finetuning_umberto.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "9fd83767e3764536ac06a3c099f30cc5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_fac0daf86949426dbf0edc272f4a3ca4",
              "IPY_MODEL_020fe982e66f46c08371bacc8a046e3e",
              "IPY_MODEL_a36cbbe53d434189961df880717a4042"
            ],
            "layout": "IPY_MODEL_83468652d6ed4030adddf76731d43663"
          }
        },
        "fac0daf86949426dbf0edc272f4a3ca4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_7eaca10846fd4b938e11f51b5d3d0a57",
            "placeholder": "​",
            "style": "IPY_MODEL_9b192d87a27149cabdc6b63d95f4c1c1",
            "value": "Map: 100%"
          }
        },
        "020fe982e66f46c08371bacc8a046e3e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a5966aea5ab24ba5a365a7cee75186bb",
            "max": 11227,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_3f9d018ef93b4af59575675f18b98c5e",
            "value": 11227
          }
        },
        "a36cbbe53d434189961df880717a4042": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_369d0fbfc1a74eefa49a19f4b4cb2146",
            "placeholder": "​",
            "style": "IPY_MODEL_17eec740b2084606bc3c9ba444e76ad9",
            "value": " 11227/11227 [00:03&lt;00:00, 3564.69 examples/s]"
          }
        },
        "83468652d6ed4030adddf76731d43663": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "7eaca10846fd4b938e11f51b5d3d0a57": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "9b192d87a27149cabdc6b63d95f4c1c1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a5966aea5ab24ba5a365a7cee75186bb": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "3f9d018ef93b4af59575675f18b98c5e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "369d0fbfc1a74eefa49a19f4b4cb2146": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "17eec740b2084606bc3c9ba444e76ad9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "6137c2f6db3142618f685621d9b09804": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_8731e1eb5b7c49e79161d846d07eed0d",
              "IPY_MODEL_99cb5ba6845a49bcac6b12cb9666b17d",
              "IPY_MODEL_3e1145b7d6e8496997c3451f7df9c4a4"
            ],
            "layout": "IPY_MODEL_bc9db475c807460d81730dc3612217ae"
          }
        },
        "8731e1eb5b7c49e79161d846d07eed0d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_5658e3cd5a004b1e88b6ea7473fe9699",
            "placeholder": "​",
            "style": "IPY_MODEL_774199a56c1343048db07366b504ae31",
            "value": "Map: 100%"
          }
        },
        "99cb5ba6845a49bcac6b12cb9666b17d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_1bdbfc0f5f754ada8baa3f5a28d114b0",
            "max": 4136,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_7af97b92cdf34092b2a84490fb6d4dcd",
            "value": 4136
          }
        },
        "3e1145b7d6e8496997c3451f7df9c4a4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_93ad5556af63459685c7b4c274ccb988",
            "placeholder": "​",
            "style": "IPY_MODEL_8fe8b3ee0c8540f4bcb1ddeb25d8e414",
            "value": " 4136/4136 [00:01&lt;00:00, 2202.51 examples/s]"
          }
        },
        "bc9db475c807460d81730dc3612217ae": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "5658e3cd5a004b1e88b6ea7473fe9699": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "774199a56c1343048db07366b504ae31": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "1bdbfc0f5f754ada8baa3f5a28d114b0": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "7af97b92cdf34092b2a84490fb6d4dcd": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "93ad5556af63459685c7b4c274ccb988": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "8fe8b3ee0c8540f4bcb1ddeb25d8e414": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a080c30672fd47d3b0205f7773b5a647": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_2678085f75334c1e9c62f42532cbcb3e",
              "IPY_MODEL_d0fefb69769e4f29bbc8d6c717ac2eb8",
              "IPY_MODEL_61a6285d8f41488da48223b0405c6408"
            ],
            "layout": "IPY_MODEL_2b2efe792cfd4588be8a4d8be84e8722"
          }
        },
        "2678085f75334c1e9c62f42532cbcb3e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_6c3abca81ea545bd92d34e8f7a59f487",
            "placeholder": "​",
            "style": "IPY_MODEL_0efd6fb1a2384e23acbf0179aa3c18b8",
            "value": "Downloading builder script: "
          }
        },
        "d0fefb69769e4f29bbc8d6c717ac2eb8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_811ff612cb9846d6bff939325b41fafc",
            "max": 2472,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_13665db9d71e410ca040e51c0c61dcff",
            "value": 2472
          }
        },
        "61a6285d8f41488da48223b0405c6408": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_730e46d12b7b48f281cc4cdf5eb4bf2d",
            "placeholder": "​",
            "style": "IPY_MODEL_d13394903c804c228becf1bdb683d9c8",
            "value": " 6.33k/? [00:00&lt;00:00, 93.7kB/s]"
          }
        },
        "2b2efe792cfd4588be8a4d8be84e8722": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6c3abca81ea545bd92d34e8f7a59f487": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0efd6fb1a2384e23acbf0179aa3c18b8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "811ff612cb9846d6bff939325b41fafc": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "13665db9d71e410ca040e51c0c61dcff": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "730e46d12b7b48f281cc4cdf5eb4bf2d": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d13394903c804c228becf1bdb683d9c8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/lucia1299/BERTolde/blob/main/finetuning_umberto.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wkVbkFwj66F7"
      },
      "outputs": [],
      "source": [
        "!pip install transformers datasets tokenizers seqeval -q"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import datasets\n",
        "import numpy as np\n",
        "from transformers import CamembertTokenizerFast, DataCollatorForTokenClassification, AutoModelForTokenClassification\n",
        "import torch\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import confusion_matrix\n"
      ],
      "metadata": {
        "id": "fz26Mt5M7Kdt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from datasets import Dataset, ClassLabel, Features, Sequence, Value, DatasetDict\n",
        "\n",
        "def creationdictionary(data, posdict, nerdict, pos_tags_list, ner_tags_list):\n",
        "    container_list = []\n",
        "\n",
        "    while data:\n",
        "        container_dict = {}\n",
        "        pos_list = []\n",
        "        ner_list = []\n",
        "        words_list = []\n",
        "\n",
        "        index_of_empty_list = data.index([]) if [] in data else len(data)\n",
        "        tokenized_data1 = data[0: index_of_empty_list]\n",
        "        for n in tokenized_data1:\n",
        "            if 'id' not in container_dict:\n",
        "                container_dict['id'] = 0\n",
        "            else:\n",
        "                container_dict['id'] += 1\n",
        "            if n[1] in posdict:\n",
        "                pos_list.append(posdict[n[1]])\n",
        "            if n[3] in nerdict:\n",
        "                ner_list.append(nerdict[n[3]])\n",
        "            words_list.append(n[0])\n",
        "\n",
        "        container_dict['pos'] = pos_list\n",
        "        container_dict['ner_tags'] = ner_list\n",
        "        container_dict['tokens'] = words_list\n",
        "        container_dict['issueid'] = tokenized_data1[0][2]\n",
        "        container_dict['id'] = str(container_dict['id'])\n",
        "\n",
        "        container_list.append(container_dict)\n",
        "\n",
        "        data = data[index_of_empty_list + 1:]\n",
        "\n",
        "    container_dict = {key: [d[key] for d in container_list] for key in container_list[0]}\n",
        "\n",
        "    ds = Dataset.from_dict(container_dict, features=Features({\n",
        "        'pos': Sequence(ClassLabel(names=pos_tags_list)),\n",
        "        'ner_tags': Sequence(ClassLabel(names=ner_tags_list)),\n",
        "        'tokens': Sequence(Value(dtype='string')),\n",
        "        'issueid': Value(dtype='string'),\n",
        "        'id': Value(dtype='string')\n",
        "        }))\n",
        "\n",
        "    return ds\n",
        "\n",
        "def posconversion(data):\n",
        "    posdict = {}\n",
        "    set1 = set()\n",
        "    for i in data:\n",
        "        if len(i) >= 1:\n",
        "            pos_tag = i[1]\n",
        "            set1.add(pos_tag)\n",
        "    pos_tags_list = list(set1)\n",
        "    v = 0\n",
        "    for i in set1:\n",
        "        posdict[i] = v\n",
        "        v +=1\n",
        "    return posdict, pos_tags_list\n",
        "\n",
        "def nerconversion(data):\n",
        "    nerdict = {}\n",
        "    set1 = set()\n",
        "    for i in data:\n",
        "        if len(i) >= 4:\n",
        "            ner_tag = i[3]\n",
        "            set1.add(ner_tag)\n",
        "    ner_tags_list = list(set1)\n",
        "    v = 0\n",
        "    for i in set1:\n",
        "        nerdict[i] = v\n",
        "        v +=1\n",
        "    return nerdict, ner_tags_list\n",
        "\n",
        "file_path_training = '/content/I-CAB-evalita09-NER-training.iob2'\n",
        "file_path_test = '/content/I-CAB-evalita09-NER-test.iob2'\n",
        "\n",
        "def read_iob_data(file_path):\n",
        "     with open(file_path, 'r', encoding='ISO-8859-1') as file:\n",
        "         iob_data = file.readlines()\n",
        "\n",
        "     data = [line.strip().split() for line in iob_data]\n",
        "     return data\n",
        "\n",
        "data_training = read_iob_data(file_path_training)\n",
        "posdict, pos_tags_list = posconversion(data_training)\n",
        "nerdict, ner_tags_list = nerconversion(data_training)\n",
        "result_training = creationdictionary(data_training, posdict, nerdict, pos_tags_list, ner_tags_list)\n",
        "\n",
        "data_test = read_iob_data(file_path_test)\n",
        "posdict, pos_tags_list = posconversion(data_test)\n",
        "nerdict, ner_tags_list = nerconversion(data_test)\n",
        "result_test = creationdictionary(data_test, posdict, nerdict, pos_tags_list, ner_tags_list)\n",
        "\n",
        "main_dataset = DatasetDict({\"train\": result_training, \"test\": result_test})\n",
        "print(main_dataset)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YkAtprgH7dJ6",
        "outputId": "b22ee1db-cc55-4d29-d013-543e25dfd6ce"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "DatasetDict({\n",
            "    train: Dataset({\n",
            "        features: ['id', 'pos', 'ner_tags', 'tokens', 'issueid'],\n",
            "        num_rows: 11227\n",
            "    })\n",
            "    test: Dataset({\n",
            "        features: ['id', 'pos', 'ner_tags', 'tokens', 'issueid'],\n",
            "        num_rows: 4136\n",
            "    })\n",
            "})\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "main_dataset.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D_gzLGfR79Sb",
        "outputId": "337c9123-24ab-43a1-b51d-28272d6fd407"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'train': (11227, 5), 'test': (4136, 5)}"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "main_dataset['train'][1]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ltsXMzzE8DAN",
        "outputId": "a8461ea3-6def-4aec-e324-eb1b26a44cd7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'id': '6',\n",
              " 'pos': [12, 10, 12, 40, 0, 49, 11],\n",
              " 'ner_tags': [8, 8, 8, 8, 8, 8, 6],\n",
              " 'tokens': ['Latte', 'al', 'seno', ',', 'sos', 'di', 'Pedrotti'],\n",
              " 'issueid': 'adige20041007_id413942'}"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "main_dataset['train'].features['ner_tags']"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PWHpzENC8L81",
        "outputId": "0165e318-1095-433c-f416-40707fcf4fc3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Sequence(feature=ClassLabel(names=['I-PER', 'B-GPE', 'B-LOC', 'I-ORG', 'I-LOC', 'B-ORG', 'B-PER', 'I-GPE', 'O'], id=None), length=-1, id=None)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tokenizer = CamembertTokenizerFast.from_pretrained(\"Musixmatch/umberto-commoncrawl-cased-v1\")\n",
        "\n",
        "#add_prefix_space needs to be used because this is how the instatiation of RobertaTokenizerFast works"
      ],
      "metadata": {
        "id": "ACQAL-EI8TZi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "example_text = main_dataset['train'][1]\n",
        "\n",
        "tokenized_input = tokenizer(example_text['tokens'], is_split_into_words=True)\n",
        "\n",
        "tokens = tokenizer.convert_ids_to_tokens(tokenized_input['input_ids'])\n",
        "\n",
        "word_ids = tokenized_input.word_ids()\n",
        "\n",
        "#ValueError: word_ids() is not available when using non-fast tokenizers (e.g. instance of a `XxxTokenizerFast` class).\n",
        "#this is why we need to use TokenizersFast also for Roberta\n",
        "\n",
        "word_ids\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XhXTKUH28pgm",
        "outputId": "4fe2380a-7b2a-42ce-e85c-b463ed9dd195"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[None, 0, 1, 2, 3, 4, 5, 6, 6, None]"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tokens = tokenizer.convert_ids_to_tokens(tokenized_input['input_ids'])\n",
        "tokens\n",
        "\n",
        "#<s> and </s> are mapped to NONE because the they are \"special tokens\""
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kwgNOOGZ_tfM",
        "outputId": "7fa38a4d-cc6f-4725-bba2-3a5b5b5d7c35"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['<s>', '▁Latte', '▁al', '▁seno', '▁,', '▁sos', '▁di', '▁Pedro', 'tti', '</s>']"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(example_text['ner_tags']), len(tokenized_input['input_ids'])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rpE2Vuj79SWb",
        "outputId": "491d4a02-0b6d-4ab0-e5f1-2b5883bd2d23"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(7, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def tokenize_and_align_labels(example, label_all_tokens =True):\n",
        "    tokenized_input = tokenizer(example['tokens'], truncation=True, is_split_into_words=True)\n",
        "\n",
        "    labels = []\n",
        "\n",
        "    for i, label in enumerate(example['ner_tags']):\n",
        "        word_ids = tokenized_input.word_ids(batch_index=i)\n",
        "        #it returns a list indicating the word corresponding to each token\n",
        "        previous_word_idx = None\n",
        "\n",
        "        label_ids =[]\n",
        "\n",
        "        for word_idx in word_ids:\n",
        "            if word_idx is None:\n",
        "                #special tokens\n",
        "                label_ids.append(-100)\n",
        "            elif word_idx != previous_word_idx:\n",
        "                #regular case\n",
        "                label_ids.append(label[word_idx])\n",
        "            else:\n",
        "                #subword case\n",
        "                label_ids.append(label[word_idx] if label_all_tokens else -100)\n",
        "            previous_word_idx = word_idx\n",
        "        labels.append(label_ids)\n",
        "    tokenized_input['labels'] = labels\n",
        "    return tokenized_input"
      ],
      "metadata": {
        "id": "wHIlfAFg-jSe"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "q = tokenize_and_align_labels(main_dataset['train'][4:5])\n",
        "print(q)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WP5L8ha9Bee-",
        "outputId": "c4061f43-1924-4975-8eb6-65b771f11b0e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Asking to truncate to max_length but no maximum length is provided and the model has no predefined maximum length. Default to no truncation.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'input_ids': [[5, 2856, 339, 29391, 56, 3101, 3294, 76, 31917, 8973, 98, 20480, 21, 31220, 36, 11780, 801, 21, 9985, 49, 1665, 6801, 801, 21, 20270, 13, 337, 126, 448, 83, 11613, 59, 1809, 216, 2000, 8173, 1048, 6]], 'attention_mask': [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]], 'labels': [[-100, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 1, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, -100]]}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for token, label in zip(tokenizer.convert_ids_to_tokens(q['input_ids'][0]), (q['labels'][0])):\n",
        "    print(f\"{token:_<40} {label}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "POVz5UwuB1sH",
        "outputId": "a3c7a8a8-1b4c-4ff4-9dc0-b01f11eacd36"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<s>_____________________________________ -100\n",
            "▁Oggi___________________________________ 8\n",
            "▁alle___________________________________ 8\n",
            "▁17.30__________________________________ 8\n",
            "▁al_____________________________________ 8\n",
            "▁Centro_________________________________ 8\n",
            "▁culturale______________________________ 8\n",
            "▁S______________________________________ 8\n",
            "._______________________________________ 8\n",
            "▁Chiara_________________________________ 8\n",
            "▁si_____________________________________ 8\n",
            "▁parlerà________________________________ 8\n",
            "▁di_____________________________________ 8\n",
            "▁allattamento___________________________ 8\n",
            "▁in_____________________________________ 8\n",
            "▁Trentino_______________________________ 1\n",
            "▁,______________________________________ 8\n",
            "▁di_____________________________________ 8\n",
            "▁mamme__________________________________ 8\n",
            "▁con____________________________________ 8\n",
            "▁tanti__________________________________ 8\n",
            "▁dubbi__________________________________ 8\n",
            "▁,______________________________________ 8\n",
            "▁di_____________________________________ 8\n",
            "▁neonati________________________________ 8\n",
            "▁a______________________________________ 8\n",
            "▁cui____________________________________ 8\n",
            "▁non____________________________________ 8\n",
            "▁sempre_________________________________ 8\n",
            "▁è______________________________________ 8\n",
            "▁garantito______________________________ 8\n",
            "▁il_____________________________________ 8\n",
            "▁diritto________________________________ 8\n",
            "▁alla___________________________________ 8\n",
            "▁migliore_______________________________ 8\n",
            "▁alimentazione__________________________ 8\n",
            "▁.______________________________________ 8\n",
            "</s>____________________________________ -100\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tokenized_dataset = main_dataset.map(tokenize_and_align_labels, batched=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81,
          "referenced_widgets": [
            "9fd83767e3764536ac06a3c099f30cc5",
            "fac0daf86949426dbf0edc272f4a3ca4",
            "020fe982e66f46c08371bacc8a046e3e",
            "a36cbbe53d434189961df880717a4042",
            "83468652d6ed4030adddf76731d43663",
            "7eaca10846fd4b938e11f51b5d3d0a57",
            "9b192d87a27149cabdc6b63d95f4c1c1",
            "a5966aea5ab24ba5a365a7cee75186bb",
            "3f9d018ef93b4af59575675f18b98c5e",
            "369d0fbfc1a74eefa49a19f4b4cb2146",
            "17eec740b2084606bc3c9ba444e76ad9",
            "6137c2f6db3142618f685621d9b09804",
            "8731e1eb5b7c49e79161d846d07eed0d",
            "99cb5ba6845a49bcac6b12cb9666b17d",
            "3e1145b7d6e8496997c3451f7df9c4a4",
            "bc9db475c807460d81730dc3612217ae",
            "5658e3cd5a004b1e88b6ea7473fe9699",
            "774199a56c1343048db07366b504ae31",
            "1bdbfc0f5f754ada8baa3f5a28d114b0",
            "7af97b92cdf34092b2a84490fb6d4dcd",
            "93ad5556af63459685c7b4c274ccb988",
            "8fe8b3ee0c8540f4bcb1ddeb25d8e414"
          ]
        },
        "id": "sCTG0BXkCosA",
        "outputId": "ccfef762-55e5-4e7a-e73e-3ca285145cdf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Map:   0%|          | 0/11227 [00:00<?, ? examples/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "9fd83767e3764536ac06a3c099f30cc5"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Map:   0%|          | 0/4136 [00:00<?, ? examples/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "6137c2f6db3142618f685621d9b09804"
            }
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "num_labels = main_dataset['train'].features['ner_tags'].feature.num_classes\n",
        "num_labels"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KhrHaitQDJ7u",
        "outputId": "8631f131-8789-406a-b8ce-11ab3a6f4afc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "model = AutoModelForTokenClassification.from_pretrained(\"Musixmatch/umberto-commoncrawl-cased-v1\", num_labels = num_labels)"
      ],
      "metadata": {
        "id": "dedA-YxfC7GA",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2451bd31-87ea-42f2-aad9-b2e4ccc14afe"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Some weights of CamembertForTokenClassification were not initialized from the model checkpoint at Musixmatch/umberto-commoncrawl-cased-v1 and are newly initialized: ['classifier.bias', 'classifier.weight']\n",
            "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install accelerate -U"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3Ym83mLclFQp",
        "outputId": "1cb30ec5-3649-4e7a-8778-71d1eb914301"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: accelerate in /usr/local/lib/python3.10/dist-packages (0.25.0)\n",
            "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.10/dist-packages (from accelerate) (1.23.5)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from accelerate) (23.2)\n",
            "Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from accelerate) (5.9.5)\n",
            "Requirement already satisfied: pyyaml in /usr/local/lib/python3.10/dist-packages (from accelerate) (6.0.1)\n",
            "Requirement already satisfied: torch>=1.10.0 in /usr/local/lib/python3.10/dist-packages (from accelerate) (2.1.0+cu118)\n",
            "Requirement already satisfied: huggingface-hub in /usr/local/lib/python3.10/dist-packages (from accelerate) (0.19.4)\n",
            "Requirement already satisfied: safetensors>=0.3.1 in /usr/local/lib/python3.10/dist-packages (from accelerate) (0.4.1)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (3.13.1)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (4.5.0)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (1.12)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (3.2.1)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (3.1.2)\n",
            "Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (2023.6.0)\n",
            "Requirement already satisfied: triton==2.1.0 in /usr/local/lib/python3.10/dist-packages (from torch>=1.10.0->accelerate) (2.1.0)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->accelerate) (2.31.0)\n",
            "Requirement already satisfied: tqdm>=4.42.1 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub->accelerate) (4.66.1)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch>=1.10.0->accelerate) (2.1.3)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->huggingface-hub->accelerate) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->huggingface-hub->accelerate) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->huggingface-hub->accelerate) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->huggingface-hub->accelerate) (2023.11.17)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch>=1.10.0->accelerate) (1.3.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pip install transformers[torch]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oCWT7BUKlTfq",
        "outputId": "cc4d1467-fdf5-4880-b141-9a3959962699"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: transformers[torch] in /usr/local/lib/python3.10/dist-packages (4.35.2)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (3.13.1)\n",
            "Requirement already satisfied: huggingface-hub<1.0,>=0.16.4 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (0.19.4)\n",
            "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (1.23.5)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (23.2)\n",
            "Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (6.0.1)\n",
            "Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (2023.6.3)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (2.31.0)\n",
            "Requirement already satisfied: tokenizers<0.19,>=0.14 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (0.15.0)\n",
            "Requirement already satisfied: safetensors>=0.3.1 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (0.4.1)\n",
            "Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (4.66.1)\n",
            "Requirement already satisfied: torch!=1.12.0,>=1.10 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (2.1.0+cu118)\n",
            "Requirement already satisfied: accelerate>=0.20.3 in /usr/local/lib/python3.10/dist-packages (from transformers[torch]) (0.25.0)\n",
            "Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from accelerate>=0.20.3->transformers[torch]) (5.9.5)\n",
            "Requirement already satisfied: fsspec>=2023.5.0 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.16.4->transformers[torch]) (2023.6.0)\n",
            "Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.16.4->transformers[torch]) (4.5.0)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch!=1.12.0,>=1.10->transformers[torch]) (1.12)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch!=1.12.0,>=1.10->transformers[torch]) (3.2.1)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch!=1.12.0,>=1.10->transformers[torch]) (3.1.2)\n",
            "Requirement already satisfied: triton==2.1.0 in /usr/local/lib/python3.10/dist-packages (from torch!=1.12.0,>=1.10->transformers[torch]) (2.1.0)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->transformers[torch]) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->transformers[torch]) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->transformers[torch]) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->transformers[torch]) (2023.11.17)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch!=1.12.0,>=1.10->transformers[torch]) (2.1.3)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch!=1.12.0,>=1.10->transformers[torch]) (1.3.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from transformers import TrainingArguments, AdamW, get_linear_schedule_with_warmup\n",
        "\n",
        "optimizer = AdamW(model.parameters(), lr=2e-5)\n",
        "\n",
        "scheduler = get_linear_schedule_with_warmup(optimizer, 1400, 17400)"
      ],
      "metadata": {
        "id": "Dnh5Vc-DLJJW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9110e28f-6bad-445a-beaa-67507a572cb5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/transformers/optimization.py:411: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
            "  warnings.warn(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#define training arguments parameters\n",
        "from transformers import TrainingArguments, Trainer\n",
        "\n",
        "args = TrainingArguments(\n",
        "    \"test-ner\",\n",
        "    evaluation_strategy=\"epoch\",\n",
        "    per_device_train_batch_size=4,\n",
        "    per_device_eval_batch_size=4,\n",
        "    num_train_epochs=6,\n",
        "    weight_decay=0.01,\n",
        ")"
      ],
      "metadata": {
        "id": "gZGaI19ADSCu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "data_collator = DataCollatorForTokenClassification(tokenizer)\n",
        "#objects that will form a batch by using a list of dataset elements as input - they are the same type of the elements of the trian dataset"
      ],
      "metadata": {
        "id": "kJ_-UpQTEwCv"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "metric = datasets.load_metric('seqeval')\n",
        "#utureWarning: load_metric is deprecated and will be removed in the next major version of datasets.\n",
        "# Use 'evaluate.load' instead, from the new library 🤗\n",
        "example = main_dataset['train'][0]"
      ],
      "metadata": {
        "id": "r4THUkMiFL98",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104,
          "referenced_widgets": [
            "a080c30672fd47d3b0205f7773b5a647",
            "2678085f75334c1e9c62f42532cbcb3e",
            "d0fefb69769e4f29bbc8d6c717ac2eb8",
            "61a6285d8f41488da48223b0405c6408",
            "2b2efe792cfd4588be8a4d8be84e8722",
            "6c3abca81ea545bd92d34e8f7a59f487",
            "0efd6fb1a2384e23acbf0179aa3c18b8",
            "811ff612cb9846d6bff939325b41fafc",
            "13665db9d71e410ca040e51c0c61dcff",
            "730e46d12b7b48f281cc4cdf5eb4bf2d",
            "d13394903c804c228becf1bdb683d9c8"
          ]
        },
        "outputId": "01d40429-337f-4f11-e4aa-829ca42bfae2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-22-1a25687c91d7>:1: FutureWarning: load_metric is deprecated and will be removed in the next major version of datasets. Use 'evaluate.load' instead, from the new library 🤗 Evaluate: https://huggingface.co/docs/evaluate\n",
            "  metric = datasets.load_metric('seqeval')\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Downloading builder script:   0%|          | 0.00/2.47k [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "a080c30672fd47d3b0205f7773b5a647"
            }
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_list = main_dataset['train'].features['ner_tags'].feature.names\n",
        "label_list"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mbqYW3FKFTkF",
        "outputId": "bf48adb3-4a2e-490b-cd9d-664f344cb4ec"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['I-PER', 'B-GPE', 'B-LOC', 'I-ORG', 'I-LOC', 'B-ORG', 'B-PER', 'I-GPE', 'O']"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "example = main_dataset['train'][1]"
      ],
      "metadata": {
        "id": "1xrheOc6GvMC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#calculate the metric on a single example (the first of the train dataset)\n",
        "labels = [label_list[i] for i in example['ner_tags']]\n",
        "labels"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KGkApBvuGMli",
        "outputId": "18cf55b4-640c-47c0-c8fd-94543f2596f6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['O', 'O', 'O', 'O', 'O', 'O', 'B-PER']"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#check metric.compute works\n",
        "metric.compute(predictions=[labels], references=[labels])\n",
        "#the score of this is very high because I pass the same list to both my predictions and references"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "T-ZEzI6OGz11",
        "outputId": "a0f0434e-8498-4b50-e97d-7c1f1d74ddfa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'PER': {'precision': 1.0, 'recall': 1.0, 'f1': 1.0, 'number': 1},\n",
              " 'overall_precision': 1.0,\n",
              " 'overall_recall': 1.0,\n",
              " 'overall_f1': 1.0,\n",
              " 'overall_accuracy': 1.0}"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def compute_metrics(eval_preds):\n",
        "    pred_logits, labels = eval_preds\n",
        "\n",
        "    # pred_logits are obtained by passing numpy.argmax\n",
        "    pred_logits = np.argmax(pred_logits, axis=2)\n",
        "    # the logits and the probabilities are in the same order,\n",
        "    # so we don’t need to apply the softmax\n",
        "\n",
        "    # We remove all the values where the label is -100\n",
        "    # define the variable prediction here\n",
        "    predictions = [\n",
        "        [label_list[eval_preds] for (eval_preds, l) in zip(prediction, label) if l != -100]\n",
        "        for prediction, label in zip(pred_logits, labels)\n",
        "    ]\n",
        "    # define the variable true_labels here\n",
        "    true_labels = [\n",
        "      [label_list[l] for (eval_preds, l) in zip(prediction, label) if l != -100]\n",
        "       for prediction, label in zip(pred_logits, labels)\n",
        "    ]\n",
        "    results = metric.compute(predictions=predictions, references=true_labels)\n",
        "\n",
        "    global stored_predictions, stored_labels\n",
        "    stored_predictions.extend(predictions)\n",
        "    stored_labels.extend(true_labels)\n",
        "\n",
        "    return {\n",
        "   \"precision\": results[\"overall_precision\"],\n",
        "   \"recall\": results[\"overall_recall\"],\n",
        "   \"f1\": results[\"overall_f1\"],\n",
        "    \"accuracy\": results[\"overall_accuracy\"],\n",
        "  }"
      ],
      "metadata": {
        "id": "JAn7-31bHMEG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "stored_predictions = []\n",
        "stored_labels = []"
      ],
      "metadata": {
        "id": "1RuPc7OgTFPJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "trainer = Trainer (\n",
        "    model,\n",
        "    args,\n",
        "    optimizers=(optimizer, scheduler),\n",
        "    train_dataset=tokenized_dataset[\"train\"],\n",
        "    eval_dataset=tokenized_dataset[\"test\"],\n",
        "    data_collator=data_collator,\n",
        "    tokenizer=tokenizer,\n",
        "    compute_metrics=compute_metrics\n",
        ")"
      ],
      "metadata": {
        "id": "SvQZcRLpHMJ4"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "trainer.train()\n",
        "\n",
        "model.save_pretrained('ner_model')\n",
        "\n",
        "tokenizer.save_pretrained('tokenizer')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        },
        "id": "y0z5_wItJftt",
        "outputId": "6d53377d-11b8-4baf-ce68-ffcdd44a6cee"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "You're using a CamembertTokenizerFast tokenizer. Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the text followed by a call to the `pad` method to get a padded encoding.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "    <div>\n",
              "      \n",
              "      <progress value='16842' max='16842' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
              "      [16842/16842 27:28, Epoch 6/6]\n",
              "    </div>\n",
              "    <table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              " <tr style=\"text-align: left;\">\n",
              "      <th>Epoch</th>\n",
              "      <th>Training Loss</th>\n",
              "      <th>Validation Loss</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1</th>\n",
              "      <th>Accuracy</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.128200</td>\n",
              "      <td>0.109127</td>\n",
              "      <td>0.855883</td>\n",
              "      <td>0.834511</td>\n",
              "      <td>0.845062</td>\n",
              "      <td>0.977828</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.072700</td>\n",
              "      <td>0.096681</td>\n",
              "      <td>0.868974</td>\n",
              "      <td>0.857715</td>\n",
              "      <td>0.863308</td>\n",
              "      <td>0.979422</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.041300</td>\n",
              "      <td>0.108930</td>\n",
              "      <td>0.862666</td>\n",
              "      <td>0.880448</td>\n",
              "      <td>0.871466</td>\n",
              "      <td>0.980323</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.024400</td>\n",
              "      <td>0.113405</td>\n",
              "      <td>0.865528</td>\n",
              "      <td>0.877150</td>\n",
              "      <td>0.871300</td>\n",
              "      <td>0.980521</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.019900</td>\n",
              "      <td>0.118694</td>\n",
              "      <td>0.866543</td>\n",
              "      <td>0.877974</td>\n",
              "      <td>0.872221</td>\n",
              "      <td>0.980284</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.009200</td>\n",
              "      <td>0.119215</td>\n",
              "      <td>0.873647</td>\n",
              "      <td>0.884452</td>\n",
              "      <td>0.879017</td>\n",
              "      <td>0.981413</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table><p>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "('tokenizer/tokenizer_config.json',\n",
              " 'tokenizer/special_tokens_map.json',\n",
              " 'tokenizer/tokenizer.json')"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from conlleval import evaluate\n",
        "\n",
        "true_flat = [item for sublist in stored_labels for item in sublist]\n",
        "pred_flat = [item for sublist in stored_predictions for item in sublist]\n",
        "\n",
        "# print out the table as above\n",
        "evaluate(true_flat, pred_flat, verbose=True)\n",
        "\n",
        "# calculate overall metrics\n",
        "prec, rec, f1 = evaluate(true_flat, pred_flat, verbose=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cY2yJPnfbop9",
        "outputId": "57b1fd63-65f8-48d8-ea72-05f8c1acd2d6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "processed 605892 tokens with 50940 phrases; found: 51124 phrases; correct: 44252.\n",
            "accuracy:  89.31%; (non-O)\n",
            "accuracy:  98.00%; precision:  86.56%; recall:  86.87%; FB1:  86.71\n",
            "              GPE: precision:  85.93%; recall:  88.08%; FB1:  86.99  10769\n",
            "              LOC: precision:  66.08%; recall:  49.07%; FB1:  56.32  1194\n",
            "              ORG: precision:  76.55%; recall:  84.41%; FB1:  80.29  13431\n",
            "              PER: precision:  93.00%; recall:  89.80%; FB1:  91.37  25730\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "conf_matrix = confusion_matrix(np.concatenate(stored_labels), np.concatenate(stored_predictions))\n",
        "labels = sorted(list(set(np.concatenate(stored_labels))))\n",
        "\n",
        "plt.figure(figsize=(10, 8))\n",
        "sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues', xticklabels=labels, yticklabels=labels)\n",
        "plt.xlabel('Predicted labels')\n",
        "plt.ylabel('True labels')\n",
        "plt.title('Confusion Matrix')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "bH_evwsjTN04",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 592
        },
        "outputId": "d30d6725-94ba-4982-f915-11deb3a922f7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "id2label= {\n",
        "    str(i):label for i, label in enumerate(label_list)\n",
        "}\n",
        "\n",
        "label2id = {\n",
        "    label:str(i) for i, label in enumerate(label_list)\n",
        "}"
      ],
      "metadata": {
        "id": "N-dS1d5uJp7w"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import json\n",
        "\n",
        "config = json.load(open('ner_model/config.json'))"
      ],
      "metadata": {
        "id": "jp24MUoyJ6hH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "config['id2label'] = id2label\n",
        "config['label2id'] = label2id\n",
        "\n",
        "json.dump(config, open('ner_model/config.json', 'w'))\n",
        "\n",
        "model_fine_tuned = AutoModelForTokenClassification.from_pretrained(\"ner_model\")"
      ],
      "metadata": {
        "id": "u__Uew-AKQ5i"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "test on evaluation text"
      ],
      "metadata": {
        "id": "aSUh33sgKPsR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from transformers import pipeline"
      ],
      "metadata": {
        "id": "Kfe0ZK-lKqAn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "nlp = pipeline(\"ner\", model=model_fine_tuned, tokenizer=tokenizer)\n",
        "\n",
        "example = \"Ieri ho trovato un fiore in Trentino e l'ho portato all'ONU ma a loro non è piaciuto quindi Pedrotti mi ha sgridato\"\n",
        "\n",
        "ner_results = nlp(example)\n",
        "\n",
        "print(ner_results)"
      ],
      "metadata": {
        "id": "PTNWb2YYKwfS",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "28c132b1-953a-4e4a-f4f3-18fa0e5d32b3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[{'entity': 'B-GPE', 'score': 0.99392563, 'index': 7, 'word': '▁Trentino', 'start': 27, 'end': 36}, {'entity': 'B-ORG', 'score': 0.9823934, 'index': 15, 'word': 'ONU', 'start': 56, 'end': 59}, {'entity': 'B-PER', 'score': 0.9975248, 'index': 23, 'word': '▁Pedro', 'start': 91, 'end': 97}, {'entity': 'B-PER', 'score': 0.9975235, 'index': 24, 'word': 'tti', 'start': 97, 'end': 100}]\n"
          ]
        }
      ]
    }
  ]
}