HopeSpeechEDI_BERT_Multi_Tamil.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "HopeSpeechEDI_BERT_Multi_Tamil.ipynb",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "39d5d213247a45889d94c0bcf7e37471": {
          "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_78cda3805a334760bde56ccf7ab57f66",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_b816cc0d02e44aaeb3f5667dfd6c7329",
              "IPY_MODEL_28953f91395847889714b924eb482e54"
            ]
          }
        },
        "78cda3805a334760bde56ccf7ab57f66": {
          "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
          }
        },
        "b816cc0d02e44aaeb3f5667dfd6c7329": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_50f7f4ee9ca84bdaa0c87e12a1638307",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 625,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 625,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_5a899497f844486ab55e3675132b4c55"
          }
        },
        "28953f91395847889714b924eb482e54": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_4336817f9d3045dfa892542fe88f0ba6",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 625/625 [00:00&lt;00:00, 2.68kB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_93b4903d9f924d1d9284dc325e09cbbc"
          }
        },
        "50f7f4ee9ca84bdaa0c87e12a1638307": {
          "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"
          }
        },
        "5a899497f844486ab55e3675132b4c55": {
          "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
          }
        },
        "4336817f9d3045dfa892542fe88f0ba6": {
          "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"
          }
        },
        "93b4903d9f924d1d9284dc325e09cbbc": {
          "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
          }
        },
        "2a6e7a1a11b244989bc5d872cb403e9c": {
          "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_fd82134abc54471eb92ebe85b3f5ede7",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_be7600553b50442aa3b78ef614ba6354",
              "IPY_MODEL_1f41a1c682b8428a95b0cded026b8ac4"
            ]
          }
        },
        "fd82134abc54471eb92ebe85b3f5ede7": {
          "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
          }
        },
        "be7600553b50442aa3b78ef614ba6354": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_79f401e04c8e4bb791f8ffa364b65b8b",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 995526,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 995526,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_172cd0115d4a43529192740f2a14240c"
          }
        },
        "1f41a1c682b8428a95b0cded026b8ac4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_a4078358fa354b8faa2c9f2883c98e64",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 996k/996k [00:00&lt;00:00, 7.28MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_111825f438d44a319fc0e19d6083c907"
          }
        },
        "79f401e04c8e4bb791f8ffa364b65b8b": {
          "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"
          }
        },
        "172cd0115d4a43529192740f2a14240c": {
          "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
          }
        },
        "a4078358fa354b8faa2c9f2883c98e64": {
          "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"
          }
        },
        "111825f438d44a319fc0e19d6083c907": {
          "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
          }
        },
        "e09f0a05ba194cdfb2c818ac6a930c83": {
          "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_99b5a4dd6da74fdba0bdfc0c2cf7286b",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_76f68633bded4c389ad4f65dbb9ac5a4",
              "IPY_MODEL_303796ca1b3b41beb12bbd29681b594f"
            ]
          }
        },
        "99b5a4dd6da74fdba0bdfc0c2cf7286b": {
          "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
          }
        },
        "76f68633bded4c389ad4f65dbb9ac5a4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_c8e2ff73d59e413db0381f00900e2d8e",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 1961828,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1961828,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_5a765abb6a36424e8251c7a15764a418"
          }
        },
        "303796ca1b3b41beb12bbd29681b594f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_67d0c6d2791d467196ca0b061a95c5ff",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 1.96M/1.96M [00:09&lt;00:00, 216kB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_1efeb4e6f6304fcd990d6287e3d8f808"
          }
        },
        "c8e2ff73d59e413db0381f00900e2d8e": {
          "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"
          }
        },
        "5a765abb6a36424e8251c7a15764a418": {
          "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
          }
        },
        "67d0c6d2791d467196ca0b061a95c5ff": {
          "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"
          }
        },
        "1efeb4e6f6304fcd990d6287e3d8f808": {
          "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
          }
        },
        "2e42662b3e614771aa91a05ae6383aa1": {
          "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_d88c83a08500413ba75a7c66dd839810",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_0e99df9b99e34d419cfa3be6ac056d3a",
              "IPY_MODEL_884be1eb95fe42d387ad5fb5c77a3763"
            ]
          }
        },
        "d88c83a08500413ba75a7c66dd839810": {
          "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
          }
        },
        "0e99df9b99e34d419cfa3be6ac056d3a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_7cc5c4fd743a412896c57147bd2e740b",
            "_dom_classes": [],
            "description": "Downloading: 100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 714314041,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 714314041,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_5d3f4f509b7a494f9b4e5461c46b4049"
          }
        },
        "884be1eb95fe42d387ad5fb5c77a3763": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_663cae3b673e4e27958a1217ee7dc1eb",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 714M/714M [00:10&lt;00:00, 71.4MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_092570b23ba948ba89290c70583a95fe"
          }
        },
        "7cc5c4fd743a412896c57147bd2e740b": {
          "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"
          }
        },
        "5d3f4f509b7a494f9b4e5461c46b4049": {
          "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
          }
        },
        "663cae3b673e4e27958a1217ee7dc1eb": {
          "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"
          }
        },
        "092570b23ba948ba89290c70583a95fe": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "elqe9PQrVH6P"
      },
      "source": [
        "### Performing checks for the resources available"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fSELLChxa5su",
        "outputId": "219f4c5d-2f5e-4f1e-f58e-2f5c39eb6086"
      },
      "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 T4\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": "b33c17b8-53bc-41e6-896d-67165a17bc35"
      },
      "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 16.7MB/s \n",
            "\u001b[?25hRequirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.6/dist-packages (from transformers) (2019.12.20)\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: filelock in /usr/local/lib/python3.6/dist-packages (from transformers) (3.0.12)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from transformers) (2.23.0)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.6/dist-packages (from transformers) (20.9)\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 51.0MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from transformers) (1.19.5)\n",
            "Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.6/dist-packages (from transformers) (4.41.1)\n",
            "Collecting 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 55.2MB/s \n",
            "\u001b[?25hRequirement 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: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (3.0.4)\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: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->transformers) (2020.12.5)\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: pyparsing>=2.0.2 in /usr/local/lib/python3.6/dist-packages (from packaging->transformers) (2.4.7)\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",
            "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=f0ee7731548064fab12fdad3dbc64c3b0fccef2372121af56673194a0516ee6d\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": "1b75c5b2-51ef-423c-8144-b72fb4c42f96"
      },
      "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": "23af2ca8-5522-4257-dffc-0d47a0ce630e"
      },
      "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']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "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": "b10f77b3-b199-4b9f-ca86-df64c498def3"
      },
      "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": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w4VW351225YK",
        "outputId": "332e8adb-edd3-4334-fe74-80704756adc2"
      },
      "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": 9
        }
      ]
    },
    {
      "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": "5151f07b-5b11-4b26-ea94-1f82cee2f31d"
      },
      "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": 11
        }
      ]
    },
    {
      "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": "5a4721c5-0f94-42fa-e8e8-5b6d05a335ad"
      },
      "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": 13
        }
      ]
    },
    {
      "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": "15ec6100-ca1f-48fd-95de-d424833239df"
      },
      "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": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AypCmtBcDV6v",
        "outputId": "cb256540-03d3-45eb-f6e6-3a3d08c7f3fb"
      },
      "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": 15
        }
      ]
    },
    {
      "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": "ff9bc9cf-de7c-474d-9414-d7bff90f3330"
      },
      "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": 17
        }
      ]
    },
    {
      "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": "5afedda0-5188-4815-e05a-16b651732f4b"
      },
      "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 validation 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": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wkt73boTL0NO",
        "outputId": "fbfc92fc-dd31-4fe3-ab41-9b97442ec544"
      },
      "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": 20
        }
      ]
    },
    {
      "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": "dcb24313-028c-4a67-ed71-668d964a17da"
      },
      "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": 22
        }
      ]
    },
    {
      "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": "xo1_bDriha4q"
      },
      "source": [
        "### 4) Make Smart Batches (On Test)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ggH8AkAEaV9N"
      },
      "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": "x6EHYpUbz8fo"
      },
      "source": [
        "### 5) Function for calculating accuracy per class"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Knuw_DOei8RO"
      },
      "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": "sOWk_8lpWO17"
      },
      "source": [
        "### 6) Function for plotting training history"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R7-FAyfnWM65"
      },
      "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": "A7QOTORh_JHa"
      },
      "source": [
        "### 7) Function to check accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yzRDobQ6MwaR"
      },
      "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": "D7TtFkV5_Olz"
      },
      "source": [
        "### 8) Function to evaluate model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TSBNdYHYMA7Q"
      },
      "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": "Lrt_FyJIbC8j"
      },
      "source": [
        "### 9) Function for making predictions on our test dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eylCkH2Zipq5"
      },
      "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": "eiRNCS-EbZmy"
      },
      "source": [
        "### 10) Function for making predictions on test dataset(without labels)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RIuwhVSmZvkd"
      },
      "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": "4IM8r_ddYZZS"
      },
      "source": [
        "### 11) Confusion Matrix Helper Function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DqNaS8SrBox5"
      },
      "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": "OVs3Jb3Vb89i"
      },
      "source": [
        "## Smart Batching of the training data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 180,
          "referenced_widgets": [
            "39d5d213247a45889d94c0bcf7e37471",
            "78cda3805a334760bde56ccf7ab57f66",
            "b816cc0d02e44aaeb3f5667dfd6c7329",
            "28953f91395847889714b924eb482e54",
            "50f7f4ee9ca84bdaa0c87e12a1638307",
            "5a899497f844486ab55e3675132b4c55",
            "4336817f9d3045dfa892542fe88f0ba6",
            "93b4903d9f924d1d9284dc325e09cbbc",
            "2a6e7a1a11b244989bc5d872cb403e9c",
            "fd82134abc54471eb92ebe85b3f5ede7",
            "be7600553b50442aa3b78ef614ba6354",
            "1f41a1c682b8428a95b0cded026b8ac4",
            "79f401e04c8e4bb791f8ffa364b65b8b",
            "172cd0115d4a43529192740f2a14240c",
            "a4078358fa354b8faa2c9f2883c98e64",
            "111825f438d44a319fc0e19d6083c907",
            "e09f0a05ba194cdfb2c818ac6a930c83",
            "99b5a4dd6da74fdba0bdfc0c2cf7286b",
            "76f68633bded4c389ad4f65dbb9ac5a4",
            "303796ca1b3b41beb12bbd29681b594f",
            "c8e2ff73d59e413db0381f00900e2d8e",
            "5a765abb6a36424e8251c7a15764a418",
            "67d0c6d2791d467196ca0b061a95c5ff",
            "1efeb4e6f6304fcd990d6287e3d8f808"
          ]
        },
        "id": "fIlwEvcMb4pb",
        "outputId": "81d378aa-ef30-458d-b375-c7bf99d216f8"
      },
      "source": [
        "model_name = 'bert-base-multilingual-cased'\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 bert-base-multilingual-cased tokenizer...\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "39d5d213247a45889d94c0bcf7e37471",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=625.0, 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": "2a6e7a1a11b244989bc5d872cb403e9c",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=995526.0, style=ProgressStyle(descripti…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "e09f0a05ba194cdfb2c818ac6a930c83",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=1961828.0, style=ProgressStyle(descript…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "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": "c1de7754-1ca6-4414-8d9c-ccb5ea357564"
      },
      "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": "8e91e360-891f-449e-e0aa-4e33767f8acf"
      },
      "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": "3855f7fe-ceb0-427a-9482-fcb0502dfcc0"
      },
      "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": "2d4f04ee-4301-40c8-abaf-033f6cc0e0d5"
      },
      "source": [
        "train_samples[0:5]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[([101, 30580, 102], 0),\n",
              " ([101, 10808, 102], 2),\n",
              " ([101, 137, 102], 2),\n",
              " ([101, 108554, 30580, 102], 2),\n",
              " ([101, 38668, 30580, 102], 2)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rBxJhEieetFh",
        "outputId": "93f86488-fb33-414e-b133-ea0a4579800c"
      },
      "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": "ab54d0aa-c258-4cae-81fb-290fff7cbac5"
      },
      "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": "iVBORw0KGgoAAAANSUhEUgAAAuoAAAGaCAYAAABZrWb6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde1RU5f4/8PdwR0RBHW8ooCmDgYCQKWrmBRXJCyQImhTqIT3q+SVqCZXnlJlXvKYn85qSppkgmoma9e2mqXkjDTVvBZIyYNxhBmb27w8Ok+MA7qkZZoD3a63Wied59uzPfGKt8+FZn/1siSAIAoiIiIiIyKxYmDoAIiIiIiLSxUKdiIiIiMgMsVAnIiIiIjJDLNSJiIiIiMwQC3UiIiIiIjPEQp2IiIiIyAyxUCciMlPR0dEYMmSIqcMQJTMzEzNmzEDfvn0hk8kQHx9v6pDMTlZWFmQyGd577z1Th0JEDYSVqQMgIjKkzMxMbNq0CWfPnsXvv/8OGxsbtGnTBj4+PggLC0Pfvn1NHWKjlJCQgGvXrmH69Olo06YNXF1dcfr0aZw5cwYvvfQSWrRoUe8xFRUVISkpCUePHsXdu3ehUqng7OwMT09PDB48GBEREQa/Z1ZWFlJSUhAUFIQePXoY/POJqGlhoU5EjcZPP/2E6OhoWFlZITQ0FN26dUN5eTl+/fVXfP/993BwcGChbgRKpRI//vgjJk2ahKlTp2rG33vvPaxfvx5hYWH1XqgXFxcjPDwcmZmZGDFiBMaNGwdra2tkZmbi/Pnz2Llzp1EK9bt372L9+vVwcXHRKdRdXFyQnp4OS0tLg9+XiBonFupE1Ghs2LABZWVlSE1Nhaenp868XC43QVSNX25uLgRBQMuWLevtniqVCkqlEvb29jXOf/LJJ7hz5w5ef/11vPTSSzrzhv5dKC4uRvPmzetcI5FIYGtra9D7ElHjxh51Imo07ty5AycnpxqLdACQSqVaP3/++eeYPn06Bg0aBG9vb/Tp0wczZszA1atXda4dMmQIoqOjcfXqVcTExKBXr14IDAzE0qVLUVlZCYVCgWXLluGZZ55Bz5498cILL+DmzZtan5GcnAyZTIaTJ0/ivffew+DBg+Ht7Y3Ro0fj8OHDen3PV199FQMGDIC3tzeGDBmCZcuWobS0VGvd77//joSEBM19AgMDERUVhZSUlMfe4/79+1i6dCnGjh2L3r17o2fPnggJCcGmTZugUqk06+Lj4zF48GAAwPr16yGTySCTyTBkyBCsX78eADB06FDN+MP92UVFRVixYgWGDRsGb29v9O3bF3PmzEFmZmateduwYQOCgoLg4+ODI0eO1JkjAAgMDKxx/tHfBQA4e/YsJk+ejICAAE2r1L59+3TWVT87kJmZif/3//4fnn76aQQEBCA5ORkvvvgigKpWoOrvHB0dDaDmHvWHx7766iuMGzcOPXv2xIABA7Bs2TJUVlbq3P/o0aMYM2YMevbsiUGDBmH9+vU4efIkZDIZkpOTa80JETU83FEnokbD1dUVt2/fxrFjxzB8+PDHrv/oo4/g5OSE8ePHQyqV4rfffsMnn3yCCRMmICUlBe7u7lrr7927h8mTJyMkJAQjRozA999/j+3bt8PS0hI3btxAeXk5Xn75Zfzxxx/Ytm0bZsyYgSNHjsDCQntPJDExEaWlpZgwYQKAqkJ0zpw5UCgUeP755+uM+fLly5qe78jISLRr1w5Xr15FUlISLly4gKSkJFhbW6OyshKTJ0/G/fv3MXHiRLi7u6O4uBjXrl3Djz/+iLCwsDrvc+3aNRw7dgzDhg2Dq6srKioq8O2332LlypXIysrCwoULAQCRkZHw9PTEkiVLMGzYMAwbNgwA0KFDB+zcuRPHjx9HQkICnJ2dAQAymQxAVZEeFRWF7OxsjBs3Dt27d4dcLsfu3bsRERGB/fv3w8XFRSum6sJ1/PjxcHBwQJcuXWqN39XVVZPbefPmwcqq7v+7+/LLLzFr1iy0adMGkydPRvPmzXH48GG8+eabyMrKQlxcnNb6kpISTJo0Cf7+/pg9ezYePHiA3r17Y/r06di4cSMiIyMREBAAAGjTpk2d9waAr7/+Grt370ZUVBTGjRuHEydOYNu2bWjZsiWmT5+uWff5559jzpw5cHV1xaxZs2BpaYkDBw7gyy+/fOw9iKgBEoiIGonz588LXl5egoeHhzB8+HAhPj5e2LVrl3Djxo0a15eUlOiM3bhxQ/Dy8hL+85//aI0PHjxY8PDwED7//HOt8bCwMEEmkwnTp08X1Gq1ZnzHjh2Ch4eH8M0332jG9u/fL3h4eAiDBg0SCgsLNeOFhYXCoEGDhN69ewtlZWWa8UmTJgmDBw/Wut/o0aOFESNGCEVFRVrjx44dEzw8PIT9+/cLgiAIGRkZgoeHh7Bp06Yav/vjlJWVaX2favPmzRM8PT2F+/fva8YyMzMFDw8PYd26dVpr161bJ3h4eAiZmZk6n/POO+8IPXv2FDIyMrTGs7KyhF69egnz58/XjFXnbfjw4UJpaamo+PPz84Vnn31W8PDwEAIDA4V//etfwgcffCCcPXtWUKlUWmsrKyuFQYMGCQEBAcK9e/c04wqFQoiMjBQ8PT2F27dva8YnTZokeHh4CKtWrdK57w8//KD13+FhNeWpeszX11crT2q1WnjuueeE/v37a8YqKiqEAQMGCIGBgUJ+fr5mvLi4WBgyZEit9yWihoutL0TUaPTq1Qv79+9HWFgYioqKkJycjLfffhshISF44YUXdFoqmjVrBgAQBAHFxcV48OABnJ2d0aVLF6Snp+t8frt27TBy5EitMX9/fwiCgOjoaEgkEs34U089BQD49ddfdT5nwoQJcHR01Pzs6OiIqKgoFBQU4PTp07V+v2vXruHatWsYNWoUlEolHjx4oPknICAAzZo1w/fff6/5TAA4ffo08vLy6sxbTezs7DTfR6lUIj8/Hw8ePMCAAQOgVqtx+fJlvT+zmiAIOHToEHr37o22bdtqfQ97e3v4+fnhu+++07luwoQJtfakP6ply5ZITk5GbGwsHB0dcfToUaxcuRIvvPACgoKCtD7/ypUrmp39du3aacZtbGzwj3/8A2q1GidOnNC5x8MPzv5dQ4cORadOnTQ/SyQS9OnTB3K5HCUlJZo4c3JyEBYWpvU8gIODA6KiogwWCxGZD7a+EFGjIpPJsHTpUgBVJ3CcPXsW+/btw48//ogZM2Zg//79sLGxAQD8/PPPWLt2Lc6cOaPT3/1w0VTXWHXB9Ohc9Skn+fn5Otd07dpVZ+yJJ54AUNWzXJvqnvf33nuv1rO4c3NzAVSdMDJ9+nRs2rQJAwYMQI8ePdC3b18EBwfDx8en1ntUq6ysxKZNm5Camopff/0VgiBozRcWFj72M2rz4MED5Ofn47vvvqu1h/zRdiEAdba61KRVq1aYN28e5s2bhz/++AMXL17EkSNHcPDgQcyaNQupqalwc3PT5Lxbt246n9G9e3cA0Pkjr1WrVgY9yaZz5846Y05OTgCqfoccHBw0cdaUB31zQ0QNAwt1Imq0XFxc4OLigrFjx2LixIk4f/480tPT8dRTTyE7OxsvvPACmjdvjn/+85/o2rUr7O3tIZFIsHjxYp3CHUCdx+rVVFgC0ClwDWHKlCl45plnapx7uHiMi4tDeHg4/u///g8//vgjPv30U2zduhX/+Mc/8Oqrr9Z5j6VLlyIpKQkhISGYPn06WrVqBWtra1y5cgWJiYlQq9V/Of7qnPTr1w+xsbGir7Ozs/vL93R2dsbgwYMxePBgdOjQARs3bsThw4cxY8aMv/R5Ynf2xarrd8sYv0NE1DCwUCeiRk8ikcDX1xfnz59HTk4OAOD48eMoLS3F+++/r3O2en5+vmbX3Rhu3bqlM1a9W17Trn01Nzc3AFV/FPTr10/UvTp37ozo6GhER0dDoVBg6tSp2LJlC6ZMmYLWrVvXel1qaip69+6N1atXa43X1MpTm4dbgR5WvRtdXFws+nsYkq+vL4Cqk22AP3N+48YNnbXVYzXteNektu9sCNUP196+fVtnrqYxImr42KNORI3G999/X+NxduXl5Zre7eoWk+odzEd3Kz/55BOjn7f+8ccfo6ioSPNzUVER9uzZgxYtWuDpp5+u9bonn3wSHh4e2LNnj04rBlDVrlLdalNUVISKigqteVtbW03bTUFBQZ0xWlhY6OSmtLQUH374YZ3XPaz6GYBH72VhYYHRo0cjPT0daWlpNV77V/rqH3bhwoVa23O++OILAH+2unh5eaFjx45ITk7W+m9fUVGBrVu3QiKRYOjQoaLuW9t3NgRvb29IpVKkpKRofX5JSQn27Nlj8PsRkelxR52IGo0lS5YgPz8fQ4YMgYeHB+zs7HDv3j0cOnQId+7cQWhoqOZ4wIEDB8Le3h6vvfYaJk2ahBYtWuD8+fP45ptv4OrqqnVWuKE5OzsjIiJCcxRjcnIysrOzsWjRojpbKiQSCZYvX46XXnoJY8aMwbhx47Tevnr8+HHMmTMHzz//PE6fPo0FCxZg+PDh6NKlCxwcHHD58mV8+umn8PX1rbFP/mEjRozA3r17MXv2bPTr1w+5ubnYv3+/pm9ajOqd68TERIwePRq2trbo3r07PDw8EBcXh/Pnz2P27NkYOXIkfH19YW1tjezsbHzzzTfw8vLSPGvwVxw6dAjJycl49tln4ePjAycnJ+Tn5+Prr7/G6dOn0a1bN4wbNw5A1R9tCxYswKxZsxAeHq45/vHIkSO4ePEipk+frnNUZ226desGBwcH7N69G3Z2dmjRogVatWpVay++PqysrDB//nzMmzcPERERCA8Ph6WlJVJSUuDk5ISsrCyj7ugTUf1joU5EjUZ8fDxOnDiBc+fO4ejRoygqKoKjoyM8PDwQGxurdUa5q6srNm/ejFWrVmHjxo2wtLSEv78/kpKS8M477+Du3btGi3PevHn48ccfsXv3buTm5qJLly6aYvZxevTogZSUFHzwwQf48ssvsWfPHjg4OMDFxQVhYWGaglAmk2HYsGE4c+YMDh06BLVajQ4dOmDatGmYMmXKY++TkJAABwcHpKWl4cSJE+jQoQMiIyPRs2dPxMTEiPqeAQEBmDdvHvbs2YMFCxagsrISs2bNgoeHBxwdHfHxxx9j27ZtmntYWlqiffv2CAgIQEREhKh71CYqKgqOjo44ffo0tm/fjvz8fFhbW8PNzQ2zZs3C5MmTNbvfQNULrT788EO8//772Lp1KyoqKvDEE09g0aJFesViZ2eH1atXY82aNVi8eDGUSiWefvppgxTqADB69GhYWVnhv//9L9atW4c2bdogPDwcMpkMs2bN4ptPiRoZicCnVIiI6kVycjISEhKwc+dO9OnTx9ThUCOybds2LFu2DHv37oWfn5+pwyEiA2GPOhERUQOhVCp12rJKSkqwa9cuODk54cknnzRRZERkDGx9ISIiaiAyMzMRGxuL5557Dp06dYJcLkdKSgqysrLw1ltvGfW0IiKqfyzUiYiIGohWrVrBz88Phw4dQl5eHqysrODh4YG5c+ciJCTE1OERkYGxR52IiIiIyAyxR52IiIiIyAyxUCciIiIiMkPsUa/DH3+UQK2u/86g1q2bIy+vuN7v2xAxV+IxV/phvsRjrsRjrvTDfInHXIlnbrmysJDA2dmhxjkW6nVQqwWTFOrV9yZxmCvxmCv9MF/iMVfiMVf6Yb7EY67Eayi5YusLEREREZEZYqFORERERGSGWKgTEREREZkhFupERERERGaIhToRERERkRlioU5EREREZIbMqlDfvHkzZDIZxo4dqzN3/vx5TJgwAb6+vujfvz8WLVqEsrIynXVKpRIrVqzAgAED4OPjg/Hjx+PUqVP1ET4RERERkcGYTaEul8vx/vvvo1mzZjpzGRkZiImJgUKhQHx8PMLDw7F3717ExcXprI2Pj8eOHTswZswYvPHGG7CwsEBsbCwuXLhQH1+DiIiIiMggzOaFRytXroS3tzcEQUBhYaHW3KpVq+Dk5ISkpCQ4OFS9ualTp0548803cerUKQQGBgIA0tPTcfjwYSQkJCAmJgYAEBoailGjRiExMRG7du2q1+9ERERERPRXmcWOenp6Og4ePIiEhASdueLiYpw8eRKhoaGaIh0Axo4di2bNmuHIkSOasbS0NFhbWyMiIkIzZmtri/DwcJw7dw45OTnG/SJERERERAZi8kJdEAS88847CA0NRY8ePXTmr127hsrKSnh7e2uN29jYoEePHsjIyNCMZWRkoEuXLloFPQD4+PhAEASttURERERE5szkhfqBAwdw48YNzJ49u8Z5uVwOAJBKpTpzUqlUa5dcLpejbdu2Na4DwB11IiIiItJx5fYDvH/gMtSCYOpQtJi0R724uBgrV67Eyy+/XGOBDQDl5eUAqnbQH2Vra6uZr15rbW1d4zoAUCgUesXXunVzvdYbklTqaLJ7NzTMlXjMlX6YL/GYK/GYK/0wX+IxV+I9mqtbP/yGs1dz8ObUvrCwkJgoKl0mLdTff/99WFtbY/LkybWusbOzA1B17OKjFAqFZr56bUVFRY3rgD8LdrHy8oqhVtf/X1ZSqSPk8qJ6v29DxFyJx1zph/kSj7kSj7nSD/MlHnMlXk25KilRwNbGEnl5xfUej4WFpNbNYZMV6jk5OdixYwdeeeUV5ObmasYVCgUqKiqQlZUFR0dHTdtKdQvMwx5tdXm0FebhdQBq3bUnIiIioqZLLQiwkJjPTno1k/Wo5+XloaKiAomJiRg6dKjmn0uXLuHmzZsYOnQoNm/eDA8PD1hZWeHy5cta1yuVSmRkZGg9gOrp6Ynbt2+jpKREa+2lS5c080REREREDxPUgBl1vGiYbEe9U6dO2LBhg874mjVrUFpaitdffx3u7u5wdHREYGAgUlNTMW3aNM2JLqmpqSgtLUVwcLDm2uDgYGzbtg379u3TnKOuVCqRnJwMf39/tGvXrl6+GxERERE1HGoIkJjhjrrJCnVHR0cEBQXpjO/YsQOWlpZac3FxcYiKikJ0dDQiIiJw7949bN++HQMHDkS/fv0063x9fREcHIzExETI5XK4uroiJSUF2dnZWLJkSb18LyIiIiJqWAS1wB31v8rLywvbt29HYmIilixZgubNm2P8+PGYM2eOztrly5djzZo1SE1NRUFBAWQyGTZt2oSAgAATRE5ERERE5k4tgDvqYiQlJdU4/tRTT2HPnj2Pvd7W1hbz58/H/PnzDR0aERERETVCgiCY1bGM1Uz+wiMiIiIiIlNSCwLMcEOdhToRERERNW2CAB7PSERERERkbrijTkRERERkhrijTkRERERkhtRq8zxHnYU6ERERETVpPPWFiIiIiMgMVZ2jbuoodLFQJyIiIqImTRAESGB+lToLdSIiIiJq0qoeJjV1FLpYqBMRERFRkyYIAsxwQ52FOhERERERW1+IiIiIiMyMYOoAasFCnYiIiIjI/DbUWagTERERUdNmpi3qLNSJiIiIqGkTIPAcdSIiIiIisyMA5rinzkKdiIiIiJo0AXwzKRERERGRWTLDOp2FOhERERE1cYJ5HtDIQp2IiIiImjQzbVFnoU5ERERETVvV8YzmV6mzUCciIiIiMr86nYU6ERERETVtgiCYY50OK1Pd+KeffsLGjRvx888/Iy8vD46OjvD09MTMmTPh7++vWRcdHY0zZ87oXB8SEoLVq1drjSmVSqxduxapqakoLCyEp6cn4uLiEBgYaPTvQ0REREQNFwv1h2RmZkKlUiEiIgJSqRRFRUU4dOgQJk2ahM2bN6N///6atR07dsTs2bO1rndxcdH5zPj4eBw7dgwvvvgi3NzckJKSgtjYWCQlJaFXr15G/05ERERE1ECZ4UHqJivUQ0JCEBISojU2YcIEBAUFYefOnVqFeosWLTB27Ng6Py89PR2HDx9GQkICYmJiAAChoaEYNWoUEhMTsWvXLoN/ByIiIiJq+Mz0dEbz6lG3t7dHq1atUFhYqDNXWVmJkpKSWq9NS0uDtbU1IiIiNGO2trYIDw/HuXPnkJOTY5SYiYiIiKhhE2CerS8mL9SLi4vx4MED3Lp1C6tWrcL169d1espv3rwJPz8/+Pv7Y8CAAdi4cSPUarXWmoyMDHTp0gUODg5a4z4+PhAEARkZGUb/LkRERETUAAmCOXa+mK71pdrrr7+Oo0ePAgCsra0RFRWF6dOna+Y7d+6MPn36QCaTobi4GJ999hlWr16N7OxsLFy4ULNOLpejXbt2Op8vlUoBgDvqRERERFQjM+18MX2hPnPmTERGRuLevXtITU2FUqlERUUFbGxsAACLFy/WWh8WFoZXXnkFn3zyCWJiYtC1a1cAQHl5OaytrXU+39bWFgCgUCj0jq116+Z6X2MoUqmjye7d0DBX4jFX+mG+xGOuxGOu9MN8icdcifdorqysLGFtZWF2OTR5oS6TySCTyQAAY8aMwbhx45CQkIB169bVes2UKVOQlpaG06dPawp1Ozs7VFRU6KytLtCrC3Z95OUVQ62u/7+xpFJHyOVF9X7fhoi5Eo+50g/zJR5zJR5zpR/mSzzmSryaclVRoQIEwSQ5tLCQ1Lo5bPIe9YdZW1tj6NChOHbsGMrLy2td1759ewBAQUGBZkwqldbY3iKXywEAbdu2NXC0RERERNRYmGOPulkV6kBVC4sgCHWe8JKZmQkAaNWqlWbM09MTt2/f1rnu0qVLmnkiIiIiokcJZtqlbrJC/cGDBzpjxcXFOHr0KDp06IDWrVujuLgYSqVSa41KpcIHH3wACwsLrdNhgoODUVFRgX379mnGlEolkpOT4e/vX+ODpkREREREEMzzeEaT9ajPnj0btra26NWrF6RSKX7//XckJyfj3r17WLVqFQDgypUrmDt3LkaNGgVXV1eUlpbiyJEjuHz5MmJjY9G5c2fN5/n6+iI4OBiJiYmQy+VwdXVFSkoKsrOzsWTJElN9TSIiIiIycwJglr0vJivUx4wZg9TUVCQlJaGwsBCOjo7w8/PD8uXL8fTTTwMAOnbsCH9/fxw7dgy5ubmwsLBA9+7dsXTpUoSFhel85vLly7FmzRqkpqaioKAAMpkMmzZtQkBAQH1/PSIiIiJqIAQz3VGXCIK5vjTV9Hjqi/ljrsRjrvTDfInHXInHXOmH+RKPuRKvply9s+MsHOytMWe8X73H02BOfSEiIiIiqm9VO+rmt6fOQp2IiIiImjwzbFFnoU5ERERETZu59oGzUCciIiKips1MK3UW6kRERETUpAkQzLBDnYU6ERERETV1AiAxwyZ1FupERERE1KSZaecLC3UiIiIiatoEgae+EBERERGRSCzUiYiIiKiJM8/mFxbqRERERNSkCeDDpERERERE5kcAj2ckIiIiIjI3AmCWlToLdSIiIiJq0gSBLzwiIiIiIjJP7FEnIiIiIjI/5lems1AnIiIioiZOMM/TGVmoExEREVHTVnU8o6mj0MVCnYiIiIiaNjPdUmehTkRERERNGl94RERERERkjvjCIyIiIiIi8yOYaaXOQp2IiIiImjwzrNNZqBMRERFR02amz5KarlD/6aefMHPmTAwePBg+Pj7o378/pk6divPnz+usPX/+PCZMmABfX1/0798fixYtQllZmc46pVKJFStWYMCAAfDx8cH48eNx6tSp+vg6RERERNSgmd+euskK9czMTKhUKkRERGDBggWYOnUqHjx4gEmTJuH777/XrMvIyEBMTAwUCgXi4+MRHh6OvXv3Ii4uTucz4+PjsWPHDowZMwZvvPEGLCwsEBsbiwsXLtTnVyMiIiKiBkQQzPMcdStT3TgkJAQhISFaYxMmTEBQUBB27tyJ/v37AwBWrVoFJycnJCUlwcHBAQDQqVMnvPnmmzh16hQCAwMBAOnp6Th8+DASEhIQExMDAAgNDcWoUaOQmJiIXbt21d+XIyIiIqIGRDDD/XQ9C/ULFy7go48+wq+//or8/HwIjzT0SCQSfPHFF385GHt7e7Rq1QqFhYUAgOLiYpw8eRJTp07VFOkAMHbsWCxevBhHjhzRFOppaWmwtrZGRESEZp2trS3Cw8OxevVq5OTkoG3btn85NiIiIiJqnATAHDtfxBfqBw4cQEJCAqysrODu7o4OHToYJIDi4mIolUrk5+fjwIEDuH79OmbOnAkAuHbtGiorK+Ht7a11jY2NDXr06IGMjAzNWEZGBrp06aJV0AOAj48PBEFARkYGC3UiIiIi0iEIgMQMK3XRhfr777+PLl26YPv27WjXrp3BAnj99ddx9OhRAIC1tTWioqIwffp0AIBcLgcASKVSneukUikuXryo+Vkul9cYV/W1OTk5BouZiIiIiBoZ86vTxRfq2dnZeO211wxapAPAzJkzERkZiXv37iE1NRVKpRIVFRWwsbFBeXk5gKod9EfZ2tpq5gGgvLwc1tbWNa4DAIVCoXdsrVs31/saQ5FKHU1274aGuRKPudIP8yUecyUec6Uf5ks85kq8R3NlYSGBvZ212eVQdKHevn17KJVKgwcgk8kgk8kAAGPGjMG4ceOQkJCAdevWwc7ODgBqvK9CodDMA4CdnR0qKipqXAf8WbDrIy+vGGp1/R+sKZU6Qi4vqvf7NkTMlXjMlX6YL/GYK/GYK/0wX+IxV+LVlCuVSo3y8gqT5NDCQlLr5rDo4xmjoqJw6NAhqFQqgwX2KGtrawwdOhTHjh1DeXm5pm2lugXmYXK5XKvnXCqV1tjeUn0t+9OJiIiIqCYCGtjxjGfPntX62dvbG8eOHUNERAQmTpyITp06wdLSUue63r17/62AysvLIQgCSkpK4OHhASsrK1y+fBnDhw/XrFEqlcjIyMDo0aM1Y56enkhKSkJJSYnWA6WXLl3SzBMRERER6TDTY19qLdSjo6MheeRPi+rjGN98880a5yQSidZJLHV58OABWrVqpTVWXFyMo0ePokOHDmjdujUAIDAwEKmpqZg2bZqmAE9NTUVpaSmCg4M11wYHB2Pbtm3Yt2+f5hx1pVKJ5ORk+Pv7G7y3noiIiIgahwa3o75kyRKj3nj27NmwtbVFr169IJVK8fvvvyM5ORn37t3DqnjAce8AACAASURBVFWrNOvi4uIQFRWF6OhoRERE4N69e9i+fTsGDhyIfv36adb5+voiODgYiYmJkMvlcHV1RUpKCrKzs43+XYiIiIioARMa2AuPwsLCjHrjMWPGIDU1FUlJSSgsLISjoyP8/PywfPlyPP3005p1Xl5e2L59OxITE7FkyRI0b94c48ePx5w5c3Q+c/ny5VizZg1SU1NRUFAAmUyGTZs2ISAgwKjfhYiIiIgarqrOF/Mr1UWf+pKQkICoqCj4+vrWOJ+eno6PP/5Y9O51eHg4wsPDRa196qmnsGfPnseus7W1xfz58zF//nxRn0tEREREZK5En/qSkpKC3377rdb5rKwsHDhwwCBBERERERHVl6o3k5of0YX645SWlsLKSvQGPRERERGR+TDDSr3Oyjo7Oxt3797V/Hzr1i2dYxsBoKCgAB9//DHc3NwMHyERERERkREJDe1hUgBITk7G+vXrIZFIIJFIsHHjRmzcuFFnnSAIsLCwwOLFi40WKBERERGRsUjMsFSvs1APCgqCi4sLBEHA66+/jvHjx6NXr15aayQSCZo1a4aePXuiQ4cORg2WiIiIiMjQBAENr/XF09NT80bP7OxsDB8+HB4eHvUSGBERERFRfTDTOl388YyzZs0yZhxERERERPQQ0YW6mKMX7ezs0LFjRzz55JM8AYaIiIiIGgjzPJ9RdDUdHx8PyUNvbBIEAQB0xiQSCZycnBAXF4fx48cbMFQiIiIiIsOrOkfd/Cp10YX69u3bkZiYiIKCAkRFRaFLly4Aqo5s3Lt3L5ydnTFt2jT89ttv2LVrF/7zn/+gZcuWGDFihNGCJyIiIiL6uwQAEvOr08UX6ufOnYNSqcShQ4dgb2+vGR86dCgmTpyIqKgoXL9+HTNmzEBkZCTGjh2L7du3s1AnIiIiIvMmmDqAmol+M+n+/fsRFhamVaRXc3BwQFhYGD799FPNz6Ghobh+/brhIiUiIiIiMgIBglnuqIsu1PPy8qBSqWqdr6ysRG5urubntm3b1rmeiIiIiMgcVFYKsLIUXRbXG9ERubu749NPP0VxcbHOXFFREfbv36/pWweArKwstG7d2jBREhEREREZgSAIUAsCLC3Mb0tddI/6zJkzMXv2bAQHB+P555+Hu7s7AOD27dtISUlBXl4e1qxZAwBQq9U4fPiwzltMiYiIiIjMifp/JxlaNORCfcSIEVi5ciWWLFmCTZs2ac1JpVKsWLFC8+CoSqXC5s2b0apVK8NGS0RERERkQGp11f826B11AAgJCcGIESNw5coVZGVlAQBcXFzg7e0NS0tLzTpra2t07drVsJESERERERmYWv2/HXUzfJpU79eHWlpawsfHBz4+PsaIh4iIiIio3qhreImnudC7UAeAsrIy5Ofna95O+rCOHTv+7aCIiIiIiOpDo+hRV6vV2LJlC5KSkrSOYXxURkaGQQIjIiIiIjK26taXBt2jnpiYiG3btqF79+4YMWIEnJycjBkXEREREZHR/dmjbuJAaiC6UD948CCeeeYZbN682ZjxEBERERHVm//V6ZCYYaUu+oVHhYWFGDp0qDFjISIiIiKqV43i1BcPDw/I5XKD3Tg9PR0pKSk4ffo0srOz4eTkhF69emH27Nlwc3PTrIuOjsaZM2d0rg8JCcHq1au1xpRKJdauXYvU1FQUFhbC09MTcXFxCAwMNFjcRERERNR4VD9M2qB71GfNmoU33ngD4eHh6NChw9++8ZYtW3D+/HkEBwdDJpNBLpdj165dCA0NxaeffoonnnhCs7Zjx46YPXu21vUuLi46nxkfH49jx47hxRdfhJubG1JSUhAbG4ukpCS+JZWIiIiIdDSKHfXLly+jY8eOCAkJwbBhw9CpUydYWGh3zkgkEsycOVPU58XExCAxMRE2NjaasZCQEIwePRqbN2/G0qVLNeMtWrTA2LFj6/y89PR0HD58GAkJCYiJiQEAhIaGYtSoUUhMTMSuXbtEflMiIiIiaio056iLbgivP6IL9fXr12v+/eDBgzWu0adQ9/f31xlzd3dH9+7dcfPmTZ25yspKKBQKODg41Ph5aWlpsLa2RkREhGbM1tYW4eHhWL16NXJyctC2bVtRsRERERFR09AodtRPnDhhzDgAAIIgIDc3F56enlrjN2/ehJ+fHyoqKiCVSjFp0iS8/PLLWjv6GRkZ6NKli04h7+PjA0EQkJGRwUKdiIiIiLRUn/rSoHvUa+oJN7SDBw/i/v37iIuL04x17twZffr0gUwmQ3FxMT777DOsXr0a2dnZWLhwoWadXC5Hu3btdD5TKpUCAHJycowePxERERE1LI1iR/1hv/76K3Jzc+Hh4QFHR0eDBHLz5k0sXLgQAQEBWv3oixcv1loXFhaGV155BZ988gliYmLQtWtXAEB5eTmsra11PtfW1hYAoFAo9I6pdevmel9jKFKpYfLaFDBX4jFX+mG+xGOuxGOu9MN8icdcifdwrv4oqwQAODk3M7sc6lWof/XVV3j33Xdx9+5dAMC2bdsQGBiIvLw8REVFYe7cuQgODtY7CLlcjmnTpqFly5ZYu3atzkOqj5oyZQrS0tJw+vRpTaFuZ2eHiooKnbXVBXp1wa6PvLxizV9Z9UkqdYRcXlTv922ImCvxmCv9MF/iMVfiMVf6Yb7EY67EezRXDx6UAACKCstNkkMLC0mtm8Oin289ffo0Zs2ahZYtW2LmzJkQhD8L2NatW8PV1RWff/653sEVFRUhNjYWRUVF2LJli6ZVpS7t27cHABQUFGjGpFJpje0t1We/sz+diIiIiB5lzueoiy7UN2zYAJlMhn379uGFF17Qmffz88OVK1f0urlCocD06dNx584dfPDBB5rd8cfJzMwEALRq1Uoz5unpidu3b6OkpERr7aVLlzTzREREREQPU6mqe9RNHEgNRBfqP/30E8aMGVNrW0r79u2Rm5sr+sYqlQqzZ8/GxYsXsXbtWvj5+emsKS4uhlKp1Lnugw8+gIWFhdYbR4ODg1FRUYF9+/ZpxpRKJZKTk+Hv71/jg6ZERERE1LQVllbVmioTtDs/jugedUEQanxYs9off/xR5/yjli5dii+//BKDBw9Gfn4+UlNTNXMODg4ICgrClStXMHfuXIwaNQqurq4oLS3FkSNHcPnyZcTGxqJz586aa3x9fREcHIzExETI5XK4uroiJSUF2dnZWLJkiei4iIiIiKjpsLKs2oR2bGbzmJX1T3Sh3rVrV5w7d67Gtheg6kFTfdpLrl69qrnuq6++0ppzcXFBUFAQOnbsCH9/fxw7dgy5ubmwsLBA9+7dsXTpUoSFhel85vLly7FmzRqkpqaioKAAMpkMmzZtQkBAgOi4iIiIiKjpqN5Jt7Q0v94X0YV6eHg43n33Xezbtw9Dhw4FUPUm0rKyMqxcuRIXL17EsmXLRN84KSnpsWs6d+6MdevWif5MW1tbzJ8/H/Pnzxd9DRERERE1XSqVGoB5PkwqulCfOHEizp8/jwULFmDZsmWQSCSYO3cu8vPzoVKp8Pzzz2PMmDHGjJWIiIiIyKD+3FEX/ehmvdHrHPXExESMGDECBw8exK1btyAIAnx8fBAaGooRI0YYK0YiIiIiIqPQFOqN4c2kw4YNw7Bhw3TGb968iWvXriEkJMQggRERERERGZs596gbbI//2LFjmDt3rqE+joiIiIjI6My5R938mnGIiIiIiOqJZke9lncFmZL5RUREREREVE+aROsLEREREVFDw9YXIiIiIiIz9GfrCwt1IiIiIiKzoVILsJBIIGloxzP26tVLdNCVlZUGCYiIiIiIqL4oK9SwsjK/Ih14TKHu7e1dX3EQEREREdW72/cKzXI3HXhMoZ6UlFRfcRARERER1TtHe2tYmWF/OsAedSIiIiJqwlRqAW2c7E0dRo1YqBMRERFRk6VSqbmjTkRERERkblRqwSyPZgRYqBMRERFRE1apFmBpaZ4lsXlGRURERERUD1Qq7qgTEREREZkdlVrNQp2IiIiIyNyoGkvri0qlwoEDBzBv3jxMnjwZP//8MwCgoKAABw4cwP37940SJBERERGRMZhz60udLzx6WFlZGaZMmYILFy7A3t4e5eXlKCgoAAA0b94ciYmJGDduHOLi4owWLBERERGRIVWq1LCyNM9CXfSO+nvvvYfLly9j/fr1OHHiBARB0MxZWlpi+PDh+O6774wSJBERERGRMSgr1bC2sjR1GDUSXainpaUhMjISQUFBkEh0/+pwdXXF3bt3DRocEREREZExqRrDjnpOTg5kMlmt8/b29igpKRF94/T0dLz99tsICQmBn58fBg0ahLi4OPz66686a8+fP48JEybA19cX/fv3x6JFi1BWVqazTqlUYsWKFRgwYAB8fHwwfvx4nDp1SnRMRERERNS0VKoFWDX0h0mdnJzqfFj0l19+Qdu2bUXfeMuWLTh+/Dj69euHN954A+PHj8eZM2cQGhqKmzdvatZlZGQgJiYGCoUC8fHxCA8Px969e2vshY+Pj8eOHTswZswYvPHGG7CwsEBsbCwuXLggOi4iIiIiajoaxcOkgYGBSE5OxtSpU3XmMjMzsX//fowdO1b0jWNiYpCYmAgbGxvNWEhICEaPHo3Nmzdj6dKlAIBVq1bByckJSUlJcHBwAAB06tQJb775Jk6dOoXAwEAAVTv0hw8fRkJCAmJiYgAAoaGhGDVqFBITE7Fr1y7RsRERERFR06BSq2HZ0FtfZs2ahcLCQoSHh+Pjjz+GRCLBt99+i5UrV+L555+HjY0Npk2bJvrG/v7+WkU6ALi7u6N79+6aHfXi4mKcPHkSoaGhmiIdAMaOHYtmzZrhyJEjmrG0tDRYW1sjIiJCM2Zra4vw8HCcO3cOOTk5omMjIiIiosZPrRYgCICVRQNvfXFzc8OHH34IS0tLrFu3DoIgYNu2bdi8eTPat2+PHTt2oEOHDn8rGEEQkJubC2dnZwDAtWvXUFlZCW9vb611NjY26NGjBzIyMjRjGRkZ6NKli1ZBDwA+Pj4QBEFrLRERERGRSq0GALPdURfd+gIA3t7eOHjwIK5fv46bN29CEAS4u7vjySefNEgwBw8exP379zX953K5HAAglUp11kqlUly8eFHzs1wuR7t27WpcB4A76kRERESkpVJVddy4pZnuqOtVqFfz8PCAh4eHQQO5efMmFi5ciICAAE2ve3l5OQDotMgAVW0t1fPVa62trWtcBwAKhULvmFq3bq73NYYilTqa7N4NDXMlHnOlH+ZLPOZKPOZKP8yXeMyVeNW5KixRAgBatrQzy/yJLtRPnTqFkydPYu7cuTXOr1y5Ev3790ffvn31DkIul2PatGlo2bIl1q5dC4v//VVjZ2cHoOrYxUcpFArNfPXaioqKGtcBfxbs+sjLK4ZaLTx+oYFJpY6Qy4vq/b4NEXMlHnOlH+ZLPOZKPOZKP8yXeMyVeA/n6o+iqjpRUVZhsvxZWEhq3RwWvc+/efPmGs84r5aVlYXNmzfrHVxRURFiY2NRVFSELVu2aLW5VP97dQvMw+RyudZxkFKptMb2lupr9Tk6koiIiIgav3JlJQDAzqaBv5n06tWr8PPzq3Xe19cX165d0+vmCoUC06dPx507d/DBBx+ga9euWvMeHh6wsrLC5cuXtcaVSiUyMjLQo0cPzZinpydu376t89KlS5cuaeaJiIiIiKqVK1UAADubv9QNbnSiC/WioiLY29vXOm9ra4uCggLRN1apVJg9ezYuXryItWvX1vhHgKOjIwIDA5GamqpVgKempqK0tBTBwcGaseDgYFRUVGDfvn2aMaVSieTkZPj7+9f4oCkRERERNV2K/xXqtma6oy76z4d27drhypUrtc5fuXKlxtNZarN06VJ8+eWXGDx4MPLz85GamqqZc3BwQFBQEAAgLi4OUVFRiI6ORkREBO7du4ft27dj4MCB6Nevn+YaX19fBAcHIzExEXK5HK6urkhJSUF2djaWLFkiOi4iIiIiahrKK6p31Bt4oT5o0CDs2bMHISEhWgUyUPWg6YEDBxAeHi76xlevXgUAfPXVV/jqq6+05lxcXDSFupeXF7Zv347ExEQsWbIEzZs3x/jx4zFnzhydz1y+fDnWrFmD1NRUFBQUQCaTYdOmTQgICBAdFxERERE1Debeoy4RBEHUsSa5ubkICwtDbm4uBg4cqOn5vnr1Kr755hu0adMG+/fvb1QPbfLUF/PHXInHXOmH+RKPuRKPudIP8yUecyXew7n6+uJd7Ei7hsQZ/dCqhd1jrjSOuk59Eb2j3qZNG+zZswdvvfUWvvnmG3z99dcAAIlEgoEDB2LBggWNqkgnIiIiosZNYeYPk+oVlYuLCzZv3oyCggLNUY1ubm5o2bKlUYIjIiIiIjKW6h51W5tG9GbSli1bwsfHx9CxEBERERHVm3KlCtZWFrC0aCSFellZGe7evYv8/HzU1N7eu3dvgwRGRERERGRMCqXKbB8kBfQo1MvKyrBkyRIkJydDpVLpzAuCAIlEgoyMDIMGSERERERkDOVKFWytG0Gh/u677+LTTz/Fs88+i759+8LJycmYcRERERERGVW5srJx7KgfP34czz33HFauXGnMeIiIiIiI6oWiQmW2J74AgOjOeaVSiT59+hgzFiIiIiKieqNQqmBrxjvqogt1b29v3Llzx4ihEBERERHVn/xiJezMuEdddKE+d+5cJCcn46effjJmPERERERE9eKPIgXUNZxiaC5EN+Xs3bsX7du3R2RkJPz8/NC5c2dYPHLmpEQiweLFiw0eJBERERGRoUkkgFNzW1OHUSvRhXpKSorm38+fP4/z58/rrGGhTkREREQNQUWlGiq1ACfHRlCoX7161ZhxEBERERHVmzJlJQCgmW0jOPWFiIiIiKixKFNUFeqN4hz1aqWlpbh48SJyc3PRr18/tGnTxhhxEREREREZTblCBaAR7ajv3r0bAwcOxJQpUzB//nz88ssvAIC8vDz07NkTn3zyiVGCJCIiIiIypNLqHfXGUKgfPXoUCxcuRJ8+fbBo0SIIDx1l07p1azzzzDP44osvjBIkEREREZEhlSsaUY/61q1b0adPH2zYsAFDhw7Vmff29tbssBMRERERmbPqh0ntbM23R110oX79+nUMGzas1nmpVIq8vDyDBEVEREREZExl/+tRt28MO+oWFhZQq9W1zufk5MDe3t4gQRERERERGVP1qS/2No2gUPf09MR3331X45xarUZaWhp69uxpsMCIiIiIiIylTFEJK0sLWFuZ72nloiObNGkSvvnmG6xZswYFBQUAAEEQcOvWLbzyyiu4ceMGoqOjjRYoEREREZGh3M0tMesz1AE9zlEPCQnBtWvXsHHjRmzatAkA8I9//AOCIEAQBMyaNQvPPvus0QIlIiIiIjIUC4kEygqVqcOok15NOXFxcRg+fDgOHTqEW7duQRAEuLm5YezYsWx7ISIiIqIGo0xRCff2jqYOo056d897eXnBy8vLIDfPycnBzp07cenSJVy+fBmlpaXYuXMn+vTpo7VuyJAhuHv3rs71sbGxmDdvntZYYWEhVqxYgePHj6O8vBw+Pj5ISEhAjx49DBIzERERETV8pYpKtG5hZ+ow6mTSx1xv376NzZs3w83NDTKZDBcuXKh1rZeXF1566SWtMQ8PD62f1Wo1Xn75ZVy/fh1TpkyBs7Mzdu/ejejoaCQnJ8PV1dUo34OIiIiIGpbS8kp0bmu+J74AehTqCQkJj10jkUiwePFi0Tf38vLCDz/8AGdnZ3zxxReYOXNmrWvbt2+PsWPH1vl5aWlpuHDhAjZs2ICgoCAAwMiRIzFixAisX78ey5cvFx0bERERETVepYpKs34rKaBHoZ6SkvLYNfoW6s2bNxe9FgCUSiVUKlWt57UfPXoUbdu21XpzaqtWrTBy5Eh89tlnqKiogLW1tV73JCIiIqLGRa0WUKaoRDO7RlKoX716VWdMpVIhMzMT27Ztw/Xr17FlyxaDBvew77//Hn5+flCpVOjcuTNiY2MRGRmptSYjIwNeXl6QSCRa4z179sTevXvx22+/4YknnjBajERERERk/uQFZQCqTn4xZ3/rhHdLS0u4u7tj4cKFcHJywooVKwwVlxYPDw/861//wrp167Bo0SI4Ozvj3//+t+aYyGpyuRxt27bVub56LCcnxyjxEREREVHDUVpe9VZS18Z26kttnnnmGaxfvx5vv/22oT5SY+PGjVo/P//885g4cSL++9//YsKECXB0rEpyeXk5bGxsdK6vHisvL9frvq1b69eaY0hSqXn/4pgT5ko85ko/zJd4zJV4zJV+mC/xmCvxrGyrWqFd2rcw67wZrFDPz89HaWmpoT6uTpaWlnjppZcQFxeHCxcuYODAgQAAOzs7KJVKnfXVY3Z2+h3Bk5dXDLVa+PsB60kqdYRcXlTv922ImCvxmCv9MF/iMVfiMVf6Yb7EY67Ek0odkX2vEACgLK8wed4sLCS1bg7/7UK9sLAQJ0+exI4dOwx2vroY7du3BwAUFBRoxqRSaY3tLdVjNbXFEBEREVHTkiUvBgA0bywPk3p6euo8pFlNEAS0bNkS8fHxBgvscTIzMwFUnepSzdPTExcuXIAgCFqxpqeno1mzZjxHnYiIiIg0D5E6Oui2TJsT0YV6aGhojYW6k5MT3N3d8dxzz+l93KIY+fn5aNGiBSws/nzuVaFQYOvWrXBwcICfn59mPDg4GEePHsWJEyc056g/ePAAaWlpGDp0KI9mJCIiIiKUKSphb2tl9qe+iC7Uly5dapQA/vvf/wIAbt68CQBITU3FuXPn0KJFC0yaNAlffvklNm7ciBEjRsDFxQX5+flISUnBnTt38NZbb8HBwUHzWSNGjICfnx9ee+01zZtJP/74Y6jVavzrX/8ySvxERERE1LBUvezI0tRhPJbJG3PWrl2r9fP+/fsBAC4uLpg0aRI8PDzQtWtXpKam4sGDB7CxsYGXlxfi4+MxePBgrWstLS2xadMmLF++HElJSVAoFOjZsyeWLVsGNze3evtORERERGS+qnbUzb/TQnShfvbs2b90g969e9c5f+3atTrnvb29dY5nrEvLli3x7rvv4t133xV9DRERERE1HWWNbUc9Ojq61odJa1L9QGdGRsZfCoyIiIiIyBhKFZVwbm5r6jAeS3ShvnjxYnz00Ue4ffs2Ro8ejW7dugEAbty4gc8++wzu7u6YNGmSXsU8EREREVF9K1NUomMbh8cvNDHRhXpRUZHmBJV27dppzc2YMQORkZEoKSlBdHS0wYMkIiIiIjIElUoNeX45vLu2NnUoj2Xx+CVVdu3ahcjISJ0iHah6+VBkZCQ++ugjgwZHRERERGRI9x6UAgCa2Zr8TJXHEl2o//7777C3t691vlmzZvj9998NEhQRERERkTGcOPsbAEDW2cnEkTye6EK9U6dOOHjwIBQKhc6cQqHAgQMH4OLiYtDgiIiIiIgM6YfLVRvLnm7OJo7k8UTv+U+ePBn//ve/ER4ejokTJ6JLly4AgFu3bmH37t24efMm3n77baMFSkRERET0d2XeL4ZbO0dYWYrerzYZ0YX6+PHjUVpairVr1+Ltt9/WnO4iCALs7Ozw2muvYfz48UYLlIiIiIjo78iSFwMAnnBpYeJIxNGriz4mJgbjxo3Dt99+i6ysLABA586d0b9/f7Ro0TC+MBERERE1Tb9k5gMAenu2NXEk4uj9uKujoyNCQkKMEQsRERERkdGUKioBAO7tG8YGs96FelZWFk6dOoXc3FyMHj0anTp1glKpRG5uLtq0aQMbGxtjxElERERE9LfcyCqArY0lbG0sTR2KKHoV6itWrMCHH34IlUoFiUQCPz8/TaH+3HPP4ZVXXkFMTIyRQiUiIiIi+uvu3CuCrXXDKNIBPY5n3LNnD7Zu3YqJEydi27ZtEARBM9e8eXMMGTIEX331lVGCJCIiIiL6Oyoq1SgoUaJbJ/M/P72a6B313bt3Y9iwYXjjjTfwxx9/6MzLZDKcPXvWoMERERERERnClTsPAAA9u7UxcSTiid5Rv3PnDvr161frvLOzc40FPBERERGRqWXnlgAABvh2NHEk4oku1G1tbVFWVlbrfHZ2No9oJCIiIiKz9F3675AAaNeqmalDEU10oe7j44Pjx4/XOKdQKJCamgp/f3+DBUZEREREZCiVKjWa2VlpXtrZEIgu1KdOnYqLFy/i1VdfxbVr1wAAubm5+PbbbxEdHY379+9jypQpRguUiIiIiOivyissx9M92pk6DL2Ifpi0X79+eOutt/Duu+/is88+AwC89tprAABra2u888476NWrl3GiJCIiIiL6i37PK4EgAHYN5Pz0anqdox4ZGYkhQ4YgLS0Nt27dgiAIcHd3x8iRI9GuXcP6C4WIiIiImoZz1+QAgCe7tDJxJPrR+82kUqkU0dHRxoiFiIiIiMjgvrmUDQDo4eps4kj0o3ehXq2yshLp6em4f/8+unXrhu7duxsyLiIiIiKiv02hVCG3oBxtne1hYdFwHiQFHlOonz59GsePH8c///lPtG7dWjOemZmJmTNn4pdfftGMhYaGYsmSJcaLlIiIiIhIT7d/LwQABAV0MnEk+qvz1JeUlBR8++23WkU6ACQkJOD69evo1asXYmJi0K1bNxw4cAApKSl63TwnJweJiYmIjo5Gr169IJPJcPr06RrXnjhxAmFhYejZsycGDRqE9evXo7KyUmddYWEhFixYgL59+8LPzw8vvvgiMjIy9IqLiIiIiBqHO/eKAADdOrU0cST6q7NQT09Px4ABA7TGbt68iR9//BG9e/fG7t27MX/+fOzbtw9ubm44cOCAXje/ffs2Nm/ejPv370Mmk9W67uuvv8bMmTPRsmVLLFiwAEFBQdiwYYPODr5arcbLL7+Mw4cPY9KkSXj11VeRl5eH6Oho/Pbb0OXHDQAAIABJREFUb3rFRkREREQN38Vfqh4k7dy2uYkj0V+drS+5ublwc3PTGjtz5gwkEgnCw8M1Y3Z2dhg1ahQ++ugjvW7u5eWFH374Ac7Ozvjiiy8wc+bMGtctX74cTz75JLZu3QpLy6pjdRwcHLBp0yZER0fD3d0dAJCWloYLFy5gw4YNCAoKAgCMHDkSI0aMwPr167F8+XK94iMiIiKihu3XnGK4tHGApYXo1weZjTojViqVsLOz0xr76aefAABPP/201niHDh1QXFys182bN28OZ+e6n769ceMGbty4gcjISE2RDgATJ06EWq3GsWPHNGNHjx5F27ZtMXToUM1Yq1atMHLkSHzxxReoqKjQKz4iIiIiarhKyyuhUKrQsY2DqUP5S+os1Dt06KD1wCgAnDt3Dq1bt0aHDh20xsvLy+Ho6GjwAH/++WcAgLe3t9Z4u3bt0L59e808AGRkZMDLy0vn1bA9e/ZESUkJ21+IiIiImpAvzmUCALwb2Pnp1eos1J966imkpv7/9u48Lqpy/wP4h2GVHXREBVFAZ1BEQE1Fc8UUyY3rvqBet9SbP80y7aa3ulqakb8MzdJSi9wTJa9LqZX+rutVU1QQA0VFZZN9mwHm/P6gOddxAGcQmAE+79eL16t5zjPPec6X4/TlzPc8Jxq3b98GABw/fhz37t1Dnz59tPrGx8fXykOP0tPL64qkUqnWNqlUirS0NI2+zZs31+qnbnu6LxERERE1bFdul+eRgZ1aGHgm1VNljfqcOXNw6NAhjBw5Eo6OjsjOzoa5uTlmzJih0a+srAy//PILhgwZUuMTLC4uBgBYWFhobbO0tERRUZFG34r6qdvUY+mqaVPD3XQgldb8txMNFWOlO8ZKP4yX7hgr3TFW+mG8dMdYaUp6nIv7qflwb2GHli00V3ypL7GqMlFv3bo1IiMjsXHjRty7dw+dO3fGvHnztB5udOHCBTg5OWnUhtcUdY28UqnU2qZQKDRq6K2srCrsp257tt7+eZ48yYdKJej1npogldohPT2vzvdbHzFWumOs9MN46Y6x0h1jpR/GS3eMlbZvD90AAIzt56URG2OLlURiUunF4ec+mdTX1xdffvlllX169eqFQ4cOVW92z6EueamorCU9PR0BAQEafSsqb1G3VVQWQ0REREQNz6X48rKXjm2rXrjEmBn9OjUdOnQAANy4cUOjPTU1FSkpKeJ2APD29sbNmzchCJpXwWNiYmBtbQ13d/fanzARERERGVRadnlp9IAAV61FRuoTo0/U27dvD09PT+zZswdlZWVi+65duyCRSDB48GCxLTg4GGlpaTh58qTYlpmZiWPHjiEoKAjm5uZ1OnciIiIiqnvfHbsFAPBr19TAM3kxzy19qW1ffPEFgPInngJAdHQ0Ll++DHt7e0yZMgUA8Pbbb2PevHmYOXMmQkJCcPv2bezYsQPjx4+Hh4eHONaQIUPg7++Pt99+GzNmzICTkxN27doFlUqFBQsW1P3BEREREVGdyi8qQWxSFgDA17N+J+omwrN1InVMLpdX2O7q6opffvlFfH3ixAls2LABiYmJcHZ2xujRozF//nyYmWn+rZGTk4O1a9fixIkTUCgU8PX1xbJly+Dj46P33HgzqfFjrHTHWOmH8dIdY6U7xko/jJfuGKv/+njHFcQ/yMb4ge0wpLt22bOxxeqFbiatbfHx8Tr1GzRoEAYNGvTcfg4ODvjwww/x4YcfvujUiIiIiKgeKSguQfyDbADAKy+1NvBsXpzR16gTEREREeliz8kEAMD4ge0gqcc3kaoxUSciIiKieu+P5Gz8+/pjAMDgBnA1HWCiTkREREQNwGf7YgAAr//Ft14vyfg0JupEREREVK+du5mCIkUpmjs1QReZ1NDTqTFM1ImIiIio3lIoy7DlUCwA4K0J/gaeTc1iok5ERERE9dane64CAPr6tUIzhyYGnk3NYqJORERERPVScno+Eh7mAACmDqn42Tz1GRN1IiIiIqp3FCVl+Mc3FwEAyyZ3gUTSMG4gfRoTdSIiIiKqdz7dXV7yEtC+GWStHQ08m9rBRJ2IiIiI6pXM3GKx5OVvob4Gnk3tYaJORERERPVGSakK7359AQDwP2M6N8iSFzUm6kRERERUL5SWqfDP7f+BQlkGJztL+Hk1NfSUapWZoSdARERERPQ8ipIy/H3zeWTlKQAAH83p2WCeQFoZJupEREREZNRyCpR4I+LfAICm9pZYNbsnLM1NDTyr2sdEnYiIiIiMVkmpCku+OAMA6OTpjNdDfWHRCJJ0gIk6ERERERmpklIV3tl8DqVlAtyktnhjrF+DL3d5GhN1IiIiIjIqpWUqXL/zBBH7rwMAHGwtsHxq10aVpANM1ImIiIjIiKhUAj7ecQWJj3IBAB3bOuFvjajc5WlM1ImIiIjIKChKyrDi6wvIyCkGAKyc2R2tmtk0uivpakzUiYiIiMigBEHA9TtP8Nm+GACAo60FVs7qARsrcwPPzLCYqBMRERGRwagEAZ//EIOYxCcAAG93Rywc69coll98HibqRERERFTnSkrLcOZGCvacTICipAwA8O7UrvBsad9oS12exUSdiIiIiOqMSiXgakIGNkRdF9v6dG6JcQPbNfpSl2fVi0T9woULmDp1aoXbjhw5Ai8vL/H1lStX8MknnyA2Nha2trYYOnQo3nzzTTRp0qSupktEREREFbgcn4YD/3cXjzIKAACerewxcVB7eLVyMPDMjFO9SNTVpk2bBh8fH402FxcX8b/j4uIwffp0tGvXDsuWLUNKSgq2bt2K5ORkfPnll3U9XSIiIqJGTxAEPEjLx6bom0jNLAQANHOwwv+M7gxXaeNd0UUX9SpR7969OwYNGlTp9nXr1sHR0RGRkZGwsbEBALi5uWH58uU4d+4cAgMD62qqRERERI2asqQMv/+RgROXHohrordxscOEoHaQuzsZeHb1Q71K1AEgPz8fVlZWMDMz02o/e/YsZs6cKSbpADBy5Eh89NFHOHr0KBN1IiIiolomCAIO/t9d/Hb1IfIKS8T2+aM6oatcyivoeqhXifqSJUtQWFgIMzMz9OjRA0uXLoVcLgcAxMfHo7S0FJ06ddJ4j4WFBTp06IC4uDhDTJmIiIiowRMEAVl5Cpy+9gjHLz1AkaJ8FZfWzW0xP7QTHG0tudxiNdSLRN3c3BxDhgxB37594eTkhPj4eGzduhWTJk3CDz/8AA8PD6SnpwMApFKp1vulUimuXr1a19MmIiIiatBKy1S4HJ+Ok5eTkfAwR2zv598KoX08YW9jYcDZ1X/1IlHv0qULunTpIr4OCgrCwIEDMXr0aGzYsAGffvopiovLHzVrYaF9QlhaWorb9dG0qW31J/2CpFI7g+27vmGsdMdY6Yfx0h1jpTvGSj+Ml+7qMlYZ2UXY+dMt/H47HRnZRWL7wvEBkLdxQmsX4/691Zfzql4k6hXx9vZGYGAgzp8/DwCwsrICACiVSq2+CoVC3K6PJ0/yoVIJLzbRapBK7ZCenlfn+62PGCvdMVb6Ybx0x1jpjrHSD+Olu9qOVWFxKfKLlDgfm4pTVx8hK08BALC3Noer1AZ/C/WFk91/y1uM+fdmbOeVRGJS6cXhepuoA0DLli3FRF1d8qIugXlaeno6mjdvXqdzIyIiIqrvfv8jHTkFSnx3LF6j/eXOLeHi1ASvBrY1zMQaiXqdqD948ABOTuXL+8hkMpiZmeHGjRsYPHiw2EepVCIuLg7Dhw831DSJiIiI6o3C4hLs/TURKU8KcDv5v3Xn/u2aoZu3FO4udnCTGq48uDGpF4l6ZmYmnJ2dNdouXbqECxcuYNSoUQAAOzs7BAYGIjo6Gq+99pq4RGN0dDQKCwsRHBxc5/MmIiIiMmY5+QoUKkpx6342fjxzFxCAnILyMmIzUwlcnK3x16HecHG2hgNvDK1z9SJRX7RoEZo0aYKAgAA4OTnhjz/+wJ49e+Dk5IQFCxaI/d544w1MmDABYWFhGDt2LFJSUrBt2zb07dsXvXr1MuAREBERERmHuHtZSM0sRFaeAofOJmlse7lzS0hMTGBnbY7QPp6QSLjmuSHVi0R90KBBOHToELZt24b8/Hw4Oztj2LBhWLBgAVq1aiX28/HxwbZt2xAeHo7Vq1fD1tYW48aNw+LFiw04eyIiIiLDyspTYP+pRChKynA5XvN+vlcD28BVagMXJ2t4tLQ30AypIvUiUZ86dSqmTp2qU99u3bph9+7dtTwjIiIiIuOTW6BEXlH500APnr6D28nZMAGQ++cTQu2tzdGyqTVC+3iinZsDLMwksLYyN+CMqSr1IlEnIiIiIk0qlYBzN1NQrCyDna0lnmQVYt9viVr9+ge4AgCa2lsipGcbmJiwnKW+YKJOREREVA/kFiqx52QCSkrLAAD3U/OR9tTDhtT6+7eCd5vyVfHk7k68CbQeY6JOREREZCQyc4tRWFwKAMjKV2DzjzcBACYmJsj/s6TFxsoMDraWMDOToE0LO8wb6QO3Vo7IeJIPM4kE1lZM7xoK/iaJiIiIDEChLMO5mykoKVMBALLzFTh6/r5Wvw5tnNCyqTUAwMHWEsMCtctXHGwtoSzSfjo71W9M1ImIiIhqWdy9LJy8nAxBEMS22KQsKErKtPqO6N1WfKCQtZUZOrZ11upDjQMTdSIiIqJqeJJTjILiEq32f51Nws2kLDx9zbtQUV7O8vQTPaWOTWBvY445I3wg+fMKubmpBJYWprU6b6o/mKgTERERVUKlEvDv649R/GeirZZToMTRC9plKk8b1NVN43WHtk4IaC+t8TlSw8VEnYiIiBq9O49ycfhcEp6qTAEAPEjLw5NcRaXvG9G7Ldxd7LTavVrZw8HWsoZnSY0NE3UiIiJqEARBwOMnhVCWatd9FyvKEBEVA5UKQAXLiCuU5e9xd7HVaLdpYg4neyvMHeEDKwvNtMnU1ASW5ixTodrDRJ2IiIiMXmpmIX7/I6PKPgkPc3DldnqVfTxb2aOdq0Ol27p3cKn2HIlqGhN1IiIiMqjk9HxEnboD1TN1JxYWZlAqy2vDYxKf6DzezFc7wMbKXKvdwlwC7zZO4o2bRMaOiToRERFVS0FxCVIyC6vsk5WrwOZDNwGYQFJJfqwsLV9H3N3FViOJNlOWofTP5QvbtrBDZ6+mCO7hXuX+zEwlMDOV6H4QREaMiToRERGJihSl+L9rj1CqEp7b94ffEnUe18/LGS2b2VS63bWZDXr7ttRok0rtkJ6ep/M+iBoaJupEREQN1KVbaTh97ZFe74lNytIqQalKhzZOGNK96qvc1lZmldaFE1HlmKgTERHVsZwCJdKzivR+3+ZDN5GVp4CuJdalZeUJt2cre5330balHZztrTDz1Q4VLY6ixYKrnhDVGibqRERE1XQ5Pg2PnlRdo/00GxsLFBQoceD0nWrv097GAi8/UyJSlc5eTSFr7Vjt/RGR4TBRJyKiRufk5WT8/kfVy/g9j0ol4Nb97Gq/v6tcin7+rfR6j8TEBO1cHXgVm6iRYKJOREQvRBAE3EvJg6JE+yEzL0pZWob1+2KgUgkwqcEl9dQ12C9aNy1r7YhxA9ppPSSnMk/fHMmVSYjoeZioExE1cOdupCA9W/96aF09yizCxdiUWhsfKK+x7tjWqQZHNEGPDs3hKtUtwa4pXDqQiPTBRJ2IyADOXH+MM9cf1/p+ipVlSEqpm+XtXhvhAztr7YfMvChzMwm8XB34kBoianSYqBOR0Xj8pAA5+co63+++3xJx93Fune8XQK3f5GdhboqObZ0w+RUZXJyta2Uf0mZ2yMjIq9HSFCIiYqJORH+6n5qHK7df7Oa6F6EoKcNPFx8YbP+mEhO8GthGfG1tbYnCQkWt7tOvXTN4tNR92TxjJZGYMEknIqoFTNSJ/iQIAvR4xkeNu5uSi/2/JUKlw9MAq8PcwgwlytJKt99OzqmV/epr1MseBllKrk0LOzSx/O9HIp+ISEREhsZEnYxKaZkKiQ9zUKZjsuqYVYzsHN3XMK7KlkOxyCmo+7KLZ8lbO+r8MBN9mEpMUCapfGBvd0f06OiCfv6uNb9zIiIi0luDS9SVSiXWr1+P6Oho5ObmwtvbG2+88QYCAwMNPbV6Ky4pE3H3s+pkX+dvpiIjp7hO9lURJztLvdc1rkmuzWzRVS6tlbF5hZiIiKh+aXCJ+rJly/Dzzz9j6tSpaNOmDQ4cOIDZs2cjMjISAQEBhp7ec6kEASWlKpSWqfR+77/OJiE2qeYT6oSH5SURdbHigiAIsLEyw4LRnXXq7+hojezsmrmibmICtG1hD3MzLp1GREREhtegEvWYmBgcPnwY77zzDqZPnw4AGDVqFIYNG4bw8HDs2LHDsBPUwYqvL+CxHo+jrkjNrjUM+LR1QlC31vBv16xGx60J5VeJLQw9DSIiIqIa16AS9WPHjsHc3Bxjx44V2ywtLTFmzBj87//+L9LS0tC8eXMDzrBqZSoVHj8phKvUBj06uOj9fhMT4KUOLmju2KQWZkdEREREdalBJepxcXHw8PCAjY2NRnvnzp0hCALi4uKMOlE3lUgQPr8XvNo2RVZmgaGnQ0REREQG1KCKcdPT0ytMxKXS8pvz0tLS6npKenO2t+LjpYmIiIioYV1RLy4uhrm59uOrLS0tAQAKhX4PL2na1LZG5lUdUqmdwfZd3zBWumOs9MN46Y6x0h1jpR/GS3eMle7qS6waVKJuZWWFkpISrXZ1gq5O2HX15El+rT18pipcRk93jJXuGCv9MF66Y6x0x1jph/HSHWOlO2OLlURiUunF4QZVYyGVSissb0lPL38sujHXpxMRERERPa1BJere3t64e/cuCgo0b8S8du2auJ2IiIiIqD5oUIl6cHAwSkpKsG/fPrFNqVQiKioKXbp0gYuL/kseEhEREREZQoOqUffz80NwcDDCw8ORnp4Od3d3HDhwAI8ePcLq1asNPT0iIiIiIp01qEQdANauXYvPPvsM0dHRyMnJgVwux+bNm9G1a1dDT42IiIiISGcNLlG3tLTE0qVLsXTpUkNPhYiIiIio2hpUjToRERERUUPBRJ2IiIiIyAgxUSciIiIiMkINrka9JkkkJo1y3/UNY6U7xko/jJfuGCvdMVb6Ybx0x1jpzphiVdVcTARBEOpwLkREREREpAOWvhARERERGSEm6kRERERERoiJOhERERGREWKiTkRERERkhJioExEREREZISbqRERERERGiIk6EREREZERYqJORERERGSEmKgTERERERkhJupEREREREaIibqRUCqV+OSTT/Dyyy+jc+fOGDduHM6dO2foadW4mJgYfPDBBwgJCYG/vz/69++PN954A/fu3dPqe+XKFUycOBF+fn7o3bs3Vq1ahaKiIq1++sRO1zGN1ZYtWyCXyzFy5EitbYxXuZiYGMyZMwcvvfQSAgICMGLECERFRWn0OXnyJEJDQ+Hr64v+/ftjw4YNKC0t1RorNzcXK1asQM+ePeHv74+pU6ciLi6uwv3qOqaxSEpKwqJFi9C3b1/4+/sjJCQEmzdvhlKp1OjX2M6rtLQ0hIeHIywsDAEBAZDL5bhw4UKFfQ15HukzZm3RJVZZWVn4+uuvMWnSJPTs2RPdunXD+PHjcfTo0QrHbKixAvQ7t9QePnwIPz8/yOXyCufcUOOlT6zy8vKwZs0aDBgwAJ06dUK/fv2wePFirX6pqalYuHAhunXrhi5dumD+/Pl48OBBhWPu27cPQ4cOha+vL4YMGYIdO3ZU2E+fMavL9P3333+/RkekalmyZAmioqIwbtw4DB8+HPHx8fjmm28QGBiIli1bGnp6NebDDz/EmTNnMGDAAISGhsLDwwPHjh1DZGQkXnnlFTg7OwMA4uLiMGXKFDg4OOC1116Du7s7vv/+e8TGxmLYsGEaY+oaO33GNEbp6elYuHAhzM3N4eDggIkTJ4rbGK9yp06dwsyZM9GyZUtMnDgRffv2hZ2dHZRKJbp37y72mTdvHtq1a4dZs2bBwcEB33zzDXJyctCvXz9xLJVKhenTp+P8+fOYNm0agoKCcPHiRURGRiI4OBgODg4a+9VlTGORmpqK0NBQZGdnY9KkSRg0aBBKS0uxfft2PHz4EIMHDwbQOM+rGzduYMWKFTAzM0Pr1q2RkpKC0NBQuLm5afQz5Hmkz5i1SZdYnT9/Hu+99x7kcjlGjRqF3r17IykpCd988w0kEon477KhxwrQ/dx62vLly/HgwQOUlJRgwoQJkEql4raGHC9dY5Wbm4vJkyfj5s2bGDNmDIYNGwZPT0/cuXNH44JWQUEBJkyYgHv37mHWrFkIDAzE8ePHcfDgQYSGhsLKykrsu3v3bvzjH/9Ajx49MGXKFKhUKmzevBk2NjYICAio1pgvRCCDu3btmiCTyYRt27aJbcXFxcKgQYOESZMmGW5iteDy5cuCQqHQaLt7967QqVMnYenSpWLbrFmzhD59+gj5+fli2969ewWZTCacPXtWbNMndrqOaayWLl0qhIWFCVOmTBFGjBihsY3xEoTc3FwhMDBQWLlyZZX9QkJChNDQUKG0tFRsW7duneDt7S3cvXtXbDt8+LAgk8mE48ePi21PnjwRunXrJixZsqRaYxqLr776SpDJZMLt27c12hcsWCB07NhRUCqVgiA0zvMqLy9PyMzMFARBEI4fPy7IZDLh/PnzWv0MeR7pM2Zt0iVW9+/fF5KTkzXaVCqVMHXqVKFz585CUVGR2N6QYyUIup9baufPnxd8fHyEdevWCTKZTIiNjdXY3pDjpWusVqxYIQwcOFDsW5nNmzcLcrlcuHnzptiWkJAgdOjQQfjss8/EtqKiIqF79+7CvHnzNN7/5ptvCgEBAUJubq7eY74olr4YgWPHjsHc3Bxjx44V2ywtLTFmzBhcvnwZaWlpBpxdzerSpQssLCw02tq2bYv27dsjMTERAJCfn4+zZ89i1KhRsLGxEfuNHDkS1tbWGl+Z6ho7fcY0RjExMfjxxx/xzjvvaG1jvModOnQIubm5WLhwIYDyYxAEQaNPQkICEhISMH78eJiamortkyZNgkqlws8//yy2/fTTT2jevDmCgoLENmdnZwwdOhQnTpxASUmJ3mMai4KCAgBA06ZNNdqbNWsGMzMzmJqaNtrzytbWFk5OTlX2MfR5pOuYtU2XWLVu3Rqurq4abSYmJhg0aBCKi4vx8OFDsb0hxwrQLV5qZWVl+PDDDzFlyhS0adOmwj4NOV66xCo3NxcHDhzAzJkz4eTkBIVCoVW6p/bTTz/B398fHTt2FNu8vLwQGBio8blz4cIF8ZvGp02ePBkFBQU4ffq03mO+KCbqRiAuLg4eHh4a/+MCgM6dO0MQBIPU0tUlQRCQkZEh/qOMj49HaWkpOnXqpNHPwsICHTp00IiHrrHTZ0xjIwgCVq5ciVGjRqFDhw5a2xmvcufOnYOnpydOnTqFfv36oWvXrujevTvCw8NRVlYGAIiNjQUAreNycXFBixYtxO1Aeax8fHxgYmKi0dfX1xcFBQW4f/++3mMai5deegkA8O677+LWrVt4/PgxfvzxRxw4cACzZ8+GRCLheVUFQ59Huo5pzDIyMgBAIxljrP5r9+7dSE1Nxfz58yvt09jjdenSJSiVSjRr1gzTp0+Hn58f/P39MWPGDI15qlQqxMfHax0/UH5cSUlJ4j0ylcXKx8cHEolE3K7PmC+KiboRSE9PR/PmzbXa1bVoDemKekV+/PFHpKamYujQoQDK4wFAoxZPTSqVasRD19jpM6axOXjwIBISErBo0aIKtzNe5e7du4eUlBQsW7YMoaGhiIiIwKBBg7BlyxasWbMGQM3ESt1Wn2P18ssvY+HChTh79ixGjhyJ/v37Y8mSJZg1axZef/11ADyvqmLo80jXMY1VdnY29u3bh+7du4v3JQGMlVp2djY+//xzLFiwAPb29pX2a+zxUifjK1asgKmpKdatW4e3334bMTExmDZtGvLz8wGUx1OpVFZ6/IIgiDFKT0+HhYUFHB0dNfqp29THr8+YL8qsRkahF1JcXAxzc3OtdktLSwCAQqGo6ynVmcTERPzzn/9E165dxRs/iouLAUCrRAYoj4l6u7qvLrHTZ0xjkp+fj08//RRz5syp8MMTYLzUCgsLkZOTgzfffBNz5swBAAwePBiFhYXYtWsX5s2b99zjevoKSHFxcYX91G3qsfQZ05i4ubmhe/fueOWVV+Do6IjffvsNERERcHZ2xsSJE3leVcHQ55GuYxojlUqFt956C3l5eVi+fLnGNsaq3Oeffw5nZ2dMmDChyn6NPV7qEj6pVIotW7ZAIim/9uzh4YE5c+Zg//79mDZtmvj5U9nxA5qxquizTN1XPZY+Y74oJupGwMrKqsK6L/WJoP6lNzTp6el47bXX4ODggPXr14v/yNR3SldUa6ZQKDTupNY1dvqMaUw2bdoEc3Nz/PWvf620D+NVTj2nZ1cOGT58OI4dO4br16/rHauK+qnb1H3rY6wOHz6M9957D8eOHYOLiwuA8j9qBEHA2rVrERISwvOqCoY+j3Qd0xitXLkS//73vxEeHg65XK6xjbECbt++jd27d2PTpk0wM6s6RWvs8VLPJTg4WMwfAKBfv35wcHDAlStXMG3aNPHzp7Ljf3qsyo5f3Vc9lj5jviiWvhiByr7yVX9tUtmV1PosLy8Ps2fPRl5eHr7++muNr4/U/13R10bPfi2na+z0GdNYpKWl4dtvv8WkSZOQkZGB5ORkJCcnQ6FQoKSkBMnJycjJyWG8/qSec7NmzTTa1a9rKlbqtvocq507d8LHx0dM0tUGDhyIwsJC3Lp1i+dVFQx9Huk6prHZsGEDdu7ciSVLllS4FCdjBaxbtw4dO3aEl5eX+JmflZUFoHzOjx8/Fvs29nhV9pkPlN8Am5ubCwBwdHSEhYVFpcdvYmIijiWVSlF5RLs+AAAQGUlEQVRSUoLs7GyNfkqlEtnZ2eLx6zPmi2KibgS8vb1x9+5d8WsctWvXronbGxKFQoG5c+ciKSkJX331FTw9PTW2y2QymJmZ4caNGxrtSqUScXFxGjdU6ho7fcY0Fk+ePEFJSQnCw8MRFBQk/ly7dg2JiYkICgrCli1bGK8/+fj4AChfI/xpKSkpAMo/uNXzfva4UlNTkZKSohWrmzdvaq0cExMTA2tra7i7uwOAXmMai4yMDPEG26epr4qXlZXxvKqCoc8jXcc0Jjt27EBERASmT5+OmTNnVtiHsQIeP36M69eva3zmr127FgAwZ84cjBkzRuzb2ONV2We+SqVCenq6eP+DRCKBTCbTOn6g/LjatGmDJk2aAKg8Vjdu3IBKpRK36zPmi2KibgSCg4NRUlKCffv2iW1KpRJRUVHo0qWL1lWv+qysrAyLFi3C1atXsX79evj7+2v1sbOzQ2BgIKKjozX+xx8dHY3CwkIEBweLbbrGTp8xjYWbmxs2btyo9dO+fXu4urpi48aNGDVqFOP1J/WcfvjhB7FNEATs27cP1tbW8Pf3R/v27eHp6Yk9e/ZoJKq7du2CRCIRH/SjHi8tLQ0nT54U2zIzM3Hs2DEEBQWJdYz6jGksPDw8cOPGDa0VHA4fPgxTU1PI5XKeV1Uw9Hmk65jG4siRI1i1ahWGDx+OZcuWVdqPsQLeeecdrc/8sLAwcdvq1avFvo09Xl5eXpDJZDh06JDGvXxHjhxBfn4+AgMDxbYhQ4bg6tWrGivc3LlzB+fPn9f43OnZsyccHR2xc+dOjX3t2rUL1tbW6Nu3r95jvig+mdQItGjRAgkJCdixYwcKCgqQnJyM1atXIzExEZ988glatWpl6CnWmNWrV+PgwYPo168fWrdujfj4ePEnOTlZvLru5eWFyMhInDp1CiqVCidOnMD69evRu3dv/O1vfxPH0yd2uo5pLCwtLeHp6an1o16f9d133xWvGDBe5V/JJicnY8eOHUhJSUFKSgo2btyI06dPY9GiRejZsycAwNXVFdu3b8eVK1egVCpx4MABbNu2DePHj0doaKg4nqenJ86cOYM9e/agpKQEf/zxB1auXIm8vDysW7dOY1UAXcc0Fi4uLoiKisLhw4ehUChw+/ZtRERE4Ndff8X48eMREhICoPGeV1988QX+85//4OLFi7h9+7a4XGV8fDw6d+4MwLDnkT5jGjpWMTExmDt3LhwcHDB+/Hj88ccfGp/7zZo1g7W1daOIlS7xcnNz0/rMz8jIwMmTJ/HWW2+hW7du1Tq2+hgvXf4dtmnTBjt37sTp06ehUChw4sQJrFu3DnK5HMuXLxfXjZfL5Th69CgOHDgAQRAQExODDz74ANbW1lizZo149dvMzAzW1tbYvn07EhISkJ+fj++++w7R0dFYuHAhevXqJc5P1zFflInw7PcbZBAKhQKfffYZDh06hJycHMjlcixevFjjpGgIwsLCcPHixQq3ubq64pdffhFfX7p0CeHh4YiNjYWtrS1CQkKwePFi8UNdTZ/Y6TqmMQsLC0Nubi6io6M12hmv8qu4X3zxBQ4ePIiMjAy4ublh+vTpWqsnnDhxAhs2bEBiYiKcnZ0xevRozJ8/X+vmrZycHKxduxYnTpyAQqGAr68vli1bJn7lWp0xjUVMTAwiIiIQFxeH7OxsuLq6YvTo0Zg5c6bGQ1Ea43n17E2Oas9+RhnyPNJnzNr0vFhFRUVV+KA2te+++w49evQQXzfkWAG6n1tPU8fw4MGDWuVhDTleusbq9OnTiIiIQHx8PKytrREUFIS33npL64FJKSkp+Oijj3DmzBmoVCr06NED7777Llq3bq21j71792Lr1q1ITk5Gy5YtERYWhqlTp2r102fM6mKiTkRERERkhFijTkRERERkhJioExEREREZISbqRERERERGiIk6EREREZERYqJORERERGSEmKgTERERERkhJupEREREREaIiToRUSMWFhaGgQMHGnoaBhMREQG5XI7k5GRDT4WISAsTdSIiPT148AArVqxAcHAw/Pz88NJLL2Ho0KFYunQpzp8/b+jpGYWwsDDI5XIEBQVBqVRqbVcnyNevXzfA7IiI6gfjfLY1EZGRun79OsLCwmBmZoZRo0ahXbt2KC4uxr1793DmzBnY2NigZ8+ehp6m0UhOTsauXbswbdo0Q0+FiKjeYaJORKSHjRs3oqioCNHR0fD29tbanp6eboBZGScrKyu4ublh06ZNGD16NGxtbQ09JYPLz89nHIhIZyx9ISLSQ1JSEhwdHStM0gFAKpVqvD5y5Ajmzp2L/v37o1OnTujRowfmz5+PW7duab134MCBCAsLw61btzB9+nQEBAQgMDAQa9asQWlpKRQKBT7++GP06dMHvr6+mDx5MhITEzXGiIqKglwux9mzZxEREYEBAwagU6dOGD58OA4fPqzXcS5ZsgQvv/wyOnXqhIEDB+Ljjz9GYWGhzmNIJBIsXrwYWVlZ+Prrr5/bv6p6cXVsniaXy7Fs2TKcO3cO48ePh5+fH/r27YvNmzcDAHJycvD3v/8dgYGB8PPzw2uvvYbU1NQK911UVIRVq1ahd+/e6Ny5M8aOHYtz585V2Pfs2bOYMWMGunXrBl9fXwwfPhy7du2qdM6xsbGYOXMmunbtihEjRjw3DkREaryiTkSkB3d3d9y9exc///wzBg8e/Nz+33//PRwdHTFu3DhIpVLcv38fe/fuxcSJE3HgwAG0bdtWo39KSgr++te/IiQkBEOGDMGZM2ewbds2mJqaIiEhAcXFxZgzZw6ysrKwdetWzJ8/H0ePHoVEonndJTw8HIWFhZg4cSKA8gR+8eLFUCgU+Mtf/lLlnG/cuIFp06bB3t4e48ePh4uLC27duoXIyEj8/vvviIyMhLm5uU7xCgoKQteuXfHtt99i8uTJWn/IvKjY2Fj8+uuvGDduHEaOHImjR4/i008/haWlJQ4ePAhXV1e8/vrruH//PiIjI7F06VJs375da5ylS5dCIpFg9uzZyM/Px549ezBr1ixs2bIFvXr1Evvt2bMH7733Hvz9/TF37lw0adIEZ8+exfvvv4/79+9j6dKlGuM+evQI06ZNQ3BwMAYPHqzXHzpERBCIiEhnV65cEXx8fASZTCYMHjxYWLZsmbBjxw4hISGhwv4FBQVabQkJCYKPj4/w3nvvabQPGDBAkMlkwpEjRzTaQ0NDBblcLsydO1dQqVRi+7fffivIZDLh9OnTYtv+/fsFmUwm9O/fX8jNzRXbc3Nzhf79+wsvvfSSUFRUJLZPmTJFGDBggMb+hg8fLgwZMkTIy8vTaP/5558FmUwm7N+/v5Lo/NeUKVMEf39/QRAE4fLly4JMJhNWrFghbv/8888FmUwmxMTEaLU9ePBAa7wBAwYIU6ZM0WiTyWSCXC4Xrl69KrYpFAqhd+/eglwuF1auXKnR/6OPPhJkMpmQmJiotc8xY8YICoVCbH/8+LHg7+8vBAcHi22pqalCp06dhMWLF2vNb+XKlYK3t7dw//59jTnLZDJh7969lQeKiKgKLH0hItJDQEAA9u/fj9DQUOTl5SEqKgoffPABQkJCMHnyZDx48ECjv7W1NQBAEATk5+cjMzMTTk5O8PDwQExMjNb4Li4uGDp0qEZbly5dIAgCwsLCYGJiIrZ369YNAHDv3j2tcSZOnAg7OzvxtZ2dHSZMmICcnBxcuHCh0uOLj49HfHw8hg0bBqVSiczMTPGna9eusLa2xpkzZ3SIlOb8Bw0ahP379+Pu3bt6vfd5/P394efnJ762sLCAr6+vGK+nVRWv6dOnw8LCQnzdokULDB8+HHfu3BHLi3766ScolUqMGTNGIy6ZmZkYOHAgVCoVzp49qzGuo6Pjc7/BICKqDEtfiIj0JJfLsWbNGgDAw4cP8Z///Af79u3DpUuXMH/+fOzfv19M+mJjY7F+/XpcvHhRq+zBzc1Na+yK2hwcHCrcZm9vDwDIzs7Weo+np6dWm5eXFwBUuWa4OimNiIhAREREhX0yMjIqfX9lFi9ejF9//RXr1q2rdNzqaN26tVZbdeKljk1FbQ8ePICXl5cYm+nTp1c6n2dj07p1a5iamlZxBERElWOiTkT0AlxdXeHq6oqRI0di0qRJuHLlCmJiYtCtWzc8evQIkydPhq2tLebNmwdPT080adIEJiYm+OijjyqsV64qqXu2Dl1NEIQaOx61GTNmoE+fPhVuUye8+vDy8sJf/vIX7Nu3D9euXauwz9PfFjyrtLS0wvaq4lXZturGS/2+jz/+GM2bN6+wz7N/ODRp0qRa+yIiApioExHVCBMTE/j5+eHKlStIS0sDABw/fhyFhYXYtGmT1trq2dnZGqUWNe3OnTtabeorwhVdtVdr06YNgPI/Cp6+ibImLFiwAP/6178QHh6O7t27a21XXwnPycnRmKNCoUB6ero4t9qQmJiotZKPOl7q5Ft946+Tk1ONx4aIqCKsUSci0sOZM2cqvLpbXFws1m6rSybUV3SfvYK7d+/eWl9vfdeuXcjLyxNf5+XlYffu3bC3t68wSVbr2LEjZDIZdu/erVVvD5Rf2a6odEQXLi4umDp1Ki5evIhTp05pbVcnws/WeW/fvh0qlapa+9TV9u3bNZ6gmpKSgkOHDsHDw0P8fQ4dOhQWFhaIiIhAcXGx1hh5eXkVPoWViKi6eEWdiEgPq1evRnZ2NgYOHAiZTAYrKysxqUtKSsKoUaMgl8sBAH379kWTJk3w9ttvY8qUKbC3t8eVK1dw+vRpuLu7o6ysrNbm6eTkhLFjx4o3MkZFReHRo0dYtWpVleUYJiYmWLt2LaZNm4YRI0Zg9OjRGk9fPX78OBYvXlztGyRnz56NPXv24Pr161rbevXqBQ8PD3z++efIzs6Gm5sbLl++jGvXrsHJyala+9NVWVkZJk+ejFdffRUFBQXYvXs3FAoFli9fLvZp0aIF3n//fSxfvhwhISEYMWIEXF1dkZmZidu3b+PEiRM4fPhwld9YEBHpg4k6EZEeli1bhpMnT+Ly5cv46aefkJeXBzs7O8hkMsyePVsjgXV3d8eWLVuwbt06fPnllzA1NUWXLl0QGRmJlStX4uHDh7U2z7feeguXLl3Czp07kZGRAQ8PD4SHh2P48OHPfW+HDh1w4MABfPXVV/jll1+we/du2NjYwNXVFaGhoQgMDKz2vOzs7DBv3jysXr1aa5upqSk2bdqEVatW4fvvv4e5uTl69+6N77//XlwPvrZ8/PHH2L17N7Zs2YLc3FzxhuHevXtr9Bs9ejTatm2LrVu3Ys+ePcjLy4OjoyM8PDywcOHCGl8nnogaNxOhNu5CIiIig4iKisI777yD7777Dj169DD0dIiI6AWwRp2IiIiIyAgxUSciIiIiMkJM1ImIiIiIjBBr1ImIiIiIjBCvqBMRERERGSEm6kRERERERoiJOhERERGREWKiTkRERERkhJioExEREREZISbqRERERERG6P8BNExdA9WWvbQAAAAASUVORK5CYII=\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": "49b6c7b0-93d1-4cb9-946f-3d3531cc84e1"
      },
      "source": [
        "train_samples[0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "([101, 30580, 102], 0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "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": "02cc656f-b0cc-48ae-af7f-36e2a94312f6"
      },
      "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": "e149d569-c9d6-4ce3-9781-58a26e34ddc9"
      },
      "source": [
        "print(batch_ordered_sentences[0])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[101, 16040, 29748, 30580, 11085, 10312, 10219, 11850, 11202, 13723, 16297, 13322, 10380, 88982, 11850, 11202, 30095, 10113, 38668, 102], [101, 1173, 91073, 13693, 73623, 65329, 1154, 15472, 12217, 52481, 33144, 24517, 1142, 65329, 37076, 21060, 44019, 28065, 92265, 102], [101, 144, 10410, 12361, 11143, 10138, 10539, 10758, 36627, 24960, 86668, 18969, 10840, 10921, 37013, 10116, 59411, 89957, 100, 102], [101, 1144, 39361, 54398, 1159, 18827, 59245, 12217, 46168, 39383, 19948, 54974, 1155, 12217, 77625, 49327, 49947, 74170, 119, 102], [101, 11520, 10824, 12146, 93742, 119, 10392, 13227, 60494, 94836, 10147, 28250, 10216, 24960, 24141, 13322, 93742, 136, 119, 102], [101, 32470, 15153, 10133, 11019, 10106, 11098, 172, 64666, 15050, 40310, 10138, 119, 119, 119, 48932, 10107, 17375, 89957, 102], [101, 10469, 14990, 34597, 55170, 79313, 27225, 18161, 15359, 10330, 10148, 16678, 13060, 10369, 11926, 27020, 35637, 10113, 100, 102], [101, 106583, 31510, 10123, 17255, 10465, 11084, 10132, 11760, 24960, 16419, 10115, 21602, 10483, 10312, 10219, 72894, 10109, 119, 102], [101, 12300, 87597, 10525, 13351, 10109, 69308, 12022, 12078, 11202, 104939, 10113, 86945, 10537, 10132, 10415, 17657, 87150, 13723, 102], [101, 137, 175, 54779, 10371, 100, 24721, 10924, 20657, 25151, 10147, 10478, 13362, 10692, 10824, 119, 119, 119, 119, 102], [101, 1160, 46492, 1154, 111313, 13693, 30746, 72894, 10109, 1162, 88567, 20242, 1155, 49327, 26225, 72894, 10228, 61233, 19932, 102], [101, 10167, 13264, 11290, 72894, 11760, 24960, 99778, 10369, 39873, 119, 119, 119, 119, 11593, 72894, 10931, 10131, 13723, 102], [101, 11593, 10115, 10321, 10800, 22469, 10113, 24256, 11534, 19180, 10921, 37013, 10245, 21602, 106333, 10371, 15566, 10716, 10138, 102], [101, 87183, 19187, 18065, 1163, 15472, 18902, 12278, 1162, 26225, 35186, 1170, 18827, 84807, 15220, 70312, 119, 119, 119, 102], [101, 12786, 93742, 119, 119, 119, 182, 11090, 76889, 15794, 10815, 15694, 22596, 10478, 24078, 93742, 119, 119, 119, 102], [101, 1160, 39123, 39112, 31829, 15472, 15220, 119, 119, 182, 11281, 40300, 30297, 69191, 40300, 18995, 10230, 24627, 18436, 102]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qmkQK_gVgiTu",
        "outputId": "d12ae2b5-5d05-4a63-c5cb-f34f2bfcc6ed"
      },
      "source": [
        "batch_ordered_labels[0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 52
        }
      ]
    },
    {
      "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": "235093e9-1ac3-47a6-91de-04062f6f028f"
      },
      "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: 375,840  (94.2% 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": "2157d9dc-e557-4dc0-be45-530fcd2d4a95"
      },
      "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=len(class_names))\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.bert.configuration_bert.BertConfig'> \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170,
          "referenced_widgets": [
            "2e42662b3e614771aa91a05ae6383aa1",
            "d88c83a08500413ba75a7c66dd839810",
            "0e99df9b99e34d419cfa3be6ac056d3a",
            "884be1eb95fe42d387ad5fb5c77a3763",
            "7cc5c4fd743a412896c57147bd2e740b",
            "5d3f4f509b7a494f9b4e5461c46b4049",
            "663cae3b673e4e27958a1217ee7dc1eb",
            "092570b23ba948ba89290c70583a95fe"
          ]
        },
        "id": "bWvc9Ikv4HtD",
        "outputId": "e80f23ba-caa1-4e28-e4bf-f84623a67e47"
      },
      "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": "2e42662b3e614771aa91a05ae6383aa1",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, description='Downloading', max=714314041.0, style=ProgressStyle(descri…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Some weights of the model checkpoint at bert-base-multilingual-cased were not used when initializing BertForSequenceClassification: ['cls.predictions.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.transform.dense.bias', 'cls.predictions.decoder.weight', 'cls.seq_relationship.weight', 'cls.seq_relationship.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.transform.LayerNorm.bias']\n",
            "- This IS expected if you are initializing BertForSequenceClassification 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 BertForSequenceClassification 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 BertForSequenceClassification were not initialized from the model checkpoint at bert-base-multilingual-cased and are newly initialized: ['classifier.weight', 'classifier.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": "a01484d1-ceed-46c1-f09e-b598fe079aad"
      },
      "source": [
        "model.cuda()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "BertForSequenceClassification(\n",
              "  (bert): BertModel(\n",
              "    (embeddings): BertEmbeddings(\n",
              "      (word_embeddings): Embedding(119547, 768, padding_idx=0)\n",
              "      (position_embeddings): Embedding(512, 768)\n",
              "      (token_type_embeddings): Embedding(2, 768)\n",
              "      (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "      (dropout): Dropout(p=0.1, inplace=False)\n",
              "    )\n",
              "    (encoder): BertEncoder(\n",
              "      (layer): ModuleList(\n",
              "        (0): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (1): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (2): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (3): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (4): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (5): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (6): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (7): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (8): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (9): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (10): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (11): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (pooler): BertPooler(\n",
              "      (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "      (activation): Tanh()\n",
              "    )\n",
              "  )\n",
              "  (dropout): Dropout(p=0.1, inplace=False)\n",
              "  (classifier): Linear(in_features=768, out_features=3, bias=True)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "da0cbVHRxYP2"
      },
      "source": [
        "## Load the model from the transformers library using AutoModel"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2IrUaVSwxYP4",
        "outputId": "06fbcfa8-205f-4eb5-f1d4-67259da6dbba"
      },
      "source": [
        "# Load the pre-trained model for classification, passing in the `config` from\r\n",
        "# above.\r\n",
        "model = 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(model)))\r\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Model type: <class 'transformers.models.xlm_roberta.modeling_xlm_roberta.XLMRobertaModel'>\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GVrbCan-xYP6"
      },
      "source": [
        "# freeze all the parameters\r\n",
        "for param in model.parameters():\r\n",
        "    param.requires_grad = False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xYXUSxyLxYP7"
      },
      "source": [
        "## Creating a custom BERT model "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mym-gU1yxYP8"
      },
      "source": [
        "\r\n",
        "class Custom_Model(nn.Module):\r\n",
        "    ## defining the constructor for the class\r\n",
        "    def __init__(self, model,num_labels):\r\n",
        "      ## calling the super class constructor\r\n",
        "      super(Custom_Model, self).__init__()\r\n",
        "      ## adding the bert layer to the model\r\n",
        "      self.model = model        \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.model(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": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mCi0iCvnxYP9",
        "outputId": "e79ece3b-4ed1-4f65-b676-cde5e5cc8634"
      },
      "source": [
        "\r\n",
        "\r\n",
        "print('\\nLoading model ...')\r\n",
        "# pass the pre-trained BERT to our define architecture\r\n",
        "model = Custom_Model(model,num_labels=len(class_names))\r\n",
        "\r\n",
        "model.cuda()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Loading model ...\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Custom_Model(\n",
              "  (model): XLMRobertaModel(\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",
              "    (pooler): RobertaPooler(\n",
              "      (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "      (activation): Tanh()\n",
              "    )\n",
              "  )\n",
              "  (relu): ReLU()\n",
              "  (fc1): Linear(in_features=768, out_features=512, bias=True)\n",
              "  (fc2): Linear(in_features=512, out_features=6, bias=True)\n",
              "  (softmax): LogSoftmax(dim=1)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 146
        }
      ]
    },
    {
      "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": "8291e0eb-1e18-4c14-f013-91ed8357161c"
      },
      "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: 0:00:13.  Remaining: 0:01:57\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:26.  Remaining: 0:01:46\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:40.  Remaining: 0:01:34\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:53.  Remaining: 0:01:22\n",
            "  Batch     500  of    1,010.    Elapsed: 0:01:06.  Remaining: 0:01:08\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:22.  Remaining: 0:00:56\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:37.  Remaining: 0:00:43\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:51.  Remaining: 0:00:29\n",
            "  Batch     900  of    1,010.    Elapsed: 0:02:05.  Remaining: 0:00:15\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:02:22.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.80\n",
            "  Training Accuracy: 0.56\n",
            "  Training epoch took: 0:02:24\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: 0:00:03\n",
            "  Batch      20  of      127.    Elapsed: 0:00:00.  Remaining: 0:00:03\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\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: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     110  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     120  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.75\n",
            "  Validation Accuracy: 0.62\n",
            "  Validation epoch took: 0:00:04\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: 0:00:15.  Remaining: 0:02:16\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:29.  Remaining: 0:01:58\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:44.  Remaining: 0:01:44\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:59.  Remaining: 0:01:30\n",
            "  Batch     500  of    1,010.    Elapsed: 0:01:13.  Remaining: 0:01:14\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:26.  Remaining: 0:00:59\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:41.  Remaining: 0:00:45\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:56.  Remaining: 0:00:30\n",
            "  Batch     900  of    1,010.    Elapsed: 0:02:10.  Remaining: 0:00:16\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:02:25.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.71\n",
            "  Training Accuracy: 0.63\n",
            "  Training epoch took: 0:02:28\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: 0:00:04\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: 0:00:03\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: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     110  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     120  of      127.    Elapsed: 0:00:04.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.75\n",
            "  Validation Accuracy: 0.62\n",
            "  Validation epoch took: 0:00:04\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: 0:00:15.  Remaining: 0:02:14\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:29.  Remaining: 0:01:56\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:44.  Remaining: 0:01:43\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:59.  Remaining: 0:01:30\n",
            "  Batch     500  of    1,010.    Elapsed: 0:01:13.  Remaining: 0:01:15\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:27.  Remaining: 0:01:00\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:41.  Remaining: 0:00:45\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:55.  Remaining: 0:00:30\n",
            "  Batch     900  of    1,010.    Elapsed: 0:02:09.  Remaining: 0:00:16\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:02:25.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.62\n",
            "  Training Accuracy: 0.70\n",
            "  Training epoch took: 0:02:28\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: 0:00:05\n",
            "  Batch      20  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:04\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:03\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:03\n",
            "  Batch      50  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:03\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: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     110  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:00\n",
            "  Batch     120  of      127.    Elapsed: 0:00:04.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.78\n",
            "  Validation Accuracy: 0.63\n",
            "  Validation epoch took: 0:00:04\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: 0:00:13.  Remaining: 0:01:59\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:29.  Remaining: 0:01:57\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:44.  Remaining: 0:01:44\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:59.  Remaining: 0:01:30\n",
            "  Batch     500  of    1,010.    Elapsed: 0:01:13.  Remaining: 0:01:15\n",
            "  Batch     600  of    1,010.    Elapsed: 0:01:28.  Remaining: 0:01:00\n",
            "  Batch     700  of    1,010.    Elapsed: 0:01:42.  Remaining: 0:00:45\n",
            "  Batch     800  of    1,010.    Elapsed: 0:01:57.  Remaining: 0:00:31\n",
            "  Batch     900  of    1,010.    Elapsed: 0:02:11.  Remaining: 0:00:16\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:02:26.  Remaining: 0:00:01\n",
            "\n",
            "  Average training loss: 0.52\n",
            "  Training Accuracy: 0.77\n",
            "  Training epoch took: 0:02:30\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: 0:00:04\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: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      70  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\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: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     110  of      127.    Elapsed: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     120  of      127.    Elapsed: 0:00:04.  Remaining: 0:00:00\n",
            "\n",
            "  Average validation loss: 0.87\n",
            "  Validation Accuracy: 0.61\n",
            "  Validation epoch took: 0:00:04\n",
            "Best val accuracy: 0.6303270564915758\n",
            "\n",
            "Training complete!\n",
            "Total training took 0:10:19 (h:mm:ss)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 450
        },
        "id": "GOmua_S1eXNM",
        "outputId": "37d52ed6-d8d7-45ca-82e8-737dc2552c2d"
      },
      "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": "4c4ddbab-98ca-4986-ba5b-20a0d12a9013"
      },
      "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: 0:00:03.  Remaining: 0:00:23\n",
            "  Batch     200  of    1,010.    Elapsed: 0:00:05.  Remaining: 0:00:22\n",
            "  Batch     300  of    1,010.    Elapsed: 0:00:07.  Remaining: 0:00:18\n",
            "  Batch     400  of    1,010.    Elapsed: 0:00:10.  Remaining: 0:00:15\n",
            "  Batch     500  of    1,010.    Elapsed: 0:00:12.  Remaining: 0:00:12\n",
            "  Batch     600  of    1,010.    Elapsed: 0:00:14.  Remaining: 0:00:10\n",
            "  Batch     700  of    1,010.    Elapsed: 0:00:18.  Remaining: 0:00:08\n",
            "  Batch     800  of    1,010.    Elapsed: 0:00:20.  Remaining: 0:00:05\n",
            "  Batch     900  of    1,010.    Elapsed: 0:00:23.  Remaining: 0:00:03\n",
            "  Batch   1,000  of    1,010.    Elapsed: 0:00:25.  Remaining: 0:00:00\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uXsVn0wngZhv",
        "outputId": "6fd92524-603a-4341-b0bd-be6b034941a3"
      },
      "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       0.88      0.59      0.70      6327\n",
            "Non_hope_speech       0.74      0.92      0.82      7872\n",
            "      not-Tamil       0.78      0.89      0.83      1961\n",
            "\n",
            "       accuracy                           0.78     16160\n",
            "      macro avg       0.80      0.80      0.78     16160\n",
            "   weighted avg       0.80      0.78      0.78     16160\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "guClUdRhYPql",
        "outputId": "57b27b7f-4737-4d6f-d54f-1004eb3bbd04"
      },
      "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": "c5388c92-2ef4-4bbf-dae2-f66053831e80"
      },
      "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: 0:00:03\n",
            "  Batch      20  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:03\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      60  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:02\n",
            "  Batch      70  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\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": "wAjlATBPgh7c",
        "outputId": "84e8f6fa-1518-4c60-af76-960444e4c709"
      },
      "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       0.62      0.42      0.50       757\n",
            "Non_hope_speech       0.64      0.78      0.70       998\n",
            "      not-Tamil       0.62      0.67      0.64       263\n",
            "\n",
            "       accuracy                           0.63      2018\n",
            "      macro avg       0.63      0.62      0.62      2018\n",
            "   weighted avg       0.63      0.63      0.62      2018\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "7nELY57NYshD",
        "outputId": "c1ef0983-cb78-4605-d0f9-8fa562b92a55"
      },
      "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": "33bae7c3-8d51-49bd-f091-5f2b926d4979"
      },
      "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: 0:00:04\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: 0:00:03\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: 0:00:03.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:03.  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": "l6EJfj0TNT6R",
        "outputId": "c8c8730d-b7a1-4ab6-d3d3-e98bcd8e62b7"
      },
      "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       0.65      0.40      0.50       815\n",
            "Non_hope_speech       0.60      0.78      0.68       946\n",
            "      not-Tamil       0.57      0.65      0.61       259\n",
            "\n",
            "       accuracy                           0.61      2020\n",
            "      macro avg       0.61      0.61      0.60      2020\n",
            "   weighted avg       0.62      0.61      0.60      2020\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "ZE3EG9RTNT6S",
        "outputId": "8b5eac33-84eb-4b12-d668-ad35c80dd39c"
      },
      "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": "2fOJyH_c3ON2"
      },
      "source": [
        "## Saving the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3PXqG-VpurVS"
      },
      "source": [
        "torch.save(model,\"./drive/MyDrive/Codalab/HopeSpeech Detection For Equality,DiversityAnd Inclusion/Models/BERT_Multilingual_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/BERTForSequenceClassification_Multilingual_Tamil_v2',map_location=device)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sAaHpZdIzsnT",
        "outputId": "5e2f78b2-c091-442f-b789-59da4efac7d8"
      },
      "source": [
        "model.cuda()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "BertForSequenceClassification(\n",
              "  (bert): BertModel(\n",
              "    (embeddings): BertEmbeddings(\n",
              "      (word_embeddings): Embedding(105879, 768, padding_idx=0)\n",
              "      (position_embeddings): Embedding(512, 768)\n",
              "      (token_type_embeddings): Embedding(2, 768)\n",
              "      (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "      (dropout): Dropout(p=0.1, inplace=False)\n",
              "    )\n",
              "    (encoder): BertEncoder(\n",
              "      (layer): ModuleList(\n",
              "        (0): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (1): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (2): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (3): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (4): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (5): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (6): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (7): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (8): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (9): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (10): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "        (11): BertLayer(\n",
              "          (attention): BertAttention(\n",
              "            (self): BertSelfAttention(\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): BertSelfOutput(\n",
              "              (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "              (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "              (dropout): Dropout(p=0.1, inplace=False)\n",
              "            )\n",
              "          )\n",
              "          (intermediate): BertIntermediate(\n",
              "            (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          )\n",
              "          (output): BertOutput(\n",
              "            (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
              "            (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
              "            (dropout): Dropout(p=0.1, inplace=False)\n",
              "          )\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (pooler): BertPooler(\n",
              "      (dense): Linear(in_features=768, out_features=768, bias=True)\n",
              "      (activation): Tanh()\n",
              "    )\n",
              "  )\n",
              "  (dropout): Dropout(p=0.1, inplace=False)\n",
              "  (classifier): Linear(in_features=768, out_features=3, bias=True)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_crczjW6V-QK"
      },
      "source": [
        "## Making Predictions on Test Set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lETtjTatLo91"
      },
      "source": [
        "test_df_path = join(dataset_dir,'tamil_hope_test.csv')\r\n",
        "test_df = pd.read_csv(test_df_path)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d_83fOUwqn0w",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "outputId": "e32ebf45-a0bf-4082-bf93-253e966e98f6"
      },
      "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": 41
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7jLPpljYQcsQ"
      },
      "source": [
        "## Loading `test_sentences`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QEVoAhgIQcsR"
      },
      "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": "slovGuCRThW1",
        "outputId": "ecce9c35-92a8-4e17-9385-709393e5648f"
      },
      "source": [
        "test_ids"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([   0,    1,    2, ..., 2017, 2018, 2019])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gL7aCG2HTjv6",
        "outputId": "8e735584-7035-4051-ad6b-f0a2aa0a6ea1"
      },
      "source": [
        "test_sentences"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Konjam methuva pesuninganna nalarukum bro...',\n",
              "       'Sema bro............. தமிழன் என்று சொல்லடா தலை நிமிர்ந்து நில்லடா',\n",
              "       '9.52 mg bro  plz crrt pannunga bro', ...,\n",
              "       'Enna nadu vanthu vaccine kudaka matadragalo anatha nadu mala matha nadu yela attract pannu',\n",
              "       'Yes..ipdii thaan avangaluku bathil kudukanu...app banned was correct',\n",
              "       'No views .. but 390 likes 🤣... MG squad'], dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b9dgO1m3QcsU",
        "outputId": "210eac2e-e7c6-4dec-d007-76fd16295105"
      },
      "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": "942587bb-c759-477f-9598-c7fa1c5544a3"
      },
      "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: 0:00:04\n",
            "  Batch      20  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:03\n",
            "  Batch      30  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      40  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\n",
            "  Batch      50  of      127.    Elapsed: 0:00:01.  Remaining: 0:00:02\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: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch      90  of      127.    Elapsed: 0:00:02.  Remaining: 0:00:01\n",
            "  Batch     100  of      127.    Elapsed: 0:00:03.  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": "c6afd5c1-6f99-4b55-9335-2411362a97d1"
      },
      "source": [
        "print(y_preds)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[1 2 0 ... 0 0 1]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kWWWjz7eUPc7",
        "outputId": "e4031923-bc24-47e6-fd7c-9c7372d3cb44"
      },
      "source": [
        "print(y_ids)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[1488 1513 1521 ...   37   59 1264]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u2ybbsAoQcsY",
        "outputId": "e42cdb1d-c623-4af9-df03-3e96b4ca64f7"
      },
      "source": [
        "le.inverse_transform(y_preds)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Non_hope_speech', 'not-Tamil', 'Hope_speech', ..., 'Hope_speech',\n",
              "       'Hope_speech', 'Non_hope_speech'], dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "kSMb6sCLQcsZ",
        "outputId": "f5b738bf-fd71-48a9-de3a-0ad8f8ff3957"
      },
      "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 0x7f5850dee6a0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        },
        {
          "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": "ef907303-487e-454c-fefd-2be926dc3cfd"
      },
      "source": [
        "len(y_ids)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2020"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9-fkBgOxU4bX",
        "outputId": "091059ee-e9a7-4fa9-faf6-12a7458d4843"
      },
      "source": [
        "len(le.inverse_transform(y_preds))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2020"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 52
        }
      ]
    },
    {
      "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": "91c2ab12-806a-433e-f548-850cf55e45af"
      },
      "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>1488</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1513</td>\n",
              "      <td>not-Tamil</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1521</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1571</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1643</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>1189</td>\n",
              "      <td>Non_hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2016</th>\n",
              "      <td>16</td>\n",
              "      <td>not-Tamil</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2017</th>\n",
              "      <td>37</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2018</th>\n",
              "      <td>59</td>\n",
              "      <td>Hope_speech</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019</th>\n",
              "      <td>1264</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     1488  Non_hope_speech\n",
              "1     1513        not-Tamil\n",
              "2     1521      Hope_speech\n",
              "3     1571  Non_hope_speech\n",
              "4     1643  Non_hope_speech\n",
              "...    ...              ...\n",
              "2015  1189  Non_hope_speech\n",
              "2016    16        not-Tamil\n",
              "2017    37      Hope_speech\n",
              "2018    59      Hope_speech\n",
              "2019  1264  Non_hope_speech\n",
              "\n",
              "[2020 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "z0CEApHmqdLr",
        "outputId": "e0797217-b2ca-46ca-dda9-0c4dbb8cf91c"
      },
      "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": 56
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "vxReWWmxn7Hj",
        "outputId": "d6c926bc-aee8-496d-8c1e-fed3c7b46637"
      },
      "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>Non_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>Non_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>Non_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  Non_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  Non_hope_speech\n",
              "2017  Enna nadu vanthu vaccine kudaka matadragalo an...  2017  Non_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": 57
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "id": "1kuASymYn9aV",
        "outputId": "a3d0d04c-c9a2-4c50-f806-df0f6d05421a"
      },
      "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>Non_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>Non_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>Non_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...  Non_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...  Non_hope_speech\n",
              "2017  2017  Enna nadu vanthu vaccine kudaka matadragalo an...  Non_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": 58
        }
      ]
    },
    {
      "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,'bert_multi_tamil.csv'),index=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jW9g_qgMv7mq"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}