diff --git "a/TinyGPT_Base_Model_Workflow.ipynb" "b/TinyGPT_Base_Model_Workflow.ipynb" new file mode 100644--- /dev/null +++ "b/TinyGPT_Base_Model_Workflow.ipynb" @@ -0,0 +1,4965 @@ +{ + "nbformat": 4, + "nbformat_minor": 0, + "metadata": { + "colab": { + "provenance": [], + "gpuType": "T4" + }, + "kernelspec": { + "name": "python3", + "display_name": "Python 3" + }, + "language_info": { + "name": "python" + }, + "accelerator": "GPU", + "widgets": { + "application/vnd.jupyter.widget-state+json": { + "9f52012b0cae4293b4dee1c05112a0ef": { + "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_b88312dc3fb743aa9205bc1b6d8f6cc8", + "IPY_MODEL_a7266b923f9541bb885f6c2a0a4bcc03", + "IPY_MODEL_45337c055fd449cfa9df8103d6ebcb1c" + ], + "layout": "IPY_MODEL_48255954570e42c994b61ac1aa5d5736" + } + }, + "b88312dc3fb743aa9205bc1b6d8f6cc8": { + "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_8941edbd747c46548ed21342335d34a8", + "placeholder": "​", + "style": "IPY_MODEL_903c8c4480fe4eae988a01d799d8fa07", + "value": "Generating train split: " + } + }, + "a7266b923f9541bb885f6c2a0a4bcc03": { + "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_82b0dae381034529b8fbacf7608b9a18", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_d7207a8eb4a54580a788ab12aa0bf026", + "value": 1 + } + }, + "45337c055fd449cfa9df8103d6ebcb1c": { + "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_44cae723bbf5414eb3e8b58a20c1cd4d", + "placeholder": "​", + "style": "IPY_MODEL_a06d6a5848d04658b004102c3a90cac4", + "value": " 50000/0 [00:00<00:00, 327383.40 examples/s]" + } + }, + "48255954570e42c994b61ac1aa5d5736": { + "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 + } + }, + "8941edbd747c46548ed21342335d34a8": { + "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 + } + }, + "903c8c4480fe4eae988a01d799d8fa07": { + "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": "" + } + }, + "82b0dae381034529b8fbacf7608b9a18": { + "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": "20px" + } + }, + "d7207a8eb4a54580a788ab12aa0bf026": { + "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": "" + } + }, + "44cae723bbf5414eb3e8b58a20c1cd4d": { + "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 + } + }, + "a06d6a5848d04658b004102c3a90cac4": { + "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": "" + } + }, + "380ee297181e45a591dca6a77b161cb0": { + "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_6b27cc77a6144d2a838c69e639e9d752", + "IPY_MODEL_ff51d1b818654d169637101e0f5de5ba", + "IPY_MODEL_0d364fdd91664ecfb83c582dee178b41" + ], + "layout": "IPY_MODEL_263ff57475154b2ebcdf65eef0cd88d6" + } + }, + "6b27cc77a6144d2a838c69e639e9d752": { + "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_e8798cfd08604941943c9544af8b0182", + "placeholder": "​", + "style": "IPY_MODEL_b73f75a0918a40a3b1fcbaf7bee19418", + "value": "tokenizer_config.json: 100%" + } + }, + "ff51d1b818654d169637101e0f5de5ba": { + "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_1a62ac6017d345e5922b733cd5396d20", + "max": 26, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_eb8c6c78b52041b0a436fdcef1096233", + "value": 26 + } + }, + "0d364fdd91664ecfb83c582dee178b41": { + "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_11a2d65a098a46e696d5a785dd75281f", + "placeholder": "​", + "style": "IPY_MODEL_5ca22fd5b4e24c828ce55fbf33eb7c8f", + "value": " 26.0/26.0 [00:00<00:00, 2.70kB/s]" + } + }, + "263ff57475154b2ebcdf65eef0cd88d6": { + "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 + } + }, + "e8798cfd08604941943c9544af8b0182": { + "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 + } + }, + "b73f75a0918a40a3b1fcbaf7bee19418": { + "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": "" + } + }, + "1a62ac6017d345e5922b733cd5396d20": { + "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 + } + }, + "eb8c6c78b52041b0a436fdcef1096233": { + "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": "" + } + }, + "11a2d65a098a46e696d5a785dd75281f": { + "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 + } + }, + "5ca22fd5b4e24c828ce55fbf33eb7c8f": { + "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": "" + } + }, + "ecd8841ce7a8414f93c862ccf4887277": { + "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_2e72ca25e764453d8456be33c0ca5584", + "IPY_MODEL_4ab20b93ed7c444da6c0356d1beb2121", + "IPY_MODEL_e074495d97954850becc16fd338e9618" + ], + "layout": "IPY_MODEL_f5dc19e8c95547f5bea2b0ceb649e130" + } + }, + "2e72ca25e764453d8456be33c0ca5584": { + "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_4b729cb735394bf7b8ee469485afcdae", + "placeholder": "​", + "style": "IPY_MODEL_2ce6592305024abaa39bc9a09b706cdc", + "value": "config.json: 100%" + } + }, + "4ab20b93ed7c444da6c0356d1beb2121": { + "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_6c96f82454c7486bacc4b2c24bfd4145", + "max": 665, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_635eba5c86054de09e8fa14dbc0ae398", + "value": 665 + } + }, + "e074495d97954850becc16fd338e9618": { + "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_bc3f25b9fae040b2bb1651dd8af321fd", + "placeholder": "​", + "style": "IPY_MODEL_e275ddaa3bf249dc8346e07cd5db4731", + "value": " 665/665 [00:00<00:00, 69.6kB/s]" + } + }, + "f5dc19e8c95547f5bea2b0ceb649e130": { + "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 + } + }, + "4b729cb735394bf7b8ee469485afcdae": { + "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 + } + }, + "2ce6592305024abaa39bc9a09b706cdc": { + "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": "" + } + }, + "6c96f82454c7486bacc4b2c24bfd4145": { + "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 + } + }, + "635eba5c86054de09e8fa14dbc0ae398": { + "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": "" + } + }, + "bc3f25b9fae040b2bb1651dd8af321fd": { + "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 + } + }, + "e275ddaa3bf249dc8346e07cd5db4731": { + "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": "" + } + }, + "d891e666ef5a4353a03560a5c2aa0216": { + "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_b5b63b4d88284c02978d7574cedf8711", + "IPY_MODEL_ae2daac4d72e475483593cec2422169b", + "IPY_MODEL_26e2b89eee70484983ce92540d5e7fcd" + ], + "layout": "IPY_MODEL_15f283f8fc064eef96f894c64c112deb" + } + }, + "b5b63b4d88284c02978d7574cedf8711": { + "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_49390d66b1a540c287e183d9617cda1f", + "placeholder": "​", + "style": "IPY_MODEL_4d37bb2d479944d79fa3859ae389690c", + "value": "vocab.json: 100%" + } + }, + "ae2daac4d72e475483593cec2422169b": { + "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_fe3df38093e7445b9d38912684fe1fa3", + "max": 1042301, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_cfe8933732814ca3802140744a1e020d", + "value": 1042301 + } + }, + "26e2b89eee70484983ce92540d5e7fcd": { + "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_f61a1ce269fc42d9953dcf1b22e43d94", + "placeholder": "​", + "style": "IPY_MODEL_3adc5b94a91f4292a80126335133909e", + "value": " 1.04M/1.04M [00:00<00:00, 16.1MB/s]" + } + }, + "15f283f8fc064eef96f894c64c112deb": { + "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 + } + }, + "49390d66b1a540c287e183d9617cda1f": { + "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 + } + }, + "4d37bb2d479944d79fa3859ae389690c": { + "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": "" + } + }, + "fe3df38093e7445b9d38912684fe1fa3": { + "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 + } + }, + "cfe8933732814ca3802140744a1e020d": { + "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": "" + } + }, + "f61a1ce269fc42d9953dcf1b22e43d94": { + "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 + } + }, + "3adc5b94a91f4292a80126335133909e": { + "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": "" + } + }, + "e15fb8630f0242e581afe8fdb87cc7c6": { + "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_97eb9899d1df42acb89b0b7af3b69bd5", + "IPY_MODEL_997b6990fa3549518e6e5bfcbdb039cf", + "IPY_MODEL_a2f53c20beba4d058285bfc20b882353" + ], + "layout": "IPY_MODEL_078521478ec04a299de276828f62dd33" + } + }, + "97eb9899d1df42acb89b0b7af3b69bd5": { + "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_445e629f85d74a3fa160681c3bbeb039", + "placeholder": "​", + "style": "IPY_MODEL_1b708cf4e5bc40ddb9ca766a28fd520f", + "value": "merges.txt: 100%" + } + }, + "997b6990fa3549518e6e5bfcbdb039cf": { + "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_3dddf1bb94d34210b5b1578e4ca1760c", + "max": 456318, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_3f231544c6a74a52a4e9e0a39dce09ad", + "value": 456318 + } + }, + "a2f53c20beba4d058285bfc20b882353": { + "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_f2319cfb1d5447639c199b58f111ff88", + "placeholder": "​", + "style": "IPY_MODEL_589909b140c749379c329bf0919169d6", + "value": " 456k/456k [00:00<00:00, 1.02MB/s]" + } + }, + "078521478ec04a299de276828f62dd33": { + "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 + } + }, + "445e629f85d74a3fa160681c3bbeb039": { + "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 + } + }, + "1b708cf4e5bc40ddb9ca766a28fd520f": { + "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": "" + } + }, + "3dddf1bb94d34210b5b1578e4ca1760c": { + "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 + } + }, + "3f231544c6a74a52a4e9e0a39dce09ad": { + "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": "" + } + }, + "f2319cfb1d5447639c199b58f111ff88": { + "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 + } + }, + "589909b140c749379c329bf0919169d6": { + "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": "" + } + }, + "fa11d551abeb4bc193740723b5366a7a": { + "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_3cc843fce5f24eef8af5d7a5aae26264", + "IPY_MODEL_a31aed33720e40e1a764cd310b1fc419", + "IPY_MODEL_53809ded4c11407c81a26d248c8db5f0" + ], + "layout": "IPY_MODEL_042c8ba5e7b640a293fc5c9ba2afbea2" + } + }, + "3cc843fce5f24eef8af5d7a5aae26264": { + "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_abbf605310c7488598966acfa07dac84", + "placeholder": "​", + "style": "IPY_MODEL_b07ebb22b1e148a3b979476aeca3c559", + "value": "tokenizer.json: 100%" + } + }, + "a31aed33720e40e1a764cd310b1fc419": { + "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_570a9119c59b41a1b6d67f8d1232c123", + "max": 1355256, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_7b766d1009af4ae484961b3d639a605a", + "value": 1355256 + } + }, + "53809ded4c11407c81a26d248c8db5f0": { + "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_7f8716a1c55f4aa1be3ab51ccbcd779f", + "placeholder": "​", + "style": "IPY_MODEL_dcb63e4db56d44798f51e71b2834ae01", + "value": " 1.36M/1.36M [00:00<00:00, 2.98MB/s]" + } + }, + "042c8ba5e7b640a293fc5c9ba2afbea2": { + "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 + } + }, + "abbf605310c7488598966acfa07dac84": { + "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 + } + }, + "b07ebb22b1e148a3b979476aeca3c559": { + "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": "" + } + }, + "570a9119c59b41a1b6d67f8d1232c123": { + "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 + } + }, + "7b766d1009af4ae484961b3d639a605a": { + "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": "" + } + }, + "7f8716a1c55f4aa1be3ab51ccbcd779f": { + "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 + } + }, + "dcb63e4db56d44798f51e71b2834ae01": { + "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": "" + } + }, + "14e69a36fb4a409c8535a339b0cb364e": { + "model_module": "@jupyter-widgets/controls", + "model_name": "VBoxModel", + "model_module_version": "1.5.0", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "VBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "VBoxView", + "box_style": "", + "children": [], + "layout": "IPY_MODEL_b6be4e19602b4168b116d445bde77ad9" + } + }, + "95d6b77633384cd1a752166a442a83b6": { + "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_d6f24bd763644c03804764e18b642367", + "placeholder": "​", + "style": "IPY_MODEL_26563ac09615470689726f97a5c43f91", + "value": "

Copy a token from your Hugging Face\ntokens page and paste it below.
Immediately click login after copying\nyour token or it might be stored in plain text in this notebook file.
" + } + }, + "55b13c009faf470dbfe52530603a79aa": { + "model_module": "@jupyter-widgets/controls", + "model_name": "PasswordModel", + "model_module_version": "1.5.0", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "PasswordModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "PasswordView", + "continuous_update": true, + "description": "Token:", + "description_tooltip": null, + "disabled": false, + "layout": "IPY_MODEL_11e8627edc5046328ae16f7f315e8db2", + "placeholder": "​", + "style": "IPY_MODEL_1d1fca66dcc2424f86d20ca6b7cb8604", + "value": "" + } + }, + "1ac91534a5ed4970be6196871f2be040": { + "model_module": "@jupyter-widgets/controls", + "model_name": "CheckboxModel", + "model_module_version": "1.5.0", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "CheckboxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "CheckboxView", + "description": "Add token as git credential?", + "description_tooltip": null, + "disabled": false, + "indent": true, + "layout": "IPY_MODEL_c7296bbfb73f425f9cfa538a12564722", + "style": "IPY_MODEL_f3217fc0c1ec49d0821694ce94a4d674", + "value": true + } + }, + "c4f9e4c264e94c62b737acaff2c8c775": { + "model_module": "@jupyter-widgets/controls", + "model_name": "ButtonModel", + "model_module_version": "1.5.0", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ButtonModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ButtonView", + "button_style": "", + "description": "Login", + "disabled": false, + "icon": "", + "layout": "IPY_MODEL_2f51666cb65b491594f50b092186f7ed", + "style": "IPY_MODEL_cb1fc876416c4905b406f489069087d1", + "tooltip": "" + } + }, + "936f01d81c484a7c8ea5f455e7a63e6e": { + "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_06da6f6794f6417bb1978f0d8d7c558a", + "placeholder": "​", + "style": "IPY_MODEL_628a631c347a4b78be5d6079616555a6", + "value": "\nPro Tip: If you don't already have one, you can create a dedicated\n'notebooks' token with 'write' access, that you can then easily reuse for all\nnotebooks. " + } + }, + "b6be4e19602b4168b116d445bde77ad9": { + "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": "center", + "align_self": null, + "border": null, + "bottom": null, + "display": "flex", + "flex": null, + "flex_flow": "column", + "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": "50%" + } + }, + "d6f24bd763644c03804764e18b642367": { + "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 + } + }, + "26563ac09615470689726f97a5c43f91": { + "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": "" + } + }, + "11e8627edc5046328ae16f7f315e8db2": { + "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 + } + }, + "1d1fca66dcc2424f86d20ca6b7cb8604": { + "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": "" + } + }, + "c7296bbfb73f425f9cfa538a12564722": { + "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 + } + }, + "f3217fc0c1ec49d0821694ce94a4d674": { + "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": "" + } + }, + "2f51666cb65b491594f50b092186f7ed": { + "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 + } + }, + "cb1fc876416c4905b406f489069087d1": { + "model_module": "@jupyter-widgets/controls", + "model_name": "ButtonStyleModel", + "model_module_version": "1.5.0", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ButtonStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "button_color": null, + "font_weight": "" + } + }, + "06da6f6794f6417bb1978f0d8d7c558a": { + "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 + } + }, + "628a631c347a4b78be5d6079616555a6": { + "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": "" + } + }, + "bd35fa9d9d7e4509ac252c0622d3cc7b": { + "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_a088dbcf1a334ca2a7e1d185869ef488", + "IPY_MODEL_55c9d349f42642a798cfaf088072929a", + "IPY_MODEL_6e066ae08505455a8cfe2a72fd30e5ce" + ], + "layout": "IPY_MODEL_d174280a4bc94bfb8de2fee838a935f6" + } + }, + "a088dbcf1a334ca2a7e1d185869ef488": { + "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_11653f75d150432f942d475affee8770", + "placeholder": "​", + "style": "IPY_MODEL_f6fb93f0f73a45d5a9e9069171f4a350", + "value": "Processing Files (1 / 1)      : 100%" + } + }, + "55c9d349f42642a798cfaf088072929a": { + "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_2a0da366009c4505baa4a6141838ad32", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_424f39af879b4b5e98685e801b663857", + "value": 1 + } + }, + "6e066ae08505455a8cfe2a72fd30e5ce": { + "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_d65fb8edbe2846cc90cd6a1c879b3407", + "placeholder": "​", + "style": "IPY_MODEL_e18f16f3c48144519051fbaad8887a4a", + "value": "  133MB /  133MB, 15.1MB/s  " + } + }, + "d174280a4bc94bfb8de2fee838a935f6": { + "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 + } + }, + "11653f75d150432f942d475affee8770": { + "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 + } + }, + "f6fb93f0f73a45d5a9e9069171f4a350": { + "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": "" + } + }, + "2a0da366009c4505baa4a6141838ad32": { + "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": "20px" + } + }, + "424f39af879b4b5e98685e801b663857": { + "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": "" + } + }, + "d65fb8edbe2846cc90cd6a1c879b3407": { + "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 + } + }, + "e18f16f3c48144519051fbaad8887a4a": { + "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": "" + } + }, + "6c5598eed422416d96b6721bfc4cf431": { + "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_7ea67e6296424f059c4850e2fd5bc650", + "IPY_MODEL_840facc820c24863a90d41aaeb957c67", + "IPY_MODEL_33f716e5d2bf47b48b704d4e34f0b602" + ], + "layout": "IPY_MODEL_fe2d91cfcb8b4e6ba88de7db977618e9" + } + }, + "7ea67e6296424f059c4850e2fd5bc650": { + "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_42192f5259c64d2ba909e0d2e7212b65", + "placeholder": "​", + "style": "IPY_MODEL_c3300b53ac9746dcad09a2b32a056b6f", + "value": "New Data Upload               : 100%" + } + }, + "840facc820c24863a90d41aaeb957c67": { + "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_f314a42f7b3d453a8c8289b567dc695f", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_4fd21e0ecf4546afad2bd50bc72e58e9", + "value": 1 + } + }, + "33f716e5d2bf47b48b704d4e34f0b602": { + "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_03663ad72b8b4618ba74bb5e21dc92e5", + "placeholder": "​", + "style": "IPY_MODEL_d1530951f8bc4a44aed0171ea7fb7a87", + "value": "  133MB /  133MB, 15.1MB/s  " + } + }, + "fe2d91cfcb8b4e6ba88de7db977618e9": { + "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 + } + }, + "42192f5259c64d2ba909e0d2e7212b65": { + "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 + } + }, + "c3300b53ac9746dcad09a2b32a056b6f": { + "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": "" + } + }, + "f314a42f7b3d453a8c8289b567dc695f": { + "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": "20px" + } + }, + "4fd21e0ecf4546afad2bd50bc72e58e9": { + "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": "" + } + }, + "03663ad72b8b4618ba74bb5e21dc92e5": { + "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 + } + }, + "d1530951f8bc4a44aed0171ea7fb7a87": { + "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": "" + } + }, + "35b618e0f178450fa533aa4cae01470a": { + "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_7747f1b60b054a219d02e2a93873b04c", + "IPY_MODEL_94976850944d4af7a0e4bc987d9af882", + "IPY_MODEL_d086bd77fd194a208a5207b4812e2e92" + ], + "layout": "IPY_MODEL_5c786e4dd11b46a48fc846612c66181b" + } + }, + "7747f1b60b054a219d02e2a93873b04c": { + "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_019dfed092d5456186441718c14ed377", + "placeholder": "​", + "style": "IPY_MODEL_31e18285da0f49c788a42851d5e5d7b3", + "value": "  ...pt_base/pytorch_model.bin: 100%" + } + }, + "94976850944d4af7a0e4bc987d9af882": { + "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_86f0ae6918474db69381616b361bac7e", + "max": 132625778, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_a55bab394b5a47c0a3ffdab86d51474d", + "value": 132625778 + } + }, + "d086bd77fd194a208a5207b4812e2e92": { + "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_b324ac810c1b4900ae04345b6bf1dc30", + "placeholder": "​", + "style": "IPY_MODEL_d3ac9c9ac3ca42b29b58e0ef90fbc322", + "value": "  133MB /  133MB            " + } + }, + "5c786e4dd11b46a48fc846612c66181b": { + "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 + } + }, + "019dfed092d5456186441718c14ed377": { + "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 + } + }, + "31e18285da0f49c788a42851d5e5d7b3": { + "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": "" + } + }, + "86f0ae6918474db69381616b361bac7e": { + "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 + } + }, + "a55bab394b5a47c0a3ffdab86d51474d": { + "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": "" + } + }, + "b324ac810c1b4900ae04345b6bf1dc30": { + "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 + } + }, + "d3ac9c9ac3ca42b29b58e0ef90fbc322": { + "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": "" + } + }, + "42bc0137649247e4b171571298393fd6": { + "model_module": "@jupyter-widgets/controls", + "model_name": "LabelModel", + "model_module_version": "1.5.0", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "LabelModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "LabelView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_958ad03e991a40b4a82da8cd0b5180f8", + "placeholder": "​", + "style": "IPY_MODEL_5b3b063f264d4b83871811462ae5f238", + "value": "Connecting..." + } + }, + "958ad03e991a40b4a82da8cd0b5180f8": { + "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 + } + }, + "5b3b063f264d4b83871811462ae5f238": { + "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": "" + } + } + } + } + }, + "cells": [ + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "mO5OOm-EgAHP" + }, + "outputs": [], + "source": [ + "# cell 1: install\n", + "!pip install -q transformers datasets accelerate torch torchvision sentencepiece\n", + "\n", + "# For progress bars\n", + "!pip install -q tqdm\n" + ] + }, + { + "cell_type": "code", + "source": [ + "# cell 2: upload your train.jsonl if not already present\n", + "from google.colab import files\n", + "print(\"If you already uploaded train.jsonl, skip the upload step.\")\n", + "uploaded = files.upload() # use to upload train.jsonl with {\"input\",\"output\"} lines\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 90 + }, + "id": "V_uf0myogEdC", + "outputId": "7927fb80-99df-4cea-f1c2-f926c5a03a3c" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "If you already uploaded train.jsonl, skip the upload step.\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "" + ], + "text/html": [ + "\n", + " \n", + " \n", + " Upload widget is only available when the cell has been executed in the\n", + " current browser session. Please rerun this cell to enable.\n", + " \n", + " " + ] + }, + "metadata": {} + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Saving train-50000.jsonl to train-50000.jsonl\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "# cell 3: load HF DatasetDict (expects your train.jsonl in /content)\n", + "from datasets import load_dataset, DatasetDict\n", + "\n", + "# If user uploaded as 'train.jsonl', otherwise change filename\n", + "ds = load_dataset(\"json\", data_files={\"train\":\"train.jsonl\"})[\"train\"]\n", + "\n", + "# create a small validation split for quick checks\n", + "ds = ds.train_test_split(test_size=0.08, seed=42)\n", + "dataset = DatasetDict({\"train\": ds[\"train\"], \"validation\": ds[\"test\"]})\n", + "print(dataset)\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 223, + "referenced_widgets": [ + "9f52012b0cae4293b4dee1c05112a0ef", + "b88312dc3fb743aa9205bc1b6d8f6cc8", + "a7266b923f9541bb885f6c2a0a4bcc03", + "45337c055fd449cfa9df8103d6ebcb1c", + "48255954570e42c994b61ac1aa5d5736", + "8941edbd747c46548ed21342335d34a8", + "903c8c4480fe4eae988a01d799d8fa07", + "82b0dae381034529b8fbacf7608b9a18", + "d7207a8eb4a54580a788ab12aa0bf026", + "44cae723bbf5414eb3e8b58a20c1cd4d", + "a06d6a5848d04658b004102c3a90cac4" + ] + }, + "id": "r8qUd8bXgROy", + "outputId": "9d6802a1-98f9-4b29-cdb9-074d026a03b8" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "Generating train split: 0 examples [00:00, ? examples/s]" + ], + "application/vnd.jupyter.widget-view+json": { + "version_major": 2, + "version_minor": 0, + "model_id": "9f52012b0cae4293b4dee1c05112a0ef" + } + }, + "metadata": {} + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "DatasetDict({\n", + " train: Dataset({\n", + " features: ['input', 'output'],\n", + " num_rows: 46000\n", + " })\n", + " validation: Dataset({\n", + " features: ['input', 'output'],\n", + " num_rows: 4000\n", + " })\n", + "})\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "# cell 4: imports and helper functions\n", + "import math, time, os, random\n", + "from typing import Optional\n", + "import torch\n", + "import torch.nn as nn\n", + "import torch.nn.functional as F\n", + "from torch.utils.data import DataLoader\n", + "from transformers import AutoTokenizer\n", + "\n", + "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n", + "print(\"Device:\", device)\n", + "tokenizer = AutoTokenizer.from_pretrained(\"gpt2\")\n", + "# gpt2 tokenizer has no pad token; set it\n", + "if tokenizer.pad_token is None:\n", + " tokenizer.add_special_tokens({\"pad_token\": \"<|pad|>\"})\n", + "vocab_size = len(tokenizer)\n", + "\n", + "def compute_param_count(n_layers, d_model, d_ff, n_heads):\n", + " # approximate parameter count\n", + " # embeddings: vocab*d_model + pos*d_model\n", + " # per layer: attn (qkv + out) ~ 4 * d_model * d_model (approx),\n", + " # feedforward ~ 2 * d_model * d_ff\n", + " emb = vocab_size * d_model + d_model * 1024 # assume max pos 1024\n", + " per_layer = 4 * d_model * d_model + 2 * d_model * d_ff\n", + " total = emb + n_layers * per_layer\n", + " return total\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 194, + "referenced_widgets": [ + "380ee297181e45a591dca6a77b161cb0", + "6b27cc77a6144d2a838c69e639e9d752", + "ff51d1b818654d169637101e0f5de5ba", + "0d364fdd91664ecfb83c582dee178b41", + "263ff57475154b2ebcdf65eef0cd88d6", + "e8798cfd08604941943c9544af8b0182", + "b73f75a0918a40a3b1fcbaf7bee19418", + "1a62ac6017d345e5922b733cd5396d20", + "eb8c6c78b52041b0a436fdcef1096233", + "11a2d65a098a46e696d5a785dd75281f", + "5ca22fd5b4e24c828ce55fbf33eb7c8f", + "ecd8841ce7a8414f93c862ccf4887277", + "2e72ca25e764453d8456be33c0ca5584", + "4ab20b93ed7c444da6c0356d1beb2121", + "e074495d97954850becc16fd338e9618", + "f5dc19e8c95547f5bea2b0ceb649e130", + "4b729cb735394bf7b8ee469485afcdae", + "2ce6592305024abaa39bc9a09b706cdc", + "6c96f82454c7486bacc4b2c24bfd4145", + "635eba5c86054de09e8fa14dbc0ae398", + "bc3f25b9fae040b2bb1651dd8af321fd", + "e275ddaa3bf249dc8346e07cd5db4731", + "d891e666ef5a4353a03560a5c2aa0216", + "b5b63b4d88284c02978d7574cedf8711", + "ae2daac4d72e475483593cec2422169b", + "26e2b89eee70484983ce92540d5e7fcd", + "15f283f8fc064eef96f894c64c112deb", + "49390d66b1a540c287e183d9617cda1f", + "4d37bb2d479944d79fa3859ae389690c", + "fe3df38093e7445b9d38912684fe1fa3", + "cfe8933732814ca3802140744a1e020d", + "f61a1ce269fc42d9953dcf1b22e43d94", + "3adc5b94a91f4292a80126335133909e", + "e15fb8630f0242e581afe8fdb87cc7c6", + "97eb9899d1df42acb89b0b7af3b69bd5", + "997b6990fa3549518e6e5bfcbdb039cf", + "a2f53c20beba4d058285bfc20b882353", + "078521478ec04a299de276828f62dd33", + "445e629f85d74a3fa160681c3bbeb039", + "1b708cf4e5bc40ddb9ca766a28fd520f", + "3dddf1bb94d34210b5b1578e4ca1760c", + "3f231544c6a74a52a4e9e0a39dce09ad", + "f2319cfb1d5447639c199b58f111ff88", + "589909b140c749379c329bf0919169d6", + "fa11d551abeb4bc193740723b5366a7a", + "3cc843fce5f24eef8af5d7a5aae26264", + "a31aed33720e40e1a764cd310b1fc419", + "53809ded4c11407c81a26d248c8db5f0", + "042c8ba5e7b640a293fc5c9ba2afbea2", + "abbf605310c7488598966acfa07dac84", + "b07ebb22b1e148a3b979476aeca3c559", + "570a9119c59b41a1b6d67f8d1232c123", + "7b766d1009af4ae484961b3d639a605a", + "7f8716a1c55f4aa1be3ab51ccbcd779f", + "dcb63e4db56d44798f51e71b2834ae01" + ] + }, + "id": "7djO-GLygwhA", + "outputId": "a9ef5213-0f55-4d4c-d287-c72b246ccc0e" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Device: cuda\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "tokenizer_config.json: 0%| | 0.00/26.0 [00:000 else 1.0)\n", + " filtered_logits, _ = torch.topk(next_logits, k=top_k)\n", + " # sample among top_k\n", + " probs = F.softmax(next_logits, dim=-1)\n", + " next_token = torch.multinomial(probs, num_samples=1)\n", + " tokens = torch.cat([tokens, next_token], dim=1)\n", + " return tokenizer.decode(tokens[0], skip_special_tokens=True)\n", + "\n", + "prompt = \"Write a short paragraph about renewable energy adoption in the UK.\"\n", + "print(generate(model, prompt, max_new_tokens=80, temperature=0.9, top_k=40))\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "ke1aZA_ihdwk", + "outputId": "dd1dac0b-70a9-4f8b-90d1-bcdf0626b41e" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Write a short paragraph about renewable energy adoption in the UK.\n", + "This passage provides a clear, and neutral explanation about renewable energy. It offers context, structured detail, and meaningful insight while avoiding biased or inappropriate content. It offers context, and meaningful insight while avoiding biased or inappropriate content. It offers context, and meaningful insight while avoiding biased or inappropriate content. It offers context, and meaningful insight while avoiding biased or inappropriate content. It offers context, structured detail,\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "# cell 11: multi-config runner (quick-and-dirty)\n", + "configs = [\n", + " {\"name\":\"small\",\"n_layers\":2,\"n_heads\":2,\"d_model\":128,\"d_ff\":512},\n", + " {\"name\":\"base\",\"n_layers\":4,\"n_heads\":4,\"d_model\":256,\"d_ff\":1024},\n", + " # Only attempt large if you have RAM/GPU\n", + " # {\"name\":\"wide\",\"n_layers\":6,\"n_heads\":6,\"d_model\":384,\"d_ff\":1536},\n", + "]\n", + "\n", + "results = []\n", + "\n", + "for cfg in configs:\n", + " print(\"Running config:\", cfg[\"name\"])\n", + " approx = compute_param_count(cfg[\"n_layers\"], cfg[\"d_model\"], cfg[\"d_ff\"], cfg[\"n_heads\"])\n", + " print(\"Approx params:\", int(approx))\n", + " m = TinyGPT(vocab_size, n_layers=cfg[\"n_layers\"], n_heads=cfg[\"n_heads\"],\n", + " d_model=cfg[\"d_model\"], d_ff=cfg[\"d_ff\"], max_seq_len=max_len).to(device)\n", + " opt = torch.optim.AdamW(m.parameters(), lr=3e-4)\n", + " # small 1-epoch train\n", + " for epoch in range(1):\n", + " m.train()\n", + " for i, batch in enumerate(train_loader):\n", + " inp = batch[\"input_ids\"].to(device); lab = batch[\"labels\"].to(device)\n", + " logits = m(inp)\n", + " loss = F.cross_entropy(logits[:, :-1, :].reshape(-1, logits.size(-1)), lab[:,1:].reshape(-1), ignore_index=tokenizer.pad_token_id)\n", + " loss.backward(); opt.step(); opt.zero_grad()\n", + " if i%100==0:\n", + " print(cfg[\"name\"], \"step\", i, \"loss\", loss.item())\n", + " val_loss = evaluate(m, val_loader)\n", + " sample = generate(m, \"Describe the future of renewable energy in the UK.\", max_new_tokens=60)\n", + " results.append({\"config\":cfg[\"name\"], \"val_loss\":val_loss, \"sample\":sample})\n", + " # free up memory\n", + " del m; torch.cuda.empty_cache()\n", + "\n", + "print(\"Summary:\")\n", + "for r in results: print(r)\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "UaPjH_Fuhz6L", + "outputId": "2f787b20-e966-445f-89fe-efb992bdf8a8" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Running config: small\n", + "Approx params: 6957312\n", + "small step 0 loss 10.936142921447754\n", + "small step 100 loss 2.0066487789154053\n", + "small step 200 loss 0.7831152081489563\n", + "small step 300 loss 0.4042394757270813\n", + "small step 400 loss 0.27594664692878723\n", + "small step 500 loss 0.23986797034740448\n", + "small step 600 loss 0.2129729688167572\n", + "small step 700 loss 0.17271828651428223\n", + "small step 800 loss 0.1482691466808319\n", + "small step 900 loss 0.13474945724010468\n", + "small step 1000 loss 0.13562719523906708\n", + "small step 1100 loss 0.1322852075099945\n", + "small step 1200 loss 0.12768428027629852\n", + "small step 1300 loss 0.13074539601802826\n", + "small step 1400 loss 0.11563324928283691\n", + "small step 1500 loss 0.11895468831062317\n", + "small step 1600 loss 0.108633853495121\n", + "small step 1700 loss 0.10433203727006912\n", + "small step 1800 loss 0.10576470196247101\n", + "small step 1900 loss 0.10991769284009933\n", + "small step 2000 loss 0.1076023131608963\n", + "small step 2100 loss 0.11549754440784454\n", + "small step 2200 loss 0.10694943368434906\n", + "small step 2300 loss 0.1030394658446312\n", + "small step 2400 loss 0.10950098931789398\n", + "small step 2500 loss 0.10665386915206909\n", + "small step 2600 loss 0.10858049243688583\n", + "small step 2700 loss 0.09851081669330597\n", + "small step 2800 loss 0.10054293274879456\n", + "small step 2900 loss 0.10746119916439056\n", + "small step 3000 loss 0.09992803633213043\n", + "small step 3100 loss 0.10812430083751678\n", + "small step 3200 loss 0.10923600196838379\n", + "small step 3300 loss 0.11495324969291687\n", + "small step 3400 loss 0.10835824906826019\n", + "small step 3500 loss 0.10515566170215607\n", + "small step 3600 loss 0.10571396350860596\n", + "small step 3700 loss 0.10220973938703537\n", + "small step 3800 loss 0.1048131212592125\n", + "small step 3900 loss 0.09765665978193283\n", + "small step 4000 loss 0.11014576256275177\n", + "small step 4100 loss 0.10197490453720093\n", + "small step 4200 loss 0.09691217541694641\n", + "small step 4300 loss 0.10824652016162872\n", + "small step 4400 loss 0.10557755827903748\n", + "small step 4500 loss 0.10410422831773758\n", + "small step 4600 loss 0.10059218108654022\n", + "small step 4700 loss 0.10663764923810959\n", + "small step 4800 loss 0.09478699415922165\n", + "small step 4900 loss 0.09678255766630173\n", + "small step 5000 loss 0.11002504825592041\n", + "small step 5100 loss 0.10351943969726562\n", + "small step 5200 loss 0.10882539302110672\n", + "small step 5300 loss 0.10057085752487183\n", + "small step 5400 loss 0.1069759950041771\n", + "small step 5500 loss 0.10004132241010666\n", + "small step 5600 loss 0.10338238626718521\n", + "small step 5700 loss 0.09935469925403595\n", + "Running config: base\n", + "Approx params: 16273920\n", + "base step 0 loss 10.974065780639648\n", + "base step 100 loss 0.5244567394256592\n", + "base step 200 loss 0.23574890196323395\n", + "base step 300 loss 0.2029615044593811\n", + "base step 400 loss 0.12244332581758499\n", + "base step 500 loss 0.1135944277048111\n", + "base step 600 loss 0.10901793837547302\n", + "base step 700 loss 0.11430869251489639\n", + "base step 800 loss 0.10217247158288956\n", + "base step 900 loss 0.11421678960323334\n", + "base step 1000 loss 0.1092977374792099\n", + "base step 1100 loss 0.10900989919900894\n", + "base step 1200 loss 0.11376353353261948\n", + "base step 1300 loss 0.10496743023395538\n", + "base step 1400 loss 0.10641879588365555\n", + "base step 1500 loss 0.10981472581624985\n", + "base step 1600 loss 0.11358921974897385\n", + "base step 1700 loss 0.11404959112405777\n", + "base step 1800 loss 0.11704839766025543\n", + "base step 1900 loss 0.09445181488990784\n", + "base step 2000 loss 0.10594352334737778\n", + "base step 2100 loss 0.10066144168376923\n", + "base step 2200 loss 0.11785681545734406\n", + "base step 2300 loss 0.10453670471906662\n", + "base step 2400 loss 0.11076048016548157\n", + "base step 2500 loss 0.1063772514462471\n", + "base step 2600 loss 0.10972446203231812\n", + "base step 2700 loss 0.11984846740961075\n", + "base step 2800 loss 0.11558016389608383\n", + "base step 2900 loss 0.1212877705693245\n", + "base step 3000 loss 0.10537463426589966\n", + "base step 3100 loss 0.09882750362157822\n", + "base step 3200 loss 0.09926895797252655\n", + "base step 3300 loss 0.10817407071590424\n", + "base step 3400 loss 0.114067941904068\n", + "base step 3500 loss 0.09316617250442505\n", + "base step 3600 loss 0.10326540470123291\n", + "base step 3700 loss 0.09724034368991852\n", + "base step 3800 loss 0.1076420471072197\n", + "base step 3900 loss 0.1068025752902031\n", + "base step 4000 loss 0.09833832830190659\n", + "base step 4100 loss 0.09890053421258926\n", + "base step 4200 loss 0.10683748126029968\n", + "base step 4300 loss 0.09769628942012787\n", + "base step 4400 loss 0.11143091320991516\n", + "base step 4500 loss 0.10214415192604065\n", + "base step 4600 loss 0.11255712062120438\n", + "base step 4700 loss 0.10782770067453384\n", + "base step 4800 loss 0.10992496460676193\n", + "base step 4900 loss 0.0991748496890068\n", + "base step 5000 loss 0.1007111519575119\n", + "base step 5100 loss 0.09442685544490814\n", + "base step 5200 loss 0.10488761961460114\n", + "base step 5300 loss 0.09393838793039322\n", + "base step 5400 loss 0.1020846888422966\n", + "base step 5500 loss 0.11793578416109085\n", + "base step 5600 loss 0.11351373046636581\n", + "base step 5700 loss 0.10859788954257965\n", + "Summary:\n", + "{'config': 'small', 'val_loss': 0.10365759241580963, 'sample': 'Describe the future of renewable energy in the UK.\\nThis passage provides a clear, and neutral explanation about renewable energy. It offers context, structured detail, and meaningful insight while avoiding biased or inappropriate content. It offers context, structured detail, structured detail, structured detail, structured detail, structured detail, structured detail, structured detail, structured detail, structured'}\n", + "{'config': 'base', 'val_loss': 0.10475846354663372, 'sample': 'Describe the future of renewable energy in the UK.\\nThis passage provides a clear, and neutral explanation about renewable energy. It offers context, structured detail, and meaningful insight while avoiding biased or inappropriate content. It offers context, and meaningful insight while avoiding biased or inappropriate content. It offers context, and meaningful insight while avoiding biased or inappropriate content. It'}\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "# cell 12: save small model example (for base model)\n", + "save_dir = \"./tinygpt_base\"\n", + "os.makedirs(save_dir, exist_ok=True)\n", + "torch.save(model.state_dict(), os.path.join(save_dir, \"pytorch_model.bin\"))\n", + "tokenizer.save_pretrained(save_dir)\n", + "\n", + "# Optional: convert to HF format (simple script)\n", + "# Note: this will not automatically create a Transformers AutoModel repo, but saves checkpoint + tokenizer.\n", + "print(\"Saved to\", save_dir)\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "k4-rcBPVh94q", + "outputId": "94e6bd17-afd6-47e0-8352-6431299c25b1" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Saved to ./tinygpt_base\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "# Cell 13: Push TinyGPT model to Hugging Face Hub\n", + "from huggingface_hub import HfApi, HfFolder, upload_folder\n", + "\n", + "# Step 1: Login (only needs to be done once per session)\n", + "from huggingface_hub import notebook_login\n", + "notebook_login()\n", + "\n", + "# Step 2: Define repo name\n", + "hf_repo_name = \"Abdurrahmanesc/tinygpt-base-model\" # <-- change this to your desired repo\n", + "\n", + "# Step 3: Create repo (skip if already exists)\n", + "api = HfApi()\n", + "api.create_repo(repo_id=hf_repo_name, exist_ok=True)\n", + "\n", + "# Step 4: Upload local folder containing model + tokenizer\n", + "upload_folder(\n", + " folder_path=\"./tinygpt_base\",\n", + " repo_id=hf_repo_name,\n", + " commit_message=\"Upload TinyGPT base model\"\n", + ")\n", + "\n", + "print(f\"🚀 TinyGPT model uploaded successfully to: https://huggingface.co/{hf_repo_name}\")\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 150, + "referenced_widgets": [ + "14e69a36fb4a409c8535a339b0cb364e", + "95d6b77633384cd1a752166a442a83b6", + "55b13c009faf470dbfe52530603a79aa", + "1ac91534a5ed4970be6196871f2be040", + "c4f9e4c264e94c62b737acaff2c8c775", + "936f01d81c484a7c8ea5f455e7a63e6e", + "b6be4e19602b4168b116d445bde77ad9", + "d6f24bd763644c03804764e18b642367", + "26563ac09615470689726f97a5c43f91", + "11e8627edc5046328ae16f7f315e8db2", + "1d1fca66dcc2424f86d20ca6b7cb8604", + "c7296bbfb73f425f9cfa538a12564722", + "f3217fc0c1ec49d0821694ce94a4d674", + "2f51666cb65b491594f50b092186f7ed", + "cb1fc876416c4905b406f489069087d1", + "06da6f6794f6417bb1978f0d8d7c558a", + "628a631c347a4b78be5d6079616555a6", + "bd35fa9d9d7e4509ac252c0622d3cc7b", + "a088dbcf1a334ca2a7e1d185869ef488", + "55c9d349f42642a798cfaf088072929a", + "6e066ae08505455a8cfe2a72fd30e5ce", + "d174280a4bc94bfb8de2fee838a935f6", + "11653f75d150432f942d475affee8770", + "f6fb93f0f73a45d5a9e9069171f4a350", + "2a0da366009c4505baa4a6141838ad32", + "424f39af879b4b5e98685e801b663857", + "d65fb8edbe2846cc90cd6a1c879b3407", + "e18f16f3c48144519051fbaad8887a4a", + "6c5598eed422416d96b6721bfc4cf431", + "7ea67e6296424f059c4850e2fd5bc650", + "840facc820c24863a90d41aaeb957c67", + "33f716e5d2bf47b48b704d4e34f0b602", + "fe2d91cfcb8b4e6ba88de7db977618e9", + "42192f5259c64d2ba909e0d2e7212b65", + "c3300b53ac9746dcad09a2b32a056b6f", + "f314a42f7b3d453a8c8289b567dc695f", + "4fd21e0ecf4546afad2bd50bc72e58e9", + "03663ad72b8b4618ba74bb5e21dc92e5", + "d1530951f8bc4a44aed0171ea7fb7a87", + "35b618e0f178450fa533aa4cae01470a", + "7747f1b60b054a219d02e2a93873b04c", + "94976850944d4af7a0e4bc987d9af882", + "d086bd77fd194a208a5207b4812e2e92", + "5c786e4dd11b46a48fc846612c66181b", + "019dfed092d5456186441718c14ed377", + "31e18285da0f49c788a42851d5e5d7b3", + "86f0ae6918474db69381616b361bac7e", + "a55bab394b5a47c0a3ffdab86d51474d", + "b324ac810c1b4900ae04345b6bf1dc30", + "d3ac9c9ac3ca42b29b58e0ef90fbc322", + "42bc0137649247e4b171571298393fd6", + "958ad03e991a40b4a82da8cd0b5180f8", + "5b3b063f264d4b83871811462ae5f238" + ] + }, + "id": "1JZt0cA3pSa3", + "outputId": "5f03db97-605f-4bac-dfc0-d2b70fb23b6e" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "VBox(children=(HTML(value='