Similar_Document_Template_Matching_for_Medical_Dataset.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ydvaayan/MedMatcher/blob/main/Similar_Document_Template_Matching_for_Medical_Dataset.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchvision"
      ],
      "metadata": {
        "id": "E489pXjCPhpb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "\n",
        "# Mount Google Drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KRt5c7nhygLJ",
        "outputId": "01fce1d9-bba9-41b0-87ba-bc7a3ace488f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "import shutil\n",
        "\n",
        "def copy_images(source_folders, destination_folder):\n",
        "    # Create the destination folder if it doesn't exist\n",
        "    if not os.path.exists(destination_folder):\n",
        "        os.makedirs(destination_folder)\n",
        "\n",
        "    # Iterate over source folders\n",
        "    for source_folder in source_folders:\n",
        "        # Iterate over files in the source folder\n",
        "        for filename in os.listdir(source_folder):\n",
        "            source_path = os.path.join(source_folder, filename)\n",
        "            destination_path = os.path.join(destination_folder, filename)\n",
        "\n",
        "            # Check if the path is a file (not a directory)\n",
        "            if os.path.isfile(source_path):\n",
        "                # Copy the file to the destination folder\n",
        "                shutil.copy2(source_path, destination_path)\n",
        "            else:\n",
        "                print(f\"Skipping directory: {source_path}\")\n",
        "\n",
        "# Example usage\n",
        "source_folders = ['/content/drive/MyDrive/sihdata/handwritten/bgremoved']\n",
        "destination_folder = '/content/drive/MyDrive/sihdata/combined'\n",
        "\n",
        "copy_images(source_folders, destination_folder)"
      ],
      "metadata": {
        "id": "hBLEmjC4DxYP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "import shutil\n",
        "\n",
        "def copy_images(source_folders, destination_folder):\n",
        "    # Create the destination folder if it doesn't exist\n",
        "    if not os.path.exists(destination_folder):\n",
        "        os.makedirs(destination_folder)\n",
        "\n",
        "    # Iterate over source folders\n",
        "    for source_folder in source_folders:\n",
        "        # Iterate over files in the source folder\n",
        "        for filename in os.listdir(source_folder):\n",
        "            source_path = os.path.join(source_folder, filename)\n",
        "            destination_path = os.path.join(destination_folder, filename)\n",
        "\n",
        "            # Copy the file to the destination folder\n",
        "            shutil.copy2(source_path, destination_path)\n",
        "\n",
        "# Example usage\n",
        "source_folders = ['/content/drive/MyDrive/sihdata/invoice', '/content/drive/MyDrive/sihdata/lab report', '/content/drive/MyDrive/sihdata/prescription']\n",
        "destination_folder = '/content/drive/MyDrive/sihdata/combined'\n",
        "\n",
        "copy_images(source_folders, destination_folder)\n"
      ],
      "metadata": {
        "id": "9uEds5oxNb9F"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "\n",
        "folder_path = '/content/drive/MyDrive/sihdata/combined'\n",
        "\n",
        "# Get the list of files in the folder\n",
        "files = [f for f in os.listdir(folder_path) if os.path.isfile(os.path.join(folder_path, f))]\n",
        "\n",
        "# Count the number of files\n",
        "num_files = len(files)\n",
        "\n",
        "print(f'The number of files in the folder is: {num_files}')\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-Q6pe3D3Pq-q",
        "outputId": "0481e5e5-d468-49df-cf6d-3aa5fe999ce1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The number of files in the folder is: 228\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "main_folder_path = '/content/drive/MyDrive/sihdata'  # Replace with the actual path to your main folder\n",
        "subfolders = ['invoice', 'lab report', 'prescription']  # Replace with the actual subfolder names\n",
        "\n",
        "# Dictionary to store filenames for each subfolder\n",
        "filenames_dict = {}\n",
        "\n",
        "for subfolder in subfolders:\n",
        "    folder_path = os.path.join(main_folder_path, subfolder)\n",
        "\n",
        "    # Get a list of filenames in the subfolder\n",
        "    filenames = os.listdir(folder_path)\n",
        "\n",
        "    # Sort the list of filenames\n",
        "    sorted_filenames = sorted(filenames)\n",
        "\n",
        "    # Store the sorted filenames in the dictionary\n",
        "    filenames_dict[subfolder] = sorted_filenames\n",
        "\n",
        "# Print the filenames for each subfolder\n",
        "for subfolder, filenames in filenames_dict.items():\n",
        "    print(f\"Filenames in {subfolder}:\")\n",
        "    for filename in filenames:\n",
        "        print(filename)\n",
        "    print(\"\\n\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U8JxQVbQwUyI",
        "outputId": "7b3025ce-b260-42ad-d354-7a460b33f520"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Filenames in invoice:\n",
            ".ipynb_checkpoints\n",
            "I1-A.jpg\n",
            "I1-B.jpg\n",
            "I1-C.jpg\n",
            "I1.jpg\n",
            "I10-A.jpg\n",
            "I10-B.jpg\n",
            "I10-C.jpg\n",
            "I10.jpg\n",
            "I11-A.jpg\n",
            "I11-B.jpg\n",
            "I11-C.jpg\n",
            "I11.jpg\n",
            "I2-A.jpg\n",
            "I2-B.jpg\n",
            "I2-C.jpg\n",
            "I2.jpg\n",
            "I3-A.jpg\n",
            "I3-B.jpg\n",
            "I3-C.jpg\n",
            "I3.jpg\n",
            "I5-A.jpg\n",
            "I5-B.jpg\n",
            "I5-C.jpg\n",
            "I5.jpg\n",
            "I6-A.jpg\n",
            "I6-B.jpg\n",
            "I6-C.jpg\n",
            "I6.jpg\n",
            "I7-A.jpg\n",
            "I7-B.jpg\n",
            "I7-C.jpg\n",
            "I7.jpg\n",
            "I8-A.jpg\n",
            "I8-B.jpg\n",
            "I8-C.jpg\n",
            "I8.jpg\n",
            "I9-A.jpg\n",
            "I9-B.jpg\n",
            "I9-C.jpg\n",
            "I9.jpg\n",
            "\n",
            "\n",
            "Filenames in lab report:\n",
            "L1-A.jpg\n",
            "L1-B.jpg\n",
            "L1-C.jpg\n",
            "L1.jpg\n",
            "L10-A.jpg\n",
            "L10-B.jpg\n",
            "L10-C.jpg\n",
            "L10.jpg\n",
            "L11-A.jpg\n",
            "L11-B.jpg\n",
            "L11-C.jpg\n",
            "L11.jpg\n",
            "L12-A.jpg\n",
            "L12-B.jpg\n",
            "L12-C.jpg\n",
            "L12.jpg\n",
            "L13-A.jpg\n",
            "L13-B.jpg\n",
            "L13-C.jpg\n",
            "L13.jpg\n",
            "L14-A.jpg\n",
            "L14-B.jpg\n",
            "L14-C.jpg\n",
            "L14.jpg\n",
            "L16-A.jpg\n",
            "L16-B.jpg\n",
            "L16-C.jpg\n",
            "L16.jpg\n",
            "L17-A.jpg\n",
            "L17-B.jpg\n",
            "L17-C.jpg\n",
            "L17.jpg\n",
            "L18-A.jpg\n",
            "L18-B.jpg\n",
            "L18-C.jpg\n",
            "L18.jpg\n",
            "L19-A.jpg\n",
            "L19-B.jpg\n",
            "L19-C.jpg\n",
            "L19.jpg\n",
            "L2-A.jpg\n",
            "L2-B.jpg\n",
            "L2-C.jpg\n",
            "L2.jpg\n",
            "L3-A.jpg\n",
            "L3-B.jpg\n",
            "L3-C.jpg\n",
            "L3.jpg\n",
            "L4-A.jpg\n",
            "L4-B.jpg\n",
            "L4-C.jpg\n",
            "L4.jpg\n",
            "L5-A.jpg\n",
            "L5-B.jpg\n",
            "L5-C.jpg\n",
            "L5.jpg\n",
            "L6-A.jpg\n",
            "L6-B.jpg\n",
            "L6-C.jpg\n",
            "L6.jpg\n",
            "L7-A.jpg\n",
            "L7-B.jpg\n",
            "L7-C.jpg\n",
            "L7.jpg\n",
            "L8-A.jpg\n",
            "L8-B.jpg\n",
            "L8-C.jpg\n",
            "L8.jpg\n",
            "L9-A.jpg\n",
            "L9-B.jpg\n",
            "L9-C.jpg\n",
            "L9.jpg\n",
            "\n",
            "\n",
            "Filenames in prescription:\n",
            "P1-A.jpg\n",
            "P1-B.jpg\n",
            "P1-C.jpg\n",
            "P1.jpg\n",
            "P2-A.jpg\n",
            "P2-B.jpg\n",
            "P2-C.jpg\n",
            "P2.jpg\n",
            "P3-A.jpg\n",
            "P3-B.jpg\n",
            "P3-C.jpg\n",
            "P3.jpg\n",
            "P4-A.jpg\n",
            "P4-B.jpg\n",
            "P4-C.jpg\n",
            "P4.jpg\n",
            "P5-A.jpg\n",
            "P5-B.jpg\n",
            "P5-C.jpg\n",
            "P5.jpg\n",
            "\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "'/handwritten/bgremoved'"
      ],
      "metadata": {
        "id": "uPtoisM2E4mo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "v0mGL6u1E6Hh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "main_folder_path = '/content/drive/MyDrive/sihdata/handwritten'  # Replace with the actual path to your main folder\n",
        "subfolders = ['bgremoved']  # Replace with the actual subfolder names\n",
        "\n",
        "# Dictionary to store filenames for each subfolder\n",
        "\n",
        "for subfolder in subfolders:\n",
        "    folder_path = os.path.join(main_folder_path, subfolder)\n",
        "\n",
        "    # Get a list of filenames in the subfolder\n",
        "    filenames = os.listdir(folder_path)\n",
        "\n",
        "    # Sort the list of filenames\n",
        "    sorted_filenames = sorted(filenames)\n",
        "\n",
        "    # Store the sorted filenames in the dictionary\n",
        "    filenames_dict[subfolder] = sorted_filenames\n",
        "\n",
        "# Print the filenames for each subfolder\n",
        "for subfolder, filenames in filenames_dict.items():\n",
        "    print(f\"Filenames in {subfolder}:\")\n",
        "    for filename in filenames:\n",
        "        print(filename)\n",
        "    print(\"\\n\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "96dadff6-7b67-4ef6-98bf-5292ece2155a",
        "id": "n7lyVOXME8pl"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Filenames in invoice:\n",
            ".ipynb_checkpoints\n",
            "I1-A.jpg\n",
            "I1-B.jpg\n",
            "I1-C.jpg\n",
            "I1.jpg\n",
            "I10-A.jpg\n",
            "I10-B.jpg\n",
            "I10-C.jpg\n",
            "I10.jpg\n",
            "I11-A.jpg\n",
            "I11-B.jpg\n",
            "I11-C.jpg\n",
            "I11.jpg\n",
            "I2-A.jpg\n",
            "I2-B.jpg\n",
            "I2-C.jpg\n",
            "I2.jpg\n",
            "I3-A.jpg\n",
            "I3-B.jpg\n",
            "I3-C.jpg\n",
            "I3.jpg\n",
            "I5-A.jpg\n",
            "I5-B.jpg\n",
            "I5-C.jpg\n",
            "I5.jpg\n",
            "I6-A.jpg\n",
            "I6-B.jpg\n",
            "I6-C.jpg\n",
            "I6.jpg\n",
            "I7-A.jpg\n",
            "I7-B.jpg\n",
            "I7-C.jpg\n",
            "I7.jpg\n",
            "I8-A.jpg\n",
            "I8-B.jpg\n",
            "I8-C.jpg\n",
            "I8.jpg\n",
            "I9-A.jpg\n",
            "I9-B.jpg\n",
            "I9-C.jpg\n",
            "I9.jpg\n",
            "\n",
            "\n",
            "Filenames in lab report:\n",
            "L1-A.jpg\n",
            "L1-B.jpg\n",
            "L1-C.jpg\n",
            "L1.jpg\n",
            "L10-A.jpg\n",
            "L10-B.jpg\n",
            "L10-C.jpg\n",
            "L10.jpg\n",
            "L11-A.jpg\n",
            "L11-B.jpg\n",
            "L11-C.jpg\n",
            "L11.jpg\n",
            "L12-A.jpg\n",
            "L12-B.jpg\n",
            "L12-C.jpg\n",
            "L12.jpg\n",
            "L13-A.jpg\n",
            "L13-B.jpg\n",
            "L13-C.jpg\n",
            "L13.jpg\n",
            "L14-A.jpg\n",
            "L14-B.jpg\n",
            "L14-C.jpg\n",
            "L14.jpg\n",
            "L16-A.jpg\n",
            "L16-B.jpg\n",
            "L16-C.jpg\n",
            "L16.jpg\n",
            "L17-A.jpg\n",
            "L17-B.jpg\n",
            "L17-C.jpg\n",
            "L17.jpg\n",
            "L18-A.jpg\n",
            "L18-B.jpg\n",
            "L18-C.jpg\n",
            "L18.jpg\n",
            "L19-A.jpg\n",
            "L19-B.jpg\n",
            "L19-C.jpg\n",
            "L19.jpg\n",
            "L2-A.jpg\n",
            "L2-B.jpg\n",
            "L2-C.jpg\n",
            "L2.jpg\n",
            "L3-A.jpg\n",
            "L3-B.jpg\n",
            "L3-C.jpg\n",
            "L3.jpg\n",
            "L4-A.jpg\n",
            "L4-B.jpg\n",
            "L4-C.jpg\n",
            "L4.jpg\n",
            "L5-A.jpg\n",
            "L5-B.jpg\n",
            "L5-C.jpg\n",
            "L5.jpg\n",
            "L6-A.jpg\n",
            "L6-B.jpg\n",
            "L6-C.jpg\n",
            "L6.jpg\n",
            "L7-A.jpg\n",
            "L7-B.jpg\n",
            "L7-C.jpg\n",
            "L7.jpg\n",
            "L8-A.jpg\n",
            "L8-B.jpg\n",
            "L8-C.jpg\n",
            "L8.jpg\n",
            "L9-A.jpg\n",
            "L9-B.jpg\n",
            "L9-C.jpg\n",
            "L9.jpg\n",
            "\n",
            "\n",
            "Filenames in prescription:\n",
            "P1-A.jpg\n",
            "P1-B.jpg\n",
            "P1-C.jpg\n",
            "P1.jpg\n",
            "P2-A.jpg\n",
            "P2-B.jpg\n",
            "P2-C.jpg\n",
            "P2.jpg\n",
            "P3-A.jpg\n",
            "P3-B.jpg\n",
            "P3-C.jpg\n",
            "P3.jpg\n",
            "P4-A.jpg\n",
            "P4-B.jpg\n",
            "P4-C.jpg\n",
            "P4.jpg\n",
            "P5-A.jpg\n",
            "P5-B.jpg\n",
            "P5-C.jpg\n",
            "P5.jpg\n",
            "\n",
            "\n",
            "Filenames in bgremoved:\n",
            "H1-A.jpg\n",
            "H1-B.jpg\n",
            "H1-C.jpg\n",
            "H1.jpg\n",
            "H10-A.jpg\n",
            "H10-B.jpg\n",
            "H10-C.jpg\n",
            "H10.jpg\n",
            "H11-A.jpg\n",
            "H11-B.jpg\n",
            "H11-C.jpg\n",
            "H11.jpg\n",
            "H12-A.jpg\n",
            "H12-B.jpg\n",
            "H12-C.jpg\n",
            "H12.jpg\n",
            "H13-A.jpg\n",
            "H13-B.jpg\n",
            "H13-C.jpg\n",
            "H13.jpg\n",
            "H14-A.jpg\n",
            "H14-B.jpg\n",
            "H14-C.jpg\n",
            "H14.jpg\n",
            "H15.jpg\n",
            "H15A.jpg\n",
            "H15B.jpg\n",
            "H15C.jpg\n",
            "H16.jpg\n",
            "H16A.jpg\n",
            "H16B.jpg\n",
            "H16C.jpg\n",
            "H17.jpg\n",
            "H17A.jpg\n",
            "H17B.jpg\n",
            "H17C.jpg\n",
            "H18.jpg\n",
            "H18A.jpg\n",
            "H18B.jpg\n",
            "H18C.jpg\n",
            "H19.jpg\n",
            "H19A.jpg\n",
            "H19B.jpg\n",
            "H19C.jpg\n",
            "H2-A.jpg\n",
            "H2-B.jpg\n",
            "H2-C.jpg\n",
            "H2.jpg\n",
            "H20.jpg\n",
            "H20A.jpg\n",
            "H20B.jpg\n",
            "H20C.jpg\n",
            "H21.jpg\n",
            "H21A.jpg\n",
            "H21B.jpg\n",
            "H21C.jpg\n",
            "H22.jpg\n",
            "H22A.jpg\n",
            "H22B.jpg\n",
            "H22C.jpg\n",
            "H23.jpg\n",
            "H23A.jpg\n",
            "H23B.jpg\n",
            "H23C.jpg\n",
            "H24.jpg\n",
            "H24A.jpg\n",
            "H24B.jpg\n",
            "H24C.jpg\n",
            "H3-A.jpg\n",
            "H3-B.jpg\n",
            "H3-C.jpg\n",
            "H3.jpg\n",
            "H4-A.jpg\n",
            "H4-B.jpg\n",
            "H4-C.jpg\n",
            "H4.jpg\n",
            "H5-A.jpg\n",
            "H5-B.jpg\n",
            "H5-C.jpg\n",
            "H5.jpg\n",
            "H6-A.jpg\n",
            "H6-B.jpg\n",
            "H6-C.jpg\n",
            "H6.jpg\n",
            "H7-A.jpg\n",
            "H7-B.jpg\n",
            "H7-C.jpg\n",
            "H7.jpg\n",
            "H8-A.jpg\n",
            "H8-B.jpg\n",
            "H8-C.jpg\n",
            "H8.jpg\n",
            "H9-A.jpg\n",
            "H9-B.jpg\n",
            "H9-C.jpg\n",
            "H9.jpg\n",
            "\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "image_filenames_list = []\n",
        "for subfolder_filenames in filenames_dict.values():\n",
        "    image_filenames_list.extend(subfolder_filenames)\n",
        "for filename in image_filenames_list:\n",
        "    print(filename)                          #i4,L15"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CO9NvW8vzBIR",
        "outputId": "137329ad-5d44-421c-d3bb-5cbbcbb13ab2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ".ipynb_checkpoints\n",
            "I1-A.jpg\n",
            "I1-B.jpg\n",
            "I1-C.jpg\n",
            "I1.jpg\n",
            "I10-A.jpg\n",
            "I10-B.jpg\n",
            "I10-C.jpg\n",
            "I10.jpg\n",
            "I11-A.jpg\n",
            "I11-B.jpg\n",
            "I11-C.jpg\n",
            "I11.jpg\n",
            "I2-A.jpg\n",
            "I2-B.jpg\n",
            "I2-C.jpg\n",
            "I2.jpg\n",
            "I3-A.jpg\n",
            "I3-B.jpg\n",
            "I3-C.jpg\n",
            "I3.jpg\n",
            "I5-A.jpg\n",
            "I5-B.jpg\n",
            "I5-C.jpg\n",
            "I5.jpg\n",
            "I6-A.jpg\n",
            "I6-B.jpg\n",
            "I6-C.jpg\n",
            "I6.jpg\n",
            "I7-A.jpg\n",
            "I7-B.jpg\n",
            "I7-C.jpg\n",
            "I7.jpg\n",
            "I8-A.jpg\n",
            "I8-B.jpg\n",
            "I8-C.jpg\n",
            "I8.jpg\n",
            "I9-A.jpg\n",
            "I9-B.jpg\n",
            "I9-C.jpg\n",
            "I9.jpg\n",
            "L1-A.jpg\n",
            "L1-B.jpg\n",
            "L1-C.jpg\n",
            "L1.jpg\n",
            "L10-A.jpg\n",
            "L10-B.jpg\n",
            "L10-C.jpg\n",
            "L10.jpg\n",
            "L11-A.jpg\n",
            "L11-B.jpg\n",
            "L11-C.jpg\n",
            "L11.jpg\n",
            "L12-A.jpg\n",
            "L12-B.jpg\n",
            "L12-C.jpg\n",
            "L12.jpg\n",
            "L13-A.jpg\n",
            "L13-B.jpg\n",
            "L13-C.jpg\n",
            "L13.jpg\n",
            "L14-A.jpg\n",
            "L14-B.jpg\n",
            "L14-C.jpg\n",
            "L14.jpg\n",
            "L16-A.jpg\n",
            "L16-B.jpg\n",
            "L16-C.jpg\n",
            "L16.jpg\n",
            "L17-A.jpg\n",
            "L17-B.jpg\n",
            "L17-C.jpg\n",
            "L17.jpg\n",
            "L18-A.jpg\n",
            "L18-B.jpg\n",
            "L18-C.jpg\n",
            "L18.jpg\n",
            "L19-A.jpg\n",
            "L19-B.jpg\n",
            "L19-C.jpg\n",
            "L19.jpg\n",
            "L2-A.jpg\n",
            "L2-B.jpg\n",
            "L2-C.jpg\n",
            "L2.jpg\n",
            "L3-A.jpg\n",
            "L3-B.jpg\n",
            "L3-C.jpg\n",
            "L3.jpg\n",
            "L4-A.jpg\n",
            "L4-B.jpg\n",
            "L4-C.jpg\n",
            "L4.jpg\n",
            "L5-A.jpg\n",
            "L5-B.jpg\n",
            "L5-C.jpg\n",
            "L5.jpg\n",
            "L6-A.jpg\n",
            "L6-B.jpg\n",
            "L6-C.jpg\n",
            "L6.jpg\n",
            "L7-A.jpg\n",
            "L7-B.jpg\n",
            "L7-C.jpg\n",
            "L7.jpg\n",
            "L8-A.jpg\n",
            "L8-B.jpg\n",
            "L8-C.jpg\n",
            "L8.jpg\n",
            "L9-A.jpg\n",
            "L9-B.jpg\n",
            "L9-C.jpg\n",
            "L9.jpg\n",
            "P1-A.jpg\n",
            "P1-B.jpg\n",
            "P1-C.jpg\n",
            "P1.jpg\n",
            "P2-A.jpg\n",
            "P2-B.jpg\n",
            "P2-C.jpg\n",
            "P2.jpg\n",
            "P3-A.jpg\n",
            "P3-B.jpg\n",
            "P3-C.jpg\n",
            "P3.jpg\n",
            "P4-A.jpg\n",
            "P4-B.jpg\n",
            "P4-C.jpg\n",
            "P4.jpg\n",
            "P5-A.jpg\n",
            "P5-B.jpg\n",
            "P5-C.jpg\n",
            "P5.jpg\n",
            "H1-A.jpg\n",
            "H1-B.jpg\n",
            "H1-C.jpg\n",
            "H1.jpg\n",
            "H10-A.jpg\n",
            "H10-B.jpg\n",
            "H10-C.jpg\n",
            "H10.jpg\n",
            "H11-A.jpg\n",
            "H11-B.jpg\n",
            "H11-C.jpg\n",
            "H11.jpg\n",
            "H12-A.jpg\n",
            "H12-B.jpg\n",
            "H12-C.jpg\n",
            "H12.jpg\n",
            "H13-A.jpg\n",
            "H13-B.jpg\n",
            "H13-C.jpg\n",
            "H13.jpg\n",
            "H14-A.jpg\n",
            "H14-B.jpg\n",
            "H14-C.jpg\n",
            "H14.jpg\n",
            "H15.jpg\n",
            "H15A.jpg\n",
            "H15B.jpg\n",
            "H15C.jpg\n",
            "H16.jpg\n",
            "H16A.jpg\n",
            "H16B.jpg\n",
            "H16C.jpg\n",
            "H17.jpg\n",
            "H17A.jpg\n",
            "H17B.jpg\n",
            "H17C.jpg\n",
            "H18.jpg\n",
            "H18A.jpg\n",
            "H18B.jpg\n",
            "H18C.jpg\n",
            "H19.jpg\n",
            "H19A.jpg\n",
            "H19B.jpg\n",
            "H19C.jpg\n",
            "H2-A.jpg\n",
            "H2-B.jpg\n",
            "H2-C.jpg\n",
            "H2.jpg\n",
            "H20.jpg\n",
            "H20A.jpg\n",
            "H20B.jpg\n",
            "H20C.jpg\n",
            "H21.jpg\n",
            "H21A.jpg\n",
            "H21B.jpg\n",
            "H21C.jpg\n",
            "H22.jpg\n",
            "H22A.jpg\n",
            "H22B.jpg\n",
            "H22C.jpg\n",
            "H23.jpg\n",
            "H23A.jpg\n",
            "H23B.jpg\n",
            "H23C.jpg\n",
            "H24.jpg\n",
            "H24A.jpg\n",
            "H24B.jpg\n",
            "H24C.jpg\n",
            "H3-A.jpg\n",
            "H3-B.jpg\n",
            "H3-C.jpg\n",
            "H3.jpg\n",
            "H4-A.jpg\n",
            "H4-B.jpg\n",
            "H4-C.jpg\n",
            "H4.jpg\n",
            "H5-A.jpg\n",
            "H5-B.jpg\n",
            "H5-C.jpg\n",
            "H5.jpg\n",
            "H6-A.jpg\n",
            "H6-B.jpg\n",
            "H6-C.jpg\n",
            "H6.jpg\n",
            "H7-A.jpg\n",
            "H7-B.jpg\n",
            "H7-C.jpg\n",
            "H7.jpg\n",
            "H8-A.jpg\n",
            "H8-B.jpg\n",
            "H8-C.jpg\n",
            "H8.jpg\n",
            "H9-A.jpg\n",
            "H9-B.jpg\n",
            "H9-C.jpg\n",
            "H9.jpg\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(len(image_filenames_list))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3XuSHtsR_5vn",
        "outputId": "ed0e2806-dfa7-4e4c-fb03-94207e785e7a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "228\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "image_filenames_list=image_filenames_list[1:229]"
      ],
      "metadata": {
        "id": "esB-X3lmFRh_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#make y matrix... and write code to extract small y given some images names\n",
        "#make y 132*132 filled with 0s initially\n",
        "#make a mapping from index to doc name\n",
        "#then just run through dataset and turn that subsection 1"
      ],
      "metadata": {
        "id": "2BZGu5SGo9TH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# Size of the tensor\n",
        "tensor_size = (228, 228)\n",
        "\n",
        "# Create a tensor with all values initialized to 0\n",
        "tensor = torch.zeros(tensor_size, dtype=torch.float32)\n",
        "# Loop to set 4x4 subgrids to 2\n",
        "for i in range(0,33):\n",
        "    for j in range(4*i,4*i+4):\n",
        "      for k in range(4*i,4*i+4):\n",
        "        tensor[j][k]=1\n",
        "\n",
        "# Print the updated tensor\n",
        "print(tensor)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "y6P3hTI-qJoR",
        "outputId": "75914baf-1d61-4298-acfe-173c052b1f92"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([[1., 1., 1.,  ..., 0., 0., 0.],\n",
            "        [1., 1., 1.,  ..., 0., 0., 0.],\n",
            "        [1., 1., 1.,  ..., 0., 0., 0.],\n",
            "        ...,\n",
            "        [0., 0., 0.,  ..., 0., 0., 0.],\n",
            "        [0., 0., 0.,  ..., 0., 0., 0.],\n",
            "        [0., 0., 0.,  ..., 0., 0., 0.]])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "from PIL import Image\n",
        "from torch.utils.data import Dataset, DataLoader\n",
        "\n",
        "class CustomDataset(Dataset):\n",
        "    def __init__(self, folder_path, file_list):\n",
        "        self.folder_path = folder_path\n",
        "        self.file_list = file_list\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.file_list)\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        filename = self.file_list[idx]\n",
        "        img_path = os.path.join(self.folder_path, filename)\n",
        "\n",
        "        # Load image using PIL (modify based on your needs)\n",
        "        image = Image.open(img_path).convert('RGB')\n",
        "\n",
        "        return {'image': image, 'filename': filename}\n",
        "\n",
        "def custom_collate(batch):\n",
        "    images = [item['image'] for item in batch]\n",
        "    filenames = [item['filename'] for item in batch]\n",
        "\n",
        "    return {'images': images, 'filenames': filenames}\n",
        "\n",
        "# Example usage\n",
        "folder_path = '/content/drive/MyDrive/sihdata/combined'\n",
        "# file_list = os.listdir(folder_path)[:4]  # Take the first 4 files\n",
        "\n",
        "\n",
        "\n",
        "# Example usage\n",
        "# for batch in custom_dataloader:\n",
        "#     tensor_size = (4, 4)\n",
        "#     small_y = torch.full(tensor_size, fill_value=-1, dtype=torch.float32)\n",
        "#     images = batch['images']\n",
        "#     filenames = batch['filenames']\n",
        "#     for x in range(4):\n",
        "#      x1 = image_filenames_list.index(filenames[x])\n",
        "#      for y in range(4):\n",
        "#       y1 = image_filenames_list.index(filenames[y])\n",
        "#       small_y[x][y]=tensor[x1][y1]\n",
        "#     print(filenames,images,small_y)\n",
        "#     # Your training/validation loop here"
      ],
      "metadata": {
        "id": "UXbiC8XTQaEZ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "! apt install tesseract-ocr\n",
        "! apt install libtesseract-dev"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "84-EEJc8_h4R",
        "outputId": "5b4f55ee-b5a0-4a63-f9e3-7d4a417cf5c2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree... Done\n",
            "Reading state information... Done\n",
            "The following additional packages will be installed:\n",
            "  tesseract-ocr-eng tesseract-ocr-osd\n",
            "The following NEW packages will be installed:\n",
            "  tesseract-ocr tesseract-ocr-eng tesseract-ocr-osd\n",
            "0 upgraded, 3 newly installed, 0 to remove and 24 not upgraded.\n",
            "Need to get 4,816 kB of archives.\n",
            "After this operation, 15.6 MB of additional disk space will be used.\n",
            "Get:1 http://archive.ubuntu.com/ubuntu jammy/universe amd64 tesseract-ocr-eng all 1:4.00~git30-7274cfa-1.1 [1,591 kB]\n",
            "Get:2 http://archive.ubuntu.com/ubuntu jammy/universe amd64 tesseract-ocr-osd all 1:4.00~git30-7274cfa-1.1 [2,990 kB]\n",
            "Get:3 http://archive.ubuntu.com/ubuntu jammy/universe amd64 tesseract-ocr amd64 4.1.1-2.1build1 [236 kB]\n",
            "Fetched 4,816 kB in 3s (1,832 kB/s)\n",
            "Selecting previously unselected package tesseract-ocr-eng.\n",
            "(Reading database ... 121658 files and directories currently installed.)\n",
            "Preparing to unpack .../tesseract-ocr-eng_1%3a4.00~git30-7274cfa-1.1_all.deb ...\n",
            "Unpacking tesseract-ocr-eng (1:4.00~git30-7274cfa-1.1) ...\n",
            "Selecting previously unselected package tesseract-ocr-osd.\n",
            "Preparing to unpack .../tesseract-ocr-osd_1%3a4.00~git30-7274cfa-1.1_all.deb ...\n",
            "Unpacking tesseract-ocr-osd (1:4.00~git30-7274cfa-1.1) ...\n",
            "Selecting previously unselected package tesseract-ocr.\n",
            "Preparing to unpack .../tesseract-ocr_4.1.1-2.1build1_amd64.deb ...\n",
            "Unpacking tesseract-ocr (4.1.1-2.1build1) ...\n",
            "Setting up tesseract-ocr-eng (1:4.00~git30-7274cfa-1.1) ...\n",
            "Setting up tesseract-ocr-osd (1:4.00~git30-7274cfa-1.1) ...\n",
            "Setting up tesseract-ocr (4.1.1-2.1build1) ...\n",
            "Processing triggers for man-db (2.10.2-1) ...\n",
            "Reading package lists... Done\n",
            "Building dependency tree... Done\n",
            "Reading state information... Done\n",
            "The following additional packages will be installed:\n",
            "  libarchive-dev libleptonica-dev\n",
            "The following NEW packages will be installed:\n",
            "  libarchive-dev libleptonica-dev libtesseract-dev\n",
            "0 upgraded, 3 newly installed, 0 to remove and 24 not upgraded.\n",
            "Need to get 3,743 kB of archives.\n",
            "After this operation, 16.0 MB of additional disk space will be used.\n",
            "Get:1 http://archive.ubuntu.com/ubuntu jammy/main amd64 libarchive-dev amd64 3.6.0-1ubuntu1 [581 kB]\n",
            "Get:2 http://archive.ubuntu.com/ubuntu jammy/universe amd64 libleptonica-dev amd64 1.82.0-3build1 [1,562 kB]\n",
            "Get:3 http://archive.ubuntu.com/ubuntu jammy/universe amd64 libtesseract-dev amd64 4.1.1-2.1build1 [1,600 kB]\n",
            "Fetched 3,743 kB in 2s (1,740 kB/s)\n",
            "Selecting previously unselected package libarchive-dev:amd64.\n",
            "(Reading database ... 121705 files and directories currently installed.)\n",
            "Preparing to unpack .../libarchive-dev_3.6.0-1ubuntu1_amd64.deb ...\n",
            "Unpacking libarchive-dev:amd64 (3.6.0-1ubuntu1) ...\n",
            "Selecting previously unselected package libleptonica-dev.\n",
            "Preparing to unpack .../libleptonica-dev_1.82.0-3build1_amd64.deb ...\n",
            "Unpacking libleptonica-dev (1.82.0-3build1) ...\n",
            "Selecting previously unselected package libtesseract-dev:amd64.\n",
            "Preparing to unpack .../libtesseract-dev_4.1.1-2.1build1_amd64.deb ...\n",
            "Unpacking libtesseract-dev:amd64 (4.1.1-2.1build1) ...\n",
            "Setting up libleptonica-dev (1.82.0-3build1) ...\n",
            "Setting up libarchive-dev:amd64 (3.6.0-1ubuntu1) ...\n",
            "Setting up libtesseract-dev:amd64 (4.1.1-2.1build1) ...\n",
            "Processing triggers for man-db (2.10.2-1) ...\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "! pip install transformers"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w6kYXL5P2z2b",
        "outputId": "c1be3865-7def-402f-e76b-36f0f9341df3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: transformers in /usr/local/lib/python3.10/dist-packages (4.35.2)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from transformers) (3.13.1)\n",
            "Requirement already satisfied: huggingface-hub<1.0,>=0.16.4 in /usr/local/lib/python3.10/dist-packages (from transformers) (0.19.4)\n",
            "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.10/dist-packages (from transformers) (1.23.5)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from transformers) (23.2)\n",
            "Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from transformers) (6.0.1)\n",
            "Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from transformers) (2023.6.3)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from transformers) (2.31.0)\n",
            "Requirement already satisfied: tokenizers<0.19,>=0.14 in /usr/local/lib/python3.10/dist-packages (from transformers) (0.15.0)\n",
            "Requirement already satisfied: safetensors>=0.3.1 in /usr/local/lib/python3.10/dist-packages (from transformers) (0.4.1)\n",
            "Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.10/dist-packages (from transformers) (4.66.1)\n",
            "Requirement already satisfied: fsspec>=2023.5.0 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.16.4->transformers) (2023.6.0)\n",
            "Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.16.4->transformers) (4.5.0)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (2023.11.17)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "! pip install Pillow\n",
        "! pip install pytesseract"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dc08kvON_xrU",
        "outputId": "26cc0dd4-fbae-443e-f6ee-23e611cac042"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: Pillow in /usr/local/lib/python3.10/dist-packages (9.4.0)\n",
            "Collecting pytesseract\n",
            "  Downloading pytesseract-0.3.10-py3-none-any.whl (14 kB)\n",
            "Requirement already satisfied: packaging>=21.3 in /usr/local/lib/python3.10/dist-packages (from pytesseract) (23.2)\n",
            "Requirement already satisfied: Pillow>=8.0.0 in /usr/local/lib/python3.10/dist-packages (from pytesseract) (9.4.0)\n",
            "Installing collected packages: pytesseract\n",
            "Successfully installed pytesseract-0.3.10\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchvision\n",
        "import torch.nn as nn\n",
        "from PIL import Image\n",
        "from transformers import LayoutLMv3Model, LayoutLMv3Processor, LayoutLMv3Config,LayoutLMv3Tokenizer,  LayoutLMv3ImageProcessor\n",
        "import torch.nn.functional as F\n",
        "from tqdm import tqdm\n",
        "from torchvision import transforms"
      ],
      "metadata": {
        "id": "U74nteHuVxkk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_name = \"microsoft/layoutlmv3-base\"\n",
        "original_config=LayoutLMv3Config.from_pretrained(model_name)\n",
        "\n",
        "# Load the original configuration\n",
        "original_config = LayoutLMv3Config.from_pretrained(model_name)\n",
        "\n",
        "# # Modify the configuration\n",
        "# original_config.text_embed = False\n",
        "\n",
        "# Load the model with the modified configuration\n",
        "model = LayoutLMv3Model(config=original_config)\n",
        "\n",
        "# Load the tokenizer associated with the model\n",
        "tokenizer = LayoutLMv3Tokenizer.from_pretrained(model_name)\n",
        "\n",
        "# Create an image processor with OCR disabled\n",
        "image_processor = LayoutLMv3ImageProcessor(apply_ocr=True)\n",
        "\n",
        "# Create the processor by combining the image processor and tokenizer\n",
        "processor = LayoutLMv3Processor(image_processor=image_processor, tokenizer=tokenizer)"
      ],
      "metadata": {
        "id": "IyVNiNNHHNqW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def calc_embeddings(images):\n",
        "    device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
        "    inputs = processor(images=images, return_tensors=\"pt\", padding=True, truncation=True)\n",
        "    inputs.to(device)\n",
        "    outputs = model(**inputs)\n",
        "    embeddings = outputs.last_hidden_state\n",
        "\n",
        "    return embeddings\n",
        "\n"
      ],
      "metadata": {
        "id": "KhqftWIq1Wqj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class Loss(nn.Module):\n",
        "  def __init__(self):\n",
        "    super(Loss,self).__init__()\n",
        "  def forward(self,y_pred,y):\n",
        "    eps=1e-6\n",
        "    #normalising vectors in y_pred\n",
        "    for i in range(len(y_pred)):\n",
        "      row = y_pred[i]\n",
        "      y_pred[i] = row / (float(torch.sqrt(torch.matmul(row, row.T)))+eps)\n",
        "\n",
        "    m=y.shape[0]\n",
        "    wt=torch.sum(y,axis=-1)/m\n",
        "    temp=torch.matmul(y_pred,y_pred.T)\n",
        "\n",
        "    temp=(temp+1)/2\n",
        "\n",
        "    #Weighted Binary Cross Entropy Loss\n",
        "    bin_loss=-(1-wt)*y*torch.log(temp)-wt*(1-y)*torch.log(1-temp+eps)\n",
        "\n",
        "    cost=torch.sum(bin_loss)/m\n",
        "\n",
        "    #Focal Loss\n",
        "    # foc_loss=torchvision.ops.sigmoid_focal_loss(temp, y)\n",
        "    # cost=torch.sum(foc_loss)/m\n",
        "\n",
        "    return cost"
      ],
      "metadata": {
        "id": "Edmivve5kz1x"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Training"
      ],
      "metadata": {
        "id": "vBcdVKokKtyL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
        "\n",
        "model_name = \"microsoft/layoutlmv3-base\"\n",
        "model = LayoutLMv3Model.from_pretrained(model_name)\n",
        "processor = LayoutLMv3Processor.from_pretrained(model_name)\n",
        "model.to(device)"
      ],
      "metadata": {
        "id": "C9388mCRMW4c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2369f3c1-d6eb-4afb-a3c9-f96006859618"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LayoutLMv3Model(\n",
              "  (embeddings): LayoutLMv3TextEmbeddings(\n",
              "    (word_embeddings): Embedding(50265, 768, padding_idx=1)\n",
              "    (token_type_embeddings): Embedding(1, 768)\n",
              "    (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "    (dropout): Dropout(p=0.1, inplace=False)\n",
              "    (position_embeddings): Embedding(514, 768, padding_idx=1)\n",
              "    (x_position_embeddings): Embedding(1024, 128)\n",
              "    (y_position_embeddings): Embedding(1024, 128)\n",
              "    (h_position_embeddings): Embedding(1024, 128)\n",
              "    (w_position_embeddings): Embedding(1024, 128)\n",
              "  )\n",
              "  (patch_embed): LayoutLMv3PatchEmbeddings(\n",
              "    (proj): Conv2d(3, 768, kernel_size=(16, 16), stride=(16, 16))\n",
              "  )\n",
              "  (pos_drop): Dropout(p=0.0, inplace=False)\n",
              "  (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "  (dropout): Dropout(p=0.1, inplace=False)\n",
              "  (norm): LayerNorm((768,), eps=1e-06, elementwise_affine=True)\n",
              "  (encoder): LayoutLMv3Encoder(\n",
              "    (layer): ModuleList(\n",
              "      (0-11): 12 x LayoutLMv3Layer(\n",
              "        (attention): LayoutLMv3Attention(\n",
              "          (self): LayoutLMv3SelfAttention(\n",
              "            (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "          (output): LayoutLMv3SelfOutput(\n",
              "            (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (intermediate): LayoutLMv3Intermediate(\n",
              "          (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (intermediate_act_fn): GELUActivation()\n",
              "        )\n",
              "        (output): LayoutLMv3Output(\n",
              "          (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (dropout): Dropout(p=0.1, inplace=False)\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (rel_pos_bias): Linear(in_features=32, out_features=12, bias=False)\n",
              "    (rel_pos_x_bias): Linear(in_features=64, out_features=12, bias=False)\n",
              "    (rel_pos_y_bias): Linear(in_features=64, out_features=12, bias=False)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.load_state_dict(torch.load('/content/drive/MyDrive/sih/model_save/checkpoint11.pth'))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A8LZ_jCiAvw-",
        "outputId": "cfefed65-ba43-4558-a96e-1c0ea1536f6b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_losses =[]\n",
        "batchwise=[]\n"
      ],
      "metadata": {
        "id": "gEvNwRP4Dfk5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from transformers import AdamW\n",
        "\n",
        "import os\n",
        "os.environ['CUDA_LAUNCH_BLOCKING'] = '1'\n",
        "transform = transforms.ToTensor()\n",
        "lr = 1e-5\n",
        "num_epoch=1\n",
        "\n",
        "optimizer = AdamW(model.parameters(),lr)\n",
        "\n",
        "loss_fn=Loss()\n",
        "device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
        "\n",
        "\n",
        "\n",
        "model.train()\n",
        "\n",
        "\n",
        "for epoch in tqdm(range(11,12)):\n",
        "  #model.load_state_dict(torch.load(f'/content/drive/MyDrive/sih/model_save/checkpoint{epoch}.pth'))\n",
        "\n",
        "  losses=[]\n",
        "  i=0\n",
        "  while i<=224:\n",
        "\n",
        "    file_list = os.listdir(folder_path)[i:i+4]\n",
        "    # Create a custom dataset\n",
        "    custom_dataset = CustomDataset(folder_path, file_list)\n",
        "\n",
        "    # Create a custom DataLoader with the custom collate function\n",
        "    custom_dataloader = DataLoader(\n",
        "        custom_dataset,\n",
        "        batch_size=4,\n",
        "        shuffle=True,\n",
        "        num_workers=4,  # Adjust the number of workers based on your system\n",
        "        collate_fn=custom_collate,  # Use the custom collate function\n",
        "    )\n",
        "    for batch in custom_dataloader:\n",
        "\n",
        "      optimizer.zero_grad()\n",
        "      tensor_size = (4, 4)\n",
        "      small_y = torch.full(tensor_size, fill_value=-1, dtype=torch.float32)\n",
        "      images = batch['images']\n",
        "\n",
        "      filenames = batch['filenames']\n",
        "      print(filenames)\n",
        "      for x in range(4):\n",
        "        x1 = image_filenames_list.index(filenames[x])\n",
        "        for y in range(4):\n",
        "          y1 = image_filenames_list.index(filenames[y])\n",
        "          small_y[x][y]=tensor[x1][y1]\n",
        "      # print(filenames,images,small_y)\n",
        "      images= [transform(image).to(device) for image in images]\n",
        "\n",
        "      embeddings=calc_embeddings(images)\n",
        "      embeddings = torch.reshape(embeddings, (embeddings.shape[0], -1))\n",
        "      small_y = small_y.to(device)\n",
        "      embeddings = embeddings.to(device)\n",
        "      loss = loss_fn(embeddings,small_y)\n",
        "\n",
        "      losses.append(loss.item())\n",
        "      loss.backward()\n",
        "      optimizer.step()\n",
        "\n",
        "    i=i+4\n",
        "\n",
        "  train_losses.append(losses[-1])\n",
        "  batchwise.append(losses)\n",
        "\n",
        "  #checkpointing and saving\n",
        "  torch.save(model.state_dict(), f'/content/drive/MyDrive/sih/model_save/checkpoint{epoch+1}.pth')\n",
        "  import pickle\n",
        "\n",
        "  with open(f'/content/drive/MyDrive/sih/model_save/training_losses{epoch+1}.pkl', 'wb') as file:\n",
        "    pickle.dump(train_losses, file)\n",
        "\n",
        "  with open(f'/content/drive/MyDrive/sih/model_save/training_losses_batchwise{epoch+1}.pkl', 'wb') as file:\n",
        "      pickle.dump(batchwise, file)\n"
      ],
      "metadata": {
        "id": "R-BX_5dTadrG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0b123c5d-3d77-4091-bab6-c195f1f9b764"
      },
      "execution_count": null,
      "outputs": [
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/optimization.py:411: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
            "  warnings.warn(\n",
            "  0%|          | 0/1 [00:00<?, ?it/s]/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['P2.jpg', 'P4-A.jpg', 'P4.jpg', 'P4-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L3-C.jpg', 'L19-B.jpg', 'P4-C.jpg', 'L19-C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['P2-C.jpg', 'L19-A.jpg', 'P1-A.jpg', 'P2-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['P5.jpg', 'P5-A.jpg', 'L1-B.jpg', 'L17-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['P5-C.jpg', 'P3-B.jpg', 'P1-C.jpg', 'P3-C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L17-C.jpg', 'L1-A.jpg', 'P3-A.jpg', 'P1.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['P3.jpg', 'P1-B.jpg', 'P5-B.jpg', 'P2-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L5.jpg', 'L2-B.jpg', 'L2-A.jpg', 'L18-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L7.jpg', 'L16.jpg', 'L19.jpg', 'L3-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L18-C.jpg', 'L4-C.jpg', 'L18.jpg', 'L3.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L5-B.jpg', 'L16-A.jpg', 'L16-B.jpg', 'L1.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L17.jpg', 'L1-C.jpg', 'L18-B.jpg', 'L17-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L4-A.jpg', 'L16-C.jpg', 'L8-B.jpg', 'L4.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L14-B.jpg', 'L10-C.jpg', 'L6.jpg', 'L11.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L8.jpg', 'L13-B.jpg', 'L8-C.jpg', 'L12-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L3-A.jpg', 'L5-C.jpg', 'L13.jpg', 'L13-C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L14-A.jpg', 'L10-A.jpg', 'L2-C.jpg', 'L4-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L14-C.jpg', 'L12-A.jpg', 'L9.jpg', 'L6-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L6-C.jpg', 'L9-C.jpg', 'L14.jpg', 'L10.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L11-A.jpg', 'L11-C.jpg', 'I9.jpg', 'L8-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I7.jpg', 'L7-B.jpg', 'L10-B.jpg', 'L7-C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L5-A.jpg', 'L2.jpg', 'L13-A.jpg', 'L12.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L12-C.jpg', 'L9-A.jpg', 'L11-B.jpg', 'L6-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['L9-B.jpg', 'I7-A.jpg', 'I10-C.jpg', 'L7-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I7-C.jpg', 'I11.jpg', 'I6-C.jpg', 'I3-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I8-A.jpg', 'I11-C.jpg', 'I1-A.jpg', 'I6-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I10.jpg', 'I9-A.jpg', 'I5-C.jpg', 'I5.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I11-B.jpg', 'I10-B.jpg', 'I9-C.jpg', 'I2-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I11-A.jpg', 'I1-B.jpg', 'I6.jpg', 'I8-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I8.jpg', 'I7-B.jpg', 'I8-C.jpg', 'I5-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I2.jpg', 'I2-B.jpg', 'I9-B.jpg', 'I1-C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I3-B.jpg', 'I2-C.jpg', 'I3-C.jpg', 'I3.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['I6-B.jpg', 'I5-A.jpg', 'I1.jpg', 'I10-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H18.jpg', 'H16.jpg', 'H24.jpg', 'H20.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H23B.jpg', 'H19.jpg', 'H23.jpg', 'H21.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H22.jpg', 'H1-B.jpg', 'H1-A.jpg', 'H17.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H2-B.jpg', 'H2-C.jpg', 'H15.jpg', 'H2-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H15C.jpg', 'H3-A.jpg', 'H15B.jpg', 'H15A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H1-C.jpg', 'H16A.jpg', 'H3-C.jpg', 'H3-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H4-A.jpg', 'H16C.jpg', 'H16B.jpg', 'H4-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H5-A.jpg', 'H5-B.jpg', 'H5-C.jpg', 'H4-C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H6-C.jpg', 'H7-A.jpg', 'H6-B.jpg', 'H6-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H8-A.jpg', 'H7-B.jpg', 'H7-C.jpg', 'H8-B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H9-B.jpg', 'H17A.jpg', 'H8-C.jpg', 'H9-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H10-A.jpg', 'H9-C.jpg', 'H17B.jpg', 'H17C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H10-C.jpg', 'H10-B.jpg', 'H11-B.jpg', 'H11-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H11-C.jpg', 'H12-A.jpg', 'H18B.jpg', 'H18A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H12-B.jpg', 'H12-C.jpg', 'H13-A.jpg', 'H18C.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H13-B.jpg', 'H19A.jpg', 'H13-C.jpg', 'H14-A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H2.jpg', 'H14-B.jpg', 'H14-C.jpg', 'H1.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H6.jpg', 'H4.jpg', 'H5.jpg', 'H3.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H8.jpg', 'H9.jpg', 'H10.jpg', 'H7.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H14.jpg', 'H12.jpg', 'H13.jpg', 'H11.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H19C.jpg', 'H20A.jpg', 'H20B.jpg', 'H19B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H21C.jpg', 'H21B.jpg', 'H20C.jpg', 'H21A.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H24A.jpg', 'H23A.jpg', 'H23C.jpg', 'H24B.jpg']\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['H22C.jpg', 'H22A.jpg', 'H24C.jpg', 'H22B.jpg']\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n",
            "100%|██████████| 1/1 [17:16<00:00, 1036.70s/it]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_losses)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "67YFmAPoC5gN",
        "outputId": "56d7d6b9-5ee4-42ff-f40e-848d6b0eaf1b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0.9738273024559021, 0.9024503827095032, 0.8889800310134888, 0.870099663734436, 0.8839300274848938, 0.8912447690963745, 0.9033598899841309]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(losses)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_r7YuXi-uMLY",
        "outputId": "5855b780-e2d0-437f-d67e-7ac1ba8f7ca8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[1.952500581741333, 1.5884238481521606, 1.5923353433609009, 1.495632290840149, 1.0935235023498535, 1.1820510625839233, 0.9133927226066589, 0.932449221611023, 0.9348500370979309, 1.1637327671051025, 0.8863048553466797, 1.0253350734710693, 1.1297250986099243, 0.7942718267440796, 0.7566736936569214, 0.7655471563339233, 0.6903344988822937, 0.7461451292037964, 0.8879510760307312, 0.744518518447876, 0.909834623336792, 0.9726134538650513, 0.7695451378822327, 0.7142298221588135, 0.7035430073738098, 0.6933834552764893, 0.7544451951980591, 0.6802178621292114, 0.865232527256012, 0.7168481349945068, 0.8322234153747559, 0.9628356695175171, 0.9738273024559021]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install pytesseract"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SQuWCepfN6xT",
        "outputId": "51816336-4a65-4d86-c2fe-e58ff9faece0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: pytesseract in /usr/local/lib/python3.10/dist-packages (0.3.10)\n",
            "Requirement already satisfied: packaging>=21.3 in /usr/local/lib/python3.10/dist-packages (from pytesseract) (23.2)\n",
            "Requirement already satisfied: Pillow>=8.0.0 in /usr/local/lib/python3.10/dist-packages (from pytesseract) (9.4.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pickle\n",
        "\n",
        "# Example list of training losses over epochs\n",
        "\n",
        "\n",
        "# Save the list to a file using pickle\n",
        "with open('/content/drive/MyDrive/sih/model_save/training_losses_batchwise4.pkl', 'wb') as file:\n",
        "    pickle.dump(batchwise, file)\n",
        "\n",
        "# # Now, you can load the list back from the file if needed\n",
        "# with open('training_losses.pkl', 'rb') as file:\n",
        "#     loaded_training_losses = pickle.load(file)\n",
        "\n",
        "# # Print the loaded list\n",
        "# print(loaded_training_losses)"
      ],
      "metadata": {
        "id": "9rAksMZAOA8B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pickle\n",
        "with open('/content/drive/MyDrive/sih/model_save/training_losses9.pkl', 'rb') as file:\n",
        "    train_losses = pickle.load(file)"
      ],
      "metadata": {
        "id": "d4OJSbGI9nBy"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Plot the training loss over epochs\n",
        "plt.plot(range(1, 9 + 1), train_losses, label='Training Loss')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.title('Training Loss Over epochs')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "uXasHi4wElgY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "outputId": "84b0e52b-fc37-49c5-b8ad-a0d4f19011ec"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Plot the training loss over batches\n",
        "plt.plot(range(1, 33 + 1), losses, label='Training Loss')\n",
        "plt.xlabel('Mini-batch')\n",
        "plt.ylabel('Loss')\n",
        "plt.title('Training Loss Over batches')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "yAUs_iA4TaTz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(batchwise)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qMNNWA4Id3ci",
        "outputId": "1fc8b95c-3dc3-4f3c-87b5-b7a202d66f3e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[1.952500581741333, 1.5884238481521606, 1.5923353433609009, 1.495632290840149, 1.0935235023498535, 1.1820510625839233, 0.9133927226066589, 0.932449221611023, 0.9348500370979309, 1.1637327671051025, 0.8863048553466797, 1.0253350734710693, 1.1297250986099243, 0.7942718267440796, 0.7566736936569214, 0.7655471563339233, 0.6903344988822937, 0.7461451292037964, 0.8879510760307312, 0.744518518447876, 0.909834623336792, 0.9726134538650513, 0.7695451378822327, 0.7142298221588135, 0.7035430073738098, 0.6933834552764893, 0.7544451951980591, 0.6802178621292114, 0.865232527256012, 0.7168481349945068, 0.8322234153747559, 0.9628356695175171, 0.9738273024559021]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "torch.save(model.state_dict(), '/content/drive/MyDrive/sih/model_save/checkpoint2.pth')"
      ],
      "metadata": {
        "id": "grXXuMkIGGzI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Eval"
      ],
      "metadata": {
        "id": "aM9Lt4uLU1WT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# model.to(device)\n",
        "model.eval()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gqzqQ3GXUvsV",
        "outputId": "05167243-bfef-4024-efa3-55ba0b6c4c85"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LayoutLMv3Model(\n",
              "  (embeddings): LayoutLMv3TextEmbeddings(\n",
              "    (word_embeddings): Embedding(50265, 768, padding_idx=1)\n",
              "    (token_type_embeddings): Embedding(1, 768)\n",
              "    (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "    (dropout): Dropout(p=0.1, inplace=False)\n",
              "    (position_embeddings): Embedding(514, 768, padding_idx=1)\n",
              "    (x_position_embeddings): Embedding(1024, 128)\n",
              "    (y_position_embeddings): Embedding(1024, 128)\n",
              "    (h_position_embeddings): Embedding(1024, 128)\n",
              "    (w_position_embeddings): Embedding(1024, 128)\n",
              "  )\n",
              "  (patch_embed): LayoutLMv3PatchEmbeddings(\n",
              "    (proj): Conv2d(3, 768, kernel_size=(16, 16), stride=(16, 16))\n",
              "  )\n",
              "  (pos_drop): Dropout(p=0.0, inplace=False)\n",
              "  (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "  (dropout): Dropout(p=0.1, inplace=False)\n",
              "  (norm): LayerNorm((768,), eps=1e-06, elementwise_affine=True)\n",
              "  (encoder): LayoutLMv3Encoder(\n",
              "    (layer): ModuleList(\n",
              "      (0-11): 12 x LayoutLMv3Layer(\n",
              "        (attention): LayoutLMv3Attention(\n",
              "          (self): LayoutLMv3SelfAttention(\n",
              "            (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "          (output): LayoutLMv3SelfOutput(\n",
              "            (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (intermediate): LayoutLMv3Intermediate(\n",
              "          (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (intermediate_act_fn): GELUActivation()\n",
              "        )\n",
              "        (output): LayoutLMv3Output(\n",
              "          (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (dropout): Dropout(p=0.1, inplace=False)\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (rel_pos_bias): Linear(in_features=32, out_features=12, bias=False)\n",
              "    (rel_pos_x_bias): Linear(in_features=64, out_features=12, bias=False)\n",
              "    (rel_pos_y_bias): Linear(in_features=64, out_features=12, bias=False)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.load_state_dict(torch.load('/content/drive/MyDrive/sih/model_save/checkpoint9.pth', map_location='cpu'), strict=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rBW2OAWKOXib",
        "outputId": "97589b20-2abe-4e16-ade9-e47014809e08"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
        "transform = transforms.ToTensor()\n",
        "i=5\n",
        "file_list = os.listdir(folder_path)[i:i+4]\n",
        "# Create a custom dataset\n",
        "custom_dataset = CustomDataset(folder_path, file_list)\n",
        "loss_fn=Loss()\n",
        "# Create a custom DataLoader with the custom collate function\n",
        "custom_dataloader = DataLoader(\n",
        "    custom_dataset,\n",
        "    batch_size=4,\n",
        "    shuffle=True,\n",
        "    num_workers=4,  # Adjust the number of workers based on your system\n",
        "    collate_fn=custom_collate,  # Use the custom collate function\n",
        ")\n",
        "for batch in custom_dataloader:\n",
        "\n",
        "  tensor_size = (4, 4)\n",
        "  small_y = torch.full(tensor_size, fill_value=-1, dtype=torch.float32)\n",
        "  images = batch['images']\n",
        "\n",
        "  filenames = batch['filenames']\n",
        "  print(filenames)\n",
        "  for x in range(4):\n",
        "    x1 = image_filenames_list.index(filenames[x])\n",
        "    for y in range(4):\n",
        "      y1 = image_filenames_list.index(filenames[y])\n",
        "      small_y[x][y]=tensor[x1][y1]\n",
        "  # print(filenames,images,small_y)\n",
        "  images= [transform(image).to(device) for image in images]\n",
        "\n",
        "  embeddings=calc_embeddings(images)\n",
        "\n",
        "  small_y = small_y.to(device)\n",
        "  embeddings = embeddings.to(device)\n",
        "  loss = loss_fn(embeddings,small_y)\n",
        "\n",
        "  print(loss)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yDYVBeBWW_ex",
        "outputId": "f558f11d-1454-4078-f2a7-75f1397ee868"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['P3-C.jpg', 'P1-B.jpg', 'P3-A.jpg', 'P2.jpg']\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(0.8407, grad_fn=<DivBackward0>)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-16-5bb899888d0a>:9: UserWarning: The use of `x.T` on tensors of dimension other than 2 to reverse their shape is deprecated and it will throw an error in a future release. Consider `x.mT` to transpose batches of matrices or `x.permute(*torch.arange(x.ndim - 1, -1, -1))` to reverse the dimensions of a tensor. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:3614.)\n",
            "  y_pred[i] = row / (float(torch.sqrt(torch.matmul(row, row.T)))+eps)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(embeddings)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cr1VgcM1Fdsi",
        "outputId": "ce3592ad-46d3-4b87-ec56-577ca51fc62c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([[ 1.6212e-04,  6.3219e-04,  7.1054e-05,  ..., -3.3571e-05,\n",
            "          4.4848e-04,  3.9381e-04],\n",
            "        [ 1.5071e-04,  6.7878e-04,  4.0242e-05,  ..., -7.2811e-05,\n",
            "          4.5676e-04,  4.2408e-04],\n",
            "        [ 1.3702e-04,  6.3112e-04,  6.2276e-05,  ..., -6.0220e-05,\n",
            "          4.7301e-04,  3.8539e-04],\n",
            "        [ 1.1033e-04,  7.2325e-04,  7.2200e-05,  ..., -4.5409e-06,\n",
            "          4.9144e-04,  4.4455e-04]], grad_fn=<AsStridedBackward0>)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import h5py\n",
        "import numpy as np\n",
        "\n",
        "# Assuming your embeddings are stored in a variable named 'embeddings'\n",
        "embeddings = np.random.rand(100, 434688).astype(np.float32)\n",
        "\n",
        "# Create an HDF5 file with compression\n",
        "with h5py.File('/content/drive/MyDrive/sih/model_save/embeddings.h5', 'w') as hf:\n",
        "    hf.create_dataset('data', data=embeddings, compression='gzip', compression_opts=9)\n"
      ],
      "metadata": {
        "id": "fzilxeuh8D9o"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "embeddings.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "igdoKBgg7mmF",
        "outputId": "f9a0c042-5821-4ebe-e6ae-7211b54eba44"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([4, 434688])"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(hashing(embeddings))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bQr3_RsH50nh",
        "outputId": "92834e4d-d94d-47c1-c86a-904248fbc09f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'[0, 1, 0, 1]': [0, 2], '[0, 0, 1, 0]': [1], '[0, 1, 1, 0]': [3]}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "On one random invoice and one random prescription\n",
        "\n",
        "loss before training : 0.3888\n",
        "\n",
        "loss after 10 epochs : 0.1870\n"
      ],
      "metadata": {
        "id": "5tmP0DHCSTMB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import os"
      ],
      "metadata": {
        "id": "qLXhF6ReFDKS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "all_embeds=[]\n",
        "\n",
        "file_list = os.listdir(folder_path)\n",
        "import numpy as np\n",
        "from sklearn.decomposition import PCA\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "for file in files:\n",
        "  img_path=os.path.join(folder_path,file)\n",
        "  image = Image.open(img_path).convert('RGB')\n",
        "  em=calc_embeddings(image)\n",
        "\n",
        "  # Set the desired number of components (new dimensions)\n",
        "  num_components = 1\n",
        "\n",
        "  # Create a PCA model and fit_transform the input vector\n",
        "  pca = PCA(n_components=num_components)\n",
        "  em=em.detach().numpy()\n",
        "  em=em.squeeze()\n",
        "  print(em.shape)\n",
        "  em = pca.fit_transform(em)\n",
        "  print(em.shape)\n",
        "  all_embeds.append(em)\n",
        "  print(len(all_embeds))\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k-9NfuaGBFRY",
        "outputId": "b58ebb37-7eb6-49aa-9a90-2232d1457c5e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(502, 768)\n",
            "(502, 1)\n",
            "1\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(499, 768)\n",
            "(499, 1)\n",
            "2\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(469, 768)\n",
            "(469, 1)\n",
            "3\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(509, 768)\n",
            "(509, 1)\n",
            "4\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(503, 768)\n",
            "(503, 1)\n",
            "5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(566, 768)\n",
            "(566, 1)\n",
            "6\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(564, 768)\n",
            "(564, 1)\n",
            "7\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(518, 768)\n",
            "(518, 1)\n",
            "8\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(502, 768)\n",
            "(502, 1)\n",
            "9\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(518, 768)\n",
            "(518, 1)\n",
            "10\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(566, 768)\n",
            "(566, 1)\n",
            "11\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(515, 768)\n",
            "(515, 1)\n",
            "12\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(511, 768)\n",
            "(511, 1)\n",
            "13\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(515, 768)\n",
            "(515, 1)\n",
            "14\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(566, 768)\n",
            "(566, 1)\n",
            "15\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(522, 768)\n",
            "(522, 1)\n",
            "16\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(559, 768)\n",
            "(559, 1)\n",
            "17\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(568, 768)\n",
            "(568, 1)\n",
            "18\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(511, 768)\n",
            "(511, 1)\n",
            "19\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(609, 768)\n",
            "(609, 1)\n",
            "20\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(533, 768)\n",
            "(533, 1)\n",
            "21\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(620, 768)\n",
            "(620, 1)\n",
            "22\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(509, 768)\n",
            "(509, 1)\n",
            "23\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(621, 768)\n",
            "(621, 1)\n",
            "24\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(488, 768)\n",
            "(488, 1)\n",
            "25\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(681, 768)\n",
            "(681, 1)\n",
            "26\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(639, 768)\n",
            "(639, 1)\n",
            "27\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(552, 768)\n",
            "(552, 1)\n",
            "28\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "29\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(630, 768)\n",
            "(630, 1)\n",
            "30\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(569, 768)\n",
            "(569, 1)\n",
            "31\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(550, 768)\n",
            "(550, 1)\n",
            "32\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(618, 768)\n",
            "(618, 1)\n",
            "33\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(562, 768)\n",
            "(562, 1)\n",
            "34\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(572, 768)\n",
            "(572, 1)\n",
            "35\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(502, 768)\n",
            "(502, 1)\n",
            "36\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(554, 768)\n",
            "(554, 1)\n",
            "37\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "38\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(548, 768)\n",
            "(548, 1)\n",
            "39\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "40\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(569, 768)\n",
            "(569, 1)\n",
            "41\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(558, 768)\n",
            "(558, 1)\n",
            "42\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "43\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(619, 768)\n",
            "(619, 1)\n",
            "44\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "45\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(538, 768)\n",
            "(538, 1)\n",
            "46\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(579, 768)\n",
            "(579, 1)\n",
            "47\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(637, 768)\n",
            "(637, 1)\n",
            "48\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "49\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(573, 768)\n",
            "(573, 1)\n",
            "50\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(505, 768)\n",
            "(505, 1)\n",
            "51\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "52\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "53\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(570, 768)\n",
            "(570, 1)\n",
            "54\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(607, 768)\n",
            "(607, 1)\n",
            "55\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "56\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(555, 768)\n",
            "(555, 1)\n",
            "57\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "58\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "59\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(599, 768)\n",
            "(599, 1)\n",
            "60\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "61\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(637, 768)\n",
            "(637, 1)\n",
            "62\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(572, 768)\n",
            "(572, 1)\n",
            "63\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(684, 768)\n",
            "(684, 1)\n",
            "64\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(432, 768)\n",
            "(432, 1)\n",
            "65\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(607, 768)\n",
            "(607, 1)\n",
            "66\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(702, 768)\n",
            "(702, 1)\n",
            "67\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "68\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(638, 768)\n",
            "(638, 1)\n",
            "69\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "70\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(705, 768)\n",
            "(705, 1)\n",
            "71\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(602, 768)\n",
            "(602, 1)\n",
            "72\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "73\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(413, 768)\n",
            "(413, 1)\n",
            "74\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "75\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "76\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "77\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(695, 768)\n",
            "(695, 1)\n",
            "78\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(642, 768)\n",
            "(642, 1)\n",
            "79\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "80\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(458, 768)\n",
            "(458, 1)\n",
            "81\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(420, 768)\n",
            "(420, 1)\n",
            "82\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(643, 768)\n",
            "(643, 1)\n",
            "83\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(613, 768)\n",
            "(613, 1)\n",
            "84\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "85\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "86\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "87\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "88\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(684, 768)\n",
            "(684, 1)\n",
            "89\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(485, 768)\n",
            "(485, 1)\n",
            "90\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(486, 768)\n",
            "(486, 1)\n",
            "91\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(456, 768)\n",
            "(456, 1)\n",
            "92\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(627, 768)\n",
            "(627, 1)\n",
            "93\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "94\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(404, 768)\n",
            "(404, 1)\n",
            "95\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(489, 768)\n",
            "(489, 1)\n",
            "96\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(351, 768)\n",
            "(351, 1)\n",
            "97\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(628, 768)\n",
            "(628, 1)\n",
            "98\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(564, 768)\n",
            "(564, 1)\n",
            "99\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(679, 768)\n",
            "(679, 1)\n",
            "100\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(648, 768)\n",
            "(648, 1)\n",
            "101\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(682, 768)\n",
            "(682, 1)\n",
            "102\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(485, 768)\n",
            "(485, 1)\n",
            "103\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(561, 768)\n",
            "(561, 1)\n",
            "104\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(489, 768)\n",
            "(489, 1)\n",
            "105\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(670, 768)\n",
            "(670, 1)\n",
            "106\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(424, 768)\n",
            "(424, 1)\n",
            "107\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "108\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(486, 768)\n",
            "(486, 1)\n",
            "109\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "110\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(676, 768)\n",
            "(676, 1)\n",
            "111\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(490, 768)\n",
            "(490, 1)\n",
            "112\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(559, 768)\n",
            "(559, 1)\n",
            "113\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(627, 768)\n",
            "(627, 1)\n",
            "114\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(642, 768)\n",
            "(642, 1)\n",
            "115\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(482, 768)\n",
            "(482, 1)\n",
            "116\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(636, 768)\n",
            "(636, 1)\n",
            "117\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(681, 768)\n",
            "(681, 1)\n",
            "118\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(462, 768)\n",
            "(462, 1)\n",
            "119\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(644, 768)\n",
            "(644, 1)\n",
            "120\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "121\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(667, 768)\n",
            "(667, 1)\n",
            "122\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(625, 768)\n",
            "(625, 1)\n",
            "123\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(668, 768)\n",
            "(668, 1)\n",
            "124\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(669, 768)\n",
            "(669, 1)\n",
            "125\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(655, 768)\n",
            "(655, 1)\n",
            "126\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(610, 768)\n",
            "(610, 1)\n",
            "127\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(541, 768)\n",
            "(541, 1)\n",
            "128\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(652, 768)\n",
            "(652, 1)\n",
            "129\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(665, 768)\n",
            "(665, 1)\n",
            "130\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(709, 768)\n",
            "(709, 1)\n",
            "131\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py:907: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(630, 768)\n",
            "(630, 1)\n",
            "132\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Inference\n"
      ],
      "metadata": {
        "id": "UF__LbLJbKc2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#all embeddings from training are stored ,\n",
        "#when we enter a new file its embedding is calculated and reshaped and normalized\n",
        "#we calculate its dot product with existing embeddings and output red/amber/green flag based on the score\n",
        "#also output with which embeddins/image it has highest dot product"
      ],
      "metadata": {
        "id": "-eYa5Y_xSpM2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# import os\n",
        "# from torchvision.io import read_image\n",
        "# def list_form(path_data):\n",
        "#   name_list = os.listdir(path_data)\n",
        "#   return name_list"
      ],
      "metadata": {
        "id": "EJFEFY41b8Fr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# def form_data_emb(path_data):\n",
        "#   data_emb=[]\n",
        "#   name_list=list_form(path_data)\n",
        "#   for file in name_list:\n",
        "#     img_path=os.path.join(path_data, file)\n",
        "#     img = Image.open(image_path).convert(\"RGB\")\n",
        "#     em=embeddings(img)\n",
        "#     data_emb.append(em)\n",
        "#   data_emb=torch.tensor(data_emb)\n",
        "#   return data_emb"
      ],
      "metadata": {
        "id": "D1pp_l0iYMwg"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# def inference(test_emb,data_emb,name_list):\n",
        "\n",
        "#   # data_emb m x em_len\n",
        "#   sim=torch.matmul(test_emb,data_emb.T)\n",
        "#   max_sim,index=torch.max(sim,dim=0)\n",
        "#   file_name=name_list[index]\n",
        "#   return max_sim,file_name\n"
      ],
      "metadata": {
        "id": "iePJFPqsYRYU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "TGmndr7SSOvb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Hashing"
      ],
      "metadata": {
        "id": "pMQo0cdxSNuq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# def gen_hash(emb,vec):\n",
        "#         bools = (torch.matmul(emb,vec.T)> 0).to(torch.int)\n",
        "#         return bools\n",
        "# def hashing(emb):\n",
        "#   dim=emb.shape[1]\n",
        "#   vec = torch.randn(4,dim)\n",
        "\n",
        "\n",
        "#   bools=gen_hash(emb,vec)\n",
        "#   emp=[]\n",
        "#   hash = [[0]] * 16\n",
        "#   hash_table=dict()\n",
        "#   for i in range(emb.shape[0]):\n",
        "#     row=bools[i]\n",
        "#     a=str(row.tolist())\n",
        "\n",
        "#     if a not in hash_table:\n",
        "#       hash_table[a] = []\n",
        "\n",
        "#     hash_table[a].extend([i])\n",
        "#   return hash_table\n",
        "\n",
        "# # returns a dictionary of hashing index maped to index"
      ],
      "metadata": {
        "id": "t5CsJH5DSMlh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# import json\n",
        "# import os\n",
        "\n",
        "# # Directory where you want to save the JSON file\n",
        "# directory = '/content/drive/MyDrive/sih/model_save/'\n",
        "\n",
        "\n",
        "# # File path for the JSON file within the directory\n",
        "# file_path = os.path.join(directory, 'data.json')\n",
        "\n",
        "# # Save the dictionary as JSON in the file\n",
        "# with open(file_path, 'w') as json_file:\n",
        "#     json.dump(data, json_file, indent=4)  # indent=4 for pretty formatting (optional)\n",
        "#     print(f\"Data saved to {file_path}\")"
      ],
      "metadata": {
        "id": "uPimZ_xQZcdf"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}