03_pytorch_computer_vision.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPa/ohj43qOm8uajge95oDb",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "c448da73c9eb462591cd175bf6842e8a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "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_9b245e42be1048bf908ff161d148c4c4",
              "IPY_MODEL_fc7c806d065f494cafef8a28307a97c0",
              "IPY_MODEL_53380e3fd625411bba9c9faa310822dc"
            ],
            "layout": "IPY_MODEL_11006e5a014841fd97e54f13f5c8034f"
          }
        },
        "9b245e42be1048bf908ff161d148c4c4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "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_2c5414eb0cfc48d98cd4cbce2db9ff49",
            "placeholder": "​",
            "style": "IPY_MODEL_b740259d18d646a3a1f01d22dc9389ce",
            "value": "100%"
          }
        },
        "fc7c806d065f494cafef8a28307a97c0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "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_7b99bb95dd1549ddbcd411f88d5f3733",
            "max": 3,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_c656678081c142b3ba2c3123612d33e8",
            "value": 3
          }
        },
        "53380e3fd625411bba9c9faa310822dc": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "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_4e572c52d3734a4989e22149c1e70d04",
            "placeholder": "​",
            "style": "IPY_MODEL_02b341654ea340b8a64b7da9eea1c743",
            "value": " 3/3 [00:42&lt;00:00, 12.90s/it]"
          }
        },
        "11006e5a014841fd97e54f13f5c8034f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "2c5414eb0cfc48d98cd4cbce2db9ff49": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "b740259d18d646a3a1f01d22dc9389ce": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "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": ""
          }
        },
        "7b99bb95dd1549ddbcd411f88d5f3733": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "c656678081c142b3ba2c3123612d33e8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "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": ""
          }
        },
        "4e572c52d3734a4989e22149c1e70d04": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "02b341654ea340b8a64b7da9eea1c743": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "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": ""
          }
        },
        "cb400a3cde2c4e3284aab441a5cfc121": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "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_9bc4d0e7057747bead5ac5e9e1f16936",
              "IPY_MODEL_485cafa27a5446208e0d1017c7a04fc9",
              "IPY_MODEL_ed126ee10c3c43d5b81995fb6675edaa"
            ],
            "layout": "IPY_MODEL_4decda59017443508ac3f13364b31a08"
          }
        },
        "9bc4d0e7057747bead5ac5e9e1f16936": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "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_22f5ff9bc4164ef58392e36f4b9d472a",
            "placeholder": "​",
            "style": "IPY_MODEL_ef898691efc6492cbf0497748f865af8",
            "value": "100%"
          }
        },
        "485cafa27a5446208e0d1017c7a04fc9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "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_5205bb18d7bc4d719f3226270d8ed31b",
            "max": 313,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_d04fccae44494b17b92f027d525d0289",
            "value": 313
          }
        },
        "ed126ee10c3c43d5b81995fb6675edaa": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "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_b02073132dd345b496e35862314af057",
            "placeholder": "​",
            "style": "IPY_MODEL_c003bcfd4cf042cf98e64946cc012209",
            "value": " 313/313 [00:02&lt;00:00, 222.42it/s]"
          }
        },
        "4decda59017443508ac3f13364b31a08": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "22f5ff9bc4164ef58392e36f4b9d472a": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "ef898691efc6492cbf0497748f865af8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "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": ""
          }
        },
        "5205bb18d7bc4d719f3226270d8ed31b": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "d04fccae44494b17b92f027d525d0289": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "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": ""
          }
        },
        "b02073132dd345b496e35862314af057": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "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
          }
        },
        "c003bcfd4cf042cf98e64946cc012209": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "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": ""
          }
        }
      }
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/maronem/PyTorch/blob/main/03_pytorch_computer_vision.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# PyTorch Computer Vision\n",
        "\n",
        "See reference notebook: https://www.learnpytorch.io/03_pytorch_computer_vision/"
      ],
      "metadata": {
        "id": "xa3l6C4Xt1VC"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 0. Computer vision libraries in PyTorch\n",
        "\n",
        "* [`torchvision`](https://www.learnpytorch.io/03_pytorch_computer_vision/) - base domain library for PyTorch computer vision\n",
        "* `torchvision.datasets` - get datasets and data loading functions for computer vision\n",
        "* `torchvision.models` - get pretrained computer vision models \n",
        "* `torchvision.transforms` - for manipulating your vision data (images) to be suitable for use with an ML model\n",
        "* `torch.utils.data.Dataset` - Base dataset class for pytorch\n",
        "* `torch.utils.data.DataLoader` - Creates a pipeline iterable over a dataset\n"
      ],
      "metadata": {
        "id": "yLsy7d3syX8O"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Import Pytorch\n",
        "import torch\n",
        "from torch import nn\n",
        "\n",
        "# Import torchvision\n",
        "import torchvision \n",
        "from torchvision import datasets\n",
        "from torchvision.transforms import ToTensor\n",
        "\n",
        "# Import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Check version\n",
        "print(torch.__version__)\n",
        "print(torchvision.__version__)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "znMDTTMIyX5l",
        "outputId": "43cc33e7-11c9-4fd7-b140-a1cf8f6fa802"
      },
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.0.0+cu118\n",
            "0.15.1+cu118\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1. Getting a dataset\n",
        "\n",
        "We will be using [FashionMNSIT](https://pytorch.org/vision/stable/generated/torchvision.datasets.FashionMNIST.html#torchvision.datasets.FashionMNIST)\n",
        "\n",
        "To use dataset you need to:\n",
        "\n",
        "1. Access the data using imagenet using torchvision: `torchvision.datasets.ImageNet('PATH')`\n",
        "2. Turn the data into a data loader: `torch.utils.data.dataloader()`"
      ],
      "metadata": {
        "id": "I9Hiat2QyX3B"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Setup training data\n",
        "train_data = datasets.FashionMNIST(\n",
        "    root='data', #where to download data to\n",
        "    train=True, # do we want trianing dataset?\n",
        "    download=True, #do we want to download dataset?\n",
        "    transform=ToTensor(), # how do we want to transform the data?\n",
        "    target_transform=None # how do we want to transform the labels/targets?\n",
        ")\n",
        "\n",
        "test_data = datasets.FashionMNIST(\n",
        "    root='data',\n",
        "    train=False,\n",
        "    download=True,\n",
        "    transform=ToTensor(),\n",
        "    target_transform=None\n",
        ")"
      ],
      "metadata": {
        "id": "SURkxKBYyX0Z"
      },
      "execution_count": 136,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "len(train_data), len(test_data)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Io1TcqSKyXx-",
        "outputId": "1eb40413-9280-456a-d90f-a3302dc80e98"
      },
      "execution_count": 137,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 10000)"
            ]
          },
          "metadata": {},
          "execution_count": 137
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# See the first training example\n",
        "image, label=train_data[0]\n",
        "image, label"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7jZke0WFyXvZ",
        "outputId": "9e0b58e1-c968-4e73-b5bb-3df798c55851"
      },
      "execution_count": 138,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(tensor([[[0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0039, 0.0000, 0.0000, 0.0510,\n",
              "           0.2863, 0.0000, 0.0000, 0.0039, 0.0157, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0039, 0.0039, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0118, 0.0000, 0.1412, 0.5333,\n",
              "           0.4980, 0.2431, 0.2118, 0.0000, 0.0000, 0.0000, 0.0039, 0.0118,\n",
              "           0.0157, 0.0000, 0.0000, 0.0118],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0235, 0.0000, 0.4000, 0.8000,\n",
              "           0.6902, 0.5255, 0.5647, 0.4824, 0.0902, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0471, 0.0392, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.6078, 0.9255,\n",
              "           0.8118, 0.6980, 0.4196, 0.6118, 0.6314, 0.4275, 0.2510, 0.0902,\n",
              "           0.3020, 0.5098, 0.2824, 0.0588],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0039, 0.0000, 0.2706, 0.8118, 0.8745,\n",
              "           0.8549, 0.8471, 0.8471, 0.6392, 0.4980, 0.4745, 0.4784, 0.5725,\n",
              "           0.5529, 0.3451, 0.6745, 0.2588],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0039, 0.0039, 0.0039, 0.0000, 0.7843, 0.9098, 0.9098,\n",
              "           0.9137, 0.8980, 0.8745, 0.8745, 0.8431, 0.8353, 0.6431, 0.4980,\n",
              "           0.4824, 0.7686, 0.8980, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.7176, 0.8824, 0.8471,\n",
              "           0.8745, 0.8941, 0.9216, 0.8902, 0.8784, 0.8706, 0.8784, 0.8667,\n",
              "           0.8745, 0.9608, 0.6784, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.7569, 0.8941, 0.8549,\n",
              "           0.8353, 0.7765, 0.7059, 0.8314, 0.8235, 0.8275, 0.8353, 0.8745,\n",
              "           0.8627, 0.9529, 0.7922, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0039, 0.0118, 0.0000, 0.0471, 0.8588, 0.8627, 0.8314,\n",
              "           0.8549, 0.7529, 0.6627, 0.8902, 0.8157, 0.8549, 0.8784, 0.8314,\n",
              "           0.8863, 0.7725, 0.8196, 0.2039],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0235, 0.0000, 0.3882, 0.9569, 0.8706, 0.8627,\n",
              "           0.8549, 0.7961, 0.7765, 0.8667, 0.8431, 0.8353, 0.8706, 0.8627,\n",
              "           0.9608, 0.4667, 0.6549, 0.2196],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0157, 0.0000, 0.0000, 0.2157, 0.9255, 0.8941, 0.9020,\n",
              "           0.8941, 0.9412, 0.9098, 0.8353, 0.8549, 0.8745, 0.9176, 0.8510,\n",
              "           0.8510, 0.8196, 0.3608, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0039, 0.0157, 0.0235, 0.0275, 0.0078, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.9294, 0.8863, 0.8510, 0.8745,\n",
              "           0.8706, 0.8588, 0.8706, 0.8667, 0.8471, 0.8745, 0.8980, 0.8431,\n",
              "           0.8549, 1.0000, 0.3020, 0.0000],\n",
              "          [0.0000, 0.0118, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.2431, 0.5686, 0.8000, 0.8941, 0.8118, 0.8353, 0.8667,\n",
              "           0.8549, 0.8157, 0.8275, 0.8549, 0.8784, 0.8745, 0.8588, 0.8431,\n",
              "           0.8784, 0.9569, 0.6235, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0706, 0.1725, 0.3216, 0.4196,\n",
              "           0.7412, 0.8941, 0.8627, 0.8706, 0.8510, 0.8863, 0.7843, 0.8039,\n",
              "           0.8275, 0.9020, 0.8784, 0.9176, 0.6902, 0.7373, 0.9804, 0.9725,\n",
              "           0.9137, 0.9333, 0.8431, 0.0000],\n",
              "          [0.0000, 0.2235, 0.7333, 0.8157, 0.8784, 0.8667, 0.8784, 0.8157,\n",
              "           0.8000, 0.8392, 0.8157, 0.8196, 0.7843, 0.6235, 0.9608, 0.7569,\n",
              "           0.8078, 0.8745, 1.0000, 1.0000, 0.8667, 0.9176, 0.8667, 0.8275,\n",
              "           0.8627, 0.9098, 0.9647, 0.0000],\n",
              "          [0.0118, 0.7922, 0.8941, 0.8784, 0.8667, 0.8275, 0.8275, 0.8392,\n",
              "           0.8039, 0.8039, 0.8039, 0.8627, 0.9412, 0.3137, 0.5882, 1.0000,\n",
              "           0.8980, 0.8667, 0.7373, 0.6039, 0.7490, 0.8235, 0.8000, 0.8196,\n",
              "           0.8706, 0.8941, 0.8824, 0.0000],\n",
              "          [0.3843, 0.9137, 0.7765, 0.8235, 0.8706, 0.8980, 0.8980, 0.9176,\n",
              "           0.9765, 0.8627, 0.7608, 0.8431, 0.8510, 0.9451, 0.2549, 0.2863,\n",
              "           0.4157, 0.4588, 0.6588, 0.8588, 0.8667, 0.8431, 0.8510, 0.8745,\n",
              "           0.8745, 0.8784, 0.8980, 0.1137],\n",
              "          [0.2941, 0.8000, 0.8314, 0.8000, 0.7569, 0.8039, 0.8275, 0.8824,\n",
              "           0.8471, 0.7255, 0.7725, 0.8078, 0.7765, 0.8353, 0.9412, 0.7647,\n",
              "           0.8902, 0.9608, 0.9373, 0.8745, 0.8549, 0.8314, 0.8196, 0.8706,\n",
              "           0.8627, 0.8667, 0.9020, 0.2627],\n",
              "          [0.1882, 0.7961, 0.7176, 0.7608, 0.8353, 0.7725, 0.7255, 0.7451,\n",
              "           0.7608, 0.7529, 0.7922, 0.8392, 0.8588, 0.8667, 0.8627, 0.9255,\n",
              "           0.8824, 0.8471, 0.7804, 0.8078, 0.7294, 0.7098, 0.6941, 0.6745,\n",
              "           0.7098, 0.8039, 0.8078, 0.4510],\n",
              "          [0.0000, 0.4784, 0.8588, 0.7569, 0.7020, 0.6706, 0.7176, 0.7686,\n",
              "           0.8000, 0.8235, 0.8353, 0.8118, 0.8275, 0.8235, 0.7843, 0.7686,\n",
              "           0.7608, 0.7490, 0.7647, 0.7490, 0.7765, 0.7529, 0.6902, 0.6118,\n",
              "           0.6549, 0.6941, 0.8235, 0.3608],\n",
              "          [0.0000, 0.0000, 0.2902, 0.7412, 0.8314, 0.7490, 0.6863, 0.6745,\n",
              "           0.6863, 0.7098, 0.7255, 0.7373, 0.7412, 0.7373, 0.7569, 0.7765,\n",
              "           0.8000, 0.8196, 0.8235, 0.8235, 0.8275, 0.7373, 0.7373, 0.7608,\n",
              "           0.7529, 0.8471, 0.6667, 0.0000],\n",
              "          [0.0078, 0.0000, 0.0000, 0.0000, 0.2588, 0.7843, 0.8706, 0.9294,\n",
              "           0.9373, 0.9490, 0.9647, 0.9529, 0.9569, 0.8667, 0.8627, 0.7569,\n",
              "           0.7490, 0.7020, 0.7137, 0.7137, 0.7098, 0.6902, 0.6510, 0.6588,\n",
              "           0.3882, 0.2275, 0.0000, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.1569,\n",
              "           0.2392, 0.1725, 0.2824, 0.1608, 0.1373, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000],\n",
              "          [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n",
              "           0.0000, 0.0000, 0.0000, 0.0000]]]),\n",
              " 9)"
            ]
          },
          "metadata": {},
          "execution_count": 138
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "image.shape #C, H, W, C=1 means image is grayscale "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t2mKB6iDJahQ",
        "outputId": "a4a4a55b-cf99-4f88-96fa-5fb3bfedfd11"
      },
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([1, 28, 28])"
            ]
          },
          "metadata": {},
          "execution_count": 139
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class_names = train_data.classes\n",
        "class_names"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "klLFxl5DyXsz",
        "outputId": "cce8a7b4-2227-4f67-aa68-815de73f08f4"
      },
      "execution_count": 140,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['T-shirt/top',\n",
              " 'Trouser',\n",
              " 'Pullover',\n",
              " 'Dress',\n",
              " 'Coat',\n",
              " 'Sandal',\n",
              " 'Shirt',\n",
              " 'Sneaker',\n",
              " 'Bag',\n",
              " 'Ankle boot']"
            ]
          },
          "metadata": {},
          "execution_count": 140
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class_to_idx = train_data.class_to_idx\n",
        "class_to_idx"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3zvb-A_cyXqR",
        "outputId": "3582615c-21f5-45d5-9d9c-04cff6b6de88"
      },
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'T-shirt/top': 0,\n",
              " 'Trouser': 1,\n",
              " 'Pullover': 2,\n",
              " 'Dress': 3,\n",
              " 'Coat': 4,\n",
              " 'Sandal': 5,\n",
              " 'Shirt': 6,\n",
              " 'Sneaker': 7,\n",
              " 'Bag': 8,\n",
              " 'Ankle boot': 9}"
            ]
          },
          "metadata": {},
          "execution_count": 141
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check the shape of our iamge\n",
        "print(f\"Image Shape: {image.shape} -> C,H,W\")\n",
        "print(f\"Image Label: {class_names[label]}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GmriW3TCyXnw",
        "outputId": "c70c615c-3d63-4d23-93ca-b894d83015c7"
      },
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Image Shape: torch.Size([1, 28, 28]) -> C,H,W\n",
            "Image Label: Ankle boot\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 1.2 Visualize our data "
      ],
      "metadata": {
        "id": "Iy8izX8kyXk-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "image, label = train_data[0]\n",
        "print(f\"Image Shape: {image.shape}\")\n",
        "plt.imshow(image.squeeze()) #squeeze() removes singular dimension of color, matplotlib expects only H+W\n",
        "plt.title(f\"{class_names[label]} - {label}\");"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 469
        },
        "id": "qdel7Q2HE7wr",
        "outputId": "f08aa283-c6bd-49b1-b2e2-8b32d197188c"
      },
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Image Shape: torch.Size([1, 28, 28])\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot in greyscale\n",
        "plt.imshow(image.squeeze(), cmap=\"gray\");\n",
        "plt.title(f\"{class_names[label]} - {label}\");"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "id": "_XEV0rqNE7t4",
        "outputId": "4993c8d4-975e-404f-e306-abb69471f548"
      },
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot more images\n",
        "torch.manual_seed(42)\n",
        "fig = plt.figure(figsize=(9,9)) #image size 9x9\n",
        "rows, cols = 4,4 # set a 4x4 grid of images\n",
        "\n",
        "for i in range(1, rows*cols+1):\n",
        "  random_idx = torch.randint(0, len(train_data), size=[1]).item()\n",
        "  img, label = train_data[random_idx] # set img/label to the random indexes\n",
        "\n",
        "  fig.add_subplot(rows, cols, i)\n",
        "  plt.imshow(img.squeeze(), cmap=\"gray\")\n",
        "  plt.title(f\"{class_names[label]} - {label}\")\n",
        "  plt.axis(False);\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 752
        },
        "id": "4tRoiha0E7rH",
        "outputId": "c3aa93ad-cd87-4c45-fcb6-0028d054cc37"
      },
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 900x900 with 16 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2. Prepare DataLoader\n",
        "\n",
        "Right now our data is in the form of PyTorch DataSets\n",
        "\n",
        "DataLoader turns our dataset into a Python iterable.\n",
        "\n",
        "More specifically, we want to turn our data into batches (or mini batches) Why?\n",
        "\n",
        "1. It is more computationally efficient and memory compatible. So we break it down into batches such as 32 images/batch.\n",
        "2. It gives our neural network more chances to update its gradients per epoch (updates parameters 1x every 32 images instead of 1x per 60,000) - [minibatch gradient descent](https://towardsdatascience.com/batch-mini-batch-stochastic-gradient-descent-7a62ecba642a)\n",
        "\n"
      ],
      "metadata": {
        "id": "oMLjOAHXE7oZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_data, test_data"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "olJPLOw9E7lL",
        "outputId": "46448a8d-71bc-4cb3-9dec-3c6f6c2d59ee"
      },
      "execution_count": 79,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(Dataset FashionMNIST\n",
              "     Number of datapoints: 60000\n",
              "     Root location: data\n",
              "     Split: Train\n",
              "     StandardTransform\n",
              " Transform: ToTensor(),\n",
              " Dataset FashionMNIST\n",
              "     Number of datapoints: 10000\n",
              "     Root location: data\n",
              "     Split: Test\n",
              "     StandardTransform\n",
              " Transform: ToTensor())"
            ]
          },
          "metadata": {},
          "execution_count": 79
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Turn datasets into DataLoaders\n",
        "from torch.utils.data import DataLoader\n",
        "\n",
        "# Set batch size hyperparameter\n",
        "BATCH_SIZE = 32\n",
        "\n",
        "train_dataloader = DataLoader(dataset = train_data,\n",
        "                              batch_size=BATCH_SIZE,\n",
        "                              shuffle=True)\n",
        "\n",
        "test_dataloader = DataLoader(dataset=test_data,\n",
        "                             batch_size = BATCH_SIZE,\n",
        "                             shuffle=False) # Don't necessarily need to shuffle test data\n",
        "\n",
        "# Let's check out what we've created\n",
        "print(f\"Dataloaders: {train_dataloader, test_dataloader}\") \n",
        "print(f\"Length of train dataloader: {len(train_dataloader)} batches of {BATCH_SIZE}\")\n",
        "print(f\"Length of test dataloader: {len(test_dataloader)} batches of {BATCH_SIZE}\")                         "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lTVsf4fYE7iR",
        "outputId": "856e7a8b-6df1-434e-e38c-369469d021ed"
      },
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Dataloaders: (<torch.utils.data.dataloader.DataLoader object at 0x7f21c68dc0d0>, <torch.utils.data.dataloader.DataLoader object at 0x7f21c71c57f0>)\n",
            "Length of train dataloader: 1875 batches of 32\n",
            "Length of test dataloader: 313 batches of 32\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check what is inside the training dataloader\n",
        "train_features_batch, train_labels_batch = next(iter(train_dataloader))\n",
        "\n",
        "train_features_batch.shape, train_labels_batch.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-B8QgOOPE7fQ",
        "outputId": "cd28f3c2-e31c-486e-e039-901c072c00c1"
      },
      "execution_count": 81,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(torch.Size([32, 1, 28, 28]), torch.Size([32]))"
            ]
          },
          "metadata": {},
          "execution_count": 81
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Visualize single image from a batch\n",
        "#torch.manual_seed(42)\n",
        "random_idx = torch.randint(0, len(train_features_batch), size=[1]).item()\n",
        "img, label = train_features_batch[random_idx], train_labels_batch[random_idx]\n",
        "\n",
        "plt.imshow(img.squeeze(), cmap=\"gray\")\n",
        "plt.title(class_names[label])\n",
        "plt.axis(False)\n",
        "print(f\"Image Size: {img.shape}\")\n",
        "print(f\"Label: {label}, label size: {label.shape}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 463
        },
        "id": "T1I_N4cLM67s",
        "outputId": "fc0ab40c-927b-496d-9287-f4587e17583c"
      },
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Image Size: torch.Size([1, 28, 28])\n",
            "Label: 8, label size: torch.Size([])\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3. Model 0: Build a baseline model\n",
        "\n",
        "When starting to build a series of machine learning modeling experiments, its best practice to start with a baseline model.\n",
        "\n",
        "A baseline model is a simple model you will try and improve upon with subsequent models/experiments.\n",
        "\n",
        "In other words: start simply and add complexity when needed"
      ],
      "metadata": {
        "id": "aXrYTxokQeyn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a flatten layer\n",
        "flatten_model = nn.Flatten()\n",
        "\n",
        "# Get a single sample\n",
        "x = train_features_batch[0]\n",
        "output = flatten_model(x)\n",
        "\n",
        "print(f\"Original shape: {x.shape}\") # -> [C,H,W] (3 dim)\n",
        "print(f\"Flattened shape: {output.shape}\") # -> [C, H*W] (2 dim)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sKBFyXpLQewC",
        "outputId": "afc0daa9-878a-41a1-bb76-8c1ceccc3f86"
      },
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Original shape: torch.Size([1, 28, 28])\n",
            "Flattened shape: torch.Size([1, 784])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from torch import nn\n",
        "\n",
        "class FashionMNISTModelV0(nn.Module):\n",
        "  def __init__(self,\n",
        "               input_shape: int,\n",
        "               hidden_units: int,\n",
        "               output_shape: int):\n",
        "    super().__init__()\n",
        "\n",
        "    self.layer_stack = nn.Sequential(\n",
        "        nn.Flatten(),\n",
        "        nn.Linear(in_features=input_shape,\n",
        "                  out_features=hidden_units),\n",
        "        nn.Linear(in_features=hidden_units,\n",
        "                  out_features=output_shape)\n",
        "    )\n",
        "\n",
        "  def forward(self,x):\n",
        "    return self.layer_stack(x)\n"
      ],
      "metadata": {
        "id": "VUVqwuAtQeth"
      },
      "execution_count": 84,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "torch.manual_seed(42)\n",
        "\n",
        "# setup model with input parameters\n",
        "model_0 = FashionMNISTModelV0(\n",
        "    input_shape=784, # or 28x28\n",
        "    hidden_units=10, # how many units in hidden layer\n",
        "    output_shape=len(class_names) # one for every class\n",
        ").to(\"cpu\")\n",
        "\n",
        "model_0"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0HDgah0TQeq4",
        "outputId": "fbe6e1ce-2439-4202-bd75-bf7162bd6470"
      },
      "execution_count": 85,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "FashionMNISTModelV0(\n",
              "  (layer_stack): Sequential(\n",
              "    (0): Flatten(start_dim=1, end_dim=-1)\n",
              "    (1): Linear(in_features=784, out_features=10, bias=True)\n",
              "    (2): Linear(in_features=10, out_features=10, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 85
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# test on dummy data\n",
        "dummy_x = torch.rand([1,1,28,28])\n",
        "\n",
        "model_0(dummy_x), model_0(dummy_x).shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LSJz0XGN6jri",
        "outputId": "46b8384c-4887-4938-a953-c05b3f91043c"
      },
      "execution_count": 86,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(tensor([[-0.0315,  0.3171,  0.0531, -0.2525,  0.5959,  0.2112,  0.3233,  0.2694,\n",
              "          -0.1004,  0.0157]], grad_fn=<AddmmBackward0>),\n",
              " torch.Size([1, 10]))"
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_0.state_dict()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZYtkmEY4-8GP",
        "outputId": "bfe30250-7682-43d0-e3d2-63ffdeb83077"
      },
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "OrderedDict([('layer_stack.1.weight',\n",
              "              tensor([[ 0.0273,  0.0296, -0.0084,  ..., -0.0142,  0.0093,  0.0135],\n",
              "                      [-0.0188, -0.0354,  0.0187,  ..., -0.0106, -0.0001,  0.0115],\n",
              "                      [-0.0008,  0.0017,  0.0045,  ..., -0.0127, -0.0188,  0.0059],\n",
              "                      ...,\n",
              "                      [-0.0116,  0.0273, -0.0344,  ...,  0.0176,  0.0283, -0.0011],\n",
              "                      [-0.0230,  0.0257,  0.0291,  ..., -0.0187, -0.0087,  0.0001],\n",
              "                      [ 0.0176, -0.0147,  0.0053,  ..., -0.0336, -0.0221,  0.0205]])),\n",
              "             ('layer_stack.1.bias',\n",
              "              tensor([-0.0093,  0.0283, -0.0033,  0.0255,  0.0017,  0.0037, -0.0302, -0.0123,\n",
              "                       0.0018,  0.0163])),\n",
              "             ('layer_stack.2.weight',\n",
              "              tensor([[ 0.0614, -0.0687,  0.0021,  0.2718,  0.2109,  0.1079, -0.2279, -0.1063,\n",
              "                        0.2019,  0.2847],\n",
              "                      [-0.1495,  0.1344, -0.0740,  0.2006, -0.0475, -0.2514, -0.3130, -0.0118,\n",
              "                        0.0932, -0.1864],\n",
              "                      [ 0.2488,  0.1500,  0.1907,  0.1457, -0.3050, -0.0580,  0.1643,  0.1565,\n",
              "                       -0.2877, -0.1792],\n",
              "                      [ 0.2305, -0.2618,  0.2397, -0.0610,  0.0232,  0.1542,  0.0851, -0.2027,\n",
              "                        0.1030, -0.2715],\n",
              "                      [-0.1596, -0.0555, -0.0633,  0.2302, -0.1726,  0.2654,  0.1473,  0.1029,\n",
              "                        0.2252, -0.2160],\n",
              "                      [-0.2725,  0.0118,  0.1559,  0.1596,  0.0132,  0.3024,  0.1124,  0.1366,\n",
              "                       -0.1533,  0.0965],\n",
              "                      [-0.1184, -0.2555, -0.2057, -0.1909, -0.0477, -0.1324,  0.2905,  0.1307,\n",
              "                       -0.2629,  0.0133],\n",
              "                      [ 0.2727, -0.0127,  0.0513,  0.0863, -0.1043, -0.2047, -0.1185, -0.0825,\n",
              "                        0.2488, -0.2571],\n",
              "                      [ 0.0425, -0.1209, -0.0336, -0.0281, -0.1227,  0.0730,  0.0747, -0.1816,\n",
              "                        0.1943,  0.2853],\n",
              "                      [-0.1310,  0.0645, -0.1171,  0.2168, -0.0245, -0.2820,  0.0736,  0.2621,\n",
              "                        0.0012, -0.0810]])),\n",
              "             ('layer_stack.2.bias',\n",
              "              tensor([-0.0087,  0.1791,  0.2712, -0.0791,  0.1685,  0.1762,  0.2825,  0.2266,\n",
              "                      -0.2612, -0.2613]))])"
            ]
          },
          "metadata": {},
          "execution_count": 87
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 3.1 Setup loss fn, optimizer, and evaluation metrics\n",
        "\n",
        "* loss fn - since we are working with mutliclass data, our loss fn will be `nn.CrossEntropyLoss()`\n",
        "* Optimizer - our optimizer `torch.optim.SGD()` for stochastic gradient descent\n",
        "* Evaluation metrics - since we are working on classification, lets use accuracy"
      ],
      "metadata": {
        "id": "sNHX7XJmpAZa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import requests\n",
        "from pathlib import Path\n",
        "\n",
        "# Download helper functions from Learn PyTorch repo\n",
        "\n",
        "if Path(\"helper_functions.py\").is_file():\n",
        "  print(\"helper_functions.py  already exists, skipping download.\")\n",
        "else:\n",
        "  print(\"Downloading helper_functions.py ...\")\n",
        "  request = requests.get(\"https://raw.githubusercontent.com/mrdbourke/pytorch-deep-learning/main/helper_functions.py\")\n",
        "  with open(\"helper_functions.py\", \"wb\") as f:\n",
        "    f.write(request.content)\n",
        "\n",
        "from helper_functions import accuracy_fn"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Lm70zJy-pBWL",
        "outputId": "f4e557ea-404f-4871-89a0-d743e399ef3c"
      },
      "execution_count": 88,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "helper_functions.py  already exists, skipping download.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Setup loss function and optimizer\n",
        "loss_fn = nn.CrossEntropyLoss()\n",
        "\n",
        "optimizer = torch.optim.SGD(params=model_0.parameters(),\n",
        "                            lr=0.1)"
      ],
      "metadata": {
        "id": "S2_U6KKBsy3M"
      },
      "execution_count": 89,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 3.2 Creating a function to time our experiments\n",
        "\n",
        "Machine learning is very experimental, two of the main things you'll often want to track is:\n",
        "\n",
        "1. Your model's performance (loss and accuracy, etc.)\n",
        "2. How fast it runs \n",
        "\n",
        "`timeit` - https://docs.python.org/3/library/timeit.html"
      ],
      "metadata": {
        "id": "b2Os1oyQucyM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from timeit import default_timer as timer\n",
        "\n",
        "def print_train_time(start: float,\n",
        "                     end: float,\n",
        "                     device: torch.device = None):\n",
        "  \"\"\"Prints difference between start and end time\"\"\"\n",
        "  total_time = end-start\n",
        "  print(f\"Train time on {device}: {total_time:.3f} seconds\")\n",
        "  return total_time"
      ],
      "metadata": {
        "id": "QYWedw45ucvd"
      },
      "execution_count": 90,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Timer example\n",
        "\n",
        "start_time = timer()\n",
        "# some code...\n",
        "end_time = timer()\n",
        "print_train_time(start=start_time, \n",
        "                 end=end_time,\n",
        "                 device=\"cpu\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MGg-wGSUucsl",
        "outputId": "dda1db3f-d162-433a-f875-ed304b57e17e"
      },
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train time on cpu: 0.000 seconds\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4.426300029081176e-05"
            ]
          },
          "metadata": {},
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 3.3 Creating a training and testing loop on batches of data\n",
        "\n",
        "**NOTE: The optimizer will update our model's parameters once per batch instead of per epoch**\n",
        "\n",
        "1. Loop through epochs\n",
        "2. Loop through training batches, perform training steps, calculate the trian loss *per batch*\n",
        "3. Loop through testing batches, perform test steps, calculate the test loss *per batch*\n",
        "4. Print out what's happening \n",
        "5. Time it all (for fun)\n",
        "\n",
        "For more on tqdm progress bar: https://tqdm.github.io/\n"
      ],
      "metadata": {
        "id": "U_zlR3YrwLPW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install torchmetrics"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Z_2PJN465_c4",
        "outputId": "787929d5-42db-46a6-e44f-a1242840d83b"
      },
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: torchmetrics in /usr/local/lib/python3.9/dist-packages (0.11.4)\n",
            "Requirement already satisfied: torch>=1.8.1 in /usr/local/lib/python3.9/dist-packages (from torchmetrics) (2.0.0+cu118)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.9/dist-packages (from torchmetrics) (23.0)\n",
            "Requirement already satisfied: numpy>=1.17.2 in /usr/local/lib/python3.9/dist-packages (from torchmetrics) (1.22.4)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.9/dist-packages (from torch>=1.8.1->torchmetrics) (4.5.0)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.9/dist-packages (from torch>=1.8.1->torchmetrics) (1.11.1)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.9/dist-packages (from torch>=1.8.1->torchmetrics) (3.10.7)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.9/dist-packages (from torch>=1.8.1->torchmetrics) (3.1.2)\n",
            "Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.9/dist-packages (from torch>=1.8.1->torchmetrics) (2.0.0)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.9/dist-packages (from torch>=1.8.1->torchmetrics) (3.0)\n",
            "Requirement already satisfied: cmake in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.8.1->torchmetrics) (3.25.2)\n",
            "Requirement already satisfied: lit in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.8.1->torchmetrics) (16.0.0)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.9/dist-packages (from jinja2->torch>=1.8.1->torchmetrics) (2.1.2)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.9/dist-packages (from sympy->torch>=1.8.1->torchmetrics) (1.3.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from torchmetrics import Accuracy\n",
        "\n",
        "# Import tqdm for progress bar\n",
        "from tqdm.auto import tqdm\n",
        "\n",
        "# Set the manual seed and start timer\n",
        "torch.manual_seed(42)\n",
        "train_time_start_cpu = timer()\n",
        "\n",
        "# Set # epochs (keep this small for faster computing time)\n",
        "epochs = 3\n",
        "\n",
        "# Create a training loop\n",
        "for epoch in tqdm(range(epochs)):\n",
        "  print(f\"Epoch: {epoch}\\n---\")\n",
        "\n",
        "  ### TRAINING\n",
        "  train_loss=0\n",
        "\n",
        "  # Add loop to loop through batches\n",
        "  for batch, (X,y) in enumerate(train_dataloader):\n",
        "    model_0.train()\n",
        "   \n",
        "    # 1. Forward pass\n",
        "    y_pred = model_0(X) \n",
        "\n",
        "    # 2. Calc loss\n",
        "    loss = loss_fn(y_pred, y) \n",
        "    train_loss += loss # accumulate loss each batch\n",
        "\n",
        "    # 3. Optimizer zero grad\n",
        "    optimizer.zero_grad()\n",
        "\n",
        "    # 4. Back propagation\n",
        "    loss.backward()\n",
        "\n",
        "    # 5. Step optimizer\n",
        "    optimizer.step()\n",
        "\n",
        "    # Print out whats happening\n",
        "    if batch % 400 == 0:\n",
        "      print(f\"Looked at {batch * len(X)} / {len(train_dataloader.dataset)} samples\")\n",
        "\n",
        "  # Divide total train loss by length of train dataloader\n",
        "  train_loss /= len(train_dataloader)\n",
        "\n",
        "  ### TESTING\n",
        "  test_loss, test_acc = 0,0\n",
        "  model_0.eval()\n",
        "  with torch.inference_mode():\n",
        "    for X_test,y_test in test_dataloader:\n",
        "      # 1. Foward pass\n",
        "      test_pred = model_0(X_test) #logits \n",
        "\n",
        "      # 2. Calculate loss (cumulatively)\n",
        "      test_loss += loss_fn(test_pred, y_test)\n",
        "\n",
        "      # 3. Calculate accuracy\n",
        "      acc_fn = Accuracy(task=\"multiclass\", num_classes=10)\n",
        "      test_acc += acc_fn(test_pred.argmax(dim=1), y_test) #argmax to find logit value with highest index\n",
        "\n",
        "    # Calculate test loss average per batch\n",
        "    test_loss /= len(test_dataloader)\n",
        "\n",
        "    # Calculate test acc average per batch\n",
        "    test_acc /= len(test_dataloader)\n",
        "\n",
        "  # Print out whats happening\n",
        "  print(f\"\\nTrain loss: {train_loss:.4f}, Test loss {test_loss:.4f}, Test Acc: {test_acc:.4f}\")\n",
        "\n",
        "# Calculate training time\n",
        "train_time_end_cpu = timer()\n",
        "total_train_time_model_0 = print_train_time(start=train_time_start_cpu,\n",
        "                                            end=train_time_end_cpu,\n",
        "                                            device=str(next(model_0.parameters()))) #provides str of device - see exmple below"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 674,
          "referenced_widgets": [
            "c448da73c9eb462591cd175bf6842e8a",
            "9b245e42be1048bf908ff161d148c4c4",
            "fc7c806d065f494cafef8a28307a97c0",
            "53380e3fd625411bba9c9faa310822dc",
            "11006e5a014841fd97e54f13f5c8034f",
            "2c5414eb0cfc48d98cd4cbce2db9ff49",
            "b740259d18d646a3a1f01d22dc9389ce",
            "7b99bb95dd1549ddbcd411f88d5f3733",
            "c656678081c142b3ba2c3123612d33e8",
            "4e572c52d3734a4989e22149c1e70d04",
            "02b341654ea340b8a64b7da9eea1c743"
          ]
        },
        "id": "iJx4cV7_wLM8",
        "outputId": "b749bcc6-71ab-4bf3-ea7e-bca817d1825e"
      },
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/3 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "c448da73c9eb462591cd175bf6842e8a"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0\n",
            "---\n",
            "Looked at 0 / 60000 samples\n",
            "Looked at 12800 / 60000 samples\n",
            "Looked at 25600 / 60000 samples\n",
            "Looked at 38400 / 60000 samples\n",
            "Looked at 51200 / 60000 samples\n",
            "\n",
            "Train loss: 0.5904, Test loss 0.5095, Test Acc: 0.8204\n",
            "Epoch: 1\n",
            "---\n",
            "Looked at 0 / 60000 samples\n",
            "Looked at 12800 / 60000 samples\n",
            "Looked at 25600 / 60000 samples\n",
            "Looked at 38400 / 60000 samples\n",
            "Looked at 51200 / 60000 samples\n",
            "\n",
            "Train loss: 0.4763, Test loss 0.4799, Test Acc: 0.8320\n",
            "Epoch: 2\n",
            "---\n",
            "Looked at 0 / 60000 samples\n",
            "Looked at 12800 / 60000 samples\n",
            "Looked at 25600 / 60000 samples\n",
            "Looked at 38400 / 60000 samples\n",
            "Looked at 51200 / 60000 samples\n",
            "\n",
            "Train loss: 0.4550, Test loss 0.4766, Test Acc: 0.8343\n",
            "Train time on Parameter containing:\n",
            "tensor([[ 0.0273,  0.0298, -0.0065,  ...,  0.0071,  0.0091,  0.0170],\n",
            "        [-0.0189, -0.0355,  0.0155,  ...,  0.0161,  0.0120,  0.0117],\n",
            "        [-0.0008,  0.0020,  0.0032,  ...,  0.0058, -0.0146,  0.0084],\n",
            "        ...,\n",
            "        [-0.0116,  0.0274, -0.0323,  ...,  0.0527,  0.0434,  0.0006],\n",
            "        [-0.0231,  0.0252,  0.0236,  ..., -0.0892, -0.0385, -0.0071],\n",
            "        [ 0.0175, -0.0148,  0.0027,  ..., -0.1155, -0.0674,  0.0125]],\n",
            "       requires_grad=True): 42.424 seconds\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# How to print out str of what device our model is on\n",
        "str(next(model_0.parameters()).device)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "3EaYMTNDwLKa",
        "outputId": "e9f6a8db-5451-4365-ddc1-7bcc7f69f091"
      },
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'cpu'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 4. Make predictions and get model_0 results"
      ],
      "metadata": {
        "id": "MQQEeUuq2yDK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "## 4. Setup device agnostic code (for using GPU if there is one)\n",
        "\n",
        "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
        "device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "ykCNJQOR3jL3",
        "outputId": "2f9e8949-e300-4d2d-cf94-c9d9987f7c8a"
      },
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'cuda'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 95
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "torch.manual_seed(42)\n",
        "\n",
        "def eval_model(model: torch.nn.Module,\n",
        "               data_loader: torch.utils.data.DataLoader,\n",
        "               loss_fn: torch.nn.Module,\n",
        "               accuracy_fn,\n",
        "               device=device):\n",
        "  \"\"\"Returns a dictionary containing the results of model predicting on dataloader\"\"\"\n",
        "  loss, acc = 0,0\n",
        "  model.eval()\n",
        "\n",
        "  with torch.inference_mode():\n",
        "\n",
        "    for X,y in tqdm(data_loader):\n",
        "\n",
        "      # Make our data device agnostic\n",
        "      X, y = X.to(device), y.to(device)\n",
        "\n",
        "      # Make predictions\n",
        "      y_pred = model(X)\n",
        "\n",
        "      #Accumulate loss and acc values\n",
        "      loss += loss_fn(y_pred, y)\n",
        "      acc += accuracy_fn(y_true=y,\n",
        "                         y_pred=y_pred.argmax(dim=1))\n",
        "      \n",
        "    # Scale loss and acc to find average loss/acc per batch\n",
        "    loss /= len(data_loader)\n",
        "    acc /= len(data_loader)\n",
        "\n",
        "  return {\"model_name\": model.__class__.__name__, # only works when model was created with a class\n",
        "          \"model_loss\": loss.item(),\n",
        "          \"model_acc\": acc} \n",
        "\n",
        "# Calculate model_0 results on test dataset\n",
        "model_0_results = eval_model(model=model_0.to(device),\n",
        "                             data_loader=test_dataloader,\n",
        "                             loss_fn=loss_fn,\n",
        "                             accuracy_fn=accuracy_fn,\n",
        "                             device=device)\n",
        "\n",
        "model_0_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101,
          "referenced_widgets": [
            "cb400a3cde2c4e3284aab441a5cfc121",
            "9bc4d0e7057747bead5ac5e9e1f16936",
            "485cafa27a5446208e0d1017c7a04fc9",
            "ed126ee10c3c43d5b81995fb6675edaa",
            "4decda59017443508ac3f13364b31a08",
            "22f5ff9bc4164ef58392e36f4b9d472a",
            "ef898691efc6492cbf0497748f865af8",
            "5205bb18d7bc4d719f3226270d8ed31b",
            "d04fccae44494b17b92f027d525d0289",
            "b02073132dd345b496e35862314af057",
            "c003bcfd4cf042cf98e64946cc012209"
          ]
        },
        "id": "jrUqHTJ63fpD",
        "outputId": "ebda5bd9-717c-43a8-c27a-db0aa97e9179"
      },
      "execution_count": 96,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/313 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "cb400a3cde2c4e3284aab441a5cfc121"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'model_name': 'FashionMNISTModelV0',\n",
              " 'model_loss': 0.47663894295692444,\n",
              " 'model_acc': 83.42651757188499}"
            ]
          },
          "metadata": {},
          "execution_count": 96
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!nvidia-smi"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K0o0VA5n3jJk",
        "outputId": "116be737-614d-4d53-ba41-a137f5aac0a0"
      },
      "execution_count": 97,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mon Apr 10 14:00:01 2023       \n",
            "+-----------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 525.85.12    Driver Version: 525.85.12    CUDA Version: 12.0     |\n",
            "|-------------------------------+----------------------+----------------------+\n",
            "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
            "|                               |                      |               MIG M. |\n",
            "|===============================+======================+======================|\n",
            "|   0  Tesla T4            Off  | 00000000:00:04.0 Off |                    0 |\n",
            "| N/A   61C    P0    30W /  70W |   1125MiB / 15360MiB |      1%      Default |\n",
            "|                               |                      |                  N/A |\n",
            "+-------------------------------+----------------------+----------------------+\n",
            "                                                                               \n",
            "+-----------------------------------------------------------------------------+\n",
            "| Processes:                                                                  |\n",
            "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
            "|        ID   ID                                                   Usage      |\n",
            "|=============================================================================|\n",
            "+-----------------------------------------------------------------------------+\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 4. Recreate model using non-linearity on GPU "
      ],
      "metadata": {
        "id": "YA1uyWm93jG-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from torch import nn\n",
        "\n",
        "class FashionMNISTModelV1(nn.Module):\n",
        "  def __init__(self,\n",
        "               input_shape: int,\n",
        "               hidden_units: int,\n",
        "               output_shape: int,):\n",
        "\n",
        "    super().__init__()\n",
        "\n",
        "    self.layer_stack = nn.Sequential(\n",
        "        nn.Flatten(), # Flatten inputs into a single vector\n",
        "        nn.Linear(in_features=input_shape, \n",
        "                  out_features=hidden_units),\n",
        "        nn.ReLU(),\n",
        "        nn.Linear(in_features=hidden_units, \n",
        "                  out_features=output_shape),\n",
        "        nn.ReLU()\n",
        "    )\n",
        "\n",
        "  def forward(self, x: torch.Tensor):\n",
        "    return self.layer_stack(x)\n",
        "\n",
        "# Instantiate model to target device (if available)\n",
        "torch.manual_seed(42)\n",
        "model_1 = FashionMNISTModelV1(input_shape=28*28,\n",
        "                              hidden_units=10,\n",
        "                              output_shape=int(len(class_names))).to(device)\n",
        "model_1, device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D4axFa4n3jEI",
        "outputId": "ff28b5d6-6bb2-4f9b-9aa1-b0dd70f44ff0"
      },
      "execution_count": 98,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(FashionMNISTModelV1(\n",
              "   (layer_stack): Sequential(\n",
              "     (0): Flatten(start_dim=1, end_dim=-1)\n",
              "     (1): Linear(in_features=784, out_features=10, bias=True)\n",
              "     (2): ReLU()\n",
              "     (3): Linear(in_features=10, out_features=10, bias=True)\n",
              "     (4): ReLU()\n",
              "   )\n",
              " ),\n",
              " 'cuda')"
            ]
          },
          "metadata": {},
          "execution_count": 98
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "next(model_1.parameters()).device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oU9lIKr3f3kY",
        "outputId": "db048b2c-911e-4abd-f19b-b1090be737a0"
      },
      "execution_count": 99,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "device(type='cuda', index=0)"
            ]
          },
          "metadata": {},
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Setup loss function, optimizer, and evaluation metrics for model_1\n",
        "\n",
        "loss_fn = nn.CrossEntropyLoss()\n",
        "\n",
        "optimizer = torch.optim.SGD(params=model_1.parameters(),\n",
        "                            lr=0.1)\n",
        "\n",
        "# Pull accuracy_fn from .py file\n",
        "from helper_functions import accuracy_fn"
      ],
      "metadata": {
        "id": "cy0KV_7Wgvax"
      },
      "execution_count": 100,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 6.2 Functionizing training and evaluation loops\n",
        "\n",
        "* training loop - `train_step()`\n",
        "* testing loop - `test_step()`"
      ],
      "metadata": {
        "id": "hDhvNnh6sM6Q"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Turn training loop into a function \n",
        "\n",
        "def train_step(model: torch.nn.Module,\n",
        "               data_loader: torch.utils.data.DataLoader,\n",
        "               loss_fn: torch.nn.Module,\n",
        "               optimizer: torch.optim.Optimizer,\n",
        "               accuracy_fn,\n",
        "               device: torch.device = device):\n",
        "  \"\"\"Performs a training loop with model trying to learn on a data loader\"\"\"\n",
        "\n",
        "  # Set cumulative loss & acc == 0\n",
        "  train_loss, train_acc = 0,0\n",
        "\n",
        "  # Put model into training mode\n",
        "  model.train()\n",
        "\n",
        "  for batch, (X,y) in enumerate(data_loader):\n",
        "\n",
        "    #Send data to target device\n",
        "    X,y = X.to(device), y.to(device)\n",
        "\n",
        "    # Forward pass\n",
        "    y_pred = model(X)\n",
        "\n",
        "    # Calculate loss\n",
        "    loss = loss_fn(y_pred, y)\n",
        "    train_loss += loss\n",
        "    train_acc += accuracy_fn(y_true=y,\n",
        "                             y_pred=y_pred.argmax(dim=1))\n",
        "    \n",
        "    # Optimizer to zero grad\n",
        "    optimizer.zero_grad()\n",
        "\n",
        "    # Backpropagation\n",
        "    loss.backward()\n",
        "\n",
        "    # Step optimizer (gradient descent)\n",
        "    optimizer.step()\n",
        "  \n",
        "  # Calculate loss and accuracy per epoch\n",
        "  train_loss /= len(data_loader)\n",
        "  train_acc /= len(data_loader)\n",
        "  print(f\"Train Loss: {train_loss:.4f} | Train accuracy: {train_acc:.2f}%\")\n",
        "\n",
        "\n",
        "def test_step(model: torch.nn.Module,\n",
        "              data_loader: torch.utils.data.DataLoader,\n",
        "              loss_fn: torch.nn.Module,\n",
        "              optimizer: torch.optim.Optimizer,\n",
        "              accuracy_fn,\n",
        "              device: torch.device = device):\n",
        "  \"\"\"Performs a testing loop on trained data from a data loader\"\"\"\n",
        "  \n",
        "  test_loss, test_acc = 0,0\n",
        "\n",
        "  model.eval()\n",
        "  with torch.inference_mode():\n",
        "    for X,y in data_loader:\n",
        "\n",
        "      # Send data to target device\n",
        "      X,y = X.to(device), y.to(device)\n",
        "\n",
        "      # 1. Foward pass\n",
        "      test_pred = model(X)\n",
        "\n",
        "      # 2. Calculate test loss / accuracy\n",
        "      test_loss += loss_fn(test_pred, y)\n",
        "      test_acc += accuracy_fn(y_true=y,\n",
        "                              y_pred=test_pred.argmax(dim=1)) # logits -> pred prob\n",
        "\n",
        "    # Adjust metrics and print out results\n",
        "    test_loss /= len(data_loader)\n",
        "    test_acc /= len(data_loader)\n",
        "    print(f\"Test Loss: {test_loss:.4f} | Test accuracy: {test_acc:.2f}%\")\n"
      ],
      "metadata": {
        "id": "VXnS0rSzgvXB"
      },
      "execution_count": 101,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Use functionalized training/testing loops on data\n",
        "from tqdm import tqdm\n",
        "from timeit import default_timer as timer\n",
        "\n",
        "torch.manual_seed(42)\n",
        "torch.cuda.manual_seed(42)\n",
        "\n",
        "train_start_gpu = timer()\n",
        "epochs = 3\n",
        "\n",
        "for epoch in tqdm(range(epochs)):\n",
        "  print(f\"Epoch: {epoch}\\n----\")\n",
        "\n",
        "  train_step(model=model_1,\n",
        "             data_loader=train_dataloader,\n",
        "             loss_fn=loss_fn,\n",
        "             optimizer=optimizer,\n",
        "             accuracy_fn=accuracy_fn,\n",
        "             device=device)\n",
        "  test_step(model=model_1,\n",
        "            data_loader=test_dataloader,\n",
        "            loss_fn=loss_fn,\n",
        "            optimizer=optimizer,\n",
        "            accuracy_fn=accuracy_fn,\n",
        "            device=device)\n",
        "  \n",
        "# Measure train time\n",
        "train_end_gpu = timer()\n",
        "total_train_time_model_1 = print_train_time(start=train_start_gpu,\n",
        "                                            end=train_end_gpu,\n",
        "                                            device=str(next(model_1.parameters())))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ga7orSJMlwvg",
        "outputId": "0bb47d8c-6be5-439d-f301-e1b674eada66"
      },
      "execution_count": 102,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r  0%|          | 0/3 [00:00<?, ?it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0\n",
            "----\n",
            "Train Loss: 1.0920 | Train accuracy: 61.34%\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r 33%|███▎      | 1/3 [00:20<00:41, 20.55s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.9564 | Test accuracy: 65.00%\n",
            "Epoch: 1\n",
            "----\n",
            "Train Loss: 0.7810 | Train accuracy: 71.93%\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r 67%|██████▋   | 2/3 [00:32<00:15, 15.22s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.7223 | Test accuracy: 73.91%\n",
            "Epoch: 2\n",
            "----\n",
            "Train Loss: 0.6703 | Train accuracy: 75.94%\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 3/3 [00:43<00:00, 14.58s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.6850 | Test accuracy: 75.02%\n",
            "Train time on Parameter containing:\n",
            "tensor([[ 0.0273,  0.0296, -0.0087,  ..., -0.0171,  0.0077,  0.0135],\n",
            "        [-0.0190, -0.0359,  0.0193,  ...,  0.0192,  0.0259,  0.0185],\n",
            "        [-0.0009,  0.0020,  0.0007,  ...,  0.0185, -0.0024,  0.0076],\n",
            "        ...,\n",
            "        [-0.0115,  0.0276, -0.0308,  ...,  0.0693,  0.0573,  0.0039],\n",
            "        [-0.0231,  0.0255,  0.0238,  ..., -0.0829, -0.0392, -0.0057],\n",
            "        [ 0.0179, -0.0148,  0.0052,  ..., -0.1301, -0.0557,  0.0184]],\n",
            "       device='cuda:0', requires_grad=True): 43.739 seconds\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**NOTE:** Sometimes, depending on your data/hardware, your model may tarin faster on CPU than GPU.\n",
        "\n",
        "Why is this?\n",
        "\n",
        "1. It could be that the overhead for copying data/model to and from GPU outweighs the compute benefits offered by the GPU.\n",
        "2. The hardware your using has a better CPU in terms of compute capability than the GPU.\n",
        "\n",
        "For more info on this: https://horace.io/brrr_intro.html"
      ],
      "metadata": {
        "id": "758t44k9yC1W"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model_0_results, total_train_time_model_0\n",
        "\n",
        "# Our model trained faster on GPU \n",
        "# BUT did not perform as well with non-linear layers, interestingly"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Rex4vEZxlws8",
        "outputId": "2683bf44-4dcd-45b8-ad61-0656ec60cdcf"
      },
      "execution_count": 103,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "({'model_name': 'FashionMNISTModelV0',\n",
              "  'model_loss': 0.47663894295692444,\n",
              "  'model_acc': 83.42651757188499},\n",
              " 42.42429682300008)"
            ]
          },
          "metadata": {},
          "execution_count": 103
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Get model_1 results dictionary\n",
        "\n",
        "model_1_results = eval_model(model=model_1,\n",
        "                             data_loader=test_dataloader,\n",
        "                             loss_fn=loss_fn,\n",
        "                             accuracy_fn=accuracy_fn,\n",
        "                             device=device)\n",
        "model_1_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nOwybneUt8AF",
        "outputId": "d0907169-0b7c-450d-8ddb-4a2748a3e1b0"
      },
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 313/313 [00:01<00:00, 234.08it/s]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'model_name': 'FashionMNISTModelV1',\n",
              " 'model_loss': 0.6850008964538574,\n",
              " 'model_acc': 75.01996805111821}"
            ]
          },
          "metadata": {},
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model 2: Build a Convolutional Neural Network (CNN)\n",
        "\n",
        "CNN's are also known as ConvNets.\n",
        "\n",
        "CNN's are known for their capabilities to find patterns in visual data"
      ],
      "metadata": {
        "id": "PlP3DbRJt787"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np"
      ],
      "metadata": {
        "id": "e41NCkyi8m-C"
      },
      "execution_count": 105,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "np.convolve((1,2,3), (4,5,6))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jAeVFez28m7W",
        "outputId": "13413554-dd03-4701-8c41-80d912046158"
      },
      "execution_count": 106,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 4, 13, 28, 27, 18])"
            ]
          },
          "metadata": {},
          "execution_count": 106
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a convolutional neural network \n",
        "\n",
        "class FashionMNISTModelV2(nn.Module):\n",
        "  \"\"\"\n",
        "  Model arcitecture that replicates TinyVGG:\n",
        "  model from CNN explainer website.\n",
        "  \"\"\"\n",
        "\n",
        "  def __init__(self,\n",
        "               input_shape: int,\n",
        "               hidden_units: int,\n",
        "               output_shape: int):\n",
        "    super().__init__()\n",
        "\n",
        "    self.conv_block_1 = nn.Sequential(\n",
        "        nn.Conv2d(in_channels=input_shape,\n",
        "                  out_channels=hidden_units,\n",
        "                  kernel_size=3, #hyperparam 1\n",
        "                  stride=1, # hyperparam 2\n",
        "                  padding=1), # hyperparam 3\n",
        "        nn.ReLU(),\n",
        "        nn.Conv2d(in_channels=hidden_units,\n",
        "                  out_channels=hidden_units,\n",
        "                  kernel_size=3,\n",
        "                  stride=1,\n",
        "                  padding=1),\n",
        "        nn.ReLU(),\n",
        "        nn.MaxPool2d(kernel_size=2)\n",
        "    )\n",
        "\n",
        "    self.conv_block_2 = nn.Sequential(\n",
        "        nn.Conv2d(in_channels=hidden_units,\n",
        "                  out_channels=hidden_units,\n",
        "                  kernel_size=3,\n",
        "                  stride=1,\n",
        "                  padding=1),\n",
        "        nn.ReLU(),\n",
        "        nn.Conv2d(in_channels=hidden_units,\n",
        "                  out_channels=hidden_units,\n",
        "                  kernel_size=3,\n",
        "                  stride=1,\n",
        "                  padding=1),\n",
        "        nn.ReLU(),\n",
        "        nn.MaxPool2d(kernel_size=2)\n",
        "    )\n",
        "\n",
        "    self.classifier = nn.Sequential(\n",
        "        nn.Flatten(), # flatten mutli-dim tensor into a single-dim feature vector\n",
        "        nn.Linear(in_features=hidden_units*7*7, # There is a trick to calculating this...\n",
        "                  out_features=output_shape)\n",
        "    )\n",
        "\n",
        "\n",
        "  def forward(self,x):\n",
        "    x = self.conv_block_1(x)\n",
        "    #print(f\"conv_block_1: {x.shape}\")\n",
        "    x = self.conv_block_2(x)\n",
        "    #print(f\"conv_block_2: {x.shape}\")\n",
        "    x = self.classifier(x)\n",
        "    #print(f\"classifier shape: {x.shape}\")\n",
        "    return x\n"
      ],
      "metadata": {
        "id": "1EQQYIiU8m1a"
      },
      "execution_count": 107,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "torch.manual_seed(42)\n",
        "model_2 = FashionMNISTModelV2(input_shape=1, # 1 color channel for greyscale\n",
        "                              hidden_units=10,\n",
        "                              output_shape=len(class_names)).to(device)\n",
        "model_2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1KdIzS0F8myS",
        "outputId": "fcbe72dd-ac1a-48f2-edf4-170b955fa077"
      },
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "FashionMNISTModelV2(\n",
              "  (conv_block_1): Sequential(\n",
              "    (0): Conv2d(1, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (1): ReLU()\n",
              "    (2): Conv2d(10, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (3): ReLU()\n",
              "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  )\n",
              "  (conv_block_2): Sequential(\n",
              "    (0): Conv2d(10, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (1): ReLU()\n",
              "    (2): Conv2d(10, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (3): ReLU()\n",
              "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  )\n",
              "  (classifier): Sequential(\n",
              "    (0): Flatten(start_dim=1, end_dim=-1)\n",
              "    (1): Linear(in_features=490, out_features=10, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 108
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create random tensor to test model\n",
        "\n",
        "rand_image_tensor = torch.randn(size=(1,28,28))\n",
        "rand_image_tensor.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3tEJN0ZEqTc1",
        "outputId": "60042f16-af8a-4d30-d73d-1e9a28ca7b0c"
      },
      "execution_count": 109,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([1, 28, 28])"
            ]
          },
          "metadata": {},
          "execution_count": 109
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Pass random image tensor through model\n",
        "\n",
        "model_2(rand_image_tensor.unsqueeze(0).to(device))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VNpF6S-X0J9y",
        "outputId": "b82ed2e0-6ca6-4af9-e518-4b50d0652937"
      },
      "execution_count": 110,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([[ 0.0366, -0.0940,  0.0686, -0.0485,  0.0068,  0.0290,  0.0132,  0.0084,\n",
              "         -0.0030, -0.0185]], device='cuda:0', grad_fn=<AddmmBackward0>)"
            ]
          },
          "metadata": {},
          "execution_count": 110
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 7.1 Stepping through `nn.Conv2d()`\n",
        "\n",
        "See the documentation for `nn.Conv2d` here: https://pytorch.org/docs/stable/generated/torch.nn.Conv2d.html\n",
        "\n"
      ],
      "metadata": {
        "id": "6E-h_OyPPuT7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create some dummy data\n",
        "\n",
        "torch.manual_seed(42)\n",
        "\n",
        "# Create batch of images\n",
        "images = torch.randn(size=(32, 3, 64, 64))\n",
        "test_image = images\n",
        "\n",
        "print(f\"image batch shape: {images.shape}\")\n",
        "print(f\"single image shape: {test_image.shape}\")\n",
        "print(f\"test iamge:\\n {test_image}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R1q-b4HoPuRg",
        "outputId": "5456ab30-8b0b-457e-de2e-56119c2c0110"
      },
      "execution_count": 111,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "image batch shape: torch.Size([32, 3, 64, 64])\n",
            "single image shape: torch.Size([32, 3, 64, 64])\n",
            "test iamge:\n",
            " tensor([[[[ 1.9269,  1.4873,  0.9007,  ...,  1.8446, -1.1845,  1.3835],\n",
            "          [ 1.4451,  0.8564,  2.2181,  ...,  0.3399,  0.7200,  0.4114],\n",
            "          [ 1.9312,  1.0119, -1.4364,  ..., -0.5558,  0.7043,  0.7099],\n",
            "          ...,\n",
            "          [-0.5610, -0.4830,  0.4770,  ..., -0.2713, -0.9537, -0.6737],\n",
            "          [ 0.3076, -0.1277,  0.0366,  ..., -2.0060,  0.2824, -0.8111],\n",
            "          [-1.5486,  0.0485, -0.7712,  ..., -0.1403,  0.9416, -0.0118]],\n",
            "\n",
            "         [[-0.5197,  1.8524,  1.8365,  ...,  0.8935, -1.5114, -0.8515],\n",
            "          [ 2.0818,  1.0677, -1.4277,  ...,  1.6612, -2.6223, -0.4319],\n",
            "          [-0.1010, -0.4388, -1.9775,  ...,  0.2106,  0.2536, -0.7318],\n",
            "          ...,\n",
            "          [ 0.2779,  0.7342, -0.3736,  ..., -0.4601,  0.1815,  0.1850],\n",
            "          [ 0.7205, -0.2833,  0.0937,  ..., -0.1002, -2.3609,  2.2465],\n",
            "          [-1.3242, -0.1973,  0.2920,  ...,  0.5409,  0.6940,  1.8563]],\n",
            "\n",
            "         [[-0.7978,  1.0261,  1.1465,  ...,  1.2134,  0.9354, -0.0780],\n",
            "          [-1.4647, -1.9571,  0.1017,  ..., -1.9986, -0.7409,  0.7011],\n",
            "          [-1.3938,  0.8466, -1.7191,  ..., -1.1867,  0.1320,  0.3407],\n",
            "          ...,\n",
            "          [ 0.8206, -0.3745,  1.2499,  ..., -0.0676,  0.0385,  0.6335],\n",
            "          [-0.5589, -0.3393,  0.2347,  ...,  2.1181,  2.4569,  1.3083],\n",
            "          [-0.4092,  1.5199,  0.2401,  ..., -0.2558,  0.7870,  0.9924]]],\n",
            "\n",
            "\n",
            "        [[[ 1.1462, -0.9252,  0.0457,  ..., -1.6060, -0.3585, -0.4588],\n",
            "          [ 1.5696,  2.1210,  0.4030,  ..., -1.1395,  0.9925, -0.7322],\n",
            "          [-1.6529, -1.4641,  0.0971,  ...,  0.2152, -1.5102,  0.4840],\n",
            "          ...,\n",
            "          [ 0.7408, -0.7165,  2.1898,  ...,  0.4568,  0.0514, -0.9807],\n",
            "          [-0.2803, -0.3592, -1.1238,  ..., -0.5225, -0.1409,  0.6916],\n",
            "          [-0.5811, -1.3816,  1.0387,  ..., -0.6358,  0.9728, -0.2934]],\n",
            "\n",
            "         [[-1.3047,  1.4507,  0.2443,  ..., -0.1408, -1.4311, -0.1850],\n",
            "          [ 0.8868, -0.2988, -0.5001,  ..., -0.0105,  0.8178,  0.1980],\n",
            "          [ 0.9566, -0.6351,  0.5435,  ...,  0.3815, -0.5045,  0.6647],\n",
            "          ...,\n",
            "          [-1.8424, -0.3769,  0.4985,  ...,  1.0519, -1.0642, -0.7167],\n",
            "          [ 0.6115,  0.4193, -0.9970,  ...,  0.1066, -0.7907,  0.2683],\n",
            "          [-0.5824, -0.3426,  0.4904,  ...,  2.5234,  1.3809,  0.4820]],\n",
            "\n",
            "         [[-0.2932, -0.3666, -1.6757,  ..., -0.7696, -1.8347,  0.5975],\n",
            "          [-0.6698, -0.3543, -1.5089,  ...,  0.7316,  0.2406, -0.2326],\n",
            "          [-0.6608,  0.4298,  0.1429,  ..., -0.2031,  1.0480, -1.2931],\n",
            "          ...,\n",
            "          [-0.0854,  1.1046, -1.1372,  ..., -1.5342,  0.3418,  0.4373],\n",
            "          [ 1.5311, -0.2290, -0.7734,  ..., -0.6297, -0.1439, -0.6415],\n",
            "          [-0.5336,  0.8814,  0.2079,  ..., -0.8355, -1.9863, -0.9132]]],\n",
            "\n",
            "\n",
            "        [[[ 1.8723, -0.4072, -0.2770,  ...,  0.7319,  0.5414, -0.4704],\n",
            "          [-1.1442, -0.8629, -0.0325,  ..., -0.6706, -0.5498, -1.0136],\n",
            "          [-0.8476, -0.0746,  1.6060,  ...,  0.0655,  1.2327,  1.5566],\n",
            "          ...,\n",
            "          [ 0.1831, -1.1265,  1.0662,  ..., -1.6199,  0.8893,  1.8041],\n",
            "          [ 2.2873,  0.2870,  0.5738,  ...,  0.3699,  0.5207,  1.7841],\n",
            "          [-0.9507, -0.6412,  0.6762,  ...,  1.6346,  1.7163,  0.5287]],\n",
            "\n",
            "         [[ 0.3617,  0.0284,  0.7862,  ..., -0.4626,  0.0907,  0.4144],\n",
            "          [-0.3127,  1.0867,  1.2147,  ...,  0.4430,  1.6820,  1.9812],\n",
            "          [-1.5535,  1.2849, -0.2383,  ..., -0.6326, -0.7753, -0.0257],\n",
            "          ...,\n",
            "          [ 0.3711,  0.0639, -0.1340,  ...,  1.0498,  1.2419,  0.7912],\n",
            "          [ 0.7199, -1.6989,  0.5186,  ...,  1.3509,  0.2584, -0.9650],\n",
            "          [ 0.6124,  1.5963, -0.8839,  ...,  1.5568,  1.9061, -0.2654]],\n",
            "\n",
            "         [[-0.7838, -0.3569, -1.3868,  ...,  0.0293,  0.3965, -0.0459],\n",
            "          [-1.2732,  1.3058,  1.1782,  ..., -0.5917, -0.1135, -0.0979],\n",
            "          [-0.4611, -1.2840,  1.2376,  ...,  0.2198, -0.9789,  1.8790],\n",
            "          ...,\n",
            "          [ 1.2557,  0.2726,  0.4869,  ..., -0.9088, -1.0899,  1.8497],\n",
            "          [ 0.4721, -0.5464,  0.0127,  ..., -0.8402,  0.3470, -0.5938],\n",
            "          [ 0.3543, -0.7542,  0.8093,  ..., -0.0075,  1.2261, -0.3360]]],\n",
            "\n",
            "\n",
            "        ...,\n",
            "\n",
            "\n",
            "        [[[-2.1682, -0.7886, -0.3461,  ...,  0.3142, -0.3933, -0.4687],\n",
            "          [ 0.9872,  0.0910, -0.2333,  ...,  0.5540,  0.9231, -1.7830],\n",
            "          [-0.3090,  0.5961, -1.9058,  ..., -0.8920,  0.2649,  0.5361],\n",
            "          ...,\n",
            "          [ 1.4019,  1.5822,  1.9109,  ...,  1.4598,  1.4994,  0.0590],\n",
            "          [ 0.2599,  0.8737, -1.8681,  ..., -1.0228, -3.4304, -1.0239],\n",
            "          [-0.1505,  0.4495,  1.4101,  ...,  0.8395,  1.4997, -0.5686]],\n",
            "\n",
            "         [[ 0.8834,  1.1450,  0.5449,  ..., -0.2867,  0.2139, -1.3341],\n",
            "          [-1.2310, -0.3959,  0.3044,  ..., -0.4019, -0.7961,  0.0661],\n",
            "          [ 1.4644, -0.7622,  0.7954,  ..., -0.3784, -0.1177,  1.8225],\n",
            "          ...,\n",
            "          [ 1.4499,  0.9205,  0.8025,  ..., -0.9204, -0.2336,  1.0684],\n",
            "          [-0.8314,  0.2011,  0.4146,  ..., -1.3222, -0.6220,  0.0374],\n",
            "          [-0.6306, -1.0083,  0.4322,  ...,  0.9623, -0.2101, -0.8070]],\n",
            "\n",
            "         [[ 0.9766,  1.8551,  0.8079,  ...,  0.0236, -0.7548,  0.2805],\n",
            "          [-1.5609, -0.9496, -1.2076,  ...,  0.6664,  0.4597,  1.2509],\n",
            "          [-0.3669,  1.0052, -0.1830,  ...,  2.1652, -0.3830, -0.6441],\n",
            "          ...,\n",
            "          [ 0.0621,  0.0039,  1.0790,  ...,  1.0980,  1.3147,  1.7881],\n",
            "          [ 1.5639,  0.2737,  0.0368,  ...,  0.3102,  0.5700,  2.1521],\n",
            "          [-0.2677,  0.9526, -1.7323,  ..., -0.5503,  0.5012,  0.4050]]],\n",
            "\n",
            "\n",
            "        [[[ 0.6869, -0.4450, -0.2497,  ...,  1.1245,  2.1253,  0.5647],\n",
            "          [ 1.3971,  1.0697, -0.8996,  ...,  1.0537,  1.3428, -0.2854],\n",
            "          [ 0.6388, -0.5320, -1.5274,  ..., -1.8715,  0.3396,  0.9248],\n",
            "          ...,\n",
            "          [-1.3018,  0.4129,  0.0119,  ..., -1.9978, -0.1642, -2.0266],\n",
            "          [ 0.3821, -0.6324, -0.3748,  ...,  0.3744, -0.5984, -0.1415],\n",
            "          [ 1.5145, -0.9476, -0.6816,  ..., -1.4079, -0.6893, -2.5450]],\n",
            "\n",
            "         [[ 1.0765,  1.2431, -0.3601,  ...,  1.4529, -1.9261,  0.2059],\n",
            "          [ 0.9315,  0.2448,  0.9260,  ..., -1.4899, -0.2039,  0.1974],\n",
            "          [-0.1349, -0.1783, -0.5261,  ..., -0.4463, -0.9531, -1.2018],\n",
            "          ...,\n",
            "          [-1.5692, -0.5791,  0.0771,  ..., -0.3282, -0.6795,  1.5796],\n",
            "          [-0.7103, -0.1010,  0.3765,  ..., -1.9444, -0.1589,  0.9398],\n",
            "          [-0.1365, -0.6480, -0.5398,  ...,  0.3594, -0.5389,  0.6653]],\n",
            "\n",
            "         [[-0.8764, -0.5104, -1.2873,  ..., -1.7659,  1.0042,  1.0264],\n",
            "          [-0.7540,  0.3738, -0.0059,  ..., -1.9595,  1.0695,  0.5267],\n",
            "          [ 0.6567,  0.8677,  0.9835,  ...,  0.7666, -0.7829, -0.3841],\n",
            "          ...,\n",
            "          [-0.1040,  0.6071, -0.5816,  ...,  0.9904,  0.8115, -0.5291],\n",
            "          [-0.9045, -0.5544, -0.4637,  ..., -0.0552,  0.0711,  0.0146],\n",
            "          [-0.6846, -1.0256, -0.6969,  ...,  0.5858, -0.9769,  0.0943]]],\n",
            "\n",
            "\n",
            "        [[[-1.7576, -0.4235,  0.4978,  ..., -1.5653, -0.0935, -0.8620],\n",
            "          [ 0.0648,  0.9873, -0.6721,  ..., -1.2957, -0.5044,  0.7429],\n",
            "          [ 0.2712, -0.8809, -0.8739,  ...,  1.0423,  0.8242,  2.0000],\n",
            "          ...,\n",
            "          [ 0.7474, -0.0822, -0.8413,  ...,  0.1025,  0.6660,  0.3744],\n",
            "          [-0.1608,  0.7354, -0.7628,  ..., -0.3506,  2.7400, -0.7302],\n",
            "          [ 0.1863,  0.3572, -0.0457,  ..., -0.2024, -0.0339, -0.6260]],\n",
            "\n",
            "         [[ 1.1211, -0.3509,  0.2602,  ..., -2.0339,  0.0289,  0.0932],\n",
            "          [-1.0552,  0.5302,  0.7214,  ...,  0.6793, -1.6776,  1.1767],\n",
            "          [-0.5452,  0.7653,  0.2258,  ...,  0.1321, -0.1846, -0.1664],\n",
            "          ...,\n",
            "          [-1.3320,  0.6906,  0.2853,  ...,  1.7544,  0.6776,  0.1974],\n",
            "          [-0.8164,  1.5707, -1.0187,  ...,  0.7140,  1.9848, -0.6659],\n",
            "          [ 1.3224,  1.0515,  1.0870,  ..., -0.0260,  1.2395,  0.9925]],\n",
            "\n",
            "         [[ 1.6056,  0.8208,  1.0230,  ...,  0.0335,  1.5842,  0.8418],\n",
            "          [-0.4198,  0.3207, -0.1954,  ..., -0.3185,  1.3290,  0.5406],\n",
            "          [ 1.0659, -1.7458, -0.2608,  ..., -1.1151, -0.8994,  0.0370],\n",
            "          ...,\n",
            "          [-0.1177,  0.4560, -1.3883,  ..., -1.1566, -1.2233,  0.4873],\n",
            "          [-0.8383,  1.2338,  0.2898,  ...,  1.0237,  1.6105, -0.6264],\n",
            "          [ 0.1180, -0.9001,  2.4965,  ...,  0.4437, -0.3335,  1.3220]]]])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        " # Create a single conv2d layer\n",
        "torch.manual_seed(42)\n",
        "conv_layer = nn.Conv2d(in_channels=3,\n",
        "                        out_channels=10, # hidden units\n",
        "                        kernel_size=3, # 3x3 pixel grid\n",
        "                        stride=1,\n",
        "                        padding=0)\n",
        " \n",
        "# pass the data through the conv layer\n",
        "conv_output = conv_layer(test_image)\n",
        "conv_output.shape # conv layer adds extra dim (out_channel) and shape will change depending on hyperparams"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VfUSVdboPuNv",
        "outputId": "9d635f35-f8ce-4215-a7ba-f48d452e2787"
      },
      "execution_count": 112,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([32, 10, 62, 62])"
            ]
          },
          "metadata": {},
          "execution_count": 112
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 7.2 Stepping through `nn.MaxPool2d()`\n",
        "\n",
        "https://pytorch.org/docs/stable/generated/torch.nn.MaxPool2d.html"
      ],
      "metadata": {
        "id": "l8FlrU-APuLS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(f\"original shape: {test_image.shape}\")\n",
        "print(f\"convolved shape: {conv_output.shape}\")\n",
        "\n",
        "# Create a sample max2d layer\n",
        "max_pool_layer = nn.MaxPool2d(kernel_size=2) #2x2 grid\n",
        "\n",
        "# Pass data through maxpool layer\n",
        "conv_output_max_pool = max_pool_layer(conv_output)\n",
        "print(f\"shape after conv image through max pool: {conv_output_max_pool.shape}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "abJNz1pZ0KDL",
        "outputId": "739ebb13-838a-42a6-e076-b47ad72e7ee0"
      },
      "execution_count": 113,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "original shape: torch.Size([32, 3, 64, 64])\n",
            "convolved shape: torch.Size([32, 10, 62, 62])\n",
            "shape after conv image through max pool: torch.Size([32, 10, 31, 31])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.imshow(image.squeeze(), cmap=\"gray\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "id": "IWFWbkeZ0KAf",
        "outputId": "194529ae-b5b9-49d6-bd98-f7bec7905080"
      },
      "execution_count": 114,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f228ad4de20>"
            ]
          },
          "metadata": {},
          "execution_count": 114
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "image.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nPyg-cO9qF3u",
        "outputId": "61b179bf-a6a8-4b9d-8171-dc3f496a3758"
      },
      "execution_count": 115,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([1, 28, 28])"
            ]
          },
          "metadata": {},
          "execution_count": 115
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "10*7*7"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mT2YlCO80J7H",
        "outputId": "9912d630-2157-452b-fd90-2625a05cb8b1"
      },
      "execution_count": 116,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "490"
            ]
          },
          "metadata": {},
          "execution_count": 116
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 7.3 Setup a loss fn and optimizer for model_2"
      ],
      "metadata": {
        "id": "CF7DcOeHv-ae"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from helper_functions import accuracy_fn\n",
        "\n",
        "# loss fn\n",
        "loss_fn = nn.CrossEntropyLoss()\n",
        "\n",
        "# optimizer\n",
        "optimizer = torch.optim.SGD(params=model_2.parameters(),\n",
        "                            lr=0.1)"
      ],
      "metadata": {
        "id": "HpdoS8TPv-Xz"
      },
      "execution_count": 117,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
        "device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "6krbi9wsyy8o",
        "outputId": "49809968-2f6f-4af7-bb43-602e366a6baa"
      },
      "execution_count": 118,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'cuda'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 118
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 7.4 Training and Testing loops"
      ],
      "metadata": {
        "id": "GQIIaS01y-W_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Run our image data through train_step() and test_step() functions\n",
        "\n",
        "torch.manual_seed(42)\n",
        "torch.cuda.manual_seed(42)\n",
        "epochs = 3\n",
        "\n",
        "# Measure time\n",
        "from timeit import default_timer\n",
        "from tqdm import tqdm\n",
        "\n",
        "train_start_model_2 = timer()\n",
        "\n",
        "for epoch in tqdm(range(epochs)):\n",
        "  print(f\"Epoch: {epoch}\\n----\")\n",
        "  ### TRAIN\n",
        "  train_step(model=model_2,\n",
        "             data_loader=train_dataloader,\n",
        "             loss_fn=loss_fn,\n",
        "             optimizer=optimizer,\n",
        "             accuracy_fn=accuracy_fn,\n",
        "             device=device) \n",
        "  ### TEST\n",
        "  test_step(model=model_2,\n",
        "            data_loader=test_dataloader,\n",
        "            loss_fn=loss_fn,\n",
        "            optimizer=optimizer,\n",
        "            accuracy_fn=accuracy_fn,\n",
        "            device=device)\n",
        "    \n",
        "train_end_model_2 = timer()\n",
        "total_train_time_model_2 = print_train_time(start=train_start_model_2,\n",
        "                                            end=train_end_model_2,\n",
        "                                            device=str(next(model_2.parameters())))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ty16g1UCv-VZ",
        "outputId": "c609a45e-5fd3-4be8-ef9d-9631ff87d883"
      },
      "execution_count": 119,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r  0%|          | 0/3 [00:00<?, ?it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0\n",
            "----\n",
            "Train Loss: 0.5942 | Train accuracy: 78.46%\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r 33%|███▎      | 1/3 [00:13<00:26, 13.32s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.3946 | Test accuracy: 85.64%\n",
            "Epoch: 1\n",
            "----\n",
            "Train Loss: 0.3589 | Train accuracy: 87.16%\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r 67%|██████▋   | 2/3 [00:26<00:13, 13.46s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.3549 | Test accuracy: 86.83%\n",
            "Epoch: 2\n",
            "----\n",
            "Train Loss: 0.3230 | Train accuracy: 88.33%\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 3/3 [00:41<00:00, 13.91s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.3187 | Test accuracy: 88.55%\n",
            "Train time on Parameter containing:\n",
            "tensor([[[[ 0.2827,  0.3208, -0.4290],\n",
            "          [-0.0171,  0.1484,  0.1188],\n",
            "          [-0.3890,  0.1404,  0.3381]]],\n",
            "\n",
            "\n",
            "        [[[-0.4041,  0.5335, -0.2793],\n",
            "          [ 0.1707,  0.4015,  0.1131],\n",
            "          [-0.0809,  0.3309,  0.0470]]],\n",
            "\n",
            "\n",
            "        [[[-0.1935, -0.0813, -0.0520],\n",
            "          [-0.6085, -0.6247,  0.2437],\n",
            "          [-0.7832, -0.6580, -0.1273]]],\n",
            "\n",
            "\n",
            "        [[[-0.7054, -1.0715, -0.6158],\n",
            "          [ 0.2071, -0.8341,  0.1286],\n",
            "          [ 0.3720,  0.0162,  0.4658]]],\n",
            "\n",
            "\n",
            "        [[[-0.0132,  0.3882, -0.0068],\n",
            "          [-0.0658,  0.2226, -0.0283],\n",
            "          [ 0.1108,  0.2005,  0.1560]]],\n",
            "\n",
            "\n",
            "        [[[-0.1465,  0.1863,  0.0569],\n",
            "          [ 0.1743, -0.2037, -0.3298],\n",
            "          [-0.1291, -0.2571,  0.2550]]],\n",
            "\n",
            "\n",
            "        [[[ 0.1661,  0.1501, -0.0064],\n",
            "          [ 0.1586,  0.2451, -0.2439],\n",
            "          [ 0.0062, -0.5142, -0.0614]]],\n",
            "\n",
            "\n",
            "        [[[-0.1787,  0.0269, -0.2410],\n",
            "          [ 0.3123, -0.2267, -0.2750],\n",
            "          [-0.2251,  0.3565,  0.2475]]],\n",
            "\n",
            "\n",
            "        [[[ 0.4817, -0.2849, -0.4332],\n",
            "          [-0.2242, -0.3000,  0.0413],\n",
            "          [ 0.2869,  0.4345, -0.1713]]],\n",
            "\n",
            "\n",
            "        [[[-0.2268,  0.1757, -0.1356],\n",
            "          [ 0.2017, -0.0805,  0.1898],\n",
            "          [-0.2590, -0.1686,  0.1009]]]], device='cuda:0', requires_grad=True): 41.727 seconds\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Get model_2 results\n",
        "model_2_results = eval_model(\n",
        "    model=model_2,\n",
        "    data_loader=test_dataloader,\n",
        "    loss_fn=loss_fn,\n",
        "    accuracy_fn=accuracy_fn,\n",
        "    device=device\n",
        ")\n",
        "\n",
        "model_2_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YYUh7xZ1v-Sy",
        "outputId": "62420576-8ec1-4204-f86a-84e949b9a1de"
      },
      "execution_count": 120,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 313/313 [00:01<00:00, 216.61it/s]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'model_name': 'FashionMNISTModelV2',\n",
              " 'model_loss': 0.31867489218711853,\n",
              " 'model_acc': 88.54832268370608}"
            ]
          },
          "metadata": {},
          "execution_count": 120
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 8. Comparing model results and training time "
      ],
      "metadata": {
        "id": "7mmAf4d4r6GL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "compare_results = pd.DataFrame([model_0_results,\n",
        "                                model_1_results,\n",
        "                                model_2_results])\n",
        "compare_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "X547lDs71hYO",
        "outputId": "fd2e42ec-917e-4b6f-bdfd-55b0f5dd0b3d"
      },
      "execution_count": 121,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            model_name  model_loss  model_acc\n",
              "0  FashionMNISTModelV0    0.476639  83.426518\n",
              "1  FashionMNISTModelV1    0.685001  75.019968\n",
              "2  FashionMNISTModelV2    0.318675  88.548323"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-c6b5d7fd-c806-429c-8504-e856221e93f8\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>model_name</th>\n",
              "      <th>model_loss</th>\n",
              "      <th>model_acc</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>FashionMNISTModelV0</td>\n",
              "      <td>0.476639</td>\n",
              "      <td>83.426518</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>FashionMNISTModelV1</td>\n",
              "      <td>0.685001</td>\n",
              "      <td>75.019968</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>FashionMNISTModelV2</td>\n",
              "      <td>0.318675</td>\n",
              "      <td>88.548323</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c6b5d7fd-c806-429c-8504-e856221e93f8')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-c6b5d7fd-c806-429c-8504-e856221e93f8 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-c6b5d7fd-c806-429c-8504-e856221e93f8');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 121
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Add training time to results comparison\n",
        "\n",
        "compare_results[\"train_time\"] = [total_train_time_model_0,\n",
        "                                 total_train_time_model_1,\n",
        "                                 total_train_time_model_2]"
      ],
      "metadata": {
        "id": "Y0NDEqeH1hVI"
      },
      "execution_count": 122,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "compare_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "nul2jaw01hSs",
        "outputId": "7aa8d1eb-b16c-4dcd-b821-fb244fb1bf8d"
      },
      "execution_count": 123,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            model_name  model_loss  model_acc  train_time\n",
              "0  FashionMNISTModelV0    0.476639  83.426518   42.424297\n",
              "1  FashionMNISTModelV1    0.685001  75.019968   43.739281\n",
              "2  FashionMNISTModelV2    0.318675  88.548323   41.726735"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-c624ad4b-ba47-44de-95a8-89877e3bc66a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>model_name</th>\n",
              "      <th>model_loss</th>\n",
              "      <th>model_acc</th>\n",
              "      <th>train_time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>FashionMNISTModelV0</td>\n",
              "      <td>0.476639</td>\n",
              "      <td>83.426518</td>\n",
              "      <td>42.424297</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>FashionMNISTModelV1</td>\n",
              "      <td>0.685001</td>\n",
              "      <td>75.019968</td>\n",
              "      <td>43.739281</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>FashionMNISTModelV2</td>\n",
              "      <td>0.318675</td>\n",
              "      <td>88.548323</td>\n",
              "      <td>41.726735</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c624ad4b-ba47-44de-95a8-89877e3bc66a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-c624ad4b-ba47-44de-95a8-89877e3bc66a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-c624ad4b-ba47-44de-95a8-89877e3bc66a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 123
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Visualize our model results\n",
        "compare_results.set_index(\"model_name\")[\"model_acc\"].plot(kind=\"barh\")\n",
        "plt.xlabel(\"accuracy (%)\")\n",
        "plt.ylabel(\"model\")\n",
        "plt.show();"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "NufSFnqy1hQH",
        "outputId": "c66544a5-275e-4969-892b-737f4fe90fc3"
      },
      "execution_count": 124,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "## 9. Make and evaluate random predictions with the best model\n",
        "\n",
        "def make_predictions(model: torch.nn.Module,\n",
        "                     data: list,\n",
        "                     device:torch.device = device):\n",
        "  pred_probs = []\n",
        "  model.to(device)\n",
        "  model.eval()\n",
        "  with torch.inference_mode():\n",
        "    for sample in data:\n",
        "      #Prepare sample (add a batch dimension and pass to target device)\n",
        "      sample = torch.unsqueeze(sample, dim=0).to(device)\n",
        "\n",
        "      # Forward pass (model outputs raw logits)\n",
        "      pred_logits = model(sample)\n",
        "\n",
        "      # Get pred probs (logit -> pred prob)\n",
        "      pred_prob = torch.softmax(pred_logits.squeeze(), dim=0)\n",
        "\n",
        "      # Get pred_prob off GPU for future calculations\n",
        "      pred_probs.append(pred_prob.cpu())\n",
        "\n",
        "  # Stack pred probs to turn list to a tensor\n",
        "  return torch.stack(pred_probs)\n"
      ],
      "metadata": {
        "id": "WqIWL8U91hNk"
      },
      "execution_count": 125,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "img, label = test_data[0][:10]\n",
        "img.shape, label"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9NTj8yf_7OQB",
        "outputId": "d3ed6cb4-4695-47d6-e6d3-d5e021303efc"
      },
      "execution_count": 126,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(torch.Size([1, 28, 28]), 9)"
            ]
          },
          "metadata": {},
          "execution_count": 126
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "#random.seed(42)\n",
        "test_samples = []\n",
        "test_labels = []\n",
        "\n",
        "for sample, label in random.sample(list(test_data), k=9):\n",
        "  test_samples.append(sample)\n",
        "  test_labels.append(label)\n",
        "\n",
        "  # View first sample shape\n",
        "  test_samples[0].shape\n"
      ],
      "metadata": {
        "id": "MESIQGz-3mdk"
      },
      "execution_count": 127,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.imshow(test_samples[0].squeeze(), cmap=\"gray\")\n",
        "plt.title(class_names[test_labels[0]])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 469
        },
        "id": "JUQmZCpO3mbC",
        "outputId": "3b680351-e22a-465e-f0c3-24921b36b8bb"
      },
      "execution_count": 128,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Ankle boot')"
            ]
          },
          "metadata": {},
          "execution_count": 128
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make predictions\n",
        "pred_probs = make_predictions(model=model_2,\n",
        "                              data = test_samples,\n",
        "                              device=device)\n",
        "\n",
        "# View first two prediction probs\n",
        "pred_probs[:2]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kYYXzPvJ3mX5",
        "outputId": "3df546ae-3190-43d5-d082-6f28a2e0b467"
      },
      "execution_count": 129,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([[2.8463e-06, 9.7615e-07, 6.7623e-07, 1.4796e-06, 1.5954e-06, 7.5700e-04,\n",
              "         2.6356e-05, 2.7201e-05, 1.6268e-05, 9.9917e-01],\n",
              "        [6.2210e-05, 9.9944e-01, 4.2794e-07, 4.6952e-04, 2.1787e-05, 3.5705e-07,\n",
              "         5.2375e-06, 5.8554e-07, 1.1206e-07, 2.8106e-08]])"
            ]
          },
          "metadata": {},
          "execution_count": 129
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Convert prediction probs to labels\n",
        "pred_classes = pred_probs.argmax(dim=1)\n",
        "pred_classes"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6WaAwEGN71aR",
        "outputId": "fe5d2b9c-d3fb-40a9-8cc1-6343d5f086f5"
      },
      "execution_count": 130,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([9, 1, 8, 5, 1, 9, 2, 5, 0])"
            ]
          },
          "metadata": {},
          "execution_count": 130
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot images with predictions\n",
        "\n",
        "plt.figure(figsize=(9,9))\n",
        "nrows=3\n",
        "ncols=3\n",
        "\n",
        "for i, sample in enumerate(test_samples):\n",
        "  # Create subplot for each sample\n",
        "  plt.subplot(nrows, ncols, i+1)\n",
        "\n",
        "  # Plot target image\n",
        "  plt.imshow(sample.squeeze(), cmap=\"gray\")\n",
        "\n",
        "  # Find pred label in text form\n",
        "  pred_label = class_names[pred_classes[i]]\n",
        "\n",
        "  # Get truth label in text form\n",
        "  truth_label = class_names[test_labels[i]]\n",
        "\n",
        "  # Create title \n",
        "  title_text= f\"Pred: {pred_label} | Test label: {truth_label}\"\n",
        "\n",
        "  if pred_label == truth_label:\n",
        "    plt.title(title_text, fontsize=10, c=\"g\")\n",
        "  else:\n",
        "    plt.title(title_text, fontsize=10, c=\"r\")\n",
        "  \n",
        "  plt.axis(False);"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 749
        },
        "id": "X8HrzWOf71WX",
        "outputId": "0b7b07a4-bf3a-4522-a4d8-485625bbcd5f"
      },
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 900x900 with 9 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 10. Making a confusion matrix for further prediction evaluation\n",
        "\n",
        "1. Make predictions with trained model on the test data\n",
        "2. Make confusion matrix using `torchmetrics.ConfusionMatrix` - https://torchmetrics.readthedocs.io/en/stable/classification/confusion_matrix.html \n",
        "3. Plot confusion matrix using `mlxtend.plotting.plot_confusion_matrix` - https://rasbt.github.io/mlxtend/user_guide/evaluate/confusion_matrix/ "
      ],
      "metadata": {
        "id": "RyEF_nXE824t"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Make predictions with trained model\n",
        "from tqdm import tqdm\n",
        "\n",
        "# 1. Make predictions with trained model\n",
        "y_preds = []\n",
        "\n",
        "model_2.eval()\n",
        "with torch.inference_mode():\n",
        "  for X,y in tqdm(test_dataloader, desc=\"Making predictions\"):\n",
        "\n",
        "    # Set data to target device\n",
        "    X, y = X.to(device), y.to(device)\n",
        "\n",
        "    # Forward pass\n",
        "    y_logits = model_2(X)\n",
        "\n",
        "    # Convert logits to pred probs -> pred label\n",
        "    y_pred = torch.softmax(y_logits.squeeze(), dim=0).argmax(dim=1)\n",
        "\n",
        "    # Put pred on CPU for evaluation\n",
        "    y_preds.append(y_pred.cpu())\n",
        "\n",
        "# Concatonate list of preds in to tensors\n",
        "y_pred_tensor = torch.cat(y_preds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "W851J0JV71Tf",
        "outputId": "6f2b86a2-3f5e-4563-f335-be6d28fde626"
      },
      "execution_count": 132,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Making predictions: 100%|██████████| 313/313 [00:01<00:00, 217.45it/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred_tensor[:10], len(y_pred_tensor)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BGpvCkCR71Q2",
        "outputId": "77111e3e-4805-4a48-9c30-a3e46044024e"
      },
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(tensor([9, 2, 1, 1, 6, 1, 4, 6, 5, 7]), 10000)"
            ]
          },
          "metadata": {},
          "execution_count": 133
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Import packages if not installed\n",
        "try:\n",
        "  import torchmetrics, mlxtend\n",
        "  print(f\"mlxtend version: {mlxtend.__version__}\")\n",
        "  assert int(mlxtend.__version__.split(\".\")[1] >= 19, \"mlxtend version should be 0.19 or higher\")\n",
        "except:\n",
        "  !pip install -q torchmetrics -U mlxtend\n",
        "  import torchmetrics, mlxtend\n",
        "  print(f\"mlxtend version: {mlxtend.__version__}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KbQuVa6371N-",
        "outputId": "f58bf7fa-bc0c-423c-efde-0a6771ac8efe"
      },
      "execution_count": 134,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "mlxtend version: 0.14.0\n",
            "mlxtend version: 0.14.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from torchmetrics import ConfusionMatrix\n",
        "from mlxtend.plotting import plot_confusion_matrix\n",
        "\n",
        "# 2. Setup confusion matrix instance and compare predictions to targets\n",
        "confmat = ConfusionMatrix(num_classes=len(class_names), task='multiclass')\n",
        "confmat_tensor = confmat(preds=y_pred_tensor,\n",
        "                         target=test_data.targets)\n",
        "\n",
        "# 3. Plot our confusion matrix\n",
        "fig, ax = plot_confusion_matrix(\n",
        "    conf_mat = confmat_tensor.numpy(), # matplotlib likes working with numpy\n",
        "    class_names = class_names,\n",
        "    figsize=(10,7)\n",
        ")"
      ],
      "metadata": {
        "id": "n2mb5SE2_Fjd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 667
        },
        "outputId": "b7bf2687-bf3c-4fe5-cf63-50ed70636765"
      },
      "execution_count": 153,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confmat_tensor"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZF7_cr4UblcW",
        "outputId": "da9b6cc0-4c68-4301-fd57-cff1a0885db4"
      },
      "execution_count": 152,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([[644,   5,  45,  58,  10,   0, 234,   0,   4,   0],\n",
              "        [ 13, 867,   8,  83,  23,   0,   6,   0,   0,   0],\n",
              "        [  8,   2, 809,  12,  97,   0,  69,   0,   3,   0],\n",
              "        [ 23,  11,  33, 849,  45,   0,  34,   0,   5,   0],\n",
              "        [  3,   6, 126,  38, 712,   0, 113,   0,   2,   0],\n",
              "        [  8,   4,   4,   9,   1, 823,   2,  53,  72,  24],\n",
              "        [ 73,   6, 129,  60,  84,   0, 641,   0,   7,   0],\n",
              "        [  1,   1,   0,   0,   0,  25,   0, 903,   7,  63],\n",
              "        [  8,   2,  33,  16,   9,   5,  35,   7, 884,   1],\n",
              "        [  1,   0,   1,   4,   0,  17,   6,  74,   6, 891]])"
            ]
          },
          "metadata": {},
          "execution_count": 152
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 11. Save and load best performing model "
      ],
      "metadata": {
        "id": "kA8NwKIfbkiP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create model directory path\n",
        "MODEL_PATH = Path(\"models\")\n",
        "MODEL_PATH.mkdir(parents=True,\n",
        "                 exist_ok=True)\n",
        "\n",
        "# Create model save\n",
        "MODEL_NAME = \"03_pytorch_computer_vision_model_2.pth\"\n",
        "MODEL_SAVE_PATH = MODEL_PATH / MODEL_NAME\n",
        "\n",
        "MODEL_SAVE_PATH"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pLb-yyXqkEDU",
        "outputId": "fbd27043-83c7-4982-83c2-c08b06fabd39"
      },
      "execution_count": 157,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "PosixPath('models/03_pytorch_computer_vision_model_2.pth')"
            ]
          },
          "metadata": {},
          "execution_count": 157
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Save model state dict\n",
        "print(f\"Saving model to {MODEL_SAVE_PATH}\")\n",
        "torch.save(obj=model_2.state_dict(),\n",
        "           f=MODEL_SAVE_PATH)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "d5k7K5DnbkgA",
        "outputId": "8ff2a7b1-94d9-4c95-e7d9-6e0d6795a9c8"
      },
      "execution_count": 162,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saving model to models/03_pytorch_computer_vision_model_2.pth\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create new instance of model_2\n",
        "\n",
        "torch.manual_seed(42)\n",
        "\n",
        "loaded_model_2 = FashionMNISTModelV2(input_shape=1,\n",
        "                                     hidden_units=10,\n",
        "                                     output_shape=len(class_names))\n",
        "\n",
        "# Load in saved state_dict\n",
        "loaded_model_2.load_state_dict(torch.load(f=MODEL_SAVE_PATH))\n",
        "\n",
        "# Send model to target device\n",
        "loaded_model_2.to(device)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lom2aYKVbkdQ",
        "outputId": "dc8b8aa1-16ad-4370-af06-9034aadf2ac7"
      },
      "execution_count": 163,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "FashionMNISTModelV2(\n",
              "  (conv_block_1): Sequential(\n",
              "    (0): Conv2d(1, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (1): ReLU()\n",
              "    (2): Conv2d(10, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (3): ReLU()\n",
              "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  )\n",
              "  (conv_block_2): Sequential(\n",
              "    (0): Conv2d(10, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (1): ReLU()\n",
              "    (2): Conv2d(10, 10, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (3): ReLU()\n",
              "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  )\n",
              "  (classifier): Sequential(\n",
              "    (0): Flatten(start_dim=1, end_dim=-1)\n",
              "    (1): Linear(in_features=490, out_features=10, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 163
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate loaded model\n",
        "torch.manual_seed(42)\n",
        "\n",
        "loaded_model_2_results = eval_model(\n",
        "    model=loaded_model_2,\n",
        "    data_loader = test_dataloader,\n",
        "    loss_fn = loss_fn,\n",
        "    accuracy_fn = accuracy_fn\n",
        ")\n",
        "\n",
        "loaded_model_2_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cdhaNlX6bkaS",
        "outputId": "4665d6ac-a479-4d5b-b04e-5e7cd2414bff"
      },
      "execution_count": 166,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 313/313 [00:02<00:00, 118.56it/s]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'model_name': 'FashionMNISTModelV2',\n",
              " 'model_loss': 0.31867489218711853,\n",
              " 'model_acc': 88.54832268370608}"
            ]
          },
          "metadata": {},
          "execution_count": 166
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fTo4bqrMmAeA",
        "outputId": "63c110c5-c78b-4492-bab2-1981e50d29a6"
      },
      "execution_count": 164,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'model_name': 'FashionMNISTModelV2',\n",
              " 'model_loss': 0.31867489218711853,\n",
              " 'model_acc': 88.54832268370608}"
            ]
          },
          "metadata": {},
          "execution_count": 164
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check to see if results are close to each other (if they are very far away, there may be an error)\n",
        "torch.isclose(torch.tensor(model_2_results[\"model_loss\"]), \n",
        "              torch.tensor(loaded_model_2_results[\"model_loss\"]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "74GR6TqsbkWl",
        "outputId": "c9906d12-c700-46f9-a569-9c4c79c25ed4"
      },
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor(True)"
            ]
          },
          "metadata": {},
          "execution_count": 168
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "ZMoGRQfL_FgF"
      },
      "execution_count": 145,
      "outputs": []
    }
  ]
}