lncc_dl_2020.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "lncc-dl-2020.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "936c5f6b16174f92bc5efe92401c31b2": {
          "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_ffda42f0b76a4db58b59b953219716ed",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_19eca837d3c543759e2c146a5fb5ed92",
              "IPY_MODEL_9b34d7b8e31a483582351ee1bb294577"
            ]
          }
        },
        "ffda42f0b76a4db58b59b953219716ed": {
          "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
          }
        },
        "19eca837d3c543759e2c146a5fb5ed92": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_1e011ddc05304e0fb442929f0070684f",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 1,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_3da36fb56d8b42c7934f89721ce347f9"
          }
        },
        "9b34d7b8e31a483582351ee1bb294577": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_e6153d76a3a644db9696d9299a8ea256",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "9920512it [00:01, 8236460.53it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_e3884fed39d6477498ef425f2809e865"
          }
        },
        "1e011ddc05304e0fb442929f0070684f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_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"
          }
        },
        "3da36fb56d8b42c7934f89721ce347f9": {
          "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
          }
        },
        "e6153d76a3a644db9696d9299a8ea256": {
          "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"
          }
        },
        "e3884fed39d6477498ef425f2809e865": {
          "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
          }
        },
        "f25f4a4612034e92b5f5f8990a92e5bc": {
          "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_eecb7599d2494199832dd3ae4d31e8a9",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_789d2a7bfae74fc9845f0bc303fa26e1",
              "IPY_MODEL_403d0274e94345c5bdfabab1cedf42a6"
            ]
          }
        },
        "eecb7599d2494199832dd3ae4d31e8a9": {
          "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
          }
        },
        "789d2a7bfae74fc9845f0bc303fa26e1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_604643cbd9b14fad95b78c7e76b9bcee",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 1,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_9a26f52f059f4386a6182c90018b4eb6"
          }
        },
        "403d0274e94345c5bdfabab1cedf42a6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_ab6b42e326c54ae099e873361a3f571d",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "32768it [00:00, 119095.40it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_087a96d4328740bb8968e0a9cd0e49d4"
          }
        },
        "604643cbd9b14fad95b78c7e76b9bcee": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_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"
          }
        },
        "9a26f52f059f4386a6182c90018b4eb6": {
          "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
          }
        },
        "ab6b42e326c54ae099e873361a3f571d": {
          "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"
          }
        },
        "087a96d4328740bb8968e0a9cd0e49d4": {
          "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
          }
        },
        "5c65d631f07c4f80834e8a6e46341509": {
          "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_8997e1cd30444cb3abc698533cfdffcc",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_6d93bb3d00be434abd346a4912a35760",
              "IPY_MODEL_d9937e2602bf4052914513168cb3ac07"
            ]
          }
        },
        "8997e1cd30444cb3abc698533cfdffcc": {
          "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
          }
        },
        "6d93bb3d00be434abd346a4912a35760": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_43a432073d6b4812bf90914fdc76a588",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 1,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_913cf542a769493083d0c41a4abf9147"
          }
        },
        "d9937e2602bf4052914513168cb3ac07": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_af48058b3ac848eba95d001e52cc5bbd",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "1654784it [00:00, 2068399.99it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_c2c42d75ea0c40b7be02c93778fa042b"
          }
        },
        "43a432073d6b4812bf90914fdc76a588": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_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"
          }
        },
        "913cf542a769493083d0c41a4abf9147": {
          "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
          }
        },
        "af48058b3ac848eba95d001e52cc5bbd": {
          "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"
          }
        },
        "c2c42d75ea0c40b7be02c93778fa042b": {
          "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
          }
        },
        "d490c1ac5ccf4b6696338331cbdbd24f": {
          "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_9f0c4f030eb5487d8ff05f8947b8bde0",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_311e748f5c744c60b08b9f9b123cb575",
              "IPY_MODEL_577c299761804717b0840d2ebf3a6dd9"
            ]
          }
        },
        "9f0c4f030eb5487d8ff05f8947b8bde0": {
          "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
          }
        },
        "311e748f5c744c60b08b9f9b123cb575": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_9ee4d088f3324c69985880e56917e1f1",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 1,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_7b9e0a0a549244879b8b910f6018e0b8"
          }
        },
        "577c299761804717b0840d2ebf3a6dd9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_9494c4b93a534e9c9a1a23852545e42e",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "8192it [00:00, 41136.04it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_65f015210c38460b95c2b4f51cc92549"
          }
        },
        "9ee4d088f3324c69985880e56917e1f1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_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"
          }
        },
        "7b9e0a0a549244879b8b910f6018e0b8": {
          "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
          }
        },
        "9494c4b93a534e9c9a1a23852545e42e": {
          "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"
          }
        },
        "65f015210c38460b95c2b4f51cc92549": {
          "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
          }
        },
        "3db05372ed66410e815168f2bfe0ec62": {
          "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_e46604b9e7d746219bebc1232d2aec09",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_c6dd7987abf54ad6b1236adb9bd55ac6",
              "IPY_MODEL_70b439215cbc4a2baa68156e0a515fc5"
            ]
          }
        },
        "e46604b9e7d746219bebc1232d2aec09": {
          "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
          }
        },
        "c6dd7987abf54ad6b1236adb9bd55ac6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_784998df19e7464dab23482b84d06ba6",
            "_dom_classes": [],
            "description": "",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 1,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_72b5534c96884c948f2532e531926702"
          }
        },
        "70b439215cbc4a2baa68156e0a515fc5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_cc0b376f286245d7acd8895f53966f53",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "170500096it [00:06, 26974951.16it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_164e013aa6844fe59349115576fd1fcb"
          }
        },
        "784998df19e7464dab23482b84d06ba6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_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"
          }
        },
        "72b5534c96884c948f2532e531926702": {
          "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
          }
        },
        "cc0b376f286245d7acd8895f53966f53": {
          "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"
          }
        },
        "164e013aa6844fe59349115576fd1fcb": {
          "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/JonatasFontele/DataScience/blob/main/lncc_dl_2020.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bp-q47-BRoTD"
      },
      "source": [
        "# Créditos\n",
        "\n",
        "Os exemplos de código aqui apresentados são retirados/adaptados de pytorch.org/tutorials/beginner/blitz/"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vTNimckuQret"
      },
      "source": [
        "# Tensores"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wh7q41n4Q38v"
      },
      "source": [
        "import torch"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "68OrHRHeQuDx",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "515ff959-2d1e-43ff-f2d3-01711c3e1a67"
      },
      "source": [
        "x = torch.empty(5, 3)\n",
        "print(x)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[5.1973e-36, 0.0000e+00, 0.0000e+00],\n",
            "        [0.0000e+00, 0.0000e+00, 0.0000e+00],\n",
            "        [0.0000e+00, 0.0000e+00, 2.8026e-45],\n",
            "        [0.0000e+00, 1.1210e-44, 0.0000e+00],\n",
            "        [1.4013e-45, 0.0000e+00, 4.4025e-26]])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_EArdAgNQzT1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "59289e52-1ec1-44c1-a452-222114125e9c"
      },
      "source": [
        "x = torch.rand(5, 3)\n",
        "print(x)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[0.0050, 0.7604, 0.0583],\n",
            "        [0.5998, 0.0466, 0.6768],\n",
            "        [0.8027, 0.3929, 0.4355],\n",
            "        [0.2873, 0.6156, 0.2184],\n",
            "        [0.9358, 0.8216, 0.6143]])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZpLPhqKrQ-3O",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2a5d3328-d530-4c37-c9b7-3bb2ceb51a4a"
      },
      "source": [
        "x = torch.rand(5, 32, 32, 3)\n",
        "print(x)\n",
        "print(x.size())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[[[0.0141, 0.1697, 0.1743],\n",
            "          [0.7074, 0.9593, 0.9388],\n",
            "          [0.3803, 0.6356, 0.7127],\n",
            "          ...,\n",
            "          [0.5693, 0.1630, 0.2542],\n",
            "          [0.7553, 0.1998, 0.6171],\n",
            "          [0.3442, 0.6049, 0.5913]],\n",
            "\n",
            "         [[0.0294, 0.9776, 0.7344],\n",
            "          [0.4349, 0.1854, 0.6400],\n",
            "          [0.3670, 0.9748, 0.5032],\n",
            "          ...,\n",
            "          [0.4587, 0.1876, 0.9926],\n",
            "          [0.0395, 0.0345, 0.2065],\n",
            "          [0.8996, 0.9260, 0.2149]],\n",
            "\n",
            "         [[0.2029, 0.0884, 0.4296],\n",
            "          [0.4700, 0.9035, 0.4185],\n",
            "          [0.0337, 0.6239, 0.8301],\n",
            "          ...,\n",
            "          [0.4344, 0.4465, 0.8549],\n",
            "          [0.8318, 0.2083, 0.2321],\n",
            "          [0.3393, 0.1259, 0.5632]],\n",
            "\n",
            "         ...,\n",
            "\n",
            "         [[0.0734, 0.1134, 0.8877],\n",
            "          [0.8312, 0.6103, 0.7075],\n",
            "          [0.8019, 0.1647, 0.5110],\n",
            "          ...,\n",
            "          [0.3776, 0.4290, 0.7715],\n",
            "          [0.4257, 0.3264, 0.3878],\n",
            "          [0.0249, 0.4002, 0.2310]],\n",
            "\n",
            "         [[0.6827, 0.0823, 0.6353],\n",
            "          [0.3979, 0.1556, 0.9912],\n",
            "          [0.6121, 0.5812, 0.4786],\n",
            "          ...,\n",
            "          [0.1264, 0.9218, 0.9514],\n",
            "          [0.2323, 0.5041, 0.7539],\n",
            "          [0.2439, 0.3248, 0.9866]],\n",
            "\n",
            "         [[0.3238, 0.8880, 0.0986],\n",
            "          [0.4575, 0.3045, 0.5004],\n",
            "          [0.2945, 0.5043, 0.3447],\n",
            "          ...,\n",
            "          [0.1448, 0.2302, 0.0719],\n",
            "          [0.1476, 0.3340, 0.3139],\n",
            "          [0.5977, 0.1812, 0.2688]]],\n",
            "\n",
            "\n",
            "        [[[0.5647, 0.2702, 0.1403],\n",
            "          [0.0970, 0.5285, 0.5360],\n",
            "          [0.4220, 0.1542, 0.6040],\n",
            "          ...,\n",
            "          [0.8749, 0.4862, 0.0155],\n",
            "          [0.5363, 0.0718, 0.9822],\n",
            "          [0.6756, 0.1584, 0.0357]],\n",
            "\n",
            "         [[0.7571, 0.1838, 0.3001],\n",
            "          [0.5569, 0.1224, 0.6557],\n",
            "          [0.0057, 0.4196, 0.3900],\n",
            "          ...,\n",
            "          [0.7500, 0.8641, 0.4795],\n",
            "          [0.7599, 0.2581, 0.3363],\n",
            "          [0.7544, 0.2399, 0.4024]],\n",
            "\n",
            "         [[0.6824, 0.4968, 0.1259],\n",
            "          [0.2330, 0.6070, 0.3028],\n",
            "          [0.2798, 0.5970, 0.3725],\n",
            "          ...,\n",
            "          [0.0668, 0.4859, 0.6704],\n",
            "          [0.3136, 0.2676, 0.3893],\n",
            "          [0.3622, 0.2972, 0.2342]],\n",
            "\n",
            "         ...,\n",
            "\n",
            "         [[0.4718, 0.2311, 0.2780],\n",
            "          [0.5810, 0.9075, 0.5002],\n",
            "          [0.0387, 0.9133, 0.4094],\n",
            "          ...,\n",
            "          [0.3784, 0.6653, 0.9973],\n",
            "          [0.7120, 0.9099, 0.4271],\n",
            "          [0.7628, 0.8804, 0.6461]],\n",
            "\n",
            "         [[0.8135, 0.9989, 0.3030],\n",
            "          [0.0118, 0.7220, 0.1826],\n",
            "          [0.1510, 0.3909, 0.8649],\n",
            "          ...,\n",
            "          [0.4757, 0.2544, 0.7363],\n",
            "          [0.8139, 0.5933, 0.4359],\n",
            "          [0.9391, 0.8513, 0.9727]],\n",
            "\n",
            "         [[0.9940, 0.0853, 0.8983],\n",
            "          [0.5353, 0.7838, 0.9062],\n",
            "          [0.6533, 0.8078, 0.4153],\n",
            "          ...,\n",
            "          [0.1591, 0.3943, 0.6093],\n",
            "          [0.3339, 0.2405, 0.4081],\n",
            "          [0.5171, 0.3402, 0.8664]]],\n",
            "\n",
            "\n",
            "        [[[0.3154, 0.9379, 0.7429],\n",
            "          [0.2048, 0.7177, 0.0040],\n",
            "          [0.4375, 0.4504, 0.1926],\n",
            "          ...,\n",
            "          [0.8093, 0.4962, 0.2806],\n",
            "          [0.2408, 0.4063, 0.7957],\n",
            "          [0.1213, 0.0986, 0.5275]],\n",
            "\n",
            "         [[0.8984, 0.3890, 0.1551],\n",
            "          [0.9877, 0.0429, 0.7639],\n",
            "          [0.3881, 0.6312, 0.7246],\n",
            "          ...,\n",
            "          [0.7995, 0.7343, 0.2180],\n",
            "          [0.2599, 0.6546, 0.3640],\n",
            "          [0.2713, 0.3847, 0.3020]],\n",
            "\n",
            "         [[0.7195, 0.2820, 0.4872],\n",
            "          [0.4672, 0.8540, 0.1650],\n",
            "          [0.5968, 0.9238, 0.8211],\n",
            "          ...,\n",
            "          [0.8689, 0.7807, 0.8055],\n",
            "          [0.5585, 0.1854, 0.1611],\n",
            "          [0.2207, 0.5949, 0.6608]],\n",
            "\n",
            "         ...,\n",
            "\n",
            "         [[0.0995, 0.3390, 0.9328],\n",
            "          [0.8288, 0.4485, 0.1249],\n",
            "          [0.2314, 0.9492, 0.8125],\n",
            "          ...,\n",
            "          [0.2276, 0.2176, 0.4147],\n",
            "          [0.2113, 0.0309, 0.3689],\n",
            "          [0.7314, 0.3254, 0.9308]],\n",
            "\n",
            "         [[0.3885, 0.5125, 0.3854],\n",
            "          [0.0020, 0.8281, 0.5767],\n",
            "          [0.9125, 0.4099, 0.5149],\n",
            "          ...,\n",
            "          [0.7856, 0.4602, 0.0946],\n",
            "          [0.8055, 0.1380, 0.2125],\n",
            "          [0.8044, 0.3114, 0.4581]],\n",
            "\n",
            "         [[0.4901, 0.9146, 0.9264],\n",
            "          [0.0193, 0.9309, 0.6372],\n",
            "          [0.9927, 0.9927, 0.8032],\n",
            "          ...,\n",
            "          [0.0225, 0.8057, 0.6500],\n",
            "          [0.1851, 0.1744, 0.0431],\n",
            "          [0.6675, 0.3458, 0.4437]]],\n",
            "\n",
            "\n",
            "        [[[0.5627, 0.1379, 0.4582],\n",
            "          [0.0661, 0.1478, 0.5375],\n",
            "          [0.6665, 0.0684, 0.0665],\n",
            "          ...,\n",
            "          [0.0881, 0.6878, 0.1994],\n",
            "          [0.0346, 0.8402, 0.9854],\n",
            "          [0.5754, 0.0020, 0.3134]],\n",
            "\n",
            "         [[0.2591, 0.3827, 0.4916],\n",
            "          [0.8874, 0.0167, 0.9967],\n",
            "          [0.7881, 0.0692, 0.8853],\n",
            "          ...,\n",
            "          [0.6570, 0.9085, 0.9343],\n",
            "          [0.4751, 0.8999, 0.7039],\n",
            "          [0.2549, 0.2337, 0.3212]],\n",
            "\n",
            "         [[0.3984, 0.9884, 0.0471],\n",
            "          [0.2517, 0.5574, 0.4909],\n",
            "          [0.0087, 0.8613, 0.9300],\n",
            "          ...,\n",
            "          [0.7852, 0.9460, 0.1055],\n",
            "          [0.0225, 0.8918, 0.4500],\n",
            "          [0.3461, 0.4091, 0.8374]],\n",
            "\n",
            "         ...,\n",
            "\n",
            "         [[0.8467, 0.7563, 0.5295],\n",
            "          [0.9131, 0.0120, 0.7754],\n",
            "          [0.1275, 0.4226, 0.4265],\n",
            "          ...,\n",
            "          [0.1711, 0.2056, 0.0933],\n",
            "          [0.2994, 0.5824, 0.3281],\n",
            "          [0.1689, 0.1152, 0.2976]],\n",
            "\n",
            "         [[0.1967, 0.4656, 0.7013],\n",
            "          [0.9358, 0.2815, 0.3853],\n",
            "          [0.2677, 0.0080, 0.0735],\n",
            "          ...,\n",
            "          [0.7798, 0.8529, 0.3558],\n",
            "          [0.1504, 0.7750, 0.7109],\n",
            "          [0.5807, 0.0336, 0.0400]],\n",
            "\n",
            "         [[0.5100, 0.4680, 0.1328],\n",
            "          [0.4028, 0.2414, 0.8025],\n",
            "          [0.7427, 0.5825, 0.0849],\n",
            "          ...,\n",
            "          [0.9313, 0.0918, 0.4180],\n",
            "          [0.2130, 0.0558, 0.6845],\n",
            "          [0.0849, 0.0801, 0.5001]]],\n",
            "\n",
            "\n",
            "        [[[0.3217, 0.0796, 0.5862],\n",
            "          [0.8688, 0.9746, 0.3293],\n",
            "          [0.1583, 0.8226, 0.6779],\n",
            "          ...,\n",
            "          [0.8321, 0.9644, 0.3064],\n",
            "          [0.0390, 0.2483, 0.5427],\n",
            "          [0.7947, 0.9476, 0.7813]],\n",
            "\n",
            "         [[0.8599, 0.1004, 0.0977],\n",
            "          [0.2740, 0.8063, 0.6938],\n",
            "          [0.0883, 0.1062, 0.8336],\n",
            "          ...,\n",
            "          [0.7106, 0.3484, 0.3263],\n",
            "          [0.3119, 0.7313, 0.7361],\n",
            "          [0.7827, 0.0375, 0.5220]],\n",
            "\n",
            "         [[0.4957, 0.3275, 0.1357],\n",
            "          [0.3926, 0.9314, 0.1936],\n",
            "          [0.0646, 0.6119, 0.0545],\n",
            "          ...,\n",
            "          [0.4933, 0.0162, 0.3977],\n",
            "          [0.4440, 0.8802, 0.4118],\n",
            "          [0.0709, 0.1432, 0.9629]],\n",
            "\n",
            "         ...,\n",
            "\n",
            "         [[0.4360, 0.0046, 0.3293],\n",
            "          [0.7371, 0.3421, 0.4211],\n",
            "          [0.7934, 0.3253, 0.9995],\n",
            "          ...,\n",
            "          [0.2500, 0.5250, 0.9344],\n",
            "          [0.5766, 0.1838, 0.2592],\n",
            "          [0.0369, 0.0596, 0.2751]],\n",
            "\n",
            "         [[0.9038, 0.3508, 0.5141],\n",
            "          [0.8530, 0.7175, 0.9975],\n",
            "          [0.0673, 0.7675, 0.8819],\n",
            "          ...,\n",
            "          [0.7576, 0.0508, 0.6564],\n",
            "          [0.0128, 0.7275, 0.0118],\n",
            "          [0.6538, 0.8005, 0.8378]],\n",
            "\n",
            "         [[0.6460, 0.7482, 0.7273],\n",
            "          [0.3777, 0.5160, 0.2080],\n",
            "          [0.1113, 0.4474, 0.1049],\n",
            "          ...,\n",
            "          [0.9052, 0.0155, 0.2758],\n",
            "          [0.6885, 0.8493, 0.7601],\n",
            "          [0.6497, 0.3009, 0.3738]]]])\n",
            "torch.Size([5, 32, 32, 3])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x0ngvqPmLZVU"
      },
      "source": [
        "# AutoGrad"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L8bU2YdKLNvW"
      },
      "source": [
        "import torch"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TrfW8DUyLcMu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "baa935b3-3fd6-40f4-eeab-9eb8103d4626"
      },
      "source": [
        "x = torch.ones(2, 2, requires_grad=True)\n",
        "print(x)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[1., 1.],\n",
            "        [1., 1.]], requires_grad=True)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IlnFReWmLgCf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 86
        },
        "outputId": "a1310ea0-432d-4c45-abcf-1eebacbc7a0b"
      },
      "source": [
        "y = x + 2\n",
        "print(y)\n",
        "\n",
        "z = y * y * 3\n",
        "out = z.mean()\n",
        "\n",
        "print(z, out)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[3., 3.],\n",
            "        [3., 3.]], grad_fn=<AddBackward0>)\n",
            "tensor([[27., 27.],\n",
            "        [27., 27.]], grad_fn=<MulBackward0>) tensor(27., grad_fn=<MeanBackward0>)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IWemyBcKP5R_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "7564c9c7-e584-42df-d438-7d7464fd3712"
      },
      "source": [
        "out.backward()\n",
        "print(x.grad)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[4.5000, 4.5000],\n",
            "        [4.5000, 4.5000]])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EdANy-fILuQe"
      },
      "source": [
        "$$\n",
        "o = \\frac{1}{4}\\sum_i z_i\n",
        "$$\n",
        "\n",
        "$$\n",
        "z_i = 3(x_i+2)^2\n",
        "$$\n",
        "\n",
        "$$\n",
        "\\frac{\\partial o}{\\partial x_i} = \\frac{3}{2}(x_i+2)\n",
        "$$\n",
        "\n",
        "$$\n",
        "\\left(z_i\\bigr\\rvert_{x_i=1} = 27\\right) \\Rightarrow \\left(\\frac{\\partial o}{\\partial x_i}\\bigr\\rvert_{x_i=1} = \\frac{9}{2} = 4.5\\right)\n",
        "$$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PvBnin2nLUho"
      },
      "source": [
        "# CrossEntropyLoss"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RGFv7iRdh5Kv",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 225
        },
        "outputId": "8dbde150-6550-465a-e568-5e4eb2570f87"
      },
      "source": [
        "loss = nn.CrossEntropyLoss()\n",
        "input = torch.randn(10, 3, requires_grad=True)\n",
        "target = torch.empty(10, dtype=torch.long).random_(3)\n",
        "output = loss(input, target)\n",
        "output.backward()\n",
        "print('input:', input)\n",
        "print('target:', target)\n",
        "print('output:', output)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "input: tensor([[-0.2203, -0.4656,  0.1692],\n",
            "        [ 0.2702, -0.2993, -1.9377],\n",
            "        [-0.4649, -0.5994, -1.2975],\n",
            "        [-0.4820,  1.5314,  1.0974],\n",
            "        [ 2.5178,  1.9212, -0.3888],\n",
            "        [ 0.8948,  2.1120, -0.9386],\n",
            "        [ 1.0620,  0.6840,  1.2253],\n",
            "        [-0.0890, -0.5516, -1.5667],\n",
            "        [ 1.5453,  1.2641, -0.8904],\n",
            "        [-1.6637,  2.5074, -0.5465]], requires_grad=True)\n",
            "target: tensor([0, 1, 0, 1, 1, 2, 1, 0, 1, 1])\n",
            "output: tensor(1.1099, grad_fn=<NllLossBackward>)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TPb7XULGmAKd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "aaaf51a9-1710-4b9f-98d7-b59c11c394cc"
      },
      "source": [
        "input = torch.FloatTensor(\n",
        "       [[0.2046, -0.3144, 0.2438],\n",
        "        [-0.0641,  0.2295, -0.1240],\n",
        "        [ 1.4380,  1.6523,  0.3243],\n",
        "        [ 1.4235,  0.5854, -0.9747],\n",
        "        [ 3.1441, -1.2237, 10.2169],\n",
        "        [ 0.2066,  0.1723,  1.0045],\n",
        "        [-0.2114,  1.1299,  1.3391],\n",
        "        [ 1.1068,  0.5069,  0.8790],\n",
        "        [-0.7555,  0.0312, -0.2468],\n",
        "        [ 0.1276,  0.3208, -1.2589]])\n",
        "target = torch.LongTensor([2, 1, 1, 0, 2, 0, 2, 1, 1, 2])\n",
        "output = loss(input, target)\n",
        "print('output:', output)\n",
        "\n",
        "madeup_target = torch.LongTensor([2, 1, 1, 0, 2, 2, 2, 0, 1, 1])\n",
        "madeup_output = loss(input, madeup_target)\n",
        "print('output:', madeup_output)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "output: tensor(0.9646)\n",
            "output: tensor(0.6669)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "24RlUhLfrsIe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "cee143e4-53ef-4373-dabc-3e03d5d07cc2"
      },
      "source": [
        "import numpy as np\n",
        "\n",
        "l1 = [[0.2046, 0.3144, 0.2438],\n",
        "      [0.1876, 0.3208, 1.2589]]\n",
        "\n",
        "l2 = [1, 2]\n",
        "\n",
        "input = torch.FloatTensor(l1)\n",
        "madeup_target = torch.LongTensor(l2)\n",
        "madeup_output = loss(input, madeup_target)\n",
        "print('output:', madeup_output)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "output: tensor(0.7950)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gf8odD2TTkvF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "34d8aa4c-7709-4d91-971a-32cc0c3cf17e"
      },
      "source": [
        "0.7950*2"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.59"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 72
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MTDe2gpF1woc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 155
        },
        "outputId": "1fec602c-617d-46dd-c088-e8c9cac1de3f"
      },
      "source": [
        "def celoss(x,a_class):\n",
        "  denominator = np.sum(np.exp(x))\n",
        "  return -np.log(np.exp(x[a_class])/denominator)\n",
        "\n",
        "l1 = [[0.2046, 0.3144, 0.2438],\n",
        "      [0.1876, 0.3208, 1.2589]]\n",
        "a1 = np.asarray(l1)\n",
        "print(a1[0])\n",
        "print(a1[1])\n",
        "\n",
        "print(compute_celoss(a1[0],0) + celoss(a1[1],0))\n",
        "print(compute_celoss(a1[0],0) + celoss(a1[1],1))\n",
        "print(celoss(a1[0],0) + celoss(a1[1],2))\n",
        "\n",
        "print(celoss(a1[0],1) + celoss(a1[1],0))\n",
        "print(celoss(a1[0],1) + celoss(a1[1],1))\n",
        "print(celoss(a1[0],1) + celoss(a1[1],2)) # classificação perfeita\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.2046 0.3144 0.2438]\n",
            "[0.1876 0.3208 1.2589]\n",
            "2.771008554769684\n",
            "2.637808554769684\n",
            "1.6997085547696842\n",
            "2.661208554769684\n",
            "2.5280085547696842\n",
            "1.5899085547696843\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-RHFdRVj6YX4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 225
        },
        "outputId": "94d4a4f9-8f15-4948-b61d-c030759fb359"
      },
      "source": [
        "m = nn.Softmax(dim=1)\n",
        "input = torch.randn(2, 3)\n",
        "print('input:\\n', input)\n",
        "output = m(input)\n",
        "print('output:\\n', output)\n",
        "print('output[0]:\\n', output[0])\n",
        "print('output[1]:\\n', output[1])\n",
        "\n",
        "print(torch.sum(output[0]))\n",
        "print(torch.sum(output[1]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "input:\n",
            " tensor([[ 0.6527,  0.8684, -0.4120],\n",
            "        [-0.0815, -0.3312, -0.7317]])\n",
            "output:\n",
            " tensor([[0.3868, 0.4799, 0.1334],\n",
            "        [0.4346, 0.3386, 0.2268]])\n",
            "output[0]:\n",
            " tensor([0.3868, 0.4799, 0.1334])\n",
            "output[1]:\n",
            " tensor([0.4346, 0.3386, 0.2268])\n",
            "tensor(1.0000)\n",
            "tensor(1.)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v3n6Tn1G-uIA"
      },
      "source": [
        "![alt text](https://ml-cheatsheet.readthedocs.io/en/latest/_images/cross_entropy.png)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lQgB1fkBQGon"
      },
      "source": [
        "# Dataset Iris"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rBTk8IJJH_WN",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 407
        },
        "outputId": "55bd05ba-a1d3-480f-9371-977abd36f14c"
      },
      "source": [
        "#load\n",
        "dataset = pd.read_csv('./iris.csv')\n",
        "dataset.head(150)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>sepal_length</th>\n",
              "      <th>sepal_width</th>\n",
              "      <th>petal_length</th>\n",
              "      <th>petal_width</th>\n",
              "      <th>species</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5.1</td>\n",
              "      <td>3.5</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "      <td>Iris-setosa</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4.9</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "      <td>Iris-setosa</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>4.7</td>\n",
              "      <td>3.2</td>\n",
              "      <td>1.3</td>\n",
              "      <td>0.2</td>\n",
              "      <td>Iris-setosa</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4.6</td>\n",
              "      <td>3.1</td>\n",
              "      <td>1.5</td>\n",
              "      <td>0.2</td>\n",
              "      <td>Iris-setosa</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5.0</td>\n",
              "      <td>3.6</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "      <td>Iris-setosa</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>145</th>\n",
              "      <td>6.7</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.2</td>\n",
              "      <td>2.3</td>\n",
              "      <td>Iris-virginica</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>146</th>\n",
              "      <td>6.3</td>\n",
              "      <td>2.5</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.9</td>\n",
              "      <td>Iris-virginica</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>147</th>\n",
              "      <td>6.5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.2</td>\n",
              "      <td>2.0</td>\n",
              "      <td>Iris-virginica</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>148</th>\n",
              "      <td>6.2</td>\n",
              "      <td>3.4</td>\n",
              "      <td>5.4</td>\n",
              "      <td>2.3</td>\n",
              "      <td>Iris-virginica</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>149</th>\n",
              "      <td>5.9</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.1</td>\n",
              "      <td>1.8</td>\n",
              "      <td>Iris-virginica</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>150 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     sepal_length  sepal_width  petal_length  petal_width         species\n",
              "0             5.1          3.5           1.4          0.2     Iris-setosa\n",
              "1             4.9          3.0           1.4          0.2     Iris-setosa\n",
              "2             4.7          3.2           1.3          0.2     Iris-setosa\n",
              "3             4.6          3.1           1.5          0.2     Iris-setosa\n",
              "4             5.0          3.6           1.4          0.2     Iris-setosa\n",
              "..            ...          ...           ...          ...             ...\n",
              "145           6.7          3.0           5.2          2.3  Iris-virginica\n",
              "146           6.3          2.5           5.0          1.9  Iris-virginica\n",
              "147           6.5          3.0           5.2          2.0  Iris-virginica\n",
              "148           6.2          3.4           5.4          2.3  Iris-virginica\n",
              "149           5.9          3.0           5.1          1.8  Iris-virginica\n",
              "\n",
              "[150 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-J5zBE--bj9l"
      },
      "source": [
        "# Demo: rede neural simples"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1fpbykKcfCvC"
      },
      "source": [
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score, precision_score, recall_score\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "from torch.autograd import Variable\n",
        "\n",
        "class Net(nn.Module):\n",
        "    # define nn\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.fc1 = nn.Linear(4, 100)\n",
        "        self.fc2 = nn.Linear(100, 3)\n",
        "        self.softmax = nn.Softmax(dim=1)\n",
        "\n",
        "    def forward(self, X):\n",
        "        X = F.relu(self.fc1(X))\n",
        "        X = F.relu(self.fc2(X))\n",
        "        X = self.softmax(X)\n",
        "\n",
        "        return X\n",
        "    \n",
        "# load IRIS dataset\n",
        "dataset = pd.read_csv('./iris.csv')\n",
        "\n",
        "# transform species to numerics\n",
        "dataset.loc[dataset.species=='Iris-setosa', 'species'] = 0\n",
        "dataset.loc[dataset.species=='Iris-versicolor', 'species'] = 1\n",
        "dataset.loc[dataset.species=='Iris-virginica', 'species'] = 2\n",
        "\n",
        "train_X, test_X, train_y, test_y = train_test_split(dataset[dataset.columns[0:4]].values,\n",
        "                                                    dataset.species.values, \n",
        "                                                    test_size=0.8, \n",
        "                                                    stratify=dataset.species)\n",
        "\n",
        "# wrap up with Variable in pytorch\n",
        "train_X = Variable(torch.Tensor(train_X).float())\n",
        "test_X = Variable(torch.Tensor(test_X).float())\n",
        "train_y = Variable(torch.Tensor(train_y).long())\n",
        "test_y = Variable(torch.Tensor(test_y).long())\n",
        "\n",
        "\n",
        "net = Net()\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()# cross entropy loss\n",
        "\n",
        "optimizer = torch.optim.SGD(net.parameters(), lr=0.001)\n",
        "\n",
        "for epoch in range(100):\n",
        "    optimizer.zero_grad()\n",
        "    out = net(train_X)\n",
        "    loss = criterion(out, train_y)\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "    \n",
        "    if epoch % 10 == 0:\n",
        "        print ('number of epoch', epoch, 'loss', loss.data.item())\n",
        "\n",
        "predict_out = net(test_X)\n",
        "_, predict_y = torch.max(predict_out, 1)\n",
        "\n",
        "print ('prediction accuracy', accuracy_score(test_y.data, predict_y.data))\n",
        "\n",
        "print ('macro precision', precision_score(test_y.data, predict_y.data, average='macro'))\n",
        "print ('micro precision', precision_score(test_y.data, predict_y.data, average='micro'))\n",
        "print ('macro recall', recall_score(test_y.data, predict_y.data, average='macro'))\n",
        "print ('micro recall', recall_score(test_y.data, predict_y.data, average='micro'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I0NlcDEsjg7i",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 898
        },
        "outputId": "6bb19a8a-0c57-44b1-a0e8-d821a2b5c9bf"
      },
      "source": [
        "from __future__ import print_function\n",
        "from builtins import range\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "\"\"\"\n",
        "SECTION 1 : Load and setup data for training\n",
        "the datasets separated in two files from originai datasets:\n",
        "iris_train.csv = datasets for training purpose, 80% from the original data\n",
        "iris_test.csv  = datasets for testing purpose, 20% from the original data\n",
        "\"\"\"\n",
        "import pandas as pd\n",
        "\n",
        "#load\n",
        "dataset = pd.read_csv('./iris.csv')\n",
        "\n",
        "#change string value to numeric\n",
        "dataset.loc[dataset['species']=='Iris-setosa', 'species']=0\n",
        "dataset.loc[dataset['species']=='Iris-versicolor', 'species']=1\n",
        "dataset.loc[dataset['species']=='Iris-virginica', 'species']=2\n",
        "dataset = dataset.apply(pd.to_numeric)\n",
        "\n",
        "#change dataframe to array\n",
        "datatrain_array = dataset.values\n",
        "\n",
        "#split x and y (feature and target)\n",
        "xtrain = datatrain_array[:,:4]\n",
        "ytrain = datatrain_array[:,4]\n",
        "\n",
        "xtrain, xtest, ytrain, ytest = train_test_split(dataset[dataset.columns[0:4]].values,\n",
        "                                                    dataset.species.values, test_size=0.8, \n",
        "                                                    stratify=dataset.species)\n",
        "\n",
        "\"\"\"\n",
        "SECTION 2 : Build and Train Model\n",
        "Multilayer perceptron model, with one hidden layer.\n",
        "input layer : 4 neuron, represents the feature of Iris\n",
        "hidden layer : 3 neuron, activation using ReLU\n",
        "output layer : 3 neuron, represents the class of Iris\n",
        "optimizer = stochastic gradient descent with no batch-size\n",
        "loss function = categorical cross entropy\n",
        "learning rate = 0.01\n",
        "epoch = 50\n",
        "\"\"\"\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "torch.manual_seed(1234)\n",
        "\n",
        "#hyperparameters\n",
        "hl = 3\n",
        "lr = 0.01\n",
        "num_epoch = 50\n",
        "\n",
        "#build model\n",
        "class Net(nn.Module):\n",
        "\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.fc1 = nn.Linear(4, hl)\n",
        "        self.fc2 = nn.Linear(hl, 3)\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = self.fc2(x)\n",
        "        return x\n",
        "        \n",
        "net = Net()\n",
        "\n",
        "#choose optimizer and loss function\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.SGD(net.parameters(), lr=lr)\n",
        "\n",
        "#train\n",
        "for epoch in range(num_epoch):\n",
        "    X = torch.Tensor(xtrain).float()\n",
        "    Y = torch.Tensor(ytrain).long()\n",
        "\n",
        "    #feedforward - backprop\n",
        "    optimizer.zero_grad()\n",
        "    out = net(X)\n",
        "    loss = criterion(out, Y)\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "    acc = 100 * torch.sum(Y==torch.max(out.data, 1)[1]).double() / len(Y)\n",
        "    print ('Epoch [%d/%d] Loss: %.4f   Acc: %.4f' \n",
        "                   %(epoch+1, num_epoch, loss.item(), acc.item()))\n",
        "\n",
        "\n",
        "\"\"\"\n",
        "SECTION 3 : Testing model\n",
        "\"\"\"\n",
        "\n",
        "#get prediction\n",
        "X = torch.Tensor(xtest).float()\n",
        "Y = torch.Tensor(ytest).long()\n",
        "out = net(X)\n",
        "_, predicted = torch.max(out.data, 1)\n",
        "\n",
        "#get accuration\n",
        "print('Accuracy of the network %.4f %%' % (100 * torch.sum(Y==predicted).double() / len(Y)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch [1/50] Loss: 1.1443   Acc: 33.3333\n",
            "Epoch [2/50] Loss: 1.1406   Acc: 33.3333\n",
            "Epoch [3/50] Loss: 1.1371   Acc: 33.3333\n",
            "Epoch [4/50] Loss: 1.1336   Acc: 33.3333\n",
            "Epoch [5/50] Loss: 1.1302   Acc: 33.3333\n",
            "Epoch [6/50] Loss: 1.1269   Acc: 33.3333\n",
            "Epoch [7/50] Loss: 1.1237   Acc: 33.3333\n",
            "Epoch [8/50] Loss: 1.1205   Acc: 33.3333\n",
            "Epoch [9/50] Loss: 1.1174   Acc: 33.3333\n",
            "Epoch [10/50] Loss: 1.1143   Acc: 33.3333\n",
            "Epoch [11/50] Loss: 1.1113   Acc: 33.3333\n",
            "Epoch [12/50] Loss: 1.1083   Acc: 33.3333\n",
            "Epoch [13/50] Loss: 1.1053   Acc: 33.3333\n",
            "Epoch [14/50] Loss: 1.1024   Acc: 33.3333\n",
            "Epoch [15/50] Loss: 1.0995   Acc: 33.3333\n",
            "Epoch [16/50] Loss: 1.0966   Acc: 33.3333\n",
            "Epoch [17/50] Loss: 1.0938   Acc: 33.3333\n",
            "Epoch [18/50] Loss: 1.0910   Acc: 33.3333\n",
            "Epoch [19/50] Loss: 1.0882   Acc: 33.3333\n",
            "Epoch [20/50] Loss: 1.0854   Acc: 33.3333\n",
            "Epoch [21/50] Loss: 1.0827   Acc: 33.3333\n",
            "Epoch [22/50] Loss: 1.0799   Acc: 33.3333\n",
            "Epoch [23/50] Loss: 1.0772   Acc: 33.3333\n",
            "Epoch [24/50] Loss: 1.0745   Acc: 33.3333\n",
            "Epoch [25/50] Loss: 1.0718   Acc: 33.3333\n",
            "Epoch [26/50] Loss: 1.0692   Acc: 33.3333\n",
            "Epoch [27/50] Loss: 1.0665   Acc: 33.3333\n",
            "Epoch [28/50] Loss: 1.0639   Acc: 33.3333\n",
            "Epoch [29/50] Loss: 1.0613   Acc: 33.3333\n",
            "Epoch [30/50] Loss: 1.0587   Acc: 33.3333\n",
            "Epoch [31/50] Loss: 1.0561   Acc: 33.3333\n",
            "Epoch [32/50] Loss: 1.0536   Acc: 33.3333\n",
            "Epoch [33/50] Loss: 1.0510   Acc: 33.3333\n",
            "Epoch [34/50] Loss: 1.0485   Acc: 33.3333\n",
            "Epoch [35/50] Loss: 1.0460   Acc: 36.6667\n",
            "Epoch [36/50] Loss: 1.0436   Acc: 36.6667\n",
            "Epoch [37/50] Loss: 1.0411   Acc: 40.0000\n",
            "Epoch [38/50] Loss: 1.0387   Acc: 40.0000\n",
            "Epoch [39/50] Loss: 1.0363   Acc: 43.3333\n",
            "Epoch [40/50] Loss: 1.0340   Acc: 53.3333\n",
            "Epoch [41/50] Loss: 1.0316   Acc: 56.6667\n",
            "Epoch [42/50] Loss: 1.0293   Acc: 63.3333\n",
            "Epoch [43/50] Loss: 1.0271   Acc: 66.6667\n",
            "Epoch [44/50] Loss: 1.0248   Acc: 76.6667\n",
            "Epoch [45/50] Loss: 1.0226   Acc: 80.0000\n",
            "Epoch [46/50] Loss: 1.0204   Acc: 90.0000\n",
            "Epoch [47/50] Loss: 1.0182   Acc: 90.0000\n",
            "Epoch [48/50] Loss: 1.0161   Acc: 93.3333\n",
            "Epoch [49/50] Loss: 1.0140   Acc: 90.0000\n",
            "Epoch [50/50] Loss: 1.0120   Acc: 90.0000\n",
            "Accuracy of the network 83.3333 %\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v6dOz4thRc5j"
      },
      "source": [
        "import torch\n",
        "import numpy as np"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pOa7YacsQ_sz"
      },
      "source": [
        "class Feedforward(torch.nn.Module):\n",
        "        def __init__(self, input_size, hidden_size):\n",
        "            super(Feedforward, self).__init__()\n",
        "            self.input_size = input_size\n",
        "            self.hidden_size  = hidden_size\n",
        "            self.fc1 = torch.nn.Linear(self.input_size, self.hidden_size)\n",
        "            self.relu = torch.nn.ReLU()\n",
        "            self.fc2 = torch.nn.Linear(self.hidden_size, 1)\n",
        "            self.sigmoid = torch.nn.Sigmoid()\n",
        "        def forward(self, x):\n",
        "            hidden = self.fc1(x)\n",
        "            relu = self.relu(hidden)\n",
        "            output = self.fc2(relu)\n",
        "            output = self.sigmoid(output)\n",
        "            return output"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HZLT4dxiRU6S"
      },
      "source": [
        "# CREATE RANDOM DATA POINTS\n",
        "from sklearn.datasets import make_blobs\n",
        "def blob_label(y, label, loc): # assign labels\n",
        "    target = np.copy(y)\n",
        "    for l in loc:\n",
        "        target[y == l] = label\n",
        "    return target\n",
        "x_train, y_train = make_blobs(n_samples=40, n_features=2, cluster_std=1.5, shuffle=True)\n",
        "x_train = torch.FloatTensor(x_train)\n",
        "y_train = torch.FloatTensor(blob_label(y_train, 0, [0]))\n",
        "y_train = torch.FloatTensor(blob_label(y_train, 1, [1,2,3]))\n",
        "x_test, y_test = make_blobs(n_samples=10, n_features=2, cluster_std=1.5, shuffle=True)\n",
        "print(x_test)\n",
        "print(y_test)\n",
        "x_test = torch.FloatTensor(x_test)\n",
        "y_test = torch.FloatTensor(blob_label(y_test, 0, [0]))\n",
        "y_test = torch.FloatTensor(blob_label(y_test, 1, [1,2,3]))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0vX3iQKvWE3o",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "ff912ff3-1e8e-4624-f027-20d69e5fd663"
      },
      "source": [
        "print(y_test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([0., 0., 1., 1., 1., 1., 1., 0., 1., 0.])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ui1wALOdRq99"
      },
      "source": [
        "model = Feedforward(2, 4)\n",
        "criterion = torch.nn.BCELoss()\n",
        "optimizer = torch.optim.SGD(model.parameters(), lr = 0.001)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2cW86lHoSRhC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "c328be41-e740-4777-c625-505299d75465"
      },
      "source": [
        "model.eval()\n",
        "y_pred = model(x_test)\n",
        "before_train = criterion(y_pred.squeeze(), y_test)\n",
        "print('Test loss before training' , before_train.item())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test loss before training 0.8403955698013306\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n2YqQa--SYJh"
      },
      "source": [
        "model.train()\n",
        "epoch = 200\n",
        "for epoch in range(epoch):\n",
        "    optimizer.zero_grad()\n",
        "    # Forward pass\n",
        "    y_pred = model(x_train)\n",
        "    # Compute Loss\n",
        "    loss = criterion(y_pred.squeeze(), y_train)\n",
        "   \n",
        "    print('Epoch {}: train loss: {}'.format(epoch, loss.item()))\n",
        "    # Backward pass\n",
        "    loss.backward()\n",
        "    optimizer.step()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cLJ844YpSia-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "13459f41-14e1-4a07-db52-df64f2cc8c89"
      },
      "source": [
        "model.eval()\n",
        "y_pred = model(x_test)\n",
        "after_train = criterion(y_pred.squeeze(), y_test) \n",
        "print('Test loss after Training' , after_train.item())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test loss after Training 0.7146216034889221\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jD0rytKmUR6d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 381
        },
        "outputId": "afea9709-e8c1-4b52-da53-ffecce7ba7af"
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "y_pred_np = y_pred.detach().numpy().round()\n",
        "y_test_np = y_test.detach().numpy()\n",
        "print(y_pred_np)\n",
        "print(y_test_np)\n",
        "print(classification_report(y_test_np, y_pred_np))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[1.]\n",
            " [1.]\n",
            " [0.]\n",
            " [1.]\n",
            " [1.]\n",
            " [0.]\n",
            " [0.]\n",
            " [1.]\n",
            " [1.]\n",
            " [1.]]\n",
            "[0. 0. 1. 1. 1. 1. 1. 0. 1. 0.]\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.00      0.00      0.00         4\n",
            "         1.0       0.43      0.50      0.46         6\n",
            "\n",
            "    accuracy                           0.30        10\n",
            "   macro avg       0.21      0.25      0.23        10\n",
            "weighted avg       0.26      0.30      0.28        10\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MIxdOeL9bn_T"
      },
      "source": [
        "# Demo: Batch training, Learning curves, Early Stopping, Dropout\n",
        "\n",
        "O exemplo de código a seguir é uma adaptação do encontrado em https://github.com/Bjarten/early-stopping-pytorch/blob/master/MNIST_Early_Stopping_example.ipynb"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eg5nVORWbp_4"
      },
      "source": [
        "# import libraries\n",
        "import torch\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6oaHxa7ydyRI"
      },
      "source": [
        "class EarlyStopping:\n",
        "    \"\"\"Early stops the training if validation loss doesn't improve after a given patience.\"\"\"\n",
        "    def __init__(self, patience=7, verbose=False, delta=0):\n",
        "        \"\"\"\n",
        "        Args:\n",
        "            patience (int): How long to wait after last time validation loss improved.\n",
        "                            Default: 7\n",
        "            verbose (bool): If True, prints a message for each validation loss improvement. \n",
        "                            Default: False\n",
        "            delta (float): Minimum change in the monitored quantity to qualify as an improvement.\n",
        "                            Default: 0\n",
        "        \"\"\"\n",
        "        self.patience = patience\n",
        "        self.verbose = verbose\n",
        "        self.counter = 0\n",
        "        self.best_score = None\n",
        "        self.early_stop = False\n",
        "        self.val_loss_min = np.Inf\n",
        "        self.delta = delta\n",
        "\n",
        "    def __call__(self, val_loss, model):\n",
        "\n",
        "        score = -val_loss\n",
        "\n",
        "        if self.best_score is None:\n",
        "            self.best_score = score\n",
        "            self.save_checkpoint(val_loss, model)\n",
        "        elif score < self.best_score + self.delta:\n",
        "            self.counter += 1\n",
        "            print(f'EarlyStopping counter: {self.counter} out of {self.patience}')\n",
        "            if self.counter >= self.patience:\n",
        "                self.early_stop = True\n",
        "        else:\n",
        "            self.best_score = score\n",
        "            self.save_checkpoint(val_loss, model)\n",
        "            self.counter = 0\n",
        "\n",
        "    def save_checkpoint(self, val_loss, model):\n",
        "        '''Saves model when validation loss decrease.'''\n",
        "        if self.verbose:\n",
        "            print(f'Validation loss decreased ({self.val_loss_min:.6f} --> {val_loss:.6f}).  Saving model ...')\n",
        "        torch.save(model.state_dict(), 'checkpoint.pt')\n",
        "        self.val_loss_min = val_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-3wySJxhb0_-"
      },
      "source": [
        "from torchvision import datasets\n",
        "import torchvision.transforms as transforms\n",
        "from torch.utils.data.sampler import SubsetRandomSampler\n",
        "\n",
        "def create_datasets(batch_size):\n",
        "\n",
        "    # percentage of training set to use as validation\n",
        "    valid_size = 0.2\n",
        "\n",
        "    # convert data to torch.FloatTensor\n",
        "    transform = transforms.ToTensor()\n",
        "\n",
        "    # choose the training and test datasets\n",
        "    train_data = datasets.MNIST(root='data', \n",
        "                                train=True,\n",
        "                                download=True, \n",
        "                                transform=transform)\n",
        "\n",
        "    test_data = datasets.MNIST(root='data',\n",
        "                               train=False,\n",
        "                               download=True,\n",
        "                               transform=transform)\n",
        "\n",
        "    # obtain training indices that will be used for validation\n",
        "    num_train = len(train_data)\n",
        "    indices = list(range(num_train))\n",
        "    np.random.shuffle(indices)\n",
        "    split = int(np.floor(valid_size * num_train))\n",
        "    train_idx, valid_idx = indices[split:], indices[:split]\n",
        "    \n",
        "    # define samplers for obtaining training and validation batches\n",
        "    train_sampler = SubsetRandomSampler(train_idx)\n",
        "    valid_sampler = SubsetRandomSampler(valid_idx)\n",
        "    \n",
        "    # load training data in batches\n",
        "    train_loader = torch.utils.data.DataLoader(train_data,\n",
        "                                               batch_size=batch_size,\n",
        "                                               sampler=train_sampler,\n",
        "                                               num_workers=0)\n",
        "    \n",
        "    # load validation data in batches\n",
        "    valid_loader = torch.utils.data.DataLoader(train_data,\n",
        "                                               batch_size=batch_size,\n",
        "                                               sampler=valid_sampler,\n",
        "                                               num_workers=0)\n",
        "    \n",
        "    # load test data in batches\n",
        "    test_loader = torch.utils.data.DataLoader(test_data,\n",
        "                                              batch_size=batch_size,\n",
        "                                              num_workers=0)\n",
        "    \n",
        "    return train_loader, test_loader, valid_loader"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zaqZqBVHb7xY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        },
        "outputId": "86f0a8b9-1e13-4014-f82c-6efcb5d26b48"
      },
      "source": [
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "\n",
        "class Net(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.fc1 = nn.Linear(28 * 28, 128)\n",
        "        self.fc2 = nn.Linear(128, 128)\n",
        "        self.fc3 = nn.Linear(128, 10)\n",
        "        self.dropout = nn.Dropout(0.5)\n",
        "    def forward(self, x):\n",
        "        # flatten image input\n",
        "        x = x.view(-1, 28 * 28)\n",
        "        # add hidden layer, with relu activation function\n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = self.dropout(x)\n",
        "        # add hidden layer, with relu activation function\n",
        "        x = F.relu(self.fc2(x))\n",
        "        x = self.dropout(x)\n",
        "        # add output layer\n",
        "        x = self.fc3(x)\n",
        "        return x\n",
        "\n",
        "# initialize the NN\n",
        "model = Net()\n",
        "print(model)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Net(\n",
            "  (fc1): Linear(in_features=784, out_features=128, bias=True)\n",
            "  (fc2): Linear(in_features=128, out_features=128, bias=True)\n",
            "  (fc3): Linear(in_features=128, out_features=10, bias=True)\n",
            "  (dropout): Dropout(p=0.5, inplace=False)\n",
            ")\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jp9DMq_ocK2p"
      },
      "source": [
        "# specify loss function\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "\n",
        "# specify optimizer\n",
        "optimizer = torch.optim.Adam(model.parameters())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4_qhj0xUcR_F"
      },
      "source": [
        "def train_model(model, batch_size, patience, n_epochs):\n",
        "    \n",
        "    # to track the training loss as the model trains\n",
        "    train_losses = []\n",
        "    # to track the validation loss as the model trains\n",
        "    valid_losses = []\n",
        "    # to track the average training loss per epoch as the model trains\n",
        "    avg_train_losses = []\n",
        "    # to track the average validation loss per epoch as the model trains\n",
        "    avg_valid_losses = [] \n",
        "    \n",
        "    # initialize the early_stopping object\n",
        "    early_stopping = EarlyStopping(patience=patience, verbose=True)\n",
        "    \n",
        "    for epoch in range(1, n_epochs + 1):\n",
        "\n",
        "        ###################\n",
        "        # train the model #\n",
        "        ###################\n",
        "        model.train() # prep model for training\n",
        "        for batch, (data, target) in enumerate(train_loader, 1):\n",
        "            # clear the gradients of all optimized variables\n",
        "            optimizer.zero_grad()\n",
        "            # forward pass: compute predicted outputs by passing inputs to the model\n",
        "            output = model(data)\n",
        "            # calculate the loss\n",
        "            loss = criterion(output, target)\n",
        "            # backward pass: compute gradient of the loss with respect to model parameters\n",
        "            loss.backward()\n",
        "            # perform a single optimization step (parameter update)\n",
        "            optimizer.step()\n",
        "            # record training loss\n",
        "            train_losses.append(loss.item())\n",
        "\n",
        "        ######################    \n",
        "        # validate the model #\n",
        "        ######################\n",
        "        model.eval() # prep model for evaluation\n",
        "        for data, target in valid_loader:\n",
        "            # forward pass: compute predicted outputs by passing inputs to the model\n",
        "            output = model(data)\n",
        "            # calculate the loss\n",
        "            loss = criterion(output, target)\n",
        "            # record validation loss\n",
        "            valid_losses.append(loss.item())\n",
        "\n",
        "        # print training/validation statistics \n",
        "        # calculate average loss over an epoch\n",
        "        train_loss = np.average(train_losses)\n",
        "        valid_loss = np.average(valid_losses)\n",
        "        avg_train_losses.append(train_loss)\n",
        "        avg_valid_losses.append(valid_loss)\n",
        "        \n",
        "        epoch_len = len(str(n_epochs))\n",
        "        \n",
        "        print_msg = (f'[{epoch:>{epoch_len}}/{n_epochs:>{epoch_len}}] ' +\n",
        "                     f'train_loss: {train_loss:.5f} ' +\n",
        "                     f'valid_loss: {valid_loss:.5f}')\n",
        "        \n",
        "        print(print_msg)\n",
        "        \n",
        "        # clear lists to track next epoch\n",
        "        train_losses = []\n",
        "        valid_losses = []\n",
        "        \n",
        "        # early_stopping needs the validation loss to check if it has decresed, \n",
        "        # and if it has, it will make a checkpoint of the current model\n",
        "        early_stopping(valid_loss, model)\n",
        "        \n",
        "        if early_stopping.early_stop:\n",
        "            print(\"Early stopping\")\n",
        "            break\n",
        "        \n",
        "    # load the last checkpoint with the best model\n",
        "    model.load_state_dict(torch.load('checkpoint.pt'))\n",
        "\n",
        "    return  model, avg_train_losses, avg_valid_losses"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pfZG-4Klctgb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000,
          "referenced_widgets": [
            "936c5f6b16174f92bc5efe92401c31b2",
            "ffda42f0b76a4db58b59b953219716ed",
            "19eca837d3c543759e2c146a5fb5ed92",
            "9b34d7b8e31a483582351ee1bb294577",
            "1e011ddc05304e0fb442929f0070684f",
            "3da36fb56d8b42c7934f89721ce347f9",
            "e6153d76a3a644db9696d9299a8ea256",
            "e3884fed39d6477498ef425f2809e865",
            "f25f4a4612034e92b5f5f8990a92e5bc",
            "eecb7599d2494199832dd3ae4d31e8a9",
            "789d2a7bfae74fc9845f0bc303fa26e1",
            "403d0274e94345c5bdfabab1cedf42a6",
            "604643cbd9b14fad95b78c7e76b9bcee",
            "9a26f52f059f4386a6182c90018b4eb6",
            "ab6b42e326c54ae099e873361a3f571d",
            "087a96d4328740bb8968e0a9cd0e49d4",
            "5c65d631f07c4f80834e8a6e46341509",
            "8997e1cd30444cb3abc698533cfdffcc",
            "6d93bb3d00be434abd346a4912a35760",
            "d9937e2602bf4052914513168cb3ac07",
            "43a432073d6b4812bf90914fdc76a588",
            "913cf542a769493083d0c41a4abf9147",
            "af48058b3ac848eba95d001e52cc5bbd",
            "c2c42d75ea0c40b7be02c93778fa042b",
            "d490c1ac5ccf4b6696338331cbdbd24f",
            "9f0c4f030eb5487d8ff05f8947b8bde0",
            "311e748f5c744c60b08b9f9b123cb575",
            "577c299761804717b0840d2ebf3a6dd9",
            "9ee4d088f3324c69985880e56917e1f1",
            "7b9e0a0a549244879b8b910f6018e0b8",
            "9494c4b93a534e9c9a1a23852545e42e",
            "65f015210c38460b95c2b4f51cc92549"
          ]
        },
        "outputId": "48f0ada1-f8be-4af5-f571-f24f6993476b"
      },
      "source": [
        "n_epochs = 100\n",
        "batch_size = 128\n",
        "\n",
        "train_loader, test_loader, valid_loader = create_datasets(batch_size)\n",
        "\n",
        "# early stopping patience; how long to wait after last time validation loss improved.\n",
        "patience = 20\n",
        "\n",
        "model, train_loss, valid_loss = train_model(model, batch_size, patience, n_epochs)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "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": "936c5f6b16174f92bc5efe92401c31b2",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting data/MNIST/raw/train-images-idx3-ubyte.gz to data/MNIST/raw\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": "f25f4a4612034e92b5f5f8990a92e5bc",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting data/MNIST/raw/train-labels-idx1-ubyte.gz to data/MNIST/raw\n",
            "Downloading http://yann.lecun.com/exdb/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": "5c65d631f07c4f80834e8a6e46341509",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting data/MNIST/raw/t10k-images-idx3-ubyte.gz to data/MNIST/raw\n",
            "Downloading http://yann.lecun.com/exdb/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": "d490c1ac5ccf4b6696338331cbdbd24f",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting data/MNIST/raw/t10k-labels-idx1-ubyte.gz to data/MNIST/raw\n",
            "Processing...\n",
            "Done!\n",
            "[  1/100] train_loss: 0.67765 valid_loss: 0.26069\n",
            "Validation loss decreased (inf --> 0.260689).  Saving model ...\n",
            "[  2/100] train_loss: 0.32582 valid_loss: 0.19368\n",
            "Validation loss decreased (0.260689 --> 0.193679).  Saving model ...\n",
            "[  3/100] train_loss: 0.26599 valid_loss: 0.16314\n",
            "Validation loss decreased (0.193679 --> 0.163142).  Saving model ...\n",
            "[  4/100] train_loss: 0.23241 valid_loss: 0.14661\n",
            "Validation loss decreased (0.163142 --> 0.146615).  Saving model ...\n",
            "[  5/100] train_loss: 0.21010 valid_loss: 0.13294\n",
            "Validation loss decreased (0.146615 --> 0.132936).  Saving model ...\n",
            "[  6/100] train_loss: 0.19619 valid_loss: 0.12647\n",
            "Validation loss decreased (0.132936 --> 0.126467).  Saving model ...\n",
            "[  7/100] train_loss: 0.18322 valid_loss: 0.12248\n",
            "Validation loss decreased (0.126467 --> 0.122484).  Saving model ...\n",
            "[  8/100] train_loss: 0.17486 valid_loss: 0.11344\n",
            "Validation loss decreased (0.122484 --> 0.113437).  Saving model ...\n",
            "[  9/100] train_loss: 0.16227 valid_loss: 0.11376\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 10/100] train_loss: 0.16060 valid_loss: 0.10991\n",
            "Validation loss decreased (0.113437 --> 0.109910).  Saving model ...\n",
            "[ 11/100] train_loss: 0.15156 valid_loss: 0.10249\n",
            "Validation loss decreased (0.109910 --> 0.102488).  Saving model ...\n",
            "[ 12/100] train_loss: 0.14758 valid_loss: 0.10445\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 13/100] train_loss: 0.14077 valid_loss: 0.10115\n",
            "Validation loss decreased (0.102488 --> 0.101154).  Saving model ...\n",
            "[ 14/100] train_loss: 0.14149 valid_loss: 0.10163\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 15/100] train_loss: 0.13406 valid_loss: 0.09753\n",
            "Validation loss decreased (0.101154 --> 0.097529).  Saving model ...\n",
            "[ 16/100] train_loss: 0.12960 valid_loss: 0.10101\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 17/100] train_loss: 0.12909 valid_loss: 0.09999\n",
            "EarlyStopping counter: 2 out of 20\n",
            "[ 18/100] train_loss: 0.12547 valid_loss: 0.09752\n",
            "Validation loss decreased (0.097529 --> 0.097525).  Saving model ...\n",
            "[ 19/100] train_loss: 0.12434 valid_loss: 0.09541\n",
            "Validation loss decreased (0.097525 --> 0.095406).  Saving model ...\n",
            "[ 20/100] train_loss: 0.12260 valid_loss: 0.09703\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 21/100] train_loss: 0.12347 valid_loss: 0.10077\n",
            "EarlyStopping counter: 2 out of 20\n",
            "[ 22/100] train_loss: 0.11555 valid_loss: 0.09411\n",
            "Validation loss decreased (0.095406 --> 0.094111).  Saving model ...\n",
            "[ 23/100] train_loss: 0.11396 valid_loss: 0.09662\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 24/100] train_loss: 0.11567 valid_loss: 0.09389\n",
            "Validation loss decreased (0.094111 --> 0.093886).  Saving model ...\n",
            "[ 25/100] train_loss: 0.11081 valid_loss: 0.10088\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 26/100] train_loss: 0.10629 valid_loss: 0.08981\n",
            "Validation loss decreased (0.093886 --> 0.089813).  Saving model ...\n",
            "[ 27/100] train_loss: 0.10945 valid_loss: 0.09163\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 28/100] train_loss: 0.10679 valid_loss: 0.08793\n",
            "Validation loss decreased (0.089813 --> 0.087934).  Saving model ...\n",
            "[ 29/100] train_loss: 0.10497 valid_loss: 0.08686\n",
            "Validation loss decreased (0.087934 --> 0.086856).  Saving model ...\n",
            "[ 30/100] train_loss: 0.09959 valid_loss: 0.09086\n",
            "EarlyStopping counter: 1 out of 20\n",
            "[ 31/100] train_loss: 0.10108 valid_loss: 0.09420\n",
            "EarlyStopping counter: 2 out of 20\n",
            "[ 32/100] train_loss: 0.10028 valid_loss: 0.08981\n",
            "EarlyStopping counter: 3 out of 20\n",
            "[ 33/100] train_loss: 0.09779 valid_loss: 0.09101\n",
            "EarlyStopping counter: 4 out of 20\n",
            "[ 34/100] train_loss: 0.09818 valid_loss: 0.09235\n",
            "EarlyStopping counter: 5 out of 20\n",
            "[ 35/100] train_loss: 0.09773 valid_loss: 0.09572\n",
            "EarlyStopping counter: 6 out of 20\n",
            "[ 36/100] train_loss: 0.09386 valid_loss: 0.09594\n",
            "EarlyStopping counter: 7 out of 20\n",
            "[ 37/100] train_loss: 0.09805 valid_loss: 0.10171\n",
            "EarlyStopping counter: 8 out of 20\n",
            "[ 38/100] train_loss: 0.09227 valid_loss: 0.09615\n",
            "EarlyStopping counter: 9 out of 20\n",
            "[ 39/100] train_loss: 0.09783 valid_loss: 0.09589\n",
            "EarlyStopping counter: 10 out of 20\n",
            "[ 40/100] train_loss: 0.09329 valid_loss: 0.09282\n",
            "EarlyStopping counter: 11 out of 20\n",
            "[ 41/100] train_loss: 0.09157 valid_loss: 0.09168\n",
            "EarlyStopping counter: 12 out of 20\n",
            "[ 42/100] train_loss: 0.09071 valid_loss: 0.10140\n",
            "EarlyStopping counter: 13 out of 20\n",
            "[ 43/100] train_loss: 0.09166 valid_loss: 0.09683\n",
            "EarlyStopping counter: 14 out of 20\n",
            "[ 44/100] train_loss: 0.08673 valid_loss: 0.10054\n",
            "EarlyStopping counter: 15 out of 20\n",
            "[ 45/100] train_loss: 0.08645 valid_loss: 0.10083\n",
            "EarlyStopping counter: 16 out of 20\n",
            "[ 46/100] train_loss: 0.09165 valid_loss: 0.09750\n",
            "EarlyStopping counter: 17 out of 20\n",
            "[ 47/100] train_loss: 0.08525 valid_loss: 0.10349\n",
            "EarlyStopping counter: 18 out of 20\n",
            "[ 48/100] train_loss: 0.08863 valid_loss: 0.09497\n",
            "EarlyStopping counter: 19 out of 20\n",
            "[ 49/100] train_loss: 0.08988 valid_loss: 0.09812\n",
            "EarlyStopping counter: 20 out of 20\n",
            "Early stopping\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ii_jN4ZFczHo",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 585
        },
        "outputId": "dd4916d1-8a74-4f32-f555-77d5e054427d"
      },
      "source": [
        "# visualize the loss as the network trained\n",
        "fig = plt.figure(figsize=(10,8))\n",
        "plt.plot(range(1,len(train_loss)+1),train_loss, label='Training Loss')\n",
        "plt.plot(range(1,len(valid_loss)+1),valid_loss,label='Validation Loss')\n",
        "\n",
        "# find position of lowest validation loss\n",
        "minposs = valid_loss.index(min(valid_loss))+1 \n",
        "plt.axvline(minposs, linestyle='--', color='r',label='Early Stopping Checkpoint')\n",
        "\n",
        "plt.xlabel('epochs')\n",
        "plt.ylabel('loss')\n",
        "plt.ylim(0, 0.5) # consistent scale\n",
        "plt.xlim(0, len(train_loss)+1) # consistent scale\n",
        "plt.grid(True)\n",
        "plt.legend()\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "fig.savefig('loss_plot.png', bbox_inches='tight')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAI4CAYAAAB3OR9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/w/RbIEe2vt2ZgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 720x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C9ZQio2Kc41P",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 259
        },
        "outputId": "6afc179f-bab3-49f9-d72d-49b87a7a3ad0"
      },
      "source": [
        "# initialize lists to monitor test loss and accuracy\n",
        "test_loss = 0.0\n",
        "class_correct = list(0. for i in range(10))\n",
        "class_total = list(0. for i in range(10))\n",
        "\n",
        "model.eval() # prep model for evaluation\n",
        "\n",
        "for data, target in test_loader:\n",
        "    if len(target.data) != batch_size:\n",
        "        break\n",
        "    # forward pass: compute predicted outputs by passing inputs to the model\n",
        "    output = model(data)\n",
        "    # calculate the loss\n",
        "    loss = criterion(output, target)\n",
        "    # update test loss \n",
        "    test_loss += loss.item()*data.size(0)\n",
        "    # convert output probabilities to predicted class\n",
        "    _, pred = torch.max(output, 1)\n",
        "    # compare predictions to true label\n",
        "    correct = np.squeeze(pred.eq(target.data.view_as(pred)))\n",
        "    # calculate test accuracy for each object class\n",
        "    for i in range(batch_size):\n",
        "        label = target.data[i]\n",
        "        class_correct[label] += correct[i].item()\n",
        "        class_total[label] += 1\n",
        "\n",
        "# calculate and print avg test loss\n",
        "test_loss = test_loss/len(test_loader.dataset)\n",
        "print('Test Loss: {:.6f}\\n'.format(test_loss))\n",
        "\n",
        "for i in range(10):\n",
        "    if class_total[i] > 0:\n",
        "        print('Test Accuracy of %5s: %2d%% (%2d/%2d)' % (\n",
        "            str(i), 100 * class_correct[i] / class_total[i],\n",
        "            np.sum(class_correct[i]), np.sum(class_total[i])))\n",
        "    else:\n",
        "        print('Test Accuracy of %5s: N/A (no training examples)' % (classes[i]))\n",
        "\n",
        "print('\\nTest Accuracy (Overall): %2d%% (%2d/%2d)' % (\n",
        "    100. * np.sum(class_correct) / np.sum(class_total),\n",
        "    np.sum(class_correct), np.sum(class_total)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test Loss: 0.082543\n",
            "\n",
            "Test Accuracy of     0: 99% (970/979)\n",
            "Test Accuracy of     1: 99% (1122/1133)\n",
            "Test Accuracy of     2: 97% (1007/1030)\n",
            "Test Accuracy of     3: 98% (992/1008)\n",
            "Test Accuracy of     4: 97% (957/980)\n",
            "Test Accuracy of     5: 97% (868/890)\n",
            "Test Accuracy of     6: 97% (936/956)\n",
            "Test Accuracy of     7: 97% (999/1027)\n",
            "Test Accuracy of     8: 96% (940/973)\n",
            "Test Accuracy of     9: 96% (972/1008)\n",
            "\n",
            "Test Accuracy (Overall): 97% (9763/9984)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r0pozTUTc58u",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "outputId": "88a77f7e-7369-4d04-c86a-5a573ade865c"
      },
      "source": [
        "# obtain one batch of test images\n",
        "dataiter = iter(test_loader)\n",
        "images, labels = dataiter.next()\n",
        "\n",
        "# get sample outputs\n",
        "output = model(images)\n",
        "# convert output probabilities to predicted class\n",
        "_, preds = torch.max(output, 1)\n",
        "# prep images for display\n",
        "images = images.numpy()\n",
        "\n",
        "# plot the images in the batch, along with predicted and true labels\n",
        "fig = plt.figure(figsize=(25, 4))\n",
        "for idx in np.arange(20):\n",
        "    ax = fig.add_subplot(2, 20/2, idx+1, xticks=[], yticks=[])\n",
        "    ax.imshow(np.squeeze(images[idx]), cmap='gray')\n",
        "    ax.set_title(\"{} ({})\".format(str(preds[idx].item()), str(labels[idx].item())),\n",
        "                 color=(\"green\" if preds[idx]==labels[idx] else \"red\"))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABXEAAAD7CAYAAAAsAtcsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
            "text/plain": [
              "<Figure size 1800x288 with 20 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ChmuCgnXlt-e"
      },
      "source": [
        "# convnets"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V0rF_8QGnU5y"
      },
      "source": [
        "## filters"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zMvYd_WhlvtA"
      },
      "source": [
        "from skimage import exposure\n",
        "from skimage import data  # most functions are in subpackages\n",
        "camera = data.camera()\n",
        "camera_equalized = exposure.equalize_hist(camera)\n",
        "\n",
        "import os\n",
        "filename = os.path.join(skimage.data_dir, 'camera.png')\n",
        "camera = io.imread(filename)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QvoTf8nembPr",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "outputId": "ca6cc830-d5e7-4584-d549-198b27e4c79a"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "plt.imshow(camera, cmap='gray', interpolation='nearest') "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f18f7e01b70>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQYAAAD8CAYAAACVSwr3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9RMjpubhgCli",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 225
        },
        "outputId": "192d0b0e-8f95-4a5c-b700-8218033253ad"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from scipy import ndimage as ndi\n",
        "\n",
        "import skimage\n",
        "import skimage.feature\n",
        "from skimage.viewer import ImageViewer\n",
        "\n",
        "from skimage import feature\n",
        "\n",
        "# Generate noisy image of a square\n",
        "im = np.zeros((128, 128))\n",
        "im[32:-32, 32:-32] = 1\n",
        "\n",
        "im = ndi.rotate(im, 15, mode='constant')\n",
        "im = ndi.gaussian_filter(im, 4)\n",
        "im += 0.2 * np.random.random(im.shape)\n",
        "\n",
        "im = camera\n",
        "\n",
        "# Compute the Canny filter\n",
        "edges = feature.canny(im, sigma=3)\n",
        "\n",
        "# display results\n",
        "fig, (ax1, ax2) = plt.subplots(nrows=1, ncols=2, figsize=(8, 3),\n",
        "                                    sharex=True, sharey=True)\n",
        "\n",
        "ax1.imshow(im, cmap=plt.cm.gray)\n",
        "ax1.axis('off')\n",
        "ax1.set_title('noisy image', fontsize=20)\n",
        "\n",
        "ax2.imshow(edges, cmap=plt.cm.gray)\n",
        "ax2.axis('off')\n",
        "ax2.set_title(r'Canny filter, $\\sigma=3$', fontsize=20)\n",
        "\n",
        "fig.tight_layout()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcgAAADQCAYAAACDU/GoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x216 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9GnLYElf5fOk"
      },
      "source": [
        "## Demo: image classification on CIFAR10"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RKoih4rI56Q6"
      },
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DoMTUn915a7y",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 138,
          "referenced_widgets": [
            "3db05372ed66410e815168f2bfe0ec62",
            "e46604b9e7d746219bebc1232d2aec09",
            "c6dd7987abf54ad6b1236adb9bd55ac6",
            "70b439215cbc4a2baa68156e0a515fc5",
            "784998df19e7464dab23482b84d06ba6",
            "72b5534c96884c948f2532e531926702",
            "cc0b376f286245d7acd8895f53966f53",
            "164e013aa6844fe59349115576fd1fcb"
          ]
        },
        "outputId": "96af9fdd-4dc1-412c-98b9-25398e13e2eb"
      },
      "source": [
        "#The compose function allows for multiple transforms\n",
        "#transforms.ToTensor() converts our PILImage to a tensor of shape (C x H x W) in the range [0,1]\n",
        "#transforms.Normalize(mean,std) normalizes a tensor to a (mean, std) for (R, G, B)\n",
        "transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])\n",
        "\n",
        "train_set = torchvision.datasets.CIFAR10(root='./cifardata', train=True, download=True, transform=transform)\n",
        "\n",
        "test_set = torchvision.datasets.CIFAR10(root='./cifardata', train=False, download=True, transform=transform)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./cifardata/cifar-10-python.tar.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "3db05372ed66410e815168f2bfe0ec62",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Extracting ./cifardata/cifar-10-python.tar.gz to ./cifardata\n",
            "Files already downloaded and verified\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p2SZhOi06Oga"
      },
      "source": [
        "from torch.autograd import Variable\n",
        "import torch.nn.functional as F\n",
        "\n",
        "class SimpleCNN(torch.nn.Module):\n",
        "    \n",
        "    #Our batch shape for input x is (3, 32, 32)\n",
        "    \n",
        "    def __init__(self):\n",
        "        super(SimpleCNN, self).__init__()\n",
        "        \n",
        "        #Input channels = 3, output channels = 18\n",
        "        self.conv1 = torch.nn.Conv2d(3, 18, kernel_size=3, stride=1, padding=1)\n",
        "        self.pool = torch.nn.MaxPool2d(kernel_size=2, stride=2, padding=0)\n",
        "        \n",
        "        #4608 input features, 64 output features (see sizing flow below)\n",
        "        self.fc1 = torch.nn.Linear(18 * 16 * 16, 64)\n",
        "        \n",
        "        #64 input features, 10 output features for our 10 defined classes\n",
        "        self.fc2 = torch.nn.Linear(64, 10)\n",
        "        \n",
        "    def forward(self, x):\n",
        "        \n",
        "        #Computes the activation of the first convolution\n",
        "        #Size changes from (3, 32, 32) to (18, 32, 32)\n",
        "        x = F.relu(self.conv1(x))\n",
        "        \n",
        "        #Size changes from (18, 32, 32) to (18, 16, 16)\n",
        "        x = self.pool(x)\n",
        "        \n",
        "        #Reshape data to input to the input layer of the neural net\n",
        "        #Size changes from (18, 16, 16) to (1, 4608)\n",
        "        #Recall that the -1 infers this dimension from the other given dimension\n",
        "        x = x.view(-1, 18 * 16 *16)\n",
        "        \n",
        "        #Computes the activation of the first fully connected layer\n",
        "        #Size changes from (1, 4608) to (1, 64)\n",
        "        x = F.relu(self.fc1(x))\n",
        "        \n",
        "        #Computes the second fully connected layer (activation applied later)\n",
        "        #Size changes from (1, 64) to (1, 10)\n",
        "        x = self.fc2(x)\n",
        "        return(x)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QqG8MMb86VPZ"
      },
      "source": [
        "import time\n",
        "\n",
        "def trainNet(net, batch_size, n_epochs, learning_rate):\n",
        "    \n",
        "    #Print all of the hyperparameters of the training iteration:\n",
        "    print(\"===== HYPERPARAMETERS =====\")\n",
        "    print(\"batch_size=\", batch_size)\n",
        "    print(\"epochs=\", n_epochs)\n",
        "    print(\"learning_rate=\", learning_rate)\n",
        "    print(\"=\" * 30)\n",
        "    \n",
        "    #Get training data\n",
        "    train_loader = get_train_loader(batch_size)\n",
        "    n_batches = len(train_loader)\n",
        "    \n",
        "    #Create our loss and optimizer functions\n",
        "    loss, optimizer = createLossAndOptimizer(net, learning_rate)\n",
        "    \n",
        "    #Time for printing\n",
        "    training_start_time = time.time()\n",
        "    \n",
        "    #Loop for n_epochs\n",
        "    for epoch in range(n_epochs):\n",
        "        \n",
        "        running_loss = 0.0\n",
        "        print_every = n_batches // 10\n",
        "        start_time = time.time()\n",
        "        total_train_loss = 0\n",
        "        \n",
        "        for i, data in enumerate(train_loader, 0):\n",
        "            \n",
        "            #Get inputs\n",
        "            inputs, labels = data\n",
        "            \n",
        "            #Wrap them in a Variable object\n",
        "            inputs, labels = Variable(inputs), Variable(labels)\n",
        "            \n",
        "            #Set the parameter gradients to zero\n",
        "            optimizer.zero_grad()\n",
        "            \n",
        "            #Forward pass, backward pass, optimize\n",
        "            outputs = net(inputs)\n",
        "            loss_size = loss(outputs, labels)\n",
        "            loss_size.backward()\n",
        "            optimizer.step()\n",
        "            \n",
        "            #Print statistics\n",
        "            running_loss += loss_size.data\n",
        "            total_train_loss += loss_size.data\n",
        "            \n",
        "            #Print every 10th batch of an epoch\n",
        "            if (i + 1) % (print_every + 1) == 0:\n",
        "                print(\"Epoch {}, {:d}% \\t train_loss: {:.2f} took: {:.2f}s\".format(\n",
        "                        epoch+1, int(100 * (i+1) / n_batches), running_loss / print_every, time.time() - start_time))\n",
        "                #Reset running loss and time\n",
        "                running_loss = 0.0\n",
        "                start_time = time.time()\n",
        "            \n",
        "        #At the end of the epoch, do a pass on the validation set\n",
        "        total_val_loss = 0\n",
        "        for inputs, labels in val_loader:\n",
        "            \n",
        "            #Wrap tensors in Variables\n",
        "            inputs, labels = Variable(inputs), Variable(labels)\n",
        "            \n",
        "            #Forward pass\n",
        "            val_outputs = net(inputs)\n",
        "            val_loss_size = loss(val_outputs, labels)\n",
        "            total_val_loss += val_loss_size.data\n",
        "            \n",
        "        print(\"Validation loss = {:.2f}\".format(total_val_loss / len(val_loader)))\n",
        "        \n",
        "    print(\"Training finished, took {:.2f}s\".format(time.time() - training_start_time))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "moG5XTiH6oPo"
      },
      "source": [
        "#DataLoader takes in a dataset and a sampler for loading (num_workers deals with system level memory) \n",
        "def get_train_loader(batch_size):\n",
        "    train_loader = torch.utils.data.DataLoader(train_set, batch_size=batch_size,\n",
        "                                           sampler=train_sampler, num_workers=2)\n",
        "    return(train_loader)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C4Q2hngE6wUe"
      },
      "source": [
        "from torch.utils.data.sampler import SubsetRandomSampler\n",
        "\n",
        "#Training\n",
        "n_training_samples = 20000\n",
        "train_sampler = SubsetRandomSampler(np.arange(n_training_samples, dtype=np.int64))\n",
        "\n",
        "#Validation\n",
        "n_val_samples = 5000\n",
        "val_sampler = SubsetRandomSampler(np.arange(n_training_samples, n_training_samples + n_val_samples, dtype=np.int64))\n",
        "\n",
        "#Test\n",
        "n_test_samples = 5000\n",
        "test_sampler = SubsetRandomSampler(np.arange(n_test_samples, dtype=np.int64))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nGg8PAQR63hl"
      },
      "source": [
        "import torch.optim as optim\n",
        "\n",
        "def createLossAndOptimizer(net, learning_rate=0.001):\n",
        "    \n",
        "    #Loss function\n",
        "    loss = torch.nn.CrossEntropyLoss()\n",
        "    \n",
        "    #Optimizer\n",
        "    optimizer = optim.Adam(net.parameters(), lr=learning_rate)\n",
        "    \n",
        "    return(loss, optimizer)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "znuUUb-67oTh"
      },
      "source": [
        "#Test and validation loaders have constant batch sizes, so we can define them directly\n",
        "test_loader = torch.utils.data.DataLoader(test_set, batch_size=4, sampler=test_sampler, num_workers=2)\n",
        "val_loader = torch.utils.data.DataLoader(train_set, batch_size=128, sampler=val_sampler, num_workers=2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5b_sQ6Of6U7S",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 985
        },
        "outputId": "a4d628cd-160a-4be2-ee0b-61391ca4dfec"
      },
      "source": [
        "CNN = SimpleCNN()\n",
        "trainNet(CNN, batch_size=32, n_epochs=5, learning_rate=0.001)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "===== HYPERPARAMETERS =====\n",
            "batch_size= 32\n",
            "epochs= 5\n",
            "learning_rate= 0.001\n",
            "==============================\n",
            "Epoch 1, 10% \t train_loss: 2.08 took: 1.35s\n",
            "Epoch 1, 20% \t train_loss: 1.82 took: 1.25s\n",
            "Epoch 1, 30% \t train_loss: 1.68 took: 1.31s\n",
            "Epoch 1, 40% \t train_loss: 1.64 took: 1.26s\n",
            "Epoch 1, 50% \t train_loss: 1.56 took: 1.31s\n",
            "Epoch 1, 60% \t train_loss: 1.52 took: 1.28s\n",
            "Epoch 1, 70% \t train_loss: 1.47 took: 1.27s\n",
            "Epoch 1, 80% \t train_loss: 1.44 took: 1.28s\n",
            "Epoch 1, 90% \t train_loss: 1.39 took: 1.29s\n",
            "Validation loss = 1.33\n",
            "Epoch 2, 10% \t train_loss: 1.32 took: 1.42s\n",
            "Epoch 2, 20% \t train_loss: 1.29 took: 1.44s\n",
            "Epoch 2, 30% \t train_loss: 1.30 took: 1.44s\n",
            "Epoch 2, 40% \t train_loss: 1.26 took: 1.39s\n",
            "Epoch 2, 50% \t train_loss: 1.24 took: 1.33s\n",
            "Epoch 2, 60% \t train_loss: 1.26 took: 1.39s\n",
            "Epoch 2, 70% \t train_loss: 1.26 took: 1.43s\n",
            "Epoch 2, 80% \t train_loss: 1.28 took: 1.44s\n",
            "Epoch 2, 90% \t train_loss: 1.23 took: 1.41s\n",
            "Validation loss = 1.22\n",
            "Epoch 3, 10% \t train_loss: 1.08 took: 1.45s\n",
            "Epoch 3, 20% \t train_loss: 1.09 took: 1.38s\n",
            "Epoch 3, 30% \t train_loss: 1.15 took: 1.39s\n",
            "Epoch 3, 40% \t train_loss: 1.12 took: 1.35s\n",
            "Epoch 3, 50% \t train_loss: 1.10 took: 1.35s\n",
            "Epoch 3, 60% \t train_loss: 1.14 took: 1.37s\n",
            "Epoch 3, 70% \t train_loss: 1.15 took: 1.44s\n",
            "Epoch 3, 80% \t train_loss: 1.15 took: 1.40s\n",
            "Epoch 3, 90% \t train_loss: 1.12 took: 1.45s\n",
            "Validation loss = 1.18\n",
            "Epoch 4, 10% \t train_loss: 1.01 took: 1.46s\n",
            "Epoch 4, 20% \t train_loss: 1.02 took: 1.36s\n",
            "Epoch 4, 30% \t train_loss: 1.01 took: 1.34s\n",
            "Epoch 4, 40% \t train_loss: 1.00 took: 1.49s\n",
            "Epoch 4, 50% \t train_loss: 1.03 took: 1.41s\n",
            "Epoch 4, 60% \t train_loss: 1.02 took: 1.40s\n",
            "Epoch 4, 70% \t train_loss: 1.07 took: 1.36s\n",
            "Epoch 4, 80% \t train_loss: 1.01 took: 1.40s\n",
            "Epoch 4, 90% \t train_loss: 1.02 took: 1.32s\n",
            "Validation loss = 1.15\n",
            "Epoch 5, 10% \t train_loss: 0.88 took: 1.48s\n",
            "Epoch 5, 20% \t train_loss: 0.95 took: 1.35s\n",
            "Epoch 5, 30% \t train_loss: 0.90 took: 1.38s\n",
            "Epoch 5, 40% \t train_loss: 0.94 took: 1.34s\n",
            "Epoch 5, 50% \t train_loss: 0.92 took: 1.39s\n",
            "Epoch 5, 60% \t train_loss: 0.94 took: 1.40s\n",
            "Epoch 5, 70% \t train_loss: 0.91 took: 1.39s\n",
            "Epoch 5, 80% \t train_loss: 0.94 took: 1.40s\n",
            "Epoch 5, 90% \t train_loss: 1.02 took: 1.43s\n",
            "Validation loss = 1.16\n",
            "Training finished, took 77.70s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "feZW1olE18HO"
      },
      "source": [
        "# Demo: LSTMs for POS tagging\n",
        "\n",
        "O código desta seção é adaptado de https://pytorch.org/tutorials/beginner/nlp/sequence_models_tutorial.html"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jkUA-15F2CWK",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "61d09dcb-3cf9-4bb4-c7c2-6ec19e827ddd"
      },
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "\n",
        "torch.manual_seed(1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<torch._C.Generator at 0x7f580ac3e450>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ji_biIgp2K9O",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "8c6596b1-dfe5-4d2c-e455-28be4561b107"
      },
      "source": [
        "def prepare_sequence(seq, to_ix):\n",
        "    idxs = [to_ix[w] for w in seq]\n",
        "    return torch.tensor(idxs, dtype=torch.long)\n",
        "\n",
        "\n",
        "training_data = [\n",
        "    (\"The dog ate the apple\".split(), [\"DET\", \"NN\", \"V\", \"DET\", \"NN\"]),\n",
        "    (\"Everybody read that book\".split(), [\"NN\", \"V\", \"DET\", \"NN\"])\n",
        "]\n",
        "word_to_ix = {}\n",
        "for sent, tags in training_data:\n",
        "    for word in sent:\n",
        "        if word not in word_to_ix:\n",
        "            word_to_ix[word] = len(word_to_ix)\n",
        "print(word_to_ix)\n",
        "tag_to_ix = {\"DET\": 0, \"NN\": 1, \"V\": 2}\n",
        "\n",
        "# These will usually be more like 32 or 64 dimensional.\n",
        "# We will keep them small, so we can see how the weights change as we train.\n",
        "EMBEDDING_DIM = 6\n",
        "HIDDEN_DIM = 6"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'The': 0, 'dog': 1, 'ate': 2, 'the': 3, 'apple': 4, 'Everybody': 5, 'read': 6, 'that': 7, 'book': 8}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V3Lj0qPk2d2X"
      },
      "source": [
        "class LSTMTagger(nn.Module):\n",
        "\n",
        "    def __init__(self, embedding_dim, hidden_dim, vocab_size, tagset_size):\n",
        "        super(LSTMTagger, self).__init__()\n",
        "        self.hidden_dim = hidden_dim\n",
        "\n",
        "        self.word_embeddings = nn.Embedding(vocab_size, embedding_dim)\n",
        "\n",
        "        # The LSTM takes word embeddings as inputs, and outputs hidden states\n",
        "        # with dimensionality hidden_dim.\n",
        "        self.lstm = nn.LSTM(embedding_dim, hidden_dim)\n",
        "\n",
        "        # The linear layer that maps from hidden state space to tag space\n",
        "        self.hidden2tag = nn.Linear(hidden_dim, tagset_size)\n",
        "\n",
        "    def forward(self, sentence):\n",
        "        embeds = self.word_embeddings(sentence)\n",
        "        lstm_out, _ = self.lstm(embeds.view(len(sentence), 1, -1))\n",
        "        tag_space = self.hidden2tag(lstm_out.view(len(sentence), -1))\n",
        "        tag_scores = F.log_softmax(tag_space, dim=1)\n",
        "        return tag_scores"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S2OpUMnN2kS5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 190
        },
        "outputId": "b296c655-f0ff-4106-da67-3fa95259cccf"
      },
      "source": [
        "model = LSTMTagger(EMBEDDING_DIM, HIDDEN_DIM, len(word_to_ix), len(tag_to_ix))\n",
        "loss_function = nn.NLLLoss()\n",
        "optimizer = optim.SGD(model.parameters(), lr=0.1)\n",
        "\n",
        "# See what the scores are before training\n",
        "# Note that element i,j of the output is the score for tag j for word i.\n",
        "# Here we don't need to train, so the code is wrapped in torch.no_grad()\n",
        "with torch.no_grad():\n",
        "    inputs = prepare_sequence(training_data[0][0], word_to_ix)\n",
        "    tag_scores = model(inputs)\n",
        "    print(tag_scores)\n",
        "\n",
        "for epoch in range(300):  # again, normally you would NOT do 300 epochs, it is toy data\n",
        "    for sentence, tags in training_data:\n",
        "        # Step 1. Remember that Pytorch accumulates gradients.\n",
        "        # We need to clear them out before each instance\n",
        "        model.zero_grad()\n",
        "\n",
        "        # Step 2. Get our inputs ready for the network, that is, turn them into\n",
        "        # Tensors of word indices.\n",
        "        sentence_in = prepare_sequence(sentence, word_to_ix)\n",
        "        targets = prepare_sequence(tags, tag_to_ix)\n",
        "\n",
        "        # Step 3. Run our forward pass.\n",
        "        tag_scores = model(sentence_in)\n",
        "\n",
        "        # Step 4. Compute the loss, gradients, and update the parameters by\n",
        "        #  calling optimizer.step()\n",
        "        loss = loss_function(tag_scores, targets)\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "# See what the scores are after training\n",
        "with torch.no_grad():\n",
        "    inputs = prepare_sequence(training_data[0][0], word_to_ix)\n",
        "    tag_scores = model(inputs)\n",
        "\n",
        "    # The sentence is \"the dog ate the apple\".  i,j corresponds to score for tag j\n",
        "    # for word i. The predicted tag is the maximum scoring tag.\n",
        "    # Here, we can see the predicted sequence below is 0 1 2 0 1\n",
        "    # since 0 is index of the maximum value of row 1,\n",
        "    # 1 is the index of maximum value of row 2, etc.\n",
        "    # Which is DET NOUN VERB DET NOUN, the correct sequence!\n",
        "    print(tag_scores)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[-1.0500, -0.9256, -1.3714],\n",
            "        [-1.0188, -0.9739, -1.3419],\n",
            "        [-1.1330, -0.9662, -1.2126],\n",
            "        [-1.1818, -0.9763, -1.1501],\n",
            "        [-1.0766, -0.9916, -1.2439]])\n",
            "tensor([[-0.3892, -1.2426, -3.3890],\n",
            "        [-2.1082, -0.1328, -5.8464],\n",
            "        [-3.0852, -5.9469, -0.0495],\n",
            "        [-0.0499, -3.4414, -4.0961],\n",
            "        [-2.4540, -0.0929, -5.8799]])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tCJtE2bZoLkp"
      },
      "source": [
        "    (\"The dog ate the apple\", [\"DET\", \"NN\", \"V\", \"DET\", \"NN\"]),\n",
        "\n",
        "    {\"DET\": 0, \"NN\": 1, \"V\": 2}\n"
      ]
    }
  ]
}