mnist_classifier.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 5,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "mnist_classifier.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "850b16e28828496ea6bde339b332a6e8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_64c6f8c15b4f405980f25685d833824c",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_73636e0504074da4b0363601a3e62bba",
              "IPY_MODEL_fc2d372dccb04753b12c5ad9162dbe64"
            ]
          }
        },
        "64c6f8c15b4f405980f25685d833824c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "73636e0504074da4b0363601a3e62bba": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_a241563dac3c4dd89c8e9f2869eb13dc",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 9912422,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 9912422,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_4307245d28194b43b0226b381f521ee2"
          }
        },
        "fc2d372dccb04753b12c5ad9162dbe64": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_89b9e6d6407643ecb2120a4c41856280",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 9913344/? [04:55&lt;00:00, 33578.82it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_e1e46e606bb14f5e94d4f2e1c6f282ab"
          }
        },
        "a241563dac3c4dd89c8e9f2869eb13dc": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "4307245d28194b43b0226b381f521ee2": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "89b9e6d6407643ecb2120a4c41856280": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "e1e46e606bb14f5e94d4f2e1c6f282ab": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "dba49ae1492f411db7e22700cc91735c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_a5c26d88b3a846f39afb07af94bf7850",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_1d8107e20feb424295284b3bcfc910a6",
              "IPY_MODEL_f8165ba165ce4c6ea7adc900aea578ce"
            ]
          }
        },
        "a5c26d88b3a846f39afb07af94bf7850": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "1d8107e20feb424295284b3bcfc910a6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_6a10123f4d154ed081f56bcd696292d8",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 28881,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 28881,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_0b94b662f0254bb68b575b6f36e67c8c"
          }
        },
        "f8165ba165ce4c6ea7adc900aea578ce": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_80d7e64a674843b4b4178a8d151cb6e6",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 29696/? [00:00&lt;00:00, 29761.07it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_cb4af30634b346cd943838c265c2976a"
          }
        },
        "6a10123f4d154ed081f56bcd696292d8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "0b94b662f0254bb68b575b6f36e67c8c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "80d7e64a674843b4b4178a8d151cb6e6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "cb4af30634b346cd943838c265c2976a": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "b7217129b38143e691bb47aa77395432": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_01228e0b79f143dcb461e86e6dd46f25",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_96404a60a0564d3e8dbef4af11efbcd7",
              "IPY_MODEL_bff912f8429944279b124b55227aea16"
            ]
          }
        },
        "01228e0b79f143dcb461e86e6dd46f25": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "96404a60a0564d3e8dbef4af11efbcd7": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_2559f668304d4c468d6c63fc90489214",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 1648877,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1648877,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_0ada3f0b530d4a4a95e34e7dda631dfe"
          }
        },
        "bff912f8429944279b124b55227aea16": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_653be3df4b854be28127b315fe04cdce",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 1649664/? [00:00&lt;00:00, 3089439.59it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_8d96b09e7008492daad733e64ccf3475"
          }
        },
        "2559f668304d4c468d6c63fc90489214": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "0ada3f0b530d4a4a95e34e7dda631dfe": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "653be3df4b854be28127b315fe04cdce": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "8d96b09e7008492daad733e64ccf3475": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "e0182f82ae06405488d593536c817c37": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_3fd6e0ebc9c54511bac01b05ae57a190",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_165b9b4acca849339d28b75a062980d3",
              "IPY_MODEL_a635125654bd4b08a8c39ec57bb71212"
            ]
          }
        },
        "3fd6e0ebc9c54511bac01b05ae57a190": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "165b9b4acca849339d28b75a062980d3": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_85196cdc87da4da79f1dcc5f971f41cb",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 4542,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 4542,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_0c252e852c7b4683a635d3ff6524f65f"
          }
        },
        "a635125654bd4b08a8c39ec57bb71212": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_338589824ca74c868ac62b0e85432e11",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 5120/? [00:00&lt;00:00, 12376.82it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_b7438b60ac7b4a159124167146ddbe05"
          }
        },
        "85196cdc87da4da79f1dcc5f971f41cb": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "0c252e852c7b4683a635d3ff6524f65f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "338589824ca74c868ac62b0e85432e11": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "b7438b60ac7b4a159124167146ddbe05": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hywkim-private/MNIST_Classifier/blob/main/mnist_classifier.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5133dee1"
      },
      "source": [
        "import torch \n",
        "import torchvision\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.manifold import TSNE\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "from torchvision.datasets import MNIST\n",
        "from torchvision.transforms import ToTensor\n",
        "from torchvision.utils import make_grid\n",
        "from torch.utils.data.dataloader import DataLoader\n",
        "from torch.utils.data import random_split"
      ],
      "id": "5133dee1",
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "842a5488"
      },
      "source": [
        "#define hyper parameters \n",
        "USE_CUDA = torch.cuda.is_available()\n",
        "DEVICE = torch.device(\"cuda\" if USE_CUDA else \"cpu\")\n",
        "EPOCHS = 20\n",
        "BATCH_SIZE = 20\n",
        "LEARNING_RATE = 0.1\n",
        "VAL_SIZE = 1000\n",
        "N_CLASS = 10\n",
        "TEST_SIZE = 1000"
      ],
      "id": "842a5488",
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 754,
          "referenced_widgets": [
            "850b16e28828496ea6bde339b332a6e8",
            "64c6f8c15b4f405980f25685d833824c",
            "73636e0504074da4b0363601a3e62bba",
            "fc2d372dccb04753b12c5ad9162dbe64",
            "a241563dac3c4dd89c8e9f2869eb13dc",
            "4307245d28194b43b0226b381f521ee2",
            "89b9e6d6407643ecb2120a4c41856280",
            "e1e46e606bb14f5e94d4f2e1c6f282ab",
            "dba49ae1492f411db7e22700cc91735c",
            "a5c26d88b3a846f39afb07af94bf7850",
            "1d8107e20feb424295284b3bcfc910a6",
            "f8165ba165ce4c6ea7adc900aea578ce",
            "6a10123f4d154ed081f56bcd696292d8",
            "0b94b662f0254bb68b575b6f36e67c8c",
            "80d7e64a674843b4b4178a8d151cb6e6",
            "cb4af30634b346cd943838c265c2976a",
            "b7217129b38143e691bb47aa77395432",
            "01228e0b79f143dcb461e86e6dd46f25",
            "96404a60a0564d3e8dbef4af11efbcd7",
            "bff912f8429944279b124b55227aea16",
            "2559f668304d4c468d6c63fc90489214",
            "0ada3f0b530d4a4a95e34e7dda631dfe",
            "653be3df4b854be28127b315fe04cdce",
            "8d96b09e7008492daad733e64ccf3475",
            "e0182f82ae06405488d593536c817c37",
            "3fd6e0ebc9c54511bac01b05ae57a190",
            "165b9b4acca849339d28b75a062980d3",
            "a635125654bd4b08a8c39ec57bb71212",
            "85196cdc87da4da79f1dcc5f971f41cb",
            "0c252e852c7b4683a635d3ff6524f65f",
            "338589824ca74c868ac62b0e85432e11",
            "b7438b60ac7b4a159124167146ddbe05"
          ]
        },
        "id": "68c4114b",
        "outputId": "2356ad1e-9afd-4901-bc04-750af07854da"
      },
      "source": [
        "#download and define data sets using random split\n",
        "ds = MNIST(root='./data/', download=True, transform=ToTensor())"
      ],
      "id": "68c4114b",
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./data/MNIST/raw/train-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "850b16e28828496ea6bde339b332a6e8",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=9912422.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting ./data/MNIST/raw/train-images-idx3-ubyte.gz to ./data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to ./data/MNIST/raw/train-labels-idx1-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "dba49ae1492f411db7e22700cc91735c",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=28881.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting ./data/MNIST/raw/train-labels-idx1-ubyte.gz to ./data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to ./data/MNIST/raw/t10k-images-idx3-ubyte.gz\n",
            "Failed to download (trying next):\n",
            "HTTP Error 503: Service Unavailable\n",
            "\n",
            "Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-images-idx3-ubyte.gz\n",
            "Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-images-idx3-ubyte.gz to ./data/MNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "b7217129b38143e691bb47aa77395432",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=1648877.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting ./data/MNIST/raw/t10k-images-idx3-ubyte.gz to ./data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz\n",
            "Failed to download (trying next):\n",
            "HTTP Error 503: Service Unavailable\n",
            "\n",
            "Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-labels-idx1-ubyte.gz\n",
            "Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-labels-idx1-ubyte.gz to ./data/MNIST/raw/t10k-labels-idx1-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "e0182f82ae06405488d593536c817c37",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=4542.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting ./data/MNIST/raw/t10k-labels-idx1-ubyte.gz to ./data/MNIST/raw\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torchvision/datasets/mnist.py:498: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at  /pytorch/torch/csrc/utils/tensor_numpy.cpp:180.)\n",
            "  return torch.from_numpy(parsed.astype(m[2], copy=False)).view(*s)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f99f6fad"
      },
      "source": [
        "train_ds, val_ds, test_ds = random_split(ds, [len(ds)-VAL_SIZE-TEST_SIZE, VAL_SIZE, TEST_SIZE])"
      ],
      "id": "f99f6fad",
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0b3e5245"
      },
      "source": [
        "#define the dataloader \n",
        "train_dl = DataLoader(train_ds,batch_size=BATCH_SIZE,shuffle=True)\n",
        "val_dl = DataLoader(val_ds,batch_size=BATCH_SIZE,shuffle=True)\n",
        "test_dl = DataLoader(test_ds,batch_size=BATCH_SIZE,shuffle=True)\n",
        "#this is to validate model with samples from train_ds (since the original set is too large)\n",
        "train_sample, dump = random_split(train_ds, [VAL_SIZE, len(train_ds)-VAL_SIZE])\n",
        "train_sample_dl = DataLoader(train_sample, batch_size=BATCH_SIZE, shuffle=True)"
      ],
      "id": "0b3e5245",
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "c91f6488",
        "outputId": "654f5dd4-8d0d-4ffe-a049-2cb93ade6279"
      },
      "source": [
        "\"\"\"visualize/check if we have the datasets rights \n",
        "for img_set, label_set in train_dl:\n",
        "    plt.figure(figsize=(5,3))\n",
        "    plt.imshow(img.permute(1,2,0))\"\"\"\n"
      ],
      "id": "c91f6488",
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'visualize/check if we have the datasets rights \\nfor img_set, label_set in train_dl:\\n    plt.figure(figsize=(5,3))\\n    plt.imshow(img.permute(1,2,0))'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7ad444a5"
      },
      "source": [
        ""
      ],
      "id": "7ad444a5",
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bfa53102"
      },
      "source": [
        "\n",
        "        \n",
        "#A class to run and evaluate the network\n",
        "class network(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(network, self).__init__()\n",
        "        #A simple network built for testing purposes\n",
        "        #The architecture for the network has been archieved from A wiki documentation\n",
        "        self.model = nn.Sequential(\n",
        "            #w=24\n",
        "            nn.Conv2d(in_channels=1, out_channels=6, kernel_size=5, stride=1, padding=2),\n",
        "            nn.Tanh(),\n",
        "            #w=12\n",
        "            nn.MaxPool2d(kernel_size=2, stride=2),\n",
        "            #w=8\n",
        "            nn.Conv2d(in_channels=6, out_channels=16, kernel_size=5, stride=1),\n",
        "            nn.Tanh(),\n",
        "            #w=4\n",
        "            nn.MaxPool2d(kernel_size=2, stride=2),\n",
        "            nn.Conv2d(in_channels=16, out_channels=120, kernel_size=5, stride=1),\n",
        "            nn.Tanh(),\n",
        "            nn.Flatten(),\n",
        "            nn.Linear(in_features=120, out_features=84, bias=True),\n",
        "            nn.Tanh(),\n",
        "            nn.Linear(in_features=84, out_features=N_CLASS, bias=True),\n",
        "        )\n",
        "        self.val_loss = []\n",
        "        self.val_accuracy = []\n",
        "        self.test_loss = []\n",
        "        self.test_accuracy = []\n",
        "        self.train_loss = []\n",
        "        self.train_accuracy = []\n",
        "        self.failed_tests_train =[]\n",
        "        self.failed_tests_valid =[]\n",
        "        self.failed_tests_test =[]\n",
        "    #propagate a network for one cycle\n",
        "    #input: batch of images\n",
        "    #return: the result value of the network, the loss value\n",
        "    #NOTE: We will use the cross-entropy validation\n",
        "    def forward_propagate(self, batch):\n",
        "        result = self.model(batch).to(DEVICE)\n",
        "        return result\n",
        "    \n",
        "    #backward propoagate and update the parameters using gradients\n",
        "    #input: output object from the forward_propagate function\n",
        "    #input: and optimizer object that contains parameters\n",
        "    def backward_propagate(self, loss, optimizer):\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "        optimizer.zero_grad()\n",
        "    \n",
        "    #an helper function for fit that trains for a single batch\n",
        "    def fit_batch(self, optimizer):\n",
        "        error = []\n",
        "        #take each batch of examples and feed it to the propagation loop\n",
        "        for index, (image, label) in enumerate(train_dl):\n",
        "            image, label = image.to(DEVICE), label.to(DEVICE)\n",
        "            #forward propagate (single iteration)\n",
        "            result = self.forward_propagate(image)\n",
        "            loss = F.cross_entropy(result, label)\n",
        "            #backpropagate\n",
        "            self.backward_propagate(loss, optimizer)\n",
        "            error.append(loss)\n",
        "        return error\n",
        "    \n",
        "    \n",
        "    #train the model and record validation test results\n",
        "    def fit(self):\n",
        "        TRAIN_SET = 0\n",
        "        VALIDATION_SET = 1\n",
        "        TEST_SET = 2\n",
        "        #define the optimizer to pass into backward propagation--we will use SGD for the purpose of running semi-batch descent\n",
        "        optimizer = optim.SGD(self.model.parameters(),lr=LEARNING_RATE)\n",
        "        #an array of loss and aggregates for train set\n",
        "        loss_aggregate_t = []\n",
        "        accuracy_aggregate_t = []\n",
        "        #an array of loss and aggregates for validation set\n",
        "        loss_aggregate_v = []\n",
        "        accuracy_aggregate_v = []\n",
        "        #an array of loss and aggregates for test set\n",
        "        loss_aggregate_te = []\n",
        "        accuracy_aggregate_te = []\n",
        "        for epoch in range(EPOCHS):\n",
        "            #empty cuda memory\n",
        "            if USE_CUDA:\n",
        "              torch.cuda.empty_cache()\n",
        "            error = self.fit_batch(optimizer)\n",
        "            train_accuracy, train_loss, failed_tests_train = self.validate(TRAIN_SET)\n",
        "            val_accuracy, val_loss, failed_tests_val = self.validate(VALIDATION_SET)\n",
        "            test_accuracy, test_loss, failed_tests_test = self.validate(TEST_SET)\n",
        "            loss_aggregate_t.append(train_loss.tolist())\n",
        "            accuracy_aggregate_t.append(train_accuracy.tolist())\n",
        "            loss_aggregate_v.append(val_loss.tolist())\n",
        "            accuracy_aggregate_v.append(val_accuracy.tolist())\n",
        "            loss_aggregate_te.append(test_loss.tolist())\n",
        "            accuracy_aggregate_te.append(test_accuracy.tolist())\n",
        "            print(f\"----------Epoch {epoch}----------\")\n",
        "            #print(f\"The loss for the TRAINSET is: {train_loss}\")\n",
        "            #print(f\"The accuracy of the model on TRAINSET is {train_accuracy}%\")\n",
        "            print(f\"The loss for the VALIDSET is: {val_loss}\")\n",
        "            print(f\"The accuracy of the model on VALIDSET is {val_accuracy}%\")\n",
        "            print(\"----------------------------------\")\n",
        "        #update the self parameters\n",
        "        self.val_loss = loss_aggregate_v\n",
        "        self.val_accuracy = accuracy_aggregate_v\n",
        "        self.test_loss = loss_aggregate_te\n",
        "        self.test_accuracy = accuracy_aggregate_te\n",
        "        self.train_loss = loss_aggregate_t\n",
        "        self.train_accuracy = accuracy_aggregate_t\n",
        "        self.failed_tests_train = failed_tests_train\n",
        "        self.failed_tests_val = failed_tests_val\n",
        "        self.failed_tests_test = failed_tests_test\n",
        "        \"\"\"validate for a batch of inputs\n",
        "          def validate_batch(self, batch ):\n",
        "            image, label = batch\n",
        "            result = self.forward_propagate(image)\n",
        "            loss = F.cross_entropy(result, label)\n",
        "            prediction = torch.argmax(result, dim=1)\n",
        "            accurate = torch.sum(label == prediction)\n",
        "            accuracy = (accurate * 100)/ (len(label))\n",
        "            avg_loss = sum(loss) / len(test_loss)\n",
        "            return avg_loss, accuracy\"\"\"\n",
        "\n",
        "    #predict the results of the trained model using validation sets\n",
        "    #if set index ==0, validate for the train set, if set indexx ==1, do it for the validation set \n",
        "    def validate(self, set_index):\n",
        "        #We will initialize with a 0 to avoid divide by zero error\n",
        "        test_loss = [0]\n",
        "        #has 1 if the iteration yielded correct data, 0 if not\n",
        "        correct_or_wrong = 0\n",
        "        #set which data loader to use\n",
        "        if set_index == 0:\n",
        "            set_dl = train_sample_dl\n",
        "        elif set_index == 1:\n",
        "            set_dl = val_dl\n",
        "        elif set_index == 2:\n",
        "            set_dl = test_dl\n",
        "        #1xnx2 matrix with predicted fail values and its true labels\n",
        "        failed_tests = []\n",
        "        #iterate over the batch of validation datasets and check the accuracy\n",
        "        for index, (image, label) in enumerate(set_dl):\n",
        "            image, label = image.to(DEVICE), label.to(DEVICE)\n",
        "            result = self.forward_propagate(image)\n",
        "            loss = F.cross_entropy(result, label)\n",
        "            prediction = torch.argmax(result, dim=1)\n",
        "            accurate = torch.sum(label == prediction)\n",
        "            #we will record the failed label and its predicted value\n",
        "            for i, lab in enumerate(label):\n",
        "              if lab != prediction[i]:\n",
        "                failed_tests.append([lab.tolist(), prediction[i].tolist()])\n",
        "            correct_or_wrong += accurate\n",
        "            test_loss.append(loss)\n",
        "        accuracy = (correct_or_wrong * 100)/ ((index+1) * BATCH_SIZE)\n",
        "        avg_loss = sum(test_loss) / len(test_loss)\n",
        "        return accuracy, avg_loss, failed_tests\n",
        "   "
      ],
      "id": "bfa53102",
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5ceea6cd",
        "outputId": "7c7af466-2083-4a97-a47a-b266197b54bf"
      },
      "source": [
        "torch.cuda.empty_cache()\n",
        "Net = network().to(DEVICE)\n",
        "Net.fit()"
      ],
      "id": "5ceea6cd",
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "----------Epoch 0----------\n",
            "The loss for the VALIDSET is: 0.03446299210190773\n",
            "The accuracy of the model on VALIDSET is 98.9000015258789%\n",
            "----------------------------------\n",
            "----------Epoch 1----------\n",
            "The loss for the VALIDSET is: 0.02647792175412178\n",
            "The accuracy of the model on VALIDSET is 99.10000610351562%\n",
            "----------------------------------\n",
            "----------Epoch 2----------\n",
            "The loss for the VALIDSET is: 0.016649732366204262\n",
            "The accuracy of the model on VALIDSET is 99.4000015258789%\n",
            "----------------------------------\n",
            "----------Epoch 3----------\n",
            "The loss for the VALIDSET is: 0.02372211031615734\n",
            "The accuracy of the model on VALIDSET is 99.10000610351562%\n",
            "----------------------------------\n",
            "----------Epoch 4----------\n",
            "The loss for the VALIDSET is: 0.020116614177823067\n",
            "The accuracy of the model on VALIDSET is 99.4000015258789%\n",
            "----------------------------------\n",
            "----------Epoch 5----------\n",
            "The loss for the VALIDSET is: 0.021811947226524353\n",
            "The accuracy of the model on VALIDSET is 99.60000610351562%\n",
            "----------------------------------\n",
            "----------Epoch 6----------\n",
            "The loss for the VALIDSET is: 0.021854504942893982\n",
            "The accuracy of the model on VALIDSET is 99.30000305175781%\n",
            "----------------------------------\n",
            "----------Epoch 7----------\n",
            "The loss for the VALIDSET is: 0.017783982679247856\n",
            "The accuracy of the model on VALIDSET is 99.30000305175781%\n",
            "----------------------------------\n",
            "----------Epoch 8----------\n",
            "The loss for the VALIDSET is: 0.02011740207672119\n",
            "The accuracy of the model on VALIDSET is 99.4000015258789%\n",
            "----------------------------------\n",
            "----------Epoch 9----------\n",
            "The loss for the VALIDSET is: 0.019479045644402504\n",
            "The accuracy of the model on VALIDSET is 99.4000015258789%\n",
            "----------------------------------\n",
            "----------Epoch 10----------\n",
            "The loss for the VALIDSET is: 0.01741885580122471\n",
            "The accuracy of the model on VALIDSET is 99.70000457763672%\n",
            "----------------------------------\n",
            "----------Epoch 11----------\n",
            "The loss for the VALIDSET is: 0.017364012077450752\n",
            "The accuracy of the model on VALIDSET is 99.70000457763672%\n",
            "----------------------------------\n",
            "----------Epoch 12----------\n",
            "The loss for the VALIDSET is: 0.016284838318824768\n",
            "The accuracy of the model on VALIDSET is 99.50000762939453%\n",
            "----------------------------------\n",
            "----------Epoch 13----------\n",
            "The loss for the VALIDSET is: 0.016471976414322853\n",
            "The accuracy of the model on VALIDSET is 99.60000610351562%\n",
            "----------------------------------\n",
            "----------Epoch 14----------\n",
            "The loss for the VALIDSET is: 0.01655721664428711\n",
            "The accuracy of the model on VALIDSET is 99.50000762939453%\n",
            "----------------------------------\n",
            "----------Epoch 15----------\n",
            "The loss for the VALIDSET is: 0.015093924477696419\n",
            "The accuracy of the model on VALIDSET is 99.50000762939453%\n",
            "----------------------------------\n",
            "----------Epoch 16----------\n",
            "The loss for the VALIDSET is: 0.016185348853468895\n",
            "The accuracy of the model on VALIDSET is 99.60000610351562%\n",
            "----------------------------------\n",
            "----------Epoch 17----------\n",
            "The loss for the VALIDSET is: 0.01628507487475872\n",
            "The accuracy of the model on VALIDSET is 99.50000762939453%\n",
            "----------------------------------\n",
            "----------Epoch 18----------\n",
            "The loss for the VALIDSET is: 0.015819275751709938\n",
            "The accuracy of the model on VALIDSET is 99.60000610351562%\n",
            "----------------------------------\n",
            "----------Epoch 19----------\n",
            "The loss for the VALIDSET is: 0.016184544190764427\n",
            "The accuracy of the model on VALIDSET is 99.60000610351562%\n",
            "----------------------------------\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "cellView": "form",
        "id": "97d90161"
      },
      "source": [
        "#@title\n",
        "#plot the loss of test and validation set as we train data for each epoch\n",
        "def plot_loss(train_loss, val_loss):\n",
        "    plt.plot(train_loss, 'b-x', linewidth=4, markersize=12, markerwidth=4)\n",
        "    plt.plot(val_loss, 'r--o', linewidth=4, markersize=12, markerwidth=4)\n",
        "    plt.legend(['Train Loss', 'Val Loss'])\n",
        "    plt.ylabel('Cross Entropy Loss')\n",
        "    plt.xlabel('Epoch')\n",
        "    \n",
        "\n",
        "    "
      ],
      "id": "97d90161",
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "55f019ec"
      },
      "source": [
        "class visualization (network): \n",
        "  def __init__(self,network):\n",
        "          super().__init__()\n",
        "\n",
        "  #A function that extracts weights from the cnn linear function layer \n",
        "  def extract_linear_weights(self, layer_index):\n",
        "    target_layer = self.model[layer_index]\n",
        "    #initiate extraction only when the layer index indicates a Linear layer\n",
        "    if isinstance(target_layer, nn.Linear):\n",
        "      weight = target_layer.weight.data\n",
        "      print({weight})\n",
        "    else:\n",
        "      print(\"ERROR: Input layer_index of a linear function\")\n",
        "      return\n",
        "    return weight\n",
        "  \n",
        "  \n",
        "  #Extract the feature map of the input data after it passed through the given layers\n",
        "  def extract_input_data(self, layer_index, image):\n",
        "    output = image\n",
        "    #iterate over the model and run the input image\n",
        "    for i in range(layer_index):\n",
        "      target_layer = self.model[i]\n",
        "      output = target_layer(output)\n",
        "      output = output.data\n",
        "      #print(output.shape)\n",
        "    return output\n",
        "\n",
        "  #visualize the input feature as it is represented by the convolutional layer\n",
        "  def visualize_feature(self, image, label, layer_index):\n",
        "    image = torch.unsqueeze(image, 0)\n",
        "    output = vis_net.extract_input_data(layer_index, image)\n",
        "\n",
        "    #the pseudo code for visualizing the feature vectors\n",
        "    squeezed_output = output.squeeze(0)\n",
        "    for index, image in enumerate(squeezed_output):\n",
        "      plt.figure(figsize=(10, 10))\n",
        "      plt.subplot(4,4, index+1)\n",
        "      plt.imshow(image, cmap='gray')\n",
        "      plt.show\n",
        "\n",
        "  #visualize the linear layer using dimension reduction (tsne)\n",
        "  def visualize_tsne(self):\n",
        "    #Compute all the necessary values for the tsne operation\n",
        "    labels = torch.Tensor\n",
        "    for index, (image, label) in enumerate(val_dl):\n",
        "      output = self.extract_input_data(11, image)\n",
        "      output = np.array(output)\n",
        "      n,m = output.shape\n",
        "      if index == 0:\n",
        "        feature = np.zeros((BATCH_SIZE*len(val_dl)-1,m))\n",
        "        labels = np.zeros((BATCH_SIZE*len(val_dl)-1,1))\n",
        "      for i in range(BATCH_SIZE-1):\n",
        "          feature[BATCH_SIZE*(index) + i,:] = output[i]\n",
        "          labels[BATCH_SIZE*(index) + i,:] = label[i]\n",
        "    #run tsne using the sikit learn library\n",
        "    feature_embedded = TSNE(n_components = 2, n_iter=7000,early_exaggeration=40,perplexity=100,learning_rate=6).fit_transform(feature)\n",
        "    #plot the representation using scatter plot\n",
        "    feature_embedded_df = pd.DataFrame(data=feature_embedded, columns=['dim1','dim2'])\n",
        "    label_df = pd.DataFrame(data=labels,columns=['label'])\n",
        "    feature_label_df = pd.concat((feature_embedded_df,label_df),axis=1)\n",
        "    plt.title(\"T-SNE Representation of the 84 Dimensional Linear Layer\")\n",
        "    sns.scatterplot(data=feature_label_df, hue='label', palette=sns.color_palette(\"hls\", 10),x='dim1', y='dim2',size=6)\n",
        "  \n",
        "  #visualize convergence for train, valid, and test sets\n",
        "  def visualize_convergence(self,val_loss, train_loss, test_loss):\n",
        "    plt.title(\"Loss Convergence\")\n",
        "    plt.xlabel(\"Iteration\")\n",
        "    plt.ylabel(\"Loss\")\n",
        "    X = np.arange(EPOCHS)\n",
        "    plt.plot(X,val_loss,label=\"val_loss\") \n",
        "    plt.plot(X,train_loss,label=\"train_loss\") \n",
        "    plt.plot(X,test_loss,label=\"test_loss\") \n",
        "    plt.legend()\n",
        "    plt.show()\n"
      ],
      "id": "55f019ec",
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "3bdc16a6",
        "outputId": "ea222641-10fe-4983-e053-76a13c263cda"
      },
      "source": [
        "vis_net = visualization(Net)\n",
        "print(Net.val_loss)\n",
        "vis_net.visualize_convergence(Net.val_loss, Net.train_loss, Net.test_loss)"
      ],
      "id": "3bdc16a6",
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.03446299210190773, 0.02647792175412178, 0.016649732366204262, 0.02372211031615734, 0.020116614177823067, 0.021811947226524353, 0.021854504942893982, 0.017783982679247856, 0.02011740207672119, 0.019479045644402504, 0.01741885580122471, 0.017364012077450752, 0.016284838318824768, 0.016471976414322853, 0.01655721664428711, 0.015093924477696419, 0.016185348853468895, 0.01628507487475872, 0.015819275751709938, 0.016184544190764427]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6398ade6"
      },
      "source": [
        ""
      ],
      "id": "6398ade6",
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cm4qBrTAobFV",
        "outputId": "491ab808-65b6-4731-a0ce-679fe8830025"
      },
      "source": [
        "print(Net.failed_tests_val)\n",
        "print(Net.failed_tests_train)\n",
        "print(Net.failed_tests_test)"
      ],
      "id": "cm4qBrTAobFV",
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[7, 9], [9, 3], [8, 6], [8, 0], [0, 6], [5, 3]]\n",
            "[]\n",
            "[[2, 7], [8, 6], [3, 9], [8, 6], [9, 5], [9, 4], [9, 5], [7, 5], [4, 1], [3, 8]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "SJv26eiTxBZn",
        "outputId": "6bf052df-9cef-4c1b-f795-155a19a4fe58"
      },
      "source": [
        "vis_net.visualize_tsne()\n",
        "  "
      ],
      "id": "SJv26eiTxBZn",
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "itWYNzs4NdIN",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 923
        },
        "outputId": "82a7f85a-1e38-4769-b5b3-b58112c2def3"
      },
      "source": [
        "image, label = ds[19]\n",
        "vis_net.visualize_feature(image, label, 2)"
      ],
      "id": "itWYNzs4NdIN",
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 923
        },
        "id": "Uf9azuvIe-nt",
        "outputId": "cf8fc5fd-8167-408c-bf6e-6ac025ffbb3e"
      },
      "source": [
        "image, label = ds[31]\n",
        "vis_net.visualize_feature(image, label, 2)"
      ],
      "id": "Uf9azuvIe-nt",
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KLCuSaF3um4W"
      },
      "source": [
        ""
      ],
      "id": "KLCuSaF3um4W",
      "execution_count": null,
      "outputs": []
    }
  ]
}