{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "md43cb4daa",
   "metadata": {},
   "source": [
    "# BitBullet:Lessons: Multi-Class Classification\n",
    "\n",
    "Agricultural research, quality control, and supply chain optimisation all depend on accurately identifying biological varieties from measurable features. In this tutorial we build a **dry bean variety classifier** — a model that predicts which of seven bean species a sample belongs to based on morphological measurements extracted from digital images.\n",
    "\n",
    "This is a real-world problem from the UCI Machine Learning Repository: a computer vision pipeline images each bean, extracts 16 shape and texture measurements, and your model determines the variety. With seven classes ranging from the dominant DERMASON (26%) to the rare BOMBAY (4%), this dataset tests everything BitBullet's multiclass handling can do.\n",
    "\n",
    "**Dataset**: `Dry_Bean_Dataset.xlsx` — 13,611 bean samples, 16 morphological features.  \n",
    "**Source**: [UCI ML Repository — Dry Bean Dataset](https://archive.uci.edu/dataset/602/dry+bean+dataset)  \n",
    "**Task**: Predict bean variety (`DERMASON` / `SIRA` / `SEKER` / `HOROZ` / `CALI` / `BARBUNYA` / `BOMBAY`) from shape features.\n",
    "\n",
    "---\n",
    "\n",
    "### What You Will Build\n",
    "\n",
    "| Step | Component | What BitBullet Handles For You |\n",
    "|------|-----------|--------------------------------|\n",
    "| 1 | Feature profiling | `generate_feature_stats` — full audit in one call |\n",
    "| 2 | Class imbalance analysis | Visual distribution across all seven varieties |\n",
    "| 3 | Hold-out test split | Stratified 80/20 — all seven class proportions preserved |\n",
    "| 4 | Target encoding | `LabelEncoder` — integer codes for the trainer, decoded labels for humans |\n",
    "| 5 | Transform pipeline | Fit on train only, serialise, reload, apply to test — zero leakage |\n",
    "| 6 | Hyperparameter search | Optuna TPE + Stratified K-Fold + early stopping, softmax objective |\n",
    "| 7 | Confusion matrix | Which varieties are confused with which — the full 7×7 picture |\n",
    "| 8 | Per-class metrics | Precision, recall, F1 for every variety separately |\n",
    "| 9 | Per-class ROC AUC | One-vs-rest decomposition |\n",
    "| 10 | SHAP explainability | Global feature attributions — which shape measurements matter most |\n",
    "| 11 | Model serialisation | `ModelSerializer` + `ModelMetadata` — provenance, class label mapping, schema, preprocessing, and structured metric payloads |"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f2e1c4d",
   "metadata": {},
   "source": [
    "> **Configure the complete classification lifecycle without writing the orchestration.**\n",
    "> [BitBullet Platform](https://bitbullet.co.uk/platform/classification) centralises your data, managed compute, storage, configurations, and results in one guided environment. Use the AI assistant to explore authorised data and prepare a draft, compare candidate models and diagnostic evidence, then export the fitted model, preprocessing, metadata, and generated inference code. This lesson shows the same workflow directly with the BitBullet SDK."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mdeb536e48",
   "metadata": {},
   "source": [
    "## 1. Environment & Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "cd43e940fa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bitbullet  : vdev\n",
      "All imports successful.\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "import os\n",
    "import warnings\n",
    "from datetime import datetime\n",
    "\n",
    "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 LabelEncoder, label_binarize\n",
    "from sklearn.metrics import (\n",
    "    classification_report,\n",
    "    confusion_matrix,\n",
    "    roc_auc_score,\n",
    "    accuracy_score,\n",
    "    balanced_accuracy_score,\n",
    "    f1_score,\n",
    ")\n",
    "\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# When running this notebook from bitbullet/lessons in a local clone, prefer the local SDK source.\n",
    "sdk_root = os.path.abspath(\"..\")\n",
    "if sdk_root not in sys.path:\n",
    "    sys.path.insert(0, sdk_root)\n",
    "\n",
    "import bitbullet\n",
    "print(f\"bitbullet  : v{getattr(bitbullet, '__version__', 'dev')}\")\n",
    "\n",
    "from bitbullet.transform import TransformPipeline\n",
    "from bitbullet.transform.eda import generate_feature_stats\n",
    "from bitbullet.train import TrainConfig, OptunaTrainer, TrainingReportGenerator\n",
    "from bitbullet.model import ModelMetadata, ModelSerializer\n",
    "\n",
    "print(\"All imports successful.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md1c1eefe4",
   "metadata": {},
   "source": [
    "## 2. Load the Dataset\n",
    "\n",
    "Each row represents a single dry bean sample imaged under controlled conditions. The 16 features are morphological measurements — area, perimeter, axis lengths, eccentricity, convex hull statistics, and shape factors — computed automatically from the digital image. The target `Class` is the true bean variety assigned by agricultural experts.\n",
    "\n",
    "**Dataset:** Dry Bean Dataset  \n",
    "**Source:** [UCI ML Repository — Dry Bean Dataset](https://archive.uci.edu/dataset/602/dry+bean+dataset)  \n",
    "**File:** `Dry_Bean_Dataset.xlsx`\n",
    "\n",
    "> **Before running this cell:** download `Dry_Bean_Dataset.xlsx` from the link above and\n",
    "> place it in the **same directory as this notebook**. If you store it elsewhere,\n",
    "> update `data_path` in the code cell below to match your chosen location."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cd5b103eb9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset loaded — Shape: (13611, 17)\n",
      "Features : 16\n",
      "Target   : {'DERMASON': 3546, 'SIRA': 2636, 'SEKER': 2027, 'HOROZ': 1928, 'CALI': 1630, 'BARBUNYA': 1322, 'BOMBAY': 522}\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Perimeter</th>\n",
       "      <th>MajorAxisLength</th>\n",
       "      <th>MinorAxisLength</th>\n",
       "      <th>AspectRation</th>\n",
       "      <th>Eccentricity</th>\n",
       "      <th>ConvexArea</th>\n",
       "      <th>EquivDiameter</th>\n",
       "      <th>Extent</th>\n",
       "      <th>Solidity</th>\n",
       "      <th>roundness</th>\n",
       "      <th>Compactness</th>\n",
       "      <th>ShapeFactor1</th>\n",
       "      <th>ShapeFactor2</th>\n",
       "      <th>ShapeFactor3</th>\n",
       "      <th>ShapeFactor4</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>28395</td>\n",
       "      <td>610.291</td>\n",
       "      <td>208.178117</td>\n",
       "      <td>173.888747</td>\n",
       "      <td>1.197191</td>\n",
       "      <td>0.549812</td>\n",
       "      <td>28715</td>\n",
       "      <td>190.141097</td>\n",
       "      <td>0.763923</td>\n",
       "      <td>0.988856</td>\n",
       "      <td>0.958027</td>\n",
       "      <td>0.913358</td>\n",
       "      <td>0.007332</td>\n",
       "      <td>0.003147</td>\n",
       "      <td>0.834222</td>\n",
       "      <td>0.998724</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28734</td>\n",
       "      <td>638.018</td>\n",
       "      <td>200.524796</td>\n",
       "      <td>182.734419</td>\n",
       "      <td>1.097356</td>\n",
       "      <td>0.411785</td>\n",
       "      <td>29172</td>\n",
       "      <td>191.272750</td>\n",
       "      <td>0.783968</td>\n",
       "      <td>0.984986</td>\n",
       "      <td>0.887034</td>\n",
       "      <td>0.953861</td>\n",
       "      <td>0.006979</td>\n",
       "      <td>0.003564</td>\n",
       "      <td>0.909851</td>\n",
       "      <td>0.998430</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>29380</td>\n",
       "      <td>624.110</td>\n",
       "      <td>212.826130</td>\n",
       "      <td>175.931143</td>\n",
       "      <td>1.209713</td>\n",
       "      <td>0.562727</td>\n",
       "      <td>29690</td>\n",
       "      <td>193.410904</td>\n",
       "      <td>0.778113</td>\n",
       "      <td>0.989559</td>\n",
       "      <td>0.947849</td>\n",
       "      <td>0.908774</td>\n",
       "      <td>0.007244</td>\n",
       "      <td>0.003048</td>\n",
       "      <td>0.825871</td>\n",
       "      <td>0.999066</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>30008</td>\n",
       "      <td>645.884</td>\n",
       "      <td>210.557999</td>\n",
       "      <td>182.516516</td>\n",
       "      <td>1.153638</td>\n",
       "      <td>0.498616</td>\n",
       "      <td>30724</td>\n",
       "      <td>195.467062</td>\n",
       "      <td>0.782681</td>\n",
       "      <td>0.976696</td>\n",
       "      <td>0.903936</td>\n",
       "      <td>0.928329</td>\n",
       "      <td>0.007017</td>\n",
       "      <td>0.003215</td>\n",
       "      <td>0.861794</td>\n",
       "      <td>0.994199</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>30140</td>\n",
       "      <td>620.134</td>\n",
       "      <td>201.847882</td>\n",
       "      <td>190.279279</td>\n",
       "      <td>1.060798</td>\n",
       "      <td>0.333680</td>\n",
       "      <td>30417</td>\n",
       "      <td>195.896503</td>\n",
       "      <td>0.773098</td>\n",
       "      <td>0.990893</td>\n",
       "      <td>0.984877</td>\n",
       "      <td>0.970516</td>\n",
       "      <td>0.006697</td>\n",
       "      <td>0.003665</td>\n",
       "      <td>0.941900</td>\n",
       "      <td>0.999166</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Area  Perimeter  MajorAxisLength  MinorAxisLength  AspectRation  \\\n",
       "0  28395    610.291       208.178117       173.888747      1.197191   \n",
       "1  28734    638.018       200.524796       182.734419      1.097356   \n",
       "2  29380    624.110       212.826130       175.931143      1.209713   \n",
       "3  30008    645.884       210.557999       182.516516      1.153638   \n",
       "4  30140    620.134       201.847882       190.279279      1.060798   \n",
       "\n",
       "   Eccentricity  ConvexArea  EquivDiameter    Extent  Solidity  roundness  \\\n",
       "0      0.549812       28715     190.141097  0.763923  0.988856   0.958027   \n",
       "1      0.411785       29172     191.272750  0.783968  0.984986   0.887034   \n",
       "2      0.562727       29690     193.410904  0.778113  0.989559   0.947849   \n",
       "3      0.498616       30724     195.467062  0.782681  0.976696   0.903936   \n",
       "4      0.333680       30417     195.896503  0.773098  0.990893   0.984877   \n",
       "\n",
       "   Compactness  ShapeFactor1  ShapeFactor2  ShapeFactor3  ShapeFactor4  Class  \n",
       "0     0.913358      0.007332      0.003147      0.834222      0.998724  SEKER  \n",
       "1     0.953861      0.006979      0.003564      0.909851      0.998430  SEKER  \n",
       "2     0.908774      0.007244      0.003048      0.825871      0.999066  SEKER  \n",
       "3     0.928329      0.007017      0.003215      0.861794      0.994199  SEKER  \n",
       "4     0.970516      0.006697      0.003665      0.941900      0.999166  SEKER  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Update this path if you stored the file in a different location.\n",
    "data_path = \"Dry_Bean_Dataset.xlsx\"\n",
    "\n",
    "df = pd.read_excel(data_path)\n",
    "print(f\"Dataset loaded — Shape: {df.shape}\")\n",
    "\n",
    "target_col = 'Class'\n",
    "X = df.drop(columns=[target_col])\n",
    "y = df[target_col]\n",
    "\n",
    "print(f\"Features : {X.shape[1]}\")\n",
    "print(f\"Target   : {y.value_counts().to_dict()}\")\n",
    "display(df.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cd9f622a6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Column names and dtypes:\n",
      "Area                 int64\n",
      "Perimeter          float64\n",
      "MajorAxisLength    float64\n",
      "MinorAxisLength    float64\n",
      "AspectRation       float64\n",
      "Eccentricity       float64\n",
      "ConvexArea           int64\n",
      "EquivDiameter      float64\n",
      "Extent             float64\n",
      "Solidity           float64\n",
      "roundness          float64\n",
      "Compactness        float64\n",
      "ShapeFactor1       float64\n",
      "ShapeFactor2       float64\n",
      "ShapeFactor3       float64\n",
      "ShapeFactor4       float64\n",
      "Class               object\n",
      "\n",
      "Missing values across all columns: 0\n",
      "\n",
      "All features are numerical — no categorical encoding required.\n"
     ]
    }
   ],
   "source": [
    "print(\"Column names and dtypes:\")\n",
    "print(df.dtypes.to_string())\n",
    "print(f\"\\nMissing values across all columns: {df.isnull().sum().sum()}\")\n",
    "print(\"\\nAll features are numerical — no categorical encoding required.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md11d3908a",
   "metadata": {},
   "source": [
    "## 3. Understanding Multi-Class Imbalance\n",
    "\n",
    "Seven classes with very different prevalences create a challenging prediction problem. DERMASON accounts for 26% of samples; BOMBAY accounts for fewer than 4%. A naive majority-class baseline achieves only 26% accuracy — but a model that ignores BOMBAY samples will still appear to perform well on aggregate metrics. We use **stratified splitting** and **balanced accuracy** to guard against this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "cd3bccaea3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class distribution:\n",
      "  DERMASON   3,546  ( 26.1%)  █████████████████\n",
      "  SIRA       2,636  ( 19.4%)  ████████████\n",
      "  SEKER      2,027  ( 14.9%)  █████████\n",
      "  HOROZ      1,928  ( 14.2%)  █████████\n",
      "  CALI       1,630  ( 12.0%)  ████████\n",
      "  BARBUNYA   1,322  (  9.7%)  ██████\n",
      "  BOMBAY       522  (  3.8%)  ██\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Majority-class baseline accuracy : 26.1%  (always predicting DERMASON)\n"
     ]
    }
   ],
   "source": [
    "BEAN_ORDER = ['DERMASON', 'SIRA', 'SEKER', 'HOROZ', 'CALI', 'BARBUNYA', 'BOMBAY']\n",
    "BEAN_COLORS = {\n",
    "    'DERMASON' : '#2563EB',\n",
    "    'SIRA'     : '#10B981',\n",
    "    'SEKER'    : '#F59E0B',\n",
    "    'HOROZ'    : '#8B5CF6',\n",
    "    'CALI'     : '#EF4444',\n",
    "    'BARBUNYA' : '#EC4899',\n",
    "    'BOMBAY'   : '#6B7280',\n",
    "}\n",
    "\n",
    "bean_counts = y.value_counts().reindex(BEAN_ORDER)\n",
    "bean_pct    = (bean_counts / len(y) * 100).round(1)\n",
    "\n",
    "print(\"Class distribution:\")\n",
    "for bean, count, pct in zip(BEAN_ORDER, bean_counts, bean_pct):\n",
    "    bar = '\\u2588' * int(pct / 1.5)\n",
    "    print(f\"  {bean:<10} {count:>5,}  ({pct:>5.1f}%)  {bar}\")\n",
    "\n",
    "fig, axes = plt.subplots(1, 2, figsize=(13, 4))\n",
    "fig.suptitle(\"Dry Bean Class Distribution — 13,611 Samples\", fontsize=13, fontweight=\"bold\")\n",
    "\n",
    "ax = axes[0]\n",
    "bars = ax.bar(\n",
    "    BEAN_ORDER, bean_counts.values,\n",
    "    color=[BEAN_COLORS[b] for b in BEAN_ORDER],\n",
    "    edgecolor='#555555', linewidth=0.8\n",
    ")\n",
    "ax.set_title(\"Sample Count per Variety\", fontweight=\"bold\")\n",
    "ax.set_ylabel(\"Count\")\n",
    "ax.set_xlabel(\"Bean Variety\")\n",
    "ax.tick_params(axis='x', rotation=30)\n",
    "for bar, count in zip(bars, bean_counts.values):\n",
    "    ax.text(bar.get_x() + bar.get_width() / 2, bar.get_height() + 20,\n",
    "            f\"{count:,}\", ha='center', va='bottom', fontsize=8)\n",
    "ax.grid(axis='y', alpha=0.3)\n",
    "\n",
    "ax = axes[1]\n",
    "wedges, texts, autotexts = ax.pie(\n",
    "    bean_counts.values,\n",
    "    labels=BEAN_ORDER,\n",
    "    colors=[BEAN_COLORS[b] for b in BEAN_ORDER],\n",
    "    autopct='%1.1f%%', startangle=90,\n",
    "    wedgeprops=dict(edgecolor='#555555', linewidth=0.8)\n",
    ")\n",
    "for at in autotexts:\n",
    "    at.set_fontsize(8)\n",
    "ax.set_title(\"Proportion per Variety\", fontweight=\"bold\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "majority_acc = bean_counts.max() / len(y)\n",
    "print(f\"\\nMajority-class baseline accuracy : {majority_acc:.1%}  (always predicting DERMASON)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md1bff264c",
   "metadata": {},
   "source": [
    "## 4. Exploratory Feature Analysis\n",
    "\n",
    "`generate_feature_stats` audits every feature column. All 16 features here are continuous numerical measurements — area in pixels, lengths in pixels, ratios and shape coefficients — which simplifies the preprocessing decision: we need scaling, not encoding."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "cd52fdc7e8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Numerical   (16): ['Area', 'Perimeter', 'MajorAxisLength', 'MinorAxisLength', 'AspectRation', 'Eccentricity', 'ConvexArea', 'EquivDiameter', 'Extent', 'Solidity', 'roundness', 'Compactness', 'ShapeFactor1', 'ShapeFactor2', 'ShapeFactor3', 'ShapeFactor4']\n",
      "Categorical (0): []\n",
      "------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dtype</th>\n",
       "      <th>missing_count</th>\n",
       "      <th>missing_percent</th>\n",
       "      <th>unique_values</th>\n",
       "      <th>zero_count</th>\n",
       "      <th>negative_count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>skewness</th>\n",
       "      <th>kurtosis</th>\n",
       "      <th>mean_percentile</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Area</th>\n",
       "      <td>int64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12011</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>53048.284549</td>\n",
       "      <td>29324.095717</td>\n",
       "      <td>2.952931</td>\n",
       "      <td>10.800814</td>\n",
       "      <td>66.497686</td>\n",
       "      <td>20420.000000</td>\n",
       "      <td>36328.000000</td>\n",
       "      <td>44652.000000</td>\n",
       "      <td>61332.000000</td>\n",
       "      <td>254616.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Perimeter</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13416</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>855.283459</td>\n",
       "      <td>214.289696</td>\n",
       "      <td>1.626124</td>\n",
       "      <td>3.588123</td>\n",
       "      <td>60.737639</td>\n",
       "      <td>524.736000</td>\n",
       "      <td>703.523500</td>\n",
       "      <td>794.941000</td>\n",
       "      <td>977.213000</td>\n",
       "      <td>1985.370000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MajorAxisLength</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>320.141867</td>\n",
       "      <td>85.694186</td>\n",
       "      <td>1.357815</td>\n",
       "      <td>2.531902</td>\n",
       "      <td>58.695173</td>\n",
       "      <td>183.601165</td>\n",
       "      <td>253.303633</td>\n",
       "      <td>296.883367</td>\n",
       "      <td>376.495012</td>\n",
       "      <td>738.860153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MinorAxisLength</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>202.270714</td>\n",
       "      <td>44.970091</td>\n",
       "      <td>2.238211</td>\n",
       "      <td>6.651067</td>\n",
       "      <td>64.998898</td>\n",
       "      <td>122.512653</td>\n",
       "      <td>175.848170</td>\n",
       "      <td>192.431733</td>\n",
       "      <td>217.031741</td>\n",
       "      <td>460.198497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AspectRation</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.583242</td>\n",
       "      <td>0.246678</td>\n",
       "      <td>0.582573</td>\n",
       "      <td>0.113814</td>\n",
       "      <td>56.432297</td>\n",
       "      <td>1.024868</td>\n",
       "      <td>1.432307</td>\n",
       "      <td>1.551124</td>\n",
       "      <td>1.707109</td>\n",
       "      <td>2.430306</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Eccentricity</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.750895</td>\n",
       "      <td>0.092002</td>\n",
       "      <td>-1.062824</td>\n",
       "      <td>1.387456</td>\n",
       "      <td>41.716259</td>\n",
       "      <td>0.218951</td>\n",
       "      <td>0.715928</td>\n",
       "      <td>0.764441</td>\n",
       "      <td>0.810466</td>\n",
       "      <td>0.911423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ConvexArea</th>\n",
       "      <td>int64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12066</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>53768.200206</td>\n",
       "      <td>29774.915817</td>\n",
       "      <td>2.941821</td>\n",
       "      <td>10.743640</td>\n",
       "      <td>66.409522</td>\n",
       "      <td>20684.000000</td>\n",
       "      <td>36714.500000</td>\n",
       "      <td>45178.000000</td>\n",
       "      <td>62294.000000</td>\n",
       "      <td>263261.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EquivDiameter</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12011</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>253.064220</td>\n",
       "      <td>59.177120</td>\n",
       "      <td>1.948958</td>\n",
       "      <td>5.192057</td>\n",
       "      <td>62.302549</td>\n",
       "      <td>161.243764</td>\n",
       "      <td>215.068003</td>\n",
       "      <td>238.438026</td>\n",
       "      <td>279.446467</td>\n",
       "      <td>569.374358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Extent</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13535</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.749733</td>\n",
       "      <td>0.049086</td>\n",
       "      <td>-0.895348</td>\n",
       "      <td>0.643319</td>\n",
       "      <td>43.060760</td>\n",
       "      <td>0.555315</td>\n",
       "      <td>0.718634</td>\n",
       "      <td>0.759859</td>\n",
       "      <td>0.786851</td>\n",
       "      <td>0.866195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Solidity</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13526</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.987143</td>\n",
       "      <td>0.004660</td>\n",
       "      <td>-2.550093</td>\n",
       "      <td>12.799621</td>\n",
       "      <td>36.536625</td>\n",
       "      <td>0.919246</td>\n",
       "      <td>0.985670</td>\n",
       "      <td>0.988283</td>\n",
       "      <td>0.990013</td>\n",
       "      <td>0.994677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roundness</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.873282</td>\n",
       "      <td>0.059520</td>\n",
       "      <td>-0.635749</td>\n",
       "      <td>0.374306</td>\n",
       "      <td>43.942400</td>\n",
       "      <td>0.489618</td>\n",
       "      <td>0.832096</td>\n",
       "      <td>0.883157</td>\n",
       "      <td>0.916869</td>\n",
       "      <td>0.990685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Compactness</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.799864</td>\n",
       "      <td>0.061713</td>\n",
       "      <td>0.037115</td>\n",
       "      <td>-0.223459</td>\n",
       "      <td>48.886930</td>\n",
       "      <td>0.640577</td>\n",
       "      <td>0.762469</td>\n",
       "      <td>0.801277</td>\n",
       "      <td>0.834270</td>\n",
       "      <td>0.987303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ShapeFactor1</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.006564</td>\n",
       "      <td>0.001128</td>\n",
       "      <td>-0.534141</td>\n",
       "      <td>0.714355</td>\n",
       "      <td>45.720373</td>\n",
       "      <td>0.002778</td>\n",
       "      <td>0.005900</td>\n",
       "      <td>0.006645</td>\n",
       "      <td>0.007271</td>\n",
       "      <td>0.010451</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ShapeFactor2</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.001716</td>\n",
       "      <td>0.000596</td>\n",
       "      <td>0.301226</td>\n",
       "      <td>-0.859254</td>\n",
       "      <td>51.186540</td>\n",
       "      <td>0.000564</td>\n",
       "      <td>0.001154</td>\n",
       "      <td>0.001694</td>\n",
       "      <td>0.002170</td>\n",
       "      <td>0.003665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ShapeFactor3</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.643590</td>\n",
       "      <td>0.098996</td>\n",
       "      <td>0.242481</td>\n",
       "      <td>-0.144475</td>\n",
       "      <td>50.811843</td>\n",
       "      <td>0.410339</td>\n",
       "      <td>0.581359</td>\n",
       "      <td>0.642044</td>\n",
       "      <td>0.696006</td>\n",
       "      <td>0.974767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ShapeFactor4</th>\n",
       "      <td>float64</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13543</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.995063</td>\n",
       "      <td>0.004366</td>\n",
       "      <td>-2.759483</td>\n",
       "      <td>13.038067</td>\n",
       "      <td>35.265594</td>\n",
       "      <td>0.947687</td>\n",
       "      <td>0.993703</td>\n",
       "      <td>0.996386</td>\n",
       "      <td>0.997883</td>\n",
       "      <td>0.999733</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   dtype  missing_count  missing_percent  unique_values  \\\n",
       "Area               int64              0              0.0          12011   \n",
       "Perimeter        float64              0              0.0          13416   \n",
       "MajorAxisLength  float64              0              0.0          13543   \n",
       "MinorAxisLength  float64              0              0.0          13543   \n",
       "AspectRation     float64              0              0.0          13543   \n",
       "Eccentricity     float64              0              0.0          13543   \n",
       "ConvexArea         int64              0              0.0          12066   \n",
       "EquivDiameter    float64              0              0.0          12011   \n",
       "Extent           float64              0              0.0          13535   \n",
       "Solidity         float64              0              0.0          13526   \n",
       "roundness        float64              0              0.0          13543   \n",
       "Compactness      float64              0              0.0          13543   \n",
       "ShapeFactor1     float64              0              0.0          13543   \n",
       "ShapeFactor2     float64              0              0.0          13543   \n",
       "ShapeFactor3     float64              0              0.0          13543   \n",
       "ShapeFactor4     float64              0              0.0          13543   \n",
       "\n",
       "                 zero_count  negative_count          mean           std  \\\n",
       "Area                      0               0  53048.284549  29324.095717   \n",
       "Perimeter                 0               0    855.283459    214.289696   \n",
       "MajorAxisLength           0               0    320.141867     85.694186   \n",
       "MinorAxisLength           0               0    202.270714     44.970091   \n",
       "AspectRation              0               0      1.583242      0.246678   \n",
       "Eccentricity              0               0      0.750895      0.092002   \n",
       "ConvexArea                0               0  53768.200206  29774.915817   \n",
       "EquivDiameter             0               0    253.064220     59.177120   \n",
       "Extent                    0               0      0.749733      0.049086   \n",
       "Solidity                  0               0      0.987143      0.004660   \n",
       "roundness                 0               0      0.873282      0.059520   \n",
       "Compactness               0               0      0.799864      0.061713   \n",
       "ShapeFactor1              0               0      0.006564      0.001128   \n",
       "ShapeFactor2              0               0      0.001716      0.000596   \n",
       "ShapeFactor3              0               0      0.643590      0.098996   \n",
       "ShapeFactor4              0               0      0.995063      0.004366   \n",
       "\n",
       "                 skewness   kurtosis  mean_percentile           min  \\\n",
       "Area             2.952931  10.800814        66.497686  20420.000000   \n",
       "Perimeter        1.626124   3.588123        60.737639    524.736000   \n",
       "MajorAxisLength  1.357815   2.531902        58.695173    183.601165   \n",
       "MinorAxisLength  2.238211   6.651067        64.998898    122.512653   \n",
       "AspectRation     0.582573   0.113814        56.432297      1.024868   \n",
       "Eccentricity    -1.062824   1.387456        41.716259      0.218951   \n",
       "ConvexArea       2.941821  10.743640        66.409522  20684.000000   \n",
       "EquivDiameter    1.948958   5.192057        62.302549    161.243764   \n",
       "Extent          -0.895348   0.643319        43.060760      0.555315   \n",
       "Solidity        -2.550093  12.799621        36.536625      0.919246   \n",
       "roundness       -0.635749   0.374306        43.942400      0.489618   \n",
       "Compactness      0.037115  -0.223459        48.886930      0.640577   \n",
       "ShapeFactor1    -0.534141   0.714355        45.720373      0.002778   \n",
       "ShapeFactor2     0.301226  -0.859254        51.186540      0.000564   \n",
       "ShapeFactor3     0.242481  -0.144475        50.811843      0.410339   \n",
       "ShapeFactor4    -2.759483  13.038067        35.265594      0.947687   \n",
       "\n",
       "                          25%           50%           75%            max  \n",
       "Area             36328.000000  44652.000000  61332.000000  254616.000000  \n",
       "Perimeter          703.523500    794.941000    977.213000    1985.370000  \n",
       "MajorAxisLength    253.303633    296.883367    376.495012     738.860153  \n",
       "MinorAxisLength    175.848170    192.431733    217.031741     460.198497  \n",
       "AspectRation         1.432307      1.551124      1.707109       2.430306  \n",
       "Eccentricity         0.715928      0.764441      0.810466       0.911423  \n",
       "ConvexArea       36714.500000  45178.000000  62294.000000  263261.000000  \n",
       "EquivDiameter      215.068003    238.438026    279.446467     569.374358  \n",
       "Extent               0.718634      0.759859      0.786851       0.866195  \n",
       "Solidity             0.985670      0.988283      0.990013       0.994677  \n",
       "roundness            0.832096      0.883157      0.916869       0.990685  \n",
       "Compactness          0.762469      0.801277      0.834270       0.987303  \n",
       "ShapeFactor1         0.005900      0.006645      0.007271       0.010451  \n",
       "ShapeFactor2         0.001154      0.001694      0.002170       0.003665  \n",
       "ShapeFactor3         0.581359      0.642044      0.696006       0.974767  \n",
       "ShapeFactor4         0.993703      0.996386      0.997883       0.999733  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "stats_df, numerical_cols, categorical_cols = generate_feature_stats(X)\n",
    "\n",
    "print(f\"Numerical   ({len(numerical_cols)}): {numerical_cols}\")\n",
    "print(f\"Categorical ({len(categorical_cols)}): {categorical_cols}\")\n",
    "print(\"-\" * 60)\n",
    "display(stats_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mda784f53e",
   "metadata": {},
   "source": [
    "## 5. The Golden Rule: Split Before You Transform\n",
    "\n",
    "Fitting scalers on the full dataset leaks test-set statistics into training — an artificially lower training loss that does not generalise. We split first, then fit all transforms exclusively on `X_train`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "cdd8820b49",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class encoding: {'BARBUNYA': np.int64(0), 'BOMBAY': np.int64(1), 'CALI': np.int64(2), 'DERMASON': np.int64(3), 'HOROZ': np.int64(4), 'SEKER': np.int64(5), 'SIRA': np.int64(6)}\n",
      "\n",
      "Training : 10,888 samples\n",
      "Test     : 2,723 samples\n",
      "\n",
      "Class proportions preserved across splits:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Train %</th>\n",
       "      <th>Test %</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Class</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DERMASON</th>\n",
       "      <td>26.1</td>\n",
       "      <td>26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SIRA</th>\n",
       "      <td>19.4</td>\n",
       "      <td>19.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SEKER</th>\n",
       "      <td>14.9</td>\n",
       "      <td>14.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HOROZ</th>\n",
       "      <td>14.2</td>\n",
       "      <td>14.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CALI</th>\n",
       "      <td>12.0</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BARBUNYA</th>\n",
       "      <td>9.7</td>\n",
       "      <td>9.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BOMBAY</th>\n",
       "      <td>3.8</td>\n",
       "      <td>3.8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Train %  Test %\n",
       "Class                    \n",
       "DERMASON     26.1    26.0\n",
       "SIRA         19.4    19.4\n",
       "SEKER        14.9    14.9\n",
       "HOROZ        14.2    14.2\n",
       "CALI         12.0    12.0\n",
       "BARBUNYA      9.7     9.7\n",
       "BOMBAY        3.8     3.8"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "le = LabelEncoder()\n",
    "le.fit(y)\n",
    "print(f\"Class encoding: {dict(zip(le.classes_, le.transform(le.classes_)))}\")\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, test_size=0.20, random_state=42, stratify=y\n",
    ")\n",
    "\n",
    "y_train_encoded = le.transform(y_train)\n",
    "y_test_encoded  = le.transform(y_test)\n",
    "\n",
    "print(f\"\\nTraining : {X_train.shape[0]:,} samples\")\n",
    "print(f\"Test     : {X_test.shape[0]:,} samples\")\n",
    "\n",
    "print(\"\\nClass proportions preserved across splits:\")\n",
    "train_dist = pd.Series(y_train).value_counts(normalize=True).reindex(BEAN_ORDER)\n",
    "test_dist  = pd.Series(y_test).value_counts(normalize=True).reindex(BEAN_ORDER)\n",
    "display(pd.DataFrame({'Train %': (train_dist * 100).round(1), 'Test %': (test_dist * 100).round(1)}))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md82f1621e",
   "metadata": {},
   "source": [
    "## 6. Feature Engineering with `TransformPipeline`\n",
    "\n",
    "All 16 features are numerical morphological measurements. Several (Area, Perimeter, ConvexArea, EquivDiameter) span large absolute ranges — standard scaling brings them onto a common scale so the Optuna search isn't biased by feature magnitude differences. Shape factors and ratios (Eccentricity, Solidity, Compactness) are already bounded to [0, 1] but scaling them consistently with the rest of the pipeline is still best practice."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "cd374bbf15",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TransformPipeline(name='dry_bean_preprocessing', steps=1, status=not fitted)\n",
      "\n",
      "Steps: 1\n",
      "  [numerical   ] standard_scale       → 16 columns\n"
     ]
    }
   ],
   "source": [
    "pipeline = TransformPipeline(name=\"dry_bean_preprocessing\")\n",
    "\n",
    "pipeline.add(\n",
    "    transformer_type=\"numerical\",\n",
    "    method=\"standard_scale\",\n",
    "    columns=numerical_cols\n",
    ")\n",
    "\n",
    "print(pipeline)\n",
    "print(f\"\\nSteps: {len(pipeline)}\")\n",
    "for step in pipeline.list_steps():\n",
    "    print(f\"  [{step['type']:12s}] {step['method']:20s} \\u2192 {len(step['columns'])} columns\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "cdd93cd93a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training features : 16 raw → 16 post-transform\n",
      "Training samples  : 10,888\n",
      "Pipeline fitted   : True\n"
     ]
    }
   ],
   "source": [
    "X_train_transformed = pipeline.fit_transform(X_train)\n",
    "\n",
    "print(f\"Training features : {X_train.shape[1]} raw \\u2192 {X_train_transformed.shape[1]} post-transform\")\n",
    "print(f\"Training samples  : {X_train_transformed.shape[0]:,}\")\n",
    "print(f\"Pipeline fitted   : {pipeline.is_fitted}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md1ff1c709",
   "metadata": {},
   "source": [
    "### 6.1 Serialise, Reload, and Apply to Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "cd284c2310",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pipeline serialised → pipeline_dry_bean.pkl\n",
      "Pipeline reloaded  — fitted: True  |  steps: 1\n",
      "\n",
      "Test set  : (2723, 16) → (2723, 16)\n",
      "Column alignment: True\n"
     ]
    }
   ],
   "source": [
    "pipeline_path = \"pipeline_dry_bean.pkl\"\n",
    "pipeline.save(pipeline_path)\n",
    "print(f\"Pipeline serialised \\u2192 {pipeline_path}\")\n",
    "\n",
    "loaded_pipeline = TransformPipeline.load(pipeline_path)\n",
    "print(f\"Pipeline reloaded  — fitted: {loaded_pipeline.is_fitted}  |  steps: {len(loaded_pipeline)}\")\n",
    "\n",
    "X_test_transformed = loaded_pipeline.transform(X_test)\n",
    "\n",
    "print(f\"\\nTest set  : {X_test.shape} \\u2192 {X_test_transformed.shape}\")\n",
    "print(f\"Column alignment: {list(X_train_transformed.columns) == list(X_test_transformed.columns)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mdea92f323",
   "metadata": {},
   "source": [
    "## 7. Model Training\n",
    "\n",
    "`task=\"multiclass_classification\"` is the only configuration change from a binary setup. BitBullet selects the softmax objective, sets `num_class=7`, and uses macro OvR ROC AUC for CV scoring automatically."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "cd7ef3d4c6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[I 2026-05-05 19:30:55,838] A new study created in memory with name: dry_bean_lgbm\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "======================================================================\n",
      "BitBullet Train - dry_bean_lgbm\n",
      "======================================================================\n",
      "\n",
      "Training samples: 10888\n",
      "Features: 16\n",
      "Class distribution: {3: 2837, 6: 2109, 5: 1621, 4: 1542, 2: 1304, 0: 1057, 1: 418}\n",
      "\n",
      "Starting hyperparameter optimization (optuna)...\n",
      "Trials: 50, CV Folds: 5\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[I 2026-05-05 19:31:28,572] Trial 14 finished with value: 0.995310276978414 and parameters: {'num_leaves': 226, 'max_depth': 6, 'min_child_samples': 22, 'lambda_l1': 1.2216850427777404e-06, 'lambda_l2': 0.0010876504729103167, 'min_gain_to_split': 0.9747751133352428, 'feature_fraction': 0.6299670136520754, 'bagging_fraction': 0.8756671376675983, 'bagging_freq': 3, 'learning_rate': 0.1941119140046595, 'max_bin': 127}. Best is trial 14 with value: 0.995310276978414.\n",
      "[I 2026-05-05 19:31:37,186] Trial 8 finished with value: 0.9953818643124024 and parameters: {'num_leaves': 272, 'max_depth': 6, 'min_child_samples': 99, 'lambda_l1': 1.871504203941811, 'lambda_l2': 0.9360771370400477, 'min_gain_to_split': 0.28316456072493756, 'feature_fraction': 0.9830631328426769, 'bagging_fraction': 0.8336729089723773, 'bagging_freq': 2, 'learning_rate': 0.21662223946853565, 'max_bin': 159}. Best is trial 8 with value: 0.9953818643124024.\n",
      "[I 2026-05-05 19:31:53,954] Trial 3 finished with value: 0.9953734344613032 and parameters: {'num_leaves': 159, 'max_depth': 12, 'min_child_samples': 85, 'lambda_l1': 1.0919512215206316e-06, 'lambda_l2': 0.04153387470971465, 'min_gain_to_split': 0.6452881694106882, 'feature_fraction': 0.8229312148073686, 'bagging_fraction': 0.8360152405610404, 'bagging_freq': 7, 'learning_rate': 0.1228022691746178, 'max_bin': 191}. Best is trial 8 with value: 0.9953818643124024.\n",
      "[I 2026-05-05 19:31:58,570] Trial 16 finished with value: 0.9948545150751364 and parameters: {'num_leaves': 190, 'max_depth': 7, 'min_child_samples': 9, 'lambda_l1': 0.0021313194206088394, 'lambda_l2': 0.0002874269935870234, 'min_gain_to_split': 0.8525614960734657, 'feature_fraction': 0.9170216913779051, 'bagging_fraction': 0.9450397546635948, 'bagging_freq': 5, 'learning_rate': 0.2862005631876366, 'max_bin': 191}. Best is trial 8 with value: 0.9953818643124024.\n",
      "[I 2026-05-05 19:32:02,574] Trial 15 finished with value: 0.9951398657461912 and parameters: {'num_leaves': 77, 'max_depth': 6, 'min_child_samples': 12, 'lambda_l1': 1.7367689769838533e-07, 'lambda_l2': 8.910796233056972, 'min_gain_to_split': 0.26169215905309307, 'feature_fraction': 0.6708159328692289, 'bagging_fraction': 0.6879137533482832, 'bagging_freq': 6, 'learning_rate': 0.23003748594584889, 'max_bin': 255}. Best is trial 8 with value: 0.9953818643124024.\n",
      "[I 2026-05-05 19:32:12,110] Trial 5 finished with value: 0.9956361287982063 and parameters: {'num_leaves': 127, 'max_depth': 3, 'min_child_samples': 5, 'lambda_l1': 8.123958352840565e-06, 'lambda_l2': 2.8535369367142184e-07, 'min_gain_to_split': 0.4556522046080762, 'feature_fraction': 0.6822762582029247, 'bagging_fraction': 0.9266798024551328, 'bagging_freq': 3, 'learning_rate': 0.05554330574588168, 'max_bin': 127}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:32:16,860] Trial 11 finished with value: 0.9954258870794515 and parameters: {'num_leaves': 269, 'max_depth': 7, 'min_child_samples': 62, 'lambda_l1': 1.7623715062726464e-07, 'lambda_l2': 3.816916790295202e-06, 'min_gain_to_split': 0.3994373664746038, 'feature_fraction': 0.6261894368750877, 'bagging_fraction': 0.6516346847885612, 'bagging_freq': 1, 'learning_rate': 0.11502193511625303, 'max_bin': 223}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:32:29,602] Trial 17 finished with value: 0.9952076336925344 and parameters: {'num_leaves': 280, 'max_depth': 9, 'min_child_samples': 19, 'lambda_l1': 2.47310047752173e-06, 'lambda_l2': 1.3032865346649805, 'min_gain_to_split': 0.605826026421758, 'feature_fraction': 0.6539251536694162, 'bagging_fraction': 0.6286154646910316, 'bagging_freq': 3, 'learning_rate': 0.2272222499239823, 'max_bin': 63}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:32:30,377] Trial 0 finished with value: 0.9953139318202204 and parameters: {'num_leaves': 182, 'max_depth': 11, 'min_child_samples': 40, 'lambda_l1': 0.0006103267655106399, 'lambda_l2': 3.498651930754746e-05, 'min_gain_to_split': 0.343623957487574, 'feature_fraction': 0.7173300810515699, 'bagging_fraction': 0.9015053633887007, 'bagging_freq': 1, 'learning_rate': 0.10257266406482887, 'max_bin': 191}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:32:33,717] Trial 2 finished with value: 0.9952585133575349 and parameters: {'num_leaves': 160, 'max_depth': 9, 'min_child_samples': 19, 'lambda_l1': 0.01136523740913846, 'lambda_l2': 0.017695911249944563, 'min_gain_to_split': 0.5348665691067385, 'feature_fraction': 0.8523110708754085, 'bagging_fraction': 0.6639276974504479, 'bagging_freq': 5, 'learning_rate': 0.08836663832621919, 'max_bin': 127}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:32:55,614] Trial 1 finished with value: 0.9955870310031235 and parameters: {'num_leaves': 243, 'max_depth': 12, 'min_child_samples': 38, 'lambda_l1': 0.38392323942083634, 'lambda_l2': 0.0015021131212307544, 'min_gain_to_split': 0.640351888018897, 'feature_fraction': 0.6171314536249424, 'bagging_fraction': 0.8064233081954508, 'bagging_freq': 7, 'learning_rate': 0.055026066386206704, 'max_bin': 63}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:33:06,495] Trial 21 finished with value: 0.9953370755502229 and parameters: {'num_leaves': 253, 'max_depth': 8, 'min_child_samples': 7, 'lambda_l1': 9.970061843021717e-08, 'lambda_l2': 0.037321252337759236, 'min_gain_to_split': 0.9812969784957366, 'feature_fraction': 0.852800587999297, 'bagging_fraction': 0.8604282459512955, 'bagging_freq': 3, 'learning_rate': 0.11576724844422992, 'max_bin': 255}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:33:33,129] Trial 23 finished with value: 0.9953356542708924 and parameters: {'num_leaves': 252, 'max_depth': 6, 'min_child_samples': 18, 'lambda_l1': 1.6772476711376047e-05, 'lambda_l2': 2.668568740913787e-07, 'min_gain_to_split': 0.36356208676359403, 'feature_fraction': 0.9181548507546556, 'bagging_fraction': 0.7452366129375917, 'bagging_freq': 2, 'learning_rate': 0.15253202203429342, 'max_bin': 191}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:33:40,539] Trial 19 finished with value: 0.9955104169905375 and parameters: {'num_leaves': 47, 'max_depth': 12, 'min_child_samples': 87, 'lambda_l1': 1.2626387164401162e-07, 'lambda_l2': 3.107357252858523e-08, 'min_gain_to_split': 0.5643898639661518, 'feature_fraction': 0.5021381705044634, 'bagging_fraction': 0.8740561567988078, 'bagging_freq': 1, 'learning_rate': 0.06762619373967828, 'max_bin': 63}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:34:13,702] Trial 25 finished with value: 0.9955833398548126 and parameters: {'num_leaves': 90, 'max_depth': 3, 'min_child_samples': 5, 'lambda_l1': 0.005688963183156376, 'lambda_l2': 0.0003111689308915312, 'min_gain_to_split': 0.47013392736473386, 'feature_fraction': 0.6730078537052165, 'bagging_fraction': 0.9250136530050982, 'bagging_freq': 5, 'learning_rate': 0.03960811164214711, 'max_bin': 127}. Best is trial 5 with value: 0.9956361287982063.\n",
      "[I 2026-05-05 19:34:42,138] Trial 13 finished with value: 0.9956555926591975 and parameters: {'num_leaves': 20, 'max_depth': 3, 'min_child_samples': 6, 'lambda_l1': 0.002516124230903166, 'lambda_l2': 0.3827011242967604, 'min_gain_to_split': 0.8206578991885104, 'feature_fraction': 0.8088061247982747, 'bagging_fraction': 0.7892724280696003, 'bagging_freq': 2, 'learning_rate': 0.014237658587585604, 'max_bin': 255}. Best is trial 13 with value: 0.9956555926591975.\n",
      "[I 2026-05-05 19:34:56,350] Trial 27 finished with value: 0.9955277702655879 and parameters: {'num_leaves': 190, 'max_depth': 10, 'min_child_samples': 25, 'lambda_l1': 0.15129120305015978, 'lambda_l2': 4.142011223697147e-06, 'min_gain_to_split': 0.5355710079321002, 'feature_fraction': 0.6631207580537366, 'bagging_fraction': 0.8124986603030669, 'bagging_freq': 7, 'learning_rate': 0.06771183072897428, 'max_bin': 63}. Best is trial 13 with value: 0.9956555926591975.\n",
      "[I 2026-05-05 19:35:12,147] Trial 18 finished with value: 0.9956685603663644 and parameters: {'num_leaves': 164, 'max_depth': 4, 'min_child_samples': 17, 'lambda_l1': 0.0005680384347629229, 'lambda_l2': 0.06257156333511303, 'min_gain_to_split': 0.6594220793024006, 'feature_fraction': 0.9708626504899072, 'bagging_fraction': 0.8296862031231932, 'bagging_freq': 5, 'learning_rate': 0.017329411323994393, 'max_bin': 95}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:35:12,472] Trial 28 finished with value: 0.9954586626805673 and parameters: {'num_leaves': 244, 'max_depth': 12, 'min_child_samples': 51, 'lambda_l1': 0.30439558725401356, 'lambda_l2': 5.0160849798922706e-05, 'min_gain_to_split': 0.9002431745570263, 'feature_fraction': 0.5492128351180758, 'bagging_fraction': 0.6334373085907152, 'bagging_freq': 7, 'learning_rate': 0.05106811007830518, 'max_bin': 95}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:35:17,258] Trial 26 finished with value: 0.9954967654610417 and parameters: {'num_leaves': 271, 'max_depth': 10, 'min_child_samples': 57, 'lambda_l1': 0.2656312980785257, 'lambda_l2': 8.544145509728189e-07, 'min_gain_to_split': 0.3769087808172218, 'feature_fraction': 0.5734899918373039, 'bagging_fraction': 0.7143542416552957, 'bagging_freq': 7, 'learning_rate': 0.05159302904304038, 'max_bin': 95}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:35:48,704] Trial 7 finished with value: 0.9955074508689826 and parameters: {'num_leaves': 275, 'max_depth': 12, 'min_child_samples': 46, 'lambda_l1': 1.3595189100408551, 'lambda_l2': 0.38260513593057655, 'min_gain_to_split': 0.14396719307639927, 'feature_fraction': 0.8249454099067652, 'bagging_fraction': 0.5799206025599142, 'bagging_freq': 2, 'learning_rate': 0.0218168743142418, 'max_bin': 127}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:36:10,639] Trial 32 finished with value: 0.9954469358190412 and parameters: {'num_leaves': 85, 'max_depth': 5, 'min_child_samples': 9, 'lambda_l1': 5.264654778577544e-06, 'lambda_l2': 2.82253584836318e-08, 'min_gain_to_split': 0.33879973330683716, 'feature_fraction': 0.5644775765248936, 'bagging_fraction': 0.7983503562144308, 'bagging_freq': 2, 'learning_rate': 0.10317802292536961, 'max_bin': 127}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:36:15,841] Trial 6 finished with value: 0.995546994717435 and parameters: {'num_leaves': 146, 'max_depth': 9, 'min_child_samples': 35, 'lambda_l1': 0.00589570142711223, 'lambda_l2': 0.004902666014276008, 'min_gain_to_split': 0.3335612667225477, 'feature_fraction': 0.6468647062932555, 'bagging_fraction': 0.9423889466410409, 'bagging_freq': 3, 'learning_rate': 0.023619722711650387, 'max_bin': 223}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:36:47,366] Trial 4 finished with value: 0.9956253437625413 and parameters: {'num_leaves': 42, 'max_depth': 3, 'min_child_samples': 16, 'lambda_l1': 0.13841416664805856, 'lambda_l2': 0.02157762035311752, 'min_gain_to_split': 0.9131397465846877, 'feature_fraction': 0.9623831630979394, 'bagging_fraction': 0.532879300496352, 'bagging_freq': 7, 'learning_rate': 0.006910974989570394, 'max_bin': 63}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:37:09,806] Trial 30 finished with value: 0.9955327502126504 and parameters: {'num_leaves': 234, 'max_depth': 12, 'min_child_samples': 26, 'lambda_l1': 0.00018466739969569116, 'lambda_l2': 4.368623711979059e-05, 'min_gain_to_split': 0.7538064755055006, 'feature_fraction': 0.5940086652813431, 'bagging_fraction': 0.8054161226314384, 'bagging_freq': 7, 'learning_rate': 0.038942161158439834, 'max_bin': 95}. Best is trial 18 with value: 0.9956685603663644.\n",
      "[I 2026-05-05 19:37:16,215] Trial 36 finished with value: 0.9956713950728802 and parameters: {'num_leaves': 172, 'max_depth': 3, 'min_child_samples': 5, 'lambda_l1': 1.2158568318093246e-06, 'lambda_l2': 1.1182701662877744e-07, 'min_gain_to_split': 0.21527568422880633, 'feature_fraction': 0.6104217815601655, 'bagging_fraction': 0.8822936933674879, 'bagging_freq': 3, 'learning_rate': 0.04405118445144062, 'max_bin': 95}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:37:32,238] Trial 31 finished with value: 0.9955172610541378 and parameters: {'num_leaves': 28, 'max_depth': 6, 'min_child_samples': 6, 'lambda_l1': 0.0014792805245907436, 'lambda_l2': 0.022072627906426303, 'min_gain_to_split': 0.8496604049685154, 'feature_fraction': 0.8843185702987948, 'bagging_fraction': 0.6326147912930222, 'bagging_freq': 4, 'learning_rate': 0.02941932373780329, 'max_bin': 223}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:37:36,961] Trial 35 finished with value: 0.9956190743402684 and parameters: {'num_leaves': 175, 'max_depth': 4, 'min_child_samples': 18, 'lambda_l1': 0.0008883435453461586, 'lambda_l2': 0.19495255453953178, 'min_gain_to_split': 0.6133456242430714, 'feature_fraction': 0.975246937685352, 'bagging_fraction': 0.8858886248513176, 'bagging_freq': 4, 'learning_rate': 0.030213368421795873, 'max_bin': 127}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:37:43,573] Trial 33 finished with value: 0.9956409475573956 and parameters: {'num_leaves': 95, 'max_depth': 3, 'min_child_samples': 9, 'lambda_l1': 0.00032804576470198967, 'lambda_l2': 0.026723850514405973, 'min_gain_to_split': 0.7791333557331799, 'feature_fraction': 0.6541977017224149, 'bagging_fraction': 0.7590563270855434, 'bagging_freq': 1, 'learning_rate': 0.023004719746161674, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:38:00,920] Trial 9 finished with value: 0.9954081879664092 and parameters: {'num_leaves': 60, 'max_depth': 8, 'min_child_samples': 9, 'lambda_l1': 2.2424571195781314e-06, 'lambda_l2': 0.00016476190757852137, 'min_gain_to_split': 0.42623275395241733, 'feature_fraction': 0.766871749061885, 'bagging_fraction': 0.5451496276954018, 'bagging_freq': 3, 'learning_rate': 0.01631839864949372, 'max_bin': 159}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:38:36,868] Trial 38 finished with value: 0.9956356148811285 and parameters: {'num_leaves': 36, 'max_depth': 3, 'min_child_samples': 7, 'lambda_l1': 0.000151833732367096, 'lambda_l2': 0.4058817034830009, 'min_gain_to_split': 0.9324339023007762, 'feature_fraction': 0.827840279697759, 'bagging_fraction': 0.8720456555454066, 'bagging_freq': 3, 'learning_rate': 0.021218368111687393, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:38:51,768] Trial 44 finished with value: 0.9956112738213119 and parameters: {'num_leaves': 251, 'max_depth': 4, 'min_child_samples': 12, 'lambda_l1': 5.590482981515547e-08, 'lambda_l2': 3.173429083044305e-06, 'min_gain_to_split': 0.1631546053289697, 'feature_fraction': 0.7101936604024353, 'bagging_fraction': 0.827968357613296, 'bagging_freq': 3, 'learning_rate': 0.07647809955376825, 'max_bin': 63}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:39:07,203] Trial 34 finished with value: 0.995638505216002 and parameters: {'num_leaves': 86, 'max_depth': 4, 'min_child_samples': 9, 'lambda_l1': 9.240130430981063e-05, 'lambda_l2': 1.5797454533593067, 'min_gain_to_split': 0.8189762885384846, 'feature_fraction': 0.9225221852689603, 'bagging_fraction': 0.8615298740793836, 'bagging_freq': 1, 'learning_rate': 0.015830476153469226, 'max_bin': 191}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:40:06,688] Trial 29 finished with value: 0.9956230279861428 and parameters: {'num_leaves': 248, 'max_depth': 11, 'min_child_samples': 30, 'lambda_l1': 0.03402009097711596, 'lambda_l2': 0.4147147208875178, 'min_gain_to_split': 0.8018620141004191, 'feature_fraction': 0.5752237627822289, 'bagging_fraction': 0.9117553322962343, 'bagging_freq': 7, 'learning_rate': 0.013448611372494104, 'max_bin': 95}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:40:12,375] Trial 40 finished with value: 0.9956566649074514 and parameters: {'num_leaves': 86, 'max_depth': 4, 'min_child_samples': 5, 'lambda_l1': 0.02663464090095706, 'lambda_l2': 0.01010547832950847, 'min_gain_to_split': 0.8396551254683601, 'feature_fraction': 0.8229446226606594, 'bagging_fraction': 0.7365012357893111, 'bagging_freq': 2, 'learning_rate': 0.019928452370191963, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:40:18,701] Trial 12 finished with value: 0.9955489491179638 and parameters: {'num_leaves': 61, 'max_depth': 8, 'min_child_samples': 46, 'lambda_l1': 3.2003222761323713e-06, 'lambda_l2': 0.00015099222057593005, 'min_gain_to_split': 0.7277316798355786, 'feature_fraction': 0.930526361812167, 'bagging_fraction': 0.7081052764007847, 'bagging_freq': 4, 'learning_rate': 0.008670791028174397, 'max_bin': 95}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:40:36,214] Trial 43 finished with value: 0.9956019072721827 and parameters: {'num_leaves': 21, 'max_depth': 4, 'min_child_samples': 10, 'lambda_l1': 0.022737548638373652, 'lambda_l2': 2.426629962690583, 'min_gain_to_split': 0.9467568591545423, 'feature_fraction': 0.6544175745365872, 'bagging_fraction': 0.9309291801332682, 'bagging_freq': 2, 'learning_rate': 0.0196557158071682, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:41:09,696] Trial 39 finished with value: 0.995627328906749 and parameters: {'num_leaves': 51, 'max_depth': 4, 'min_child_samples': 10, 'lambda_l1': 0.08575932766745756, 'lambda_l2': 0.2638405213677287, 'min_gain_to_split': 0.894729611988991, 'feature_fraction': 0.7947539655958209, 'bagging_fraction': 0.6212906944314172, 'bagging_freq': 3, 'learning_rate': 0.011624739258513733, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:41:36,018] Trial 47 finished with value: 0.995645956819514 and parameters: {'num_leaves': 95, 'max_depth': 3, 'min_child_samples': 6, 'lambda_l1': 0.0013960598170063613, 'lambda_l2': 0.035849231020539746, 'min_gain_to_split': 0.7512243793185198, 'feature_fraction': 0.5163651325158564, 'bagging_fraction': 0.7074232194948445, 'bagging_freq': 1, 'learning_rate': 0.017548141465371744, 'max_bin': 223}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:41:41,258] Trial 48 finished with value: 0.9956646644625499 and parameters: {'num_leaves': 127, 'max_depth': 3, 'min_child_samples': 13, 'lambda_l1': 0.00015740195100974314, 'lambda_l2': 0.0010573406148057915, 'min_gain_to_split': 0.9342193347904526, 'feature_fraction': 0.568948060390206, 'bagging_fraction': 0.7427385541453286, 'bagging_freq': 2, 'learning_rate': 0.018296906498919865, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:41:44,881] Trial 41 finished with value: 0.9955700081230183 and parameters: {'num_leaves': 216, 'max_depth': 4, 'min_child_samples': 10, 'lambda_l1': 3.707673861022298e-05, 'lambda_l2': 2.6945899667402085, 'min_gain_to_split': 0.6925881857380494, 'feature_fraction': 0.9964910241878071, 'bagging_fraction': 0.7669188654008343, 'bagging_freq': 4, 'learning_rate': 0.013250869522095157, 'max_bin': 95}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:41:49,079] Trial 24 finished with value: 0.9955791951590699 and parameters: {'num_leaves': 135, 'max_depth': 9, 'min_child_samples': 63, 'lambda_l1': 1.026882126971342e-07, 'lambda_l2': 3.086492836327151e-08, 'min_gain_to_split': 0.9604131143614552, 'feature_fraction': 0.538396909819014, 'bagging_fraction': 0.6706899552595833, 'bagging_freq': 5, 'learning_rate': 0.007147697832930898, 'max_bin': 191}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:41:54,558] Trial 49 finished with value: 0.9956465686283529 and parameters: {'num_leaves': 118, 'max_depth': 3, 'min_child_samples': 17, 'lambda_l1': 2.985454169673534e-05, 'lambda_l2': 0.35845161754931415, 'min_gain_to_split': 0.9305767017076033, 'feature_fraction': 0.5717375255992563, 'bagging_fraction': 0.613795002851741, 'bagging_freq': 2, 'learning_rate': 0.02404570871539765, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:42:03,728] Trial 37 finished with value: 0.9956161353651923 and parameters: {'num_leaves': 39, 'max_depth': 5, 'min_child_samples': 8, 'lambda_l1': 0.00045277764307307165, 'lambda_l2': 0.0032561683368737067, 'min_gain_to_split': 0.6412745382950861, 'feature_fraction': 0.7949446844058535, 'bagging_fraction': 0.791432919967867, 'bagging_freq': 1, 'learning_rate': 0.011169386096426951, 'max_bin': 223}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:42:10,066] Trial 42 finished with value: 0.9956241934583195 and parameters: {'num_leaves': 57, 'max_depth': 3, 'min_child_samples': 15, 'lambda_l1': 0.00015083086325649457, 'lambda_l2': 4.837059975919953, 'min_gain_to_split': 0.6694321172225927, 'feature_fraction': 0.6385187055769579, 'bagging_fraction': 0.8128063932966648, 'bagging_freq': 1, 'learning_rate': 0.01161120404536803, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:42:15,937] Trial 22 finished with value: 0.9955025861750144 and parameters: {'num_leaves': 27, 'max_depth': 8, 'min_child_samples': 7, 'lambda_l1': 2.2557261494972876e-05, 'lambda_l2': 4.8356779871171527e-08, 'min_gain_to_split': 0.10845473773908221, 'feature_fraction': 0.7388537732589076, 'bagging_fraction': 0.6170052506634711, 'bagging_freq': 7, 'learning_rate': 0.008483357468458816, 'max_bin': 63}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:42:30,851] Trial 10 finished with value: 0.9954729261798221 and parameters: {'num_leaves': 72, 'max_depth': 12, 'min_child_samples': 8, 'lambda_l1': 0.0066493397303571235, 'lambda_l2': 0.0008077575700217869, 'min_gain_to_split': 0.8632181780342428, 'feature_fraction': 0.8455187902343829, 'bagging_fraction': 0.5289519109581469, 'bagging_freq': 6, 'learning_rate': 0.00805479526442681, 'max_bin': 95}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:42:43,347] Trial 45 finished with value: 0.995625359569873 and parameters: {'num_leaves': 40, 'max_depth': 5, 'min_child_samples': 11, 'lambda_l1': 0.02916061816423291, 'lambda_l2': 2.2046951849591094, 'min_gain_to_split': 0.875573964203656, 'feature_fraction': 0.9832217063932372, 'bagging_fraction': 0.7207201732733017, 'bagging_freq': 1, 'learning_rate': 0.010122726138422817, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:42:54,058] Trial 46 finished with value: 0.9956490987019002 and parameters: {'num_leaves': 50, 'max_depth': 3, 'min_child_samples': 9, 'lambda_l1': 0.00040222918971568857, 'lambda_l2': 0.08894744981895228, 'min_gain_to_split': 0.6608491477900419, 'feature_fraction': 0.518381529915495, 'bagging_fraction': 0.7834812269098982, 'bagging_freq': 2, 'learning_rate': 0.007613503205830335, 'max_bin': 255}. Best is trial 36 with value: 0.9956713950728802.\n",
      "[I 2026-05-05 19:49:26,887] Trial 20 finished with value: 0.9954068552244439 and parameters: {'num_leaves': 137, 'max_depth': 12, 'min_child_samples': 14, 'lambda_l1': 0.0010142807444471253, 'lambda_l2': 8.727529950711935e-05, 'min_gain_to_split': 0.43985086856848343, 'feature_fraction': 0.5097266361720776, 'bagging_fraction': 0.6301814663187999, 'bagging_freq': 5, 'learning_rate': 0.0058931075710505374, 'max_bin': 63}. Best is trial 36 with value: 0.9956713950728802.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "======================================================================\n",
      "Optuna Optimization Summary\n",
      "======================================================================\n",
      "Number of finished trials: 50\n",
      "Best trial: 36\n",
      "Best value: 0.995671\n",
      "\n",
      "Best hyperparameters:\n",
      "  num_leaves: 172\n",
      "  max_depth: 3\n",
      "  min_child_samples: 5\n",
      "  lambda_l1: 0.000001\n",
      "  lambda_l2: 0.000000\n",
      "  min_gain_to_split: 0.215276\n",
      "  feature_fraction: 0.610422\n",
      "  bagging_fraction: 0.882294\n",
      "  bagging_freq: 3\n",
      "  learning_rate: 0.044051\n",
      "  max_bin: 95\n",
      "\n",
      "Additional attributes:\n",
      "  cv_scores: [0.9956319391345688, 0.9971030946093824, 0.9950893041314529, 0.9952388175351814, 0.9952938199538153]\n",
      "  cv_std: 0.0007375413879422372\n",
      "  mean_training_loss: 0.10258775495897347\n",
      "  mean_validation_loss: 0.19172524925699416\n",
      "  best_iteration: 195\n",
      "======================================================================\n",
      "\n",
      "\n",
      "Optimization complete! Best score: 0.9957\n",
      "Time: 1111.1s\n",
      "\n",
      "Training final model with best parameters...\n",
      "[LightGBM] [Warning] feature_fraction is set=0.6104217815601655, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.6104217815601655\n",
      "[LightGBM] [Warning] lambda_l2 is set=1.1182701662877744e-07, reg_lambda=0.0 will be ignored. Current value: lambda_l2=1.1182701662877744e-07\n",
      "[LightGBM] [Warning] min_gain_to_split is set=0.21527568422880633, min_split_gain=0.0 will be ignored. Current value: min_gain_to_split=0.21527568422880633\n",
      "[LightGBM] [Warning] lambda_l1 is set=1.2158568318093246e-06, reg_alpha=0.0 will be ignored. Current value: lambda_l1=1.2158568318093246e-06\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8822936933674879, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8822936933674879\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n",
      "[LightGBM] [Warning] feature_fraction is set=0.6104217815601655, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.6104217815601655\n",
      "[LightGBM] [Warning] lambda_l2 is set=1.1182701662877744e-07, reg_lambda=0.0 will be ignored. Current value: lambda_l2=1.1182701662877744e-07\n",
      "[LightGBM] [Warning] min_gain_to_split is set=0.21527568422880633, min_split_gain=0.0 will be ignored. Current value: min_gain_to_split=0.21527568422880633\n",
      "[LightGBM] [Warning] lambda_l1 is set=1.2158568318093246e-06, reg_alpha=0.0 will be ignored. Current value: lambda_l1=1.2158568318093246e-06\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8822936933674879, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8822936933674879\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n",
      "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.001779 seconds.\n",
      "You can set `force_col_wise=true` to remove the overhead.\n",
      "[LightGBM] [Info] Total Bins 1520\n",
      "[LightGBM] [Info] Number of data points in the train set: 10888, number of used features: 16\n",
      "[LightGBM] [Info] Start training from score -2.332227\n",
      "[LightGBM] [Info] Start training from score -3.259935\n",
      "[LightGBM] [Info] Start training from score -2.122225\n",
      "[LightGBM] [Info] Start training from score -1.344914\n",
      "[LightGBM] [Info] Start training from score -1.954581\n",
      "[LightGBM] [Info] Start training from score -1.904618\n",
      "[LightGBM] [Info] Start training from score -1.641447\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
      "\n",
      "Generating SHAP explanations...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|===================| 3497/3500 [01:05<00:00]        "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SHAP values computed for 500 samples\n",
      "Top 5 SHAP features:\n",
      "  ShapeFactor1: 0.3931\n",
      "  ConvexArea: 0.3822\n",
      "  Perimeter: 0.3613\n",
      "  roundness: 0.3609\n",
      "  MinorAxisLength: 0.3506\n",
      "\n",
      "======================================================================\n",
      "Training Complete!\n",
      "======================================================================\n",
      "=== Training Summary ===\n",
      "Model: LGBMClassifier\n",
      "Best Score: 0.9957\n",
      "CV Score: 0.0000 ± 0.0000\n",
      "Features: 16/16\n",
      "Training Time: 1186.3s\n",
      "Optimization Time: 1111.1s\n",
      "\n",
      "Best Hyperparameters:\n",
      "  num_leaves: 172\n",
      "  max_depth: 3\n",
      "  min_child_samples: 5\n",
      "  lambda_l1: 0.0000\n",
      "  lambda_l2: 0.0000\n",
      "  min_gain_to_split: 0.2153\n",
      "  feature_fraction: 0.6104\n",
      "  bagging_fraction: 0.8823\n",
      "  bagging_freq: 3\n",
      "  learning_rate: 0.0441\n",
      "  max_bin: 95\n",
      "  n_estimators: 195\n",
      "======================================================================\n",
      "\n"
     ]
    }
   ],
   "source": [
    "config = TrainConfig(\n",
    "    name=\"dry_bean_lgbm\",\n",
    "    model_type=\"lgbm\",\n",
    "    task=\"multiclass_classification\",\n",
    "    optimization_metric=\"roc_auc\",\n",
    "    optuna_sampler=\"tpe\",\n",
    "    n_trials=50,\n",
    "    cv_folds=5,\n",
    "    optimize_threshold=False,\n",
    "    save_feature_importance=True,\n",
    "    generate_shap=True,\n",
    "    shap_n_samples=500,\n",
    "    shap_n_background=100,\n",
    "    verbose=True,\n",
    "    optuna_show_progress=False,\n",
    "    random_state=42\n",
    ")\n",
    "\n",
    "trainer = OptunaTrainer(config=config)\n",
    "trainer.fit(X_train_transformed, y_train_encoded)\n",
    "state = trainer.state"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mdc0c8fdcf",
   "metadata": {},
   "source": [
    "## 8. Unpacking the Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "cd1b74c1b8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== Training Summary ===\n",
      "Model: LGBMClassifier\n",
      "Best Score: 0.9957\n",
      "CV Score: 0.0000 ± 0.0000\n",
      "Features: 16/16\n",
      "Training Time: 1186.3s\n",
      "Optimization Time: 1111.1s\n",
      "\n",
      "Best Hyperparameters:\n",
      "  num_leaves: 172\n",
      "  max_depth: 3\n",
      "  min_child_samples: 5\n",
      "  lambda_l1: 0.0000\n",
      "  lambda_l2: 0.0000\n",
      "  min_gain_to_split: 0.2153\n",
      "  feature_fraction: 0.6104\n",
      "  bagging_fraction: 0.8823\n",
      "  bagging_freq: 3\n",
      "  learning_rate: 0.0441\n",
      "  max_bin: 95\n",
      "  n_estimators: 195\n",
      "\n",
      "Best trial — per-fold AUC : [0.9956, 0.9971, 0.9951, 0.9952, 0.9953]\n",
      "Mean ± Std              : 0.9957 ± 0.0007\n",
      "\n",
      "Training state snapshot:\n",
      "  Model               : LGBMClassifierWrapper\n",
      "  SHAP values         : Available\n",
      "  Training time       : 1186.3s\n"
     ]
    }
   ],
   "source": [
    "print(state.summary())\n",
    "\n",
    "fold_scores = state.study.best_trial.user_attrs.get('cv_scores', [])\n",
    "fold_std    = state.study.best_trial.user_attrs.get('cv_std', 0.0)\n",
    "\n",
    "if fold_scores:\n",
    "    print(f\"\\nBest trial — per-fold AUC : {[round(s, 4) for s in fold_scores]}\")\n",
    "    print(f\"Mean \\u00b1 Std              : {np.mean(fold_scores):.4f} \\u00b1 {fold_std:.4f}\")\n",
    "\n",
    "print(f\"\\nTraining state snapshot:\")\n",
    "print(f\"  Model               : {type(state.model).__name__}\")\n",
    "print(f\"  SHAP values         : {'Available' if state.shap_values is not None else 'Not computed'}\")\n",
    "print(f\"  Training time       : {state.training_time_seconds:.1f}s\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "cd1e35fe04",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA90AAAJOCAYAAACqS2TfAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjYsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvq6yFwwAAAAlwSFlzAAAPYQAAD2EBqD+naQAApzpJREFUeJzs3XmcTvX///HnmTGuWa7ZzGAsw5jP2HcNZfmQkC0hsoTMx1YkZUmRbRIiEi2UhpGUbKkIITRRoUyU+egTDT5fMolmbM0M1/n90c/5uMyOq2vkcb/dzu12nXPe7/d5nXO9h+t13mcxTNM0BQAAAAAAbjoPdwcAAAAAAMDfFUk3AAAAAAAuQtINAAAAAICLkHQDAAAAAOAiJN0AAAAAALgISTcAAAAAAC5C0g0AAAAAgIuQdAMAAAAA4CIk3QAAAAAAuAhJNwDAcvfdd8swDBmGofj4eHeHAwCSpN9++03BwcEyDEO9evVydziFzrhx42QYhry8vPTjjz+6OxwA1yDpBgA3io+Pt5JcwzDUunXrLGXCwsJuWiI8adIka/r9999vqC1Xu/oEQH6mwnySYNu2bXnGHxQU9JfGdCv1hexMmjTJOnYRERHuDsfl4uPjre8rMTHR3eH85WJjY/X777/LMAyNGTPGaV1hPzb79u3Tk08+qcaNG8vX1zff/dbhcGjJkiVq1aqVQkNDVbRoUZUsWVKNGjXS1KlTnco++eST8vX11aVLlzR69GgX7g2A61HE3QEAAP7n008/1eeff66mTZu6pP3Y2Fjrc0xMTJZE75VXXlFqaqokqVKlSi6JAYVDXn0BhUt8fLy2b98uSYqIiFCdOnXcG9Bf6OTJk3rjjTck/XkyrkaNGk7rC/ux+eyzzzRnzpwC1Tl37py6dOmiTz/91Gl5SkqKUlJSdPz4cY0dO9ZaHhoaqh49emjhwoX68MMP9d1336l27do3JX4AN46kGwAKmWeffVYJCQlu2XbNmjXdst3sXH0CQJLWr1/vNLpz7THK6SSBw+FQenq6fHx8XBPodVixYoXCwsKclhUp8vf9L7kwfge3Ao7bn+Li4pSRkSFJeuihh9wcTcEFBQWpTZs2io6O1qlTpzR//vw86zz22GNWwl2+fHkNGzZM1atXV2Zmpg4dOqSDBw9mqdOzZ08tXLhQkjR//nzNmzfv5u4IgOtnAgDcZtGiRaakLNP69eutMiVLlrSWL1q0yFq+e/dus3fv3mbNmjXN4sWLm15eXqafn59ZvXp1c+TIkeavv/5qle3bt2+227m23WbNmmVZ1qNHD2vZ5MmTneJPT083g4KCrPXfffedte7AgQNm//79zQoVKpg2m8309/c3GzVqZC5atMh0OBw3fKyudfW67777zhw2bJhZunRp08PDw/zggw+schs3bjTvv/9+s2TJkqaXl5cZGhpqdujQwfz888+z3e7N2I+tW7c6xffzzz/nWefYsWPmE088YVauXNn09vY2/fz8zHr16pkvvfSSmZGR4VTWFX2hfPny1rKtW7da9X/++eccvwdXfQfZmThxorWt8uXL5xrHo48+aoaGhpp2u9287777zJ9//tl0OBzmzJkzzcjISLNo0aJm5cqVzbffftupnau/t/Lly5vJyclmz549zWLFipk+Pj7mP//5T/OLL77IEltGRob5yiuvmI0aNTIDAwNNLy8vs3Tp0maXLl3MhISELOVzO26zZ8/O9fvq27evaZoF6wOm6fz31KxZM3P//v3m/fffbwYEBJi+vr5mmzZtzB9//DFLrGlpaebUqVPN+vXrmwEBAWbRokXNsmXLmp06dTL37t3rVPZm/RtQo0YNK9bjx49nuw+5HRvTNM2zZ8+aU6ZMMe+44w7TbrebRYsWNcuXL28+/PDD5r59+5y2d20fP3XqlPnII4+YJUuWNG02m3nHHXeYH374Yb7jv9rVMV/bb6/44YcfTMMwTElmiRIlzBMnTuSr7UuXLpm+vr6mJLNYsWLm5cuXrytGADcfSTcAuNHVP8CCgoLMSpUqmZLM6Ohoq0xOSfe8efNy/cFZoUIF88yZM6Zp3ljSvXnzZmtZlSpVnOL/4IMPrHX169d3Wu7t7Z3j9nr16lXgxLsgSXfFihWd5q8kfE8//XSOMXl4eJjz5s3Lsn83Yz8KmnR/+eWXTiczrp2aN29u/vHHH1Z5V/SFG026b9Z3kJP8Jt1X/qauPR6PPPJItjFcnURf/b0FBQWZpUqVylK+aNGi5rZt26w658+fNxs3bpzjPhqGYc6ePTvfxy2/SXdB+oBpOv89lSpVyvTz88tSp0qVKualS5esOsnJyWZkZGSefcc0b97fzq+//moloOXKlXNal9+k++TJk2blypVzLFe0aFFz+fLlVrvX9vGqVatm+z0uXbo0z/ivlZ+ke/LkyVaZ1q1bm7179zbLlClj2mw2s0qVKuaUKVOynHi74p///KdV99qTIADch6QbANzo6h9gISEh5nvvvWfNr1692jTNnJPuHTt2mDNnzjRXr15tbtq0ydy2bZv5wQcfmK1bt7bKv/jii6ZpmuaPP/5oJiQkOP1oXLFihZmQkGAmJCSYJ0+eNE0z+6Tb4XCYFSpUsJZ/8803Vgxdu3a1ls+fP980TdNMSUkx7Xa7tfzRRx81N2zYYL799ttmuXLlrOVxcXHXfayuTfZM0zlx8fDwMEeOHGmuX7/eXLJkiblv3z7zk08+sdb7+PiYM2bMMDdt2mTOnDnTtNlspiSzSJEi5sGDB2/6flybdGc3PfHEE6ZpmuYff/zh1H6XLl3MdevWmStWrDBr1aplLR8/frxL+8KNJt034zvITX6T7uDgYDMuLs585513rFHAK9PQoUPNTz75xLzrrrusZd27d8/xe6tataq5fPlyc9myZU7JcaVKlawEcuTIkdZyPz8/86WXXjLXrVtn9urVy+nYXH1VSG7Hbdu2bWZCQoJZp04dq8zYsWOt7+vKaHRB+oBpZv17qlevnrlq1Srz5ZdfNr28vKzlV19106hRI2u53W43Y2NjzQ0bNphLly41e/fubb7zzjumabrub6dly5ZO606ePJmvY9OlSxdrffHixc0FCxaYH330kdOx8fPzM3/55RfTNLP28dKlS5uLFy82P/jgA/POO++0lgcFBZnnzp3Lcx+ulp+k+8EHH8zz34t7773X6YTIFQMGDLDKLF68uECxAXAdkm4AcKNrk26Hw2HWrFnTlGRWr17dvHz5co5Jd2ZmpvnKK6+YTZo0MYsVK2Z6enpm+WH2wAMPOG3v6nXZjbZml3Sbpmk+99xz1vIRI0aYpmmaqamp1kiWr6+vmZqaapqmab7yyitW2Ro1alg/gBMSEsxnn33WWnfXXXdd97HKK+m+EuPVrv7h3adPH6e42rVrZ6175plnbvp+FCTp/vjjj50ShM8//9za7tUxlSpVymrfFX3hRpPum/Ed5Ca/Sffrr79uLW/btq21vEGDBtbyFStWOCWfV1z7vX3//ffWuj179jit+/bbb02Hw2GGhoZay6ZPn26VdzgcZvXq1a11I0eOzPdxM82c/zavKGgfuPrvycvLy/y///s/a93VyejcuXNN0zTN77//3qmtVatW5fTV3NS/neXLl1vlrz4hkt9jc+bMGadj8f7771vrzp8/7/R9vfLKK6ZpZu3j69ats+ocP37cLFq0qLXuysnR/MpP0t2yZUun7Xfq1Mn85JNPzKlTp5oeHh7W8vj4+Cx1R48eba2fMWNGgWID4Dp/36e2AMAtyDAMPffcc+rcubN++OEHLVu2LMey/fr105IlS3Jt78yZMzclrn/961+aNGmSHA6Hli1bphdffFGrVq3SH3/8IUnq2rWrAgICJEkHDhyw6n3//ff65z//mW2bP/zww02JLTtdunTJsuzquJYsWZLjsbsSlyv3I7sHqZUpUybLdn/99dccn2R/4sQJnT59WsWKFftL+0J+3Yzv4GZo2LCh9Tk0NDTP5adPn862neDgYFWvXt2av+OOO+Tj46OLFy9Kkv7zn/+oTJkyOnXqlFWmSZMm1mfDMPTPf/7T2rekpKRst5PdccuPG+kDVapUUenSpa357I7H1d9dkSJF1KFDhxy3czP/dkzTzPZzfv3444+6fPmyNX/1d+Lr66vo6Ght2LBBUs7fydV1SpUqpcjISP373/+W9Of3frN5e3tbn728vPT222/L399fbdu21ZdffqmPP/5YkvTJJ5+ob9++TnWvPkaGYdz02ABcH5JuAChkOnXqpPr162v37t2aOHGiLl26lKXM//3f/zn9wB42bJjat28vX19fffzxx5oxY4akP59+fDOULVtW9957rzZs2KDjx49r69atWrp0qbV+wIABBW7z7NmzNyW27JQqVeq66xY0ruvZj+jo6JvybumzZ8/q4sWLLukLV/9gv7oP/vrrr/mq/1d+B7m5+lVoHh4e2S6/2vUkdvmVn7av57jd6L8HxYoVc5q/+kn6V2J2RTKXn++5RIkS1uecTojcCFd+39erfPny1ueQkBD5+/tb8//4xz+sz7///nuWulcfo+LFi7smQAAF5pF3EQDAX23y5MmSpJ9++km//fZblvXHjh2zPhcrVkxz5szRvffeqyZNmujkyZM5tnv1j+WCJmFXJ9YzZ87U1q1bJUkVK1Z0GsmqWrWq9blRo0Yy/7yVKcuUlpZWoO0XRHZJwdVxjRkzJtuYMjMz9cknn7h1P67ebrly5ZSZmZnjdsuXL++yvhAcHGx9/u9//2t9vjLKlpeb8R0UJmfOnHEaCf3222+tUW7pz7+D4sWLO40S79ixw/psmqbTfJUqVbLdTk4J7dUnDK79vq63DxTE1aP8mZmZWrt2bZYyVxLYm/m3U6NGDeuYZPeaLCn3Y1OpUiV5enpa81d/BxcvXtS3335rzef0nVxd55dfftHhw4et+YoVK+a5DwV19913W59/++03nTt3zpq/etvZnbi7+hgVpldAArc7RroBoBBq3bq1/vnPf+b4vu7IyEjr8+nTp/Xcc8/prrvu0saNG/X222/n2G5ISIh1+ev8+fN13333ycPDQw0aNFDRokVzjen+++9X8eLF9euvv1qXY0pS//79ncp1795dY8eO1blz57Rz50517dpVDz30kAIDA/V///d/SkpK0urVq9WzZ09NmjQpr0Nx0/Tv31+rV6+WJL344ou6fPmymjVrJg8PDx07dkx79uzR6tWrtWLFCt19991u249WrVopPDxcx44d09GjR9W6dWsNHDhQJUqU0C+//KL//Oc/WrNmjerUqaNFixa5rC9UqlRJe/fulSSNGzdOZ8+e1c8//6y5c+de974V9DsobB588EHrNosJEyZYyytXrqw6derIMAz17dtXs2bNkiQ999xz1rF877339P3330uSVa4gQkJCrM8rVqxQRESEihYtqsqVK193HyiI6tWrq2HDhvryyy8lSX379tUPP/yg+vXr68yZM9q4caNatmypXr163dS/ndDQUFWvXl3ff/+9jh07phMnTmS5GiC3Y1O8eHF16tRJq1atkiQ9/vjjOnv2rEqUKKHXX3/dunLDz89P3bp1yzaGQYMGaerUqQoICND06dOtd4YHBwfr3nvvzfPYHTlyRLt375Yk7dmzx1p+4cIFrVy50tp+27ZtJf35b21ERISSk5OVmZmpvn37auDAgfruu++0bt06q/617yy/fPmydRIhMDBQderUyTM2AH8RF90rDgDIh2sfpHa17du3Z3kQ0tUPCXrooYeyfSBX8+bNrc/NmjVzarNnz57Z1jl27Jhpmnk/rGnEiBFO9YoUKZLtO2RXr16d6+uCJJkTJ0687mOV3X9fV6/L6ZVcVz9kKKfp6oeG3az9KOgrw3bu3JnrK8Mk53cQu6IvfPHFF9muv/Kgv+y+B1d8BznJ74PUro7j6telXf29Xfs+7uyWFytWzOnhclemokWLOsV7o68My+m4vfHGG9m2t2TJEtM0C94Hrn1P99VyOk6HDx/O9hhk92/Gzfw34Pnnn7fqLFiwoMDH5pdffrmhV4bVrVs32+/xytPa85LXq82y68NffvmlGRAQkGP5p59+Ost2Nm3aZK0fOHBgvmID8Nfg8nIAKKSaNm2qVq1a5bh+wYIFGjVqlMqXLy9vb2/VqVNH77//vh5++OEc68yZM0fdu3dXsWLFruu+zGtHtdu1a5flgWCS1LlzZ+3du1eDBg1SVFSUvL295efnp6ioKN13332aP3++hgwZUuDt36jp06fr008/VefOnVWqVCl5eXkpODhY1apV08MPP6yVK1fqrrvucvt+NGzYUPv379eIESNUvXp1+fr6ysfHRxUqVFCrVq00e/ZsPffcc1Z5V/SFxo0ba/HixapcubK8vLxUrlw5TZw4UStWrLihfSvod1BY+Pv7a+fOnerdu7eKFSsmb29v/fOf/9SWLVucRuV9fX21detWzZ07Vw0bNlRAQICKFCmisLAwdenSRZ9//rmefPLJAm+/f//+GjNmjMqWLet0OfUV19MHCqpChQrat2+fJk+erDvuuEP+/v7y8vJSmTJldP/99zuNrN7Mv50BAwbIy8tLkvTuu+9mWZ/XsSlZsqT27Nmj559/XvXq1ZOfn5+8vLwUHh6uPn36aPfu3XrwwQdz3P7WrVv12GOPKSwsTDabTfXq1dMHH3ygXr165Sv+63HXXXdp79696tevn8qWLSsvLy8FBQWpZcuW+vDDD/XCCy9kqfPee+9ZnwcPHuyy2AAUnGGahfAJEgAAAG62bds2NW/eXNKfD7dKTk52b0C3sWHDhumVV16RYRjav3+/0z3mN1tycrIqVKhgzd8KP5VPnTql8uXL68KFC7r//vv14YcfujskAFdhpBsAAACF2sSJExUUFCTTNDVlyhR3h1PozJkzRxcuXFCRIkWsp9UDKDwY6QYAAMgGI923p1txpBtA4cZINwAAAAAALsJINwAAAAAALsJINwAAAAAALkLSDQAAAACAixRxdwAonBwOh44fPy5/f//repcvAAAAAPydmaaps2fPqnTp0vLwyHk8m6Qb2Tp+/LjCw8PdHQYAAAAAFGrHjh1T2bJlc1xP0o1s+fv7S5KOHDmioKAg9wYD5IPD4dCvv/6q4sWL53qmESgM6K+41dBncSuhv+KvkpaWpvDwcCt3yglJN7J15ZLygIAABQQEuDkaIG8Oh0N//PGHAgIC+A8WhR79Fbca+ixuJfRX/NXyuh2XXggAAAAAgIuQdAMAAAAA4CIk3QAAAAAAuAhJNwAAAAAALkLSDQAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICLkHQDAAAAAOAiJN0AAAAAALgISTcAAAAAAC5C0g0AAAAAgIuQdAMAAAAA4CIk3QAAAAAAuAhJNwAAAAAALkLSDQAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuUsTdAaBwO3baodRLDneHAeTJNB1KS3PoD0+HDMPd0QC5o7/iVkOfxa2E/vr3YveWQuy39lgxSTdyNSguXZ7e6e4OA8iTIYci/TN1+Gy6TC7iQSFHf8Wthj6LWwn99e9l6WCbQuzujuLG0AsBAAAAAHARkm4AAAAAAFyEpBsAAAAAABch6QYAAAAAwEVIuv8i27Ztk2EY+v33390dCgAAAADgL0LSDQAAAACAi9xWSXdGRoa7QwAAAAAA3Eb+1kn33XffraFDh2rEiBEKDQ1Vq1attH37djVo0EA2m02lSpXSM888o0uXLll1IiIi9PLLLzu1U6dOHU2aNMmaNwxDb731ljp37ixfX19VrFhRH330kVOdTz75RJUqVZKPj4+aN2+u5ORkp/Xx8fEKCgrSxo0bVbVqVdntdrVp00YnTpxwKrdo0SJVrVpV3t7eqlKlil5//XVrXUZGhoYOHapSpUrJ29tbERERmjZtmrV+0qRJKleunGw2m0qXLq1hw4Zd55EEAAAAAFyPv3XSLUmLFy9WkSJFtGPHDk2dOlXt2rVT/fr19d1332nevHmKi4vT888/X+B2Y2Nj1a1bN+3bt0/t2rVTr169dPr0aUnSsWPH9MADD6hdu3ZKTEzUgAED9Mwzz2Rp48KFC5o5c6aWLFmizz//XEePHtWoUaOs9QsWLNCzzz6rKVOmKCkpSVOnTtX48eO1ePFiSdLcuXP10Ucfafny5Tp48KDeeecdRURESJJWrlyp2bNn64033tB//vMfrVmzRjVr1ryOIwgAAAAAuF5F3B2Aq0VFRWnGjBmSpLffflvh4eF69dVXZRiGqlSpouPHj+vpp5/WhAkT5OGR/3MQMTEx6tmzpyRp6tSpeuWVV7Rr1y61adNG8+bNU2RkpGbPni3DMFS5cmXt379f06dPd2ojMzNT8+fP1z/+8Q9J0tChQ/Xcc89Z6ydPnqxZs2bpgQcekCRVqFBBBw4c0BtvvKG+ffvq6NGjqlixopo0aSLDMFS+fHmr7tGjRxUWFqaWLVvKy8tL5cqVU4MGDXLcn/T0dKWnp1vzaWlp+T4WAAAAAIDs/e1HuqOjo63PSUlJatiwoQzDsJY1btxY586d03//+98CtVurVi3rs5+fn/z9/ZWSkmJt56677nLaTsOGDbO04evrayXcklSqVCmrjV9//VXHjh1T//79Zbfbren555/XoUOHJP2Z+CcmJqpy5coaNmyYPv30U6utBx98UBcvXlRkZKQGDhyoDz74wOky+mtNmzZNgYGB1hQeHl6g4wEAAAAAyOpvn3T7+flZn03TdEqEryyTZC338PCwll2RmZmZpV0vLy+necMw5HA4nNrMS3ZtXKl7pa0FCxYoMTHRmr7//nt99dVXkqR69erp559/1uTJk3Xx4kV169ZNXbt2lSSFh4fr4MGDeu211+Tj46MhQ4aoadOm2e6LJI0ZM0apqanWdOzYsXztAwAAAAAgZ3/7y8uvVq1aNa1atcop+d65c6f8/f1VpkwZSVLx4sWdHmaWlpamn3/+ucDbWbNmjdOyK4lyfpUsWVJlypTR4cOH1atXrxzLBQQEqHv37urevbu6du2qNm3a6PTp0ypWrJh8fHx0//336/7779djjz2mKlWqaP/+/apXr16Wdmw2m2w2W4FiBAAAAADk7rZKuocMGaKXX35Zjz/+uIYOHaqDBw9q4sSJGjFihHU/9z333KP4+Hh16NBBwcHBGj9+vDw9PQu0nUcffVSzZs3SiBEj9Mgjj+ibb75RfHx8geOdNGmShg0bpoCAALVt21bp6enas2ePzpw5oxEjRmj27NkqVaqU6tSpIw8PD61YsUJhYWEKCgpSfHy8Ll++rDvvvFO+vr5asmSJfHx8nO77BgAAAAC41t/+8vKrlSlTRp988ol27dql2rVr69FHH1X//v01btw4q8yYMWPUtGlT3XfffWrXrp06derkdN91fpQrV06rVq3Sxx9/rNq1a2v+/PmaOnVqgeMdMGCA3nrrLcXHx6tmzZpq1qyZ4uPjVaFCBUmS3W7X9OnTFR0drfr16ys5OVmffPKJPDw8FBQUpAULFqhx48aqVauWtmzZoo8//lghISEFjgMAAAAAcH0MM783IOO2kpaWpsDAQN019rg8vYPcHQ6QJ0MORfqf0uGzoTJvr/OJuAXRX3Groc/iVkJ//XtZOtim8qGF83u8kjOlpqYqICAgx3KFM3oAAAAAAP4GSLoBAAAAAHARkm4AAAAAAFyEpBsAAAAAABch6QYAAAAAwEVuq/d0o+De7G9TQIDN3WEAeTJNh9LOeCkg2CbD4HwiCjf6K2419FncSuivfy92b3dHcONIupGr8GIeCgriHysUfg6HlHLZQyVCPOThQZ9F4UZ/xa2GPotbCf0VhQ29EAAAAAAAFyHpBgAAAADARUi6AQAAAABwEZJuAAAAAABchAepIVfHTjuUesnh7jCAPJmmQ2lpDv3h6ZBhuDsaIHf0V9xq6LO4lfxd+qvdWwqxM0b6d0DSjVwNikuXp3e6u8MA8mTIoUj/TB0+my6Ti3hQyNFfcauhz+JW8nfpr0sH2xRid3cUuBlu3V4IAAAAAEAhR9INAAAAAICLkHQDAAAAAOAiJN0AAAAAALhIoU+6DcPQmjVr3B0GAAAAAAAF5vakOyUlRY888ojKlSsnm82msLAwtW7dWl9++aW7Q3OSnJwswzCyTL17977htrdt2ybDMPT777/feKDX2LFjh4oUKaI6derc9LYBAAAAALlz+yvDunTposzMTC1evFiRkZE6efKktmzZotOnT7s7tGxt3rxZ1atXt+Z9fHzcGI0z0zR1+fJlFSny59eampqqhx9+WC1atNDJkyfdHB0AAAAA3H7cOtL9+++/64svvtD06dPVvHlzlS9fXg0aNNCYMWPUvn17q9ypU6fUuXNn+fr6qmLFivroo4+sdZcvX1b//v1VoUIF+fj4qHLlypozZ47TdmJiYtSpUyfFxsaqRIkSCggI0COPPKKMjAyrjGmamjFjhiIjI+Xj46PatWtr5cqVWWIOCQlRWFiYNQUGBurQoUPq2LGjSpYsKbvdrvr162vz5s1O9dLT0zV69GiFh4fLZrOpYsWKiouLU3Jyspo3by5JCg4OlmEYiomJseoMGzZMJUqUkLe3t5o0aaLdu3dbbV4ZId+4caOio6Nls9mUkJBgrX/kkUf00EMPqWHDhtfx7QAAAAAAbpRbk2673S673a41a9YoPT09x3KxsbHq1q2b9u3bp3bt2qlXr17WSLjD4VDZsmW1fPlyHThwQBMmTNDYsWO1fPlypza2bNmipKQkbd26Ve+9954++OADxcbGWuvHjRunRYsWad68efrhhx80fPhw9e7dW9u3b89zP86dO6d27dpp8+bN2rt3r1q3bq0OHTro6NGjVpmHH35Yy5Yt09y5c5WUlKT58+fLbrcrPDxcq1atkiQdPHhQJ06csE4ajB49WqtWrdLixYv17bffKioqSq1bt85yFcDo0aM1bdo0JSUlqVatWpKkRYsW6dChQ5o4cWKe8QMAAAAAXMMwTdN0ZwCrVq3SwIEDdfHiRdWrV0/NmjVTjx49rOTRMAyNGzdOkydPliSdP39e/v7++uSTT9SmTZts23zsscd08uRJa6Q6JiZGH3/8sY4dOyZfX19J0vz58/XUU08pNTVVFy9eVGhoqD777DOnUeEBAwbowoULevfdd5WcnGyNpnt4/O9cRUJCgurWrZslhurVq2vw4MEaOnSofvzxR1WuXFmbNm1Sy5Yts5Tdtm2bmjdvrjNnzigoKMjaz+DgYMXHx+uhhx6SJGVmZioiIkJPPvmknnrqKavemjVr1LFjR6u9//znP2rSpIkSEhJUqVIlTZo0SWvWrFFiYmKO30N6errTiY+0tDSFh4frrrHH5ekdlGM9oLAw5FCk/ykdPhsq0/2PqwByRX/FrYY+i1vJ36W/Lh1sU/nQWzf+20FaWpoCAwOVmpqqgICAHMsVinu627dvr4SEBH355ZfasGGDZsyYobfeesu6zPpKAi5Jfn5+8vf3V0pKirVs/vz5euutt3TkyBFdvHhRGRkZWR4cVrt2bSvhlqSGDRvq3LlzOnbsmFJSUvTHH3+oVatWTnUyMjKyJNTvv/++qlatas2Hh4fr/Pnzio2N1dq1a3X8+HFdunRJFy9etEa6ExMT5enpqWbNmuX7uBw6dEiZmZlq3LixtczLy0sNGjRQUlKSU9no6Gjr8+XLl/XQQw8pNjZWlSpVyvf2pk2b5jTyDwAAAAC4cW5PuiXJ29tbrVq1UqtWrTRhwgQNGDBAEydOtJJuLy8vp/KGYcjhcEiSli9fruHDh2vWrFlq2LCh/P399eKLL+rrr7/O17avbmvdunUqU6aM03qbzeY0Hx4erqioKKdlw4cP18aNGzVz5kxFRUXJx8dHXbt2te4Zv56HrV25AMEwjCzLr13m5+dnfT579qz27NmjvXv3aujQoZL+vATfNE0VKVJEn376qe65554s2xszZoxGjBhhzV8Z6QYAAAAAXL9CkXRfq1q1avl+N3dCQoIaNWqkIUOGWMsOHTqUpdx3332nixcvWgnwV199JbvdrrJlyyo4OFg2m01Hjx4t0Gj01THExMSoc+fOkv68xzs5OdlaX7NmTTkcDm3fvj3by8uLFi0q6c9R6iuioqJUtGhRffHFF06Xl+/Zs0dPPvlkjrEEBARo//79Tstef/11ffbZZ1q5cqUqVKiQbT2bzZblBAMAAAAA4Ma4Nen+7bff9OCDD6pfv36qVauW/P39tWfPHs2YMcPpHuXcREVF6e2339bGjRtVoUIFLVmyRLt3786SXGZkZKh///4aN26cjhw5ookTJ2ro0KHy8PCQv7+/Ro0apeHDh8vhcKhJkyZKS0vTzp07Zbfb1bdv3zxjWL16tTp06CDDMDR+/Hhr9FySIiIi1LdvX/Xr109z585V7dq1deTIEaWkpKhbt24qX768DMPQ2rVr1a5dO/n4+Mhut2vw4MF66qmnVKxYMZUrV04zZszQhQsX1L9//xxj8fDwUI0aNZyWXXn6+bXLAQAAAACu5dak2263684779Ts2bOte5jDw8M1cOBAjR07Nl9tPProo0pMTFT37t1lGIZ69uypIUOGaP369U7lWrRooYoVK6pp06ZKT09Xjx49NGnSJGv95MmTVaJECU2bNk2HDx9WUFCQ6tWrl684Zs+erX79+qlRo0YKDQ3V008/rbS0NKcy8+bN09ixYzVkyBD99ttvKleunNV2mTJlFBsbq2eeeUb/+te/9PDDDys+Pl4vvPCCHA6H+vTpo7Nnzyo6OlobN25UcHBwvo4NAAAAAMC93P708r9CTEyMfv/993xfso7/PYmPp5fjVvF3eVIpbg/0V9xq6LO4lfxd+itPLy/88vv0cr5FAAAAAABchKQbAAAAAAAXKZRPL7/Z4uPj3R0CAAAAAOA2xEg3AAAAAAAuQtINAAAAAICL3BaXl+P6vdnfpoAAm7vDAPJkmg6lnfFSQLBNhsH5RBRu9FfcauizuJX8Xfqr3dvdEeBmIelGrsKLeSgo6Nb9xwq3D4dDSrnsoRIhHvLwoM+icKO/4lZDn8WthP6KwoZeCAAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuwoPUkKtjpx1KveRwdxhAnkzTobQ0h/7wdMgw3B0NkDv6K2419Nk/nyQdYme8CkDBkXQjV4Pi0uXpne7uMIA8GXIo0j9Th8+my+QiHhRy9Ffcauiz0tLBNoXY3R0FgFvR7fmvJgAAAAAAfwGSbgAAAAAAXISkGwAAAAAAFyHpBgAAAADARUi6/2KTJk1SnTp1rPmYmBh16tQp1zp33323nnzySWs+IiJCL7/8skviAwAAAADcPCTdBZSSkqJHHnlE5cqVk81mU1hYmFq3bq0vv/zyutqbM2eO4uPjC1Rn9+7dGjRokDVvGIbWrFlzXdsHAAAAALgOrwwroC5duigzM1OLFy9WZGSkTp48qS1btuj06dPX1V5gYGCB6xQvXvy6tgUAAAAA+Gsx0l0Av//+u7744gtNnz5dzZs3V/ny5dWgQQONGTNG7du3lyQdPXpUHTt2lN1uV0BAgLp166aTJ0/m2Oa1l5efP39eDz/8sOx2u0qVKqVZs2ZlqXP15eURERGSpM6dO8swDEVERCg5OVkeHh7as2ePU71XXnlF5cuXl2maN3YgAAAAAAD5QtJdAHa7XXa7XWvWrFF6enqW9aZpqlOnTjp9+rS2b9+uTZs26dChQ+revXu+t/HUU09p69at+uCDD/Tpp59q27Zt+uabb3Isv3v3bknSokWLdOLECe3evVsRERFq2bKlFi1a5FR20aJFiomJkWEY+Y4HAAAAAHD9SLoLoEiRIoqPj9fixYsVFBSkxo0ba+zYsdq3b58kafPmzdq3b5/effdd3XHHHbrzzju1ZMkSbd++3UqOc3Pu3DnFxcVp5syZatWqlWrWrKnFixfr8uXLOda5cql5UFCQwsLCrPkBAwbovffes04OfPfdd0pMTNS//vWvbNtJT09XWlqa0wQAAAAAuDEk3QXUpUsXHT9+XB999JFat26tbdu2qV69eoqPj1dSUpLCw8MVHh5ula9WrZqCgoKUlJSUZ9uHDh1SRkaGGjZsaC0rVqyYKleuXOA4O3XqpCJFiuiDDz6QJC1cuFDNmze3Lke/1rRp0xQYGGhNV+8DAAAAAOD6kHRfB29vb7Vq1UoTJkzQzp07FRMTo4kTJ8o0zWwv3c5peXblbpaiRYuqT58+WrRokTIyMvTuu++qX79+OZYfM2aMUlNTrenYsWM3LRYAAAAAuF2RdN8E1apV0/nz51WtWjUdPXrUKWE9cOCAUlNTVbVq1TzbiYqKkpeXl7766itr2ZkzZ/Tjjz/mWs/LyyvbS9AHDBigzZs36/XXX1dmZqYeeOCBHNuw2WwKCAhwmgAAAAAAN4akuwB+++033XPPPXrnnXe0b98+/fzzz1qxYoVmzJihjh07qmXLlqpVq5Z69eqlb7/9Vrt27dLDDz+sZs2aKTo6Os/27Xa7+vfvr6eeekpbtmzR999/r5iYGHl45P41RUREaMuWLfrll1905swZa3nVqlV111136emnn1bPnj3l4+Nzw8cAAAAAAJB/JN0FYLfbdeedd2r27Nlq2rSpatSoofHjx2vgwIF69dVXZRiG1qxZo+DgYDVt2lQtW7ZUZGSk3n///Xxv48UXX1TTpk11//33q2XLlmrSpInuuOOOXOvMmjVLmzZtUnh4uOrWreu0rn///srIyMj10nIAAAAAgGsYJi9t/lubMmWKli1bpv379xeoXlpamgIDA3XX2OPy9A5yTXDATWTIoUj/Uzp8NlQm5xNRyNFfcauhz0pLB9tUPvT23PdbjcPhUEpKikqUKJHnFaPAjbiSM6WmpuZ6ey698G/q3Llz2r17t1555RUNGzbM3eEAAAAAwG2JpPtvaujQoWrSpImaNWvGpeUAAAAA4CZF3B0AXCM+Pl7x8fHuDgMAAAAAbmuMdAMAAAAA4CIk3QAAAAAAuAhJNwAAAAAALsI93cjVm/1tCgiwuTsMIE+m6VDaGS8FBNtkGJxPROFGf8Wthj4r2b3dHQGAWxVJN3IVXsxDQUG353+uuLU4HFLKZQ+VCPHgnZwo9OivuNXQZwHg+vGvJgAAAAAALkLSDQAAAACAi5B0AwAAAADgItzTjVwdO+1Q6iWHu8MA8mSaDqWlOfSHp0OG4e5ogNzRX3GryU+ftXtLIXbGcwDgWiTdyNWguHR5eqe7OwwgT4YcivTP1OGz6TK5iAeFHP0Vt5r89Nmlg20Ksf/FgQHALYD/6QEAAAAAcBGSbgAAAAAAXISkGwAAAAAAFyHpBgAAAADARQp90m0YhtasWePuMAAAAAAAKDC3J90pKSl65JFHVK5cOdlsNoWFhal169b68ssv3R2ak+TkZBmGkWXq3bv3Dbe9bds2GYah33///cYDlXTixAk99NBDqly5sjw8PPTkk0/elHYBAAAAAAXj9leGdenSRZmZmVq8eLEiIyN18uRJbdmyRadPn3Z3aNnavHmzqlevbs37+Pi4MRpnpmnq8uXLSk9PV/HixfXss89q9uzZ7g4LAAAAAG5bbh3p/v333/XFF19o+vTpat68ucqXL68GDRpozJgxat++vVXu1KlT6ty5s3x9fVWxYkV99NFH1rrLly+rf//+qlChgnx8fFS5cmXNmTPHaTsxMTHq1KmTYmNjVaJECQUEBOiRRx5RRkaGVcY0Tc2YMUORkZHy8fFR7dq1tXLlyiwxh4SEKCwszJoCAwN16NAhdezYUSVLlpTdblf9+vW1efNmp3rp6ekaPXq0wsPDZbPZVLFiRcXFxSk5OVnNmzeXJAUHB8swDMXExFh1hg0bphIlSsjb21tNmjTR7t27rTavjJBv3LhR0dHRstlsSkhIUEREhObMmaOHH35YgYGB1/8FAQAAAABuiFuTbrvdLrvdrjVr1ig9PT3HcrGxserWrZv27dundu3aqVevXtZIuMPhUNmyZbV8+XIdOHBAEyZM0NixY7V8+XKnNrZs2aKkpCRt3bpV7733nj744APFxsZa68eNG6dFixZp3rx5+uGHHzR8+HD17t1b27dvz3M/zp07p3bt2mnz5s3au3evWrdurQ4dOujo0aNWmYcffljLli3T3LlzlZSUpPnz58tutys8PFyrVq2SJB08eFAnTpywThqMHj1aq1at0uLFi/Xtt98qKipKrVu3znIVwOjRozVt2jQlJSWpVq1aecYLAAAAAPhrGKZpmu4MYNWqVRo4cKAuXryoevXqqVmzZurRo4eVPBqGoXHjxmny5MmSpPPnz8vf31+ffPKJ2rRpk22bjz32mE6ePGmNVMfExOjjjz/WsWPH5OvrK0maP3++nnrqKaWmpurixYsKDQ3VZ599poYNG1rtDBgwQBcuXNC7776r5ORkazTdw+N/5yoSEhJUt27dLDFUr15dgwcP1tChQ/Xjjz+qcuXK2rRpk1q2bJml7LZt29S8eXOdOXNGQUFB1n4GBwcrPj5eDz30kCQpMzNTERERevLJJ/XUU09Z9dasWaOOHTtmeyzuvvtu1alTRy+//HJuX4PS09OdTnykpaUpPDxcd409Lk/voFzrAoWBIYci/U/p8NlQme5/XAWQK/orbjX56bNLB9tUPpT+DPdzOBxKSUlRiRIlnH63AzdbWlqaAgMDlZqaqoCAgBzLFYp7utu3b6+EhAR9+eWX2rBhg2bMmKG33nrLusz66tFbPz8/+fv7KyUlxVo2f/58vfXWWzpy5IguXryojIwM1alTx2k7tWvXthJuSWrYsKHOnTunY8eOKSUlRX/88YdatWrlVCcjIyNLQv3++++ratWq1nx4eLjOnz+v2NhYrV27VsePH9elS5d08eJFa6Q7MTFRnp6eatasWb6Py6FDh5SZmanGjRtby7y8vNSgQQMlJSU5lY2Ojs53uzmZNm2a08g/AAAAAODGuT3pliRvb2+1atVKrVq10oQJEzRgwABNnDjRSrq9vLycyhuGIYfDIUlavny5hg8frlmzZqlhw4by9/fXiy++qK+//jpf2766rXXr1qlMmTJO6202m9N8eHi4oqKinJYNHz5cGzdu1MyZMxUVFSUfHx917drVumf8eh62duUCBMMwsiy/dpmfn1+B27/WmDFjNGLECGv+ykg3AAAAAOD6FYqk+1rVqlXL97u5ExIS1KhRIw0ZMsRadujQoSzlvvvuO128eNFKgL/66ivZ7XaVLVtWwcHBstlsOnr0aIFGo6+OISYmRp07d5b05z3eycnJ1vqaNWvK4XBo+/bt2V5eXrRoUUl/PhTuiqioKBUtWlRffPGF0+Xle/bscckrwGw2W5YTDAAAAACAG+PWpPu3337Tgw8+qH79+qlWrVry9/fXnj17NGPGjBzvUb5WVFSU3n77bW3cuFEVKlTQkiVLtHv3blWoUMGpXEZGhvr3769x48bpyJEjmjhxooYOHSoPDw/5+/tr1KhRGj58uBwOh5o0aaK0tDTt3LlTdrtdffv2zTOG1atXq0OHDjIMQ+PHj7dGzyUpIiJCffv2Vb9+/TR37lzVrl1bR44cUUpKirp166by5cvLMAytXbtW7dq1k4+Pj+x2uwYPHqynnnpKxYoVU7ly5TRjxgxduHBB/fv3z/O4JCYmSvrzBMCvv/6qxMREFS1aVNWqVcvXcQUAAAAA3Di3Jt12u1133nmnZs+ebd3DHB4eroEDB2rs2LH5auPRRx9VYmKiunfvLsMw1LNnTw0ZMkTr1693KteiRQtVrFhRTZs2VXp6unr06KFJkyZZ6ydPnqwSJUpo2rRpOnz4sIKCglSvXr18xTF79mz169dPjRo1UmhoqJ5++mmlpaU5lZk3b57Gjh2rIUOG6LffflO5cuWstsuUKaPY2Fg988wz+te//qWHH35Y8fHxeuGFF+RwONSnTx+dPXtW0dHR2rhxo4KDg/OM6ep70b/55hu9++67Kl++vNMIPAAAAADAtdz+9PK/QkxMjH7//fd8X7KO/z2Jj6eX41bB06BxK6G/4lbD08txK+Hp5fir5Pfp5fRCAAAAAABchKQbAAAAAAAXKZRPL7/Z4uPj3R0CAAAAAOA2xEg3AAAAAAAuQtINAAAAAICLkHQDAAAAAOAit8U93bh+b/a3KSDA5u4wgDyZpkNpZ7wUEGyTYXA+EYUb/RW3mvz0Wbv3XxwUANwiSLqRq/BiHgoK4gchCj+HQ0q57KESIR68kxOFHv0Vtxr6LABcP/7VBAAAAADARUi6AQAAAABwEZJuAAAAAABchHu6katjpx1KveRwdxhAnkzTobQ0h/7wdMgw3B0NkDv6K2411/ZZu7cUYmfsBgDyg6QbuRoUly5P73R3hwHkyZBDkf6ZOnw2XSYX8aCQo7/iVnNtn1062KYQu7ujAoBbA//TAwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICLkHQDAAAAAOAiJN03SUxMjAzDyDK1adMmX/XvvvtuPfnkky6JzTAMrVmzxiVtAwAAAAByxivDbqI2bdpo0aJFTstsNpubogEAAAAAuBsj3TeRzWZTWFiY0xQcHKxt27apaNGiSkhIsMrOmjVLoaGhOnHihGJiYrR9+3bNmTPHGiFPTk6WJB04cEDt2rWT3W5XyZIl1adPH506dcpq5+6779awYcM0evRoFStWTGFhYZo0aZK1PiIiQpLUuXNnGYZhzQMAAAAAXI+k+y9w5dLxPn36KDU1Vd99952effZZLViwQKVKldKcOXPUsGFDDRw4UCdOnNCJEycUHh6uEydOqFmzZqpTp4727NmjDRs26OTJk+rWrZtT+4sXL5afn5++/vprzZgxQ88995w2bdokSdq9e7ckadGiRTpx4oQ1DwAAAABwPS4vv4nWrl0ru93utOzpp5/W+PHj9fzzz2vz5s0aNGiQfvjhB/Xp00edO3eWJAUGBqpo0aLy9fVVWFiYVXfevHmqV6+epk6dai1buHChwsPD9eOPP6pSpUqSpFq1amnixImSpIoVK+rVV1/Vli1b1KpVKxUvXlySFBQU5NT2tdLT05Wenm7Np6Wl3eDRAAAAAACQdN9EzZs317x585yWFStWTJJUtGhRvfPOO6pVq5bKly+vl19+Oc/2vvnmG23dujVLIi9Jhw4dckq6r1aqVCmlpKQUKPZp06YpNja2QHUAAAAAALkj6b6J/Pz8FBUVleP6nTt3SpJOnz6t06dPy8/PL9f2HA6HOnTooOnTp2dZV6pUKeuzl5eX0zrDMORwOAoSusaMGaMRI0ZY82lpaQoPDy9QGwAAAAAAZyTdf5FDhw5p+PDhWrBggZYvX66HH35YW7ZskYfHn7fVFy1aVJcvX3aqU69ePa1atUoREREqUuT6vyovL68sbV/LZrPxpHUAAAAAuMl4kNpNlJ6erl9++cVpOnXqlC5fvqw+ffro3nvv1b/+9S8tWrRI33//vWbNmmXVjYiI0Ndff63k5GSdOnVKDodDjz32mE6fPq2ePXtq165dOnz4sD799FP169cvzyT6ahEREdqyZYt++eUXnTlzxhW7DgAAAADIBkn3TbRhwwaVKlXKaWrSpImmTJmi5ORkvfnmm5KksLAwvfXWWxo3bpwSExMlSaNGjZKnp6eqVaum4sWL6+jRoypdurR27Nihy5cvq3Xr1qpRo4aeeOIJBQYGWiPk+TFr1ixt2rRJ4eHhqlu3rit2HQAAAACQDcM0TdPdQaDwSUtLU2BgoO4ae1ye3kHuDgfIkyGHIv1P6fDZUJmcT0QhR3/FrebaPrt0sE3lQ+m7KJwcDodSUlJUokSJAg1UAQV1JWdKTU1VQEBAjuXohQAAAAAAuAhJNwAAAAAALkLSDQAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuUsTdAaBwe7O/TQEBNneHAeTJNB1KO+OlgGCbDIPziSjc6K+41VzbZ+3e7o4IAG4dJN3IVXgxDwUF8YMQhZ/DIaVc9lCJEA/eyYlCj/6KWw19FgCuH/9qAgAAAADgIiTdAAAAAAC4CEk3AAAAAAAuwj3dyNWx0w6lXnK4OwwgT6bpUFqaQ394OmQY7o4GyB39FYWN3VsKsTMWAwCuQNKNXA2KS5end7q7wwDyZMihSP9MHT6bLpOLeFDI0V9R2CwdbFOI3d1RAMDfE//TAwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICLuCzpvvvuu/Xkk0+6qvmbLjk5WYZhKDEx0d2h3DDDMLRmzRp3hwEAAAAAt70CJd0xMTEyDEOPPvpolnVDhgyRYRiKiYmRJK1evVqTJ0++KUHm17vvvitPT89s48tLeHi4Tpw4oRo1auSrfGFIbCdNmqQ6deq4NQYAAAAAQM4KPNIdHh6uZcuW6eLFi9ayP/74Q++9957KlStnLStWrJj8/f1vTpQ5yMjIcJpfuHChRo8erWXLlunChQsFasvT01NhYWEqUoS3qAEAAAAAbo4CJ9316tVTuXLltHr1amvZ6tWrFR4errp161rLrr28PCIiQlOnTlW/fv3k7++vcuXK6c0333Rqe//+/brnnnvk4+OjkJAQDRo0SOfOnbPWx8TEqFOnTpo2bZpKly6tSpUqWeuSk5O1c+dOPfPMM6pSpYpWrlzp1Ha/fv1Uq1Ytpaf/+c7pzMxM3XHHHerVq5dV/+rLy8+cOaNevXqpePHi8vHxUcWKFbVo0aJ8H6dFixapatWq8vb2VpUqVfT66687xWoYhlavXq3mzZvL19dXtWvX1pdffunUxoIFCxQeHi5fX1917txZL730koKCgiRJ8fHxio2N1XfffSfDMGQYhuLj4626p06dUufOneXr66uKFSvqo48+ynfsAAAAAICb47ru6f7Xv/7llIAuXLhQ/fr1y7PerFmzFB0drb1792rIkCEaPHiw/v3vf0uSLly4oDZt2ig4OFi7d+/WihUrtHnzZg0dOtSpjS1btigpKUmbNm3S2rVrnWJo3769AgMD1bt3b8XFxTnVmzt3rs6fP69nnnlGkjR+/HidOnXKKRm+2vjx43XgwAGtX79eSUlJmjdvnkJDQ/N1fBYsWKBnn31WU6ZMUVJSkqZOnarx48dr8eLFTuWeffZZjRo1SomJiapUqZJ69uypS5cuSZJ27NihRx99VE888YQSExPVqlUrTZkyxarbvXt3jRw5UtWrV9eJEyd04sQJde/e3VofGxurbt26ad++fWrXrp169eql06dP5yt+AAAAAMDNcV3XUvfp00djxoyxRmx37NihZcuWadu2bbnWa9eunYYMGSJJevrppzV79mxt27ZNVapU0dKlS3Xx4kW9/fbb8vPzkyS9+uqr6tChg6ZPn66SJUtKkvz8/PTWW2+paNGiVrsOh0Px8fF65ZVXJEk9evTQiBEj9NNPPykqKkqSZLfb9c4776hZs2by9/fXrFmztGXLFgUGBmYb69GjR1W3bl1FR0dL+nOkPr8mT56sWbNm6YEHHpAkVahQQQcOHNAbb7yhvn37WuVGjRql9u3bS/ozSa5evbp++uknValSRa+88oratm2rUaNGSZIqVaqknTt3WicafHx8ZLfbVaRIEYWFhWWJISYmRj179pQkTZ06Va+88op27dqlNm3aZBtzenq6dRWAJKWlpeV7fwEAAAAA2buuke7Q0FC1b99eixcv1qJFi9S+fft8jQLXqlXL+mwYhsLCwpSSkiJJSkpKUu3ata2EW5IaN24sh8OhgwcPWstq1qzplHBL0qeffqrz58+rbdu2Vnz33nuvFi5c6FSuYcOGGjVqlCZPnqyRI0eqadOmOcY6ePBgLVu2THXq1NHo0aO1c+fOPPdPkn799VcdO3ZM/fv3l91ut6bnn39ehw4dyvF4lCpVSpKs43Hw4EE1aNDAqfy187m5um0/Pz/5+/tbbWdn2rRpCgwMtKbw8PB8bwsAAAAAkL3rfmpYv379rEu/X3vttXzV8fLycpo3DEMOh0OSZJqmDMPItt7Vy69Oyq9YuHChTp8+LV9fX2uZw+HQ3r17NXnyZHl6elrLduzYIU9PT/3nP//JNda2bdvqyJEjWrdunTZv3qwWLVroscce08yZM3Otd2V/FixYoDvvvNNp3ZU4rrj6eFzZx9yOh2mauW47p7avtH+l7eyMGTNGI0aMsObT0tJIvAEAAADgBl33e7rbtGmjjIwMZWRkqHXr1jccSLVq1ZSYmKjz589by3bs2CEPDw+nB6Zd67ffftOHH36oZcuWKTEx0Wk6d+6c1q9fb5V98cUXlZSUpO3bt2vjxo15PhitePHiiomJ0TvvvKOXX345y4PfslOyZEmVKVNGhw8fVlRUlNNUoUKFfByJP1WpUkW7du1yWrZnzx6n+aJFi+ry5cv5bjM3NptNAQEBThMAAAAA4MZc90i3p6enkpKSrM83qlevXpo4caL69u2rSZMm6ddff9Xjjz+uPn36WPdzZ2fJkiUKCQnRgw8+KA8P53MI9913n+Li4nTfffcpMTFREyZM0MqVK9W4cWPNmTNHTzzxhJo1a6bIyMgs7U6YMEF33HGHqlevrvT0dK1du1ZVq1Z1KvPzzz9bTzu/IioqSpMmTdKwYcMUEBCgtm3bKj09XXv27NGZM2ecRpNz8/jjj6tp06Z66aWX1KFDB3322Wdav3690+h3RESEFUPZsmXl7+8vm82Wr/YBAAAAAK533SPdkm7qiKivr682btyo06dPq379+uratatatGihV199Ndd6CxcuVOfOnbMk3JLUpUsXrV27VkeOHFGvXr0UExOjDh06SJL69++vli1bqk+fPtmOFhctWlRjxoxRrVq11LRpU3l6emrZsmVOZUaMGKG6des6TXv27NGAAQP01ltvKT4+XjVr1lSzZs0UHx9foJHuxo0ba/78+XrppZdUu3ZtbdiwQcOHD5e3t7fT/rVp00bNmzdX8eLF9d577+W7fQAAAACA6xlmQW4UhlsNHDhQ//73v5WQkODybaWlpSkwMFB3jT0uT+8gl28PuFGGHIr0P6XDZ0Nl3tj5RMDl6K8obJYOtql8aM590eFwKCUlRSVKlMh2oAMoTOiv+KtcyZlSU1NzHYy+7svL4XozZ85Uq1at5Ofnp/Xr12vx4sU5vlccAAAAAFD4kHQXYrt27dKMGTN09uxZRUZGau7cuRowYIC7wwIAAAAA5BNJdyG2fPlyd4cAAAAAALgB3OQAAAAAAICLkHQDAAAAAOAiJN0AAAAAALgI93QjV2/2tykgwObuMIA8maZDaWe8FBBsk2FwPhGFG/0VhY3d290RAMDfF0k3chVezENBQfwgROHncEgplz1UIsSDd3Ki0KO/AgBw++B/egAAAAAAXISkGwAAAAAAFyHpBgAAAADARbinG7k6dtqh1EsOd4cB5Mk0HUpLc+gPT4cMw93RALmjv94Yu7cUYmfcAABwayDpRq4GxaXL0zvd3WEAeTLkUKR/pg6fTZfJRTwo5OivN2bpYJtC7O6OAgCA/OF/egAAAAAAXISkGwAAAAAAFyHpBgAAAADARUi6AQAAAABwEZLuGzRp0iTVqVPH3WEAAAAAAAqh2yrpjomJkWEYMgxDXl5eioyM1KhRo3T+/PnrbnPUqFHasmXLTYwyezExMerUqZPLtwMAAAAAuHluu1eGtWnTRosWLVJmZqYSEhI0YMAAnT9/XvPmzStQO6Zp6vLly7Lb7bLbb533lmRkZKho0aLuDgMAAAAAbgu31Ui3JNlsNoWFhSk8PFwPPfSQevXqpTVr1sg0Tc2YMUORkZHy8fFR7dq1tXLlSqvetm3bZBiGNm7cqOjoaNlsNiUkJGS5vPzKiPTUqVNVsmRJBQUFKTY2VpcuXdJTTz2lYsWKqWzZslq4cKFTXP/3f/+n7t27Kzg4WCEhIerYsaOSk5Ml/XkJ++LFi/Xhhx9aI/Xbtm3Ls97V8UybNk2lS5dWpUqVXHVoAQAAAADXuO2S7mv5+PgoMzNT48aN06JFizRv3jz98MMPGj58uHr37q3t27c7lR89erSmTZumpKQk1apVK9s2P/vsMx0/flyff/65XnrpJU2aNEn33XefgoOD9fXXX+vRRx/Vo48+qmPHjkmSLly4oObNm8tut+vzzz/XF198IbvdrjZt2igjI0OjRo1St27d1KZNG504cUInTpxQo0aN8qx3xZYtW5SUlKRNmzZp7dq1rjuYAAAAAAAnt93l5VfbtWuX3n33XTVv3lwvvfSSPvvsMzVs2FCSFBkZqS+++EJvvPGGmjVrZtV57rnn1KpVq1zbLVasmObOnSsPDw9VrlxZM2bM0IULFzR27FhJ0pgxY/TCCy9ox44d6tGjh5YtWyYPDw+99dZbMgxDkrRo0SIFBQVp27Ztuvfee+Xj46P09HSFhYVZ23nnnXfyrCdJfn5+euutt3K9rDw9PV3p6enWfFpaWkEOJQAAAAAgG7dd0r127VrZ7XZdunRJmZmZ6tixo0aNGqWVK1dmSaYzMjJUt25dp2XR0dF5bqN69ery8PjfRQQlS5ZUjRo1rHlPT0+FhIQoJSVFkvTNN9/op59+kr+/v1M7f/zxhw4dOpTjdvJbr2bNmnnexz1t2jTFxsbmuW8AAAAAgPy77ZLu5s2ba968efLy8lLp0qXl5eWlr7/+WpK0bt06lSlTxqm8zWZzmvfz88tzG15eXk7zV56Wfu0yh8MhSXI4HLrjjju0dOnSLG0VL148x+3kt15+Yh4zZoxGjBhhzaelpSk8PDzPegAAAACAnN12Sbefn5+ioqKcllWrVk02m01Hjx51upT8r1KvXj29//77KlGihAICArItU7RoUV2+fLnA9fLLZrNlOcEAAAAAALgxt/2D1CTJ399fo0aN0vDhw7V48WIdOnRIe/fu1WuvvabFixe7fPu9evVSaGioOnbsqISEBP3888/avn27nnjiCf33v/+VJEVERGjfvn06ePCgTp06pczMzHzVAwAAAAC4D0n3/zd58mRNmDBB06ZNU9WqVdW6dWt9/PHHqlChgsu37evrq88//1zlypXTAw88oKpVq6pfv366ePGiNYI9cOBAVa5cWdHR0SpevLh27NiRr3oAAAAAAPcxTNM03R0ECp+0tDQFBgbqrrHH5ekd5O5wgDwZcijS/5QOnw2VyflEFHL01xuzdLBN5UM5bn8lh8OhlJQUlShRwulhsUBhRH/FX+VKzpSamprroCe9EAAAAAAAFyHpBgAAAADARUi6AQAAAABwEZJuAAAAAABchKQbAAAAAAAXIekGAAAAAMBFirg7ABRub/a3KSDA5u4wgDyZpkNpZ7wUEGyTYXA+EYUb/fXG2L3dHQEAAPlH0o1chRfzUFAQPwhR+DkcUsplD5UI8eCdnCj06K8AANw++J8eAAAAAAAXIekGAAAAAMBFSLoBAAAAAHAR7ulGro6ddij1ksPdYQB5Mk2H0tIc+sPTIcNwdzRA7v7O/dXuLYXYOacPAMAVJN3I1aC4dHl6p7s7DCBPhhyK9M/U4bPpMrmIB4Xc37m/Lh1sU4jd3VEAAFB4/L3+pwcAAAAAoBAh6QYAAAAAwEVIugEAAAAAcBGSbgAAAAAAXISkGwAAAAAAF3Fr0m0YhtasWePOECzJyckyDEOJiYnuDuWGFabjCgAAAAC3swIl3TExMTIMQ48++miWdUOGDJFhGIqJicl3eydOnFDbtm0LEkKu3n33XXl6emYbX17Cw8N14sQJ1ahRI1/lC0NiO2nSJNWpU8etMQAAAAAAclbgke7w8HAtW7ZMFy9etJb98ccfeu+991SuXLkCtRUWFiabzVbQECwZGRlO8wsXLtTo0aO1bNkyXbhwoUBteXp6KiwsTEWK8OpyAAAAAMDNUeCku169eipXrpxWr15tLVu9erXCw8NVt25da9mGDRvUpEkTBQUFKSQkRPfdd58OHTrk1Na1o8X79+/XPffcIx8fH4WEhGjQoEE6d+6ctT4mJkadOnXStGnTVLp0aVWqVMlal5ycrJ07d+qZZ55RlSpVtHLlSqdt9evXT7Vq1VJ6erokKTMzU3fccYd69epl1b/68vIzZ86oV69eKl68uHx8fFSxYkUtWrQo38dp0aJFqlq1qry9vVWlShW9/vrrTrEahqHVq1erefPm8vX1Ve3atfXll186tbFgwQKFh4fL19dXnTt31ksvvaSgoCBJUnx8vGJjY/Xdd9/JMAwZhqH4+Hir7qlTp9S5c2f5+vqqYsWK+uijj/IdOwAAAADg5riue7r/9a9/OSWgCxcuVL9+/ZzKnD9/XiNGjNDu3bu1ZcsWeXh4qHPnznI4HNm2eeHCBbVp00bBwcHavXu3VqxYoc2bN2vo0KFO5bZs2aKkpCRt2rRJa9eudYqhffv2CgwMVO/evRUXF+dUb+7cuTp//ryeeeYZSdL48eN16tQpp2T4auPHj9eBAwe0fv16JSUlad68eQoNDc3X8VmwYIGeffZZTZkyRUlJSZo6darGjx+vxYsXO5V79tlnNWrUKCUmJqpSpUrq2bOnLl26JEnasWOHHn30UT3xxBNKTExUq1atNGXKFKtu9+7dNXLkSFWvXl0nTpzQiRMn1L17d2t9bGysunXrpn379qldu3bq1auXTp8+nWPM6enpSktLc5oAAAAAADfmuq6l7tOnj8aMGWON2O7YsUPLli3Ttm3brDJdunRxqhMXF6cSJUrowIED2d43vXTpUl28eFFvv/22/Pz8JEmvvvqqOnTooOnTp6tkyZKSJD8/P7311lsqWrSoVdfhcCg+Pl6vvPKKJKlHjx4aMWKEfvrpJ0VFRUmS7Ha73nnnHTVr1kz+/v6aNWuWtmzZosDAwGz38ejRo6pbt66io6MlSREREfk+PpMnT9asWbP0wAMPSJIqVKigAwcO6I033lDfvn2tcqNGjVL79u0l/ZkkV69eXT/99JOqVKmiV155RW3bttWoUaMkSZUqVdLOnTutEw0+Pj6y2+0qUqSIwsLCssQQExOjnj17SpKmTp2qV155Rbt27VKbNm2yjXnatGmKjY3N9z4CAAAAAPJ2XSPdoaGhat++vRYvXqxFixapffv2WUaBDx06pIceekiRkZEKCAhQhQoVJP2ZzGYnKSlJtWvXthJuSWrcuLEcDocOHjxoLatZs6ZTwi1Jn376qc6fP289lC00NFT33nuvFi5c6FSuYcOGGjVqlCZPnqyRI0eqadOmOe7j4MGDtWzZMtWpU0ejR4/Wzp0783FkpF9//VXHjh1T//79Zbfbren555/Pcnl9rVq1rM+lSpWSJKWkpEiSDh48qAYNGjiVv3Y+N1e37efnJ39/f6vt7IwZM0apqanWdOzYsXxvCwAAAACQvet+ali/fv2sS79fe+21LOs7dOig8PBwLViwQKVLl5bD4VCNGjWyPPzsCtM0ZRhGtuuuXn51Un7FwoULdfr0afn6+lrLHA6H9u7dq8mTJ8vT09NatmPHDnl6euo///lPrvvXtm1bHTlyROvWrdPmzZvVokULPfbYY5o5c2au9a5cPr9gwQLdeeedTuuuxHGFl5dXln28Uj+742GaZq7bzqntK+3ndGm/JNlstht6qB0AAAAAIKvrfk93mzZtlJGRoYyMDLVu3dpp3W+//aakpCSNGzdOLVq0UNWqVXXmzJlc26tWrZoSExN1/vx5a9mOHTvk4eHh9MC0a/3222/68MMPtWzZMiUmJjpN586d0/r1662yL774opKSkrR9+3Zt3LgxzwejFS9eXDExMXrnnXf08ssv680338y1vCSVLFlSZcqU0eHDhxUVFeU0XRntz48qVapo165dTsv27NnjNF+0aFFdvnw5320CAAAAAP5a1z3S7enpqaSkJOvz1YKDgxUSEqI333xTpUqV0tGjR60HmOWkV69emjhxovr27atJkybp119/1eOPP64+ffpY93NnZ8mSJQoJCdGDDz4oDw/ncwj33Xef4uLidN999ykxMVETJkzQypUr1bhxY82ZM0dPPPGEmjVrpsjIyCztTpgwQXfccYeqV6+u9PR0rV27VlWrVnUq8/PPP1tPO78iKipKkyZN0rBhwxQQEKC2bdsqPT1de/bs0ZkzZzRixIhcj8MVjz/+uJo2baqXXnpJHTp00Geffab169c7jX5HRERYMZQtW1b+/v6MVgMAAABAIXLdI92SFBAQoICAgKyNenho2bJl+uabb1SjRg0NHz5cL774Yq5t+fr6auPGjTp9+rTq16+vrl27qkWLFnr11Vdzrbdw4UJ17tw5S8It/fkwt7Vr1+rIkSPq1auXYmJi1KFDB0lS//791bJlS/Xp0yfb0eKiRYtqzJgxqlWrlpo2bSpPT08tW7bMqcyIESNUt25dp2nPnj0aMGCA3nrrLcXHx6tmzZpq1qyZ4uPjCzTS3bhxY82fP18vvfSSateurQ0bNmj48OHy9vZ22r82bdqoefPmKl68uN577718tw8AAAAAcD3DLMiNwjdRenq6vL29tWnTJrVs2dIdIdxyBg4cqH//+99KSEhw+bbS0tIUGBiou8Yel6d3kMu3B9woQw5F+p/S4bOhMm/sfCLgcn/n/rp0sE3lQ/9e+4Q/nzmTkpKiEiVKZDvQARQm9Ff8Va7kTKmpqdkORl9x3ZeX34i0tDStXr1aHh4eqlKlijtCuCXMnDlTrVq1kp+fn9avX6/Fixfn+F5xAAAAAEDh45ake+LEiXr33Xc1ffp0lS1b1h0h3BJ27dqlGTNm6OzZs4qMjNTcuXM1YMAAd4cFAAAAAMgntyTds2fP1uzZs92x6VvK8uXL3R0CAAAAAOAGcJMDAAAAAAAuQtINAAAAAICLuOXyctw63uxvU0AA7/5G4WeaDqWd8VJAsE2GwflEFG5/5/5q9867DAAAtxOSbuQqvJiHgoL+Xj8I8ffkcEgplz1UIsSD14Og0KO/AgBw++B/egAAAAAAXISkGwAAAAAAFyHpBgAAAADARbinG7k6dtqh1EsOd4cB5Mk0HUpLc+gPT4cMw93R4HZl95ZC7JzPBgAA/0PSjVwNikuXp3e6u8MA8mTIoUj/TB0+my6Ti3jgJksH2xRid3cUAACgMOGXKQAAAAAALkLSDQAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICLFJqk+5dfftHjjz+uyMhI2Ww2hYeHq0OHDtqyZYu7Q7thERERevnll90dBgAAAADgL1Yo3tOdnJysxo0bKygoSDNmzFCtWrWUmZmpjRs36rHHHtO///1vd4cIAAAAAECBFYqR7iFDhsgwDO3atUtdu3ZVpUqVVL16dY0YMUJfffWVJOno0aPq2LGj7Ha7AgIC1K1bN508edJqY9KkSapTp44WLlyocuXKyW63a/Dgwbp8+bJmzJihsLAwlShRQlOmTHHatmEYmjdvntq2bSsfHx9VqFBBK1ascCrz9NNPq1KlSvL19VVkZKTGjx+vzMxMpzIfffSRoqOj5e3trdDQUD3wwAOSpLvvvltHjhzR8OHDZRiGDMOQJMXHxysoKEgbN25U1apVZbfb1aZNG504ccKp3UWLFqlq1ary9vZWlSpV9Prrr1vrMjIyNHToUJUqVUre3t6KiIjQtGnTnI5JuXLlZLPZVLp0aQ0bNux6vyIAAAAAwHVw+0j36dOntWHDBk2ZMkV+fn5Z1gcFBck0TXXq1El+fn7avn27Ll26pCFDhqh79+7atm2bVfbQoUNav369NmzYoEOHDqlr1676+eefValSJW3fvl07d+5Uv3791KJFC911111WvfHjx+uFF17QnDlztGTJEvXs2VM1atRQ1apVJUn+/v6Kj49X6dKltX//fg0cOFD+/v4aPXq0JGndunV64IEH9Oyzz2rJkiXKyMjQunXrJEmrV69W7dq1NWjQIA0cONBp3y5cuKCZM2dqyZIl8vDwUO/evTVq1CgtXbpUkrRgwQJNnDhRr776qurWrau9e/dq4MCB8vPzU9++fTV37lx99NFHWr58ucqVK6djx47p2LFjkqSVK1dq9uzZWrZsmapXr65ffvlF3333XY7fQ3p6utLT0635tLS0gnyNAAAAAIBsuD3p/umnn2SapqpUqZJjmc2bN2vfvn36+eefFR4eLklasmSJqlevrt27d6t+/fqSJIfDoYULF8rf31/VqlVT8+bNdfDgQX3yySfy8PBQ5cqVNX36dG3bts0p6X7wwQc1YMAASdLkyZO1adMmvfLKK9ao8rhx46yyERERGjlypN5//30r6Z4yZYp69Oih2NhYq1zt2rUlScWKFZOnp6f8/f0VFhbmtF+ZmZmaP3++/vGPf0iShg4dqueee85aP3nyZM2aNcsaNa9QoYIOHDigN954Q3379tXRo0dVsWJFNWnSRIZhqHz58lbdo0ePKiwsTC1btpSXl5fKlSunBg0a5HiMp02b5hQ/AAAAAODGuf3yctM0Jcm67Do7SUlJCg8PtxJuSapWrZqCgoKUlJRkLYuIiJC/v781X7JkSVWrVk0eHh5Oy1JSUpzab9iwYZb5q9tduXKlmjRporCwMNntdo0fP15Hjx611icmJqpFixb53WWLr6+vlXBLUqlSpazYfv31Vx07dkz9+/eX3W63pueff16HDh2SJMXExCgxMVGVK1fWsGHD9Omnn1ptPfjgg7p48aIiIyM1cOBAffDBB7p06VKOsYwZM0apqanWdGXEHAAAAABw/dyedFesWFGGYTgludcyTTPbpPza5V5eXk7rDcPIdpnD4cgzrivtfvXVV+rRo4fatm2rtWvXau/evXr22WeVkZFhlfXx8cmzvexkF9uVkxBXYlywYIESExOt6fvvv7fuc69Xr55+/vlnTZ48WRcvXlS3bt3UtWtXSVJ4eLgOHjyo1157TT4+PhoyZIiaNm2a5V70K2w2mwICApwmAAAAAMCNcXvSXaxYMbVu3Vqvvfaazp8/n2X977//rmrVquno0aNOo68HDhxQamqqdd/1jbiSxF49f+Vy9x07dqh8+fJ69tlnFR0drYoVK+rIkSNO5WvVqpXrq82KFi2qy5cvFyimkiVLqkyZMjp8+LCioqKcpgoVKljlAgIC1L17dy1YsEDvv/++Vq1apdOnT0v682TA/fffr7lz52rbtm368ssvtX///gLFAQAAAAC4fm6/p1uSXn/9dTVq1EgNGjTQc889p1q1aunSpUvatGmT5s2bpwMHDqhWrVrq1auXXn75ZetBas2aNVN0dPQNb3/FihWKjo5WkyZNtHTpUu3atUtxcXGSpKioKB09elTLli1T/fr1tW7dOn3wwQdO9SdOnKgWLVroH//4h3r06KFLly5p/fr11j3fERER+vzzz9WjRw/ZbDaFhobmK65JkyZp2LBhCggIUNu2bZWenq49e/bozJkzGjFihGbPnq1SpUqpTp068vDw0IoVKxQWFqagoCDFx8fr8uXLuvPOO+Xr66slS5bIx8fH6b5vAAAAAIBruX2kW/rzAWHffvutmjdvrpEjR6pGjRpq1aqVtmzZonnz5skwDK1Zs0bBwcFq2rSpWrZsqcjISL3//vs3ZfuxsbFatmyZatWqpcWLF2vp0qWqVq2aJKljx44aPny4hg4dqjp16mjnzp0aP368U/27775bK1as0EcffaQ6deronnvu0ddff22tf+6555ScnKx//OMfKl68eL7jGjBggN566y3Fx8erZs2aatasmeLj462RbrvdrunTpys6Olr169dXcnKy9dC4oKAgLViwQI0bN7ZG4j/++GOFhITchCMGAAAAAMgPw7xyE/FtyjAMffDBB+rUqZO7QylU0tLSFBgYqLvGHpend5C7wwHyZMihSP9TOnw2VGbhOJ+I29DSwTaVD827/zkcDqWkpKhEiRJOD/sECiv6LG4l9Ff8Va7kTKmpqbk+E4teCAAAAACAi5B0AwAAAADgIoXiQWrudJtfXQ8AAAAAcCFGugEAAAAAcBGSbgAAAAAAXOS2v7wcuXuzv00BATZ3hwHkyTQdSjvjpYBgmwyD84lwD7u3uyMAAACFDUk3chVezENBQSQwKPwcDinlsodKhHjwehAAAAAUGvwyBQAAAADARUi6AQAAAABwEZJuAAAAAABchHu6katjpx1KveRwdxhAnkzTobQ0h/7wdMgw3B0NbjV2bynEznloAABw85F0I1eD4tLl6Z3u7jCAPBlyKNI/U4fPpsvkIh4U0NLBNoXY3R0FAAD4O+KXKQAAAAAALkLSDQAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICL3HJJ9y+//KLHH39ckZGRstlsCg8PV4cOHbRlyxZ3h3Zdpk6dKk9PT73wwgvuDgUAAAAAcJPdUkl3cnKy7rjjDn322WeaMWOG9u/frw0bNqh58+Z67LHH3B3edVm0aJFGjx6thQsX5lk2MzPzL4gIAAAAAHCz3FJJ95AhQ2QYhnbt2qWuXbuqUqVKql69ukaMGKGvvvpKknT06FF17NhRdrtdAQEB6tatm06ePGm1MWnSJNWpU0dLlixRRESEAgMD1aNHD509e1aS9MYbb6hMmTJyOBxO277//vvVt29fa/7jjz/WHXfcIW9vb0VGRio2NlaXLl2SJD333HMqXbq0fvvtN6f6TZs2dWp3+/btunjxop577jmdP39en3/+udM2r8S6cOFCa2TfNE2lpqZq0KBBKlGihAICAnTPPffou+++s+odOnRIHTt2VMmSJWW321W/fn1t3rz5Rg8/AAAAAKCAbpmk+/Tp09qwYYMee+wx+fn5ZVkfFBQk0zTVqVMnnT59Wtu3b9emTZt06NAhde/e3ansoUOHtGbNGq1du1Zr167V9u3brcu7H3zwQZ06dUpbt261yp85c0YbN25Ur169JEkbN25U7969NWzYMB04cEBvvPGG4uPjNWXKFEnSs88+q4iICA0YMECSNH/+fH3++edasmSJPDz+d8jj4uLUs2dPeXl5qWfPnoqLi8uyXz/99JOWL1+uVatWKTExUZLUvn17/fLLL/rkk0/0zTffqF69emrRooVOnz4tSTp37pzatWunzZs3a+/evWrdurU6dOigo0eP5nh809PTlZaW5jQBAAAAAG7MLZN0//TTTzJNU1WqVMmxzObNm7Vv3z69++67uuOOO3TnnXdqyZIl2r59u3bv3m2Vczgcio+PV40aNfTPf/5Tffr0se4JL1asmNq0aaN3333XKr9ixQoVK1ZMLVq0kCRNmTJFzzzzjPr27avIyEi1atVKkydP1htvvCFJ8vT01DvvvKMtW7bomWee0ciRI/Xaa6+pfPnyVptpaWlatWqVevfuLUnq3bu3Vq5cmSXZzcjI0JIlS1S3bl3VqlVLW7du1f79+7VixQpFR0erYsWKmjlzpoKCgrRy5UpJUu3atfXII4+oZs2aqlixop5//nlFRkbqo48+yvHYTZs2TYGBgdYUHh6er+8FAAAAAJCzWybpNk1TkmQYRo5lkpKSFB4e7pQwVqtWTUFBQUpKSrKWRUREyN/f35ovVaqUUlJSrPlevXpp1apVSk9PlyQtXbpUPXr0kKenpyTpm2++0XPPPSe73W5NAwcO1IkTJ3ThwgVJUmRkpGbOnKnp06erQ4cO1ij5Fe+++64iIyNVu3ZtSVKdOnUUGRmpZcuWOZUrX768ihcvbs1/8803OnfunEJCQpy2//PPP+vQoUOSpPPnz2v06NHWvtvtdv373//OdaR7zJgxSk1NtaZjx47lWBYAAAAAkD9F3B1AflWsWFGGYSgpKUmdOnXKtoxpmtkm5dcu9/LyclpvGIbTvdYdOnSQw+HQunXrVL9+fSUkJOill16y1jscDsXGxuqBBx7Isi1vb2/r8+effy5PT08lJyfr0qVLKlLkf4d74cKF+uGHH5yWORwOxcXFadCgQdayay+ldzgcKlWqlLZt25Zl20FBQZKkp556Shs3btTMmTMVFRUlHx8fde3aVRkZGVnqXGGz2WSz2XJcDwAAAAAouFsm6S5WrJhat26t1157TcOGDcuSjP7++++qVq2ajh49qmPHjlmj3QcOHFBqaqqqVq2a7235+PjogQce0NKlS/XTTz+pUqVKuuOOO6z19erV08GDBxUVFZVjG++//75Wr16tbdu2qXv37po8ebJiY2MlSfv379eePXu0bds2FStWzGkfmjZtqu+//141atTItt169erpl19+UZEiRRQREZFtmYSEBMXExKhz586S/rzHOzk5Od/7DwAAAAC4OW6ZpFuSXn/9dTVq1EgNGjTQc889p1q1aunSpUvatGmT5s2bpwMHDqhWrVrq1auXXn75ZV26dElDhgxRs2bNFB0dXaBt9erVSx06dNAPP/xg3Xd9xYQJE3TfffcpPDxcDz74oDw8PLRv3z7t379fzz//vP773/9q8ODBmj59upo0aaL4+Hi1b99ebdu21V133aW4uDg1aNBATZs2zbLdhg0bKi4uTrNnz842rpYtW6phw4bq1KmTpk+frsqVK+v48eP65JNP1KlTJ0VHRysqKkqrV69Whw4dZBiGxo8fn+Vp7AAAAAAA17tl7umWpAoVKujbb79V8+bNNXLkSNWoUUOtWrXSli1bNG/ePBmGoTVr1ig4OFhNmzZVy5YtFRkZqffff7/A27rnnntUrFgxHTx4UA899JDTutatW2vt2rXatGmT6tevr7vuuksvvfSSypcvL9M0FRMTowYNGmjo0KGSpFatWmno0KHq3bu30tLS9M4776hLly7ZbrdLly565513crwU3DAMffLJJ2ratKn69eunSpUqqUePHkpOTlbJkiUlSbNnz1ZwcLAaNWqkDh06qHXr1qpXr16BjwEAAAAA4MYY5pUnlAFXSUtLU2BgoO4ae1ye3kHuDgfIkyGHIv1P6fDZUJm31vlEFAJLB9tUPvSv6zcOh0MpKSkqUaKE06skgcKKPotbCf0Vf5UrOVNqaqoCAgJyLEcvBAAAAADARUi6AQAAAABwEZJuAAAAAABchKQbAAAAAAAXIekGAAAAAMBFbqn3dOOv92Z/mwICbO4OA8iTaTqUdsZLAcE2GQbnE1Ewdm93RwAAAP6uSLqRq/BiHgoKIoFB4edwSCmXPVQixIPXgwAAAKDQ4JcpAAAAAAAuQtINAAAAAICLkHQDAAAAAOAi3NONXB077VDqJYe7wwDyZJoOpaU59IenQ4bh7mhQWNm9pRA755sBAMBfh6QbuRoUly5P73R3hwHkyZBDkf6ZOnw2XSYX8SAHSwfbFGJ3dxQAAOB2wi9TAAAAAABchKQbAAAAAAAXIekGAAAAAMBFSLoBAAAAAHARkm4AAAAAAFyk0CfdhmFozZo17g4DAAAAAIACc3vSnZKSokceeUTlypWTzWZTWFiYWrdurS+//NLdoTlJTk6WYRhZpt69e99w29u2bZNhGPr9999vPFBJX3zxhRo3bqyQkBD5+PioSpUqmj179k1pGwAAAACQf25/T3eXLl2UmZmpxYsXKzIyUidPntSWLVt0+vRpd4eWrc2bN6t69erWvI+PjxujcWaapi5fviw/Pz8NHTpUtWrVkp+fn7744gs98sgj8vPz06BBg9wdJgAAAADcNtw60v3777/riy++0PTp09W8eXOVL19eDRo00JgxY9S+fXur3KlTp9S5c2f5+vqqYsWK+uijj6x1ly9fVv/+/VWhQgX5+PiocuXKmjNnjtN2YmJi1KlTJ8XGxqpEiRIKCAjQI488ooyMDKuMaZqaMWOGIiMj5ePjo9q1a2vlypVZYg4JCVFYWJg1BQYG6tChQ+rYsaNKliwpu92u+vXra/PmzU710tPTNXr0aIWHh8tms6lixYqKi4tTcnKymjdvLkkKDg6WYRiKiYmx6gwbNkwlSpSQt7e3mjRpot27d1ttXhkh37hxo6Kjo2Wz2ZSQkKC6deuqZ8+eql69uiIiItS7d2+1bt1aCQkJ1/9lAQAAAAAKzK1Jt91ul91u15o1a5Senp5judjYWHXr1k379u1Tu3bt1KtXL2sk3OFwqGzZslq+fLkOHDigCRMmaOzYsVq+fLlTG1u2bFFSUpK2bt2q9957Tx988IFiY2Ot9ePGjdOiRYs0b948/fDDDxo+fLh69+6t7du357kf586dU7t27bR582bt3btXrVu3VocOHXT06FGrzMMPP6xly5Zp7ty5SkpK0vz582W32xUeHq5Vq1ZJkg4ePKgTJ05YJw1Gjx6tVatWafHixfr2228VFRWl1q1bZ7kKYPTo0Zo2bZqSkpJUq1atLPHt3btXO3fuVLNmzXLch/T0dKWlpTlNAAAAAIAbY5imabozgFWrVmngwIG6ePGi6tWrp2bNmqlHjx5W8mgYhsaNG6fJkydLks6fPy9/f3998sknatOmTbZtPvbYYzp58qQ1Uh0TE6OPP/5Yx44dk6+vryRp/vz5euqpp5SamqqLFy8qNDRUn332mRo2bGi1M2DAAF24cEHvvvuukpOTrdF0D4//nau4MrJ8rerVq2vw4MEaOnSofvzxR1WuXFmbNm1Sy5Yts5Tdtm2bmjdvrjNnzigoKMjaz+DgYMXHx+uhhx6SJGVmZioiIkJPPvmknnrqKavemjVr1LFjxyztli1bVr/++qsuXbqkSZMmafz48Tl+D5MmTXI6CXHFXWOPy9M7KMd6QGFhyKFI/1M6fDZUpvsfV4FCaulgm8qHur9/OBwOpaSkqESJEk7/pwCFFX0WtxL6K/4qaWlpCgwMVGpqqgICAnIsVyju6W7fvr0SEhL05ZdfasOGDZoxY4beeust6zLrq0dv/fz85O/vr5SUFGvZ/Pnz9dZbb+nIkSO6ePGiMjIyVKdOHaft1K5d20q4Jalhw4Y6d+6cjh07ppSUFP3xxx9q1aqVU52MjIwsCfX777+vqlWrWvPh4eE6f/68YmNjtXbtWh0/flyXLl3SxYsXrZHuxMREeXp65jrSfK1Dhw4pMzNTjRs3tpZ5eXmpQYMGSkpKciobHR2dbRsJCQk6d+6cvvrqKz3zzDOKiopSz549sy07ZswYjRgxwppPS0tTeHh4vuMFAAAAAGTl9qRbkry9vdWqVSu1atVKEyZM0IABAzRx4kQr6fby8nIqbxiGHA6HJGn58uUaPny4Zs2apYYNG8rf318vvviivv7663xt++q21q1bpzJlyjitt9lsTvPh4eGKiopyWjZ8+HBt3LhRM2fOVFRUlHx8fNS1a1frnvHredjalQsQDMPIsvzaZX5+ftm2UaFCBUlSzZo1dfLkSU2aNCnHpNtms2XZVwAAAADAjSmU11tUq1ZN58+fz1fZhIQENWrUSEOGDFHdunUVFRWlQ4cOZSn33Xff6eLFi9b8V199JbvdrrJly6patWqy2Ww6evSooqKinKb8jPYmJCQoJiZGnTt3Vs2aNRUWFqbk5GRrfc2aNeVwOHK8P7xo0aKS/nwo3BVRUVEqWrSovvjiC2tZZmam9uzZ4zTSnl+maeZ63zwAAAAA4OZz60j3b7/9pgcffFD9+vVTrVq15O/vrz179mjGjBnZ3qOcnaioKL399tvauHGjKlSooCVLlmj37t3WKO8VGRkZ6t+/v8aNG6cjR45o4sSJGjp0qDw8POTv769Ro0Zp+PDhcjgcatKkidLS0rRz507Z7Xb17ds3zxhWr16tDh06yDAMjR8/3ho9l6SIiAj17dtX/fr109y5c1W7dm0dOXJEKSkp6tatm8qXLy/DMLR27Vq1a9dOPj4+stvtGjx4sJ566ikVK1ZM5cqV04wZM3ThwgX1798/13hee+01lStXTlWqVJH053u7Z86cqccffzxfxxQAAAAAcHO4Nem22+268847NXv2bOse5vDwcA0cOFBjx47NVxuPPvqoEhMT1b17dxmGoZ49e2rIkCFav369U7kWLVqoYsWKatq0qdLT09WjRw9NmjTJWj958mSVKFFC06ZN0+HDhxUUFKR69erlK47Zs2erX79+atSokUJDQ/X0009nefr3vHnzNHbsWA0ZMkS//fabypUrZ7VdpkwZxcbG6plnntG//vUvPfzww4qPj9cLL7wgh8OhPn366OzZs4qOjtbGjRsVHBycazwOh0NjxozRzz//rCJFiugf//iHXnjhBT3yyCP5OqYAAAAAgJvD7U8v/yvExMTo999/15o1a9wdyi3jypP4eHo5bhU8vRz5wdPLgetDn8WthP6Kv0p+n15OLwQAAAAAwEVIugEAAAAAcJFC8cowV4uPj3d3CAAAAACA2xAj3QAAAAAAuAhJNwAAAAAALnJbXF6O6/dmf5sCAmzuDgPIk2k6lHbGSwHBNhkG5xORPbu3uyMAAAC3G5Ju5Cq8mIeCgkhgUPg5HFLKZQ+VCPHg9SAAAAAoNPhlCgAAAACAi5B0AwAAAADgIiTdAAAAAAC4CPd0I1fHTjuUesnh7jCAPJmmQ2lpDv3h6ZBhuDsauJPdWwqxc04ZAAAUDiTdyNWguHR5eqe7OwwgT4YcivTP1OGz6TK5iOe2tnSwTSF2d0cBAADwJ36ZAgAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICLkHQDAAAAAOAiJN23AMMwtGbNGneHAQAAAAAooEKZdO/cuVOenp5q06aNu0NxkpycLMMwlJiY6LR80qRJMgxDhmHIw8NDpUuXVq9evXTs2LECtT9p0iTVqVMny/ITJ06obdu2NxA5AAAAAMAdCmXSvXDhQj3++OP64osvdPToUXeHky/Vq1fXiRMn9N///lfvv/++9u/fr27dut2UtsPCwmSz2W5KWwAAAACAv06hS7rPnz+v5cuXa/DgwbrvvvsUHx9vrTtz5ox69eql4sWLy8fHRxUrVtSiRYsk/W8UetmyZWrUqJG8vb1VvXp1bdu2zan9AwcOqF27drLb7SpZsqT69OmjU6dOWesdDoemT5+uqKgo2Ww2lStXTlOmTJEkVahQQZJUt25dGYahu+++26pXpEgRhYWFqXTp0vrnP/+pgQMH6quvvlJaWppV5umnn1alSpXk6+uryMhIjR8/XpmZmZKk+Ph4xcbG6rvvvrNGza/s+7WXl+/fv1/33HOPfHx8FBISokGDBuncuXPW+piYGHXq1EkzZ85UqVKlFBISoscee8zaFgAAAADgr1Hoku73339flStXVuXKldW7d28tWrRIpmlKksaPH68DBw5o/fr1SkpK0rx58xQaGupU/6mnntLIkSO1d+9eNWrUSPfff79+++03SX9ept2sWTPVqVNHe/bs0YYNG3Ty5EmnEekxY8Zo+vTp1rbeffddlSxZUpK0a9cuSdLmzZt14sQJrV69Ott9+OWXX7R69Wp5enrK09PTWu7v76/4+HgdOHBAc+bM0YIFCzR79mxJUvfu3TVy5EhrxPzEiRPq3r17lrYvXLigNm3aKDg4WLt379aKFSu0efNmDR061Knc1q1bdejQIW3dulWLFy9WfHy80wmMa6WnpystLc1pAgAAAADcmCLuDuBacXFx6t27tySpTZs2OnfunLZs2aKWLVvq6NGjqlu3rqKjoyVJERERWeoPHTpUXbp0kSTNmzdPGzZsUFxcnEaPHq158+apXr16mjp1qlV+4cKFCg8P148//qhSpUppzpw5evXVV9W3b19J0j/+8Q81adJEklS8eHFJUkhIiMLCwpy2u3//ftntdjkcDl28eFGSNGzYMPn5+Vllxo0bZ32OiIjQyJEj9f7772v06NHy8fGR3W63RsxzsnTpUl28eFFvv/221farr76qDh06aPr06dYJguDgYL366qvy9PRUlSpV1L59e23ZskUDBw7Mtt1p06YpNjY2x+0CAAAAAAquUI10Hzx4ULt27VKPHj0k/XnJdvfu3bVw4UJJ0uDBg7Vs2TLVqVNHo0eP1s6dO7O00bBhQ+tzkSJFFB0draSkJEnSN998o61bt8put1tTlSpVJEmHDh1SUlKS0tPT1aJFiwLHXrlyZSUmJmr37t2aMmWK6tSpY12WfsXKlSvVpEkThYWFyW63a/z48QW+Zz0pKUm1a9d2SuYbN24sh8OhgwcPWsuqV6/uNMpeqlQppaSk5NjumDFjlJqaak0FfQgcAAAAACCrQjXSHRcXp0uXLqlMmTLWMtM05eXlpTNnzqht27Y6cuSI1q1bp82bN6tFixZ67LHHNHPmzFzbNQxD0p/3a18ZEb5WqVKldPjw4euOvWjRooqKipL0Z8L7n//8R4MHD9aSJUskSV999ZV69Oih2NhYtW7dWoGBgVq2bJlmzZpVoO2Ypmntz7WuXu7l5ZVlncPhyLFdm83Gw9oAAAAA4CYrNCPdly5d0ttvv61Zs2YpMTHRmr777juVL19eS5culfTnJd4xMTF655139PLLL+vNN990auerr75yavObb76xRrPr1aunH374QREREYqKinKa/Pz8VLFiRfn4+GjLli3Zxli0aFFJ0uXLl/Pcn/Hjx+u9997Tt99+K0nasWOHypcvr2effVbR0dGqWLGijhw5kqX9vNquVq2aEhMTdf78eWvZjh075OHhoUqVKuUZFwAAAADgr1Noku61a9fqzJkz6t+/v2rUqOE0de3aVXFxcZowYYI+/PBD/fTTT/rhhx+0du1aVa1a1amd1157TR988IH+/e9/67HHHtOZM2fUr18/SdJjjz2m06dPq2fPntq1a5cOHz6sTz/9VP369dPly5fl7e2tp59+WqNHj9bbb7+tQ4cO6auvvlJcXJwkqUSJEvLx8bEewJaamprj/kRGRqpjx46aMGGCJCkqKkpHjx7VsmXLdOjQIc2dO1cffPCBU52IiAj9/PPPSkxM1KlTp5Senp6l3V69esnb21t9+/bV999/r61bt+rxxx9Xnz59rPu5AQAAAACFQ6FJuuPi4tSyZUsFBgZmWdelSxclJiaqSJEiGjNmjGrVqqWmTZvK09NTy5Ytcyr7wgsvaPr06apdu7YSEhL04YcfWk84L126tHbs2KHLly+rdevWqlGjhp544gkFBgbKw+PPQzF+/HiNHDlSEyZMUNWqVdW9e3frXugiRYpo7ty5euONN1S6dGl17Ngx130aOXKk1q1bp6+//lodO3bU8OHDNXToUNWpU0c7d+7U+PHjs+xnmzZt1Lx5cxUvXlzvvfdeljZ9fX21ceNGnT59WvXr11fXrl3VokULvfrqq/k/2AAAAACAv4RhXnkf1y0uOTlZFSpU0N69e1WnTh13h3PLS0tLU2BgoO4ae1ye3kHuDgfIkyGHIv1P6fDZUJmF53wi3GDpYJvKhxbuPuBwOJSSkqISJUpYJ32Bwow+i1sJ/RV/lSs5U2pqqgICAnIsRy8EAAAAAMBFSLoBAAAAAHCRQvXKsBsRERGhv8mV8gAAAACAvwlGugEAAAAAcBGSbgAAAAAAXORvc3k5XOPN/jYFBNjcHQaQJ9N0KO2MlwKCbTIMzifezuze7o4AAADgf0i6kavwYh4KCiKBQeHncEgplz1UIsSD14MAAACg0OCXKQAAAAAALkLSDQAAAACAi5B0AwAAAADgIiTdAAAAAAC4CA9SQ66OnXYo9ZLD3WEAeTJNh9LSHPrD0yHDcHc0cBe7txRi53wyAAAoPEi6katBceny9E53dxhAngw5FOmfqcNn02VyEc9ta+lgm0Ls7o4CAADgf/hlCgAAAACAi5B0AwAAAADgIiTdAAAAAAC4CEk3AAAAAAAuQtINAAAAAICLkHS72c6dO+Xp6ak2bdq4OxQAAAAAwE1G0u1mCxcu1OOPP64vvvhCR48ezbGcaZq6dOnSXxgZAAAAAOBGkXS70fnz57V8+XINHjxY9913n+Lj461127Ztk2EY2rhxo6Kjo2Wz2ZSQkCDTNDVjxgxFRkbKx8dHtWvX1sqVK616ly9fVv/+/VWhQgX5+PiocuXKmjNnjhv2DgAAAABQxN0B3M7ef/99Va5cWZUrV1bv3r31+OOPa/z48TIMwyozevRozZw5U5GRkQoKCtK4ceO0evVqzZs3TxUrVtTnn3+u3r17q3jx4mrWrJkcDofKli2r5cuXKzQ0VDt37tSgQYNUqlQpdevWLcdY0tPTlZ6ebs2npaW5dN8BAAAA4HZA0u1GcXFx6t27tySpTZs2OnfunLZs2aKWLVtaZZ577jm1atVK0p8j4y+99JI+++wzNWzYUJIUGRmpL774Qm+88YaaNWsmLy8vxcbGWvUrVKignTt3avny5bkm3dOmTXOqBwAAAAC4cVxe7iYHDx7Url271KNHD0lSkSJF1L17dy1cuNCpXHR0tPX5wIED+uOPP9SqVSvZ7XZrevvtt3Xo0CGr3Pz58xUdHa3ixYvLbrdrwYIFud4vLkljxoxRamqqNR07duwm7i0AAAAA3J4Y6XaTuLg4Xbp0SWXKlLGWmaYpLy8vnTlzxlrm5+dnfXY4HJKkdevWOdWTJJvNJklavny5hg8frlmzZqlhw4by9/fXiy++qK+//jrXeGw2m9UGAAAAAODmIOl2g0uXLuntt9/WrFmzdO+99zqt69Kli5YuXaoaNWpkqVetWjXZbDYdPXpUzZo1y7bthIQENWrUSEOGDLGWXT0KDgAAAAD465B0u8HatWt15swZ9e/fX4GBgU7runbtqri4OM2ePTtLPX9/f40aNUrDhw+Xw+FQkyZNlJaWpp07d8put6tv376KiorS22+/rY0bN6pChQpasmSJdu/erQoVKvxVuwcAAAAA+P+4p9sN4uLi1LJlyywJt/TnSHdiYqK+/fbbbOtOnjxZEyZM0LRp01S1alW1bt1aH3/8sZVUP/roo3rggQfUvXt33Xnnnfrtt9+cRr0BAAAAAH8dwzRN091BoPBJS/t/7d17XFTlvsfx74A4IMNVQrwgwkZTvCteUBOpCNPalu7Ku6SWmealnde8ZqbZ1srTSa3dBjNv7TSP2im3Yrr1qGkYSkrqMU1LjDQFlERk1vmjl+s0cRG1cUA/79drvWKe9VvP+s3ieZG/9ax5Jkd+fn5qO/GU3D39XZ0OcE0W2RXhc0bf5gbJ4H7iHWvpUKvCgsr/799utysrK0vBwcFycyv/+QKMWVQkjFfcKldrpuzsbPn6+pYYxygEAAAAAMBJKLoBAAAAAHASim4AAAAAAJyEohsAAAAAACeh6AYAAAAAwEn4nm6U6p1BVvn6Wl2dBnBNhmFXzjkP+QZYZbFwP/FOZfN0dQYAAACOKLpRqtBAN/n7U8Cg/LPbpaxCNwVXdePrQQAAAFBu8C9TAAAAAACchKIbAAAAAAAnoegGAAAAAMBJKLoBAAAAAHASFlJDqU7+bFf2Fbur0wCuyTDsysmx65K7XRaLq7PBtdg8pao27vsCAIDbH0U3SvX0e/ly98x3dRrANVlkV4RPgb7NzZfBQzzl3tKhVlW1uToLAAAA5+NfpgAAAAAAOAlFNwAAAAAATkLRDQAAAACAk1B0AwAAAADgJOW+6LZYLFqzZo2r0wAAAAAA4Lq5vOjOysrSkCFDVLt2bVmtVoWEhCghIUE7d+50dWoOjh8/LovFUmTr27fvTfe9ZcsWWSwWnT9//uYTlbR69WrFx8frrrvukq+vr2JiYrRhw4Y/pG8AAAAAQNm5/CvDevTooYKCAi1evFgRERH68ccflZKSop9//tnVqRVr06ZNatiwofnay8vLhdk4MgxDhYWF+ve//634+Hi98sor8vf3V1JSkh5++GF98cUXat68uavTBAAAAIA7hktnus+fP6/t27fr1VdfVVxcnMLCwtS6dWtNmDBBXbt2NePOnDmjRx99VFWqVFHdunW1du1ac19hYaEGDRqk8PBweXl56e6779abb77pcJ7ExEQ98sgjmj59uoKDg+Xr66shQ4bo8uXLZoxhGJozZ44iIiLk5eWlpk2b6qOPPiqSc9WqVRUSEmJufn5+Onr0qLp166Zq1arJZrOpVatW2rRpk8Nx+fn5Gjt2rEJDQ2W1WlW3bl299957On78uOLi4iRJAQEBslgsSkxMNI8ZMWKEgoOD5enpqQ4dOmjPnj1mn1dnyDds2KDo6GhZrVZt27ZNb7zxhsaOHatWrVqpbt26euWVV1S3bl2tW7fuxn9ZAAAAAIDr5tKi22azyWazac2aNcrPzy8xbvr06Xr88ce1f/9+denSRX369DFnwu12u2rVqqUPP/xQBw8e1JQpUzRx4kR9+OGHDn2kpKQoIyNDn3/+uZYvX66PP/5Y06dPN/dPmjRJSUlJWrBggQ4cOKDRo0erb9++2rp16zXfx4ULF9SlSxdt2rRJX331lRISEvTwww/rxIkTZkz//v21YsUKzZ8/XxkZGVq4cKFsNptCQ0O1atUqSdKhQ4eUmZlp3jQYO3asVq1apcWLF2vv3r2KjIxUQkJCkacAxo4dq1mzZikjI0NNmjQpkp/dbldubq4CAwNLfA/5+fnKyclx2AAAAAAAN8diGIbhygRWrVqlp556Sr/88otatGih2NhY9ezZ0yweLRaLJk2apBkzZkiSLl68KB8fH/33f/+3OnfuXGyfw4YN048//mjOVCcmJmrdunU6efKkqlSpIklauHChxowZo+zsbP3yyy8KCgrS5s2bFRMTY/YzePBg5eXladmyZTp+/Lg5m+7m9v/3KrZt21bsI9sNGzbU0KFDNXz4cB0+fFh33323Nm7cqPvvv79I7JYtWxQXF6dz587J39/ffJ8BAQFKTk5W7969JUkFBQWqU6eORo0apTFjxpjHrVmzRt26dSvxGr/22muaPXu2MjIyFBwcXGzMtGnTHG5CXNV24im5e/qX2DdQXlhkV4TPGX2bGyTD9ctV4BqWDrUqLOjO/T3Z7XZlZWUpODjY4f8pQHnFmEVFwnjFrZKTkyM/Pz9lZ2fL19e3xDiXj8IePXro1KlTWrt2rRISErRlyxa1aNFCycnJZsxvZ2+9vb3l4+OjrKwss23hwoWKjo7WXXfdJZvNpnfffddhllmSmjZtahbckhQTE6MLFy7o5MmTOnjwoC5duqT4+Hhz9t1ms+n999/X0aNHHfpZuXKl0tLSzC0qKkoXL17U2LFjFRUVJX9/f9lsNn3zzTdmDmlpaXJ3d1dsbGyZr8vRo0dVUFCg9u3bm20eHh5q3bq1MjIyHGKjo6NL7Gf58uWaNm2aVq5cWWLBLUkTJkxQdna2uZ08ebLMuQIAAAAAiufyhdQkydPTU/Hx8YqPj9eUKVM0ePBgTZ061fxss4eHh0O8xWKR3W6XJH344YcaPXq05s6dq5iYGPn4+Oi1117TF198UaZz/7avTz75RDVr1nTYb7VaHV6HhoYqMjLSoW306NHasGGD/va3vykyMlJeXl76y1/+Yn5m/EYWW7v6AILFYinS/vs2b2/vYvtYuXKlBg0apH/+85/FzrD/ltVqLfJeAQAAAAA3x+Uz3cW5OntcFtu2bVO7du307LPPqnnz5oqMjCwyOy1J+/bt0y+//GK+3rVrl2w2m2rVqqWoqChZrVadOHFCkZGRDltoaGiZckhMTNSjjz6qxo0bKyQkRMePHzf3N27cWHa7vcTPh1euXFnSr4vCXRUZGanKlStr+/btZltBQYG+/PJLNWjQ4Jo5LV++XImJiVq2bJnDonQAAAAAgFvHpTPdZ8+e1WOPPaaBAweqSZMm8vHx0Zdffqk5c+aU+hnl34qMjNT777+vDRs2KDw8XEuWLNGePXsUHh7uEHf58mUNGjRIkyZN0nfffaepU6dq+PDhcnNzk4+Pj1544QWNHj1adrtdHTp0UE5Ojnbs2CGbzaYBAwZcM4fVq1fr4YcflsVi0eTJk83Zc0mqU6eOBgwYoIEDB2r+/Plq2rSpvvvuO2VlZenxxx9XWFiYLBaL1q9fry5dusjLy0s2m01Dhw7VmDFjFBgYqNq1a2vOnDnKy8vToEGDSs1n+fLl6t+/v9588021bdtWp0+flvTrjLufn1+ZrisAAAAA4Oa5tOi22Wxq06aNXn/9dfMzzKGhoXrqqac0ceLEMvXxzDPPKC0tTU888YQsFot69eqlZ599Vp9++qlD3H333ae6deuqY8eOys/PV8+ePTVt2jRz/4wZMxQcHKxZs2bp22+/lb+/v1q0aFGmPF5//XUNHDhQ7dq1U1BQkMaNG1dk9e8FCxZo4sSJevbZZ3X27FnVrl3b7LtmzZqaPn26xo8fryeffFL9+/dXcnKyZs+eLbvdrn79+ik3N1fR0dHasGGDAgICSs1n0aJFunLlioYNG6Zhw4aZ7QMGDHD4rDwAAAAAwLlcvnr5rZCYmKjz589rzZo1rk6lwri6Eh+rl6OiYPXyioXVy1lZFxULYxYVCeMVt0qFWb0cAAAAAIDbFUU3AAAAAABOUi6+MszZ+BwzAAAAAMAVmOkGAAAAAMBJKLoBAAAAAHCSO+Lxcty4dwZZ5etrdXUawDUZhl055zzkG2CVxcL9xPLO5unqDAAAAG4Nim6UKjTQTf7+FDAo/+x2KavQTcFV3fh6EAAAAJQb/MsUAAAAAAAnoegGAAAAAMBJKLoBAAAAAHASim4AAAAAAJyEhdRQqpM/25V9xe7qNHCHsnlKVW3cGwQAAEDFRdGNUj39Xr7cPfNdnQbuUEuHWlXV5uosAAAAgBvHFBIAAAAAAE5C0Q0AAAAAgJNQdAMAAAAA4CQU3QAAAAAAOAlFt4tt2bJFFotF58+fL1N8p06dNGrUKKfmBAAAAAD4Y1TIojsxMVEWi6XI1rlzZ1enJkmaNm2amjVrVqbYdu3aKTMzU35+fmWKX716tWbMmGG+rlOnjt54440byBIAAAAA4GwV9ivDOnfurKSkJIc2q9XqomxuTEFBgSpXrqyQkJAyHxMYGOjEjAAAAAAAf6QKOdMt/Vpgh4SEOGwBAQGSpPPnz+vpp59WtWrV5OnpqUaNGmn9+vXmsf/zP/+j2NhYValSRQEBAUpISNC5c+ckSYZhaM6cOYqIiJCXl5eaNm2qjz76yDz26uPgKSkpio6OVpUqVdSuXTsdOnRIkpScnKzp06dr37595gx8cnKyJMlisWjhwoXq1q2bvL299fLLLxf7eHlp+f328fJOnTrpu+++0+jRo81zXbx4Ub6+vg45S9K6devk7e2t3NzcP/T3AAAAAAAoWYUtuktit9v14IMPaseOHfrggw908OBBzZ49W+7u7pKktLQ03XfffWrYsKF27typ7du36+GHH1ZhYaEkadKkSUpKStKCBQt04MABjR49Wn379tXWrVsdzvPiiy9q7ty5+vLLL1WpUiUNHDhQkvTEE0/or3/9qxo2bKjMzExlZmbqiSeeMI+bOnWqunXrpvT0dPOY37pWfr+1evVq1apVSy+99JJ5Lm9vb/Xs2bPIUwBJSUn6y1/+Ih8fn2KvW35+vnJychw2AAAAAMDNqbCPl69fv142m82hbdy4cWrTpo12796tjIwM1atXT5IUERFhxsyZM0fR0dF6++23zbaGDRtKki5evKh58+Zp8+bNiomJMY/dvn27Fi1apNjYWPOYmTNnmq/Hjx+vrl276tKlS/Ly8pLNZlOlSpWKfWy8d+/eDsX2sWPHHPaXlt/vBQYGyt3dXT4+Pg7nGjx4sNq1a6dTp06pRo0aOnPmjNavX6+NGzcW248kzZo1S9OnTy9xPwAAAADg+lXYme64uDilpaU5bMOGDVNaWppq1aplFty/d3UmuTgHDx7UpUuXFB8fL5vNZm7vv/++jh496hDbpEkT8+fq1atLkrKysq6Zd3R0dKn7S8uvrFq3bq2GDRvq/ffflyQtWbJEtWvXVseOHUs8ZsKECcrOzja3kydP3lQOAAAAAIAKPNPt7e2tyMjIIu1eXl6lHlfafrvdLkn65JNPVLNmTYd9v1+kzcPDw/zZYrE4HF8ab2/vG87vegwePFhvvfWWxo8fr6SkJD355JNmnsWxWq0VbiE6AAAAACjvKuxMd0maNGmi77//XocPHy5xf0pKSrH7oqKiZLVadeLECUVGRjpsoaGhZc6hcuXKxX4Gu6z5l5Tf9Zyrb9++OnHihObPn68DBw5owIABN5QPAAAAAODGVdiZ7vz8fJ0+fdqhrVKlSoqNjVXHjh3Vo0cPzZs3T5GRkfrmm2/M7/GeMGGCGjdurGeffVbPPPOMKleurM8//1yPPfaYgoKC9MILL2j06NGy2+3q0KGDcnJytGPHDtlstjIXrnXq1NGxY8fMR919fHzKPIt8rfyKO9e///1v9ezZU1ar1YwJCAhQ9+7dNWbMGD3wwAOqVatWmc4PAAAAAPjjVNiZ7s8++0zVq1d32Dp06CBJWrVqlVq1aqVevXopKipKY8eONWeD69Wrp3/961/at2+fWrdurZiYGP3Xf/2XKlX69f7DjBkzNGXKFM2aNUsNGjRQQkKC1q1bp/Dw8DLn1qNHD3Xu3FlxcXG66667tHz58jIfe638fu+ll17S8ePH9ac//Ul33XWXw75Bgwbp8uXLxa6SDgAAAABwPothGIark4BzLF26VCNHjtSpU6dUuXLl6zo2JydHfn5+ajvxlNw9/Z2TIHANS4daFRZUtnuDdrtdWVlZCg4Olptbhb2fiDsE4xUVDWMWFQnjFbfK1ZopOztbvr6+JcZV2MfLUbK8vDwdO3ZMs2bN0pAhQ6674AYAAAAA/DG49XMbmjNnjpo1a6Zq1appwoQJrk4HAAAAAO5YFN23oWnTpqmgoEApKSmy2WyuTgcAAAAA7lgU3QAAAAAAOAlFNwAAAAAATsJCaijVO4Os8vUt23eMA380m6erMwAAAABuDkU3ShUa6CZ/fx6IAAAAAIAbQTUFAAAAAICTUHQDAAAAAOAkFN0AAAAAADgJRTcAAAAAAE7CQmoo1cmf7cq+Ynd1GriN2Dylqjbu9wEAAODOQNGNUj39Xr7cPfNdnQZuI0uHWlXV5uosAAAAgFuD6SYAAAAAAJyEohsAAAAAACeh6AYAAAAAwEkougEAAAAAcJIKX3QnJyfL39/fKX1PmzZNzZo1c0rfAAAAAIDbn1OL7sTERFksliJb586d/7BzPPHEEzp8+HCZ448fP+6Qi4+Pjxo2bKhhw4bpyJEjDrEvvPCCUlJS/rBcb4bFYtGaNWtcnQYAAAAA4Do4/SvDOnfurKSkJIc2q9X6h/Xv5eUlLy+v6z5u06ZNatiwofLy8pSenq4333xTTZs21bp163TfffdJkmw2m2y22+u7jQoKCuTh4eHqNAAAAADgjuD0x8utVqtCQkIctoCAAEnSkSNH1LFjR3l6eioqKkobN250mNHdsmWLLBaLzp8/b/aXlpYmi8Wi48ePS3J8vPzQoUOyWCz65ptvHHKYN2+e6tSpI8MwzLaqVasqJCREERER6tatmzZt2qQ2bdpo0KBBKiwslFT08fI9e/YoPj5eQUFB8vPzU2xsrPbu3etwLovFokWLFumhhx5SlSpV1KBBA+3cuVP/+7//q06dOsnb21sxMTE6evSow3Hr1q1Ty5Yt5enpqYiICE2fPl1XrlyRJNWpU0eS9Oijj8pisZivr3Xc1XwWLlyobt26ydvbWy+//PK1f2kAAAAAgD+Eyz7Tbbfb1b17d7m7u2vXrl1auHChxo0bd1N93n333WrZsqWWLl3q0L5s2TL17t1bFoulxGPd3Nw0cuRIfffdd0pNTS02Jjc3VwMGDNC2bdu0a9cu1a1bV126dFFubq5D3IwZM9S/f3+lpaWpfv366t27t4YMGaIJEyboyy+/lCQNHz7cjN+wYYP69u2rESNG6ODBg1q0aJGSk5M1c+ZMSb8W+5KUlJSkzMxM8/W1jrtq6tSp6tatm9LT0zVw4MBi31t+fr5ycnIcNgAAAADAzXF60b1+/XrzMe2r24wZM7Rp0yZlZGRoyZIlatasmTp27KhXXnnlps/Xp08fLVu2zHx9+PBhpaamqm/fvtc8tn79+pJkzqL/3r333qu+ffuqQYMGatCggRYtWqS8vDxt3brVIe7JJ5/U448/rnr16mncuHE6fvy4+vTpo4SEBDVo0EAjR47Uli1bzPiZM2dq/PjxGjBggCIiIhQfH68ZM2Zo0aJFkqS77rpLkuTv76+QkBDz9bWOu6p3794aOHCgIiIiFBYWVux7mzVrlvz8/MwtNDT0mtcLAAAAAFA6p3+mOy4uTgsWLHBoCwwM1JIlS1S7dm3VqlXLbI+Jibnp8/Xs2VNjxozRrl271LZtWy1dulTNmjVTVFTUNY+9+vh5STPiWVlZmjJlijZv3qwff/xRhYWFysvL04kTJxzimjRpYv5crVo1SVLjxo0d2i5duqScnBz5+voqNTVVe/bscZihLiws1KVLl5SXl6cqVaoUm09Zj4uOjr7me58wYYKef/5583VOTg6FNwAAAADcJKcX3d7e3oqMjCzS/tvPV1/1+2LXzc2tSGxBQUGp56tevbri4uK0bNkytW3bVsuXL9eQIUPKlGtGRoYkKTw8vNj9iYmJ+umnn/TGG28oLCxMVqtVMTExunz5skPcbxcqu/qeimuz2+3mf6dPn67u3bsXOaenp2eJ+Zb1OG9v7xL7uMpqtf6hC9wBAAAAAG5B0V2SqKgonThxQqdOnVKNGjUkSTt37nSIufoYdWZmprn4Wlpa2jX77tOnj8aNG6devXrp6NGj6tmz5zWPsdvtmj9/vsLDw9W8efNiY7Zt26a3335bXbp0kSSdPHlSZ86cuWbf19KiRQsdOnSo2JsTV3l4eJgLvF3PcQAAAAAA13F60Z2fn6/Tp087nrRSJd1///26++671b9/f82dO1c5OTl68cUXHeIiIyMVGhqqadOm6eWXX9aRI0c0d+7ca56ze/fuGjp0qIYOHaq4uDjVrFmzSMzZs2d1+vRp5eXl6euvv9Ybb7yh3bt365NPPpG7u3ux/UZGRmrJkiWKjo5WTk6OxowZc0NfV/Z7U6ZM0UMPPaTQ0FA99thjcnNz0/79+5Wenm6uNl6nTh2lpKSoffv2slqtCggIKNNxAAAAAADXcfpCap999pmqV6/usHXo0EFubm76+OOPlZ+fr9atW2vw4MFFVt328PDQ8uXL9c0336hp06Z69dVXy1RM+vr66uGHH9a+ffvUp0+fYmPuv/9+Va9eXY0bN9b48ePVoEED7d+/X3FxcSX2+49//EPnzp1T8+bN1a9fP40YMULBwcHXd0GKkZCQoPXr12vjxo1q1aqV2rZtq3nz5jksejZ37lxt3LhRoaGh5kx8WY4DAAAAALiOxSjuw9UuZLFY9PHHH+uRRx5xdSp3tJycHPn5+antxFNy9/R3dTq4jSwdalVY0B9/v89utysrK0vBwcHmehBAecV4RUXDmEVFwnjFrXK1ZsrOzpavr2+JcYxCAAAAAACchKIbAAAAAAAncdnq5SUpZ0+7AwAAAABww5jpBgAAAADASSi6AQAAAABwknL3eDnKl3cGWeXra3V1GriN2DxdnQEAAABw61B0o1ShgW7y9+eBCAAAAAC4EVRTAAAAAAA4CUU3AAAAAABOQtENAAAAAICTUHQDAAAAAOAkFN0AAAAAADgJRTcAAAAAAE5C0Q0AAAAAgJNQdAMAAAAA4CQU3QAAAAAAOAlFNwAAAAAATkLRDQAAAACAk1B0AwAAAADgJBTdAAAAAAA4CUU3AAAAAABOQtENAAAAAICTUHQDAAAAAOAkFN0AAAAAADgJRTcAAAAAAE5C0Q0AAAAAgJNQdAMAAAAA4CSVXJ0AyifDMCRJOTk5cnPj3gzKP7vdrtzcXHl6ejJmUe4xXlHRMGZRkTBecavk5ORI+v/aqSQU3SjW2bNnJUlhYWEuzgQAAAAAyq/c3Fz5+fmVuJ+iG8UKDAyUJJ04caLUAQSUFzk5OQoNDdXJkyfl6+vr6nSAUjFeUdEwZlGRMF5xqxiGodzcXNWoUaPUOIpuFOvqozh+fn78sUKF4uvry5hFhcF4RUXDmEVFwnjFrVCWCUo+5AAAAAAAgJNQdAMAAAAA4CQU3SiW1WrV1KlTZbVaXZ0KUCaMWVQkjFdUNIxZVCSMV5Q3FuNa65sDAAAAAIAbwkw3AAAAAABOQtENAAAAAICTUHQDAAAAAOAkFN0o1ttvv63w8HB5enqqZcuW2rZtm6tTwh1m1qxZatWqlXx8fBQcHKxHHnlEhw4dcogxDEPTpk1TjRo15OXlpU6dOunAgQMOMfn5+XruuecUFBQkb29v/fnPf9b3339/K98K7lCzZs2SxWLRqFGjzDbGLMqTH374QX379lXVqlVVpUoVNWvWTKmpqeZ+xivKkytXrmjSpEkKDw+Xl5eXIiIi9NJLL8lut5sxjFmUVxTdKGLlypUaNWqUXnzxRX311Ve655579OCDD+rEiROuTg13kK1bt2rYsGHatWuXNm7cqCtXruiBBx7QxYsXzZg5c+Zo3rx5euutt7Rnzx6FhIQoPj5eubm5ZsyoUaP08ccfa8WKFdq+fbsuXLighx56SIWFha54W7hD7NmzR++8846aNGni0M6YRXlx7tw5tW/fXh4eHvr000918OBBzZ07V/7+/mYM4xXlyauvvqqFCxfqrbfeUkZGhubMmaPXXntN//Ef/2HGMGZRbhnA77Ru3dp45plnHNrq169vjB8/3kUZAYaRlZVlSDK2bt1qGIZh2O12IyQkxJg9e7YZc+nSJcPPz89YuHChYRiGcf78ecPDw8NYsWKFGfPDDz8Ybm5uxmeffXZr3wDuGLm5uUbdunWNjRs3GrGxscbIkSMNw2DMonwZN26c0aFDhxL3M15R3nTt2tUYOHCgQ1v37t2Nvn37GobBmEX5xkw3HFy+fFmpqal64IEHHNofeOAB7dixw0VZAVJ2drYkKTAwUJJ07NgxnT592mGsWq1WxcbGmmM1NTVVBQUFDjE1atRQo0aNGM9wmmHDhqlr1666//77HdoZsyhP1q5dq+joaD322GMKDg5W8+bN9e6775r7Ga8obzp06KCUlBQdPnxYkrRv3z5t375dXbp0kcSYRflWydUJoHw5c+aMCgsLVa1aNYf2atWq6fTp0y7KCnc6wzD0/PPPq0OHDmrUqJEkmeOxuLH63XffmTGVK1dWQEBAkRjGM5xhxYoV2rt3r/bs2VNkH2MW5cm3336rBQsW6Pnnn9fEiRO1e/dujRgxQlarVf3792e8otwZN26csrOzVb9+fbm7u6uwsFAzZ85Ur169JPE3FuUbRTeKZbFYHF4bhlGkDbhVhg8frv3792v79u1F9t3IWGU8wxlOnjypkSNH6l//+pc8PT1LjGPMojyw2+2Kjo7WK6+8Iklq3ry5Dhw4oAULFqh///5mHOMV5cXKlSv1wQcfaNmyZWrYsKHS0tI0atQo1ahRQwMGDDDjGLMoj3i8HA6CgoLk7u5e5G5fVlZWkTuHwK3w3HPPae3atfr8889Vq1Ytsz0kJESSSh2rISEhunz5ss6dO1diDPBHSU1NVVZWllq2bKlKlSqpUqVK2rp1q+bPn69KlSqZY44xi/KgevXqioqKcmhr0KCBuWgqf2NR3owZM0bjx49Xz5491bhxY/Xr10+jR4/WrFmzJDFmUb5RdMNB5cqV1bJlS23cuNGhfePGjWrXrp2LssKdyDAMDR8+XKtXr9bmzZsVHh7usD88PFwhISEOY/Xy5cvaunWrOVZbtmwpDw8Ph5jMzEx9/fXXjGf84e677z6lp6crLS3N3KKjo9WnTx+lpaUpIiKCMYtyo3379kW+hvHw4cMKCwuTxN9YlD95eXlyc3MsXdzd3c2vDGPMolxz0QJuKMdWrFhheHh4GO+9955x8OBBY9SoUYa3t7dx/PhxV6eGO8jQoUMNPz8/Y8uWLUZmZqa55eXlmTGzZ882/Pz8jNWrVxvp6elGr169jOrVqxs5OTlmzDPPPGPUqlXL2LRpk7F3717j3nvvNZo2bWpcuXLFFW8Ld5jfrl5uGIxZlB+7d+82KlWqZMycOdM4cuSIsXTpUqNKlSrGBx98YMYwXlGeDBgwwKhZs6axfv1649ixY8bq1auNoKAgY+zYsWYMYxblFUU3ivWf//mfRlhYmFG5cmWjRYsW5tc0AbeKpGK3pKQkM8ZutxtTp041QkJCDKvVanTs2NFIT0936OeXX34xhg8fbgQGBhpeXl7GQw89ZJw4ceIWvxvcqX5fdDNmUZ6sW7fOaNSokWG1Wo369esb77zzjsN+xivKk5ycHGPkyJFG7dq1DU9PTyMiIsJ48cUXjfz8fDOGMYvyymIYhuHKmXYAAAAAAG5XfKYbAAAAAAAnoegGAAAAAMBJKLoBAAAAAHASim4AAAAAAJyEohsAAAAAACeh6AYAAAAAwEkougEAAAAAcBKKbgAAAAAAnISiGwAAAAAAJ6HoBgCggkpMTNQjjzzi6jRKdPz4cVksFqWlpbk6lTLJysrSkCFDVLt2bVmtVoWEhCghIUE7d+50dWoAgAqskqsTAAAAt5/Lly+7OoXr1qNHDxUUFGjx4sWKiIjQjz/+qJSUFP38889OO+fly5dVuXJlp/UPAHA9ZroBALhNdOrUSc8995xGjRqlgIAAVatWTe+8844uXryoJ598Uj4+PvrTn/6kTz/91Dxmy5Ytslgs+uSTT9S0aVN5enqqTZs2Sk9Pd+h71apVatiwoaxWq+rUqaO5c+c67K9Tp45efvllJSYmys/PT0899ZTCw8MlSc2bN5fFYlGnTp0kSXv27FF8fLyCgoLk5+en2NhY7d2716E/i8Wiv//973r00UdVpUoV1a1bV2vXrnWIOXDggLp27SpfX1/5+Pjonnvu0dGjR839SUlJatCggTw9PVW/fn29/fbbJV678+fPa/v27Xr11VcVFxensLAwtW7dWhMmTFDXrl0d4p5++mlVq1ZNnp6eatSokdavX39T10mSduzYoY4dO8rLy0uhoaEaMWKELl68WGK+AICKg6IbAIDbyOLFixUUFKTdu3frueee09ChQ/XYY4+pXbt22rt3rxISEtSvXz/l5eU5HDdmzBj97W9/0549exQcHKw///nPKigokCSlpqbq8ccfV8+ePZWenq5p06Zp8uTJSk5OdujjtddeU6NGjZSamqrJkydr9+7dkqRNmzYpMzNTq1evliTl5uZqwIAB2rZtm3bt2qW6deuqS5cuys3Ndehv+vTpevzxx7V//3516dJFffr0MWedf/jhB3Xs2FGenp7avHmzUlNTNXDgQF25ckWS9O677+rFF1/UzJkzlZGRoVdeeUWTJ0/W4sWLi71uNptNNptNa9asUX5+frExdrtdDz74oHbs2KEPPvhABw8e1OzZs+Xu7n5T1yk9PV0JCQnq3r279u/fr5UrV2r79u0aPnx4ab9qAEBFYQAAgAppwIABRrdu3czXsbGxRocOHczXV65cMby9vY1+/fqZbZmZmYYkY+fOnYZhGMbnn39uSDJWrFhhxpw9e9bw8vIyVq5caRiGYfTu3duIj493OPeYMWOMqKgo83VYWJjxyCOPOMQcO3bMkGR89dVXpb6PK1euGD4+Psa6devMNknGpEmTzNcXLlwwLBaL8emnnxqGYRgTJkwwwsPDjcuXLxfbZ2hoqLFs2TKHthkzZhgxMTEl5vHRRx8ZAQEBhqenp9GuXTtjwoQJxr59+8z9GzZsMNzc3IxDhw4Ve/yNXqd+/foZTz/9tEPbtm3bDDc3N+OXX34pMV8AQMXATDcAALeRJk2amD+7u7uratWqaty4sdlWrVo1Sb8uGvZbMTEx5s+BgYG6++67lZGRIUnKyMhQ+/btHeLbt2+vI0eOqLCw0GyLjo4uU45ZWVl65plnVK9ePfn5+cnPz08XLlzQiRMnSnwv3t7e8vHxMfNOS0vTPffcIw8PjyL9//TTTzp58qQGDRpkzmDbbDa9/PLLDo+f/16PHj106tQprV27VgkJCdqyZYtatGhhzlSnpaWpVq1aqlevXrHH3+h1Sk1NVXJyskOuCQkJstvtOnbsWIn5AgAqBhZSAwDgNvL7ItRisTi0WSwWSb8+Kn0tV2MNwzB/vsowjCLx3t7eZcoxMTFRP/30k9544w2FhYXJarUqJiamyOJrxb2Xq3l7eXmV2P/VmHfffVdt2rRx2Hf1UfCSeHp6Kj4+XvHx8ZoyZYoGDx6sqVOnKjExsdRzSjd+nex2u4YMGaIRI0YUia1du3ap5wQAlH8U3QAAQLt27TILvHPnzunw4cOqX7++JCkqKkrbt293iN+xY4fq1atXahF7dVXu387yStK2bdv09ttvq0uXLpKkkydP6syZM9eVb5MmTbR48WIVFBQUKc6rVaummjVr6ttvv1WfPn2uq9/fi4qK0po1a8xzfv/99zp8+HCxs903ep1atGihAwcOKDIy8qZyBQCUTzxeDgAA9NJLLyklJUVff/21EhMTFRQUZH4H+F//+lelpKRoxowZOnz4sBYvXqy33npLL7zwQql9BgcHy8vLS5999pl+/PFHZWdnS5IiIyO1ZMkSZWRk6IsvvlCfPn2uOYv8e8OHD1dOTo569uypL7/8UkeOHNGSJUt06NAhSdK0adM0a9Ysvfnmmzp8+LDS09OVlJSkefPmFdvf2bNnde+99+qDDz7Q/v37dezYMf3zn//UnDlz1K1bN0lSbGysOnbsqB49emjjxo06duyYPv30U3322Wc3dZ3GjRunnTt3atiwYUpLS9ORI0e0du1aPffcc9d1TQAA5RNFNwAA0OzZszVy5Ei1bNlSmZmZWrt2rTlT3aJFC3344YdasWKFGjVqpClTpuill15SYmJiqX1WqlRJ8+fP16JFi1SjRg2zeP3HP/6hc+fOqXnz5urXr59GjBih4ODg68q3atWq2rx5sy5cuKDY2Fi1bNlS7777rjnrPXjwYP39739XcnKyGjdurNjYWCUnJ5tfY/Z7NptNbdq00euvv66OHTuqUaNGmjx5sp566im99dZbZtyqVavUqlUr9erVS1FRURo7dqw5k3+j16lJkybaunWrjhw5onvuuUfNmzfX5MmTVb169eu6JgCA8sliFPdhIwAAcEfYsmWL4uLidO7cOfn7+7s6HQAAbjvMdAMAAAAA4CQU3QAAAAAAOAmPlwMAAAAA4CTMdAMAAAAA4CQU3QAAAAAAOAlFNwAAAAAATkLRDQAAAACAk1B0AwAAAADgJBTdAAAAAAA4CUU3AAAAAABOQtENAAAAAICTUHQDAAAAAOAk/wfbjAsg/gZoPQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "if state.feature_importance is not None:\n",
    "    top_n = min(16, len(state.feature_importance))\n",
    "    top_features = state.feature_importance.head(top_n)\n",
    "\n",
    "    fig, ax = plt.subplots(figsize=(10, 6))\n",
    "    ax.barh(top_features['feature'][::-1], top_features['importance'][::-1],\n",
    "            color=\"#3B82F6\", edgecolor='white', linewidth=0.5)\n",
    "    ax.set_title(f\"Native Tree Feature Importance (top {top_n})\", fontweight=\"bold\", fontsize=13)\n",
    "    ax.set_xlabel(\"Importance Score\")\n",
    "    ax.grid(axis='x', alpha=0.3)\n",
    "    plt.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md7b126e73",
   "metadata": {},
   "source": [
    "## 9. Evaluating on Unseen Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "cd2f0b275a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] feature_fraction is set=0.6104217815601655, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.6104217815601655\n",
      "[LightGBM] [Warning] lambda_l2 is set=1.1182701662877744e-07, reg_lambda=0.0 will be ignored. Current value: lambda_l2=1.1182701662877744e-07\n",
      "[LightGBM] [Warning] min_gain_to_split is set=0.21527568422880633, min_split_gain=0.0 will be ignored. Current value: min_gain_to_split=0.21527568422880633\n",
      "[LightGBM] [Warning] lambda_l1 is set=1.2158568318093246e-06, reg_alpha=0.0 will be ignored. Current value: lambda_l1=1.2158568318093246e-06\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8822936933674879, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8822936933674879\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n",
      "Prediction shape : (2723, 7)  (one probability per class per row)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAAKyCAYAAADvmf4QAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjYsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvq6yFwwAAAAlwSFlzAAAPYQAAD2EBqD+naQAAzkJJREFUeJzs3XdUFFcbBvBnaUvvyIIiooCoiL33ghV7b9HYSDQmxI4aW1SMiS2xRUWxIVhi7xo1GkvsvSZ2QWnSBYT5/mCZj3VBF2VZYJ/fOXMOO3Nn5p27w+zefefekQiCIICIiIiIiIigo+kAiIiIiIiICgs2kIiIiIiIiOTYQCIiIiIiIpJjA4mIiIiIiEiODSQiIiIiIiI5NpCIiIiIiIjk2EAiIiIiIiKSYwOJiIiIiIhIjg0kIiIiIiIiOTaQiD7R/v37Ub9+fVhZWUEikUAikWDRokVq32/Tpk3F/QUFBal9f9qAdUrFTVBQkHhON23aVNPhqM3jx4/F45RIJJoOR8nH3of169ejevXqMDc3F8vt3Lmz0L5/vFaStmADiQq9jIwM7Nq1Cz179kSZMmVgbGwMCwsLVKpUCQMHDsTevXshCEKBxnTlyhV07NgRZ8+exZs3bwp034XR+19SJBIJfv75Z6VyUVFRMDIyUig3ffr0z9r3okWLMH36dEyfPh2PHz/+rG0VVSdOnFCoUx0dHUilUtjZ2aFq1ar48ssvcezYMU2HqfClL/tkbGyMsmXLonfv3vj77781HeYn8/X1FY+pRo0auZb7+uuvxXKVK1cuwAg/zc6dO8X/sRMnThTIPo8fP46BAwfCzc0NpqamMDU1hYeHB3r06IHQ0FCkpqYWSBzqtGfPHgwcOBBXrlxBfHy8psMBAPF9nj59Oj/bSLsJRIVYeHi40LhxYwHAB6eYmJgCjWvq1KnivmvVqiUcPXpUOHXqlPDixQu17/v69evCqVOnhFOnTgmvXr1S+/5U8ejRI6X3pEyZMkJ6erpCuYCAAKVy06ZN+6x9Ozs7i9s6fvz4J22jMNZpXhw/fvyj/yMAhFatWgmvX7/WWJxr1679aIy6urrCnj17NBbj5zhz5ozCsdy6dUupTEpKimBtbS2W+eWXX9QSy6tXr8Rz+vr165+1rYEDB+bb/+vHxMfHC926dfvoeXLlyhVBEJSvPYXNh96HL774Qoy7Xbt2wvHjx4VTp04JUVFR+fr+5UX2unz06JHS8qJ+rSRSlV7+NLOI8l9ycjJat26Na9euAQB0dHQwaNAg+Pj4wMLCAs+ePcOBAwfwxx9/FHhsz549E/9u27YtWrRoUWD7Lgq/OAOZWaU9e/agU6dOAID09HQsX75cw1EpSkhIgKmpaZGpU1Vt3boVtra2ePnyJfbs2YPQ0FAIgoDDhw+jbdu2OHXqFIyMjD66naz6UZdTp04BAJ4+fYrRo0fj1atXSE9Px4IFC+Dj46O2/apLvXr1UL58edy7dw8AsGHDBgQEBCiU2b9/P6KjowEAenp66N+/f77GkJycDKlUihIlSqBEiRL5um11EwQB3bt3x6FDh8R53bp1Q8+ePWFnZ4fw8HAcO3YMmzdv1mCUefOh9yH750j37t2VbqUrjO9fcbtWEuVK0y00oty8n23YvHlzjuXu3bsnpKSkiK8zMjKEDRs2CM2bNxesra0FfX19wc7OTmjXrl2Ov0xnz0AcOXJEmDdvnuDm5iYYGBgIZcqUEX7++Wex7Md+qX/06JHCr+RNmjRR2Fduv8S+efNGGD9+vFChQgXByMhIMDAwEGQymVCvXj3h22+/FcLCwsSyTZo0Ebexdu1ahe2/fv1amDBhglCpUiXB2NhYkEqlgqurqzBixAilXwOzH4uzs7Pw5MkToV+/foK1tbVgaGgoNGjQQDh//vxH3qVM7/+Ka25uLgAQWrRoIZbZtm2b0vL36+Hhw4fC4MGDherVqwv29vaCgYGBYGRkJLi7uwu+vr4KxzBt2rQPvhdZ232/zjdu3ChUrVpVkEqlQqdOnXKt0507d4rzbG1tFX4tHTx4sMIvvxkZGSrVk7q8f16+/14HBwcrLJ87d6647P3z9cKFC0LLli0FMzMzwdLSUhg0aJC4fPLkyQrbfffunWBraysuv3jx4gfjfD+DlN2YMWPE+eXLl1da99mzZ8J3330nlC9fXjA0NBRMTEyE6tWrCwsWLBBSU1MVyl64cEHo37+/ULlyZcHOzk7Q19cXTExMhEqVKgljxowRIiIico2rSZMmwo0bN4SOHTsK5ubmgrGxsdCmTRvh/v37Hzy2LHPmzBG35eTkpJRF7dq1q7jcx8dHEIS8nfeCoPy/e//+faFLly6CpaWlmFH/0HVIEAQhJCRE8Pb2FmxsbAR9fX1BJpMJvXv3Fq5du5bjfnKamjRpIkyfPl183a9fP6X9VK1aVVy+bdu2D9bd5s2bcz1Ps3vx4oUQGRkpCMKHM0jTpk0TWrRoIZQuXVowNTUV9PX1BXt7e6Ft27bCrl27lLb7zz//CF26dBEcHBwEfX19wdTUVChTpozQoUMH4ddff1UoGxwcLDRp0kSwtrYWdHV1BSsrK6FChQpC3759hf3794vlcnofPpZJzW297Pbs2SN06NBBkMlkgr6+vmBtbS3UqVNHWLp0qVgmL/8H2a+TOU1Z18XC+PlDpA5sIFGhVb58efECmv2L9oekp6cL3bt3/+CF/vvvv1dYJ3sDyc3NLcd1Nm3aJAiC+hpITZs2/eB2z549K5bN7QPq7t27gkwmy3Ub5ubmwqlTp8Ty2Y/F3NxcKFGihNI6NjY2Qmxs7Efr/f0vKd9++6349+3btxXirlKlisIxZK+HAwcOfLAerK2txQ/aT2kgvf/+fqiBJAiC8M033yiV3b17tzjP0dFR6cu2JnysgSQIgtCyZUtxuYeHhzg/+/lasmRJwdjYWHxtYWEhXLx4UWH5u3fvxHWPHj0qLqtcufJH48ytgfTs2TOhWrVq4vyePXsqrHf27Fnxy39OU7NmzYS3b9+K5ZcvX/7Bc8PFxUXhttzscTk4OAgmJiZK63h4eCgce26eP38u6OjoiOv9+eef4rLo6GhBKpWKy7IaDHk57wVB8f22sLAQ7OzsFMp/qIGUnp4u9O3bN9d9SaVSYffu3Ur7yWlq0qSJEBYWJujr6wsABENDQ4V6ffjwocK1JPsPWTlp3bq1WN7V1VWlHx4+1ECyt7f/YPyLFy8Wy969e1fhvXl/yt5oDwoK+uB2fX19xbL53UDKyMgQhg4d+sH3JEte/g8+t4Gkyc8fInXgIA1UKCUmJoq3qQCAt7e3SustW7YM27ZtA5B5+8q0adNw4MABfPfdd2KZhQsXYu/evTmu/99//2HatGnYu3cvGjduLM7/9ddfAQDVqlXDqVOn0LZtW3HZl19+iVOnTuHUqVNwcHBQ/SDlIiMjxY7PpUqVQkhICI4dO4bg4GD88MMPqFmzJnR0Pv6v2r9/f4SHhwMAypUrh02bNuGPP/5AzZo1AQBxcXHo06cP3r59q7RuXFwcTExMEBwcjLVr18Lc3BxA5qAKwcHBeT6mfv36wcrKCgCwZMkS3LhxAydPngQAfPPNN7mu5+zsjICAAGzduhWHDh3CiRMnsHv3bvTr1w8AEB0djV9++QUAMHjwYJw6dQoymUxc/9dffxXfi8GDBytt/8GDB2jQoAFCQ0Oxb98+9O7d+4PH8csvv6BKlSoAgF27dmHBggUYNmwYgMxbPjdu3AhbW1tVq0WjWrZsKf599+5dJCUlKZV58eIFrKyssGrVKhw+fBgzZ85EjRo1UKdOHXF59tuftm7dKv795Zdf5jmmrIEKnJyccOXKFQCZt/BkH+AjJSUFvXr1EjuMd+vWDfv27cPWrVvh5eUFILND/+zZs8V1vLy88Msvv+CPP/7AkSNHcOLECezYsQOtW7cGADx69AirV6/OMaawsDCUL18e27dvx6JFi6Cvry/W2ZEjRz56TCVLllSo6w0bNoh/b9myBSkpKQAAa2trdOjQAUDezvv3xcbGIi0tDYsWLcLhw4exePFiSKXSXOP7/fffxf9pW1tbLF26FEeOHMGUKVMgkUiQkpKCAQMGICYm5qPXu99++w0ymQxdu3YFALx9+1bhepH9/OjXrx8MDAw+WHeXLl0S/27ZsuVnj0rn5+eHdevWYe/evThx4gSOHDmC3377TayfadOm4d27dwAyB0zIem+6d++OgwcP4sCBA1i9ejUGDx6MkiVL5nhc06ZNw7Fjx7Bz5078+uuv6NSpE8zMzD4YV7t27XDq1ClUrVpVnDdp0iSxXj9k9erVCudu165dsXXrVuzZswczZsyAo6OjuCwv/weTJ09W2vfWrVvFmNq1a/fBuArT5w9RvtB0C40oJ8+fP1f4JWnVqlUqrVe9enVxna+//lphWbt27cRl3bp1E+dnzyCNHDlSnH/27FmFX3Cz+1Cn5bxmkJKTkwVdXV0ByMyuXL58WUhKSsr1GHP6Be/69esK9ZX91oTnz5+Lv/ACEG8zfP/X4QsXLojr+Pr6ivNHjx6dayxZ3v8V98aNG+ItU6ampkLPnj3FekxKSso1gyQIgrBhwwahZcuWQokSJQQ9PT2lXxWrV6+uUP5jgzRkr/OSJUsKycnJKtVpljt37ihkVbKmH3744aP1IgiZt09mdWr+lCn7+5IbVTJIq1atUiiTNaBI9vNVIpEo3GKVZcOGDWKZrl27CoKQeXtd1q+++vr6Kg3+oMogDbVr1xYuX74srrNnzx5xmZ2dnfDXX3+JdfPbb7+JyxwcHMR10tLShN9++01o2LCheAvU+/vJOo7349LX11cYbCV7VuP926xyk/2WRjMzM/H/uUGDBuL8b775RqmOVT3v33+/szI+udV19utQjRo1xPnjxo1TONeyZ/FWrFghrvOxQRpOnTqVY5zZr8dZgyp8SPbjfv92ztx8KIN08+ZNYdCgQUK5cuUEQ0PDHM+3rAEQVq5cqVAvjx8/Vro9Mkv2DFxoaOgHz/0PfR586LqT23o1a9YU52fdopmbvP4fCMLHB2kojJ8/ROrAQRqoULK0tFR4HRUVpdJ6d+7cEf9u2LChwrLGjRtj//79SuWya968ufh39sxAVqdqdTA0NMTgwYOxatUqXLt2DdWrVxd/Va9Vqxb69OmDbt26fXAb2Y/HwMAAtWrVEl+XLFkS5cqVw927d8Wy73eANzMzE3/pA/Ln2EeOHImFCxciISEBW7ZsAQAMGTLkg4MDTJ06FT/++OMHtxsTE/NJ8QCZv9waGhrmaR0PDw8sWrQIw4cPF+fVrVsX06ZNU2n9K1euoFmzZnnaZ3bOzs75Mnx5RESEwuv3/8cAwNXVVczKZNejRw+MHj0aERER2LNnDyIiInDz5k28fv0aANC+fXvY2dnlOaasX6xjY2OxevVq7Ny5E//88w9atGiBf//9F1ZWVrh9+7bCMWTP7GYXFhaG6OhoWFtbY/DgwQqZm5zkdh55eHgo/Ar/Kf8LXbp0gaWlJd68eYP4+Hjs2rULtWvXVhjCfNCgQeLfn3PeS6XSPA1okb0+f/755xyH4weAW7duqbzNhg0bwsvLC9evX8fly5dx7do1mJub4/LlywCAqlWrKmRLcmNpaYnIyEgAql/zc3Pjxg3UrVs3x0xpdln12rlzZ0yfPh0vX74U68XQ0BDly5dHkyZNMGLECJQvXx5A5nDuoaGhSE9PR69evcTYPT090bp1a4wcOVLMoOe37O/fxz4XPuf/IC8K4+cP0efiLXZUKJmYmIgfRgBw9OjRz96moMKzkqytrcW/9fQ+7feD7LeFZN2+keX9L6lZVqxYgZCQEPTr1w9VqlSBsbExnj59iu3bt6N79+747bffPmn/WT527NmPG1A8dlXqLScuLi5o3769+FpHRwcjRozItXxaWhoWLFggvu7Tpw/27duHU6dOKczPyMj4pHgAfNItkABw8+ZNhddPnjzJly8WBSn7rXEeHh4wNjZWKpNb/UilUgwdOhRA5vu0YcOGz769Dsj8Yt2wYUO0b98eW7ZsgYmJCYDML207duzI8/bi4+Px4sULhS+F3377LQ4dOoRTp05h/Pjx4vzczqP8+F8wNDREz549xdcbNmzAxo0bxdeenp7ic5I+97y3t7dXywNS8/pcnpEjR4p/BwYGftL5kf3ZUceOHfus59v99ttvYuPI1dUVGzZswMmTJ3Hq1CmFL+BZ9WpnZ4fLly9j9uzZaNOmDVxcXJCamopr167h119/Rd26dcVR5xo3bowrV65g9OjRaNSoEWQyGd68eYPTp0/jhx9+QOvWrZGenv7JsavqQ+/75/4f5GdMmvj8IfpcbCBRoZX9F9ajR48qfOBmd//+ffGhgRUqVBDnv//AydOnT4t/e3h45GOkirL/cvjixQvx76wP0JxIJBL06tULGzduxNWrVxEfH4+QkBBx+aZNmz64z+zHnZKSgosXL4qvX758if/++098rc5jf9+oUaPEvzt06IAyZcrkWjYqKgqJiYni699//x3t2rVDw4YNERsbm+t62ftnfezD/lO+SO7du1fsg5b1wR0WFqbyl76mTZtCyBwQ55Om/MgeBQUFiX3AAGDgwIE5lvtQ/Xz99dfQ1dUFkNkPImt4/RIlSny0f8KnyPoxIfu5Xbp0aaSlpeVYT3FxcXB2dlYYOtna2hqLFy9Gq1at0LBhQ7x69Srf48xN9uvX4cOHsWrVqhyXfep5nyWv53T2+vz9999zrMvk5GSsXLlSLKfK/1j//v3FrOSmTZvEviMGBgbo27evSrFlr5cHDx5g/vz5OZZ7+fLlRzNMT58+Ff/+9ttv0b9/fzRu3BglS5bMcV1BEGBvb49JkybhwIED+O+//xAXF4cuXboAyLx+79u3TyxbuXJlzJ8/H3/99RfCwsLw8uVL8fp24cIFPHjwQKVjzquKFSuKf+f0iIusBsWn/h9kP59UbTwV1s8fos/BW+yo0Pruu+8QEhIiPgepT58+OHz4MHx8fGBubi4+B2n79u14/fo1DAwM8OWXX4q3daxatQoymQy1atXC4cOHxQ834NN/8VaFu7u7+Pfjx48xaNAg1KxZE4GBgYiLi8txnXLlyqFNmzaoWbMmHB0doaOjg8OHD4vLc+rYml3lypVRs2ZN8YOpX79+mDFjBgwNDTF37lykpaUByLzdQdUBL/KDt7c3AgIC8PbtW/GLRm7s7e1hYmIifln09/dHp06dcO7cOaVnyWRnY2ODR48eAQDWrVsHHR0d6OnpwcvLS+zs+6levnwpnis6OjrYv38/xowZgxs3bmDv3r1YvHixwgAghcXFixfx5MkTvHjxArt27VL4caF69eqfFLOTkxM6dOiAnTt3KtxS079//0/Otmb9YBAXF4fAwECFhkLWly5vb284OTnh2bNnePr0KVq3bo1hw4ahRIkSCA8Px4MHD7Bz505UrVoVa9euRdmyZcVtREdHY+bMmahbty4OHTqE9evXf1KcnyL7M5HevXuH58+fA1B+9tGnnvefasiQIeI1csyYMYiIiECtWrWQmpqKZ8+e4e+//8bu3btx/fp18Qu/jY2NuP7+/fvRsGFDGBsbw9nZGU5OTgAAY2NjDBo0CIsWLUJ0dLR4a5SPj4/KA5n06tULQUFBYrZz3Lhx+Oeff9CjRw/xOUhHjx5FSEgITp8+rRDX+7KfB6tXr0aZMmWQmJiI6dOn55iV2LJlC+bPn4+OHTvC1dUV9vb2iIqKUjjXs67Dfn5+ePjwIVq1agUnJydYWVnhwYMHCncIfOya/amGDRsmXud3796NHj16oE+fPpBKpbh27Rpu3ryJ4ODgT/4/sLGxEW9zXLFiBXx8fKCjo4PatWvnOshGYf38Ifos6u3iRPR5wsLChMaNG3+0Y3fWUKXv3r376FPY/fz8FPaRWyf/D3X+/VinZW9vb6X9GhgYCO7u7jmu96HhZQEIixYtEsvm1rH3zp07Hxxm1czMLNdhVp2dnRXizz6E9sCBAz/6PuU0SMOH5DZIw6RJk3KMvXnz5rnG6u/vn+M6Wcf6sffq/Xiy6jQ9PV1hvxMnThQEIbNDctb7JZVKVep8rm4fG445a2rZsqVSh/KPPW8luyNHjiht82PvdW77+tDUtGlThSG1z5w588Fhvt8/T3MbxrpZs2Y5HuunDM2viuzPRMqacupYn9fz/kP/ux87pvT0dKFPnz4ffQ+yd9A/dOhQjmV+/PFHhX3ev39fkEgkCmVyevbch8THx3/0Gg78f9CH3K7TN2/eFAwMDJTWq1SpksKQ0lnX/PefwfT+ZG5uLjx58kQQBMVBBHKaqlevLg7wkN+DNGRkZCg8m+z9KXvZvP4fCIKQ67nx7NmzD8asyc8fInXgLXZUqMlkMhw/fhw7duxA9+7dUbp0aRgaGsLU1BTly5dH3759sWvXLlhYWAAAdHV1sXXrVqxfvx7NmjWDlZUV9PT0YGtri7Zt22L37t1YuHCh2uNev349evbsCXNzcxgbG6Nly5Y4deoU6tWrl2P5gIAAdOrUCS4uLjAzM4Ouri5sbGzQvHlzBAcHq/SLv4eHB65fv47x48ejYsWKMDIyglQqRbly5fD111/j2rVrSgNXFDYzZszA3Llz4erqCqlUCg8PDyxZsgQ//PBDrutMmTIFvr6+KFGiRL72xQgICMCff/4JILOT+YwZMwBk/lo6Z84cAJm3k/Tp00ch81FY6Ovrw8bGBpUrV8aAAQNw6NAhHD58+JMGU8jSsmVLhVtkatasCU9Pz8+OVUdHB1ZWVqhfvz4WLFiAgwcPirfzAZnZmBs3bmD06NGoVKkSjI2NYWRkBBcXF3h7e2PhwoWYOXOmWH7VqlUYO3YsnJ2dYWhoiKpVqyI0NBRffPHFZ8eaF1988YXSEP3ZbyPL8inn/afS0dFBcHAwtmzZgjZt2sDOzk68Rnp5ecHX1xf79+8XM0MA0KpVKyxYsADlypVTeF/e5+bmhlatWomvZTIZ2rRpk6f4TE1NsW3bNhw7dgwDBgxAuXLlYGxsDGNjY5QrVw6dO3dGcHCwwq1mOalUqRL+/PNPNGjQAKamprC1tcWXX36J48eP5zhQTJ06dTB27FjUr18fDg4OMDAwgIGBAcqUKYOBAwfi/PnzKF26NACgb9++GDZsGLy8vGBraws9PT0YGxvD09MTEyZMwLFjx1R6NMOnkEgkWLt2LXbu3AkfHx/Y29tDT08PVlZWqF27tkLft0/5P1i8eDF69eoFa2vrPF1Pi8PnD1F2EkFgDzgiIlLNxIkT8dNPPwEAli5d+sGBN0j7rFixAl9//TWAzFvk5s2bp+GIiIjyjg0kIiL6oPT0dCQnJ+Px48fw8fHBkydPYGpqiufPn4vZW9JegiAgMTERr169Qu/evXHx4kXo6Ojg7t27cHNz03R4RER5xkEaiIjog06dOqX0LCd/f382jghA5rD3Li4uCvOGDBnCxhERFVlsIBERkUp0dHTg7OwMX19fhWepEAGZ/WMcHR3Rt2/fjz74loioMOMtdkRERERERHIcxY6IiIiIiEiODSQiIiIiIiI5NpCIiIiIiIjkOEgDEREREZEWMKr2TYHuL/nKkgLdX35hA0lLrL/4TNMhFEtf1HTC23eajqJ4MtQDElI4hkx+M5VKeM6qiaEeWLdqYqgHJKdpOorix0gfSErjdVYdjPUlmg6BPgMbSERERERE2kDC3jWqYC0RERERERHJsYFEREREREQkx1vsiIiIiIi0gYR9o1TBDBIREREREZEcM0hERERERNqAgzSohLVEREREREQkxwwSEREREZE2YB8klTCDREREREREJMcMEhERERGRNmAfJJWwloiIiIiIiOSYQSIiIiIi0gbsg6QSZpCIiIiIiIjkmEEiIiIiItIG7IOkEtYSERERERGRHBtIREREREREcrzFjoiIiIhIG3CQBpUwg0RERERERCTHDBIRERERkTbgIA0qYS0RERERERHJMYNERERERKQN2AdJJcwgERERERERyTGDRERERESkDdgHSSWsJSIiIiIiIjlmkIiIiIiItAH7IKmEGSQiIiIiIiI5ZpCIiIiIiLQB+yCphLVEREREREQkxwwSEREREZE2YAZJJawlIiIiIiIiOTaQiIiIiIiI5HiLHRERERGRNtDhMN+qKJQZpEGDBkEikYiTjY0N2rRpg+vXryuVHT58OHR1dRESEqK0bPr06eI2dHR04OjoiH79+uHZs2cK5Zo2bapQzt7eHj169MCTJ0/EMidOnIBEIsGbN2+U9lO1alVMnz5dfF2mTBlIJBKcO3dOoZyfnx+aNm0KABgyZAgqV66M1NRUhTL79++Hvr4+Ll68KM4rX748DAwM8OLFi1zrrLD6e1cw1vwwAj8P6YCFX3fH1gVTEfVSsf7/2r4OK8Z+iXmDfTB/WGdsmjMOLx7eUSgT8+olti6choVfdcPPQzrij19nIiE2piAPpUgL3bwJbVs1R61qldG7R1dcvnTx4yuRgssXL8Dvm6/QukUj1PDywPE/jyosnzZlImp4eShMA/v10lC0RdulixcwasRXaNm0IapUKo8/jx39+EqkMl4P1Ctw1e+o6lke8+bO1nQoxcLrV68wecI4NG1QB/VqVkWvbp1x+9ZNTYdFxVyhbCABQJs2bRAWFoawsDAcO3YMenp68PHxUSiTlJSE0NBQjBs3DoGBgTlup1KlSggLC8Pz588RGhqKGzduoGfPnkrlhg0bhrCwMLx48QK7du3Cs2fP0L9//0+O39DQEBMmTMh1+aJFixAfH49p06aJ8968eYPhw4dj8uTJqFmzJgDg9OnTePv2LXr06IGgoKBPjkdTnt69jhotO2HQjN/Qd+JPyEhPR/DcCUh9myyWsZaVQutB32DY3JX4YtoiWNjJsHnuBCTGvQEApL5NRvDcCZBAgn6TfsbAaYuQ/u4dtvwyBUJGhoaOrOg4eGA/5s0NwLDhXyN0205Ur14DI3yHIezlS02HVqQkJyfDvbwHJvj/kGuZ+g0a4dCfp8Tp12W/F2CExUdychLKly+PiZOnajqUYofXA/W6eeM6tm8Lhbt7eU2HUizExcZi0IA+0NPXw5IVq7B9116MHjcBZmbmmg6t6JLoFOxURBXayKVSKWQyGWQyGapWrYoJEybg2bNniIiIEMts3boVFStWhL+/P/7++288fvxYaTt6enqQyWRwdHREo0aNMGzYMJw7dw5xcXEK5YyNjSGTyeDg4IC6deti5MiRuHz58ifH7+vri3PnzmH//v05LjczM0NQUBDmz5+P8+fPA8jMMDk4OGDKlCliucDAQPTt2xcDBgzAmjVrIAjCJ8ekCX0mzEWVJq1hV6oM7J3Lwcd3HOKiXiP80QOxjGeDFnDxrAGrEo6wK1UG3v2+QkpyEl4//Q8A8Pz+LcRGvEIH33EoUbosSpQuCx/fcQj77x4e376iqUMrMjasW4su3bqha/ceKFuuHMb7T4bMQYYtoZs1HVqR0qBRY4wY5YfmLVvlWkbfwAC2tnbiZGFhWXABFiMNGzXBN999j5beudc1fRpeD9QnKSkRkyaOw9Tps2BmbqHpcIqFtWtWQyZzwIxZAfCs7AXHkqVQp249OJUurenQqJgrtA2k7BISErBp0ya4urrCxsZGnB8YGIj+/fvDwsIC7dq1w9q1az+4nfDwcPzxxx/Q1dWFrq5uruWio6OxdetW1KlT55NjLlOmDL766iv4+/sjI5csR9OmTTFixAgMHDgQW7duxZYtW7B+/Xro6WV2DYuPj8fWrVvRv39/eHt7IzExESdOnPjkmAqDlKREAIChqVmOy9PfpeHK8X2QGpvA3rkcAODduzRAAujq64vl9PQNIJHo4Nk9ptk/JC01FXdu30K9+g0V5ter3wDXrrJxmd8uXfwHLZvUR5cOrfHj9B8QHRWl6ZCIRLweqNecWTPRqHET1K1XX9OhFBsnj/+JipU8MW70d2jeuD56d++CP7Zt0XRYRZtEUrBTEVVoG0h79+6FqakpTE1NYWZmht27dyM0NBQ6OpkhP3jwAOfOnUOvXpn3+Pfv3x9r165VaozcuHEDpqamMDY2hoODA06cOIGRI0fCxMREodyyZctgamoKExMT2NjY4N69e1izZs1nHcOUKVPw6NEjbNq0KdcyAQEBkEgk6N27N+bMmYMKFSqIy0JCQuDm5oZKlSpBV1cXvXv3zvVWwqJAEAQc3bQCTuU9UcLJRWHZg8vnMG+wD+YOaofzB7aj78SfYGyW+QtcSdcKMJAa4s+Q1UhLeYvUt8k4FrwSgpCBhDfRmjiUIiPmTQzS09MVflgAABsbW0RGRuSyFn2KBg0bY1bAz1ixOgjfj5mA27du4Kuhg5T6GRJpCq8H6nNw/z7cvXMb3/qN0XQoxcqL58+wNXQzSpd2xrLfV6N7z16YFzAbe3bt1HRoVMwV2gZSs2bNcPXqVVy9ehXnz59Hq1at0LZtW3HghMDAQLRu3Rq2trYAgHbt2iExMRFHjyp25i1fvjyuXr2KCxcuYPbs2ahatSpmz1buONmvXz9cvXoV165dw+nTp+Hq6opWrVohPj7+k4/Bzs4OY8eOxdSpU3P9kmRkZIQxY8bA2NgY3333ncKyrAxZlv79++OPP/7IcaCILCkpKYiLi1OYUlJSPvkY8tOhoN/w+ul/6DxystIy54pVMHTO7xg0bTHKedXCH7/NQqJ8EAYTc0t0/XYqHlw+i3lDOuCXYZ2QkpQIWRk3SHQK7SlcqEje+xVHEASlefR5WrVph0aNm8LVzR2NmzbHr8tW4smTxzj91wlNh0akgNeD/BUeFoZ5c2djdsDPkEqlmg6nWMnIEOBRoSJG+Y2GR4WK6N6zN7p064GtW3hL6CdjHySVFNrITUxM4OrqCldXV9SuXRuBgYFITEzEqlWrkJ6ejvXr12Pfvn3Q09ODnp4ejI2NER0drZRhMTAwgKurKypVqoRJkyahatWq+Prrr5X2Z2FhIe6vQYMGCAwMxIMHDxAaGgoAMDfP7BAYGxurtO6bN29gYZHz/cajR49GcnIyli1bluux6unpQVdXV+ED6vbt2zh//jzGjx8vHmPdunWRnJyMzZtzvzAEBATAwsJCYQoICMi1fEE5tO433L98Fv0n/wJzGzul5QaGRrCWlURJt4rwGT4WOjq6uHrigLi8rFdNjFy4Ad8v34bRK/5ApxETER8TCUs7WUEeRpFjZWkFXV1dREZGKsyPjo6CjY2thqLSDnZ2JeDg6IinT598vDBRAeD1QD1u376F6Ogo9O3VFTWqVESNKhVx6eI/2LxpA2pUqYj09HRNh1hk2drZoWw5V4V5LmXLITwsTEMRkbYoMs9ByhqCOzk5Gfv370d8fDyuXLmi0Jfo7t276NevH6KiopRuIcjyww8/wN3dHd9//z2qV6+e6/6ytpucnDnampubG3R0dHDhwgU4OzuL5bJGvitfPucRa0xNTfHDDz9g+vTp6NChg8rHGxgYiMaNG2Pp0qUK8zds2IDAwMAcG3kA4O/vj9GjRyvMk0qlCL3xWuV95ydBEHBo3RLcu3gaA6bMh2UJB1XXRPq7NKW5WbfdPb51BYlxb+Benfd6f4i+gQEqVKyEc2f+RouW3uL8c2fOoGnzFhqMrPh78yYGr8LDYGur/IMAkSbweqAederWxbYdexTmTZ3iDxeXsvhyyLAP9nmmD6tarRqePH6kMO/pk8dwcHDUUETFALPFKim0DaSUlBSEh4cDAGJiYrBkyRIkJCSgQ4cOWLRoEdq3b48qVaoorFOpUiX4+flh48aNSrerZSlbtiw6deqEqVOnYu/eveL8pKQkcX+vXr3CrFmzYGhoiFatMkdRMjMzg6+vL8aMGQM9PT1UqVIFL1++xOTJk1GhQgWxXE6GDx+OhQsXYvPmzSoN/JCWloYNGzZg5syZ8PT0VFg2dOhQzJs3D9euXVM6fiCzMVSYUvwHg37FrTN/osfomTAwNBb7DEmNTaBvIEXq22T8vSsY7tXrwdTSBkkJcbh0dDfioiNQoU4TcTvXTh6ErWNpGJtb4vmD2ziyYSnqtOkGG0cnTR1akTFg4JeYPHE8Knp6okqVati+NRRhYWHo0au3pkMrUpKSEvHs6VPx9csXz3Hv7h2YyzO1vy9bghberWBra4eXL19g6a8LYWlphWYtWmow6qIpKTERT7PV9Yvnz3H3zh1YWFjAwZFfjD4Hrwf5z8TEFK5u7grzjIyMYWFpqTSf8qb/gEEYNKAPAleugHebtrh14zq2b9uCH6bN1HRoVMwV2gbSwYMH4eCQmW0wMzODh4cHtm7digoVKmDfvn0IDg5WWkcikaBr164IDAzMtYEEAGPGjEGDBg1w/vx5scGyatUqrFq1CgBgZWUFLy8v7N+/XyEztHDhQjg4OGDSpEl4/PgxSpQogWbNmiEkJEQceS4n+vr6+PHHH9G3b1+Vjn337t2IiopCly5dlJa5ubmhcuXKCAwMxK+//qrS9jTp8tHMX9U2zlLsuOozfByqNGkNHR1dRL18hm2nDiM5Pg5GpuZwKOuOL35YCLtSZcTyUWHPcDw0EMkJ8bC0s0eDTv1Qu223gjyUIqtN23aIfRODlcuXISLiNVzd3LF0xUo4OpbUdGhFyu1bN+E7ZKD4esHPcwEAPh07w3/KdDx8eB/79uxCfHw8bO3sULNWbQT8vBAmJqaaCrnIunXrJoZ++YX4+pd5mbcJd+zUBT/OmaupsIoFXg+oKKlUuTLmL/oNvy1egJUrlqFkyVIYN8Ef7XxUvyOH3lOE+wUVJIlQ1B6sQ59k/cVnmg6hWPqiphPevtN0FMWToR6QkMLLU34zlUp4zqqJoR5Yt2piqAckK991TZ/JSB9ISuN1Vh2M9QvnrWxGrX4u0P0lHx6nctkXL15gwoQJOHDgQObD2d3dERgYiBo1agDI7LYxY8YMrFy5EjExMahTpw6WLl2KSpUqidtISUnB2LFjsXnzZiQnJ6NFixZYtmwZSpUqlae42YwkIiIiItIGhfQ5SDExMWjQoAH09fVx4MAB3L59G/Pnz4elpaVYZt68eViwYAGWLFmCCxcuQCaTwdvbW2HEaT8/P+zYsQMhISE4ffo0EhIS4OPjk+fBUgrtLXZERERERFT8/fTTT3BycsLatWvFeWXKlBH/FgQBixYtwuTJk9G1a1cAwLp162Bvb4/g4GD4+voiNjYWgYGB2LBhA1q2zOz/u3HjRjg5OeHo0aNo3bq1yvEwg0RERERERPlO1edz7t69GzVr1kSPHj1QokQJVKtWTRwbAAAePXqE8PBwhUHRpFIpmjRpgjNnzgAALl26hLS0NIUyjo6O8PT0FMuoig0kIiIiIiJtUMAPilX1+Zz//fcfli9fDjc3Nxw6dAhfffUVvv32W6xfvx4AxJGm7e3tFdazt7cXl4WHh8PAwABWVla5llEVb7EjIiIiIqJ8l9vzOd+XkZGBmjVrYs6cOQCAatWq4datW1i+fDm++OL/o5pK3uvXJAiC0rz3qVLmfcwgERERERFpgwIepEEqlcLc3FxhyqmB5ODggIoVKyrMq1ChgvhMPJlMBgBKmaDXr1+LWSWZTIbU1FTExMTkWkZVbCAREREREZHGNGjQAPfu3VOYd//+fTg7OwMAXFxcIJPJcOTIEXF5amoqTp48ifr16wMAatSoAX19fYUyYWFhuHnzplhGVbzFjoiIiIhIGxTSB8V+//33qF+/PubMmYOePXvin3/+wcqVK7Fy5UoAmbfW+fn5Yc6cOXBzc4ObmxvmzJkDY2Nj9O3bFwBgYWGBIUOGYMyYMbCxsYG1tTXGjh2LypUri6PaqYoNJCIiIiIi0phatWphx44d8Pf3x8yZM+Hi4oJFixahX79+Ypnx48cjOTkZI0aMEB8Ue/jwYZiZmYllFi5cCD09PfTs2VN8UGxQUBB0dXXzFI9EEAQ+QlkLrL/4TNMhFEtf1HTC23eajqJ4MtQDElJ4ecpvplIJz1k1MdQD61ZNDPWA5DRNR1H8GOkDSWm8zqqDsX7eBgUoKEbtfy3Q/SXv+7ZA95dfCmeejYiIiIiISAN4ix0RERERkTYopH2QChvWEhERERERkRwzSERERERE2oAZJJWwloiIiIiIiOTYQCIiIiIiIpLjLXZERERERNpAUjiHHy9smEEiIiIiIiKSYwaJiIiIiEgbcJAGlbCWiIiIiIiI5JhBIiIiIiLSBuyDpBJmkIiIiIiIiOSYQSIiIiIi0gbsg6QS1hIREREREZEcM0hERERERNqAfZBUwgwSERERERGRHDNIRERERERaQMIMkkqYQSIiIiIiIpJjBomIiIiISAswg6QaZpCIiIiIiIjk2EAiIiIiIiKS4y12RERERETagHfYqYQZJCIiIiIiIjlmkIiIiIiItAAHaVANG0ha4ouaTpoOodgy5H+R2phKeSFXB56z6sO6VR8jfU1HUDwZ6/M6S/Q+Xsq1xNt3mo6geDLUA+6FJ2k6jGKpvMyY560aGOrxeqAurFv1Yd2qB+tVfQrrjyXMIKmGfZCIiIiIiIjkCmn7loiIiIiI8hMzSKphBomIiIiIiEiOGSQiIiIiIi3ADJJqmEEiIiIiIiKSYwaJiIiIiEgbMIGkEmaQiIiIiIiI5JhBIiIiIiLSAuyDpBpmkIiIiIiIiOTYQCIiIiIiIpLjLXZERERERFqAt9iphhkkIiIiIiIiOWaQiIiIiIi0ADNIqmEGiYiIiIiISI4ZJCIiIiIiLcAMkmqYQSIiIiIiIpJjBomIiIiISBswgaQSZpCIiIiIiIjkmEEiIiIiItIC7IOkGmaQiIiIiIiI5JhBIiIiIiLSAswgqYYZJCIiIiIiIjk2kIiIiIiIiOR4ix0RERERkRbgLXaqYQaJiIiIiIhIjhkkIiIiIiJtwASSSphBIiIiIiIikmMGiYiIiIhIC7APkmqYQSIiIiIiIpJjBomIiIiISAswg6QaZpCIiIiIiIjkim0DadCgQZBIJOJkY2ODNm3a4Pr162KZ9PR0LFy4EF5eXjA0NISlpSXatm2Lv//+W2FbQUFBkEgkqFChgtJ+tmzZAolEgjJlyiiVz5pMTU1Ro0YN/PHHHznGGhwcDF1dXXz11VfivB9//BEODg6Ijo5WKHvt2jUYGBhg165dn1IthVro5k1o26o5alWrjN49uuLypYuaDqlQu3ntEn6c+B0GdfVGxybVcO7UcYXlgiAgeO0KDOrqje7edTHpu6F4+ujfHLclCAKmjxuZ43Yodzxn1Yd1qz6sW/Vh3aoH6zX/ZP9+WhBTUVVsG0gA0KZNG4SFhSEsLAzHjh2Dnp4efHx8AGR+IezduzdmzpyJb7/9Fnfu3MHJkyfh5OSEpk2bYufOnQrbMjExwevXr3H27FmF+WvWrEHp0qWV9m1ubi7u+8qVK2jdujV69uyJe/fuKZVds2YNxo8fj5CQECQlJQEA/P394eTkhJEjR4rl0tLSMGjQIPTt2xedOnX63OopVA4e2I95cwMwbPjXCN22E9Wr18AI32EIe/lS06EVWinJyXBxdcdwv4k5Lv9jcxB2bdmI4X4TMf/3jbCytsHUMV8hKSlRqezurZuK9IVME3jOqg/rVn1Yt+rDulUP1itpQrFuIEmlUshkMshkMlStWhUTJkzAs2fPEBERgS1btmDbtm1Yv349hg4dChcXF1SpUgUrV65Ex44dMXToUCQm/v+LpJ6eHvr27Ys1a9aI854/f44TJ06gb9++SvuWSCTivt3c3DBr1izo6OgoZLAA4PHjxzhz5gwmTpwIDw8PbNu2Tdzf+vXrsWvXLnHe7NmzER0djV9//VUd1aVRG9atRZdu3dC1ew+ULVcO4/0nQ+Ygw5bQzZoOrdCqUbch+g8difqNWygtEwQBu7cGo+eAIajfuAWcy7rCz/9HpKS8xV9HDyiUffTwHnZt2YhvJ0wvoMiLB56z6sO6VR/WrfqwbtWD9Zq/mEFSTbFuIGWXkJCATZs2wdXVFTY2NggODoa7uzs6dOigVHbMmDGIiorCkSNHFOYPGTIEoaGhYpYnKCgIbdq0gb29/Qf3nZ6ejnXr1gEAqlevrrBszZo1aN++PSwsLNC/f38EBgaKyzw8PDBnzhx8/fXXOHToEAICArB27VqYm5t/Uh0UVmmpqbhz+xbq1W+oML9e/Qa4dvWKhqIq2l6FvUBMdCSq1qwnztM3MEClKjVw5+Y1cV7K22T8MtMfw/0mwMrGVhOhFkk8Z9WHdas+rFv1Yd2qB+uVNKVYN5D27t0LU1NTmJqawszMDLt370ZoaCh0dHRw//79HPsUARDn379/X2F+1apVUa5cOWzbtg2CICAoKAiDBw/OcRuxsbHivg0MDPD1119j5cqVKFeunFgmIyMDQUFB6N+/PwCgd+/eOHv2LB4+fCiW+e677+Dp6Yl27drh66+/RvPmzT+rTgqjmDcxSE9Ph42NjcJ8GxtbREZGaCiqoi0mOhIAYGltrTDf0soGb6KjxNerl8yHh2cV1G3YrEDjK+p4zqoP61Z9WLfqw7pVD9arGkgKeCqiinUDqVmzZrh69SquXr2K8+fPo1WrVmjbti2ePHmi0vo5pQYHDx6MtWvX4uTJk0hISEC7du1yXNfMzEzc95UrVzBnzhz4+vpiz549YpnDhw8jMTERbdu2BQDY2tqiVatWCrfxSSQSTJ48GRkZGZgyZcpHY05JSUFcXJzClJKSotLxatr79S0IQpFOzxYGSvUnCIB83vm/T+D65X8w9JtxGoiseOA5qz6sW/Vh3aoP61Y9WK9U0Ir1c5BMTEzg6uoqvq5RowYsLCywatUquLu74/bt2zmud+fOHQCAm5ub0rJ+/fph/PjxmD59Or744gvo6eVchTo6Ogr79vLywuHDh/HTTz+Jt/WtWbMG0dHRMDY2FstlZGTgypUr+PHHH6GrqwsA4j5y21d2AQEBmDFjhsK8adOmYeKU6R9dV1OsLK2gq6uLyMhIhfnR0VGw4W1fn8TKOrPeYqKiYG1jJ85/8yYallaZWaXrly8g/OVz9PFprLDu3KljUdGrGuYsXl1wARcxPGfVh3WrPqxb9WHdqgfrlTSlWGeQ3ieRSKCjo4Pk5GT07t0bDx48UMjoZJk/fz5sbGzg7e2ttMza2hodO3bEyZMnc729Lje6urpITk4GAERFRWHXrl0ICQkRM01ZU0JCAg4cOPCRreXM398fsbGxCpO/v/8nbaug6BsYoELFSjh3RnF49XNnzqBK1Woaiqpos3coCStrW1y9eE6cl5aWhlvXLqGCZxUAQPe+X+LXNVuweHWIOAHAkJFj8O3EGTlulzLxnFUf1q36sG7Vh3WrHqzX/MdBGlRTrDNIKSkpCA8PBwDExMRgyZIlSEhIQIcOHdCkSRNs3boVAwcOxM8//4wWLVogLi4OS5cuxe7du7F161aYmJjkuN2goCAsW7ZM6Z7Y7ARBEPednJyMI0eO4NChQ5g6dSoAYMOGDbCxsUGPHj2go6PYTvXx8UFgYKA4JHleSKVSSKVSpflv3+V5UwVqwMAvMXnieFT09ESVKtWwfWsowsLC0KNXb02HVmglJyUh7MUz8fWrsBf478E9mJmbw87eAR179MW2TYFwLFUajqVKY+vGQEilhmjcMvOWTisb2xwHZrCzd4DMoWSBHUdRxXNWfVi36sO6VR/WrXqwXkkTinUD6eDBg3BwcACQ2SfIw8MDW7duRdOmTQFkPuR18eLFWLhwIUaOHAmpVIp69erh+PHjaNiwYa7bNTIygpGR0Qf3HRcXJ+5bKpXC2dkZM2fOxIQJEwBk3l7XpUsXpcYRAHTr1g29evXCq1evPjpCXnHRpm07xL6JwcrlyxAR8Rqubu5YumIlHB35RT03D+/dxmS/YeLrwKXzAQDN23SAn/9MdO0zCCkpKVixMAAJCXFwr+CJGb8sh7Fxzg1/yhues+rDulUf1q36sG7Vg/Wav4pyVqcgSQRBEDQdBKlfYc8gFVWGesC98CRNh1EslZcZ87xVA0M9Xg/UhXWrPqxb9WC9qo9hIU1BlBqxs0D393xZ5wLdX34ppG8fERERERHlJ2aQVKNVgzQQERERERF9CDNIRERERETagAkklTCDREREREREJMcGEhERERGRFiisz0GaPn260roymUxcLggCpk+fDkdHRxgZGaFp06a4deuWwjZSUlIwatQo2NrawsTEBB07dsTz588/qZ7YQCIiIiIiIo2qVKkSwsLCxOnGjRvisnnz5mHBggVYsmQJLly4AJlMBm9vb8THx4tl/Pz8sGPHDoSEhOD06dNISEiAj48P0tPT8xwL+yAREREREWmBwjyKnZ6enkLWKIsgCFi0aBEmT56Mrl27AgDWrVsHe3t7BAcHw9fXF7GxsQgMDMSGDRvQsmVLAMDGjRvh5OSEo0ePonXr1nmKhRkkIiIiIiLSqAcPHsDR0REuLi7o3bs3/vvvPwDAo0ePEB4ejlatWollpVIpmjRpgjNnzgAALl26hLS0NIUyjo6O8PT0FMvkBTNIRERERERaoKAzSCkpKUhJSVGYJ5VKIZVKFebVqVMH69evh7u7O169eoVZs2ahfv36uHXrFsLDwwEA9vb2CuvY29vjyZMnAIDw8HAYGBjAyspKqUzW+nnBDBIREREREeW7gIAAWFhYKEwBAQFK5dq2bYtu3bqhcuXKaNmyJfbt2wcg81a6LO837gRB+GiDT5UyOWEDiYiIiIiI8p2/vz9iY2MVJn9//4+uZ2JigsqVK+PBgwdiv6T3M0GvX78Ws0oymQypqamIiYnJtUxesIFERERERKQFCnqYb6lUCnNzc4Xp/dvrcpKSkoI7d+7AwcEBLi4ukMlkOHLkiLg8NTUVJ0+eRP369QEANWrUgL6+vkKZsLAw3Lx5UyyTF+yDREREREREGjN27Fh06NABpUuXxuvXrzFr1izExcVh4MCBkEgk8PPzw5w5c+Dm5gY3NzfMmTMHxsbG6Nu3LwDAwsICQ4YMwZgxY2BjYwNra2uMHTtWvGUvr9hAIiIiIiLSBoV0lO/nz5+jT58+iIyMhJ2dHerWrYtz587B2dkZADB+/HgkJydjxIgRiImJQZ06dXD48GGYmZmJ21i4cCH09PTQs2dPJCcno0WLFggKCoKurm6e45EIgiDk29FRofX2naYjKJ4M9YB74UmaDqNYKi8z5nmrBoZ6vB6oC+tWfVi36sF6VR/DQpqCcPl+X4Hu79HC9gW6v/xSSN8+IiIiIiLKT4X5QbGFCQdpICIiIiIikmMGiYiIiIhICzCDpBpmkIiIiIiIiOSYQSIiIiIi0gJMIKmGGSQiIiIiIiI5ZpCIiIiIiLQA+yCphhkkIiIiIiIiOTaQiIiIiIiI5HiLHRERERGRFuAddqphBomIiIiIiEiOGSQiIiIiIi3AQRpUwwwSERERERGRHDNIRERERERagAkk1TCDREREREREJMcMEhERERGRFtDRYQpJFcwgERERERERyTGDRERERESkBdgHSTXMIBEREREREckxg0REREREpAX4HCTVsIGkJQz5TqtNeZmxpkMotnjeqgfrVX1Yt+rDulUP1iuRMv5baImkVEHTIRRLxgYSvH2n6SiKJ0M9YNL++5oOo9iZ086d56yaGOoByWmajqJ4MtIHz1s14DmrPkb6mo4gZ0wgqYZ9kIiIiIiIiOTYQCIiIiIiIpLjLXZERERERFqAgzSohhkkIiIiIiIiOWaQiIiIiIi0ADNIqmEGiYiIiIiISI4ZJCIiIiIiLcAEkmqYQSIiIiIiIpJjBomIiIiISAuwD5JqmEEiIiIiIiKSYwaJiIiIiEgLMIGkGmaQiIiIiIiI5JhBIiIiIiLSAuyDpBpmkIiIiIiIiOSYQSIiIiIi0gJMIKmGGSQiIiIiIiI5NpCIiIiIiIjkeIsdEREREZEW4CANqmEGiYiIiIiISI4ZJCIiIiIiLcAEkmqYQSIiIiIiIpJjBomIiIiISAuwD5JqmEEiIiIiIiKSYwaJiIiIiEgLMIGkGmaQiIiIiIiI5JhBIiIiIiLSAuyDpBpmkIiIiIiIiOSYQSIiIiIi0gJMIKmGGSQiIiIiIiI5NpCIiIiIiIjkeIsdEREREZEW4CANqmEDKR+Fh4dj9uzZ2LdvH168eIESJUqgatWq8PPzQ4sWLcRyc+bMwQ8//IDZs2dj4sSJCtsICgqCn58f3rx5k+M+Bg0ahDdv3mDnzp1qPJKCl5iYgGVLfsWfx44iJjoK5T0qYPzEyajkWVnToRULoZs3IWhtICIjIlDO1Q3jJ05C9Ro1NR1WofXv3/vx398HkBT9CgBgLiuNCq17Q1Yhs84EQcCdQ5vx6OwhpCYnwLq0O6p1+wrmDs4AgMToVzj449Act11n4ASUqtqwYA6kCOM5m/+WL/0Nvy9fojDPxsYWx07+raGIio/AVb/j2JHDePToP0gNDVG1ajX4jR6LMi5lNR1akbYlJBhbQzfj5csXAIByrm4Y/tUINGzURMORUXHHBlI+efz4MRo0aABLS0vMmzcPXl5eSEtLw6FDhzBy5EjcvXtXLLt27VqMHz8ea9asUWogaauZ037Aw4cPMGvOT7ArUQL79+7GV8O+xPad+1DC3l7T4RVpBw/sx7y5AZj8wzRUrVYd27aEYITvMOzYvQ8Ojo6aDq9QMrKwhafPQJjaOgAAnlw4hjOBs9FyzCKYOzjj/p/b8eDETtTs6wdTu5K4eyQUp1ZMRSv/5dA3NIaxpS3az1ivsM1HZw/i3p9/QFahhiYOqUjhOas+5Vzd8PvqteJrHR1dDUZTfFy88A969emHSpUrI/1dOn77dSG+GjYEf+zeB2NjY02HV2TZy2T49vuxKF26NABg966d8Bs1EiHbdsDV1U3D0RVNTCCphn2Q8smIESMgkUjwzz//oHv37nB3d0elSpUwevRonDt3Tix38uRJJCcnY+bMmUhMTMRff/2lwagLh7dv3+LY0cPwGz0WNWrWQunSzvhqxCg4liyFraGbNR1ekbdh3Vp06dYNXbv3QNly5TDefzJkDjJsYd3mytGzNhwq1oRZiZIwK1ESnu2/gJ7UEFFP7kEQBDw8uRse3j1R0qs+LBycUbPv90hPTcGzyycBABIdXRiaWylML26cg1O1RtCTGmn46Ao/nrPqo6urC1tbO3GytrbWdEjFwvKVgejUpStcXd1Q3sMDM2cFICzsJe7cvqXp0Iq0Jk2bo1HjJnAu4wLnMi4Y9d33MDY2xo1rVzUdGhVzbCDlg+joaBw8eBAjR46EiYmJ0nJLS0vx78DAQPTp0wf6+vro06cPAgMDCzDSwik9/R3S09NhYCBVmC+VSnHlyiUNRVU8pKWm4s7tW6hXX/GWrnr1G+Da1SsaiqpoETLS8ezyX0hPeQubMh5IjHqFt/ExsC9fTSyjq6cPW1dPRD26m+M2Yp49ROyL/1CmjndBhV1k8ZxVr6dPn8C7WUO0a90cE8Z+j+fPnmk6pGIpIT4eAGBuYaHhSIqP9PR0HNy/D8nJSfCqWu3jK1COJBJJgU5FFW+xywcPHz6EIAjw8PD4YLm4uDhs374dZ86cAQD0798fDRo0wG+//QZzc/OCCLVQMjExhVeVqlj1+zK4lC0LGxtbHNy/DzdvXEdpZ2dNh1ekxbyJQXp6OmxsbBTm29jYIjIyQkNRFQ2xLx/j+OJxyHiXCj0DI9QdPBnmstKIenQHACA1s1QoLzW1RFLM6xy39fj8YZjZO8HGpYK6wy7yeM6qT2UvL8ya8xOcncsgKioKq35fjoH9e2P7rr2wtLTSdHjFhiAI+GVeAKpVrwE3N3dNh1PkPbh/D1/0643U1BQYGRtjweKlKFfOVdNhUTHHDFI+EAQBwMdHBgkODkbZsmVRpUoVAEDVqlVRtmxZhISE5FssKSkpiIuLU5hSUlLybfvqMitgHgRBQOsWTVCnhhc2B29A23Y+vD8+n7x/bgqCUKR/2SkIZiVKouXYxWj23S8o26AtLgYvRFz402wl3q+/nOs0PTUFzy79xexRHvGczX8NGzVBS+/WcHMvj7r16mPJst8BAHt27dRsYMVMwKyZeHD/Pn76eYGmQykWyri4IHT7TqzfFIqePftg6uQJ+Pffh5oOq8hiBkk1bCDlAzc3N0gkEty5c+eD5dasWYNbt25BT09PnG7dupWvt9kFBATAwsJCYQoICMi37auLk1NpBAZtxJnzl3HgyHFs3LwV7969Q8mSpTQdWpFmZWkFXV1dREZGKsyPjo6CjY2thqIqGnT09GFq5wir0m7w9BkIC0cXPPxrN6Rmmb+0p8THKJRPSYiF1NRSaTvPr/2Nd2kpcK7VvCDCLvJ4zhYcI2NjuLq54+mTx5oOpdgImP0jTpz4E6vWroO9TKbpcIoFfX0DlC7tjEqelfHt92PgXt4DwRvXf3xFos/ABlI+sLa2RuvWrbF06VIkJiYqLX/z5g1u3LiBixcv4sSJE7h69ao4/fXXX7hw4QJu3ryZL7H4+/sjNjZWYfL398+XbRcEI2Nj2NmVQFxsLM6cOY2mzfil8nPoGxigQsVKOHdGcRjfc2fOoArv4c4jARnv0mBiYw9DMyu8undVXJLxLg2RD2/CxkX5NtvH54/AsVJtSE3ZF0EVPGcLTmpqKh49+he2dnaaDqXIEwQBc2bNxLGjh7FqzTqUKuWk6ZCKLUEQkJqaqukwiiyJpGCnoop9kPLJsmXLUL9+fdSuXRszZ86El5cX3r17hyNHjmD58uVo3bo1ateujcaNGyutW69ePQQGBmLhwoUAMjsiXr16VaGMgYEBKlas+NE4pFIppFKp0vykVOHTDqyAnPn7FAQBKFPGBc+ePsHCBT+jTBkXdOzcVdOhFXkDBn6JyRPHo6KnJ6pUqYbtW0MRFhaGHr16azq0QuvmvvWQedSAkZUt3r1NxrMrfyHi4U009J0OiUQC1yYdce/oVpjaOcLUzhH3jm6BroEUTtUVn82REPESkf/dQoNh0zR0JEUTz1n1WPDzT2jctBkcHBwQHR2NVb8vR2JCAjp06qLp0Iq8OT/OwIH9e7Hot2UwMTZBZERmfzlTMzMYGhpqOLqi69dFC9CwUWPYy2RISkzEwQP7cfHCP1i6YrWmQ6Nijg2kfOLi4oLLly9j9uzZGDNmDMLCwmBnZ4caNWpg8eLF6Nu3LyZMmJDjut26dUNAQAB++uknAEBCQgKqVVP8pdTZ2RmPHz9W92FoTEJ8An5bvACvXoXDwsISLVp6Y+S330NfX1/ToRV5bdq2Q+ybGKxcvgwREa/h6uaOpStWwtGxpKZDK7RS4t/gwqYFeBsXDX0jE5g7lEFD3+niyHXuzbshPS0VV7ctz3xQrLM7Gn41E/qGis87efzPURhZ2CiMeEcfx3NWPV69Cof/+NGIiXkDK2sreHlVxfrgLazXfJA1BP2QQQMU5s+cFYBOXfhD36eKjorEZP/xiIx4DVMzM7i7l8fSFatRr34DTYdWZBXlfkEFSSJkjTBAxVphzyAVVcYGErx9p+koiidDPWDS/vuaDqPYmdPOneesmhjqAclpmo6ieDLSB89bNeA5qz5GhfT33aaLzhTo/k741S/Q/eUXZpCIiIiIiLQAE0iq4SANREREREREcmwgERERERERyfEWOyIiIiIiLcBBGlTDDBIREREREZEcM0hERERERFqACSTVMINEREREREQkxwwSEREREZEW0GEKSSXMIBEREREREcmxgUREREREpAUkkoKdPlVAQAAkEgn8/PzEeYIgYPr06XB0dISRkRGaNm2KW7duKayXkpKCUaNGwdbWFiYmJujYsSOeP3+e5/2zgURERERERIXChQsXsHLlSnh5eSnMnzdvHhYsWIAlS5bgwoULkMlk8Pb2Rnx8vFjGz88PO3bsQEhICE6fPo2EhAT4+PggPT09TzGwgUREREREpAUkEkmBTnmVkJCAfv36YdWqVbCyshLnC4KARYsWYfLkyejatSs8PT2xbt06JCUlITg4GAAQGxuLwMBAzJ8/Hy1btkS1atWwceNG3LhxA0ePHs1THGwgERERERGRxo0cORLt27dHy5YtFeY/evQI4eHhaNWqlThPKpWiSZMmOHPmDADg0qVLSEtLUyjj6OgIT09PsYyqOIodEREREZEW0CngQexSUlKQkpKiME8qlUIqlSqVDQkJweXLl3HhwgWlZeHh4QAAe3t7hfn29vZ48uSJWMbAwEAh85RVJmt9VTGDRERERERE+S4gIAAWFhYKU0BAgFK5Z8+e4bvvvsPGjRthaGiY6/bev21PEISP3sqnSpn3MYNERERERKQFPqVf0Ofw9/fH6NGjFebllD26dOkSXr9+jRo1aojz0tPT8ddff2HJkiW4d+8egMwskYODg1jm9evXYlZJJpMhNTUVMTExClmk169fo379+nmKmxkkIiIiIiLKd1KpFObm5gpTTg2kFi1a4MaNG7h69ao41axZE/369cPVq1dRtmxZyGQyHDlyRFwnNTUVJ0+eFBs/NWrUgL6+vkKZsLAw3Lx5M88NJGaQiIiIiIhIY8zMzODp6akwz8TEBDY2NuJ8Pz8/zJkzB25ubnBzc8OcOXNgbGyMvn37AgAsLCwwZMgQjBkzBjY2NrC2tsbYsWNRuXJlpUEfPoYNJCIiIiIiLVDAd9jlq/HjxyM5ORkjRoxATEwM6tSpg8OHD8PMzEwss3DhQujp6aFnz55ITk5GixYtEBQUBF1d3TztSyIIgpDfB0CFT1Iq32Z1MDaQ4O07TUdRPBnqAZP239d0GMXOnHbuPGfVxFAPSE7TdBTFk5E+eN6qAc9Z9THS13QEOWv/+z8Fur99vrULdH/5hRkkIiIiIiItIEERTiEVIA7SQEREREREJMcMEhERERGRFijoB8UWVcwgERERERERyTGDRERERESkBQr6QbFFFTNIREREREREcswgERERERFpASaQVMMMEhERERERkRwzSEREREREWkCHKSSVMINEREREREQkxwYSERERERGRHG+xIyIiIiLSArzDTjXMIBEREREREckxg0REREREpAX4oFjVsIGkJYwN+A+hLob8L1KbOe3cNR1CscRzVn2M9DUdQfHF81Y9eM4SKePlRku8fafpCIonQz3WrbqwbtXDUA8wqvaNpsMolpKvLOE5qyaGekBiqqDpMIodEwMJIhN40qqDrWnh/IrNBJJq2AeJiIiIiIhIrnA2b4mIiIiIKF/xQbGqYQaJiIiIiIhIjhkkIiIiIiItwPyRaphBIiIiIiIikmMGiYiIiIhIC/A5SKphBomIiIiIiEiOGSQiIiIiIi2gwwSSSphBIiIiIiIikmMDiYiIiIiISI632BERERERaQEO0qAaZpCIiIiIiIjkmEEiIiIiItICTCCphhkkIiIiIiIiOWaQiIiIiIi0APsgqUalBlLz5s1V3qBEIsGxY8c+OSAiIiIiIiJNUamBlJGRoXKLUxCEzwqIiIiIiIjyHx8UqxqVGkgnTpxQcxhERERERESaxz5IRERERERagH2QVPNZDaSIiAgkJycrzS9duvTnbJaIiIiIiEgjPqmBNGvWLPz666+IiorKcXl6evpnBUVERERERPmL+SPV5Pk5SGvWrMHcuXPx7bffQhAETJo0Cf7+/ihVqhTc3NywevVqdcRJRERERESkdnluIC1dulRsFAFAly5dMGvWLNy9exdmZmaIjIzM9yCJiIiIiOjz6EgkBToVVXluID18+BB169aFjk7mqqmpqQAAIyMjjBkzBitXrszfCImIiIiIiApInhtIenqZ3ZYkEgnMzc3x/PlzcZmtrS1evHiRf9EREREREREVoDw3kNzc3PDs2TMAQK1atbBq1SqkpaUhPT0dK1euRJkyZfI7RiIiIiIi+kwSScFORVWeR7Fr164d/vrrLwwcOBD+/v5o3bo1LC0toaenh4SEBKxZs0YdcRIREREREaldnhtIU6dOFf9u3rw5zpw5g5CQEEgkErRv3x7NmjXL1wCJiIiIiOjz8UGxqvmsB8UCmbfZ1apVKz9iISIiIiIi0qjPbiAREREREVHhxwSSavLcQHJxcflgek4ikeDff//9rKCIiIiIiIg0Ic8NpCZNmig1kCIjI3HmzBmYm5ujSZMm+RYcERERERHlj6L88NaClOdhvoOCgrB27VqFac+ePbh//z6srKzQvn37PG1v0KBBkEgkkEgk0NfXh729Pby9vbFmzRpkZGSI5cqUKSOWyz7NnTsXAPD48WOF+RYWFqhbty727NmjFL9EIkGFChWUYtmyZQskEkmOQ5UnJyfDysoK1tbWSE5OVlp+5coV+Pj4oESJEjA0NESZMmXQq1cvREZGKpRbt24dateuDRMTE5iZmaFx48bYu3evQpkTJ05AIpHA09MT6enpCsssLS0RFBT0wTotqkI3b0LbVs1Rq1pl9O7RFZcvXdR0SMUG61Y9WK9552hngTWzvsDz4z8h6swCnAuZiGoVnMTlK2f0R/KVJQrTyXVjxOWlHayVlmdNXVtW08QhFTk8bz/fpYsX8N03X6FV80aoXtkDx48dFZelpaVh8YJf0LNLB9SvXQ2tmjfCD5MmIOL1Kw1GXDStX7MKDWpUwqJfAsR5giAg8Pel6Ni6KZrVr45vhg/Cf/8+1GCUVBzluYGUGxsbG4wbNw4zZszI87pt2rRBWFgYHj9+jAMHDqBZs2b47rvv4OPjg3fv3onlZs6cibCwMIVp1KhRCts6evQowsLCcP78edSuXRvdunXDzZs3FcqYmJjg9evXOHv2rML8NWvWoHTp0jnGuH37dnh6eqJixYr4448/FJa9fv0aLVu2hK2tLQ4dOoQ7d+5gzZo1cHBwQFJSklhu7Nix8PX1Rc+ePXHt2jX8888/aNSoETp16oQlS5Yo7fPff//F+vXrVavEIu7ggf2YNzcAw4Z/jdBtO1G9eg2M8B2GsJcvNR1akce6VQ/Wa95Zmhnhz6DRSHuXgc7fLEO1brMwccEfeBOv+KPTob9voUxLf3HqPGq5uOz5qxiFZWVa+mPm8r1ISErBob9vFfQhFTk8b/PH2+RkuLt7YMKkH5SXvX2Lu3duY6jvCASHbscvC3/DkyeP4TdqhAYiLbru3LqB3Tu2wtXNXWH+pnWBCNm0DqMnTEbg+lBY29jCb8RQJCYmaijSooXPQVJNvjWQAMDW1hb//fdfnteTSqWQyWQoWbIkqlevjkmTJmHXrl04cOCAQrbEzMwMMplMYTIxMVHYlo2NDWQyGTw8PDB79mykpaXh+PHjCmX09PTQt29fhWc2PX/+HCdOnEDfvn1zjDEwMBD9+/dH//79ERgYqLDszJkziIuLw+rVq1GtWjW4uLigefPmWLRokdjgOnfuHObPn4+ff/4ZY8eOhaurKypUqIDZs2fDz88Po0ePFh/Am2XUqFGYNm0a3r59m+c6LWo2rFuLLt26oWv3HihbrhzG+0+GzEGGLaGbNR1akce6VQ/Wa96N+dIbz8Nj4Dt9Iy7eeoKnYdE48c99PHqumGlPTX2HV1Hx4hQT9/8fmjIyBIVlr6Li0bFZFWw7fAmJyakFfUhFDs/b/NGgUWOM/NYPLVq2UlpmZmaG5avWoFWbtijjUhZeVapigv8U3Ll9C2FhbIiqIikpETOmTMCEKTNgZm4hzhcEAVuCN2Dg4OFo2twbZV3dMGXGHKS8fYsjB/dpMGIqbvKtgZSWloZVq1bBxcUlX7bXvHlzVKlSRSlbk9d4AEBfX19p+ZAhQxAaGipmeIKCgtCmTRvY29srlf33339x9uxZ9OzZEz179sSZM2cUGoIymQzv3r3Djh07IAhCjvFs3rwZpqam8PX1VVo2ZswYpKWlYfv27Qrz/fz88O7duxyzS8VJWmoq7ty+hXr1GyrMr1e/Aa5dvaKhqIoH1q16sF4/TfsmlXH59lNsmjcYT44F4OzmCfiyS32lco1quuHJsQBc3zkVS3/oAzsr01y3Wa2CE6p6OGHdzrO5lqFMPG81JyE+HhKJBGZm5poOpUiYP3cW6jVsjFp16inMf/niOaKiIlG7bgNxnoGBAarWqIkb13gOqyKn7irqnIqqPDeQmjdvrjQ1aNAAjo6O2L59OyZOnJhvwXl4eODx48fi6wkTJsDU1FRhOnHihMI69evXh6mpKQwNDTFmzBiUKVMGPXv2VNp21apVUa5cOWzbtg2CICAoKAiDBw/OMY41a9agbdu2Yh+kNm3aKGSf6tati0mTJqFv376wtbVF27Zt8fPPP+PVq//fb3z//n2UK1cOBgYGStt3dHSEhYUF7t+/rzDf2NgY06ZNQ0BAAGJjY1WpsiIp5k0M0tPTYWNjozDfxsYWkZERGoqqeGDdqgfr9dO4lLTFsB6N8PBpBDqOWIrV205j/vju6OtTWyxz+O/b+HLSOrQd/ismLvgDNSo548DKb2Ggn/OYQgM718Od/8Jw7tqjgjqMIovnrWakpKTg10Xz0aadD0xNc2/sU6ajh/bj/t07+Oqb75WWRUdlZput3juHra1txGVE+SHPDaSMjAwIgqAwmZubo3v37jh16hT69++fb8EJgqDQ+hw3bhyuXr2qMNWpU0dhndDQUFy5cgW7d++Gq6srVq9eDWtr6xy3P3jwYKxduxYnT55EQkIC2rVrp1QmPT0d69atUziu/v37Y926dQoDKMyePRvh4eFYsWIFKlasiBUrVsDDwwM3btz4pGPNMmTIENja2uKnn35SaTspKSmIi4tTmFJSUlRaV9PeP/7c6oTyjnWrHqzXvNHRkeDq3WeYtmQPrt17jsDtf2PtjjMY3qORWGbb4cs4ePoWbv8bhv1/3UTnb5bBzbkE2jaqpLQ9Q6k+erWtyexRHvG8LThpaWnwHzcagiDAf8o0TYdT6L0KD8OiX+Zi6qy5kEqluZaTgOcwqVeeh/l+P2OjTnfu3FG4Zc/W1haurq4fXMfJyQlubm5wc3ODqakpunXrhtu3b6NEiRJKZfv164fx48dj+vTp+OKLL6Cnp1wdhw4dwosXL9CrVy+F+enp6Th8+DDatm0rzrOxsUGPHj3Qo0cPBAQEoFq1avjll1+wbt06uLu74/Tp00hNTVXKIr18+RJxcXFwc3NT2r+enh5mzZqFQYMG4ZtvvvngsQNAQECA0kAZ06ZNw8Qp0z+6rqZYWVpBV1dXacS/6Ogo2NjYaiiq4oF1qx6s108THhmHO/+FK8y7+ygcnVtU/eA6T8Oi4VraTmlZl5ZVYWxogE17/8nvUIslnrcFKy0tDRPHfo8XL57j98AgZo9UcO/ObcRER2FI///f+ZOeno6rly/ijy2bEbw9c9Tf6KhI2Nr9/5oQExMNK2sbpe2RsnwdfKAYy3M9zZw5Ey9zGe0mLCwMM2fO/OygAODPP//EjRs30K1bt0/eRpMmTeDp6YnZs2fnuNza2hodO3bEyZMnc729LjAwEL1791bKXPXr109psIbsDAwMUK5cOXFUld69eyMhIQG///67UtlffvkF+vr6uR5rjx49UKlSJZVGCPT390dsbKzC5O/v/9H1NEnfwAAVKlbCuTN/K8w/d+YMqlTlsL2fg3WrHqzXT3P26n9wd1b8scqtdAk8DYvOdR1rCxOUsrdCWGSc0rJBnetj38kbiIxJyPdYiyOetwUnq3H09OkTrFi1FpaWVpoOqUioUbsuNoTuRFDwdnHyqFgJrdr6ICh4O0qWcoKNjS0unD8jrpOWloqrly6ichWew5R/8pxBmjFjBtq0aQNHR0elZS9fvsSMGTMwderUPG0zJSUF4eHhSE9Px6tXr3Dw4EEEBATAx8cHX3zxhVguPj4e4eGKvz4aGxvD3Dz3To9jxoxBjx49MH78eJQsWVJpeVBQEJYtW6Z0TzYAREREYM+ePdi9ezc8PT0Vlg0cOBDt27dHREQEzp8/j5CQEPTu3Rvu7u4QBAF79uzB/v37sXbtWgBAvXr18N1332HcuHFITU1F586dkZaWho0bN2Lx4sVYtGgRnJyclGLIMnfuXLRu3TrX5VmkUmmOaem373IoXIgMGPglJk8cj4qenqhSpRq2bw1FWFgYevTqrenQijzWrXqwXvPut41/4njQGIwb3Arbj1xGrUplMLhbA3zzY+YIaiZGBpjyVXvsPHYVYRGxcHa0wcxRHRD1JgG7/7ymsK2yTrZoWL2cwhDg9HE8b/NHUlIinj19Kr5+8eI57t29A3MLC9jZlcD40d/h7p3bWLx0BdIz0sU+XhYWFtDXV+6LTJlMTExQ1lXxbhojI2OYW1iI83v2HYD1a1ahlJMznEo7Y/2alZAaGsK7Td6ew6mteCuiavLcQMptlDYASEhIyHHEuI85ePAgHBwcoKenBysrK1SpUgW//vorBg4cCB2d/ye5pk6dqtT48vX1xYoVK3Ldto+PD8qUKYPZs2dj2bJlSsuNjIxgZGSU47rr16+HiYkJWrRoobSsWbNmMDMzw4YNG9C5c2cYGxtjzJgxePbsGaRSKdzc3LB69WoMGDBAXGfRokXw8vLC8uXL8cMPP0AikaB69erYuXMnOnTo8ME6yhoQ4/Dhwx8sV1S1adsOsW9isHL5MkREvIarmzuWrlgJR0flRi3lDetWPViveXfp9lP0GrMKM0d1xKThbfH4RRTG/bwdIQcyH1SaniGgkqsj+vrUhqWZEcIj43Dywn0MmLAGCUmKfSkHdqqHl69jcfTsXU0cSpHF8zZ/3L51E8MHDxRfL/g586H1HTp2hu+Ib3DyxJ8AgN7dOyust3LNOtSspdh3mvKm38AhSElJwfy5PyI+Pg4VPb2waOkqpce+EH0OifChFo/c9evXcfXqVQDAoEGDMHXqVJQtW1ahTHJyMjZt2oSoqCjcusWH9RU2hT2DVFQZ6rFu1YV1qx6GeoBRtY/3Z6S8S76yhOesmhjqAYmpH/26QnlkYiBBZAJPWnWwNc1zDqJA+O0q2B+VFnXyKND95ReV3r0dO3aI/V8kEkmu/YyMjIzEW8qIiIiIiIiKGpUaSMOHD4ePjw8EQUDt2rWxdu1apT45UqkU5cqVy/V2NSIiIiIi0hwddkFSiUoNJAcHBzg4OAAAjh8/jho1anC4SiIiIiIiKnbyPMx3xYoVcx3m+/79+0rPVyAiIiIiIs2TSCQFOhVVee5BNnLkSFhYWGDVqlVKy+bPn4+4uDhs3rw5X4IjIiIiIiIqSHnOIP3999+5Po+ndevWOH369GcHRURERERE+UtHUrBTUZXnBlJkZGSOD1UFACsrK0RERHx2UERERERERJqQ5waSvb09bty4keOyGzdu5Np4IiIiIiIizZFICnYqqvLcQGrTpg1mz56N+/fvK8x/8OABAgIC0K5du3wLjoiIiIiIqCDleZCG6dOnY+/evfDy8kKzZs1QqlQpPH/+HMePH4eNjY34QFkiIiIiIqKiJs8ZJEdHR1y8eBH9+vXD9evXsW7dOly/fh39+/fHxYsXoa+vr444iYiIiIjoM+hIJAU6FVV5ziABmY2kwMBA8XVGRgYOHjyIb775Bnv37kVKSkq+BUhERERERFRQPqmBlOXff//FmjVrsG7dOoSFhcHAwADdunXLr9iIiIiIiCif5PnWMS2V5wbS27dvsXXrVgQGBuLUqVMQBAESiQSjR4/GxIkTOYodEREREREVWSo3JC9cuICvvvoKMpkMgwYNwuXLlzFo0CDs3bsXgiCgQ4cObBwRERERERVShXWY7+XLl8PLywvm5uYwNzdHvXr1cODAAXG5IAiYPn06HB0dYWRkhKZNm+LWrVsK20hJScGoUaNga2sLExMTdOzYEc+fP/+kelKpgeTl5YW6deti1apV8PT0xKpVqxAWFobAwEA0aNDgk3ZMRERERERUqlQpzJ07FxcvXsTFixfRvHlzdOrUSWwEzZs3DwsWLMCSJUtw4cIFyGQyeHt7Iz4+XtyGn58fduzYgZCQEJw+fRoJCQnw8fFBenp6nuNR6Ra7mzdvQiKRoH379pg7dy4qVqyY5x0REREREZHmFNaR5Tp06KDwevbs2Vi+fDnOnTuHihUrYtGiRZg8eTK6du0KAFi3bh3s7e0RHBwMX19fxMbGIjAwEBs2bEDLli0BABs3boSTkxOOHj2K1q1b5ykelTJIixYtgpeXF/bu3YvKlSujXr16WL16tUKrjYiIiIiIKEtKSgri4uIUpo+Ndp2eno6QkBAkJiaiXr16ePToEcLDw9GqVSuxjFQqRZMmTXDmzBkAwKVLl5CWlqZQxtHREZ6enmKZvFCpgfTtt9/iypUr+OeffzB8+HDcvXsXw4cPh4ODA4YPHw6JRAJJIW2REhERERFRwfdBCggIgIWFhcIUEBCQY2w3btyAqakppFIpvvrqK+zYsQMVK1ZEeHg4AMDe3l6hvL29vbgsPDwcBgYGsLKyyrVMXuRptL+aNWti+fLlCAsLw7p161CzZk1s27YNgiBgyJAhmD9/PqKiovIcBBERERERFS/+/v6IjY1VmPz9/XMsW758eVy9ehXnzp3D119/jYEDB+L27dvi8veTMVkjaX+IKmVy8knDoRsaGmLAgAE4ceIE7t+/j4kTJyIpKQnjxo2Dk5PTp2ySiIiIiIjUSEdSsJNUKhVHpsuapFJpjrEZGBjA1dUVNWvWREBAAKpUqYLFixdDJpMBgFIm6PXr12JWSSaTITU1FTExMbmWyVM95XmN95QrVw5z5szB06dPsXv3brRp0+ZzN0lERERERFpMEASkpKTAxcUFMpkMR44cEZelpqbi5MmTqF+/PgCgRo0a0NfXVygTFhaGmzdvimXyIs8Pis2Njo4OfHx84OPjk1+bJCIiIiKifFJYR7GbNGkS2rZtCycnJ8THxyMkJAQnTpzAwYMHIZFI4Ofnhzlz5sDNzQ1ubm6YM2cOjI2N0bdvXwCAhYUFhgwZgjFjxsDGxgbW1tYYO3YsKleuLI5qlxf51kAiIiIiIiLKq1evXmHAgAEICwuDhYUFvLy8cPDgQXh7ewMAxo8fj+TkZIwYMQIxMTGoU6cODh8+DDMzM3EbCxcuhJ6eHnr27Ink5GS0aNECQUFB0NXVzXM8EkEQhHw7Oiq03r7TdATFk6Ee61ZdWLfqYagHGFX7RtNhFEvJV5bwnFUTQz0gMZVfV/KbiYEEkQk8adXB1rRw5iBmHnlYoPub6u1aoPvLL4Xz3SMiIiIionxVSO+wK3Q+e5AGIiIiIiKi4oIZJCIiIiIiLaDDDJJKmEEiIiIiIiKSYwaJiIiIiEgLSMAUkiqYQSIiIiIiIpJjBomIiIiISAuwD5JqmEEiIiIiIiKSYwaJiIiIiEgLMIOkGjaQtIQh32m1Yd2qD+tWPZKvLNF0CMUWz1n1MTHgNzt1sDXlSUv0Pv5XaIm37zQdQfFkqMe6VRfWrXoY6gFJaYKmwyiWjPUl6LXuiqbDKJZCB1bj9UANeD1QH2P9wtmgl0gKZ1yFDfsgERERERERybGBREREREREJMdb7IiIiIiItAAHaVANM0hERERERERyzCAREREREWkBjtGgGmaQiIiIiIiI5JhBIiIiIiLSAjpMIamEGSQiIiIiIiI5ZpCIiIiIiLQAR7FTDTNIREREREREcswgERERERFpAXZBUg0zSERERERERHLMIBERERERaQEdMIWkCmaQiIiIiIiI5JhBIiIiIiLSAuyDpBpmkIiIiIiIiOTYQCIiIiIiIpLjLXZERERERFqAD4pVDTNIREREREREcswgERERERFpAR2O0qASZpCIiIiIiIjkmEEiIiIiItICTCCphhkkIiIiIiIiOWaQiIiIiIi0APsgqYYZJCIiIiIiIjlmkIiIiIiItAATSKphBomIiIiIiEiOGSQiIiIiIi3AzIhqWE9ERERERERyzCAREREREWkBCTshqYQZJCIiIiIiIjmtbCANGjQInTt3Vpp/4sQJSCQSvHnzBgCQnp6OhQsXwsvLC4aGhrC0tETbtm3x999/K6wXFBQEiUQiTvb29ujQoQNu3bqltI9nz55hyJAhcHR0hIGBAZydnfHdd98hKioq1+29P3355Zf5Wh+FwaWLFzBqxFdo2bQhqlQqjz+PHdV0SMVK6OZNaNuqOWpVq4zePbri8qWLmg6pSAtc9Tv69uyGerWqoWmjevAbNQKPH/2n6bCKjdevXmHyhHFo2qAO6tWsil7dOuP2rZuaDqtQ8y5vi3kdPLC2jxfW9vHCj23dUbWkubg8dGC1HKcOlUqIZVq42WBqa1es7eOF0IHVYKyvq4lDKdJ4rc1/vB6QJmhlA0kVgiCgd+/emDlzJr799lvcuXMHJ0+ehJOTE5o2bYqdO3cqlDc3N0dYWBhevnyJffv2ITExEe3bt0dqaqpY5r///kPNmjVx//59bN68GQ8fPsSKFStw7Ngx1KtXD9HR0QCAXr16ISwsTGn64YcfYGBggGHDhhVkVRSI5OQklC9fHhMnT9V0KMXOwQP7MW9uAIYN/xqh23aievUaGOE7DGEvX2o6tCLr4oV/0KtPP2zYvAW/r1qLd+np+GrYECQlJWk6tCIvLjYWgwb0gZ6+HpasWIXtu/Zi9LgJMDMz//jKWiwqMRXBl19i0r57mLTvHm6Gx2NcMxeUsjQEAAwPvaEwLf/7CTIEAeefvBG3IdXTwbUX8dh545WGjqJo47U2//F6kP8kBTwVVeyDlIstW7Zg27Zt2L17Nzp06CDOX7lyJaKiojB06FB4e3vDxMQEQOY9nTKZDADg4OCA77//Hh07dsS9e/dQuXJlAMDIkSNhYGCAw4cPw8jICABQunRpVKtWDeXKlcPkyZOxfPlyGBkZicuznDx5EgEBAVi+fDnq169fEFVQoBo2aoKGjZpoOoxiacO6tejSrRu6du8BABjvPxlnzpzGltDN+O77MRqOrmhavjJQ4fXMWQFo1qge7ty+hRo1a2koquJh7ZrVkMkcMGNWgDjPsWQpDUZUNFx+HqfwOvRKGFqVt4WbrTGev3mL2LfvFJbXdLLArfAEvE74/494++9EAAAq2puqP+BiiNfa/MfrAWkKM0i5CA4Ohru7u0LjKMuYMWMQFRWFI0eO5LjumzdvEBwcDADQ19cHAERHR+PQoUMYMWKEUuNHJpOhX79+CA0NhSAIStt78uQJevToAV9fXwwdOvRzD420SFpqKu7cvoV69RsqzK9XvwGuXb2ioaiKn4T4eACAuYWFhiMp+k4e/xMVK3li3Ojv0LxxffTu3gV/bNui6bCKFIkEqF/GElI9HdyPUM5qWhjqoVopCxx/EJXD2vQpeK1VD14P8p+ORFKgU1GltRmkvXv3wtRU8Vey9PR08e/79++jQoUKOa6bNf/+/fvivNjYWJiamkIQBPE2m44dO8LDwwMA8ODBAwiC8MFtxsTEICIiAiVK/P+e8KSkJHTp0gWVKlXCokWL8n6gpNVi3sQgPT0dNjY2CvNtbGwRGRmhoaiKF0EQ8Mu8AFSrXgNubu6aDqfIe/H8GbaGbkb/LwZhyDBf3LxxHfMCZkNf3wAdOnXWdHiFmpOlIWa1c4e+rg7evkvHL8cf4UXsW6VyTcpZ421aOv7JdnsdfR5ea9WD1wPSFK1tIDVr1gzLly9XmHf+/Hn0799f5W1kHyrRzMwMly9fxrt373Dy5En8/PPPWLFihcrbysocvT/84pAhQxATE4PDhw9DT+/jb1dKSgpSUlIU5kmlUkBXqnIsVPy8f14JgsChPvNJwKyZeHD/PoI2BGs6lGIhI0NAxUqVMMpvNADAo0JF/PvwIbZu2cwvRB/xMi4F4/fchYmBLuo4W2Jkw9KYfvChUiOpqZsNTv8Xg7QM5TsW6PPwWpu/eD3IfzwbVaO1t9iZmJjA1dVVYSpZsqS43N3dHbdv385x3Tt37gAA3NzcxHk6OjpwdXWFh4cHfH19MWDAAPTq1Utc7urqColEkus27969CysrK9ja2orzfvrpJ+zevRs7d+5UmP8hAQEBsLCwUJgCAgI+viIVS1aWVtDV1UVkZKTC/OjoKNjYqHZOUe4CZv+IEyf+xKq162Av74NIn8fWzg5ly7kqzHMpWw7hYWEaiqjoSM8Q8Co+Ff9FJWPz5TA8iX6LdhXsFMp4lDBBSQtD/Mnb6/IVr7XqwesBaYrWNpA+pnfv3njw4AH27NmjtGz+/PmwsbGBt7d3rut///33uHbtGnbs2AEAYvlly5YhOTlZoWx4eDg2bdqEXr16ib80HTx4EJMnT0ZQUBCqVKmictz+/v6IjY1VmPz9/VVen4oXfQMDVKhYCefOKA5Nf+7MGVSpWk1DURV9giBgzqyZOHb0MFatWYdSpZw0HVKxUbVaNTx5/Ehh3tMnj+Hg4KihiIowCaCnq/h7cTM3G/wbmYQnMcm5rESfgtda9eD1IP9JJAU7FVVsIOWid+/e6NKlCwYOHIjAwEA8fvwY169fh6+vL3bv3o3Vq1eLI9jlxNzcHEOHDsW0adPE2+eWLFmClJQUtG7dGn/99ReePXuGgwcPwtvbGyVLlsTs2bMBZPZX6tOnD4YOHYpGjRohPDxcYcoaDjwnUqkU5ubmCpNUWvhvr0tKTMTdO3dwV56de/H8Oe7eucPhUfPBgIFf4o/t27Djj234799/8fPcOQgLC0OPXr01HVqRNefHGdi/dzfmzpsPE2MTREZEIDIiAm/fKvf3oLzpP2AQbly/hsCVK/D06RMc2LcH27dtQa8+/TQdWqHWu5oDPEqYwM7EAE6WhuhVzQGV7E1x+r8YsYyRvg7qOlvizweROW7DwlAPzlZGkJlnfmaUtjKEs5URTAz4PCRV8Fqb/3g9IE2RCDkNm1bMDRo0CG/evFF6ltGJEyfQrFkzxMTEwNLSEu/evcPixYuxdu1aPHz4EFKpFPXq1cOUKVPQsOH/R6oJCgqCn5+f+IDZLE+fPkW5cuWwadMm9OzZE0DmiHTTp0/HwYMHERUVBZlMhs6dO2PatGli584ZM2Zg+vTpucbfpEkTnDhxIk/H/N4Ir4XOhX/OY+iXXyjN79ipC36cM1cDEanGUK/w1y2Q+fDCoDWBiIh4DVc3d4yb4F/oh6MuzHVbpVL5HOfPnBWATl26FnA0eWOoBySlFe7L/l8njuO3xQvw9MkTlCxZCv0HDkLX7j01HdZHGetL0GudZkYs861fGp4OprAy0kdSajqexrzFrpuvcCMsXizTws0GA2uXgu+WG0hOy1DaRvcqMvSo6qA0f9npJzj5b+4/zBWE0IHVCu31ILuidq3l9UB9jPULZ/pk85UXBbq/PtVKfrxQIaSVDSRtVBQ+WIqiwvwlvqhj3apHUfhCVFRpsoFU3BWVBlJRw+uB+rCBlKmoNpC0dhQ7IiIiIiJtwr41qmE9ERERERERybGBREREREREJMdb7IiIiIiItAAfXKwaZpCIiIiIiIjkmEEiIiIiItICzB+phhkkIiIiIiIiOWaQiIiIiIi0APsgqYYZJCIiIiIiIjlmkIiIiIiItAAzI6phPREREREREckxg0REREREpAXYB0k1zCARERERERHJMYNERERERKQFmD9SDTNIREREREREcswgERERERFpAXZBUg0zSERERERERHJsIBEREREREcnxFjsiIiIiIi2gw2EaVMIMEhERERERkRwbSEREREREWkAiKdhJVQEBAahVqxbMzMxQokQJdO7cGffu3VMoIwgCpk+fDkdHRxgZGaFp06a4deuWQpmUlBSMGjUKtra2MDExQceOHfH8+fM81xMbSEREREREpDEnT57EyJEjce7cORw5cgTv3r1Dq1atkJiYKJaZN28eFixYgCVLluDChQuQyWTw9vZGfHy8WMbPzw87duxASEgITp8+jYSEBPj4+CA9PT1P8UgEQRDy7eio0Hr7TtMRFE+GeqxbdWHdqoehHpCUxsu+OhjrS9Br3RVNh1EshQ6sxuuBGvB6oD7G+oWzr8++m68LdH/tPUt80noREREoUaIETp48icaNG0MQBDg6OsLPzw8TJkwAkJktsre3x08//QRfX1/ExsbCzs4OGzZsQK9evQAAL1++hJOTE/bv34/WrVurvH9mkIiIiIiIKN+lpKQgLi5OYUpJSfnoerGxsQAAa2trAMCjR48QHh6OVq1aiWWkUimaNGmCM2fOAAAuXbqEtLQ0hTKOjo7w9PQUy6iKDSQiIiIiIi1Q0H2QAgICYGFhoTAFBAR8MEZBEDB69Gg0bNgQnp6eAIDw8HAAgL29vUJZe3t7cVl4eDgMDAxgZWWVaxlVcZhvIiIiIiLKd/7+/hg9erTCPKlU+sF1vvnmG1y/fh2nT59WWiZ5b+QHQRCU5r1PlTLvYwOJiIiIiEgLFPRzkKRS6UcbRNmNGjUKu3fvxl9//YVSpUqJ82UyGYDMLJGDg4M4//Xr12JWSSaTITU1FTExMQpZpNevX6N+/fp5ipu32BERERERkcYIgoBvvvkGf/zxB/7880+4uLgoLHdxcYFMJsORI0fEeampqTh58qTY+KlRowb09fUVyoSFheHmzZt5biAxg0REREREpAXyeKdZgRk5ciSCg4Oxa9cumJmZiX2GLCwsYGRkBIlEAj8/P8yZMwdubm5wc3PDnDlzYGxsjL59+4plhwwZgjFjxsDGxgbW1tYYO3YsKleujJYtW+YpHjaQiIiIiIhIY5YvXw4AaNq0qcL8tWvXYtCgQQCA8ePHIzk5GSNGjEBMTAzq1KmDw4cPw8zMTCy/cOFC6OnpoWfPnkhOTkaLFi0QFBQEXV3dPMXD5yBpCT4/Qj34rB71Yd2qB597oj58DpL68DlI6sHrgfoU1ucgHb4TUaD7a1XBrkD3l1/YB4mIiIiIiEiODSQiIiIiIiI59kEiIiIiItICkgIe5ruoYh8kIiIiIiItcOROZIHuz7uCbYHuL78wg6Ql2LlVPTiQgPqwbtWDnbLVx1hfwnNWTQz1gI4rL2g6jGJn9/BaSErl9UAdjA0KZ6ZGp3CGVeiwDxIREREREZEcM0hERERERFqAfZBUwwwSERERERGRHDNIRERERERaQMIEkkqYQSIiIiIiIpJjBomIiIiISAuwD5JqmEEiIiIiIiKSYwaJiIiIiEgL8DlIqmEGiYiIiIiISI4NJCIiIiIiIjneYkdEREREpAU4SINqmEEiIiIiIiKSYwaJiIiIiEgL8EGxqmEGiYiIiIiISI4ZJCIiIiIiLcAEkmqYQSIiIiIiIpJjBomIiIiISAvosBOSSphBIiIiIiIikmMGiYiIiIhICzB/pBpmkIiIiIiIiOSYQSIiIiIi0gZMIamEGSQiIiIiIiI5ZpCIiIiIiLSAhCkklTCDREREREREJMcGEhERERERkRxvsSMiIiIi0gJ8TqxqmEEiIiIiIiKSYwaJiIiIiEgLMIGkGmaQiIiIiIiI5JhBIiIiIiLSBkwhqYQZJCIiIiIiIjlmkIiIiIiItAAfFKsaZpDe8/r1a/j6+qJ06dKQSqWQyWRo3bo1zp49CwAoU6YMJBKJ0jR37lwAwOPHjyGRSHD16lVxm/Hx8WjatCk8PDzw7NkzAMhxGxKJBCEhIQCAEydOKMy3sbFB8+bN8ffffxdshRSg0M2b0LZVc9SqVhm9e3TF5UsXNR1SscG6VQ/Wa/5r16o5qnl6KE0Bs2ZqOrRig+dt3rStYIdfu1VCyKDqCBlUHfM6VUB1JwtxuaGeDnwblMaavlWwdXANLO3hibYV7BS2ITOTwt/bFRsGVEXIoOoY36IcLI34G3VOLl28gO+++QrezRuhWmUPHD92VGG5IAhYsew3eDdvhLo1q2DolwPw78MHGoqWiis2kN7TrVs3XLt2DevWrcP9+/exe/duNG3aFNHR0WKZmTNnIiwsTGEaNWpUjtuLiIhAs2bNkJCQgNOnT8PJyUlctnbtWqXtdO7cWWH9e/fuISwsDCdOnICdnR3at2+P169fq+XYNenggf2YNzcAw4Z/jdBtO1G9eg2M8B2GsJcvNR1akce6VQ/Wq3psDNmGIydOidPyVWsAAN6tWms4suKB523eRSamYt0/zzF6xy2M3nEL11/GYXIrVzhZGQIAhtQrjeqlLLDg+H8YueUGdt94heENnFHH2RIAINXTwYz27gCAKXvvYcKuO9DTlWBKazf+lp+D5ORkuLt7YOKkH3JcHrRmNTauD8LEST9g4+atsLG1w1fDByMxMaGAIy2aJJKCnYoqNpCyefPmDU6fPo2ffvoJzZo1g7OzM2rXrg1/f3+0b99eLGdmZgaZTKYwmZiYKG3v2bNnaNSoEczMzHD8+HHY2toqLLe0tFTajqGhoUKZEiVKQCaToXLlypgyZQpiY2Nx/vx59VSABm1YtxZdunVD1+49ULZcOYz3nwyZgwxbQjdrOrQij3WrHqxX9bC2toatrZ04nTp5Ak5OpVGjVm1Nh1Ys8LzNuwtPY3HpWSxexqbgZWwKNl54gbdpGfAoYQoA8LA3wZ/3I3EzLB6vE1Jx6G4EHkUlwdUu83tBBXtTlDCVYtGJ//AkJhlPYpKx+MQjuJcwhVdJc00eWqHUsFFjjPzWDy1atlJaJggCgjeux5BhX6FFy1ZwdXPHj7Pn4u3btziwb68GoqXiig2kbExNTWFqaoqdO3ciJSXls7Z17949NGjQAB4eHjh48CDMzMw+a3tJSUlYu3YtAEBfX/+ztlXYpKWm4s7tW6hXv6HC/Hr1G+Da1Ssaiqp4YN2qB+u1YKSlpWL/3t3o1KUrJEX5p8hCguft59ORAI3KWcNQXwd3X2VmLG6HJ6C2sxWsjTM/mys7mMHRwhCXn8UCAPR1M8/dtHRB3E5aegbSMwRUlJkW8BEUbS+eP0dkZATq1W8gzjMwMECNGrVw7RrPYVVICngqqngDbDZ6enoICgrCsGHDsGLFClSvXh1NmjRB79694eXlJZabMGECpkyZorDu3r170bRpU/H1F198gfr162P79u3Q1dXNcX99+vRRWnb9+nWULVtWfF2qVCkAmQ0kQRBQo0YNtGjR4nMPtVCJeROD9PR02NjYKMy3sbFFZGSEhqIqHli36sF6LRjHjx1DfHw8OnTuoulQigWet5/O2coI8zpXgIGuDpLT0jHn8EM8e/MWALDqzFN807gMgvpXxbuMDAgC8Ntfj3FH3oC69zoRb9+lY1CdUlj/zwtIJMCgOqWgqyOBlXHx+sFT3SKjMs9Ta6Vz2AZhYbxNlPIPG0jv6datG9q3b49Tp07h7NmzOHjwIObNm4fVq1dj0KBBAIBx48aJf2cpWbKkwutOnTphx44d2L59O3r27JnjvhYuXIiWLVsqzMveRwkATp06BRMTE1y5cgUTJkxAUFDQBzNIKSkpStkvqVQK6Eo/dNiFwvu/EAuCwF+N8wnrVj1Yr+q1849taNCwEUqUsNd0KMUKz9u8exH7Fn7bb8HEQBf1Xazh19QFk/bcxbM3b+HjaQ/3Eib48eB9RCSkopKDGb5q4IyYpDRcexGHuLfv8NORf/F1I2f4eNpDEIC//o3Cw4hEZGRo+siKJqVzOId5lAtWk0rYQMqBoaEhvL294e3tjalTp2Lo0KGYNm2a2CiytbWFq6vrB7cxadIkeHl5oV+/fhAEAb169VIqI5PJProdFxcXWFpawt3dHW/fvkWXLl1w8+bNzEZPDgICAjBjxgyFedOmTcPEKdM/uB9NsrK0gq6uLiIjIxXmR0dHwcbGNpe1SBWsW/Vgvarfy5cvcP7cWfyy6DdNh1Js8Lz9dO8yBITFZf74+DAyCa52xuhQ2R6rzzzFgFolEXD4IS7Kb6l7HJ0MFxtjdPGS4dqLOADA1Rdx8A25ATOpHjIEAYmp6VjXvypexUfnuk9SZmuTOTpgVGQk7OxKiPOjo6KUskpEn4N9kFRQsWJFJCYm5nm9KVOm4Mcff0S/fv2wefPnd4AdMGAAMjIysGzZslzL+Pv7IzY2VmHy9/f/7H2rk76BASpUrIRzZxSHMD935gyqVK2moaiKB9aterBe1W/3jj9gbW2DRo2baDqUYoPnbf6RSCTQ19GBro4E+ro6eD8RlCEIOY7gFZ/yDomp6fByNIOFkR7+efKmIMItNkqWKgVbWzucO3tGnJeWlopLly6gShWew5R/mEHKJioqCj169MDgwYPh5eUFMzMzXLx4EfPmzUOnTp3EcvHx8QgPD1dY19jYGObmyqPRTJw4Ebq6umLjpl+/fuKyN2/eKG3HzMwsxxHxAEBHRwd+fn6YNWsWfH19YWxsrFRGKpXmmF16++7Dx65pAwZ+ickTx6OipyeqVKmG7VtDERYWhh69ems6tCKPdaserFf1ycjIwK6dO+DTqTP09PgxlZ943ubdgFolcelZLCITUmGkr4tGrtbwdDDDjAP3kZyWgRsv4/BlnVJIfZeBiIQUVHIwQzM3W6w5+1TcRgt3Wzx/k4zY5HfwsDfF0PqlsfvGK7yIfavBIyuckpIS8ezp/+vuxYvnuHf3DswtLODg4Ii+/b9A4OrfUdrZGaVLOyNw1e8wNDRE2/Y+Goy66OCDYlXDT55sTE1NUadOHSxcuBD//vsv0tLS4OTkhGHDhmHSpEliualTp2Lq1KkK6/r6+mLFihU5bnfcuHHQ1dXFwIEDkZGRgQEDBgAAvvzyS6WyAQEBmDhxYq4xDh48GNOmTcOSJUswfvz4TznMQqlN23aIfRODlcuXISLiNVzd3LF0xUo4Opb8+Mr0Qaxb9WC9qs/5s2cQHvYSnbt01XQoxQ7P27yzNNLH983KwtpYH4mp6XgclYQZB+7jqvz2uZ+P/YsvapfCmOZlYSrVQ0RCCjZeeI4Dd/4/8EVJS0N8UbsUTKW6eB2fiq1XXmLXjVeaOqRC7fatmxg2eKD4ev7PcwEAHTp2xszZczFo8FCkpLxFwKyZiIuLhWdlLyz/PRAmJhwRkPKPRBAE4ePFqKgr7BmkospQj3WrLqxb9TDUA5LSeNlXB2N9Cc9ZNTHUAzquvKDpMIqd3cNrISmV1wN1MDYonJmaq0/jC3R/VUt/3mNuNIV9kIiIiIiIiOR4ix0RERERkRYonHmtwocZJCIiIiIiIjlmkIiIiIiItAFTSCphBomIiIiIiEiOGSQiIiIiIi3A5yCphhkkIiIiIiIiOWaQiIiIiIi0gIQJJJUwg0RERERERCTHBhIREREREZEcb7EjIiIiItICvMNONcwgERERERERyTGDRERERESkDZhCUgkzSERERERERHLMIBERERERaQE+KFY1zCARERERERHJMYNERERERKQF+KBY1TCDREREREREJMcMEhERERGRFmACSTXMIBEREREREckxg0REREREpA2YQlIJM0hERERERERyzCAREREREWkBPgdJNcwgERERERERybGBREREREREGvXXX3+hQ4cOcHR0hEQiwc6dOxWWC4KA6dOnw9HREUZGRmjatClu3bqlUCYlJQWjRo2Cra0tTExM0LFjRzx//jzPsbCBRERERESkBSSSgp3yIjExEVWqVMGSJUtyXD5v3jwsWLAAS5YswYULFyCTyeDt7Y34+HixjJ+fH3bs2IGQkBCcPn0aCQkJ8PHxQXp6ep5iYR8kIiIiIiLSqLZt26Jt27Y5LhMEAYsWLcLkyZPRtWtXAMC6detgb2+P4OBg+Pr6IjY2FoGBgdiwYQNatmwJANi4cSOcnJxw9OhRtG7dWuVYmEEiIiIiItICkgKe8sujR48QHh6OVq1aifOkUimaNGmCM2fOAAAuXbqEtLQ0hTKOjo7w9PQUy6iKGSQiIiKi/7V33/E13f8fwF9XbmTIENkRm4TYQkuIUKu1V1CjoqRqE4QYtcVWrVHUbuxVbc0Sav1siZGYkRhByBARuRnv3x+SU7eJNu2X3IzX0+M+2nzOuTfve1znnvd5fwYRvXdJSUlISkrSajMwMICBgcG/ep3Hjx8DAGxtbbXabW1tER4eruxTuHBhWFhYZNon4/nZxQoSEREREVFBkMMlJH9/f5ibm2s9/P39/3v4fxnYJCKZ2v4qO/v8FRMkIiIiIiJ67/z8/BAXF6f18PPz+9evY2dnBwCZKkFPnz5Vqkp2dnbQaDSIiYl55z7ZxQSJiIiIiKgAUOXwHwMDA5iZmWk9/m33OgAoU6YM7OzscOjQIaVNo9Hg2LFjcHNzAwC4urpCX19fa5/IyEhcvXpV2Se7OAaJiIiIiIh06uXLl7h9+7byc1hYGC5fvoxixYqhZMmSGD58OGbOnIkKFSqgQoUKmDlzJoyNjdG9e3cAgLm5Ofr27YuRI0fC0tISxYoVw6hRo1C1alVlVrvsYoJERERERFQA/Nu1iXLS+fPn0bhxY+VnHx8fAEDv3r2xdu1a+Pr6IjExEQMHDkRMTAw+/vhjHDx4EKampspzFi5cCLVajS5duiAxMRFNmjTB2rVroaen969iUYmIvJ+3RUREREREudXtp4k5+vvK2xjl6O97X1hBKiBep+g6gvzJUM1j+6Hw2H4YhmogMk6j6zDyJXvzwniVzHuOH4KxvgqJybqOIv8x0gdqTjmi6zDypUuTPtF1CFnKxQWkXIWTNBAREREREaVjBYmIiIiIqCBgCSlbWEEiIiIiIiJKxwSJiIiIiIgoHbvYEREREREVACr2scsWVpCIiIiIiIjSsYJERERERFQA5OaFYnMTVpCIiIiIiIjSsYJERERERFQAsICUPawgERERERERpWMFiYiIiIioIGAJKVtYQSIiIiIiIkrHChIRERERUQHAdZCyhxUkIiIiIiKidKwgEREREREVAFwHKXtYQSIiIiIiIkrHBImIiIiIiCgdu9gRERERERUA7GGXPawgERERERERpWMFiYiIiIioAOAkDdnDChIREREREVE6VpCIiIiIiAoElpCygxUkIiIiIiKidKwgEREREREVAByDlD2sIBEREREREaVjBYmIiIiIqABgASl7WEEiIiIiIiJKxwoSEREREVEBwDFI2cMKEhERERERUTpWkIiIiIiICgAVRyFlCytIRERERERE6VhBek+ePn2KiRMnYt++fXjy5AksLCxQvXp1TJ48GfXq1UPp0qUxfPhwDB8+HABQunRphIeHAwAMDQ1RqlQp9O3bF6NGjYLqLx1ET506BXd3dzRr1gz79+/P6beWIy6cP4e1q1ch5PpVREVFYeF3S/BJk6a6Divf2LIpAGvXrMKzqCiUK18BvmPHoZZrbV2HlafxM/t+rFmxFOt+XKbVZlHMErv2HwUANPqoapbP+3qID7r16vOhw8tXWjb/BJGPHmVq79KtO/wmfKODiPKPZUu+x/Jli7XaLC2tcPjYSR1FlPd82aAUhjQph4D/u495B24BAKa0q4S2Ney19gt+EIfeqy5otVVzNMOgT8qhanEzpKSl4cbjlxgcEISklLQci5/yFyZI70mnTp2QnJyMdevWoWzZsnjy5AkOHz6M6Ojodz5n6tSp8Pb2xuvXr/H7779jwIABMDMzQ//+/bX2W716NYYMGYIff/wRERERKFmy5Id+OzkuMfEVnJ2d0a5DR4wcPkTX4eQr+/ftxZxZ/hg/cRJq1KyF7Vs3Y2B/b+za8xvsHRx0HV6exc/s+1O6bHnMX7xS+VlP78/ODTv2Bmrte/b0ccyZPgkNP2Ey+m/9tHk70tJSlZ9v37qFAd5folnzFjqMKv8oV74Clv+4Rvm5UCE9HUaTt7g4mKJjLQfcfByfadvJW88x6ecQ5efkVO2kp5qjGRb3qIE1J8Ixe99NpKSmwcnWBGkiHzzuPIk97LKFCdJ7EBsbixMnTuDo0aPw8PAAAJQqVQofffTR3z7P1NQUdnZ2AIB+/fph2bJlOHjwoFaClJCQgK1bt+LcuXN4/Pgx1q5di2++yX93+hq4e6CBu4euw8iXNqxbgw6dOqFjZ08AgK/feJw6dQJbt2zCsBEjdRxd3sXP7Pujp6cHSyurLLf9tf3EsUDUdP0IDsVL5ERo+UqxYsW0fl7z40qUKFESrnX+/ruKskdPTw9WVta6DiPPMdLXw8yOlTHtl1D0a1g603ZNahqeJ2je+fyRLSpg89n7WHMyXGmLiE78EKFSAcIxSO+BiYkJTExMsHv3biQlJf3r54sIjh49ipCQEOjr62tt27JlC5ydneHs7IyePXtizZo1EN4VoWxK1mgQcv0a6rk10Gqv51YfQZcv6SgqIm0P70egU8tP0K3dp5gyfjQePbyf5X7Rz5/h/04eR8u2HXI4wvwnOVmDvb/uQbsOHTN166b/JiIiHM0aN0DLFp9gzKgReHA/688xafNr6YTjt57hTFhMlttrly6Kw6MaYPfgupjYpiIsjP+8TrIw1kc1R3NEJyRj7Zeu+H1kA/zYuyZqlDDPqfDzHFUOP/IqJkjvgVqtxtq1a7Fu3ToULVoU9evXx7hx4xAcHPy3zxszZgxMTExgYGCAxo0bQ0QwdOhQrX1WrVqFnj17AgA+/fRTvHz5EocPH/5g74Xyl5jYGKSmpsLS0lKr3dLSCs+eRekoKqI/uVSpCr/JMzD3ux8wavwkRD9/hkF9eyEuNjbTvgd+2wPjIsZwb8zudf+rwMOHER8fjzbtmWy+D1WrVcP0mbOxdPkqfDN5Op49e4bePbshNjbri356o0VlG1S0N8X3v9/NcvvJ288xbud1fLXuEhYcvIXKDqZY0bsm9PXeXHo7WhgBAPp7lMHOi48wKOAyQh7HY/kXNVGymFGOvQ/Kf5ggvSedOnXCo0ePsGfPHrRo0QJHjx5FrVq1sHbt2nc+Z/To0bh8+TKOHTuGxo0bY/z48XBzc1O237hxA2fPnkW3bt0AvEnEunbtitWrV7/zNZOSkvDixQutx3+palH+8tc7xCLCu8aUK3zs5g6PT5qhbHkn1P6oHmYtXAIAOPDbz5n23fvLLjRt0QoGBgY5HWa+s3vndtRv4A4bG1tdh5IvNHD3QNNmLVDByRl167lh8dLlAIBfft6t28ByMVszA4z+1AkTdl2HJjXryRQOXnuKE7ee405UAv64+RyDA4JQytIY7hXedL0tlP41tuPCQ+y5HIkbj19i/oHbuPf8FdrVtM/yNQs6lSpnH3kVE6T3yNDQEM2aNcM333yDU6dOwcvLC5MmTXrn/lZWVihfvjzq1auHHTt2YOHChfj999+V7atWrUJKSgqKFy8OtVoNtVqNZcuWYefOnYiJyfqulL+/P8zNzbUe/v7+7/29Ut5gUdQCenp6ePbsmVZ7dPRzWFpmPeaDSJeMjIxRtnwFPLgfodUefOkC7offQ6t2nXQUWf7x6NFDnPm/02jfyVPXoeRbRsbGKF/BCRHh93QdSq5Vyd4UliaFEfBVbZyb2AjnJjZC7dIW+PxjR5yb2EhJft727KUGkbGvUdLyTXUo6uWbsUl3oxK09guLSoCdmeEHfw+UfzFB+oBcXFyQkJDwzzsCsLCwwJAhQzBq1CiICFJSUrB+/XrMnz8fly9fVh5BQUEoVaoUAgICsnwdPz8/xMXFaT38/Pze59uiPES/cGFUcqmM/zulPdXs/506heo1auooKqJ302g0CL93N9PkDL/t2Qmnii4o7+Sso8jyjz27dqJYMUu4N+QkIx+KRqNBWNgdWFlz0oZ3ORsWg85Lz6DbD+eUx7WHL7A3+Am6/XAOaVkMtzY3UsPW3ADP4t8kRo9iX+PpiySUtjLW2q+UpTEi417nxNvIc1Q5/Cev4ix278Hz58/h6emJL7/8EtWqVYOpqSnOnz+POXPmoF27dtl+nUGDBmH27NnYsWMH1Go1YmJi0LdvX5ibaw827Ny5M1atWoXBgwdneg0DA4Msu5+8Tvn37ysnvUpIQETEn3eMHz54gNCQEJibm3Mq6v9Rr959MH6sL1yqVEH16jWxY9sWREZGwrNrN12HlqfxM/t+LF00D27uHrC1tUdMTDQ2rF6BVwkJaNHqz3NnwsuXOHb4EAYMG6XDSPOHtLQ0/Lx7F1q3aw+1mpcA78uCubPRsFFj2NvbIzo6GiuXL0PCy5do045jvN7llSYVd/5S+UlMTkVcYjLuRCXASF8PXzcqg8MhTxEVr4FDUUMMaVIOsa+ScST0zzG0606F4+tGZXHz8UvcePwSbWrYobSVMUZvu5rTb4nyEZ4d3wMTExN8/PHHWLhwIe7cuYPk5GSUKFEC3t7eGDduXLZfx9raGr169cLkyZNRpkwZNG3aNFNyBLwZ7zRz5kxcvHgRtWrVep9vRWeuXbuKfn2+UH6eN+dNt8C27Tpg2sxZugorX/j0s5aIi43BimVLERX1FOUrOGHJDyvg4FBc16HlafzMvh9RT59g2oQxiIuNQVGLYnCpUg1LVwXAzv7PJPPIoX0QETRp8ZkOI80fzpw+hceRj9C+Q0ddh5KvPHnyGH6+PoiJiYVFMQtUq1YD6zdu5Xn2f5AmgvI2RdC6ejWYGqrxLF6Dc/diMGb7VbzS/Lme18YzD2Cg1sPIFhVgbqSPm09eYsCGy3gQw6m+s5R3izo5SiWcM7pAyO0VpLzKUM1j+6Hw2H4YhmogMu7da4rQf2dvXhivkvmV+iEY66uQmKzrKPIfI32g5pQjug4jX7o06RNdh5ClqJc5+8VqbZI3azF5M2oiIiIiIvpXWEDKHk7SQERERERElI4VJCIiIiKiAiAvr02Uk1hBIiIiIiIiSscEiYiIiIiIKB272BERERERFQB5efHWnMQKEhERERERUTpWkIiIiIiICgBO0pA9rCARERERERGlY4JERERERESUjgkSERERERFROo5BIiIiIiIqADgGKXtYQSIiIiIiIkrHChIRERERUQHAdZCyhxUkIiIiIiKidKwgEREREREVAByDlD2sIBEREREREaVjgkRERERERJSOXeyIiIiIiAoA9rDLHlaQiIiIiIiI0rGCRERERERUELCElC2sIBEREREREaVjBYmIiIiIqADgQrHZwwoSERERERFROlaQiIiIiIgKAC4Umz2sIBEREREREaVjBYmIiIiIqABgASl7WEEiIiIiIiJKxwoSEREREVFBwBJStrCCRERERERElI4JEhERERFRAaDK4T//1tKlS1GmTBkYGhrC1dUVx48f/wBH4Z8xQSIiIiIiIp3asmULhg8fjvHjx+PSpUtwd3fHZ599hoiIiByPhQkSERERERHp1IIFC9C3b1/069cPlSpVwrfffosSJUpg2bJlOR4LEyQiIiIiogJApcrZR3ZpNBpcuHABzZs312pv3rw5Tp069Z6Pwj/jLHZERERERPTeJSUlISkpSavNwMAABgYGWm3Pnj1DamoqbG1ttdptbW3x+PHjDx7nXzFBKiAM88DfdFJSEvz9/eHn55fpH05uxmP74fDYfhj25oV1HUK25MVja6yf++fQzYvHFQCM9HUdwT/Li8f20qRPdB1CtuTFY5sb5fT36uTp/pgyZYpW26RJkzB58uQs91f9pewkIpnacoJKRCTHfytRFl68eAFzc3PExcXBzMxM1+HkKzy2Hw6P7YfDY/th8Lh+ODy2Hw6Pbd6U3QqSRqOBsbExtm3bhg4dOijtw4YNw+XLl3Hs2LEciTcDxyAREREREdF7Z2BgADMzM61HVhXAwoULw9XVFYcOHdJqP3ToENzc3HIqXEUe6MBCRERERET5mY+PD3r16oXatWujXr16WLFiBSIiIvD111/neCxMkIiIiIiISKe6du2K58+fY+rUqYiMjESVKlWwd+9elCpVKsdjYYJEuYaBgQEmTZrEwZcfAI/th8Nj++Hw2H4YPK4fDo/th8NjWzAMHDgQAwcO1HUYnKSBiIiIiIgoAydpICIiIiIiSscEiYiIiIiIKB0TJCIiIiIionRMkIiIiIiIiNIxQSIies9evnyp6xAomzhP0b/D40VEBQETJCL6R7du3cLRo0cBAGlpaboNJpeLjIzE559/jg0bNug6FHqHBw8eICgoCACgUql40Z8Nz58/B/DmeBHlJvxOog+BCRIR/aMBAwZg6dKlAIBChXja+DtPnz5FSkoK1qxZg61bt+o6HPqLlJQUtG3bFv3798fFixcBMEn6Jzdv3kTr1q2VmyT0v7l37x6WLl2K0aNH4/z583j9+rWuQ8pz7ty5g/Xr1wP48zuJiRK9T7zSoXzr2bNnAP48aaampuoynDytTp06PH7ZVL16dcyYMQPW1tZYvHgxk6RcRq1WY8uWLYiOjsa4ceNw4cIFAEyS3iUoKAi1a9fGmTNncOPGDV2Hk+cFBwfDw8MDGzduxNatW/HJJ59gy5YtAHiBn10vXryAm5sbvLy80Lx5c8ybNw93797VunnHY0n/KyZIlC/duXMHtra2GDBgACZNmoS4uDjo6enpOqw8468Xih999BHOnTuH6OhofvFkQ61ateDr6wt7e3smSblMSkoKKlSogP379+PGjRsYP348k6R3CAoKQt26dfHVV1+hXbt2iImJAcCLz/8qKCgI9erVQ69evfDbb78hKCgI5cuXx6xZs5CUlIRChQrx2GaDkZERevTogZUrV6JNmza4ePEiatWqBX9/fxw4cAAAq0r0v2OCRPlSREQERASmpqa4d+8eqlatijlz5uDYsWNa+7EqkrXk5GStn0uWLInExES8evWKXeyykNVFtaurK5OkXOLevXs4efIkEhISoFarAQBly5bFwYMHcePGDfj5+WklSfTnxbyPjw/mzZuHwoULK+O2mET+e48ePUKtWrXg7e2N6dOnw8zMDEWLFoWtrS2ePXumTOzC8+s/09fXh4ODA7799lt4e3tj48aNWLVqFe7cuYMuXbpg0KBB2LdvH1JTU3k86T/jJ4fylYyEp3Hjxhg6dCgsLCywYcMGDB8+HOHh4ejcuTNGjBiBn3/+GQCUqhK/8P8UHh6OWrVqwcvLC2PHjsWxY8dgbGyMsmXLKheRvCv3xu3bt3H8+HGoVKosj8lfk6Tt27frIMqCLTIyEk5OTnB3d8fnn38OT09PHDx4ELdu3UKFChVw4sQJ3Lt3DxMmTMDZs2d1HW6ucPHiRXz88cfw8fHBjBkzAABmZmZKt+WsqvE8h/69ly9fomrVqjh69Ciio6OhUqkwe/ZspeIxdOhQODk5YdasWTh27Bji4+N1HHHulHGeHTVqFIoXL47p06cDADp16oQnT56gWLFiCAoKwpQpU1CmTBkcOXJEl+FSHsYEifKN0NBQDBkyRPm5ZMmS2LNnD9LS0uDj44MlS5agdOnSCAgIwOTJk9GwYUNs3LgRjx8/5l3jt9y/fx/du3eHsbExDh48iCFDhqBevXo4d+4c9u7dC+DNXc6CfkGUnJyMRYsWwcPDA0ePHn1n95iMJMnR0RFTpkzB7t27cz7YAiwpKQmfffYZAKBSpUoQEQwdOhS1a9dGr169cPjwYaxatQrnz5/HkiVLcOrUKR1HrFsJCQmYOHEiBg8erFx8AkDlypWVLnYpKSmZnsdzaNYiIyPx4MEDODk5YcuWLdDX10fjxo3xzTffYOHChdi1axcuXryImTNnomfPnjh+/DgaN26M7t27M0nKQqFChZCcnAwRQePGjXHr1i0AQO/evXHu3DkcPnwYgYGBmDNnDho3bgx7e3sdR0x5lhDlE5s2bZKyZcvKq1evlLYqVarIzJkzRUSkd+/eUqJECfnjjz/k8uXL0qRJE6lZs6Y8fPhQVyHnKq9fv86yPS4uTs6cOSNLliyRIkWKyLBhw5RtaWlpORRd7hQSEiJ9+/YVCwsLOXz4sIiIpKamZrnvtWvXpHfv3hIWFpaDEZKIyK1bt6RVq1ZStmxZiYqKktjYWAkICJCvvvpKbG1tpUmTJmJsbCwqlUq++OILSUxM1HXIOpOSkiK3b9/O1P7TTz+Jra2txMfHS0pKitLu7+8vEyZMyMkQ84yLFy+KiYmJ7N+/X2kLDQ0VDw8PUalUsn379kzPSUxMlAsXLsitW7dyMtRcKywsTBYsWCBTpkyR3bt3a22LjIwUGxsbKVGihDg4OMjFixd1FCXlR0yQKN8IDAwUc3NzefTokdK2aNEi6dOnj7Rq1Urs7Ozk3LlzWs+JiorK6TBzpWvXrkm7du1k1qxZkpSUJMnJySKinQClpaXJTz/9JAYGBjJy5EhdhZorvJ0E3bx5U7y8vMTCwkJ+//13ERGtC8jXr1+Ln5+fXLp0SaudPryMz7GISHh4uNSvX19Kliwpd+/eFZE3f09JSUmyf/9+mTZtmri7u8u1a9d0FW6uk5ycrJwDfvvtNyVByjBx4kQpXLiwXL58WVch5lqXL18WExMTGTVqVKZtV69eFTc3N3FxcZFnz56JiEhSUpKIvPsGS0EUFBQkDg4O0qxZM7GyspJy5crJjBkzROTP47RgwQIpXbq0HDlyRJehUj7EBInyvIyLoJiYGHFxcZHTp08r265evSoWFhZibm6udeeeF6p/Sk1NlREjRkilSpXE2dlZPv30UxkxYoTcu3cv074ajUY2bdokKpVKxo0bp4NodUuj0Sj///bF99tJ0qFDh0TkTUL5+vVrGTJkiBQqVEiCg4NzPN6C6M6dO7JgwQLx9PSUXr16ybJly5TP8sOHD6Vhw4bi4OCQZSWvoFaO7t+/LwEBAeLj4yNz586VXbt2KdsyLkRDQ0PFxsZGQkJCRERkwoQJYmRkJOfPn9dFyLlaUFCQGBkZZTpHvp18h4aGSu3ataVixYry/PlzEWFy9Lbg4GAxMjKSiRMnikajkQcPHkjlypXlo48+0krSjx07JtbW1rJ3714R4TGk94cJEuVZkZGREh8fr9x5S0tLE1dXV5kyZYqIaN9hql+/vjx48EBnseZ2P/30k7Rs2VJevnwpmzZtko4dO4qNjY2MGzdOueDP8Pr1a9m2bZtyoVRQXLt2Tbp27Sr+/v6SmpqaqXvhjRs3MlWSBg4cKEZGRuz6kUOCgoLE3t5e2rZtK02aNJFPPvlEVCqV1KtXT37++WcREXnw4IE0atRIHB0dJTw8XEQky4ppQREUFCRly5aVevXqSY0aNcTc3FyMjY3F09NTq9ttWFiYGBgYyOnTp8Xf318MDQ2ZHGXh+vXrolarM1XZJ0+eLDY2Nlo9HEJDQ6Vu3bpia2sr0dHROR1qrhUWFiZFihSR7t27a7W3bdtWzMzMMn33DBgwQOk+S/S+MEGiPOnp06dSvXp1sbCwEDc3N+nfv7/MmjVLevTooXRpyLjoOXTokDg7Oyt3mChrtWvXlvHjxysXib/++quYm5uLnp6e9OzZUwICArTGdxUkKSkpMnjwYHFxcZHGjRtLzZo1Zfr06ZmqQiEhIeLl5SXW1tbSrFkzKVKkiFy4cEFHURcsd+/eFQcHB5kwYYLWhf2RI0fE2tpaqlWrpnTDCQsLk2bNmomxsbFEREToKmSdu3nzplhZWcnYsWOVKkZYWJjMmDFDTExMpGXLlsq+sbGxUqNGDXF1dRVjY+NM3ZXpjR07dohKpRJ/f3+JiYkRkTfjtGxsbOS3334TEe1E/Pr16+Lh4ZHluK+CKjQ0VGxtbcXT01NOnTolIiKzZ88WlUolFStWlGbNmkmLFi2kX79+EhoaKv7+/tK2bVt5/PixjiOn/IQJEuU5GV86N2/elF27dsny5culY8eO0rBhQzEwMBBjY2OtbnYiIh4eHlK3bt0CeYf4n2Qkkps3b5a2bdtKbGysiIj069dPnJycZMeOHdKuXTspUaKEuLm5FdjuievXrxcXFxfRaDSyevVq6dWrl5ibm8ukSZO0ku+IiAjp0qWLWFhYsHKUg+bNmycdOnTQGkOX8d9Tp06Jqamp9OjRQ9n/9u3b0rZtW7l586ZO4tW1tLQ0GTZsmHTr1i3TttjYWFmyZIkYGBhojaH5+OOPxdjYWIKCgnIy1Dzh7t27yti2lStXikqlkgULFsjkyZOlWLFicvDgwUzPybigf7vrbkH24sULpfvcpUuXxNnZWbp16yZff/21FCtWTA4cOCDh4eFy584dmT9/vjRo0EDs7OykTp06TDDpvWOCRHlKVFSUuLu7v3P8y8WLF8Xb21ssLS3l2LFjSvvJkyc5K1A6jUaT5ViL8PBwKVmypGzevFkGDx4s9vb2cvbsWRERefnypZw9e7bAfwl99tlnMnXqVOWCJjAwUAoXLizm5ubSpk0bOXDggMTFxUlUVJRERkbqONqCpWvXrtK0adNM7Rldbb/77jspVKiQ1jiQt8eRFUQeHh7i6+srIpnHbkRFRUnHjh2lWrVq8vLlS0lNTZX169fLnTt3dBFqrvb06VNxdHTUGrv1ww8/iEqlEpVKJTt37sz0nEmTJkmfPn0KbFX+rx49eiRubm6yYsUKefHihYi8+T6vVKmSqFQqWbhwYZbP++WXX+T+/fs5GCkVFFwHifKUtLQ01KhRA7///jsmT56stCclJQEAatasCR8fH7Rq1QpdunTBwYMHAQBubm4oX768LkLOVUJDQ+Ht7Y2mTZti/PjxuH//PoA3C+yWLFkSY8aMweeff47du3dj7969qFOnDkQERYoUQZ06dVCuXDkdv4OcExUVhfPnzyuL4wJA69atcfr0aejr6wMAtm7dCgcHB2zcuBGJiYno378/WrZsCQsLC9jZ2ekq9AIlLS1NebzdJunrdGWsz1OrVi0ULlwYcXFxyn5qtTpng81lXrx4gSdPngB4s75MhrS0NFhZWcHLywvXrl3DkydPUKhQIfTq1Qtly5bVVbi5lr6+PjQaDcqUKaOsEdW/f3/89NNPAIDr168jOjpa2X/SpEmYOnUqBg8eDCMjI53EnNvY29vDwsICixcvxq5duxAfH4+aNWti27ZtcHZ2xh9//KG1RllycjKAN+dkR0dHXYVN+ZmuMzSifysyMlLGjh0rNWvWlMmTJyvtf51VrEOHDlK2bFlJSEhg1zp5M+1ssWLF5PPPP5cRI0aIoaGhjB8/XmufCxcuiJOTk/zwww8iUnBn+7t27ZrUr19fPv30U+nYsaPy2YqOjpYSJUrIypUr5euvvxZ7e3tlLEZqaqocPnxY6WZDH1ZsbKwkJCQoP69atSrT3fqUlBTl3/6JEyfExcWlwFdBRd6cK1NTU6Vr167i5OQkV69eVba9fa5ct26duLi4FNjZ/bIjNTVVbt26JVZWVsqkH28fw+XLl4tKpZJJkyaJRqORyZMni4GBAccmpgsLC5OlS5cqlUlPT0+pXLmyrFu3TquS5OzsLO3atcvUfZ7oQ2GCRLlefHx8pj7a9+/fl7Fjx0r16tXfmSTdvn2bi8Cmu3z5shgbG4ufn5/SNmvWLOncubM8ffpUGdclItK3b19xdnYusP3ir169KkWLFpVx48ZJeHi40vUo47OVMTajXLlyyixenFo2Z4WEhEjz5s3Fz89P+exev35d6tatK87Ozspg+Lf5+PhI/fr1lTF2BY1Go8n0b/rcuXOir68vPXv21JrWP+PGyLBhw6Rdu3ZaiSi98eTJE2UykNDQUDE3N1cSJBHtJGnFihWir68v1atXF1NTU87+ly44OFicnJykQ4cOsmfPHqW9c+fO4uLikilJqly5sjRq1EjOnDmjq5CpAGGCRLlaSEiIWFtbS5MmTWTEiBFy6tQpZWG9uLg4GTdunNSqVUtrJfeCemH/Lk+ePBELCwvp3LmziPx58fPll19KxYoVxd7eXurWraskT7dv3xZ7e3v58ccfdRazrjx//lwaNGggQ4YM0Wp/+2Ln4sWLYmNjo1TZmBzlrODgYLGyspLBgwfLtm3btLbt2LFDqlatKlZWVjJ79mw5d+6cHD16VEaOHCkmJiYFdkHT0NBQ6datm7i7u0v79u21kqEVK1aIgYGBtG7dWjme165dkwkTJoiZmZlcuXJFV2HnWgkJCeLg4CC9evWS1NRUCQsLE1tbW4mIiHjn+WD58uViYGDAiVvShYSEiIWFhYwdOzbLG5ndunUTZ2dnrSTp7NmzUqdOHY45ohzBBIlytYypPStXriyOjo7i6uoq1tbWMnjwYNm+fbtcuXJFRo8eLR4eHuLv76/rcHOlq1evSufOncXe3l6OHz8uIiIzZ86UIkWKyA8//CAbN26UHj16iJ2dnWzfvl0SExOlc+fOBXIw9rVr16RcuXJy9OjRLC90MhIlHx8fcXFxkadPn+Z0iAVaRESElC9fPlPX0Lf98ccf0r17dylcuLCYmJiIs7OzNGzYsMAmR5cuXZJixYrJF198ISNGjJASJUpIvXr1lMkBNBqN7Ny5U2xtbaVw4cJiYGAgVapUkerVq8ulS5d0G3wulHFe2LVrl5iZmcnXX38tgYGBUqZMGbl+/bokJCTIixcv5MmTJxIdHS0vXrxQZv17e4HTguzVq1fSuXNnGTRokFa7RqORsLAwZYKbr776SpycnGTDhg0SFxcnIqI1hT/Rh8QEiXKlt/u8jxs3TipWrCgLFy6UkydPyo8//ihdu3YVc3NzqV+/vpQtW1acnJxEpVLJ4sWLdRh17vJ2V6Jr165Jjx49xNraWvr27St2dnayb98+ZfutW7eUKatFCm5VJCAgQNRqtZIIZXUcEhIS5LvvvpPq1avL5s2bczrEAm3r1q3i7u4uUVFRyt9NaGio7N69W7y9vWXNmjXKYpGhoaFy5MgRuXnzplYX0oIkODhYjI2N5ZtvvhGRN9Vjf39/UalUsnXrVq1979+/L4GBgbJy5Uo5d+4c15TJwvXr12XJkiVKcrl3714xMDCQGjVqiLGxsRgaGkqpUqWkRIkSYmpqKra2tuLo6CiOjo5cxPQtGo1GGjRoIN9//73Stn//fhk+fLiYmZmJo6OjdOzYUUREvL29xdbWVjZv3ixpaWkcT0w5pmBP4UO50r179+Dv749u3bqhcePGmDFjBuLi4rBkyRKoVCoMGjQIffv2RVRUFI4ePYo//vgDhw8fho2NDZo0aaLr8HOF0NBQuLu7Y/To0fD19YWLiwv8/PxQuHBhrF69GnPmzMGnn36K5ORkFCpUCMWLF0e1atVgZWWl69B1qnTp0lCr1di5cyc6deqkNbNXhg0bNmDPnj1wdHRErVq1dBBlwXX79m2Eh4crn9ONGzfip59+QkhICNRqNQ4ePIgjR45g0aJFcHZ2hrOzs44j1p0XL16gb9++sLS0xJQpUwAAenp6ePToEQAgPj4eFy5cgKurKzQaDRwdHeHo6IhGjRrpMOrcKygoCDVr1sTcuXOVmec+++wz7NmzBz169IC1tTWmTJkCU1NT6OvrIykpCQYGBjAwMEC5cuUK/Ln1bYmJiXj27BmCg4MRGhqKXbt2Yd26dahSpQqmTZsGExMTTJ06FVOnTsWKFSug0WhQp04dZUZKohyh6wyN6K+OHj0qxYsXl169eildwkREhg4dKiVLlpRFixZlurv5+PHjAnuX+K9SU1Nl9OjRolKpxN7eXqZOnapsCwoKkt69e4uVlZXWOlETJkyQ4sWLF/gZ2B48eCA2NjbStm1brXEab9+1HDZsmMyYMaPAVtl0KSQkRKysrMTd3V3atWsnJiYmMnr0aDlx4oSIiMyfP18cHR0lJCREx5Hq3osXL2Tp0qVSvHhx6d+/v4i8WUzXyMhIOnXqJJ6enlKxYkWpXLmyeHt7y5YtW+TJkyc6jjp3unTpkhgbG2dafy/jHHDo0CExMjKS0aNH87yQTYcPHxa1Wi2lSpUSU1NT+eGHH5S1CjUajTRv3ly6d++u4yipIGOCRLnS77//LlWqVJHPP/9cufgReZMklS5dWhYtWiTPnz/XYYS52+bNm6VMmTLi5+cnFSpU0EqSrl69Kj169BBLS0u5cOGCLFy4UAwNDTl4ON2OHTvEwMBAevXqpbWoaEJCgvj5+UnJkiUlNDRUhxEWLBnJaVpamiQlJcnBgwela9eu4unpKSdOnJCXL18q+x47dkwqVKjABCldbGysrF69WqytraV69epibW0tR48eFZE33e0SExNlxowZ0r59e7GysuLixlm4cuWKGBsbK92PM6xbt07Onj2rfD5/++03MTQ0lAEDBmh9JundIiIi5Pz585m6H6ampoqnp6dMmDCB3epIZ5ggUa7x1xPhwYMHlSTp7UrSsGHDpEKFCuLv7y/R0dG6CDXXenua8/r160vXrl3F399fSpQoITNmzFC2Xb16Vb744gtRqVSip6fHaWffkpqaKj/88IOo1WpxdnaWPn36yIABA6Rt27ZiY2PDRDKH3L59Wy5fvvzONXje/qxnGDVqlLi7uxfY84JGo5FXr15JTEyMJCUlicibxH7NmjVSpkwZadasmbJvxvYMrMBn9vjxYzE1NZXmzZtrfTfNmjVLVCpVpumm9+/fLyqVSnx8fHI61HwjKSlJJkyYIA4ODnLz5k1dh0MFGBMk0rm3p+xMTU3NlCRVrVpVevXqpbWYobe3t1SvXr3AXgj9VVZru2zevFl69+4twcHBMnnyZHFwcNBKki5duiQ+Pj5ax5X+dObMGencubPUrFlTGjRoIGPGjOEXdg55/PixqFQqsbOzEx8fH1m4cKGIiNbkGW+fJx49eiS+vr5iYWEhwcHBughZ50JCQuSLL76QmjVrStmyZaVGjRqybds2iY2NFY1GI2vWrBEbGxvx9vZWnvP2kgi8S5+1Dh06SI0aNWTDhg0i8qaboqWlpRw8eFBEMh+3AwcOsIL5H23YsEGGDh0qtra2vBFFOscEiXQqMTFRqlSpInXr1lXa/nrxs2/fPrG2tlZmYsrAWZbeCA0NFRsbG+nbt68cPnxY6a5w69YtKVWqlGzdulVSU1NlypQpUrx4ca3p0Dll6t/jeALd6dChg3Tr1k2WLFkiZcqUkaZNm8qMGTMydcf59ttvpWnTpuLi4lJgp6UODg4WCwsL8fLykoULF8q0adOkUaNGoqenJ8OHD5fHjx8rSZKtra18/fXXug45VwsLC5OlS5cqSx14enpK9erVpUuXLmJhYaF0U/zr9xS7KP53oaGh0qhRI+nQoYNcv35d1+EQMUEi3QsMDJTSpUtLixYtlLaMJCnjC2jx4sViaWkpUVFRWXatKag0Go34+vqKSqUSCwsLGThwoDg7O8uePXvk5cuXsnHjRmnSpInExsbKw4cPZfr06WJkZCTz58/Xdeh5wtsXQLzDnjMyktL58+fL4MGDRUQkOjpaZs2aJZ6enmJnZydz586VwMBAEXlzx37FihUSFhamo4h1KzIyUlxcXMTX11erPS0tTZmsJWMMYkxMjKxbt07UarUMGzZMB9HmfsHBweLk5CQdOnSQPXv2KO3du3cXlUolvr6+ymc045wwbtw4KVGiBBcw/R89efIky94QRLrABIl0TqPRyIkTJ6R48eKZkqSML6JNmzZJ7dq1lfUn6E/BwcHi4+MjRYsWlR9//FFWrlwptWrVkqZNm8pnn30mtWvXVu6s379/X+bMmcOuYpQrvX1TJDQ0VMzNzWXVqlXK9vbt24uZmZm0bNlSbG1tpWvXrnL48OECXenbt2+f1K5dWyIiIkRE+7wpIjJw4EApUqSIMkNlXFycBAQEyI0bN3QSb24WEhIiFhYWMnbsWHn48GGm7T169JCKFSvKunXrJCEhQUREJk6cKIaGhnLu3LmcDpeIPiAmSJTjwsPDJSAgQJYsWSKPHj0SkTcDrt9Okt7uGy8iMnr0aGndujVnB3qHq1evypdffikWFhZy5coViYuLkz179kjVqlVFpVLJzz//rOybkpKiw0iJMrtz546cPXtWRN4kSRmf0Tlz5kjv3r1FRMTLy0scHBzkxo0bEhkZKT///LO4uroq3aAKqgULFoijo6PEx8drtWckmteuXZOiRYvKihUrMm2jP7169Uo6d+4sgwYN0mrXaDQSFhamdJ/76quvxMnJSbZv3y5jx44VQ0NDTnJDlA9xoVjKUcHBwWjfvj2KFi2K58+fY86cOTh9+jTs7e3h5uaGLVu2oHv37mjYsCHmzJmDtLQ0HDhwAKtWrcKxY8dQpEgRXb8FnYuJicGzZ89w//59mJiY4KOPPkLlypXxzTffICkpCQ0bNsT27dvRpk0bNGzYEJGRkahYsSJEBCqVCnp6erp+C0SKhw8fonz58jAyMsLevXvh4eGhLNBbrVo1rF27Fm5ubnj06BF2794NJycnAEDbtm3RqlWrAvl5vnfvHiwsLGBubg4LCws8ffoU4eHhqFy5MtLS0lCoUCFlUc1KlSpBrVYjOjpaeT4X3MxMrVbj8ePH8PDwUNoOHDiA/fv3Y/Xq1TAzM8NHH32EHTt2wNvbG56enihSpAhOnjzJBaOJ8iNdZ2hUcFy+fFmMjIxk3LhxEhUVJRcuXJCKFSvK7t27tfa7deuWuLq6SokSJcTFxUUaN24sQUFBOoo6d7ly5YrUrVtXqlWrpkzR3aRJEzl06JCkpaVJRESEfPHFF2JhYSEHDhwQkczTpxPlJmfOnJGSJUuKq6urlC5dWvbv36+1vU+fPmJoaJhpSmWRglkJ0Wg00rhxY7G1tZXo6GiJiIiQ0qVLS4cOHZSFXjMq8MnJyfLo0SNxc3PLdFxJW1xcnFSsWFG8vb0lJCREZs6cKc7OztKpUydZtGiRrFq1SkqVKiVTpkwREZFBgwbJlStXdBw1EX0oTJAoR9y4cUNZ9f5t9erVk5EjR0qPHj0kICBAwsPDlW2XL1+Wu3fvcirvdFevXhUzMzPx8fGRM2fOyJUrV2TdunVStmxZcXR0lG3btonIm/VjvLy8xMbGRvbt26fjqIn+3suXL6VRo0bSvHlzGTlypJQqVUqZQlnkzRgbNzc3ZVKGgjzeKMOVK1ekTp06Uq1aNYmJiZFJkyZJ0aJFpX///vL06VOtfSdOnCjly5fnBALZcPjwYVGr1VKqVCkxNTWVH374QW7duiUib5LO5s2bS/fu3XUcJRHlBHaxoxyxcuVK6Ovro2zZstBoNChcuDBmzZqF8+fPo3jx4oiOjkbPnj0xcuRITJs2DYaGhqhevbquw8414uPjMWrUKHh5eWH+/PlKe5UqVdCsWTM0btwYEydOxEcffYRy5crBz88PL1++xMCBA3H16lUYGRmxWw3lOqmpqShSpAimTZuGkSNHokqVKnj+/Dm8vb2xcuVKNGvWDPXr14dGo8HixYvRqFEjpftdQSTp3WRdXFywfv16eHl5oX379vj5558RGxuL9evX4+jRoxg0aBBiY2Px4MEDbN26FUeOHIGjo6Ouw8/1PvnkE9y9exdPnz5FqVKlYGVlpWzT09ODubk5ypYtCxEBwK6KRPmarjM0KhgSEhKkX79+8vHHH8uaNWtk2rRpYm1tLfv371em7R4/fryo1eoCP+g6K5GRkeLk5CQ7d+4UEe0FM0XezL701wrdnTt3lEkwiHKLu3fvKuvIvN3WunVr2bJli9y/f18+//xzKVWqlNIt7KeffpLixYvL8+fPC2S3usTEROX/357AZuTIkaJSqcTd3V1iY2Nl8+bN0rJlS3FwcJAaNWqIl5eXXLt2TRch5ytJSUkyYcIEcXBw4AygRAUEEyTKMQkJCeLl5SXly5cXIyMj+fXXX0Xkzy/8n3/+WcqVK6d0aaA/nTp1SkxMTJRxGG+vBZWSkiIpKSnSpk0bad++PWepo1zr/v37YmBgICqVSsaOHSsLFixQtn3//fdSpkwZiY+Pl5CQEOnVq5eUL19e9u/fLykpKQU22X/w4IF4enrKkSNHtNpnz54tlpaW8uOPP0qNGjXE1dVVYmJiRETk4cOHkpaWJklJSTqIOH/ZsGGDDB06VGxtbeXixYu6DoeIckjB7atAH9SjR49w4MABBAQEIDExEQBgbGyMJUuWoEmTJihXrhzu3buH169fQ19fHwBw7NgxWFlZwdLSUpeh5xry5gYGgDczURkbG2PNmjUA3sy4lJaWBuBN1w89PT2YmJggMTGxQM7qRXlDREQE6tatCwMDA8TFxeHo0aOoXLkyVq5ciTp16sDDwwMHDhxAxYoVlS53o0ePRnJyMuzt7XUdvk4kJSXhwYMHmDdvHk6ePAkAmDVrFmbPno3Nmzejb9++2LhxI1JTU+Hh4YHo6Gg4ODhApVIp51b6b27cuIFVq1bh/v37CAwMRM2aNXUdEhHlEJVkXIERvSdXrlxB165dAQD3799HiRIlcPz4cSXxefXqFQYNGoSQkBB8/vnnGDZsGGbMmAF/f3+cOnUK1apV02X4ucLNmzfx/fff4+HDh6hXrx5Gjx6N/v3745dffsHMmTPh5eWl7JuWlgYRgaenJ6pXr45JkybpLnCiv5GWloY//vgD/v7+ePjwIY4cOYJt27bhjz/+QGBgIJ49e4b27dtj586dAICrV6+iWLFicHBw0HHkunXr1i0MHToUBgYGsLGxwe7du/HTTz+hefPmyj6hoaFo2bIlbG1tcfLkyQI9Vut9evr0KQwMDGBubq7rUIgoBzFBovcqKCgI9erVw7Bhw/Dll1/i+vXr6NChAzw9PbFlyxakpKRArVYrSdLdu3chIjh37hxOnDgBV1dXXb8FnQsKClIGpxsaGmL79u1YunQpmjZtilatWiEtLQ39+/fHiBEjALy5wzxz5kwsW7YMJ06cUNaJIcqNUlNTceLECfj6+qJQoUI4fPgwjI2NsXv3bqxevRqenp7o1auXrsPMdW7evInBgwfjxIkTyqQWAJR1jzL20dfXR5kyZXQZKhFRnscEid6bsLAwVKxYEaNHj8b06dMBACkpKahSpQqKFy+Ow4cPa+2fkJCAvn374sSJE/jtt984ax3eLKRbt25djBgxAjNmzEBqaiqGDRuGQoUK4bvvvsOlS5cwePBg3LhxA46OjrCwsECRIkVw6dIl/PLLL1ywkHKVu3fvYvv27QgLC0Pbtm1Rp04dWFlZIS0tDSdPnsSoUaOQlJSEI0eOoFixYnjx4gXMzMx0HXaudefOHQwcOBB6enoYN24cGjRoAEA7SSIiov8dz6j0XogIfv/9d1hZWSE2NlZpnz9/Pm7evIm7d+9i2LBh6NWrF86dO4f79++jSJEiWLduHc6fP8/kCG+6IzZp0gStW7fGjBkzALwZX/Ts2TMcO3YMTk5O+Pbbb9G8eXN8//33qFWrFkqUKIGmTZvi+PHjTI4oVwkKCoK7uzsOHDiAwMBAtGvXDsuXL0dSUhIKFSqEBg0aYN68eTAyMkKjRo0QGxsLMzMzpKam6jr0XKtcuXJYvHgxRATTp09XxiQxOSIier9YQaL35sWLF/jpp5+wYsUKeHh4wMHBAfPmzcOMGTNQvXp1XLlyBXv27MH169fx9OlTDBkyREkECLh37x66dOkCe3t7+Pr6on79+pg1axamTZuGsWPHws7ODvPnz4darca2bdtQqVIlXYdMlKXg4GC4ubnBx8cHfn5+MDIyUtaYCQ4OVqpEaWlpOHHiBMaPH4/w8HBcuXKFYz2y4datW/Dx8cGzZ8+wcOFC1K1bV9chERHlK0yQ6L1ITU2Fnp4e4uPjsXbtWqxYsQLXrl3DgQMH0KxZM619L1y4gPPnz6NBgwaoXLmyjiLOnTIGYxcuXBg2NjbYs2cPNmzYoAzGDg8PR5kyZbBkyRIMGDAAwJ+LRxLlBpGRkShevDi8vLywevVqpb1Lly749ddfERwcjNKlS0OtfrNOeWpqKo4fP45p06Zh5cqVKFu2rK5Cz1NCQ0MxceJEzJ8/HyVLltR1OERE+QoTJPrPEhISoFKpYGxsDODPC/WXL18qSVK9evWwfPlyAG8mEzAwMNBlyHlCVoOxRQQpKSl4+vQpWrVqhQkTJqBz5866DpUok5cvX6JFixZ4/vw5tm/fjipVqmDu3LkYM2YMbGxs0LhxY5w4cQIdO3ZEkyZNUKdOHdjb2+P169cwNDTUdfh5ikajQeHChXUdBhFRvsOOy/Sf3L59G66urhg2bBi2bdsGAEoVw8TEBF988QW8vb1x5swZ9OvXDwBgYGDA8QXZ4OTkhGXLlsHd3R2HDx/G8ePHlTVNli9fjhcvXuDjjz/WdZhEWu7fv4/Lly/DxMQEBw4cgIODAzp16oThw4dj3rx52LdvH4KCgrB69Wr4+vrixYsXaN++PT7//HPExsYyOfoPmBwREX0YrCDRf7Js2TKMGTMG8+bNw5gxY9ChQwc4OztjzJgxSiUpNjYWGzZswNq1a1G+fHls2bJF12HnKRnd7UQE/v7+OHToECZNmoRTp05xwULKVS5dugQPDw+sXbsWHTt2BPCmktS1a1fs27cPy5YtQ//+/TM9LyQkBMbGxihVqlROh0xERPROTJDoP4mNjUXNmjXx7bffwtXVFWvWrMGvv/6K1NRU9OzZE5999hmcnZ2RmJiIZcuWYfv27dixYwfs7e11HXqekjEY++zZs4iJicHp06e5VhTlKkFBQWjQoAEGDhyI2bNna2178eIFPD09cfv2bezatQvVqlVTbqBwamoiIsqtmCDRv5acnAx9fX0sWLAAV69exapVq5TudRYWFjA1NUVsbCx8fX1Ru3ZtfPrpp4iNjUXRokV1G3gedePGDfj6+mLmzJmc1IJylYx1u3x8fJS1zwDg/PnzqFChAszNzREfH4927dohLCwMe/bsQdWqVXUYMRER0T9jgkTZkvExeXu2tGPHjqFLly44cOAAatSogX79+uG3337Dnj17cPnyZSxatEjZz9LSUidx5xcZSSlRbnH79m3UrFkT3bt3VyZiAYCpU6di2bJlOHv2LEqUKAEAiI+PR6dOnXDmzBmcPn0aLi4uugqbiIjoH6l1HQDlfjdv3sT333+Phw8fon79+hg5ciQAwMPDA127dsXcuXORlJSEEydOYN++fahZsybq1KmDJk2awMTEhMnRe8DkiHKbGzduIDExEebm5rh9+zbKly+P2bNnY/HixVi3bp2SHAGAqakpdu7ciW7dunFiASIiyvWYINHfCgoKQrNmzVC/fn0YGhrCz88PIoJRo0YBAJo3b47evXvD1tYWgYGBWouXcj0TovyrVatWWLt2LcaMGQO1Wo3k5GSsWbMGmzZtyrT2WUYC9euvv+ooWiIiouxjgkTvFBwcjHr16mHEiBGYMWMGUlNTYWVlhYcPHyIxMRFGRkZo3bo1GjZsiPj4eK3kiIjyn1evXuHVq1cICgpCxYoV0bNnT5ibm+PLL79ETEwMfvzxRyU5ypiMYfLkyTh37hw2b94MU1NTHb8DIiKif8YphChL9+/fR5MmTdC6dWvMmDEDAKCnp4eoqCgEBgaiVq1aaNasGbZt24avv/4aMTEx+P3333UcNRF9KDdv3sSAAQPg7u6O1q1bo1KlSujRowdq1KiB3bt3w9LSEkFBQbh58yaAN+MVJ02ahGnTpmHatGlMjoiIKM9ggkRZSk1NRZkyZZCUlISTJ08CAGbNmoVffvkFnTp1wqhRo/DgwQNMmjQJqampiIyMxL59+8A5P4jyn+DgYDRq1AjGxsYYO3YsLl26hIEDB+L06dNo1qwZ7O3tsX79emzZsgWLFy9GZGQkpk+fjtmzZ+PcuXOoVauWrt8CERFRtnEWO3qnjIVKCxcuDBsbG+zZswcbNmxA8+bNAQDh4eEoU6YMtm3bBrVajfLly3MaaqJ8JqOr7bBhwzB16lSo1X/2zN66dStmzpwJAwMDHDlyBHv37oWvry+MjIwQHh6OP/74g+t2ERFRnsMKEr1ThQoVsGjRIiQmJiIgIAC+vr5o3rw5RATJyclQq9WoWrUq0tLS0K5dOyZHRPlMRlfbVq1aYebMmVCr1RARpKSkAAC6dOmCQYMG4dq1a9i4cSM8PT0xceJEvH79mosaExFRnsUEif6Wk5MTli1bBnd3dxw+fBjHjx+HSqWCvr4+li9fjvj4eNStW1fXYRLRB/B2V9sTJ04AeDO2KCNRAgBvb2+4urpi7969AIAvv/wSwcHBqFatms7iJiIi+l8wQaJ/VK5cOSxevBgighkzZuDSpUuYM2cO5s6dix07dmitd0JE+Ufp0qUREBAAjUaD6dOnK0nSXxUqVAjGxsbKzyYmJjkVIhER0XvHBImypUKFCvjuu++gr6+PTz/9FBMmTMCJEydQs2ZNXYdGRB9Qxr99lUqF6dOnK5O2qFQqpKWl4cGDBzAyMlLGJnJYKxER5XVMkCjbKlSogHnz5qFu3bq4dOkSxxcQFRBvJ0nTpk1TKkmFChXC4sWL8ejRIzRp0gTAm8SJiIgoL+MsdvSvJScnQ19fX9dhEFEOy5jZUkTg7++PQ4cOKQlT9erVdR0eERHRe8EEiYiIsu3WrVvw8fHB2bNnERMTw9nqiIgo32EXOyIiyjZ2tSUiovyOFSQiIvrX2NWWiIjyKyZIRERERERE6djFjoiIiIiIKB0TJCIiIiIionRMkIiIiIiIiNIxQSIiIiIiIkrHBImIiIiIiCgdEyQiIiIiIqJ0TJCIiAhr166FSqVSHmq1Go6OjujTpw8ePnz4wX9/6dKl4eXlpfx89OhRqFQqHD169F+9zqlTpzB58mTExsa+1/gAwMvLC6VLl37vr0tERLkLEyQiIlKsWbMGp0+fxqFDh+Dt7Y1NmzbB3d0dCQkJORpHrVq1cPr0adSqVetfPe/UqVOYMmXKB0mQiIioYFDrOgAiIso9qlSpgtq1awMAGjdujNTUVEybNg27d+9Gjx49Mu3/6tUrGBsbv/c4zMzMULdu3ff+ukRERP+EFSQiInqnjCQlPDwcXl5eMDExwZUrV9C8eXOYmpqiSZMmAACNRoPp06ejYsWKMDAwgLW1Nfr06YOoqCit10tOToavry/s7OxgbGyMBg0a4OzZs5l+77u62J05cwZt2rSBpaUlDA0NUa5cOQwfPhwAMHnyZIwePRoAUKZMGaW74NuvsWXLFtSrVw9FihSBiYkJWrRogUuXLmX6/WvXroWzszMMDAxQqVIlrF+//r8eQiIiymNYQSIione6ffs2AMDa2ho3b96ERqNB27Zt0b9/f4wdOxYpKSlIS0tDu3btcPz4cfj6+sLNzQ3h4eGYNGkSGjVqhPPnz8PIyAgA4O3tjfXr12PUqFFo1qwZrl69io4dOyI+Pv4fYzlw4ADatGmDSpUqYcGCBShZsiTu3buHgwcPAgD69euH6OhofP/999i5cyfs7e0BAC4uLgCAmTNnYsKECejTpw8mTJgAjUaDuXPnwt3dHWfPnlX2W7t2Lfr06YN27dph/vz5iIuLw+TJk5GUlIRChXhfkYgo3xMiIirw1qxZIwDk//7v/yQ5OVni4+Pl119/FWtrazE1NZXHjx9L7969BYCsXr1a67mbNm0SALJjxw6t9nPnzgkAWbp0qYiIhISECAAZMWKE1n4BAQECQHr37q20BQYGCgAJDAxU2sqVKyflypWTxMTEd76PuXPnCgAJCwvTao+IiBC1Wi1DhgzRao+Pjxc7Ozvp0qWLiIikpqaKg4OD1KpVS9LS0pT97t27J/r6+lKqVKl3/m4iIsofeCuMiIgUdevWhb6+PkxNTdG6dWvY2dlh3759sLW1Vfbp1KmT1nN+/fVXFC1aFG3atEFKSoryqFGjBuzs7JQuboGBgQCQaSxTly5doFb/fYeGmzdv4s6dO+jbty8MDQ3/9fs6cOAAUlJS8MUXX2jFaGhoCA8PDyXGGzdu4NGjR+jevTtUKpXy/FKlSsHNze1f/14iIsp72MWOiIgU69evR6VKlaBWq2Fra6t0U8tgbGwMMzMzrbYnT54gNjYWhQsXzvI1nz17BgB4/vw5AMDOzk5ru1qthqWl5d/GlTGWydHRMftv5i8xAkCdOnWy3J7Rde5dMWa03bt37z/9fiIiyjuYIBERkaJSpUrKLHZZebuqksHKygqWlpbYv39/ls8xNTUFACUJevz4MYoXL65sT0lJURKTd7G2tgYAPHjw4O/fwDtYWVkBALZv345SpUq9c7+3Y/yrrNqIiCj/YYJERET/k9atW2Pz5s1ITU3Fxx9//M79GjVqBAAICAiAq6ur0r5161akpKT87e9wcnJCuXLlsHr1avj4+MDAwCDL/TLaExMTtdpbtGgBtVqNO3fuZOoi+DZnZ2fY29tj06ZN8PHxURLC8PBwnDp1Cg4ODn8bJxER5X1MkIiI6H/SrVs3BAQEoGXLlhg2bBg++ugj6Ovr48GDBwgMDES7du3QoUMHVKpUCT179sS3334LfX19NG3aFFevXsW8efMyddvLypIlS9CmTRvUrVsXI0aMQMmSJREREYEDBw4gICAAAFC1alUAwKJFi9C7d2/o6+vD2dkZpUuXxtSpUzF+/HjcvXsXn376KSwsLPDkyROcPXsWRYoUwZQpU1CoUCFMmzYN/fr1Q4cOHeDt7Y3Y2FhMnjw5y253RESU/zBBIiKi/4menh727NmDRYsWYcOGDfD394darYajoyM8PDyUpAUAVq1aBVtbW6xduxbfffcdatSogR07dqBbt27/+HtatGiBP/74A1OnTsXQoUPx+vVrODo6om3btso+jRo1gp+fH9atW4eVK1ciLS0NgYGBSruLiwsWLVqETZs2ISkpCXZ2dqhTpw6+/vpr5TX69u0LAJg9ezY6duyI0qVLY9y4cTh27FimdZmIiCj/UYmI6DoIIiIiIiKi3IDTfBMREREREaVjgkRERERERJSOCRIREREREVE6JkhERERERETpmCARERERERGlY4JERERERESUjgkSERERERFROiZIRERERERE6ZggERERERERpWOCRERERERElI4JEhERERERUTomSEREREREROn+H5m3Vwrx8Sq8AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 900x700 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred_proba  = state.model.predict_proba(X_test_transformed)\n",
    "y_pred        = np.argmax(y_pred_proba, axis=1)\n",
    "y_pred_labels = le.inverse_transform(y_pred)\n",
    "\n",
    "# SDK-level structured metrics: serialisable numbers for metadata and reports.\n",
    "test_report = evaluate_classification(\n",
    "    y_true=y_test_encoded,\n",
    "    y_pred_proba=y_pred_proba,\n",
    "    labels=list(range(len(le.classes_))),\n",
    ")\n",
    "\n",
    "print(f\"Prediction shape : {y_pred_proba.shape}  (one probability per class per row)\")\n",
    "print(\"Structured metric keys:\", sorted(test_report.metrics.keys()))\n",
    "\n",
    "cm = confusion_matrix(y_test_encoded, y_pred)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(9, 7))\n",
    "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues',\n",
    "            xticklabels=le.classes_, yticklabels=le.classes_,\n",
    "            ax=ax, linewidths=0.5)\n",
    "ax.set_xlabel('Predicted', fontsize=12)\n",
    "ax.set_ylabel('Actual', fontsize=12)\n",
    "ax.set_title('Confusion Matrix \\u2014 Dry Bean Variety Classification', fontweight=\"bold\", fontsize=13)\n",
    "ax.tick_params(axis='x', rotation=45)\n",
    "ax.tick_params(axis='y', rotation=0)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "cdd5443966",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Per-Class Classification Report\n",
      "=================================================================\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    BARBUNYA       0.96      0.90      0.93       265\n",
      "      BOMBAY       1.00      1.00      1.00       104\n",
      "        CALI       0.94      0.94      0.94       326\n",
      "    DERMASON       0.90      0.93      0.91       709\n",
      "       HOROZ       0.97      0.96      0.96       386\n",
      "       SEKER       0.94      0.96      0.95       406\n",
      "        SIRA       0.88      0.87      0.87       527\n",
      "\n",
      "    accuracy                           0.93      2723\n",
      "   macro avg       0.94      0.94      0.94      2723\n",
      "weighted avg       0.93      0.93      0.93      2723\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(\"Per-Class Classification Report\")\n",
    "print(\"=\" * 65)\n",
    "print(classification_report(y_test_encoded, y_pred, target_names=le.classes_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "cddbc30265",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=======================================================\n",
      "  TEST SET — AGGREGATE METRICS\n",
      "=======================================================\n",
      "  Accuracy             : 0.9266\n",
      "  Balanced Accuracy    : 0.9364  ← accounts for class imbalance\n",
      "  Macro F1             : 0.9388  ← treats each variety equally\n",
      "  Weighted F1          : 0.9265\n",
      "  ROC AUC (OvR macro)  : 0.9952\n",
      "=======================================================\n",
      "\n",
      "  Majority-class baseline accuracy : 0.2605\n",
      "  Improvement over baseline        : +0.6660\n"
     ]
    }
   ],
   "source": [
    "accuracy     = accuracy_score(y_test_encoded, y_pred)\n",
    "balanced_acc = balanced_accuracy_score(y_test_encoded, y_pred)\n",
    "macro_f1     = f1_score(y_test_encoded, y_pred, average='macro')\n",
    "weighted_f1  = f1_score(y_test_encoded, y_pred, average='weighted')\n",
    "roc_auc      = roc_auc_score(y_test_encoded, y_pred_proba, multi_class='ovr', average='macro')\n",
    "\n",
    "print(\"=\" * 55)\n",
    "print(\"  TEST SET \\u2014 AGGREGATE METRICS\")\n",
    "print(\"=\" * 55)\n",
    "print(f\"  Accuracy             : {accuracy:.4f}\")\n",
    "print(f\"  Balanced Accuracy    : {balanced_acc:.4f}  \\u2190 accounts for class imbalance\")\n",
    "print(f\"  Macro F1             : {macro_f1:.4f}  \\u2190 treats each variety equally\")\n",
    "print(f\"  Weighted F1          : {weighted_f1:.4f}\")\n",
    "print(f\"  ROC AUC (OvR macro)  : {roc_auc:.4f}\")\n",
    "print(\"=\" * 55)\n",
    "print(f\"\\n  Majority-class baseline accuracy : {majority_acc:.4f}\")\n",
    "print(f\"  Improvement over baseline        : +{accuracy - majority_acc:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mde06527d9",
   "metadata": {},
   "source": [
    "### 9.1 Per-Class ROC AUC (One-vs-Rest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cd2c10965e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Per-Class ROC AUC (One-vs-Rest):\n",
      "-----------------------------------\n",
      "  BARBUNYA   AUC = 0.9944  █████████████████████████████\n",
      "  BOMBAY     AUC = 1.0000  ██████████████████████████████\n",
      "  CALI       AUC = 0.9963  █████████████████████████████\n",
      "  DERMASON   AUC = 0.9928  █████████████████████████████\n",
      "  HOROZ      AUC = 0.9986  █████████████████████████████\n",
      "  SEKER      AUC = 0.9974  █████████████████████████████\n",
      "  SIRA       AUC = 0.9867  █████████████████████████████\n",
      "  Macro      AUC = 0.9952\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_test_bin = label_binarize(y_test_encoded, classes=list(range(len(le.classes_))))\n",
    "\n",
    "print(\"Per-Class ROC AUC (One-vs-Rest):\")\n",
    "print(\"-\" * 35)\n",
    "per_class_aucs = {}\n",
    "for i, class_name in enumerate(le.classes_):\n",
    "    auc_i = roc_auc_score(y_test_bin[:, i], y_pred_proba[:, i])\n",
    "    per_class_aucs[class_name] = auc_i\n",
    "    bar = '\\u2588' * int(auc_i * 30)\n",
    "    print(f\"  {class_name:<10} AUC = {auc_i:.4f}  {bar}\")\n",
    "print(f\"  {'Macro':<10} AUC = {roc_auc:.4f}\")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(9, 4))\n",
    "auc_values  = [per_class_aucs[c] for c in le.classes_]\n",
    "bar_colors  = [BEAN_COLORS.get(c, '#6B7280') for c in le.classes_]\n",
    "bars = ax.bar(list(le.classes_), auc_values, color=bar_colors, edgecolor='#555555', linewidth=0.8)\n",
    "ax.axhline(roc_auc, color='#DC2626', ls='--', lw=1.5, label=f'Macro avg = {roc_auc:.4f}')\n",
    "ax.set_ylim(0.5, 1.05)\n",
    "ax.set_ylabel(\"ROC AUC (OvR)\")\n",
    "ax.set_title(\"Per-Class ROC AUC \\u2014 One-vs-Rest\", fontweight=\"bold\", fontsize=13)\n",
    "ax.legend(fontsize=10)\n",
    "ax.grid(axis='y', alpha=0.3)\n",
    "ax.tick_params(axis='x', rotation=30)\n",
    "for bar, auc_val in zip(bars, auc_values):\n",
    "    ax.text(bar.get_x() + bar.get_width() / 2, bar.get_height() + 0.004,\n",
    "            f\"{auc_val:.4f}\", ha='center', va='bottom', fontsize=8)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mdf4ca751b",
   "metadata": {},
   "source": [
    "## 10. SHAP Explainability — Which Shape Features Drive Variety?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "cd988652b7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Feature ranking: tree importance vs SHAP (sorted by SHAP rank):\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>tree_rank</th>\n",
       "      <th>shap_rank</th>\n",
       "      <th>rank_delta</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ShapeFactor1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ConvexArea</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Perimeter</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>roundness</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>MinorAxisLength</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Area</td>\n",
       "      <td>13</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ShapeFactor4</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ShapeFactor3</td>\n",
       "      <td>11</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MajorAxisLength</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Compactness</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Extent</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>ShapeFactor2</td>\n",
       "      <td>14</td>\n",
       "      <td>12</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            feature  tree_rank  shap_rank  rank_delta\n",
       "3      ShapeFactor1          4          1           3\n",
       "9        ConvexArea         10          2           8\n",
       "6         Perimeter          7          3           4\n",
       "0         roundness          1          4           3\n",
       "5   MinorAxisLength          6          5           1\n",
       "12             Area         13          6           7\n",
       "1      ShapeFactor4          2          7           5\n",
       "10     ShapeFactor3         11          8           3\n",
       "7   MajorAxisLength          8          9           1\n",
       "8       Compactness          9         10           1\n",
       "4            Extent          5         11           6\n",
       "13     ShapeFactor2         14         12           2"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "if state.shap_feature_importance is not None:\n",
    "    shap_importance = state.shap_feature_importance\n",
    "    top_n_shap = min(16, len(shap_importance))\n",
    "    top_shap   = shap_importance.head(top_n_shap)\n",
    "\n",
    "    fig, ax = plt.subplots(figsize=(10, 6))\n",
    "    ax.barh(top_shap['feature'][::-1], top_shap['importance'][::-1],\n",
    "            color=\"#10B981\", edgecolor='white', linewidth=0.5)\n",
    "    ax.set_title(f\"SHAP Global Feature Importance (top {top_n_shap})\", fontweight=\"bold\", fontsize=13)\n",
    "    ax.set_xlabel(\"Mean |SHAP value| (aggregated across all classes)\")\n",
    "    ax.grid(axis='x', alpha=0.3)\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "\n",
    "    if state.feature_importance is not None:\n",
    "        tree_rank = state.feature_importance.assign(\n",
    "            tree_rank=range(1, len(state.feature_importance) + 1)\n",
    "        ).rename(columns={'importance': 'tree_importance'})\n",
    "        shap_rank = shap_importance.assign(\n",
    "            shap_rank=range(1, len(shap_importance) + 1)\n",
    "        ).rename(columns={'importance': 'shap_importance'})\n",
    "        comparison = (\n",
    "            tree_rank.merge(shap_rank, on='feature')\n",
    "            .assign(rank_delta=lambda d: (d['tree_rank'] - d['shap_rank']).abs())\n",
    "            .sort_values('shap_rank').head(12)\n",
    "        )\n",
    "        print(\"\\nFeature ranking: tree importance vs SHAP (sorted by SHAP rank):\")\n",
    "        display(comparison[['feature', 'tree_rank', 'shap_rank', 'rank_delta']])\n",
    "else:\n",
    "    print(\"SHAP importance not available.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "md914778ae",
   "metadata": {},
   "source": [
    "## 11. Packaging for Production"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "cd21f855a1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: dry_bean_classifier_v1 (v1.0.0)\n",
      "Type: LGBMClassifier (lightgbm)\n",
      "Task: multiclass_classification\n",
      "Features: 16\n",
      "Trained: 2026-05-05 19:59\n",
      "\n",
      "Performance Metrics:\n",
      "  test_roc_auc_macro_ovr: 0.9952\n",
      "  test_macro_f1: 0.9388\n",
      "  test_weighted_f1: 0.9265\n",
      "  test_accuracy: 0.9266\n",
      "  test_balanced_accuracy: 0.9364\n",
      "  test_roc_auc_barbunya: 0.9944\n",
      "  test_roc_auc_bombay: 1.0000\n",
      "  test_roc_auc_cali: 0.9963\n",
      "  test_roc_auc_dermason: 0.9928\n",
      "  test_roc_auc_horoz: 0.9986\n",
      "  test_roc_auc_seker: 0.9974\n",
      "  test_roc_auc_sira: 0.9867\n",
      "\n",
      "Cross-Validation:\n",
      "  roc_auc: 0.9957 ± 0.0007\n",
      "\n",
      "Training Data: 10888 samples\n",
      "Test Data: 2723 samples\n"
     ]
    }
   ],
   "source": [
    "metadata = ModelMetadata(\n",
    "    name=\"dry_bean_classifier_v1\",\n",
    "    version=\"1.0.0\",\n",
    "    model_type=state.model.model_type,\n",
    "    framework=state.model.framework,\n",
    "    task=\"multiclass_classification\",\n",
    "    trained_at=datetime.now(),\n",
    "    feature_names=list(X_train_transformed.columns),\n",
    "    n_features=X_train_transformed.shape[1],\n",
    "    hyperparameters=state.best_params,\n",
    "    training_time_seconds=state.training_time_seconds,\n",
    "    tags=[\"multiclass\", \"dry_bean\", \"morphology\", \"lightgbm\", \"uci\"],\n",
    "    notes=\"7-class dry bean variety classifier. Stratified 80/20 split. Standard scaling.\",\n",
    "    classes=list(le.classes_)\n",
    ")\n",
    "\n",
    "metadata.add_cv_scores(\"roc_auc\", fold_scores if fold_scores else [state.best_score])\n",
    "metadata.add_metric(\"test_roc_auc_macro_ovr\", roc_auc)\n",
    "metadata.add_metric(\"test_macro_f1\",          macro_f1)\n",
    "metadata.add_metric(\"test_weighted_f1\",       weighted_f1)\n",
    "metadata.add_metric(\"test_accuracy\",          accuracy)\n",
    "metadata.add_metric(\"test_balanced_accuracy\", balanced_acc)\n",
    "for class_name, auc_val in per_class_aucs.items():\n",
    "    metadata.add_metric(f\"test_roc_auc_{class_name.lower()}\", auc_val)\n",
    "metadata.add_dataset(X_train_transformed, y_train, \"train\")\n",
    "metadata.add_dataset(X_test_transformed,  y_test,  \"test\")\n",
    "\n",
    "print(metadata.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "cdf61d017d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model package → models/dry_bean_lgbm_v1.pkl\n"
     ]
    }
   ],
   "source": [
    "model_path = \"models/dry_bean_lgbm_v1.pkl\"\n",
    "os.makedirs(\"models\", exist_ok=True)\n",
    "\n",
    "ModelSerializer.save(\n",
    "    model=state.model,\n",
    "    path=model_path,\n",
    "    metadata=metadata,\n",
    "    train_data=(X_train_transformed, y_train_encoded),\n",
    "    test_data=(X_test_transformed,   y_test_encoded),\n",
    "    include_datasets=True\n",
    ")\n",
    "print(f\"Model package \\u2192 {model_path}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "cd1ce4489a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] feature_fraction is set=0.6104217815601655, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.6104217815601655\n",
      "[LightGBM] [Warning] lambda_l2 is set=1.1182701662877744e-07, reg_lambda=0.0 will be ignored. Current value: lambda_l2=1.1182701662877744e-07\n",
      "[LightGBM] [Warning] min_gain_to_split is set=0.21527568422880633, min_split_gain=0.0 will be ignored. Current value: min_gain_to_split=0.21527568422880633\n",
      "[LightGBM] [Warning] lambda_l1 is set=1.2158568318093246e-06, reg_alpha=0.0 will be ignored. Current value: lambda_l1=1.2158568318093246e-06\n",
      "[LightGBM] [Warning] bagging_fraction is set=0.8822936933674879, subsample=1.0 will be ignored. Current value: bagging_fraction=0.8822936933674879\n",
      "[LightGBM] [Warning] bagging_freq is set=3, subsample_freq=0 will be ignored. Current value: bagging_freq=3\n",
      "=================================================================\n",
      "  INFERENCE RESULTS (6 sample beans)\n",
      "=================================================================\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Perimeter</th>\n",
       "      <th>MajorAxisLength</th>\n",
       "      <th>Compactness</th>\n",
       "      <th>predicted_variety</th>\n",
       "      <th>confidence</th>\n",
       "      <th>ground_truth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12803</th>\n",
       "      <td>36002</td>\n",
       "      <td>700.889</td>\n",
       "      <td>257.888986</td>\n",
       "      <td>0.830205</td>\n",
       "      <td>DERMASON</td>\n",
       "      <td>0.8596</td>\n",
       "      <td>DERMASON</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3713</th>\n",
       "      <td>183825</td>\n",
       "      <td>1594.423</td>\n",
       "      <td>582.225240</td>\n",
       "      <td>0.830934</td>\n",
       "      <td>BOMBAY</td>\n",
       "      <td>0.9997</td>\n",
       "      <td>BOMBAY</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8974</th>\n",
       "      <td>45595</td>\n",
       "      <td>801.717</td>\n",
       "      <td>301.092828</td>\n",
       "      <td>0.800227</td>\n",
       "      <td>SIRA</td>\n",
       "      <td>0.9897</td>\n",
       "      <td>SIRA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1743</th>\n",
       "      <td>45084</td>\n",
       "      <td>763.811</td>\n",
       "      <td>261.138204</td>\n",
       "      <td>0.917478</td>\n",
       "      <td>SEKER</td>\n",
       "      <td>0.9997</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2303</th>\n",
       "      <td>61645</td>\n",
       "      <td>952.199</td>\n",
       "      <td>335.139136</td>\n",
       "      <td>0.835947</td>\n",
       "      <td>BARBUNYA</td>\n",
       "      <td>0.9065</td>\n",
       "      <td>BARBUNYA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>34321</td>\n",
       "      <td>669.531</td>\n",
       "      <td>227.995845</td>\n",
       "      <td>0.916871</td>\n",
       "      <td>SEKER</td>\n",
       "      <td>0.9990</td>\n",
       "      <td>SEKER</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Area  Perimeter  MajorAxisLength  Compactness predicted_variety  \\\n",
       "12803   36002    700.889       257.888986     0.830205          DERMASON   \n",
       "3713   183825   1594.423       582.225240     0.830934            BOMBAY   \n",
       "8974    45595    801.717       301.092828     0.800227              SIRA   \n",
       "1743    45084    763.811       261.138204     0.917478             SEKER   \n",
       "2303    61645    952.199       335.139136     0.835947          BARBUNYA   \n",
       "192     34321    669.531       227.995845     0.916871             SEKER   \n",
       "\n",
       "       confidence ground_truth  \n",
       "12803      0.8596     DERMASON  \n",
       "3713       0.9997       BOMBAY  \n",
       "8974       0.9897         SIRA  \n",
       "1743       0.9997        SEKER  \n",
       "2303       0.9065     BARBUNYA  \n",
       "192        0.9990        SEKER  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Inference complete: two artefacts (pipeline + model), zero leakage, full reproducibility.\n"
     ]
    }
   ],
   "source": [
    "package           = ModelSerializer.load(model_path)\n",
    "inference_model   = package.model\n",
    "inference_pipeline = TransformPipeline.load(pipeline_path)\n",
    "\n",
    "saved_classes   = package.metadata.classes\n",
    "inference_le    = LabelEncoder()\n",
    "inference_le.fit(saved_classes)\n",
    "\n",
    "new_beans_raw         = X_test.iloc[:6].copy()\n",
    "new_beans_transformed = inference_pipeline.transform(new_beans_raw)\n",
    "new_proba             = inference_model.predict_proba(new_beans_transformed)\n",
    "new_pred_labels       = inference_le.inverse_transform(np.argmax(new_proba, axis=1))\n",
    "\n",
    "results_df = new_beans_raw[['Area', 'Perimeter', 'MajorAxisLength', 'Compactness']].assign(\n",
    "    predicted_variety=new_pred_labels,\n",
    "    confidence=new_proba.max(axis=1).round(4),\n",
    "    ground_truth=y_test.iloc[:6].values\n",
    ")\n",
    "\n",
    "print(\"=\" * 65)\n",
    "print(\"  INFERENCE RESULTS (6 sample beans)\")\n",
    "print(\"=\" * 65)\n",
    "display(results_df)\n",
    "print(\"\\nInference complete: two artefacts (pipeline + model), zero leakage, full reproducibility.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mddc5ca6bc",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## What You Built\n",
    "\n",
    "| You Wrote | BitBullet Handled |\n",
    "|-----------|-------------------|\n",
    "| `generate_feature_stats(X)` | Full statistical audit — all 16 numerical features in one call |\n",
    "| `LabelEncoder().fit(y)` before splitting | Invertible integer coding covering all 7 classes |\n",
    "| `train_test_split(stratify=y)` | All seven class proportions preserved |\n",
    "| Declared `standard_scale` in `pipeline.add()` | Scaling fit on train only, applied consistently to test |\n",
    "| Set `task=\"multiclass_classification\"` in `TrainConfig` | Softmax objective, multiclass CV scoring, correct SHAP aggregation |\n",
    "| `trainer.fit(X_train_transformed, y_train_encoded)` | Optuna TPE, Stratified K-Fold CV, early stopping per fold |\n",
    "| `np.argmax(predict_proba(...), axis=1)` | Argmax decoding from seven-class softmax probability vectors |\n",
    "| `ModelMetadata(classes=list(le.classes_))` | Class label map plus target mapping embedded — inference can decode predictions without a separate notebook variable |\n",
    "\n",
    "**Continue the Academy:**\n",
    "- `03_Clustering.ipynb` — unsupervised segmentation with `bitbullet.cluster`\n",
    "- `04_Data_Transformations.ipynb` — deep dive into every transform in the library\n",
    "- `05_Model_Management.ipynb` — versioning, comparison, and model governance"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "bitbullet",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}