaining: 40.2s\n",
            "2100:\tlearn: 0.0812060\ttotal: 1m 25s\tremaining: 36.6s\n",
            "2200:\tlearn: 0.0777770\ttotal: 1m 29s\tremaining: 32.3s\n",
            "2300:\tlearn: 0.0744607\ttotal: 1m 32s\tremaining: 28.2s\n",
            "2400:\tlearn: 0.0713980\ttotal: 1m 37s\tremaining: 24.4s\n",
            "2500:\tlearn: 0.0685676\ttotal: 1m 41s\tremaining: 20.3s\n",
            "2600:\tlearn: 0.0657970\ttotal: 1m 45s\tremaining: 16.2s\n",
            "2700:\tlearn: 0.0632032\ttotal: 1m 50s\tremaining: 12.2s\n",
            "2800:\tlearn: 0.0609313\ttotal: 1m 54s\tremaining: 8.12s\n",
            "2900:\tlearn: 0.0585855\ttotal: 1m 57s\tremaining: 4.03s\n",
            "2999:\tlearn: 0.0564846\ttotal: 2m 3s\tremaining: 0us\n",
            "0:\tlearn: 0.6873200\ttotal: 37.2ms\tremaining: 1m 51s\n",
            "100:\tlearn: 0.3861556\ttotal: 2.83s\tremaining: 1m 21s\n",
            "200:\tlearn: 0.2962213\ttotal: 6.14s\tremaining: 1m 25s\n",
            "300:\tlearn: 0.2548701\ttotal: 10.2s\tremaining: 1m 31s\n",
            "400:\tlearn: 0.2313588\ttotal: 14.5s\tremaining: 1m 33s\n",
            "500:\tlearn: 0.2134847\ttotal: 18.1s\tremaining: 1m 30s\n",
            "600:\tlearn: 0.1981134\ttotal: 21.9s\tremaining: 1m 27s\n",
            "700:\tlearn: 0.1842945\ttotal: 26.9s\tremaining: 1m 28s\n",
            "800:\tlearn: 0.1719717\ttotal: 30.6s\tremaining: 1m 24s\n",
            "900:\tlearn: 0.1612883\ttotal: 34.3s\tremaining: 1m 19s\n",
            "1000:\tlearn: 0.1511487\ttotal: 39.4s\tremaining: 1m 18s\n",
            "1100:\tlearn: 0.1426445\ttotal: 43.1s\tremaining: 1m 14s\n",
            "1200:\tlearn: 0.1342949\ttotal: 46.8s\tremaining: 1m 10s\n",
            "1300:\tlearn: 0.1263694\ttotal: 51.9s\tremaining: 1m 7s\n",
            "1400:\tlearn: 0.1194790\ttotal: 55.6s\tremaining: 1m 3s\n",
            "1500:\tlearn: 0.1132208\ttotal: 59.3s\tremaining: 59.2s\n",
            "1600:\tlearn: 0.1074794\ttotal: 1m 4s\tremaining: 56.1s\n",
            "1700:\tlearn: 0.1019938\ttotal: 1m 7s\tremaining: 51.9s\n",
            "1800:\tlearn: 0.0971448\ttotal: 1m 11s\tremaining: 47.7s\n",
            "1900:\tlearn: 0.0928321\ttotal: 1m 16s\tremaining: 44s\n",
            "2000:\tlearn: 0.0887653\ttotal: 1m 20s\tremaining: 40.2s\n",
            "2100:\tlearn: 0.0849580\ttotal: 1m 24s\tremaining: 36s\n",
            "2200:\tlearn: 0.0812312\ttotal: 1m 28s\tremaining: 32s\n",
            "2300:\tlearn: 0.0779560\ttotal: 1m 33s\tremaining: 28.3s\n",
            "2400:\tlearn: 0.0748578\ttotal: 1m 36s\tremaining: 24.1s\n",
            "2500:\tlearn: 0.0719138\ttotal: 1m 40s\tremaining: 20.1s\n",
            "2600:\tlearn: 0.0691353\ttotal: 1m 45s\tremaining: 16.2s\n",
            "2700:\tlearn: 0.0667195\ttotal: 1m 49s\tremaining: 12.1s\n",
            "2800:\tlearn: 0.0644734\ttotal: 1m 53s\tremaining: 8.03s\n",
            "2900:\tlearn: 0.0622526\ttotal: 1m 58s\tremaining: 4.03s\n",
            "2999:\tlearn: 0.0600973\ttotal: 2m 1s\tremaining: 0us\n",
            "0:\tlearn: 0.6870521\ttotal: 12.2ms\tremaining: 24.5s\n",
            "100:\tlearn: 0.3718630\ttotal: 1.09s\tremaining: 20.6s\n",
            "200:\tlearn: 0.2889148\ttotal: 2.25s\tremaining: 20.2s\n",
            "300:\tlearn: 0.2548081\ttotal: 3.45s\tremaining: 19.5s\n",
            "400:\tlearn: 0.2337375\ttotal: 4.69s\tremaining: 18.7s\n",
            "500:\tlearn: 0.2177216\ttotal: 7.13s\tremaining: 21.3s\n",
            "600:\tlearn: 0.2061095\ttotal: 8.91s\tremaining: 20.7s\n",
            "700:\tlearn: 0.1951439\ttotal: 10.1s\tremaining: 18.8s\n",
            "800:\tlearn: 0.1859570\ttotal: 11.4s\tremaining: 17.1s\n",
            "900:\tlearn: 0.1773362\ttotal: 12.6s\tremaining: 15.4s\n",
            "1000:\tlearn: 0.1696604\ttotal: 13.9s\tremaining: 13.8s\n",
            "1100:\tlearn: 0.1625505\ttotal: 15.1s\tremaining: 12.3s\n",
            "1200:\tlearn: 0.1559986\ttotal: 16.4s\tremaining: 10.9s\n",
            "1300:\tlearn: 0.1498398\ttotal: 17.6s\tremaining: 9.47s\n",
            "1400:\tlearn: 0.1442031\ttotal: 19.8s\tremaining: 8.45s\n",
            "1500:\tlearn: 0.1386029\ttotal: 21.9s\tremaining: 7.26s\n",
            "1600:\tlearn: 0.1336207\ttotal: 23.1s\tremaining: 5.76s\n",
            "1700:\tlearn: 0.1285794\ttotal: 24.3s\tremaining: 4.28s\n",
            "1800:\tlearn: 0.1239130\ttotal: 25.6s\tremaining: 2.83s\n",
            "1900:\tlearn: 0.1192380\ttotal: 26.8s\tremaining: 1.4s\n",
            "1999:\tlearn: 0.1148696\ttotal: 28.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6871162\ttotal: 12.1ms\tremaining: 24.3s\n",
            "100:\tlearn: 0.3762849\ttotal: 1.12s\tremaining: 21s\n",
            "200:\tlearn: 0.2921564\ttotal: 2.29s\tremaining: 20.5s\n",
            "300:\tlearn: 0.2556317\ttotal: 3.99s\tremaining: 22.5s\n",
            "400:\tlearn: 0.2346940\ttotal: 6.21s\tremaining: 24.8s\n",
            "500:\tlearn: 0.2196354\ttotal: 7.47s\tremaining: 22.4s\n",
            "600:\tlearn: 0.2071306\ttotal: 8.76s\tremaining: 20.4s\n",
            "700:\tlearn: 0.1971736\ttotal: 10s\tremaining: 18.6s\n",
            "800:\tlearn: 0.1879399\ttotal: 11.3s\tremaining: 16.9s\n",
            "900:\tlearn: 0.1801303\ttotal: 12.5s\tremaining: 15.3s\n",
            "1000:\tlearn: 0.1731042\ttotal: 13.8s\tremaining: 13.8s\n",
            "1100:\tlearn: 0.1661613\ttotal: 15.1s\tremaining: 12.3s\n",
            "1200:\tlearn: 0.1594857\ttotal: 16.9s\tremaining: 11.3s\n",
            "1300:\tlearn: 0.1534432\ttotal: 19.3s\tremaining: 10.4s\n",
            "1400:\tlearn: 0.1477283\ttotal: 20.5s\tremaining: 8.77s\n",
            "1500:\tlearn: 0.1422667\ttotal: 21.8s\tremaining: 7.24s\n",
            "1600:\tlearn: 0.1371533\ttotal: 23s\tremaining: 5.74s\n",
            "1700:\tlearn: 0.1320757\ttotal: 24.3s\tremaining: 4.27s\n",
            "1800:\tlearn: 0.1275730\ttotal: 25.6s\tremaining: 2.82s\n",
            "1900:\tlearn: 0.1232790\ttotal: 26.8s\tremaining: 1.4s\n",
            "1999:\tlearn: 0.1192001\ttotal: 28s\tremaining: 0us\n",
            "0:\tlearn: 0.6867492\ttotal: 12.2ms\tremaining: 24.4s\n",
            "100:\tlearn: 0.3727338\ttotal: 1.44s\tremaining: 27.1s\n",
            "200:\tlearn: 0.2914114\ttotal: 3.86s\tremaining: 34.5s\n",
            "300:\tlearn: 0.2543703\ttotal: 5.13s\tremaining: 29s\n",
            "400:\tlearn: 0.2329630\ttotal: 6.39s\tremaining: 25.5s\n",
            "500:\tlearn: 0.2179842\ttotal: 7.64s\tremaining: 22.8s\n",
            "600:\tlearn: 0.2057205\ttotal: 8.88s\tremaining: 20.7s\n",
            "700:\tlearn: 0.1951377\ttotal: 10.1s\tremaining: 18.8s\n",
            "800:\tlearn: 0.1866081\ttotal: 11.4s\tremaining: 17s\n",
            "900:\tlearn: 0.1784528\ttotal: 12.6s\tremaining: 15.4s\n",
            "1000:\tlearn: 0.1710893\ttotal: 13.9s\tremaining: 13.9s\n",
            "1100:\tlearn: 0.1640140\ttotal: 16.5s\tremaining: 13.5s\n",
            "1200:\tlearn: 0.1574815\ttotal: 18s\tremaining: 12s\n",
            "1300:\tlearn: 0.1513703\ttotal: 19.3s\tremaining: 10.4s\n",
            "1400:\tlearn: 0.1453327\ttotal: 20.6s\tremaining: 8.79s\n",
            "1500:\tlearn: 0.1395798\ttotal: 21.8s\tremaining: 7.26s\n",
            "1600:\tlearn: 0.1346524\ttotal: 23.1s\tremaining: 5.75s\n",
            "1700:\tlearn: 0.1296564\ttotal: 24.3s\tremaining: 4.28s\n",
            "1800:\tlearn: 0.1250879\ttotal: 25.6s\tremaining: 2.83s\n",
            "1900:\tlearn: 0.1211504\ttotal: 26.8s\tremaining: 1.4s\n",
            "1999:\tlearn: 0.1172623\ttotal: 29.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6869888\ttotal: 36.7ms\tremaining: 1m 13s\n",
            "100:\tlearn: 0.3671916\ttotal: 1.23s\tremaining: 23.2s\n",
            "200:\tlearn: 0.2866126\ttotal: 2.35s\tremaining: 21.1s\n",
            "300:\tlearn: 0.2531556\ttotal: 3.53s\tremaining: 19.9s\n",
            "400:\tlearn: 0.2315881\ttotal: 4.77s\tremaining: 19s\n",
            "500:\tlearn: 0.2156324\ttotal: 6.02s\tremaining: 18s\n",
            "600:\tlearn: 0.2039811\ttotal: 7.24s\tremaining: 16.9s\n",
            "700:\tlearn: 0.1935114\ttotal: 8.48s\tremaining: 15.7s\n",
            "800:\tlearn: 0.1849604\ttotal: 9.74s\tremaining: 14.6s\n",
            "900:\tlearn: 0.1765556\ttotal: 11.8s\tremaining: 14.3s\n",
            "1000:\tlearn: 0.1692553\ttotal: 13.8s\tremaining: 13.8s\n",
            "1100:\tlearn: 0.1622111\ttotal: 15.1s\tremaining: 12.3s\n",
            "1200:\tlearn: 0.1556426\ttotal: 16.3s\tremaining: 10.9s\n",
            "1300:\tlearn: 0.1497275\ttotal: 17.6s\tremaining: 9.45s\n",
            "1400:\tlearn: 0.1439305\ttotal: 18.9s\tremaining: 8.06s\n",
            "1500:\tlearn: 0.1379786\ttotal: 20.1s\tremaining: 6.7s\n",
            "1600:\tlearn: 0.1324985\ttotal: 21.4s\tremaining: 5.33s\n",
            "1700:\tlearn: 0.1274488\ttotal: 22.6s\tremaining: 3.98s\n",
            "1800:\tlearn: 0.1225240\ttotal: 24.6s\tremaining: 2.72s\n",
            "1900:\tlearn: 0.1180232\ttotal: 26.9s\tremaining: 1.4s\n",
            "1999:\tlearn: 0.1138484\ttotal: 28.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6867347\ttotal: 12.6ms\tremaining: 25.2s\n",
            "100:\tlearn: 0.3700745\ttotal: 1.08s\tremaining: 20.3s\n",
            "200:\tlearn: 0.2881306\ttotal: 2.17s\tremaining: 19.4s\n",
            "300:\tlearn: 0.2548475\ttotal: 3.42s\tremaining: 19.3s\n",
            "400:\tlearn: 0.2330775\ttotal: 4.67s\tremaining: 18.6s\n",
            "500:\tlearn: 0.2185586\ttotal: 5.92s\tremaining: 17.7s\n",
            "600:\tlearn: 0.2056791\ttotal: 7.2s\tremaining: 16.8s\n",
            "700:\tlearn: 0.1953076\ttotal: 8.74s\tremaining: 16.2s\n",
            "800:\tlearn: 0.1865724\ttotal: 11.4s\tremaining: 17s\n",
            "900:\tlearn: 0.1785620\ttotal: 12.6s\tremaining: 15.4s\n",
            "1000:\tlearn: 0.1711392\ttotal: 13.9s\tremaining: 13.9s\n",
            "1100:\tlearn: 0.1639162\ttotal: 15.1s\tremaining: 12.4s\n",
            "1200:\tlearn: 0.1571145\ttotal: 16.4s\tremaining: 10.9s\n",
            "1300:\tlearn: 0.1511049\ttotal: 17.6s\tremaining: 9.47s\n",
            "1400:\tlearn: 0.1452282\ttotal: 18.9s\tremaining: 8.07s\n",
            "1500:\tlearn: 0.1397573\ttotal: 20.1s\tremaining: 6.7s\n",
            "1600:\tlearn: 0.1344870\ttotal: 21.5s\tremaining: 5.36s\n",
            "1700:\tlearn: 0.1293630\ttotal: 24.1s\tremaining: 4.23s\n",
            "1800:\tlearn: 0.1247107\ttotal: 25.5s\tremaining: 2.82s\n",
            "1900:\tlearn: 0.1203493\ttotal: 26.8s\tremaining: 1.4s\n",
            "1999:\tlearn: 0.1163677\ttotal: 28s\tremaining: 0us\n",
            "0:\tlearn: 0.6870156\ttotal: 5.51ms\tremaining: 11s\n",
            "100:\tlearn: 0.3654410\ttotal: 556ms\tremaining: 10.5s\n",
            "200:\tlearn: 0.2934643\ttotal: 1.07s\tremaining: 9.6s\n",
            "300:\tlearn: 0.2588266\ttotal: 1.62s\tremaining: 9.16s\n",
            "400:\tlearn: 0.2404602\ttotal: 2.16s\tremaining: 8.62s\n",
            "500:\tlearn: 0.2273255\ttotal: 2.73s\tremaining: 8.15s\n",
            "600:\tlearn: 0.2172926\ttotal: 3.28s\tremaining: 7.63s\n",
            "700:\tlearn: 0.2090602\ttotal: 3.85s\tremaining: 7.13s\n",
            "800:\tlearn: 0.2024823\ttotal: 4.43s\tremaining: 6.63s\n",
            "900:\tlearn: 0.1969522\ttotal: 4.99s\tremaining: 6.08s\n",
            "1000:\tlearn: 0.1919841\ttotal: 5.56s\tremaining: 5.54s\n",
            "1100:\tlearn: 0.1871166\ttotal: 6.1s\tremaining: 4.98s\n",
            "1200:\tlearn: 0.1827457\ttotal: 7.18s\tremaining: 4.78s\n",
            "1300:\tlearn: 0.1787958\ttotal: 8.41s\tremaining: 4.52s\n",
            "1400:\tlearn: 0.1751221\ttotal: 9.61s\tremaining: 4.11s\n",
            "1500:\tlearn: 0.1714767\ttotal: 10.2s\tremaining: 3.38s\n",
            "1600:\tlearn: 0.1680032\ttotal: 10.7s\tremaining: 2.68s\n",
            "1700:\tlearn: 0.1641694\ttotal: 11.3s\tremaining: 1.98s\n",
            "1800:\tlearn: 0.1603756\ttotal: 11.8s\tremaining: 1.31s\n",
            "1900:\tlearn: 0.1566628\ttotal: 12.4s\tremaining: 645ms\n",
            "1999:\tlearn: 0.1530174\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6871021\ttotal: 5.55ms\tremaining: 11.1s\n",
            "100:\tlearn: 0.3694632\ttotal: 523ms\tremaining: 9.83s\n",
            "200:\tlearn: 0.2937717\ttotal: 1.06s\tremaining: 9.54s\n",
            "300:\tlearn: 0.2633884\ttotal: 1.59s\tremaining: 9.01s\n",
            "400:\tlearn: 0.2426596\ttotal: 2.17s\tremaining: 8.65s\n",
            "500:\tlearn: 0.2296921\ttotal: 2.72s\tremaining: 8.15s\n",
            "600:\tlearn: 0.2199937\ttotal: 3.28s\tremaining: 7.64s\n",
            "700:\tlearn: 0.2120870\ttotal: 3.85s\tremaining: 7.13s\n",
            "800:\tlearn: 0.2058247\ttotal: 4.39s\tremaining: 6.57s\n",
            "900:\tlearn: 0.2003369\ttotal: 4.95s\tremaining: 6.04s\n",
            "1000:\tlearn: 0.1957692\ttotal: 5.49s\tremaining: 5.48s\n",
            "1100:\tlearn: 0.1915238\ttotal: 6.06s\tremaining: 4.95s\n",
            "1200:\tlearn: 0.1869963\ttotal: 7.03s\tremaining: 4.68s\n",
            "1300:\tlearn: 0.1829565\ttotal: 8.37s\tremaining: 4.5s\n",
            "1400:\tlearn: 0.1791936\ttotal: 9.57s\tremaining: 4.09s\n",
            "1500:\tlearn: 0.1754072\ttotal: 10.2s\tremaining: 3.38s\n",
            "1600:\tlearn: 0.1717659\ttotal: 10.7s\tremaining: 2.67s\n",
            "1700:\tlearn: 0.1681709\ttotal: 11.3s\tremaining: 1.98s\n",
            "1800:\tlearn: 0.1643941\ttotal: 11.8s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.1608279\ttotal: 12.4s\tremaining: 645ms\n",
            "1999:\tlearn: 0.1574538\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6870476\ttotal: 5.81ms\tremaining: 11.6s\n",
            "100:\tlearn: 0.3662281\ttotal: 513ms\tremaining: 9.64s\n",
            "200:\tlearn: 0.2923539\ttotal: 1.05s\tremaining: 9.4s\n",
            "300:\tlearn: 0.2607107\ttotal: 1.6s\tremaining: 9.03s\n",
            "400:\tlearn: 0.2403798\ttotal: 2.18s\tremaining: 8.68s\n",
            "500:\tlearn: 0.2266474\ttotal: 2.73s\tremaining: 8.16s\n",
            "600:\tlearn: 0.2167856\ttotal: 3.3s\tremaining: 7.68s\n",
            "700:\tlearn: 0.2091569\ttotal: 3.85s\tremaining: 7.13s\n",
            "800:\tlearn: 0.2027237\ttotal: 4.42s\tremaining: 6.61s\n",
            "900:\tlearn: 0.1975293\ttotal: 4.98s\tremaining: 6.08s\n",
            "1000:\tlearn: 0.1926506\ttotal: 5.54s\tremaining: 5.53s\n",
            "1100:\tlearn: 0.1879267\ttotal: 6.1s\tremaining: 4.98s\n",
            "1200:\tlearn: 0.1836367\ttotal: 7.17s\tremaining: 4.77s\n",
            "1300:\tlearn: 0.1798678\ttotal: 8.43s\tremaining: 4.53s\n",
            "1400:\tlearn: 0.1762370\ttotal: 9.61s\tremaining: 4.11s\n",
            "1500:\tlearn: 0.1725823\ttotal: 10.2s\tremaining: 3.38s\n",
            "1600:\tlearn: 0.1691544\ttotal: 10.7s\tremaining: 2.67s\n",
            "1700:\tlearn: 0.1657301\ttotal: 11.3s\tremaining: 1.98s\n",
            "1800:\tlearn: 0.1622734\ttotal: 11.8s\tremaining: 1.31s\n",
            "1900:\tlearn: 0.1591254\ttotal: 12.4s\tremaining: 646ms\n",
            "1999:\tlearn: 0.1558129\ttotal: 12.9s\tremaining: 0us\n",
            "0:\tlearn: 0.6869545\ttotal: 5.43ms\tremaining: 10.9s\n",
            "100:\tlearn: 0.3610533\ttotal: 535ms\tremaining: 10.1s\n",
            "200:\tlearn: 0.2898377\ttotal: 1.08s\tremaining: 9.64s\n",
            "300:\tlearn: 0.2597321\ttotal: 1.6s\tremaining: 9.03s\n",
            "400:\tlearn: 0.2392528\ttotal: 2.15s\tremaining: 8.58s\n",
            "500:\tlearn: 0.2256803\ttotal: 2.69s\tremaining: 8.06s\n",
            "600:\tlearn: 0.2154694\ttotal: 3.25s\tremaining: 7.57s\n",
            "700:\tlearn: 0.2078135\ttotal: 3.79s\tremaining: 7.03s\n",
            "800:\tlearn: 0.2017403\ttotal: 4.34s\tremaining: 6.5s\n",
            "900:\tlearn: 0.1961259\ttotal: 4.88s\tremaining: 5.95s\n",
            "1000:\tlearn: 0.1914248\ttotal: 5.43s\tremaining: 5.42s\n",
            "1100:\tlearn: 0.1866353\ttotal: 5.96s\tremaining: 4.87s\n",
            "1200:\tlearn: 0.1824923\ttotal: 6.96s\tremaining: 4.63s\n",
            "1300:\tlearn: 0.1785353\ttotal: 8.15s\tremaining: 4.38s\n",
            "1400:\tlearn: 0.1747158\ttotal: 9.37s\tremaining: 4.01s\n",
            "1500:\tlearn: 0.1709100\ttotal: 9.92s\tremaining: 3.3s\n",
            "1600:\tlearn: 0.1674515\ttotal: 10.5s\tremaining: 2.61s\n",
            "1700:\tlearn: 0.1637324\ttotal: 11s\tremaining: 1.94s\n",
            "1800:\tlearn: 0.1598984\ttotal: 11.6s\tremaining: 1.28s\n",
            "1900:\tlearn: 0.1562596\ttotal: 12.1s\tremaining: 633ms\n",
            "1999:\tlearn: 0.1526512\ttotal: 12.7s\tremaining: 0us\n",
            "0:\tlearn: 0.6870228\ttotal: 5.67ms\tremaining: 11.3s\n",
            "100:\tlearn: 0.3644409\ttotal: 540ms\tremaining: 10.2s\n",
            "200:\tlearn: 0.2933819\ttotal: 1.06s\tremaining: 9.47s\n",
            "300:\tlearn: 0.2608964\ttotal: 1.61s\tremaining: 9.09s\n",
            "400:\tlearn: 0.2411299\ttotal: 2.15s\tremaining: 8.59s\n",
            "500:\tlearn: 0.2283285\ttotal: 2.71s\tremaining: 8.12s\n",
            "600:\tlearn: 0.2189259\ttotal: 3.25s\tremaining: 7.57s\n",
            "700:\tlearn: 0.2106883\ttotal: 3.81s\tremaining: 7.06s\n",
            "800:\tlearn: 0.2035238\ttotal: 4.36s\tremaining: 6.53s\n",
            "900:\tlearn: 0.1977561\ttotal: 4.93s\tremaining: 6.01s\n",
            "1000:\tlearn: 0.1926428\ttotal: 5.46s\tremaining: 5.45s\n",
            "1100:\tlearn: 0.1877107\ttotal: 6.02s\tremaining: 4.92s\n",
            "1200:\tlearn: 0.1833342\ttotal: 6.85s\tremaining: 4.56s\n",
            "1300:\tlearn: 0.1794201\ttotal: 8.04s\tremaining: 4.32s\n",
            "1400:\tlearn: 0.1754434\ttotal: 9.4s\tremaining: 4.02s\n",
            "1500:\tlearn: 0.1717216\ttotal: 10s\tremaining: 3.33s\n",
            "1600:\tlearn: 0.1683217\ttotal: 10.6s\tremaining: 2.63s\n",
            "1700:\tlearn: 0.1647695\ttotal: 11.1s\tremaining: 1.96s\n",
            "1800:\tlearn: 0.1612948\ttotal: 11.7s\tremaining: 1.3s\n",
            "1900:\tlearn: 0.1576870\ttotal: 12.3s\tremaining: 640ms\n",
            "1999:\tlearn: 0.1541539\ttotal: 12.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6861330\ttotal: 5.37ms\tremaining: 5.36s\n",
            "100:\tlearn: 0.3575534\ttotal: 495ms\tremaining: 4.4s\n",
            "200:\tlearn: 0.2887672\ttotal: 1.04s\tremaining: 4.12s\n",
            "300:\tlearn: 0.2563568\ttotal: 1.57s\tremaining: 3.65s\n",
            "400:\tlearn: 0.2352991\ttotal: 2.13s\tremaining: 3.19s\n",
            "500:\tlearn: 0.2223787\ttotal: 2.68s\tremaining: 2.67s\n",
            "600:\tlearn: 0.2117843\ttotal: 3.24s\tremaining: 2.15s\n",
            "700:\tlearn: 0.2042387\ttotal: 3.81s\tremaining: 1.62s\n",
            "800:\tlearn: 0.1975284\ttotal: 4.36s\tremaining: 1.08s\n",
            "900:\tlearn: 0.1919883\ttotal: 4.92s\tremaining: 541ms\n",
            "999:\tlearn: 0.1869407\ttotal: 5.46s\tremaining: 0us\n",
            "0:\tlearn: 0.6862217\ttotal: 5.41ms\tremaining: 5.41s\n",
            "100:\tlearn: 0.3612912\ttotal: 521ms\tremaining: 4.63s\n",
            "200:\tlearn: 0.2885955\ttotal: 1.22s\tremaining: 4.86s\n",
            "300:\tlearn: 0.2569369\ttotal: 2.53s\tremaining: 5.88s\n",
            "400:\tlearn: 0.2374585\ttotal: 3.95s\tremaining: 5.9s\n",
            "500:\tlearn: 0.2253277\ttotal: 4.5s\tremaining: 4.48s\n",
            "600:\tlearn: 0.2159255\ttotal: 5.04s\tremaining: 3.35s\n",
            "700:\tlearn: 0.2082953\ttotal: 5.6s\tremaining: 2.39s\n",
            "800:\tlearn: 0.2021448\ttotal: 6.15s\tremaining: 1.53s\n",
            "900:\tlearn: 0.1966724\ttotal: 6.72s\tremaining: 739ms\n",
            "999:\tlearn: 0.1915127\ttotal: 7.26s\tremaining: 0us\n",
            "0:\tlearn: 0.6861719\ttotal: 5.33ms\tremaining: 5.32s\n",
            "100:\tlearn: 0.3591330\ttotal: 513ms\tremaining: 4.57s\n",
            "200:\tlearn: 0.2865091\ttotal: 1.05s\tremaining: 4.18s\n",
            "300:\tlearn: 0.2577224\ttotal: 1.59s\tremaining: 3.7s\n",
            "400:\tlearn: 0.2367584\ttotal: 2.15s\tremaining: 3.21s\n",
            "500:\tlearn: 0.2238141\ttotal: 2.7s\tremaining: 2.69s\n",
            "600:\tlearn: 0.2140091\ttotal: 3.26s\tremaining: 2.17s\n",
            "700:\tlearn: 0.2065144\ttotal: 3.81s\tremaining: 1.63s\n",
            "800:\tlearn: 0.1997124\ttotal: 4.39s\tremaining: 1.09s\n",
            "900:\tlearn: 0.1944664\ttotal: 4.93s\tremaining: 542ms\n",
            "999:\tlearn: 0.1892082\ttotal: 5.5s\tremaining: 0us\n",
            "0:\tlearn: 0.6860623\ttotal: 5.31ms\tremaining: 5.31s\n",
            "100:\tlearn: 0.3544743\ttotal: 490ms\tremaining: 4.37s\n",
            "200:\tlearn: 0.2857916\ttotal: 1.22s\tremaining: 4.84s\n",
            "300:\tlearn: 0.2537933\ttotal: 2.52s\tremaining: 5.86s\n",
            "400:\tlearn: 0.2335729\ttotal: 3.84s\tremaining: 5.73s\n",
            "500:\tlearn: 0.2201509\ttotal: 4.53s\tremaining: 4.51s\n",
            "600:\tlearn: 0.2105346\ttotal: 5.09s\tremaining: 3.38s\n",
            "700:\tlearn: 0.2030608\ttotal: 5.64s\tremaining: 2.4s\n",
            "800:\tlearn: 0.1967244\ttotal: 6.2s\tremaining: 1.54s\n",
            "900:\tlearn: 0.1909849\ttotal: 6.76s\tremaining: 743ms\n",
            "999:\tlearn: 0.1862473\ttotal: 7.3s\tremaining: 0us\n",
            "0:\tlearn: 0.6858460\ttotal: 5.41ms\tremaining: 5.4s\n",
            "100:\tlearn: 0.3583354\ttotal: 524ms\tremaining: 4.66s\n",
            "200:\tlearn: 0.2894932\ttotal: 1.04s\tremaining: 4.15s\n",
            "300:\tlearn: 0.2576911\ttotal: 1.6s\tremaining: 3.71s\n",
            "400:\tlearn: 0.2367927\ttotal: 2.14s\tremaining: 3.19s\n",
            "500:\tlearn: 0.2241297\ttotal: 2.74s\tremaining: 2.73s\n",
            "600:\tlearn: 0.2127726\ttotal: 3.29s\tremaining: 2.19s\n",
            "700:\tlearn: 0.2045384\ttotal: 3.85s\tremaining: 1.64s\n",
            "800:\tlearn: 0.1981104\ttotal: 4.43s\tremaining: 1.1s\n",
            "900:\tlearn: 0.1922727\ttotal: 4.99s\tremaining: 548ms\n",
            "999:\tlearn: 0.1876948\ttotal: 5.54s\tremaining: 0us\n",
            "0:\tlearn: 0.6818150\ttotal: 18.1ms\tremaining: 18.1s\n",
            "100:\tlearn: 0.2941215\ttotal: 1.32s\tremaining: 11.8s\n",
            "200:\tlearn: 0.2345867\ttotal: 3.98s\tremaining: 15.8s\n",
            "300:\tlearn: 0.2083761\ttotal: 5.26s\tremaining: 12.2s\n",
            "400:\tlearn: 0.1896073\ttotal: 6.51s\tremaining: 9.73s\n",
            "500:\tlearn: 0.1748030\ttotal: 7.77s\tremaining: 7.74s\n",
            "600:\tlearn: 0.1613723\ttotal: 9.01s\tremaining: 5.98s\n",
            "700:\tlearn: 0.1498900\ttotal: 10.3s\tremaining: 4.38s\n",
            "800:\tlearn: 0.1396249\ttotal: 11.5s\tremaining: 2.87s\n",
            "900:\tlearn: 0.1303849\ttotal: 12.8s\tremaining: 1.41s\n",
            "999:\tlearn: 0.1223597\ttotal: 14.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6819469\ttotal: 24.8ms\tremaining: 24.8s\n",
            "100:\tlearn: 0.3006940\ttotal: 2.35s\tremaining: 20.9s\n",
            "200:\tlearn: 0.2410250\ttotal: 3.77s\tremaining: 15s\n",
            "300:\tlearn: 0.2137906\ttotal: 5.04s\tremaining: 11.7s\n",
            "400:\tlearn: 0.1956296\ttotal: 6.3s\tremaining: 9.4s\n",
            "500:\tlearn: 0.1813736\ttotal: 7.56s\tremaining: 7.53s\n",
            "600:\tlearn: 0.1676198\ttotal: 8.82s\tremaining: 5.86s\n",
            "700:\tlearn: 0.1563287\ttotal: 10.1s\tremaining: 4.3s\n",
            "800:\tlearn: 0.1458310\ttotal: 11.4s\tremaining: 2.83s\n",
            "900:\tlearn: 0.1362602\ttotal: 12.7s\tremaining: 1.4s\n",
            "999:\tlearn: 0.1275449\ttotal: 15.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6812135\ttotal: 28.2ms\tremaining: 28.2s\n",
            "100:\tlearn: 0.2961465\ttotal: 1.14s\tremaining: 10.2s\n",
            "200:\tlearn: 0.2362328\ttotal: 2.39s\tremaining: 9.49s\n",
            "300:\tlearn: 0.2113432\ttotal: 3.65s\tremaining: 8.47s\n",
            "400:\tlearn: 0.1927883\ttotal: 4.92s\tremaining: 7.36s\n",
            "500:\tlearn: 0.1779320\ttotal: 6.21s\tremaining: 6.19s\n",
            "600:\tlearn: 0.1649111\ttotal: 7.46s\tremaining: 4.96s\n",
            "700:\tlearn: 0.1534316\ttotal: 8.71s\tremaining: 3.71s\n",
            "800:\tlearn: 0.1434525\ttotal: 9.99s\tremaining: 2.48s\n",
            "900:\tlearn: 0.1340899\ttotal: 12.6s\tremaining: 1.38s\n",
            "999:\tlearn: 0.1256820\ttotal: 14.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6817004\ttotal: 12.4ms\tremaining: 12.4s\n",
            "100:\tlearn: 0.2923181\ttotal: 1.13s\tremaining: 10.1s\n",
            "200:\tlearn: 0.2336439\ttotal: 2.35s\tremaining: 9.34s\n",
            "300:\tlearn: 0.2082831\ttotal: 3.6s\tremaining: 8.36s\n",
            "400:\tlearn: 0.1908519\ttotal: 4.85s\tremaining: 7.24s\n",
            "500:\tlearn: 0.1751942\ttotal: 6.12s\tremaining: 6.09s\n",
            "600:\tlearn: 0.1617002\ttotal: 7.39s\tremaining: 4.9s\n",
            "700:\tlearn: 0.1503896\ttotal: 8.65s\tremaining: 3.69s\n",
            "800:\tlearn: 0.1403139\ttotal: 11s\tremaining: 2.73s\n",
            "900:\tlearn: 0.1306170\ttotal: 12.8s\tremaining: 1.41s\n",
            "999:\tlearn: 0.1222140\ttotal: 14.1s\tremaining: 0us\n",
            "0:\tlearn: 0.6811784\ttotal: 14.8ms\tremaining: 14.8s\n",
            "100:\tlearn: 0.2922185\ttotal: 1.09s\tremaining: 9.74s\n",
            "200:\tlearn: 0.2374522\ttotal: 2.3s\tremaining: 9.16s\n",
            "300:\tlearn: 0.2110109\ttotal: 3.55s\tremaining: 8.25s\n",
            "400:\tlearn: 0.1922852\ttotal: 4.79s\tremaining: 7.16s\n",
            "500:\tlearn: 0.1768471\ttotal: 6.03s\tremaining: 6.01s\n",
            "600:\tlearn: 0.1636135\ttotal: 7.32s\tremaining: 4.86s\n",
            "700:\tlearn: 0.1516421\ttotal: 9.42s\tremaining: 4.02s\n",
            "800:\tlearn: 0.1414424\ttotal: 11.5s\tremaining: 2.85s\n",
            "900:\tlearn: 0.1320537\ttotal: 12.7s\tremaining: 1.4s\n",
            "999:\tlearn: 0.1239082\ttotal: 14s\tremaining: 0us\n",
            "0:\tlearn: 0.6584063\ttotal: 3.79ms\tremaining: 3.79s\n",
            "100:\tlearn: 0.2082680\ttotal: 3.41s\tremaining: 30.4s\n",
            "200:\tlearn: 0.1474006\ttotal: 7.5s\tremaining: 29.8s\n",
            "300:\tlearn: 0.1072592\ttotal: 12.4s\tremaining: 28.9s\n",
            "400:\tlearn: 0.0811582\ttotal: 16.3s\tremaining: 24.3s\n",
            "500:\tlearn: 0.0628792\ttotal: 20.1s\tremaining: 20s\n",
            "600:\tlearn: 0.0515411\ttotal: 25.1s\tremaining: 16.7s\n",
            "700:\tlearn: 0.0421521\ttotal: 28.9s\tremaining: 12.3s\n",
            "800:\tlearn: 0.0351791\ttotal: 32.9s\tremaining: 8.17s\n",
            "900:\tlearn: 0.0292608\ttotal: 38.1s\tremaining: 4.18s\n",
            "999:\tlearn: 0.0250784\ttotal: 41.8s\tremaining: 0us\n",
            "0:\tlearn: 0.6584063\ttotal: 3.85ms\tremaining: 3.85s\n",
            "100:\tlearn: 0.2082680\ttotal: 4.89s\tremaining: 43.5s\n",
            "200:\tlearn: 0.1474006\ttotal: 8.88s\tremaining: 35.3s\n",
            "300:\tlearn: 0.1072592\ttotal: 12.8s\tremaining: 29.8s\n",
            "400:\tlearn: 0.0811582\ttotal: 18.3s\tremaining: 27.3s\n",
            "500:\tlearn: 0.0628792\ttotal: 22.2s\tremaining: 22.2s\n",
            "600:\tlearn: 0.0515411\ttotal: 26.3s\tremaining: 17.5s\n",
            "700:\tlearn: 0.0421521\ttotal: 31.8s\tremaining: 13.6s\n",
            "800:\tlearn: 0.0351791\ttotal: 35.8s\tremaining: 8.89s\n",
            "900:\tlearn: 0.0292608\ttotal: 39.8s\tremaining: 4.38s\n",
            "999:\tlearn: 0.0250784\ttotal: 45.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6592837\ttotal: 6.83ms\tremaining: 6.82s\n",
            "100:\tlearn: 0.1541093\ttotal: 4.91s\tremaining: 43.7s\n",
            "200:\tlearn: 0.0999807\ttotal: 8.65s\tremaining: 34.4s\n",
            "300:\tlearn: 0.0684183\ttotal: 12.5s\tremaining: 29s\n",
            "400:\tlearn: 0.0491210\ttotal: 17.8s\tremaining: 26.5s\n",
            "500:\tlearn: 0.0367037\ttotal: 21.7s\tremaining: 21.7s\n",
            "600:\tlearn: 0.0283833\ttotal: 25.6s\tremaining: 17s\n",
            "700:\tlearn: 0.0222339\ttotal: 31.2s\tremaining: 13.3s\n",
            "800:\tlearn: 0.0180875\ttotal: 35.2s\tremaining: 8.74s\n",
            "900:\tlearn: 0.0149153\ttotal: 39.2s\tremaining: 4.31s\n",
            "999:\tlearn: 0.0124888\ttotal: 44.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6590616\ttotal: 2.88ms\tremaining: 2.88s\n",
            "100:\tlearn: 0.2137231\ttotal: 3.33s\tremaining: 29.7s\n",
            "200:\tlearn: 0.1497320\ttotal: 7.09s\tremaining: 28.2s\n",
            "300:\tlearn: 0.1054376\ttotal: 12.2s\tremaining: 28.4s\n",
            "400:\tlearn: 0.0778441\ttotal: 16s\tremaining: 23.9s\n",
            "500:\tlearn: 0.0593587\ttotal: 19.7s\tremaining: 19.6s\n",
            "600:\tlearn: 0.0478640\ttotal: 24.9s\tremaining: 16.5s\n",
            "700:\tlearn: 0.0384044\ttotal: 28.7s\tremaining: 12.2s\n",
            "800:\tlearn: 0.0319618\ttotal: 32.4s\tremaining: 8.06s\n",
            "900:\tlearn: 0.0271017\ttotal: 37.6s\tremaining: 4.14s\n",
            "999:\tlearn: 0.0229608\ttotal: 41.4s\tremaining: 0us\n",
            "0:\tlearn: 0.6590005\ttotal: 2.98ms\tremaining: 2.98s\n",
            "100:\tlearn: 0.2203240\ttotal: 3.38s\tremaining: 30.1s\n",
            "200:\tlearn: 0.1531368\ttotal: 8.38s\tremaining: 33.3s\n",
            "300:\tlearn: 0.1089968\ttotal: 12.2s\tremaining: 28.4s\n",
            "400:\tlearn: 0.0804868\ttotal: 15.9s\tremaining: 23.8s\n",
            "500:\tlearn: 0.0631099\ttotal: 20.4s\tremaining: 20.3s\n",
            "600:\tlearn: 0.0508970\ttotal: 24.7s\tremaining: 16.4s\n",
            "700:\tlearn: 0.0414975\ttotal: 28.4s\tremaining: 12.1s\n",
            "800:\tlearn: 0.0342631\ttotal: 32.4s\tremaining: 8.06s\n",
            "900:\tlearn: 0.0287294\ttotal: 37.3s\tremaining: 4.1s\n",
            "999:\tlearn: 0.0245127\ttotal: 41s\tremaining: 0us\n",
            "0:\tlearn: 0.6587525\ttotal: 2.83ms\tremaining: 2.83s\n",
            "100:\tlearn: 0.2167030\ttotal: 3.23s\tremaining: 28.8s\n",
            "200:\tlearn: 0.1512523\ttotal: 8.35s\tremaining: 33.2s\n",
            "300:\tlearn: 0.1080047\ttotal: 12.1s\tremaining: 28.2s\n",
            "400:\tlearn: 0.0811677\ttotal: 15.8s\tremaining: 23.6s\n",
            "500:\tlearn: 0.0619834\ttotal: 20.9s\tremaining: 20.9s\n",
            "600:\tlearn: 0.0500057\ttotal: 24.7s\tremaining: 16.4s\n",
            "700:\tlearn: 0.0411428\ttotal: 28.4s\tremaining: 12.1s\n",
            "800:\tlearn: 0.0339802\ttotal: 33.7s\tremaining: 8.36s\n",
            "900:\tlearn: 0.0283463\ttotal: 37.4s\tremaining: 4.11s\n",
            "999:\tlearn: 0.0242559\ttotal: 41.2s\tremaining: 0us\n",
            "0:\tlearn: 0.6590461\ttotal: 2.92ms\tremaining: 2.92s\n",
            "100:\tlearn: 0.2217481\ttotal: 4.17s\tremaining: 37.1s\n",
            "200:\tlearn: 0.1536310\ttotal: 8.37s\tremaining: 33.3s\n",
            "300:\tlearn: 0.1093686\ttotal: 12.2s\tremaining: 28.2s\n",
            "400:\tlearn: 0.0820734\ttotal: 16.4s\tremaining: 24.4s\n",
            "500:\tlearn: 0.0633632\ttotal: 21s\tremaining: 20.9s\n",
            "600:\tlearn: 0.0514841\ttotal: 24.8s\tremaining: 16.4s\n",
            "700:\tlearn: 0.0423600\ttotal: 28.7s\tremaining: 12.2s\n",
            "800:\tlearn: 0.0352449\ttotal: 33.7s\tremaining: 8.38s\n",
            "900:\tlearn: 0.0298789\ttotal: 37.4s\tremaining: 4.11s\n",
            "999:\tlearn: 0.0257141\ttotal: 41.2s\tremaining: 0us\n",
            "Validation Set Metrics:\n",
            "Accuracy: 0.812457221081451\n",
            "Precision: 0.8588235294117647\n",
            "Recall: 0.9209419680403701\n",
            "F1-score: 0.8887987012987013\n",
            "AUC-ROC: 0.6295886310790086\n",
            "Submission file 'Enhance_Cat-svm-rf_ensemble_submission.csv' has been created.\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split, RandomizedSearchCV, StratifiedKFold\n",
        "from sklearn.preprocessing import StandardScaler, OneHotEncoder, MinMaxScaler\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from imblearn.over_sampling import SMOTE,ADASYN\n",
        "from catboost import CatBoostClassifier\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score\n",
        "from sklearn.ensemble import VotingClassifier, StackingClassifier\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "\n",
        "# Load the datasets\n",
        "train_df = pd.read_csv('/content/Train Dataset.csv')\n",
        "test_df = pd.read_csv('/content/Test Dataset.csv')\n",
        "\n",
        "# Separate features and target in the training data\n",
        "X = train_df.drop('target', axis=1)\n",
        "y = train_df['target']\n",
        "\n",
        "# Split the data into training and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)\n",
        "\n",
        "# Identify numeric and categorical columns\n",
        "numeric_features = X.select_dtypes(include=[np.number]).columns\n",
        "categorical_features = X.select_dtypes(exclude=[np.number]).columns\n",
        "\n",
        "\n",
        "# Create preprocessing pipelines\n",
        "numeric_transformer = Pipeline(steps=[\n",
        "    ('scaler', StandardScaler())\n",
        "])\n",
        "\n",
        "categorical_transformer = Pipeline(steps=[\n",
        "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
        "\n",
        "])\n",
        "\n",
        "# Combine preprocessing steps\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', numeric_transformer, numeric_features),\n",
        "        ('cat', categorical_transformer, categorical_features)\n",
        "\n",
        "    ]\n",
        ")\n",
        "\n",
        "# Create the full preprocessing\n",
        "full_pipeline = Pipeline([\n",
        "    ('preprocessor', preprocessor)\n",
        "])\n",
        "\n",
        "# Fit the pipeline and transform the data\n",
        "X_train_processed = full_pipeline.fit_transform(X_train, y_train)\n",
        "X_val_processed = full_pipeline.transform(X_val)\n",
        "\n",
        "# Change the column names in test_df to match the training data\n",
        "test_df.rename(columns={'id': 'Id', 'age': 'Age', 'sex': 'Sex'}, inplace=True)\n",
        "\n",
        "# Transform the test data\n",
        "X_test_processed = full_pipeline.transform(test_df)\n",
        "\n",
        "# Apply SMOTE or ADASYN to handle class imbalance\n",
        "resampling_strategy = 'SMOTE'  # or 'ADASYN'\n",
        "if resampling_strategy == 'SMOTE':\n",
        "    smote = SMOTE(random_state=42)\n",
        "elif resampling_strategy == 'ADASYN':\n",
        "    adasyn = ADASYN(random_state=42)\n",
        "\n",
        "if resampling_strategy == 'SMOTE':\n",
        "    X_train_processed, y_train_balanced = smote.fit_resample(X_train_processed, y_train)\n",
        "elif resampling_strategy == 'ADASYN':\n",
        "    X_train_processed, y_train_balanced = adasyn.fit_resample(X_train_processed, y_train)\n",
        "\n",
        "# Define individual models\n",
        "catboost_model = CatBoostClassifier(verbose=100)\n",
        "svm_model = SVC(probability=True)\n",
        "rf_model = RandomForestClassifier(random_state=42)\n",
        "\n",
        "# Define hyperparameter grids for each model\n",
        "catboost_param_grid = {\n",
        "    'iterations': [1000, 2000, 3000],\n",
        "    'learning_rate': [0.005, 0.01, 0.02],\n",
        "    'depth': [6, 8, 10],\n",
        "    'l2_leaf_reg': [2, 4, 6],\n",
        "    'random_seed': [42, 123, 987]\n",
        "}\n",
        "\n",
        "svm_param_grid = {\n",
        "    'C': [0.1, 1, 10],\n",
        "    'kernel': ['linear', 'rbf', 'poly'],\n",
        "    'gamma': ['scale', 'auto'],\n",
        "    'class_weight': ['balanced']\n",
        "}\n",
        "\n",
        "rf_param_grid = {\n",
        "    'n_estimators': [200, 300, 400],\n",
        "    'max_depth': [None, 10, 15],\n",
        "    'min_samples_split': [2, 5, 10],\n",
        "    'min_samples_leaf': [1, 2, 4],\n",
        "    'class_weight': ['balanced']\n",
        "}\n",
        "\n",
        "# Perform randomized search hyperparameter tuning for each model\n",
        "catboost_random_search = RandomizedSearchCV(catboost_model, param_distributions=catboost_param_grid, n_iter=18, cv=StratifiedKFold(n_splits=5, shuffle=True, random_state=42), scoring='f1_macro')\n",
        "svm_random_search = RandomizedSearchCV(svm_model, param_distributions=svm_param_grid, n_iter=18, cv=StratifiedKFold(n_splits=5, shuffle=True, random_state=42), scoring='f1_macro')\n",
        "rf_random_search = RandomizedSearchCV(rf_model, param_distributions=rf_param_grid, n_iter=18, cv=StratifiedKFold(n_splits=5, shuffle=True, random_state=42), scoring='f1_macro')\n",
        "\n",
        "# Fit the random searches to the training data\n",
        "catboost_random_search.fit(X_train_processed, y_train_balanced)\n",
        "svm_random_search.fit(X_train_processed, y_train_balanced)\n",
        "rf_random_search.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Get the best models from each random search\n",
        "best_catboost_model = catboost_random_search.best_estimator_\n",
        "best_svm_model = svm_random_search.best_estimator_\n",
        "best_rf_model = rf_random_search.best_estimator_\n",
        "\n",
        "# Create the ensemble model using StackingClassifier\n",
        "estimators = [('catboost', best_catboost_model), ('svm', best_svm_model), ('rf', best_rf_model)]\n",
        "final_estimator = LogisticRegression()\n",
        "ensemble_model = StackingClassifier(estimators=estimators, final_estimator=final_estimator, cv=5)\n",
        "\n",
        "# Fit the ensemble model to the training data\n",
        "ensemble_model.fit(X_train_processed, y_train_balanced)\n",
        "\n",
        "# Evaluate the ensemble model on the validation set\n",
        "y_val_pred = ensemble_model.predict(X_val_processed)\n",
        "\n",
        "# Calculate accuracy and other metrics\n",
        "accuracy = accuracy_score(y_val, y_val_pred)\n",
        "precision = precision_score(y_val, y_val_pred)\n",
        "recall = recall_score(y_val, y_val_pred)\n",
        "f1_score = f1_score(y_val, y_val_pred)\n",
        "\n",
        "auc_roc = roc_auc_score(y_val, y_val_pred)\n",
        "\n",
        "print(\"Validation Set Metrics:\")\n",
        "print(\"Accuracy:\", accuracy)\n",
        "print(\"Precision:\", precision)\n",
        "print(\"Recall:\", recall)\n",
        "print(\"F1-score:\", f1_score)\n",
        "print(\"AUC-ROC:\", auc_roc)\n",
        "\n",
        "# Make predictions on the test set\n",
        "test_predictions = ensemble_model.predict(X_test_processed)\n",
        "\n",
        "\n",
        "# Create a submission DataFrame\n",
        "submission = pd.DataFrame({\n",
        "    'ID': test_df['Id'],\n",
        "    'target': test_predictions\n",
        "})\n",
        "\n",
        "# Save the submission file\n",
        "submission.to_csv('Enhance_Cat-svm-rf_ensemble_submission.csv', index=False)\n",
        "print(\"Submission file 'Enhance_Cat-svm-rf_ensemble_submission.csv' has been created.\")"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}