HopeSpeechEDI_XLMR_Tamil.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "HopeSpeechEDI_XLMR_Tamil.ipynb",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "69259d26d2a946abb70e8fe6aaec116f": {
          "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_3e8c2d279a724af8afefe735f9b234df",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_0cbffe1248184c6a969719e642230b37",
              "IPY_MODEL_6dcde407594a4c30857cf94ed09f93f7"
            ]
          }
        },
        "3e8c2d279a724af8afefe735f9b234df": {
          "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
          }
        },
        "0cbffe1248184c6a969719e642230b37": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_2af2f368b3ba4f37b90c7a5ba020ec0f",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 512,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 512,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_a3a7cccc6d5d485699a581af509c7dd9"
          }
        },
        "6dcde407594a4c30857cf94ed09f93f7": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_9d115b32f16c424a99f3d399f2195712",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 512/512 [01:01&lt;00:00, 8.34B/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_764371eb5bc347c6942e8706530b7bd8"
          }
        },
        "2af2f368b3ba4f37b90c7a5ba020ec0f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "a3a7cccc6d5d485699a581af509c7dd9": {
          "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
          }
        },
        "9d115b32f16c424a99f3d399f2195712": {
          "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"
          }
        },
        "764371eb5bc347c6942e8706530b7bd8": {
          "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
          }
        },
        "45927ca0331f41c881b769b6663ff07d": {
          "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_682ffe42c93249a0b5f9070e8924580e",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_6ab47a24a6814a38862108d247ef318e",
              "IPY_MODEL_02e714b9aa8c47f7ae083f90d5eded27"
            ]
          }
        },
        "682ffe42c93249a0b5f9070e8924580e": {
          "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
          }
        },
        "6ab47a24a6814a38862108d247ef318e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_b3df70ab91be4033837494e5aa155515",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 5069051,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 5069051,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_93a430e594ea45bd96161d1ee3fa06d4"
          }
        },
        "02e714b9aa8c47f7ae083f90d5eded27": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_4f96192d52834249b9da927cf8c8ecad",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 5.07M/5.07M [00:01&lt;00:00, 4.67MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_3ed032c36ae545fbb210f7715ea66943"
          }
        },
        "b3df70ab91be4033837494e5aa155515": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "93a430e594ea45bd96161d1ee3fa06d4": {
          "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
          }
        },
        "4f96192d52834249b9da927cf8c8ecad": {
          "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"
          }
        },
        "3ed032c36ae545fbb210f7715ea66943": {
          "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
          }
        },
        "c3f8cd0a07694469b8824b3b8e3d3f49": {
          "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_85b3438077bb4b15a90d907c1db6ee18",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_066fd44de32d4fb68a43ead9def7ac9f",
              "IPY_MODEL_678caa0cd6bc40a8893026f5a162e337"
            ]
          }
        },
        "85b3438077bb4b15a90d907c1db6ee18": {
          "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
          }
        },
        "066fd44de32d4fb68a43ead9def7ac9f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_9376f06f6be34932bad7982246ec5057",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 9096718,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 9096718,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_1cbc7e2d150e40eeaa5dd8b1ddb3428c"
          }
        },
        "678caa0cd6bc40a8893026f5a162e337": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_0180b937b0f1442382624ff5de6982af",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 9.10M/9.10M [00:59&lt;00:00, 152kB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_aae4a53b02064634b0a282c4cc36ff41"
          }
        },
        "9376f06f6be34932bad7982246ec5057": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "1cbc7e2d150e40eeaa5dd8b1ddb3428c": {
          "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
          }
        },
        "0180b937b0f1442382624ff5de6982af": {
          "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"
          }
        },
        "aae4a53b02064634b0a282c4cc36ff41": {
          "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
          }
        },
        "f1d3cd21ca724e99a0667c2f03e09685": {
          "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_5731efdc9d494f56b6eb08ba163c948b",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_abc0f7c718904b91b132879e49a9cce1",
              "IPY_MODEL_2986b0ec98e14f158817d176fc17a686"
            ]
          }
        },
        "5731efdc9d494f56b6eb08ba163c948b": {
          "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
          }
        },
        "abc0f7c718904b91b132879e49a9cce1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_9a73bb59951843f5bf45ab7957e7b034",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 1115590446,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1115590446,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_693062c4586e4f83b40a0302151b1f51"
          }
        },
        "2986b0ec98e14f158817d176fc17a686": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_6e695cb22742456b93b84761b220cef8",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 1.12G/1.12G [00:16&lt;00:00, 67.1MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_2b833a43029e485bbbf5db292b8a4e2d"
          }
        },
        "9a73bb59951843f5bf45ab7957e7b034": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "693062c4586e4f83b40a0302151b1f51": {
          "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
          }
        },
        "6e695cb22742456b93b84761b220cef8": {
          "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"
          }
        },
        "2b833a43029e485bbbf5db292b8a4e2d": {
          "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
          }
        }
      }
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "elqe9PQrVH6P"
      },
      "source": [
        "### Performing checks for the resources available"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fSELLChxa5su",
        "outputId": "763a4143-1456-41f1-bf41-7ea513d0c4f5"
      },
      "source": [
        "import torch\r\n",
        "\r\n",
        "# If there's a GPU available...\r\n",
        "if torch.cuda.is_available():    \r\n",
        "\r\n",
        "    # Tell PyTorch to use the GPU.    \r\n",
        "    device = torch.device(\"cuda\")\r\n",
        "\r\n",
        "    print('There are %d GPU(s) available.' % torch.cuda.device_count())\r\n",
        "\r\n",
        "    print('We will use the GPU:', torch.cuda.get_device_name(0))\r\n",
        "\r\n",
        "# If not...\r\n",
        "else:\r\n",
        "    print('No GPU available, using the CPU instead.')\r\n",
        "    device = torch.device(\"cpu\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "There are 1 GPU(s) available.\n",
            "We will use the GPU: Tesla P100-PCIE-16GB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8SokVowWo3cc",
        "outputId": "0bc00809-0688-4273-919f-021899aa375a"
      },
      "source": [
        "!nvidia-smi"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Fri Feb  5 05:59:35 2021       \n",
            "+-----------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 460.39       Driver Version: 418.67       CUDA Version: 10.1     |\n",
            "|-------------------------------+----------------------+----------------------+\n",
            "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
            "|                               |                      |               MIG M. |\n",
            "|===============================+======================+======================|\n",
            "|   0  Tesla P100-PCIE...  Off  | 00000000:00:04.0 Off |                    0 |\n",
            "| N/A   34C    P0    27W / 250W |     10MiB / 16280MiB |      0%      Default |\n",
            "|                               |                      |                 ERR! |\n",
            "+-------------------------------+----------------------+----------------------+\n",
            "                                                                               \n",
            "+-----------------------------------------------------------------------------+\n",
            "| Processes:                                                                  |\n",
            "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
            "|        ID   ID                                                   Usage      |\n",
            "|=============================================================================|\n",
            "|  No running processes found                                                 |\n",
            "+-----------------------------------------------------------------------------+\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "K-gbOgflVRxg"
      },
      "source": [
        "### Installing huggingface transformers library"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DbDuxKy0a9Zl",
        "outputId": "89b08927-153c-45fd-dd22-9aa70e5d07d0"
      },
      "source": [
        "!pip install transformers"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting transformers\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/88/b1/41130a228dd656a1a31ba281598a968320283f48d42782845f6ba567f00b/transformers-4.2.2-py3-none-any.whl (1.8MB)\n",
            "\u001b[K     |████████████████████████████████| 1.8MB 7.7MB/s \n",
            "\u001b[?25hRequirement already satisfied: filelock in /usr/local/lib/python3.6/dist-packages (from transformers) (3.0.12)\n",
            "Requirement already satisfied: dataclasses; python_version < \"3.7\" in /usr/local/lib/python3.6/dist-packages (from transformers) (0.8)\n",
            "Requirement already satisfied: importlib-metadata; python_version < \"3.8\" in /usr/local/lib/python3.6/dist-packages (from transformers) (3.4.0)\n",
            "Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.6/dist-packages (from transformers) (4.41.1)\n",
            "Collecting sacremoses\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/7d/34/09d19aff26edcc8eb2a01bed8e98f13a1537005d31e95233fd48216eed10/sacremoses-0.0.43.tar.gz (883kB)\n",
            "\u001b[K     |████████████████████████████████| 890kB 39.7MB/s \n",
            "\u001b[?25hCollecting tokenizers==0.9.4\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/0f/1c/e789a8b12e28be5bc1ce2156cf87cb522b379be9cadc7ad8091a4cc107c4/tokenizers-0.9.4-cp36-cp36m-manylinux2010_x86_64.whl (2.9MB)\n",
            "\u001b[K     |████████████████████████████████| 2.9MB 23.5MB/s \n",
            "\u001b[?25hRequirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from transformers) (2.23.0)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from transformers) (1.19.5)\n",
            "Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.6/dist-packages (from transformers) (2019.12.20)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.6/dist-packages (from transformers) (20.9)\n",
            "Requirement already satisfied: typing-extensions>=3.6.4; python_version < \"3.8\" in /usr/local/lib/python3.6/dist-packages (from importlib-metadata; python_version < \"3.8\"->transformers) (3.7.4.3)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.6/dist-packages (from importlib-metadata; python_version < \"3.8\"->transformers) (3.4.0)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from sacremoses->transformers) (1.15.0)\n",
            "Requirement already satisfied: click in /usr/local/lib/python3.6/dist-packages (from sacremoses->transformers) (7.1.2)\n",
            "Requirement already satisfied: joblib in /usr/local/lib/python3.6/dist-packages (from sacremoses->transformers) (1.0.0)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (1.24.3)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (2020.12.5)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (2.10)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (3.0.4)\n",
            "Requirement already satisfied: pyparsing>=2.0.2 in /usr/local/lib/python3.6/dist-packages (from packaging->transformers) (2.4.7)\n",
            "Building wheels for collected packages: sacremoses\n",
            "  Building wheel for sacremoses (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for sacremoses: filename=sacremoses-0.0.43-cp36-none-any.whl size=893261 sha256=b1c11c0986aa1f09a42d8fd7698c215b258d11aa115ddfba242655fbbe82b9ce\n",
            "  Stored in directory: /root/.cache/pip/wheels/29/3c/fd/7ce5c3f0666dab31a50123635e6fb5e19ceb42ce38d4e58f45\n",
            "Successfully built sacremoses\n",
            "Installing collected packages: sacremoses, tokenizers, transformers\n",
            "Successfully installed sacremoses-0.0.43 tokenizers-0.9.4 transformers-4.2.2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bPwaYLiRSNPC"
      },
      "source": [
        "## Importing necessary packages"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0XdFrhr8SRG0"
      },
      "source": [
        "import pandas as pd\r\n",
        "import numpy as np \r\n",
        "import random\r\n",
        "from sklearn.preprocessing import LabelEncoder\r\n",
        "from sklearn.utils.class_weight import compute_class_weight\r\n",
        "import time\r\n",
        "import datetime\r\n",
        "from sklearn.metrics import classification_report,confusion_matrix\r\n",
        "import random\r\n",
        "import time\r\n",
        "import torch.nn as nn\r\n",
        "from transformers import AutoModel , AutoModelForSequenceClassification, AutoConfig , AutoTokenizer , AdamW ,get_linear_schedule_with_warmup \r\n",
        "import torch\r\n",
        "import matplotlib.pyplot as plt\r\n",
        "import seaborn as sns\r\n",
        "from matplotlib import rc\r\n",
        "from matplotlib.ticker import MaxNLocator"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dROX3juhVCxC"
      },
      "source": [
        "## Setting some configuration values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kzafv4BzVCC9"
      },
      "source": [
        "# Use plot styling from seaborn.\r\n",
        "sns.set(style='darkgrid')\r\n",
        "\r\n",
        "# Increase the plot size and font size.\r\n",
        "sns.set(font_scale=1.5)\r\n",
        "plt.rcParams[\"figure.figsize\"] = (16,12)\r\n",
        "\r\n",
        "# Set the seed value all over the place to make this reproducible.\r\n",
        "seed_val = 42\r\n",
        "\r\n",
        "random.seed(seed_val)\r\n",
        "np.random.seed(seed_val)\r\n",
        "torch.manual_seed(seed_val)\r\n",
        "torch.cuda.manual_seed_all(seed_val)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2v1AIN_OVX4t"
      },
      "source": [
        "## Importing drive into the colaboratory"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Lofimyh2a_tp",
        "outputId": "e3ac36c9-18ae-47a8-a95a-2fdc70c096ff"
      },
      "source": [
        "from google.colab import drive \r\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yOK-M00dVeW2"
      },
      "source": [
        "## Importing dataset directories"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-y-JRPKBbCTv"
      },
      "source": [
        "from os import getcwd , listdir \r\n",
        "from os.path import join\r\n",
        "curr_dir = getcwd()\r\n",
        "drive_dir = join(curr_dir , 'drive','MyDrive')\r\n",
        "dataset_dir = join(drive_dir,'Codalab','HopeSpeech Detection For Equality,DiversityAnd Inclusion','Dataset')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7YU1h8WhmqvK",
        "outputId": "1a2913b1-eea1-43f8-9d4d-d75912a548ad"
      },
      "source": [
        "listdir(dataset_dir)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['english_hope_dev.csv',\n",
              " 'english_hope_train.csv',\n",
              " 'malayalam_hope_dev.csv',\n",
              " 'malayalam_hope_train.csv',\n",
              " 'tamil_hope_first_dev.csv',\n",
              " 'tamil_hope_first_train.csv',\n",
              " 'tamil_hope_test.csv',\n",
              " 'malayalam_hope_test.csv',\n",
              " 'english_hope_test.csv',\n",
              " 'english_hope_test _labels.csv',\n",
              " 'malayalam_hope_test _labels.csv',\n",
              " 'tamil_hope_test _labels.csv',\n",
              " 'malayalam_hope_test_labels.csv']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FUPkZxjLVjLK"
      },
      "source": [
        "## Loading training data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 377
        },
        "id": "joone3H4bINR",
        "outputId": "a61be324-fbe6-4136-b15c-29d06789ca96"
      },
      "source": [
        "train_df_path = join(dataset_dir,'tamil_hope_first_train.csv')\r\n",
        "# Load the dataset into a pandas dataframe.\r\n",
        "train_df = pd.read_csv(train_df_path)\r\n",
        "\r\n",
        "# Report the number of sentences.\r\n",
        "print('Number of training sentences: {:,}\\n'.format(train_df.shape[0]))\r\n",
        "\r\n",
        "# Display 10 random rows from the data.\r\n",
        "train_df.sample(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of training sentences: 16,160\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "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>text</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1799</th>\n",
              "      <td>Hiii madan gowri</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6624</th>\n",
              "      <td>கருவாச்சி கனவுகள்(karuvachi kanavugal)‍ என் தோ...</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13554</th>\n",
              "      <td>Gowri sir ennathu ithu views konjama irukku</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2308</th>\n",
              "      <td>Fantastic news...nalla coverage MG</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7836</th>\n",
              "      <td>Bro videoder or vidmate ku  alternative iruka</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11822</th>\n",
              "      <td>ஆனா  நண்ப யூஸ் பண்ற போன் எல்லாம் சைனா மொபைல் தான</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12830</th>\n",
              "      <td>Apo nokia kuda va</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7066</th>\n",
              "      <td>Aiyoo YouTube ilana sethuduva bro</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1530</th>\n",
              "      <td>Ippadiye Kora sollitu iruntha ethuvum maarathu...</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12335</th>\n",
              "      <td>Bro American double sun rise pathi video podun...</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                    text            label\n",
              "1799                                    Hiii madan gowri  Non_hope_speech\n",
              "6624   கருவாச்சி கனவுகள்(karuvachi kanavugal)‍ என் தோ...      Hope_speech\n",
              "13554        Gowri sir ennathu ithu views konjama irukku      Hope_speech\n",
              "2308                  Fantastic news...nalla coverage MG      Hope_speech\n",
              "7836       Bro videoder or vidmate ku  alternative iruka  Non_hope_speech\n",
              "11822   ஆனா  நண்ப யூஸ் பண்ற போன் எல்லாம் சைனா மொபைல் தான      Hope_speech\n",
              "12830                                  Apo nokia kuda va      Hope_speech\n",
              "7066                   Aiyoo YouTube ilana sethuduva bro  Non_hope_speech\n",
              "1530   Ippadiye Kora sollitu iruntha ethuvum maarathu...  Non_hope_speech\n",
              "12335  Bro American double sun rise pathi video podun...      Hope_speech"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w4VW351225YK",
        "outputId": "fd29720d-de08-4e80-ace7-e2bbdbe61be5"
      },
      "source": [
        "train_df['label'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Non_hope_speech    7872\n",
              "Hope_speech        6327\n",
              "not-Tamil          1961\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ooOv9QZtnLIF"
      },
      "source": [
        "\r\n",
        "le = LabelEncoder()\r\n",
        "train_df['label'] = le.fit_transform(train_df['label'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a89Bf4LpnigU",
        "outputId": "49f8e89e-3c41-4502-c2de-130f86cf1352"
      },
      "source": [
        "train_df['label'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1    7872\n",
              "0    6327\n",
              "2    1961\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bvNeR4lTVnK0"
      },
      "source": [
        "### Loading `train_sentences` and `train_labels`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EufQZx2dbKMT"
      },
      "source": [
        "# Get the lists of sentences and their labels.\r\n",
        "train_sentences = train_df['text'].values\r\n",
        "train_labels = train_df['label'].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wnZSg5f7ZWs_"
      },
      "source": [
        "## Saving Class Names"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "83tTT-wbZa1o",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a0cc8e29-7db8-4987-962b-66eda4105be0"
      },
      "source": [
        "class_names = le.inverse_transform(range(0,len(train_df['label'].unique())))\r\n",
        "class_names"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Hope_speech', 'Non_hope_speech', 'not-Tamil'], dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ToUbAgZ9ap63"
      },
      "source": [
        "## Loading the validation data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 377
        },
        "id": "Dle8VzoAjOLS",
        "outputId": "0ffae03c-a4a1-4568-a228-815882479190"
      },
      "source": [
        "val_df_path = join(dataset_dir,'tamil_hope_first_dev.csv')\r\n",
        "# Load the dataset into a pandas dataframe.\r\n",
        "val_df = pd.read_csv(val_df_path)\r\n",
        "\r\n",
        "# Report the number of sentences.\r\n",
        "print('Number of validation sentences: {:,}\\n'.format(val_df.shape[0]))\r\n",
        "\r\n",
        "# Display 10 random rows from the data.\r\n",
        "val_df.sample(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of validation sentences: 2,018\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "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>text</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>816</th>\n",
              "      <td>Frame  aasana panni face romba bright  aachi.n...</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1558</th>\n",
              "      <td>Bro ena aachu video kaanom</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>363</th>\n",
              "      <td>Uninstall senja pathadhu</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1093</th>\n",
              "      <td>Hi bro I am balaji  I want to talk with you</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>992</th>\n",
              "      <td>Bro video call ku google duo use pannalama</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>63</th>\n",
              "      <td>Na bio medical engineering padikiran</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1293</th>\n",
              "      <td>இதுவரைக்கும் நீங்க விளம்பர பண்ணுன பொருள்கள் எல...</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1847</th>\n",
              "      <td>Dalai Lama Pathi sollunga</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1393</th>\n",
              "      <td>Tamil pesum anithu nala ulangalukkum Vanakkam</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1413</th>\n",
              "      <td>Evlo than chinavala namukku nanmai irukkunnu n...</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                   text            label\n",
              "816   Frame  aasana panni face romba bright  aachi.n...      Hope_speech\n",
              "1558                         Bro ena aachu video kaanom  Non_hope_speech\n",
              "363                            Uninstall senja pathadhu  Non_hope_speech\n",
              "1093        Hi bro I am balaji  I want to talk with you  Non_hope_speech\n",
              "992          Bro video call ku google duo use pannalama  Non_hope_speech\n",
              "63                 Na bio medical engineering padikiran  Non_hope_speech\n",
              "1293  இதுவரைக்கும் நீங்க விளம்பர பண்ணுன பொருள்கள் எல...  Non_hope_speech\n",
              "1847                          Dalai Lama Pathi sollunga      Hope_speech\n",
              "1393      Tamil pesum anithu nala ulangalukkum Vanakkam      Hope_speech\n",
              "1413  Evlo than chinavala namukku nanmai irukkunnu n...  Non_hope_speech"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AypCmtBcDV6v",
        "outputId": "b9630ab5-b9bd-49b6-a48d-09a631dc89d1"
      },
      "source": [
        "val_df['label'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Non_hope_speech    998\n",
              "Hope_speech        757\n",
              "not-Tamil          263\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-CgK81lGDPKC"
      },
      "source": [
        "val_df['label'] = le.transform(val_df['label'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Yfgn-E6JDZQZ",
        "outputId": "845c2add-73e6-4fb1-81e3-a37129fbf169"
      },
      "source": [
        "val_df['label'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1    998\n",
              "0    757\n",
              "2    263\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DMdNNVqrDCuN"
      },
      "source": [
        "### Loading `val_sentences` and `val_labels`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LvGhjUkaDCuO"
      },
      "source": [
        "# Get the lists of sentences and their labels.\r\n",
        "val_sentences = val_df['text'].values\r\n",
        "val_labels = val_df['label'].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4mgM6b0yL0NM"
      },
      "source": [
        "## Loading the test data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 377
        },
        "id": "d7tVWiIeL0NN",
        "outputId": "efaf3a46-e69b-4486-ed20-f8c3f8b208b5"
      },
      "source": [
        "test_df_path = join(dataset_dir,'tamil_hope_test _labels.csv')\r\n",
        "# Load the dataset into a pandas dataframe.\r\n",
        "test_df = pd.read_csv(test_df_path , sep = ';')\r\n",
        "\r\n",
        "# Report the number of sentences.\r\n",
        "print('Number of test sentences: {:,}\\n'.format(test_df.shape[0]))\r\n",
        "\r\n",
        "# Display 10 random rows from the data.\r\n",
        "test_df.sample(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of test sentences: 2,020\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "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>text</th>\n",
              "      <th>label</th>\n",
              "      <th>Unnamed: 2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>257</th>\n",
              "      <td>Back poitu vandhaa varadhu</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1453</th>\n",
              "      <td>நினைத்தால் மட்டும் போதாது bro</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1607</th>\n",
              "      <td>JIO app Data va hack panna mattangala brother</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1932</th>\n",
              "      <td>First tq so much bro ... daily full day spend ...</td>\n",
              "      <td>Hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>637</th>\n",
              "      <td>Super bro</td>\n",
              "      <td>Hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>779</th>\n",
              "      <td>Chinese app Chinese app nu solringle ana phone...</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1863</th>\n",
              "      <td>MadhanGowri  army</td>\n",
              "      <td>not-Tamil</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>932</th>\n",
              "      <td>I love my India</td>\n",
              "      <td>Hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>451</th>\n",
              "      <td>India Ku China nikkuma???</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>209</th>\n",
              "      <td>Mg squad from erode</td>\n",
              "      <td>not-Tamil</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                   text  ... Unnamed: 2\n",
              "257                          Back poitu vandhaa varadhu  ...        NaN\n",
              "1453                      நினைத்தால் மட்டும் போதாது bro  ...        NaN\n",
              "1607      JIO app Data va hack panna mattangala brother  ...        NaN\n",
              "1932  First tq so much bro ... daily full day spend ...  ...        NaN\n",
              "637                                           Super bro  ...        NaN\n",
              "779   Chinese app Chinese app nu solringle ana phone...  ...        NaN\n",
              "1863                                  MadhanGowri  army  ...        NaN\n",
              "932                                     I love my India  ...        NaN\n",
              "451                           India Ku China nikkuma???  ...        NaN\n",
              "209                                 Mg squad from erode  ...        NaN\n",
              "\n",
              "[10 rows x 3 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wkt73boTL0NO",
        "outputId": "b719ee55-934f-4660-e464-f8c8aac4d8ea"
      },
      "source": [
        "test_df['label'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Non_hope_speech    946\n",
              "Hope_speech        815\n",
              "not-Tamil          259\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ProewaglL0NP"
      },
      "source": [
        "test_df['label'] = le.transform(test_df['label'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GYR2XokSL0NP",
        "outputId": "bf545635-0e09-417a-f962-800f4f90869f"
      },
      "source": [
        "test_df['label'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1    946\n",
              "0    815\n",
              "2    259\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eCCNrKNmL0NQ"
      },
      "source": [
        "### Loading `test_sentences` and `test_labels`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NvBSOGdkL0NR"
      },
      "source": [
        "# Get the lists of sentences and their labels.\r\n",
        "test_sentences = test_df['text'].values\r\n",
        "test_labels = test_df['label'].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fS2UcOiEVuhm"
      },
      "source": [
        "## Handling class imbalance using `sklearn's compute_class_weight`\r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WL4tP_YkI1mY"
      },
      "source": [
        "\r\n",
        "# #compute the class weights\r\n",
        "# class_wts = compute_class_weight('balanced', np.unique(train_labels), train_labels)\r\n",
        "\r\n",
        "# print(class_wts)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z_WjMTuibsrf"
      },
      "source": [
        "## Helper Functions"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bef28eHNV-_d"
      },
      "source": [
        "#### 1) Update Interval Function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lBK0ujfEbMir"
      },
      "source": [
        "def good_update_interval(total_iters, num_desired_updates):\r\n",
        "    '''\r\n",
        "    This function will try to pick an intelligent progress update interval \r\n",
        "    based on the magnitude of the total iterations.\r\n",
        "\r\n",
        "    Parameters:\r\n",
        "      `total_iters` - The number of iterations in the for-loop.\r\n",
        "      `num_desired_updates` - How many times we want to see an update over the \r\n",
        "                              course of the for-loop.\r\n",
        "    '''\r\n",
        "    # Divide the total iterations by the desired number of updates. Most likely\r\n",
        "    # this will be some ugly number.\r\n",
        "    exact_interval = total_iters / num_desired_updates\r\n",
        "\r\n",
        "    # The `round` function has the ability to round down a number to, e.g., the\r\n",
        "    # nearest thousandth: round(exact_interval, -3)\r\n",
        "    #\r\n",
        "    # To determine the magnitude to round to, find the magnitude of the total,\r\n",
        "    # and then go one magnitude below that.\r\n",
        "\r\n",
        "    # Get the order of magnitude of the total.\r\n",
        "    order_of_mag = len(str(total_iters)) - 1\r\n",
        "\r\n",
        "    # Our update interval should be rounded to an order of magnitude smaller. \r\n",
        "    round_mag = order_of_mag - 1\r\n",
        "\r\n",
        "    # Round down and cast to an int.\r\n",
        "    update_interval = int(round(exact_interval, -round_mag))\r\n",
        "\r\n",
        "    # Don't allow the interval to be zero!\r\n",
        "    if update_interval == 0:\r\n",
        "        update_interval = 1\r\n",
        "\r\n",
        "    return update_interval"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eMGycwGAWFyO"
      },
      "source": [
        "#### 2) Format time function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c9zVyNQLm-3q"
      },
      "source": [
        "def format_time(elapsed):\r\n",
        "    '''\r\n",
        "    Takes a time in seconds and returns a string hh:mm:ss\r\n",
        "    '''\r\n",
        "    # Round to the nearest second.\r\n",
        "    elapsed_rounded = int(round((elapsed)))\r\n",
        "    \r\n",
        "    # Format as hh:mm:ss\r\n",
        "    return str(datetime.timedelta(seconds=elapsed_rounded))\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EBAo04ENWw3n"
      },
      "source": [
        "#### 3) Make Smart Batches function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nof-AUeoiZc2"
      },
      "source": [
        "def make_smart_batches(text_samples, labels, batch_size,tokenizer,max_input_length):\r\n",
        "    '''\r\n",
        "    This function combines all of the required steps to prepare batches.\r\n",
        "    '''\r\n",
        "\r\n",
        "    print('Creating Smart Batches from {:,} examples with batch size {:,}...\\n'.format(len(text_samples), batch_size))\r\n",
        "\r\n",
        "    # =========================\r\n",
        "    #   Tokenize & Truncate\r\n",
        "    # =========================\r\n",
        "\r\n",
        "    full_input_ids = []\r\n",
        "\r\n",
        "    # Tokenize all training examples\r\n",
        "    print('Tokenizing {:,} samples...'.format(len(labels)))\r\n",
        "\r\n",
        "    # Choose an interval on which to print progress updates.\r\n",
        "    update_interval = good_update_interval(total_iters=len(labels), num_desired_updates=10)\r\n",
        "\r\n",
        "    # For each training example...\r\n",
        "    for text in text_samples:\r\n",
        "        \r\n",
        "        # Report progress.\r\n",
        "        if ((len(full_input_ids) % update_interval) == 0):\r\n",
        "            print('  Tokenized {:,} samples.'.format(len(full_input_ids)))\r\n",
        "\r\n",
        "        # Tokenize the sample.\r\n",
        "        input_ids = tokenizer.encode(text=text,              # Text to encode.\r\n",
        "                                    add_special_tokens=True, # Do add specials.\r\n",
        "                                    max_length=max_input_length,      \r\n",
        "                                    truncation=True,         # Do Truncate!\r\n",
        "                                    padding=False)           # DO NOT pad.\r\n",
        "                                    \r\n",
        "        # Add the tokenized result to our list.\r\n",
        "        full_input_ids.append(input_ids)\r\n",
        "        \r\n",
        "    print('DONE.')\r\n",
        "    print('{:>10,} samples\\n'.format(len(full_input_ids)))\r\n",
        "\r\n",
        "    # =========================\r\n",
        "    #      Select Batches\r\n",
        "    # =========================    \r\n",
        "\r\n",
        "    # Sort the two lists together by the length of the input sequence.\r\n",
        "    samples = sorted(zip(full_input_ids, labels), key=lambda x: len(x[0]))\r\n",
        "\r\n",
        "    print('{:>10,} samples after sorting\\n'.format(len(samples)))\r\n",
        "\r\n",
        "    import random\r\n",
        "\r\n",
        "    # List of batches that we'll construct.\r\n",
        "    batch_ordered_sentences = []\r\n",
        "    batch_ordered_labels = []\r\n",
        "\r\n",
        "    print('Creating batches of size {:}...'.format(batch_size))\r\n",
        "\r\n",
        "    # Choose an interval on which to print progress updates.\r\n",
        "    update_interval = good_update_interval(total_iters=len(samples), num_desired_updates=10)\r\n",
        "    \r\n",
        "    # Loop over all of the input samples...    \r\n",
        "    while len(samples) > 0:\r\n",
        "        \r\n",
        "        # Report progress.\r\n",
        "        if ((len(batch_ordered_sentences) % update_interval) == 0 \\\r\n",
        "            and not len(batch_ordered_sentences) == 0):\r\n",
        "            print('  Selected {:,} batches.'.format(len(batch_ordered_sentences)))\r\n",
        "\r\n",
        "        # `to_take` is our actual batch size. It will be `batch_size` until \r\n",
        "        # we get to the last batch, which may be smaller. \r\n",
        "        to_take = min(batch_size, len(samples))\r\n",
        "\r\n",
        "        # Pick a random index in the list of remaining samples to start\r\n",
        "        # our batch at.\r\n",
        "        select = random.randint(0, len(samples) - to_take)\r\n",
        "\r\n",
        "        # Select a contiguous batch of samples starting at `select`.\r\n",
        "        #print(\"Selecting batch from {:} to {:}\".format(select, select+to_take))\r\n",
        "        batch = samples[select:(select + to_take)]\r\n",
        "\r\n",
        "        #print(\"Batch length:\", len(batch))\r\n",
        "\r\n",
        "        # Each sample is a tuple--split them apart to create a separate list of \r\n",
        "        # sequences and a list of labels for this batch.\r\n",
        "        batch_ordered_sentences.append([s[0] for s in batch])\r\n",
        "        batch_ordered_labels.append([s[1] for s in batch])\r\n",
        "\r\n",
        "        # Remove these samples from the list.\r\n",
        "        del samples[select:select + to_take]\r\n",
        "\r\n",
        "    print('\\n  DONE - Selected {:,} batches.\\n'.format(len(batch_ordered_sentences)))\r\n",
        "\r\n",
        "    # =========================\r\n",
        "    #        Add Padding\r\n",
        "    # =========================    \r\n",
        "\r\n",
        "    print('Padding out sequences within each batch...')\r\n",
        "\r\n",
        "    py_inputs = []\r\n",
        "    py_attn_masks = []\r\n",
        "    py_labels = []\r\n",
        "\r\n",
        "    # For each batch...\r\n",
        "    for (batch_inputs, batch_labels) in zip(batch_ordered_sentences, batch_ordered_labels):\r\n",
        "\r\n",
        "        # New version of the batch, this time with padded sequences and now with\r\n",
        "        # attention masks defined.\r\n",
        "        batch_padded_inputs = []\r\n",
        "        batch_attn_masks = []\r\n",
        "        \r\n",
        "        # First, find the longest sample in the batch. \r\n",
        "        # Note that the sequences do currently include the special tokens!\r\n",
        "        max_size = max([len(sen) for sen in batch_inputs])\r\n",
        "\r\n",
        "        # For each input in this batch...\r\n",
        "        for sen in batch_inputs:\r\n",
        "            \r\n",
        "            # How many pad tokens do we need to add?\r\n",
        "            num_pads = max_size - len(sen)\r\n",
        "\r\n",
        "            # Add `num_pads` padding tokens to the end of the sequence.\r\n",
        "            padded_input = sen + [tokenizer.pad_token_id]*num_pads\r\n",
        "\r\n",
        "            # Define the attention mask--it's just a `1` for every real token\r\n",
        "            # and a `0` for every padding token.\r\n",
        "            attn_mask = [1] * len(sen) + [0] * num_pads\r\n",
        "\r\n",
        "            # Add the padded results to the batch.\r\n",
        "            batch_padded_inputs.append(padded_input)\r\n",
        "            batch_attn_masks.append(attn_mask)\r\n",
        "\r\n",
        "        # Our batch has been padded, so we need to save this updated batch.\r\n",
        "        # We also need the inputs to be PyTorch tensors, so we'll do that here.\r\n",
        "        # Todo - Michael's code specified \"dtype=torch.long\"\r\n",
        "        py_inputs.append(torch.tensor(batch_padded_inputs))\r\n",
        "        py_attn_masks.append(torch.tensor(batch_attn_masks))\r\n",
        "        py_labels.append(torch.tensor(batch_labels))\r\n",
        "    \r\n",
        "    print('  DONE.')\r\n",
        "\r\n",
        "    # Return the smart-batched dataset!\r\n",
        "    return (py_inputs, py_attn_masks, py_labels)\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5816-I6miXJ1"
      },
      "source": [
        "### 4) Make Smart Batches (On Test Set)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ssv3g8tzEdWa"
      },
      "source": [
        "def make_smart_batches_on_test(text_samples, text_ids, batch_size,tokenizer,max_input_length):\r\n",
        "    '''\r\n",
        "    This function combines all of the required steps to prepare batches.\r\n",
        "    '''\r\n",
        "\r\n",
        "    print('Creating Smart Batches from {:,} examples with batch size {:,}...\\n'.format(len(text_samples), batch_size))\r\n",
        "\r\n",
        "    # =========================\r\n",
        "    #   Tokenize & Truncate\r\n",
        "    # =========================\r\n",
        "\r\n",
        "    full_input_ids = []\r\n",
        "\r\n",
        "    # Tokenize all training examples\r\n",
        "    print('Tokenizing {:,} samples...'.format(len(text_samples)))\r\n",
        "\r\n",
        "    # Choose an interval on which to print progress updates.\r\n",
        "    update_interval = good_update_interval(total_iters=len(text_samples), num_desired_updates=10)\r\n",
        "\r\n",
        "    # For each training example...\r\n",
        "    for text in text_samples:\r\n",
        "        \r\n",
        "        # Report progress.\r\n",
        "        if ((len(full_input_ids) % update_interval) == 0):\r\n",
        "            print('  Tokenized {:,} samples.'.format(len(full_input_ids)))\r\n",
        "\r\n",
        "        # Tokenize the sample.\r\n",
        "        input_ids = tokenizer.encode(text=text,              # Text to encode.\r\n",
        "                                    add_special_tokens=True, # Do add specials.\r\n",
        "                                    max_length=max_input_length,      \r\n",
        "                                    truncation=True,         # Do Truncate!\r\n",
        "                                    padding=False)           # DO NOT pad.\r\n",
        "                                    \r\n",
        "        # Add the tokenized result to our list.\r\n",
        "        full_input_ids.append(input_ids)\r\n",
        "        \r\n",
        "    print('DONE.')\r\n",
        "    print('{:>10,} samples\\n'.format(len(full_input_ids)))\r\n",
        "\r\n",
        "    # =========================\r\n",
        "    #      Select Batches\r\n",
        "    # =========================    \r\n",
        "\r\n",
        "    # Sort the two lists together by the length of the input sequence.\r\n",
        "    samples = sorted(zip(full_input_ids, text_ids), key=lambda x: len(x[0]))\r\n",
        "\r\n",
        "    print('{:>10,} samples after sorting\\n'.format(len(samples)))\r\n",
        "\r\n",
        "    import random\r\n",
        "\r\n",
        "    # List of batches that we'll construct.\r\n",
        "    batch_ordered_sentences = []\r\n",
        "    batch_ordered_ids = []\r\n",
        "\r\n",
        "    print('Creating batches of size {:}...'.format(batch_size))\r\n",
        "\r\n",
        "    # Choose an interval on which to print progress updates.\r\n",
        "    update_interval = good_update_interval(total_iters=len(samples), num_desired_updates=10)\r\n",
        "    \r\n",
        "    # Loop over all of the input samples...    \r\n",
        "    while len(samples) > 0:\r\n",
        "        \r\n",
        "        # Report progress.\r\n",
        "        if ((len(batch_ordered_sentences) % update_interval) == 0 \\\r\n",
        "            and not len(batch_ordered_sentences) == 0):\r\n",
        "            print('  Selected {:,} batches.'.format(len(batch_ordered_sentences)))\r\n",
        "\r\n",
        "        # `to_take` is our actual batch size. It will be `batch_size` until \r\n",
        "        # we get to the last batch, which may be smaller. \r\n",
        "        to_take = min(batch_size, len(samples))\r\n",
        "\r\n",
        "        # Pick a random index in the list of remaining samples to start\r\n",
        "        # our batch at.\r\n",
        "        select = random.randint(0, len(samples) - to_take)\r\n",
        "\r\n",
        "        # Select a contiguous batch of samples starting at `select`.\r\n",
        "        #print(\"Selecting batch from {:} to {:}\".format(select, select+to_take))\r\n",
        "        batch = samples[select:(select + to_take)]\r\n",
        "\r\n",
        "        #print(\"Batch length:\", len(batch))\r\n",
        "\r\n",
        "        # Each sample is a tuple--split them apart to create a separate list of \r\n",
        "        # sequences and a list of labels for this batch.\r\n",
        "        batch_ordered_sentences.append([s[0] for s in batch])\r\n",
        "        batch_ordered_ids.append([s[1] for s in batch])\r\n",
        "\r\n",
        "        # Remove these samples from the list.\r\n",
        "        del samples[select:select + to_take]\r\n",
        "\r\n",
        "    print('\\n  DONE - Selected {:,} batches.\\n'.format(len(batch_ordered_sentences)))\r\n",
        "\r\n",
        "    # =========================\r\n",
        "    #        Add Padding\r\n",
        "    # =========================    \r\n",
        "\r\n",
        "    print('Padding out sequences within each batch...')\r\n",
        "\r\n",
        "    py_inputs = []\r\n",
        "    py_attn_masks = []\r\n",
        "    py_ids = []\r\n",
        "\r\n",
        "    # For each batch...\r\n",
        "    for (batch_inputs, batch_ids) in zip(batch_ordered_sentences, batch_ordered_ids):\r\n",
        "\r\n",
        "        # New version of the batch, this time with padded sequences and now with\r\n",
        "        # attention masks defined.\r\n",
        "        batch_padded_inputs = []\r\n",
        "        batch_attn_masks = []\r\n",
        "        \r\n",
        "        # First, find the longest sample in the batch. \r\n",
        "        # Note that the sequences do currently include the special tokens!\r\n",
        "        max_size = max([len(sen) for sen in batch_inputs])\r\n",
        "\r\n",
        "        # For each input in this batch...\r\n",
        "        for sen in batch_inputs:\r\n",
        "            \r\n",
        "            # How many pad tokens do we need to add?\r\n",
        "            num_pads = max_size - len(sen)\r\n",
        "\r\n",
        "            # Add `num_pads` padding tokens to the end of the sequence.\r\n",
        "            padded_input = sen + [tokenizer.pad_token_id]*num_pads\r\n",
        "\r\n",
        "            # Define the attention mask--it's just a `1` for every real token\r\n",
        "            # and a `0` for every padding token.\r\n",
        "            attn_mask = [1] * len(sen) + [0] * num_pads\r\n",
        "\r\n",
        "            # Add the padded results to the batch.\r\n",
        "            batch_padded_inputs.append(padded_input)\r\n",
        "            batch_attn_masks.append(attn_mask)\r\n",
        "\r\n",
        "        # Our batch has been padded, so we need to save this updated batch.\r\n",
        "        # We also need the inputs to be PyTorch tensors, so we'll do that here.\r\n",
        "        # Todo - Michael's code specified \"dtype=torch.long\"\r\n",
        "        py_inputs.append(torch.tensor(batch_padded_inputs))\r\n",
        "        py_attn_masks.append(torch.tensor(batch_attn_masks))\r\n",
        "        py_ids.append(torch.tensor(batch_ids))\r\n",
        "    \r\n",
        "    print('  DONE.')\r\n",
        "\r\n",
        "    # Return the smart-batched dataset!\r\n",
        "    return (py_inputs, py_attn_masks, py_ids)\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fT7qY9gtEdWc"
      },
      "source": [
        "### 5) Function for calculating accuracy per class"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tSHVuo18EdWc"
      },
      "source": [
        "def accuracy_per_class(preds, labels , label_dict):\r\n",
        "    \r\n",
        "    preds_flat = np.argmax(preds, axis=1).flatten()\r\n",
        "    labels_flat = labels.flatten()\r\n",
        "\r\n",
        "    for label in np.unique(labels_flat):\r\n",
        "        y_preds = preds_flat[labels_flat==label]\r\n",
        "        y_true = labels_flat[labels_flat==label]\r\n",
        "        print(f'Class: {label_dict[label]}')\r\n",
        "        print(f'Accuracy: {round(100*len(y_preds[y_preds==label])/len(y_true) , 2)}\\n')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MLcm5L2MEdWd"
      },
      "source": [
        "### 6) Function for plotting training history"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rLLSWnMTEdWd"
      },
      "source": [
        "def plot_training_history(history):\r\n",
        "  fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(18, 6))\r\n",
        "\r\n",
        "  ax1.plot(history['train_loss'], label='train loss')\r\n",
        "  ax1.plot(history['val_loss'], label='validation loss')\r\n",
        "\r\n",
        "  ax1.xaxis.set_major_locator(MaxNLocator(integer=True))\r\n",
        "  ax1.set_ylim([-0.05, 1.05])\r\n",
        "  ax1.legend()\r\n",
        "  ax1.set_ylabel('Loss')\r\n",
        "  ax1.set_xlabel('Epoch')\r\n",
        "\r\n",
        "  ax2.plot(history['train_acc'], label='train accuracy')\r\n",
        "  ax2.plot(history['val_acc'], label='validation accuracy')\r\n",
        "\r\n",
        "  ax2.xaxis.set_major_locator(MaxNLocator(integer=True))\r\n",
        "  ax2.set_ylim([-0.05, 1.05])\r\n",
        "  ax2.legend()\r\n",
        "\r\n",
        "  ax2.set_ylabel('Accuracy')\r\n",
        "  ax2.set_xlabel('Epoch')\r\n",
        "\r\n",
        "  fig.suptitle('Training history')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mHcxGVmREdWd"
      },
      "source": [
        "### 7) Function to check accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VF8tJIO6EdWe"
      },
      "source": [
        "def check_accuracy(predictions,true_labels):\r\n",
        "  # Combine the results across the batches.\r\n",
        "  predictions = np.concatenate(predictions, axis=0)\r\n",
        "  true_labels = np.concatenate(true_labels, axis=0)\r\n",
        "\r\n",
        "  # Choose the label with the highest score as our prediction.\r\n",
        "  preds = np.argmax(predictions, axis=1).flatten()\r\n",
        "\r\n",
        "  # Calculate simple flat accuracy -- number correct over total number.\r\n",
        "  accuracy = (preds == true_labels).mean()\r\n",
        "\r\n",
        "  return accuracy"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uW3YFVk7EdWe"
      },
      "source": [
        "### 8) Function to evaluate model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6ukojsSXEdWe"
      },
      "source": [
        "def eval_model(model,py_inputs, py_attn_masks, py_labels):\r\n",
        "# Prediction on test set\r\n",
        "  t0 = time.time()\r\n",
        "\r\n",
        "  print('Doing validation on {:,} sentences...'.format(len(py_labels)))\r\n",
        "\r\n",
        "  # Put model in evaluation mode\r\n",
        "  model.eval()\r\n",
        "\r\n",
        "  # Tracking variables \r\n",
        "  predictions , true_labels = [], []\r\n",
        "\r\n",
        "  # Choose an interval on which to print progress updates.\r\n",
        "  update_interval = good_update_interval(total_iters=len(py_inputs), num_desired_updates=10)\r\n",
        "\r\n",
        "  # Measure elapsed time.\r\n",
        "  t0 = time.time()\r\n",
        "  total_val_loss = 0\r\n",
        "\r\n",
        "  # Put model in evaluation mode\r\n",
        "  model.eval()\r\n",
        "\r\n",
        "  # For each batch of training data...\r\n",
        "  for step in range(0, len(py_inputs)):\r\n",
        "\r\n",
        "      # Progress update every 100 batches.\r\n",
        "      if step % update_interval == 0 and not step == 0:\r\n",
        "          # Calculate elapsed time in minutes.\r\n",
        "          elapsed = format_time(time.time() - t0)\r\n",
        "          \r\n",
        "          # Calculate the time remaining based on our progress.\r\n",
        "          steps_per_sec = (time.time() - t0) / step\r\n",
        "          remaining_sec = steps_per_sec * (len(py_inputs) - step)\r\n",
        "          remaining = format_time(remaining_sec)\r\n",
        "\r\n",
        "          # Report progress.\r\n",
        "          print('  Batch {:>7,}  of  {:>7,}.    Elapsed: {:}.  Remaining: {:}'.format(step, len(py_inputs), elapsed, remaining))\r\n",
        "\r\n",
        "      # Copy the batch to the GPU.\r\n",
        "      b_input_ids = py_inputs[step].to(device)\r\n",
        "      b_input_mask = py_attn_masks[step].to(device)\r\n",
        "      b_labels = py_labels[step].to(device)\r\n",
        "    \r\n",
        "      # Telling the model not to compute or store gradients, saving memory and \r\n",
        "      # speeding up prediction\r\n",
        "      with torch.no_grad():\r\n",
        "        output = model(b_input_ids, \r\n",
        "                             token_type_ids=None, \r\n",
        "                             attention_mask=b_input_mask, \r\n",
        "                             labels=b_labels)\r\n",
        "\r\n",
        "\r\n",
        "      logits = output.logits \r\n",
        "      loss = output.loss\r\n",
        "      # Move logits and labels to CPU\r\n",
        "      logits = logits.detach().cpu().numpy()\r\n",
        "      label_ids = b_labels.to('cpu').numpy()\r\n",
        "    \r\n",
        "      # Store predictions and true labels\r\n",
        "      predictions.append(logits)\r\n",
        "      true_labels.append(label_ids)\r\n",
        "\r\n",
        "      total_val_loss += loss.item()\r\n",
        "  \r\n",
        "  val_accuracy = check_accuracy(predictions,true_labels)\r\n",
        "\r\n",
        "  validation_time = format_time(time.time() - t0)\r\n",
        "  \r\n",
        "  # Calculate the average loss over all of the batches.\r\n",
        "  avg_val_loss = total_val_loss / len(py_inputs)     \r\n",
        "    \r\n",
        "\r\n",
        "  # print('Accuracy: {:.3f}'.format(val_accuracy))\r\n",
        "  return (avg_val_loss,val_accuracy,validation_time)\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rj5Zao7WEdWe"
      },
      "source": [
        "### 9) Function for making predictions on our test dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ChLIow-bEdWf"
      },
      "source": [
        "def get_predictions(py_inputs, py_attn_masks, py_labels):\r\n",
        "\r\n",
        "  print('Predicting labels for {:,} test batches...'.format(len(py_labels)))\r\n",
        "\r\n",
        "  # Put model in evaluation mode\r\n",
        "  model.eval()\r\n",
        "\r\n",
        "  # Tracking variables \r\n",
        "  predictions , true_labels = [], []\r\n",
        "\r\n",
        "  # Choose an interval on which to print progress updates.\r\n",
        "  update_interval = good_update_interval(total_iters=len(py_inputs), num_desired_updates=10)\r\n",
        "\r\n",
        "  # Measure elapsed time.\r\n",
        "  t0 = time.time()\r\n",
        "\r\n",
        "  # Put model in evaluation mode\r\n",
        "  model.eval()\r\n",
        "\r\n",
        "  # For each batch of training data...\r\n",
        "  for step in range(0, len(py_inputs)):\r\n",
        "\r\n",
        "      # Progress update every 100 batches.\r\n",
        "      if step % update_interval == 0 and not step == 0:\r\n",
        "          # Calculate elapsed time in minutes.\r\n",
        "          elapsed = format_time(time.time() - t0)\r\n",
        "          \r\n",
        "          # Calculate the time remaining based on our progress.\r\n",
        "          steps_per_sec = (time.time() - t0) / step\r\n",
        "          remaining_sec = steps_per_sec * (len(py_inputs) - step)\r\n",
        "          remaining = format_time(remaining_sec)\r\n",
        "\r\n",
        "          # Report progress.\r\n",
        "          print('  Batch {:>7,}  of  {:>7,}.    Elapsed: {:}.  Remaining: {:}'.format(step, len(py_inputs), elapsed, remaining))\r\n",
        "\r\n",
        "      # Copy the batch to the GPU.\r\n",
        "      b_input_ids = py_inputs[step].to(device)\r\n",
        "      b_input_mask = py_attn_masks[step].to(device)\r\n",
        "      b_labels = py_labels[step].to(device)\r\n",
        "    \r\n",
        "      # Telling the model not to compute or store gradients, saving memory and \r\n",
        "      # speeding up prediction\r\n",
        "      with torch.no_grad():\r\n",
        "          # Forward pass, calculate logit predictions\r\n",
        "          outputs = model(b_input_ids, b_input_mask)\r\n",
        "\r\n",
        "      logits = outputs.logits \r\n",
        "\r\n",
        "      # Move logits and labels to CPU\r\n",
        "      logits = logits.detach().cpu().numpy()\r\n",
        "      label_ids = b_labels.to('cpu').numpy()\r\n",
        "    \r\n",
        "      # Store predictions and true labels\r\n",
        "      predictions.append(logits)\r\n",
        "      true_labels.append(label_ids)\r\n",
        "      # Combine the results across the batches.\r\n",
        "  predictions = np.concatenate(predictions, axis=0)\r\n",
        "  true_labels = np.concatenate(true_labels, axis=0)\r\n",
        "\r\n",
        "      # Choose the label with the highest score as our prediction.\r\n",
        "  preds = np.argmax(predictions, axis=1).flatten()\r\n",
        "\r\n",
        "  return (preds,true_labels)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G8-bi_GEEdWf"
      },
      "source": [
        "### 10) Function for making predictions on test dataset(without labels)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tsgswf85EdWf"
      },
      "source": [
        "def get_predictions_test(py_inputs, py_attn_masks,py_ids):\r\n",
        "\r\n",
        "  print('Predicting labels for {:,} test batches...'.format(len(py_inputs)))\r\n",
        "\r\n",
        "  # Put model in evaluation mode\r\n",
        "  model.eval()\r\n",
        "\r\n",
        "  # Tracking variables \r\n",
        "  predictions , true_labels ,ids = [], [] , []\r\n",
        "\r\n",
        "  # Choose an interval on which to print progress updates.\r\n",
        "  update_interval = good_update_interval(total_iters=len(py_inputs), num_desired_updates=10)\r\n",
        "\r\n",
        "  # Measure elapsed time.\r\n",
        "  t0 = time.time()\r\n",
        "\r\n",
        "  # Put model in evaluation mode\r\n",
        "  model.eval()\r\n",
        "\r\n",
        "  # For each batch of training data...\r\n",
        "  for step in range(0, len(py_inputs)):\r\n",
        "\r\n",
        "      # Progress update every 100 batches.\r\n",
        "      if step % update_interval == 0 and not step == 0:\r\n",
        "          # Calculate elapsed time in minutes.\r\n",
        "          elapsed = format_time(time.time() - t0)\r\n",
        "          \r\n",
        "          # Calculate the time remaining based on our progress.\r\n",
        "          steps_per_sec = (time.time() - t0) / step\r\n",
        "          remaining_sec = steps_per_sec * (len(py_inputs) - step)\r\n",
        "          remaining = format_time(remaining_sec)\r\n",
        "\r\n",
        "          # Report progress.\r\n",
        "          print('  Batch {:>7,}  of  {:>7,}.    Elapsed: {:}.  Remaining: {:}'.format(step, len(py_inputs), elapsed, remaining))\r\n",
        "\r\n",
        "      # Copy the batch to the GPU.\r\n",
        "      b_input_ids = py_inputs[step].to(device)\r\n",
        "      b_input_mask = py_attn_masks[step].to(device)\r\n",
        "      b_ids = py_ids[step].to(device)\r\n",
        "    \r\n",
        "      # Telling the model not to compute or store gradients, saving memory and \r\n",
        "      # speeding up prediction\r\n",
        "      with torch.no_grad():\r\n",
        "          # Forward pass, calculate logit predictions\r\n",
        "          outputs = model(b_input_ids, b_input_mask)\r\n",
        "\r\n",
        "      logits = outputs.logits \r\n",
        "\r\n",
        "      # Move logits and labels to CPU\r\n",
        "      logits = logits.detach().cpu().numpy()\r\n",
        "      b_ids = b_ids.detach().cpu().numpy()\r\n",
        "      # Store predictions\r\n",
        "      predictions.append(logits)\r\n",
        "      ids.append(b_ids)\r\n",
        "      # Combine the results across the batches.\r\n",
        "  predictions = np.concatenate(predictions, axis=0)\r\n",
        "  ids = np.concatenate(ids,axis=0)\r\n",
        "  # Choose the label with the highest score as our prediction.\r\n",
        "  preds = np.argmax(predictions, axis=1).flatten()\r\n",
        "\r\n",
        "  return ids,preds"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4Fh-Tb5OEdWf"
      },
      "source": [
        "### 11) Confusion Matrix Helper Function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XGsBuZ1kEdWg"
      },
      "source": [
        "def show_confusion_matrix(confusion_matrix, class_names):\r\n",
        "\r\n",
        "  cm = confusion_matrix.copy()\r\n",
        "\r\n",
        "  cell_counts = cm.flatten()\r\n",
        "\r\n",
        "  cm_row_norm = cm / cm.sum(axis=1)[:, np.newaxis]\r\n",
        "\r\n",
        "  row_percentages = [\"{0:.2f}\".format(value) for value in cm_row_norm.flatten()]\r\n",
        "\r\n",
        "  cell_labels = [f\"{cnt}\\n{per}\" for cnt, per in zip(cell_counts, row_percentages)]\r\n",
        "  cell_labels = np.asarray(cell_labels).reshape(cm.shape[0], cm.shape[1])\r\n",
        "\r\n",
        "  df_cm = pd.DataFrame(cm_row_norm, index=class_names, columns=class_names)\r\n",
        "\r\n",
        "  hmap = sns.heatmap(df_cm, annot=cell_labels, fmt=\"\", cmap=\"Blues\")\r\n",
        "  hmap.yaxis.set_ticklabels(hmap.yaxis.get_ticklabels(), rotation=0, ha='right')\r\n",
        "  hmap.xaxis.set_ticklabels(hmap.xaxis.get_ticklabels(), rotation=30, ha='right')\r\n",
        "  plt.ylabel('True Sign')\r\n",
        "  plt.xlabel('Predicted Sign');"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MBxHY4rMojd-"
      },
      "source": [
        "## Loading the tokenizer using AutoTokenizer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 180,
          "referenced_widgets": [
            "69259d26d2a946abb70e8fe6aaec116f",
            "3e8c2d279a724af8afefe735f9b234df",
            "0cbffe1248184c6a969719e642230b37",
            "6dcde407594a4c30857cf94ed09f93f7",
            "2af2f368b3ba4f37b90c7a5ba020ec0f",
            "a3a7cccc6d5d485699a581af509c7dd9",
            "9d115b32f16c424a99f3d399f2195712",
            "764371eb5bc347c6942e8706530b7bd8",
            "45927ca0331f41c881b769b6663ff07d",
            "682ffe42c93249a0b5f9070e8924580e",
            "6ab47a24a6814a38862108d247ef318e",
            "02e714b9aa8c47f7ae083f90d5eded27",
            "b3df70ab91be4033837494e5aa155515",
            "93a430e594ea45bd96161d1ee3fa06d4",
            "4f96192d52834249b9da927cf8c8ecad",
            "3ed032c36ae545fbb210f7715ea66943",
            "c3f8cd0a07694469b8824b3b8e3d3f49",
            "85b3438077bb4b15a90d907c1db6ee18",
            "066fd44de32d4fb68a43ead9def7ac9f",
            "678caa0cd6bc40a8893026f5a162e337",
            "9376f06f6be34932bad7982246ec5057",
            "1cbc7e2d150e40eeaa5dd8b1ddb3428c",
            "0180b937b0f1442382624ff5de6982af",
            "aae4a53b02064634b0a282c4cc36ff41"
          ]
        },
        "id": "fIlwEvcMb4pb",
        "outputId": "ec99e086-c6b8-43ca-d1cf-2fec2ae3e1ba"
      },
      "source": [
        "model_name = 'xlm-roberta-base'\r\n",
        "# Load the BERT tokenizer.\r\n",
        "print(f'Loading {model_name} tokenizer...')\r\n",
        "tokenizer = AutoTokenizer.from_pretrained(model_name, do_lower_case=False)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Loading xlm-roberta-base tokenizer...\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "69259d26d2a946abb70e8fe6aaec116f",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=512.0, style=ProgressStyle(description_…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "45927ca0331f41c881b769b6663ff07d",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=5069051.0, style=ProgressStyle(descript…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "c3f8cd0a07694469b8824b3b8e3d3f49",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=9096718.0, style=ProgressStyle(descript…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OVs3Jb3Vb89i"
      },
      "source": [
        "## Smart Batching of the training data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cWNt4eOLofmF"
      },
      "source": [
        "lengths = []\r\n",
        "for text in train_sentences:\r\n",
        "  lengths.append(len(text))       "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 427
        },
        "id": "hbdGRA3Bo02O",
        "outputId": "608552f9-c1cb-42c9-c3f9-ea382df07454"
      },
      "source": [
        "\r\n",
        "# Use plot styling from seaborn.\r\n",
        "sns.set(style='darkgrid')\r\n",
        "\r\n",
        "# Increase the plot size and font size.\r\n",
        "sns.set(font_scale=1.5)\r\n",
        "plt.rcParams[\"figure.figsize\"] = (12,6)\r\n",
        "\r\n",
        "plt.scatter(range(0, len(lengths)), lengths, marker=\"|\")\r\n",
        "\r\n",
        "plt.xlabel('Sample Number')\r\n",
        "plt.ylabel('Sequence Length')\r\n",
        "plt.title('Samples BEFORE Tokenizing')\r\n",
        "\r\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9rMhLNhPcZjU"
      },
      "source": [
        "### Tokenizing the sequences"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t7JsLi3EdEje"
      },
      "source": [
        "max_input_length = 400"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J968o8RJcUEd",
        "outputId": "c5667931-1d75-4f94-fda4-ee495e4943ab"
      },
      "source": [
        "full_input_ids = []\r\n",
        "labels = []\r\n",
        "\r\n",
        "# Tokenize all training examples\r\n",
        "print('Tokenizing {:,} training samples...'.format(len(train_sentences)))\r\n",
        "\r\n",
        "# Choose an interval on which to print progress updates.\r\n",
        "update_interval = good_update_interval(total_iters=len(train_sentences), num_desired_updates=10)\r\n",
        "\r\n",
        "# For each training example...\r\n",
        "for text in train_sentences:\r\n",
        "    \r\n",
        "    # Report progress.\r\n",
        "    if ((len(full_input_ids) % update_interval) == 0):\r\n",
        "        print('  Tokenized {:,} samples.'.format(len(full_input_ids)))\r\n",
        "\r\n",
        "    # Tokenize the sentence.\r\n",
        "    input_ids = tokenizer.encode(text=text,           \r\n",
        "                                 add_special_tokens=True, \r\n",
        "                                 max_length=max_input_length,  \r\n",
        "                                 truncation=True,     \r\n",
        "                                 padding=False)       \r\n",
        "                                 \r\n",
        "    # Add the tokenized result to our list.\r\n",
        "    full_input_ids.append(input_ids)\r\n",
        "    \r\n",
        "print('DONE.')\r\n",
        "print('{:>10,} samples'.format(len(full_input_ids)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Tokenizing 16,160 training samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "  Tokenized 4,000 samples.\n",
            "  Tokenized 6,000 samples.\n",
            "  Tokenized 8,000 samples.\n",
            "  Tokenized 10,000 samples.\n",
            "  Tokenized 12,000 samples.\n",
            "  Tokenized 14,000 samples.\n",
            "  Tokenized 16,000 samples.\n",
            "DONE.\n",
            "    16,160 samples\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rZzhO2tZeXur"
      },
      "source": [
        "# Get all of the lengths.\r\n",
        "unsorted_lengths = [len(x) for x in full_input_ids]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 427
        },
        "id": "x7c_mYuVed_H",
        "outputId": "5a0093b4-72c7-46d2-ddba-6badeb8d866a"
      },
      "source": [
        "\r\n",
        "# Use plot styling from seaborn.\r\n",
        "sns.set(style='darkgrid')\r\n",
        "\r\n",
        "# Increase the plot size and font size.\r\n",
        "sns.set(font_scale=1.5)\r\n",
        "plt.rcParams[\"figure.figsize\"] = (12,6)\r\n",
        "\r\n",
        "plt.scatter(range(0, len(unsorted_lengths)), unsorted_lengths, marker=\"|\")\r\n",
        "\r\n",
        "plt.xlabel('Sample Number')\r\n",
        "plt.ylabel('Sequence Length')\r\n",
        "plt.title('Samples BEFORE Sorting')\r\n",
        "\r\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7PaeohUVegmv"
      },
      "source": [
        "# Sort the two lists together by the length of the input sequence.\r\n",
        "train_samples = sorted(zip(full_input_ids, train_labels), key=lambda x: len(x[0]))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_9nyx-BtepE4",
        "outputId": "79444ac7-ecbb-4c7b-cb59-7fbe509be75e"
      },
      "source": [
        "train_samples[0:5]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[([0, 10897, 2], 1),\n",
              " ([0, 7082, 2], 2),\n",
              " ([0, 112285, 2], 2),\n",
              " ([0, 1374, 2], 2),\n",
              " ([0, 64848, 15122, 2], 2)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rBxJhEieetFh",
        "outputId": "7f9441bd-cd0f-42e3-f454-9980e0541280"
      },
      "source": [
        "print('Shortest sample:', len(train_samples[0][0]))\r\n",
        "print('Longest sample:', len(train_samples[-1][0]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shortest sample: 3\n",
            "Longest sample: 400\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P44bFOI-e7hW"
      },
      "source": [
        "# Get the new list of lengths after sorting.\r\n",
        "sorted_lengths = [len(s[0]) for s in train_samples]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 427
        },
        "id": "9P5ZX99le_G1",
        "outputId": "899e564a-a565-498d-dac5-2d1e937e753c"
      },
      "source": [
        "\r\n",
        "plt.plot(range(0, len(sorted_lengths)), sorted_lengths)\r\n",
        "\r\n",
        "plt.xlabel('Sample Number')\r\n",
        "plt.ylabel('Sequence Length')\r\n",
        "plt.title('Samples after Sorting')\r\n",
        "\r\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Eh1z0xugfF2t"
      },
      "source": [
        "### Random Batch Selections"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gPrYkFMcgm5H",
        "outputId": "aa664f86-1e38-4f09-f613-4dd567a6d065"
      },
      "source": [
        "train_samples[0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "([0, 10897, 2], 1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "isMxwgezfAvO"
      },
      "source": [
        "batch_size = 16"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "s0tjD8pIfJ0S",
        "outputId": "27c02361-5343-4e96-c54f-a75fcc3453e6"
      },
      "source": [
        "# List of batches that we'll construct.\r\n",
        "batch_ordered_sentences = []\r\n",
        "batch_ordered_labels = []\r\n",
        "\r\n",
        "print('Creating training batches of size {:}'.format(batch_size))\r\n",
        "\r\n",
        "# Loop over all of the input samples...    \r\n",
        "while len(train_samples) > 0:\r\n",
        "    \r\n",
        "    # Report progress.\r\n",
        "    if ((len(batch_ordered_sentences) % 100) == 0):\r\n",
        "        print('  Selected {:,} batches.'.format(len(batch_ordered_sentences)))\r\n",
        "\r\n",
        "    # `to_take` is our actual batch size. It will be `batch_size` until \r\n",
        "    # we get to the last batch, which may be smaller. \r\n",
        "    to_take = min(batch_size, len(train_samples))\r\n",
        "\r\n",
        "    # Pick a random index in the list of remaining samples to start\r\n",
        "    # our batch at.\r\n",
        "    select = random.randint(0, len(train_samples) - to_take)\r\n",
        "\r\n",
        "    # Select a contiguous batch of samples starting at `select`.\r\n",
        "    batch = train_samples[select:(select + to_take)]\r\n",
        "\r\n",
        "    # Each sample is a tuple--split them apart to create a separate list of \r\n",
        "    # sequences and a list of labels for this batch.\r\n",
        "    batch_ordered_sentences.append([s[0] for s in batch])\r\n",
        "    batch_ordered_labels.append([s[1] for s in batch])\r\n",
        "\r\n",
        "    # Remove these samples from the list.\r\n",
        "    del train_samples[select:select + to_take]\r\n",
        "\r\n",
        "print('\\n  DONE - {:,} batches.'.format(len(batch_ordered_sentences)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Creating training batches of size 16\n",
            "  Selected 0 batches.\n",
            "  Selected 100 batches.\n",
            "  Selected 200 batches.\n",
            "  Selected 300 batches.\n",
            "  Selected 400 batches.\n",
            "  Selected 500 batches.\n",
            "  Selected 600 batches.\n",
            "  Selected 700 batches.\n",
            "  Selected 800 batches.\n",
            "  Selected 900 batches.\n",
            "  Selected 1,000 batches.\n",
            "\n",
            "  DONE - 1,010 batches.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rf91Dd3LhBaP",
        "outputId": "74bbea54-f54f-421f-d1f2-5a744a5cc370"
      },
      "source": [
        "print(batch_ordered_sentences[0])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[0, 911, 549, 738, 5429, 14, 31534, 35686, 156, 66, 28929, 7113, 10500, 56, 2870, 27, 2], [0, 13177, 22, 76, 34391, 202986, 73, 31842, 2260, 76, 21402, 22, 76, 9519, 3408, 914, 2], [0, 22365, 14288, 7, 21, 39, 24927, 2652, 93, 8347, 2652, 35780, 1539, 33, 27, 149595, 2], [0, 6, 31756, 15202, 12928, 68489, 7963, 7114, 100710, 2798, 4930, 11920, 12928, 34044, 18890, 67336, 2], [0, 159, 291, 22, 5236, 109728, 1347, 8767, 5515, 12853, 402, 12853, 402, 10, 4464, 14, 2], [0, 9098, 8966, 509, 67, 15054, 416, 147202, 85, 51294, 13, 13269, 17, 206471, 5, 5, 2], [0, 13177, 17, 8805, 47746, 1518, 508, 3530, 52875, 7, 187, 193, 10508, 931, 32, 243691, 2], [0, 13177, 108, 33, 208, 64792, 177, 1263, 51, 80383, 2652, 93, 1916, 1539, 11, 13862, 2], [0, 227285, 80720, 34987, 53413, 14994, 6, 16027, 43542, 207091, 58981, 26437, 3686, 2650, 5, 5, 2], [0, 13177, 17, 18422, 9519, 150, 458, 60055, 429, 4527, 89174, 107, 24089, 1658, 24366, 14380, 2], [0, 92437, 272, 41392, 45146, 22, 5236, 4027, 24, 23, 933, 5584, 45146, 4911, 211, 11, 2], [0, 149978, 77773, 11, 31918, 843, 196, 37664, 77773, 11, 10832, 11993, 14233, 9000, 12152, 95424, 2], [0, 1374, 7492, 141663, 341, 10605, 1495, 9461, 973, 41368, 46653, 41368, 46653, 22162, 3769, 54174, 2], [0, 1374, 12662, 33768, 3000, 927, 6, 89625, 3984, 24, 2606, 3319, 4591, 449, 187, 1176, 2], [0, 40469, 83, 132283, 13862, 19, 73852, 307, 1601, 17, 771, 2096, 192, 6, 20446, 4070, 2], [0, 13177, 483, 927, 15393, 68360, 319, 13280, 300, 72033, 9519, 11, 19521, 33059, 7779, 243691, 2]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qmkQK_gVgiTu",
        "outputId": "36978b55-3bf9-4234-9b7f-a1f5347a9e3a"
      },
      "source": [
        "batch_ordered_labels[0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 0]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xZ0TFA8Mfn6m"
      },
      "source": [
        "### Padding"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c9rdswcpfTip"
      },
      "source": [
        "py_inputs = []\r\n",
        "py_attn_masks = []\r\n",
        "py_labels = []\r\n",
        "\r\n",
        "# For each batch...\r\n",
        "for (batch_inputs, batch_labels) in zip(batch_ordered_sentences, batch_ordered_labels):\r\n",
        "\r\n",
        "    # New version of the batch, this time with padded sequences and now with\r\n",
        "    # attention masks defined.\r\n",
        "    batch_padded_inputs = []\r\n",
        "    batch_attn_masks = []\r\n",
        "    \r\n",
        "    # First, find the longest sample in the batch. \r\n",
        "    # Note that the sequences do currently include the special tokens!\r\n",
        "    max_size = max([len(sen) for sen in batch_inputs])\r\n",
        "\r\n",
        "    #print('Max size:', max_size)\r\n",
        "\r\n",
        "    # For each input in this batch...\r\n",
        "    for sen in batch_inputs:\r\n",
        "        \r\n",
        "        # How many pad tokens do we need to add?\r\n",
        "        num_pads = max_size - len(sen)\r\n",
        "\r\n",
        "        # Add `num_pads` padding tokens to the end of the sequence.\r\n",
        "        padded_input = sen + [tokenizer.pad_token_id]*num_pads\r\n",
        "\r\n",
        "        # Define the attention mask--it's just a `1` for every real token\r\n",
        "        # and a `0` for every padding token.\r\n",
        "        attn_mask = [1] * len(sen) + [0] * num_pads\r\n",
        "\r\n",
        "        # Add the padded results to the batch.\r\n",
        "        batch_padded_inputs.append(padded_input)\r\n",
        "        batch_attn_masks.append(attn_mask)\r\n",
        "\r\n",
        "    # Our batch has been padded, so we need to save this updated batch.\r\n",
        "    # We also need the inputs to be PyTorch tensors, so we'll do that here.\r\n",
        "    py_inputs.append(torch.tensor(batch_padded_inputs))\r\n",
        "    py_attn_masks.append(torch.tensor(batch_attn_masks))\r\n",
        "    py_labels.append(torch.tensor(batch_labels))\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "roS5lLWdX-5J"
      },
      "source": [
        "### Check the number of token reductions because of smart batching"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wlh7cteVfroV",
        "outputId": "60ba11e4-d1f8-4e21-a6da-502eac97cc52"
      },
      "source": [
        "# Get the new list of lengths after sorting.\r\n",
        "\r\n",
        "padded_lengths = []\r\n",
        "\r\n",
        "# For each batch...\r\n",
        "for batch in py_inputs:\r\n",
        "    \r\n",
        "    # For each sample...\r\n",
        "    for s in batch:\r\n",
        "    \r\n",
        "        # Record its length.\r\n",
        "        padded_lengths.append(len(s))\r\n",
        "\r\n",
        "# Sum up the lengths to the get the total number of tokens after smart batching.\r\n",
        "smart_token_count = np.sum(padded_lengths)\r\n",
        "\r\n",
        "# To get the total number of tokens in the dataset using fixed padding, it's\r\n",
        "# as simple as the number of samples times our `max_len` parameter (that we\r\n",
        "# would pad everything to).\r\n",
        "fixed_token_count = len(train_sentences) * max_input_length\r\n",
        "\r\n",
        "# Calculate the percentage reduction.\r\n",
        "prcnt_reduced = (fixed_token_count - smart_token_count) / float(fixed_token_count) \r\n",
        "\r\n",
        "print('Total tokens:')\r\n",
        "print('   Fixed Padding: {:,}'.format(fixed_token_count))\r\n",
        "print('  Smart Batching: {:,}  ({:.1%} less)'.format(smart_token_count, prcnt_reduced))\r\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total tokens:\n",
            "   Fixed Padding: 6,464,000\n",
            "  Smart Batching: 316,464  (95.1% less)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8wgdhsI-YIAL"
      },
      "source": [
        "## Load the model configuration from the transformers library using AutoConfig"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9er7EHvdfudf",
        "outputId": "6c96bf98-6cfc-4708-cd58-5199eb20ede3"
      },
      "source": [
        "# Load the Config object, with an output configured for classification.\r\n",
        "config = AutoConfig.from_pretrained(pretrained_model_name_or_path=model_name,\r\n",
        "                                    num_labels=3)\r\n",
        "\r\n",
        "print('Config type:', str(type(config)), '\\n')\r\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Config type: <class 'transformers.models.xlm_roberta.configuration_xlm_roberta.XLMRobertaConfig'> \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170,
          "referenced_widgets": [
            "f1d3cd21ca724e99a0667c2f03e09685",
            "5731efdc9d494f56b6eb08ba163c948b",
            "abc0f7c718904b91b132879e49a9cce1",
            "2986b0ec98e14f158817d176fc17a686",
            "9a73bb59951843f5bf45ab7957e7b034",
            "693062c4586e4f83b40a0302151b1f51",
            "6e695cb22742456b93b84761b220cef8",
            "2b833a43029e485bbbf5db292b8a4e2d"
          ]
        },
        "id": "bWvc9Ikv4HtD",
        "outputId": "99889366-4818-4f85-b1b2-70aca8f372fb"
      },
      "source": [
        "model = AutoModelForSequenceClassification.from_pretrained(pretrained_model_name_or_path=model_name,config = config)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f1d3cd21ca724e99a0667c2f03e09685",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=1115590446.0, style=ProgressStyle(descr…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Some weights of the model checkpoint at xlm-roberta-base were not used when initializing XLMRobertaForSequenceClassification: ['lm_head.bias', 'lm_head.dense.weight', 'lm_head.dense.bias', 'lm_head.layer_norm.weight', 'lm_head.layer_norm.bias', 'lm_head.decoder.weight', 'roberta.pooler.dense.weight', 'roberta.pooler.dense.bias']\n",
            "- This IS expected if you are initializing XLMRobertaForSequenceClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
            "- This IS NOT expected if you are initializing XLMRobertaForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
            "Some weights of XLMRobertaForSequenceClassification were not initialized from the model checkpoint at xlm-roberta-base and are newly initialized: ['classifier.dense.weight', 'classifier.dense.bias', 'classifier.out_proj.weight', 'classifier.out_proj.bias']\n",
            "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xouf5M2B55ei",
        "outputId": "cd85a7ba-54b7-4a7d-89bc-d67b35395f8f"
      },
      "source": [
        "model.cuda()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "XLMRobertaForSequenceClassification(\n",
              "  (roberta): RobertaModel(\n",
              "    (embeddings): RobertaEmbeddings(\n",
              "      (word_embeddings): Embedding(250002, 768, padding_idx=1)\n",
              "      (position_embeddings): Embedding(514, 768, padding_idx=1)\n",
              "      (token_type_embeddings): Embedding(1, 768)\n",
              "      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      (dropout): Dropout(p=0.1, inplace=False)\n",
              "    )\n",
              "    (encoder): RobertaEncoder(\n",
              "      (layer): ModuleList(\n",
              "        (0): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (1): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (2): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (3): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (4): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (5): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (6): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (7): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (8): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (9): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (10): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (11): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "  )\n",
              "  (classifier): RobertaClassificationHead(\n",
              "    (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "    (dropout): Dropout(p=0.1, inplace=False)\n",
              "    (out_proj): Linear(in_features=768, out_features=3, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SXJHBVlxYR-l"
      },
      "source": [
        "## Load the model from the transformers library using AutoModel"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n8J4r6KyhveR"
      },
      "source": [
        "# # Load the pre-trained model for classification, passing in the `config` from\r\n",
        "# # above.\r\n",
        "# bert = AutoModel.from_pretrained(\r\n",
        "#     pretrained_model_name_or_path=model_name,\r\n",
        "#     config=config)\r\n",
        "\r\n",
        "# print('\\nModel type:', str(type(bert)))\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RcFMgf5EuPAF"
      },
      "source": [
        "# # freeze all the parameters\r\n",
        "# for param in bert.parameters():\r\n",
        "#     param.requires_grad = False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NS_-0rBDYeY5"
      },
      "source": [
        "## Creating a custom BERT model "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bmGtZwykJ39i"
      },
      "source": [
        "\r\n",
        "# class BERT_Model(nn.Module):\r\n",
        "#     ## defining the constructor for the class\r\n",
        "#     def __init__(self, bert,num_labels):\r\n",
        "#       ## calling the super class constructor\r\n",
        "#       super(BERT_Model, self).__init__()\r\n",
        "#       ## adding the bert layer to the model\r\n",
        "#       self.bert = bert        \r\n",
        "#       # relu activation function\r\n",
        "#       self.relu =  nn.ReLU()\r\n",
        "#       # adding a dense layer to our custom model\r\n",
        "#       self.fc1 = nn.Linear(768,512)\r\n",
        "#       # adding another dense layer to our custom model ,i.e., the Output layer\r\n",
        "#       self.fc2 = nn.Linear(512,num_labels)\r\n",
        "#       # adding a softmax  activation function for our custom model's output layer\r\n",
        "#       self.softmax = nn.LogSoftmax(dim=1)\r\n",
        "\r\n",
        "#     #defining the forward pass\r\n",
        "#     def forward(self, input_id, mask):\r\n",
        "#       #pass the inputs to the model  \r\n",
        "#       outputs = self.bert(input_id,mask)\r\n",
        "  \r\n",
        "#       last_hidden_state = outputs.last_hidden_state       ## last hidden state from the model\r\n",
        "#       pooler_output = outputs.pooler_output               ## pooler output from the model\r\n",
        "#       ## adding a fully connected layer to the BERT model\r\n",
        "#       x = self.fc1(pooler_output)\r\n",
        "#       ## applying relu activation function\r\n",
        "#       x = self.relu(x)\r\n",
        "#       # the final output layer\r\n",
        "#       x = self.fc2(x)\r\n",
        "#       # apply softmax activation to our output layer\r\n",
        "#       x = self.softmax(x)\r\n",
        "\r\n",
        "#       return x"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jvOs0G5sh02S"
      },
      "source": [
        "\r\n",
        "\r\n",
        "# print('\\nLoading model ...')\r\n",
        "# # pass the pre-trained BERT to our define architecture\r\n",
        "# model = BERT_Model(bert,num_labels=3)\r\n",
        "\r\n",
        "# model.cuda()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LjQSUqCWJTbI"
      },
      "source": [
        "## Custom Loss function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iOHVl1RRJSUB"
      },
      "source": [
        "# # convert class weights to tensor\r\n",
        "# weights= torch.tensor(class_wts,dtype=torch.float)\r\n",
        "# weights = weights.to(device)\r\n",
        "\r\n",
        "# # loss function\r\n",
        "# cross_entropy  = nn.NLLLoss(weight=weights) \r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "evYV0yrQ2u_b"
      },
      "source": [
        "## Loading Optimizer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8tAijXBEh7t8"
      },
      "source": [
        "# Note: AdamW is a class from the huggingface library (as opposed to pytorch) \r\n",
        "# I believe the 'W' stands for 'Weight Decay fix\"\r\n",
        "optimizer = AdamW(model.parameters(),\r\n",
        "                  lr = 2e-5, # This is the value Michael used.\r\n",
        "                  eps = 1e-8 # args.adam_epsilon  - default is 1e-8.\r\n",
        "                )\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s9n3_5jI2ywo"
      },
      "source": [
        "## Loading lr scheduler"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "agidP6D6iBFI"
      },
      "source": [
        "# Number of training epochs. I chose to train for 1 simply because the training\r\n",
        "# time is long. More epochs may improve the model's accuracy.\r\n",
        "epochs = 4\r\n",
        "\r\n",
        "# Total number of training steps is [number of batches] x [number of epochs]. \r\n",
        "# Note that it's the number of *batches*, not *samples*!\r\n",
        "total_steps = len(py_inputs) * epochs\r\n",
        "\r\n",
        "# Create the learning rate scheduler.\r\n",
        "scheduler = get_linear_schedule_with_warmup(optimizer, \r\n",
        "                                            num_warmup_steps = 0, # Default value in run_glue.py\r\n",
        "                                            num_training_steps = total_steps)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MDDp05Cm245e"
      },
      "source": [
        "## Training Loop"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "H4ERR8fBiEJn",
        "outputId": "9783c78f-520e-420e-e2d7-3bed269c1643"
      },
      "source": [
        "# We'll store a number of quantities such as training and validation loss, \r\n",
        "# validation accuracy, and timings.\r\n",
        "training_stats = {\r\n",
        "    'epoch':[],\r\n",
        "    'train_loss':[],\r\n",
        "    'Training Time':[],\r\n",
        "    'val_loss':[],\r\n",
        "    'Validation Time':[],\r\n",
        "    'train_acc':[],\r\n",
        "    'val_acc':[]\r\n",
        "}\r\n",
        "\r\n",
        "# Update every `update_interval` batches.\r\n",
        "update_interval = good_update_interval(total_iters=len(py_inputs), num_desired_updates=10)\r\n",
        "\r\n",
        "# Measure the total training time for the whole run.\r\n",
        "total_t0 = time.time()\r\n",
        "best_accuracy = 0\r\n",
        "# For each epoch...\r\n",
        "for epoch_i in range(0, epochs):\r\n",
        "\r\n",
        "    predictions = []\r\n",
        "    true_labels = []\r\n",
        "    \r\n",
        "    # ========================================\r\n",
        "    #               Training\r\n",
        "    # ========================================\r\n",
        "    \r\n",
        "    # Perform one full pass over the training set.\r\n",
        "\r\n",
        "    print(\"\")\r\n",
        "    print('======== Epoch {:} / {:} ========'.format(epoch_i + 1, epochs))\r\n",
        "    \r\n",
        "    # At the start of each epoch (except for the first) we need to re-randomize\r\n",
        "    # our training data.\r\n",
        "    if epoch_i > 0:\r\n",
        "        # Use our `make_smart_batches` function (from 6.1.) to re-shuffle the \r\n",
        "        # dataset into new batches.\r\n",
        "        (py_inputs, py_attn_masks, py_labels) = make_smart_batches(train_sentences, train_labels, batch_size,tokenizer,max_input_length)\r\n",
        "    \r\n",
        "    print('Training on {:,} batches...'.format(len(py_inputs)))\r\n",
        "\r\n",
        "    # Measure how long the training epoch takes.\r\n",
        "    t0 = time.time()\r\n",
        "\r\n",
        "    # Reset the total loss for this epoch.\r\n",
        "    total_train_loss = 0\r\n",
        "\r\n",
        "    # Put the model into training mode. Don't be mislead--the call to \r\n",
        "    # `train` just changes the *mode*, it doesn't *perform* the training.\r\n",
        "    # `dropout` and `batchnorm` layers behave differently during training\r\n",
        "    # vs. test (source: https://stackoverflow.com/questions/51433378/what-does-model-train-do-in-pytorch)\r\n",
        "    model.train()\r\n",
        "\r\n",
        "    # For each batch of training data...\r\n",
        "    for step in range(0, len(py_inputs)):\r\n",
        "\r\n",
        "        # Progress update every, e.g., 100 batches.\r\n",
        "        if step % update_interval == 0 and not step == 0:\r\n",
        "            # Calculate elapsed time in minutes.\r\n",
        "            elapsed = format_time(time.time() - t0)\r\n",
        "            \r\n",
        "            # Calculate the time remaining based on our progress.\r\n",
        "            steps_per_sec = (time.time() - t0) / step\r\n",
        "            remaining_sec = steps_per_sec * (len(py_inputs) - step)\r\n",
        "            remaining = format_time(remaining_sec)\r\n",
        "\r\n",
        "            # Report progress.\r\n",
        "            print('  Batch {:>7,}  of  {:>7,}.    Elapsed: {:}.  Remaining: {:}'.format(step, len(py_inputs), elapsed, remaining))\r\n",
        "\r\n",
        "        # Copy the current training batch to the GPU using the `to` method.\r\n",
        "        b_input_ids = py_inputs[step].to(device)\r\n",
        "        b_input_mask = py_attn_masks[step].to(device)\r\n",
        "        b_labels = py_labels[step].to(device)\r\n",
        "\r\n",
        "        # Always clear any previously calculated gradients before performing a\r\n",
        "        # backward pass.\r\n",
        "        model.zero_grad()        \r\n",
        "\r\n",
        "        # Perform a forward pass (evaluate the model on this training batch).\r\n",
        "        # The call returns the loss (because we provided labels) and the \r\n",
        "        # \"logits\"--the model outputs prior to activation.\r\n",
        "        output = model(b_input_ids, \r\n",
        "                             token_type_ids=None, \r\n",
        "                             attention_mask=b_input_mask, \r\n",
        "                             labels=b_labels)\r\n",
        "\r\n",
        "        # Accumulate the training loss over all of the batches so that we can\r\n",
        "        # calculate the average loss at the end. `loss` is a Tensor containing a\r\n",
        "        # single value; the `.item()` function just returns the Python value \r\n",
        "        # from the tensor.\r\n",
        "        loss = output.loss \r\n",
        "        logits = output.logits\r\n",
        "\r\n",
        "        # Move logits and labels to CPU\r\n",
        "        logits = logits.detach().cpu().numpy()\r\n",
        "        label_ids = b_labels.to('cpu').numpy()\r\n",
        "      \r\n",
        "        # Store predictions and true labels\r\n",
        "        predictions.append(logits)\r\n",
        "        true_labels.append(label_ids)\r\n",
        "        \r\n",
        "        # Accumulate the training loss over all of the batches so that we can\r\n",
        "        # calculate the average loss at the end. `loss` is a Tensor containing a\r\n",
        "        # single value; the `.item()` function just returns the Python value \r\n",
        "        # from the tensor.\r\n",
        "        total_train_loss += loss.item()\r\n",
        "\r\n",
        "        # Perform a backward pass to calculate the gradients.\r\n",
        "        loss.backward()\r\n",
        "\r\n",
        "        # Clip the norm of the gradients to 1.0.\r\n",
        "        # This is to help prevent the \"exploding gradients\" problem.\r\n",
        "        torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0)\r\n",
        "\r\n",
        "        # Update parameters and take a step using the computed gradient.\r\n",
        "        # The optimizer dictates the \"update rule\"--how the parameters are\r\n",
        "        # modified based on their gradients, the learning rate, etc.\r\n",
        "        optimizer.step()\r\n",
        "\r\n",
        "        # Update the learning rate.\r\n",
        "        scheduler.step()\r\n",
        "\r\n",
        "    # Calculate the average loss over all of the batches.\r\n",
        "    avg_train_loss = total_train_loss / len(py_inputs)     \r\n",
        "       \r\n",
        "    training_accuracy = check_accuracy(predictions,true_labels)       \r\n",
        "    \r\n",
        "    # Measure how long this epoch took.\r\n",
        "    training_time = format_time(time.time() - t0)\r\n",
        "\r\n",
        "    print(\"\")\r\n",
        "    print(\"  Average training loss: {0:.2f}\".format(avg_train_loss))\r\n",
        "    print(\"  Training Accuracy: {0:.2f}\".format(training_accuracy))\r\n",
        "    print(\"  Training epoch took: {:}\".format(training_time))\r\n",
        "        \r\n",
        "    (py_inputs, py_attn_masks, py_labels) = make_smart_batches(val_sentences, val_labels, batch_size ,tokenizer,max_input_length)\r\n",
        "    val_loss,val_accuracy,validation_time = eval_model(model,py_inputs, py_attn_masks, py_labels)  \r\n",
        "\r\n",
        "    if val_accuracy > best_accuracy:\r\n",
        "        torch.save(model.state_dict(), 'best_model_state.bin')\r\n",
        "        best_accuracy = val_accuracy\r\n",
        "    # Record all statistics from this epoch.\r\n",
        "\r\n",
        "    print(\"\")\r\n",
        "    print(\"  Average validation loss: {0:.2f}\".format(val_loss))\r\n",
        "    print(\"  Validation Accuracy: {0:.2f}\".format(val_accuracy))\r\n",
        "    print(\"  Validation epoch took: {:}\".format(validation_time))\r\n",
        "    \r\n",
        "    \r\n",
        "    training_stats['epoch'].append(epoch_i + 1)\r\n",
        "    training_stats['train_loss'].append(avg_train_loss)\r\n",
        "    training_stats['Training Time'].append(training_time)\r\n",
        "    training_stats['val_loss'].append(val_loss)\r\n",
        "    training_stats['Validation Time'].append(validation_time)\r\n",
        "    training_stats['train_acc'].append(training_accuracy)\r\n",
        "    training_stats['val_acc'].append(val_accuracy)\r\n",
        "\r\n",
        "print(f'Best val accuracy: {best_accuracy}')\r\n",
        "  \r\n",
        "model.load_state_dict(torch.load('best_model_state.bin'))\r\n",
        "\r\n",
        "print(\"\")\r\n",
        "print(\"Training complete!\")\r\n",
        "\r\n",
        "print(\"Total training took {:} (h:mm:ss)\".format(format_time(time.time()-total_t0)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "======== Epoch 1 / 4 ========\n",
            "Training on 1,010 batches...\n",
            "  Batch     100  of    1,010.    Elapsed: 0:00:10.  Remaining: 0:01:32\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:20.  Remaining: 0:01:22\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:30.  Remaining: 0:01:12\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:41.  Remaining: 0:01:02\n",
            "  Batch     500  of    1,010.    Elapsed: 0:00:51.  Remaining: 0:00:52\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:02.  Remaining: 0:00:42\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:12.  Remaining: 0:00:32\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:23.  Remaining: 0:00:22\n",
            "  Batch     900  of    1,010.    Elapsed: 0:01:33.  Remaining: 0:00:11\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:01:45.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.84\n",
            "  Training Accuracy: 0.53\n",
            "  Training epoch took: 0:01:46\n",
            "Creating Smart Batches from 2,018 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,018 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,018 samples\n",
            "\n",
            "     2,018 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Doing validation on 127 sentences...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      30  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:01\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      60  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      70  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      80  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     110  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.77\n",
            "  Validation Accuracy: 0.59\n",
            "  Validation epoch took: 0:00:02\n",
            "\n",
            "======== Epoch 2 / 4 ========\n",
            "Creating Smart Batches from 16,160 examples with batch size 16...\n",
            "\n",
            "Tokenizing 16,160 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "  Tokenized 4,000 samples.\n",
            "  Tokenized 6,000 samples.\n",
            "  Tokenized 8,000 samples.\n",
            "  Tokenized 10,000 samples.\n",
            "  Tokenized 12,000 samples.\n",
            "  Tokenized 14,000 samples.\n",
            "  Tokenized 16,000 samples.\n",
            "DONE.\n",
            "    16,160 samples\n",
            "\n",
            "    16,160 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 1,010 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Training on 1,010 batches...\n",
            "  Batch     100  of    1,010.    Elapsed: 0:00:11.  Remaining: 0:01:36\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:21.  Remaining: 0:01:25\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:31.  Remaining: 0:01:14\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:42.  Remaining: 0:01:04\n",
            "  Batch     500  of    1,010.    Elapsed: 0:00:52.  Remaining: 0:00:53\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:02.  Remaining: 0:00:42\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:12.  Remaining: 0:00:32\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:23.  Remaining: 0:00:22\n",
            "  Batch     900  of    1,010.    Elapsed: 0:01:33.  Remaining: 0:00:11\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:01:43.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.75\n",
            "  Training Accuracy: 0.60\n",
            "  Training epoch took: 0:01:45\n",
            "Creating Smart Batches from 2,018 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,018 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,018 samples\n",
            "\n",
            "     2,018 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Doing validation on 127 sentences...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      60  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      70  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      80  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     110  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.74\n",
            "  Validation Accuracy: 0.62\n",
            "  Validation epoch took: 0:00:02\n",
            "\n",
            "======== Epoch 3 / 4 ========\n",
            "Creating Smart Batches from 16,160 examples with batch size 16...\n",
            "\n",
            "Tokenizing 16,160 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "  Tokenized 4,000 samples.\n",
            "  Tokenized 6,000 samples.\n",
            "  Tokenized 8,000 samples.\n",
            "  Tokenized 10,000 samples.\n",
            "  Tokenized 12,000 samples.\n",
            "  Tokenized 14,000 samples.\n",
            "  Tokenized 16,000 samples.\n",
            "DONE.\n",
            "    16,160 samples\n",
            "\n",
            "    16,160 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 1,010 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Training on 1,010 batches...\n",
            "  Batch     100  of    1,010.    Elapsed: 0:00:10.  Remaining: 0:01:34\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:20.  Remaining: 0:01:22\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:31.  Remaining: 0:01:13\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:41.  Remaining: 0:01:03\n",
            "  Batch     500  of    1,010.    Elapsed: 0:00:51.  Remaining: 0:00:52\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:01.  Remaining: 0:00:42\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:11.  Remaining: 0:00:32\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:21.  Remaining: 0:00:21\n",
            "  Batch     900  of    1,010.    Elapsed: 0:01:31.  Remaining: 0:00:11\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:01:42.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.71\n",
            "  Training Accuracy: 0.64\n",
            "  Training epoch took: 0:01:44\n",
            "Creating Smart Batches from 2,018 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,018 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,018 samples\n",
            "\n",
            "     2,018 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Doing validation on 127 sentences...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      60  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      70  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      80  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     110  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.75\n",
            "  Validation Accuracy: 0.62\n",
            "  Validation epoch took: 0:00:02\n",
            "\n",
            "======== Epoch 4 / 4 ========\n",
            "Creating Smart Batches from 16,160 examples with batch size 16...\n",
            "\n",
            "Tokenizing 16,160 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "  Tokenized 4,000 samples.\n",
            "  Tokenized 6,000 samples.\n",
            "  Tokenized 8,000 samples.\n",
            "  Tokenized 10,000 samples.\n",
            "  Tokenized 12,000 samples.\n",
            "  Tokenized 14,000 samples.\n",
            "  Tokenized 16,000 samples.\n",
            "DONE.\n",
            "    16,160 samples\n",
            "\n",
            "    16,160 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 1,010 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Training on 1,010 batches...\n",
            "  Batch     100  of    1,010.    Elapsed: 0:00:10.  Remaining: 0:01:29\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:21.  Remaining: 0:01:23\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:31.  Remaining: 0:01:13\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:41.  Remaining: 0:01:03\n",
            "  Batch     500  of    1,010.    Elapsed: 0:00:51.  Remaining: 0:00:52\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:02.  Remaining: 0:00:42\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:11.  Remaining: 0:00:32\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:22.  Remaining: 0:00:22\n",
            "  Batch     900  of    1,010.    Elapsed: 0:01:32.  Remaining: 0:00:11\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:01:42.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.67\n",
            "  Training Accuracy: 0.67\n",
            "  Training epoch took: 0:01:45\n",
            "Creating Smart Batches from 2,018 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,018 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,018 samples\n",
            "\n",
            "     2,018 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Doing validation on 127 sentences...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      30  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:01\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      60  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      70  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      80  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     110  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.74\n",
            "  Validation Accuracy: 0.62\n",
            "  Validation epoch took: 0:00:02\n",
            "Best val accuracy: 0.6238850346878098\n",
            "\n",
            "Training complete!\n",
            "Total training took 0:07:35 (h:mm:ss)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 450
        },
        "id": "GOmua_S1eXNM",
        "outputId": "b4a2fb76-2af7-4929-d126-fe340537bff5"
      },
      "source": [
        "plot_training_history(training_stats)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1296x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3c-6bOIxVyrM"
      },
      "source": [
        "## Evaluating Performance Over Training Set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2QmKPNAeWHkZ",
        "outputId": "bfd34723-066c-4652-d863-7137fb112ef1"
      },
      "source": [
        "(py_inputs, py_attn_masks, py_labels) = make_smart_batches(train_sentences, train_labels, batch_size ,tokenizer,max_input_length)\r\n",
        "\r\n",
        "y_pred , y_true = get_predictions(py_inputs, py_attn_masks, py_labels)\r\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Creating Smart Batches from 16,160 examples with batch size 16...\n",
            "\n",
            "Tokenizing 16,160 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "  Tokenized 4,000 samples.\n",
            "  Tokenized 6,000 samples.\n",
            "  Tokenized 8,000 samples.\n",
            "  Tokenized 10,000 samples.\n",
            "  Tokenized 12,000 samples.\n",
            "  Tokenized 14,000 samples.\n",
            "  Tokenized 16,000 samples.\n",
            "DONE.\n",
            "    16,160 samples\n",
            "\n",
            "    16,160 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 1,010 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Predicting labels for 1,010 test batches...\n",
            "  Batch     100  of    1,010.    Elapsed: 0:00:02.  Remaining: 0:00:14\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:03.  Remaining: 0:00:13\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:04.  Remaining: 0:00:10\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:06.  Remaining: 0:00:09\n",
            "  Batch     500  of    1,010.    Elapsed: 0:00:07.  Remaining: 0:00:07\n",
            "  Batch     600  of    1,010.    Elapsed: 0:00:09.  Remaining: 0:00:06\n",
            "  Batch     700  of    1,010.    Elapsed: 0:00:10.  Remaining: 0:00:05\n",
            "  Batch     800  of    1,010.    Elapsed: 0:00:12.  Remaining: 0:00:03\n",
            "  Batch     900  of    1,010.    Elapsed: 0:00:13.  Remaining: 0:00:02\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:00:15.  Remaining: 0:00:00\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uXsVn0wngZhv",
        "outputId": "761b24ca-38a2-4205-f4e2-545d1ff41b4e"
      },
      "source": [
        "print(classification_report(y_true, y_pred, target_names=class_names))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                 precision    recall  f1-score   support\n",
            "\n",
            "    Hope_speech       0.73      0.56      0.64      6327\n",
            "Non_hope_speech       0.72      0.78      0.75      7872\n",
            "      not-Tamil       0.63      0.87      0.73      1961\n",
            "\n",
            "       accuracy                           0.71     16160\n",
            "      macro avg       0.69      0.74      0.71     16160\n",
            "   weighted avg       0.71      0.71      0.70     16160\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "guClUdRhYPql",
        "outputId": "4f84f01e-6b6b-45bd-ad44-952aa089050f"
      },
      "source": [
        "cm = confusion_matrix(y_true, y_pred)\r\n",
        "show_confusion_matrix(cm, class_names)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VM2N-X1tV3il"
      },
      "source": [
        "## Evaluating Performance Over Validation Set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BnadNvtxWIYw",
        "outputId": "0d2d25fb-0793-4215-8cac-e26a9e917d0a"
      },
      "source": [
        "(py_inputs, py_attn_masks, py_labels) = make_smart_batches(val_sentences, val_labels, batch_size ,tokenizer,max_input_length)\r\n",
        "\r\n",
        "y_pred , y_true = get_predictions(py_inputs, py_attn_masks, py_labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Creating Smart Batches from 2,018 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,018 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,018 samples\n",
            "\n",
            "     2,018 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Predicting labels for 127 test batches...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      30  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:01\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      60  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      70  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      80  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:00\n",
            "  Batch     110  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wAjlATBPgh7c",
        "outputId": "b6b067e7-f82b-426e-ba1a-b7e079d35817"
      },
      "source": [
        "print(classification_report(y_true, y_pred, target_names=class_names))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                 precision    recall  f1-score   support\n",
            "\n",
            "    Hope_speech       0.59      0.49      0.53       757\n",
            "Non_hope_speech       0.66      0.69      0.68       998\n",
            "      not-Tamil       0.58      0.76      0.65       263\n",
            "\n",
            "       accuracy                           0.62      2018\n",
            "      macro avg       0.61      0.65      0.62      2018\n",
            "   weighted avg       0.62      0.62      0.62      2018\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "7nELY57NYshD",
        "outputId": "2b4fa513-f092-4ac9-e62f-0284d42bfa0e"
      },
      "source": [
        "cm = confusion_matrix(y_true, y_pred)\r\n",
        "show_confusion_matrix(cm, class_names)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j_Ev8kTSNT6L"
      },
      "source": [
        "## Evaluating Performance Over Test Set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "knzcVqvONT6N",
        "outputId": "aa078c9c-1e2a-4e52-f2c5-ada01f14df0d"
      },
      "source": [
        "(py_inputs, py_attn_masks, py_labels) = make_smart_batches(test_sentences, test_labels, batch_size ,tokenizer,max_input_length)\r\n",
        "\r\n",
        "y_pred , y_true = get_predictions(py_inputs, py_attn_masks, py_labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Creating Smart Batches from 2,020 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,020 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,020 samples\n",
            "\n",
            "     2,020 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n",
            "Predicting labels for 127 test batches...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      30  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:02\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      60  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      70  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      80  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     110  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:00\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "l6EJfj0TNT6R",
        "outputId": "dd096b48-3b96-4512-d7a6-a9bc5f7b2655"
      },
      "source": [
        "print(classification_report(y_true, y_pred, target_names=class_names))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                 precision    recall  f1-score   support\n",
            "\n",
            "    Hope_speech       0.59      0.44      0.50       815\n",
            "Non_hope_speech       0.62      0.69      0.65       946\n",
            "      not-Tamil       0.56      0.78      0.65       259\n",
            "\n",
            "       accuracy                           0.60      2020\n",
            "      macro avg       0.59      0.64      0.60      2020\n",
            "   weighted avg       0.60      0.60      0.59      2020\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "ZE3EG9RTNT6S",
        "outputId": "9fba367a-9605-49f1-ddd3-3fbab9ae3c4b"
      },
      "source": [
        "cm = confusion_matrix(y_true, y_pred)\r\n",
        "show_confusion_matrix(cm, class_names)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I37UTLXjFM95"
      },
      "source": [
        "## Saving the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k4mv0TG2FM95"
      },
      "source": [
        "torch.save(model,\"./drive/MyDrive/Codalab/HopeSpeech Detection For Equality,DiversityAnd Inclusion/Models/XLMR_Tamil\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a9M_91z1yj0F"
      },
      "source": [
        "## Loading the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "42ObCUU1QcsJ"
      },
      "source": [
        "model = torch.load('./drive/MyDrive/HopeSpeech Detection For Equality,DiversityAnd Inclusion/Models/XLMRobertaForSequenceClassification_Mutlti_Tamil_v2',map_location=device)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sAaHpZdIzsnT",
        "outputId": "da55e10f-457b-4488-f3f3-97cf5c757083"
      },
      "source": [
        "model.cuda()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "XLMRobertaForSequenceClassification(\n",
              "  (roberta): RobertaModel(\n",
              "    (embeddings): RobertaEmbeddings(\n",
              "      (word_embeddings): Embedding(250002, 768, padding_idx=1)\n",
              "      (position_embeddings): Embedding(514, 768, padding_idx=1)\n",
              "      (token_type_embeddings): Embedding(1, 768)\n",
              "      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      (dropout): Dropout(p=0.1, inplace=False)\n",
              "    )\n",
              "    (encoder): RobertaEncoder(\n",
              "      (layer): ModuleList(\n",
              "        (0): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (1): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (2): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (3): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (4): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (5): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (6): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (7): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (8): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (9): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (10): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (11): RobertaLayer(\n",
              "          (attention): RobertaAttention(\n",
              "            (self): RobertaSelfAttention(\n",
              "              (query): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (key): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (value): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "            (output): RobertaSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): RobertaIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): RobertaOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "  )\n",
              "  (classifier): RobertaClassificationHead(\n",
              "    (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "    (dropout): Dropout(p=0.1, inplace=False)\n",
              "    (out_proj): Linear(in_features=768, out_features=3, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hY2dTIjnFM98"
      },
      "source": [
        "## Making Predictions on Test Set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XuUBBoiZFM98"
      },
      "source": [
        "test_df_path = join(dataset_dir,'tamil_hope_test.csv')\r\n",
        "test_df = pd.read_csv(test_df_path,sep = '\\t')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "bFOWRE24FM98",
        "outputId": "07a99dd8-e4c9-47f1-a7c3-2a81de5cfd04"
      },
      "source": [
        "test_df"
      ],
      "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>text</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Konjam methuva pesuninganna nalarukum bro...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Sema bro............. தமிழன் என்று சொல்லடா தலை...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>9.52 mg bro  plz crrt pannunga bro</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Superb anna</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Central Government kitta irunthu neraiya fund ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015</th>\n",
              "      <td>Solunga bro parvala</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2016</th>\n",
              "      <td>Madangowri anna 9.53s  time note pannunga Indi...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2017</th>\n",
              "      <td>Enna nadu vanthu vaccine kudaka matadragalo an...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2018</th>\n",
              "      <td>Yes..ipdii thaan avangaluku bathil kudukanu......</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019</th>\n",
              "      <td>No views .. but 390 likes 🤣... MG squad</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2020 rows × 1 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                   text\n",
              "0          Konjam methuva pesuninganna nalarukum bro...\n",
              "1     Sema bro............. தமிழன் என்று சொல்லடா தலை...\n",
              "2                    9.52 mg bro  plz crrt pannunga bro\n",
              "3                                           Superb anna\n",
              "4     Central Government kitta irunthu neraiya fund ...\n",
              "...                                                 ...\n",
              "2015                                Solunga bro parvala\n",
              "2016  Madangowri anna 9.53s  time note pannunga Indi...\n",
              "2017  Enna nadu vanthu vaccine kudaka matadragalo an...\n",
              "2018  Yes..ipdii thaan avangaluku bathil kudukanu......\n",
              "2019            No views .. but 390 likes 🤣... MG squad\n",
              "\n",
              "[2020 rows x 1 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DbI53tABFM99"
      },
      "source": [
        "## Loading `test_sentences`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3uoBX0l7FM99"
      },
      "source": [
        "test_sentences = test_df['text'].values \r\n",
        "test_ids = test_df.index.values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7HnVIFvZFM99",
        "outputId": "3f032f43-44a3-4e36-c1c6-e6d0475ca439"
      },
      "source": [
        "(py_inputs, py_attn_masks,py_ids) = make_smart_batches_on_test(test_sentences,test_ids, 16,tokenizer,max_input_length = 400)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Creating Smart Batches from 2,020 examples with batch size 16...\n",
            "\n",
            "Tokenizing 2,020 samples...\n",
            "  Tokenized 0 samples.\n",
            "  Tokenized 200 samples.\n",
            "  Tokenized 400 samples.\n",
            "  Tokenized 600 samples.\n",
            "  Tokenized 800 samples.\n",
            "  Tokenized 1,000 samples.\n",
            "  Tokenized 1,200 samples.\n",
            "  Tokenized 1,400 samples.\n",
            "  Tokenized 1,600 samples.\n",
            "  Tokenized 1,800 samples.\n",
            "  Tokenized 2,000 samples.\n",
            "DONE.\n",
            "     2,020 samples\n",
            "\n",
            "     2,020 samples after sorting\n",
            "\n",
            "Creating batches of size 16...\n",
            "\n",
            "  DONE - Selected 127 batches.\n",
            "\n",
            "Padding out sequences within each batch...\n",
            "  DONE.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vjsjqQpRQcsV"
      },
      "source": [
        "## Evaluating accuracy over test set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rwkx-30AQcsV",
        "outputId": "51780d7b-e2da-4530-f624-4588b45c3ab4"
      },
      "source": [
        "y_ids,y_preds  = get_predictions_test(py_inputs, py_attn_masks,py_ids)\r\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Predicting labels for 127 test batches...\n",
            "  Batch      10  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:05\n",
            "  Batch      20  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:03\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:03\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      60  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:02\n",
            "  Batch      70  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:02\n",
            "  Batch      80  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch     110  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:00\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pDM-cnDHQcsW",
        "outputId": "b04e6ccd-3747-48a1-8113-0db9b445e7ea"
      },
      "source": [
        "print(y_preds)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0 1 1 ... 0 1 1]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kWWWjz7eUPc7",
        "outputId": "55c5a3bc-c7e2-44b0-f9f9-836ea0cc7ec2"
      },
      "source": [
        "print(y_ids)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[ 138 1001 1148 ... 1882 1124 1486]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u2ybbsAoQcsY",
        "outputId": "968fbf55-2948-44d1-814c-afbacf3980c6"
      },
      "source": [
        "le.inverse_transform(y_preds)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Hope_speech', 'Non_hope_speech', 'Non_hope_speech', ...,\n",
              "       'Hope_speech', 'Non_hope_speech', 'Non_hope_speech'], dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 62
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "kSMb6sCLQcsZ",
        "outputId": "14d70af9-51a3-49c0-b33a-32a8aced8755"
      },
      "source": [
        "sns.countplot(y =le.inverse_transform(y_preds))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f18123bb128>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 63
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4KXOXwH9U2MT",
        "outputId": "63a919bc-84b9-42b8-88d7-d10a6f26fb27"
      },
      "source": [
        "len(y_ids)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2020"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9-fkBgOxU4bX",
        "outputId": "74ef1fc2-b12f-4a41-d55d-c94144725119"
      },
      "source": [
        "len(le.inverse_transform(y_preds))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2020"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sWPDqogJrjIZ"
      },
      "source": [
        "## Creating Submission DataFrame"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o4Hb7RD1yHol"
      },
      "source": [
        "df_new = pd.DataFrame({\r\n",
        "    'id':y_ids,\r\n",
        "    'label':le.inverse_transform(y_preds)\r\n",
        "})"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "twQAkVzcn4Lg"
      },
      "source": [
        "test_df['id'] = test_df.index"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "qov36eWmqawv",
        "outputId": "055797f9-5065-4226-e86d-6ddedbfb8f16"
      },
      "source": [
        "df_new"
      ],
      "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>id</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>138</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1001</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1148</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>349</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>627</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015</th>\n",
              "      <td>667</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2016</th>\n",
              "      <td>59</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2017</th>\n",
              "      <td>1882</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2018</th>\n",
              "      <td>1124</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019</th>\n",
              "      <td>1486</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2020 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        id            label\n",
              "0      138      Hope_speech\n",
              "1     1001  Non_hope_speech\n",
              "2     1148  Non_hope_speech\n",
              "3      349  Non_hope_speech\n",
              "4      627  Non_hope_speech\n",
              "...    ...              ...\n",
              "2015   667  Non_hope_speech\n",
              "2016    59      Hope_speech\n",
              "2017  1882      Hope_speech\n",
              "2018  1124  Non_hope_speech\n",
              "2019  1486  Non_hope_speech\n",
              "\n",
              "[2020 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "z0CEApHmqdLr",
        "outputId": "96f2b53d-2d6a-49d0-fe90-110d3bdc998d"
      },
      "source": [
        "test_df"
      ],
      "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>text</th>\n",
              "      <th>id</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Konjam methuva pesuninganna nalarukum bro...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Sema bro............. தமிழன் என்று சொல்லடா தலை...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>9.52 mg bro  plz crrt pannunga bro</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Superb anna</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Central Government kitta irunthu neraiya fund ...</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015</th>\n",
              "      <td>Solunga bro parvala</td>\n",
              "      <td>2015</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2016</th>\n",
              "      <td>Madangowri anna 9.53s  time note pannunga Indi...</td>\n",
              "      <td>2016</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2017</th>\n",
              "      <td>Enna nadu vanthu vaccine kudaka matadragalo an...</td>\n",
              "      <td>2017</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2018</th>\n",
              "      <td>Yes..ipdii thaan avangaluku bathil kudukanu......</td>\n",
              "      <td>2018</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019</th>\n",
              "      <td>No views .. but 390 likes 🤣... MG squad</td>\n",
              "      <td>2019</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2020 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                   text    id\n",
              "0          Konjam methuva pesuninganna nalarukum bro...     0\n",
              "1     Sema bro............. தமிழன் என்று சொல்லடா தலை...     1\n",
              "2                    9.52 mg bro  plz crrt pannunga bro     2\n",
              "3                                           Superb anna     3\n",
              "4     Central Government kitta irunthu neraiya fund ...     4\n",
              "...                                                 ...   ...\n",
              "2015                                Solunga bro parvala  2015\n",
              "2016  Madangowri anna 9.53s  time note pannunga Indi...  2016\n",
              "2017  Enna nadu vanthu vaccine kudaka matadragalo an...  2017\n",
              "2018  Yes..ipdii thaan avangaluku bathil kudukanu......  2018\n",
              "2019            No views .. but 390 likes 🤣... MG squad  2019\n",
              "\n",
              "[2020 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "vxReWWmxn7Hj",
        "outputId": "901a883f-1043-41b9-c004-1fb9a29780dd"
      },
      "source": [
        "df_f = pd.merge(test_df,df_new,on = 'id')\r\n",
        "df_f"
      ],
      "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>text</th>\n",
              "      <th>id</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Konjam methuva pesuninganna nalarukum bro...</td>\n",
              "      <td>0</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Sema bro............. தமிழன் என்று சொல்லடா தலை...</td>\n",
              "      <td>1</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>9.52 mg bro  plz crrt pannunga bro</td>\n",
              "      <td>2</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Superb anna</td>\n",
              "      <td>3</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Central Government kitta irunthu neraiya fund ...</td>\n",
              "      <td>4</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015</th>\n",
              "      <td>Solunga bro parvala</td>\n",
              "      <td>2015</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2016</th>\n",
              "      <td>Madangowri anna 9.53s  time note pannunga Indi...</td>\n",
              "      <td>2016</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2017</th>\n",
              "      <td>Enna nadu vanthu vaccine kudaka matadragalo an...</td>\n",
              "      <td>2017</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2018</th>\n",
              "      <td>Yes..ipdii thaan avangaluku bathil kudukanu......</td>\n",
              "      <td>2018</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019</th>\n",
              "      <td>No views .. but 390 likes 🤣... MG squad</td>\n",
              "      <td>2019</td>\n",
              "      <td>not-Tamil</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2020 rows × 3 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                   text    id            label\n",
              "0          Konjam methuva pesuninganna nalarukum bro...     0      Hope_speech\n",
              "1     Sema bro............. தமிழன் என்று சொல்லடா தலை...     1  Non_hope_speech\n",
              "2                    9.52 mg bro  plz crrt pannunga bro     2      Hope_speech\n",
              "3                                           Superb anna     3      Hope_speech\n",
              "4     Central Government kitta irunthu neraiya fund ...     4  Non_hope_speech\n",
              "...                                                 ...   ...              ...\n",
              "2015                                Solunga bro parvala  2015  Non_hope_speech\n",
              "2016  Madangowri anna 9.53s  time note pannunga Indi...  2016      Hope_speech\n",
              "2017  Enna nadu vanthu vaccine kudaka matadragalo an...  2017      Hope_speech\n",
              "2018  Yes..ipdii thaan avangaluku bathil kudukanu......  2018  Non_hope_speech\n",
              "2019            No views .. but 390 likes 🤣... MG squad  2019        not-Tamil\n",
              "\n",
              "[2020 rows x 3 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "1kuASymYn9aV",
        "outputId": "7203d6c8-84bf-499e-ada4-a62c43e993e0"
      },
      "source": [
        "df_f = df_f[['id','text','label']]\r\n",
        "df_f"
      ],
      "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>id</th>\n",
              "      <th>text</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>Konjam methuva pesuninganna nalarukum bro...</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>Sema bro............. தமிழன் என்று சொல்லடா தலை...</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>9.52 mg bro  plz crrt pannunga bro</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>Superb anna</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>Central Government kitta irunthu neraiya fund ...</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015</th>\n",
              "      <td>2015</td>\n",
              "      <td>Solunga bro parvala</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2016</th>\n",
              "      <td>2016</td>\n",
              "      <td>Madangowri anna 9.53s  time note pannunga Indi...</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2017</th>\n",
              "      <td>2017</td>\n",
              "      <td>Enna nadu vanthu vaccine kudaka matadragalo an...</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2018</th>\n",
              "      <td>2018</td>\n",
              "      <td>Yes..ipdii thaan avangaluku bathil kudukanu......</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019</th>\n",
              "      <td>2019</td>\n",
              "      <td>No views .. but 390 likes 🤣... MG squad</td>\n",
              "      <td>not-Tamil</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2020 rows × 3 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        id                                               text            label\n",
              "0        0       Konjam methuva pesuninganna nalarukum bro...      Hope_speech\n",
              "1        1  Sema bro............. தமிழன் என்று சொல்லடா தலை...  Non_hope_speech\n",
              "2        2                 9.52 mg bro  plz crrt pannunga bro      Hope_speech\n",
              "3        3                                        Superb anna      Hope_speech\n",
              "4        4  Central Government kitta irunthu neraiya fund ...  Non_hope_speech\n",
              "...    ...                                                ...              ...\n",
              "2015  2015                                Solunga bro parvala  Non_hope_speech\n",
              "2016  2016  Madangowri anna 9.53s  time note pannunga Indi...      Hope_speech\n",
              "2017  2017  Enna nadu vanthu vaccine kudaka matadragalo an...      Hope_speech\n",
              "2018  2018  Yes..ipdii thaan avangaluku bathil kudukanu......  Non_hope_speech\n",
              "2019  2019            No views .. but 390 likes 🤣... MG squad        not-Tamil\n",
              "\n",
              "[2020 rows x 3 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p3FtBVTVroJi"
      },
      "source": [
        "## Saving Submission DataFrame"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SzoT0p9rqrO-"
      },
      "source": [
        "path = './drive/MyDrive/Codalab/HopeSpeech Detection For Equality,DiversityAnd Inclusion/SubmissionFiles'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LZmgZFM4qi7F"
      },
      "source": [
        "df_f.to_csv(join(path,'xlmr_tamil.csv'),index=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UDZ3hJw1NUOR"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}