image_classification.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "SsKhFxmGiGMa"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/home/mrblack/pyenvs/venv_signLanguage/lib/python3.8/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
            "  from .autonotebook import tqdm as notebook_tqdm\n"
          ]
        }
      ],
      "source": [
        "#Pytorch for vision classification\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "from torch.optim import lr_scheduler\n",
        "import torchvision\n",
        "from torchvision import datasets, models\n",
        "import torchvision.transforms as transforms\n",
        "from torch.utils.data import Dataset, DataLoader\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "import time\n",
        "import os\n",
        "import copy\n",
        "\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4h3iWJ8Ziyj8"
      },
      "source": [
        "The output of torchvision datasets are PILImage images of range [0, 1]. We transform them to Tensors of normalized range [-1, 1]."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7pqQna7arxkR",
        "outputId": "28f98ef3-a92d-4707-8f2f-76670f5fba45"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "#data from google drive:\n",
        "# from google.colab import drive\n",
        "# drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iXI_bMbF3vKG",
        "outputId": "6442abc1-3cb9-4c79-cce6-40c0626ddf7e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting split-folders[full]\n",
            "  Downloading split_folders-0.5.1-py3-none-any.whl (8.4 kB)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (from split-folders[full]) (4.64.1)\n",
            "Installing collected packages: split-folders\n",
            "Successfully installed split-folders-0.5.1\n"
          ]
        }
      ],
      "source": [
        "#package for splitting the dataset\n",
        "# !pip install split-folders[full]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "EedfbMiR34So"
      },
      "outputs": [],
      "source": [
        "# import splitfolders\n",
        "\n",
        "# input_folder = '/content/drive/MyDrive/Datasets/SignLanguageDataset'\n",
        "# splitfolders.ratio(input_folder, output=\"/content/drive/MyDrive/Datasets/SignLanguageDatasetDivided\",\n",
        "#                    seed=42, ratio=(.7,.3),\n",
        "#                    group_prefix=None)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "0jkkiGi0bl-u"
      },
      "outputs": [],
      "source": [
        "mean = np.array([0.485, 0.456, 0.406])\n",
        "std = np.array([0.229, 0.224, 0.225])\n",
        "\n",
        "data_transforms = {\n",
        "    'train': transforms.Compose([\n",
        "        transforms.RandomResizedCrop(224),\n",
        "        transforms.RandomHorizontalFlip(),\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Normalize(mean,std)\n",
        "    ]),\n",
        "    'val': transforms.Compose([\n",
        "        transforms.Resize(256),\n",
        "        transforms.CenterCrop(224),\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Normalize(mean, std)\n",
        "    ]),\n",
        "}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qRAvEgpY3g_s",
        "outputId": "7e915c80-49d5-4271-e518-faa9031770f8"
      },
      "outputs": [
        {
          "ename": "FileNotFoundError",
          "evalue": "[Errno 2] No such file or directory: '/home/mrblack/Datasets/SignLanguageLetteres/train'",
          "output_type": "error",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
            "Cell \u001b[0;32mIn[3], line 5\u001b[0m\n\u001b[1;32m      3\u001b[0m data_dir \u001b[39m=\u001b[39m os\u001b[39m.\u001b[39mpath\u001b[39m.\u001b[39mexpanduser(\u001b[39m'\u001b[39m\u001b[39m~\u001b[39m\u001b[39m'\u001b[39m) \u001b[39m+\u001b[39m \u001b[39m'\u001b[39m\u001b[39m/Datasets/SignLanguageLetteres\u001b[39m\u001b[39m'\u001b[39m\n\u001b[1;32m      4\u001b[0m sets \u001b[39m=\u001b[39m [\u001b[39m'\u001b[39m\u001b[39mtrain\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mval\u001b[39m\u001b[39m'\u001b[39m]\n\u001b[0;32m----> 5\u001b[0m image_datasets \u001b[39m=\u001b[39m {x: datasets\u001b[39m.\u001b[39mImageFolder(os\u001b[39m.\u001b[39mpath\u001b[39m.\u001b[39mjoin(data_dir, x),\n\u001b[1;32m      6\u001b[0m                                           data_transforms[x])\n\u001b[1;32m      7\u001b[0m                   \u001b[39mfor\u001b[39;00m x \u001b[39min\u001b[39;00m sets}\n\u001b[1;32m      9\u001b[0m dataloaders \u001b[39m=\u001b[39m {x: DataLoader(image_datasets[x], batch_size\u001b[39m=\u001b[39m\u001b[39m4\u001b[39m,\n\u001b[1;32m     10\u001b[0m                                               shuffle\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m, num_workers\u001b[39m=\u001b[39m\u001b[39m0\u001b[39m)\n\u001b[1;32m     11\u001b[0m                   \u001b[39mfor\u001b[39;00m x \u001b[39min\u001b[39;00m sets} \n\u001b[1;32m     13\u001b[0m dataset_sizes \u001b[39m=\u001b[39m {x: \u001b[39mlen\u001b[39m(image_datasets[x]) \u001b[39mfor\u001b[39;00m x \u001b[39min\u001b[39;00m sets}\n",
            "Cell \u001b[0;32mIn[3], line 5\u001b[0m, in \u001b[0;36m<dictcomp>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m      3\u001b[0m data_dir \u001b[39m=\u001b[39m os\u001b[39m.\u001b[39mpath\u001b[39m.\u001b[39mexpanduser(\u001b[39m'\u001b[39m\u001b[39m~\u001b[39m\u001b[39m'\u001b[39m) \u001b[39m+\u001b[39m \u001b[39m'\u001b[39m\u001b[39m/Datasets/SignLanguageLetteres\u001b[39m\u001b[39m'\u001b[39m\n\u001b[1;32m      4\u001b[0m sets \u001b[39m=\u001b[39m [\u001b[39m'\u001b[39m\u001b[39mtrain\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mval\u001b[39m\u001b[39m'\u001b[39m]\n\u001b[0;32m----> 5\u001b[0m image_datasets \u001b[39m=\u001b[39m {x: datasets\u001b[39m.\u001b[39;49mImageFolder(os\u001b[39m.\u001b[39;49mpath\u001b[39m.\u001b[39;49mjoin(data_dir, x),\n\u001b[1;32m      6\u001b[0m                                           data_transforms[x])\n\u001b[1;32m      7\u001b[0m                   \u001b[39mfor\u001b[39;00m x \u001b[39min\u001b[39;00m sets}\n\u001b[1;32m      9\u001b[0m dataloaders \u001b[39m=\u001b[39m {x: DataLoader(image_datasets[x], batch_size\u001b[39m=\u001b[39m\u001b[39m4\u001b[39m,\n\u001b[1;32m     10\u001b[0m                                               shuffle\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m, num_workers\u001b[39m=\u001b[39m\u001b[39m0\u001b[39m)\n\u001b[1;32m     11\u001b[0m                   \u001b[39mfor\u001b[39;00m x \u001b[39min\u001b[39;00m sets} \n\u001b[1;32m     13\u001b[0m dataset_sizes \u001b[39m=\u001b[39m {x: \u001b[39mlen\u001b[39m(image_datasets[x]) \u001b[39mfor\u001b[39;00m x \u001b[39min\u001b[39;00m sets}\n",
            "File \u001b[0;32m~/pyenvs/venv_signLanguage/lib/python3.8/site-packages/torchvision/datasets/folder.py:309\u001b[0m, in \u001b[0;36mImageFolder.__init__\u001b[0;34m(self, root, transform, target_transform, loader, is_valid_file)\u001b[0m\n\u001b[1;32m    301\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__init__\u001b[39m(\n\u001b[1;32m    302\u001b[0m     \u001b[39mself\u001b[39m,\n\u001b[1;32m    303\u001b[0m     root: \u001b[39mstr\u001b[39m,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    307\u001b[0m     is_valid_file: Optional[Callable[[\u001b[39mstr\u001b[39m], \u001b[39mbool\u001b[39m]] \u001b[39m=\u001b[39m \u001b[39mNone\u001b[39;00m,\n\u001b[1;32m    308\u001b[0m ):\n\u001b[0;32m--> 309\u001b[0m     \u001b[39msuper\u001b[39;49m()\u001b[39m.\u001b[39;49m\u001b[39m__init__\u001b[39;49m(\n\u001b[1;32m    310\u001b[0m         root,\n\u001b[1;32m    311\u001b[0m         loader,\n\u001b[1;32m    312\u001b[0m         IMG_EXTENSIONS \u001b[39mif\u001b[39;49;00m is_valid_file \u001b[39mis\u001b[39;49;00m \u001b[39mNone\u001b[39;49;00m \u001b[39melse\u001b[39;49;00m \u001b[39mNone\u001b[39;49;00m,\n\u001b[1;32m    313\u001b[0m         transform\u001b[39m=\u001b[39;49mtransform,\n\u001b[1;32m    314\u001b[0m         target_transform\u001b[39m=\u001b[39;49mtarget_transform,\n\u001b[1;32m    315\u001b[0m         is_valid_file\u001b[39m=\u001b[39;49mis_valid_file,\n\u001b[1;32m    316\u001b[0m     )\n\u001b[1;32m    317\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mimgs \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39msamples\n",
            "File \u001b[0;32m~/pyenvs/venv_signLanguage/lib/python3.8/site-packages/torchvision/datasets/folder.py:144\u001b[0m, in \u001b[0;36mDatasetFolder.__init__\u001b[0;34m(self, root, loader, extensions, transform, target_transform, is_valid_file)\u001b[0m\n\u001b[1;32m    134\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__init__\u001b[39m(\n\u001b[1;32m    135\u001b[0m     \u001b[39mself\u001b[39m,\n\u001b[1;32m    136\u001b[0m     root: \u001b[39mstr\u001b[39m,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    141\u001b[0m     is_valid_file: Optional[Callable[[\u001b[39mstr\u001b[39m], \u001b[39mbool\u001b[39m]] \u001b[39m=\u001b[39m \u001b[39mNone\u001b[39;00m,\n\u001b[1;32m    142\u001b[0m ) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m \u001b[39mNone\u001b[39;00m:\n\u001b[1;32m    143\u001b[0m     \u001b[39msuper\u001b[39m()\u001b[39m.\u001b[39m\u001b[39m__init__\u001b[39m(root, transform\u001b[39m=\u001b[39mtransform, target_transform\u001b[39m=\u001b[39mtarget_transform)\n\u001b[0;32m--> 144\u001b[0m     classes, class_to_idx \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mfind_classes(\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mroot)\n\u001b[1;32m    145\u001b[0m     samples \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mmake_dataset(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mroot, class_to_idx, extensions, is_valid_file)\n\u001b[1;32m    147\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mloader \u001b[39m=\u001b[39m loader\n",
            "File \u001b[0;32m~/pyenvs/venv_signLanguage/lib/python3.8/site-packages/torchvision/datasets/folder.py:218\u001b[0m, in \u001b[0;36mDatasetFolder.find_classes\u001b[0;34m(self, directory)\u001b[0m\n\u001b[1;32m    191\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mfind_classes\u001b[39m(\u001b[39mself\u001b[39m, directory: \u001b[39mstr\u001b[39m) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m Tuple[List[\u001b[39mstr\u001b[39m], Dict[\u001b[39mstr\u001b[39m, \u001b[39mint\u001b[39m]]:\n\u001b[1;32m    192\u001b[0m \u001b[39m    \u001b[39m\u001b[39m\"\"\"Find the class folders in a dataset structured as follows::\u001b[39;00m\n\u001b[1;32m    193\u001b[0m \n\u001b[1;32m    194\u001b[0m \u001b[39m        directory/\u001b[39;00m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    216\u001b[0m \u001b[39m        (Tuple[List[str], Dict[str, int]]): List of all classes and dictionary mapping each class to an index.\u001b[39;00m\n\u001b[1;32m    217\u001b[0m \u001b[39m    \"\"\"\u001b[39;00m\n\u001b[0;32m--> 218\u001b[0m     \u001b[39mreturn\u001b[39;00m find_classes(directory)\n",
            "File \u001b[0;32m~/pyenvs/venv_signLanguage/lib/python3.8/site-packages/torchvision/datasets/folder.py:40\u001b[0m, in \u001b[0;36mfind_classes\u001b[0;34m(directory)\u001b[0m\n\u001b[1;32m     35\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mfind_classes\u001b[39m(directory: \u001b[39mstr\u001b[39m) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m Tuple[List[\u001b[39mstr\u001b[39m], Dict[\u001b[39mstr\u001b[39m, \u001b[39mint\u001b[39m]]:\n\u001b[1;32m     36\u001b[0m \u001b[39m    \u001b[39m\u001b[39m\"\"\"Finds the class folders in a dataset.\u001b[39;00m\n\u001b[1;32m     37\u001b[0m \n\u001b[1;32m     38\u001b[0m \u001b[39m    See :class:`DatasetFolder` for details.\u001b[39;00m\n\u001b[1;32m     39\u001b[0m \u001b[39m    \"\"\"\u001b[39;00m\n\u001b[0;32m---> 40\u001b[0m     classes \u001b[39m=\u001b[39m \u001b[39msorted\u001b[39m(entry\u001b[39m.\u001b[39mname \u001b[39mfor\u001b[39;00m entry \u001b[39min\u001b[39;00m os\u001b[39m.\u001b[39;49mscandir(directory) \u001b[39mif\u001b[39;00m entry\u001b[39m.\u001b[39mis_dir())\n\u001b[1;32m     41\u001b[0m     \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m classes:\n\u001b[1;32m     42\u001b[0m         \u001b[39mraise\u001b[39;00m \u001b[39mFileNotFoundError\u001b[39;00m(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mCouldn\u001b[39m\u001b[39m'\u001b[39m\u001b[39mt find any class folder in \u001b[39m\u001b[39m{\u001b[39;00mdirectory\u001b[39m}\u001b[39;00m\u001b[39m.\u001b[39m\u001b[39m\"\u001b[39m)\n",
            "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: '/home/mrblack/Datasets/SignLanguageLetteres/train'"
          ]
        }
      ],
      "source": [
        "#There is no val, it is only train and test\n",
        "\n",
        "data_dir = os.path.expanduser('~') + '/Datasets/SignLanguageLetteres'\n",
        "sets = ['train', 'val']\n",
        "image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),\n",
        "                                          data_transforms[x])\n",
        "                  for x in sets}\n",
        "\n",
        "dataloaders = {x: DataLoader(image_datasets[x], batch_size=4,\n",
        "                                              shuffle=True, num_workers=0)\n",
        "                  for x in sets} \n",
        "\n",
        "dataset_sizes = {x: len(image_datasets[x]) for x in sets}\n",
        "class_names = image_datasets['train'].classes\n",
        "print(class_names)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "na5EcG49qjLR"
      },
      "outputs": [],
      "source": [
        "batch_size = 4\n",
        "# num_epochs = 1\n",
        "# learning_rate = 0.01"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tInpvlQCaI7C"
      },
      "outputs": [],
      "source": [
        "train_loader = DataLoader(dataset=image_datasets['train'], batch_size=batch_size, shuffle=True)\n",
        "test_loader = DataLoader(dataset=image_datasets['val'], batch_size=batch_size, shuffle=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "id": "XCAChshRkD_A",
        "outputId": "744d8672-3a4f-480c-c93d-765836a1cd6c"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 4 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/plain": [
              "torch.Size([4, 3, 224, 224])"
            ]
          },
          "execution_count": 7,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "## Plotting some images of the dataset:\n",
        "sign_example = iter(train_loader)\n",
        "sign_data, sign_label = sign_example.next()\n",
        "\n",
        "for i in range(4):\n",
        "  plt.subplot(2,2, i+1)\n",
        "  plt.imshow(sign_data[i][0], cmap='gray')\n",
        "plt.show()\n",
        "sign_data.shape\n",
        "# image shape = (3, 224, 224)\n",
        "# the 4 is for batch size"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W0SldUoenQ0L"
      },
      "source": [
        "Another way of showing the images: \n",
        "\n",
        "[thanks to transfer learning guide from pytorch](https://pytorch.org/tutorials/beginner/transfer_learning_tutorial.html)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 148
        },
        "id": "pn0bs8Ewmh04",
        "outputId": "be802f74-7edc-4be5-d050-b77f9611ba84"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "def imshow(inp, title=None):\n",
        "    \"\"\"Imshow for Tensor.\"\"\"\n",
        "    inp = inp.numpy().transpose((1, 2, 0))\n",
        "    mean = np.array([0.485, 0.456, 0.406])\n",
        "    std = np.array([0.229, 0.224, 0.225])\n",
        "    inp = std * inp + mean\n",
        "    inp = np.clip(inp, 0, 1)\n",
        "    plt.imshow(inp)\n",
        "    if title is not None:\n",
        "        plt.title(title)\n",
        "    plt.pause(0.001)  # pause a bit so that plots are updated\n",
        "\n",
        "out = torchvision.utils.make_grid(sign_data)\n",
        "imshow(out, title=[class_names[x] for x in sign_label])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TTgM7osZpfLM",
        "outputId": "b1524eab-3a58-4d9d-e2e1-9bac781c6062"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "SignNet(\n",
            "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
            "  (pool1): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
            "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
            "  (pool2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
            "  (fc1): Linear(in_features=44944, out_features=120, bias=True)\n",
            "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
            "  (fc3): Linear(in_features=84, out_features=11, bias=True)\n",
            ")\n"
          ]
        }
      ],
      "source": [
        "# image shape = (3, 224, 224)\n",
        "#Formula for CNN = (W-F + 2P) / S + 1\n",
        "# W: input width\n",
        "# F: Filter Size\n",
        "# P: Padding\n",
        "# S: Stride\n",
        "# 224 - F + 2*(0) / 1 + 1\n",
        "class SignNet(nn.Module):\n",
        "  def __init__(self):\n",
        "    super(SignNet, self).__init__()\n",
        "    self.conv1 = nn.Conv2d(3, 6, 5) # (224 - 5 + 2*(0)) / 1 + 1 =220 -> 6, 220, 220\n",
        "    self.pool1 = nn.MaxPool2d(2,2) #220 / 2 = 110\n",
        "    self.conv2 = nn.Conv2d(6, 16, 5)  # (110 - 5 + 2*(0)) / 1 + 1 = 106 -> 16, 106, 106\n",
        "    self.pool2 = nn.MaxPool2d(2,2) #106 / 2 = 53 - > 16, 53, 53 \n",
        "    self.fc1 = nn.Linear(16*53*53, 120)\n",
        "    self.fc2 = nn.Linear(120, 84)\n",
        "    self.fc3 = nn.Linear(84, 11) # 11 classes\n",
        "\n",
        "  def forward(self, x):\n",
        "    x = F.relu(self.conv1(x))\n",
        "    x = self.pool1(x)\n",
        "    x = F.relu(self.conv2(x))\n",
        "    x = self.pool2(x)\n",
        "    x = x.view(-1, 16*53*53) #Flatten\n",
        "    x = F.relu(self.fc1(x))\n",
        "    x = F.relu(self.fc2(x))\n",
        "    x = F.relu(self.fc3(x))\n",
        "    return x\n",
        "\n",
        "\n",
        "model = SignNet().to(device) #in case I'm training on the GPU\n",
        "# x = model.forward\n",
        "# print(x)\n",
        "print(model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3aojL4DWq5xO"
      },
      "outputs": [],
      "source": [
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)\n",
        "\n",
        "# n_total_steps = len(train_loader)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dEzSYd9VrJ3D"
      },
      "outputs": [],
      "source": [
        "def train_model(model, train_loader, criterion, optimizer, num_epochs ,batch_size=4):\n",
        "  # best_model_wts = copy.deepcopy(model.state_dict())\n",
        "  # best_acc = 0.0\n",
        "  n_total_steps = len(train_loader)\n",
        "  for epoch in range(num_epochs):\n",
        "    for i, (images, labels) in enumerate(train_loader):\n",
        "\n",
        "      images = images.to(device)\n",
        "      labels = labels.to(device)\n",
        "\n",
        "      #Forward pass\n",
        "      outputs = model(images)\n",
        "      loss = criterion(outputs, labels)\n",
        "      wandb.log({\"loss\": loss}) ########## weight and biases\n",
        "\n",
        "      #Backward and optimize\n",
        "      optimizer.zero_grad()\n",
        "      loss.backward()\n",
        "      optimizer.step()\n",
        "      # print('here')\n",
        "      if (i+1) % 2000 == 0:\n",
        "        print(f'Epoch [{epoch + 1}/ {num_epochs}], Step [{i+1} / {n_total_steps}], Loss: {loss.item():.4f}')\n",
        "\n",
        "\n",
        "def eval_model(test_loader, model, batch_size=4):\n",
        "  with torch.no_grad():\n",
        "        n_correct = 0\n",
        "        n_samples = 0\n",
        "        n_class_correct = [0 for i in range(11)] # you have 10 classes 0 to 9 and the unknown\n",
        "        n_class_samples = [0 for i in range(11)]\n",
        "\n",
        "        for images, labels in test_loader:\n",
        "          images = images.to(device)\n",
        "          labels = labels.to(device)\n",
        "          outputs = model(images)\n",
        "\n",
        "          _, predicted = torch.max(outputs,1)\n",
        "          n_samples += labels.size(0)\n",
        "          n_correct += (predicted == labels).sum().item()\n",
        "          # print(len(labels))\n",
        "          for i in range(len(labels)): #range(len(labels)): #range(batch_size)\n",
        "            label = labels[i]\n",
        "            pred = predicted[i]\n",
        "            if (label == pred):\n",
        "              n_class_correct[label] += 1\n",
        "\n",
        "            n_class_samples[label] +=1\n",
        "\n",
        "        acc= 100.0 * n_correct / n_samples\n",
        "        print(f'Accuracy of the model {acc} %')\n",
        "\n",
        "\n",
        "        for i in range(11): #number of classes\n",
        "          acc = 100.0 * n_class_correct[i] / n_class_samples[i]\n",
        "          print(f'Accuracy: {class_names[i]} : {acc} %')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "5OgPAqCr3wLB",
        "outputId": "1bd6a7c5-83a3-474b-f6c6-ab072b2f7a69"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch [1/ 1], Step [2000 / 2889], Loss: 2.3956\n"
          ]
        }
      ],
      "source": [
        "train_model(model, train_loader, criterion, optimizer, 1, batch_size=4)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "EUGRa3bQQ6ni"
      },
      "outputs": [],
      "source": [
        "File_Path = '/content/drive/MyDrive/Datasets/DL_Models/sign_model.pth'\n",
        "\n",
        "# torch.save(model.state_dict(), File_Path)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "W979ORXpRTLJ"
      },
      "outputs": [],
      "source": [
        "loaded_model = model()\n",
        "loaded_model.load_state_dict(torch.load(File_Path))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "K1T2CiTKG08F"
      },
      "outputs": [],
      "source": [
        "eval_model(test_loader, model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "jV40R3BYVQB8"
      },
      "outputs": [],
      "source": [
        "## Fine tuning the network using mobilenet_v2 (training the whole model)\n",
        "\n",
        "model_2 = models.mobilenet_v2(pretrained=True)\n",
        "num_features_2 = model_2.fc.in_features\n",
        "model_2.fc = nn.Linear(num_features_2, 11) #11 classes\n",
        "model_2.to(device)\n",
        "\n",
        "\n",
        "\n",
        "print(num_features_2)\n",
        "print(model_2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000,
          "referenced_widgets": [
            "40b1de1b96e540928148a44cdf0a1bfa",
            "faab1081d00e4f6db2b22390dafeda7a",
            "dbcf3d908ff64860917fdf7d31d14a25",
            "c5b2f092d77349d1bc46017920ef912f",
            "d0927ba0eea6464ab01eaeecb554d014",
            "6a61b67593514ee88456810caf35a365",
            "a7ae4a35d40345d090b3a2a64fbc7096",
            "b659edce486d474b986f77d5111856ec",
            "007e590b8ea948a9bc3c6b910f391929",
            "46c1644284584953a7f56ec9366c2643",
            "79c0b819f7b24ca3b48d23a31a794437"
          ]
        },
        "id": "-2gLk6DRomW7",
        "outputId": "a4ad6046-0c89-43a2-8f0f-e85c21a10767"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.\n",
            "  warnings.warn(msg)\n",
            "Downloading: \"https://download.pytorch.org/models/resnet18-f37072fd.pth\" to /root/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "40b1de1b96e540928148a44cdf0a1bfa",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0.00/44.7M [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "512\n",
            "ResNet(\n",
            "  (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
            "  (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "  (relu): ReLU(inplace=True)\n",
            "  (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
            "  (layer1): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (layer2): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (downsample): Sequential(\n",
            "        (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "        (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      )\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (layer3): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (downsample): Sequential(\n",
            "        (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      )\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (layer4): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (downsample): Sequential(\n",
            "        (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "        (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      )\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
            "  (fc): Linear(in_features=512, out_features=11, bias=True)\n",
            ")\n"
          ]
        }
      ],
      "source": [
        "## Fine tuning the network using resnet18 (training the whole model)\n",
        "\n",
        "model_3 = models.resnet18(pretrained=True)\n",
        "num_features_3 = model_3.fc.in_features\n",
        "model_3.fc = nn.Linear(num_features_3, 11)\n",
        "\n",
        "model_3.to(device)\n",
        "\n",
        "print(num_features_3)\n",
        "print(model_3)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000,
          "referenced_widgets": [
            "c457f3d5132440d088a1bb41250f091e",
            "b6fdbd5ec7bc40faa59d89300d869fe9",
            "9f35d18d6885427c8af3357ac58a0b5c",
            "022ff962f15f407588450857195ce416",
            "2be713a0f9bd4f219494d269a287b4da",
            "9a784257925d42028158ad154ab3e305",
            "dfd9d992a8c5472aa5b5210435945bdd",
            "18709ae985d54153aa3173b7b6466747",
            "b312e4ce7ba845c1836e88c789a5bc89",
            "16b0fa37df0440f8b9946e056ce91d8b",
            "94f6c919190140ce891422d00cdd0af5"
          ]
        },
        "id": "VKsuEGlPxaqP",
        "outputId": "306c50ec-4caa-485c-a264-1c64a00ef1bb"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torchvision/models/_utils.py:209: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and will be removed in 0.15, please use 'weights' instead.\n",
            "  f\"The parameter '{pretrained_param}' is deprecated since 0.13 and will be removed in 0.15, \"\n",
            "/usr/local/lib/python3.7/dist-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.\n",
            "  warnings.warn(msg)\n",
            "Downloading: \"https://download.pytorch.org/models/resnet18-f37072fd.pth\" to /root/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "c457f3d5132440d088a1bb41250f091e",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0.00/44.7M [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "512\n",
            "ResNet(\n",
            "  (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
            "  (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "  (relu): ReLU(inplace=True)\n",
            "  (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
            "  (layer1): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (layer2): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (downsample): Sequential(\n",
            "        (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "        (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      )\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (layer3): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (downsample): Sequential(\n",
            "        (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      )\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (layer4): Sequential(\n",
            "    (0): BasicBlock(\n",
            "      (conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (downsample): Sequential(\n",
            "        (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "        (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      )\n",
            "    )\n",
            "    (1): BasicBlock(\n",
            "      (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "      (relu): ReLU(inplace=True)\n",
            "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    )\n",
            "  )\n",
            "  (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
            "  (fc): Linear(in_features=512, out_features=11, bias=True)\n",
            ")\n"
          ]
        }
      ],
      "source": [
        "## Freezing the layers at the begining at only train the last layers\n",
        "## Much faster\n",
        "model_4 = models.resnet18(pretrained=True)\n",
        "for param in model_4.parameters():\n",
        "  param.requires_grad = False\n",
        "\n",
        "num_features_4 = model_4.fc.in_features\n",
        "model_4.fc = nn.Linear(num_features_4, 11) #11 classes\n",
        "model_4.to(device)\n",
        "\n",
        "\n",
        "\n",
        "print(num_features_4)\n",
        "print(model_4)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qm2jJRt8osf-"
      },
      "outputs": [],
      "source": [
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.SGD(model.parameters(), lr=0.0001)\n",
        "\n",
        "step_lr_scheduler = lr_scheduler.StepLR(optimizer, step_size=7, gamma=0.1) #every 7 epochs, learning rate is multiplied by 0.1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PPkFmcwtyXW6"
      },
      "outputs": [],
      "source": [
        "#model 3 training (pretrained using resnet18)\n",
        "# train_model(model_3, train_loader, criterion, optimizer, 1, batch_size=4)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dHS_9-8sHuV0",
        "outputId": "ab1383b1-68cf-4e32-d1f9-2b7f541a55ad"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch [1/ 1], Step [2000 / 2889], Loss: 2.1943\n"
          ]
        }
      ],
      "source": [
        "#model 4 training (pretrained using resnet18) - Freeze\n",
        "# train_model(model_4, train_loader, criterion, optimizer, 1, batch_size=4)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "k_ESk-rscmM7"
      },
      "outputs": [],
      "source": [
        "File_Path = '/content/drive/MyDrive/Datasets/DL_Models/sign_model_100.pth'\n",
        "\n",
        "# torch.save(model_4.state_dict(), File_Path)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7Yk3rAYm00DZ",
        "outputId": "19f529f3-0670-458a-dfba-2a067b26cf14"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "execution_count": 26,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Loading the model\n",
        "\n",
        "model = model_4\n",
        "model.load_state_dict(torch.load(File_Path))\n",
        "# model.eval()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "1c1hApbiCle6",
        "outputId": "4274aee0-498a-4c26-ca03-bd4996c1ee64"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting wandb\n",
            "  Downloading wandb-0.13.5-py2.py3-none-any.whl (1.9 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.9 MB 26.0 MB/s \n",
            "\u001b[?25hRequirement already satisfied: six>=1.13.0 in /usr/local/lib/python3.7/dist-packages (from wandb) (1.15.0)\n",
            "Requirement already satisfied: protobuf!=4.0.*,!=4.21.0,<5,>=3.12.0 in /usr/local/lib/python3.7/dist-packages (from wandb) (3.19.6)\n",
            "Requirement already satisfied: requests<3,>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from wandb) (2.23.0)\n",
            "Collecting shortuuid>=0.5.0\n",
            "  Downloading shortuuid-1.0.11-py3-none-any.whl (10 kB)\n",
            "Requirement already satisfied: Click!=8.0.0,>=7.0 in /usr/local/lib/python3.7/dist-packages (from wandb) (7.1.2)\n",
            "Requirement already satisfied: psutil>=5.0.0 in /usr/local/lib/python3.7/dist-packages (from wandb) (5.4.8)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from wandb) (57.4.0)\n",
            "Collecting setproctitle\n",
            "  Downloading setproctitle-1.3.2-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl (30 kB)\n",
            "Collecting sentry-sdk>=1.0.0\n",
            "  Downloading sentry_sdk-1.11.0-py2.py3-none-any.whl (168 kB)\n",
            "\u001b[K     |████████████████████████████████| 168 kB 72.2 MB/s \n",
            "\u001b[?25hCollecting GitPython>=1.0.0\n",
            "  Downloading GitPython-3.1.29-py3-none-any.whl (182 kB)\n",
            "\u001b[K     |████████████████████████████████| 182 kB 64.9 MB/s \n",
            "\u001b[?25hRequirement already satisfied: PyYAML in /usr/local/lib/python3.7/dist-packages (from wandb) (6.0)\n",
            "Requirement already satisfied: promise<3,>=2.0 in /usr/local/lib/python3.7/dist-packages (from wandb) (2.3)\n",
            "Collecting pathtools\n",
            "  Downloading pathtools-0.1.2.tar.gz (11 kB)\n",
            "Collecting docker-pycreds>=0.4.0\n",
            "  Downloading docker_pycreds-0.4.0-py2.py3-none-any.whl (9.0 kB)\n",
            "Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.7/dist-packages (from GitPython>=1.0.0->wandb) (4.1.1)\n",
            "Collecting gitdb<5,>=4.0.1\n",
            "  Downloading gitdb-4.0.9-py3-none-any.whl (63 kB)\n",
            "\u001b[K     |████████████████████████████████| 63 kB 1.9 MB/s \n",
            "\u001b[?25hCollecting smmap<6,>=3.0.1\n",
            "  Downloading smmap-5.0.0-py3-none-any.whl (24 kB)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.0.0->wandb) (3.0.4)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.0.0->wandb) (2022.9.24)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.0.0->wandb) (2.10)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.0.0->wandb) (1.24.3)\n",
            "Collecting sentry-sdk>=1.0.0\n",
            "  Downloading sentry_sdk-1.10.1-py2.py3-none-any.whl (166 kB)\n",
            "\u001b[K     |████████████████████████████████| 166 kB 74.3 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.10.0-py2.py3-none-any.whl (166 kB)\n",
            "\u001b[K     |████████████████████████████████| 166 kB 72.6 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.10-py2.py3-none-any.whl (162 kB)\n",
            "\u001b[K     |████████████████████████████████| 162 kB 76.5 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.9-py2.py3-none-any.whl (162 kB)\n",
            "\u001b[K     |████████████████████████████████| 162 kB 75.1 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.8-py2.py3-none-any.whl (158 kB)\n",
            "\u001b[K     |████████████████████████████████| 158 kB 78.2 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.7-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 78.4 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.6-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 78.7 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.5-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 78.5 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.4-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 78.0 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.3-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 78.6 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.2-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 79.6 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.1-py2.py3-none-any.whl (157 kB)\n",
            "\u001b[K     |████████████████████████████████| 157 kB 76.1 MB/s \n",
            "\u001b[?25h  Downloading sentry_sdk-1.9.0-py2.py3-none-any.whl (156 kB)\n",
            "\u001b[K     |████████████████████████████████| 156 kB 75.6 MB/s \n",
            "\u001b[?25hBuilding wheels for collected packages: pathtools\n",
            "  Building wheel for pathtools (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for pathtools: filename=pathtools-0.1.2-py3-none-any.whl size=8806 sha256=21f0ee5298a20e8c5e8132cf34f51816136639335a0dd6436df8f4f6d309a83b\n",
            "  Stored in directory: /root/.cache/pip/wheels/3e/31/09/fa59cef12cdcfecc627b3d24273699f390e71828921b2cbba2\n",
            "Successfully built pathtools\n",
            "Installing collected packages: smmap, gitdb, shortuuid, setproctitle, sentry-sdk, pathtools, GitPython, docker-pycreds, wandb\n",
            "Successfully installed GitPython-3.1.29 docker-pycreds-0.4.0 gitdb-4.0.9 pathtools-0.1.2 sentry-sdk-1.9.0 setproctitle-1.3.2 shortuuid-1.0.11 smmap-5.0.0 wandb-0.13.5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "ERROR:wandb.jupyter:Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n"
          ]
        },
        {
          "data": {
            "application/javascript": "\n        window._wandbApiKey = new Promise((resolve, reject) => {\n            function loadScript(url) {\n            return new Promise(function(resolve, reject) {\n                let newScript = document.createElement(\"script\");\n                newScript.onerror = reject;\n                newScript.onload = resolve;\n                document.body.appendChild(newScript);\n                newScript.src = url;\n            });\n            }\n            loadScript(\"https://cdn.jsdelivr.net/npm/postmate/build/postmate.min.js\").then(() => {\n            const iframe = document.createElement('iframe')\n            iframe.style.cssText = \"width:0;height:0;border:none\"\n            document.body.appendChild(iframe)\n            const handshake = new Postmate({\n                container: iframe,\n                url: 'https://wandb.ai/authorize'\n            });\n            const timeout = setTimeout(() => reject(\"Couldn't auto authenticate\"), 5000)\n            handshake.then(function(child) {\n                child.on('authorize', data => {\n                    clearTimeout(timeout)\n                    resolve(data)\n                });\n            });\n            })\n        });\n    ",
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Appending key for api.wandb.ai to your netrc file: /root/.netrc\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "Tracking run with wandb version 0.13.5"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "Run data is saved locally in <code>/content/wandb/run-20221116_155526-16qazzov</code>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "Syncing run <strong><a href=\"https://wandb.ai/ahmedharbi/Hand-Gestures/runs/16qazzov\" target=\"_blank\">rosy-plasma-1</a></strong> to <a href=\"https://wandb.ai/ahmedharbi/Hand-Gestures\" target=\"_blank\">Weights & Biases</a> (<a href=\"https://wandb.me/run\" target=\"_blank\">docs</a>)<br/>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "<button onClick=\"this.nextSibling.style.display='block';this.style.display='none';\">Display W&B run</button><iframe src=\"https://wandb.ai/ahmedharbi/Hand-Gestures/runs/16qazzov?jupyter=true\" style=\"border:none;width:100%;height:420px;display:none;\"></iframe>"
            ],
            "text/plain": [
              "<wandb.sdk.wandb_run.Run at 0x7f062f9774d0>"
            ]
          },
          "execution_count": 23,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "!pip install wandb\n",
        "import wandb\n",
        "wandb.login\n",
        "wandb.init(project=\"Hand-Gestures\", entity=\"ahmedharbi\")\n",
        "# Copy\n",
        "# wandb.config = {\n",
        "#   \"learning_rate\": 0.001,\n",
        "#   \"epochs\": 100,\n",
        "#   \"batch_size\": 128\n",
        "# }\n",
        "\n",
        "# wandb.log({\"loss\": loss})"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "IZYgWhEwpMEw",
        "outputId": "25a610f9-71df-4e0e-8460-a638b084ce95"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch [1/ 20], Step [2000 / 2889], Loss: 1.1563\n",
            "Epoch [2/ 20], Step [2000 / 2889], Loss: 0.7588\n",
            "Epoch [3/ 20], Step [2000 / 2889], Loss: 1.4350\n",
            "Epoch [4/ 20], Step [2000 / 2889], Loss: 0.4897\n",
            "Epoch [5/ 20], Step [2000 / 2889], Loss: 3.7371\n",
            "Epoch [6/ 20], Step [2000 / 2889], Loss: 1.2425\n",
            "Epoch [7/ 20], Step [2000 / 2889], Loss: 1.9752\n",
            "Epoch [8/ 20], Step [2000 / 2889], Loss: 1.8471\n",
            "Epoch [9/ 20], Step [2000 / 2889], Loss: 1.6889\n",
            "Epoch [10/ 20], Step [2000 / 2889], Loss: 0.9096\n",
            "Epoch [11/ 20], Step [2000 / 2889], Loss: 1.5320\n",
            "Epoch [12/ 20], Step [2000 / 2889], Loss: 1.1999\n",
            "Epoch [13/ 20], Step [2000 / 2889], Loss: 1.6318\n",
            "Epoch [14/ 20], Step [2000 / 2889], Loss: 1.3349\n",
            "Epoch [15/ 20], Step [2000 / 2889], Loss: 2.5404\n",
            "Epoch [16/ 20], Step [2000 / 2889], Loss: 1.5569\n",
            "Epoch [17/ 20], Step [2000 / 2889], Loss: 2.3606\n",
            "Epoch [18/ 20], Step [2000 / 2889], Loss: 0.8668\n",
            "Epoch [19/ 20], Step [2000 / 2889], Loss: 1.3154\n",
            "Epoch [20/ 20], Step [2000 / 2889], Loss: 1.4108\n"
          ]
        }
      ],
      "source": [
        "model.train()\n",
        "train_model(model, train_loader, criterion, optimizer, 20, batch_size=32)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fsGmOPmxrxvt"
      },
      "outputs": [],
      "source": [
        "File_Path = '/content/drive/MyDrive/Datasets/DL_Models/sign_model_100.pth'\n",
        "\n",
        "torch.save(model.state_dict(), File_Path)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JYQB376A1cii",
        "outputId": "e1d29289-4021-4ec4-8af3-9c3d27e0e31d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy of the model 76.33757318796688 %\n",
            "Accuracy: 0 : 92.0 %\n",
            "Accuracy: 1 : 90.66666666666667 %\n",
            "Accuracy: 2 : 50.44444444444444 %\n",
            "Accuracy: 3 : 78.0 %\n",
            "Accuracy: 4 : 79.55555555555556 %\n",
            "Accuracy: 5 : 92.0 %\n",
            "Accuracy: 6 : 63.55555555555556 %\n",
            "Accuracy: 7 : 72.66666666666667 %\n",
            "Accuracy: 8 : 57.17439293598234 %\n",
            "Accuracy: 9 : 74.22222222222223 %\n",
            "Accuracy: unknown : 89.55555555555556 %\n"
          ]
        }
      ],
      "source": [
        "model.eval()\n",
        "eval_model(test_loader, model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2hu8aKBOJ9-I"
      },
      "outputs": [],
      "source": [
        "#After 11 epochs:\n",
        "\n",
        "# Accuracy of the model 12.194629517464163 %\n",
        "# Accuracy: 0 : 7.555555555555555 %\n",
        "# Accuracy: 1 : 4.444444444444445 %\n",
        "# Accuracy: 2 : 14.444444444444445 %\n",
        "# Accuracy: 3 : 8.88888888888889 %\n",
        "# Accuracy: 4 : 9.333333333333334 %\n",
        "# Accuracy: 5 : 11.11111111111111 %\n",
        "# Accuracy: 6 : 15.333333333333334 %\n",
        "# Accuracy: 7 : 26.88888888888889 %\n",
        "# Accuracy: 8 : 11.258278145695364 %\n",
        "# Accuracy: 9 : 16.666666666666668 %\n",
        "# Accuracy: unknown : 8.222222222222221 %\n",
        "\n",
        "\n",
        "#After 21 epochs:\n",
        "# Accuracy of the model 75.59055118110236 %\n",
        "# Accuracy: 0 : 89.55555555555556 %\n",
        "# Accuracy: 1 : 84.66666666666667 %\n",
        "# Accuracy: 2 : 58.0 %\n",
        "# Accuracy: 3 : 78.44444444444444 %\n",
        "# Accuracy: 4 : 72.88888888888889 %\n",
        "# Accuracy: 5 : 92.22222222222223 %\n",
        "# Accuracy: 6 : 74.88888888888889 %\n",
        "# Accuracy: 7 : 52.666666666666664 %\n",
        "# Accuracy: 8 : 68.87417218543047 %\n",
        "# Accuracy: 9 : 68.66666666666667 %\n",
        "# Accuracy: unknown : 90.66666666666667 %"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "venv_signLanguage",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.10 (default, Nov 14 2022, 12:59:47) \n[GCC 9.4.0]"
    },
    "vscode": {
      "interpreter": {
        "hash": "d4e9d720ed516f16d73fb007b435309c5c1aaf742885f7295284d0960e709c68"
      }
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "007e590b8ea948a9bc3c6b910f391929": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "022ff962f15f407588450857195ce416": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_16b0fa37df0440f8b9946e056ce91d8b",
            "placeholder": "​",
            "style": "IPY_MODEL_94f6c919190140ce891422d00cdd0af5",
            "value": " 44.7M/44.7M [00:00&lt;00:00, 99.2MB/s]"
          }
        },
        "16b0fa37df0440f8b9946e056ce91d8b": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "18709ae985d54153aa3173b7b6466747": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "2be713a0f9bd4f219494d269a287b4da": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "40b1de1b96e540928148a44cdf0a1bfa": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_faab1081d00e4f6db2b22390dafeda7a",
              "IPY_MODEL_dbcf3d908ff64860917fdf7d31d14a25",
              "IPY_MODEL_c5b2f092d77349d1bc46017920ef912f"
            ],
            "layout": "IPY_MODEL_d0927ba0eea6464ab01eaeecb554d014"
          }
        },
        "46c1644284584953a7f56ec9366c2643": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6a61b67593514ee88456810caf35a365": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "79c0b819f7b24ca3b48d23a31a794437": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "94f6c919190140ce891422d00cdd0af5": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "9a784257925d42028158ad154ab3e305": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "9f35d18d6885427c8af3357ac58a0b5c": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_18709ae985d54153aa3173b7b6466747",
            "max": 46830571,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_b312e4ce7ba845c1836e88c789a5bc89",
            "value": 46830571
          }
        },
        "a7ae4a35d40345d090b3a2a64fbc7096": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "b312e4ce7ba845c1836e88c789a5bc89": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "b659edce486d474b986f77d5111856ec": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "b6fdbd5ec7bc40faa59d89300d869fe9": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_9a784257925d42028158ad154ab3e305",
            "placeholder": "​",
            "style": "IPY_MODEL_dfd9d992a8c5472aa5b5210435945bdd",
            "value": "100%"
          }
        },
        "c457f3d5132440d088a1bb41250f091e": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_b6fdbd5ec7bc40faa59d89300d869fe9",
              "IPY_MODEL_9f35d18d6885427c8af3357ac58a0b5c",
              "IPY_MODEL_022ff962f15f407588450857195ce416"
            ],
            "layout": "IPY_MODEL_2be713a0f9bd4f219494d269a287b4da"
          }
        },
        "c5b2f092d77349d1bc46017920ef912f": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_46c1644284584953a7f56ec9366c2643",
            "placeholder": "​",
            "style": "IPY_MODEL_79c0b819f7b24ca3b48d23a31a794437",
            "value": " 44.7M/44.7M [00:00&lt;00:00, 193MB/s]"
          }
        },
        "d0927ba0eea6464ab01eaeecb554d014": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "dbcf3d908ff64860917fdf7d31d14a25": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_b659edce486d474b986f77d5111856ec",
            "max": 46830571,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_007e590b8ea948a9bc3c6b910f391929",
            "value": 46830571
          }
        },
        "dfd9d992a8c5472aa5b5210435945bdd": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "faab1081d00e4f6db2b22390dafeda7a": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_6a61b67593514ee88456810caf35a365",
            "placeholder": "​",
            "style": "IPY_MODEL_a7ae4a35d40345d090b3a2a64fbc7096",
            "value": "100%"
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}