diff --git "a/notebook/ai_vs_human/mainv3.ipynb" "b/notebook/ai_vs_human/mainv3.ipynb"
deleted file mode 100644--- "a/notebook/ai_vs_human/mainv3.ipynb"
+++ /dev/null
@@ -1,1276 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "markdown",
-   "id": "557e00db",
-   "metadata": {},
-   "source": [
-    "# AI vs Human Detector V3 - Feature Engineering Approach\n",
-    "\n",
-    "This notebook uses a feature-based ML architecture instead of transformers:\n",
-    "\n",
-    "```\n",
-    "Input Text → Preprocessing → Feature Extraction (Perplexity + Burstiness + Stylometry) → ML Classifier → Prediction\n",
-    "```\n",
-    "\n",
-    "**Features:**\n",
-    "- ✅ **Perplexity**: Language model-based text naturalness\n",
-    "- ✅ **Burstiness**: Sentence length variation patterns\n",
-    "- ✅ **Stylometry**: Writing style metrics (readability, lexical diversity, etc.)\n",
-    "- ✅ Works with all sentence lengths\n",
-    "- ✅ Fast inference (no GPU needed)\n",
-    "- ✅ Saves to `v3_model/`"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "id": "09fc6563",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Note: you may need to restart the kernel to use updated packages.\n"
-     ]
-    }
-   ],
-   "source": [
-    "%pip install -q -U scikit-learn pandas numpy nltk textstat transformers torch"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "id": "7f0d02a3",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "from __future__ import annotations\n",
-    "\n",
-    "import json\n",
-    "import pickle\n",
-    "import re\n",
-    "import warnings\n",
-    "from dataclasses import dataclass\n",
-    "from pathlib import Path\n",
-    "from typing import Any\n",
-    "\n",
-    "import numpy as np\n",
-    "import pandas as pd\n",
-    "import matplotlib.pyplot as plt\n",
-    "import seaborn as sns\n",
-    "from scipy.sparse import csr_matrix, hstack\n",
-    "from sklearn.linear_model import LogisticRegression\n",
-    "from sklearn.metrics import (\n",
-    "    accuracy_score,\n",
-    "    classification_report,\n",
-    "    confusion_matrix,\n",
-    "    f1_score,\n",
-    "    roc_auc_score,\n",
-    "    roc_curve,\n",
-    ")\n",
-    "from sklearn.model_selection import GridSearchCV, StratifiedKFold, train_test_split\n",
-    "from sklearn.preprocessing import StandardScaler\n",
-    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
-    "\n",
-    "warnings.filterwarnings('ignore')\n",
-    "sns.set_style('whitegrid')\n",
-    "plt.rcParams['figure.figsize'] = (10, 6)\n",
-    "\n",
-    "import nltk\n",
-    "try:\n",
-    "    nltk.data.find('tokenizers/punkt_tab')\n",
-    "except LookupError:\n",
-    "    nltk.download('punkt_tab', quiet=True)\n",
-    "\n",
-    "try:\n",
-    "    import textstat\n",
-    "except ImportError:\n",
-    "    import subprocess\n",
-    "    import sys\n",
-    "    subprocess.check_call([sys.executable, '-m', 'pip', 'install', '-q', 'textstat'])\n",
-    "    import textstat"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "id": "df53fee1",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "V3 Configuration:\n",
-      "Max samples: 20000\n",
-      "Output: ./v3_model\n",
-      "CV folds: 3\n"
-     ]
-    }
-   ],
-   "source": [
-    "@dataclass\n",
-    "class V3Config:\n",
-    "    max_samples: int = 20000\n",
-    "    test_size: float = 0.2\n",
-    "    output_dir: str = \"./v3_model\"\n",
-    "    random_state: int = 42\n",
-    "    cv_folds: int = 3\n",
-    "\n",
-    "\n",
-    "cfg = V3Config()\n",
-    "print(\"V3 Configuration:\")\n",
-    "print(f\"Max samples: {cfg.max_samples}\")\n",
-    "print(f\"Output: {cfg.output_dir}\")\n",
-    "print(f\"CV folds: {cfg.cv_folds}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "c04f1f0e",
-   "metadata": {},
-   "source": [
-    "## Data Loading (Recursive)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "id": "653ac0af",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Found 2 data files\n",
-      "Loaded 19940 samples\n",
-      "Human: 9970\n",
-      "AI: 9970\n"
-     ]
-    }
-   ],
-   "source": [
-    "def normalize_text(text: str) -> str:\n",
-    "    \"\"\"Clean and normalize text.\"\"\"\n",
-    "    return \" \".join(str(text).split()).strip()\n",
-    "\n",
-    "\n",
-    "def load_dataset_recursive(max_samples: int = 20000) -> pd.DataFrame:\n",
-    "    \"\"\"Load dataset from DATASET folder recursively.\"\"\"\n",
-    "    all_texts: list[str] = []\n",
-    "    all_labels: list[int] = []\n",
-    "\n",
-    "    dataset_root = Path(\"./DATASET\")\n",
-    "    if not dataset_root.exists():\n",
-    "        raise FileNotFoundError(f\"Dataset directory not found: {dataset_root}\")\n",
-    "\n",
-    "    candidates = (\n",
-    "        list(dataset_root.rglob(\"*.jsonl\"))\n",
-    "        + list(dataset_root.rglob(\"*.json\"))\n",
-    "        + list(dataset_root.rglob(\"*.csv\"))\n",
-    "    )\n",
-    "\n",
-    "    print(f\"Found {len(candidates)} data files\")\n",
-    "\n",
-    "    for file_path in candidates:\n",
-    "        try:\n",
-    "            suffix = file_path.suffix.lower()\n",
-    "            if suffix == \".jsonl\":\n",
-    "                df = pd.read_json(file_path, lines=True)\n",
-    "            elif suffix == \".json\":\n",
-    "                df = pd.read_json(file_path)\n",
-    "            elif suffix == \".csv\":\n",
-    "                df = pd.read_csv(file_path)\n",
-    "            else:\n",
-    "                continue\n",
-    "\n",
-    "            # Schema 1: human_text + ai_text\n",
-    "            if {\"human_text\", \"ai_text\"}.issubset(df.columns):\n",
-    "                human_texts = [normalize_text(x) for x in df[\"human_text\"].dropna().tolist()]\n",
-    "                ai_texts = [normalize_text(x) for x in df[\"ai_text\"].dropna().tolist()]\n",
-    "                human_texts = [x for x in human_texts if x]\n",
-    "                ai_texts = [x for x in ai_texts if x]\n",
-    "                all_texts.extend(human_texts)\n",
-    "                all_labels.extend([0] * len(human_texts))\n",
-    "                all_texts.extend(ai_texts)\n",
-    "                all_labels.extend([1] * len(ai_texts))\n",
-    "\n",
-    "            # Schema 2: text + label\n",
-    "            elif \"text\" in df.columns and (\"label\" in df.columns or \"ai_gen\" in df.columns):\n",
-    "                label_col = \"label\" if \"label\" in df.columns else \"ai_gen\"\n",
-    "                for _, row in df.iterrows():\n",
-    "                    text = normalize_text(row.get(\"text\", \"\"))\n",
-    "                    if not text:\n",
-    "                        continue\n",
-    "                    val = str(row.get(label_col, \"\")).strip().lower()\n",
-    "                    is_ai = val in {\"1\", \"true\", \"ai\", \"ai-generated\", \"ai_generated\"}\n",
-    "                    all_texts.append(text)\n",
-    "                    all_labels.append(1 if is_ai else 0)\n",
-    "\n",
-    "        except Exception:\n",
-    "            pass\n",
-    "\n",
-    "    if not all_texts:\n",
-    "        raise ValueError(\"No data loaded. Check DATASET folder.\")\n",
-    "\n",
-    "    df = pd.DataFrame({\"text\": all_texts, \"label\": all_labels})\n",
-    "    df = df[df[\"text\"].str.len() > 0].reset_index(drop=True)\n",
-    "\n",
-    "    if len(df) > max_samples:\n",
-    "        df = df.sample(n=max_samples, random_state=cfg.random_state).reset_index(drop=True)\n",
-    "\n",
-    "    print(f\"Loaded {len(df)} samples\")\n",
-    "    print(f\"Human: {(df['label'] == 0).sum()}\")\n",
-    "    print(f\"AI: {(df['label'] == 1).sum()}\")\n",
-    "\n",
-    "    return df\n",
-    "\n",
-    "\n",
-    "df = load_dataset_recursive(max_samples=cfg.max_samples)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "id": "6ed0d052",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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",
-      "text/plain": [
-       "<Figure size 800x500 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "# Visualize dataset distribution\n",
-    "plt.figure(figsize=(8, 5))\n",
-    "label_counts = df['label'].value_counts()\n",
-    "plt.bar(['Human', 'AI'], [label_counts[0], label_counts[1]], color=['blue', 'red'], alpha=0.7)\n",
-    "plt.ylabel('Count')\n",
-    "plt.title('Dataset Label Distribution', fontsize=14, fontweight='bold')\n",
-    "plt.grid(axis='y', alpha=0.3)\n",
-    "for i, v in enumerate([label_counts[0], label_counts[1]]):\n",
-    "    plt.text(i, v + 100, str(v), ha='center', fontweight='bold')\n",
-    "plt.tight_layout()\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "9f1d9b58",
-   "metadata": {},
-   "source": [
-    "## Feature Extraction: Perplexity"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "id": "e25cc72e",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Loading perplexity model: distilgpt2\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Loading weights: 100%|██████████| 76/76 [00:00<00:00, 926.77it/s]\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Loaded on: cuda\n"
-     ]
-    }
-   ],
-   "source": [
-    "import torch\n",
-    "from transformers import AutoModelForCausalLM, AutoTokenizer\n",
-    "from torch.optim import AdamW\n",
-    "from torch.utils.data import DataLoader, Dataset\n",
-    "import numpy as np\n",
-    "import random\n",
-    "import os\n",
-    "\n",
-    "class PerplexityCalculator:\n",
-    "    \"\"\"Calculate perplexity using a small language model.\"\"\"\n",
-    "\n",
-    "    def __init__(self, model_name: str = \"distilgpt2\", learning_rate: float = 5e-5):\n",
-    "        print(f\"Loading perplexity model: {model_name}\")\n",
-    "        self.device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
-    "        self.tokenizer = AutoTokenizer.from_pretrained(model_name)\n",
-    "        self.model = AutoModelForCausalLM.from_pretrained(model_name).to(self.device)\n",
-    "        self.model.eval()\n",
-    "        self.optimizer = AdamW(self.model.parameters(), lr=learning_rate)\n",
-    "        print(f\"Loaded on: {self.device}\")\n",
-    "\n",
-    "    def calculate(self, text: str, max_length: int = 512) -> float:\n",
-    "        \"\"\"Calculate perplexity for given text.\"\"\"\n",
-    "        try:\n",
-    "            encodings = self.tokenizer(text, return_tensors=\"pt\", truncation=True, max_length=max_length)\n",
-    "            input_ids = encodings.input_ids.to(self.device)\n",
-    "\n",
-    "            with torch.no_grad():\n",
-    "                outputs = self.model(input_ids, labels=input_ids)\n",
-    "                loss = outputs.loss\n",
-    "                perplexity = torch.exp(loss).item()\n",
-    "\n",
-    "            return min(perplexity, 10000.0)  # Cap at 10k to avoid outliers\n",
-    "        except Exception:\n",
-    "            return 100.0  # Default fallback\n",
-    "\n",
-    "    def fine_tune(self, dataset: Dataset, epochs: int = 3, batch_size: int = 8):\n",
-    "        \"\"\"Fine-tune the model on a given dataset.\"\"\"\n",
-    "        self.model.train()\n",
-    "        dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)\n",
-    "        for epoch in range(epochs):\n",
-    "            epoch_loss = 0.0\n",
-    "            for batch in dataloader:\n",
-    "                inputs = self.tokenizer(batch[\"text\"], return_tensors=\"pt\", truncation=True, padding=True, max_length=512)\n",
-    "                input_ids = inputs.input_ids.to(self.device)\n",
-    "                labels = input_ids.clone()\n",
-    "                self.optimizer.zero_grad()\n",
-    "                outputs = self.model(input_ids, labels=labels)\n",
-    "                loss = outputs.loss\n",
-    "                loss.backward()\n",
-    "                self.optimizer.step()\n",
-    "                epoch_loss += loss.item()\n",
-    "            print(f\"Epoch {epoch + 1}/{epochs}, Loss: {epoch_loss / len(dataloader)}\")\n",
-    "        self.model.eval()\n",
-    "\n",
-    "# Set random seeds for reproducibility\n",
-    "def set_seed(seed: int = 42):\n",
-    "    random.seed(seed)\n",
-    "    np.random.seed(seed)\n",
-    "    torch.manual_seed(seed)\n",
-    "    if torch.cuda.is_available():\n",
-    "        torch.cuda.manual_seed_all(seed)\n",
-    "\n",
-    "set_seed()\n",
-    "perplexity_calc = PerplexityCalculator()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "8311ff0c",
-   "metadata": {},
-   "source": [
-    "## Feature Extraction: Burstiness"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "id": "23cb7fb6",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def extract_burstiness_features(text: str) -> dict[str, float]:\n",
-    "    \"\"\"Extract burstiness features (sentence length variation).\"\"\"\n",
-    "    sentences = nltk.sent_tokenize(text)\n",
-    "    if len(sentences) == 0:\n",
-    "        return {\"burst_mean\": 0, \"burst_std\": 0, \"burst_max\": 0, \"burst_min\": 0, \"burst_range\": 0}\n",
-    "\n",
-    "    lengths = [len(s.split()) for s in sentences]\n",
-    "\n",
-    "    return {\n",
-    "        \"burst_mean\": np.mean(lengths),\n",
-    "        \"burst_std\": np.std(lengths),\n",
-    "        \"burst_max\": np.max(lengths),\n",
-    "        \"burst_min\": np.min(lengths),\n",
-    "        \"burst_range\": np.max(lengths) - np.min(lengths),\n",
-    "    }"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "ef03f84f",
-   "metadata": {},
-   "source": [
-    "## Feature Extraction: Stylometry"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "id": "b90a7af4",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def extract_stylometry_features(text: str) -> dict[str, float]:\n",
-    "    \"\"\"Extract stylometry features (writing style metrics).\"\"\"\n",
-    "    words = text.split()\n",
-    "    num_words = len(words)\n",
-    "    num_chars = len(text)\n",
-    "    num_sentences = max(len(nltk.sent_tokenize(text)), 1)\n",
-    "\n",
-    "    # Basic metrics\n",
-    "    avg_word_len = np.mean([len(w) for w in words]) if words else 0\n",
-    "    avg_sent_len = num_words / num_sentences\n",
-    "\n",
-    "    # Lexical diversity\n",
-    "    unique_words = len(set(words))\n",
-    "    lexical_diversity = unique_words / num_words if num_words > 0 else 0\n",
-    "\n",
-    "    # Punctuation\n",
-    "    num_punct = sum(1 for c in text if c in \".,!?;:\")\n",
-    "    punct_ratio = num_punct / num_chars if num_chars > 0 else 0\n",
-    "\n",
-    "    # Capitalization\n",
-    "    num_caps = sum(1 for c in text if c.isupper())\n",
-    "    caps_ratio = num_caps / num_chars if num_chars > 0 else 0\n",
-    "\n",
-    "    # Readability scores\n",
-    "    try:\n",
-    "        flesch_reading = textstat.flesch_reading_ease(text)\n",
-    "        flesch_grade = textstat.flesch_kincaid_grade(text)\n",
-    "    except Exception:\n",
-    "        flesch_reading = 50.0\n",
-    "        flesch_grade = 8.0\n",
-    "\n",
-    "    return {\n",
-    "        \"num_words\": num_words,\n",
-    "        \"num_chars\": num_chars,\n",
-    "        \"num_sentences\": num_sentences,\n",
-    "        \"avg_word_len\": avg_word_len,\n",
-    "        \"avg_sent_len\": avg_sent_len,\n",
-    "        \"lexical_diversity\": lexical_diversity,\n",
-    "        \"punct_ratio\": punct_ratio,\n",
-    "        \"caps_ratio\": caps_ratio,\n",
-    "        \"flesch_reading\": flesch_reading,\n",
-    "        \"flesch_grade\": flesch_grade,\n",
-    "    }"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "dbe0b8f9",
-   "metadata": {},
-   "source": [
-    "## Combined Feature Extraction"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 8,
-   "id": "6cad5be8",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "`loss_type=None` was set in the config but it is unrecognized. Using the default loss: `ForCausalLMLoss`.\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Sample features extracted: 16 features\n",
-      "Feature names: ['perplexity', 'burst_mean', 'burst_std', 'burst_max', 'burst_min', 'burst_range', 'num_words', 'num_chars', 'num_sentences', 'avg_word_len', 'avg_sent_len', 'lexical_diversity', 'punct_ratio', 'caps_ratio', 'flesch_reading', 'flesch_grade']\n"
-     ]
-    }
-   ],
-   "source": [
-    "def extract_all_features(text: str, calc_perplexity: bool = True) -> dict[str, float]:\n",
-    "    \"\"\"Extract all features (perplexity + burstiness + stylometry).\"\"\"\n",
-    "    features = {}\n",
-    "\n",
-    "    # Perplexity\n",
-    "    if calc_perplexity:\n",
-    "        features[\"perplexity\"] = perplexity_calc.calculate(text)\n",
-    "\n",
-    "    # Burstiness\n",
-    "    burst_features = extract_burstiness_features(text)\n",
-    "    features.update(burst_features)\n",
-    "\n",
-    "    # Stylometry\n",
-    "    style_features = extract_stylometry_features(text)\n",
-    "    features.update(style_features)\n",
-    "\n",
-    "    return features\n",
-    "\n",
-    "\n",
-    "# Test feature extraction on a sample\n",
-    "sample_text = df.iloc[0][\"text\"]\n",
-    "sample_features = extract_all_features(sample_text)\n",
-    "print(f\"Sample features extracted: {len(sample_features)} features\")\n",
-    "print(f\"Feature names: {list(sample_features.keys())}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "21e48aac",
-   "metadata": {},
-   "source": [
-    "## Extract Features for All Data"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 9,
-   "id": "ca490c06",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Extracting features for all samples...\n",
-      "This may take a few minutes for perplexity calculation.\n",
-      "Progress: 0/19940\n"
-     ]
-    },
-    {
-     "ename": "KeyboardInterrupt",
-     "evalue": "",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[31m---------------------------------------------------------------------------\u001b[39m",
-      "\u001b[31mKeyboardInterrupt\u001b[39m                         Traceback (most recent call last)",
-      "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[9]\u001b[39m\u001b[32m, line 8\u001b[39m\n\u001b[32m      6\u001b[39m     \u001b[38;5;28;01mif\u001b[39;00m idx % \u001b[32m500\u001b[39m == \u001b[32m0\u001b[39m:\n\u001b[32m      7\u001b[39m         \u001b[38;5;28mprint\u001b[39m(\u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[33mProgress: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00midx\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m/\u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mlen\u001b[39m(df)\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m\"\u001b[39m)\n\u001b[32m----> \u001b[39m\u001b[32m8\u001b[39m     features = \u001b[43mextract_all_features\u001b[49m\u001b[43m(\u001b[49m\u001b[43mrow\u001b[49m\u001b[43m[\u001b[49m\u001b[33;43m\"\u001b[39;49m\u001b[33;43mtext\u001b[39;49m\u001b[33;43m\"\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcalc_perplexity\u001b[49m\u001b[43m=\u001b[49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m)\u001b[49m\n\u001b[32m      9\u001b[39m     feature_list.append(features)\n\u001b[32m     11\u001b[39m \u001b[38;5;28mprint\u001b[39m(\u001b[33m\"\u001b[39m\u001b[33mFeature extraction complete\u001b[39m\u001b[33m\"\u001b[39m)\n",
-      "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[8]\u001b[39m\u001b[32m, line 7\u001b[39m, in \u001b[36mextract_all_features\u001b[39m\u001b[34m(text, calc_perplexity)\u001b[39m\n\u001b[32m      5\u001b[39m \u001b[38;5;66;03m# Perplexity\u001b[39;00m\n\u001b[32m      6\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m calc_perplexity:\n\u001b[32m----> \u001b[39m\u001b[32m7\u001b[39m     features[\u001b[33m\"\u001b[39m\u001b[33mperplexity\u001b[39m\u001b[33m\"\u001b[39m] = \u001b[43mperplexity_calc\u001b[49m\u001b[43m.\u001b[49m\u001b[43mcalculate\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtext\u001b[49m\u001b[43m)\u001b[49m\n\u001b[32m      9\u001b[39m \u001b[38;5;66;03m# Burstiness\u001b[39;00m\n\u001b[32m     10\u001b[39m burst_features = extract_burstiness_features(text)\n",
-      "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[5]\u001b[39m\u001b[32m, line 30\u001b[39m, in \u001b[36mPerplexityCalculator.calculate\u001b[39m\u001b[34m(self, text, max_length)\u001b[39m\n\u001b[32m     28\u001b[39m         outputs = \u001b[38;5;28mself\u001b[39m.model(input_ids, labels=input_ids)\n\u001b[32m     29\u001b[39m         loss = outputs.loss\n\u001b[32m---> \u001b[39m\u001b[32m30\u001b[39m         perplexity = \u001b[43mtorch\u001b[49m\u001b[43m.\u001b[49m\u001b[43mexp\u001b[49m\u001b[43m(\u001b[49m\u001b[43mloss\u001b[49m\u001b[43m)\u001b[49m\u001b[43m.\u001b[49m\u001b[43mitem\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[32m     32\u001b[39m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mmin\u001b[39m(perplexity, \u001b[32m10000.0\u001b[39m)  \u001b[38;5;66;03m# Cap at 10k to avoid outliers\u001b[39;00m\n\u001b[32m     33\u001b[39m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m:\n",
-      "\u001b[31mKeyboardInterrupt\u001b[39m: "
-     ]
-    }
-   ],
-   "source": [
-    "print(\"Extracting features for all samples...\")\n",
-    "print(\"This may take a few minutes for perplexity calculation.\")\n",
-    "\n",
-    "feature_list = []\n",
-    "for idx, row in df.iterrows():\n",
-    "    if idx % 500 == 0:\n",
-    "        print(f\"Progress: {idx}/{len(df)}\")\n",
-    "    features = extract_all_features(row[\"text\"], calc_perplexity=True)\n",
-    "    feature_list.append(features)\n",
-    "\n",
-    "print(\"Feature extraction complete\")\n",
-    "\n",
-    "# Convert to DataFrame\n",
-    "X_df = pd.DataFrame(feature_list)\n",
-    "y = df[\"label\"].values\n",
-    "\n",
-    "print(f\"Feature matrix shape: {X_df.shape}\")\n",
-    "print(f\"Labels shape: {y.shape}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "37c71deb",
-   "metadata": {},
-   "source": [
-    "## Train-Test Split & Scaling"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 17,
-   "id": "5aa74c74",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Train set: 15952 samples\n",
-      "Test set: 3988 samples\n",
-      "Engineered features scaled using StandardScaler\n"
-     ]
-    }
-   ],
-   "source": [
-    "indices = np.arange(len(df))\n",
-    "train_idx, test_idx = train_test_split(\n",
-    "    indices,\n",
-    "    test_size=cfg.test_size,\n",
-    "    random_state=cfg.random_state,\n",
-    "    stratify=y,\n",
-    ")\n",
-    "\n",
-    "X_train = X_df.iloc[train_idx].reset_index(drop=True)\n",
-    "X_test = X_df.iloc[test_idx].reset_index(drop=True)\n",
-    "y_train = y[train_idx]\n",
-    "y_test = y[test_idx]\n",
-    "\n",
-    "text_train = df.iloc[train_idx][\"text\"].tolist()\n",
-    "text_test = df.iloc[test_idx][\"text\"].tolist()\n",
-    "\n",
-    "print(f\"Train set: {len(X_train)} samples\")\n",
-    "print(f\"Test set: {len(X_test)} samples\")\n",
-    "\n",
-    "# Standardize engineered numeric features\n",
-    "scaler = StandardScaler()\n",
-    "X_train_scaled = scaler.fit_transform(X_train)\n",
-    "X_test_scaled = scaler.transform(X_test)\n",
-    "\n",
-    "print(\"Engineered features scaled using StandardScaler\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "377f9f9e",
-   "metadata": {},
-   "source": [
-    "## Train Classifier"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "bf27b65b",
-   "metadata": {},
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 18,
-   "id": "0d79acfa",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Training hybrid model (word TF-IDF + char TF-IDF + engineered features)...\n",
-      "Hybrid train matrix shape: (15952, 166972)\n",
-      "Hybrid test matrix shape: (3988, 166972)\n",
-      "Best CV params: {'C': 4.0, 'class_weight': 'balanced'}\n",
-      "Best CV F1: 0.8594\n"
-     ]
-    }
-   ],
-   "source": [
-    "print(\"Training hybrid model (word TF-IDF + char TF-IDF + engineered features)...\")\n",
-    "\n",
-    "word_vectorizer = TfidfVectorizer(\n",
-    "    analyzer=\"word\",\n",
-    "    ngram_range=(1, 2),\n",
-    "    min_df=3,\n",
-    "    max_df=0.98,\n",
-    "    max_features=120000,\n",
-    "    sublinear_tf=True,\n",
-    ")\n",
-    "\n",
-    "char_vectorizer = TfidfVectorizer(\n",
-    "    analyzer=\"char_wb\",\n",
-    "    ngram_range=(3, 5),\n",
-    "    min_df=3,\n",
-    "    max_df=0.98,\n",
-    "    max_features=80000,\n",
-    "    sublinear_tf=True,\n",
-    ")\n",
-    "\n",
-    "X_train_word = word_vectorizer.fit_transform(text_train)\n",
-    "X_test_word = word_vectorizer.transform(text_test)\n",
-    "\n",
-    "X_train_char = char_vectorizer.fit_transform(text_train)\n",
-    "X_test_char = char_vectorizer.transform(text_test)\n",
-    "\n",
-    "X_train_num_sparse = csr_matrix(X_train_scaled)\n",
-    "X_test_num_sparse = csr_matrix(X_test_scaled)\n",
-    "\n",
-    "X_train_hybrid = hstack([X_train_word, X_train_char, X_train_num_sparse], format=\"csr\")\n",
-    "X_test_hybrid = hstack([X_test_word, X_test_char, X_test_num_sparse], format=\"csr\")\n",
-    "\n",
-    "print(f\"Hybrid train matrix shape: {X_train_hybrid.shape}\")\n",
-    "print(f\"Hybrid test matrix shape: {X_test_hybrid.shape}\")\n",
-    "\n",
-    "cv = StratifiedKFold(n_splits=cfg.cv_folds, shuffle=True, random_state=cfg.random_state)\n",
-    "\n",
-    "grid = GridSearchCV(\n",
-    "    estimator=LogisticRegression(\n",
-    "        solver=\"saga\",\n",
-    "        penalty=\"l2\",\n",
-    "        max_iter=2500,\n",
-    "        random_state=cfg.random_state,\n",
-    "        n_jobs=-1,\n",
-    "    ),\n",
-    "    param_grid={\n",
-    "        \"C\": [0.5, 1.0, 2.0, 4.0],\n",
-    "        \"class_weight\": [None, \"balanced\"],\n",
-    "    },\n",
-    "    scoring=\"f1\",\n",
-    "    cv=cv,\n",
-    "    n_jobs=-1,\n",
-    "    verbose=0,\n",
-    ")\n",
-    "\n",
-    "grid.fit(X_train_hybrid, y_train)\n",
-    "classifier = grid.best_estimator_\n",
-    "best_model_name = \"hybrid_tfidf_logistic\"\n",
-    "cv_best_f1 = float(grid.best_score_)\n",
-    "\n",
-    "print(f\"Best CV params: {grid.best_params_}\")\n",
-    "print(f\"Best CV F1: {cv_best_f1:.4f}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "e3d8ef30",
-   "metadata": {},
-   "source": [
-    "## Evaluate Model"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 19,
-   "id": "4048d5cc",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "MODEL EVALUATION RESULTS\n",
-      "Selected Model: hybrid_tfidf_logistic\n",
-      "Best CV F1: 0.8594\n",
-      "Train Accuracy: 0.9803\n",
-      "Train F1 Score: 0.9802\n",
-      "Test Accuracy:  0.8714\n",
-      "Test F1 Score:  0.8707\n",
-      "\n",
-      "              precision    recall  f1-score   support\n",
-      "\n",
-      "       Human       0.87      0.88      0.87      1994\n",
-      "          AI       0.87      0.87      0.87      1994\n",
-      "\n",
-      "    accuracy                           0.87      3988\n",
-      "   macro avg       0.87      0.87      0.87      3988\n",
-      "weighted avg       0.87      0.87      0.87      3988\n",
-      "\n"
-     ]
-    }
-   ],
-   "source": [
-    "# Train predictions\n",
-    "y_train_pred = classifier.predict(X_train_hybrid)\n",
-    "train_acc = accuracy_score(y_train, y_train_pred)\n",
-    "train_f1 = f1_score(y_train, y_train_pred)\n",
-    "\n",
-    "# Test predictions\n",
-    "y_test_pred = classifier.predict(X_test_hybrid)\n",
-    "test_acc = accuracy_score(y_test, y_test_pred)\n",
-    "test_f1 = f1_score(y_test, y_test_pred)\n",
-    "\n",
-    "if hasattr(classifier, \"predict_proba\"):\n",
-    "    y_test_proba = classifier.predict_proba(X_test_hybrid)[:, 1]\n",
-    "else:\n",
-    "    decision_scores = classifier.decision_function(X_test_hybrid)\n",
-    "    y_test_proba = (decision_scores - decision_scores.min()) / (decision_scores.max() - decision_scores.min() + 1e-8)\n",
-    "\n",
-    "print(\"MODEL EVALUATION RESULTS\")\n",
-    "print(f\"Selected Model: {best_model_name}\")\n",
-    "print(f\"Best CV F1: {cv_best_f1:.4f}\")\n",
-    "print(f\"Train Accuracy: {train_acc:.4f}\")\n",
-    "print(f\"Train F1 Score: {train_f1:.4f}\")\n",
-    "print(f\"Test Accuracy:  {test_acc:.4f}\")\n",
-    "print(f\"Test F1 Score:  {test_f1:.4f}\")\n",
-    "\n",
-    "print(f\"\\n{classification_report(y_test, y_test_pred, target_names=['Human', 'AI'])}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "c9588b0a",
-   "metadata": {},
-   "source": [
-    "## Visualizations"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 20,
-   "id": "a7c1bb26",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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",
-      "text/plain": [
-       "<Figure size 2000x1200 with 7 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "All visualizations completed\n"
-     ]
-    }
-   ],
-   "source": [
-    "# Create comprehensive visualizations\n",
-    "fig = plt.figure(figsize=(20, 12))\n",
-    "\n",
-    "# 1. Confusion Matrix\n",
-    "plt.subplot(2, 3, 1)\n",
-    "cm = confusion_matrix(y_test, y_test_pred)\n",
-    "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', xticklabels=['Human', 'AI'], yticklabels=['Human', 'AI'])\n",
-    "plt.title('Confusion Matrix', fontsize=14, fontweight='bold')\n",
-    "plt.ylabel('True Label')\n",
-    "plt.xlabel('Predicted Label')\n",
-    "\n",
-    "# 2. Engineered Feature Importance (from hybrid model coefficients)\n",
-    "plt.subplot(2, 3, 2)\n",
-    "engineered_names = list(X_df.columns)\n",
-    "engineered_count = len(engineered_names)\n",
-    "if hasattr(classifier, 'coef_') and classifier.coef_.shape[1] >= engineered_count:\n",
-    "    engineered_coef = np.abs(classifier.coef_[0][-engineered_count:])\n",
-    "    top_k = min(10, engineered_count)\n",
-    "    sorted_idx = np.argsort(engineered_coef)[-top_k:]\n",
-    "    plt.barh(range(top_k), engineered_coef[sorted_idx])\n",
-    "    plt.yticks(range(top_k), [engineered_names[i] for i in sorted_idx])\n",
-    "    plt.xlabel('Absolute Coefficient')\n",
-    "    plt.title('Top Engineered Feature Importance', fontsize=14, fontweight='bold')\n",
-    "else:\n",
-    "    plt.text(0.5, 0.5, 'Engineered feature importance unavailable', ha='center', va='center')\n",
-    "    plt.axis('off')\n",
-    "\n",
-    "# 3. ROC Curve\n",
-    "plt.subplot(2, 3, 3)\n",
-    "fpr, tpr, _ = roc_curve(y_test, y_test_proba)\n",
-    "roc_auc = roc_auc_score(y_test, y_test_proba)\n",
-    "plt.plot(fpr, tpr, color='darkorange', lw=2, label=f'ROC curve (AUC = {roc_auc:.3f})')\n",
-    "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--', label='Random')\n",
-    "plt.xlim([0.0, 1.0])\n",
-    "plt.ylim([0.0, 1.05])\n",
-    "plt.xlabel('False Positive Rate')\n",
-    "plt.ylabel('True Positive Rate')\n",
-    "plt.title('ROC Curve', fontsize=14, fontweight='bold')\n",
-    "plt.legend(loc='lower right')\n",
-    "\n",
-    "# 4. Perplexity Distribution by Class\n",
-    "plt.subplot(2, 3, 4)\n",
-    "human_perp = X_df[y == 0]['perplexity']\n",
-    "ai_perp = X_df[y == 1]['perplexity']\n",
-    "plt.hist(human_perp, bins=30, alpha=0.6, label='Human', color='blue', density=True)\n",
-    "plt.hist(ai_perp, bins=30, alpha=0.6, label='AI', color='red', density=True)\n",
-    "plt.xlabel('Perplexity')\n",
-    "plt.ylabel('Density')\n",
-    "plt.title('Perplexity Distribution', fontsize=14, fontweight='bold')\n",
-    "plt.legend()\n",
-    "\n",
-    "# 5. Lexical Diversity by Class\n",
-    "plt.subplot(2, 3, 5)\n",
-    "human_lex = X_df[y == 0]['lexical_diversity']\n",
-    "ai_lex = X_df[y == 1]['lexical_diversity']\n",
-    "plt.hist(human_lex, bins=30, alpha=0.6, label='Human', color='blue', density=True)\n",
-    "plt.hist(ai_lex, bins=30, alpha=0.6, label='AI', color='red', density=True)\n",
-    "plt.xlabel('Lexical Diversity')\n",
-    "plt.ylabel('Density')\n",
-    "plt.title('Lexical Diversity Distribution', fontsize=14, fontweight='bold')\n",
-    "plt.legend()\n",
-    "\n",
-    "# 6. Burstiness STD by Class\n",
-    "plt.subplot(2, 3, 6)\n",
-    "human_burst = X_df[y == 0]['burst_std']\n",
-    "ai_burst = X_df[y == 1]['burst_std']\n",
-    "plt.hist(human_burst, bins=30, alpha=0.6, label='Human', color='blue', density=True)\n",
-    "plt.hist(ai_burst, bins=30, alpha=0.6, label='AI', color='red', density=True)\n",
-    "plt.xlabel('Burstiness STD')\n",
-    "plt.ylabel('Density')\n",
-    "plt.title('Burstiness STD Distribution', fontsize=14, fontweight='bold')\n",
-    "plt.legend()\n",
-    "\n",
-    "plt.tight_layout()\n",
-    "plt.show()\n",
-    "\n",
-    "print('All visualizations completed')"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "cbbafff1",
-   "metadata": {},
-   "source": [
-    "## Save Model"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 21,
-   "id": "891cc54a",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model saved to: /mnt/linux-data/Work/aiapi/notebook/ai_vs_human/v3_model\n",
-      "Files: classifier.pkl, scaler.pkl, word_vectorizer.pkl, char_vectorizer.pkl, feature_names.json, metadata.json\n"
-     ]
-    }
-   ],
-   "source": [
-    "save_dir = Path(cfg.output_dir)\n",
-    "save_dir.mkdir(parents=True, exist_ok=True)\n",
-    "\n",
-    "# Save classifier and preprocessing artifacts\n",
-    "with open(save_dir / \"classifier.pkl\", \"wb\") as f:\n",
-    "    pickle.dump(classifier, f)\n",
-    "\n",
-    "with open(save_dir / \"scaler.pkl\", \"wb\") as f:\n",
-    "    pickle.dump(scaler, f)\n",
-    "\n",
-    "with open(save_dir / \"word_vectorizer.pkl\", \"wb\") as f:\n",
-    "    pickle.dump(word_vectorizer, f)\n",
-    "\n",
-    "with open(save_dir / \"char_vectorizer.pkl\", \"wb\") as f:\n",
-    "    pickle.dump(char_vectorizer, f)\n",
-    "\n",
-    "# Save engineered feature names\n",
-    "with open(save_dir / \"feature_names.json\", \"w\") as f:\n",
-    "    json.dump(list(X_df.columns), f, indent=2)\n",
-    "\n",
-    "metadata = {\n",
-    "    \"selected_model\": best_model_name,\n",
-    "    \"cv_best_f1\": cv_best_f1,\n",
-    "    \"num_engineered_features\": X_df.shape[1],\n",
-    "    \"num_word_tfidf_features\": int(X_train_word.shape[1]),\n",
-    "    \"num_char_tfidf_features\": int(X_train_char.shape[1]),\n",
-    "    \"train_samples\": len(X_train),\n",
-    "    \"test_samples\": len(X_test),\n",
-    "    \"train_accuracy\": float(train_acc),\n",
-    "    \"train_f1\": float(train_f1),\n",
-    "    \"test_accuracy\": float(test_acc),\n",
-    "    \"test_f1\": float(test_f1),\n",
-    "}\n",
-    "\n",
-    "with open(save_dir / \"metadata.json\", \"w\") as f:\n",
-    "    json.dump(metadata, f, indent=2)\n",
-    "\n",
-    "print(f\"Model saved to: {save_dir.resolve()}\")\n",
-    "print(\"Files: classifier.pkl, scaler.pkl, word_vectorizer.pkl, char_vectorizer.pkl, feature_names.json, metadata.json\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "b693d7d6",
-   "metadata": {},
-   "source": [
-    "## Prediction Function"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 22,
-   "id": "a1ad73cb",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Prediction function ready\n"
-     ]
-    }
-   ],
-   "source": [
-    "def predict_v3(text: str) -> dict[str, Any]:\n",
-    "    \"\"\"Predict if text is AI or human using hybrid TF-IDF + engineered-feature model.\"\"\"\n",
-    "    cleaned = normalize_text(text)\n",
-    "    if not cleaned:\n",
-    "        raise ValueError(\"Input text is empty.\")\n",
-    "\n",
-    "    # Engineered features\n",
-    "    features = extract_all_features(cleaned, calc_perplexity=True)\n",
-    "    feature_vector = np.array([features[name] for name in X_df.columns]).reshape(1, -1)\n",
-    "    feature_scaled = scaler.transform(feature_vector)\n",
-    "\n",
-    "    # Text features\n",
-    "    word_vec = word_vectorizer.transform([cleaned])\n",
-    "    char_vec = char_vectorizer.transform([cleaned])\n",
-    "    num_vec = csr_matrix(feature_scaled)\n",
-    "\n",
-    "    # Hybrid feature vector\n",
-    "    hybrid_vec = hstack([word_vec, char_vec, num_vec], format=\"csr\")\n",
-    "\n",
-    "    # Predict\n",
-    "    pred_label = int(classifier.predict(hybrid_vec)[0])\n",
-    "    if hasattr(classifier, \"predict_proba\"):\n",
-    "        pred_proba = classifier.predict_proba(hybrid_vec)[0]\n",
-    "    else:\n",
-    "        score = classifier.decision_function(hybrid_vec)[0]\n",
-    "        p_ai = 1 / (1 + np.exp(-score))\n",
-    "        pred_proba = np.array([1 - p_ai, p_ai])\n",
-    "\n",
-    "    return {\n",
-    "        \"text\": cleaned[:100] + \"...\" if len(cleaned) > 100 else cleaned,\n",
-    "        \"word_count\": len(cleaned.split()),\n",
-    "        \"predicted_label\": pred_label,\n",
-    "        \"predicted_name\": \"ai\" if pred_label == 1 else \"human\",\n",
-    "        \"probability_human\": float(pred_proba[0]),\n",
-    "        \"probability_ai\": float(pred_proba[1]),\n",
-    "        \"features\": features,\n",
-    "    }\n",
-    "\n",
-    "\n",
-    "print(\"Prediction function ready\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "0d55d414",
-   "metadata": {},
-   "source": [
-    "## Test Predictions on Various Sentence Types"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "961ec8dc",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "TESTING V3 MODEL ON VARIOUS SENTENCE TYPES\n",
-      "Running predictions...\n"
-     ]
-    },
-    {
-     "ename": "NameError",
-     "evalue": "name 'predict_v3' is not defined",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[31m---------------------------------------------------------------------------\u001b[39m",
-      "\u001b[31mNameError\u001b[39m                                 Traceback (most recent call last)",
-      "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[30]\u001b[39m\u001b[32m, line 18\u001b[39m\n\u001b[32m     16\u001b[39m \u001b[38;5;28mprint\u001b[39m(\u001b[33m\"\u001b[39m\u001b[33mRunning predictions...\u001b[39m\u001b[33m\"\u001b[39m)\n\u001b[32m     17\u001b[39m \u001b[38;5;28;01mfor\u001b[39;00m category, text \u001b[38;5;129;01min\u001b[39;00m test_cases:\n\u001b[32m---> \u001b[39m\u001b[32m18\u001b[39m     result = \u001b[43mpredict_v3\u001b[49m(text)\n\u001b[32m     19\u001b[39m     \u001b[38;5;28mprint\u001b[39m(\u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[33m[\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mcategory\u001b[38;5;132;01m:\u001b[39;00m\u001b[33m12\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m] \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mresult[\u001b[33m'\u001b[39m\u001b[33mpredicted_name\u001b[39m\u001b[33m'\u001b[39m]\u001b[38;5;132;01m:\u001b[39;00m\u001b[33m6\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m (AI: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mresult[\u001b[33m'\u001b[39m\u001b[33mprobability_ai\u001b[39m\u001b[33m'\u001b[39m]\u001b[38;5;132;01m:\u001b[39;00m\u001b[33m.1%\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m)\u001b[39m\u001b[33m\"\u001b[39m)\n\u001b[32m     20\u001b[39m     \u001b[38;5;28mprint\u001b[39m(\u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[33mText: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mtext[:\u001b[32m60\u001b[39m]\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m...\u001b[39m\u001b[33m\"\u001b[39m)\n",
-      "\u001b[31mNameError\u001b[39m: name 'predict_v3' is not defined"
-     ]
-    }
-   ],
-   "source": [
-    "print(\"TESTING V3 MODEL ON VARIOUS SENTENCE TYPES\")\n",
-    "\n",
-    "test_cases = [\n",
-    "    (\"Very short\", \"Hello world!\"),\n",
-    "    (\"Short\", \"I love programming and building cool AI projects.\"),\n",
-    "    (\n",
-    "        \"Medium\",\n",
-    "        \"Artificial intelligence has revolutionized many industries by enabling automation, improving decision-making, and creating new opportunities for innovation across sectors.\",\n",
-    "    ),\n",
-    "    (\n",
-    "        \"Long\",\n",
-    "        \"Machine learning represents a subset of artificial intelligence that enables computer systems to automatically learn and improve from experience without being explicitly programmed. The fundamental idea is to develop algorithms that can receive input data and use statistical analysis to predict outputs while updating as new data becomes available. This field has grown exponentially, driven by increases in computational power and breakthroughs in algorithmic approaches.\",\n",
-    "    ),\n",
-    "]\n",
-    "\n",
-    "print(\"Running predictions...\")\n",
-    "for category, text in test_cases:\n",
-    "    result = predict_v3(text)\n",
-    "    print(f\"[{category:12}] {result['predicted_name']:6} (AI: {result['probability_ai']:.1%})\")\n",
-    "    print(f\"Text: {text[:60]}...\")\n",
-    "    print(f\"Key features: perplexity={result['features'].get('perplexity', 0):.1f}, \"\n",
-    "          f\"burst_std={result['features'].get('burst_std', 0):.1f}, \"\n",
-    "          f\"lexical_div={result['features'].get('lexical_diversity', 0):.2f}\")\n",
-    "    print()\n",
-    "\n",
-    "print(\"All test cases completed\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "33925c05",
-   "metadata": {},
-   "source": [
-    "## Load and Test Saved Model"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "id": "fb5f483c",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "TESTING SAVED V3 MODEL\n",
-      "Loaded model from: v3_model\n",
-      "Selected Model: hybrid_tfidf_logistic\n",
-      "CV Best F1: 0.8594\n",
-      "Engineered features: 16\n",
-      "Word TF-IDF features: 86956\n",
-      "Char TF-IDF features: 80000\n",
-      "Test Accuracy: 0.8714\n",
-      "Test F1: 0.8707\n",
-      "Test inference: 'What life even is when there is not a single will power bone in your body? Are you living or is it you who is trying to just get by? She often questions herself. But never seems to find an answer. She's been the eldest daughter all her life and  a break is never in her life book.'\n",
-      "Prediction: Human\n",
-      "Probabilities: Human=99.94%, AI=0.06%\n",
-      "Saved model works correctly\n"
-     ]
-    }
-   ],
-   "source": [
-    "print(\"TESTING SAVED V3 MODEL\")\n",
-    "\n",
-    "model_dir = Path(cfg.output_dir)\n",
-    "if model_dir.exists():\n",
-    "    with open(model_dir / \"classifier.pkl\", \"rb\") as f:\n",
-    "        loaded_classifier = pickle.load(f)\n",
-    "\n",
-    "    with open(model_dir / \"scaler.pkl\", \"rb\") as f:\n",
-    "        loaded_scaler = pickle.load(f)\n",
-    "\n",
-    "    with open(model_dir / \"word_vectorizer.pkl\", \"rb\") as f:\n",
-    "        loaded_word_vectorizer = pickle.load(f)\n",
-    "\n",
-    "    with open(model_dir / \"char_vectorizer.pkl\", \"rb\") as f:\n",
-    "        loaded_char_vectorizer = pickle.load(f)\n",
-    "\n",
-    "    with open(model_dir / \"feature_names.json\", \"r\") as f:\n",
-    "        loaded_features = json.load(f)\n",
-    "\n",
-    "    with open(model_dir / \"metadata.json\", \"r\") as f:\n",
-    "        loaded_metadata = json.load(f)\n",
-    "\n",
-    "    print(f\"Loaded model from: {model_dir}\")\n",
-    "    print(f\"Selected Model: {loaded_metadata.get('selected_model', 'unknown')}\")\n",
-    "    print(f\"CV Best F1: {loaded_metadata.get('cv_best_f1', 0.0):.4f}\")\n",
-    "    print(f\"Engineered features: {loaded_metadata.get('num_engineered_features', len(loaded_features))}\")\n",
-    "    print(f\"Word TF-IDF features: {loaded_metadata.get('num_word_tfidf_features', 0)}\")\n",
-    "    print(f\"Char TF-IDF features: {loaded_metadata.get('num_char_tfidf_features', 0)}\")\n",
-    "    print(f\"Test Accuracy: {loaded_metadata['test_accuracy']:.4f}\")\n",
-    "    print(f\"Test F1: {loaded_metadata['test_f1']:.4f}\")\n",
-    "\n",
-    "    test_text = \"Mistral is a family of large language models (LLMs) developed by Mistral AI, a French artificial intelligence company founded in 2023. Designed as open and efficient alternatives to proprietary frontier models, Mistral LLMs are optimized for reasoning, multilingual understanding, and enterprise deployment across a variety of contexts.\"\n",
-    "    # test_text = \"By converting documents into numerical vectors TF-IDF enables comparison and grouping of related texts. \"\n",
-    "    test_text  = \"sensitive skin needs lotion, lotion nourishes the skin and prevents it from cracking and being powdery and itchy. It is a lifesaver in winter season specially cause the season is really dry.\"\n",
-    "    test_text = \"analyze url input box will also be highlighted in the can-web instead of free tools section, and on clicking goes to another route like tools in sms\"\n",
-    "    test_text = \"i eat rice because i am hungry but i like roti more. Rice is easy to cook roti is not, so i have to go with rice everytime. But roti wins anytime any day any week or any place in this entire universe. Roti is the sole winner of this competition rice doesnt even come close to the superiority of rotiiii!!\"\n",
-    "    test_text = \"The cat sat on the mat. It was a sunny day. The cat was happy.\"\n",
-    "    test_text = \"Web Application (commonly known as Web Apps) is an application software that runs on the web server and is accessed via the web browser over the internet\"\n",
-    "    test_text = \"I've been living a very dull life. Every time I try to pursue my dreams, some dark unimaginable force tries to stop me. The whole process itself is very draining but what life even is if there is nothing to challenge you to your fullest. I think more the failures are in your door step, the more allurious  your success will be.\"\n",
-    "    test_text = \"What life even is when there is not a single will power bone in your body? Are you living or is it you who is trying to just get by? She often questions herself. But never seems to find an answer. She's been the eldest daughter all her life and  a break is never in her life book.\"\n",
-    "    features = extract_all_features(test_text, calc_perplexity=True)\n",
-    "    feature_vector = np.array([features[name] for name in loaded_features]).reshape(1, -1)\n",
-    "    feature_scaled = loaded_scaler.transform(feature_vector)\n",
-    "\n",
-    "    word_vec = loaded_word_vectorizer.transform([test_text])\n",
-    "    char_vec = loaded_char_vectorizer.transform([test_text])\n",
-    "    num_vec = csr_matrix(feature_scaled)\n",
-    "    hybrid_vec = hstack([word_vec, char_vec, num_vec], format=\"csr\")\n",
-    "\n",
-    "    prediction = loaded_classifier.predict(hybrid_vec)[0]\n",
-    "    if hasattr(loaded_classifier, \"predict_proba\"):\n",
-    "        proba = loaded_classifier.predict_proba(hybrid_vec)[0]\n",
-    "    else:\n",
-    "        score = loaded_classifier.decision_function(hybrid_vec)[0]\n",
-    "        p_ai = 1 / (1 + np.exp(-score))\n",
-    "        proba = np.array([1 - p_ai, p_ai])\n",
-    "\n",
-    "    print(f\"Test inference: '{test_text}'\")\n",
-    "    print(f\"Prediction: {'AI' if prediction == 1 else 'Human'}\")\n",
-    "    print(f\"Probabilities: Human={proba[0]:.2%}, AI={proba[1]:.2%}\")\n",
-    "\n",
-    "    print(\"Saved model works correctly\")\n",
-    "else:\n",
-    "    print(f\"Model not found at: {model_dir}\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "da4780eb",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "[{'text': 'This is a simple sentence.',\n",
-       "  'error': \"name 'predict_v3' is not defined\"},\n",
-       " {'text': 'The cat sat on the mat.',\n",
-       "  'error': \"name 'predict_v3' is not defined\"},\n",
-       " {'text': 'Artificial intelligence is transforming industries worldwide.',\n",
-       "  'error': \"name 'predict_v3' is not defined\"},\n",
-       " {'text': 'I had a great day at the park with my friends!',\n",
-       "  'error': \"name 'predict_v3' is not defined\"},\n",
-       " {'text': \"Hello, Today's environment is perfect for a picnic and the sun is shining in the sky.\",\n",
-       "  'error': \"name 'predict_v3' is not defined\"}]"
-      ]
-     },
-     "execution_count": 25,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "def predict_v3_batch(texts: list[str]) -> list[dict[str, Any]]:\n",
-    "    \"\"\"Predict if texts are AI or human in batch.\"\"\"\n",
-    "    results = []\n",
-    "    for text in texts:\n",
-    "        try:\n",
-    "            result = predict_v3(text)\n",
-    "            results.append(result)\n",
-    "        except Exception as e:\n",
-    "            results.append({\"text\": text[:100] + \"...\" if len(text) > 100 else text, \"error\": str(e)})\n",
-    "    return results\n",
-    "word_list = [\n",
-    "    \"This is a simple sentence.\",\n",
-    "    \"The cat sat on the mat.\",\n",
-    "    \"Artificial intelligence is transforming industries worldwide.\",\n",
-    "    \"I had a great day at the park with my friends!\",\n",
-    "    \"Hello, Today's environment is perfect for a picnic and the sun is shining in the sky.\",\n",
-    "    \"In the world of machine and robots, we are the only human alive\"\n",
-    "]\n",
-    "predict_v3_batch(word_list)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "b0acb82a",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "ml",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.11.14"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 5
-}