train-motor-sentinel.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.7.6"
    },
    "colab": {
      "name": "backup_good.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "edef493ee0ea4b80a8c729c3cea33419": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_17177e56634746e595cf655dc2fa01b7",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_12655d444a9b4cc3a41b4621735a0cf6",
              "IPY_MODEL_4bbda383308547fabd02d4a4a767eb38"
            ]
          }
        },
        "17177e56634746e595cf655dc2fa01b7": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "12655d444a9b4cc3a41b4621735a0cf6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_297340d8a19a443eac8c89ad4d3593f4",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 1,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_1b29176b93f84395ad6b386037e7fd11"
          }
        },
        "4bbda383308547fabd02d4a4a767eb38": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_d616bfe4a4794e468df9fd369ba57b4f",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 1/1 [15:43&lt;00:00, 943.99s/it]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_cc47c72ca3634e9abda3f0695c6f877b"
          }
        },
        "297340d8a19a443eac8c89ad4d3593f4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "1b29176b93f84395ad6b386037e7fd11": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "d616bfe4a4794e468df9fd369ba57b4f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "cc47c72ca3634e9abda3f0695c6f877b": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "0badd4b7904a45c1b8e380084e7c41d9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_07843d5f5dce4e7e8c7cb6f361188267",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_64497f2c755843348a5b7292b800c7b6",
              "IPY_MODEL_4b86d282383b41718890d1c6eba2e557"
            ]
          }
        },
        "07843d5f5dce4e7e8c7cb6f361188267": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "64497f2c755843348a5b7292b800c7b6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_c0dcd182ccea46ca90785bfd56f204bc",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 747,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 747,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_b72183dc641843e38810f7287c71bc9b"
          }
        },
        "4b86d282383b41718890d1c6eba2e557": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_264882b62d654c8f895c07c65bf843e8",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 747/747 [15:33&lt;00:00,  1.25s/it]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_43acd0102e3d43c4960a8852a38eee67"
          }
        },
        "c0dcd182ccea46ca90785bfd56f204bc": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "b72183dc641843e38810f7287c71bc9b": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "264882b62d654c8f895c07c65bf843e8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "43acd0102e3d43c4960a8852a38eee67": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Mainakdeb/eyes-on-the-road/blob/master/train-motor-sentinel.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vPM3JELLfNyq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AsHaumoKmi30",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "from tqdm.notebook import tnrange, tqdm\n",
        "import torch\n",
        "import torchvision\n",
        "from torch.utils.data import Dataset, TensorDataset\n",
        "from torchvision.transforms import ToTensor\n",
        "import torchvision.transforms as transforms\n",
        "from torch.utils.data import DataLoader\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import torch.nn.functional as F\n",
        "from torchvision import models\n",
        "from PIL import Image"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b-qZV8xae9dv",
        "colab_type": "code",
        "outputId": "1d69d20c-bc53-488a-df3a-c5a68ae81e47",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "torch.cuda.is_available()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XjqfO6ytmpMJ",
        "colab_type": "code",
        "outputId": "13fd8194-1011-45dd-db7b-f5f7cfa3ca60",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "device"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "device(type='cuda', index=0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2kaaErrkffQj",
        "colab_type": "text"
      },
      "source": [
        "## Unzip the training data into runtime storage from Google Drive :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UuV164LlxxPD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%%bash \n",
        "mkdir foldername\n",
        "tar -xvf '/content/drive/My Drive/train.tar.xz' -C '/content/'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "njL0NArMx7BB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def image_to_nparray(path):\n",
        "    i = plt.imread(path)    \n",
        "    return(np.array(i))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IJMRoDfnZmkV",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "## A Custom class for this dataset:\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LOuAhn1ie9fg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class C_Dataset(Dataset):\n",
        "    \"\"\"custom\"\"\"\n",
        "\n",
        "    def __init__(self, csv_file_path, root_dir, transform=None):\n",
        "        \n",
        "        \"\"\"\n",
        "        Args:\n",
        "            csv_file (string): Path to the csv file with annotations.\n",
        "            root_dir (string): Directory with all the images.\n",
        "            transform (callable, optional): Optional transform to be applied\n",
        "                on a sample.\n",
        "        \"\"\"\n",
        "        \n",
        "        self.transform = transform\n",
        "        self.root_dir = root_dir\n",
        "        self.d = pd.read_csv(csv_file_path, header= None)[1:]\n",
        "    \n",
        "\n",
        "        \n",
        "    def __getitem__(self, idx): \n",
        "\n",
        "        smol_img_path = str(self.d.iloc[idx][3])\n",
        "        class_name = str(self.d.iloc[idx][2])[-1] \n",
        "        full_img_path = self.root_dir + '/' + 'c'+ class_name + '/' + smol_img_path\n",
        "        img_arr = image_to_nparray(full_img_path)\n",
        "\n",
        "        if self.transform is not None:\n",
        "            img_arr = self.transform(img_arr)\n",
        "        \n",
        "        return  img_arr, torch.tensor(int(class_name))\n",
        "        \n",
        "        \n",
        "\n",
        "    def __len__(self):\n",
        "      \n",
        "        return (len(self.d))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fezgstzxZuEs",
        "colab_type": "text"
      },
      "source": [
        "## Custom class for adding Gausian noise :\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "id": "RdJujVMTe9fp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class AddGaussianNoise(object):\n",
        "    def __init__(self, mean=0., std=1.):\n",
        "        self.std = std\n",
        "        self.mean = mean\n",
        "        \n",
        "    def __call__(self, tensor):\n",
        "        return tensor + torch.randn(tensor.size()) * self.std + self.mean\n",
        "    \n",
        "    def __repr__(self):\n",
        "        return self.__class__.__name__ + '(mean={0}, std={1})'.format(self.mean, self.std)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VPqlkYtBe9f0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data = C_Dataset('/content/drive/My Drive/balanced.csv', \n",
        "               '/content/train/',\n",
        "                transform = transforms.Compose([\n",
        "                                                transforms.ToPILImage(),\n",
        "                                                transforms.Resize((224,224), interpolation=2),\n",
        "                                                transforms.RandomRotation(20),\n",
        "                                                transforms.RandomPerspective(distortion_scale=0.1, p=0.5, interpolation=3, fill=0),\n",
        "                                                transforms.ToTensor(),\n",
        "                                                #AddGaussianNoise(0., 0.05)\n",
        "                                                ])\n",
        "                )\n",
        "\n",
        "\n",
        "val_data = C_Dataset('/content/drive/My Drive/val.csv', \n",
        "               '/content/train/',\n",
        "                transform = transforms.Compose([\n",
        "                                                transforms.ToPILImage(),\n",
        "                                                transforms.Resize((224,224), interpolation=2),\n",
        "                                                #transforms.RandomRotation(10),\n",
        "                                                transforms.ToTensor(),\n",
        "                                                #AddGaussianNoise(0., 0.5)\n",
        "                                                ])\n",
        "                )\n",
        "\n",
        "train_set, val_set = data, val_data\n",
        "\n",
        "print(len(train_set), len(val_set))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rdM_P6K0Z8xP",
        "colab_type": "text"
      },
      "source": [
        "## Plot Augmented Samples from the Train-Loader :\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h21Iw8d5e9f-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_loader_1 = torch.utils.data.DataLoader(\n",
        "    train_set\n",
        "    ,batch_size=10\n",
        "    ,shuffle=True\n",
        ")\n",
        "\n",
        "class_list = ['safe', 'text-R', 'talk on phone-R', 'text-L', 'talk on phone-L', 'radio', 'drink'\n",
        ", 'reach behind', 'hair/make-up', 'talk to passenger']\n",
        "fig, axs = plt.subplots(5, 5, figsize=(17,17))\n",
        "fig = plt.figure(figsize=(10,3))\n",
        "trans = transforms.ToPILImage()\n",
        "plt.figure(figsize=(50,50))\n",
        "\n",
        "for i in range(5):\n",
        "    for j in range(5):\n",
        "        for images, labels in (train_loader_1):\n",
        "            \n",
        "            x=np.transpose(images[0], (1, 2, 0))  \n",
        "            class_number = labels[0].item()\n",
        "            im_label=class_list[class_number]\n",
        "            axs[j,i].imshow(x, interpolation='none')\n",
        "            axs[j,i].title.set_text(im_label)\n",
        "            break\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5gcwpkUuaRuO",
        "colab_type": "text"
      },
      "source": [
        "## A function to compute the number of correct predictions :\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "etfLGxlAaOPG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_num_correct(preds, labels):\n",
        "    return preds.argmax(dim=1).eq(labels).sum().item()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X1Lk_1ukbm08",
        "colab_type": "text"
      },
      "source": [
        "## A function to plot the training metrics :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ceIo-TinblsI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot_metric(arg, frm):\n",
        "    \n",
        "    if (arg)=='loss':\n",
        "        a,b = loss_list, val_loss_list\n",
        "    elif (arg)=='accuracy':\n",
        "        a,b=acc_list, val_acc_list\n",
        "            \n",
        "    width_height = (11,3)\n",
        "    plt.figure(figsize=width_height)\n",
        "\n",
        "    #plt.axvspan(11,13, color='red', alpha=0.2)\n",
        "    \n",
        "    plt.plot(a[frm:], label='Training ' + str(arg))\n",
        "    plt.plot(b[frm:], label='Validation ' + str(arg))\n",
        "    plt.xlabel(str(arg))\n",
        "    plt.ylabel(\"Epochs\")\n",
        "    plt.grid()\n",
        "    plt.legend()\n",
        "    plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2KydeRraaj2Z",
        "colab_type": "text"
      },
      "source": [
        "## The Model :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OETqXh4EwYK9",
        "colab_type": "code",
        "outputId": "2a504fe6-4717-4110-bd02-f2385a20d6ee",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "def get_num_correct(preds, labels):\n",
        "    return preds.argmax(dim=1).eq(labels).sum().item()\n",
        "\n",
        "class Net(nn.Module):\n",
        "    \n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.conv1 = nn.Conv2d(3, 6, 7)\n",
        "        self.pool = nn.MaxPool2d(2, 2)\n",
        "        self.conv2 = nn.Conv2d(6, 16, 5)\n",
        "        self.fc1 = nn.Linear(16*34*34, 500)\n",
        "        self.fc2 = nn.Linear(500, 50)\n",
        "        self.fc3 = nn.Linear(50, 2)\n",
        "\n",
        "        \n",
        "    def forward(self, x):\n",
        "        x = self.pool(F.relu(self.conv1(x)))\n",
        "        x = self.pool(F.relu(self.conv2(x)))\n",
        "        \n",
        "        print(x.shape)\n",
        "        x = x.view(-1, 16*34*34)\n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = F.relu(self.fc2(x))\n",
        "        x = self.fc3(x)\n",
        "        return x\n",
        "\n",
        "        \n",
        "\n",
        "network = model_ft.to(device)\n",
        "print(network)\n"
      ],
      "execution_count": 120,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "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=10, bias=True)\n",
            ")\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qeTbOEtUyzsM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "batch_size=32\n",
        "learning_rate = 0.0002\n",
        "\n",
        "\n",
        "train_loader_2 = torch.utils.data.DataLoader(\n",
        "    train_set\n",
        "    ,batch_size=batch_size\n",
        "    ,shuffle=True\n",
        "\n",
        ")\n",
        "\n",
        "val_loader = torch.utils.data.DataLoader(\n",
        "    val_set, \n",
        "    batch_size=batch_size, \n",
        "    shuffle=True,\n",
        ")\n",
        "\n",
        "loss_list=[]\n",
        "acc_list=[]\n",
        "val_loss_list=[]\n",
        "val_acc_list=[]\n",
        "\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = optim.Adam(network.parameters(), lr=learning_rate) #lr=0.00003)\n",
        "scheduler = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, 'min', factor=0.1, patience=4)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Xc8-NmAHbTFS",
        "colab_type": "text"
      },
      "source": [
        "## Training Loop :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bAAig7HxyzkW",
        "colab_type": "code",
        "outputId": "598c82f3-d064-402d-87a5-8884f9694d29",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 116,
          "referenced_widgets": [
            "edef493ee0ea4b80a8c729c3cea33419",
            "17177e56634746e595cf655dc2fa01b7",
            "12655d444a9b4cc3a41b4621735a0cf6",
            "4bbda383308547fabd02d4a4a767eb38",
            "297340d8a19a443eac8c89ad4d3593f4",
            "1b29176b93f84395ad6b386037e7fd11",
            "d616bfe4a4794e468df9fd369ba57b4f",
            "cc47c72ca3634e9abda3f0695c6f877b",
            "0badd4b7904a45c1b8e380084e7c41d9",
            "07843d5f5dce4e7e8c7cb6f361188267",
            "64497f2c755843348a5b7292b800c7b6",
            "4b86d282383b41718890d1c6eba2e557",
            "c0dcd182ccea46ca90785bfd56f204bc",
            "b72183dc641843e38810f7287c71bc9b",
            "264882b62d654c8f895c07c65bf843e8",
            "43acd0102e3d43c4960a8852a38eee67"
          ]
        }
      },
      "source": [
        "for epoch in tnrange(1):\n",
        "  \n",
        "    total_loss = 0\n",
        "    total_correct = 0\n",
        "    total_loss2 = 0\n",
        "    total_correct2 = 0\n",
        "    \n",
        "    \n",
        "    with torch.no_grad():\n",
        "        network.eval()\n",
        "        for batch2 in (val_loader): # Get Batch\n",
        "\n",
        "            images2, labels2 = batch2\n",
        "            images2=images2.to(device)\n",
        "            labels2=labels2.to(device)\n",
        "\n",
        "            preds2 = network(images2) # Pass Batch\n",
        "        \n",
        "            loss2 = F.cross_entropy(preds2, labels2)\n",
        "            \n",
        "            total_loss2 += loss2.item()\n",
        "            total_correct2 += get_num_correct(preds2, labels2)\n",
        "\n",
        "    for batch in tqdm(train_loader_2): # Get Batch\n",
        "\n",
        "        optimizer.zero_grad()\n",
        "        images, labels = batch\n",
        "        images=images.to(device)\n",
        "        labels=labels.to(device)\n",
        "\n",
        "        preds = network(images)\n",
        "        \n",
        "        loss = F.cross_entropy(preds, labels)\n",
        "        loss.backward() # Calculate Gradients\n",
        "        optimizer.step()  \n",
        "        total_loss += loss.item()\n",
        "        total_correct += get_num_correct(preds, labels)\n",
        "            \n",
        "\n",
        "    loss_list.append(total_loss/(len(train_set)))\n",
        "    acc_list.append(total_correct/(len(train_set)))\n",
        "    val_loss_list.append(total_loss2/(len(val_set)))\n",
        "    val_acc_list.append(total_correct2/(len(val_set)))"
      ],
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "edef493ee0ea4b80a8c729c3cea33419",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=1.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "0badd4b7904a45c1b8e380084e7c41d9",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=747.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AmNL9l2Gby7J",
        "colab_type": "text"
      },
      "source": [
        "## The Metrics :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HFRpgfww0qHx",
        "colab_type": "code",
        "outputId": "4b320016-266c-48c6-d9d0-ae01df9278ac",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 433
        }
      },
      "source": [
        "plot_metric('loss', 0)\n",
        "plot_metric('accuracy', 0)"
      ],
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 792x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 792x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8uJYGjDFG9Zy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "1fd4d594-2008-44f0-bf39-6f604f391eae"
      },
      "source": [
        "print(\"Accuracy:\", max(val_acc_list)*100, \"%\")"
      ],
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy: 84.8 %\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CSCuqL5e33fP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "torch.save(network.state_dict(), 'model_custom.pt') "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RDxR2EyYFbUm",
        "colab_type": "text"
      },
      "source": [
        "## To do:\n",
        "* ~~train the nn~~\n",
        "* use pytorch-lr-finder"
      ]
    }
  ]
}