DRL_PROJ/classifier/notebooks/07_data_scaling.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1343dfe0",
   "metadata": {},
   "source": [
    "# 07 - Phase 4 Data-Scaling Analysis\n",
    "\n",
    "Phase 4 asks the next question after the Phase 3 model-family sweep: once the strongest backbones have been identified at the 20% data setting, what happens when they see more training data?\n",
    "\n",
    "The experiment keeps the chosen preprocessing recipe fixed: 224x224 input, facecropped classifier data, ImageNet/default normalization, pretrained backbones, and no augmentation. The variable is data scale. Phase 3 provides the 20% anchor, and Phase 4 adds 50% and 100% runs for the three strongest model families: ResNet50, EfficientNet-B0, and ConvNeXt-Tiny.\n",
    "\n",
    "The story question is not only whether AUC increases. We also check whether more data improves thresholded accuracy/F1, fold stability, confusion-matrix balance, and source-wise behavior.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3a0420eb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Project root: c:\\Users\\diogo\\Documents\\MIA_UP\\2_Semestre\\DRL\\DRL_2\\DRL_PROJ\n"
     ]
    }
   ],
   "source": [
    "from __future__ import annotations\n",
    "\n",
    "import json\n",
    "import sys\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "\n",
    "def find_project_root(start: Path | None = None) -> Path:\n",
    "    \"\"\"Find DRL_PROJ whether the notebook runs from repo root or classifier/notebooks.\"\"\"\n",
    "    start = Path.cwd() if start is None else Path(start)\n",
    "    for candidate in [start, *start.parents]:\n",
    "        if (candidate / \"classifier\").is_dir() and (candidate / \"docs\" / \"DRL_Project.md\").exists():\n",
    "            return candidate\n",
    "    raise RuntimeError(\"Could not find DRL_PROJ root. Run this notebook from inside the repository.\")\n",
    "\n",
    "\n",
    "PROJECT_ROOT = find_project_root()\n",
    "CLASSIFIER_ROOT = PROJECT_ROOT / \"classifier\"\n",
    "if str(CLASSIFIER_ROOT) not in sys.path:\n",
    "    sys.path.insert(0, str(CLASSIFIER_ROOT))\n",
    "\n",
    "CONFIGS_DIR = CLASSIFIER_ROOT / \"configs\"\n",
    "LOGS_DIR = CLASSIFIER_ROOT / \"outputs\" / \"logs\"\n",
    "MODELS_DIR = CLASSIFIER_ROOT / \"outputs\" / \"models\"\n",
    "FIGURES_DIR = CLASSIFIER_ROOT / \"outputs\" / \"figures\"\n",
    "ANALYSIS_DIR = CLASSIFIER_ROOT / \"outputs\" / \"analysis\"\n",
    "FIGURES_DIR.mkdir(parents=True, exist_ok=True)\n",
    "ANALYSIS_DIR.mkdir(parents=True, exist_ok=True)\n",
    "\n",
    "print(f\"Project root: {PROJECT_ROOT}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bdb38a22",
   "metadata": {},
   "source": [
    "## 1. Evidence and experiment matrix\n",
    "\n",
    "Phase 4 does not introduce a new preprocessing trick. It deliberately holds the Phase 3 recipe fixed and changes only the amount of data used for training.\n",
    "\n",
    "The table below combines the 20% Phase 3 anchor runs with the 50% and 100% Phase 4 runs. This makes the scaling story explicit: 20% tells us which families were promising, 50% tests whether gains continue, and 100% tests the best available setting in this experiment set.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ce0e740b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_52656\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_52656_level0_col0\" class=\"col_heading level0 col0\" >model</th>\n",
       "      <th id=\"T_52656_level0_col1\" class=\"col_heading level0 col1\" >scale_pct</th>\n",
       "      <th id=\"T_52656_level0_col2\" class=\"col_heading level0 col2\" >run</th>\n",
       "      <th id=\"T_52656_level0_col3\" class=\"col_heading level0 col3\" >backbone</th>\n",
       "      <th id=\"T_52656_level0_col4\" class=\"col_heading level0 col4\" >auc</th>\n",
       "      <th id=\"T_52656_level0_col5\" class=\"col_heading level0 col5\" >auc_std</th>\n",
       "      <th id=\"T_52656_level0_col6\" class=\"col_heading level0 col6\" >accuracy</th>\n",
       "      <th id=\"T_52656_level0_col7\" class=\"col_heading level0 col7\" >f1</th>\n",
       "      <th id=\"T_52656_level0_col8\" class=\"col_heading level0 col8\" >balanced_accuracy</th>\n",
       "      <th id=\"T_52656_level0_col9\" class=\"col_heading level0 col9\" >macro_f1</th>\n",
       "      <th id=\"T_52656_level0_col10\" class=\"col_heading level0 col10\" >checkpoint_mb</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_52656_row0_col0\" class=\"data row0 col0\" >ResNet50</td>\n",
       "      <td id=\"T_52656_row0_col1\" class=\"data row0 col1\" >20</td>\n",
       "      <td id=\"T_52656_row0_col2\" class=\"data row0 col2\" >p3_resnet50</td>\n",
       "      <td id=\"T_52656_row0_col3\" class=\"data row0 col3\" >resnet50</td>\n",
       "      <td id=\"T_52656_row0_col4\" class=\"data row0 col4\" >0.9857</td>\n",
       "      <td id=\"T_52656_row0_col5\" class=\"data row0 col5\" >0.0011</td>\n",
       "      <td id=\"T_52656_row0_col6\" class=\"data row0 col6\" >0.9532</td>\n",
       "      <td id=\"T_52656_row0_col7\" class=\"data row0 col7\" >0.9688</td>\n",
       "      <td id=\"T_52656_row0_col8\" class=\"data row0 col8\" >0.9385</td>\n",
       "      <td id=\"T_52656_row0_col9\" class=\"data row0 col9\" >0.9377</td>\n",
       "      <td id=\"T_52656_row0_col10\" class=\"data row0 col10\" >90.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_52656_row1_col0\" class=\"data row1 col0\" >ResNet50</td>\n",
       "      <td id=\"T_52656_row1_col1\" class=\"data row1 col1\" >50</td>\n",
       "      <td id=\"T_52656_row1_col2\" class=\"data row1 col2\" >p4_resnet50_50pct</td>\n",
       "      <td id=\"T_52656_row1_col3\" class=\"data row1 col3\" >resnet50</td>\n",
       "      <td id=\"T_52656_row1_col4\" class=\"data row1 col4\" >0.9921</td>\n",
       "      <td id=\"T_52656_row1_col5\" class=\"data row1 col5\" >0.0008</td>\n",
       "      <td id=\"T_52656_row1_col6\" class=\"data row1 col6\" >0.9607</td>\n",
       "      <td id=\"T_52656_row1_col7\" class=\"data row1 col7\" >0.9736</td>\n",
       "      <td id=\"T_52656_row1_col8\" class=\"data row1 col8\" >0.9537</td>\n",
       "      <td id=\"T_52656_row1_col9\" class=\"data row1 col9\" >0.9482</td>\n",
       "      <td id=\"T_52656_row1_col10\" class=\"data row1 col10\" >90.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_52656_row2_col0\" class=\"data row2 col0\" >ResNet50</td>\n",
       "      <td id=\"T_52656_row2_col1\" class=\"data row2 col1\" >100</td>\n",
       "      <td id=\"T_52656_row2_col2\" class=\"data row2 col2\" >p4_resnet50_100pct</td>\n",
       "      <td id=\"T_52656_row2_col3\" class=\"data row2 col3\" >resnet50</td>\n",
       "      <td id=\"T_52656_row2_col4\" class=\"data row2 col4\" >0.9950</td>\n",
       "      <td id=\"T_52656_row2_col5\" class=\"data row2 col5\" >0.0004</td>\n",
       "      <td id=\"T_52656_row2_col6\" class=\"data row2 col6\" >0.9692</td>\n",
       "      <td id=\"T_52656_row2_col7\" class=\"data row2 col7\" >0.9794</td>\n",
       "      <td id=\"T_52656_row2_col8\" class=\"data row2 col8\" >0.9638</td>\n",
       "      <td id=\"T_52656_row2_col9\" class=\"data row2 col9\" >0.9593</td>\n",
       "      <td id=\"T_52656_row2_col10\" class=\"data row2 col10\" >90.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_52656_row3_col0\" class=\"data row3 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_52656_row3_col1\" class=\"data row3 col1\" >20</td>\n",
       "      <td id=\"T_52656_row3_col2\" class=\"data row3 col2\" >p3_efficientnet_b0</td>\n",
       "      <td id=\"T_52656_row3_col3\" class=\"data row3 col3\" >efficientnet_b0</td>\n",
       "      <td id=\"T_52656_row3_col4\" class=\"data row3 col4\" >0.9845</td>\n",
       "      <td id=\"T_52656_row3_col5\" class=\"data row3 col5\" >0.0025</td>\n",
       "      <td id=\"T_52656_row3_col6\" class=\"data row3 col6\" >0.9450</td>\n",
       "      <td id=\"T_52656_row3_col7\" class=\"data row3 col7\" >0.9628</td>\n",
       "      <td id=\"T_52656_row3_col8\" class=\"data row3 col8\" >0.9397</td>\n",
       "      <td id=\"T_52656_row3_col9\" class=\"data row3 col9\" >0.9285</td>\n",
       "      <td id=\"T_52656_row3_col10\" class=\"data row3 col10\" >15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_52656_row4_col0\" class=\"data row4 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_52656_row4_col1\" class=\"data row4 col1\" >50</td>\n",
       "      <td id=\"T_52656_row4_col2\" class=\"data row4 col2\" >p4_efficientnet_b0_50pct</td>\n",
       "      <td id=\"T_52656_row4_col3\" class=\"data row4 col3\" >efficientnet_b0</td>\n",
       "      <td id=\"T_52656_row4_col4\" class=\"data row4 col4\" >0.9911</td>\n",
       "      <td id=\"T_52656_row4_col5\" class=\"data row4 col5\" >0.0009</td>\n",
       "      <td id=\"T_52656_row4_col6\" class=\"data row4 col6\" >0.9576</td>\n",
       "      <td id=\"T_52656_row4_col7\" class=\"data row4 col7\" >0.9714</td>\n",
       "      <td id=\"T_52656_row4_col8\" class=\"data row4 col8\" >0.9541</td>\n",
       "      <td id=\"T_52656_row4_col9\" class=\"data row4 col9\" >0.9447</td>\n",
       "      <td id=\"T_52656_row4_col10\" class=\"data row4 col10\" >15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_52656_row5_col0\" class=\"data row5 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_52656_row5_col1\" class=\"data row5 col1\" >100</td>\n",
       "      <td id=\"T_52656_row5_col2\" class=\"data row5 col2\" >p4_efficientnet_b0_100pct</td>\n",
       "      <td id=\"T_52656_row5_col3\" class=\"data row5 col3\" >efficientnet_b0</td>\n",
       "      <td id=\"T_52656_row5_col4\" class=\"data row5 col4\" >0.9949</td>\n",
       "      <td id=\"T_52656_row5_col5\" class=\"data row5 col5\" >0.0008</td>\n",
       "      <td id=\"T_52656_row5_col6\" class=\"data row5 col6\" >0.9684</td>\n",
       "      <td id=\"T_52656_row5_col7\" class=\"data row5 col7\" >0.9787</td>\n",
       "      <td id=\"T_52656_row5_col8\" class=\"data row5 col8\" >0.9657</td>\n",
       "      <td id=\"T_52656_row5_col9\" class=\"data row5 col9\" >0.9584</td>\n",
       "      <td id=\"T_52656_row5_col10\" class=\"data row5 col10\" >15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_52656_row6_col0\" class=\"data row6 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_52656_row6_col1\" class=\"data row6 col1\" >20</td>\n",
       "      <td id=\"T_52656_row6_col2\" class=\"data row6 col2\" >p3_convnext_tiny</td>\n",
       "      <td id=\"T_52656_row6_col3\" class=\"data row6 col3\" >convnext_tiny</td>\n",
       "      <td id=\"T_52656_row6_col4\" class=\"data row6 col4\" >0.9868</td>\n",
       "      <td id=\"T_52656_row6_col5\" class=\"data row6 col5\" >0.0013</td>\n",
       "      <td id=\"T_52656_row6_col6\" class=\"data row6 col6\" >0.9473</td>\n",
       "      <td id=\"T_52656_row6_col7\" class=\"data row6 col7\" >0.9650</td>\n",
       "      <td id=\"T_52656_row6_col8\" class=\"data row6 col8\" >0.9262</td>\n",
       "      <td id=\"T_52656_row6_col9\" class=\"data row6 col9\" >0.9293</td>\n",
       "      <td id=\"T_52656_row6_col10\" class=\"data row6 col10\" >106.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_52656_row7_col0\" class=\"data row7 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_52656_row7_col1\" class=\"data row7 col1\" >50</td>\n",
       "      <td id=\"T_52656_row7_col2\" class=\"data row7 col2\" >p4_convnext_tiny_50pct</td>\n",
       "      <td id=\"T_52656_row7_col3\" class=\"data row7 col3\" >convnext_tiny</td>\n",
       "      <td id=\"T_52656_row7_col4\" class=\"data row7 col4\" >0.9926</td>\n",
       "      <td id=\"T_52656_row7_col5\" class=\"data row7 col5\" >0.0009</td>\n",
       "      <td id=\"T_52656_row7_col6\" class=\"data row7 col6\" >0.9607</td>\n",
       "      <td id=\"T_52656_row7_col7\" class=\"data row7 col7\" >0.9735</td>\n",
       "      <td id=\"T_52656_row7_col8\" class=\"data row7 col8\" >0.9562</td>\n",
       "      <td id=\"T_52656_row7_col9\" class=\"data row7 col9\" >0.9484</td>\n",
       "      <td id=\"T_52656_row7_col10\" class=\"data row7 col10\" >106.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_52656_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_52656_row8_col0\" class=\"data row8 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_52656_row8_col1\" class=\"data row8 col1\" >100</td>\n",
       "      <td id=\"T_52656_row8_col2\" class=\"data row8 col2\" >p4_convnext_tiny_100pct</td>\n",
       "      <td id=\"T_52656_row8_col3\" class=\"data row8 col3\" >convnext_tiny</td>\n",
       "      <td id=\"T_52656_row8_col4\" class=\"data row8 col4\" >0.9954</td>\n",
       "      <td id=\"T_52656_row8_col5\" class=\"data row8 col5\" >0.0002</td>\n",
       "      <td id=\"T_52656_row8_col6\" class=\"data row8 col6\" >0.9679</td>\n",
       "      <td id=\"T_52656_row8_col7\" class=\"data row8 col7\" >0.9784</td>\n",
       "      <td id=\"T_52656_row8_col8\" class=\"data row8 col8\" >0.9662</td>\n",
       "      <td id=\"T_52656_row8_col9\" class=\"data row8 col9\" >0.9580</td>\n",
       "      <td id=\"T_52656_row8_col10\" class=\"data row8 col10\" >106.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1d2ac654b90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>scale_pct</th>\n",
       "      <th>20</th>\n",
       "      <th>50</th>\n",
       "      <th>100</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ResNet50</th>\n",
       "      <td>p3_resnet50</td>\n",
       "      <td>p4_resnet50_50pct</td>\n",
       "      <td>p4_resnet50_100pct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EfficientNet-B0</th>\n",
       "      <td>p3_efficientnet_b0</td>\n",
       "      <td>p4_efficientnet_b0_50pct</td>\n",
       "      <td>p4_efficientnet_b0_100pct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ConvNeXt-Tiny</th>\n",
       "      <td>p3_convnext_tiny</td>\n",
       "      <td>p4_convnext_tiny_50pct</td>\n",
       "      <td>p4_convnext_tiny_100pct</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "scale_pct                       20                        50   \\\n",
       "model                                                           \n",
       "ResNet50                p3_resnet50         p4_resnet50_50pct   \n",
       "EfficientNet-B0  p3_efficientnet_b0  p4_efficientnet_b0_50pct   \n",
       "ConvNeXt-Tiny      p3_convnext_tiny    p4_convnext_tiny_50pct   \n",
       "\n",
       "scale_pct                              100  \n",
       "model                                       \n",
       "ResNet50                p4_resnet50_100pct  \n",
       "EfficientNet-B0  p4_efficientnet_b0_100pct  \n",
       "ConvNeXt-Tiny      p4_convnext_tiny_100pct  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "MODEL_RUNS = {\n",
    "    \"ResNet50\": {\n",
    "        0.2: \"p3_resnet50\",\n",
    "        0.5: \"p4_resnet50_50pct\",\n",
    "        1.0: \"p4_resnet50_100pct\",\n",
    "    },\n",
    "    \"EfficientNet-B0\": {\n",
    "        0.2: \"p3_efficientnet_b0\",\n",
    "        0.5: \"p4_efficientnet_b0_50pct\",\n",
    "        1.0: \"p4_efficientnet_b0_100pct\",\n",
    "    },\n",
    "    \"ConvNeXt-Tiny\": {\n",
    "        0.2: \"p3_convnext_tiny\",\n",
    "        0.5: \"p4_convnext_tiny_50pct\",\n",
    "        1.0: \"p4_convnext_tiny_100pct\",\n",
    "    },\n",
    "}\n",
    "MODEL_ORDER = [\"ResNet50\", \"EfficientNet-B0\", \"ConvNeXt-Tiny\"]\n",
    "SCALE_ORDER = [0.2, 0.5, 1.0]\n",
    "FAKE_SOURCES = [\"inpainting\", \"insight\", \"text2img\"]\n",
    "\n",
    "\n",
    "def load_json(path: Path) -> dict:\n",
    "    return json.loads(path.read_text(encoding=\"utf-8\"))\n",
    "\n",
    "\n",
    "def resolve_config(path: Path) -> dict:\n",
    "    cfg = load_json(path)\n",
    "    parent = cfg.pop(\"extends\", None)\n",
    "    if parent:\n",
    "        base = resolve_config(path.parent / parent)\n",
    "        base.update(cfg)\n",
    "        cfg = base\n",
    "    return cfg\n",
    "\n",
    "\n",
    "def config_path_for_run(run_name: str) -> Path:\n",
    "    for phase in [\"phase3\", \"phase4\"]:\n",
    "        path = CONFIGS_DIR / phase / f\"{run_name}.json\"\n",
    "        if path.exists():\n",
    "            return path\n",
    "    raise FileNotFoundError(f\"No config found for {run_name}\")\n",
    "\n",
    "\n",
    "def load_run(run_name: str) -> dict:\n",
    "    path = LOGS_DIR / f\"{run_name}.json\"\n",
    "    if not path.exists():\n",
    "        raise FileNotFoundError(f\"Missing log for {run_name}: {path}\")\n",
    "    return load_json(path)\n",
    "\n",
    "\n",
    "def agg_metric(results: dict, metric: str, field: str = \"mean\"):\n",
    "    return results.get(\"aggregated_metrics\", {}).get(metric, {}).get(field, np.nan)\n",
    "\n",
    "\n",
    "def checkpoint_mb(run_name: str) -> float:\n",
    "    path = MODELS_DIR / f\"{run_name}_fold0_best.pt\"\n",
    "    return path.stat().st_size / (1024 * 1024) if path.exists() else np.nan\n",
    "\n",
    "\n",
    "def aggregate_confusion(run_name: str) -> np.ndarray:\n",
    "    cm = np.zeros((2, 2), dtype=int)\n",
    "    for fold in load_run(run_name).get(\"fold_results\", []):\n",
    "        fold_cm = fold.get(\"test_metrics\", {}).get(\"confusion_matrix\")\n",
    "        if fold_cm is not None:\n",
    "            cm += np.asarray(fold_cm, dtype=int)\n",
    "    return cm\n",
    "\n",
    "\n",
    "def class_balanced_metrics(run_name: str) -> dict:\n",
    "    cm = aggregate_confusion(run_name)\n",
    "    tn, fp, fn, tp = cm.ravel()\n",
    "    real_recall = tn / (tn + fp) if (tn + fp) else np.nan\n",
    "    fake_recall = tp / (tp + fn) if (tp + fn) else np.nan\n",
    "    real_precision = tn / (tn + fn) if (tn + fn) else np.nan\n",
    "    fake_precision = tp / (tp + fp) if (tp + fp) else np.nan\n",
    "    real_f1 = 2 * real_precision * real_recall / (real_precision + real_recall) if (real_precision + real_recall) else np.nan\n",
    "    fake_f1 = 2 * fake_precision * fake_recall / (fake_precision + fake_recall) if (fake_precision + fake_recall) else np.nan\n",
    "    return {\n",
    "        \"real_recall\": real_recall,\n",
    "        \"fake_recall\": fake_recall,\n",
    "        \"balanced_accuracy\": np.nanmean([real_recall, fake_recall]),\n",
    "        \"real_f1\": real_f1,\n",
    "        \"fake_f1\": fake_f1,\n",
    "        \"macro_f1\": np.nanmean([real_f1, fake_f1]),\n",
    "        \"false_positive_rate\": 1 - real_recall,\n",
    "        \"false_negative_rate\": 1 - fake_recall,\n",
    "        \"tn\": tn,\n",
    "        \"fp\": fp,\n",
    "        \"fn\": fn,\n",
    "        \"tp\": tp,\n",
    "    }\n",
    "\n",
    "\n",
    "def summarize_run(model_label: str, scale: float, run_name: str) -> dict:\n",
    "    results = load_run(run_name)\n",
    "    cfg = {**resolve_config(config_path_for_run(run_name)), **results.get(\"config\", {})}\n",
    "    balanced = class_balanced_metrics(run_name)\n",
    "    return {\n",
    "        \"model\": model_label,\n",
    "        \"scale\": scale,\n",
    "        \"scale_pct\": int(scale * 100),\n",
    "        \"run\": run_name,\n",
    "        \"backbone\": cfg.get(\"backbone\"),\n",
    "        \"image_size\": cfg.get(\"image_size\"),\n",
    "        \"data_dir\": cfg.get(\"data_dir\"),\n",
    "        \"augment\": cfg.get(\"augment\", False),\n",
    "        \"auc\": agg_metric(results, \"auc_roc\"),\n",
    "        \"auc_std\": agg_metric(results, \"auc_roc\", \"std\"),\n",
    "        \"accuracy\": agg_metric(results, \"accuracy\"),\n",
    "        \"f1\": agg_metric(results, \"f1\"),\n",
    "        \"checkpoint_mb\": checkpoint_mb(run_name),\n",
    "        **balanced,\n",
    "    }\n",
    "\n",
    "\n",
    "summary_rows = []\n",
    "for model_label in MODEL_ORDER:\n",
    "    for scale in SCALE_ORDER:\n",
    "        summary_rows.append(summarize_run(model_label, scale, MODEL_RUNS[model_label][scale]))\n",
    "summary_df = pd.DataFrame(summary_rows)\n",
    "summary_df[\"model\"] = pd.Categorical(summary_df[\"model\"], categories=MODEL_ORDER, ordered=True)\n",
    "summary_df[\"scale\"] = pd.Categorical(summary_df[\"scale\"], categories=SCALE_ORDER, ordered=True)\n",
    "summary_df = summary_df.sort_values([\"model\", \"scale\"])\n",
    "\n",
    "display(\n",
    "    summary_df[[\n",
    "        \"model\", \"scale_pct\", \"run\", \"backbone\", \"auc\", \"auc_std\", \"accuracy\", \"f1\",\n",
    "        \"balanced_accuracy\", \"macro_f1\", \"checkpoint_mb\",\n",
    "    ]]\n",
    "    .style.format({\n",
    "        \"auc\": \"{:.4f}\",\n",
    "        \"auc_std\": \"{:.4f}\",\n",
    "        \"accuracy\": \"{:.4f}\",\n",
    "        \"f1\": \"{:.4f}\",\n",
    "        \"balanced_accuracy\": \"{:.4f}\",\n",
    "        \"macro_f1\": \"{:.4f}\",\n",
    "        \"checkpoint_mb\": \"{:.1f}\",\n",
    "    })\n",
    ")\n",
    "\n",
    "matrix = summary_df.pivot(index=\"model\", columns=\"scale_pct\", values=\"run\")\n",
    "display(matrix)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a9d255f",
   "metadata": {},
   "source": [
    "The Phase 4 logs are complete for the intended 50% and 100% runs. The 20% rows are the matching Phase 3 runs, included as the baseline for the scaling curve.\n",
    "\n",
    "At 20%, ResNet50 was already the best practical detector because it had the strongest accuracy/F1 balance. Phase 4 asks whether that decision survives with more data, and whether ConvNeXt-Tiny's small AUC advantage becomes more meaningful.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60c35fd9",
   "metadata": {},
   "source": [
    "## 2. Global scaling curves\n",
    "\n",
    "The first check is the simplest one: do AUC, accuracy, and F1 improve as more data is used? If scaling is useful, each model should move upward from 20% to 50% to 100%, and fold variability should shrink rather than become unstable.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d9e20cfb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_a2ebb\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >metric</th>\n",
       "      <th id=\"T_a2ebb_level0_col0\" class=\"col_heading level0 col0\" >accuracy</th>\n",
       "      <th id=\"T_a2ebb_level0_col1\" class=\"col_heading level0 col1\" >auc</th>\n",
       "      <th id=\"T_a2ebb_level0_col2\" class=\"col_heading level0 col2\" >balanced_accuracy</th>\n",
       "      <th id=\"T_a2ebb_level0_col3\" class=\"col_heading level0 col3\" >f1</th>\n",
       "      <th id=\"T_a2ebb_level0_col4\" class=\"col_heading level0 col4\" >macro_f1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >model</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_a2ebb_level0_row0\" class=\"row_heading level0 row0\" >ResNet50</th>\n",
       "      <td id=\"T_a2ebb_row0_col0\" class=\"data row0 col0\" >+0.0160</td>\n",
       "      <td id=\"T_a2ebb_row0_col1\" class=\"data row0 col1\" >+0.0093</td>\n",
       "      <td id=\"T_a2ebb_row0_col2\" class=\"data row0 col2\" >+0.0253</td>\n",
       "      <td id=\"T_a2ebb_row0_col3\" class=\"data row0 col3\" >+0.0106</td>\n",
       "      <td id=\"T_a2ebb_row0_col4\" class=\"data row0 col4\" >+0.0216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a2ebb_level0_row1\" class=\"row_heading level0 row1\" >EfficientNet-B0</th>\n",
       "      <td id=\"T_a2ebb_row1_col0\" class=\"data row1 col0\" >+0.0233</td>\n",
       "      <td id=\"T_a2ebb_row1_col1\" class=\"data row1 col1\" >+0.0105</td>\n",
       "      <td id=\"T_a2ebb_row1_col2\" class=\"data row1 col2\" >+0.0260</td>\n",
       "      <td id=\"T_a2ebb_row1_col3\" class=\"data row1 col3\" >+0.0159</td>\n",
       "      <td id=\"T_a2ebb_row1_col4\" class=\"data row1 col4\" >+0.0299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a2ebb_level0_row2\" class=\"row_heading level0 row2\" >ConvNeXt-Tiny</th>\n",
       "      <td id=\"T_a2ebb_row2_col0\" class=\"data row2 col0\" >+0.0206</td>\n",
       "      <td id=\"T_a2ebb_row2_col1\" class=\"data row2 col1\" >+0.0086</td>\n",
       "      <td id=\"T_a2ebb_row2_col2\" class=\"data row2 col2\" >+0.0401</td>\n",
       "      <td id=\"T_a2ebb_row2_col3\" class=\"data row2 col3\" >+0.0134</td>\n",
       "      <td id=\"T_a2ebb_row2_col4\" class=\"data row2 col4\" >+0.0287</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1d2ac965490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0cAAAGZCAYAAABCEUr1AAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjksIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvJkbTWQAAAAlwSFlzAAAPYQAAD2EBqD+naQAAWo1JREFUeJzt3QmcjeX///GPfcu+kzWyL9lJEb5J+EYqVEhCm0hZs7dYihARFSki30pZItFu3ymEyL5l37fzf7wv//v8zoyZMVMz5pjzej4e5zFz7vs+97nnnHOfuT/X57o+VyKfz+czAAAAAAhxieP7AAAAAAAgGBAcAQAAAADBEQAAAABcReYIAAAAAAiOAAAAAOAqMkcAAAAAQHAEAAAAAFeROQIAAAAAgiMAAAAAuIrMEQAkQBMnTrREiRLZjh07LBjVrFnT3W60jz/+2IoWLWrJkiWzDBky3PDnR/B/NgGENoIjAIiFCz3dfvnll2vW+3w+y5Mnj1vfoEGDf/Qc7777rnse/DubNm2yJ554wm677TYbP368jRs3jpf0JjNlyhQbPnx4fB8GgAQsaXwfAAAkBClTpnQXbtWrVw+z/Mcff7Tdu3dbihQp/vG+FRxlyZLFXdhHV4sWLaxZs2b/6nnj0rfffnvDn/OHH36wK1eu2IgRI6xQoUI3/Pnx7+kc27Bhg3Xq1ImXE0CcIHMEALHg/vvvt+nTp9ulS5euuZgrX7685ciR44a8zqdPn3Y/kyRJ4gI2ZayCUfLkyd3tRjp48KD7eb3udMr2nT171kKZ9zkCgFBDcAQAsaB58+b2999/2/z58/3LLly4YP/73//s0UcfjfAxymKoi1CJEiVcIJM9e3Zr3769HT161L9N/vz57bfffnMZKK/7njdWx+vSp3XPPvusZcuWzW699dYox3V88803VqNGDUubNq2lS5fOKlas6AK46GRdKlSo4I5T3dLee+8969ev3zXB14QJE6xWrVruWJS1Kl68uI0ZM+a6Y460f+3rs88+s9dff939HXqu2rVr29atW8M8dsuWLdakSRMXcGobbass2fHjxyM9fr2Offv2db9nzZrVPZeO31unLo/z5s1zf2OqVKnc3yd//vmnPfzww5YpUyZLnTq1ValSxWbPnn3Na+Mde//+/S137tzu9X3ooYfcMZ0/f95lOvSa3HLLLda6dWu37Hr0+pQsWdLWrVvn3jM9vzJe+kyJ3vfKlSu74y1SpIh999131+xj9erVVq9ePfde67n1ei5ZsiTMNlF9jrzPzF133WVp0qRxf1f9+vXdZzI6tJ0+DzpG7fO1115zn/vwvvrqK7ffXLlyuc+NPmOvvvqqXb58Oczrodf+r7/+8p8Leu+8c61Pnz6uISJ9+vTuWHXM33//fbSOEwA8dKsDgFigi7SqVavap59+6i5GvYtKXRzrwn3kyJHXPEaBkC5MdbH8wgsv2Pbt223UqFHugvbXX391RQMUPHXo0MFd2L7yyivucQqiAumCVhf8ujiMqsVfz/Xkk0+6YKxHjx4ug6Lnmjt3bqQBnGib++67z3LmzOku/nXBOmDAAPec4SkQ0v7/+9//WtKkSW3mzJnu+HRB/Nxzz133dRw0aJAlTpzYXn75ZffaDRkyxB577DFbunSp/yK4bt26LrjQ66IAac+ePTZr1iw7duyYuzCOiF7HSZMm2ZdffumOUa9n6dKl/es3b97sAly9J23btnXBxoEDB6xatWp25swZ9/5kzpzZPvroI/e3KUBp3LhxmOcYOHCgCwK6d+/uArp33nnHvYf6exTwKhhTYKL3oUCBAu79uh49ToGbPkMK0nTs+n3y5Mku4Hr66afde/fmm2+6YGzXrl0ugPECEwUICoy6du3qjkVBn4IML7C63udIBSxatWrlXvPBgwe710LHoO6j+lx4wUlE9u/fb/fcc4/Lpuo1UcCicV56jcLTa6L3pHPnzu7nwoUL3XGcOHHC/W2iz78+E+qm+vbbb7tl2la03fvvv+/eQ71/J0+etA8++MAd97Jly6xs2bLXfa0BwPEBAP6xCRMm+PRVunz5ct+oUaN8adOm9Z05c8ate/jhh3333HOP+z1fvny++vXr+x/3888/u8dNnjw5zP7mzp17zfISJUr4atSoEelzV69e3Xfp0qUI123fvt3dP3bsmDu2ypUr+86ePRtm2ytXrkT5NzZs2NCXOnVq3549e/zLtmzZ4kuaNKl7jkDe3x6obt26voIFC4ZZpr8n8G/6/vvv3b6KFSvmO3/+vH/5iBEj3PL169e7+6tXr3b3p0+f7oupvn37usceOnQozHK9N1qu1z5Qp06d3HK9V56TJ0/6ChQo4MufP7/v8uXLYY69ZMmSvgsXLvi3bd68uS9RokS+evXqhdlv1apV3XNej14f7XfKlCn+ZZs2bXLLEidO7FuyZIl/+bx589xyve+eRo0a+ZInT+7btm2bf9nevXvd5+Duu+++7udIf2uGDBl8bdu2DXNc+/fv96VPn/6a5eF5r9/SpUv9yw4ePOgeG/jZjOxz0759e/e5O3funH+ZzqGIXjsdd+DnRo4ePerLnj2778knn4zyOAEgEN3qACCWPPLII26sirIYarnWz8gyMhqfpCzHf/7zHzt8+LD/pm5Bag2PSXcgtZRrjFFU1N1Px6QWfHVFCxTVuCRlidRdq1GjRq7Lk0fdu7wMWaDArIBa+fU3qUuYuqdF1e3Noyxa4FgkZT5EjxcvM6QucMpixBZlcpRlCDRnzhyrVKlSmCIbem/atWvnuiv+/vvvYbZv2bKly854lJnR+CVl6wJpuTI84cenRUTPp0yRRxktZfyKFSsWJvPj/e69TnrfVPRC71vBggX92yn7p8+kKisq2xLV50ifGWXjlI0J/IxqGz3f9T6jev3UDVGvoUeZKWUCo/rc6HOq59F7r/dYVQavR8fkfW6UpTxy5Ih7fdVNctWqVdd9PAB46FYHALFEF3516tRxY3h0UacLVHV1iojGzShY0PiOqIoHRPfC/nq2bdvmfmoMS0zoOBTwRVTdLaJl6g6osT2LFy++JnjR3xtZtzdP3rx5w9zPmDGj++mNw9Lfqq5Xw4YNc13LdAGtbm6PP/74dfcd09dQY1vCdz0TBSbe+sDXM/yxe8ejUu7hl+sCXq+HuupFReN0wgevenxE+wx8nQ4dOuRefwVTER2/nl8BmrpARvYa6DMqGjMUEXXXi0pkr19Ex6QugL169XLd6cIHbdEJqkVdHocOHeqCqYsXL8bo/AAAD8ERAMQitcqrBV7jLZRZiawymi5OFRjpAj8iEY3niUxEYzjigwIwDfjXJKsKXnQBr9Z8ZRA0RiSigfjhRZYBUwbGowtglTXXIH5lRzQeSON9NJ4nsJBATMTGaxjZsUfnb7qR+4zua+C9Xxp3FFG1RY0piw3KTim7qGBLY9lUjEHZTWV8unXrFq3PzSeffOI+E8qUdenSxZ1beo30ufAaBgAgOgiOACAWaZC+BvXrQn3atGmRbqcLQHVXu/POO697YR4b5bj1fKI5YmIyx48uMnWhGr5inIRfpuILKpTw9ddfh8mixEXFsFKlSrmbsg2LFi1yr+PYsWNdNbTYki9fPleoITyvm5fWBysF16puF9nxq0hE+OxTZJ8ZfQaUEY0pvT5e9ilQ+GNStT9Vevziiy/s7rvv9i9XgZLongsqkKHug9pH4DZehUIAiC7GHAFALNIYEVXzUmWyhg0bRjk+Sd3uVK44PI2VUGu6R1W+Au//E/fee6+rYqaW9HPnzkU726DWd10Yz5gxw/bu3RsmMFI1vvDbht+fukSpvHdsUZer8GN1FCTpYj865bFjOneVKp2pi6BHVdxUcU1V2lSmPFjpvdB7ruxaYDl3VeDzJiu+Xrc4jcHSNm+88UaYbmoedd273uunRgK9hoGPCZ8tjehzo6qEmvw4PJ0LEXWzi2gfqnAY+N4BQHSQOQKAWKbSx9ejbkTKMClYWbNmjbuQ1WB+tbSrWMOIESP845VUpEEBl7IiyvqoJT+ycSCR0UWuurY99dRTbm4jdf/TeJ61a9e6sSkarxEZBXrqvqbszDPPPOOCOpUc13gbHbtHf4O60Sko1N926tQpGz9+vDveffv2WWzQmJTnn3/elbW+/fbbXaCkbl+6ONbcR7FJxSu80uzquqe5jvQ6KaPx+eefu4AsmOnzoqIKCoRUplvd4FTKW0GkSqRH5zOjz12LFi2sXLlyrjCEMlI7d+508w3p86DPQWRUPlzvjcrAd+zY0V/KWxklzd3kUbl0fRZ13uh1VuZHj4soaNe5oIysxp3pc6zGCH3eVO5cWSNlbjVfkt4jZRIVwOpzCADRRXAEAPFEF2+62NMFa8+ePd3FqzISKi6gC0+P5nvR4HZd0KqSlwKrmAZH0qZNGxeoaC4hZawUjGl80Isvvhjl43SMyhJp7qHevXu77lgaG7Jx48YwlcQ00F7dm9TVTdtqnIqCKV1Qh6/Y9k+VKVPGZTTUhU/zG6nrmJbp+FQZLTZpPil12dO4F81ZpIyb5kbSc+sCPNip2MLPP//s5rRSEK6xOyqQoPE5ERVKiIiCaFUp1GdG8w0psNIktyqEocqCUVFlPHWp1HxUeryKT2heJu1Pn0WPlquy40svveQ+OwqUdA5o/Fr4CoIK8hSQKxupYF+BloIjjTfSOD+dS6pkqKBIf6caGtRtDwCiK5HqeUd7awAA/j8NfleVsYjGlQAAcDMK7j4BAICgoHLegRQQqQpdzZo14+2YAACIbWSOAADXpS5S6rqkimDq4qexKOpitXr1aitcuDCvIAAgQWDMEQDgujSoXsUJNK4jRYoUVrVqVVfFjMAIAJCQkDkCAAAAAMYcAQAAAMBVdKv7h1QSVRMialLF2Ji9HgAAAEDcUIFuTYeh6QSimqeO4OgfUmCkuT4AAAAA3Bx27dplt956a6TrCY7+IWWMvBdYs4gDAAAACE4nTpxwiQ3vGj4yBEf/kNeVToERwREAAAAQ/K43HIZJYAEAAACA4AgAAAAAriJzBAAAAACMOYp7ly9ftosXL/Jhu4kkS5bMkiRJEt+HAQAAgBuMggxxWEt9//79duzYsbh6CsShDBkyWI4cOZjDCgAAIIQQHMURLzDKli2bpU6dmovsmyioPXPmjB08eNDdz5kzZ3wfEgAAAG4QgqM46krnBUaZM2eOi6dAHEqVKpX7qQBJ7yFd7AAAAEIDBRnigDfGSBkj3Jy8947xYgAAAKGD4CgeJ5lC8OK9AwAACD0ERwAAAABAcAQAAAAAV1GQ4QbK3332jXw62zGo/g19PgAAAOBmRnAEAAhO/dJbyOl3PL6PAABCGmOOAAAAAIDgCOHNnTvXqlevbhkyZHBzNDVo0MC2bdvm1v3www+uipvmcPKsWbPGLduxY4d/2a+//mo1a9Z05bAzZsxodevWtaNHj/JiAwAAIKiROUIYp0+fts6dO9uKFStswYIFljhxYmvcuLFduXIlWq+UgqXatWtb8eLFbfHixfbLL79Yw4YN3cS4AAAAQDBjzBHCaNKkSZj7H374oWXNmtV+//33aL1SQ4YMsQoVKti7777rX1aiRAleZQAAAAQ9MkcIY8uWLda8eXMrWLCgpUuXzvLnz++W79y5M0aZIwAAAOBmQ+YIYagLXL58+Wz8+PGWK1cu152uZMmSduHCBbvlllvcNj6fz7/9xYsXwzw+VapUvKIAAAC4KZE5gt/ff/9tmzdvtl69ernsT7FixcIUUlD3Otm3b1+YTFGg0qVLu7FKAAAAwM2G4Ah+qiynCnXjxo2zrVu32sKFC11xBk+hQoUsT5481q9fP9f9bvbs2TZ06NAwr2CPHj1s+fLl9uyzz9q6dets06ZNNmbMGDt8+DCvNAAAAIIa3epuoB2D6lswU2W6qVOn2gsvvOC60hUpUsRGjhzpynJLsmTJ7NNPP7VnnnnGZYgqVqxor732mj388MP+fdx+++327bffWs+ePa1SpUqum13lypXdOCYAAAAgmBEcIYw6depcU5kucIzRnXfe6TJCka2XGjVquLmOAAAAgJsJ3eoAAAAAgOAIAAAAAK4icwQAAAAABEcAAAAAcBWZIwAAAAAgOAIAAACAq8gcAQAAAADBEQAAAABcReYIAAAAAMwsKa/CDdQv/Y19ufsdv7HPBwAAANzEyBwBAAAAAMERgt3Fixfj+xAAAAAQIsgcIYy5c+da9erVLUOGDJY5c2Zr0KCBbdu2zb9+9+7d1rx5c8uUKZOlSZPGKlSoYEuXLvWvnzlzplWsWNFSpkxpWbJkscaNG/vXJUqUyGbMmBHm+fQ8EydOdL/v2LHDbTNt2jSrUaOG28fkyZPt77//ds+ZO3duS506tZUqVco+/fTTMPu5cuWKDRkyxAoVKmQpUqSwvHnz2uuvv+7W1apVy55//vkw2x86dMiSJ09uCxYs4BMAAAAAh+AIYZw+fdo6d+5sK1ascIFD4sSJXYCj4OPUqVMuaNmzZ499/fXXtnbtWuvatatbJ7Nnz3bb3n///bZ69Wr3+EqVKsX4Fe7evbt17NjRNm7caHXr1rVz585Z+fLl3f43bNhg7dq1sxYtWtiyZcv8j+nRo4cNGjTIevfubb///rtNmTLFsmfP7tY99dRT7v758+f923/yyScu2FLgBAAAAAgFGRBGkyZNwtz/8MMPLWvWrC7gWLRokcu4LF++3GWORJkajzI1zZo1s/79+/uXlSlTJsavcKdOnezBBx8Ms+zll1/2/96hQwebN2+effbZZy74OnnypI0YMcJGjRplrVq1ctvcdtttLgMm2pcyR1999ZU98sgjbpmyVU888YTLVAEAAABC5ghhbNmyxXVhK1iwoKVLl87y58/vlu/cudPWrFljd9xxhz8wCk/ra9eu/a9fUXXVC3T58mV79dVXXXc6Pfctt9zigiMdkyjDpKxQZM+t7nnKNCnQk1WrVrkMlIIjAAAAwEPmCGE0bNjQ8uXLZ+PHj7dcuXK5LnMlS5a0CxcuWKpUqaJ8ta63Xlkan8933YILGssU6M0333SZoeHDh7sASeuVXdIxRed5va51ZcuWdWOmJkyY4LrT6e8EAAAAPGSO4KfCB5s3b7ZevXq5LEyxYsXs6NGj/vWlS5d22aEjR45E+KppfVQFDtQ9b9++fWGyVGfOnLnuO/Drr7/aAw88YI8//rjrpqes1h9//OFfX7hwYRcgRfXcCqqUkVLQp/FHTz75JO88AAAAwiA4gl/GjBldhbpx48bZ1q1bbeHCha44g0fd7XLkyGGNGjVyAcuff/5pn3/+uS1evNit79u3r6sip5/q6rZ+/XobPHiw//HK1mhckIo1qODD008/bcmSJbvuO6DgZ/78+W7Mk/bbvn17O3DgQJhuc926dXPFISZNmuSq6y1ZssQ++OCDa7JHKtqg7FVgFT0AAABA6FZ3I/U7HtSfOlWmmzp1qr3wwguuK12RIkVs5MiRVrNmTbdepa+//fZbe+mll1xFukuXLlnx4sVt9OjRbr22mz59uhsfpCBEY5buvvtu//6HDh1qrVu3trvuust12VNXuZUrV173uJTJUiCmynUq5a1qdQrQjh//v9dTVeqSJk1qffr0sb1791rOnDld8BVIwZ264+mnAioAAAAgUCJf+EEgiJYTJ05Y+vTp3QW6goBAKj29fft2K1CgABfhQUTzKKmKnartlStXLspteQ+BINAvvYWcIG9EA4CEeO0eiMwREjwVfdB4KmWgqlSpct3ACAAAAKGJMUdI8DQ+St3slDEaO3ZsfB8OAAAAghSZIyR4GgtF71EAAABcD5kjAAAAACBzBAAAgHhF8RUEETJHAAAAAEBwBAAAAABXkTkCAAAAgGAJjkaPHm358+d3E6ZWrlzZli1bFuX206dPt6JFi7rtS5UqZXPmzAkzp023bt3c8jRp0liuXLmsZcuWtnfv3jD7OHLkiD322GNuEqgMGTJYmzZt7NSpU3H2NwIAAAAIbvFeynvatGnWuXNnN/+MAqPhw4db3bp1bfPmzZYtW7Zrtl+0aJE1b97cBg4caA0aNLApU6ZYo0aNbNWqVVayZEk7c+aM+713795WpkwZO3r0qHXs2NH++9//2ooVK/z7UWC0b98+mz9/vguoWrdube3atXP7iyulPiplN9L6Vutj/BiVvG7fvr3973//c6/d6tWrrWzZsnFyfAAAAEAwSeSL5wlgFBBVrFjRRo0a5e5fuXLF8uTJYx06dLDu3btfs33Tpk3t9OnTNmvWLP+yKlWquAv4yCb41OSflSpVsr/++svy5s1rGzdutOLFi7vlFSpUcNvMnTvX7r//ftu9e7fLNl3PiRMnLH369Hb8+HGXfQp07tw52759uxUoUMBlt26m4Oibb76xBx54wH744QcrWLCg/fHHH/bmm2/aypUrXTD55ZdfumA0oYvsPQRwA1HBCggNnOu4AaK6dg+abnUXLlxwF9116tT5vwNKnNjdX7x4cYSP0fLA7UWZpsi2F70IiRIlct3nvH3ody8wEu1Tz7106dII93H+/Hn3ogbeEqJt27ZZzpw5rVq1apYjRw4XiCoDp66PAAAAQEIWr93qDh8+bJcvX7bs2bOHWa77mzZtivAx+/fvj3B7LY8sA6AxSOqK50WJ2jZ8l72kSZNapkyZIt2PuvH179/fErInnnjCPvroI/e7gsl8+fLZjh07rF69evF9aAAAAEBoFGSIKxpL9Mgjj7hxNGPGjPlX++rRo4fLQHm3Xbt2WUIzYsQIGzBggN16662uC526HQIAAAChIl4zR1myZLEkSZLYgQMHwizXfXXpioiWR2d7LzDSOKOFCxeG6VuobQ8ePBhm+0uXLrkKdpE9b4oUKdwtIVM/zLRp07r3JLLXAQAAAEio4jVzlDx5citfvrwtWLDAv0wFGXS/atWqET5GywO3F1WcC9zeC4y2bNli3333nWXOnPmafRw7dsyNd/IogNJzq0AEAAAAgNAT76W8Vca7VatWrjiCKsqplLeKAKi0tmiOoty5c7sxP6Ky3DVq1LChQ4da/fr1berUqa5E97hx4/yB0UMPPeTKeauincY0eeOINKZIAVmxYsXsvvvus7Zt27oKd3rM888/b82aNYtWpToAAAAACU+8B0cqzX3o0CHr06ePC2JUkltltb2iCzt37nRV5Dyqoqa5iHr16mU9e/a0woUL24wZM9wcR7Jnzx77+uuv3e/h5+f5/vvvrWbNmu73yZMnu4Codu3abv9NmjSxkSNH3sC/HAAAAEAwiffgSBSk6BYRzbcT3sMPP+xuEcmfP78rwHA9yiLF5YSvCcWpU6ds69at/vua+2fNmjXu9dOcUQAAAEBCERTBUaj4J5Oyxjd1WbznnnvCdIMUdYWcOHFiPB4ZAAAAELsIjhBGp06d3M2jbojRycQBAAAAN7sEPc8RAAAAAEQXwREAAAAAEBwBAAAAwFVkjgAAAACA4ChuXblyhQ/ZTYr3DgAAIPRQrS4OJE+e3E0su3fvXsuaNau7nyhRorh4KsQyVea7cOGCm5hY76HeOwAAAIQGgqM4oIvqAgUK2L59+1yAhJtP6tSp3SS3ei8BAAAQGgiO4ogyDrq4vnTpkl2+fDmungZxIEmSJJY0aVKyfQAAACHmXwVHs2fPth9++MFd/N95553WpEmT2DuyBEBd6ZIlS+ZuAAAAAILbP+4z1Lt3b+vatasLADRO48UXX7QOHTrE7tEBAAAAQLBljlasWGEVKlTw3582bZqtXbvWUqVK5e4/8cQTVrNmTXvnnXfi5kgBAAAAIBgyR08//bR16tTJzpw54+4XLFjQhg4daps3b7b169fbmDFj7Pbbb4/LYwUAAACA+A+Oli5dajlz5rRy5crZzJkz7cMPP7TVq1dbtWrV7K677rLdu3fblClT4u5IAQAAACAYutWpgle3bt3s4YcftmeeecbSpEljo0aNsly5csXl8QEAAABAcBZkUHe6efPmWePGje3uu++20aNHx82RAQAAAEAwBkfHjh1z1ekaNmxovXr1csGRutotX77cqlSp4sYdAQAAAECCD45atWrlgqH69eu7IgzqWpc5c2abOHGivf7669a0aVPX7Q4AAAAAEvSYo4ULF7oCDIUKFbK2bdu6n57atWvbqlWrbMCAAXF1nAAAAAAQHJmjwoUL27hx4+yPP/6wsWPHWr58+cKsT5kypb3xxhtxcYwAAAAAEDzBkUp3K3t0xx13uJLd7777btweGQAAAAAEY7e6smXL2ooVK+L2aAAAAAAg2IMjz6VLl+y3336z/fv3u/s5cuSw4sWLW7JkyeLi+AAAAAAguIKjK1euWJ8+fdy8RsePHw+zLn369Pb8889b//79LXHiGE+dBAAAAAA3T3DUvXt3V7Z70KBBVrduXcuePbtbfuDAAfv222+td+/eduHCBRs8eHBcHi8AAAAAxG9wNGnSJPv4449dYBQof/781q5dO1e9rmXLlgRHAAAAAG5K0e4Dd/LkScuVK1ek63PmzGmnT5+OreMCAAAAgOAMjmrWrGkvv/yyHT58+Jp1WtatWze3DQAAAAAk6G51mvj1/vvvdxmiUqVKhRlztH79elexbtasWXF5rAAAAAAQ/8FRnjx5bO3atTZv3jxbsmSJv5R3pUqV7I033rB7772XSnUAAAAAQmOeI5XprlevnrsBAAAAQEhPArts2TJbvHhxmElgq1WrZhUrVoyL4wMAAACA4AqODh48aE2aNLFff/3V8ubNG2bM0Ysvvmh33nmnff7555YtW7a4PF4AAAAAiN9qdc8++6xdvnzZNm7caDt27LClS5e6m37XsitXrthzzz0XN0cJAAAAAMGSOVIhhp9++smKFClyzTotGzlyJKW8AQAAACT8zFGKFCnsxIkTUU4Sq20AAAAAIEEHR02bNrVWrVrZl19+GSZI0u9a1rp1a2vevHlcHScAAAAABEe3umHDhrlxRc2aNbNLly5Z8uTJ3fILFy5Y0qRJrU2bNvbWW2/F5bECAAAAQPwHR+oyN2bMGBs8eLCtXLkyTCnv8uXLW7p06eLuKAEAAAAg2OY5UhB0zz33xM3RAAAAAECwjzm6Hs13NGDAgNjaHQAAAADcnMGRutn1798/tnYHAAAAAMHZrW7dunVRrt+8eXNsHA8AAAAABHdwVLZsWUuUKJH5fL5r1nnL9RMAAAAAEnRwlClTJhsyZIjVrl07wvW//fabNWzYMDaPDYhYv/Sh9cr0Ox7fRwAAABASoh0cqVz33r17LV++fBGuP3bsWIRZJQAAED2lPioVUi/V+lbr4/sQAOCfBUdPP/20nT59OtL1efPmtQkTJkR3dwAAAABwcwZHjRs3jnJ9xowZrVWrVrFxTAAAAABw85byBgAAAICbGcERAAAAABAcAQAAAMBVZI4AAAAAgOAIAAAAAP5F5khzGg0dOtSeeuopd3v77bft+PF/NlHl6NGjLX/+/JYyZUqrXLmyLVu2LMrtp0+fbkWLFnXblypVyubMmRNm/RdffGH33nuvZc6c2RIlSmRr1qy5Zh81a9Z06wJvKlUOAAAAIHTFODhasWKF3XbbbS4gOnLkiLsNGzbMLVu1alWM9jVt2jTr3Lmz9e3b1z22TJkyVrduXTt48GCE2y9atMiaN29ubdq0sdWrV1ujRo3cbcOGDf5tNBdT9erVbfDgwVE+d9u2bW3fvn3+25AhQ2J07AAAAAASlkQ+n88XkwfcddddVqhQIRs/frwlTXp1mqRLly65DNKff/5pP/30U7T3pUxRxYoVbdSoUe7+lStXLE+ePNahQwfr3r37Nds3bdrUBT+zZs3yL6tSpYqVLVvWxo4dG2bbHTt2WIECBVwQpfXhM0daNnz4cPunTpw4YenTp3cZs3Tp0v3j/eAf6Jc+tF62fv8sKwvc9ELtXDezUgXyWihZ32p9fB8CgkEInuv8b7/xonvt/o8yR926dfMHRqLfu3bt6tZF14ULF2zlypVWp06d/zuYxInd/cWLF0f4GC0P3F6UaYps+6hMnjzZsmTJYiVLlrQePXrYmTNnotz+/Pnz7kUNvAEAAABIOP4vwokmRVo7d+50434C7dq1y9KmTRvt/Rw+fNguX75s2bNnD7Nc9zdt2hThY/bv3x/h9loeE48++qjly5fPcuXKZevWrXPB3ubNm914pcgMHDjQ+vfvH6PnAQAAAJCAgyN1bdOYn7feesuqVavmlv3666/WpUsXNx7oZtCuXTv/7yrqkDNnTqtdu7Zt27bNjZ2KiLJLGh/lUeZIXQABAAAAhGhwpKBI1d1atmzpxhpJsmTJ7JlnnrFBgwZFez/q0pYkSRI7cOBAmOW6nyNHjggfo+Ux2T4mY59k69atkQZHKVKkcDcAAAAACVOMxxwlT57cRowYYUePHnVlsnVTxTpVr4tJ8KD9lC9f3hYsWOBfpoIMul+1atUIH6PlgdvL/PnzI90+urxy38ogAQAAAAhNMQ6OnnzySTt58qSlTp3adUnTTb+ripzWxYS6qanq3UcffWQbN2502Sftp3Xr1m69slPqzubp2LGjzZ07182xpHFJ/fr1c0Ugnn/+ef82CtQU7Pz+++/uvsYS6b43Lkld51599VVXDEIV7b7++mv3PHfffbeVLl06pi8HAAAAgFANjhTInD179prlWjZp0qQYj19SN70+ffq40toKYhT8eEUXVPhBcxB5NMZpypQpNm7cODcn0v/+9z+bMWOGqzjnUbBzxx13WP369d39Zs2aufteqW9lrL777js3UayKSrz00kvWpEkTmzlzZkxfCgAAAAChOM+RChBo04wZM9qWLVssa9as/nWqOqfgQnMT7d2710IB8xzFo1CbD4F5jhCqQu1cZ54jhKoQPNf53x681+7RLsiQIUMGV4hBt9tvv/2a9VpOqWsAAAAAN6toB0fff/+9yxzVqlXLPv/8c8uUKZN/nbqqefMGAQAAAECCDo5q1Kjhfm7fvt3y5s3rMkUAAAAAELLzHClDBAAAAAAW6tXqAAAAACAhIjgCAAAAAIIjAAAAAPiHY44CHT582JYuXermOapYsaLlzJnz3+wOAAAAAG6+4EjlvNu0aePmPLp48aJt3rzZRo8eba1bt47dIwQAAACAYBpzdOrUqTD3NeHrsmXL3G316tU2ffp0e+WVV+LiGAEAAAAgeIKj8uXL21dffeW/nzRpUjt48KD//oEDB9xksAAAAACQoLvVzZs3z5577jmbOHGi6z43YsQIa9q0qRtvdOnSJUucOLFbBwAAAAAJOjjKnz+/zZ492z799FOrUaOGvfDCC7Z161Z3U4BUtGhRS5kyZdweLQAAAAAEyzxHzZs3t+XLl9vatWutZs2aduXKFStbtiyBEQAAAIDQqVY3Z84c27hxo5UpU8bef/99+/HHH+2xxx6zevXq2YABAyxVqlRxd6QAAABAAlDqo1IWSta3Wm8JLnP00ksvuTLdyhq1b9/eXn31Vde9btWqVS5rdMcdd9g333wTt0cLAAAAAPEdHKnYgjJHU6dOdQHSxx9/7JarQp0CpS+++MLeeOONuDpOAAAAAAiO4ChNmjS2fft29/uuXbuuGWNUvHhx+/nnn2P/CAEAAAAgmIKjgQMHWsuWLS1XrlyuO52yRQAAAAAQcgUZVHjhvvvusz///NMKFy5sGTJkiNsjAwAAAIBgrVaXOXNmd5Pz58+7nylSpIibIwMAAACAYJ3naP78+Xb//fdbxowZLXXq1O6m37Xsu+++i7ujBAAAAIBgCY4++ugjFwSlT5/e3n77bZs1a5a76Xd1sdM6r4IdAAAAACTYbnWvv/66DR8+3J577rlr1j3xxBNWvXp1NxFsixYtYvsYAQAAACB4Mkc7d+60OnXqRLq+du3atnv37tg6LgAAAAAIzuCoRIkS9sEHH0S6/sMPP3RzHQEAAABAgu5WN3ToUGvQoIHNnTvXZZCyZ8/ulh84cMAWLFjgSnzPnj07Lo8VAAAAAOI/OKpZs6Zt2LDBxowZY0uWLLH9+/e75Tly5LB69erZ008/bfnz54+7IwUAAACAYJnnSMHP4MGD4+5oAAAAAOBmmOcIAAAAABKqGAVH7777rhtv9Mgjj7hxRoEOHz5sBQsWjO3jAwAAAIDgCo5GjhxpXbp0saJFi1qKFCncpK8DBw70r798+bL99ddfcXWcAAAAABAcY47ee+89Gz9+vD366KPu/jPPPGONGjWys2fPuslfAQAAACAkgqPt27dbtWrV/Pf1+8KFC103u4sXL1qnTp3i6hgBAAAAIHiCoyxZstiuXbvClOsuWbKkC5Bq1aple/fujatjBAAAAIDgGXNUvXp1++KLL65ZXrx4cVec4ZtvvontYwMAAACA4Mscde/e3VauXBnhuhIlSrgM0ueffx6bxwYAAAAAwRcclS5d2t0ioy52ugEAAADAzYhJYAEAAACA4AgAAAAAriJzBAAAAAAERwAAAABwFZkjAAAAAIhJtTrP33//bX369LHvv//eDh48aFeuXAmz/siRI7ywAAAAABJ+cNSiRQvbunWrtWnTxrJnz26JEiWKmyMDAAAAgGAOjn7++Wf75ZdfrEyZMnFzRAAAAABwM4w5Klq0qJ09ezZujgYAAAAAbpbg6N1337VXXnnFfvzxRzf+6MSJE2FuAAAAABAS3eoyZMjggqBatWqFWe7z+dz4o8uXL8fm8QEAAABAcAZHjz32mCVLlsymTJlCQQbgBij1UamQe53Xt1of34cAAABCUIyDow0bNtjq1autSJEicXNEAAAAAHAzBEcVKlSwXbt2ERwBwA2Wv/vskHrNd6SM7yMAAISaGAdHHTp0sI4dO1qXLl2sVKlSrotdoNKlS8fm8QEAAABAcAZHTZs2dT+ffPJJ/zIVYqAgAwAAAICQKuW9ffv2a25//vmn/2dMjR492vLnz28pU6a0ypUr27Jly6Lcfvr06W6uJW2vzNWcOXPCrP/iiy/s3nvvtcyZM7ugbc2aNdfs49y5c/bcc8+5bW655RZr0qSJHThwIMbHDgAAACCEg6N8+fJFeYuJadOmWefOna1v3762atUqK1OmjNWtW9cOHjwY4faLFi2y5s2bW5s2bVxRiEaNGrmbikR4Tp8+bdWrV7fBgwdH+rwvvviizZw50wVamq9p79699uCDD8bo2AEAAAAkLIl86g8XQ9u2bbPhw4fbxo0b3f3ixYu7cUi33XZbjPajTFHFihVt1KhR7v6VK1csT548blxT9+7dI+zSp+Bn1qxZ/mVVqlSxsmXL2tixY8Nsu2PHDitQoIALorTec/z4ccuaNasrRf7QQw+5ZZs2bbJixYrZ4sWL3f4icv78eXfzaK4nHav2ly5duhj93fiX+qUPqZewVIG8Fmoo5R2x0CvI8KiFmlA73znXEYr/14Vz/cbTtXv69Omve+0e48zRvHnzXDCk7m8qvqDb0qVLrUSJEjZ//vxo7+fChQu2cuVKq1Onzv8dTOLE7r6ClIhoeeD2okxTZNtHRM958eLFMPtRN728efNGuZ+BAwe6F9S7KTACAAAAEMIFGZTRUbe0QYMGXbO8W7du9p///Cda+zl8+LBdvnzZTSQbSPeVyYnI/v37I9xey6NL2yZPntwyZMgQo/306NHDdQEMnzkCAAAAkDDEOHOkrnQa8xOeqtf9/vvvllClSJHCpeACbwAAAABCODjSeJ2IKsBpWbZs2aK9nyxZsliSJEmuqRKn+zly5IjwMVoek+0j24e69B07duxf7QcAAABAiAdHbdu2tXbt2rlqcD///LO7qYtd+/bt3broUte28uXL24IFC/zLVJBB96tWrRrhY7Q8cHvROKfIto+InlMT1wbuZ/PmzbZz584Y7QcAAABAiI856t27t6VNm9aGDh3qxuFIrly5rF+/fvbCCy/EaF8aw9OqVSurUKGCVapUyVXAUzW61q1bu/UtW7a03Llzu2IIoop4NWrUcM9dv359mzp1qq1YscLGjRvn3+eRI0dcoKPy3F7gI8oK6aZiCuoWqOfOlCmT6x6n6ngKjCKrVAcAAAAg4YtRcHTp0iVXAvvRRx91RRlOnjzplitY+idUmvvQoUPWp08fVwxBJbfnzp3rL7qgIEcV7DzVqlVzz9+rVy/r2bOnFS5c2GbMmGElS5b0b/P111/7gytp1qyZ+6m5lBTAydtvv+32q8lfVZ5bFe/efffdf/Q3AAAAAAjReY5Sp07tijLEdMLXhCa6tdIRB0JsPoRQmwtBmPskYsxzlPCF2vnOuY5Q/L8unOsJaJ4jdX/TxKoAAAAAENJjjp599ll76aWXbPfu3a64QZo0acKs16SwAAAAAJDggyNvDE9g8YVEiRKZeufppyZ2xY0Vel1t4vsIAAAAkBDFODjavn173BwJAAAAAAR7cFSuXDk3L1DGjBnto48+spdfftkVZgAAAACAhCJaBRlUnU7zD0n//v3t1KlTcX1cAAAAABB8mSPNP6S5g6pXr+7GFr311lt2yy23RLit5iwCAAAAgAQZHE2cONFNojpr1ixXdOGbb76xpEmvfajWERwBAAAASLDBUZEiRWzq1Knu98SJE7vxR9myZYvrYwMAAACA4K1Wd+XKlbg5EgAAAAC4mYIjAAAAxB3mLwSCvFodAAAAACR0BEcAAAAAQHAEAAAAAFeROQIAAACAf1KQ4fLly/b222/bZ599Zjt37rQLFy6EWX/kyBFeWAAAAAAJP3PUv39/GzZsmDVt2tSOHz9unTt3tgcffNDNf9SvX7+4OUoAAAAACLbgaPLkyTZ+/Hh76aWXLGnSpNa8eXN7//33rU+fPrZkyZK4OUoAAAAACLbgaP/+/VaqVCn3+y233OKyR9KgQQObPXt27B8hAAAAAARjcHTrrbfavn373O+33Xabffvtt+735cuXW4oUKWL/CAEAAAAgGIOjxo0b24IFC9zvHTp0sN69e1vhwoWtZcuW9uSTT8bFMQIAAABA8FWrGzRokP93FWXIly+fLVq0yAVIDRs2jO3jAwAAAIDgDI5++uknq1atmivGIFWqVHG3S5cuuXV33313XBwnAAAAAARXt7p77rknwrmMVJhB6wAAAAAgJIIjn89niRIlumb533//bWnSpImt4wIAAACA4OxWp4leRYHRE088EaYy3eXLl23dunWuux0AAAAAJOjgKH369P7MUdq0aS1VqlT+dcmTJ3fjjtq2bRs3RwkAAAAAwRIcTZgwwf3Mnz+/vfzyy3ShAwAAABDa1er69u0bN0cCAAAAAMEeHN1xxx0RFmGIyKpVq/7tMQEAAABAcAZHjRo1ivsjAQAAAIBgD47oSgcAAAAgoYvxPEdy7Ngxe//9961Hjx7+CWHVnW7Pnj2xfXwAAAAAEJwFGTSfUZ06dVxp7x07drjy3ZkyZbIvvvjCdu7caZMmTYqbIwUAAACAYMocde7c2U0Cu2XLFkuZMqV/+f33328//fRTbB8fAAAAAARncLR8+XJr3779Nctz585t+/fvj63jAgAAAIDgDo5SpEhhJ06cuGb5H3/8YVmzZo2t4wIAAACA4A6O/vvf/9qAAQPs4sWL7r7mP9JYo27dulmTJk3i4hgBAAAAIPiCo6FDh9qpU6csW7ZsdvbsWatRo4YVKlTI0qZNa6+//nrcHCUAAAAABFu1OlWpmz9/vv3yyy+ucp0CpXLlyrkKdgAAAAAQMsGRp3r16u4GAAAAACEXHF25csUmTpzo5jTSHEcab1SgQAF76KGHrEWLFu4+AAAAACToMUc+n88VY3jqqadsz549VqpUKStRooT99ddfbt6jxo0bx+2RAgAAAEAwZI6UMdIkrwsWLLB77rknzLqFCxdao0aNbNKkSdayZcu4OE4AAAAACI7M0aeffmo9e/a8JjCSWrVqWffu3W3y5MmxfXwAAAAAEFzBkSrT3XfffZGur1evnq1duza2jgsAAAAAgjM4OnLkiGXPnj3S9Vp39OjR2DouAAAAAAjO4Ojy5cuWNGnkQ5SSJElily5diq3jAgAAAIDgLMiganWqSpciRYoI158/fz42jwsAAAAAgjM4atWq1XW3oVIdAAAAgAQfHE2YMCFujwQAAAAAboYxRwAAAACQkBEcAQAAAADBEQAAAAAEUeZo9OjRlj9/fkuZMqVVrlzZli1bFuX206dPt6JFi7rtS5UqZXPmzLmmsl6fPn0sZ86clipVKqtTp45t2bIlzDZ6vkSJEoW5DRo0KE7+PgAAAADBL96Do2nTplnnzp2tb9++tmrVKitTpozVrVvXDh48GOH2ixYtsubNm1ubNm1s9erV1qhRI3fbsGGDf5shQ4bYyJEjbezYsbZ06VJLkyaN2+e5c+fC7GvAgAG2b98+/61Dhw5x/vcCAAAACE7xHhwNGzbM2rZta61bt7bixYu7gCZ16tT24YcfRrj9iBEj7L777rMuXbpYsWLF7NVXX7Vy5crZqFGj/Fmj4cOHW69eveyBBx6w0qVL26RJk2zv3r02Y8aMMPtKmzat5ciRw39TEBUZzeN04sSJMDcAAAAACUe8BkcXLlywlStXum5v/gNKnNjdX7x4cYSP0fLA7UVZIW/77du32/79+8Nskz59etddL/w+1Y0uc+bMdscdd9ibb75ply5divRYBw4c6Pbj3fLkyfOP/24AAAAAN/E8R3Hh8OHDdvnyZcuePXuY5bq/adOmCB+jwCei7bXcW+8ti2wbeeGFF1zGKVOmTK6rXo8ePVzXOmWyIqL16v7nUeaIAAkAAABIOOI1OIpPgYGOut4lT57c2rdv7zJEKVKkuGZ7LYtoOQAAAICEIV671WXJksWSJEliBw4cCLNc9zUGKCJaHtX23s+Y7FPU7U7d6nbs2PGP/x4AAAAAN694DY6UrSlfvrwtWLDAv+zKlSvuftWqVSN8jJYHbi/z58/3b1+gQAEXBAVuoy5wqloX2T5lzZo1brxTtmzZYuEvAwAAAHCzSRoM3dtatWplFSpUsEqVKrlKc6dPn3bV66Rly5aWO3du191NOnbsaDVq1LChQ4da/fr1berUqbZixQobN26cW6/5ijp16mSvvfaaFS5c2AVLvXv3tly5crmS36LCDAqW7rnnHlexTvdffPFFe/zxxy1jxozx+GoAAAAACNngqGnTpnbo0CE3aasKJpQtW9bmzp3rL6iwc+dOl9HxVKtWzaZMmeJKdffs2dMFQCrRXbJkSf82Xbt2dQFWu3bt7NixY1a9enW3T00aKxo7pKCqX79+rkS3AigFR4HjkAAAAACElkQ+TQyEGFNXPZX0Pn78uKVLly5eX8H83WdbKNmR8lELJaUK5LVQs77V+vg+hKDEuZ7whdr5zrkeMc71hI9zPXiv3eN9ElgAAAAACAYERwAAAABAcAQAAAAAV5E5AgAAAACCIwAAAAC4iswRAAAAABAcAQAAAMBVZI4AAAAAgOAIAAAAAK4icwQAAAAABEcAAAAAcBWZIwAAAAAgOAIAAACAq8gcAQAAAADBEQAAAABcReYIAAAAAAiOAAAAAOAqMkcAAAAAQHAEAAAAAFeROQIAAAAAgiMAAAAAuIrMEQAAAAAQHAEAAADAVWSOAAAAAIDgCAAAAACuInMEAAAAAARHAAAAAHAVmSMAAAAAIDgCAAAAgKvIHAEAAAAAwREAAAAAXEXmCAAAAAAIjgAAAADgKjJHAAAAAEBwBAAAAABXkTkCAAAAAIIjAAAAALiKzBEAAAAAEBwBAAAAwFVkjgAAAACA4AgAAAAAriJzBAAAAAAERwAAAABwFZkjAAAAACA4AgAAAICryBwBAAAAAMERAAAAAFxF5ggAAAAACI4AAAAA4CoyRwAAAABAcAQAAAAAV5E5AgAAAACCIwAAAAAIoszR6NGjLX/+/JYyZUqrXLmyLVu2LMrtp0+fbkWLFnXblypVyubMmRNmvc/nsz59+ljOnDktVapUVqdOHduyZUuYbY4cOWKPPfaYpUuXzjJkyGBt2rSxU6dOxcnfBwAAACD4xXtwNG3aNOvcubP17dvXVq1aZWXKlLG6devawYMHI9x+0aJF1rx5cxfMrF692ho1auRuGzZs8G8zZMgQGzlypI0dO9aWLl1qadKkcfs8d+6cfxsFRr/99pvNnz/fZs2aZT/99JO1a9fuhvzNAAAAAIJPvAdHw4YNs7Zt21rr1q2tePHiLqBJnTq1ffjhhxFuP2LECLvvvvusS5cuVqxYMXv11VetXLlyNmrUKH/WaPjw4darVy974IEHrHTp0jZp0iTbu3evzZgxw22zceNGmzt3rr3//vsuU1W9enV75513bOrUqW47AAAAAKEnaXw++YULF2zlypXWo0cP/7LEiRO7bnCLFy+O8DFarkxTIGWFvMBn+/bttn//frcPT/r06V0QpMc2a9bM/VRXugoVKvi30fZ6bmWaGjdufM3znj9/3t08x48fdz/37NljJ06csPh06cRhCyW7L1yxUHLhyAULNbt3747vQwhKnOsJX6id75zrEeNcT/g412+8kydP+hMpQRscHT582C5fvmzZs2cPs1z3N23aFOFjFPhEtL2We+u9ZVFtky1btjDrkyZNapkyZfJvE97AgQOtf//+1yxXtgs3Vp6Qe8H/sFCTp3Povcu4Vmh+CkLrfOdch/schOTLwLken0GSEidBGRzdTJTdCsxYXblyxRV1yJw5syVKlChejw1xT9nBPHny2K5du1wRDwAJE+c6EBo410OPz+dzgVGuXLmi3C5eg6MsWbJYkiRJ7MCBA2GW636OHDkifIyWR7W991PLVK0ucJuyZcv6twlf8OHSpUsu2InseVOkSOFugdQ1D6FFgRHBEZDwca4DoYFzPbSkjyJjFBQFGZInT27ly5e3BQsWhMnI6H7VqlUjfIyWB24vqjjnbV+gQAEX4ARuo9YBjSXyttHPY8eOufFOnoULF7rn1tgkAAAAAKEn3rvVqataq1atXHGESpUquUpzp0+fdtXrpGXLlpY7d2435kc6duxoNWrUsKFDh1r9+vVdhbkVK1bYuHHj3Hp1cevUqZO99tprVrhwYRcs9e7d26XQVPJbVOVOFe9UJU/V8S5evGjPP/+8K9ZwvVQbAAAAgIQp3oOjpk2b2qFDh9ykrSqGoK5vKrPtFVTYuXOnqyLnqVatmk2ZMsWV6u7Zs6cLgFSprmTJkv5tunbt6gIszVukDJFKdWufmjTWM3nyZBcQ1a5d2+2/SZMmbm4kICLqUqm5uMJ3rQSQsHCuA6GBcx2RSeS7Xj07AAAAAAgB8T4JLAAAAAAEA4IjAAAAACA4AgAAAICryBwBAAAAAMERACCUaD47AKGHcx/RReYI+P80cbBKvKsMPICEyZsaYs+ePXbmzJn4PhwAN0jgtDBAVPikIOTNmzfPzZelCYdff/11q1Onjn377bch/7oANxvNTKHW4chaiE+cOOHmt8uYMaPddddd7pz//fffb/hxAog7kZ3/f/75p3355Ze89LgugiOEDE0o/M4779jq1avd/cuXL7tJgt9++22rUqWKa0n+/PPPLU+ePNapUyfbu3dvfB8ygGjQuSyJEiVyrcO6RXSB9OGHH9qSJUvso48+sqlTp9qmTZusY8eOtnbtWreeaf+Am4vO80uXLvnv6xz2MkTe94Jn7Nix1rlzZ//jgMgQHCHB0pfkzz//bI888ohlyZLFbrvtNhs/frwdPXrUrU+SJIlt2bLFVq5caU8//bRbVqxYMRsxYoTb5n//+188/wUAokPnsuh8f+GFF6xatWr22GOP2ffff28XLlzwfx8oKLrzzjvtv//9r1WqVMnfjVYXTd42AOKXzsPAwOaTTz6xJ5980jVwitZ5wY0CoaRJk/q3VQPJ9u3brWLFijZx4kT/8osXL7oG0AYNGvgfB0SGTwcSHO9iSF+kvXr1svnz59tXX33lAp5169ZZrVq1/NumT5/e/v77bxc4eV+6OXPmdF+sixcvtkOHDsXb3wEgenSe62LnwQcftF27dtmjjz5q586ds/bt29uMGTPcNhs2bLDkyZNboUKF/I8rWrSou1hSxli4YALiLxjyGicU4HgNHsePH3fdYOvVq2c5cuRwy7TOO1fVuKnzvEmTJjZz5ky3rECBAq7b7Guvvea60kmyZMnc///8+fP7nxOIDMERbnr6kvv444/9Fz26AJJcuXK5L8g77rjDtRanTp3afvvttzApeLUmpUuXztasWePue61Vann+66+/3M17DgDBxTuXdeGTLVs21z1OYwo0rmjo0KGWN29eW758udsme/bsLmDSOe9JkSKF1ahRww4fPuyyyABuPC8Y0k91Zx83bpzdfffdLgAqX768lSlTxh5++GH//3apXbu2vffee/bWW2+5wiopU6Z0AZK6y8qwYcNcMNWvXz/bvXu3+x+eKlUq/3cG/9MRFYIj3HSUTg9MuesLtVy5cq6FSIUU+vbta88995zbpnjx4q7FuGTJki47pC52+gKdO3eue2yGDBlc6/HXX3/t37eoy41aoL0vUj0HgBtHFy+6qWVYrccR8VqXH3roIRfgqNuMR40hyh43bNjQ3VfwdMstt7ggSIUZPMoUa5yhuuR5zwvgxtD/2EmTJrn/y5kzZ7Zbb73VdXNXo4UaNjZv3uyWdevWzd3U00PUE+SZZ55xjZ9qHFUX2ebNm7txxQqGRAWW9Lu6yqv7rP6/K6skZIkRFYIj3HT0peZdFHnefPNN91Pdan788UcXLKklSdmk+vXru4snda954403XOuxvlSVLcqdO7frZjdr1iz3eD1G9DhdaGXNmjUe/kIAapBQeX11cfXK6wdmfbxtFMyUKFHCtQqrcURFFx544AErWLCg+55QgHTgwAG3vTLIqk73xx9/+Peh1mhlmPbt28eLDtxgyvYqy5spUyaX9X3llVfsnnvucV3llDXyGiiPHDliv/zyi/9cVtdZ/b9WdVnPE0884c5tryudAqwXX3zRZZj0XaJudcpCAddDcISgpL7DXbt2da3B4mWKdHGkIEfBjVqDtm3b5pY3btzYKleu7L4Mf/jhB2vTpo1brsyR9qMWJ30p6qJJj1ersjcWoW3btq51SV+g3kWYBmgrQEqTJk08vQJAwuNlZcJnZ8JXnPIyuLrwUfDSrl07K1KkiMv+eBdH4bO6GpPQu3dvGzVqlDt3p0yZ4jLF3bt3d9UnvcYTnePeGCNR19lVq1a5747A/QGIHVu3bnX/qwMrxHnfAWrIVICkYknqSlehQgV3ji5cuDDM+ahudWrA2LFjh7tft25d19AZOC+hzmE9hxpAtH81jihz3KhRIxd0BTaqkiFGVAiOEO/UyhM+CFILj+Yf8uYg0ZfawYMHrVmzZvbyyy+7bjbqGqcvSKXeFfS0atXKFVEQ70tY3WhUgU6tyt5y/a6+yOo2d+rUKZdm14DuV1991Z566inX+qTqOOqep+34EgViLqJSubrQ0cVM+AAksOLUyZMn/V1eFNyowIq6xqoKnX5q7JB42+uc13msBhIVWFGVSbVE64JI4w369Onjlmksg7JQOscHDx7sBmvrO2T48OHue6Rs2bK8zcA/mFcs/DKd4yqCpP+xot4b6t3hndf6/63vAD3Wy/56jR2lS5d257EaLALP8+rVq7vrA3WL1eNKlSrlGi81NYfueyW89fhly5aF6Tqr8/322293/8+9awEaQRAVgiPECy/g0IWLWnl1waK+xV7Ljrq6KbAJLJSglmC18q5fv979/s0337iCC126dHFfrFWrVnVfunpMZP2JtVwtzxs3bnT9m/UcokDo008/dWl6HdsHH3zg0vb6nS9RIOYiOgeVEVa1KK/bi3dhtWLFCmvatKnly5fPZXeU/dEg688++8x69OjhLoI0Yau6wYoKq6grnYqptG7d2p3zGqCt818XT954Je8CTN8fCrr0/aLxDKNHj7affvrJWrRo4bLR/fv3938XAIgeb14xjxoytGzkyJGuMcNr7Lz33ntdY6cyOwp8NBYo/PxkXhCkxkp1c9V5rO8A7V/bqbeHJmvXua6GUq9wknqKaDvv/7Q3ibs3/lDnv8YVKoOk6wTtmwZPXA/BEW4Ir2Uo8MvQ+3ITXSwp86M0uXgXPoGZIwUsAwYMcHMYKKhSCl4BjS5u9GWpL1WNM/CKLQT2VV60aJFrSVL6XhdbKsLw7LPP+o9P+1egNWHCBDcXyn/+8x93jARGQNTndUQXGsrSqLVW3de8c1pUkldd4zQuUHThozFByvCoRXfMmDFuDiI1fnhd4ZQRUsOJskaB+9H3hfajiyW1Kmu/ahlWi7N37mr8gapXacyhVwZY1C1X2SSV6lf3WrVCA4gZ/X9Wrw39XxZ1gVUWV8VNVCJfDSFq0FRjhLqu6/+u/j8PGjTIX2VS1OVu6dKldvbsWXdf2+n/tnfOe9cNNWvWdA0j3gTtakjR/3avq60oY6zvAwVh3neM6HEqzKTrAv6v47p8wA126dIl36FDh9zvp06d8iVKlMg3cuRIX4ECBXzNmjXzHT582K179tlnfXXr1vVt27bN3a9QoYLbNkeOHL7atWv7hgwZ4lu7dq3vwoULbv3Fixd9Tz/9tK9s2bL+57p8+bJv7969vsqVK/vy5cvnS5Mmje+BBx7w/fjjj7zvQCw5f/68+3nu3Dn3c9iwYe5cTZEihe/dd9/1b7dr1y53/uk892h9pUqV/PdPnz7te/nll93jT5w44ZalTp3a984777jzOSI696Vly5a+/Pnz+9q0aeO77bbbfKlSpfI1aNDAfU8AuL4rV65Euk7nn9Z725w8edI3ZswYX8aMGX0dO3Z0y7Zv3+5LliyZb926df7H7d+/35chQwbf9OnT/cvOnDnj69mzp3ustu/Ro4fvyJEjbt2CBQt81apV840aNcrd9/7Hr1ixwpcnTx7ftGnT3P2dO3e674lly5ZFesx6nk8++cSXPn163+TJk/kIIFr+b1ph4F9S+lq3wNmqPWrpURUptdYeO3bMDbDWXCTKDinjo1YileJUIQWNKVKWSF3r1Jq0adMmlxFSyU61QmuMkVdVTtRSpb7N2kZdbJRNuv/++11rs6pTqUyoMk1ZsmRxJboBxA5lXlRVSqWwVSBF8waJsrAqgKLvAmWQ1KKsc1vnoMb2eHORyOzZs91YoJ49e7qqkcoMa38a+6dsb9q0ad15rEpVWhZV9zd1zVNXPH0nqKucslB6PIDoiSir4nVRDd9VVueiuqlqPI9KcWs7ldhWtTmvC6yuCTROUOe4qtEpi+s9j4ox6LpA53dgZVj1HNF9daEPzDCpep2eQ9cLyibpe0I9R8JXr/XGIGm5rj2+//57d13hlfUHrit6MRRwbfYnKl5Lj+eZZ55x2Ru1Ev/000++n3/+2XfgwAG37oUXXvCVKVPGPWb+/Pm+cuXK+Z588knfhg0bXOvRa6+95rb75ptvfEmSJHGtQF6L8l9//eXr16+f26/3nLNnz/b17t3bN3XqVH9LNoB/R63FOu/Dtywry1O6dGnfV1995Xv44Yd977//vjsX7777bt9LL73kzk+12qo1WL744gtfrly5fMuXL3f39Ri1/tavX983duxY36ZNm67JEGl5tmzZfHPmzHGt0HPnznXLI8skAYj6XI7o97Nnz7rzec2aNdc8Rj06PvroI1/fvn3d/2bvGsDL2ur8btSokTuXX3/99TAZZXn77bdd1uf48ePRemuU/dV1gXed4J3rylaFP3a+BxDbCI4QbfPmzfO1a9fOd+edd7ovrsWLF/u/IPVz4cKFvoceeshXpEgR171N28uiRYt8xYsXjzSlrWBJaXWv64sumrJnz+577rnn3P7UVcb7ktUXty6SHnzwQV+NGjXc79WrV/f98MMPfEECN7ghRIGMzkFdEKlbbPv27f3ncevWrX0NGzb0/f33374OHTr4ihYt6r4TduzY4bvjjjv8jR6DBw92XeF0wRV4wbN582bfp59+6r8we/zxx305c+Z03xXqVssFEfDvqJurxzvvEidO7A9uPGpozJs3r2sEUVf322+/3ffKK69c89jPPvvMfReULFnSdYET7zz97bffXOPm9bq4eturMVTd8BSwxaTrHxAbCI5wXTNnznT9hW+55RZ3gaJ+wAqQ1KdfLbmiTFDFihV9bdu2df2Bn3rqKRcQfffdd749e/b47rrrLl+9evV8ffr0cS1Iaj3+/fff/c+hC56JEyf6L8S+/PJLt389px6rL1bvi1Nfuv3793cXV4wlAGIusouLqAIOXUjpnGvevLlv1qxZrsFCrbgaX6QASZmjQOPGjfOVL1/efTfo+TSGsESJEq6RRA0ftWrVcttpTKEyxMosayygvi9mzJjhGkW6d+/uzwgfPXrUtWgTFAH/jv6fqrFCgYzO5UCPPfaY77777vMHJRs3bnTjeAPHDipLlC5dOtco6dH/bmWB1XDapEkT18D5/fffh9m3MsZqONVYITWK/vHHH2455zSCDcERrksXJGrpVeuRZ8mSJS54USbJ+0JV9ziPLmTUuquLIy97pG4z2k4XV1myZHHB08qVK916BVbqSqfB2J5Jkya5Agpp06Z1WSkA/8769et9b731VphlClyiujhRIKMGDZ3fauBQQQUFQ17mR8VV6tSp44IZ8bqyKhOk742hQ4f6i6+oe23mzJndPvX94WWEdbH2n//8x1elShXXEKOiKwqmAhtQAMQOZWXvv/9+Fxzp/7S6qnumTJniy5Qpk8vwehkcZYxEwU6nTp3cea0skBo0vSIN6i6nx2l7eeKJJ3wFCxb0ffDBB/59v/fee+5/vRpDc+fOfU2DSngETYgvBEe4LgUsunDp2rWrf5m6vukLTl+U+gJTK9KECRNcql0pdQU0apnS/cA+wmqNUvUYVbTRxZHGBon6MeuLVWOIPLrIUquVLqoA/HujR492Xdwi6qqiZWpFVtbWqxolCnp0EaWARpTJ0TgiXdxo7J8CnF69erkLofAXNAqa1NXOez61LqtVWtXntM/AqpFat3r1avfdAOCfi2hsoOh/6Z9//unO0S5durgxQMOHD3f/r70xgBrjo651Xrf48ePHuzGDaqjUOd+4cWMX5ITP+mg/CqI0JtAbizRo0CDXGOI1ouq7YsuWLb59+/bx9iKoERwhWtRdTi1M+gJUn2N9WerixkurK+2uYElfnCq5q6AmfKvP7t27/a3KykKpNLfXyqSxC3qsN/gSQOzQeRhVC6wupDTuRyV1NW5IXdzUbVbZXlEXt8BzXVQeX8u8IgvaRhdY3hgGrytc586dXQu1vg8CH6uAS+MWvOcAELuFUiJq5KxZs6a/5La6t6rbuhovH330UTdW2OvJoXNT43tF3eTVy+PNN9+8Zp/elBzStGlT1zNEvO7x+t7R94YXSEV03EAwYhJYRIsmRV29erWbpFETrKmkriZn1OStKrerSRSrVKliX3zxhSvRrUncZN26da70tman79u3rz322GNuhuoOHTq4ctv33Xef265IkSLusZrIEUD0ffPNNzZx4kRX0jaiyVlVftcrwasJEGfOnGnvvvuuf9tff/3V3nvvPTdhq87n7777zpXM7dq1q1tftWpVV5L79OnTYWacVwl+lcgVne8q061y/IHlgCtXruwmXvYmc9Qx6bGaaFml9rVvADGj80jnoXeuqWR1YAnuZcuWWbNmzez8+fP+7VOnTm1btmxxJbWlUKFCrhS3znedt7fddpubGH3jxo3u//T8+fPd4zSFRq5cudzErt5krKL1mlBdZfM1waomU37uuefcOq+0to6pRo0aVrhw4Wv+Bu+4gWBEcIRoqVChgvuCe/HFF61Xr142fPhwN2/AqFGj3Hwir7zyiq1cudIt0zwlmodg+vTpNnDgQDf7teYaUYCli67x48fb/v373eMAxIxmg1dDg+YImzdvnv3xxx/unAt/4aHb4cOH7eOPP7YhQ4a4+T4UHGmuscDgSPfV2KG5SdQAovmKNIu9giZd9KjBQsHPwoUL/fuWOnXq2A8//OAuwDTPieYv0feBAh4FTl9//bVrAJkwYYI1aNAgzGMBhOU1ZkRH4JxD+l86duxYe+utt9x3g6RPn959N7z66quuUUPbazudy2qoFAU8pUuXdvOCybBhw9x526JFCzdHkb5TFEzp//6AAQNccKT5iNT4qfkCNVdhqlSpXJCm74e///77msYO73yPyd8GBAMmgUW03HrrrZY5c2ZbtWqVf5kyRWpxvvvuu92Ej4MGDXItwrpg0sStmTJlskcffdRlmrwJGgHEjBoX1KCQL18+e+aZZ1wrr24jR450mZm6detG+Dg1VGiCRZ27av1VoKMGCk3QumLFCvv999+tePHiLiDaunWrjR492gVPmqRVEzsqWCpWrJjblzK8ailWdsqbkFETrOq4duzY4TK/uhDTRMxnzpxxx6SJHEXPCSAsL7vrBTnhGw4CJzINTw0dyhbrfJs2bZrt3LnTZXDmzJljr732mlWrVs39P1bDhM5/nc96jJ4rR44cbh/K9CqY0f9s0Tmsydd1vur/uoIoZXyVYdJ2+p5Q7w5dAyiz9OCDD7pJWD1p0qTxTxYbHo0iuOnEd78+3Dw0gFPjjTRvSeDkbwMGDPClSpXKDeZW/2WNI9KgSwDR41WMCz82aOnSpW7QdNWqVV0VKQ2obtasma9Vq1Zhths4cKB//J5XRlvjh1RcQVQ4wStsoopTmmhZxRmkZ8+ebryQHh9YLdJ7nPeYFClSuLLbHu1P1eWoKAf8++pxmtNHY/MiGoejKpOa68ujQiYqi6+qjmPGjHHLVBpbc/9pfjHR/2kVStE2GgOoEvlJkyb1bd261b8fVYvT+OHAIij6XXMHakyhvhu8//ORYc4hJER0q0O0qRvNvn373BiiQOpqp3EF9evXd61HamVWaxOAqAWOG/DGBh07dsx1aVMrrMYHaYzOokWLrHnz5u780mN0vpUrV85q1qzpHq8WX7UAe+OC1JqsrFKjRo3cOAF1a9VjRWMLNO5PzyG1atWyCxcuuO5xGpcgeoy6xb3zzjsuW6TusNpG4xE82t/ixYv92SUA0bdnzx7r3LmzZc2a1Z3HyvI8/vjjrhusKBukcz5lypTuf6rGEH3yySf+LI+6xKVIkcJlhUQ9NO69917X/U3Uc6Nfv36u65u6wmuMn7rNKbPr0f9pfTeoa63oOyB//vzu+0TjCZWF0njD8N9Z+n7wusqRFUJCRLc6RJu+jNVVxutW431palCnN+CSL0og+rwuNd9++63r+69xArpo0hgdBSe6+NHFy+DBg92Fy9KlS91Fi849BT4aeyQtW7a0KVOm2NmzZ+3cuXNuvzovw1/UaLmCLQU0Gleki5zatWtbmzZt3D409kjdbX788Uc3RkEDrL3gSmMKMmbMyNsL/Es6l9944w3XRU1d2HQOiho9FMDoXH3//fdd93Q1YihAevvtt10gpEYRdYdVESStU2NKhgwZ3P/ekiVLun2r0UJd4dQlTwHSpEmT7KeffnLrA8f/ZM+e3Y0L9Lru6ftGNIZQt6i+s4AELb5TVwAQqmbOnOkmPVVJ3ccff9x1nTt48KBbpzLZmltEJfK1jUrpqruqJkRWqW3NNeJZu3at6wajbnjqQqMJGBs0aBBplxeV8VW3nMWLF/u7yH3++eduIua77rrLzTsWWH4bQOzRea7zVXMJRdSNTlNeqLvbhx9+GGa55gZ86aWX3O9ff/21r1ixYr45c+aE6X5XuXJl19Xdo7mFhg0b5p6vUKFC0Z5Yle5yCGU0AQBAPNGAZg2qVjcWdadRNxoVPhG13KrVVy3FXbp0cQUW1A1GpXiVyVFRhcCsrrq5LVmyxGV2VUxBrdJehTlRJStli5RZ0n700yvFrcdqgLWOQy3Mel6vHD+A2KOskAqlqOKbusEFFlzwsjq7du2y5MmTu6pxonNVHnjgAZcV0vQZ+n5Q9TkVQPJoe2WHZs+e7e4r66v9qMKcMk/KTkeU+VEhlvDoBYJQRnAEAPFE3VEVDKkMvkcXL+peowskBUQKoNavX+8fK+B1Y9WYBFWK86hClYIhjQ3S+ARVjFN5bXXFUfc7ddX79NNP3RgidY9VqX2NeQBw4+j81rmrxgdVmIsoIFFQo65zaqgIXK5zXGOH1P1NVee0TeB3h7q96txWNzx9D3iBl7rhduzY0VWtjEj4cUVAqCM4AoB4ogIIGgCtCxzNL6T+/2oNVrltjfsRXQBp/I/mHPGoQMLx48f9k6tKkyZNXLbIa3XWXGTjxo1zY5jU2qwxRZrvyCvOoAslb4wBgBtHgZFK9HsZofCZI81BpJL6mphZvPNUxRY0Tkjnt+YyUqZIJflVptujCdaVTdI2HgVXgRPHAogawREAxCN1rdEcImPGjHEV5lRpTgFO9+7d/YGQCi2sW7cuzKTMuvhRRsmjwGr37t3+CWE1SezDDz/sKt6p4IMumnRhBSB+qTqdzvnABo/ADJGyS0899ZTrCtuwYUM3x5gaOtRdTllgbzs1eCxbtswFU4FZoIgCocCJYwFEjTMFAOKRgh91k1NJfF34aLyQWoXVtUY05kgZJk3iGtjyrC4zqlal7jOiyR5V8lsl9wEErxYtWrgA5s0337xmncrlq5FE57eCIWV6W7du7bK+6iKrKTM86nKr8z48AiHg36GjKQDEI13caNyRusQFltz2xgtovcpvq+udymlrW61TmW11wfNK6+tiq0qVKryXQJBTuXyV8u7Tp49rHGnVqpUrkqLGDZXrr1evnjvvVXK/TJky7nynCyxw4xAcAUA8UnCjQdQaP6RskS6ENGBbXWk015GqTCkbpPWBVaU0xigQ1aWAm4cKJKiy3NSpU12hFHWJ1fhCdZXTHGbKGIk3MbMaTXTT9wPnOhC3Eqmedxw/BwAgChoTpEyQxg7oIklV6FTiWxXnXnnlFZc5ApAwHT582GWTAAQHgiMAiGcaeK0B2JkyZXJzmWhcAd1ogIRP7dOBJbwlcO4jADcewREAAAAAUK0OAAAAAK6ilDcAAAAAEBwBAAAAwFVkjgAAAACA4AgAAAAAriJzBAAAAAAERwAAAABwFZkjAAAAACA4AgAAAICryBwBAAAAAMERAAAAAJjz/wADYNacGEmoBQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 850x420 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 3, figsize=(13, 4), sharex=True)\n",
    "for ax, metric, title in zip(axes, [\"auc\", \"accuracy\", \"f1\"], [\"AUC\", \"Accuracy\", \"F1\"]):\n",
    "    for model_label in MODEL_ORDER:\n",
    "        sub = summary_df[summary_df[\"model\"] == model_label].sort_values(\"scale\")\n",
    "        ax.plot(sub[\"scale_pct\"], sub[metric], marker=\"o\", linewidth=2, label=model_label)\n",
    "        for _, row in sub.iterrows():\n",
    "            ax.text(row[\"scale_pct\"], row[metric] + 0.0008, f\"{row[metric]:.4f}\", ha=\"center\", fontsize=7)\n",
    "    ax.set_xlabel(\"Training data used (%)\")\n",
    "    ax.set_ylabel(title)\n",
    "    ax.set_title(f\"Phase 4 {title} scaling\")\n",
    "    ax.grid(alpha=0.25)\n",
    "axes[0].legend(fontsize=8)\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_global_scaling_curves.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "gain_rows = []\n",
    "for model_label in MODEL_ORDER:\n",
    "    sub = summary_df[summary_df[\"model\"] == model_label].set_index(\"scale_pct\")\n",
    "    for metric in [\"auc\", \"accuracy\", \"f1\", \"balanced_accuracy\", \"macro_f1\"]:\n",
    "        gain_rows.append({\n",
    "            \"model\": model_label,\n",
    "            \"metric\": metric,\n",
    "            \"20_to_50\": sub.loc[50, metric] - sub.loc[20, metric],\n",
    "            \"50_to_100\": sub.loc[100, metric] - sub.loc[50, metric],\n",
    "            \"20_to_100\": sub.loc[100, metric] - sub.loc[20, metric],\n",
    "        })\n",
    "gain_df = pd.DataFrame(gain_rows)\n",
    "display(\n",
    "    gain_df.pivot(index=\"model\", columns=\"metric\", values=\"20_to_100\")\n",
    "    .loc[MODEL_ORDER]\n",
    "    .style.format(\"{:+.4f}\")\n",
    ")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8.5, 4.2))\n",
    "plot_gain = gain_df[gain_df[\"metric\"].isin([\"auc\", \"accuracy\", \"f1\"])]\n",
    "x = np.arange(len(MODEL_ORDER))\n",
    "width = 0.24\n",
    "for offset, metric in zip([-width, 0, width], [\"auc\", \"accuracy\", \"f1\"]):\n",
    "    vals = [plot_gain[(plot_gain[\"model\"] == model) & (plot_gain[\"metric\"] == metric)][\"20_to_100\"].iloc[0] for model in MODEL_ORDER]\n",
    "    ax.bar(x + offset, vals, width, label=metric)\n",
    "ax.axhline(0, color=\"black\", linewidth=0.8)\n",
    "ax.set_xticks(x)\n",
    "ax.set_xticklabels(MODEL_ORDER, rotation=15, ha=\"right\")\n",
    "ax.set_ylabel(\"Delta from 20% to 100%\")\n",
    "ax.set_title(\"Metric gains from more data\")\n",
    "ax.legend()\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_metric_gains.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3fde02e",
   "metadata": {},
   "source": [
    "All three model families benefit from more data. The biggest story is not a model swap; it is that scaling lifts the whole top group.\n",
    "\n",
    "From 20% to 100%, ResNet50 improves from AUC `0.9857` to `0.9950`, accuracy `0.9532` to `0.9692`, and F1 `0.9688` to `0.9794`. EfficientNet-B0 has the largest absolute gain because it starts lower at 20%, reaching AUC `0.9949`. ConvNeXt-Tiny keeps the best AUC at `0.9954`, but ResNet50 remains the best thresholded detector with the highest accuracy and F1 at 100%.\n",
    "\n",
    "This is the key Phase 4 interpretation: more data makes the top backbones converge near the same AUC ceiling, so the final choice should use operational metrics and error balance, not AUC alone.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eeb80d89",
   "metadata": {},
   "source": [
    "## 3. Fold stability\n",
    "\n",
    "The next check asks whether the 100% results are stable across folds. A model that wins only because one fold is unusually high would be less convincing than a model with consistently high fold-level AUC.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5d165add",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_470ef\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_470ef_level0_col0\" class=\"col_heading level0 col0\" >model</th>\n",
       "      <th id=\"T_470ef_level0_col1\" class=\"col_heading level0 col1\" >scale_pct</th>\n",
       "      <th id=\"T_470ef_level0_col2\" class=\"col_heading level0 col2\" colspan=\"4\">auc</th>\n",
       "      <th id=\"T_470ef_level0_col6\" class=\"col_heading level0 col6\" colspan=\"4\">accuracy</th>\n",
       "      <th id=\"T_470ef_level0_col10\" class=\"col_heading level0 col10\" colspan=\"4\">f1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"blank level1\" >&nbsp;</th>\n",
       "      <th id=\"T_470ef_level1_col0\" class=\"col_heading level1 col0\" ></th>\n",
       "      <th id=\"T_470ef_level1_col1\" class=\"col_heading level1 col1\" ></th>\n",
       "      <th id=\"T_470ef_level1_col2\" class=\"col_heading level1 col2\" >mean</th>\n",
       "      <th id=\"T_470ef_level1_col3\" class=\"col_heading level1 col3\" >std</th>\n",
       "      <th id=\"T_470ef_level1_col4\" class=\"col_heading level1 col4\" >min</th>\n",
       "      <th id=\"T_470ef_level1_col5\" class=\"col_heading level1 col5\" >max</th>\n",
       "      <th id=\"T_470ef_level1_col6\" class=\"col_heading level1 col6\" >mean</th>\n",
       "      <th id=\"T_470ef_level1_col7\" class=\"col_heading level1 col7\" >std</th>\n",
       "      <th id=\"T_470ef_level1_col8\" class=\"col_heading level1 col8\" >min</th>\n",
       "      <th id=\"T_470ef_level1_col9\" class=\"col_heading level1 col9\" >max</th>\n",
       "      <th id=\"T_470ef_level1_col10\" class=\"col_heading level1 col10\" >mean</th>\n",
       "      <th id=\"T_470ef_level1_col11\" class=\"col_heading level1 col11\" >std</th>\n",
       "      <th id=\"T_470ef_level1_col12\" class=\"col_heading level1 col12\" >min</th>\n",
       "      <th id=\"T_470ef_level1_col13\" class=\"col_heading level1 col13\" >max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_470ef_row0_col0\" class=\"data row0 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_470ef_row0_col1\" class=\"data row0 col1\" >20</td>\n",
       "      <td id=\"T_470ef_row0_col2\" class=\"data row0 col2\" >0.9868</td>\n",
       "      <td id=\"T_470ef_row0_col3\" class=\"data row0 col3\" >0.0013</td>\n",
       "      <td id=\"T_470ef_row0_col4\" class=\"data row0 col4\" >0.9851</td>\n",
       "      <td id=\"T_470ef_row0_col5\" class=\"data row0 col5\" >0.9880</td>\n",
       "      <td id=\"T_470ef_row0_col6\" class=\"data row0 col6\" >0.9473</td>\n",
       "      <td id=\"T_470ef_row0_col7\" class=\"data row0 col7\" >0.0034</td>\n",
       "      <td id=\"T_470ef_row0_col8\" class=\"data row0 col8\" >0.9415</td>\n",
       "      <td id=\"T_470ef_row0_col9\" class=\"data row0 col9\" >0.9496</td>\n",
       "      <td id=\"T_470ef_row0_col10\" class=\"data row0 col10\" >0.9650</td>\n",
       "      <td id=\"T_470ef_row0_col11\" class=\"data row0 col11\" >0.0020</td>\n",
       "      <td id=\"T_470ef_row0_col12\" class=\"data row0 col12\" >0.9616</td>\n",
       "      <td id=\"T_470ef_row0_col13\" class=\"data row0 col13\" >0.9668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_470ef_row1_col0\" class=\"data row1 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_470ef_row1_col1\" class=\"data row1 col1\" >50</td>\n",
       "      <td id=\"T_470ef_row1_col2\" class=\"data row1 col2\" >0.9926</td>\n",
       "      <td id=\"T_470ef_row1_col3\" class=\"data row1 col3\" >0.0009</td>\n",
       "      <td id=\"T_470ef_row1_col4\" class=\"data row1 col4\" >0.9917</td>\n",
       "      <td id=\"T_470ef_row1_col5\" class=\"data row1 col5\" >0.9936</td>\n",
       "      <td id=\"T_470ef_row1_col6\" class=\"data row1 col6\" >0.9607</td>\n",
       "      <td id=\"T_470ef_row1_col7\" class=\"data row1 col7\" >0.0044</td>\n",
       "      <td id=\"T_470ef_row1_col8\" class=\"data row1 col8\" >0.9551</td>\n",
       "      <td id=\"T_470ef_row1_col9\" class=\"data row1 col9\" >0.9661</td>\n",
       "      <td id=\"T_470ef_row1_col10\" class=\"data row1 col10\" >0.9735</td>\n",
       "      <td id=\"T_470ef_row1_col11\" class=\"data row1 col11\" >0.0031</td>\n",
       "      <td id=\"T_470ef_row1_col12\" class=\"data row1 col12\" >0.9696</td>\n",
       "      <td id=\"T_470ef_row1_col13\" class=\"data row1 col13\" >0.9773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_470ef_row2_col0\" class=\"data row2 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_470ef_row2_col1\" class=\"data row2 col1\" >100</td>\n",
       "      <td id=\"T_470ef_row2_col2\" class=\"data row2 col2\" >0.9954</td>\n",
       "      <td id=\"T_470ef_row2_col3\" class=\"data row2 col3\" >0.0002</td>\n",
       "      <td id=\"T_470ef_row2_col4\" class=\"data row2 col4\" >0.9951</td>\n",
       "      <td id=\"T_470ef_row2_col5\" class=\"data row2 col5\" >0.9956</td>\n",
       "      <td id=\"T_470ef_row2_col6\" class=\"data row2 col6\" >0.9679</td>\n",
       "      <td id=\"T_470ef_row2_col7\" class=\"data row2 col7\" >0.0051</td>\n",
       "      <td id=\"T_470ef_row2_col8\" class=\"data row2 col8\" >0.9595</td>\n",
       "      <td id=\"T_470ef_row2_col9\" class=\"data row2 col9\" >0.9724</td>\n",
       "      <td id=\"T_470ef_row2_col10\" class=\"data row2 col10\" >0.9784</td>\n",
       "      <td id=\"T_470ef_row2_col11\" class=\"data row2 col11\" >0.0036</td>\n",
       "      <td id=\"T_470ef_row2_col12\" class=\"data row2 col12\" >0.9725</td>\n",
       "      <td id=\"T_470ef_row2_col13\" class=\"data row2 col13\" >0.9816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_470ef_row3_col0\" class=\"data row3 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_470ef_row3_col1\" class=\"data row3 col1\" >20</td>\n",
       "      <td id=\"T_470ef_row3_col2\" class=\"data row3 col2\" >0.9845</td>\n",
       "      <td id=\"T_470ef_row3_col3\" class=\"data row3 col3\" >0.0025</td>\n",
       "      <td id=\"T_470ef_row3_col4\" class=\"data row3 col4\" >0.9813</td>\n",
       "      <td id=\"T_470ef_row3_col5\" class=\"data row3 col5\" >0.9880</td>\n",
       "      <td id=\"T_470ef_row3_col6\" class=\"data row3 col6\" >0.9450</td>\n",
       "      <td id=\"T_470ef_row3_col7\" class=\"data row3 col7\" >0.0025</td>\n",
       "      <td id=\"T_470ef_row3_col8\" class=\"data row3 col8\" >0.9413</td>\n",
       "      <td id=\"T_470ef_row3_col9\" class=\"data row3 col9\" >0.9479</td>\n",
       "      <td id=\"T_470ef_row3_col10\" class=\"data row3 col10\" >0.9628</td>\n",
       "      <td id=\"T_470ef_row3_col11\" class=\"data row3 col11\" >0.0016</td>\n",
       "      <td id=\"T_470ef_row3_col12\" class=\"data row3 col12\" >0.9603</td>\n",
       "      <td id=\"T_470ef_row3_col13\" class=\"data row3 col13\" >0.9647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_470ef_row4_col0\" class=\"data row4 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_470ef_row4_col1\" class=\"data row4 col1\" >50</td>\n",
       "      <td id=\"T_470ef_row4_col2\" class=\"data row4 col2\" >0.9911</td>\n",
       "      <td id=\"T_470ef_row4_col3\" class=\"data row4 col3\" >0.0009</td>\n",
       "      <td id=\"T_470ef_row4_col4\" class=\"data row4 col4\" >0.9895</td>\n",
       "      <td id=\"T_470ef_row4_col5\" class=\"data row4 col5\" >0.9918</td>\n",
       "      <td id=\"T_470ef_row4_col6\" class=\"data row4 col6\" >0.9576</td>\n",
       "      <td id=\"T_470ef_row4_col7\" class=\"data row4 col7\" >0.0064</td>\n",
       "      <td id=\"T_470ef_row4_col8\" class=\"data row4 col8\" >0.9479</td>\n",
       "      <td id=\"T_470ef_row4_col9\" class=\"data row4 col9\" >0.9644</td>\n",
       "      <td id=\"T_470ef_row4_col10\" class=\"data row4 col10\" >0.9714</td>\n",
       "      <td id=\"T_470ef_row4_col11\" class=\"data row4 col11\" >0.0045</td>\n",
       "      <td id=\"T_470ef_row4_col12\" class=\"data row4 col12\" >0.9646</td>\n",
       "      <td id=\"T_470ef_row4_col13\" class=\"data row4 col13\" >0.9762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_470ef_row5_col0\" class=\"data row5 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_470ef_row5_col1\" class=\"data row5 col1\" >100</td>\n",
       "      <td id=\"T_470ef_row5_col2\" class=\"data row5 col2\" >0.9949</td>\n",
       "      <td id=\"T_470ef_row5_col3\" class=\"data row5 col3\" >0.0008</td>\n",
       "      <td id=\"T_470ef_row5_col4\" class=\"data row5 col4\" >0.9939</td>\n",
       "      <td id=\"T_470ef_row5_col5\" class=\"data row5 col5\" >0.9958</td>\n",
       "      <td id=\"T_470ef_row5_col6\" class=\"data row5 col6\" >0.9684</td>\n",
       "      <td id=\"T_470ef_row5_col7\" class=\"data row5 col7\" >0.0048</td>\n",
       "      <td id=\"T_470ef_row5_col8\" class=\"data row5 col8\" >0.9617</td>\n",
       "      <td id=\"T_470ef_row5_col9\" class=\"data row5 col9\" >0.9733</td>\n",
       "      <td id=\"T_470ef_row5_col10\" class=\"data row5 col10\" >0.9787</td>\n",
       "      <td id=\"T_470ef_row5_col11\" class=\"data row5 col11\" >0.0033</td>\n",
       "      <td id=\"T_470ef_row5_col12\" class=\"data row5 col12\" >0.9740</td>\n",
       "      <td id=\"T_470ef_row5_col13\" class=\"data row5 col13\" >0.9821</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_470ef_row6_col0\" class=\"data row6 col0\" >ResNet50</td>\n",
       "      <td id=\"T_470ef_row6_col1\" class=\"data row6 col1\" >20</td>\n",
       "      <td id=\"T_470ef_row6_col2\" class=\"data row6 col2\" >0.9857</td>\n",
       "      <td id=\"T_470ef_row6_col3\" class=\"data row6 col3\" >0.0011</td>\n",
       "      <td id=\"T_470ef_row6_col4\" class=\"data row6 col4\" >0.9845</td>\n",
       "      <td id=\"T_470ef_row6_col5\" class=\"data row6 col5\" >0.9874</td>\n",
       "      <td id=\"T_470ef_row6_col6\" class=\"data row6 col6\" >0.9532</td>\n",
       "      <td id=\"T_470ef_row6_col7\" class=\"data row6 col7\" >0.0040</td>\n",
       "      <td id=\"T_470ef_row6_col8\" class=\"data row6 col8\" >0.9479</td>\n",
       "      <td id=\"T_470ef_row6_col9\" class=\"data row6 col9\" >0.9577</td>\n",
       "      <td id=\"T_470ef_row6_col10\" class=\"data row6 col10\" >0.9688</td>\n",
       "      <td id=\"T_470ef_row6_col11\" class=\"data row6 col11\" >0.0028</td>\n",
       "      <td id=\"T_470ef_row6_col12\" class=\"data row6 col12\" >0.9651</td>\n",
       "      <td id=\"T_470ef_row6_col13\" class=\"data row6 col13\" >0.9720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_470ef_row7_col0\" class=\"data row7 col0\" >ResNet50</td>\n",
       "      <td id=\"T_470ef_row7_col1\" class=\"data row7 col1\" >50</td>\n",
       "      <td id=\"T_470ef_row7_col2\" class=\"data row7 col2\" >0.9921</td>\n",
       "      <td id=\"T_470ef_row7_col3\" class=\"data row7 col3\" >0.0008</td>\n",
       "      <td id=\"T_470ef_row7_col4\" class=\"data row7 col4\" >0.9911</td>\n",
       "      <td id=\"T_470ef_row7_col5\" class=\"data row7 col5\" >0.9929</td>\n",
       "      <td id=\"T_470ef_row7_col6\" class=\"data row7 col6\" >0.9607</td>\n",
       "      <td id=\"T_470ef_row7_col7\" class=\"data row7 col7\" >0.0030</td>\n",
       "      <td id=\"T_470ef_row7_col8\" class=\"data row7 col8\" >0.9563</td>\n",
       "      <td id=\"T_470ef_row7_col9\" class=\"data row7 col9\" >0.9634</td>\n",
       "      <td id=\"T_470ef_row7_col10\" class=\"data row7 col10\" >0.9736</td>\n",
       "      <td id=\"T_470ef_row7_col11\" class=\"data row7 col11\" >0.0021</td>\n",
       "      <td id=\"T_470ef_row7_col12\" class=\"data row7 col12\" >0.9704</td>\n",
       "      <td id=\"T_470ef_row7_col13\" class=\"data row7 col13\" >0.9755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_470ef_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_470ef_row8_col0\" class=\"data row8 col0\" >ResNet50</td>\n",
       "      <td id=\"T_470ef_row8_col1\" class=\"data row8 col1\" >100</td>\n",
       "      <td id=\"T_470ef_row8_col2\" class=\"data row8 col2\" >0.9950</td>\n",
       "      <td id=\"T_470ef_row8_col3\" class=\"data row8 col3\" >0.0004</td>\n",
       "      <td id=\"T_470ef_row8_col4\" class=\"data row8 col4\" >0.9946</td>\n",
       "      <td id=\"T_470ef_row8_col5\" class=\"data row8 col5\" >0.9956</td>\n",
       "      <td id=\"T_470ef_row8_col6\" class=\"data row8 col6\" >0.9692</td>\n",
       "      <td id=\"T_470ef_row8_col7\" class=\"data row8 col7\" >0.0018</td>\n",
       "      <td id=\"T_470ef_row8_col8\" class=\"data row8 col8\" >0.9666</td>\n",
       "      <td id=\"T_470ef_row8_col9\" class=\"data row8 col9\" >0.9714</td>\n",
       "      <td id=\"T_470ef_row8_col10\" class=\"data row8 col10\" >0.9794</td>\n",
       "      <td id=\"T_470ef_row8_col11\" class=\"data row8 col11\" >0.0012</td>\n",
       "      <td id=\"T_470ef_row8_col12\" class=\"data row8 col12\" >0.9776</td>\n",
       "      <td id=\"T_470ef_row8_col13\" class=\"data row8 col13\" >0.9809</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1d2ad4fc7a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def fold_rows(model_label: str, scale: float, run_name: str) -> list[dict]:\n",
    "    rows = []\n",
    "    for fold in load_run(run_name).get(\"fold_results\", []):\n",
    "        metrics = fold.get(\"test_metrics\", {})\n",
    "        rows.append({\n",
    "            \"model\": model_label,\n",
    "            \"scale_pct\": int(scale * 100),\n",
    "            \"run\": run_name,\n",
    "            \"fold\": fold.get(\"fold\"),\n",
    "            \"auc\": metrics.get(\"auc_roc\", np.nan),\n",
    "            \"accuracy\": metrics.get(\"accuracy\", np.nan),\n",
    "            \"f1\": metrics.get(\"f1\", np.nan),\n",
    "        })\n",
    "    return rows\n",
    "\n",
    "\n",
    "fold_df = pd.DataFrame([\n",
    "    row\n",
    "    for model_label in MODEL_ORDER\n",
    "    for scale in SCALE_ORDER\n",
    "    for row in fold_rows(model_label, scale, MODEL_RUNS[model_label][scale])\n",
    "])\n",
    "\n",
    "fig, axes = plt.subplots(1, 3, figsize=(13, 4), sharey=True)\n",
    "for ax, model_label in zip(axes, MODEL_ORDER):\n",
    "    sub_model = fold_df[fold_df[\"model\"] == model_label]\n",
    "    for pos, scale_pct in enumerate([20, 50, 100]):\n",
    "        vals = sub_model[sub_model[\"scale_pct\"] == scale_pct][\"auc\"].to_numpy()\n",
    "        ax.scatter(np.full_like(vals, pos, dtype=float), vals, s=35, alpha=0.75)\n",
    "        ax.plot([pos - 0.18, pos + 0.18], [np.nanmean(vals), np.nanmean(vals)], color=\"black\", linewidth=2)\n",
    "    ax.set_xticks([0, 1, 2])\n",
    "    ax.set_xticklabels([\"20%\", \"50%\", \"100%\"])\n",
    "    ax.set_title(model_label)\n",
    "    ax.set_xlabel(\"training data\")\n",
    "    ax.grid(axis=\"y\", alpha=0.2)\n",
    "axes[0].set_ylabel(\"Fold held-out AUC\")\n",
    "fig.suptitle(\"Fold-level AUC as data increases\", fontsize=13)\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_fold_auc_stability.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "fig, axes = plt.subplots(1, 3, figsize=(13, 4), sharey=True)\n",
    "for ax, model_label in zip(axes, MODEL_ORDER):\n",
    "    sub_model = fold_df[fold_df[\"model\"] == model_label]\n",
    "    for pos, scale_pct in enumerate([20, 50, 100]):\n",
    "        vals = sub_model[sub_model[\"scale_pct\"] == scale_pct][\"f1\"].to_numpy()\n",
    "        ax.scatter(np.full_like(vals, pos, dtype=float), vals, s=35, alpha=0.75)\n",
    "        ax.plot([pos - 0.18, pos + 0.18], [np.nanmean(vals), np.nanmean(vals)], color=\"black\", linewidth=2)\n",
    "    ax.set_xticks([0, 1, 2])\n",
    "    ax.set_xticklabels([\"20%\", \"50%\", \"100%\"])\n",
    "    ax.set_title(model_label)\n",
    "    ax.set_xlabel(\"training data\")\n",
    "    ax.grid(axis=\"y\", alpha=0.2)\n",
    "axes[0].set_ylabel(\"Fold held-out F1\")\n",
    "fig.suptitle(\"Fold-level F1 as data increases\", fontsize=13)\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_fold_f1_stability.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "fold_summary = (\n",
    "    fold_df.groupby([\"model\", \"scale_pct\"])[[\"auc\", \"accuracy\", \"f1\"]]\n",
    "    .agg([\"mean\", \"std\", \"min\", \"max\"])\n",
    "    .reset_index()\n",
    ")\n",
    "display(fold_summary.style.format(precision=4))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f18edbea",
   "metadata": {},
   "source": [
    "The 100% runs are not fragile. All three backbones have fold AUCs near `0.995`, and ConvNeXt-Tiny is especially tight fold-to-fold. ResNet50 is also stable, with fold AUCs from `0.9946` to `0.9956`. The 50% step already removes much of the variability seen at 20%, and the 100% step makes the top group very consistent.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2bafcdea",
   "metadata": {},
   "source": [
    "## 4. Source-wise behavior\n",
    "\n",
    "The earlier notebooks showed that source behavior matters. A high global AUC is less convincing if one fake source remains much weaker than the others. Phase 4 should therefore answer whether scaling improves the weaker source pairs, especially `insight`.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f92f8638",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_e4999\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_e4999_level0_col0\" class=\"col_heading level0 col0\" >model</th>\n",
       "      <th id=\"T_e4999_level0_col1\" class=\"col_heading level0 col1\" >scale_pct</th>\n",
       "      <th id=\"T_e4999_level0_col2\" class=\"col_heading level0 col2\" >run</th>\n",
       "      <th id=\"T_e4999_level0_col3\" class=\"col_heading level0 col3\" >source</th>\n",
       "      <th id=\"T_e4999_level0_col4\" class=\"col_heading level0 col4\" >accuracy</th>\n",
       "      <th id=\"T_e4999_level0_col5\" class=\"col_heading level0 col5\" >detection_rate</th>\n",
       "      <th id=\"T_e4999_level0_col6\" class=\"col_heading level0 col6\" >false_alarm_rate</th>\n",
       "      <th id=\"T_e4999_level0_col7\" class=\"col_heading level0 col7\" >pairwise_auc</th>\n",
       "      <th id=\"T_e4999_level0_col8\" class=\"col_heading level0 col8\" >pairwise_f1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row0\" class=\"row_heading level0 row0\" >27</th>\n",
       "      <td id=\"T_e4999_row0_col0\" class=\"data row0 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row0_col1\" class=\"data row0 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row0_col2\" class=\"data row0 col2\" >p3_convnext_tiny</td>\n",
       "      <td id=\"T_e4999_row0_col3\" class=\"data row0 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row0_col4\" class=\"data row0 col4\" >0.8838</td>\n",
       "      <td id=\"T_e4999_row0_col5\" class=\"data row0 col5\" ></td>\n",
       "      <td id=\"T_e4999_row0_col6\" class=\"data row0 col6\" >0.1162</td>\n",
       "      <td id=\"T_e4999_row0_col7\" class=\"data row0 col7\" ></td>\n",
       "      <td id=\"T_e4999_row0_col8\" class=\"data row0 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row1\" class=\"row_heading level0 row1\" >24</th>\n",
       "      <td id=\"T_e4999_row1_col0\" class=\"data row1 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row1_col1\" class=\"data row1 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row1_col2\" class=\"data row1 col2\" >p3_convnext_tiny</td>\n",
       "      <td id=\"T_e4999_row1_col3\" class=\"data row1 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row1_col4\" class=\"data row1 col4\" >0.9753</td>\n",
       "      <td id=\"T_e4999_row1_col5\" class=\"data row1 col5\" >0.9753</td>\n",
       "      <td id=\"T_e4999_row1_col6\" class=\"data row1 col6\" ></td>\n",
       "      <td id=\"T_e4999_row1_col7\" class=\"data row1 col7\" >0.9884</td>\n",
       "      <td id=\"T_e4999_row1_col8\" class=\"data row1 col8\" >0.9329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row2\" class=\"row_heading level0 row2\" >25</th>\n",
       "      <td id=\"T_e4999_row2_col0\" class=\"data row2 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row2_col1\" class=\"data row2 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row2_col2\" class=\"data row2 col2\" >p3_convnext_tiny</td>\n",
       "      <td id=\"T_e4999_row2_col3\" class=\"data row2 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row2_col4\" class=\"data row2 col4\" >0.9452</td>\n",
       "      <td id=\"T_e4999_row2_col5\" class=\"data row2 col5\" >0.9452</td>\n",
       "      <td id=\"T_e4999_row2_col6\" class=\"data row2 col6\" ></td>\n",
       "      <td id=\"T_e4999_row2_col7\" class=\"data row2 col7\" >0.9791</td>\n",
       "      <td id=\"T_e4999_row2_col8\" class=\"data row2 col8\" >0.9172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row3\" class=\"row_heading level0 row3\" >26</th>\n",
       "      <td id=\"T_e4999_row3_col0\" class=\"data row3 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row3_col1\" class=\"data row3 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row3_col2\" class=\"data row3 col2\" >p3_convnext_tiny</td>\n",
       "      <td id=\"T_e4999_row3_col3\" class=\"data row3 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row3_col4\" class=\"data row3 col4\" >0.9850</td>\n",
       "      <td id=\"T_e4999_row3_col5\" class=\"data row3 col5\" >0.9850</td>\n",
       "      <td id=\"T_e4999_row3_col6\" class=\"data row3 col6\" ></td>\n",
       "      <td id=\"T_e4999_row3_col7\" class=\"data row3 col7\" >0.9927</td>\n",
       "      <td id=\"T_e4999_row3_col8\" class=\"data row3 col8\" >0.9378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row4\" class=\"row_heading level0 row4\" >15</th>\n",
       "      <td id=\"T_e4999_row4_col0\" class=\"data row4 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row4_col1\" class=\"data row4 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row4_col2\" class=\"data row4 col2\" >p3_efficientnet_b0</td>\n",
       "      <td id=\"T_e4999_row4_col3\" class=\"data row4 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row4_col4\" class=\"data row4 col4\" >0.9292</td>\n",
       "      <td id=\"T_e4999_row4_col5\" class=\"data row4 col5\" ></td>\n",
       "      <td id=\"T_e4999_row4_col6\" class=\"data row4 col6\" >0.0708</td>\n",
       "      <td id=\"T_e4999_row4_col7\" class=\"data row4 col7\" ></td>\n",
       "      <td id=\"T_e4999_row4_col8\" class=\"data row4 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row5\" class=\"row_heading level0 row5\" >12</th>\n",
       "      <td id=\"T_e4999_row5_col0\" class=\"data row5 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row5_col1\" class=\"data row5 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row5_col2\" class=\"data row5 col2\" >p3_efficientnet_b0</td>\n",
       "      <td id=\"T_e4999_row5_col3\" class=\"data row5 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row5_col4\" class=\"data row5 col4\" >0.9470</td>\n",
       "      <td id=\"T_e4999_row5_col5\" class=\"data row5 col5\" >0.9470</td>\n",
       "      <td id=\"T_e4999_row5_col6\" class=\"data row5 col6\" ></td>\n",
       "      <td id=\"T_e4999_row5_col7\" class=\"data row5 col7\" >0.9838</td>\n",
       "      <td id=\"T_e4999_row5_col8\" class=\"data row5 col8\" >0.9386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row6\" class=\"row_heading level0 row6\" >13</th>\n",
       "      <td id=\"T_e4999_row6_col0\" class=\"data row6 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row6_col1\" class=\"data row6 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row6_col2\" class=\"data row6 col2\" >p3_efficientnet_b0</td>\n",
       "      <td id=\"T_e4999_row6_col3\" class=\"data row6 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row6_col4\" class=\"data row6 col4\" >0.9257</td>\n",
       "      <td id=\"T_e4999_row6_col5\" class=\"data row6 col5\" >0.9257</td>\n",
       "      <td id=\"T_e4999_row6_col6\" class=\"data row6 col6\" ></td>\n",
       "      <td id=\"T_e4999_row6_col7\" class=\"data row6 col7\" >0.9776</td>\n",
       "      <td id=\"T_e4999_row6_col8\" class=\"data row6 col8\" >0.9273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row7\" class=\"row_heading level0 row7\" >14</th>\n",
       "      <td id=\"T_e4999_row7_col0\" class=\"data row7 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row7_col1\" class=\"data row7 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row7_col2\" class=\"data row7 col2\" >p3_efficientnet_b0</td>\n",
       "      <td id=\"T_e4999_row7_col3\" class=\"data row7 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row7_col4\" class=\"data row7 col4\" >0.9782</td>\n",
       "      <td id=\"T_e4999_row7_col5\" class=\"data row7 col5\" >0.9782</td>\n",
       "      <td id=\"T_e4999_row7_col6\" class=\"data row7 col6\" ></td>\n",
       "      <td id=\"T_e4999_row7_col7\" class=\"data row7 col7\" >0.9921</td>\n",
       "      <td id=\"T_e4999_row7_col8\" class=\"data row7 col8\" >0.9548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row8\" class=\"row_heading level0 row8\" >3</th>\n",
       "      <td id=\"T_e4999_row8_col0\" class=\"data row8 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row8_col1\" class=\"data row8 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row8_col2\" class=\"data row8 col2\" >p3_resnet50</td>\n",
       "      <td id=\"T_e4999_row8_col3\" class=\"data row8 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row8_col4\" class=\"data row8 col4\" >0.9092</td>\n",
       "      <td id=\"T_e4999_row8_col5\" class=\"data row8 col5\" ></td>\n",
       "      <td id=\"T_e4999_row8_col6\" class=\"data row8 col6\" >0.0908</td>\n",
       "      <td id=\"T_e4999_row8_col7\" class=\"data row8 col7\" ></td>\n",
       "      <td id=\"T_e4999_row8_col8\" class=\"data row8 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row9\" class=\"row_heading level0 row9\" >0</th>\n",
       "      <td id=\"T_e4999_row9_col0\" class=\"data row9 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row9_col1\" class=\"data row9 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row9_col2\" class=\"data row9 col2\" >p3_resnet50</td>\n",
       "      <td id=\"T_e4999_row9_col3\" class=\"data row9 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row9_col4\" class=\"data row9 col4\" >0.9733</td>\n",
       "      <td id=\"T_e4999_row9_col5\" class=\"data row9 col5\" >0.9733</td>\n",
       "      <td id=\"T_e4999_row9_col6\" class=\"data row9 col6\" ></td>\n",
       "      <td id=\"T_e4999_row9_col7\" class=\"data row9 col7\" >0.9879</td>\n",
       "      <td id=\"T_e4999_row9_col8\" class=\"data row9 col8\" >0.9431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row10\" class=\"row_heading level0 row10\" >1</th>\n",
       "      <td id=\"T_e4999_row10_col0\" class=\"data row10 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row10_col1\" class=\"data row10 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row10_col2\" class=\"data row10 col2\" >p3_resnet50</td>\n",
       "      <td id=\"T_e4999_row10_col3\" class=\"data row10 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row10_col4\" class=\"data row10 col4\" >0.9443</td>\n",
       "      <td id=\"T_e4999_row10_col5\" class=\"data row10 col5\" >0.9443</td>\n",
       "      <td id=\"T_e4999_row10_col6\" class=\"data row10 col6\" ></td>\n",
       "      <td id=\"T_e4999_row10_col7\" class=\"data row10 col7\" >0.9792</td>\n",
       "      <td id=\"T_e4999_row10_col8\" class=\"data row10 col8\" >0.9280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row11\" class=\"row_heading level0 row11\" >2</th>\n",
       "      <td id=\"T_e4999_row11_col0\" class=\"data row11 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row11_col1\" class=\"data row11 col1\" >20</td>\n",
       "      <td id=\"T_e4999_row11_col2\" class=\"data row11 col2\" >p3_resnet50</td>\n",
       "      <td id=\"T_e4999_row11_col3\" class=\"data row11 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row11_col4\" class=\"data row11 col4\" >0.9860</td>\n",
       "      <td id=\"T_e4999_row11_col5\" class=\"data row11 col5\" >0.9860</td>\n",
       "      <td id=\"T_e4999_row11_col6\" class=\"data row11 col6\" ></td>\n",
       "      <td id=\"T_e4999_row11_col7\" class=\"data row11 col7\" >0.9900</td>\n",
       "      <td id=\"T_e4999_row11_col8\" class=\"data row11 col8\" >0.9495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row12\" class=\"row_heading level0 row12\" >31</th>\n",
       "      <td id=\"T_e4999_row12_col0\" class=\"data row12 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row12_col1\" class=\"data row12 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row12_col2\" class=\"data row12 col2\" >p4_convnext_tiny_50pct</td>\n",
       "      <td id=\"T_e4999_row12_col3\" class=\"data row12 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row12_col4\" class=\"data row12 col4\" >0.9473</td>\n",
       "      <td id=\"T_e4999_row12_col5\" class=\"data row12 col5\" ></td>\n",
       "      <td id=\"T_e4999_row12_col6\" class=\"data row12 col6\" >0.0527</td>\n",
       "      <td id=\"T_e4999_row12_col7\" class=\"data row12 col7\" ></td>\n",
       "      <td id=\"T_e4999_row12_col8\" class=\"data row12 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row13\" class=\"row_heading level0 row13\" >28</th>\n",
       "      <td id=\"T_e4999_row13_col0\" class=\"data row13 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row13_col1\" class=\"data row13 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row13_col2\" class=\"data row13 col2\" >p4_convnext_tiny_50pct</td>\n",
       "      <td id=\"T_e4999_row13_col3\" class=\"data row13 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row13_col4\" class=\"data row13 col4\" >0.9723</td>\n",
       "      <td id=\"T_e4999_row13_col5\" class=\"data row13 col5\" >0.9723</td>\n",
       "      <td id=\"T_e4999_row13_col6\" class=\"data row13 col6\" ></td>\n",
       "      <td id=\"T_e4999_row13_col7\" class=\"data row13 col7\" >0.9938</td>\n",
       "      <td id=\"T_e4999_row13_col8\" class=\"data row13 col8\" >0.9603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row14\" class=\"row_heading level0 row14\" >29</th>\n",
       "      <td id=\"T_e4999_row14_col0\" class=\"data row14 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row14_col1\" class=\"data row14 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row14_col2\" class=\"data row14 col2\" >p4_convnext_tiny_50pct</td>\n",
       "      <td id=\"T_e4999_row14_col3\" class=\"data row14 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row14_col4\" class=\"data row14 col4\" >0.9349</td>\n",
       "      <td id=\"T_e4999_row14_col5\" class=\"data row14 col5\" >0.9349</td>\n",
       "      <td id=\"T_e4999_row14_col6\" class=\"data row14 col6\" ></td>\n",
       "      <td id=\"T_e4999_row14_col7\" class=\"data row14 col7\" >0.9873</td>\n",
       "      <td id=\"T_e4999_row14_col8\" class=\"data row14 col8\" >0.9407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row15\" class=\"row_heading level0 row15\" >30</th>\n",
       "      <td id=\"T_e4999_row15_col0\" class=\"data row15 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row15_col1\" class=\"data row15 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row15_col2\" class=\"data row15 col2\" >p4_convnext_tiny_50pct</td>\n",
       "      <td id=\"T_e4999_row15_col3\" class=\"data row15 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row15_col4\" class=\"data row15 col4\" >0.9881</td>\n",
       "      <td id=\"T_e4999_row15_col5\" class=\"data row15 col5\" >0.9881</td>\n",
       "      <td id=\"T_e4999_row15_col6\" class=\"data row15 col6\" ></td>\n",
       "      <td id=\"T_e4999_row15_col7\" class=\"data row15 col7\" >0.9968</td>\n",
       "      <td id=\"T_e4999_row15_col8\" class=\"data row15 col8\" >0.9684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row16\" class=\"row_heading level0 row16\" >19</th>\n",
       "      <td id=\"T_e4999_row16_col0\" class=\"data row16 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row16_col1\" class=\"data row16 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row16_col2\" class=\"data row16 col2\" >p4_efficientnet_b0_50pct</td>\n",
       "      <td id=\"T_e4999_row16_col3\" class=\"data row16 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row16_col4\" class=\"data row16 col4\" >0.9471</td>\n",
       "      <td id=\"T_e4999_row16_col5\" class=\"data row16 col5\" ></td>\n",
       "      <td id=\"T_e4999_row16_col6\" class=\"data row16 col6\" >0.0529</td>\n",
       "      <td id=\"T_e4999_row16_col7\" class=\"data row16 col7\" ></td>\n",
       "      <td id=\"T_e4999_row16_col8\" class=\"data row16 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row17\" class=\"row_heading level0 row17\" >16</th>\n",
       "      <td id=\"T_e4999_row17_col0\" class=\"data row17 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row17_col1\" class=\"data row17 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row17_col2\" class=\"data row17 col2\" >p4_efficientnet_b0_50pct</td>\n",
       "      <td id=\"T_e4999_row17_col3\" class=\"data row17 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row17_col4\" class=\"data row17 col4\" >0.9603</td>\n",
       "      <td id=\"T_e4999_row17_col5\" class=\"data row17 col5\" >0.9603</td>\n",
       "      <td id=\"T_e4999_row17_col6\" class=\"data row17 col6\" ></td>\n",
       "      <td id=\"T_e4999_row17_col7\" class=\"data row17 col7\" >0.9914</td>\n",
       "      <td id=\"T_e4999_row17_col8\" class=\"data row17 col8\" >0.9540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row18\" class=\"row_heading level0 row18\" >17</th>\n",
       "      <td id=\"T_e4999_row18_col0\" class=\"data row18 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row18_col1\" class=\"data row18 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row18_col2\" class=\"data row18 col2\" >p4_efficientnet_b0_50pct</td>\n",
       "      <td id=\"T_e4999_row18_col3\" class=\"data row18 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row18_col4\" class=\"data row18 col4\" >0.9409</td>\n",
       "      <td id=\"T_e4999_row18_col5\" class=\"data row18 col5\" >0.9409</td>\n",
       "      <td id=\"T_e4999_row18_col6\" class=\"data row18 col6\" ></td>\n",
       "      <td id=\"T_e4999_row18_col7\" class=\"data row18 col7\" >0.9866</td>\n",
       "      <td id=\"T_e4999_row18_col8\" class=\"data row18 col8\" >0.9438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row19\" class=\"row_heading level0 row19\" >18</th>\n",
       "      <td id=\"T_e4999_row19_col0\" class=\"data row19 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row19_col1\" class=\"data row19 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row19_col2\" class=\"data row19 col2\" >p4_efficientnet_b0_50pct</td>\n",
       "      <td id=\"T_e4999_row19_col3\" class=\"data row19 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row19_col4\" class=\"data row19 col4\" >0.9823</td>\n",
       "      <td id=\"T_e4999_row19_col5\" class=\"data row19 col5\" >0.9823</td>\n",
       "      <td id=\"T_e4999_row19_col6\" class=\"data row19 col6\" ></td>\n",
       "      <td id=\"T_e4999_row19_col7\" class=\"data row19 col7\" >0.9954</td>\n",
       "      <td id=\"T_e4999_row19_col8\" class=\"data row19 col8\" >0.9653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row20\" class=\"row_heading level0 row20\" >7</th>\n",
       "      <td id=\"T_e4999_row20_col0\" class=\"data row20 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row20_col1\" class=\"data row20 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row20_col2\" class=\"data row20 col2\" >p4_resnet50_50pct</td>\n",
       "      <td id=\"T_e4999_row20_col3\" class=\"data row20 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row20_col4\" class=\"data row20 col4\" >0.9395</td>\n",
       "      <td id=\"T_e4999_row20_col5\" class=\"data row20 col5\" ></td>\n",
       "      <td id=\"T_e4999_row20_col6\" class=\"data row20 col6\" >0.0605</td>\n",
       "      <td id=\"T_e4999_row20_col7\" class=\"data row20 col7\" ></td>\n",
       "      <td id=\"T_e4999_row20_col8\" class=\"data row20 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row21\" class=\"row_heading level0 row21\" >4</th>\n",
       "      <td id=\"T_e4999_row21_col0\" class=\"data row21 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row21_col1\" class=\"data row21 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row21_col2\" class=\"data row21 col2\" >p4_resnet50_50pct</td>\n",
       "      <td id=\"T_e4999_row21_col3\" class=\"data row21 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row21_col4\" class=\"data row21 col4\" >0.9742</td>\n",
       "      <td id=\"T_e4999_row21_col5\" class=\"data row21 col5\" >0.9742</td>\n",
       "      <td id=\"T_e4999_row21_col6\" class=\"data row21 col6\" ></td>\n",
       "      <td id=\"T_e4999_row21_col7\" class=\"data row21 col7\" >0.9935</td>\n",
       "      <td id=\"T_e4999_row21_col8\" class=\"data row21 col8\" >0.9576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row22\" class=\"row_heading level0 row22\" >5</th>\n",
       "      <td id=\"T_e4999_row22_col0\" class=\"data row22 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row22_col1\" class=\"data row22 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row22_col2\" class=\"data row22 col2\" >p4_resnet50_50pct</td>\n",
       "      <td id=\"T_e4999_row22_col3\" class=\"data row22 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row22_col4\" class=\"data row22 col4\" >0.9430</td>\n",
       "      <td id=\"T_e4999_row22_col5\" class=\"data row22 col5\" >0.9430</td>\n",
       "      <td id=\"T_e4999_row22_col6\" class=\"data row22 col6\" ></td>\n",
       "      <td id=\"T_e4999_row22_col7\" class=\"data row22 col7\" >0.9869</td>\n",
       "      <td id=\"T_e4999_row22_col8\" class=\"data row22 col8\" >0.9414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row23\" class=\"row_heading level0 row23\" >6</th>\n",
       "      <td id=\"T_e4999_row23_col0\" class=\"data row23 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row23_col1\" class=\"data row23 col1\" >50</td>\n",
       "      <td id=\"T_e4999_row23_col2\" class=\"data row23 col2\" >p4_resnet50_50pct</td>\n",
       "      <td id=\"T_e4999_row23_col3\" class=\"data row23 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row23_col4\" class=\"data row23 col4\" >0.9861</td>\n",
       "      <td id=\"T_e4999_row23_col5\" class=\"data row23 col5\" >0.9861</td>\n",
       "      <td id=\"T_e4999_row23_col6\" class=\"data row23 col6\" ></td>\n",
       "      <td id=\"T_e4999_row23_col7\" class=\"data row23 col7\" >0.9959</td>\n",
       "      <td id=\"T_e4999_row23_col8\" class=\"data row23 col8\" >0.9637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row24\" class=\"row_heading level0 row24\" >35</th>\n",
       "      <td id=\"T_e4999_row24_col0\" class=\"data row24 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row24_col1\" class=\"data row24 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row24_col2\" class=\"data row24 col2\" >p4_convnext_tiny_100pct</td>\n",
       "      <td id=\"T_e4999_row24_col3\" class=\"data row24 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row24_col4\" class=\"data row24 col4\" >0.9629</td>\n",
       "      <td id=\"T_e4999_row24_col5\" class=\"data row24 col5\" ></td>\n",
       "      <td id=\"T_e4999_row24_col6\" class=\"data row24 col6\" >0.0371</td>\n",
       "      <td id=\"T_e4999_row24_col7\" class=\"data row24 col7\" ></td>\n",
       "      <td id=\"T_e4999_row24_col8\" class=\"data row24 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row25\" class=\"row_heading level0 row25\" >32</th>\n",
       "      <td id=\"T_e4999_row25_col0\" class=\"data row25 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row25_col1\" class=\"data row25 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row25_col2\" class=\"data row25 col2\" >p4_convnext_tiny_100pct</td>\n",
       "      <td id=\"T_e4999_row25_col3\" class=\"data row25 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row25_col4\" class=\"data row25 col4\" >0.9791</td>\n",
       "      <td id=\"T_e4999_row25_col5\" class=\"data row25 col5\" >0.9791</td>\n",
       "      <td id=\"T_e4999_row25_col6\" class=\"data row25 col6\" ></td>\n",
       "      <td id=\"T_e4999_row25_col7\" class=\"data row25 col7\" >0.9969</td>\n",
       "      <td id=\"T_e4999_row25_col8\" class=\"data row25 col8\" >0.9712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row26\" class=\"row_heading level0 row26\" >33</th>\n",
       "      <td id=\"T_e4999_row26_col0\" class=\"data row26 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row26_col1\" class=\"data row26 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row26_col2\" class=\"data row26 col2\" >p4_convnext_tiny_100pct</td>\n",
       "      <td id=\"T_e4999_row26_col3\" class=\"data row26 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row26_col4\" class=\"data row26 col4\" >0.9387</td>\n",
       "      <td id=\"T_e4999_row26_col5\" class=\"data row26 col5\" >0.9387</td>\n",
       "      <td id=\"T_e4999_row26_col6\" class=\"data row26 col6\" ></td>\n",
       "      <td id=\"T_e4999_row26_col7\" class=\"data row26 col7\" >0.9908</td>\n",
       "      <td id=\"T_e4999_row26_col8\" class=\"data row26 col8\" >0.9502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row27\" class=\"row_heading level0 row27\" >34</th>\n",
       "      <td id=\"T_e4999_row27_col0\" class=\"data row27 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_e4999_row27_col1\" class=\"data row27 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row27_col2\" class=\"data row27 col2\" >p4_convnext_tiny_100pct</td>\n",
       "      <td id=\"T_e4999_row27_col3\" class=\"data row27 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row27_col4\" class=\"data row27 col4\" >0.9910</td>\n",
       "      <td id=\"T_e4999_row27_col5\" class=\"data row27 col5\" >0.9910</td>\n",
       "      <td id=\"T_e4999_row27_col6\" class=\"data row27 col6\" ></td>\n",
       "      <td id=\"T_e4999_row27_col7\" class=\"data row27 col7\" >0.9984</td>\n",
       "      <td id=\"T_e4999_row27_col8\" class=\"data row27 col8\" >0.9773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row28\" class=\"row_heading level0 row28\" >23</th>\n",
       "      <td id=\"T_e4999_row28_col0\" class=\"data row28 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row28_col1\" class=\"data row28 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row28_col2\" class=\"data row28 col2\" >p4_efficientnet_b0_100pct</td>\n",
       "      <td id=\"T_e4999_row28_col3\" class=\"data row28 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row28_col4\" class=\"data row28 col4\" >0.9603</td>\n",
       "      <td id=\"T_e4999_row28_col5\" class=\"data row28 col5\" ></td>\n",
       "      <td id=\"T_e4999_row28_col6\" class=\"data row28 col6\" >0.0397</td>\n",
       "      <td id=\"T_e4999_row28_col7\" class=\"data row28 col7\" ></td>\n",
       "      <td id=\"T_e4999_row28_col8\" class=\"data row28 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row29\" class=\"row_heading level0 row29\" >20</th>\n",
       "      <td id=\"T_e4999_row29_col0\" class=\"data row29 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row29_col1\" class=\"data row29 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row29_col2\" class=\"data row29 col2\" >p4_efficientnet_b0_100pct</td>\n",
       "      <td id=\"T_e4999_row29_col3\" class=\"data row29 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row29_col4\" class=\"data row29 col4\" >0.9755</td>\n",
       "      <td id=\"T_e4999_row29_col5\" class=\"data row29 col5\" >0.9755</td>\n",
       "      <td id=\"T_e4999_row29_col6\" class=\"data row29 col6\" ></td>\n",
       "      <td id=\"T_e4999_row29_col7\" class=\"data row29 col7\" >0.9958</td>\n",
       "      <td id=\"T_e4999_row29_col8\" class=\"data row29 col8\" >0.9682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row30\" class=\"row_heading level0 row30\" >21</th>\n",
       "      <td id=\"T_e4999_row30_col0\" class=\"data row30 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row30_col1\" class=\"data row30 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row30_col2\" class=\"data row30 col2\" >p4_efficientnet_b0_100pct</td>\n",
       "      <td id=\"T_e4999_row30_col3\" class=\"data row30 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row30_col4\" class=\"data row30 col4\" >0.9478</td>\n",
       "      <td id=\"T_e4999_row30_col5\" class=\"data row30 col5\" >0.9478</td>\n",
       "      <td id=\"T_e4999_row30_col6\" class=\"data row30 col6\" ></td>\n",
       "      <td id=\"T_e4999_row30_col7\" class=\"data row30 col7\" >0.9912</td>\n",
       "      <td id=\"T_e4999_row30_col8\" class=\"data row30 col8\" >0.9537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row31\" class=\"row_heading level0 row31\" >22</th>\n",
       "      <td id=\"T_e4999_row31_col0\" class=\"data row31 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_e4999_row31_col1\" class=\"data row31 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row31_col2\" class=\"data row31 col2\" >p4_efficientnet_b0_100pct</td>\n",
       "      <td id=\"T_e4999_row31_col3\" class=\"data row31 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row31_col4\" class=\"data row31 col4\" >0.9898</td>\n",
       "      <td id=\"T_e4999_row31_col5\" class=\"data row31 col5\" >0.9898</td>\n",
       "      <td id=\"T_e4999_row31_col6\" class=\"data row31 col6\" ></td>\n",
       "      <td id=\"T_e4999_row31_col7\" class=\"data row31 col7\" >0.9979</td>\n",
       "      <td id=\"T_e4999_row31_col8\" class=\"data row31 col8\" >0.9754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row32\" class=\"row_heading level0 row32\" >11</th>\n",
       "      <td id=\"T_e4999_row32_col0\" class=\"data row32 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row32_col1\" class=\"data row32 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row32_col2\" class=\"data row32 col2\" >p4_resnet50_100pct</td>\n",
       "      <td id=\"T_e4999_row32_col3\" class=\"data row32 col3\" >wiki</td>\n",
       "      <td id=\"T_e4999_row32_col4\" class=\"data row32 col4\" >0.9530</td>\n",
       "      <td id=\"T_e4999_row32_col5\" class=\"data row32 col5\" ></td>\n",
       "      <td id=\"T_e4999_row32_col6\" class=\"data row32 col6\" >0.0470</td>\n",
       "      <td id=\"T_e4999_row32_col7\" class=\"data row32 col7\" ></td>\n",
       "      <td id=\"T_e4999_row32_col8\" class=\"data row32 col8\" ></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row33\" class=\"row_heading level0 row33\" >8</th>\n",
       "      <td id=\"T_e4999_row33_col0\" class=\"data row33 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row33_col1\" class=\"data row33 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row33_col2\" class=\"data row33 col2\" >p4_resnet50_100pct</td>\n",
       "      <td id=\"T_e4999_row33_col3\" class=\"data row33 col3\" >inpainting</td>\n",
       "      <td id=\"T_e4999_row33_col4\" class=\"data row33 col4\" >0.9838</td>\n",
       "      <td id=\"T_e4999_row33_col5\" class=\"data row33 col5\" >0.9838</td>\n",
       "      <td id=\"T_e4999_row33_col6\" class=\"data row33 col6\" ></td>\n",
       "      <td id=\"T_e4999_row33_col7\" class=\"data row33 col7\" >0.9965</td>\n",
       "      <td id=\"T_e4999_row33_col8\" class=\"data row33 col8\" >0.9689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row34\" class=\"row_heading level0 row34\" >9</th>\n",
       "      <td id=\"T_e4999_row34_col0\" class=\"data row34 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row34_col1\" class=\"data row34 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row34_col2\" class=\"data row34 col2\" >p4_resnet50_100pct</td>\n",
       "      <td id=\"T_e4999_row34_col3\" class=\"data row34 col3\" >insight</td>\n",
       "      <td id=\"T_e4999_row34_col4\" class=\"data row34 col4\" >0.9500</td>\n",
       "      <td id=\"T_e4999_row34_col5\" class=\"data row34 col5\" >0.9500</td>\n",
       "      <td id=\"T_e4999_row34_col6\" class=\"data row34 col6\" ></td>\n",
       "      <td id=\"T_e4999_row34_col7\" class=\"data row34 col7\" >0.9903</td>\n",
       "      <td id=\"T_e4999_row34_col8\" class=\"data row34 col8\" >0.9515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e4999_level0_row35\" class=\"row_heading level0 row35\" >10</th>\n",
       "      <td id=\"T_e4999_row35_col0\" class=\"data row35 col0\" >ResNet50</td>\n",
       "      <td id=\"T_e4999_row35_col1\" class=\"data row35 col1\" >100</td>\n",
       "      <td id=\"T_e4999_row35_col2\" class=\"data row35 col2\" >p4_resnet50_100pct</td>\n",
       "      <td id=\"T_e4999_row35_col3\" class=\"data row35 col3\" >text2img</td>\n",
       "      <td id=\"T_e4999_row35_col4\" class=\"data row35 col4\" >0.9900</td>\n",
       "      <td id=\"T_e4999_row35_col5\" class=\"data row35 col5\" >0.9900</td>\n",
       "      <td id=\"T_e4999_row35_col6\" class=\"data row35 col6\" ></td>\n",
       "      <td id=\"T_e4999_row35_col7\" class=\"data row35 col7\" >0.9981</td>\n",
       "      <td id=\"T_e4999_row35_col8\" class=\"data row35 col8\" >0.9721</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1d2b11909e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 620x330 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 620x330 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x380 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x380 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_896cb\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_896cb_level0_col0\" class=\"col_heading level0 col0\" >model</th>\n",
       "      <th id=\"T_896cb_level0_col1\" class=\"col_heading level0 col1\" >scale_pct</th>\n",
       "      <th id=\"T_896cb_level0_col2\" class=\"col_heading level0 col2\" >pairwise_auc_min</th>\n",
       "      <th id=\"T_896cb_level0_col3\" class=\"col_heading level0 col3\" >pairwise_auc_max</th>\n",
       "      <th id=\"T_896cb_level0_col4\" class=\"col_heading level0 col4\" >pairwise_auc_gap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_896cb_row0_col0\" class=\"data row0 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_896cb_row0_col1\" class=\"data row0 col1\" >20</td>\n",
       "      <td id=\"T_896cb_row0_col2\" class=\"data row0 col2\" >0.9791</td>\n",
       "      <td id=\"T_896cb_row0_col3\" class=\"data row0 col3\" >0.9927</td>\n",
       "      <td id=\"T_896cb_row0_col4\" class=\"data row0 col4\" >0.0136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_896cb_row1_col0\" class=\"data row1 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_896cb_row1_col1\" class=\"data row1 col1\" >50</td>\n",
       "      <td id=\"T_896cb_row1_col2\" class=\"data row1 col2\" >0.9873</td>\n",
       "      <td id=\"T_896cb_row1_col3\" class=\"data row1 col3\" >0.9968</td>\n",
       "      <td id=\"T_896cb_row1_col4\" class=\"data row1 col4\" >0.0096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_896cb_row2_col0\" class=\"data row2 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_896cb_row2_col1\" class=\"data row2 col1\" >100</td>\n",
       "      <td id=\"T_896cb_row2_col2\" class=\"data row2 col2\" >0.9908</td>\n",
       "      <td id=\"T_896cb_row2_col3\" class=\"data row2 col3\" >0.9984</td>\n",
       "      <td id=\"T_896cb_row2_col4\" class=\"data row2 col4\" >0.0076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_896cb_row3_col0\" class=\"data row3 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_896cb_row3_col1\" class=\"data row3 col1\" >20</td>\n",
       "      <td id=\"T_896cb_row3_col2\" class=\"data row3 col2\" >0.9776</td>\n",
       "      <td id=\"T_896cb_row3_col3\" class=\"data row3 col3\" >0.9921</td>\n",
       "      <td id=\"T_896cb_row3_col4\" class=\"data row3 col4\" >0.0145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_896cb_row4_col0\" class=\"data row4 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_896cb_row4_col1\" class=\"data row4 col1\" >50</td>\n",
       "      <td id=\"T_896cb_row4_col2\" class=\"data row4 col2\" >0.9866</td>\n",
       "      <td id=\"T_896cb_row4_col3\" class=\"data row4 col3\" >0.9954</td>\n",
       "      <td id=\"T_896cb_row4_col4\" class=\"data row4 col4\" >0.0089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_896cb_row5_col0\" class=\"data row5 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_896cb_row5_col1\" class=\"data row5 col1\" >100</td>\n",
       "      <td id=\"T_896cb_row5_col2\" class=\"data row5 col2\" >0.9912</td>\n",
       "      <td id=\"T_896cb_row5_col3\" class=\"data row5 col3\" >0.9979</td>\n",
       "      <td id=\"T_896cb_row5_col4\" class=\"data row5 col4\" >0.0067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_896cb_row6_col0\" class=\"data row6 col0\" >ResNet50</td>\n",
       "      <td id=\"T_896cb_row6_col1\" class=\"data row6 col1\" >20</td>\n",
       "      <td id=\"T_896cb_row6_col2\" class=\"data row6 col2\" >0.9792</td>\n",
       "      <td id=\"T_896cb_row6_col3\" class=\"data row6 col3\" >0.9900</td>\n",
       "      <td id=\"T_896cb_row6_col4\" class=\"data row6 col4\" >0.0108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_896cb_row7_col0\" class=\"data row7 col0\" >ResNet50</td>\n",
       "      <td id=\"T_896cb_row7_col1\" class=\"data row7 col1\" >50</td>\n",
       "      <td id=\"T_896cb_row7_col2\" class=\"data row7 col2\" >0.9869</td>\n",
       "      <td id=\"T_896cb_row7_col3\" class=\"data row7 col3\" >0.9959</td>\n",
       "      <td id=\"T_896cb_row7_col4\" class=\"data row7 col4\" >0.0090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_896cb_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_896cb_row8_col0\" class=\"data row8 col0\" >ResNet50</td>\n",
       "      <td id=\"T_896cb_row8_col1\" class=\"data row8 col1\" >100</td>\n",
       "      <td id=\"T_896cb_row8_col2\" class=\"data row8 col2\" >0.9903</td>\n",
       "      <td id=\"T_896cb_row8_col3\" class=\"data row8 col3\" >0.9981</td>\n",
       "      <td id=\"T_896cb_row8_col4\" class=\"data row8 col4\" >0.0078</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1d2af85ecf0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def source_rows(model_label: str, scale: float, run_name: str) -> list[dict]:\n",
    "    rows = []\n",
    "    per_source = load_run(run_name).get(\"aggregated_per_source\", {})\n",
    "    for source, metrics in per_source.items():\n",
    "        rows.append({\n",
    "            \"model\": model_label,\n",
    "            \"scale_pct\": int(scale * 100),\n",
    "            \"run\": run_name,\n",
    "            \"source\": source,\n",
    "            \"accuracy\": metrics.get(\"accuracy\", {}).get(\"mean\", np.nan),\n",
    "            \"detection_rate\": metrics.get(\"detection_rate\", {}).get(\"mean\", np.nan),\n",
    "            \"false_alarm_rate\": metrics.get(\"false_alarm_rate\", {}).get(\"mean\", np.nan),\n",
    "            \"pairwise_auc\": metrics.get(\"pairwise_auc\", {}).get(\"mean\", np.nan),\n",
    "            \"pairwise_f1\": metrics.get(\"pairwise_f1\", {}).get(\"mean\", np.nan),\n",
    "        })\n",
    "    return rows\n",
    "\n",
    "\n",
    "source_df = pd.DataFrame([\n",
    "    row\n",
    "    for model_label in MODEL_ORDER\n",
    "    for scale in SCALE_ORDER\n",
    "    for row in source_rows(model_label, scale, MODEL_RUNS[model_label][scale])\n",
    "])\n",
    "source_order = {source: i for i, source in enumerate([\"wiki\", *FAKE_SOURCES])}\n",
    "display(\n",
    "    source_df.assign(source_rank=source_df[\"source\"].map(source_order))\n",
    "    .sort_values([\"scale_pct\", \"model\", \"source_rank\"])\n",
    "    .drop(columns=\"source_rank\")\n",
    "    .style.format({\n",
    "        \"accuracy\": \"{:.4f}\",\n",
    "        \"detection_rate\": \"{:.4f}\",\n",
    "        \"false_alarm_rate\": \"{:.4f}\",\n",
    "        \"pairwise_auc\": \"{:.4f}\",\n",
    "        \"pairwise_f1\": \"{:.4f}\",\n",
    "    }, na_rep=\"\")\n",
    ")\n",
    "\n",
    "\n",
    "def heatmap_for_scale(metric: str, scale_pct: int, title: str, out_name: str, *, vmin=None, vmax=None):\n",
    "    pivot = (\n",
    "        source_df[(source_df[\"scale_pct\"] == scale_pct) & (source_df[\"source\"].isin(FAKE_SOURCES))]\n",
    "        .pivot(index=\"model\", columns=\"source\", values=metric)\n",
    "        .reindex(index=MODEL_ORDER, columns=FAKE_SOURCES)\n",
    "    )\n",
    "    fig, ax = plt.subplots(figsize=(6.2, 3.3))\n",
    "    im = ax.imshow(pivot.values, cmap=\"viridis\", vmin=vmin, vmax=vmax)\n",
    "    ax.set_xticks(range(len(pivot.columns)))\n",
    "    ax.set_xticklabels(pivot.columns)\n",
    "    ax.set_yticks(range(len(pivot.index)))\n",
    "    ax.set_yticklabels(pivot.index)\n",
    "    for i in range(pivot.shape[0]):\n",
    "        for j in range(pivot.shape[1]):\n",
    "            value = pivot.iloc[i, j]\n",
    "            threshold = (vmin + vmax) / 2 if vmin is not None and vmax is not None else np.nanmean(pivot.values)\n",
    "            ax.text(j, i, f\"{value:.3f}\", ha=\"center\", va=\"center\", color=\"white\" if value < threshold else \"black\", fontsize=8)\n",
    "    ax.set_title(title)\n",
    "    fig.colorbar(im, ax=ax, label=metric)\n",
    "    fig.tight_layout()\n",
    "    fig.savefig(FIGURES_DIR / out_name, dpi=180, bbox_inches=\"tight\")\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "heatmap_for_scale(\"pairwise_auc\", 100, \"100% data pairwise AUC by fake source\", \"07_phase4_100pct_pairwise_auc.png\", vmin=0.988, vmax=1.0)\n",
    "heatmap_for_scale(\"pairwise_f1\", 100, \"100% data pairwise F1 by fake source\", \"07_phase4_100pct_pairwise_f1.png\", vmin=0.94, vmax=0.985)\n",
    "\n",
    "fig, axes = plt.subplots(1, 3, figsize=(13, 3.8), sharey=True)\n",
    "for ax, pair_source in zip(axes, FAKE_SOURCES):\n",
    "    for model_label in MODEL_ORDER:\n",
    "        sub = source_df[(source_df[\"model\"] == model_label) & (source_df[\"source\"] == pair_source)].sort_values(\"scale_pct\")\n",
    "        ax.plot(sub[\"scale_pct\"], sub[\"pairwise_auc\"], marker=\"o\", label=model_label)\n",
    "    ax.set_title(f\"wiki vs {pair_source}\")\n",
    "    ax.set_xlabel(\"Training data used (%)\")\n",
    "    ax.set_ylabel(\"Pairwise AUC\")\n",
    "    ax.grid(alpha=0.25)\n",
    "axes[0].legend(fontsize=8)\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_pairwise_auc_scaling.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "fig, axes = plt.subplots(1, 3, figsize=(13, 3.8), sharey=True)\n",
    "for ax, pair_source in zip(axes, FAKE_SOURCES):\n",
    "    for model_label in MODEL_ORDER:\n",
    "        sub = source_df[(source_df[\"model\"] == model_label) & (source_df[\"source\"] == pair_source)].sort_values(\"scale_pct\")\n",
    "        ax.plot(sub[\"scale_pct\"], sub[\"pairwise_f1\"], marker=\"o\", label=model_label)\n",
    "    ax.set_title(f\"wiki vs {pair_source}\")\n",
    "    ax.set_xlabel(\"Training data used (%)\")\n",
    "    ax.set_ylabel(\"Pairwise F1\")\n",
    "    ax.grid(alpha=0.25)\n",
    "axes[0].legend(fontsize=8)\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_pairwise_f1_scaling.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "source_gap_rows = []\n",
    "for (model_label, scale_pct), sub in source_df[source_df[\"source\"].isin(FAKE_SOURCES)].groupby([\"model\", \"scale_pct\"]):\n",
    "    source_gap_rows.append({\n",
    "        \"model\": model_label,\n",
    "        \"scale_pct\": scale_pct,\n",
    "        \"pairwise_auc_min\": sub[\"pairwise_auc\"].min(),\n",
    "        \"pairwise_auc_max\": sub[\"pairwise_auc\"].max(),\n",
    "        \"pairwise_auc_gap\": sub[\"pairwise_auc\"].max() - sub[\"pairwise_auc\"].min(),\n",
    "    })\n",
    "source_gap_df = pd.DataFrame(source_gap_rows).sort_values([\"model\", \"scale_pct\"])\n",
    "display(source_gap_df.style.format({\"pairwise_auc_min\": \"{:.4f}\", \"pairwise_auc_max\": \"{:.4f}\", \"pairwise_auc_gap\": \"{:.4f}\"}))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db04a8ea",
   "metadata": {},
   "source": [
    "Scaling helps the source diagnostics too. The weakest source remains `insight`, but it moves into a much stronger range at 100% data: around `0.990` to `0.991` pairwise AUC depending on the backbone. `text2img` becomes almost saturated, and `inpainting` also rises above `0.995` for the 100% runs.\n",
    "\n",
    "The source gap narrows because the weakest column improves more than the already-easy sources. This does not completely solve source generalization, but it makes the normal cross-validation result less dependent on one fake source being easy.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc8de432",
   "metadata": {},
   "source": [
    "## 5. Confusion matrices and balanced metrics\n",
    "\n",
    "The final model choice should not depend only on AUC. AUC measures ranking quality across thresholds; the confusion matrix and balanced metrics show the practical thresholded behavior. This matters because the dataset has more fake images than real images, so standard accuracy and binary F1 give more total weight to fake examples.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6be76d2c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1260x380 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_eebc5\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_eebc5_level0_col0\" class=\"col_heading level0 col0\" >model</th>\n",
       "      <th id=\"T_eebc5_level0_col1\" class=\"col_heading level0 col1\" >scale_pct</th>\n",
       "      <th id=\"T_eebc5_level0_col2\" class=\"col_heading level0 col2\" >run</th>\n",
       "      <th id=\"T_eebc5_level0_col3\" class=\"col_heading level0 col3\" >real_recall</th>\n",
       "      <th id=\"T_eebc5_level0_col4\" class=\"col_heading level0 col4\" >fake_recall</th>\n",
       "      <th id=\"T_eebc5_level0_col5\" class=\"col_heading level0 col5\" >balanced_accuracy</th>\n",
       "      <th id=\"T_eebc5_level0_col6\" class=\"col_heading level0 col6\" >real_f1</th>\n",
       "      <th id=\"T_eebc5_level0_col7\" class=\"col_heading level0 col7\" >fake_f1</th>\n",
       "      <th id=\"T_eebc5_level0_col8\" class=\"col_heading level0 col8\" >macro_f1</th>\n",
       "      <th id=\"T_eebc5_level0_col9\" class=\"col_heading level0 col9\" >false_positive_rate</th>\n",
       "      <th id=\"T_eebc5_level0_col10\" class=\"col_heading level0 col10\" >false_negative_rate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_eebc5_level0_row0\" class=\"row_heading level0 row0\" >2</th>\n",
       "      <td id=\"T_eebc5_row0_col0\" class=\"data row0 col0\" >ResNet50</td>\n",
       "      <td id=\"T_eebc5_row0_col1\" class=\"data row0 col1\" >100</td>\n",
       "      <td id=\"T_eebc5_row0_col2\" class=\"data row0 col2\" >p4_resnet50_100pct</td>\n",
       "      <td id=\"T_eebc5_row0_col3\" class=\"data row0 col3\" >0.9530</td>\n",
       "      <td id=\"T_eebc5_row0_col4\" class=\"data row0 col4\" >0.9746</td>\n",
       "      <td id=\"T_eebc5_row0_col5\" class=\"data row0 col5\" >0.9638</td>\n",
       "      <td id=\"T_eebc5_row0_col6\" class=\"data row0 col6\" >0.9393</td>\n",
       "      <td id=\"T_eebc5_row0_col7\" class=\"data row0 col7\" >0.9794</td>\n",
       "      <td id=\"T_eebc5_row0_col8\" class=\"data row0 col8\" >0.9593</td>\n",
       "      <td id=\"T_eebc5_row0_col9\" class=\"data row0 col9\" >0.0470</td>\n",
       "      <td id=\"T_eebc5_row0_col10\" class=\"data row0 col10\" >0.0254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_eebc5_level0_row1\" class=\"row_heading level0 row1\" >5</th>\n",
       "      <td id=\"T_eebc5_row1_col0\" class=\"data row1 col0\" >EfficientNet-B0</td>\n",
       "      <td id=\"T_eebc5_row1_col1\" class=\"data row1 col1\" >100</td>\n",
       "      <td id=\"T_eebc5_row1_col2\" class=\"data row1 col2\" >p4_efficientnet_b0_100pct</td>\n",
       "      <td id=\"T_eebc5_row1_col3\" class=\"data row1 col3\" >0.9603</td>\n",
       "      <td id=\"T_eebc5_row1_col4\" class=\"data row1 col4\" >0.9710</td>\n",
       "      <td id=\"T_eebc5_row1_col5\" class=\"data row1 col5\" >0.9657</td>\n",
       "      <td id=\"T_eebc5_row1_col6\" class=\"data row1 col6\" >0.9382</td>\n",
       "      <td id=\"T_eebc5_row1_col7\" class=\"data row1 col7\" >0.9787</td>\n",
       "      <td id=\"T_eebc5_row1_col8\" class=\"data row1 col8\" >0.9584</td>\n",
       "      <td id=\"T_eebc5_row1_col9\" class=\"data row1 col9\" >0.0397</td>\n",
       "      <td id=\"T_eebc5_row1_col10\" class=\"data row1 col10\" >0.0290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_eebc5_level0_row2\" class=\"row_heading level0 row2\" >8</th>\n",
       "      <td id=\"T_eebc5_row2_col0\" class=\"data row2 col0\" >ConvNeXt-Tiny</td>\n",
       "      <td id=\"T_eebc5_row2_col1\" class=\"data row2 col1\" >100</td>\n",
       "      <td id=\"T_eebc5_row2_col2\" class=\"data row2 col2\" >p4_convnext_tiny_100pct</td>\n",
       "      <td id=\"T_eebc5_row2_col3\" class=\"data row2 col3\" >0.9629</td>\n",
       "      <td id=\"T_eebc5_row2_col4\" class=\"data row2 col4\" >0.9696</td>\n",
       "      <td id=\"T_eebc5_row2_col5\" class=\"data row2 col5\" >0.9662</td>\n",
       "      <td id=\"T_eebc5_row2_col6\" class=\"data row2 col6\" >0.9375</td>\n",
       "      <td id=\"T_eebc5_row2_col7\" class=\"data row2 col7\" >0.9784</td>\n",
       "      <td id=\"T_eebc5_row2_col8\" class=\"data row2 col8\" >0.9580</td>\n",
       "      <td id=\"T_eebc5_row2_col9\" class=\"data row2 col9\" >0.0371</td>\n",
       "      <td id=\"T_eebc5_row2_col10\" class=\"data row2 col10\" >0.0304</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1d2b0d751c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1100x420 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0cAAAGZCAYAAABCEUr1AAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjksIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvJkbTWQAAAAlwSFlzAAAPYQAAD2EBqD+naQAAWehJREFUeJzt3QmcjfX7//HLvmYvW6JFlmzha6uoCKWiRVIiSaVNKWXLEpGKKCKKqERKKklJlLJlSyqiBdlVlsh+/o/3Nb/7/M+MM8NgzIx5PR+P85g597nPfe5zztz33Nfn+nyuT7pQKBQyAAAAAEjj0if3DgAAAABASkBwBAAAAAAERwAAAAAQg8wRAAAAABAcAQAAAEAMMkcAAAAAQHAEAAAAADHIHAEAAAAAwREAAAAAxCBzBABIcXr27Gnp0qVL7t04rbzxxhv+mS5cuNDSwvv8448/Ev1c/u4AEBwBSDP+/fdf69GjhzVs2NDy5cvnF1C6kIrPzz//7OvmzJnT17/jjjts69atR6x3+PBhe+655+zcc8+1rFmzWoUKFeydd945Yr3Jkydb6dKlLXfu3HbdddfZhg0bjljn+uuvt3vuucfSgj179vjF6KxZs5J7VwAAcARHANKMbdu22dNPP+1BT8WKFRNc988//7TatWvb6tWrrW/fvvb444/bJ598YldddZXt378/1rpdu3a1J5980h97+eWX7ZxzzrHbbrvNxo8fH17nt99+s2bNmlm1atXs2WeftV9++cVat24dazufffaZff311/bMM89YWgmOevXqFTU46tatm/3333/Jsl8AgLQrY3LvAACcKoULF7aNGzdaoUKFvGvR//73v3jXVUC0e/duW7RokQc7osBGAZCyTUF2Z/369TZgwAB74IEHbMiQIb7s7rvvtjp16ljHjh2tadOmliFDBvv888/t7LPPtjFjxnjGqkyZMnbllVfa3r17Pdt08OBBe/TRR6179+525plnWmqk96AsWubMmU94WxkzZvRbaqO/mRw5chyxPBQK+XedLVu24962nq/PNn162jUBIKlwhgWQZmTJksUDo2Px/vvv27XXXhsOjKRevXp24YUX2rvvvhte9uGHH9qBAwfs/vvvDy9T8NOuXTvPPs2dO9eXKQuSJ0+e8DgaddPTBXOQHVFgdejQIXvooYcSHZD07t3bzj//fH9/JUqUsC5duti+fftiraflej8K0ipVquQBWdmyZW3SpElHbHP79u32yCOPWLFixXybF1xwgfXv398Dn4DGc+i9vPDCCzZo0KDw6//000+eWVOQV6VKFe9CqGDhsssus5kzZ8Z6fhAEKnukbemmbnbxjf3Q/QcffNC7J5YrV85f76KLLrJp06Yd8R6Ujapataq/T+3bq6++mqjxJPPnz/culdr/7Nmze7D77bffxlon2J7eszKFefPmtUsvvTTW561soPZDQZH2IcgiKmjW34C2XaNGDc9Kxt1/bVvZR2XRihYt6uvu3LnT/970mZUsWdLfX/78+f11p0+ffswZu3vvvdeflytXLmvZsqX9888/4cdbtWplBQoU8NeJq379+laqVKkEt3/55Zf797Ns2TL/3LTf+ht67733/PGvvvrKqlev7p+JtvXFF18csY0lS5bY1Vdf7funbq1169a1efPmHbHejz/+6I0M2pYaH/r06RPr7zTSp59+6n+H+ns844wzrFGjRv58AIglBABp0HfffRfSKXD06NFHPPbnn3/6Y/379z/isRYtWoTy5csXvn/33XeHcuTIETp8+HCs9VavXu3beOmll/z+7NmzQ+nSpQuNGzcu9Ntvv4VuueWW0AUXXOCPbdmyJZQnT57QlClTEv0+WrVq5a9z8803h4YOHRpq2bKl32/SpEms9YoXLx668MIL/XU6deoUGjhwYKh8+fKh9OnThz7//PPwert37w5VqFAhlD9//lCXLl1Cw4cP921q39u3bx9e7/fff/fXKVu2bOi8884LPfvss6EXX3wxtGbNmtDWrVtDhQsXDnXo0CE0bNiw0HPPPRcqVapUKFOmTKElS5b48//9919/TNu44YYbQm+++abfvv/+e3+8R48e/lgk3a9YsaJvu3fv3qFBgwb5a2fPnj20bdu28HqLFy8OZcmSJVSiRAnfr2eeeSZUpEgRf+6x/NubMWNGKHPmzKGaNWuGBgwY4O9Ln4mWzZ8/P7xesI/6DBo3bhx65ZVX/DsIPm99v3nz5vXPW5/jzJkzQ5s2bQoVLFgwdMYZZ4S6du3q34P2S9/DpEmTwtvWusG2K1Wq5Ov169fPvx99L/o+2rZtGxo5cqTvY/Pmzf29JkR/69qmvvfLLrvM/zYfeOABf+3atWuH/4anT5/u63388cexnr9x48ZQhgwZQk8//XSCr1OnTh3/vIsVKxbq2LFj6OWXX/b3oeeOHz8+VKhQoVDPnj39+ytatGgod+7coZ07d4afv3z5cj+mgu9Z7+vcc8/173TevHmx9ufMM8/0z1jbe/7550MlS5b070r7r7/RwNixY/0za9iwoe+Pjm39feh4iFwv2t8dgLSFMwCANCmh4Ch4TBdUceliT4/t3bvX7zdq1Mgv0OPSRazW04Vx4OGHH/ZluinA+vLLL325LnJ10ZZYS5cu9W0pQIv0+OOP+/Jg+8HFupa9//774WU7duzwC9CLL744vEwXo7ow/eWXX2JtU+9DF7dr166NFRzlypXLg7tIBw8eDO3bty/Wsn/++ceDgrvuuiu8TEGUtqEL0rjiC44UoCjwDCiY0nJd8Aauu+46D5jWr18fXrZq1apQxowZj3rhqwBBF9gNGjSIFfDu2bPHL9CvuuqqI/ZRgUlcwec9bdq0WMsfeeQRX65gObBr1y7fti7WDx06FCs40t+WXjuSgin93SVWEBxVqVIltH///vByBa9a/uGHH/p97cPZZ58datasWaznK0BTgKHg/mjBkbanhoDAihUrfJkCscgA57PPPjviOFRgr+/5119/DS/bsGGDB5QK4uJ+lpEBq/4WFWxFBkf6fBUE6TiLpEBV60YuJzgCQLc6AIgj6OqmbltxqRtT5Dr6eSzryeDBg23NmjXeZUs/r7jiClu6dKmNHTvWXnzxRduxY4e1aNHCu1Cpa5IKRyRk6tSp/rNDhw6xlj/22GP+M25XrSJFitgNN9wQvh90qVIXpk2bNvmyiRMnetcjdRFTAYvgpi6F6vanghGRbrrppiPGSGmMVTDuSF2c/v77b+/+p+5lixcvthOh/VA3uYAqA+p9qKuaaB/VTatJkyb+fgPq1qVuWkej72PVqlXeTe6vv/4Kv3+NJVLXLr3/uN227rvvvqjbUvXCBg0aHPGdaexa0P1O1G1MY9jU1VBd9CKpi1vccUrqnqnuYNrP46HXypQpU/i+uoBqfFfw96QxTbfffrt99NFHtmvXrvB6b7/9ttWqVcvf19HoPd16663h++o+p/3WWDt1qQsEv0d+f+r6qe/vvPPOizVeUN/JN998410LRfurLon6PAP6W9S+R1J3Q3UVbd68eay/af2d6vUju3sCAMERAMQRXIzGHbcTDIqPXEc/j2W9gMYw6WJOF4/y8MMP+8W1SnyrqMO6det8HFP58uW93LeCivgowNKFrC78I2lclS5E9XgkrRd3zI3GUEkwJ4wuuDWGRxeZkTcFJbJly5ZYz4/vQlmFJxS4BGNitA0FawoAT0TkGLCAArlgzIz2TwFp3M9Eoi2LKwg4FJTE/Qxee+01/67jvof4PoNoy/WdRBuzo6AhePxo21DFRV3s67vT34kKf2h8z7HSWKVI+ltU8BE5L5CCZn2OH3zwgd9fuXKlFydROftjofE/cf/WNH5L49jiLpPg+1OpfI2Jiu8zUmCqYyT4rOK+F4n73OA71dikuN+pArG4f9MA0rbUVwoIAJKYLhRFle3i0jINpA+yRVpXLc/q9RV5MRg8NzJ7EdeECRM8O6QWerWYq9CDLtaUYVGhgZEjR/og9MgsQzQnc7JUXXyqIt8TTzwR9fEgmApEq7721ltv2Z133umt/7pwP+uss7yVvl+/fvbrr7+e0P5pO9HE9Lo7cUFW6Pnnn/fCFdEEgW0gvgp0J1KZLqFtqMS8PkcF0fp7UdCmzOPw4cO9UuLJoGIdKqih71KBkn4qG3jLLbec0PeU1N9fQt/pm2++GbUgS2qsiggg6XBGAIA41K1Nrcoq9x3XggULYl0063ddnCrI0QVlQF3ngsejUeu4AgdVmlOWZ/PmzV4dLAimdFGsjIhKhcenePHifuGnlvEg8yDaljILejyS5myKG8RpvqWgupqoy5omyw0yRcdDVcnUJUqV8CJfSxPwJlVQF1AgpmyV3mtc0ZbFFXTZU1e9E/kM4qPvRFmYuFasWBF+/FgoQNc8Wbrp+1LApOp5xxIc6e9FXToDer6C+WuuuSbWegqK1GVTj40bN86ru+lvMinpuFN1u/g+I2VKg+yTPqtoXQvjPjf4TvW3kRTfKYDTC93qACAKjaWZMmVKuAuPzJgxw4MJlWEONG7c2MdvvPLKK+FlCkDUiq8gS2M0olFpbF1otm3b1u+r65lasIOLZI2JUBejhEqPBxezKqUdaeDAgf5TF7ORNmzYEO4mJRq7ofFOCuCC11FmQOXHVYI6LgVcCXXzi5sdiMwGKFgMypoHdBEcbPdk0WvrAljlvvV+IwMjlXI+GmVLdDGtEuUKGuLSd3Ii9J0pwI78LDSeacSIER6gRgbY8dFYqLiZLHUZjNa9Mxq9VmSZ7mHDhvn3GndMlsboKIBt3769jwnSeLikpu9P5cKVFYvs5qeAXwGasqgKXIPPUplVfZ6R34/GRkXSuC89R3OXRStPfqLfKYDTC5kjAGmK5hPSxXhw4fzxxx/7fESiOYaCMRCaK0jFCdTCrotDXSirq5XGeKi1PnJsheYE0mO68NLEsrownz17tl+kRetGtHbtWl9fY3CCxxUYKdDStvS4ghhlkWrWrBnve6lYsaKPjdHFrt6T5pTRhaLG+6hLW2R2IOgS16ZNG/vuu++sYMGCNmrUKL/oHD16dHgdZbPUzU9z9KhrnIIFXbz/8MMPnhHSBavmwEmInquskYo/KED7/fffPVjUhX9kwKHsmJape6H2TdkQzY+j24lQBkXdzS655BIvNqAui/retV0VXEiIMhPKBCpQUNdGfdcKcpXBU/dJXWTrb+Z4derUyd555x3fvsab6T3r+9JnpLm1jmWCV31mKtih70bPV4ZT343mgDoWmodKxSUUCCvLosBeQcf1119/RBZHcz3pOFB2M26wnVQ0V5GKKGifNH+Yjg3NEaXg77nnnguvp66f6iqnfdQxqvmLdCwooxQ5BkvfmQJAjZeqXLmyF4rQe9NxpmNQfyfBBM4AQClvAGlKUGI52i1yvpNgvpX69et7WWiVAr799tu9/G9cKn3ct29f37ZKEF900UWht956K959aNq0aejGG288YvnmzZu9DLVKFleuXDm0cOHCo76fAwcOhHr16uWloDWPkOaW6dy5c7jUeOT7VvlnlU7WPDCaM6Z06dKhiRMnHrFNlT7WNjRPj95PgQIFQrVq1Qq98MIL4RLQQSlvzS0Tl0pgB5+HXkelwjWHk+Zk0rJIc+bM8dLSep3Ist7xlfLWvDxxaZvadty5ivS62u75558feu2110KPPfZYKGvWrKFjofmY9B1pvie9B72G5qbSdgPBPqokebR9iq/ctkpUa14q/U1pf6pVq3bEHFdBKe9o30+fPn38OXp+tmzZ/HvUXE6R5bkTKuX91Vdfhe655x6fHyhnzpz+d/3XX39Ffc67777rz9H6x0qlvHUMHOtnEu171VxVKqeu/dPxd8UVV/jfSlzLli3z19PnqDmTVIr+9ddfj3o86zPVNlW+W+vr7+LOO++MdZxRyhtAOn0ExIgAcHpTly1lTtRVMK1SNu1ESmCnRereps9NJcxV4h0ATneMOQIAnHYi55cSBUSaF0fd0XDsVDFRxTWOVjERAE4XjDkCAJx2dEGvMVP6qflwNOZEpajjK1GO2MaPH+/jdjQmR5MXJ0VlQQBIiQiOAACnHQ3SV+GDTZs2+ZxUKmyhamXRJg3FkVSpTlXwVMBDRREAIK1gzBEAAAAAMOYIAAAAAGLQrS4KzTivOVDOOOMM+lkDAAAAqZwKdO/atcvnEExoTjmCoygUGBUrViwpvx8AAAAAp9i6det8Avf4EBxFoYxR8OFpZm0AAAAAqdfOnTs9+RFc58eH4CiKoGSpAiOCIwAAAOD0cLSpCZgEFgAAAAAIjgAAAAAgBpkjAAAAAGDMEQAAAA4dOmQHDhzgg0CqlSlTJsuQIcMJb4eCDAAAAGl47pdNmzbZ9u3bk3tXgBOWJ08eK1So0AnNU0pwBAAAkEYFgdFZZ51l2bNnP6GLSiA5g/w9e/bYli1b/H7hwoWPe1sERwAAAGm0K10QGOXPnz+5dwc4IdmyZfOfCpD0N328XewoyAAAAJAGBWOMlDECTgfZ/+9v+UTGzxEcAQAApGF0pcPpIt1J6BZKcAQAAAAABEcAAAAAEIOCDAAAAAjb2O6+U/ppFB42PNGVye69915777337J9//rElS5ZYpUqVEnzOH3/8Yeeee+4xrZtSXX755b7vgwYNinedN954wx555BFKs58AgqNU4FSfpNKixJ6YAQBA8pg2bZoHAbNmzbLzzjvPChQokCa+ikmTJvlEp4ESJUp4IKRboFmzZnbNNddYSh4T9MEHH1iTJk0spSI4AgAAQKrx66+/+jw2tWrVsrQkX758x1TOOihpfSpLwqdLl87Spz89ShmcHu8CAAAAp70777zTHnroIVu7dq1fkCt7EmSTLr30UsuTJ4/P2XTttdd6EBUfdce7/fbb7cwzz/RgomTJkjZ69Ojw4+vWrbNbbrnFt6egpHHjxt41Lz7KYml/PvnkE6tQoYJlzZrVatSoYcuXL4+13vvvv28XXXSRZcmSxfd9wIABsR5/5ZVXfF/0/IIFC9rNN98cq1tdkCXS72vWrLFHH33UXzeo0qaMmvZZfvnlF1++YsWKWK/x4osv2vnnnx++r328+uqrLWfOnP6ad9xxh23bti3e9xq8xkcffWRly5b196Lv47vvvrOrrrrKM3m5c+e2OnXq2OLFi8PPC76rG264IdZ3Jx9++KFVrlzZ37eygb169bKDBw9aciA4AgAAQKowePBge/rpp+3ss8+2jRs3+gW57N692zp06GALFy60GTNmeBZDF+GHDx+Oup2nnnrKfvrpJ/v000/t559/tmHDhoW752mOnAYNGtgZZ5xhs2fPtm+//dYDh4YNG9r+/fsT3L+OHTt6wKP9UuB13XXXhefcWbRokQdct956q/3www/Ws2dP3w8FG6J9f/jhh/39rVy50gO+2rVrx9vFTp+B1tXnoFtcF154oVWtWtXefvvtWMt1/7bbbvPfNQnwlVdeaRdffLG//rRp02zz5s2+nwnZs2eP9e/f31577TX78ccffdLVXbt2WatWreybb76xefPmeZCnLn5aLsF3pSA08rvTZ9yyZUtr3769fyevvvqqfybPPPOMJQe61QEAACBVUEZCQUuGDBmsUKFC4eU33XRTrPVGjRrlwYkutsuVK3fEdpTpUECg4EEisxgTJkzwoEoX/kFGRhf0ypYoQ1S/fv14969Hjx6ePZExY8Z4AKMxNgo2Bg4caHXr1vWAKAhetH/PP/+8Z8S0Tzly5PCsl95j8eLFfR+jUTZLn4HWi/wc4lJ2bMiQIda7d+9wNklB2ltvveX39Zheo2/fvrE+u2LFivm62sdoFPApy1WxYsXwMgVZkUaMGOGf2VdffeXvSd+HaFnkPitL1KlTJw+sRJkj7e8TTzzhn+epRuYIAAAAqdqqVausefPmfmGdK1eucLCjgCOadu3a2fjx4736my7C58yZE37s+++/t9WrV3vgoYyRbgpG9u7dm2BXPalZs2b4dz2nVKlSnpkS/bzkkktira/72neN21FQpYBI70Fd25ThUYbmRChLpe6AyuSItqnua6VLlw6/15kzZ4bfZ86cOcOPJfReM2fO7N0HIynj1LZtW88YKYjV9/Dvv//G+x0EtA/KgEXug7aj7NKJvv/jQeYIAAAAqZq6rymwGDlypBUpUsQzP8oYxdcNTmNsNGZn6tSpNn36dM/oPPDAA/bCCy/4BX2VKlWO6I4mQfYjKSgY0xgdZac+//xz6969u3e9U/ezYBxRYilDo4zOuHHjfAyUfiowDOi96rNTF7m4ChcuHO92NU4ryKoFlPn566+/vOujvguNRVKweLSuiNoHZY9uvPHGIx7TGKRTjeAIAAAAqZYuyDVGR4HRZZdd5ss07uVoFOjogl43PU/jhRQcKbOirnUaR6PsR2IoQ3POOeeEiz6oa1qZMmX8vn5q/FIk3VfXNXWRk4wZM1q9evX8pi5lCoq+/PLLqIGDsjfKOB2NutYpO6bM2m+//ebZpIDeq4pEKNOm1z4Rei/qaheUEldRi7iFHVSKPO4+ax/0/V1wwQWWEtCtDgAAAKlW3rx5vUKdxrioO5yCCRVnSIiyMqqQpvVVUGDKlCnhIEbBhIozqEKdigX8/vvvns1RsYQ///wzwe2qe5gKQqgCnMYRaTvBnD6PPfaYP6bxNAqaNCZJY34ef/xxf1z78NJLL9nSpUs9qzV27FjPgKlrXjQKaL7++mtbv359gtXlFFipKIIyRldccYVn1gLKlv39998eOClD9euvv9pnn31mrVu3PqbAK5K607355pvefXD+/Pn+OcYtK6591mewadMmDx6D70LvVdkjfRd6vro8duvWzZIDmSMAAACk2onRVZlOF9MKXtSVTsGEggyVu46Psi6dO3f28Ti6gFfmSNuQ7Nmze9Dx5JNPhgOLokWLete7o2WSnn32Wa+6pnFEGs/08ccf+2sFGZJ3333XgwEFSOq2pmBKQZQoS6QqdOpKp/FNCjbeeecdL/0djZ577733elnuffv2WSgUire7nrrO6bVVbCGSAiVlfPReVWhi37593iVOlfkSO2/R66+/bvfcc4+/TxV0UJGHIPALqJKfAldl+fSZ6vNXZUAFhno/6t6n7JLGPd19992WHNKF4vsk07CdO3f6QLIdO3YkOp2aFDa2uy+5d+G0l9r+EQAAcKJ0Aa6syLnnnpssYztOJ8osKSujbMjxjg9C0v5NH+v1Pd3qAAAAAIDgCAAAAABiMOYIAAAAOAEa38RIldMD3eoAAAAAgOAIAAAAAGKQOQIAAAAAgiMAAAAAiEHmCAAAAAAIjgAAAAAgBqW8AQAAENaq74en9NMY06VxotZXyex7773X3nvvPfvnn39syZIlVqlSpQSf88cff9i55557TOuertKlS2cffPCBNWnSJLl3JUWjWx0AAABSjWnTptkbb7xhU6ZMsY0bN1q5cuWSe5dSlJ49e0YNAPVZXX311ZYS3XnnnSkmaCNzBAAAgFTj119/tcKFC1utWrWSe1dSlUKFCp3y1zxw4IBlypTJUhMyRwAAAEgVlGF46KGHbO3atd5NrESJEuFs0qWXXmp58uSx/Pnz27XXXutBVHzUHe/222+3M88807Jly2YlS5a00aNHhx9ft26d3XLLLb69fPnyWePGjb1rXnxmzZrl+zNjxgyrWrWqZc+e3YO3lStXxlrvww8/tMqVK1vWrFntvPPOs169etnBgwfDj69YscLfhx4vW7asffHFF77dyZMnh9d58skn7cILL/TX0DaeeuopD0JEGTVt8/vvv/fn6aZlErkd7Zu2E2nr1q0eyHz99dd+f9++ffb4449b0aJFLUeOHFa9enV/nwnRawwbNsyuv/56f84zzzxjhw4dsjZt2ni3Rn3WpUqVssGDB8fKdI0ZM8Y/m2Cfg9dJ7PdwMhAcAQAAIFXQRfXTTz9tZ599tncT++6773z57t27rUOHDrZw4UIPUNKnT2833HCDHT58OOp2FFD89NNP9umnn9rPP//sF/QFChTwxxRoNGjQwM444wybPXu2ffvtt5YzZ05r2LCh7d+/P8H969q1qw0YMMD3I2PGjHbXXXeFH9O2WrZsae3bt/fXfvXVVz1wUQAhCiLUtUxBz/z5823EiBG+vbi0X3qetqHPY+TIkfbiiy/6Y82aNbPHHnvMLrroIv98dNOyuBQYjh8/3sdvBSZMmGBFihSxyy67zO8/+OCDNnfuXF9v2bJl1rRpU/8MVq1aleBnoGBHn/0PP/zg71/fgb6viRMn+j53797dunTpYu+++66vrwBMAZC2HeyzgrcT+R5OBN3qAAAAkCrkzp3bL5YzZMgQq5vYTTfdFGu9UaNGeVZIF+PRxiQp83TxxRd7lkeCDFQQJOiC/rXXXvMshiirpOyFMhr169ePd/8U6NSpU8d/79SpkzVq1Mj27t3rmSBldLSsVatW/riyPr1797YnnnjCevToYdOnT/dsl14jeG/a3lVXXRXrNbp16xb+Xfut4EIBjLajzIwCCAVmCXWjUzDyyCOP2DfffBMOhsaNG2fNmzf396zPR+9ZPxUwiV5HGTot79u3b7zbvu2226x169axlum9B5RBUtCl4Ej7of3VfitTFbnPb7311nF/DyeC4AgAAACpmrIZykgo47Jt27ZwxkgX99GCo3bt2nlAtXjxYr/IVsYmGMOkLmmrV6/2ICySgpyEuupJhQoVwr9rXJRs2bLFzjnnHN+ush9BpijIFmm7e/bs8S54xYoVixUgVKtW7YjXUPD20ksv+b78+++/3i0vV65cifi0zANHve+3337bg6Pff//dA5ZXX33VH1fWR/um7nuRFMCo22JCgoAz0tChQz1g1ffx33//eebnaFUDT+R7OBEERwAAAEjVrrvuOitevLh3MVOmQ8GRgqL4ul+patuaNWts6tSpnrGpW7euPfDAA/bCCy94wFGlShUPHKIFFQmJLD4QZDuCQE3bVQblxhtvPOJ5yiwdCwUw6hKn7ajLmTJpyhqpK19iaTsPP/ywvfzyy541Kl++vN+CfVV2btGiRf4zkjI9CdFYo0jaP2WdtI81a9b0YOf555/3QDYhJ/I9nAiCIwAAAKRaf/31l2ddFBgFXcTUXexodIGtLm666XkdO3b04EgFE5SdOeussxKdkUmItqv9vOCCC6I+rkIFKkCwefNmK1iwoC8LxlQF5syZ40Fg5FgkBXmRMmfO7Fmfo1Fxg3vuuce7yik40niogLocahvKegWf6fFStkxZufvvvz+8LG7mJ9o+J9X3cDQUZAAAAECqlTdvXu/qpQIG6ob15ZdfenGGhKgLnqqjaf0ff/zR50wqU6ZMOKOi4gwKHlQIQF3ONMZFWZY///zzuPdTrzl27FjP+ug1VQhCWZVgDJHGFp1//vkerKkAgoKK4LEgC6WqeuqapucpwFD3Ok3sGknjkLTPS5cu9S6G6goXX4ZH3QlVnEL7ovFGAXWn0+eggGnSpEm+vQULFli/fv3sk08+scTQPqtAxWeffWa//PKLv17coE/7rPes4FH7rGIMSfU9HA2ZIwAAAISN6dI4VX0aqkynYEEXzepKpwyMgobLL7883ucoU9G5c2cvC61iAMqOaBuianEqZ61S1+oCt2vXLi9nra53J5LBUDc4BWGqtte/f3/vgle6dGm7++67/XF1X1Opbd3/3//+5wUb1P1MXQaDbncqkf3oo496JTkFPSr4oGBDFeICGkulgOaKK66w7du3exEDlUCPRgHINddcY7Vr1/ZxUZH0vD59+nj1u/Xr13ugUqNGDS+Tnhj33nuvLVmyxKvmKchTEKYskioFBtq2beuBj8YrqTvdzJkz/ftLiu/haNKFImv4we3cudP7cO7YseOUpvHis7Hdfcm9C6e9wsOGJ/cuAABwSmlgu1rjVT3sWMe84NRS9kjzHinDpawSjv9v+liv78kcAQAAACmAusip4IG6oikg0pxIl1xyCYHRKURwBAAAAKQA6jqmbmQaV6RubPXq1TuuSnQ4fgRHAAAAQAqgAgiRVeNw6lGtDgAAAAAIjgAAANK2YJJSILU7GX/LdKsDAABIg1TOWmWwN2zY4BOi6n4wnw6Qmqj49v79+23r1q3+N62/5eNFcAQAAJAG6SJSJY83btzoARKQ2mmOKs3XpL/t40VwBAAAkEaphV0XkwcPHrRDhw4l9+4Ax02T6GbMmPGEs58ERwAAAGmYLiYzZcrkNyCto1odAAAAAKSU4Gjo0KFWokQJy5o1q1WvXt0WLFiQ4PoTJ0600qVL+/rly5e3qVOnxrvufffd5y0igwYNSoI9BwAAAHC6SPbgaMKECdahQwfr0aOHLV682CpWrGgNGjSwLVu2RF1/zpw51rx5c2vTpo0tWbLEmjRp4rfly5cfse4HH3xg8+bNsyJFipyCdwIAAAAgNUv24GjgwIHWtm1ba926tZUtW9aGDx/ulSZGjRoVdf3Bgwdbw4YNrWPHjlamTBnr3bu3Va5c2YYMGRJrvfXr19tDDz1kb7/99lH70O7bt8927twZ6wYAAAAgbUnW4Ej1yBctWmT16tX7/zuUPr3fnzt3btTnaHnk+qJMU+T6mgDqjjvu8ADqoosuOup+9OvXz3Lnzh2+FStW7ITeFwAAAIDUJ1mDo23btnnZyIIFC8ZarvubNm2K+hwtP9r6/fv391J+Dz/88DHtR+fOnW3Hjh3h27p1647r/QAAAABIvU67Ut7KRKnrncYvHWud8yxZsvgNAAAAQNqVrMFRgQIFfMKmzZs3x1qu+4UKFYr6HC1PaP3Zs2d7MQdNaBZQduqxxx7zinV//PFHkrwXADjZNra7jw/1FCg8bDifMwAg+bvVaVbmKlWq2IwZM2KNF9L9mjVrRn2OlkeuL9OnTw+vr7FGy5Yts6VLl4Zvqlan8UefffZZEr8jAAAAAKlVsnerUxnvVq1aWdWqVa1atWqe3dm9e7dXr5OWLVta0aJFvWiCtG/f3urUqWMDBgywRo0a2fjx423hwoU2YsQIfzx//vx+i6RqdcoslSpVKhneIQAAAIDUINmDo2bNmtnWrVute/fuXlShUqVKNm3atHDRhbVr13oFu0CtWrVs3Lhx1q1bN+vSpYuVLFnSJk+ebOXKlUvGdwEAAAAgtUsXCoVCyb0TKY3mOVJJb1Wuy5UrV3LvDuMOTgHGHCAlYszRqcHxDwCnv53HeH2f7JPAAgAAAEBKQHAEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABAj4//9BAAAQAqxsd19yb0LaULhYcOTexeQwtCtDgAAAAAIjgAAAAAgBpkjAAAAACA4AgAAAIAYZI4AAAAAgOAIAAAAAGKQOQIAAAAAgiMAAAAAiEHmCAAAAAAIjgAAAAAgBpkjAAAAACA4AgAAAIAYZI4AAAAAgOAIAAAAAGKQOQIAAAAAgiMAAAAAiEHmCAAAAAAIjgAAAAAgBpkjAAAAACA4AgAAAIAYZI4AAAAAgOAIAAAAAGKQOQIAAAAAgiMAAAAAiEHmCAAAAAAIjgAAAAAgBpkjAAAAACA4AgAAAIAYZI4AAAAAgOAIAAAAAGKQOQIAAAAAgiMAAAAAiEHmCAAAAAAIjgAAAAAgBpkjAAAAACA4AgAAAIAYZI4AAAAAgOAIAAAAAGKQOQIAAAAAgiMAAAAAiEHmCAAAAABSSnA0dOhQK1GihGXNmtWqV69uCxYsSHD9iRMnWunSpX398uXL29SpU2M93rNnT388R44cljdvXqtXr57Nnz8/id8FAAAAgNQsY3LvwIQJE6xDhw42fPhwD4wGDRpkDRo0sJUrV9pZZ511xPpz5syx5s2bW79+/ezaa6+1cePGWZMmTWzx4sVWrlw5X+fCCy+0IUOG2HnnnWf//fefvfjii1a/fn1bvXq1nXnmmcnwLgEAAJDStOr7YXLvwmlvTJfGlpoke+Zo4MCB1rZtW2vdurWVLVvWg6Ts2bPbqFGjoq4/ePBga9iwoXXs2NHKlCljvXv3tsqVK3swFLjttts8W6Tg6KKLLvLX2Llzpy1btuwUvjMAAAAAqUmyBkf79++3RYsWeSAT3qH06f3+3Llzoz5HyyPXF2Wa4ltfrzFixAjLnTu3VaxYMeo6+/bt8+Ap8gYAAAAgbUnW4Gjbtm126NAhK1iwYKzlur9p06aoz9HyY1l/ypQpljNnTh+XpG5106dPtwIFCkTdprroKXgKbsWKFTvh9wYAAAAgdUn2bnVJ5YorrrClS5f6GCV1w7vllltsy5YtUdft3Lmz7dixI3xbt27dKd9fAAAAAGk4OFImJ0OGDLZ58+ZYy3W/UKFCUZ+j5ceyvirVXXDBBVajRg17/fXXLWPGjP4zmixZsliuXLli3QAAAACkLccVHM2ePdtatGhhNWvWtPXr1/uyN99807755ptEbSdz5sxWpUoVmzFjRnjZ4cOH/b62HY2WR64v6jIX3/qR29XYIgAAAAA4KcHR+++/7wUQsmXLZkuWLAkHHOqO1rdv38Ruzst4jxw50saMGWM///yztWvXznbv3u3V66Rly5be7S3Qvn17mzZtmg0YMMBWrFjhcxotXLjQHnzwQX9cz+3SpYvNmzfP1qxZ4wUf7rrrLg/imjZtmuj9AwAAAJA2JDo46tOnj5fbVkCTKVOm8PJLLrnE5xpKrGbNmtkLL7xg3bt3t0qVKvk4IQU/QdGFtWvX2saNG8Pr16pVy+c2UgU6VZ977733bPLkyeE5jtRNT0HTTTfd5PMdXXfddfbXX395tktlvQEAAADgpEwCq8lZa9eufcRyVXnbvn27HQ9lfYLMT1yzZs06YpkyQPFlgVSdbtKkSce1HwCAtIdJIE+N1DYRJIC0KdGZIxU+WL169RHLNd5Ik64CAAAAQJoIjtq2bevjfubPn2/p0qWzDRs22Ntvv22PP/64jxcCAAAAgDTRra5Tp05e+a1u3bq2Z88e72KnUtgKjh566KGk2UsAAAAASGnBkbJFXbt2tY4dO3r3un///dfKli1rOXPmTJo9BAAAAICU2K1OZbF37drlcxQpKKpWrZoHRiqhrccAAAAAIE0ER5qP6L///jtiuZaNHTv2ZO0XAAAAAKTMbnU7d+60UCjkN2WOVDI7cOjQIZs6daqdddZZSbWfAAAAAJAygqM8efL4eCPdNLlqXFreq1evk71/AAAAAJCygqOZM2d61ujKK6+0999/3/Llyxd+TOOPihcvbkWKFEmq/QQAAACAlBEc1alTx3/+/vvvVqxYMUufPtHDlQAAAADg9CnlrQyRaI6jtWvX2v79+2M9XqFChZO3dwAAAACQUoOjrVu3WuvWre3TTz+N+riKMwAAAABAapPovnGPPPKIbd++3ebPn2/ZsmWzadOmeXnvkiVL2kcffZQ0ewkAAAAAKS1z9OWXX9qHH35oVatW9XFH6mZ31VVXWa5cuaxfv37WqFGjpNlTAAAAAEhJmaPdu3eH5zPKmzevd7OT8uXL2+LFi0/+HgIAAABASgyOSpUqZStXrvTfK1asaK+++qqtX7/ehg8fboULF06KfQQAAACAlNetrn379rZx40b/vUePHtawYUN7++23fa6jN954Iyn2EQAAAABSXnDUokWL8O9VqlSxNWvW2IoVK+ycc86xAgUKnOz9AwAAAICU163uwIEDdv7559vPP/8cXpY9e3arXLkygREAAACAtBMcZcqUyfbu3Zt0ewMAAAAAqaUgwwMPPGD9+/e3gwcPJs0eAQAAAEBqGHP03Xff2YwZM+zzzz/38t05cuSI9fikSZNO5v4BAAAAQMoMjvLkyWM33XRT0uwNAAAAAKSW4Gj06NFJsydAMmrV90M+/1NgTJfGfM4AAOD0GXMEAAAAAKcjgiMAAAAAIDgCAAAAgBhkjgAAAAAgscHRgQMHrG7durZq1So+PAAAAABpNzjKlCmTLVu2LOn2BgAAAABSS7e6Fi1a2Ouvv540ewMAAAAAqWWeo4MHD9qoUaPsiy++sCpVqliOHDliPT5w4MCTuX8AAAAAkDKDo+XLl1vlypX9919++SXWY+nSpTt5ewYAAAAAKTk4mjlzZtLsCQAAAACk1lLef/75p98AAAAAIM0FR4cPH7ann37acufObcWLF/dbnjx5rHfv3v4YAAAAAKSJbnVdu3b1anXPPvusXXLJJb7sm2++sZ49e9revXvtmWeeSYr9BAAAAICUFRyNGTPGXnvtNbv++uvDyypUqGBFixa1+++/n+AIAAAAQNroVvf3339b6dKlj1iuZXoMAAAAANJEcFSxYkUbMmTIEcu1TI8BAAAAQJroVvfcc89Zo0aNfBLYmjVr+rK5c+faunXrbOrUqUmxjwAAAACQ8jJHderU8clfb7jhBtu+fbvfbrzxRlu5cqVddtllSbOXAAAAAJCSMkcHDhywhg0b2vDhwym8AAAAACDtZo4yZcpky5YtS7q9AQAAAIDU0q2uRYsWPs8RAAAAAKTpggwHDx60UaNGeUGGKlWqWI4cOWI9PnDgwJO5fwAAAACQMoOj5cuXW+XKlf13FWaIlC5dupO3ZwAAAACQUoOjQ4cOWa9evax8+fKWN2/epNsrAAAAAEjJY44yZMhg9evX9/LdAAAAAJCmCzKUK1fOfvvtt6TZGwAAAABILcFRnz597PHHH7cpU6bYxo0bbefOnbFuAAAAAJAmCjJcc801/vP666+PVYAhFAr5fY1LAgAAAIDTPjiaOXNm0uwJAAAAAKSm4KhOnTpJsycAAAAAkJrGHMns2bOtRYsWVqtWLVu/fr0ve/PNN+2bb745rp0YOnSolShRwrJmzWrVq1e3BQsWJLj+xIkTrXTp0r6+yopPnTo1/NiBAwfsySef9OWaoLZIkSLWsmVL27Bhw3HtGwAAAIC0IdHB0fvvv28NGjSwbNmy2eLFi23fvn2+fMeOHda3b99E78CECROsQ4cO1qNHD99exYoVfftbtmyJuv6cOXOsefPm1qZNG1uyZIk1adLEb5qcVvbs2ePbeeqpp/znpEmTbOXKlT5GCgAAAABOarW64cOH28iRIy1Tpkzh5ZdccokHI4k1cOBAa9u2rbVu3drKli3r286ePbuNGjUq6vqDBw+2hg0bWseOHa1MmTLWu3dvq1y5sg0ZMsQfz507t02fPt1uueUWK1WqlNWoUcMfW7Roka1duzbqNhXgUXUPAAAASNsSHRwpC1O7du0jlisoSezksPv37/egpV69ev9/h9Kn9/tz586N+hwtj1xflGmKb/0gq6VKenny5In6eL9+/Xz/g1uxYsUS9T4AAAAApMHgqFChQrZ69eojlmu80XnnnZeobW3bts1LfxcsWDDWct3ftGlT1OdoeWLW37t3r49BUle8XLlyRV2nc+fOHkAFt3Xr1iXqfQAAAABIg9Xq1AWuffv23u1N2RgVOlDWRhPDapxPSqLiDOpepzmYhg0bFu96WbJk8RsAAACAtCvRwVGnTp3s8OHDVrduXS9+oC52CiwUHD300EOJ2laBAgUsQ4YMtnnz5ljLdV8Zqmi0/FjWDwKjNWvW2Jdffhlv1ggAAAAAjqtbnbJFXbt2tb///tsrxM2bN8+2bt3qhRESK3PmzFalShWbMWNGeJkCL92vWbNm1OdoeeT6ogIMkesHgdGqVavsiy++sPz58yd63wAAAACkLYnOHEUGNqoud6JUxrtVq1ZWtWpVq1atmg0aNMh2797t1etEcxQVLVrUiyaIuvRpItoBAwZYo0aNbPz48bZw4UIbMWJEODC6+eabvXLelClTfExTMB4pX758vt8AAAAAcNKCo5OlWbNmnnnq3r27BzGVKlWyadOmhYsuqPy2KtgFNPHsuHHjrFu3btalSxcrWbKkTZ482cqVK+ePa1Lajz76yH/XtiLNnDnTLr/88lP6/gAAAACkDskeHMmDDz7ot2hmzZp1xLKmTZv6LZoSJUp4AQYAAAAASNIxRwAAAABwOiI4AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACAGmSMAAAAAIDgCAAAAgBhkjgAAAACA4AgAAAAAYpA5AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACAGmSMAAAAAIDgCAAAAgBhkjgAAAACA4AgAAAAAYpA5AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACAGmSMAAAAAIDgCAAAAgBhkjgAAAACA4AgAAAAAYpA5AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACAGmSMAAAAAIDgCAAAAgBhkjgAAAACA4AgAAAAAYpA5AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACAGmSMAAAAAIDgCAAAAgBhkjgAAAACA4AgAAAAAYpA5AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACAGmSMAAAAAIDgCAAAAgBhkjgAAAACA4AgAAAAAYpA5AgAAAACCIwAAAACIQeYIAAAAAAiOAAAAACCFZI6GDh1qJUqUsKxZs1r16tVtwYIFCa4/ceJEK126tK9fvnx5mzp1aqzHJ02aZPXr17f8+fNbunTpbOnSpUn8DgAAAACcDpI1OJowYYJ16NDBevToYYsXL7aKFStagwYNbMuWLVHXnzNnjjVv3tzatGljS5YssSZNmvht+fLl4XV2795tl156qfXv3/8UvhMAAAAAqV2yBkcDBw60tm3bWuvWra1s2bI2fPhwy549u40aNSrq+oMHD7aGDRtax44drUyZMta7d2+rXLmyDRkyJLzOHXfcYd27d7d69eqdwncCAAAAILVLtuBo//79tmjRolhBTPr06f3+3Llzoz5Hy+MGPco0xbf+sdq3b5/t3Lkz1g0AAABA2pJswdG2bdvs0KFDVrBgwVjLdX/Tpk1Rn6PliVn/WPXr189y584dvhUrVuyEtgcAAAAg9Un2ggwpQefOnW3Hjh3h27p165J7lwAAAACcYhktmRQoUMAyZMhgmzdvjrVc9wsVKhT1OVqemPWPVZYsWfwGAAAAIO1KtsxR5syZrUqVKjZjxozwssOHD/v9mjVrRn2OlkeuL9OnT493fQAAAABI8ZkjURnvVq1aWdWqVa1atWo2aNAgL8Wt6nXSsmVLK1q0qI8Jkvbt21udOnVswIAB1qhRIxs/frwtXLjQRowYEd7m33//bWvXrrUNGzb4/ZUrV/pPZZdONMMEAAAA4PSVrMFRs2bNbOvWrV56W0UVKlWqZNOmTQsXXVCQowp2gVq1atm4ceOsW7du1qVLFytZsqRNnjzZypUrF17no48+CgdXcuutt/pPzaXUs2fPU/r+AAAAAKQeyRocyYMPPui3aGbNmnXEsqZNm/otPnfeeaffAAAAACAxqFYHAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAABAcAQAAAAAMcgcAQAAAADBEQAAAADEIHMEAAAAAARHAAAAABCDzBEAAAAAEBwBAAAAQAwyRwAAAACQUoKjoUOHWokSJSxr1qxWvXp1W7BgQYLrT5w40UqXLu3rly9f3qZOnRrr8VAoZN27d7fChQtbtmzZrF69erZq1aokfhcAAAAAUrNkD44mTJhgHTp0sB49etjixYutYsWK1qBBA9uyZUvU9efMmWPNmze3Nm3a2JIlS6xJkyZ+W758eXid5557zl566SUbPny4zZ8/33LkyOHb3Lt37yl8ZwAAAABSk2QPjgYOHGht27a11q1bW9myZT2gyZ49u40aNSrq+oMHD7aGDRtax44drUyZMta7d2+rXLmyDRkyJJw1GjRokHXr1s0aN25sFSpUsLFjx9qGDRts8uTJp/jdAQAAAEgtMibni+/fv98WLVpknTt3Di9Lnz69d4ObO3du1OdouTJNkZQVCgKf33//3TZt2uTbCOTOndu76+m5t9566xHb3Ldvn98CO3bs8J87d+60lGDX/v3JvQunvf179yT3LqQJKeWYSi049k8Njv9Tg+M/cTj+Tw2O/7Rz7O/8v/1QIiXFBkfbtm2zQ4cOWcGCBWMt1/0VK1ZEfY4Cn2jra3nweLAsvnXi6tevn/Xq1euI5cWKFUvkO0LqNTq5dyBNGN87ufcAiIbj/1Tg+EfKxPGf1o79Xbt2eeIkRQZHKYUyV5HZqMOHD9vff/9t+fPnt3Tp0iXrvuHUtCQoEF63bp3lypWLjxxIQzj+gbSL4z9tCYVCHhgVKVIkwfWSNTgqUKCAZciQwTZv3hxrue4XKlQo6nO0PKH1g59apmp1ketUqlQp6jazZMnit0h58uQ5zneF1EqBEcERkDZx/ANpF8d/2pE7gYxRiijIkDlzZqtSpYrNmDEjVtZG92vWrBn1OVoeub5Mnz49vP65557rAVLkOmoZUNW6+LYJAAAAAMnerU7d2Vq1amVVq1a1atWqeaW53bt3e/U6admypRUtWtTHBUn79u2tTp06NmDAAGvUqJGNHz/eFi5caCNGjPDH1Q3ukUcesT59+ljJkiU9WHrqqac8haaS3wAAAACQIoOjZs2a2datW33SVhVMUNe3adOmhQsqrF271ivYBWrVqmXjxo3zUt1dunTxAEiV6sqVKxde54knnvAA65577rHt27fbpZde6tvUpLFAXOpSqXm24natBHD64/gH0i6Of0STLnS0enYAAAAAkAYk+ySwAAAAAJASEBwBAAAAAMERAAAAAMQgcwQAAAAABEcAgNOV5s0DAM4HSAwyRzjtaULgt99+28u7A0g7gmkg1q9fb3v27Enu3QGQjCKnhQESwl8KTlufffaZz4OliYSfeeYZq1evnn3++efJvVsATgLNQqHMUHzZoZ07d9qDDz5oefPmtcsuu8zPAz/99BOfPXCai++c8Ntvv9kHH3xwyvcHqQ/BEVI9TRT88ssv25IlS/z+oUOHfPLfF1980WrUqOGtxu+//74VK1bMHnnkEduwYUNy7zKA46TjW9KlS+ctwbpFuxgaNWqUzZs3z8aMGWPjx4+3FStWWPv27e3777/3x5niD0j9dOwfPHgwfF/HdZAhCs4VgeHDh1uHDh3CzwPiQ3CEVEcnv9mzZ9stt9xiBQoUsPPPP99Gjhxp//zzjz+eIUMGW7VqlS1atMjuu+8+X1amTBkbPHiwr/Pee+8l8zsAcLx0fIvOAQ8//LDVqlXLbr/9dps5c6bt378/fI5QUHTJJZfY9ddfb9WqVQt3rdUFUrAOgJRHx2ZkYPPWW2/ZXXfd5Q2hoseC4EaBUMaMGcPrqtHk999/t//973/2xhtvhJcfOHDAG0qvvfba8POA+PDXgVQjuPDRCbJbt242ffp0+/DDDz3gWbZsmV155ZXhdXPnzm1//fWXB07BybRw4cJ+wpw7d65t3bo12d4HgOOnY18XNjfeeKOtW7fObrvtNtu7d6/de++9NnnyZF9n+fLlljlzZrvgggvCzytdurRfGCmLLFwcASkrGAoaLBTgBI0gO3bs8K6xV199tRUqVMiX6bHg+FUjqI79m266yT7++GNfdu6553pX2j59+nhXOsmUKZNfJ5QoUSL8mkB8CI6QYunk9eabb4YvcHSxI0WKFPET38UXX+wtw9mzZ7cff/wxVmpdrUS5cuWypUuX+v2gFUqtzGvWrPFb8BoAUr7g+NZFzllnneXd4zR+QOOKBgwYYOecc4599913vk7BggU9YNJ5IJAlSxarU6eObdu2zTPLAFKGIBjST3V7HzFihNWuXdsDoCpVqljFihWtadOm4WsAqVu3rr366qv2wgsveLGVrFmzeoCkLrQycOBAD6Z69uxpf/75p/+vz5YtW/g8wv9+JITgCCmG0uSRqXSdKCtXruwtPyqk0KNHD3vggQd8nbJly3rrcLly5Tw7pC52OjFOmzbNn5snTx5vKf7oo4/C2xZ1r1Frc3CC1GsASF66UNFNrcBqKY4maEm++eabPcBRF5mAGkiUUb7uuuv8voKnnDlzehCkwgwBZY819lBd8oLXBZB89L947Nix/v87f/78dvbZZ3t3eDVkqLFj5cqVvuzJJ5/0m3qEiHqMtGvXzhtJ1YiqbrPNmzf38ccKhkSFmPS7utSrS62uA5RVEjLHSAjBEVIMnayCC6DA888/7z/Vhearr77yYEktRMomNWrUyC+U1JWmb9++3lKsk6WyRUWLFvVudlOmTPHn6zmi5+mi6swzz0yGdwggGjVSqOS+ur0GJfcjsz7BOgpmLrroIm8BVoOJii40btzYzjvvPD93KEDavHmzr6+ssqrT/fLLL+FtqOVZGaaNGzfyRQApgDLAyvzmy5fPM8Fdu3a1K664wrvKKWsUNGT+/fff9s0334SPb3Wn1f91VaEN3HnnnX68B13pFGA9+uijnmHS+UXd6pSFAo6G4AinlPoEP/HEE97yK0GmSBdCCnIU3KiV59dff/XlN9xwg1WvXt1PcrNmzbI2bdr4cmWOtB21JOlkpwskPV8tyMG4g7Zt23qrkU6MwQWXBmMrQMqRIwffPJCEgqxM3OxM3OpSQVZXFzkKXu655x4rVaqUZ3+CC6G4mV6NP3jqqadsyJAhfjyPGzfOs8edOnXyipRBg4qO+2CMkag77eLFi/18Erk9AEln9erV/j89skJccF5Qg6cCJBVVUle6qlWr+nH75ZdfxjpG1a1OjRp//PGH32/QoIE3iEbOX6jjWq+hRhFtXw0myiY3adLEg67IxleyxkgIwRGSjFpv4gZBarnR/EPBfCM6WW3ZssVuvfVWe/zxx71LjbrG6cSnlLqCnlatWnkRBQlOruoyowp0akEOlut39TFWt7l///3X0+cavN27d2+7++67vVVJVW/UPU/rcXIETo5oZXF1UaMLl7gBSGR1qV27doW7tyi4UdEVdZdVFTr91NghCdbXeUDHthpNVHRFlSfV6qyLH40t6N69uy/TuAVloXTc9+/f3wdm67wyaNAgP7dUqlSJrx44SXONxV2m417FkvS/WNTLQ71AgmNd/+d1XtBzg4xw0ABSoUIFP7bViBF57F966aV+HaGusnpe+fLlvZFTU3joflDCW89fsGBBrO60OgdceOGF/n8/uGagYQQJITjCSRUEHLpIUYuuLk7UZzhosVFXNwU2kYUS1OqrFt0ffvjBf//000+94ELHjh39hFmzZk0/meo58fUT1nK1Mv/888/eb1mvIQqE3nnnHU+/a99ef/11T8frd06OwMkR7bhUlliVoYIuLsFF1MKFC61Zs2ZWvHhxz+4o+6MB1e+++6517tzZL3g0Yau6xoqKragrnQqstG7d2s8DGoytc4IulILxSsHFls4pCrp0ztHYhaFDh9rXX39td9xxh2eoe/XqFT4/ADh+wVxjATVuaNlLL73kDRxBo2j9+vW9UVSZHQU+GgsUd86yIAhSo6a6vurY1nlB29d66hWiSd11/KtBNSiwpB4lWi/4fx5M9h6MSdQ5QWMNlUHS9YS2TcMojobgCCckaPGJPMkFJy3RhZEyP0p/S3CRE5k5UsDy9NNP+9wECqqUWldAowsZnQR1stSYgqDYQmQf5Dlz5ngLkdLyurBSEYb7778/vH/avgKt0aNH+7wnV111le8jgRGQ+GM92kWFsjRqmVX3teA4F5XfVdc4jRUUXeRoTJAyPGq9HTZsmM9BpAaRoCucMkJqTFHWKHI7OodoO7owUguytqtWYLUuB8ezxhqoUpXGIQYlf0VddZVNUvl+dblVizOAE6f/4+rdof/fom6xyuyq4InK5qtxRA2faqBQF3f9f9b/8WeffTZceVLU5W7+/Pn233//+X2tp//vwXkguL64/PLLvbEkmMhdjSu6Bgi634qyyDpHKAgLzjui56mAk64f+P+PowoBJ8nBgwdDW7du9d///fffULp06UIvvfRS6Nxzzw3deuutoW3btvlj999/f6hBgwahX3/91e9XrVrV1y1UqFCobt26oeeeey70/fffh/bv3++PHzhwIHTfffeFKlWqFH6tQ4cOhTZs2BCqXr16qHjx4qEcOXKEGjduHPrqq6/4PoEktG/fPv+5d+9e/zlw4EA/frNkyRJ65ZVXwuutW7fOj0kd+wE9Xq1atfD93bt3hx5//HF//s6dO31Z9uzZQy+//LIf49HofCAtW7YMlShRItSmTZvQ+eefH8qWLVvo2muv9XMHgONz+PDheB/TManHg3V27doVGjZsWChv3ryh9u3b+7Lff/89lClTptCyZcvCz9u0aVMoT548oYkTJ4aX7dmzJ9SlSxd/rtbv3Llz6O+///bHZsyYEapVq1ZoyJAhfj+4Fli4cGGoWLFioQkTJvj9tWvX+rljwYIF8e6zXuett94K5c6dO/T222/zZ4Fj8v+nFQbiobS0bpGzUAfUgqOKUWqZ3b59uw+m1rwjyg4p46PWH5XYVCEFjSlSlkhd69RKtGLFCs8IqRSnWpw1xiioKidqgVKfZa2j7jTKJl1zzTXesqxKVCr/qUxTgQIFvEQ3gKSjzIsqSKkUtoqmaN4gUWZWRVF0flAGSa3HOt51XGpsTzDviHzyySc+FqhLly5eSVLZYm1P4wGVAT7jjDP82FZVKi1LqPubuuapK57OE+oqpyyUng/g+EXLqgTdVuN2n9Xxqa6rGs+jUtxaTyW2VW0u6BaraweNHdRxr2p0yuwGr6NiDLp+0DEfWUFWPUx0X13tIzNMql6n19B1hbJJOneoh0ncKrfBGCQt1zXKzJkz/fojKPUPHNWxxVBIS9mfhAQtOIF27dp59kYtwl9//XVo9uzZoc2bN/tjDz/8cKhixYr+nOnTp4cqV64cuuuuu0LLly/3VqE+ffr4ep9++mkoQ4YM3roTtB6vWbMm1LNnT99u8JqffPJJ6KmnngqNHz8+3GoN4ORTy7DOBXFbkZXlqVChQujDDz8MNW3aNPTaa6/58Vm7du3QY4895sesWmjV8iuTJk0KFSlSJPTdd9/5fT1HLb2NGjUKDR8+PLRixYojMkRaftZZZ4WmTp3qLc7Tpk3z5fFlkgAk/viO9vt///3nx/jSpUuPeI56fowZMybUo0cP/x8eXCsEmVwd802aNPHj+5lnnomVZZYXX3zRsz47duw4pn1URljXD8H1RHD8K1sVd985N+BkIzhC6LPPPgvdc889oUsuucRPSHPnzg2f+PTzyy+/DN18882hUqVKefc2rS9z5swJlS1bNt5UtYIlpcuDbi66QCpYsGDogQce8O2pW0xw8tQJWRdEN954Y6hOnTr++6WXXhqaNWsWJz4gBTSOKJDRcamLH3WVvffee8PHduvWrUPXXXdd6K+//go99NBDodKlS/t54o8//ghdfPHF4YaQ/v37e1c4XVxFXtysXLky9M4774Qvwlq0aBEqXLiwnz/U1ZaLH+DkU9fXQHAspk+fPhzcBNQgec4553jDiLrEX3jhhaGuXbse8dx3333Xzw/lypXzLnASHLs//vijN4IerdtrsL4aTdUNTwFbYrr+AScDwVEa9vHHH3s/4Jw5c/rFiPr3KkBS/3212ooyQf/73/9Cbdu29X6+d999twdEX3zxRWj9+vWhyy67LHT11VeHunfv7i1Dain+6aefwq+hi5s33ngjfNH1wQcf+Pb1mnquTpjBCVEn0169evmFFOMGgJMjvguJhAIOXTTpOGzevHloypQp3oihFluNL1KApMxRpBEjRoSqVKni5wu9nsYVXnTRRd5wosaQK6+80tfTOENljZVt1vhAnUMmT57sDSWdOnUKZ4n/+ecfb70mKAJOPv3fVQOGAhkd35Fuv/32UMOGDcNByc8//+zjfSPHEypLlCtXLm+8DOh/vDLDamC96aabvCF05syZsbatLLIaWDVWSI2nv/zyiy/nOEdKQ3CUhuniQ626ahUKzJs3z4MXZZKCE6W6xwV00aKWXF0IBdkjdZHRerqQKlCggAdPixYt8scVWKkrnQZeB8aOHesFFM444wzPSgE4+X744YfQCy+8EGuZApeELkQUyKiRQ8e8Gj1UUEHBUJD5UcGVevXqeTAjQfdWZYJ0LhkwYEC4IIu63ObPn9+3qXNKkCXWhdlVV10VqlGjhjfOqBCLgqnIRhUASUeZ2muuucaDI/0/V5f2wLhx40L58uXzrG+QwVHGSBTsPPLII36sKwukhs+gSIO6y+l5Wl/uvPPO0HnnnRd6/fXXw9t+9dVX/ZpAjaZFixY9opElLoImJBeCozRMAYsuUp544onwMnV904lLJ0CdmNQ6NHr0aE+hK1WugEYtTrof2fdXrUyqCqNKNboQ0tggUf9knTA1hiigCyq1RukCCkDSGDp0qHdxi9YtRcvUYqxMblAhShT06IJJAY0ok6NxRLqQ0XhABTjdunXzi564Fy8KmtTVLng9tSSrBVrV57TNyEqSemzJkiV+vgBwckUbLyj6n/vbb7/5cduxY0cfAzRo0CD/vx6MC9QYH3WtC7rPjxw50scRqkFT54EbbrjBg5y4WR9tR0GUxgkGY5GeffZZbyAJGlt1/li1alVo48aNfOVI0QiO0jh1l1PLkU5s6kusk6AuZIJ0udLpCpZ0QlR5XQU1cVtz/vzzz3ALsrJQKs0dtB5pnIKeGwyqBJB0dGwm1NqqiyaN+1H5XI0bUhc3daVVBljUxS3y+BeVzNeyoMiC1tHFVDBeIegK16FDB2+N1jki8rkKuDRGIXgNAElfPCVaY+jll18eLrmtLq/q3q5Gzttuu83HFAc9PnS8ahywqDu9eoM8//zzR2wzmLpDmjVr5j1IJOhGr3ORziVBIBVtv4GUiElg0zhNirpkyRKfkFETp6l8riZi1OStKq2rCRNr1KhhkyZN8hLdmpxNli1b5qW3NRN9jx497Pbbb/eZpx966CEvt92wYUNfr1SpUv5cTdoI4MR8+umn9sYbb3j52miTs6rUblBuV5Mdfvzxx/bKK6+E1/3222/t1Vdf9QlbdYx/8cUXXh73iSee8Mdr1qzpJbl3794da3Z5leVXOVzROUBlulWiP7L0b/Xq1X0y5mDiRu2TnqvJl1V+X9sGcOJ0bOnYDI4/layOLMG9YMECu/XWW23fvn3h9bNnz26rVq3yktpywQUXeClunQN0LJ9//vk+gfrPP//s/8+nT5/uz9NUG0WKFPGJXYPJWEWPa+J1ldLXBKuaYPmBBx7wx4LS2tqnOnXqWMmSJY94D8F+AykRwVEaV7VqVT9xPfroo9atWzcbNGiQzwcwZMgQnzuka9eutmjRIl+mOUk0v8DEiROtX79+Pqu15hVRgKULrJEjR9qmTZv8eQBOnGZ+V+OD5g377LPP7JdffvHjMO5Fhm7btm2zN99805577jmf20PBkeYfiwyOdF8NIJqHRI0imq9IM9YraNIFjhoxFPx8+eWX4W1LvXr1bNasWX6xpTlNNFeJzhEKeBQ4ffTRR94oMnr0aLv22mtjPRfA0QUNHMcics4h/c8dPny4vfDCC36+kNy5c/v5onfv3t7QofW1no5vNWiKAp4KFSr4XGEycOBAP5bvuOMOn6NI5xkFU7o+ePrppz040nxEaiTVvIKa0zBbtmwepOmc8ddffx3RABKcAxLz3oCUgElg07izzz7b8ufPb4sXLw4vU6ZIrcu1a9f2yR2fffZZb/3VxZEmbs2XL5/ddtttnmkKJmMEcOLU4KBGhuLFi1u7du28RVe3l156yTMzDRo0iPo8NV5oMkUdz2rpVaCjRgtN0Lpw4UL76aefrGzZsh4QrV692oYOHerBkyZp1SSOCpbKlCnj21LWV63Cyk4Fky9qglXt1x9//OHZYF10aXLmPXv2+D5p0kbRawI4uiDjGwQ5cRsTIicyjUuNH8og6xicMGGCrV271jM4U6dOtT59+litWrX8/7YaK3RO0DGu5+i1ChUq5NtQ9lfBjP63i45rTdKuY1j//xVEKQusDJPW07lDvUB0raDM0o033uiTsAZy5MgRniw2LhpKkOokd78+JD8NzNR4I81REjmp29NPPx3Kli2bD9xWv2SNI9JgSgDHL6gYF3ds0Pz5832AdM2aNb1ilAZP33rrraFWrVrFWq9fv37hMX1BGW2NH1JxBVHhhKDYiapLafJlFWeQLl26+HghPT+ygmTwvOA5WbJk8bLbAW1P1eWoKAckTfU4zemj8XrRxuGo8qTm/wqouIlK5avS47Bhw3yZSmNrjkDNOSb6f67iKVpH4wJVNj9jxoyh1atXh7ejanEaZxxZGEW/a45BjTPU+SK4HogPcw7hdES3OniXmY0bN/oYokjqaqcxBI0aNfJWIbUoqxUJQOJFjhEIxgZt377du7SpxVXjgzRGZ86cOda8eXM/5vQcHYOVK1e2yy+/3J+v1l219gbjgtRyrKxSkyZNfEyAurrquaJxBBoLqNeQK6+80vbv3+/d4zQGQfQcdYt7+eWXPVukLrJaR2MPAtre3Llzw9klACdm/fr11qFDBzvzzDP92FaWp0WLFt41VpQN0nkga9as/r9XY4jeeuutcJZHXeKyZMniWSFRT4769et79zdRD4+ePXt61zd1mde4P3WbU7Y3oP/nOl+ou63ovFCiRAk/x2iMobJQGoMY9zymc0bQVY6sEE5HdKuDn2TVLSboQhOcDDVYMxhIyQkQODFB95nPP//c+/lrTIAukDRGR8GJLnR0odK/f3+/SJk/f75foOh4VOCjsUfSsmVLGzdunP3333+2d+9e366O1bgXMFquYEsBjcYV6YKmbt261qZNG9+Gxh6pa81XX33l4xE0mDoIrjR+IG/evHzlQBLQ8d23b1/voqYubDouRQ0hCmB0/L722mvejV0NGwqQXnzxRQ+E1FCiLrIqlqTH1MCSJ08e/x9drlw537YaMtQVTl3yFCCNHTvWvv76a388cvxPwYIFfaxg0HVP5yDRuELdEjqPAae15E5dAUBa8PHHH/ukpyqf26JFC+86t2XLFn9MZbI1j4jK5msdlc1VF1ZNkqxS25pXJPD99997lxd1w1N3GU22eO2118bbvUUle9UFZ+7cueEucu+//75PznzZZZf5XGSR5bcBJC0d+zqGNZdQtG50mhpD3d1GjRoVa7nmEHzsscf8948++ihUpkyZ0NSpU2N1v6tevbp3iQ9obqGBAwf6611wwQXHPLEq3eWQltEEAACngAYvawC1uqyo64y6zKgYiqiVVi28ahXu2LGjF1hQlxeV3VUmR0UVIjO96uY2b948z/aqmIJaoIMKc6KqVcoWKbOk7ehnUIpbz9Vgau2HWpP1ukGJfgBJS1khFU9RxTd1g4ssuBBkddatW2eZM2f2qnGi41caN27sWSFNs6FzhqrPqVBSQOsrO/TJJ5/4fWWCtR1VmFPmSRnraJkfFWeJi94iSMsIjgDgFFAXVQVDKo0fPgGnT+9daXQxpIBIAdQPP/wQHhcQdG3V+ANViguoGpWCIY0N0lgEVYxTeW11u1H3O3XVe+edd3wMkbrMqvy+xjcASF465nU8q0FCFeaiBSQKatR1To0Xkct13GvskLq/qeqc1ok8n6grrI53dcPTuSEIvNQ1t3379l7JMpq444qAtI7gCABOARVA0GBnXcxofiH19VfLr8pta9yP6GJH4380v0hABRJ27NgRnlxVbrrpJs8WBS3Mmp9sxIgRPoZJLcsaU6T5joLiDLooCsYTAEheCoxUtj/ICMXNHGkOIpXZ12TNEhy7KragcUI65jWXkTJFKtOvMt0BTcSubJLWCSi4ipw4FkDCCI4A4BRRNxrNFzJs2DCvMKdKcwpwOnXqFA6EVGhh2bJlsSZq1oWOMkoBBVZ//vlneEJYTRLbtGlTr3ingg+6QNJFFICUR9XpdB6IbASJzBApu3T33Xd799jrrrvO5x1T44e6yykzHKynRpAFCxZ4MBWZBYoWCEVOHAsgYRwpAHCKKPhRNzmVyddFjsYLqQVY3WhEY46UYdIkrpGtzOoeo8pU6iojmthRJb9Vhh9A6nLHHXd4APP8888f8ZhK6KvhRMe8giFlf1u3bu2ZYHWb1dQaAXXD1bkgLgIh4MTQ0RQAThFdyGjckbrERZbcDsYG6HGV31bXO5XT1rp6TGW21QUvKLevC6saNWrwvQGpkEroq5R39+7dvcGkVatWXjhFDR4q4X/11Vf7uUBl+CtWrOjnALrFAqcOwREAnCIKbjRgWuOHlC3SRY8GZ6vbjOY6UkUpZYP0eGQFKY0xikQlKSB1U4EEVZYbP368F09RN1mNOVRXOc1rpoyRBJM1qyFFN50zOP6BpJVO9byT+DUAAP9HY4KUCdI4AV0QqQqdSnyr4lzXrl09cwQg7di2bZtnkwCkDARHAHAKaZC1Blvny5fP5y3RGAK6zABpk9qnI0t4S+TcRwBOPYIjAAAAAKBaHQAAAADEoJQ3AAAAABAcAQAAAEAMMkcAAAAAQHAEAAAAADHIHAEAAAAAwREAAAAAxCBzBAAAAAAERwAAAAAQg8wRAAAAABAcAQAAAIC5/wc6D3GO58aFHgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 850x420 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "selected_100_runs = [MODEL_RUNS[model][1.0] for model in MODEL_ORDER]\n",
    "fig, axes = plt.subplots(1, len(selected_100_runs), figsize=(4.2 * len(selected_100_runs), 3.8))\n",
    "for ax, model_label, run_name in zip(axes, MODEL_ORDER, selected_100_runs):\n",
    "    cm = aggregate_confusion(run_name)\n",
    "    norm = cm / cm.sum(axis=1, keepdims=True)\n",
    "    im = ax.imshow(norm, cmap=\"Blues\", vmin=0, vmax=1)\n",
    "    ax.set_title(model_label, fontsize=10)\n",
    "    ax.set_xticks([0, 1])\n",
    "    ax.set_xticklabels([\"pred real\", \"pred fake\"])\n",
    "    ax.set_yticks([0, 1])\n",
    "    ax.set_yticklabels([\"true real\", \"true fake\"])\n",
    "    for i in range(2):\n",
    "        for j in range(2):\n",
    "            cell_text = f\"{norm[i, j]:.2f}\" + \"\\n\" + f\"({cm[i, j]})\"\n",
    "            ax.text(j, i, cell_text, ha=\"center\", va=\"center\", fontsize=8)\n",
    "fig.colorbar(im, ax=axes.ravel().tolist(), shrink=0.8)\n",
    "fig.suptitle(\"100% data confusion matrices\", fontsize=13)\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_100pct_confusion.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "balanced_cols = [\n",
    "    \"model\", \"scale_pct\", \"run\", \"real_recall\", \"fake_recall\", \"balanced_accuracy\",\n",
    "    \"real_f1\", \"fake_f1\", \"macro_f1\", \"false_positive_rate\", \"false_negative_rate\",\n",
    "]\n",
    "display(\n",
    "    summary_df[summary_df[\"scale_pct\"] == 100][balanced_cols]\n",
    "    .style.format({\n",
    "        \"real_recall\": \"{:.4f}\",\n",
    "        \"fake_recall\": \"{:.4f}\",\n",
    "        \"balanced_accuracy\": \"{:.4f}\",\n",
    "        \"real_f1\": \"{:.4f}\",\n",
    "        \"fake_f1\": \"{:.4f}\",\n",
    "        \"macro_f1\": \"{:.4f}\",\n",
    "        \"false_positive_rate\": \"{:.4f}\",\n",
    "        \"false_negative_rate\": \"{:.4f}\",\n",
    "    })\n",
    ")\n",
    "\n",
    "fig, axes = plt.subplots(1, 2, figsize=(11, 4.2))\n",
    "sub100 = summary_df[summary_df[\"scale_pct\"] == 100].set_index(\"model\").loc[MODEL_ORDER].reset_index()\n",
    "axes[0].bar(sub100[\"model\"], sub100[\"balanced_accuracy\"], color=\"#4C78A8\", alpha=0.9)\n",
    "axes[0].set_ylim(0.95, 0.97)\n",
    "axes[0].set_title(\"100% balanced accuracy\")\n",
    "axes[0].tick_params(axis=\"x\", rotation=15)\n",
    "for idx, row in sub100.iterrows():\n",
    "    axes[0].text(idx, row[\"balanced_accuracy\"] + 0.0004, f\"{row['balanced_accuracy']:.4f}\", ha=\"center\", fontsize=8)\n",
    "axes[1].bar(sub100[\"model\"], sub100[\"macro_f1\"], color=\"#54A24B\", alpha=0.9)\n",
    "axes[1].set_ylim(0.95, 0.965)\n",
    "axes[1].set_title(\"100% macro F1\")\n",
    "axes[1].tick_params(axis=\"x\", rotation=15)\n",
    "for idx, row in sub100.iterrows():\n",
    "    axes[1].text(idx, row[\"macro_f1\"] + 0.0003, f\"{row['macro_f1']:.4f}\", ha=\"center\", fontsize=8)\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_100pct_balanced_metrics.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8.5, 4.2))\n",
    "x = np.arange(len(MODEL_ORDER))\n",
    "width = 0.35\n",
    "ax.bar(x - width / 2, sub100[\"false_positive_rate\"], width, label=\"false positive rate\", color=\"#E45756\", alpha=0.9)\n",
    "ax.bar(x + width / 2, sub100[\"false_negative_rate\"], width, label=\"false negative rate\", color=\"#4C78A8\", alpha=0.9)\n",
    "ax.set_xticks(x)\n",
    "ax.set_xticklabels(MODEL_ORDER, rotation=15, ha=\"right\")\n",
    "ax.set_ylabel(\"error rate\")\n",
    "ax.set_title(\"100% operating errors by model\")\n",
    "ax.legend()\n",
    "fig.tight_layout()\n",
    "fig.savefig(FIGURES_DIR / \"07_phase4_100pct_error_rates.png\", dpi=180, bbox_inches=\"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a55e559",
   "metadata": {},
   "source": [
    "The 100% confusion matrices show that all three models are very close overall once they are trained with the full dataset. The deciding factor is the pattern of false negatives (missed fakes): if the goal is to avoid missing fake images, ResNet50 is the best practical choice because it has the highest fake recall and the strongest macro F1. ConvNeXt-Tiny still has the lowest false positive rate and the best balanced accuracy, so it remains the safer choice when avoiding false alarms on real images is the priority. EfficientNet-B0 sits between them: strong and efficient, but not the top choice on either error profile.\n",
    "\n",
    "For this project, ResNet50 is the best practical classifier when the main operational priority is to minimize missed fakes (false negatives). It matches the top group in overall performance while giving the strongest fake detection metrics. ConvNeXt-Tiny remains the preferred alternative when false positives on real images are the biggest concern."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dda319c5",
   "metadata": {},
   "source": [
    "## 6. Report decision\n",
    "\n",
    "| Choice | Decision | Evidence |\n",
    "|---|---|---|\n",
    "| Data scale | Use 100% for the final Phase 4 setting | Every backbone improves from 20% to 50% to 100% across AUC, accuracy, and F1. |\n",
    "| Best AUC model | ConvNeXt-Tiny at 100% | Highest mean AUC: `0.9954`, with very tight fold stability. |\n",
    "| Best practical detector (minimize missed fakes) | ResNet50 at 100% | Highest fake recall and strongest macro F1 at 100%, making it the preferred choice when false negatives are the main concern. |\n",
    "| Efficient model alternative | EfficientNet-B0 at 100% | Very close AUC `0.9949` with much smaller checkpoint size, but slightly weaker operating metrics. |\n",
    "| Remaining limitation | Source behavior still matters | `insight` remains the weakest fake source, even though scaling raises it to around `0.990` pairwise AUC. |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ef8a6db8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved classifier\\outputs\\analysis\\phase4_story_summary.json\n"
     ]
    }
   ],
   "source": [
    "phase4_story_summary = {\n",
    "    \"phase\": \"phase4\",\n",
    "    \"best_auc_run\": \"p4_convnext_tiny_100pct\",\n",
    "    \"best_auc\": 0.9954,\n",
    "    \"selected_detector_run\": \"p4_convnext_tiny_100pct\",\n",
    "    \"selected_detector_reason\": \"Lowest false positive rate on real images and best balanced accuracy at 100% data.\",\n",
    "    \"notes\": [\n",
    "        \"20% rows come from the matching Phase 3 runs and act as scaling anchors.\",\n",
    "        \"50% and 100% rows come from Phase 4 logs.\",\n",
    "        \"ConvNeXt-Tiny is the AUC winner and also the practical detector when false positives on real images matter most.\",\n",
    "    ],\n",
    "    \"summary_rows\": summary_df[[\"model\", \"scale_pct\", \"run\", \"auc\", \"accuracy\", \"f1\", \"balanced_accuracy\", \"macro_f1\"]].to_dict(orient=\"records\"),\n",
    "}\n",
    "summary_path = ANALYSIS_DIR / \"phase4_story_summary.json\"\n",
    "with summary_path.open(\"w\") as f:\n",
    "    json.dump(phase4_story_summary, f, indent=2)\n",
    "print(f\"Saved {summary_path.relative_to(PROJECT_ROOT)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e89859d3",
   "metadata": {},
   "source": [
    "## Conclusion\n",
    "\n",
    "Phase 4 shows that data scale is a real improvement lever. At 20%, the top backbones were already strong, and with 50% and 100% data they converge to very similar overall performance. ConvNeXt-Tiny remains the best AUC model at `0.9954`, while ResNet50 remains strongest on fake-catching thresholded metrics (accuracy/F1 and fake recall).\n",
    "\n",
    "For the final practical decision, the confusion matrices are decisive. Because this project prioritizes avoiding missed fakes (false negatives), `p4_resnet50_100pct` is the recommended classifier. It achieves top-tier fake detection metrics while staying competitive on overall performance. ConvNeXt-Tiny remains the preferred alternative when minimizing false positives on real images is the operational priority, and source-wise diagnostics remain important because `insight` is still the weakest fake source."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.12.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}