{
 "metadata": {
  "kernelspec": {
   "language": "python",
   "display_name": "Python 3",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.12.13",
   "mimetype": "text/x-python",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "pygments_lexer": "ipython3",
   "nbconvert_exporter": "python",
   "file_extension": ".py"
  },
  "kaggle": {
   "accelerator": "nvidiaTeslaT4",
   "dataSources": [
    {
     "sourceType": "datasetVersion",
     "sourceId": 16583191,
     "datasetId": 10617394,
     "databundleVersionId": 17591786,
     "mountSlug": "datasets/hardikgautam24119014/datasetnew"
    }
   ],
   "dockerImageVersionId": 31401,
   "isInternetEnabled": true,
   "language": "python",
   "sourceType": "notebook",
   "isGpuEnabled": true
  }
 },
 "nbformat_minor": 4,
 "nbformat": 4,
 "cells": [
  {
   "cell_type": "code",
   "source": "# Import standard libraries, PyTorch modules, and check CUDA device compatibility\nimport os\nimport cv2\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\n\nfrom tqdm import tqdm\n\nimport torch\nimport torch.nn as nn\nfrom torch.utils.data import Dataset, DataLoader\nfrom torchvision import models\n\nfrom sklearn.model_selection import train_test_split\n\nprint(\"PyTorch Version:\", torch.__version__)\nprint(\"CUDA Available:\", torch.cuda.is_available())\n\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nprint(\"Using Device:\", device)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.342862Z",
     "iopub.execute_input": "2026-06-02T07:19:30.343648Z",
     "iopub.status.idle": "2026-06-02T07:19:30.349908Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.343613Z",
     "shell.execute_reply": "2026-06-02T07:19:30.348923Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "PyTorch Version: 2.10.0+cu128\nCUDA Available: True\nUsing Device: cuda\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 80
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Dataset Overview\n\nThe dataset contains distorted visual sequences (CAPTCHAs) used for pattern recognition tasks. It consists of training images and a CSV file (`train-labels.csv`) linking image file names to their corresponding text labels."
  },
  {
   "cell_type": "code",
   "source": "# Load the CAPTCHA training labels CSV and display its shape and preview\nlabels_df = pd.read_csv('/kaggle/input/datasets/hardikgautam24119014/datasetnew/cig_ps/train-labels.csv')\n\nprint(\"Shape:\", labels_df.shape)\n\nlabels_df.head()",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.351252Z",
     "iopub.execute_input": "2026-06-02T07:19:30.351669Z",
     "iopub.status.idle": "2026-06-02T07:19:30.393514Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.351646Z",
     "shell.execute_reply": "2026-06-02T07:19:30.392689Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Shape: (20000, 3)\n",
     "output_type": "stream"
    },
    {
     "execution_count": 81,
     "output_type": "execute_result",
     "data": {
      "text/plain": "   Unnamed: 0        image    text\n0           0  train-0.png  BU522X\n1           1  train-1.png  XQ8NE2\n2           2  train-2.png  DTZD3E\n3           3  train-3.png  SM424H\n4           4  train-4.png  6YVTQR",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Unnamed: 0</th>\n      <th>image</th>\n      <th>text</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>train-0.png</td>\n      <td>BU522X</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>train-1.png</td>\n      <td>XQ8NE2</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>train-2.png</td>\n      <td>DTZD3E</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3</td>\n      <td>train-3.png</td>\n      <td>SM424H</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>train-4.png</td>\n      <td>6YVTQR</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ],
   "execution_count": 81
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Data Exploration\n\nExploratory data analysis is performed to understand label lengths, unique character distribution, missing values, and visual variations of the images. Visualizations and distribution plots reveal consistent sequence lengths and image dimensions of 100x200 pixels."
  },
  {
   "cell_type": "code",
   "source": "# Display DataFrame columns, data types, and non-null count information\nlabels_df.info()",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.394913Z",
     "iopub.execute_input": "2026-06-02T07:19:30.395409Z",
     "iopub.status.idle": "2026-06-02T07:19:30.406130Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.395386Z",
     "shell.execute_reply": "2026-06-02T07:19:30.405417Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 20000 entries, 0 to 19999\nData columns (total 3 columns):\n #   Column      Non-Null Count  Dtype \n---  ------      --------------  ----- \n 0   Unnamed: 0  20000 non-null  int64 \n 1   image       20000 non-null  object\n 2   text        20000 non-null  object\ndtypes: int64(1), object(2)\nmemory usage: 468.9+ KB\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 82
  },
  {
   "cell_type": "code",
   "source": "# Generate summary statistics for categorical columns in the dataset\nlabels_df.describe(include='object')",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.407136Z",
     "iopub.execute_input": "2026-06-02T07:19:30.407480Z",
     "iopub.status.idle": "2026-06-02T07:19:30.441124Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.407442Z",
     "shell.execute_reply": "2026-06-02T07:19:30.440532Z"
    }
   },
   "outputs": [
    {
     "execution_count": 83,
     "output_type": "execute_result",
     "data": {
      "text/plain": "              image    text\ncount         20000   20000\nunique        20000   19999\ntop     train-0.png  S9WJPK\nfreq              1       2",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>image</th>\n      <th>text</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>count</th>\n      <td>20000</td>\n      <td>20000</td>\n    </tr>\n    <tr>\n      <th>unique</th>\n      <td>20000</td>\n      <td>19999</td>\n    </tr>\n    <tr>\n      <th>top</th>\n      <td>train-0.png</td>\n      <td>S9WJPK</td>\n    </tr>\n    <tr>\n      <th>freq</th>\n      <td>1</td>\n      <td>2</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ],
   "execution_count": 83
  },
  {
   "cell_type": "code",
   "source": "# Check for missing values across all columns\nlabels_df.isnull().sum()",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.442943Z",
     "iopub.execute_input": "2026-06-02T07:19:30.443183Z",
     "iopub.status.idle": "2026-06-02T07:19:30.451026Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.443161Z",
     "shell.execute_reply": "2026-06-02T07:19:30.450207Z"
    }
   },
   "outputs": [
    {
     "execution_count": 84,
     "output_type": "execute_result",
     "data": {
      "text/plain": "Unnamed: 0    0\nimage         0\ntext          0\ndtype: int64"
     },
     "metadata": {}
    }
   ],
   "execution_count": 84
  },
  {
   "cell_type": "code",
   "source": "# Preview the full labels DataFrame reference\nlabels_df.head",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.452090Z",
     "iopub.execute_input": "2026-06-02T07:19:30.452521Z",
     "iopub.status.idle": "2026-06-02T07:19:30.464453Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.452480Z",
     "shell.execute_reply": "2026-06-02T07:19:30.463707Z"
    }
   },
   "outputs": [
    {
     "execution_count": 85,
     "output_type": "execute_result",
     "data": {
      "text/plain": "<bound method NDFrame.head of        Unnamed: 0            image    text\n0               0      train-0.png  BU522X\n1               1      train-1.png  XQ8NE2\n2               2      train-2.png  DTZD3E\n3               3      train-3.png  SM424H\n4               4      train-4.png  6YVTQR\n...           ...              ...     ...\n19995       19995  train-19995.png  ST9ERQ\n19996       19996  train-19996.png  CCYT9D\n19997       19997  train-19997.png  9X4DMK\n19998       19998  train-19998.png  U38HBZ\n19999       19999  train-19999.png  ZU5T2N\n\n[20000 rows x 3 columns]>"
     },
     "metadata": {}
    }
   ],
   "execution_count": 85
  },
  {
   "cell_type": "code",
   "source": "# Compute label character lengths to verify the expected sequence length\nlabels_df['label_length'] = labels_df['text'].apply(len)\n\nprint(\"Maximum Length :\", labels_df['label_length'].max())\nprint(\"Average Length :\", labels_df['label_length'].mean())\nprint(\"Minimum Length :\", labels_df['label_length'].min())",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.465392Z",
     "iopub.execute_input": "2026-06-02T07:19:30.465701Z",
     "iopub.status.idle": "2026-06-02T07:19:30.482594Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.465679Z",
     "shell.execute_reply": "2026-06-02T07:19:30.481834Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Maximum Length : 9\nAverage Length : 6.00025\nMinimum Length : 6\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 86
  },
  {
   "cell_type": "code",
   "source": "# Identify and preview labels containing anomalies or length deviation\nlabels_df[labels_df['label_length'] > 6].head(20)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.483616Z",
     "iopub.execute_input": "2026-06-02T07:19:30.483992Z",
     "iopub.status.idle": "2026-06-02T07:19:30.501549Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.483957Z",
     "shell.execute_reply": "2026-06-02T07:19:30.500852Z"
    }
   },
   "outputs": [
    {
     "execution_count": 87,
     "output_type": "execute_result",
     "data": {
      "text/plain": "      Unnamed: 0           image       text  label_length\n2184        2184  train-2184.png   5.40E+12             8\n6819        6819  train-6819.png  04-Mar-54             9",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Unnamed: 0</th>\n      <th>image</th>\n      <th>text</th>\n      <th>label_length</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>2184</th>\n      <td>2184</td>\n      <td>train-2184.png</td>\n      <td>5.40E+12</td>\n      <td>8</td>\n    </tr>\n    <tr>\n      <th>6819</th>\n      <td>6819</td>\n      <td>train-6819.png</td>\n      <td>04-Mar-54</td>\n      <td>9</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "metadata": {}
    }
   ],
   "execution_count": 87
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Preprocessing\n\nImages are converted to grayscale and normalized to float32 values between 0.0 and 1.0. Corrupted or invalid labels (such as those in scientific notation or date formats) that do not match the expected length of 6 characters are filtered out from the dataset."
  },
  {
   "cell_type": "code",
   "source": "# Filter the dataset to include only valid 6-character sequences\nlabels_df = labels_df[\n    labels_df['label_length'] == 6\n].copy()\n\nlabels_df.reset_index(drop=True, inplace=True)\n\nprint(labels_df.shape)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.502526Z",
     "iopub.execute_input": "2026-06-02T07:19:30.502983Z",
     "iopub.status.idle": "2026-06-02T07:19:30.515809Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.502956Z",
     "shell.execute_reply": "2026-06-02T07:19:30.514957Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "(19998, 4)\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 88
  },
  {
   "cell_type": "code",
   "source": "# Build vocabulary of unique characters from the filtered CAPTCHA text\nall_text = ''.join(labels_df['text'])\n\nvocab = sorted(set(all_text))\n\nprint(\"Vocabulary Size:\", len(vocab))\nprint(vocab)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.516710Z",
     "iopub.execute_input": "2026-06-02T07:19:30.516979Z",
     "iopub.status.idle": "2026-06-02T07:19:30.530190Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.516957Z",
     "shell.execute_reply": "2026-06-02T07:19:30.529417Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Vocabulary Size: 31\n['2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'J', 'K', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z']\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 89
  },
  {
   "cell_type": "code",
   "source": "# Verify that all remaining labels are exactly 6 characters long\nprint(\"Labels longer than 6 chars:\",\n      len(labels_df[labels_df['label_length'] > 6]))",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.532426Z",
     "iopub.execute_input": "2026-06-02T07:19:30.532734Z",
     "iopub.status.idle": "2026-06-02T07:19:30.542015Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.532690Z",
     "shell.execute_reply": "2026-06-02T07:19:30.541391Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Labels longer than 6 chars: 0\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 90
  },
  {
   "cell_type": "code",
   "source": "# Plot a histogram to visualize the label character length distribution\nplt.figure(figsize=(10,5))\nplt.hist(labels_df['label_length'], bins=20)\nplt.title(\"Text Length Distribution\")\nplt.xlabel(\"Length\")\nplt.ylabel(\"Frequency\")\nplt.show()",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.542918Z",
     "iopub.execute_input": "2026-06-02T07:19:30.543121Z",
     "iopub.status.idle": "2026-06-02T07:19:30.709054Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.543101Z",
     "shell.execute_reply": "2026-06-02T07:19:30.708384Z"
    }
   },
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 1000x500 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {}
    }
   ],
   "execution_count": 91
  },
  {
   "cell_type": "code",
   "source": "# Set path to the training images directory\nIMAGE_FOLDER = \"/kaggle/input/datasets/hardikgautam24119014/datasetnew/cig_ps/train_images\"",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.710037Z",
     "iopub.execute_input": "2026-06-02T07:19:30.710563Z",
     "iopub.status.idle": "2026-06-02T07:19:30.714544Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.710534Z",
     "shell.execute_reply": "2026-06-02T07:19:30.713739Z"
    }
   },
   "outputs": [],
   "execution_count": 92
  },
  {
   "cell_type": "code",
   "source": "# Read a subset of images to inspect and verify uniform spatial dimensions\nsizes = []\n\nfor img_name in labels_df['image'].head(1000):\n\n    img_path = os.path.join(\n        IMAGE_FOLDER,\n        img_name\n    )\n\n    img = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)\n\n    sizes.append(img.shape)\n\nprint(\"Unique sizes:\", len(set(sizes)))\nprint(set(sizes))",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:30.715626Z",
     "iopub.execute_input": "2026-06-02T07:19:30.716170Z",
     "iopub.status.idle": "2026-06-02T07:19:32.904646Z",
     "shell.execute_reply.started": "2026-06-02T07:19:30.716142Z",
     "shell.execute_reply": "2026-06-02T07:19:32.903851Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Unique sizes: 1\n{(100, 200)}\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 93
  },
  {
   "cell_type": "code",
   "source": "# Visualize random sample images with their associated labels\nsample = labels_df.sample(9)\n\nplt.figure(figsize=(12,8))\n\nfor i, (_, row) in enumerate(sample.iterrows()):\n\n    img_path = os.path.join(\n        IMAGE_FOLDER,\n        row['image']\n    )\n\n    image = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)\n\n    plt.subplot(3,3,i+1)\n    plt.imshow(image, cmap='gray')\n    plt.title(row['text'])      # <-- FIXED\n    plt.axis('off')\n\nplt.tight_layout()\nplt.show()",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:32.905672Z",
     "iopub.execute_input": "2026-06-02T07:19:32.905947Z",
     "iopub.status.idle": "2026-06-02T07:19:33.437936Z",
     "shell.execute_reply.started": "2026-06-02T07:19:32.905925Z",
     "shell.execute_reply": "2026-06-02T07:19:33.437305Z"
    }
   },
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "<Figure size 1200x800 with 9 Axes>",
      "image/png": "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\n"
     },
     "metadata": {}
    }
   ],
   "execution_count": 94
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Label Encoding\n\nThe sequence text labels are converted into numeric representations for model training. A character-to-index vocabulary maps each of the 31 unique alphanumeric characters to a unique integer index, enabling multi-character classification."
  },
  {
   "cell_type": "code",
   "source": "# Establish dictionary mappings between characters and integer indices\nchar_to_idx = {\n    char: idx + 1\n    for idx, char in enumerate(vocab)\n}\n\nidx_to_char = {\n    idx + 1: char\n    for idx, char in enumerate(vocab)\n}\n\nBLANK_IDX = 0\n\nNUM_CLASSES = len(vocab) + 1\n\nprint(\"Vocabulary Size:\", len(vocab))\nprint(\"Number of Classes:\", NUM_CLASSES)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.438812Z",
     "iopub.execute_input": "2026-06-02T07:19:33.439281Z",
     "iopub.status.idle": "2026-06-02T07:19:33.444906Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.439256Z",
     "shell.execute_reply": "2026-06-02T07:19:33.444046Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Vocabulary Size: 31\nNumber of Classes: 32\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 95
  },
  {
   "cell_type": "code",
   "source": "# Partition the dataset into stratified training (90%) and validation (10%) splits\nfrom sklearn.model_selection import train_test_split\n\ntrain_df, val_df = train_test_split(\n    labels_df,\n    test_size=0.10,\n    random_state=42,\n    shuffle=True\n)\n\nprint(\"Train:\", train_df.shape)\nprint(\"Validation:\", val_df.shape)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.447883Z",
     "iopub.execute_input": "2026-06-02T07:19:33.448135Z",
     "iopub.status.idle": "2026-06-02T07:19:33.465147Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.448113Z",
     "shell.execute_reply": "2026-06-02T07:19:33.464534Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Train: (17998, 4)\nValidation: (2000, 4)\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 96
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Dataset & DataLoader Creation\n\nA custom PyTorch `Dataset` is defined to handle image reading, normalization, tensor conversion, and label encoding. `DataLoader` instances are then created for both training and validation splits with a batch size of 128 to facilitate mini-batch gradient descent."
  },
  {
   "cell_type": "code",
   "source": "# Create a PyTorch Dataset for loading grayscale images and mapping alphanumeric labels\nclass CaptchaDataset(Dataset):\n\n    def __init__(self, dataframe, image_folder):\n\n        self.df = dataframe.reset_index(drop=True)\n        self.image_folder = image_folder\n\n    def __len__(self):\n        return len(self.df)\n\n    def __getitem__(self, idx):\n\n        row = self.df.iloc[idx]\n\n        img_path = os.path.join(\n            self.image_folder,\n            row['image']\n        )\n\n        image = cv2.imread(\n            img_path,\n            cv2.IMREAD_GRAYSCALE\n        )\n\n        image = image.astype(np.float32) / 255.0\n\n        image = torch.tensor(\n            image,\n            dtype=torch.float32\n        ).unsqueeze(0)\n\n        label = torch.tensor(\n            [char_to_idx[ch] - 1 for ch in row['text']],\n            dtype=torch.long\n        )\n\n        return image, label",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.466091Z",
     "iopub.execute_input": "2026-06-02T07:19:33.466535Z",
     "iopub.status.idle": "2026-06-02T07:19:33.475986Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.466511Z",
     "shell.execute_reply": "2026-06-02T07:19:33.475368Z"
    }
   },
   "outputs": [],
   "execution_count": 97
  },
  {
   "cell_type": "code",
   "source": "# Instantiate CaptchaDataset for train and validation splits\ntrain_dataset = CaptchaDataset(\n    train_df,\n    IMAGE_FOLDER\n)\n\nval_dataset = CaptchaDataset(\n    val_df,\n    IMAGE_FOLDER\n)\n\nprint(len(train_dataset))\nprint(len(val_dataset))",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.476876Z",
     "iopub.execute_input": "2026-06-02T07:19:33.477235Z",
     "iopub.status.idle": "2026-06-02T07:19:33.492025Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.477197Z",
     "shell.execute_reply": "2026-06-02T07:19:33.491107Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "17998\n2000\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 98
  },
  {
   "cell_type": "code",
   "source": "# Retrieve and inspect a single dataset item to verify tensor shapes\nimage, label = train_dataset[0]\n\nprint(\"Image Shape:\", image.shape)\nprint(\"Encoded Label:\", label)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.493121Z",
     "iopub.execute_input": "2026-06-02T07:19:33.493479Z",
     "iopub.status.idle": "2026-06-02T07:19:33.508947Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.493444Z",
     "shell.execute_reply": "2026-06-02T07:19:33.508381Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Image Shape: torch.Size([1, 100, 200])\nEncoded Label: tensor([ 6, 26, 23, 19,  6,  9])\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 99
  },
  {
   "cell_type": "code",
   "source": "# Create DataLoader instances for batching, shuffling, and pinned memory access\ntrain_loader = DataLoader(\n    train_dataset,\n    batch_size=128,\n    shuffle=True,\n    num_workers=0,\n    pin_memory=True\n)\n\nval_loader = DataLoader(\n    val_dataset,\n    batch_size=128,\n    shuffle=False,\n    num_workers=0,\n    pin_memory=True\n)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.509881Z",
     "iopub.execute_input": "2026-06-02T07:19:33.510424Z",
     "iopub.status.idle": "2026-06-02T07:19:33.519047Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.510391Z",
     "shell.execute_reply": "2026-06-02T07:19:33.518363Z"
    }
   },
   "outputs": [],
   "execution_count": 100
  },
  {
   "cell_type": "code",
   "source": "# Fetch and inspect a batch of images and labels from the DataLoader\nimages, labels = next(iter(train_loader))\n\nprint(images.shape)\nprint(labels.shape)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.519983Z",
     "iopub.execute_input": "2026-06-02T07:19:33.520300Z",
     "iopub.status.idle": "2026-06-02T07:19:33.866320Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.520269Z",
     "shell.execute_reply": "2026-06-02T07:19:33.865353Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "torch.Size([128, 1, 100, 200])\ntorch.Size([128, 6])\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 101
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Model Architecture\n\nThe pattern recognition model utilizes a modified ResNet-18 backbone. The first convolutional layer is adapted to accept 1-channel grayscale input, and the spatial dimensions are pooled using `AdaptiveAvgPool2d` to extract features corresponding to 6 sequence character slots for classification."
  },
  {
   "cell_type": "code",
   "source": "# Define ResNet-18 model modified for single-channel input and 6 character slot classification\nclass ResNetCaptcha(nn.Module):\n\n    def __init__(self, num_chars):\n\n        super().__init__()\n\n        backbone = models.resnet18(weights=\"DEFAULT\")\n\n        backbone.conv1 = nn.Conv2d(\n            1,\n            64,\n            kernel_size=7,\n            stride=2,\n            padding=3,\n            bias=False\n        )\n\n        self.features = nn.Sequential(\n            backbone.conv1,\n            backbone.bn1,\n            backbone.relu,\n            backbone.maxpool,\n\n            backbone.layer1,\n            backbone.layer2,\n            backbone.layer3,\n            backbone.layer4\n        )\n\n        self.pool = nn.AdaptiveAvgPool2d((1, 6))\n\n        self.classifier = nn.Linear(\n            512,\n            num_chars\n        )\n\n        self.num_chars = num_chars\n\n    def forward(self, x):\n\n        x = self.features(x)\n\n        x = self.pool(x)\n\n        x = x.squeeze(2)\n\n        x = x.permute(0, 2, 1)\n\n        x = self.classifier(x)\n\n        return x",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.867586Z",
     "iopub.execute_input": "2026-06-02T07:19:33.867859Z",
     "iopub.status.idle": "2026-06-02T07:19:33.874565Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.867836Z",
     "shell.execute_reply": "2026-06-02T07:19:33.873775Z"
    }
   },
   "outputs": [],
   "execution_count": 102
  },
  {
   "cell_type": "code",
   "source": "# Instantiate ResNetCaptcha model and move to active device\nNUM_CHARS = len(vocab)\n\nmodel = ResNetCaptcha(\n    NUM_CHARS\n).to(device)\n\nprint(model)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:33.875538Z",
     "iopub.execute_input": "2026-06-02T07:19:33.875950Z",
     "iopub.status.idle": "2026-06-02T07:19:34.097042Z",
     "shell.execute_reply.started": "2026-06-02T07:19:33.875923Z",
     "shell.execute_reply": "2026-06-02T07:19:34.096383Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "ResNetCaptcha(\n  (features): Sequential(\n    (0): Conv2d(1, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n    (2): ReLU(inplace=True)\n    (3): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n    (4): Sequential(\n      (0): BasicBlock(\n        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n      )\n      (1): BasicBlock(\n        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n      )\n    )\n    (5): Sequential(\n      (0): BasicBlock(\n        (conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (downsample): Sequential(\n          (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)\n          (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        )\n      )\n      (1): BasicBlock(\n        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n      )\n    )\n    (6): Sequential(\n      (0): BasicBlock(\n        (conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (downsample): Sequential(\n          (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)\n          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        )\n      )\n      (1): BasicBlock(\n        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n      )\n    )\n    (7): Sequential(\n      (0): BasicBlock(\n        (conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (downsample): Sequential(\n          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        )\n      )\n      (1): BasicBlock(\n        (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n        (relu): ReLU(inplace=True)\n        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n      )\n    )\n  )\n  (pool): AdaptiveAvgPool2d(output_size=(1, 6))\n  (classifier): Linear(in_features=512, out_features=31, bias=True)\n)\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 103
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Training Strategy\n\nThe model is trained using the AdamW optimizer with a learning rate of 3e-4 and weight decay of 1e-4. The loss function is CrossEntropyLoss with 0.1 label smoothing, and a ReduceLROnPlateau scheduler dynamically decreases the learning rate on validation loss plateau."
  },
  {
   "cell_type": "code",
   "source": "# Initialize multi-label classification loss with label smoothing\ncriterion = nn.CrossEntropyLoss(\n    label_smoothing=0.1\n)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T08:09:24.039003Z",
     "iopub.execute_input": "2026-06-02T08:09:24.039844Z",
     "iopub.status.idle": "2026-06-02T08:09:24.043948Z",
     "shell.execute_reply.started": "2026-06-02T08:09:24.039813Z",
     "shell.execute_reply": "2026-06-02T08:09:24.043060Z"
    }
   },
   "outputs": [],
   "execution_count": 128
  },
  {
   "cell_type": "code",
   "source": "# Initialize AdamW optimizer with weight decay\noptimizer = torch.optim.AdamW(\n    model.parameters(),\n    lr=3e-4,\n    weight_decay=1e-4\n)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.103280Z",
     "iopub.execute_input": "2026-06-02T07:19:34.103931Z",
     "iopub.status.idle": "2026-06-02T07:19:34.115992Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.103905Z",
     "shell.execute_reply": "2026-06-02T07:19:34.115256Z"
    }
   },
   "outputs": [],
   "execution_count": 105
  },
  {
   "cell_type": "code",
   "source": "# Set up ReduceLROnPlateau learning rate scheduler\nscheduler = torch.optim.lr_scheduler.ReduceLROnPlateau(\n    optimizer,\n    mode='min',\n    factor=0.5,\n    patience=2\n)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.116942Z",
     "iopub.execute_input": "2026-06-02T07:19:34.117272Z",
     "iopub.status.idle": "2026-06-02T07:19:34.128004Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.117250Z",
     "shell.execute_reply": "2026-06-02T07:19:34.127400Z"
    }
   },
   "outputs": [],
   "execution_count": 106
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Evaluation Metrics\n\nPerformance is monitored using character-level accuracy, sequence-level accuracy (requiring all characters to match), and Character Error Rate (CER). CER measures the distance between the predicted text and ground truth using insertion, deletion, and substitution edits."
  },
  {
   "cell_type": "code",
   "source": "# Helper function to calculate average character-level classification accuracy\ndef character_accuracy(outputs, labels):\n\n    preds = outputs.argmax(dim=2)\n\n    correct = (preds == labels).sum().item()\n\n    total = labels.numel()\n\n    return 100 * correct / total",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.128990Z",
     "iopub.execute_input": "2026-06-02T07:19:34.129264Z",
     "iopub.status.idle": "2026-06-02T07:19:34.140633Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.129233Z",
     "shell.execute_reply": "2026-06-02T07:19:34.139845Z"
    }
   },
   "outputs": [],
   "execution_count": 107
  },
  {
   "cell_type": "code",
   "source": "# Helper function to compute exact sequence-level match accuracy\ndef calculate_accuracy(outputs, labels):\n\n    preds = outputs.argmax(dim=2)\n\n    correct = (preds == labels).all(dim=1)\n\n    return correct.float().mean().item() * 100",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.141638Z",
     "iopub.execute_input": "2026-06-02T07:19:34.141928Z",
     "iopub.status.idle": "2026-06-02T07:19:34.158975Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.141896Z",
     "shell.execute_reply": "2026-06-02T07:19:34.158229Z"
    }
   },
   "outputs": [],
   "execution_count": 108
  },
  {
   "cell_type": "code",
   "source": "# Helper function to prepare target labels for classification\ndef prepare_targets(labels):\n\n    target_lengths = torch.full(\n        size=(labels.size(0),),\n        fill_value=labels.size(1),\n        dtype=torch.long\n    )\n\n    targets = labels.view(-1)\n\n    return targets, target_lengths",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.159797Z",
     "iopub.execute_input": "2026-06-02T07:19:34.160093Z",
     "iopub.status.idle": "2026-06-02T07:19:34.173155Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.160062Z",
     "shell.execute_reply": "2026-06-02T07:19:34.172383Z"
    }
   },
   "outputs": [],
   "execution_count": 109
  },
  {
   "cell_type": "code",
   "source": "# Decode integer prediction indices back into characters\ndef decode_prediction(pred):\n\n    text = \"\"\n\n    for idx in pred:\n        text += vocab[idx]\n\n    return text",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.174140Z",
     "iopub.execute_input": "2026-06-02T07:19:34.174430Z",
     "iopub.status.idle": "2026-06-02T07:19:34.186784Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.174390Z",
     "shell.execute_reply": "2026-06-02T07:19:34.185968Z"
    }
   },
   "outputs": [],
   "execution_count": 110
  },
  {
   "cell_type": "code",
   "source": "# Utility function to print ground truth versus model predictions for sample validation images\ndef show_predictions(model, loader, n=5):\n\n    model.eval()\n\n    with torch.no_grad():\n\n        images, labels = next(iter(loader))\n\n        outputs = model(\n            images.to(device)\n        )\n\n        preds = outputs.argmax(dim=2)\n\n        for i in range(n):\n\n            gt = decode_prediction(\n                labels[i].numpy()\n            )\n\n            pr = decode_prediction(\n                preds[i].cpu().numpy()\n            )\n\n            print(f\"GT   : {gt}\")\n            print(f\"PRED : {pr}\")\n            print(\"-\"*50)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.187822Z",
     "iopub.execute_input": "2026-06-02T07:19:34.188114Z",
     "iopub.status.idle": "2026-06-02T07:19:34.201045Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.188085Z",
     "shell.execute_reply": "2026-06-02T07:19:34.200506Z"
    }
   },
   "outputs": [],
   "execution_count": 111
  },
  {
   "cell_type": "code",
   "source": "# Decode batch of integer labels to text sequences\ndef decode_labels(labels):\n\n    texts = []\n\n    for label in labels:\n\n        text = \"\"\n\n        for idx in label.cpu().numpy():\n\n            text += idx_to_char.get(idx, \"\")\n\n        texts.append(text)\n\n    return texts",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.203890Z",
     "iopub.execute_input": "2026-06-02T07:19:34.204483Z",
     "iopub.status.idle": "2026-06-02T07:19:34.214713Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.204461Z",
     "shell.execute_reply": "2026-06-02T07:19:34.213972Z"
    }
   },
   "outputs": [],
   "execution_count": 112
  },
  {
   "cell_type": "code",
   "source": "# Run validation evaluation loop, computing loss, sequence match accuracy, and character accuracy\ndef validate(model, loader):\n\n    model.eval()\n\n    total_loss = 0\n    total_seq_acc = 0\n    total_char_acc = 0\n\n    with torch.no_grad():\n\n        for images, labels in loader:\n\n            images = images.to(device)\n            labels = labels.to(device)\n\n            outputs = model(images)\n\n            loss = 0\n\n            for i in range(6):\n\n                loss += criterion(\n                    outputs[:, i, :],\n                    labels[:, i]\n                )\n\n            total_loss += loss.item()\n\n            seq_acc = calculate_accuracy(\n                outputs,\n                labels\n            )\n\n            char_acc = character_accuracy(\n                outputs,\n                labels\n            )\n\n            total_seq_acc += seq_acc\n            total_char_acc += char_acc\n\n    avg_loss = total_loss / len(loader)\n    avg_seq_acc = total_seq_acc / len(loader)\n    avg_char_acc = total_char_acc / len(loader)\n\n    return avg_loss, avg_seq_acc, avg_char_acc",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:19:34.215625Z",
     "iopub.execute_input": "2026-06-02T07:19:34.216107Z",
     "iopub.status.idle": "2026-06-02T07:19:34.227088Z",
     "shell.execute_reply.started": "2026-06-02T07:19:34.216073Z",
     "shell.execute_reply": "2026-06-02T07:19:34.226440Z"
    }
   },
   "outputs": [],
   "execution_count": 113
  },
  {
   "cell_type": "code",
   "source": "# Train the model for 40 epochs, tracking validation metrics and saving the best checkpoint\nfrom tqdm.auto import tqdm\nimport time\nimport torch\n\nEPOCHS = 40\n\nbest_char_acc = 0\n\nfor epoch in range(EPOCHS):\n\n    model.train()\n\n    running_loss = 0\n\n    start_time = time.time()\n\n    pbar = tqdm(\n        train_loader,\n        desc=f\"Epoch {epoch+1}/{EPOCHS}\"\n    )\n\n    for images, labels in pbar:\n\n        images = images.to(device)\n        labels = labels.to(device)\n\n        outputs = model(images)\n\n        loss = 0\n\n        for i in range(6):\n\n            loss += criterion(\n                outputs[:, i, :],\n                labels[:, i]\n            )\n\n        optimizer.zero_grad()\n\n        loss.backward()\n\n        torch.nn.utils.clip_grad_norm_(\n            model.parameters(),\n            max_norm=5.0\n        )\n\n        optimizer.step()\n\n        running_loss += loss.item()\n\n        pbar.set_postfix({\n            \"Loss\": f\"{loss.item():.4f}\",\n            \"LR\": f\"{optimizer.param_groups[0]['lr']:.6f}\"\n        })\n\n    val_loss, val_acc, char_acc = validate(\n    model,\n    val_loader\n        )\n\n    scheduler.step(val_loss)\n\n    current_lr = optimizer.param_groups[0]['lr']\n\n    epoch_time = time.time() - start_time\n\n    print(\"\\n\" + \"=\" * 70)\n    print(f\"Epoch         : {epoch+1}/{EPOCHS}\")\n    print(f\"Train Loss    : {running_loss / len(train_loader):.4f}\")\n    print(f\"Val Accuracy  : {val_acc:.2f}%\")\n    print(f\"Char Accuracy : {char_acc:.2f}%\")\n    print(f\"Learning Rate : {current_lr:.6f}\")\n    print(f\"Time          : {epoch_time:.2f}s\")\n    print(\"=\" * 70)\n\n    show_predictions(\n        model,\n        val_loader,\n        5\n    )\n\n    if char_acc > best_char_acc:\n\n        best_char_acc = char_acc\n\n        torch.save(\n            model.state_dict(),\n            \"best_resnet_captcha.pth\"\n        )\n\n        print(\"\\n✅ New Best Model Saved\")\n        print(f\"✅ Best Char Accuracy : {best_char_acc:.2f}%\")",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T08:09:39.849166Z",
     "iopub.execute_input": "2026-06-02T08:09:39.849924Z",
     "iopub.status.idle": "2026-06-02T08:49:48.280461Z",
     "shell.execute_reply.started": "2026-06-02T08:09:39.849896Z",
     "shell.execute_reply": "2026-06-02T08:49:48.279573Z"
    }
   },
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 1/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "bff36edfc9fe4d7f90a6430095f6e619"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 1/40\nTrain Loss    : 4.8876\nVal Accuracy  : 99.14%\nChar Accuracy : 99.83%\nLearning Rate : 0.000300\nTime          : 72.95s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNY5DG\n--------------------------------------------------\n\n✅ New Best Model Saved\n✅ Best Char Accuracy : 99.83%\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 2/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "df1e5b815a754d4ead7ddfb542ecb454"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 2/40\nTrain Loss    : 4.1404\nVal Accuracy  : 99.48%\nChar Accuracy : 99.90%\nLearning Rate : 0.000300\nTime          : 56.08s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n\n✅ New Best Model Saved\n✅ Best Char Accuracy : 99.90%\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 3/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "405d6883d900416e994240a010c90df9"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 3/40\nTrain Loss    : 4.0410\nVal Accuracy  : 99.61%\nChar Accuracy : 99.92%\nLearning Rate : 0.000150\nTime          : 57.05s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n\n✅ New Best Model Saved\n✅ Best Char Accuracy : 99.92%\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 4/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "62f210d9f6f04f4c9356e688b8e680ce"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 4/40\nTrain Loss    : 3.9924\nVal Accuracy  : 99.71%\nChar Accuracy : 99.94%\nLearning Rate : 0.000150\nTime          : 56.37s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n\n✅ New Best Model Saved\n✅ Best Char Accuracy : 99.94%\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 5/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "51eb49d6fd11495e923ed51d2121687b"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 5/40\nTrain Loss    : 3.9772\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000150\nTime          : 59.06s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNY5DG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 6/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "45279057053f4708a07b0f36a6e5a13d"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 6/40\nTrain Loss    : 3.9671\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000075\nTime          : 58.45s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 7/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "b0dfb5943c4d43f7a159c7023b9daf8a"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 7/40\nTrain Loss    : 3.9566\nVal Accuracy  : 99.66%\nChar Accuracy : 99.93%\nLearning Rate : 0.000075\nTime          : 58.08s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 8/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "d128110fc7b643a88a48e3fc6d97b5ac"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 8/40\nTrain Loss    : 3.9521\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000075\nTime          : 60.35s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 9/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "c26776d5c52c42f7b975044429f21c5b"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 9/40\nTrain Loss    : 3.9494\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000037\nTime          : 61.89s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 10/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "cd9492fba23e42bb84338a40182672d2"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 10/40\nTrain Loss    : 3.9447\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000037\nTime          : 61.56s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 11/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "7026be6b5b3342ea9ef50de50e4b64ad"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 11/40\nTrain Loss    : 3.9432\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000037\nTime          : 63.43s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 12/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "fe32f2f581174e2886f58e6d72348080"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 12/40\nTrain Loss    : 3.9417\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000019\nTime          : 57.98s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 13/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "c272f1d6257741edb58a802c9cd5492c"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 13/40\nTrain Loss    : 3.9398\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000019\nTime          : 59.06s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 14/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "60908d6b4fec450ab6d4b7a270a22d0f"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 14/40\nTrain Loss    : 3.9390\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000019\nTime          : 59.81s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 15/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "33b3c068b3cc496a909aef68c27ab7ac"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 15/40\nTrain Loss    : 3.9385\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000009\nTime          : 58.78s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 16/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "3d1d087f703146ec952f581de1c7fac8"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 16/40\nTrain Loss    : 3.9371\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000009\nTime          : 59.24s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 17/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "07a691b1df184a458142bd1b7d8c6587"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 17/40\nTrain Loss    : 3.9370\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000009\nTime          : 57.68s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 18/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "5f2abdda2d73441c948a932e5feff901"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 18/40\nTrain Loss    : 3.9368\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000005\nTime          : 63.38s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 19/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "4af3b44a77704968af4d2431dff5072b"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 19/40\nTrain Loss    : 3.9358\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000005\nTime          : 64.16s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 20/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "c768e7dc179f4b36808ab4bf95ef07f8"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 20/40\nTrain Loss    : 3.9355\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000005\nTime          : 65.92s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 21/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "94e6b0a977c544b18bc2dd5174d4f960"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 21/40\nTrain Loss    : 3.9356\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000002\nTime          : 60.86s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 22/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "f1ee661def3c4e1d93a42be12fc4bd12"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 22/40\nTrain Loss    : 3.9352\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000002\nTime          : 58.31s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 23/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "2fb9ce5168bd4b9fa8a655c58509b3aa"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 23/40\nTrain Loss    : 3.9349\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000002\nTime          : 57.54s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 24/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "c952e385a66e4bb1a62962b751c55e50"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 24/40\nTrain Loss    : 3.9348\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000001\nTime          : 57.73s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 25/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "5ad10258d33b4739bb9781fe7ea03580"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 25/40\nTrain Loss    : 3.9346\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000001\nTime          : 60.17s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 26/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "97c19a14d1cf4902a28fc202836dabdd"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 26/40\nTrain Loss    : 3.9346\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000001\nTime          : 61.51s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 27/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "f456eafc5f5f4f419307a00c3d92fae6"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 27/40\nTrain Loss    : 3.9344\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000001\nTime          : 59.02s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 28/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "d62618afdb5e4f288e9e10acee224815"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 28/40\nTrain Loss    : 3.9346\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000001\nTime          : 56.58s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 29/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "5cced714675d44a4999e0a01341b5d21"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 29/40\nTrain Loss    : 3.9345\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000001\nTime          : 58.92s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 30/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "aa6bdf7522214f538e11ab3f7c3ff0a4"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 30/40\nTrain Loss    : 3.9345\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 61.72s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 31/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "0beed7d9883a4e7eb389bf17ce217d3d"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 31/40\nTrain Loss    : 3.9343\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 61.65s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 32/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "26f47cf85b8a47fea83858159e2d35cd"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 32/40\nTrain Loss    : 3.9344\nVal Accuracy  : 99.76%\nChar Accuracy : 99.93%\nLearning Rate : 0.000000\nTime          : 59.38s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 33/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "07a6083120a54d7ebd46abfe057848ed"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 33/40\nTrain Loss    : 3.9341\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000000\nTime          : 57.24s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 34/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "fab808e2a63a401ba5f212e0a676a3e6"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 34/40\nTrain Loss    : 3.9343\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 57.99s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 35/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "9e90caa7f93e4f509220b1020d99ec52"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 35/40\nTrain Loss    : 3.9343\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 57.92s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 36/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "3f6f199548724dc49cca08f7606d0364"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 36/40\nTrain Loss    : 3.9341\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 58.56s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 37/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "cfa47b6e9e08450ca1b13dd43cc77161"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 37/40\nTrain Loss    : 3.9344\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000000\nTime          : 60.12s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 38/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "7fe49be09b1041f594173d2da0b5dc65"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 38/40\nTrain Loss    : 3.9341\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 58.70s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 39/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "16a081efd2374089aa98f36bcf8f094d"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 39/40\nTrain Loss    : 3.9342\nVal Accuracy  : 99.71%\nChar Accuracy : 99.93%\nLearning Rate : 0.000000\nTime          : 61.41s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": "Epoch 40/40:   0%|          | 0/141 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "6dc8fc5c1d4f4504b14e18b604005396"
      }
     },
     "metadata": {}
    },
    {
     "name": "stdout",
     "text": "\n======================================================================\nEpoch         : 40/40\nTrain Loss    : 3.9342\nVal Accuracy  : 99.76%\nChar Accuracy : 99.94%\nLearning Rate : 0.000000\nTime          : 60.17s\n======================================================================\nGT   : TDJ3F4\nPRED : TDJ3F4\n--------------------------------------------------\nGT   : SEKSCT\nPRED : SEKSCT\n--------------------------------------------------\nGT   : 9EQ4U4\nPRED : 9EQ4U4\n--------------------------------------------------\nGT   : U3RBDY\nPRED : U3RBDY\n--------------------------------------------------\nGT   : MNYSDG\nPRED : MNYSDG\n--------------------------------------------------\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 129
  },
  {
   "cell_type": "code",
   "source": "# Install jiwer library for computing Character Error Rate (CER)\n!pip install jiwer",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T07:46:03.311021Z",
     "iopub.execute_input": "2026-06-02T07:46:03.311404Z",
     "iopub.status.idle": "2026-06-02T07:46:09.997544Z",
     "shell.execute_reply.started": "2026-06-02T07:46:03.311370Z",
     "shell.execute_reply": "2026-06-02T07:46:09.996383Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Collecting jiwer\n  Downloading jiwer-4.0.0-py3-none-any.whl.metadata (3.3 kB)\nRequirement already satisfied: click>=8.1.8 in /usr/local/lib/python3.12/dist-packages (from jiwer) (8.3.2)\nCollecting rapidfuzz>=3.9.7 (from jiwer)\n  Downloading rapidfuzz-3.14.5-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.metadata (12 kB)\nDownloading jiwer-4.0.0-py3-none-any.whl (23 kB)\nDownloading rapidfuzz-3.14.5-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (3.1 MB)\n\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.1/3.1 MB\u001b[0m \u001b[31m36.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n\u001b[?25hInstalling collected packages: rapidfuzz, jiwer\nSuccessfully installed jiwer-4.0.0 rapidfuzz-3.14.5\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 117
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Results\n\nAfter training for 40 epochs, the best model checkpoint is loaded and evaluated on the validation set. The model achieves high character accuracy (99.94%), sequence accuracy (99.70%), and a low Character Error Rate (CER) of 0.0006."
  },
  {
   "cell_type": "code",
   "source": "# Load the best validation checkpoint and calculate final accuracy and CER metrics\nfrom jiwer import cer\n\n# Load Best Model\nmodel.load_state_dict(\n    torch.load(\n        \"best_resnet_captcha.pth\",\n        map_location=device\n    )\n)\n\nmodel.eval()\n\ntotal_cer = 0\ntotal_chars = 0\ntotal_seq_correct = 0\ntotal_char_correct = 0\ntotal_samples = 0\n\nwith torch.no_grad():\n\n    for images, labels in val_loader:\n\n        images = images.to(device)\n\n        outputs = model(images)\n\n        preds = outputs.argmax(dim=2).cpu()\n\n        total_char_correct += (preds == labels).sum().item()\n\n        total_chars += labels.numel()\n\n        total_seq_correct += (\n            (preds == labels).all(dim=1)\n        ).sum().item()\n\n        total_samples += labels.size(0)\n\n        for i in range(labels.size(0)):\n\n            gt = decode_prediction(\n                labels[i].numpy()\n            )\n\n            pred = decode_prediction(\n                preds[i].numpy()\n            )\n\n            total_cer += cer(gt, pred)\n\nfinal_cer = total_cer / total_samples\n\nchar_acc = 100 * total_char_correct / total_chars\n\nseq_acc = 100 * total_seq_correct / total_samples\n\nprint(\"=\" * 60)\nprint(f\"Character Accuracy : {char_acc:.2f}%\")\nprint(f\"Sequence Accuracy  : {seq_acc:.2f}%\")\nprint(f\"CER                : {final_cer:.4f}\")\nprint(f\"CER (%)            : {final_cer*100:.2f}%\")\nprint(\"=\" * 60)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T08:54:20.627697Z",
     "iopub.execute_input": "2026-06-02T08:54:20.628533Z",
     "iopub.status.idle": "2026-06-02T08:54:26.612764Z",
     "shell.execute_reply.started": "2026-06-02T08:54:20.628503Z",
     "shell.execute_reply": "2026-06-02T08:54:26.611786Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "============================================================\nCharacter Accuracy : 99.94%\nSequence Accuracy  : 99.70%\nCER                : 0.0006\nCER (%)            : 0.06%\n============================================================\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 130
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Error Analysis\n\nIncorrectly predicted validation samples are extracted and analyzed to identify common failure modes. Reviewing these rare errors helps diagnose whether distortions like overlapping characters or high noise led to misclassifications."
  },
  {
   "cell_type": "code",
   "source": "# Scan validation set and compile list of misclassified instances for error analysis\nwrong_samples = []\n\nmodel.eval()\n\nwith torch.no_grad():\n\n    for images, labels in val_loader:\n\n        images = images.to(device)\n\n        outputs = model(images)\n\n        preds = outputs.argmax(dim=2).cpu()\n\n        for i in range(labels.size(0)):\n\n            gt = decode_prediction(\n                labels[i].numpy()\n            )\n\n            pred = decode_prediction(\n                preds[i].numpy()\n            )\n\n            if gt != pred:\n\n                wrong_samples.append(\n                    (images[i].cpu(), gt, pred)\n                )\n\nprint(\"Wrong Samples:\", len(wrong_samples))",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T08:54:38.664665Z",
     "iopub.execute_input": "2026-06-02T08:54:38.665265Z",
     "iopub.status.idle": "2026-06-02T08:54:42.790257Z",
     "shell.execute_reply.started": "2026-06-02T08:54:38.665233Z",
     "shell.execute_reply": "2026-06-02T08:54:42.789376Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "text": "Wrong Samples: 6\n",
     "output_type": "stream"
    }
   ],
   "execution_count": 131
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Test Prediction Generation\n\nTest predictions are generated on unseen images using the trained ResNet-18 model. The inference pipeline loads the best model checkpoint, runs predictions in evaluation mode, and outputs the results to a submission CSV file as shown in `02_predict.ipynb`."
  },
  {
   "cell_type": "code",
   "source": "# Save the final trained model weights\ntorch.save(\n    model.state_dict(),\n    \"final_resnet18_captcha.pth\"\n)",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T08:54:48.337937Z",
     "iopub.execute_input": "2026-06-02T08:54:48.338540Z",
     "iopub.status.idle": "2026-06-02T08:54:48.430518Z",
     "shell.execute_reply.started": "2026-06-02T08:54:48.338509Z",
     "shell.execute_reply": "2026-06-02T08:54:48.429550Z"
    }
   },
   "outputs": [],
   "execution_count": 132
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "### Reproducibility\n\nTo ensure reproducibility, model weights, optimizer state, and training metrics are saved to serialized PyTorch checkpoints. This allows the complete training state to be restored and evaluated consistently across environments."
  },
  {
   "cell_type": "code",
   "source": "# Save checkpoint containing model weights, optimizer state, and final metrics\ntorch.save({\n    \"model_state_dict\": model.state_dict(),\n    \"optimizer_state_dict\": optimizer.state_dict(),\n    \"char_acc\": 99.94,\n    \"seq_acc\": 99.70,\n    \"cer\": 0.0006,\n    \"wrong_samples\": 6\n}, \"captcha_checkpoint.pth\")",
   "metadata": {
    "trusted": true,
    "execution": {
     "iopub.status.busy": "2026-06-02T08:58:33.857847Z",
     "iopub.execute_input": "2026-06-02T08:58:33.858511Z",
     "iopub.status.idle": "2026-06-02T08:58:34.126005Z",
     "shell.execute_reply.started": "2026-06-02T08:58:33.858478Z",
     "shell.execute_reply": "2026-06-02T08:58:34.125384Z"
    }
   },
   "outputs": [],
   "execution_count": 137
  }
 ]
}