CNN_MNIST_model.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of CNN_MNIST_model.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "f7bd7cd117594ace880e1d6259e4a85a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_875fb55c2c3a494592d15010362e2f1c",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_0866cc439fb84973befe5921a6ad8f6c",
              "IPY_MODEL_f6335a7685d847d1830de3d34a975c69",
              "IPY_MODEL_37c8ff9057da4114af150d8d891da8e4"
            ]
          }
        },
        "875fb55c2c3a494592d15010362e2f1c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "0866cc439fb84973befe5921a6ad8f6c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_8dd55136ea044526bb17b95713d23d27",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_4ea2aa6697d24063b725387784854e88"
          }
        },
        "f6335a7685d847d1830de3d34a975c69": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_14ea8bf162e04119ac64fe8e0b8c5e9d",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 9912422,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 9912422,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_1f1b9cdb67af481eb33a7d26d25c973c"
          }
        },
        "37c8ff9057da4114af150d8d891da8e4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_801414f9c2ba4add97867378f908ed73",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 9913344/? [00:00&lt;00:00, 22996923.60it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_583c6cc1fff844a19e39d444d6ca07ef"
          }
        },
        "8dd55136ea044526bb17b95713d23d27": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "4ea2aa6697d24063b725387784854e88": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "14ea8bf162e04119ac64fe8e0b8c5e9d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "1f1b9cdb67af481eb33a7d26d25c973c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "801414f9c2ba4add97867378f908ed73": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "583c6cc1fff844a19e39d444d6ca07ef": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "ff9d0f639c2d4ce9936f070a0dabe870": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_c7dea5630b4b49d792c14bacc85ded00",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_1c65cea52d2848f7bdd7cc52aa515b98",
              "IPY_MODEL_2d74b6557e504c729af524da890d47b1",
              "IPY_MODEL_55cf45710ba14a0b86e4658582202ffc"
            ]
          }
        },
        "c7dea5630b4b49d792c14bacc85ded00": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "1c65cea52d2848f7bdd7cc52aa515b98": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_b2df871fe1b6407fa77b3fdb1c331a68",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_75e06b6703a84646871b34c319c27ca8"
          }
        },
        "2d74b6557e504c729af524da890d47b1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_83dd4ed891b6427c98823be2956f69c8",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 28881,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 28881,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_36acb95e947e40588ae8bd0578208809"
          }
        },
        "55cf45710ba14a0b86e4658582202ffc": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_ffcf9b1005da46d6b1b3ac2203bf296c",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 29696/? [00:00&lt;00:00, 799930.97it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_a54e6896f5694d56904c375cbc9cd073"
          }
        },
        "b2df871fe1b6407fa77b3fdb1c331a68": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "75e06b6703a84646871b34c319c27ca8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "83dd4ed891b6427c98823be2956f69c8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "36acb95e947e40588ae8bd0578208809": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "ffcf9b1005da46d6b1b3ac2203bf296c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "a54e6896f5694d56904c375cbc9cd073": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "0747f0bae1894f49b9dfee9f59a65f12": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_8b41b2240a3c47f9a962fec0841c6943",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_b3e5a1a84f944c13a1c8eb27a05fb155",
              "IPY_MODEL_a221fac65b62406089c5f023a1c44a6e",
              "IPY_MODEL_991a7590cc214cc580ffb796100ecc37"
            ]
          }
        },
        "8b41b2240a3c47f9a962fec0841c6943": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "b3e5a1a84f944c13a1c8eb27a05fb155": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_24b99b801ccd48ffa34dcc32150820cb",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_36ad2cdbbaf1441b9375d490a035619e"
          }
        },
        "a221fac65b62406089c5f023a1c44a6e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_8632438401844a4992fa2c52dab4bceb",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 1648877,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 1648877,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_054ec901b41646a9a165bd12f28ec4b4"
          }
        },
        "991a7590cc214cc580ffb796100ecc37": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_3379670ddf364802bcb3a33c4b49d687",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 1649664/? [00:00&lt;00:00, 16919533.13it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_1f3cc0ef6d8d44feaa55f0bff0d0c354"
          }
        },
        "24b99b801ccd48ffa34dcc32150820cb": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "36ad2cdbbaf1441b9375d490a035619e": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "8632438401844a4992fa2c52dab4bceb": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "054ec901b41646a9a165bd12f28ec4b4": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "3379670ddf364802bcb3a33c4b49d687": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "1f3cc0ef6d8d44feaa55f0bff0d0c354": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "3373a19589654ba6a051a281d819d6bb": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_8477258038fa4e3dafaa9379a1393222",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_e406e95ecc664ef089d3948f91bc87ba",
              "IPY_MODEL_803b827cce1f41cea1b238b10c6eee68",
              "IPY_MODEL_73f7cdfcd7ba495c8f53ce90b0b9f15a"
            ]
          }
        },
        "8477258038fa4e3dafaa9379a1393222": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "e406e95ecc664ef089d3948f91bc87ba": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_1299e18a944b4208a8965f3928c10ee0",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": "",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_01ba7d6c43bf495fb7124120f72ba3f6"
          }
        },
        "803b827cce1f41cea1b238b10c6eee68": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_361b3b39e2de4669b076c0158ae97e8a",
            "_dom_classes": [],
            "description": "",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 4542,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 4542,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_ad950cdcbe9342cab9c1ca36fe052fac"
          }
        },
        "73f7cdfcd7ba495c8f53ce90b0b9f15a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_640e883c272748a592d21c701545cd4a",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 5120/? [00:00&lt;00:00, 150392.43it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_b488b760f82b4223828663e7ca86f4f2"
          }
        },
        "1299e18a944b4208a8965f3928c10ee0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "01ba7d6c43bf495fb7124120f72ba3f6": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "361b3b39e2de4669b076c0158ae97e8a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "ad950cdcbe9342cab9c1ca36fe052fac": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "640e883c272748a592d21c701545cd4a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "b488b760f82b4223828663e7ca86f4f2": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# **IMAGE CLASSIFICATION USING CNN**\n"
      ],
      "metadata": {
        "id": "9oY6VUOpjWld"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Convolutional Neural Network (CNN)**\n",
        " Convolutional Neural Network models are ubiquitous and work phenomenally well on computer vision tasks like image classification, object detection, etc. So, in this notebook, we will leverage one of the popular datasets - MNIST for applying CNN."
      ],
      "metadata": {
        "id": "K6KqBK4YuNxt"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Here's an example of Convolutional Neural Network -\n",
        "\n",
        "<img src=\"https://miro.medium.com/max/1400/1*kkyW7BR5FZJq4_oBTx3OPQ.png\">\n",
        "\n",
        "When a new image is presented to the CNN, it percolates through the hidden layers until it reaches the fully connected layer at the end."
      ],
      "metadata": {
        "id": "12MIfuIVLsDr"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **MNIST DATASET**\n",
        "\n",
        "The MNIST dataset is an acronym that stands for the Modified National Institute of Standards and Technology dataset. It is a dataset of 60,000 small square 28×28 pixel grayscale images of handwritten single digits between 0 and 9.\n",
        "\n",
        "<img src=\"https://theanets.readthedocs.io/en/stable/_images/mnist-digits-small.png\">"
      ],
      "metadata": {
        "id": "8q_iAECbzdy5"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VWvKjL2BAG4k"
      },
      "source": [
        "## IMPORTS\n",
        "\n",
        "In this notebook, we'll be using Pytorch. PyTorch is an open source machine learning framework developed by Facebook's AI Research lab. It is used for applications such as computer vision and natural language processing. Here, we'll also be using matplotlib, PIL and numpy."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k9wHM0ja9iXI"
      },
      "source": [
        "# importing necessary libraries\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as funct\n",
        "import torch.optim as optim\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import time\n",
        "from google.colab import files\n",
        "import os\n",
        "from PIL import Image"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4snA55sAASc4"
      },
      "source": [
        "## DATASET AND DATA LOADER\n",
        "\n",
        "**Dataset** : It is a collection of data. Here, the dataset has two components i.e, the image and its label.\n",
        "\n",
        "**DataLoader** : It fetches the data from a dataset and serves the data up in batches thus preventing the issue of overloading."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458,
          "referenced_widgets": [
            "f7bd7cd117594ace880e1d6259e4a85a",
            "875fb55c2c3a494592d15010362e2f1c",
            "0866cc439fb84973befe5921a6ad8f6c",
            "f6335a7685d847d1830de3d34a975c69",
            "37c8ff9057da4114af150d8d891da8e4",
            "8dd55136ea044526bb17b95713d23d27",
            "4ea2aa6697d24063b725387784854e88",
            "14ea8bf162e04119ac64fe8e0b8c5e9d",
            "1f1b9cdb67af481eb33a7d26d25c973c",
            "801414f9c2ba4add97867378f908ed73",
            "583c6cc1fff844a19e39d444d6ca07ef",
            "ff9d0f639c2d4ce9936f070a0dabe870",
            "c7dea5630b4b49d792c14bacc85ded00",
            "1c65cea52d2848f7bdd7cc52aa515b98",
            "2d74b6557e504c729af524da890d47b1",
            "55cf45710ba14a0b86e4658582202ffc",
            "b2df871fe1b6407fa77b3fdb1c331a68",
            "75e06b6703a84646871b34c319c27ca8",
            "83dd4ed891b6427c98823be2956f69c8",
            "36acb95e947e40588ae8bd0578208809",
            "ffcf9b1005da46d6b1b3ac2203bf296c",
            "a54e6896f5694d56904c375cbc9cd073",
            "0747f0bae1894f49b9dfee9f59a65f12",
            "8b41b2240a3c47f9a962fec0841c6943",
            "b3e5a1a84f944c13a1c8eb27a05fb155",
            "a221fac65b62406089c5f023a1c44a6e",
            "991a7590cc214cc580ffb796100ecc37",
            "24b99b801ccd48ffa34dcc32150820cb",
            "36ad2cdbbaf1441b9375d490a035619e",
            "8632438401844a4992fa2c52dab4bceb",
            "054ec901b41646a9a165bd12f28ec4b4",
            "3379670ddf364802bcb3a33c4b49d687",
            "1f3cc0ef6d8d44feaa55f0bff0d0c354",
            "3373a19589654ba6a051a281d819d6bb",
            "8477258038fa4e3dafaa9379a1393222",
            "e406e95ecc664ef089d3948f91bc87ba",
            "803b827cce1f41cea1b238b10c6eee68",
            "73f7cdfcd7ba495c8f53ce90b0b9f15a",
            "1299e18a944b4208a8965f3928c10ee0",
            "01ba7d6c43bf495fb7124120f72ba3f6",
            "361b3b39e2de4669b076c0158ae97e8a",
            "ad950cdcbe9342cab9c1ca36fe052fac",
            "640e883c272748a592d21c701545cd4a",
            "b488b760f82b4223828663e7ca86f4f2"
          ]
        },
        "id": "1oO2ndseAF8E",
        "outputId": "60238ee8-93a4-458e-babc-78e4152e1ae0"
      },
      "source": [
        "# defining the transformations to be applied on the images\n",
        "transform = transforms.Compose([transforms.Scale((32,32)), transforms.ToTensor(), transforms.Normalize((0.5), (0.5))])\n",
        "\n",
        "# batch size during training\n",
        "batch = 8\n",
        "\n",
        "# creating the dataset for training\n",
        "trainset = torchvision.datasets.MNIST(root='./data', train=True, download=True, transform=transform)\n",
        "\n",
        "# loading the data using pytorch's DataLoader\n",
        "trainloader = torch.utils.data.DataLoader(trainset, batch_size=batch, shuffle=True, num_workers=2)\n",
        "\n",
        "# creating the dataset for testing\n",
        "testset = torchvision.datasets.MNIST(root='./data', train=False, download=True, transform=transform)\n",
        "\n",
        "# loading the data using pytorch's DataLoader\n",
        "testloader = torch.utils.data.DataLoader(testset, batch_size=batch, shuffle=False, num_workers=2, drop_last=True)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./data/MNIST/raw/train-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torchvision/transforms/transforms.py:317: UserWarning: The use of the transforms.Scale transform is deprecated, please use transforms.Resize instead.\n",
            "  warnings.warn(\"The use of the transforms.Scale transform is deprecated, \" +\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f7bd7cd117594ace880e1d6259e4a85a",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "  0%|          | 0/9912422 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/MNIST/raw/train-images-idx3-ubyte.gz to ./data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to ./data/MNIST/raw/train-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "ff9d0f639c2d4ce9936f070a0dabe870",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "  0%|          | 0/28881 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/MNIST/raw/train-labels-idx1-ubyte.gz to ./data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to ./data/MNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "0747f0bae1894f49b9dfee9f59a65f12",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "  0%|          | 0/1648877 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/MNIST/raw/t10k-images-idx3-ubyte.gz to ./data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to ./data/MNIST/raw/t10k-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "3373a19589654ba6a051a281d819d6bb",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "  0%|          | 0/4542 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/MNIST/raw/t10k-labels-idx1-ubyte.gz to ./data/MNIST/raw\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0zFAssi0MfCb",
        "outputId": "eb6f7121-1ae7-4393-cbfc-bcec941b9a7f"
      },
      "source": [
        "# the different class of images available in training set\n",
        "classes = trainset.classes\n",
        "classes"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['0 - zero',\n",
              " '1 - one',\n",
              " '2 - two',\n",
              " '3 - three',\n",
              " '4 - four',\n",
              " '5 - five',\n",
              " '6 - six',\n",
              " '7 - seven',\n",
              " '8 - eight',\n",
              " '9 - nine']"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kUIqJgi1F0vO"
      },
      "source": [
        "## DATA VISUALIZATION"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7DWjOGEGHn1l"
      },
      "source": [
        "# function to visualize the dataset\n",
        "def imshow(img):\n",
        "    img = img / 2 + 0.5     # unnormalize\n",
        "    npimg = img.numpy()\n",
        "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
        "    plt.show()"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 112
        },
        "id": "jX6VkVnrF3uU",
        "outputId": "7f1266aa-e397-4173-d47a-5da4d1894c9a"
      },
      "source": [
        "# shows the imput of dataloader in the form of grid of batch number\n",
        "detaiter = iter(trainloader)\n",
        "img, labels = detaiter.next()\n",
        "imshow(torchvision.utils.make_grid(img))\n",
        "print(' '.join('%5s' % classes[labels[j]] for j in range(batch)))"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAABOCAYAAAA5Hk1WAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO2daXBbV3agv4ud4AaCBFeQ4C6KkihKIilS++K2ZbtdcsedTvcknu6uJM6PTmpSmama7klVKv0vM6mZJFM1NVWeJJXuiaszHSeu9hK3N0k0rc0UF3EVV3FfwAUEQYAACODND/K9pmRqaYsEQPl9VSiCDyDfue++d+69555FSJKEioqKisruQxNrAVRUVFRUvhyqAldRUVHZpagKXEVFRWWXoipwFRUVlV2KqsBVVFRUdimqAldRUVHZpTyRAhdCXBBC9AkhBoUQP9wuoVRUVFRUHo34sn7gQggt0A98DZgAmoHvSJLUs33iqaioqKg8iCeZgdcBg5IkDUuSFAT+Cbi4PWKpqKioqDwK3RP8bR4wvun3CeDo/V8SQrwGvAag1+uPZGRkPMEpVVRUVL56TE9Pz0uSZLv/+JMo8MdCkqTXgdcBcnNzpddee22nT6mioqLyVPHjH/94dKvjT6LAJ4H8Tb/bN449ku7ubkZHt5Qn5jgcDvbt24ckSXzwwQdEIpFYi7Ql58+fx2g0MjY2RldXV6zF2ZL09HSOHl1flF26dAm/3x9jibbm2LFjWCwWFhcXuXHjRqzF2ZKEhATOnj0LwI0bN1hcXIyxRFtTVVWF3W7H7/dz6dKlWIuzJRqNhueeew4hBF1dXYyNjcVapC0pKipi7969D/3OkyjwZqBMCFHEuuL+NvDvHucPR0dHaW5ufoJT7yyyAm9ubo5bBX7q1CmMRiNzc3Nxey0LCwsVBd7e3o7H44mxRFtTVVWFxWJhZWUlbq9lSkqKosB7e3vjVunk5ORgt9tZW1uL22up1Wp59tlnEUIwMjJCS0tLrEXaEo1Gs3MKXJKkkBDiD4EPAC3w95IkdX/Z/6cSH5jNZiRJIhwOEwqF4nYAU1FReUIbuCRJ/wb82zbJohJjTCYTf/Inf4LX66W3t5f29nZmZmZiLZaKisoD2PFNTJXdgU6nIyMjg/Pnz2M2m7l9+zZ6vZ7PPvsMl8sVa/FUthEhBGazmZycHCorKzl06BA6nY7R0VH6+voYGBhQB+5dgqrAVRSEEKyuruJwODhw4ACzs7PcvXuXpaUlYl34Q6/Xk56eTkFBAVarlYWFBbq7u1ldXY25bLuN9PR0Kisrqa6upra2lrq6OnQ6HYODg3z66aesra0xPz9PKBSKtagqj2BXKnCdTofJZCIxMZHExMR7PpMkiUgkQjAYJBAIKA+3JEkEAgH8fr/6wG9BJBLB5/PR19eH3W5XHnK73c6dO3di/jAbDAbsdjtnzpyhrKyMzs5ORkZGYtafQggMBgNGo5HExEQMBgNarRaAcDiMx+NBr9eztrbG6uoqgUAgLvYTdDoddrudEydOcPz4ccrLy8nLy0Ov12M0GllcXGR0dJTBwUHm5+djLe49aDQaDAYDZrMZs9mMEIJgMIjX62VlZSXW4n0BrVZLQkICKSkpaLVafD4fXq93W72xdp0CF0KQmJhIbm4uJSUllJSUIEnSFxT1wsICCwsLRCIRJEkiGAwyMzPD9PQ0a2trMW5F/BGJRPB4PFy/fp2ysjIqKyspLi4mPz8fo9FIOByO6cBnNBqx2+3U19dTU1ODEIJf/OIXuFyuqCpGjUaDVqvFZDKRlpZGZmYmDoeD9PR0EhISiEQirK6uMjAwQGpqKm63m8nJSaampvD5fFGT80GYTCYKCwupq6ujrq4Os9mM2+0mMTGRlJQUiouLqayspLu7O24UuBBCUYZWq5XCwkIKCgrQaDQsLCxw9+5dent742KA3IzZbCY/P5/9+/djMpkYGRlhaGiIycnH8rZ+LHadAtfpdOTm5nLq1CkuXrzIhQsX7uk4efbT399Pf38/4XCYcDiMy+Xiww8/xOVyEQ6H466z44G1tTWuXLlCXV0dxcXFWK1WcnNzSU5OJhAIxHQWrtfrSU5Oxmg0EgqFWF1dxWAwoNPpiEQiUelPjUaD2WwmIyMDu91OSUkJ+/btU1YFFotFudfee+89CgoKGB8f57PPPuODDz6IC9e/pKQkMjIySExMZG1tjampKXp6enA4HOTm5pKTk0NtbS0DAwO0t7fHfLWq0WgwGo1YLBby8vLYv38/p0+f5tSpU2i1Wrq7u3n//fcZGRnB6/XGVNbN6PV6srOzOXfuHN/97ncxm828++67vPPOO19tBZ6cnExhYSH79u1jz549BAIBNBoNQghgfbROSUmhtraWmpoaAILBIH19ffz85z8nFArF/KZ8XIQQaDTr6WrklcROIkkSTqeT8fFx5ubmSExMJD09HYvFgsvliqkCz8vL4+TJk5SWlmKz2Xj22WdpbW2lsbGR2dlZVldXd1SJy/dVXV0dr776Ki+//DIJCQkASh/J79PT03n11VeVY3V1dbjdbmZnZwkEAjsm4+Pg9Xrp7+/n8uXLdHR0MDAwwD/+4z9y8eJFfvu3f5uSkhIsFgtpaWkYDIaYyqvRaEhJSWHv3r2cOXOGkydPcuTIERITE4lEImi1WiorK/F6vbS1tdHe3r7j98HjkpubS01NDefOnaOsrIyOjg76+/uZnp7e1vPsOgXudru5fv06o6Oj3Lhxg9/5nd+hvLxcUeJarRaz2UxiYiI6ne6eY0VFRQwODt5jG49XNBoNFouFoqIiVldXmZmZiVr03cjICIODg2RnZ1NbW0ttbS1OpzOmD7M8CzObzRiNRvLy8vjRj34EQGNjIxMTEwSDwR05t6xITp8+zQ9+8AOqq6tJSEhQbN6b2TyRkHE4HPzZn/0ZWq2Wt99+O6YzRZ/PR3t7O0NDQ+h0Ovx+Pz6fj4WFBXw+H0ajkczMTHJzczGZTASDwZg8K/IM9uLFi3z729/GbrdjNBrxer388pe/pKOjg9LSUqqrqzlx4gTZ2dk0Nzfz13/910xNTREOh6Mus4xOp6O+vp5XXnmFhoYG3G43ra2ttLS0MDExsb3n2tb/FgXC4TBut5tgMIjP50OSJLKzs5VZUFJSEkVFRYoZQKdbb+JmO3k0EEKQmZlJTU0NaWlprK2tMTMzQ2tr6wMjEjUaDZmZmRQUFCj257m5OW7fvh1V2WdnZ5meniYQCCimAovFgsfjidn+gUajQafTKQO10WikoKAAi8WCwWC4R2FuJzqdDovFwqFDh/it3/otqqqqsFqt98y63W43CwsLilmnrKzsnv9hNBqVe7Knp4ehoaGYbbrJJsbNg0gkElGCtuTrbDQalWsdbQUuexxVVVXx/PPPU1FRwfLyMp2dnbS0tHD9+nXm5ubQarVUVFSQlpZGZWUlRqORv/3bv0Wj0cREgQshSEtLo6GhgbNnz7Jv3z6Sk5Npa2ujpaVlR1Zgu06Bw7oyXl1dZXp6mo8//lh5gHU6HZmZmdTX11NWVnaPR8rKygpLS0tR2YzT6/VYrVYaGhp48cUXsVgsOJ1OWlpa6O7u3lKBy14WBw4coKqqioqKCkwmEx9++CE+n2/HZpdbsby8zNLSEmtra1gsFrKzs0lNTWV2djamG8BCCOUF69d5s/lsJ9DpdKSmplJZWcnp06dJS0vD5/PhcrlYXFxkbW2Nubk5pqam8Hg8mM1m/H4/DocDs9mMVqtVvCf27NnD3r17cbvdBAKBmF1LOdJ2KyKRiBKFG4uZt8FgwGazsWfPHk6ePMm+fftwu920t7dz7do1rl69yvDwMKmpqYTDYTQajbI/Yrfbd3QwfxQ6nY6ysjIuXLhATU0NNpsNv9/P8PAwPT09O5JKYlcqcPiVt8nc3BzwKxtlRkYGKSkppKSkIIRgbW0Nl8vF6Ogok5OTO7oklM016enp1NfX8/3vf5+ysjLm5uYYHR1lfHz8Cy5E8sCTlZXFyy+/TENDA3l5eYTDYa5evcrt27dxOp1RTQQlu2BGIhFMJhOpqakkJSWh1+ujJsNWbFbeMnq9Hp1Ot6U5Y7vOKXudJCYmEggEuHPnDjdu3ODatWu43W7FPWxtbQ2j0cjNmzf5/d//fcrKykhJSVEGGnmAnp+fx+/3xyxYRm6TPPgJIUhISECn0xEKhfD5fLjd7pgo8YyMDKqrqzl79izPP/88Op2OS5cu8dFHH9HS0sLk5CQ+n4+cnBzS0tK+4EYcK2S30uPHj1NXV4fD4UCr1TI+Pq4k71tdXd328+5aBX4/FouFAwcOcOzYMc6cOUNeXh6SJLG4uEhXVxeNjY077kKYlJREcXExR44c4cKFCxw+fJjBwUEuXbrEtWvXuH379j2dKNvmc3NzOXv2LN/5znewWCxMTEzw2Wef8d5779HV1YXP54vqxozP52NlZYVQKKQsqfV6/Y4pycdlKzNYdXW1MjCOjIxse//KQS3Nzc38wz/8A5Ik0dPTQ0tLC3fu3PmCa6AQgv7+frxeL88++yy1tbWUlJQoLmUvvPACqampmEwm3n333W2V9XFJT0+npKSE3NxckpKSADhx4gSlpaUATE1NKdc0mgrcbDZTU1PDhQsXOHXqFFlZWVy9epU33niDjo4OxS0YwG63U1BQQGJiojJB++ijj1hYWIiJ+USr1ZKUlMTJkyfJz89Hp9MxOTnJ5cuXuXz5srL63252vQI3GAwcPnyY/fv3c+TIEaqqqigtLcVgMDAyMsKNGze4evUq165dY2VlZcduSNn2Ke+WHzp0iKWlJa5evcoHH3zAnTt3WFlZuUcRm0wmcnNzqa+v5zd+4zcoLi5mZGSEW7du8cEHH9DT04PX6436LCgQCMTcbfBxOXjwIIuLi4TDYYLB4LanKQ6FQiwtLSk2THlSsLCwsOVgIUkSbreb9957j+npaRYXF3n++ecpLS0lJSWF/fv34/f76e/v31Y5HxchBHa7na9//evU1NQoCtzhcGCxWJidncXpdDI7Oxv1/i8tLeXYsWMcPnyY9PR0enp6+NnPfkZLSwtutxtJktBoNNhsNg4cOEBJSQkGg4Hh4WHeffddfv7znzM3Nxd1L5TNjhLl5eWkpKQwNTVFW1sbly9fpqenZ8eu5a5U4DqdDrPZTFpaGna7nVdeeYUDBw6Qn5+P1WpFr9czPj5OY2MjH3/8MR0dHUxMTOzYyKzVaiksLOTYsWPU1dUpniMtLS20tbUxMTHxBeWt0WiwWq3s2bOHhoYGiouLmZ2dpb29nY6ODsbHx2MWJh7tDd8vgyRJeDweFhcXiUQiJCQkKMpou4lEIqysrDA8PAysK/WHzfQlScLn89HR0UFqaiqpqamkpKQoMzN5MzYWCCFITU2lrKyMgwcPYjKZABSvGjnS0Wq1kpaWhtvtjppCrK6uprKykrS0NFwuF+3t7TQ1NeHxeBTlnZiYSE1NDUeOHMFutxOJRJiamuLmzZuMjY3FxGtGNp+kpaUp+mdxcZG7d+8qz/FOsWsUuLyUNxgMpKamkpeXx549ezhy5AinT58mKysLg8FAKBRienqa5uZmPv74Y9rb23csCk6Wx2KxcOzYMY4ePUphYaGyzL506RKdnZ0sLS3d8xDImy7FxcVUV1dz+PBhIpEIt27dorGxkfb2dhYWFmI2A461knkcJEliamqK5uZmOjs7GR4e3tF84+Fw+Nd6EOWZuKzEDx48SH7+ev0TWVEajcaYuGaGw2H8fj+BQACDwQCA3+9X0gI4HA4OHz5Mf38/nZ2dirfXTlNeXo7D4cBoNCrh/LOzs8rEy2g0kp2dzfHjxxVFv7CwwPLyMhMTEzuqKB+GPPsuKChQAs2WlpaYn59neXl5RwfAXaHA7w9dLiwsZO/evRw+fJjjx4+TlpaGx+NRvAGGh4f55JNPaGpqUlwOtxudTkdKSgrZ2dns37+fF198kdLSUiVo6MqVK1y6dInp6Wkikcg9m2+pqamUl5fT0NBAQ0MDDoeD1tZWfvnLX9LU1BRzP9bNfvTxSiQSYXR0lMbGRjo6Opibm4vLij9jY2PcunWL2dlZYF2xy4N+dnZ21CtTSZLEwsIC7e3taLVaMjMzgfUZeH5+Punp6YotfHx8nOnpaWZmZqIy0OTm5pKenq4MfvPz80owmxwXsW/fPk6cOEFRURFGo5H5+XnW1tZimqZAp9NhtVqprq5Gr9fjcrmYnp5mdnZ2x33+4/cJ3UDeIc/Pz+fw4cOcP3+e2tpa7HY7SUlJCCGIRCIMDw9z+fJlrl69SmdnJ9PT0zv2QGs0GsW17JlnnuEP/uAPsFqtdHZ28sknn/DJJ5/Q1tZGMBjEaDR+oT2VlZV8//vf5+TJk9hsNgYGBnjzzTd56623YjaL2Exubi4FBQUkJyfHWpQHIkkSLpcLp9OpzMLiFdltT/aztlqtSgj+T37yk6jL0tfXx/j4OP/8z/+MXGQ8NzeX3/zN36S+vh6Hw8HevXu5ePEi/f39rK6u4nQ6d1w2r9dLIBBQPJ+ysrKwWq0Eg0H0ej179+7lpZdeorS0FKPRqNie5URcsfBZh18Nfi+99BIGg4G2tjYaGxu5efMmU1NTO3ruRypwIUQ+8FMgC5CA1yVJ+hshhBX4f0AhMAJ8S5KkbU8cnZycTHV1Nd/73vd45plnSEtLu8cjQnYn7Ojo4MaNG7S2tjI1NbWjHWm32/na177Gc889x5kzZ7BarQghsNlsXLhwgfr6elwuFxMTE4yNjSmzadmGd+bMGfbu3UtiYiJ+v5+1tTUyMjJi5r96P0ajEZPJFHcz8M1uhFqtluLiYjIzM7l79y4ejyemq5aH4ff7GRwcZGZmhqysLBITE8nLy6O8vDwm8sg2fZ/Pp7gy9vb2MjU1RXd3Ny+88AL19fXYbDYOHjzI4OBgVBT4m2++idFo5MSJExw5coT8/HzOnTuHVqslOTmZ9PR0cnJy7nEdTE1N5cCBA3znO9/hL//yL6MeSm80GklPT6eoqIji4mI0Gg1XrlyhtbU1KtfscZ7QEPAfJUlqFUIkAy1CiI+A7wGfSJL0F0KIHwI/BP7zdgsohECv12M2m0lKSlIixDZ/bjAYOHv2LBUVFUxOTjI8PEx3d7cS9eZ2u7dVpszMTOrq6jh48CCpqamKUklPTyc5OVnZ5FpdXb3Hi0T2+c7IyFBmt36/n0gkEpUlqsViobi4mKNHj5Keno7T6WRkZITR0VEmJiaU5V5ubi75+fkkJSWxtrampOuMZjDR/ayurjI7O4vH48FmsyGEoKysjKysLCVgJl4VeCAQYGRkhLm5OWw2GwkJCeTm5ioBW4ODgzHZ77g/Cdzw8DC3b9+moKBAib7ds2cPNpuNwcHBHQ88khNTeb1ejh49SnFxMTU1NffsfxkMBoLBIDqdDo/HQ19fH01NTbz11lvKsxRNTCYTNpsNu91OYmIiGo2GtbW1qPnQP1KBS5I0DUxvvPcIIXqBPOAicGbjaz8BrrADClxOA3vz5k2MRqOSzCghIUEJRIhEIthsNtLS0nA4HJSUlFBaWkpOTg7hcJjBwcFttUUtLy/T1dVFKBRSAoaAe7Liye/laDGLxUJ+fj6lpaUkJyfj8XiYnJxkYGCAtrY2Ojo6dvQhNhqNlJSUcOHCBaUKu2yrm5ycZHx8nLGxMUZGRigqKiI7OxuDwaAkYdqpvYTHxe12MzQ0xIEDByguLgbWV2fySiFeVi8yclh1WloaRUVF7NmzB4vFothzU1JSKCsr41vf+hZ/9Vd/taMuro+L1+tlamqKvr4+RkdHqa6uxuFw4HA46O/vx+l07qiCdLvddHZ2EggEGB8fp6ioCJvNRl1dnWLqcblcrK2tkZKSgtPppLe3l1u3bjE8PByTJFZGo5GUlBRlFe5yuVhaWmJ1dTUqE4pfa40shCgEDgE3gawN5Q4ww7qJZau/eQ14DdaXO78usgK/du0aHo+HqqoqCgsLlSTp8iaHzWZTXLaSkpKUzZn+/v5t30yYn5/n+vXr3Llz57HMDHKIrRxOHQ6HGR0d5datWzQ3NyuuhjulwOV9hMLCQo4fP051dTUajYaVlRXy8/NZWVnB4/EwNDREa2srFRUV2Gw2dDod8/PzSs7tWOXGAFhZWWFyclJZTcn2T/keiCc0Gg1JSUns37+foqIiysrKFN9mWYGbTCZycnJ48cUXuXTpEktLSywtLbG4uBizDblIJILL5WJsbIzJyUlqa2ux2Wzk5eVhs9mYn5/fUSUpp7f1er2Mjo5is9moqKhgz549pKenKznrYX2jfW5ujqGhIfr7+2M2AMob0rJ8ExMTSh/GlQIXQiQB/wL8sSRJy5sfGkmSJCHElldPkqTXgdcBcnNzf+0rLOdX7urqYmZmht7e3i/MwPV6vZJDpKysjMzMTJKSknA4HOTk5JCcnLytYctut5uurq6Hutltzkonm34kSSIUCrG4uMj169e5dOmSorx3cnkqJ3+yWq04HA5sNhsLCwtKpJ3VaqWsrIzS0lIqKiqw2+1kZmai0+mUcmo5OTmsrq4qYeDRJhgMsry8HJeeJjKbw+7Ly8v5xje+wd69e8nOzqawsFDZdAcUu25VVRXf+9738Pl8tLa20tTUxNDQ0BPLIT8bv27uH7/fr+R5kfdsMjIysFgsURkofT4fPp+P+fl5ZmdnSU9PV/K+h8NhJdEWwMLCAuPj4zu+5/UwTCYTGRkZZGdnEwqF6O3txel0xtcMXAihZ115vyFJ0r9uHJ4VQuRIkjQthMgBdsxiH4lE8Hq9ysh8n2xotVr27dvH6dOnOX36NHV1dYqdNCEh4QueIE+KXCTiUci5oY8dO8b58+cpLS1lcHCQxsZGJdLS6XRGJc+33+/H7/crtrnR0VGuX7/O7OwsRqORqqoq6uvrOXnypLJBHAwGSUpKora2FovFQmtrK21tbQwODu6ovLsReSmdkZFBYWEhL730El//+tfJycn5wiptc39rNBqys7Px+/3cvXtX8cv+ssiVaxwOBwaDgYmJiV/b/LU5aZhcYSjaQWVarZasrCzOnz9Pbm4uCQkJhEIhMjIyMBgMDA0NcefOHUZGRmLqgWQ0GklNTcVqtRIIBLh16xbz8/NRMzc+jheKAP4O6JUk6X9s+uht4LvAX2z8/MWOSPgI5FntwMCAUt6qoqICq9WK1+vF4/HEZNYmz7zr6ur4vd/7PQ4ePIjL5eLNN9/k7bffZnJyMmq5JiRJUjIMyhs9TqeT9vZ2BgcH0ev1SJLE/v37sVqtwPoqY3x8nImJCQKBAHq9fssapLEkXkwnJpNJqc5TXV3N3r17eeaZZ5RV4v0EAgGWl5eZm5vD6XTyN3/zN0ocw5N4LsiZE0tKSvjmN7+J2WxW8unMzs4+UqnI92xmZiZZWVlotVolDfLs7GzUvTsKCwt57rnnMJvNSiSmyWQiHA7T0dFBa2srY2NjMcvqKEeEyzVRV1ZWuHXrlmKnj4oMj/Gd48CrQKcQon3j2H9hXXH/XAjxu8Ao8K2dEfHhCCGwWCxYLBYKCwspLCzEYrEoGzITExMsLS3FRK7c3FzOnDmDw+FQOvfKlSuMj49HvaiEPNAFg0HC4bDiCWE2m8nMzOSZZ54hOztbCfSQfdpv3ryJy+UiEAiwsrKy7R49j8vmQtWBQEDx+401er2eoqIiTp06paz+kpKSSExM3FK+QCCguOvdvHmTwcFB2tralPSyT/Lgy3b12tpavvnNbyqFdDUaDd3d3Yo57EHIecsrKytxOBxIksTa2hp+vz+qG9hGo1FJkyFv/ALKisDlcjE0NMTExMSORt8+CoPBQEpKipIV0ev1Mjw8jNfrjZpH1ON4oXwGPOhJOb+94jweso1Pq9ViNBqVEPajR49SUlKCVqvl7t27tLe3K+kno41Wq1Uq2pjNZiW0/u7duzGz48oBGSsrK9jtdk6fPk0oFMJisVBeXo5er2d+fp62tjY+++wzrly5Qnd3d1zUGgwGgywsLDA1NYXT6SQvL0/ZENzpnOAPQi7Acfz4cc6dO0dNTQ2FhYVbfldWnE6nk7a2NpqammhqamJ2dlbJ5/KkmEwmMjMzKSsro6CgAICGhgbFI2ZmZuah55ELI8iBcsFgELfbjdfrjWqOEavVSmVlJQ0NDfesYuTi5D09PcqeWCwjMOUArVAopLy8Xm9UXULjK1LjMZA3iZKSkkhNTSU9PZ1XXnmFiooKsrOzMZlMOJ1Orl69SlNTE9PT0zFxf5NdB4uLi1ldXaWrq4tPP/00pva6hYUFOjo6yMrKory8nPz8fDQaDaFQiJWVFUZGRujp6eH999/n888/V2YT8YDf72d2dlbJ6y5XYUpOTlaWsNEcGGVPk0OHDvHiiy9SX19Pdnb2Q/8mEAgoaRauXLlCV1fXtsokL+nlItTy3kZKSgpFRUXMzMw8dGaYnZ2tuA3KXicjIyPMz89HbcVoMpkoKCigvr6e8+fPKwUa5NXAwsICTU1NtLW14XQ6Y+raKmfADAQCMZMjLhX4g2ZTQgiSk5MpLy/n4MGD1NXVceDAASoqKtDr9SwsLHDnzh2uXLnCO++8Q19fX8xq+kUiEXw+H2NjYywsLCjRbLFM0To4OMhPf/pTent7efnll6mqqsJgMCiK/ebNmzQ1NeF0OvH5fDGtvnM/8gMcDAZZW1tT+rS6uprh4WEWFxejOjgmJCRQWVnJH/3RH9HQ0PDIvQFJkpicnOTTTz/lypUr9Pb2brtMXq+X8fFxbt++zbFjx3A4HCQkJFBRUUFZWZlSGFuSpC1rd27OOyK7ur7zzjtb5j3fKfLz86mqqqK6upqcnJx75FtYWODatWu8//77DA8Px1R5w7orY3Z2Nna7naysLCYnJ6O+EowrBS5nGiwqKiIr6163cjmHRFVVFfv371dSxxqNRsbGxhgaGuLGjRs0NzfT19fH3NxcTIvwygFEr7/+upJgK9YVyf1+v+Kn+vnnn2M2m5UZuFzEQQ5Jj4fK3vfj9Xrp7OwkLy9PyUZXVVWl7HPIGeCigdlsprKykpycnMeyx0uSxPj4OCMjIzuW3N/r9TIwMKCYaU6ePMn58+cpKyvDarUSCoUYGhrC5VlfsPYAAAfTSURBVHKRlZVFTk4OFotF+fvZ2VlmZmaUgg5y/cloXVMAm81Gfn4+OTk5yrFQKMTw8DCNjY288cYbdHd3x1x5w68GPDlKVM7kGAwG8Xq9UXmG4kaBJycnU1hYyOHDh7lw4cIXgn5kl8DMzEzFN9Tr9dLa2sr169fp6upicHBQCfaIdTEC2dOjqalJCUWPtVKUZ7Eejwefz3eP0pEjR2Mt48NYXV1lYGBAqYP46quvKmaqkpISBgYGoqZsNBqNkkN7q1JvMmNjY/T393P37l2uXbtGc3MzLte2pwwCUFz+5Ox8LpeLjo4O0tPTSUhIIBKJsLi4iN/vJzExUYlklVlZWVEG8eXlZaWwQ7SUpc1mY9++fezbt4+8vDxg3S+8u7ubDz/8kMbGRnp6emJq995MMBhkaWmJhYUFVlZWSEpK4ty5c0o92WiY9OJGgZtMJrKysqioqKChoUHx3fb7/YpzvxypNTg4yOrqKi6Xi87OTrq6uhgdHcXlcsXV0n91dZWZmRlCoVBUiik/LvGuqB9EOBxmaWmJ/v5+dDodycnJpKamMjw8zNjYWFQzOcqyDA8PEwqFlLJzZrOZ1NRUWltbWVpa4u7duwwMDCiKfGpqakcf7M2D9MjICE6nE4PBcI9vvxwQc38t0c2bcWtra4rHT7TulYSEBCX9QEJCAh6PR1HeTU1N9Pb2RrXAxKOQi6XLG702m43jx49z8+ZNlpaWCAaDOy5r3ChweQYjZxeUlbDb7WZxcZGlpSWWl5cZGxtTghM8Hg8TExM4nU6Wl5fvsY3GA79uEQCVhyPfG7LNHtZNGbJP9U7NbLciGAwyOTlJc3MzIyMjyua61WolJyeHd955h/HxcWZmZhTZZHfMaNyjcsbBlZWVHT/XdhEOh/F6vUqE5fLyslJVa2BgAJfLFfOV9WZCoRDLy8tMTU0xNDSkBL7Jm+oajearo8D9fr9SYWVzJ7ndbpaWlnC73Uo+jKmpKaXobrxmoFPZOeRCw5cuXYqZDH6/n4GBAVZXV0lISEAIcU/RkbfffnvHZ9tPGx6Ph46ODvx+P21tbXg8Hj7//HNu374dl9dRkiTF7TYcDpORkUEoFFKKTMRFNsJosby8TE9PDz09Pbz11luxFkdF5aHIM/DJyclYi/LUsLy8rPjH7xacTidOp5PGxsaYnD9+ix6qqKioqDwUVYGrqKio7FJiYkIpLCyM24rncgiyEIK6urq42hTdjF6vB9arAx09ejTG0myNnBgL4NChQzH3g38QSUlJwLora7xey83ufrL/eTxis9mA9fszXq/l5tQLRUVFcVc6UMbhcDzyOyKaCio3N1d67bXXonY+FRUVlaeBH//4xy2SJNXcfzw+p8EqKioqKo8kqjNwIYQH6IvaCeOHDCB68cjxgdrmrwZqm6ODQ5Ik2/0Ho2386dtqGfC0I4S49VVrt9rmrwZqm2OLakJRUVFR2aWoClxFRUVllxJtBf56lM8XL3wV2622+auB2uYYEtVNTBUVFRWV7UM1oaioqKjsUlQFrqKiorJLiZoCF0JcEEL0CSEGhRA/jNZ5o40QYkQI0SmEaBdC3No4ZhVCfCSEGNj4mRZrOZ8EIcTfCyGcQoiuTce2bKNY539u9HuHEOJw7CT/8jygzX8uhJjc6Ot2IcQLmz770Uab+4QQz8VG6idDCJEvhLgshOgRQnQLIf7DxvGntq8f0ub47Gu5yOlOvgAtMAQUAwbgNlAZjXNH+wWMABn3HftvwA833v8Q+K+xlvMJ23gKOAx0PaqNwAvA+4AA6oGbsZZ/G9v858B/2uK7lRv3uBEo2rj3tbFuw5docw5weON9MtC/0bantq8f0ua47OtozcDrgEFJkoYlSQoC/wRcjNK544GLwE823v8EeDmGsjwxkiR9Cized/hBbbwI/FRa5wZgEULEZyamh/CANj+Ii8A/SZIUkCTpLjDI+jOwq5AkaVqSpNaN9x6gF8jjKe7rh7T5QcS0r6OlwPOA8U2/T/Dwi7KbkYAPhRAtQgg5c1eWJEnTG+9ngKzYiLajPKiNT3vf/+GGueDvN5nGnro2CyEKgUPATb4ifX1fmyEO+1rdxNx+TkiSdBh4HviBEOLU5g+l9XXXU+27+VVo4wb/GygBqoFp4L/HVpydQQiRBPwL8MeSJC1v/uxp7est2hyXfR0tBT4J5G/63b5x7KlDkqTJjZ9O4C3Wl1Oz8lJy46czdhLuGA9q41Pb95IkzUqSFJYkKQL8H361dH5q2iyE0LOuyN6QJOlfNw4/1X29VZvjta+jpcCbgTIhRJEQwgB8G3g7SueOGkKIRCFEsvweeBboYr2t39342neBX8RGwh3lQW18G/j3Gx4K9YB70/J7V3OfffcbrPc1rLf520IIoxCiCCgDPo+2fE+KWK968HdAryRJ/2PTR09tXz+ozXHb11Hc3X2B9R3dIeBPo3XeaL5Y97K5vfHqltsJpAOfAAPAx4A11rI+YTt/xvoyco11m9/vPqiNrHsk/K+Nfu8EamIt/za2+f9utKmD9Qc5Z9P3/3SjzX3A87GW/0u2+QTr5pEOoH3j9cLT3NcPaXNc9rUaSq+ioqKyS1E3MVVUVFR2KaoCV1FRUdmlqApcRUVFZZeiKnAVFRWVXYqqwFVUVFR2KaoCV1FRUdmlqApcRUVFZZfy/wGyyUKazbuD+AAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "3 - three 4 - four 2 - two 1 - one 2 - two 3 - three 8 - eight 1 - one\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lmn5NrXmNofU"
      },
      "source": [
        "## MODEL\n",
        "\n",
        "Here we define the training pipeline.\n",
        "\n",
        "<img src = \"https://miro.medium.com/max/1400/1*0NwaOkzvom6YpMZoIgWTiQ.png\">"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Convolution** : It is a simple application of a filter to an input that results in an activation. Repeated application of the same filter to an input results in a map of activations called a feature map, indicating the locations and strength of a detected feature in an input, such as an image.\n",
        "\n",
        "<img src = \"https://miro.medium.com/max/543/1*Ff6mG3aPdWFkjb6hZPy3Xw.gif\">\n",
        "\n",
        "**Pooling** : Its function is to progressively reduce the spatial size of the representation to reduce the amount of parameters and computation in the network.\n",
        "\n",
        "<img src = \"https://developers.google.com/machine-learning/practica/image-classification/images/maxpool_animation.gif\">\n",
        "\n",
        "**Linear** : It creates single layer feed forward network with input as in_features and output as out_features.\n",
        "\n",
        "\n",
        "\n",
        "**ReLU** : It helps to prevent the exponential growth in the computation required to operate the neural network.\n",
        "\n",
        "<img src = \"https://miro.medium.com/max/357/1*oePAhrm74RNnNEolprmTaQ.png\">"
      ],
      "metadata": {
        "id": "ks8lz7DncpB7"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kdpiE08aNoD0"
      },
      "source": [
        "# Model\n",
        "class NNmodel(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(NNmodel, self).__init__()\n",
        "        # MNIST datasets have 1 channel\n",
        "        \n",
        "        # Conv2d(in_channel, out_channel, kernel_size)\n",
        "        self.conv1 = nn.Conv2d(1, 6, 5)\n",
        "        \n",
        "        # MaxPool2d(kernel_size, stride)\n",
        "        self.pool = nn.MaxPool2d(2, 2)\n",
        "        \n",
        "        # Conv2d(in_channel, out_channel, kernel_size)\n",
        "        self.conv2 = nn.Conv2d(6, 16, 5)\n",
        "        \n",
        "        # Linear(in_features, out_features)\n",
        "        self.fc1 = nn.Linear(16*5*5, 120)\n",
        "        self.fc2 = nn.Linear(120, 84)\n",
        "        self.fc3 = nn.Linear(84, 10)\n",
        "    \n",
        "    # forward pass to predict the result\n",
        "    def forward(self, x):\n",
        "        # different layers\n",
        "        x = self.pool(funct.relu(self.conv1(x)))\n",
        "        x = self.pool(funct.relu(self.conv2(x)))\n",
        "        \n",
        "        # view(-1) reverses the dimension of the tensor\n",
        "        x = x.view(-1, 16*5*5)\n",
        "        x = funct.relu(self.fc1(x))\n",
        "        x = funct.relu(self.fc2(x))\n",
        "        x = self.fc3(x)\n",
        "        return x"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XGa9d6ukU6NE"
      },
      "source": [
        "## TRAINING UTILS, LOSS & OPTIMISER"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EzR40c5BU-ra",
        "outputId": "a8ea0b13-622b-44e7-8f05-c22921c88128"
      },
      "source": [
        "# check if cuda available\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "device"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "device(type='cuda', index=0)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nkHUOf7NVcRf"
      },
      "source": [
        "step = 0\n",
        "loss_train = []\n",
        "loss_val = []\n",
        "\n",
        "# import the model\n",
        "net = NNmodel()\n",
        "net = net.to(device)\n",
        "net = net.train()\n",
        "\n",
        "min_loss = 100\n",
        "patience = 5\n",
        "training_loss_store = []\n",
        "validation_loss_store = []"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Cross Entropy Loss :\n",
        "\n",
        "<img src = 'https://i.stack.imgur.com/gNip2.pnghttps://i.stack.imgur.com/gNip2.png'>"
      ],
      "metadata": {
        "id": "re9ALBUaizdo"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4pxuwYA5VmkP"
      },
      "source": [
        "# cross entropy loss\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MxX86zQfVt5D"
      },
      "source": [
        "## TRAINING"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "T1mX52-QVv9p",
        "outputId": "c4d92910-1b39-41de-f9e8-a928bc1ab3a4"
      },
      "source": [
        "file = open('logs_test4_epoch100_with_max_pool.txt', 'w')\n",
        "\n",
        "print('Training Started.............................................')\n",
        "file.write('Training Started.............................................\\n')\n",
        "start_time = time.time()\n",
        "for epoch in range(150):  # loop over the dataset multiple times\n",
        "    #file.write('##############################TRAINING###############################\\n')\n",
        "    running_loss = 0.0\n",
        "    for i, data in enumerate(trainloader, 0):\n",
        "        step+=1\n",
        "        # get the inputs; data is a list of [inputs, labels]\n",
        "        inputs, labels = data[0].to(device),data[1].to(device)\n",
        "\n",
        "        # zero the parameter gradients\n",
        "        optimizer.zero_grad()\n",
        "        # forward + backward + optimize\n",
        "        outputs = net(inputs)\n",
        "        loss = criterion(outputs, labels)\n",
        "        loss_train.append(loss.item())\n",
        "        training_loss_store.append([epoch, loss.item()])\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "        # print statistics\n",
        "        running_loss += loss.item()\n",
        "        if i % 20 == 19:    # print every 10 mini-batches\n",
        "            print('[%d, %5d] loss: %.5f' %(epoch + 1, i + 1, running_loss / 2000))\n",
        "            file.write('epoch = '+ str(epoch + 1) + '\\t' +'step = '+ str(step) +'\\t'+'train_loss = '+'\\t'+str(np.mean(loss_train)) +'\\n')\n",
        "            loss_train = []\n",
        "            running_loss = 0.0\n",
        "            \n",
        "    print('Finished training for epoch ' + str(epoch) + ' time taken = ' + str(time.time() - start_time))\n",
        "    file.write('Finished training for epoch ' + str(epoch) + ' time taken = ' + str(time.time() - start_time) + '\\n')\n",
        "    file.write('##################################evaluation##############################\\n')\n",
        "    print('################################evaluation###########################\\n')\n",
        "    with torch.no_grad():\n",
        "        val_loss = 0\n",
        "        net.eval()\n",
        "        \n",
        "        for i, data in enumerate(testloader, 0):\n",
        "            step+=1\n",
        "            inputs, labels = data[0].to(device),data[1].to(device)\n",
        "            outputs = net(inputs)\n",
        "            loss = criterion(outputs, labels)\n",
        "            loss_val.append(loss.item())\n",
        "            validation_loss_store.append([epoch, loss.item()])\n",
        "            val_loss += loss\n",
        "            \n",
        "        val_loss = val_loss/float(i + 1)\n",
        "        \n",
        "        if val_loss < min_loss:\n",
        "            min_loss = val_loss\n",
        "            no_impr_epoch = 0\n",
        "            \n",
        "            #save the best model\n",
        "            print(str(val_loss))\n",
        "            torch.save(net.state_dict(), 'xyz_epoch_'+str(epoch+1)+'.pt')\n",
        "            \n",
        "            print('performance improved with validation loss ' + str(val_loss))\n",
        "            file.write('--------------------------------------------------------------------\\n')\n",
        "            file.write('performance improved with validation loss =  ' + str(val_loss) + '\\n')\n",
        "            \n",
        "            file.write('epoch = '+ str(epoch + 1) + '\\t' +'step = '+ str(step) +'\\t'+'val_loss = '+'\\t'+str(np.mean(loss_val)) +'\\n')\n",
        "            file.write('--------------------------------------------------------------------\\n\\n')\n",
        "            loss_val = []\n",
        "        else:\n",
        "            no_impr_epoch += 1\n",
        "            print('no improvement with prev best model ' + str(no_impr_epoch) + 'th')\n",
        "            file.write('no improvement with prev best model ' + str(no_impr_epoch) + 'th \\n')\n",
        "            \n",
        "        if no_impr_epoch > patience:\n",
        "            print('stop training')\n",
        "            file.write('stop training')\n",
        "            break\n",
        "    \n",
        "print('Finished Training................................................')\n",
        "file.write('Finished Training................................................\\n')\n",
        "end_time = time.time()\n",
        "file.write('Training time:- ' + str(end_time - start_time))\n",
        "file.close()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\n",
            "[4,  6240] loss: 0.00043\n",
            "[4,  6260] loss: 0.00022\n",
            "[4,  6280] loss: 0.00019\n",
            "[4,  6300] loss: 0.00030\n",
            "[4,  6320] loss: 0.00029\n",
            "[4,  6340] loss: 0.00059\n",
            "[4,  6360] loss: 0.00029\n",
            "[4,  6380] loss: 0.00075\n",
            "[4,  6400] loss: 0.00007\n",
            "[4,  6420] loss: 0.00028\n",
            "[4,  6440] loss: 0.00048\n",
            "[4,  6460] loss: 0.00017\n",
            "[4,  6480] loss: 0.00039\n",
            "[4,  6500] loss: 0.00029\n",
            "[4,  6520] loss: 0.00030\n",
            "[4,  6540] loss: 0.00070\n",
            "[4,  6560] loss: 0.00063\n",
            "[4,  6580] loss: 0.00009\n",
            "[4,  6600] loss: 0.00024\n",
            "[4,  6620] loss: 0.00071\n",
            "[4,  6640] loss: 0.00064\n",
            "[4,  6660] loss: 0.00034\n",
            "[4,  6680] loss: 0.00064\n",
            "[4,  6700] loss: 0.00066\n",
            "[4,  6720] loss: 0.00016\n",
            "[4,  6740] loss: 0.00055\n",
            "[4,  6760] loss: 0.00025\n",
            "[4,  6780] loss: 0.00071\n",
            "[4,  6800] loss: 0.00009\n",
            "[4,  6820] loss: 0.00010\n",
            "[4,  6840] loss: 0.00050\n",
            "[4,  6860] loss: 0.00024\n",
            "[4,  6880] loss: 0.00007\n",
            "[4,  6900] loss: 0.00020\n",
            "[4,  6920] loss: 0.00011\n",
            "[4,  6940] loss: 0.00013\n",
            "[4,  6960] loss: 0.00038\n",
            "[4,  6980] loss: 0.00036\n",
            "[4,  7000] loss: 0.00031\n",
            "[4,  7020] loss: 0.00082\n",
            "[4,  7040] loss: 0.00042\n",
            "[4,  7060] loss: 0.00030\n",
            "[4,  7080] loss: 0.00030\n",
            "[4,  7100] loss: 0.00043\n",
            "[4,  7120] loss: 0.00089\n",
            "[4,  7140] loss: 0.00012\n",
            "[4,  7160] loss: 0.00130\n",
            "[4,  7180] loss: 0.00041\n",
            "[4,  7200] loss: 0.00015\n",
            "[4,  7220] loss: 0.00004\n",
            "[4,  7240] loss: 0.00012\n",
            "[4,  7260] loss: 0.00012\n",
            "[4,  7280] loss: 0.00030\n",
            "[4,  7300] loss: 0.00043\n",
            "[4,  7320] loss: 0.00007\n",
            "[4,  7340] loss: 0.00051\n",
            "[4,  7360] loss: 0.00022\n",
            "[4,  7380] loss: 0.00035\n",
            "[4,  7400] loss: 0.00021\n",
            "[4,  7420] loss: 0.00046\n",
            "[4,  7440] loss: 0.00042\n",
            "[4,  7460] loss: 0.00008\n",
            "[4,  7480] loss: 0.00039\n",
            "[4,  7500] loss: 0.00045\n",
            "Finished training for epoch 3 time taken = 248.6124610900879\n",
            "################################evaluation###########################\n",
            "\n",
            "tensor(0.0361, device='cuda:0')\n",
            "performance improved with validation loss tensor(0.0361, device='cuda:0')\n",
            "[5,    20] loss: 0.00035\n",
            "[5,    40] loss: 0.00024\n",
            "[5,    60] loss: 0.00022\n",
            "[5,    80] loss: 0.00017\n",
            "[5,   100] loss: 0.00021\n",
            "[5,   120] loss: 0.00006\n",
            "[5,   140] loss: 0.00023\n",
            "[5,   160] loss: 0.00005\n",
            "[5,   180] loss: 0.00005\n",
            "[5,   200] loss: 0.00010\n",
            "[5,   220] loss: 0.00038\n",
            "[5,   240] loss: 0.00008\n",
            "[5,   260] loss: 0.00020\n",
            "[5,   280] loss: 0.00044\n",
            "[5,   300] loss: 0.00009\n",
            "[5,   320] loss: 0.00012\n",
            "[5,   340] loss: 0.00034\n",
            "[5,   360] loss: 0.00023\n",
            "[5,   380] loss: 0.00019\n",
            "[5,   400] loss: 0.00094\n",
            "[5,   420] loss: 0.00026\n",
            "[5,   440] loss: 0.00039\n",
            "[5,   460] loss: 0.00008\n",
            "[5,   480] loss: 0.00034\n",
            "[5,   500] loss: 0.00028\n",
            "[5,   520] loss: 0.00052\n",
            "[5,   540] loss: 0.00009\n",
            "[5,   560] loss: 0.00016\n",
            "[5,   580] loss: 0.00021\n",
            "[5,   600] loss: 0.00046\n",
            "[5,   620] loss: 0.00017\n",
            "[5,   640] loss: 0.00016\n",
            "[5,   660] loss: 0.00016\n",
            "[5,   680] loss: 0.00047\n",
            "[5,   700] loss: 0.00039\n",
            "[5,   720] loss: 0.00063\n",
            "[5,   740] loss: 0.00034\n",
            "[5,   760] loss: 0.00045\n",
            "[5,   780] loss: 0.00021\n",
            "[5,   800] loss: 0.00013\n",
            "[5,   820] loss: 0.00022\n",
            "[5,   840] loss: 0.00009\n",
            "[5,   860] loss: 0.00016\n",
            "[5,   880] loss: 0.00030\n",
            "[5,   900] loss: 0.00026\n",
            "[5,   920] loss: 0.00006\n",
            "[5,   940] loss: 0.00009\n",
            "[5,   960] loss: 0.00009\n",
            "[5,   980] loss: 0.00004\n",
            "[5,  1000] loss: 0.00021\n",
            "[5,  1020] loss: 0.00034\n",
            "[5,  1040] loss: 0.00082\n",
            "[5,  1060] loss: 0.00061\n",
            "[5,  1080] loss: 0.00012\n",
            "[5,  1100] loss: 0.00015\n",
            "[5,  1120] loss: 0.00013\n",
            "[5,  1140] loss: 0.00041\n",
            "[5,  1160] loss: 0.00015\n",
            "[5,  1180] loss: 0.00037\n",
            "[5,  1200] loss: 0.00076\n",
            "[5,  1220] loss: 0.00006\n",
            "[5,  1240] loss: 0.00048\n",
            "[5,  1260] loss: 0.00030\n",
            "[5,  1280] loss: 0.00051\n",
            "[5,  1300] loss: 0.00012\n",
            "[5,  1320] loss: 0.00014\n",
            "[5,  1340] loss: 0.00069\n",
            "[5,  1360] loss: 0.00040\n",
            "[5,  1380] loss: 0.00065\n",
            "[5,  1400] loss: 0.00024\n",
            "[5,  1420] loss: 0.00037\n",
            "[5,  1440] loss: 0.00030\n",
            "[5,  1460] loss: 0.00030\n",
            "[5,  1480] loss: 0.00026\n",
            "[5,  1500] loss: 0.00045\n",
            "[5,  1520] loss: 0.00031\n",
            "[5,  1540] loss: 0.00022\n",
            "[5,  1560] loss: 0.00033\n",
            "[5,  1580] loss: 0.00008\n",
            "[5,  1600] loss: 0.00047\n",
            "[5,  1620] loss: 0.00048\n",
            "[5,  1640] loss: 0.00034\n",
            "[5,  1660] loss: 0.00005\n",
            "[5,  1680] loss: 0.00024\n",
            "[5,  1700] loss: 0.00026\n",
            "[5,  1720] loss: 0.00020\n",
            "[5,  1740] loss: 0.00023\n",
            "[5,  1760] loss: 0.00015\n",
            "[5,  1780] loss: 0.00039\n",
            "[5,  1800] loss: 0.00021\n",
            "[5,  1820] loss: 0.00045\n",
            "[5,  1840] loss: 0.00005\n",
            "[5,  1860] loss: 0.00039\n",
            "[5,  1880] loss: 0.00027\n",
            "[5,  1900] loss: 0.00023\n",
            "[5,  1920] loss: 0.00022\n",
            "[5,  1940] loss: 0.00021\n",
            "[5,  1960] loss: 0.00033\n",
            "[5,  1980] loss: 0.00016\n",
            "[5,  2000] loss: 0.00014\n",
            "[5,  2020] loss: 0.00066\n",
            "[5,  2040] loss: 0.00004\n",
            "[5,  2060] loss: 0.00009\n",
            "[5,  2080] loss: 0.00015\n",
            "[5,  2100] loss: 0.00018\n",
            "[5,  2120] loss: 0.00017\n",
            "[5,  2140] loss: 0.00065\n",
            "[5,  2160] loss: 0.00016\n",
            "[5,  2180] loss: 0.00055\n",
            "[5,  2200] loss: 0.00075\n",
            "[5,  2220] loss: 0.00025\n",
            "[5,  2240] loss: 0.00053\n",
            "[5,  2260] loss: 0.00017\n",
            "[5,  2280] loss: 0.00018\n",
            "[5,  2300] loss: 0.00023\n",
            "[5,  2320] loss: 0.00062\n",
            "[5,  2340] loss: 0.00007\n",
            "[5,  2360] loss: 0.00040\n",
            "[5,  2380] loss: 0.00010\n",
            "[5,  2400] loss: 0.00031\n",
            "[5,  2420] loss: 0.00029\n",
            "[5,  2440] loss: 0.00037\n",
            "[5,  2460] loss: 0.00017\n",
            "[5,  2480] loss: 0.00006\n",
            "[5,  2500] loss: 0.00050\n",
            "[5,  2520] loss: 0.00055\n",
            "[5,  2540] loss: 0.00072\n",
            "[5,  2560] loss: 0.00046\n",
            "[5,  2580] loss: 0.00026\n",
            "[5,  2600] loss: 0.00037\n",
            "[5,  2620] loss: 0.00031\n",
            "[5,  2640] loss: 0.00045\n",
            "[5,  2660] loss: 0.00071\n",
            "[5,  2680] loss: 0.00029\n",
            "[5,  2700] loss: 0.00103\n",
            "[5,  2720] loss: 0.00042\n",
            "[5,  2740] loss: 0.00035\n",
            "[5,  2760] loss: 0.00014\n",
            "[5,  2780] loss: 0.00044\n",
            "[5,  2800] loss: 0.00066\n",
            "[5,  2820] loss: 0.00046\n",
            "[5,  2840] loss: 0.00039\n",
            "[5,  2860] loss: 0.00011\n",
            "[5,  2880] loss: 0.00009\n",
            "[5,  2900] loss: 0.00031\n",
            "[5,  2920] loss: 0.00022\n",
            "[5,  2940] loss: 0.00016\n",
            "[5,  2960] loss: 0.00005\n",
            "[5,  2980] loss: 0.00006\n",
            "[5,  3000] loss: 0.00044\n",
            "[5,  3020] loss: 0.00051\n",
            "[5,  3040] loss: 0.00034\n",
            "[5,  3060] loss: 0.00032\n",
            "[5,  3080] loss: 0.00042\n",
            "[5,  3100] loss: 0.00085\n",
            "[5,  3120] loss: 0.00016\n",
            "[5,  3140] loss: 0.00055\n",
            "[5,  3160] loss: 0.00024\n",
            "[5,  3180] loss: 0.00071\n",
            "[5,  3200] loss: 0.00055\n",
            "[5,  3220] loss: 0.00007\n",
            "[5,  3240] loss: 0.00014\n",
            "[5,  3260] loss: 0.00005\n",
            "[5,  3280] loss: 0.00052\n",
            "[5,  3300] loss: 0.00082\n",
            "[5,  3320] loss: 0.00037\n",
            "[5,  3340] loss: 0.00046\n",
            "[5,  3360] loss: 0.00010\n",
            "[5,  3380] loss: 0.00046\n",
            "[5,  3400] loss: 0.00056\n",
            "[5,  3420] loss: 0.00041\n",
            "[5,  3440] loss: 0.00021\n",
            "[5,  3460] loss: 0.00013\n",
            "[5,  3480] loss: 0.00034\n",
            "[5,  3500] loss: 0.00079\n",
            "[5,  3520] loss: 0.00045\n",
            "[5,  3540] loss: 0.00020\n",
            "[5,  3560] loss: 0.00026\n",
            "[5,  3580] loss: 0.00030\n",
            "[5,  3600] loss: 0.00060\n",
            "[5,  3620] loss: 0.00047\n",
            "[5,  3640] loss: 0.00020\n",
            "[5,  3660] loss: 0.00022\n",
            "[5,  3680] loss: 0.00009\n",
            "[5,  3700] loss: 0.00051\n",
            "[5,  3720] loss: 0.00041\n",
            "[5,  3740] loss: 0.00058\n",
            "[5,  3760] loss: 0.00025\n",
            "[5,  3780] loss: 0.00041\n",
            "[5,  3800] loss: 0.00020\n",
            "[5,  3820] loss: 0.00011\n",
            "[5,  3840] loss: 0.00031\n",
            "[5,  3860] loss: 0.00048\n",
            "[5,  3880] loss: 0.00025\n",
            "[5,  3900] loss: 0.00047\n",
            "[5,  3920] loss: 0.00046\n",
            "[5,  3940] loss: 0.00046\n",
            "[5,  3960] loss: 0.00014\n",
            "[5,  3980] loss: 0.00007\n",
            "[5,  4000] loss: 0.00006\n",
            "[5,  4020] loss: 0.00036\n",
            "[5,  4040] loss: 0.00031\n",
            "[5,  4060] loss: 0.00023\n",
            "[5,  4080] loss: 0.00034\n",
            "[5,  4100] loss: 0.00073\n",
            "[5,  4120] loss: 0.00020\n",
            "[5,  4140] loss: 0.00021\n",
            "[5,  4160] loss: 0.00025\n",
            "[5,  4180] loss: 0.00042\n",
            "[5,  4200] loss: 0.00013\n",
            "[5,  4220] loss: 0.00026\n",
            "[5,  4240] loss: 0.00022\n",
            "[5,  4260] loss: 0.00005\n",
            "[5,  4280] loss: 0.00044\n",
            "[5,  4300] loss: 0.00004\n",
            "[5,  4320] loss: 0.00047\n",
            "[5,  4340] loss: 0.00043\n",
            "[5,  4360] loss: 0.00028\n",
            "[5,  4380] loss: 0.00098\n",
            "[5,  4400] loss: 0.00033\n",
            "[5,  4420] loss: 0.00037\n",
            "[5,  4440] loss: 0.00018\n",
            "[5,  4460] loss: 0.00005\n",
            "[5,  4480] loss: 0.00036\n",
            "[5,  4500] loss: 0.00030\n",
            "[5,  4520] loss: 0.00068\n",
            "[5,  4540] loss: 0.00005\n",
            "[5,  4560] loss: 0.00022\n",
            "[5,  4580] loss: 0.00072\n",
            "[5,  4600] loss: 0.00018\n",
            "[5,  4620] loss: 0.00013\n",
            "[5,  4640] loss: 0.00013\n",
            "[5,  4660] loss: 0.00018\n",
            "[5,  4680] loss: 0.00121\n",
            "[5,  4700] loss: 0.00006\n",
            "[5,  4720] loss: 0.00018\n",
            "[5,  4740] loss: 0.00055\n",
            "[5,  4760] loss: 0.00012\n",
            "[5,  4780] loss: 0.00015\n",
            "[5,  4800] loss: 0.00081\n",
            "[5,  4820] loss: 0.00013\n",
            "[5,  4840] loss: 0.00026\n",
            "[5,  4860] loss: 0.00020\n",
            "[5,  4880] loss: 0.00038\n",
            "[5,  4900] loss: 0.00034\n",
            "[5,  4920] loss: 0.00024\n",
            "[5,  4940] loss: 0.00006\n",
            "[5,  4960] loss: 0.00043\n",
            "[5,  4980] loss: 0.00009\n",
            "[5,  5000] loss: 0.00036\n",
            "[5,  5020] loss: 0.00014\n",
            "[5,  5040] loss: 0.00050\n",
            "[5,  5060] loss: 0.00023\n",
            "[5,  5080] loss: 0.00011\n",
            "[5,  5100] loss: 0.00021\n",
            "[5,  5120] loss: 0.00004\n",
            "[5,  5140] loss: 0.00038\n",
            "[5,  5160] loss: 0.00005\n",
            "[5,  5180] loss: 0.00006\n",
            "[5,  5200] loss: 0.00035\n",
            "[5,  5220] loss: 0.00030\n",
            "[5,  5240] loss: 0.00053\n",
            "[5,  5260] loss: 0.00048\n",
            "[5,  5280] loss: 0.00016\n",
            "[5,  5300] loss: 0.00057\n",
            "[5,  5320] loss: 0.00024\n",
            "[5,  5340] loss: 0.00017\n",
            "[5,  5360] loss: 0.00008\n",
            "[5,  5380] loss: 0.00024\n",
            "[5,  5400] loss: 0.00083\n",
            "[5,  5420] loss: 0.00009\n",
            "[5,  5440] loss: 0.00006\n",
            "[5,  5460] loss: 0.00041\n",
            "[5,  5480] loss: 0.00021\n",
            "[5,  5500] loss: 0.00114\n",
            "[5,  5520] loss: 0.00073\n",
            "[5,  5540] loss: 0.00047\n",
            "[5,  5560] loss: 0.00010\n",
            "[5,  5580] loss: 0.00024\n",
            "[5,  5600] loss: 0.00031\n",
            "[5,  5620] loss: 0.00037\n",
            "[5,  5640] loss: 0.00016\n",
            "[5,  5660] loss: 0.00055\n",
            "[5,  5680] loss: 0.00024\n",
            "[5,  5700] loss: 0.00073\n",
            "[5,  5720] loss: 0.00032\n",
            "[5,  5740] loss: 0.00035\n",
            "[5,  5760] loss: 0.00068\n",
            "[5,  5780] loss: 0.00033\n",
            "[5,  5800] loss: 0.00049\n",
            "[5,  5820] loss: 0.00060\n",
            "[5,  5840] loss: 0.00040\n",
            "[5,  5860] loss: 0.00004\n",
            "[5,  5880] loss: 0.00056\n",
            "[5,  5900] loss: 0.00014\n",
            "[5,  5920] loss: 0.00026\n",
            "[5,  5940] loss: 0.00011\n",
            "[5,  5960] loss: 0.00012\n",
            "[5,  5980] loss: 0.00033\n",
            "[5,  6000] loss: 0.00061\n",
            "[5,  6020] loss: 0.00063\n",
            "[5,  6040] loss: 0.00013\n",
            "[5,  6060] loss: 0.00050\n",
            "[5,  6080] loss: 0.00109\n",
            "[5,  6100] loss: 0.00050\n",
            "[5,  6120] loss: 0.00026\n",
            "[5,  6140] loss: 0.00023\n",
            "[5,  6160] loss: 0.00041\n",
            "[5,  6180] loss: 0.00016\n",
            "[5,  6200] loss: 0.00053\n",
            "[5,  6220] loss: 0.00019\n",
            "[5,  6240] loss: 0.00039\n",
            "[5,  6260] loss: 0.00028\n",
            "[5,  6280] loss: 0.00027\n",
            "[5,  6300] loss: 0.00052\n",
            "[5,  6320] loss: 0.00034\n",
            "[5,  6340] loss: 0.00084\n",
            "[5,  6360] loss: 0.00030\n",
            "[5,  6380] loss: 0.00054\n",
            "[5,  6400] loss: 0.00052\n",
            "[5,  6420] loss: 0.00015\n",
            "[5,  6440] loss: 0.00018\n",
            "[5,  6460] loss: 0.00010\n",
            "[5,  6480] loss: 0.00063\n",
            "[5,  6500] loss: 0.00012\n",
            "[5,  6520] loss: 0.00032\n",
            "[5,  6540] loss: 0.00018\n",
            "[5,  6560] loss: 0.00015\n",
            "[5,  6580] loss: 0.00013\n",
            "[5,  6600] loss: 0.00021\n",
            "[5,  6620] loss: 0.00016\n",
            "[5,  6640] loss: 0.00026\n",
            "[5,  6660] loss: 0.00070\n",
            "[5,  6680] loss: 0.00089\n",
            "[5,  6700] loss: 0.00018\n",
            "[5,  6720] loss: 0.00014\n",
            "[5,  6740] loss: 0.00011\n",
            "[5,  6760] loss: 0.00024\n",
            "[5,  6780] loss: 0.00021\n",
            "[5,  6800] loss: 0.00023\n",
            "[5,  6820] loss: 0.00042\n",
            "[5,  6840] loss: 0.00033\n",
            "[5,  6860] loss: 0.00008\n",
            "[5,  6880] loss: 0.00052\n",
            "[5,  6900] loss: 0.00038\n",
            "[5,  6920] loss: 0.00090\n",
            "[5,  6940] loss: 0.00005\n",
            "[5,  6960] loss: 0.00007\n",
            "[5,  6980] loss: 0.00036\n",
            "[5,  7000] loss: 0.00062\n",
            "[5,  7020] loss: 0.00022\n",
            "[5,  7040] loss: 0.00016\n",
            "[5,  7060] loss: 0.00033\n",
            "[5,  7080] loss: 0.00021\n",
            "[5,  7100] loss: 0.00010\n",
            "[5,  7120] loss: 0.00008\n",
            "[5,  7140] loss: 0.00021\n",
            "[5,  7160] loss: 0.00020\n",
            "[5,  7180] loss: 0.00045\n",
            "[5,  7200] loss: 0.00016\n",
            "[5,  7220] loss: 0.00049\n",
            "[5,  7240] loss: 0.00035\n",
            "[5,  7260] loss: 0.00051\n",
            "[5,  7280] loss: 0.00018\n",
            "[5,  7300] loss: 0.00005\n",
            "[5,  7320] loss: 0.00014\n",
            "[5,  7340] loss: 0.00061\n",
            "[5,  7360] loss: 0.00042\n",
            "[5,  7380] loss: 0.00068\n",
            "[5,  7400] loss: 0.00061\n",
            "[5,  7420] loss: 0.00038\n",
            "[5,  7440] loss: 0.00008\n",
            "[5,  7460] loss: 0.00029\n",
            "[5,  7480] loss: 0.00011\n",
            "[5,  7500] loss: 0.00011\n",
            "Finished training for epoch 4 time taken = 312.1593873500824\n",
            "################################evaluation###########################\n",
            "\n",
            "tensor(0.0325, device='cuda:0')\n",
            "performance improved with validation loss tensor(0.0325, device='cuda:0')\n",
            "[6,    20] loss: 0.00030\n",
            "[6,    40] loss: 0.00033\n",
            "[6,    60] loss: 0.00062\n",
            "[6,    80] loss: 0.00018\n",
            "[6,   100] loss: 0.00009\n",
            "[6,   120] loss: 0.00007\n",
            "[6,   140] loss: 0.00047\n",
            "[6,   160] loss: 0.00017\n",
            "[6,   180] loss: 0.00006\n",
            "[6,   200] loss: 0.00049\n",
            "[6,   220] loss: 0.00004\n",
            "[6,   240] loss: 0.00010\n",
            "[6,   260] loss: 0.00014\n",
            "[6,   280] loss: 0.00049\n",
            "[6,   300] loss: 0.00034\n",
            "[6,   320] loss: 0.00053\n",
            "[6,   340] loss: 0.00022\n",
            "[6,   360] loss: 0.00033\n",
            "[6,   380] loss: 0.00010\n",
            "[6,   400] loss: 0.00004\n",
            "[6,   420] loss: 0.00021\n",
            "[6,   440] loss: 0.00012\n",
            "[6,   460] loss: 0.00040\n",
            "[6,   480] loss: 0.00043\n",
            "[6,   500] loss: 0.00005\n",
            "[6,   520] loss: 0.00032\n",
            "[6,   540] loss: 0.00065\n",
            "[6,   560] loss: 0.00012\n",
            "[6,   580] loss: 0.00002\n",
            "[6,   600] loss: 0.00046\n",
            "[6,   620] loss: 0.00008\n",
            "[6,   640] loss: 0.00012\n",
            "[6,   660] loss: 0.00007\n",
            "[6,   680] loss: 0.00032\n",
            "[6,   700] loss: 0.00012\n",
            "[6,   720] loss: 0.00010\n",
            "[6,   740] loss: 0.00005\n",
            "[6,   760] loss: 0.00007\n",
            "[6,   780] loss: 0.00012\n",
            "[6,   800] loss: 0.00007\n",
            "[6,   820] loss: 0.00028\n",
            "[6,   840] loss: 0.00043\n",
            "[6,   860] loss: 0.00012\n",
            "[6,   880] loss: 0.00035\n",
            "[6,   900] loss: 0.00026\n",
            "[6,   920] loss: 0.00015\n",
            "[6,   940] loss: 0.00016\n",
            "[6,   960] loss: 0.00007\n",
            "[6,   980] loss: 0.00039\n",
            "[6,  1000] loss: 0.00014\n",
            "[6,  1020] loss: 0.00017\n",
            "[6,  1040] loss: 0.00013\n",
            "[6,  1060] loss: 0.00014\n",
            "[6,  1080] loss: 0.00020\n",
            "[6,  1100] loss: 0.00015\n",
            "[6,  1120] loss: 0.00063\n",
            "[6,  1140] loss: 0.00016\n",
            "[6,  1160] loss: 0.00040\n",
            "[6,  1180] loss: 0.00016\n",
            "[6,  1200] loss: 0.00023\n",
            "[6,  1220] loss: 0.00082\n",
            "[6,  1240] loss: 0.00026\n",
            "[6,  1260] loss: 0.00031\n",
            "[6,  1280] loss: 0.00023\n",
            "[6,  1300] loss: 0.00044\n",
            "[6,  1320] loss: 0.00003\n",
            "[6,  1340] loss: 0.00002\n",
            "[6,  1360] loss: 0.00023\n",
            "[6,  1380] loss: 0.00051\n",
            "[6,  1400] loss: 0.00028\n",
            "[6,  1420] loss: 0.00044\n",
            "[6,  1440] loss: 0.00078\n",
            "[6,  1460] loss: 0.00034\n",
            "[6,  1480] loss: 0.00077\n",
            "[6,  1500] loss: 0.00012\n",
            "[6,  1520] loss: 0.00039\n",
            "[6,  1540] loss: 0.00026\n",
            "[6,  1560] loss: 0.00077\n",
            "[6,  1580] loss: 0.00061\n",
            "[6,  1600] loss: 0.00008\n",
            "[6,  1620] loss: 0.00011\n",
            "[6,  1640] loss: 0.00049\n",
            "[6,  1660] loss: 0.00012\n",
            "[6,  1680] loss: 0.00027\n",
            "[6,  1700] loss: 0.00012\n",
            "[6,  1720] loss: 0.00024\n",
            "[6,  1740] loss: 0.00003\n",
            "[6,  1760] loss: 0.00014\n",
            "[6,  1780] loss: 0.00004\n",
            "[6,  1800] loss: 0.00021\n",
            "[6,  1820] loss: 0.00021\n",
            "[6,  1840] loss: 0.00011\n",
            "[6,  1860] loss: 0.00020\n",
            "[6,  1880] loss: 0.00018\n",
            "[6,  1900] loss: 0.00006\n",
            "[6,  1920] loss: 0.00028\n",
            "[6,  1940] loss: 0.00006\n",
            "[6,  1960] loss: 0.00005\n",
            "[6,  1980] loss: 0.00032\n",
            "[6,  2000] loss: 0.00015\n",
            "[6,  2020] loss: 0.00030\n",
            "[6,  2040] loss: 0.00071\n",
            "[6,  2060] loss: 0.00016\n",
            "[6,  2080] loss: 0.00026\n",
            "[6,  2100] loss: 0.00036\n",
            "[6,  2120] loss: 0.00033\n",
            "[6,  2140] loss: 0.00015\n",
            "[6,  2160] loss: 0.00009\n",
            "[6,  2180] loss: 0.00007\n",
            "[6,  2200] loss: 0.00018\n",
            "[6,  2220] loss: 0.00044\n",
            "[6,  2240] loss: 0.00035\n",
            "[6,  2260] loss: 0.00006\n",
            "[6,  2280] loss: 0.00030\n",
            "[6,  2300] loss: 0.00016\n",
            "[6,  2320] loss: 0.00048\n",
            "[6,  2340] loss: 0.00006\n",
            "[6,  2360] loss: 0.00026\n",
            "[6,  2380] loss: 0.00020\n",
            "[6,  2400] loss: 0.00030\n",
            "[6,  2420] loss: 0.00012\n",
            "[6,  2440] loss: 0.00009\n",
            "[6,  2460] loss: 0.00018\n",
            "[6,  2480] loss: 0.00052\n",
            "[6,  2500] loss: 0.00040\n",
            "[6,  2520] loss: 0.00012\n",
            "[6,  2540] loss: 0.00058\n",
            "[6,  2560] loss: 0.00019\n",
            "[6,  2580] loss: 0.00035\n",
            "[6,  2600] loss: 0.00036\n",
            "[6,  2620] loss: 0.00009\n",
            "[6,  2640] loss: 0.00026\n",
            "[6,  2660] loss: 0.00026\n",
            "[6,  2680] loss: 0.00002\n",
            "[6,  2700] loss: 0.00026\n",
            "[6,  2720] loss: 0.00043\n",
            "[6,  2740] loss: 0.00008\n",
            "[6,  2760] loss: 0.00003\n",
            "[6,  2780] loss: 0.00017\n",
            "[6,  2800] loss: 0.00015\n",
            "[6,  2820] loss: 0.00026\n",
            "[6,  2840] loss: 0.00006\n",
            "[6,  2860] loss: 0.00018\n",
            "[6,  2880] loss: 0.00074\n",
            "[6,  2900] loss: 0.00043\n",
            "[6,  2920] loss: 0.00016\n",
            "[6,  2940] loss: 0.00014\n",
            "[6,  2960] loss: 0.00020\n",
            "[6,  2980] loss: 0.00056\n",
            "[6,  3000] loss: 0.00021\n",
            "[6,  3020] loss: 0.00014\n",
            "[6,  3040] loss: 0.00019\n",
            "[6,  3060] loss: 0.00005\n",
            "[6,  3080] loss: 0.00010\n",
            "[6,  3100] loss: 0.00024\n",
            "[6,  3120] loss: 0.00004\n",
            "[6,  3140] loss: 0.00010\n",
            "[6,  3160] loss: 0.00046\n",
            "[6,  3180] loss: 0.00052\n",
            "[6,  3200] loss: 0.00085\n",
            "[6,  3220] loss: 0.00025\n",
            "[6,  3240] loss: 0.00016\n",
            "[6,  3260] loss: 0.00043\n",
            "[6,  3280] loss: 0.00055\n",
            "[6,  3300] loss: 0.00021\n",
            "[6,  3320] loss: 0.00046\n",
            "[6,  3340] loss: 0.00049\n",
            "[6,  3360] loss: 0.00015\n",
            "[6,  3380] loss: 0.00011\n",
            "[6,  3400] loss: 0.00057\n",
            "[6,  3420] loss: 0.00020\n",
            "[6,  3440] loss: 0.00052\n",
            "[6,  3460] loss: 0.00010\n",
            "[6,  3480] loss: 0.00020\n",
            "[6,  3500] loss: 0.00032\n",
            "[6,  3520] loss: 0.00025\n",
            "[6,  3540] loss: 0.00021\n",
            "[6,  3560] loss: 0.00007\n",
            "[6,  3580] loss: 0.00060\n",
            "[6,  3600] loss: 0.00049\n",
            "[6,  3620] loss: 0.00020\n",
            "[6,  3640] loss: 0.00010\n",
            "[6,  3660] loss: 0.00033\n",
            "[6,  3680] loss: 0.00018\n",
            "[6,  3700] loss: 0.00036\n",
            "[6,  3720] loss: 0.00027\n",
            "[6,  3740] loss: 0.00011\n",
            "[6,  3760] loss: 0.00031\n",
            "[6,  3780] loss: 0.00099\n",
            "[6,  3800] loss: 0.00074\n",
            "[6,  3820] loss: 0.00037\n",
            "[6,  3840] loss: 0.00013\n",
            "[6,  3860] loss: 0.00013\n",
            "[6,  3880] loss: 0.00022\n",
            "[6,  3900] loss: 0.00004\n",
            "[6,  3920] loss: 0.00137\n",
            "[6,  3940] loss: 0.00052\n",
            "[6,  3960] loss: 0.00014\n",
            "[6,  3980] loss: 0.00031\n",
            "[6,  4000] loss: 0.00020\n",
            "[6,  4020] loss: 0.00015\n",
            "[6,  4040] loss: 0.00016\n",
            "[6,  4060] loss: 0.00030\n",
            "[6,  4080] loss: 0.00050\n",
            "[6,  4100] loss: 0.00027\n",
            "[6,  4120] loss: 0.00032\n",
            "[6,  4140] loss: 0.00006\n",
            "[6,  4160] loss: 0.00009\n",
            "[6,  4180] loss: 0.00019\n",
            "[6,  4200] loss: 0.00015\n",
            "[6,  4220] loss: 0.00005\n",
            "[6,  4240] loss: 0.00069\n",
            "[6,  4260] loss: 0.00032\n",
            "[6,  4280] loss: 0.00061\n",
            "[6,  4300] loss: 0.00021\n",
            "[6,  4320] loss: 0.00009\n",
            "[6,  4340] loss: 0.00031\n",
            "[6,  4360] loss: 0.00006\n",
            "[6,  4380] loss: 0.00044\n",
            "[6,  4400] loss: 0.00038\n",
            "[6,  4420] loss: 0.00022\n",
            "[6,  4440] loss: 0.00039\n",
            "[6,  4460] loss: 0.00016\n",
            "[6,  4480] loss: 0.00056\n",
            "[6,  4500] loss: 0.00024\n",
            "[6,  4520] loss: 0.00017\n",
            "[6,  4540] loss: 0.00038\n",
            "[6,  4560] loss: 0.00064\n",
            "[6,  4580] loss: 0.00023\n",
            "[6,  4600] loss: 0.00014\n",
            "[6,  4620] loss: 0.00034\n",
            "[6,  4640] loss: 0.00009\n",
            "[6,  4660] loss: 0.00011\n",
            "[6,  4680] loss: 0.00066\n",
            "[6,  4700] loss: 0.00007\n",
            "[6,  4720] loss: 0.00028\n",
            "[6,  4740] loss: 0.00043\n",
            "[6,  4760] loss: 0.00029\n",
            "[6,  4780] loss: 0.00010\n",
            "[6,  4800] loss: 0.00024\n",
            "[6,  4820] loss: 0.00001\n",
            "[6,  4840] loss: 0.00025\n",
            "[6,  4860] loss: 0.00021\n",
            "[6,  4880] loss: 0.00054\n",
            "[6,  4900] loss: 0.00022\n",
            "[6,  4920] loss: 0.00010\n",
            "[6,  4940] loss: 0.00021\n",
            "[6,  4960] loss: 0.00039\n",
            "[6,  4980] loss: 0.00015\n",
            "[6,  5000] loss: 0.00007\n",
            "[6,  5020] loss: 0.00038\n",
            "[6,  5040] loss: 0.00042\n",
            "[6,  5060] loss: 0.00016\n",
            "[6,  5080] loss: 0.00040\n",
            "[6,  5100] loss: 0.00025\n",
            "[6,  5120] loss: 0.00006\n",
            "[6,  5140] loss: 0.00053\n",
            "[6,  5160] loss: 0.00054\n",
            "[6,  5180] loss: 0.00038\n",
            "[6,  5200] loss: 0.00031\n",
            "[6,  5220] loss: 0.00048\n",
            "[6,  5240] loss: 0.00013\n",
            "[6,  5260] loss: 0.00010\n",
            "[6,  5280] loss: 0.00046\n",
            "[6,  5300] loss: 0.00034\n",
            "[6,  5320] loss: 0.00030\n",
            "[6,  5340] loss: 0.00005\n",
            "[6,  5360] loss: 0.00007\n",
            "[6,  5380] loss: 0.00005\n",
            "[6,  5400] loss: 0.00020\n",
            "[6,  5420] loss: 0.00028\n",
            "[6,  5440] loss: 0.00015\n",
            "[6,  5460] loss: 0.00010\n",
            "[6,  5480] loss: 0.00014\n",
            "[6,  5500] loss: 0.00049\n",
            "[6,  5520] loss: 0.00050\n",
            "[6,  5540] loss: 0.00005\n",
            "[6,  5560] loss: 0.00003\n",
            "[6,  5580] loss: 0.00043\n",
            "[6,  5600] loss: 0.00062\n",
            "[6,  5620] loss: 0.00010\n",
            "[6,  5640] loss: 0.00003\n",
            "[6,  5660] loss: 0.00019\n",
            "[6,  5680] loss: 0.00044\n",
            "[6,  5700] loss: 0.00025\n",
            "[6,  5720] loss: 0.00016\n",
            "[6,  5740] loss: 0.00019\n",
            "[6,  5760] loss: 0.00020\n",
            "[6,  5780] loss: 0.00014\n",
            "[6,  5800] loss: 0.00123\n",
            "[6,  5820] loss: 0.00017\n",
            "[6,  5840] loss: 0.00008\n",
            "[6,  5860] loss: 0.00028\n",
            "[6,  5880] loss: 0.00129\n",
            "[6,  5900] loss: 0.00073\n",
            "[6,  5920] loss: 0.00055\n",
            "[6,  5940] loss: 0.00046\n",
            "[6,  5960] loss: 0.00025\n",
            "[6,  5980] loss: 0.00022\n",
            "[6,  6000] loss: 0.00032\n",
            "[6,  6020] loss: 0.00009\n",
            "[6,  6040] loss: 0.00044\n",
            "[6,  6060] loss: 0.00050\n",
            "[6,  6080] loss: 0.00006\n",
            "[6,  6100] loss: 0.00014\n",
            "[6,  6120] loss: 0.00007\n",
            "[6,  6140] loss: 0.00057\n",
            "[6,  6160] loss: 0.00030\n",
            "[6,  6180] loss: 0.00064\n",
            "[6,  6200] loss: 0.00007\n",
            "[6,  6220] loss: 0.00080\n",
            "[6,  6240] loss: 0.00021\n",
            "[6,  6260] loss: 0.00049\n",
            "[6,  6280] loss: 0.00015\n",
            "[6,  6300] loss: 0.00017\n",
            "[6,  6320] loss: 0.00031\n",
            "[6,  6340] loss: 0.00006\n",
            "[6,  6360] loss: 0.00032\n",
            "[6,  6380] loss: 0.00009\n",
            "[6,  6400] loss: 0.00023\n",
            "[6,  6420] loss: 0.00011\n",
            "[6,  6440] loss: 0.00051\n",
            "[6,  6460] loss: 0.00021\n",
            "[6,  6480] loss: 0.00029\n",
            "[6,  6500] loss: 0.00008\n",
            "[6,  6520] loss: 0.00010\n",
            "[6,  6540] loss: 0.00020\n",
            "[6,  6560] loss: 0.00013\n",
            "[6,  6580] loss: 0.00045\n",
            "[6,  6600] loss: 0.00016\n",
            "[6,  6620] loss: 0.00011\n",
            "[6,  6640] loss: 0.00017\n",
            "[6,  6660] loss: 0.00018\n",
            "[6,  6680] loss: 0.00016\n",
            "[6,  6700] loss: 0.00017\n",
            "[6,  6720] loss: 0.00027\n",
            "[6,  6740] loss: 0.00029\n",
            "[6,  6760] loss: 0.00016\n",
            "[6,  6780] loss: 0.00027\n",
            "[6,  6800] loss: 0.00004\n",
            "[6,  6820] loss: 0.00003\n",
            "[6,  6840] loss: 0.00016\n",
            "[6,  6860] loss: 0.00005\n",
            "[6,  6880] loss: 0.00052\n",
            "[6,  6900] loss: 0.00052\n",
            "[6,  6920] loss: 0.00017\n",
            "[6,  6940] loss: 0.00012\n",
            "[6,  6960] loss: 0.00022\n",
            "[6,  6980] loss: 0.00062\n",
            "[6,  7000] loss: 0.00009\n",
            "[6,  7020] loss: 0.00019\n",
            "[6,  7040] loss: 0.00004\n",
            "[6,  7060] loss: 0.00013\n",
            "[6,  7080] loss: 0.00036\n",
            "[6,  7100] loss: 0.00065\n",
            "[6,  7120] loss: 0.00042\n",
            "[6,  7140] loss: 0.00053\n",
            "[6,  7160] loss: 0.00006\n",
            "[6,  7180] loss: 0.00016\n",
            "[6,  7200] loss: 0.00021\n",
            "[6,  7220] loss: 0.00014\n",
            "[6,  7240] loss: 0.00001\n",
            "[6,  7260] loss: 0.00013\n",
            "[6,  7280] loss: 0.00055\n",
            "[6,  7300] loss: 0.00036\n",
            "[6,  7320] loss: 0.00039\n",
            "[6,  7340] loss: 0.00023\n",
            "[6,  7360] loss: 0.00057\n",
            "[6,  7380] loss: 0.00059\n",
            "[6,  7400] loss: 0.00023\n",
            "[6,  7420] loss: 0.00019\n",
            "[6,  7440] loss: 0.00077\n",
            "[6,  7460] loss: 0.00015\n",
            "[6,  7480] loss: 0.00081\n",
            "[6,  7500] loss: 0.00013\n",
            "Finished training for epoch 5 time taken = 375.80576062202454\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 1th\n",
            "[7,    20] loss: 0.00052\n",
            "[7,    40] loss: 0.00009\n",
            "[7,    60] loss: 0.00025\n",
            "[7,    80] loss: 0.00065\n",
            "[7,   100] loss: 0.00006\n",
            "[7,   120] loss: 0.00050\n",
            "[7,   140] loss: 0.00015\n",
            "[7,   160] loss: 0.00015\n",
            "[7,   180] loss: 0.00004\n",
            "[7,   200] loss: 0.00009\n",
            "[7,   220] loss: 0.00007\n",
            "[7,   240] loss: 0.00040\n",
            "[7,   260] loss: 0.00008\n",
            "[7,   280] loss: 0.00024\n",
            "[7,   300] loss: 0.00014\n",
            "[7,   320] loss: 0.00016\n",
            "[7,   340] loss: 0.00008\n",
            "[7,   360] loss: 0.00001\n",
            "[7,   380] loss: 0.00035\n",
            "[7,   400] loss: 0.00015\n",
            "[7,   420] loss: 0.00017\n",
            "[7,   440] loss: 0.00009\n",
            "[7,   460] loss: 0.00017\n",
            "[7,   480] loss: 0.00016\n",
            "[7,   500] loss: 0.00007\n",
            "[7,   520] loss: 0.00006\n",
            "[7,   540] loss: 0.00022\n",
            "[7,   560] loss: 0.00033\n",
            "[7,   580] loss: 0.00019\n",
            "[7,   600] loss: 0.00020\n",
            "[7,   620] loss: 0.00047\n",
            "[7,   640] loss: 0.00009\n",
            "[7,   660] loss: 0.00018\n",
            "[7,   680] loss: 0.00007\n",
            "[7,   700] loss: 0.00054\n",
            "[7,   720] loss: 0.00036\n",
            "[7,   740] loss: 0.00007\n",
            "[7,   760] loss: 0.00023\n",
            "[7,   780] loss: 0.00004\n",
            "[7,   800] loss: 0.00023\n",
            "[7,   820] loss: 0.00030\n",
            "[7,   840] loss: 0.00048\n",
            "[7,   860] loss: 0.00057\n",
            "[7,   880] loss: 0.00008\n",
            "[7,   900] loss: 0.00061\n",
            "[7,   920] loss: 0.00053\n",
            "[7,   940] loss: 0.00030\n",
            "[7,   960] loss: 0.00009\n",
            "[7,   980] loss: 0.00008\n",
            "[7,  1000] loss: 0.00049\n",
            "[7,  1020] loss: 0.00025\n",
            "[7,  1040] loss: 0.00009\n",
            "[7,  1060] loss: 0.00023\n",
            "[7,  1080] loss: 0.00013\n",
            "[7,  1100] loss: 0.00010\n",
            "[7,  1120] loss: 0.00005\n",
            "[7,  1140] loss: 0.00012\n",
            "[7,  1160] loss: 0.00039\n",
            "[7,  1180] loss: 0.00035\n",
            "[7,  1200] loss: 0.00003\n",
            "[7,  1220] loss: 0.00020\n",
            "[7,  1240] loss: 0.00026\n",
            "[7,  1260] loss: 0.00013\n",
            "[7,  1280] loss: 0.00014\n",
            "[7,  1300] loss: 0.00009\n",
            "[7,  1320] loss: 0.00016\n",
            "[7,  1340] loss: 0.00025\n",
            "[7,  1360] loss: 0.00023\n",
            "[7,  1380] loss: 0.00017\n",
            "[7,  1400] loss: 0.00062\n",
            "[7,  1420] loss: 0.00012\n",
            "[7,  1440] loss: 0.00013\n",
            "[7,  1460] loss: 0.00086\n",
            "[7,  1480] loss: 0.00006\n",
            "[7,  1500] loss: 0.00020\n",
            "[7,  1520] loss: 0.00005\n",
            "[7,  1540] loss: 0.00019\n",
            "[7,  1560] loss: 0.00020\n",
            "[7,  1580] loss: 0.00044\n",
            "[7,  1600] loss: 0.00074\n",
            "[7,  1620] loss: 0.00045\n",
            "[7,  1640] loss: 0.00020\n",
            "[7,  1660] loss: 0.00003\n",
            "[7,  1680] loss: 0.00003\n",
            "[7,  1700] loss: 0.00061\n",
            "[7,  1720] loss: 0.00009\n",
            "[7,  1740] loss: 0.00008\n",
            "[7,  1760] loss: 0.00003\n",
            "[7,  1780] loss: 0.00015\n",
            "[7,  1800] loss: 0.00018\n",
            "[7,  1820] loss: 0.00005\n",
            "[7,  1840] loss: 0.00019\n",
            "[7,  1860] loss: 0.00020\n",
            "[7,  1880] loss: 0.00013\n",
            "[7,  1900] loss: 0.00006\n",
            "[7,  1920] loss: 0.00008\n",
            "[7,  1940] loss: 0.00004\n",
            "[7,  1960] loss: 0.00037\n",
            "[7,  1980] loss: 0.00013\n",
            "[7,  2000] loss: 0.00025\n",
            "[7,  2020] loss: 0.00004\n",
            "[7,  2040] loss: 0.00073\n",
            "[7,  2060] loss: 0.00011\n",
            "[7,  2080] loss: 0.00016\n",
            "[7,  2100] loss: 0.00031\n",
            "[7,  2120] loss: 0.00018\n",
            "[7,  2140] loss: 0.00014\n",
            "[7,  2160] loss: 0.00044\n",
            "[7,  2180] loss: 0.00032\n",
            "[7,  2200] loss: 0.00032\n",
            "[7,  2220] loss: 0.00009\n",
            "[7,  2240] loss: 0.00062\n",
            "[7,  2260] loss: 0.00024\n",
            "[7,  2280] loss: 0.00003\n",
            "[7,  2300] loss: 0.00010\n",
            "[7,  2320] loss: 0.00046\n",
            "[7,  2340] loss: 0.00017\n",
            "[7,  2360] loss: 0.00053\n",
            "[7,  2380] loss: 0.00002\n",
            "[7,  2400] loss: 0.00017\n",
            "[7,  2420] loss: 0.00019\n",
            "[7,  2440] loss: 0.00030\n",
            "[7,  2460] loss: 0.00008\n",
            "[7,  2480] loss: 0.00051\n",
            "[7,  2500] loss: 0.00034\n",
            "[7,  2520] loss: 0.00031\n",
            "[7,  2540] loss: 0.00013\n",
            "[7,  2560] loss: 0.00044\n",
            "[7,  2580] loss: 0.00012\n",
            "[7,  2600] loss: 0.00021\n",
            "[7,  2620] loss: 0.00010\n",
            "[7,  2640] loss: 0.00009\n",
            "[7,  2660] loss: 0.00010\n",
            "[7,  2680] loss: 0.00053\n",
            "[7,  2700] loss: 0.00022\n",
            "[7,  2720] loss: 0.00013\n",
            "[7,  2740] loss: 0.00007\n",
            "[7,  2760] loss: 0.00008\n",
            "[7,  2780] loss: 0.00008\n",
            "[7,  2800] loss: 0.00014\n",
            "[7,  2820] loss: 0.00010\n",
            "[7,  2840] loss: 0.00018\n",
            "[7,  2860] loss: 0.00038\n",
            "[7,  2880] loss: 0.00029\n",
            "[7,  2900] loss: 0.00002\n",
            "[7,  2920] loss: 0.00056\n",
            "[7,  2940] loss: 0.00008\n",
            "[7,  2960] loss: 0.00006\n",
            "[7,  2980] loss: 0.00026\n",
            "[7,  3000] loss: 0.00017\n",
            "[7,  3020] loss: 0.00010\n",
            "[7,  3040] loss: 0.00008\n",
            "[7,  3060] loss: 0.00030\n",
            "[7,  3080] loss: 0.00013\n",
            "[7,  3100] loss: 0.00018\n",
            "[7,  3120] loss: 0.00014\n",
            "[7,  3140] loss: 0.00014\n",
            "[7,  3160] loss: 0.00034\n",
            "[7,  3180] loss: 0.00019\n",
            "[7,  3200] loss: 0.00022\n",
            "[7,  3220] loss: 0.00022\n",
            "[7,  3240] loss: 0.00014\n",
            "[7,  3260] loss: 0.00035\n",
            "[7,  3280] loss: 0.00079\n",
            "[7,  3300] loss: 0.00004\n",
            "[7,  3320] loss: 0.00007\n",
            "[7,  3340] loss: 0.00014\n",
            "[7,  3360] loss: 0.00010\n",
            "[7,  3380] loss: 0.00067\n",
            "[7,  3400] loss: 0.00065\n",
            "[7,  3420] loss: 0.00051\n",
            "[7,  3440] loss: 0.00010\n",
            "[7,  3460] loss: 0.00089\n",
            "[7,  3480] loss: 0.00056\n",
            "[7,  3500] loss: 0.00040\n",
            "[7,  3520] loss: 0.00020\n",
            "[7,  3540] loss: 0.00023\n",
            "[7,  3560] loss: 0.00019\n",
            "[7,  3580] loss: 0.00018\n",
            "[7,  3600] loss: 0.00020\n",
            "[7,  3620] loss: 0.00005\n",
            "[7,  3640] loss: 0.00011\n",
            "[7,  3660] loss: 0.00002\n",
            "[7,  3680] loss: 0.00032\n",
            "[7,  3700] loss: 0.00007\n",
            "[7,  3720] loss: 0.00035\n",
            "[7,  3740] loss: 0.00013\n",
            "[7,  3760] loss: 0.00005\n",
            "[7,  3780] loss: 0.00016\n",
            "[7,  3800] loss: 0.00049\n",
            "[7,  3820] loss: 0.00013\n",
            "[7,  3840] loss: 0.00060\n",
            "[7,  3860] loss: 0.00040\n",
            "[7,  3880] loss: 0.00009\n",
            "[7,  3900] loss: 0.00074\n",
            "[7,  3920] loss: 0.00039\n",
            "[7,  3940] loss: 0.00008\n",
            "[7,  3960] loss: 0.00038\n",
            "[7,  3980] loss: 0.00003\n",
            "[7,  4000] loss: 0.00006\n",
            "[7,  4020] loss: 0.00022\n",
            "[7,  4040] loss: 0.00006\n",
            "[7,  4060] loss: 0.00059\n",
            "[7,  4080] loss: 0.00004\n",
            "[7,  4100] loss: 0.00002\n",
            "[7,  4120] loss: 0.00006\n",
            "[7,  4140] loss: 0.00004\n",
            "[7,  4160] loss: 0.00012\n",
            "[7,  4180] loss: 0.00057\n",
            "[7,  4200] loss: 0.00016\n",
            "[7,  4220] loss: 0.00009\n",
            "[7,  4240] loss: 0.00014\n",
            "[7,  4260] loss: 0.00026\n",
            "[7,  4280] loss: 0.00009\n",
            "[7,  4300] loss: 0.00013\n",
            "[7,  4320] loss: 0.00020\n",
            "[7,  4340] loss: 0.00012\n",
            "[7,  4360] loss: 0.00030\n",
            "[7,  4380] loss: 0.00016\n",
            "[7,  4400] loss: 0.00011\n",
            "[7,  4420] loss: 0.00017\n",
            "[7,  4440] loss: 0.00037\n",
            "[7,  4460] loss: 0.00018\n",
            "[7,  4480] loss: 0.00045\n",
            "[7,  4500] loss: 0.00020\n",
            "[7,  4520] loss: 0.00012\n",
            "[7,  4540] loss: 0.00023\n",
            "[7,  4560] loss: 0.00014\n",
            "[7,  4580] loss: 0.00037\n",
            "[7,  4600] loss: 0.00028\n",
            "[7,  4620] loss: 0.00159\n",
            "[7,  4640] loss: 0.00021\n",
            "[7,  4660] loss: 0.00032\n",
            "[7,  4680] loss: 0.00011\n",
            "[7,  4700] loss: 0.00040\n",
            "[7,  4720] loss: 0.00027\n",
            "[7,  4740] loss: 0.00015\n",
            "[7,  4760] loss: 0.00027\n",
            "[7,  4780] loss: 0.00029\n",
            "[7,  4800] loss: 0.00023\n",
            "[7,  4820] loss: 0.00025\n",
            "[7,  4840] loss: 0.00005\n",
            "[7,  4860] loss: 0.00011\n",
            "[7,  4880] loss: 0.00013\n",
            "[7,  4900] loss: 0.00062\n",
            "[7,  4920] loss: 0.00013\n",
            "[7,  4940] loss: 0.00007\n",
            "[7,  4960] loss: 0.00028\n",
            "[7,  4980] loss: 0.00010\n",
            "[7,  5000] loss: 0.00036\n",
            "[7,  5020] loss: 0.00021\n",
            "[7,  5040] loss: 0.00012\n",
            "[7,  5060] loss: 0.00041\n",
            "[7,  5080] loss: 0.00004\n",
            "[7,  5100] loss: 0.00009\n",
            "[7,  5120] loss: 0.00051\n",
            "[7,  5140] loss: 0.00009\n",
            "[7,  5160] loss: 0.00031\n",
            "[7,  5180] loss: 0.00015\n",
            "[7,  5200] loss: 0.00019\n",
            "[7,  5220] loss: 0.00004\n",
            "[7,  5240] loss: 0.00004\n",
            "[7,  5260] loss: 0.00017\n",
            "[7,  5280] loss: 0.00027\n",
            "[7,  5300] loss: 0.00036\n",
            "[7,  5320] loss: 0.00015\n",
            "[7,  5340] loss: 0.00009\n",
            "[7,  5360] loss: 0.00009\n",
            "[7,  5380] loss: 0.00015\n",
            "[7,  5400] loss: 0.00014\n",
            "[7,  5420] loss: 0.00011\n",
            "[7,  5440] loss: 0.00007\n",
            "[7,  5460] loss: 0.00017\n",
            "[7,  5480] loss: 0.00003\n",
            "[7,  5500] loss: 0.00061\n",
            "[7,  5520] loss: 0.00012\n",
            "[7,  5540] loss: 0.00029\n",
            "[7,  5560] loss: 0.00063\n",
            "[7,  5580] loss: 0.00054\n",
            "[7,  5600] loss: 0.00006\n",
            "[7,  5620] loss: 0.00095\n",
            "[7,  5640] loss: 0.00014\n",
            "[7,  5660] loss: 0.00015\n",
            "[7,  5680] loss: 0.00037\n",
            "[7,  5700] loss: 0.00005\n",
            "[7,  5720] loss: 0.00065\n",
            "[7,  5740] loss: 0.00054\n",
            "[7,  5760] loss: 0.00008\n",
            "[7,  5780] loss: 0.00025\n",
            "[7,  5800] loss: 0.00052\n",
            "[7,  5820] loss: 0.00010\n",
            "[7,  5840] loss: 0.00017\n",
            "[7,  5860] loss: 0.00022\n",
            "[7,  5880] loss: 0.00006\n",
            "[7,  5900] loss: 0.00004\n",
            "[7,  5920] loss: 0.00007\n",
            "[7,  5940] loss: 0.00021\n",
            "[7,  5960] loss: 0.00017\n",
            "[7,  5980] loss: 0.00023\n",
            "[7,  6000] loss: 0.00036\n",
            "[7,  6020] loss: 0.00014\n",
            "[7,  6040] loss: 0.00034\n",
            "[7,  6060] loss: 0.00014\n",
            "[7,  6080] loss: 0.00017\n",
            "[7,  6100] loss: 0.00014\n",
            "[7,  6120] loss: 0.00014\n",
            "[7,  6140] loss: 0.00022\n",
            "[7,  6160] loss: 0.00029\n",
            "[7,  6180] loss: 0.00029\n",
            "[7,  6200] loss: 0.00039\n",
            "[7,  6220] loss: 0.00009\n",
            "[7,  6240] loss: 0.00014\n",
            "[7,  6260] loss: 0.00012\n",
            "[7,  6280] loss: 0.00084\n",
            "[7,  6300] loss: 0.00013\n",
            "[7,  6320] loss: 0.00032\n",
            "[7,  6340] loss: 0.00008\n",
            "[7,  6360] loss: 0.00009\n",
            "[7,  6380] loss: 0.00016\n",
            "[7,  6400] loss: 0.00018\n",
            "[7,  6420] loss: 0.00025\n",
            "[7,  6440] loss: 0.00013\n",
            "[7,  6460] loss: 0.00054\n",
            "[7,  6480] loss: 0.00020\n",
            "[7,  6500] loss: 0.00051\n",
            "[7,  6520] loss: 0.00007\n",
            "[7,  6540] loss: 0.00040\n",
            "[7,  6560] loss: 0.00013\n",
            "[7,  6580] loss: 0.00028\n",
            "[7,  6600] loss: 0.00012\n",
            "[7,  6620] loss: 0.00015\n",
            "[7,  6640] loss: 0.00042\n",
            "[7,  6660] loss: 0.00021\n",
            "[7,  6680] loss: 0.00016\n",
            "[7,  6700] loss: 0.00014\n",
            "[7,  6720] loss: 0.00015\n",
            "[7,  6740] loss: 0.00030\n",
            "[7,  6760] loss: 0.00048\n",
            "[7,  6780] loss: 0.00012\n",
            "[7,  6800] loss: 0.00006\n",
            "[7,  6820] loss: 0.00036\n",
            "[7,  6840] loss: 0.00006\n",
            "[7,  6860] loss: 0.00014\n",
            "[7,  6880] loss: 0.00020\n",
            "[7,  6900] loss: 0.00002\n",
            "[7,  6920] loss: 0.00048\n",
            "[7,  6940] loss: 0.00041\n",
            "[7,  6960] loss: 0.00060\n",
            "[7,  6980] loss: 0.00015\n",
            "[7,  7000] loss: 0.00055\n",
            "[7,  7020] loss: 0.00012\n",
            "[7,  7040] loss: 0.00016\n",
            "[7,  7060] loss: 0.00015\n",
            "[7,  7080] loss: 0.00002\n",
            "[7,  7100] loss: 0.00088\n",
            "[7,  7120] loss: 0.00030\n",
            "[7,  7140] loss: 0.00003\n",
            "[7,  7160] loss: 0.00020\n",
            "[7,  7180] loss: 0.00019\n",
            "[7,  7200] loss: 0.00008\n",
            "[7,  7220] loss: 0.00041\n",
            "[7,  7240] loss: 0.00024\n",
            "[7,  7260] loss: 0.00003\n",
            "[7,  7280] loss: 0.00023\n",
            "[7,  7300] loss: 0.00011\n",
            "[7,  7320] loss: 0.00003\n",
            "[7,  7340] loss: 0.00006\n",
            "[7,  7360] loss: 0.00026\n",
            "[7,  7380] loss: 0.00006\n",
            "[7,  7400] loss: 0.00039\n",
            "[7,  7420] loss: 0.00047\n",
            "[7,  7440] loss: 0.00021\n",
            "[7,  7460] loss: 0.00021\n",
            "[7,  7480] loss: 0.00034\n",
            "[7,  7500] loss: 0.00011\n",
            "Finished training for epoch 6 time taken = 439.6992988586426\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 2th\n",
            "[8,    20] loss: 0.00010\n",
            "[8,    40] loss: 0.00045\n",
            "[8,    60] loss: 0.00052\n",
            "[8,    80] loss: 0.00019\n",
            "[8,   100] loss: 0.00003\n",
            "[8,   120] loss: 0.00009\n",
            "[8,   140] loss: 0.00008\n",
            "[8,   160] loss: 0.00007\n",
            "[8,   180] loss: 0.00017\n",
            "[8,   200] loss: 0.00002\n",
            "[8,   220] loss: 0.00014\n",
            "[8,   240] loss: 0.00003\n",
            "[8,   260] loss: 0.00013\n",
            "[8,   280] loss: 0.00005\n",
            "[8,   300] loss: 0.00015\n",
            "[8,   320] loss: 0.00004\n",
            "[8,   340] loss: 0.00003\n",
            "[8,   360] loss: 0.00021\n",
            "[8,   380] loss: 0.00024\n",
            "[8,   400] loss: 0.00006\n",
            "[8,   420] loss: 0.00002\n",
            "[8,   440] loss: 0.00004\n",
            "[8,   460] loss: 0.00024\n",
            "[8,   480] loss: 0.00003\n",
            "[8,   500] loss: 0.00004\n",
            "[8,   520] loss: 0.00016\n",
            "[8,   540] loss: 0.00001\n",
            "[8,   560] loss: 0.00009\n",
            "[8,   580] loss: 0.00013\n",
            "[8,   600] loss: 0.00002\n",
            "[8,   620] loss: 0.00015\n",
            "[8,   640] loss: 0.00031\n",
            "[8,   660] loss: 0.00030\n",
            "[8,   680] loss: 0.00015\n",
            "[8,   700] loss: 0.00002\n",
            "[8,   720] loss: 0.00026\n",
            "[8,   740] loss: 0.00005\n",
            "[8,   760] loss: 0.00035\n",
            "[8,   780] loss: 0.00008\n",
            "[8,   800] loss: 0.00003\n",
            "[8,   820] loss: 0.00033\n",
            "[8,   840] loss: 0.00034\n",
            "[8,   860] loss: 0.00010\n",
            "[8,   880] loss: 0.00010\n",
            "[8,   900] loss: 0.00013\n",
            "[8,   920] loss: 0.00003\n",
            "[8,   940] loss: 0.00026\n",
            "[8,   960] loss: 0.00026\n",
            "[8,   980] loss: 0.00019\n",
            "[8,  1000] loss: 0.00040\n",
            "[8,  1020] loss: 0.00020\n",
            "[8,  1040] loss: 0.00022\n",
            "[8,  1060] loss: 0.00028\n",
            "[8,  1080] loss: 0.00012\n",
            "[8,  1100] loss: 0.00004\n",
            "[8,  1120] loss: 0.00006\n",
            "[8,  1140] loss: 0.00039\n",
            "[8,  1160] loss: 0.00018\n",
            "[8,  1180] loss: 0.00056\n",
            "[8,  1200] loss: 0.00036\n",
            "[8,  1220] loss: 0.00010\n",
            "[8,  1240] loss: 0.00003\n",
            "[8,  1260] loss: 0.00012\n",
            "[8,  1280] loss: 0.00001\n",
            "[8,  1300] loss: 0.00017\n",
            "[8,  1320] loss: 0.00011\n",
            "[8,  1340] loss: 0.00026\n",
            "[8,  1360] loss: 0.00024\n",
            "[8,  1380] loss: 0.00027\n",
            "[8,  1400] loss: 0.00018\n",
            "[8,  1420] loss: 0.00012\n",
            "[8,  1440] loss: 0.00015\n",
            "[8,  1460] loss: 0.00002\n",
            "[8,  1480] loss: 0.00022\n",
            "[8,  1500] loss: 0.00013\n",
            "[8,  1520] loss: 0.00005\n",
            "[8,  1540] loss: 0.00014\n",
            "[8,  1560] loss: 0.00013\n",
            "[8,  1580] loss: 0.00014\n",
            "[8,  1600] loss: 0.00039\n",
            "[8,  1620] loss: 0.00024\n",
            "[8,  1640] loss: 0.00031\n",
            "[8,  1660] loss: 0.00009\n",
            "[8,  1680] loss: 0.00003\n",
            "[8,  1700] loss: 0.00017\n",
            "[8,  1720] loss: 0.00020\n",
            "[8,  1740] loss: 0.00054\n",
            "[8,  1760] loss: 0.00059\n",
            "[8,  1780] loss: 0.00043\n",
            "[8,  1800] loss: 0.00014\n",
            "[8,  1820] loss: 0.00017\n",
            "[8,  1840] loss: 0.00023\n",
            "[8,  1860] loss: 0.00019\n",
            "[8,  1880] loss: 0.00015\n",
            "[8,  1900] loss: 0.00002\n",
            "[8,  1920] loss: 0.00002\n",
            "[8,  1940] loss: 0.00007\n",
            "[8,  1960] loss: 0.00016\n",
            "[8,  1980] loss: 0.00011\n",
            "[8,  2000] loss: 0.00010\n",
            "[8,  2020] loss: 0.00014\n",
            "[8,  2040] loss: 0.00006\n",
            "[8,  2060] loss: 0.00004\n",
            "[8,  2080] loss: 0.00003\n",
            "[8,  2100] loss: 0.00022\n",
            "[8,  2120] loss: 0.00015\n",
            "[8,  2140] loss: 0.00007\n",
            "[8,  2160] loss: 0.00002\n",
            "[8,  2180] loss: 0.00006\n",
            "[8,  2200] loss: 0.00023\n",
            "[8,  2220] loss: 0.00022\n",
            "[8,  2240] loss: 0.00013\n",
            "[8,  2260] loss: 0.00040\n",
            "[8,  2280] loss: 0.00009\n",
            "[8,  2300] loss: 0.00091\n",
            "[8,  2320] loss: 0.00003\n",
            "[8,  2340] loss: 0.00006\n",
            "[8,  2360] loss: 0.00009\n",
            "[8,  2380] loss: 0.00047\n",
            "[8,  2400] loss: 0.00047\n",
            "[8,  2420] loss: 0.00003\n",
            "[8,  2440] loss: 0.00025\n",
            "[8,  2460] loss: 0.00022\n",
            "[8,  2480] loss: 0.00053\n",
            "[8,  2500] loss: 0.00013\n",
            "[8,  2520] loss: 0.00008\n",
            "[8,  2540] loss: 0.00025\n",
            "[8,  2560] loss: 0.00015\n",
            "[8,  2580] loss: 0.00005\n",
            "[8,  2600] loss: 0.00007\n",
            "[8,  2620] loss: 0.00036\n",
            "[8,  2640] loss: 0.00015\n",
            "[8,  2660] loss: 0.00012\n",
            "[8,  2680] loss: 0.00005\n",
            "[8,  2700] loss: 0.00021\n",
            "[8,  2720] loss: 0.00006\n",
            "[8,  2740] loss: 0.00001\n",
            "[8,  2760] loss: 0.00017\n",
            "[8,  2780] loss: 0.00044\n",
            "[8,  2800] loss: 0.00073\n",
            "[8,  2820] loss: 0.00008\n",
            "[8,  2840] loss: 0.00008\n",
            "[8,  2860] loss: 0.00014\n",
            "[8,  2880] loss: 0.00003\n",
            "[8,  2900] loss: 0.00013\n",
            "[8,  2920] loss: 0.00001\n",
            "[8,  2940] loss: 0.00004\n",
            "[8,  2960] loss: 0.00006\n",
            "[8,  2980] loss: 0.00003\n",
            "[8,  3000] loss: 0.00006\n",
            "[8,  3020] loss: 0.00002\n",
            "[8,  3040] loss: 0.00005\n",
            "[8,  3060] loss: 0.00027\n",
            "[8,  3080] loss: 0.00006\n",
            "[8,  3100] loss: 0.00009\n",
            "[8,  3120] loss: 0.00018\n",
            "[8,  3140] loss: 0.00020\n",
            "[8,  3160] loss: 0.00013\n",
            "[8,  3180] loss: 0.00066\n",
            "[8,  3200] loss: 0.00061\n",
            "[8,  3220] loss: 0.00006\n",
            "[8,  3240] loss: 0.00019\n",
            "[8,  3260] loss: 0.00014\n",
            "[8,  3280] loss: 0.00065\n",
            "[8,  3300] loss: 0.00019\n",
            "[8,  3320] loss: 0.00050\n",
            "[8,  3340] loss: 0.00046\n",
            "[8,  3360] loss: 0.00006\n",
            "[8,  3380] loss: 0.00002\n",
            "[8,  3400] loss: 0.00006\n",
            "[8,  3420] loss: 0.00028\n",
            "[8,  3440] loss: 0.00001\n",
            "[8,  3460] loss: 0.00011\n",
            "[8,  3480] loss: 0.00014\n",
            "[8,  3500] loss: 0.00022\n",
            "[8,  3520] loss: 0.00009\n",
            "[8,  3540] loss: 0.00009\n",
            "[8,  3560] loss: 0.00045\n",
            "[8,  3580] loss: 0.00022\n",
            "[8,  3600] loss: 0.00061\n",
            "[8,  3620] loss: 0.00015\n",
            "[8,  3640] loss: 0.00011\n",
            "[8,  3660] loss: 0.00017\n",
            "[8,  3680] loss: 0.00054\n",
            "[8,  3700] loss: 0.00005\n",
            "[8,  3720] loss: 0.00004\n",
            "[8,  3740] loss: 0.00007\n",
            "[8,  3760] loss: 0.00050\n",
            "[8,  3780] loss: 0.00010\n",
            "[8,  3800] loss: 0.00015\n",
            "[8,  3820] loss: 0.00032\n",
            "[8,  3840] loss: 0.00005\n",
            "[8,  3860] loss: 0.00010\n",
            "[8,  3880] loss: 0.00033\n",
            "[8,  3900] loss: 0.00016\n",
            "[8,  3920] loss: 0.00008\n",
            "[8,  3940] loss: 0.00020\n",
            "[8,  3960] loss: 0.00051\n",
            "[8,  3980] loss: 0.00121\n",
            "[8,  4000] loss: 0.00016\n",
            "[8,  4020] loss: 0.00015\n",
            "[8,  4040] loss: 0.00028\n",
            "[8,  4060] loss: 0.00007\n",
            "[8,  4080] loss: 0.00016\n",
            "[8,  4100] loss: 0.00019\n",
            "[8,  4120] loss: 0.00015\n",
            "[8,  4140] loss: 0.00008\n",
            "[8,  4160] loss: 0.00006\n",
            "[8,  4180] loss: 0.00002\n",
            "[8,  4200] loss: 0.00108\n",
            "[8,  4220] loss: 0.00018\n",
            "[8,  4240] loss: 0.00005\n",
            "[8,  4260] loss: 0.00003\n",
            "[8,  4280] loss: 0.00036\n",
            "[8,  4300] loss: 0.00034\n",
            "[8,  4320] loss: 0.00002\n",
            "[8,  4340] loss: 0.00020\n",
            "[8,  4360] loss: 0.00007\n",
            "[8,  4380] loss: 0.00033\n",
            "[8,  4400] loss: 0.00022\n",
            "[8,  4420] loss: 0.00039\n",
            "[8,  4440] loss: 0.00038\n",
            "[8,  4460] loss: 0.00023\n",
            "[8,  4480] loss: 0.00007\n",
            "[8,  4500] loss: 0.00018\n",
            "[8,  4520] loss: 0.00044\n",
            "[8,  4540] loss: 0.00016\n",
            "[8,  4560] loss: 0.00052\n",
            "[8,  4580] loss: 0.00023\n",
            "[8,  4600] loss: 0.00009\n",
            "[8,  4620] loss: 0.00007\n",
            "[8,  4640] loss: 0.00029\n",
            "[8,  4660] loss: 0.00011\n",
            "[8,  4680] loss: 0.00034\n",
            "[8,  4700] loss: 0.00040\n",
            "[8,  4720] loss: 0.00008\n",
            "[8,  4740] loss: 0.00046\n",
            "[8,  4760] loss: 0.00032\n",
            "[8,  4780] loss: 0.00009\n",
            "[8,  4800] loss: 0.00004\n",
            "[8,  4820] loss: 0.00026\n",
            "[8,  4840] loss: 0.00006\n",
            "[8,  4860] loss: 0.00011\n",
            "[8,  4880] loss: 0.00014\n",
            "[8,  4900] loss: 0.00019\n",
            "[8,  4920] loss: 0.00028\n",
            "[8,  4940] loss: 0.00004\n",
            "[8,  4960] loss: 0.00029\n",
            "[8,  4980] loss: 0.00005\n",
            "[8,  5000] loss: 0.00022\n",
            "[8,  5020] loss: 0.00019\n",
            "[8,  5040] loss: 0.00013\n",
            "[8,  5060] loss: 0.00037\n",
            "[8,  5080] loss: 0.00043\n",
            "[8,  5100] loss: 0.00025\n",
            "[8,  5120] loss: 0.00017\n",
            "[8,  5140] loss: 0.00036\n",
            "[8,  5160] loss: 0.00012\n",
            "[8,  5180] loss: 0.00010\n",
            "[8,  5200] loss: 0.00013\n",
            "[8,  5220] loss: 0.00016\n",
            "[8,  5240] loss: 0.00024\n",
            "[8,  5260] loss: 0.00018\n",
            "[8,  5280] loss: 0.00047\n",
            "[8,  5300] loss: 0.00021\n",
            "[8,  5320] loss: 0.00046\n",
            "[8,  5340] loss: 0.00001\n",
            "[8,  5360] loss: 0.00036\n",
            "[8,  5380] loss: 0.00005\n",
            "[8,  5400] loss: 0.00030\n",
            "[8,  5420] loss: 0.00002\n",
            "[8,  5440] loss: 0.00003\n",
            "[8,  5460] loss: 0.00003\n",
            "[8,  5480] loss: 0.00020\n",
            "[8,  5500] loss: 0.00012\n",
            "[8,  5520] loss: 0.00013\n",
            "[8,  5540] loss: 0.00087\n",
            "[8,  5560] loss: 0.00020\n",
            "[8,  5580] loss: 0.00040\n",
            "[8,  5600] loss: 0.00042\n",
            "[8,  5620] loss: 0.00059\n",
            "[8,  5640] loss: 0.00021\n",
            "[8,  5660] loss: 0.00011\n",
            "[8,  5680] loss: 0.00003\n",
            "[8,  5700] loss: 0.00033\n",
            "[8,  5720] loss: 0.00017\n",
            "[8,  5740] loss: 0.00018\n",
            "[8,  5760] loss: 0.00010\n",
            "[8,  5780] loss: 0.00002\n",
            "[8,  5800] loss: 0.00014\n",
            "[8,  5820] loss: 0.00018\n",
            "[8,  5840] loss: 0.00035\n",
            "[8,  5860] loss: 0.00016\n",
            "[8,  5880] loss: 0.00014\n",
            "[8,  5900] loss: 0.00011\n",
            "[8,  5920] loss: 0.00045\n",
            "[8,  5940] loss: 0.00014\n",
            "[8,  5960] loss: 0.00014\n",
            "[8,  5980] loss: 0.00009\n",
            "[8,  6000] loss: 0.00019\n",
            "[8,  6020] loss: 0.00018\n",
            "[8,  6040] loss: 0.00018\n",
            "[8,  6060] loss: 0.00018\n",
            "[8,  6080] loss: 0.00034\n",
            "[8,  6100] loss: 0.00020\n",
            "[8,  6120] loss: 0.00008\n",
            "[8,  6140] loss: 0.00010\n",
            "[8,  6160] loss: 0.00026\n",
            "[8,  6180] loss: 0.00011\n",
            "[8,  6200] loss: 0.00028\n",
            "[8,  6220] loss: 0.00021\n",
            "[8,  6240] loss: 0.00021\n",
            "[8,  6260] loss: 0.00015\n",
            "[8,  6280] loss: 0.00037\n",
            "[8,  6300] loss: 0.00028\n",
            "[8,  6320] loss: 0.00039\n",
            "[8,  6340] loss: 0.00028\n",
            "[8,  6360] loss: 0.00007\n",
            "[8,  6380] loss: 0.00036\n",
            "[8,  6400] loss: 0.00016\n",
            "[8,  6420] loss: 0.00021\n",
            "[8,  6440] loss: 0.00005\n",
            "[8,  6460] loss: 0.00030\n",
            "[8,  6480] loss: 0.00025\n",
            "[8,  6500] loss: 0.00005\n",
            "[8,  6520] loss: 0.00013\n",
            "[8,  6540] loss: 0.00003\n",
            "[8,  6560] loss: 0.00013\n",
            "[8,  6580] loss: 0.00003\n",
            "[8,  6600] loss: 0.00010\n",
            "[8,  6620] loss: 0.00012\n",
            "[8,  6640] loss: 0.00013\n",
            "[8,  6660] loss: 0.00032\n",
            "[8,  6680] loss: 0.00008\n",
            "[8,  6700] loss: 0.00013\n",
            "[8,  6720] loss: 0.00027\n",
            "[8,  6740] loss: 0.00016\n",
            "[8,  6760] loss: 0.00016\n",
            "[8,  6780] loss: 0.00055\n",
            "[8,  6800] loss: 0.00015\n",
            "[8,  6820] loss: 0.00003\n",
            "[8,  6840] loss: 0.00013\n",
            "[8,  6860] loss: 0.00002\n",
            "[8,  6880] loss: 0.00004\n",
            "[8,  6900] loss: 0.00005\n",
            "[8,  6920] loss: 0.00003\n",
            "[8,  6940] loss: 0.00008\n",
            "[8,  6960] loss: 0.00053\n",
            "[8,  6980] loss: 0.00021\n",
            "[8,  7000] loss: 0.00006\n",
            "[8,  7020] loss: 0.00050\n",
            "[8,  7040] loss: 0.00010\n",
            "[8,  7060] loss: 0.00010\n",
            "[8,  7080] loss: 0.00009\n",
            "[8,  7100] loss: 0.00005\n",
            "[8,  7120] loss: 0.00007\n",
            "[8,  7140] loss: 0.00029\n",
            "[8,  7160] loss: 0.00004\n",
            "[8,  7180] loss: 0.00004\n",
            "[8,  7200] loss: 0.00069\n",
            "[8,  7220] loss: 0.00017\n",
            "[8,  7240] loss: 0.00013\n",
            "[8,  7260] loss: 0.00004\n",
            "[8,  7280] loss: 0.00009\n",
            "[8,  7300] loss: 0.00004\n",
            "[8,  7320] loss: 0.00009\n",
            "[8,  7340] loss: 0.00038\n",
            "[8,  7360] loss: 0.00013\n",
            "[8,  7380] loss: 0.00008\n",
            "[8,  7400] loss: 0.00091\n",
            "[8,  7420] loss: 0.00040\n",
            "[8,  7440] loss: 0.00014\n",
            "[8,  7460] loss: 0.00016\n",
            "[8,  7480] loss: 0.00049\n",
            "[8,  7500] loss: 0.00017\n",
            "Finished training for epoch 7 time taken = 503.083860874176\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 3th\n",
            "[9,    20] loss: 0.00004\n",
            "[9,    40] loss: 0.00003\n",
            "[9,    60] loss: 0.00033\n",
            "[9,    80] loss: 0.00015\n",
            "[9,   100] loss: 0.00024\n",
            "[9,   120] loss: 0.00019\n",
            "[9,   140] loss: 0.00003\n",
            "[9,   160] loss: 0.00029\n",
            "[9,   180] loss: 0.00004\n",
            "[9,   200] loss: 0.00017\n",
            "[9,   220] loss: 0.00004\n",
            "[9,   240] loss: 0.00039\n",
            "[9,   260] loss: 0.00004\n",
            "[9,   280] loss: 0.00011\n",
            "[9,   300] loss: 0.00024\n",
            "[9,   320] loss: 0.00020\n",
            "[9,   340] loss: 0.00025\n",
            "[9,   360] loss: 0.00017\n",
            "[9,   380] loss: 0.00009\n",
            "[9,   400] loss: 0.00017\n",
            "[9,   420] loss: 0.00013\n",
            "[9,   440] loss: 0.00032\n",
            "[9,   460] loss: 0.00003\n",
            "[9,   480] loss: 0.00033\n",
            "[9,   500] loss: 0.00010\n",
            "[9,   520] loss: 0.00021\n",
            "[9,   540] loss: 0.00024\n",
            "[9,   560] loss: 0.00009\n",
            "[9,   580] loss: 0.00013\n",
            "[9,   600] loss: 0.00019\n",
            "[9,   620] loss: 0.00012\n",
            "[9,   640] loss: 0.00004\n",
            "[9,   660] loss: 0.00017\n",
            "[9,   680] loss: 0.00019\n",
            "[9,   700] loss: 0.00004\n",
            "[9,   720] loss: 0.00043\n",
            "[9,   740] loss: 0.00006\n",
            "[9,   760] loss: 0.00031\n",
            "[9,   780] loss: 0.00014\n",
            "[9,   800] loss: 0.00024\n",
            "[9,   820] loss: 0.00013\n",
            "[9,   840] loss: 0.00004\n",
            "[9,   860] loss: 0.00020\n",
            "[9,   880] loss: 0.00010\n",
            "[9,   900] loss: 0.00005\n",
            "[9,   920] loss: 0.00009\n",
            "[9,   940] loss: 0.00042\n",
            "[9,   960] loss: 0.00023\n",
            "[9,   980] loss: 0.00008\n",
            "[9,  1000] loss: 0.00008\n",
            "[9,  1020] loss: 0.00017\n",
            "[9,  1040] loss: 0.00017\n",
            "[9,  1060] loss: 0.00006\n",
            "[9,  1080] loss: 0.00020\n",
            "[9,  1100] loss: 0.00013\n",
            "[9,  1120] loss: 0.00012\n",
            "[9,  1140] loss: 0.00038\n",
            "[9,  1160] loss: 0.00003\n",
            "[9,  1180] loss: 0.00006\n",
            "[9,  1200] loss: 0.00009\n",
            "[9,  1220] loss: 0.00030\n",
            "[9,  1240] loss: 0.00027\n",
            "[9,  1260] loss: 0.00010\n",
            "[9,  1280] loss: 0.00006\n",
            "[9,  1300] loss: 0.00027\n",
            "[9,  1320] loss: 0.00008\n",
            "[9,  1340] loss: 0.00016\n",
            "[9,  1360] loss: 0.00033\n",
            "[9,  1380] loss: 0.00014\n",
            "[9,  1400] loss: 0.00006\n",
            "[9,  1420] loss: 0.00036\n",
            "[9,  1440] loss: 0.00026\n",
            "[9,  1460] loss: 0.00064\n",
            "[9,  1480] loss: 0.00007\n",
            "[9,  1500] loss: 0.00020\n",
            "[9,  1520] loss: 0.00037\n",
            "[9,  1540] loss: 0.00020\n",
            "[9,  1560] loss: 0.00029\n",
            "[9,  1580] loss: 0.00003\n",
            "[9,  1600] loss: 0.00001\n",
            "[9,  1620] loss: 0.00045\n",
            "[9,  1640] loss: 0.00004\n",
            "[9,  1660] loss: 0.00050\n",
            "[9,  1680] loss: 0.00024\n",
            "[9,  1700] loss: 0.00003\n",
            "[9,  1720] loss: 0.00027\n",
            "[9,  1740] loss: 0.00015\n",
            "[9,  1760] loss: 0.00035\n",
            "[9,  1780] loss: 0.00008\n",
            "[9,  1800] loss: 0.00005\n",
            "[9,  1820] loss: 0.00001\n",
            "[9,  1840] loss: 0.00015\n",
            "[9,  1860] loss: 0.00036\n",
            "[9,  1880] loss: 0.00005\n",
            "[9,  1900] loss: 0.00009\n",
            "[9,  1920] loss: 0.00003\n",
            "[9,  1940] loss: 0.00015\n",
            "[9,  1960] loss: 0.00006\n",
            "[9,  1980] loss: 0.00019\n",
            "[9,  2000] loss: 0.00004\n",
            "[9,  2020] loss: 0.00000\n",
            "[9,  2040] loss: 0.00039\n",
            "[9,  2060] loss: 0.00052\n",
            "[9,  2080] loss: 0.00006\n",
            "[9,  2100] loss: 0.00002\n",
            "[9,  2120] loss: 0.00029\n",
            "[9,  2140] loss: 0.00014\n",
            "[9,  2160] loss: 0.00004\n",
            "[9,  2180] loss: 0.00015\n",
            "[9,  2200] loss: 0.00011\n",
            "[9,  2220] loss: 0.00034\n",
            "[9,  2240] loss: 0.00030\n",
            "[9,  2260] loss: 0.00035\n",
            "[9,  2280] loss: 0.00016\n",
            "[9,  2300] loss: 0.00009\n",
            "[9,  2320] loss: 0.00045\n",
            "[9,  2340] loss: 0.00007\n",
            "[9,  2360] loss: 0.00002\n",
            "[9,  2380] loss: 0.00002\n",
            "[9,  2400] loss: 0.00011\n",
            "[9,  2420] loss: 0.00018\n",
            "[9,  2440] loss: 0.00003\n",
            "[9,  2460] loss: 0.00030\n",
            "[9,  2480] loss: 0.00016\n",
            "[9,  2500] loss: 0.00011\n",
            "[9,  2520] loss: 0.00046\n",
            "[9,  2540] loss: 0.00005\n",
            "[9,  2560] loss: 0.00011\n",
            "[9,  2580] loss: 0.00009\n",
            "[9,  2600] loss: 0.00019\n",
            "[9,  2620] loss: 0.00018\n",
            "[9,  2640] loss: 0.00003\n",
            "[9,  2660] loss: 0.00004\n",
            "[9,  2680] loss: 0.00028\n",
            "[9,  2700] loss: 0.00020\n",
            "[9,  2720] loss: 0.00010\n",
            "[9,  2740] loss: 0.00052\n",
            "[9,  2760] loss: 0.00015\n",
            "[9,  2780] loss: 0.00005\n",
            "[9,  2800] loss: 0.00016\n",
            "[9,  2820] loss: 0.00002\n",
            "[9,  2840] loss: 0.00011\n",
            "[9,  2860] loss: 0.00006\n",
            "[9,  2880] loss: 0.00018\n",
            "[9,  2900] loss: 0.00008\n",
            "[9,  2920] loss: 0.00005\n",
            "[9,  2940] loss: 0.00029\n",
            "[9,  2960] loss: 0.00024\n",
            "[9,  2980] loss: 0.00014\n",
            "[9,  3000] loss: 0.00025\n",
            "[9,  3020] loss: 0.00012\n",
            "[9,  3040] loss: 0.00002\n",
            "[9,  3060] loss: 0.00002\n",
            "[9,  3080] loss: 0.00025\n",
            "[9,  3100] loss: 0.00035\n",
            "[9,  3120] loss: 0.00021\n",
            "[9,  3140] loss: 0.00017\n",
            "[9,  3160] loss: 0.00004\n",
            "[9,  3180] loss: 0.00006\n",
            "[9,  3200] loss: 0.00012\n",
            "[9,  3220] loss: 0.00007\n",
            "[9,  3240] loss: 0.00025\n",
            "[9,  3260] loss: 0.00041\n",
            "[9,  3280] loss: 0.00003\n",
            "[9,  3300] loss: 0.00020\n",
            "[9,  3320] loss: 0.00014\n",
            "[9,  3340] loss: 0.00026\n",
            "[9,  3360] loss: 0.00030\n",
            "[9,  3380] loss: 0.00032\n",
            "[9,  3400] loss: 0.00034\n",
            "[9,  3420] loss: 0.00012\n",
            "[9,  3440] loss: 0.00021\n",
            "[9,  3460] loss: 0.00021\n",
            "[9,  3480] loss: 0.00026\n",
            "[9,  3500] loss: 0.00004\n",
            "[9,  3520] loss: 0.00022\n",
            "[9,  3540] loss: 0.00036\n",
            "[9,  3560] loss: 0.00016\n",
            "[9,  3580] loss: 0.00016\n",
            "[9,  3600] loss: 0.00092\n",
            "[9,  3620] loss: 0.00016\n",
            "[9,  3640] loss: 0.00004\n",
            "[9,  3660] loss: 0.00116\n",
            "[9,  3680] loss: 0.00010\n",
            "[9,  3700] loss: 0.00008\n",
            "[9,  3720] loss: 0.00016\n",
            "[9,  3740] loss: 0.00003\n",
            "[9,  3760] loss: 0.00008\n",
            "[9,  3780] loss: 0.00003\n",
            "[9,  3800] loss: 0.00061\n",
            "[9,  3820] loss: 0.00031\n",
            "[9,  3840] loss: 0.00023\n",
            "[9,  3860] loss: 0.00009\n",
            "[9,  3880] loss: 0.00012\n",
            "[9,  3900] loss: 0.00005\n",
            "[9,  3920] loss: 0.00011\n",
            "[9,  3940] loss: 0.00006\n",
            "[9,  3960] loss: 0.00020\n",
            "[9,  3980] loss: 0.00005\n",
            "[9,  4000] loss: 0.00005\n",
            "[9,  4020] loss: 0.00005\n",
            "[9,  4040] loss: 0.00005\n",
            "[9,  4060] loss: 0.00019\n",
            "[9,  4080] loss: 0.00007\n",
            "[9,  4100] loss: 0.00014\n",
            "[9,  4120] loss: 0.00010\n",
            "[9,  4140] loss: 0.00025\n",
            "[9,  4160] loss: 0.00013\n",
            "[9,  4180] loss: 0.00001\n",
            "[9,  4200] loss: 0.00010\n",
            "[9,  4220] loss: 0.00005\n",
            "[9,  4240] loss: 0.00081\n",
            "[9,  4260] loss: 0.00024\n",
            "[9,  4280] loss: 0.00058\n",
            "[9,  4300] loss: 0.00020\n",
            "[9,  4320] loss: 0.00007\n",
            "[9,  4340] loss: 0.00003\n",
            "[9,  4360] loss: 0.00006\n",
            "[9,  4380] loss: 0.00008\n",
            "[9,  4400] loss: 0.00012\n",
            "[9,  4420] loss: 0.00015\n",
            "[9,  4440] loss: 0.00031\n",
            "[9,  4460] loss: 0.00021\n",
            "[9,  4480] loss: 0.00009\n",
            "[9,  4500] loss: 0.00014\n",
            "[9,  4520] loss: 0.00005\n",
            "[9,  4540] loss: 0.00006\n",
            "[9,  4560] loss: 0.00001\n",
            "[9,  4580] loss: 0.00032\n",
            "[9,  4600] loss: 0.00007\n",
            "[9,  4620] loss: 0.00016\n",
            "[9,  4640] loss: 0.00001\n",
            "[9,  4660] loss: 0.00002\n",
            "[9,  4680] loss: 0.00006\n",
            "[9,  4700] loss: 0.00003\n",
            "[9,  4720] loss: 0.00031\n",
            "[9,  4740] loss: 0.00017\n",
            "[9,  4760] loss: 0.00015\n",
            "[9,  4780] loss: 0.00025\n",
            "[9,  4800] loss: 0.00043\n",
            "[9,  4820] loss: 0.00023\n",
            "[9,  4840] loss: 0.00007\n",
            "[9,  4860] loss: 0.00002\n",
            "[9,  4880] loss: 0.00026\n",
            "[9,  4900] loss: 0.00028\n",
            "[9,  4920] loss: 0.00029\n",
            "[9,  4940] loss: 0.00047\n",
            "[9,  4960] loss: 0.00018\n",
            "[9,  4980] loss: 0.00033\n",
            "[9,  5000] loss: 0.00005\n",
            "[9,  5020] loss: 0.00010\n",
            "[9,  5040] loss: 0.00010\n",
            "[9,  5060] loss: 0.00007\n",
            "[9,  5080] loss: 0.00004\n",
            "[9,  5100] loss: 0.00066\n",
            "[9,  5120] loss: 0.00003\n",
            "[9,  5140] loss: 0.00025\n",
            "[9,  5160] loss: 0.00011\n",
            "[9,  5180] loss: 0.00005\n",
            "[9,  5200] loss: 0.00061\n",
            "[9,  5220] loss: 0.00013\n",
            "[9,  5240] loss: 0.00004\n",
            "[9,  5260] loss: 0.00026\n",
            "[9,  5280] loss: 0.00038\n",
            "[9,  5300] loss: 0.00063\n",
            "[9,  5320] loss: 0.00049\n",
            "[9,  5340] loss: 0.00004\n",
            "[9,  5360] loss: 0.00014\n",
            "[9,  5380] loss: 0.00019\n",
            "[9,  5400] loss: 0.00010\n",
            "[9,  5420] loss: 0.00025\n",
            "[9,  5440] loss: 0.00066\n",
            "[9,  5460] loss: 0.00017\n",
            "[9,  5480] loss: 0.00027\n",
            "[9,  5500] loss: 0.00004\n",
            "[9,  5520] loss: 0.00016\n",
            "[9,  5540] loss: 0.00003\n",
            "[9,  5560] loss: 0.00007\n",
            "[9,  5580] loss: 0.00004\n",
            "[9,  5600] loss: 0.00012\n",
            "[9,  5620] loss: 0.00037\n",
            "[9,  5640] loss: 0.00008\n",
            "[9,  5660] loss: 0.00034\n",
            "[9,  5680] loss: 0.00012\n",
            "[9,  5700] loss: 0.00017\n",
            "[9,  5720] loss: 0.00010\n",
            "[9,  5740] loss: 0.00003\n",
            "[9,  5760] loss: 0.00020\n",
            "[9,  5780] loss: 0.00003\n",
            "[9,  5800] loss: 0.00022\n",
            "[9,  5820] loss: 0.00073\n",
            "[9,  5840] loss: 0.00005\n",
            "[9,  5860] loss: 0.00026\n",
            "[9,  5880] loss: 0.00012\n",
            "[9,  5900] loss: 0.00058\n",
            "[9,  5920] loss: 0.00007\n",
            "[9,  5940] loss: 0.00018\n",
            "[9,  5960] loss: 0.00069\n",
            "[9,  5980] loss: 0.00010\n",
            "[9,  6000] loss: 0.00004\n",
            "[9,  6020] loss: 0.00004\n",
            "[9,  6040] loss: 0.00005\n",
            "[9,  6060] loss: 0.00005\n",
            "[9,  6080] loss: 0.00005\n",
            "[9,  6100] loss: 0.00006\n",
            "[9,  6120] loss: 0.00015\n",
            "[9,  6140] loss: 0.00003\n",
            "[9,  6160] loss: 0.00006\n",
            "[9,  6180] loss: 0.00010\n",
            "[9,  6200] loss: 0.00035\n",
            "[9,  6220] loss: 0.00012\n",
            "[9,  6240] loss: 0.00002\n",
            "[9,  6260] loss: 0.00003\n",
            "[9,  6280] loss: 0.00014\n",
            "[9,  6300] loss: 0.00016\n",
            "[9,  6320] loss: 0.00004\n",
            "[9,  6340] loss: 0.00012\n",
            "[9,  6360] loss: 0.00023\n",
            "[9,  6380] loss: 0.00038\n",
            "[9,  6400] loss: 0.00004\n",
            "[9,  6420] loss: 0.00008\n",
            "[9,  6440] loss: 0.00004\n",
            "[9,  6460] loss: 0.00009\n",
            "[9,  6480] loss: 0.00003\n",
            "[9,  6500] loss: 0.00021\n",
            "[9,  6520] loss: 0.00023\n",
            "[9,  6540] loss: 0.00009\n",
            "[9,  6560] loss: 0.00019\n",
            "[9,  6580] loss: 0.00010\n",
            "[9,  6600] loss: 0.00004\n",
            "[9,  6620] loss: 0.00031\n",
            "[9,  6640] loss: 0.00018\n",
            "[9,  6660] loss: 0.00014\n",
            "[9,  6680] loss: 0.00014\n",
            "[9,  6700] loss: 0.00024\n",
            "[9,  6720] loss: 0.00027\n",
            "[9,  6740] loss: 0.00017\n",
            "[9,  6760] loss: 0.00015\n",
            "[9,  6780] loss: 0.00001\n",
            "[9,  6800] loss: 0.00010\n",
            "[9,  6820] loss: 0.00007\n",
            "[9,  6840] loss: 0.00017\n",
            "[9,  6860] loss: 0.00002\n",
            "[9,  6880] loss: 0.00007\n",
            "[9,  6900] loss: 0.00008\n",
            "[9,  6920] loss: 0.00005\n",
            "[9,  6940] loss: 0.00016\n",
            "[9,  6960] loss: 0.00032\n",
            "[9,  6980] loss: 0.00008\n",
            "[9,  7000] loss: 0.00027\n",
            "[9,  7020] loss: 0.00066\n",
            "[9,  7040] loss: 0.00016\n",
            "[9,  7060] loss: 0.00022\n",
            "[9,  7080] loss: 0.00018\n",
            "[9,  7100] loss: 0.00012\n",
            "[9,  7120] loss: 0.00001\n",
            "[9,  7140] loss: 0.00001\n",
            "[9,  7160] loss: 0.00003\n",
            "[9,  7180] loss: 0.00004\n",
            "[9,  7200] loss: 0.00010\n",
            "[9,  7220] loss: 0.00007\n",
            "[9,  7240] loss: 0.00006\n",
            "[9,  7260] loss: 0.00039\n",
            "[9,  7280] loss: 0.00017\n",
            "[9,  7300] loss: 0.00059\n",
            "[9,  7320] loss: 0.00017\n",
            "[9,  7340] loss: 0.00009\n",
            "[9,  7360] loss: 0.00027\n",
            "[9,  7380] loss: 0.00006\n",
            "[9,  7400] loss: 0.00022\n",
            "[9,  7420] loss: 0.00016\n",
            "[9,  7440] loss: 0.00014\n",
            "[9,  7460] loss: 0.00003\n",
            "[9,  7480] loss: 0.00005\n",
            "[9,  7500] loss: 0.00007\n",
            "Finished training for epoch 8 time taken = 566.7755036354065\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 4th\n",
            "[10,    20] loss: 0.00017\n",
            "[10,    40] loss: 0.00011\n",
            "[10,    60] loss: 0.00004\n",
            "[10,    80] loss: 0.00004\n",
            "[10,   100] loss: 0.00002\n",
            "[10,   120] loss: 0.00003\n",
            "[10,   140] loss: 0.00001\n",
            "[10,   160] loss: 0.00007\n",
            "[10,   180] loss: 0.00003\n",
            "[10,   200] loss: 0.00006\n",
            "[10,   220] loss: 0.00009\n",
            "[10,   240] loss: 0.00002\n",
            "[10,   260] loss: 0.00021\n",
            "[10,   280] loss: 0.00004\n",
            "[10,   300] loss: 0.00004\n",
            "[10,   320] loss: 0.00001\n",
            "[10,   340] loss: 0.00008\n",
            "[10,   360] loss: 0.00011\n",
            "[10,   380] loss: 0.00003\n",
            "[10,   400] loss: 0.00003\n",
            "[10,   420] loss: 0.00002\n",
            "[10,   440] loss: 0.00028\n",
            "[10,   460] loss: 0.00001\n",
            "[10,   480] loss: 0.00005\n",
            "[10,   500] loss: 0.00007\n",
            "[10,   520] loss: 0.00010\n",
            "[10,   540] loss: 0.00022\n",
            "[10,   560] loss: 0.00004\n",
            "[10,   580] loss: 0.00020\n",
            "[10,   600] loss: 0.00007\n",
            "[10,   620] loss: 0.00038\n",
            "[10,   640] loss: 0.00022\n",
            "[10,   660] loss: 0.00027\n",
            "[10,   680] loss: 0.00007\n",
            "[10,   700] loss: 0.00011\n",
            "[10,   720] loss: 0.00050\n",
            "[10,   740] loss: 0.00009\n",
            "[10,   760] loss: 0.00007\n",
            "[10,   780] loss: 0.00008\n",
            "[10,   800] loss: 0.00002\n",
            "[10,   820] loss: 0.00024\n",
            "[10,   840] loss: 0.00030\n",
            "[10,   860] loss: 0.00012\n",
            "[10,   880] loss: 0.00003\n",
            "[10,   900] loss: 0.00008\n",
            "[10,   920] loss: 0.00035\n",
            "[10,   940] loss: 0.00012\n",
            "[10,   960] loss: 0.00003\n",
            "[10,   980] loss: 0.00003\n",
            "[10,  1000] loss: 0.00003\n",
            "[10,  1020] loss: 0.00003\n",
            "[10,  1040] loss: 0.00003\n",
            "[10,  1060] loss: 0.00002\n",
            "[10,  1080] loss: 0.00009\n",
            "[10,  1100] loss: 0.00002\n",
            "[10,  1120] loss: 0.00006\n",
            "[10,  1140] loss: 0.00002\n",
            "[10,  1160] loss: 0.00019\n",
            "[10,  1180] loss: 0.00043\n",
            "[10,  1200] loss: 0.00021\n",
            "[10,  1220] loss: 0.00010\n",
            "[10,  1240] loss: 0.00010\n",
            "[10,  1260] loss: 0.00025\n",
            "[10,  1280] loss: 0.00030\n",
            "[10,  1300] loss: 0.00022\n",
            "[10,  1320] loss: 0.00013\n",
            "[10,  1340] loss: 0.00009\n",
            "[10,  1360] loss: 0.00005\n",
            "[10,  1380] loss: 0.00006\n",
            "[10,  1400] loss: 0.00012\n",
            "[10,  1420] loss: 0.00002\n",
            "[10,  1440] loss: 0.00002\n",
            "[10,  1460] loss: 0.00017\n",
            "[10,  1480] loss: 0.00032\n",
            "[10,  1500] loss: 0.00034\n",
            "[10,  1520] loss: 0.00009\n",
            "[10,  1540] loss: 0.00032\n",
            "[10,  1560] loss: 0.00020\n",
            "[10,  1580] loss: 0.00011\n",
            "[10,  1600] loss: 0.00000\n",
            "[10,  1620] loss: 0.00001\n",
            "[10,  1640] loss: 0.00027\n",
            "[10,  1660] loss: 0.00008\n",
            "[10,  1680] loss: 0.00009\n",
            "[10,  1700] loss: 0.00018\n",
            "[10,  1720] loss: 0.00009\n",
            "[10,  1740] loss: 0.00003\n",
            "[10,  1760] loss: 0.00004\n",
            "[10,  1780] loss: 0.00004\n",
            "[10,  1800] loss: 0.00051\n",
            "[10,  1820] loss: 0.00007\n",
            "[10,  1840] loss: 0.00007\n",
            "[10,  1860] loss: 0.00009\n",
            "[10,  1880] loss: 0.00014\n",
            "[10,  1900] loss: 0.00024\n",
            "[10,  1920] loss: 0.00005\n",
            "[10,  1940] loss: 0.00002\n",
            "[10,  1960] loss: 0.00019\n",
            "[10,  1980] loss: 0.00009\n",
            "[10,  2000] loss: 0.00013\n",
            "[10,  2020] loss: 0.00015\n",
            "[10,  2040] loss: 0.00002\n",
            "[10,  2060] loss: 0.00009\n",
            "[10,  2080] loss: 0.00001\n",
            "[10,  2100] loss: 0.00018\n",
            "[10,  2120] loss: 0.00006\n",
            "[10,  2140] loss: 0.00006\n",
            "[10,  2160] loss: 0.00023\n",
            "[10,  2180] loss: 0.00011\n",
            "[10,  2200] loss: 0.00003\n",
            "[10,  2220] loss: 0.00008\n",
            "[10,  2240] loss: 0.00002\n",
            "[10,  2260] loss: 0.00002\n",
            "[10,  2280] loss: 0.00009\n",
            "[10,  2300] loss: 0.00003\n",
            "[10,  2320] loss: 0.00001\n",
            "[10,  2340] loss: 0.00064\n",
            "[10,  2360] loss: 0.00009\n",
            "[10,  2380] loss: 0.00029\n",
            "[10,  2400] loss: 0.00003\n",
            "[10,  2420] loss: 0.00024\n",
            "[10,  2440] loss: 0.00018\n",
            "[10,  2460] loss: 0.00018\n",
            "[10,  2480] loss: 0.00013\n",
            "[10,  2500] loss: 0.00004\n",
            "[10,  2520] loss: 0.00014\n",
            "[10,  2540] loss: 0.00016\n",
            "[10,  2560] loss: 0.00024\n",
            "[10,  2580] loss: 0.00020\n",
            "[10,  2600] loss: 0.00003\n",
            "[10,  2620] loss: 0.00064\n",
            "[10,  2640] loss: 0.00001\n",
            "[10,  2660] loss: 0.00027\n",
            "[10,  2680] loss: 0.00020\n",
            "[10,  2700] loss: 0.00003\n",
            "[10,  2720] loss: 0.00001\n",
            "[10,  2740] loss: 0.00006\n",
            "[10,  2760] loss: 0.00005\n",
            "[10,  2780] loss: 0.00053\n",
            "[10,  2800] loss: 0.00005\n",
            "[10,  2820] loss: 0.00022\n",
            "[10,  2840] loss: 0.00018\n",
            "[10,  2860] loss: 0.00014\n",
            "[10,  2880] loss: 0.00002\n",
            "[10,  2900] loss: 0.00019\n",
            "[10,  2920] loss: 0.00022\n",
            "[10,  2940] loss: 0.00019\n",
            "[10,  2960] loss: 0.00028\n",
            "[10,  2980] loss: 0.00005\n",
            "[10,  3000] loss: 0.00009\n",
            "[10,  3020] loss: 0.00003\n",
            "[10,  3040] loss: 0.00061\n",
            "[10,  3060] loss: 0.00002\n",
            "[10,  3080] loss: 0.00011\n",
            "[10,  3100] loss: 0.00051\n",
            "[10,  3120] loss: 0.00002\n",
            "[10,  3140] loss: 0.00001\n",
            "[10,  3160] loss: 0.00032\n",
            "[10,  3180] loss: 0.00039\n",
            "[10,  3200] loss: 0.00036\n",
            "[10,  3220] loss: 0.00026\n",
            "[10,  3240] loss: 0.00054\n",
            "[10,  3260] loss: 0.00026\n",
            "[10,  3280] loss: 0.00063\n",
            "[10,  3300] loss: 0.00009\n",
            "[10,  3320] loss: 0.00018\n",
            "[10,  3340] loss: 0.00004\n",
            "[10,  3360] loss: 0.00011\n",
            "[10,  3380] loss: 0.00003\n",
            "[10,  3400] loss: 0.00005\n",
            "[10,  3420] loss: 0.00002\n",
            "[10,  3440] loss: 0.00009\n",
            "[10,  3460] loss: 0.00074\n",
            "[10,  3480] loss: 0.00002\n",
            "[10,  3500] loss: 0.00021\n",
            "[10,  3520] loss: 0.00009\n",
            "[10,  3540] loss: 0.00008\n",
            "[10,  3560] loss: 0.00008\n",
            "[10,  3580] loss: 0.00005\n",
            "[10,  3600] loss: 0.00001\n",
            "[10,  3620] loss: 0.00023\n",
            "[10,  3640] loss: 0.00007\n",
            "[10,  3660] loss: 0.00001\n",
            "[10,  3680] loss: 0.00019\n",
            "[10,  3700] loss: 0.00005\n",
            "[10,  3720] loss: 0.00023\n",
            "[10,  3740] loss: 0.00002\n",
            "[10,  3760] loss: 0.00016\n",
            "[10,  3780] loss: 0.00009\n",
            "[10,  3800] loss: 0.00002\n",
            "[10,  3820] loss: 0.00006\n",
            "[10,  3840] loss: 0.00003\n",
            "[10,  3860] loss: 0.00019\n",
            "[10,  3880] loss: 0.00010\n",
            "[10,  3900] loss: 0.00005\n",
            "[10,  3920] loss: 0.00006\n",
            "[10,  3940] loss: 0.00022\n",
            "[10,  3960] loss: 0.00010\n",
            "[10,  3980] loss: 0.00025\n",
            "[10,  4000] loss: 0.00008\n",
            "[10,  4020] loss: 0.00010\n",
            "[10,  4040] loss: 0.00019\n",
            "[10,  4060] loss: 0.00024\n",
            "[10,  4080] loss: 0.00009\n",
            "[10,  4100] loss: 0.00002\n",
            "[10,  4120] loss: 0.00017\n",
            "[10,  4140] loss: 0.00028\n",
            "[10,  4160] loss: 0.00012\n",
            "[10,  4180] loss: 0.00010\n",
            "[10,  4200] loss: 0.00008\n",
            "[10,  4220] loss: 0.00031\n",
            "[10,  4240] loss: 0.00010\n",
            "[10,  4260] loss: 0.00031\n",
            "[10,  4280] loss: 0.00008\n",
            "[10,  4300] loss: 0.00012\n",
            "[10,  4320] loss: 0.00001\n",
            "[10,  4340] loss: 0.00016\n",
            "[10,  4360] loss: 0.00022\n",
            "[10,  4380] loss: 0.00002\n",
            "[10,  4400] loss: 0.00004\n",
            "[10,  4420] loss: 0.00025\n",
            "[10,  4440] loss: 0.00003\n",
            "[10,  4460] loss: 0.00011\n",
            "[10,  4480] loss: 0.00035\n",
            "[10,  4500] loss: 0.00018\n",
            "[10,  4520] loss: 0.00002\n",
            "[10,  4540] loss: 0.00040\n",
            "[10,  4560] loss: 0.00011\n",
            "[10,  4580] loss: 0.00004\n",
            "[10,  4600] loss: 0.00002\n",
            "[10,  4620] loss: 0.00019\n",
            "[10,  4640] loss: 0.00036\n",
            "[10,  4660] loss: 0.00015\n",
            "[10,  4680] loss: 0.00023\n",
            "[10,  4700] loss: 0.00037\n",
            "[10,  4720] loss: 0.00001\n",
            "[10,  4740] loss: 0.00008\n",
            "[10,  4760] loss: 0.00009\n",
            "[10,  4780] loss: 0.00022\n",
            "[10,  4800] loss: 0.00023\n",
            "[10,  4820] loss: 0.00017\n",
            "[10,  4840] loss: 0.00002\n",
            "[10,  4860] loss: 0.00036\n",
            "[10,  4880] loss: 0.00035\n",
            "[10,  4900] loss: 0.00011\n",
            "[10,  4920] loss: 0.00003\n",
            "[10,  4940] loss: 0.00004\n",
            "[10,  4960] loss: 0.00019\n",
            "[10,  4980] loss: 0.00003\n",
            "[10,  5000] loss: 0.00021\n",
            "[10,  5020] loss: 0.00007\n",
            "[10,  5040] loss: 0.00006\n",
            "[10,  5060] loss: 0.00003\n",
            "[10,  5080] loss: 0.00042\n",
            "[10,  5100] loss: 0.00047\n",
            "[10,  5120] loss: 0.00015\n",
            "[10,  5140] loss: 0.00002\n",
            "[10,  5160] loss: 0.00011\n",
            "[10,  5180] loss: 0.00009\n",
            "[10,  5200] loss: 0.00023\n",
            "[10,  5220] loss: 0.00016\n",
            "[10,  5240] loss: 0.00005\n",
            "[10,  5260] loss: 0.00021\n",
            "[10,  5280] loss: 0.00008\n",
            "[10,  5300] loss: 0.00021\n",
            "[10,  5320] loss: 0.00005\n",
            "[10,  5340] loss: 0.00008\n",
            "[10,  5360] loss: 0.00038\n",
            "[10,  5380] loss: 0.00003\n",
            "[10,  5400] loss: 0.00025\n",
            "[10,  5420] loss: 0.00009\n",
            "[10,  5440] loss: 0.00035\n",
            "[10,  5460] loss: 0.00055\n",
            "[10,  5480] loss: 0.00045\n",
            "[10,  5500] loss: 0.00019\n",
            "[10,  5520] loss: 0.00021\n",
            "[10,  5540] loss: 0.00008\n",
            "[10,  5560] loss: 0.00001\n",
            "[10,  5580] loss: 0.00022\n",
            "[10,  5600] loss: 0.00011\n",
            "[10,  5620] loss: 0.00004\n",
            "[10,  5640] loss: 0.00004\n",
            "[10,  5660] loss: 0.00031\n",
            "[10,  5680] loss: 0.00016\n",
            "[10,  5700] loss: 0.00003\n",
            "[10,  5720] loss: 0.00006\n",
            "[10,  5740] loss: 0.00012\n",
            "[10,  5760] loss: 0.00043\n",
            "[10,  5780] loss: 0.00055\n",
            "[10,  5800] loss: 0.00004\n",
            "[10,  5820] loss: 0.00013\n",
            "[10,  5840] loss: 0.00026\n",
            "[10,  5860] loss: 0.00016\n",
            "[10,  5880] loss: 0.00014\n",
            "[10,  5900] loss: 0.00029\n",
            "[10,  5920] loss: 0.00001\n",
            "[10,  5940] loss: 0.00023\n",
            "[10,  5960] loss: 0.00008\n",
            "[10,  5980] loss: 0.00002\n",
            "[10,  6000] loss: 0.00016\n",
            "[10,  6020] loss: 0.00021\n",
            "[10,  6040] loss: 0.00005\n",
            "[10,  6060] loss: 0.00004\n",
            "[10,  6080] loss: 0.00011\n",
            "[10,  6100] loss: 0.00040\n",
            "[10,  6120] loss: 0.00005\n",
            "[10,  6140] loss: 0.00021\n",
            "[10,  6160] loss: 0.00041\n",
            "[10,  6180] loss: 0.00006\n",
            "[10,  6200] loss: 0.00006\n",
            "[10,  6220] loss: 0.00004\n",
            "[10,  6240] loss: 0.00015\n",
            "[10,  6260] loss: 0.00006\n",
            "[10,  6280] loss: 0.00005\n",
            "[10,  6300] loss: 0.00099\n",
            "[10,  6320] loss: 0.00018\n",
            "[10,  6340] loss: 0.00001\n",
            "[10,  6360] loss: 0.00003\n",
            "[10,  6380] loss: 0.00006\n",
            "[10,  6400] loss: 0.00010\n",
            "[10,  6420] loss: 0.00009\n",
            "[10,  6440] loss: 0.00006\n",
            "[10,  6460] loss: 0.00008\n",
            "[10,  6480] loss: 0.00091\n",
            "[10,  6500] loss: 0.00042\n",
            "[10,  6520] loss: 0.00015\n",
            "[10,  6540] loss: 0.00047\n",
            "[10,  6560] loss: 0.00004\n",
            "[10,  6580] loss: 0.00006\n",
            "[10,  6600] loss: 0.00019\n",
            "[10,  6620] loss: 0.00030\n",
            "[10,  6640] loss: 0.00026\n",
            "[10,  6660] loss: 0.00003\n",
            "[10,  6680] loss: 0.00008\n",
            "[10,  6700] loss: 0.00005\n",
            "[10,  6720] loss: 0.00061\n",
            "[10,  6740] loss: 0.00063\n",
            "[10,  6760] loss: 0.00016\n",
            "[10,  6780] loss: 0.00004\n",
            "[10,  6800] loss: 0.00021\n",
            "[10,  6820] loss: 0.00010\n",
            "[10,  6840] loss: 0.00002\n",
            "[10,  6860] loss: 0.00011\n",
            "[10,  6880] loss: 0.00000\n",
            "[10,  6900] loss: 0.00007\n",
            "[10,  6920] loss: 0.00018\n",
            "[10,  6940] loss: 0.00004\n",
            "[10,  6960] loss: 0.00068\n",
            "[10,  6980] loss: 0.00021\n",
            "[10,  7000] loss: 0.00034\n",
            "[10,  7020] loss: 0.00012\n",
            "[10,  7040] loss: 0.00011\n",
            "[10,  7060] loss: 0.00008\n",
            "[10,  7080] loss: 0.00017\n",
            "[10,  7100] loss: 0.00036\n",
            "[10,  7120] loss: 0.00027\n",
            "[10,  7140] loss: 0.00006\n",
            "[10,  7160] loss: 0.00010\n",
            "[10,  7180] loss: 0.00018\n",
            "[10,  7200] loss: 0.00020\n",
            "[10,  7220] loss: 0.00013\n",
            "[10,  7240] loss: 0.00012\n",
            "[10,  7260] loss: 0.00001\n",
            "[10,  7280] loss: 0.00011\n",
            "[10,  7300] loss: 0.00022\n",
            "[10,  7320] loss: 0.00031\n",
            "[10,  7340] loss: 0.00006\n",
            "[10,  7360] loss: 0.00003\n",
            "[10,  7380] loss: 0.00058\n",
            "[10,  7400] loss: 0.00003\n",
            "[10,  7420] loss: 0.00014\n",
            "[10,  7440] loss: 0.00027\n",
            "[10,  7460] loss: 0.00026\n",
            "[10,  7480] loss: 0.00025\n",
            "[10,  7500] loss: 0.00025\n",
            "Finished training for epoch 9 time taken = 630.5840969085693\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 5th\n",
            "[11,    20] loss: 0.00007\n",
            "[11,    40] loss: 0.00006\n",
            "[11,    60] loss: 0.00005\n",
            "[11,    80] loss: 0.00002\n",
            "[11,   100] loss: 0.00012\n",
            "[11,   120] loss: 0.00004\n",
            "[11,   140] loss: 0.00004\n",
            "[11,   160] loss: 0.00005\n",
            "[11,   180] loss: 0.00008\n",
            "[11,   200] loss: 0.00007\n",
            "[11,   220] loss: 0.00003\n",
            "[11,   240] loss: 0.00009\n",
            "[11,   260] loss: 0.00033\n",
            "[11,   280] loss: 0.00002\n",
            "[11,   300] loss: 0.00007\n",
            "[11,   320] loss: 0.00015\n",
            "[11,   340] loss: 0.00003\n",
            "[11,   360] loss: 0.00004\n",
            "[11,   380] loss: 0.00013\n",
            "[11,   400] loss: 0.00005\n",
            "[11,   420] loss: 0.00010\n",
            "[11,   440] loss: 0.00004\n",
            "[11,   460] loss: 0.00016\n",
            "[11,   480] loss: 0.00034\n",
            "[11,   500] loss: 0.00013\n",
            "[11,   520] loss: 0.00007\n",
            "[11,   540] loss: 0.00027\n",
            "[11,   560] loss: 0.00003\n",
            "[11,   580] loss: 0.00005\n",
            "[11,   600] loss: 0.00002\n",
            "[11,   620] loss: 0.00004\n",
            "[11,   640] loss: 0.00046\n",
            "[11,   660] loss: 0.00002\n",
            "[11,   680] loss: 0.00019\n",
            "[11,   700] loss: 0.00002\n",
            "[11,   720] loss: 0.00007\n",
            "[11,   740] loss: 0.00004\n",
            "[11,   760] loss: 0.00007\n",
            "[11,   780] loss: 0.00005\n",
            "[11,   800] loss: 0.00004\n",
            "[11,   820] loss: 0.00006\n",
            "[11,   840] loss: 0.00002\n",
            "[11,   860] loss: 0.00033\n",
            "[11,   880] loss: 0.00022\n",
            "[11,   900] loss: 0.00001\n",
            "[11,   920] loss: 0.00010\n",
            "[11,   940] loss: 0.00010\n",
            "[11,   960] loss: 0.00008\n",
            "[11,   980] loss: 0.00013\n",
            "[11,  1000] loss: 0.00004\n",
            "[11,  1020] loss: 0.00002\n",
            "[11,  1040] loss: 0.00011\n",
            "[11,  1060] loss: 0.00004\n",
            "[11,  1080] loss: 0.00007\n",
            "[11,  1100] loss: 0.00003\n",
            "[11,  1120] loss: 0.00004\n",
            "[11,  1140] loss: 0.00035\n",
            "[11,  1160] loss: 0.00003\n",
            "[11,  1180] loss: 0.00005\n",
            "[11,  1200] loss: 0.00030\n",
            "[11,  1220] loss: 0.00038\n",
            "[11,  1240] loss: 0.00010\n",
            "[11,  1260] loss: 0.00005\n",
            "[11,  1280] loss: 0.00003\n",
            "[11,  1300] loss: 0.00050\n",
            "[11,  1320] loss: 0.00005\n",
            "[11,  1340] loss: 0.00004\n",
            "[11,  1360] loss: 0.00006\n",
            "[11,  1380] loss: 0.00008\n",
            "[11,  1400] loss: 0.00011\n",
            "[11,  1420] loss: 0.00012\n",
            "[11,  1440] loss: 0.00044\n",
            "[11,  1460] loss: 0.00007\n",
            "[11,  1480] loss: 0.00019\n",
            "[11,  1500] loss: 0.00003\n",
            "[11,  1520] loss: 0.00001\n",
            "[11,  1540] loss: 0.00016\n",
            "[11,  1560] loss: 0.00033\n",
            "[11,  1580] loss: 0.00012\n",
            "[11,  1600] loss: 0.00026\n",
            "[11,  1620] loss: 0.00021\n",
            "[11,  1640] loss: 0.00002\n",
            "[11,  1660] loss: 0.00008\n",
            "[11,  1680] loss: 0.00010\n",
            "[11,  1700] loss: 0.00014\n",
            "[11,  1720] loss: 0.00048\n",
            "[11,  1740] loss: 0.00013\n",
            "[11,  1760] loss: 0.00013\n",
            "[11,  1780] loss: 0.00002\n",
            "[11,  1800] loss: 0.00006\n",
            "[11,  1820] loss: 0.00008\n",
            "[11,  1840] loss: 0.00014\n",
            "[11,  1860] loss: 0.00001\n",
            "[11,  1880] loss: 0.00001\n",
            "[11,  1900] loss: 0.00017\n",
            "[11,  1920] loss: 0.00002\n",
            "[11,  1940] loss: 0.00001\n",
            "[11,  1960] loss: 0.00003\n",
            "[11,  1980] loss: 0.00003\n",
            "[11,  2000] loss: 0.00010\n",
            "[11,  2020] loss: 0.00009\n",
            "[11,  2040] loss: 0.00054\n",
            "[11,  2060] loss: 0.00004\n",
            "[11,  2080] loss: 0.00022\n",
            "[11,  2100] loss: 0.00029\n",
            "[11,  2120] loss: 0.00011\n",
            "[11,  2140] loss: 0.00008\n",
            "[11,  2160] loss: 0.00006\n",
            "[11,  2180] loss: 0.00007\n",
            "[11,  2200] loss: 0.00011\n",
            "[11,  2220] loss: 0.00019\n",
            "[11,  2240] loss: 0.00030\n",
            "[11,  2260] loss: 0.00001\n",
            "[11,  2280] loss: 0.00018\n",
            "[11,  2300] loss: 0.00009\n",
            "[11,  2320] loss: 0.00014\n",
            "[11,  2340] loss: 0.00015\n",
            "[11,  2360] loss: 0.00021\n",
            "[11,  2380] loss: 0.00005\n",
            "[11,  2400] loss: 0.00004\n",
            "[11,  2420] loss: 0.00004\n",
            "[11,  2440] loss: 0.00051\n",
            "[11,  2460] loss: 0.00007\n",
            "[11,  2480] loss: 0.00042\n",
            "[11,  2500] loss: 0.00003\n",
            "[11,  2520] loss: 0.00001\n",
            "[11,  2540] loss: 0.00002\n",
            "[11,  2560] loss: 0.00005\n",
            "[11,  2580] loss: 0.00001\n",
            "[11,  2600] loss: 0.00022\n",
            "[11,  2620] loss: 0.00020\n",
            "[11,  2640] loss: 0.00005\n",
            "[11,  2660] loss: 0.00011\n",
            "[11,  2680] loss: 0.00002\n",
            "[11,  2700] loss: 0.00011\n",
            "[11,  2720] loss: 0.00014\n",
            "[11,  2740] loss: 0.00035\n",
            "[11,  2760] loss: 0.00003\n",
            "[11,  2780] loss: 0.00027\n",
            "[11,  2800] loss: 0.00001\n",
            "[11,  2820] loss: 0.00016\n",
            "[11,  2840] loss: 0.00016\n",
            "[11,  2860] loss: 0.00003\n",
            "[11,  2880] loss: 0.00001\n",
            "[11,  2900] loss: 0.00011\n",
            "[11,  2920] loss: 0.00003\n",
            "[11,  2940] loss: 0.00011\n",
            "[11,  2960] loss: 0.00051\n",
            "[11,  2980] loss: 0.00011\n",
            "[11,  3000] loss: 0.00007\n",
            "[11,  3020] loss: 0.00002\n",
            "[11,  3040] loss: 0.00007\n",
            "[11,  3060] loss: 0.00004\n",
            "[11,  3080] loss: 0.00004\n",
            "[11,  3100] loss: 0.00001\n",
            "[11,  3120] loss: 0.00094\n",
            "[11,  3140] loss: 0.00004\n",
            "[11,  3160] loss: 0.00009\n",
            "[11,  3180] loss: 0.00032\n",
            "[11,  3200] loss: 0.00013\n",
            "[11,  3220] loss: 0.00016\n",
            "[11,  3240] loss: 0.00001\n",
            "[11,  3260] loss: 0.00052\n",
            "[11,  3280] loss: 0.00009\n",
            "[11,  3300] loss: 0.00012\n",
            "[11,  3320] loss: 0.00015\n",
            "[11,  3340] loss: 0.00006\n",
            "[11,  3360] loss: 0.00007\n",
            "[11,  3380] loss: 0.00008\n",
            "[11,  3400] loss: 0.00004\n",
            "[11,  3420] loss: 0.00005\n",
            "[11,  3440] loss: 0.00047\n",
            "[11,  3460] loss: 0.00011\n",
            "[11,  3480] loss: 0.00012\n",
            "[11,  3500] loss: 0.00002\n",
            "[11,  3520] loss: 0.00008\n",
            "[11,  3540] loss: 0.00007\n",
            "[11,  3560] loss: 0.00003\n",
            "[11,  3580] loss: 0.00004\n",
            "[11,  3600] loss: 0.00008\n",
            "[11,  3620] loss: 0.00012\n",
            "[11,  3640] loss: 0.00032\n",
            "[11,  3660] loss: 0.00011\n",
            "[11,  3680] loss: 0.00001\n",
            "[11,  3700] loss: 0.00010\n",
            "[11,  3720] loss: 0.00003\n",
            "[11,  3740] loss: 0.00001\n",
            "[11,  3760] loss: 0.00016\n",
            "[11,  3780] loss: 0.00001\n",
            "[11,  3800] loss: 0.00043\n",
            "[11,  3820] loss: 0.00024\n",
            "[11,  3840] loss: 0.00008\n",
            "[11,  3860] loss: 0.00002\n",
            "[11,  3880] loss: 0.00018\n",
            "[11,  3900] loss: 0.00004\n",
            "[11,  3920] loss: 0.00016\n",
            "[11,  3940] loss: 0.00006\n",
            "[11,  3960] loss: 0.00008\n",
            "[11,  3980] loss: 0.00009\n",
            "[11,  4000] loss: 0.00016\n",
            "[11,  4020] loss: 0.00005\n",
            "[11,  4040] loss: 0.00025\n",
            "[11,  4060] loss: 0.00014\n",
            "[11,  4080] loss: 0.00005\n",
            "[11,  4100] loss: 0.00037\n",
            "[11,  4120] loss: 0.00010\n",
            "[11,  4140] loss: 0.00034\n",
            "[11,  4160] loss: 0.00031\n",
            "[11,  4180] loss: 0.00006\n",
            "[11,  4200] loss: 0.00004\n",
            "[11,  4220] loss: 0.00002\n",
            "[11,  4240] loss: 0.00011\n",
            "[11,  4260] loss: 0.00002\n",
            "[11,  4280] loss: 0.00005\n",
            "[11,  4300] loss: 0.00007\n",
            "[11,  4320] loss: 0.00025\n",
            "[11,  4340] loss: 0.00001\n",
            "[11,  4360] loss: 0.00021\n",
            "[11,  4380] loss: 0.00004\n",
            "[11,  4400] loss: 0.00002\n",
            "[11,  4420] loss: 0.00003\n",
            "[11,  4440] loss: 0.00011\n",
            "[11,  4460] loss: 0.00003\n",
            "[11,  4480] loss: 0.00058\n",
            "[11,  4500] loss: 0.00022\n",
            "[11,  4520] loss: 0.00010\n",
            "[11,  4540] loss: 0.00008\n",
            "[11,  4560] loss: 0.00001\n",
            "[11,  4580] loss: 0.00002\n",
            "[11,  4600] loss: 0.00002\n",
            "[11,  4620] loss: 0.00001\n",
            "[11,  4640] loss: 0.00006\n",
            "[11,  4660] loss: 0.00047\n",
            "[11,  4680] loss: 0.00005\n",
            "[11,  4700] loss: 0.00047\n",
            "[11,  4720] loss: 0.00011\n",
            "[11,  4740] loss: 0.00019\n",
            "[11,  4760] loss: 0.00003\n",
            "[11,  4780] loss: 0.00025\n",
            "[11,  4800] loss: 0.00007\n",
            "[11,  4820] loss: 0.00002\n",
            "[11,  4840] loss: 0.00005\n",
            "[11,  4860] loss: 0.00005\n",
            "[11,  4880] loss: 0.00008\n",
            "[11,  4900] loss: 0.00020\n",
            "[11,  4920] loss: 0.00031\n",
            "[11,  4940] loss: 0.00005\n",
            "[11,  4960] loss: 0.00004\n",
            "[11,  4980] loss: 0.00013\n",
            "[11,  5000] loss: 0.00020\n",
            "[11,  5020] loss: 0.00029\n",
            "[11,  5040] loss: 0.00016\n",
            "[11,  5060] loss: 0.00001\n",
            "[11,  5080] loss: 0.00006\n",
            "[11,  5100] loss: 0.00005\n",
            "[11,  5120] loss: 0.00001\n",
            "[11,  5140] loss: 0.00002\n",
            "[11,  5160] loss: 0.00002\n",
            "[11,  5180] loss: 0.00002\n",
            "[11,  5200] loss: 0.00018\n",
            "[11,  5220] loss: 0.00011\n",
            "[11,  5240] loss: 0.00002\n",
            "[11,  5260] loss: 0.00004\n",
            "[11,  5280] loss: 0.00005\n",
            "[11,  5300] loss: 0.00017\n",
            "[11,  5320] loss: 0.00004\n",
            "[11,  5340] loss: 0.00010\n",
            "[11,  5360] loss: 0.00007\n",
            "[11,  5380] loss: 0.00012\n",
            "[11,  5400] loss: 0.00009\n",
            "[11,  5420] loss: 0.00006\n",
            "[11,  5440] loss: 0.00007\n",
            "[11,  5460] loss: 0.00005\n",
            "[11,  5480] loss: 0.00021\n",
            "[11,  5500] loss: 0.00012\n",
            "[11,  5520] loss: 0.00024\n",
            "[11,  5540] loss: 0.00010\n",
            "[11,  5560] loss: 0.00001\n",
            "[11,  5580] loss: 0.00011\n",
            "[11,  5600] loss: 0.00070\n",
            "[11,  5620] loss: 0.00057\n",
            "[11,  5640] loss: 0.00045\n",
            "[11,  5660] loss: 0.00028\n",
            "[11,  5680] loss: 0.00030\n",
            "[11,  5700] loss: 0.00007\n",
            "[11,  5720] loss: 0.00067\n",
            "[11,  5740] loss: 0.00022\n",
            "[11,  5760] loss: 0.00022\n",
            "[11,  5780] loss: 0.00052\n",
            "[11,  5800] loss: 0.00005\n",
            "[11,  5820] loss: 0.00007\n",
            "[11,  5840] loss: 0.00014\n",
            "[11,  5860] loss: 0.00022\n",
            "[11,  5880] loss: 0.00006\n",
            "[11,  5900] loss: 0.00024\n",
            "[11,  5920] loss: 0.00003\n",
            "[11,  5940] loss: 0.00008\n",
            "[11,  5960] loss: 0.00028\n",
            "[11,  5980] loss: 0.00003\n",
            "[11,  6000] loss: 0.00052\n",
            "[11,  6020] loss: 0.00001\n",
            "[11,  6040] loss: 0.00018\n",
            "[11,  6060] loss: 0.00001\n",
            "[11,  6080] loss: 0.00037\n",
            "[11,  6100] loss: 0.00016\n",
            "[11,  6120] loss: 0.00007\n",
            "[11,  6140] loss: 0.00036\n",
            "[11,  6160] loss: 0.00014\n",
            "[11,  6180] loss: 0.00003\n",
            "[11,  6200] loss: 0.00001\n",
            "[11,  6220] loss: 0.00003\n",
            "[11,  6240] loss: 0.00002\n",
            "[11,  6260] loss: 0.00005\n",
            "[11,  6280] loss: 0.00003\n",
            "[11,  6300] loss: 0.00008\n",
            "[11,  6320] loss: 0.00016\n",
            "[11,  6340] loss: 0.00006\n",
            "[11,  6360] loss: 0.00004\n",
            "[11,  6380] loss: 0.00021\n",
            "[11,  6400] loss: 0.00006\n",
            "[11,  6420] loss: 0.00012\n",
            "[11,  6440] loss: 0.00016\n",
            "[11,  6460] loss: 0.00018\n",
            "[11,  6480] loss: 0.00001\n",
            "[11,  6500] loss: 0.00010\n",
            "[11,  6520] loss: 0.00003\n",
            "[11,  6540] loss: 0.00003\n",
            "[11,  6560] loss: 0.00001\n",
            "[11,  6580] loss: 0.00005\n",
            "[11,  6600] loss: 0.00007\n",
            "[11,  6620] loss: 0.00018\n",
            "[11,  6640] loss: 0.00020\n",
            "[11,  6660] loss: 0.00003\n",
            "[11,  6680] loss: 0.00007\n",
            "[11,  6700] loss: 0.00001\n",
            "[11,  6720] loss: 0.00026\n",
            "[11,  6740] loss: 0.00014\n",
            "[11,  6760] loss: 0.00003\n",
            "[11,  6780] loss: 0.00018\n",
            "[11,  6800] loss: 0.00004\n",
            "[11,  6820] loss: 0.00002\n",
            "[11,  6840] loss: 0.00024\n",
            "[11,  6860] loss: 0.00009\n",
            "[11,  6880] loss: 0.00005\n",
            "[11,  6900] loss: 0.00014\n",
            "[11,  6920] loss: 0.00006\n",
            "[11,  6940] loss: 0.00002\n",
            "[11,  6960] loss: 0.00004\n",
            "[11,  6980] loss: 0.00027\n",
            "[11,  7000] loss: 0.00002\n",
            "[11,  7020] loss: 0.00017\n",
            "[11,  7040] loss: 0.00018\n",
            "[11,  7060] loss: 0.00045\n",
            "[11,  7080] loss: 0.00004\n",
            "[11,  7100] loss: 0.00048\n",
            "[11,  7120] loss: 0.00015\n",
            "[11,  7140] loss: 0.00003\n",
            "[11,  7160] loss: 0.00014\n",
            "[11,  7180] loss: 0.00021\n",
            "[11,  7200] loss: 0.00002\n",
            "[11,  7220] loss: 0.00016\n",
            "[11,  7240] loss: 0.00077\n",
            "[11,  7260] loss: 0.00031\n",
            "[11,  7280] loss: 0.00002\n",
            "[11,  7300] loss: 0.00022\n",
            "[11,  7320] loss: 0.00011\n",
            "[11,  7340] loss: 0.00031\n",
            "[11,  7360] loss: 0.00005\n",
            "[11,  7380] loss: 0.00022\n",
            "[11,  7400] loss: 0.00032\n",
            "[11,  7420] loss: 0.00023\n",
            "[11,  7440] loss: 0.00012\n",
            "[11,  7460] loss: 0.00002\n",
            "[11,  7480] loss: 0.00005\n",
            "[11,  7500] loss: 0.00005\n",
            "Finished training for epoch 10 time taken = 694.0684204101562\n",
            "################################evaluation###########################\n",
            "\n",
            "tensor(0.0291, device='cuda:0')\n",
            "performance improved with validation loss tensor(0.0291, device='cuda:0')\n",
            "[12,    20] loss: 0.00019\n",
            "[12,    40] loss: 0.00001\n",
            "[12,    60] loss: 0.00004\n",
            "[12,    80] loss: 0.00017\n",
            "[12,   100] loss: 0.00003\n",
            "[12,   120] loss: 0.00015\n",
            "[12,   140] loss: 0.00004\n",
            "[12,   160] loss: 0.00004\n",
            "[12,   180] loss: 0.00010\n",
            "[12,   200] loss: 0.00011\n",
            "[12,   220] loss: 0.00006\n",
            "[12,   240] loss: 0.00032\n",
            "[12,   260] loss: 0.00010\n",
            "[12,   280] loss: 0.00005\n",
            "[12,   300] loss: 0.00028\n",
            "[12,   320] loss: 0.00002\n",
            "[12,   340] loss: 0.00007\n",
            "[12,   360] loss: 0.00005\n",
            "[12,   380] loss: 0.00004\n",
            "[12,   400] loss: 0.00004\n",
            "[12,   420] loss: 0.00002\n",
            "[12,   440] loss: 0.00001\n",
            "[12,   460] loss: 0.00005\n",
            "[12,   480] loss: 0.00001\n",
            "[12,   500] loss: 0.00014\n",
            "[12,   520] loss: 0.00001\n",
            "[12,   540] loss: 0.00004\n",
            "[12,   560] loss: 0.00023\n",
            "[12,   580] loss: 0.00029\n",
            "[12,   600] loss: 0.00008\n",
            "[12,   620] loss: 0.00013\n",
            "[12,   640] loss: 0.00004\n",
            "[12,   660] loss: 0.00002\n",
            "[12,   680] loss: 0.00001\n",
            "[12,   700] loss: 0.00007\n",
            "[12,   720] loss: 0.00001\n",
            "[12,   740] loss: 0.00019\n",
            "[12,   760] loss: 0.00022\n",
            "[12,   780] loss: 0.00016\n",
            "[12,   800] loss: 0.00002\n",
            "[12,   820] loss: 0.00002\n",
            "[12,   840] loss: 0.00004\n",
            "[12,   860] loss: 0.00002\n",
            "[12,   880] loss: 0.00004\n",
            "[12,   900] loss: 0.00003\n",
            "[12,   920] loss: 0.00019\n",
            "[12,   940] loss: 0.00015\n",
            "[12,   960] loss: 0.00001\n",
            "[12,   980] loss: 0.00006\n",
            "[12,  1000] loss: 0.00006\n",
            "[12,  1020] loss: 0.00000\n",
            "[12,  1040] loss: 0.00047\n",
            "[12,  1060] loss: 0.00022\n",
            "[12,  1080] loss: 0.00001\n",
            "[12,  1100] loss: 0.00004\n",
            "[12,  1120] loss: 0.00001\n",
            "[12,  1140] loss: 0.00005\n",
            "[12,  1160] loss: 0.00004\n",
            "[12,  1180] loss: 0.00001\n",
            "[12,  1200] loss: 0.00007\n",
            "[12,  1220] loss: 0.00002\n",
            "[12,  1240] loss: 0.00003\n",
            "[12,  1260] loss: 0.00005\n",
            "[12,  1280] loss: 0.00002\n",
            "[12,  1300] loss: 0.00014\n",
            "[12,  1320] loss: 0.00002\n",
            "[12,  1340] loss: 0.00006\n",
            "[12,  1360] loss: 0.00050\n",
            "[12,  1380] loss: 0.00016\n",
            "[12,  1400] loss: 0.00002\n",
            "[12,  1420] loss: 0.00007\n",
            "[12,  1440] loss: 0.00002\n",
            "[12,  1460] loss: 0.00027\n",
            "[12,  1480] loss: 0.00004\n",
            "[12,  1500] loss: 0.00002\n",
            "[12,  1520] loss: 0.00005\n",
            "[12,  1540] loss: 0.00001\n",
            "[12,  1560] loss: 0.00001\n",
            "[12,  1580] loss: 0.00006\n",
            "[12,  1600] loss: 0.00015\n",
            "[12,  1620] loss: 0.00011\n",
            "[12,  1640] loss: 0.00004\n",
            "[12,  1660] loss: 0.00009\n",
            "[12,  1680] loss: 0.00006\n",
            "[12,  1700] loss: 0.00022\n",
            "[12,  1720] loss: 0.00010\n",
            "[12,  1740] loss: 0.00004\n",
            "[12,  1760] loss: 0.00003\n",
            "[12,  1780] loss: 0.00033\n",
            "[12,  1800] loss: 0.00055\n",
            "[12,  1820] loss: 0.00014\n",
            "[12,  1840] loss: 0.00005\n",
            "[12,  1860] loss: 0.00022\n",
            "[12,  1880] loss: 0.00015\n",
            "[12,  1900] loss: 0.00004\n",
            "[12,  1920] loss: 0.00001\n",
            "[12,  1940] loss: 0.00001\n",
            "[12,  1960] loss: 0.00006\n",
            "[12,  1980] loss: 0.00001\n",
            "[12,  2000] loss: 0.00008\n",
            "[12,  2020] loss: 0.00015\n",
            "[12,  2040] loss: 0.00014\n",
            "[12,  2060] loss: 0.00004\n",
            "[12,  2080] loss: 0.00005\n",
            "[12,  2100] loss: 0.00007\n",
            "[12,  2120] loss: 0.00009\n",
            "[12,  2140] loss: 0.00003\n",
            "[12,  2160] loss: 0.00005\n",
            "[12,  2180] loss: 0.00005\n",
            "[12,  2200] loss: 0.00011\n",
            "[12,  2220] loss: 0.00000\n",
            "[12,  2240] loss: 0.00010\n",
            "[12,  2260] loss: 0.00006\n",
            "[12,  2280] loss: 0.00011\n",
            "[12,  2300] loss: 0.00002\n",
            "[12,  2320] loss: 0.00006\n",
            "[12,  2340] loss: 0.00004\n",
            "[12,  2360] loss: 0.00007\n",
            "[12,  2380] loss: 0.00003\n",
            "[12,  2400] loss: 0.00001\n",
            "[12,  2420] loss: 0.00005\n",
            "[12,  2440] loss: 0.00003\n",
            "[12,  2460] loss: 0.00001\n",
            "[12,  2480] loss: 0.00001\n",
            "[12,  2500] loss: 0.00002\n",
            "[12,  2520] loss: 0.00024\n",
            "[12,  2540] loss: 0.00009\n",
            "[12,  2560] loss: 0.00002\n",
            "[12,  2580] loss: 0.00004\n",
            "[12,  2600] loss: 0.00001\n",
            "[12,  2620] loss: 0.00001\n",
            "[12,  2640] loss: 0.00004\n",
            "[12,  2660] loss: 0.00002\n",
            "[12,  2680] loss: 0.00001\n",
            "[12,  2700] loss: 0.00006\n",
            "[12,  2720] loss: 0.00022\n",
            "[12,  2740] loss: 0.00001\n",
            "[12,  2760] loss: 0.00010\n",
            "[12,  2780] loss: 0.00002\n",
            "[12,  2800] loss: 0.00004\n",
            "[12,  2820] loss: 0.00005\n",
            "[12,  2840] loss: 0.00009\n",
            "[12,  2860] loss: 0.00004\n",
            "[12,  2880] loss: 0.00002\n",
            "[12,  2900] loss: 0.00003\n",
            "[12,  2920] loss: 0.00001\n",
            "[12,  2940] loss: 0.00017\n",
            "[12,  2960] loss: 0.00007\n",
            "[12,  2980] loss: 0.00000\n",
            "[12,  3000] loss: 0.00023\n",
            "[12,  3020] loss: 0.00001\n",
            "[12,  3040] loss: 0.00013\n",
            "[12,  3060] loss: 0.00009\n",
            "[12,  3080] loss: 0.00006\n",
            "[12,  3100] loss: 0.00029\n",
            "[12,  3120] loss: 0.00064\n",
            "[12,  3140] loss: 0.00041\n",
            "[12,  3160] loss: 0.00006\n",
            "[12,  3180] loss: 0.00014\n",
            "[12,  3200] loss: 0.00001\n",
            "[12,  3220] loss: 0.00007\n",
            "[12,  3240] loss: 0.00009\n",
            "[12,  3260] loss: 0.00002\n",
            "[12,  3280] loss: 0.00007\n",
            "[12,  3300] loss: 0.00008\n",
            "[12,  3320] loss: 0.00014\n",
            "[12,  3340] loss: 0.00008\n",
            "[12,  3360] loss: 0.00004\n",
            "[12,  3380] loss: 0.00014\n",
            "[12,  3400] loss: 0.00005\n",
            "[12,  3420] loss: 0.00023\n",
            "[12,  3440] loss: 0.00006\n",
            "[12,  3460] loss: 0.00005\n",
            "[12,  3480] loss: 0.00001\n",
            "[12,  3500] loss: 0.00005\n",
            "[12,  3520] loss: 0.00000\n",
            "[12,  3540] loss: 0.00013\n",
            "[12,  3560] loss: 0.00004\n",
            "[12,  3580] loss: 0.00001\n",
            "[12,  3600] loss: 0.00021\n",
            "[12,  3620] loss: 0.00100\n",
            "[12,  3640] loss: 0.00095\n",
            "[12,  3660] loss: 0.00026\n",
            "[12,  3680] loss: 0.00014\n",
            "[12,  3700] loss: 0.00004\n",
            "[12,  3720] loss: 0.00042\n",
            "[12,  3740] loss: 0.00039\n",
            "[12,  3760] loss: 0.00033\n",
            "[12,  3780] loss: 0.00002\n",
            "[12,  3800] loss: 0.00035\n",
            "[12,  3820] loss: 0.00055\n",
            "[12,  3840] loss: 0.00053\n",
            "[12,  3860] loss: 0.00028\n",
            "[12,  3880] loss: 0.00021\n",
            "[12,  3900] loss: 0.00015\n",
            "[12,  3920] loss: 0.00013\n",
            "[12,  3940] loss: 0.00009\n",
            "[12,  3960] loss: 0.00019\n",
            "[12,  3980] loss: 0.00003\n",
            "[12,  4000] loss: 0.00004\n",
            "[12,  4020] loss: 0.00002\n",
            "[12,  4040] loss: 0.00019\n",
            "[12,  4060] loss: 0.00021\n",
            "[12,  4080] loss: 0.00016\n",
            "[12,  4100] loss: 0.00001\n",
            "[12,  4120] loss: 0.00016\n",
            "[12,  4140] loss: 0.00019\n",
            "[12,  4160] loss: 0.00035\n",
            "[12,  4180] loss: 0.00021\n",
            "[12,  4200] loss: 0.00059\n",
            "[12,  4220] loss: 0.00017\n",
            "[12,  4240] loss: 0.00022\n",
            "[12,  4260] loss: 0.00016\n",
            "[12,  4280] loss: 0.00030\n",
            "[12,  4300] loss: 0.00031\n",
            "[12,  4320] loss: 0.00051\n",
            "[12,  4340] loss: 0.00002\n",
            "[12,  4360] loss: 0.00026\n",
            "[12,  4380] loss: 0.00005\n",
            "[12,  4400] loss: 0.00001\n",
            "[12,  4420] loss: 0.00044\n",
            "[12,  4440] loss: 0.00002\n",
            "[12,  4460] loss: 0.00041\n",
            "[12,  4480] loss: 0.00008\n",
            "[12,  4500] loss: 0.00004\n",
            "[12,  4520] loss: 0.00007\n",
            "[12,  4540] loss: 0.00006\n",
            "[12,  4560] loss: 0.00014\n",
            "[12,  4580] loss: 0.00013\n",
            "[12,  4600] loss: 0.00005\n",
            "[12,  4620] loss: 0.00001\n",
            "[12,  4640] loss: 0.00020\n",
            "[12,  4660] loss: 0.00003\n",
            "[12,  4680] loss: 0.00020\n",
            "[12,  4700] loss: 0.00007\n",
            "[12,  4720] loss: 0.00006\n",
            "[12,  4740] loss: 0.00001\n",
            "[12,  4760] loss: 0.00006\n",
            "[12,  4780] loss: 0.00011\n",
            "[12,  4800] loss: 0.00015\n",
            "[12,  4820] loss: 0.00045\n",
            "[12,  4840] loss: 0.00002\n",
            "[12,  4860] loss: 0.00008\n",
            "[12,  4880] loss: 0.00002\n",
            "[12,  4900] loss: 0.00001\n",
            "[12,  4920] loss: 0.00001\n",
            "[12,  4940] loss: 0.00002\n",
            "[12,  4960] loss: 0.00002\n",
            "[12,  4980] loss: 0.00002\n",
            "[12,  5000] loss: 0.00005\n",
            "[12,  5020] loss: 0.00018\n",
            "[12,  5040] loss: 0.00011\n",
            "[12,  5060] loss: 0.00004\n",
            "[12,  5080] loss: 0.00002\n",
            "[12,  5100] loss: 0.00004\n",
            "[12,  5120] loss: 0.00007\n",
            "[12,  5140] loss: 0.00003\n",
            "[12,  5160] loss: 0.00004\n",
            "[12,  5180] loss: 0.00003\n",
            "[12,  5200] loss: 0.00021\n",
            "[12,  5220] loss: 0.00002\n",
            "[12,  5240] loss: 0.00015\n",
            "[12,  5260] loss: 0.00029\n",
            "[12,  5280] loss: 0.00002\n",
            "[12,  5300] loss: 0.00014\n",
            "[12,  5320] loss: 0.00009\n",
            "[12,  5340] loss: 0.00018\n",
            "[12,  5360] loss: 0.00024\n",
            "[12,  5380] loss: 0.00002\n",
            "[12,  5400] loss: 0.00004\n",
            "[12,  5420] loss: 0.00019\n",
            "[12,  5440] loss: 0.00002\n",
            "[12,  5460] loss: 0.00025\n",
            "[12,  5480] loss: 0.00013\n",
            "[12,  5500] loss: 0.00012\n",
            "[12,  5520] loss: 0.00025\n",
            "[12,  5540] loss: 0.00044\n",
            "[12,  5560] loss: 0.00003\n",
            "[12,  5580] loss: 0.00005\n",
            "[12,  5600] loss: 0.00005\n",
            "[12,  5620] loss: 0.00029\n",
            "[12,  5640] loss: 0.00009\n",
            "[12,  5660] loss: 0.00001\n",
            "[12,  5680] loss: 0.00001\n",
            "[12,  5700] loss: 0.00005\n",
            "[12,  5720] loss: 0.00021\n",
            "[12,  5740] loss: 0.00007\n",
            "[12,  5760] loss: 0.00017\n",
            "[12,  5780] loss: 0.00004\n",
            "[12,  5800] loss: 0.00003\n",
            "[12,  5820] loss: 0.00009\n",
            "[12,  5840] loss: 0.00004\n",
            "[12,  5860] loss: 0.00034\n",
            "[12,  5880] loss: 0.00002\n",
            "[12,  5900] loss: 0.00009\n",
            "[12,  5920] loss: 0.00046\n",
            "[12,  5940] loss: 0.00023\n",
            "[12,  5960] loss: 0.00002\n",
            "[12,  5980] loss: 0.00023\n",
            "[12,  6000] loss: 0.00008\n",
            "[12,  6020] loss: 0.00001\n",
            "[12,  6040] loss: 0.00012\n",
            "[12,  6060] loss: 0.00025\n",
            "[12,  6080] loss: 0.00007\n",
            "[12,  6100] loss: 0.00004\n",
            "[12,  6120] loss: 0.00005\n",
            "[12,  6140] loss: 0.00022\n",
            "[12,  6160] loss: 0.00014\n",
            "[12,  6180] loss: 0.00006\n",
            "[12,  6200] loss: 0.00004\n",
            "[12,  6220] loss: 0.00003\n",
            "[12,  6240] loss: 0.00002\n",
            "[12,  6260] loss: 0.00004\n",
            "[12,  6280] loss: 0.00005\n",
            "[12,  6300] loss: 0.00013\n",
            "[12,  6320] loss: 0.00013\n",
            "[12,  6340] loss: 0.00003\n",
            "[12,  6360] loss: 0.00003\n",
            "[12,  6380] loss: 0.00004\n",
            "[12,  6400] loss: 0.00028\n",
            "[12,  6420] loss: 0.00023\n",
            "[12,  6440] loss: 0.00021\n",
            "[12,  6460] loss: 0.00041\n",
            "[12,  6480] loss: 0.00003\n",
            "[12,  6500] loss: 0.00005\n",
            "[12,  6520] loss: 0.00023\n",
            "[12,  6540] loss: 0.00008\n",
            "[12,  6560] loss: 0.00016\n",
            "[12,  6580] loss: 0.00002\n",
            "[12,  6600] loss: 0.00001\n",
            "[12,  6620] loss: 0.00011\n",
            "[12,  6640] loss: 0.00025\n",
            "[12,  6660] loss: 0.00008\n",
            "[12,  6680] loss: 0.00013\n",
            "[12,  6700] loss: 0.00009\n",
            "[12,  6720] loss: 0.00008\n",
            "[12,  6740] loss: 0.00007\n",
            "[12,  6760] loss: 0.00009\n",
            "[12,  6780] loss: 0.00040\n",
            "[12,  6800] loss: 0.00015\n",
            "[12,  6820] loss: 0.00004\n",
            "[12,  6840] loss: 0.00007\n",
            "[12,  6860] loss: 0.00002\n",
            "[12,  6880] loss: 0.00014\n",
            "[12,  6900] loss: 0.00021\n",
            "[12,  6920] loss: 0.00001\n",
            "[12,  6940] loss: 0.00012\n",
            "[12,  6960] loss: 0.00061\n",
            "[12,  6980] loss: 0.00029\n",
            "[12,  7000] loss: 0.00006\n",
            "[12,  7020] loss: 0.00014\n",
            "[12,  7040] loss: 0.00010\n",
            "[12,  7060] loss: 0.00017\n",
            "[12,  7080] loss: 0.00025\n",
            "[12,  7100] loss: 0.00004\n",
            "[12,  7120] loss: 0.00030\n",
            "[12,  7140] loss: 0.00021\n",
            "[12,  7160] loss: 0.00007\n",
            "[12,  7180] loss: 0.00006\n",
            "[12,  7200] loss: 0.00090\n",
            "[12,  7220] loss: 0.00003\n",
            "[12,  7240] loss: 0.00017\n",
            "[12,  7260] loss: 0.00024\n",
            "[12,  7280] loss: 0.00022\n",
            "[12,  7300] loss: 0.00004\n",
            "[12,  7320] loss: 0.00002\n",
            "[12,  7340] loss: 0.00019\n",
            "[12,  7360] loss: 0.00008\n",
            "[12,  7380] loss: 0.00019\n",
            "[12,  7400] loss: 0.00011\n",
            "[12,  7420] loss: 0.00019\n",
            "[12,  7440] loss: 0.00043\n",
            "[12,  7460] loss: 0.00002\n",
            "[12,  7480] loss: 0.00055\n",
            "[12,  7500] loss: 0.00079\n",
            "Finished training for epoch 11 time taken = 757.0556921958923\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 1th\n",
            "[13,    20] loss: 0.00001\n",
            "[13,    40] loss: 0.00001\n",
            "[13,    60] loss: 0.00005\n",
            "[13,    80] loss: 0.00004\n",
            "[13,   100] loss: 0.00013\n",
            "[13,   120] loss: 0.00001\n",
            "[13,   140] loss: 0.00002\n",
            "[13,   160] loss: 0.00006\n",
            "[13,   180] loss: 0.00006\n",
            "[13,   200] loss: 0.00011\n",
            "[13,   220] loss: 0.00003\n",
            "[13,   240] loss: 0.00006\n",
            "[13,   260] loss: 0.00002\n",
            "[13,   280] loss: 0.00004\n",
            "[13,   300] loss: 0.00003\n",
            "[13,   320] loss: 0.00002\n",
            "[13,   340] loss: 0.00003\n",
            "[13,   360] loss: 0.00007\n",
            "[13,   380] loss: 0.00000\n",
            "[13,   400] loss: 0.00003\n",
            "[13,   420] loss: 0.00012\n",
            "[13,   440] loss: 0.00004\n",
            "[13,   460] loss: 0.00008\n",
            "[13,   480] loss: 0.00003\n",
            "[13,   500] loss: 0.00004\n",
            "[13,   520] loss: 0.00011\n",
            "[13,   540] loss: 0.00006\n",
            "[13,   560] loss: 0.00002\n",
            "[13,   580] loss: 0.00014\n",
            "[13,   600] loss: 0.00001\n",
            "[13,   620] loss: 0.00002\n",
            "[13,   640] loss: 0.00006\n",
            "[13,   660] loss: 0.00039\n",
            "[13,   680] loss: 0.00013\n",
            "[13,   700] loss: 0.00003\n",
            "[13,   720] loss: 0.00019\n",
            "[13,   740] loss: 0.00005\n",
            "[13,   760] loss: 0.00017\n",
            "[13,   780] loss: 0.00002\n",
            "[13,   800] loss: 0.00015\n",
            "[13,   820] loss: 0.00002\n",
            "[13,   840] loss: 0.00004\n",
            "[13,   860] loss: 0.00003\n",
            "[13,   880] loss: 0.00003\n",
            "[13,   900] loss: 0.00012\n",
            "[13,   920] loss: 0.00001\n",
            "[13,   940] loss: 0.00010\n",
            "[13,   960] loss: 0.00001\n",
            "[13,   980] loss: 0.00001\n",
            "[13,  1000] loss: 0.00000\n",
            "[13,  1020] loss: 0.00004\n",
            "[13,  1040] loss: 0.00003\n",
            "[13,  1060] loss: 0.00002\n",
            "[13,  1080] loss: 0.00001\n",
            "[13,  1100] loss: 0.00004\n",
            "[13,  1120] loss: 0.00001\n",
            "[13,  1140] loss: 0.00001\n",
            "[13,  1160] loss: 0.00001\n",
            "[13,  1180] loss: 0.00001\n",
            "[13,  1200] loss: 0.00009\n",
            "[13,  1220] loss: 0.00003\n",
            "[13,  1240] loss: 0.00001\n",
            "[13,  1260] loss: 0.00031\n",
            "[13,  1280] loss: 0.00041\n",
            "[13,  1300] loss: 0.00017\n",
            "[13,  1320] loss: 0.00002\n",
            "[13,  1340] loss: 0.00019\n",
            "[13,  1360] loss: 0.00002\n",
            "[13,  1380] loss: 0.00014\n",
            "[13,  1400] loss: 0.00006\n",
            "[13,  1420] loss: 0.00001\n",
            "[13,  1440] loss: 0.00002\n",
            "[13,  1460] loss: 0.00003\n",
            "[13,  1480] loss: 0.00011\n",
            "[13,  1500] loss: 0.00001\n",
            "[13,  1520] loss: 0.00000\n",
            "[13,  1540] loss: 0.00004\n",
            "[13,  1560] loss: 0.00006\n",
            "[13,  1580] loss: 0.00001\n",
            "[13,  1600] loss: 0.00057\n",
            "[13,  1620] loss: 0.00013\n",
            "[13,  1640] loss: 0.00005\n",
            "[13,  1660] loss: 0.00001\n",
            "[13,  1680] loss: 0.00001\n",
            "[13,  1700] loss: 0.00014\n",
            "[13,  1720] loss: 0.00002\n",
            "[13,  1740] loss: 0.00003\n",
            "[13,  1760] loss: 0.00009\n",
            "[13,  1780] loss: 0.00082\n",
            "[13,  1800] loss: 0.00003\n",
            "[13,  1820] loss: 0.00013\n",
            "[13,  1840] loss: 0.00002\n",
            "[13,  1860] loss: 0.00005\n",
            "[13,  1880] loss: 0.00005\n",
            "[13,  1900] loss: 0.00018\n",
            "[13,  1920] loss: 0.00002\n",
            "[13,  1940] loss: 0.00003\n",
            "[13,  1960] loss: 0.00015\n",
            "[13,  1980] loss: 0.00020\n",
            "[13,  2000] loss: 0.00039\n",
            "[13,  2020] loss: 0.00004\n",
            "[13,  2040] loss: 0.00007\n",
            "[13,  2060] loss: 0.00010\n",
            "[13,  2080] loss: 0.00031\n",
            "[13,  2100] loss: 0.00004\n",
            "[13,  2120] loss: 0.00000\n",
            "[13,  2140] loss: 0.00001\n",
            "[13,  2160] loss: 0.00013\n",
            "[13,  2180] loss: 0.00003\n",
            "[13,  2200] loss: 0.00012\n",
            "[13,  2220] loss: 0.00006\n",
            "[13,  2240] loss: 0.00004\n",
            "[13,  2260] loss: 0.00002\n",
            "[13,  2280] loss: 0.00003\n",
            "[13,  2300] loss: 0.00001\n",
            "[13,  2320] loss: 0.00009\n",
            "[13,  2340] loss: 0.00008\n",
            "[13,  2360] loss: 0.00002\n",
            "[13,  2380] loss: 0.00001\n",
            "[13,  2400] loss: 0.00006\n",
            "[13,  2420] loss: 0.00012\n",
            "[13,  2440] loss: 0.00003\n",
            "[13,  2460] loss: 0.00030\n",
            "[13,  2480] loss: 0.00001\n",
            "[13,  2500] loss: 0.00002\n",
            "[13,  2520] loss: 0.00010\n",
            "[13,  2540] loss: 0.00014\n",
            "[13,  2560] loss: 0.00002\n",
            "[13,  2580] loss: 0.00003\n",
            "[13,  2600] loss: 0.00008\n",
            "[13,  2620] loss: 0.00001\n",
            "[13,  2640] loss: 0.00004\n",
            "[13,  2660] loss: 0.00002\n",
            "[13,  2680] loss: 0.00007\n",
            "[13,  2700] loss: 0.00015\n",
            "[13,  2720] loss: 0.00002\n",
            "[13,  2740] loss: 0.00009\n",
            "[13,  2760] loss: 0.00002\n",
            "[13,  2780] loss: 0.00038\n",
            "[13,  2800] loss: 0.00010\n",
            "[13,  2820] loss: 0.00013\n",
            "[13,  2840] loss: 0.00041\n",
            "[13,  2860] loss: 0.00022\n",
            "[13,  2880] loss: 0.00001\n",
            "[13,  2900] loss: 0.00013\n",
            "[13,  2920] loss: 0.00019\n",
            "[13,  2940] loss: 0.00032\n",
            "[13,  2960] loss: 0.00004\n",
            "[13,  2980] loss: 0.00003\n",
            "[13,  3000] loss: 0.00010\n",
            "[13,  3020] loss: 0.00001\n",
            "[13,  3040] loss: 0.00015\n",
            "[13,  3060] loss: 0.00003\n",
            "[13,  3080] loss: 0.00037\n",
            "[13,  3100] loss: 0.00028\n",
            "[13,  3120] loss: 0.00029\n",
            "[13,  3140] loss: 0.00009\n",
            "[13,  3160] loss: 0.00003\n",
            "[13,  3180] loss: 0.00004\n",
            "[13,  3200] loss: 0.00063\n",
            "[13,  3220] loss: 0.00021\n",
            "[13,  3240] loss: 0.00010\n",
            "[13,  3260] loss: 0.00012\n",
            "[13,  3280] loss: 0.00008\n",
            "[13,  3300] loss: 0.00004\n",
            "[13,  3320] loss: 0.00003\n",
            "[13,  3340] loss: 0.00001\n",
            "[13,  3360] loss: 0.00006\n",
            "[13,  3380] loss: 0.00006\n",
            "[13,  3400] loss: 0.00004\n",
            "[13,  3420] loss: 0.00008\n",
            "[13,  3440] loss: 0.00013\n",
            "[13,  3460] loss: 0.00009\n",
            "[13,  3480] loss: 0.00013\n",
            "[13,  3500] loss: 0.00005\n",
            "[13,  3520] loss: 0.00003\n",
            "[13,  3540] loss: 0.00008\n",
            "[13,  3560] loss: 0.00015\n",
            "[13,  3580] loss: 0.00003\n",
            "[13,  3600] loss: 0.00006\n",
            "[13,  3620] loss: 0.00001\n",
            "[13,  3640] loss: 0.00010\n",
            "[13,  3660] loss: 0.00004\n",
            "[13,  3680] loss: 0.00002\n",
            "[13,  3700] loss: 0.00004\n",
            "[13,  3720] loss: 0.00028\n",
            "[13,  3740] loss: 0.00001\n",
            "[13,  3760] loss: 0.00001\n",
            "[13,  3780] loss: 0.00019\n",
            "[13,  3800] loss: 0.00004\n",
            "[13,  3820] loss: 0.00000\n",
            "[13,  3840] loss: 0.00011\n",
            "[13,  3860] loss: 0.00002\n",
            "[13,  3880] loss: 0.00005\n",
            "[13,  3900] loss: 0.00011\n",
            "[13,  3920] loss: 0.00026\n",
            "[13,  3940] loss: 0.00001\n",
            "[13,  3960] loss: 0.00001\n",
            "[13,  3980] loss: 0.00006\n",
            "[13,  4000] loss: 0.00007\n",
            "[13,  4020] loss: 0.00001\n",
            "[13,  4040] loss: 0.00036\n",
            "[13,  4060] loss: 0.00008\n",
            "[13,  4080] loss: 0.00007\n",
            "[13,  4100] loss: 0.00074\n",
            "[13,  4120] loss: 0.00008\n",
            "[13,  4140] loss: 0.00002\n",
            "[13,  4160] loss: 0.00010\n",
            "[13,  4180] loss: 0.00005\n",
            "[13,  4200] loss: 0.00009\n",
            "[13,  4220] loss: 0.00014\n",
            "[13,  4240] loss: 0.00003\n",
            "[13,  4260] loss: 0.00018\n",
            "[13,  4280] loss: 0.00081\n",
            "[13,  4300] loss: 0.00001\n",
            "[13,  4320] loss: 0.00010\n",
            "[13,  4340] loss: 0.00008\n",
            "[13,  4360] loss: 0.00031\n",
            "[13,  4380] loss: 0.00003\n",
            "[13,  4400] loss: 0.00003\n",
            "[13,  4420] loss: 0.00040\n",
            "[13,  4440] loss: 0.00024\n",
            "[13,  4460] loss: 0.00007\n",
            "[13,  4480] loss: 0.00017\n",
            "[13,  4500] loss: 0.00002\n",
            "[13,  4520] loss: 0.00001\n",
            "[13,  4540] loss: 0.00028\n",
            "[13,  4560] loss: 0.00017\n",
            "[13,  4580] loss: 0.00002\n",
            "[13,  4600] loss: 0.00005\n",
            "[13,  4620] loss: 0.00004\n",
            "[13,  4640] loss: 0.00046\n",
            "[13,  4660] loss: 0.00017\n",
            "[13,  4680] loss: 0.00001\n",
            "[13,  4700] loss: 0.00004\n",
            "[13,  4720] loss: 0.00025\n",
            "[13,  4740] loss: 0.00010\n",
            "[13,  4760] loss: 0.00034\n",
            "[13,  4780] loss: 0.00019\n",
            "[13,  4800] loss: 0.00002\n",
            "[13,  4820] loss: 0.00009\n",
            "[13,  4840] loss: 0.00006\n",
            "[13,  4860] loss: 0.00004\n",
            "[13,  4880] loss: 0.00022\n",
            "[13,  4900] loss: 0.00005\n",
            "[13,  4920] loss: 0.00002\n",
            "[13,  4940] loss: 0.00005\n",
            "[13,  4960] loss: 0.00001\n",
            "[13,  4980] loss: 0.00010\n",
            "[13,  5000] loss: 0.00001\n",
            "[13,  5020] loss: 0.00004\n",
            "[13,  5040] loss: 0.00000\n",
            "[13,  5060] loss: 0.00004\n",
            "[13,  5080] loss: 0.00002\n",
            "[13,  5100] loss: 0.00017\n",
            "[13,  5120] loss: 0.00004\n",
            "[13,  5140] loss: 0.00017\n",
            "[13,  5160] loss: 0.00003\n",
            "[13,  5180] loss: 0.00004\n",
            "[13,  5200] loss: 0.00012\n",
            "[13,  5220] loss: 0.00001\n",
            "[13,  5240] loss: 0.00020\n",
            "[13,  5260] loss: 0.00009\n",
            "[13,  5280] loss: 0.00004\n",
            "[13,  5300] loss: 0.00001\n",
            "[13,  5320] loss: 0.00006\n",
            "[13,  5340] loss: 0.00005\n",
            "[13,  5360] loss: 0.00003\n",
            "[13,  5380] loss: 0.00006\n",
            "[13,  5400] loss: 0.00017\n",
            "[13,  5420] loss: 0.00003\n",
            "[13,  5440] loss: 0.00026\n",
            "[13,  5460] loss: 0.00005\n",
            "[13,  5480] loss: 0.00003\n",
            "[13,  5500] loss: 0.00008\n",
            "[13,  5520] loss: 0.00028\n",
            "[13,  5540] loss: 0.00002\n",
            "[13,  5560] loss: 0.00012\n",
            "[13,  5580] loss: 0.00007\n",
            "[13,  5600] loss: 0.00016\n",
            "[13,  5620] loss: 0.00010\n",
            "[13,  5640] loss: 0.00001\n",
            "[13,  5660] loss: 0.00014\n",
            "[13,  5680] loss: 0.00013\n",
            "[13,  5700] loss: 0.00006\n",
            "[13,  5720] loss: 0.00037\n",
            "[13,  5740] loss: 0.00003\n",
            "[13,  5760] loss: 0.00007\n",
            "[13,  5780] loss: 0.00001\n",
            "[13,  5800] loss: 0.00002\n",
            "[13,  5820] loss: 0.00003\n",
            "[13,  5840] loss: 0.00017\n",
            "[13,  5860] loss: 0.00003\n",
            "[13,  5880] loss: 0.00012\n",
            "[13,  5900] loss: 0.00014\n",
            "[13,  5920] loss: 0.00004\n",
            "[13,  5940] loss: 0.00004\n",
            "[13,  5960] loss: 0.00000\n",
            "[13,  5980] loss: 0.00003\n",
            "[13,  6000] loss: 0.00006\n",
            "[13,  6020] loss: 0.00005\n",
            "[13,  6040] loss: 0.00001\n",
            "[13,  6060] loss: 0.00002\n",
            "[13,  6080] loss: 0.00006\n",
            "[13,  6100] loss: 0.00001\n",
            "[13,  6120] loss: 0.00004\n",
            "[13,  6140] loss: 0.00003\n",
            "[13,  6160] loss: 0.00003\n",
            "[13,  6180] loss: 0.00004\n",
            "[13,  6200] loss: 0.00001\n",
            "[13,  6220] loss: 0.00001\n",
            "[13,  6240] loss: 0.00012\n",
            "[13,  6260] loss: 0.00005\n",
            "[13,  6280] loss: 0.00005\n",
            "[13,  6300] loss: 0.00004\n",
            "[13,  6320] loss: 0.00001\n",
            "[13,  6340] loss: 0.00018\n",
            "[13,  6360] loss: 0.00020\n",
            "[13,  6380] loss: 0.00006\n",
            "[13,  6400] loss: 0.00010\n",
            "[13,  6420] loss: 0.00001\n",
            "[13,  6440] loss: 0.00005\n",
            "[13,  6460] loss: 0.00030\n",
            "[13,  6480] loss: 0.00028\n",
            "[13,  6500] loss: 0.00003\n",
            "[13,  6520] loss: 0.00002\n",
            "[13,  6540] loss: 0.00002\n",
            "[13,  6560] loss: 0.00002\n",
            "[13,  6580] loss: 0.00001\n",
            "[13,  6600] loss: 0.00024\n",
            "[13,  6620] loss: 0.00000\n",
            "[13,  6640] loss: 0.00003\n",
            "[13,  6660] loss: 0.00008\n",
            "[13,  6680] loss: 0.00025\n",
            "[13,  6700] loss: 0.00003\n",
            "[13,  6720] loss: 0.00009\n",
            "[13,  6740] loss: 0.00009\n",
            "[13,  6760] loss: 0.00002\n",
            "[13,  6780] loss: 0.00029\n",
            "[13,  6800] loss: 0.00006\n",
            "[13,  6820] loss: 0.00003\n",
            "[13,  6840] loss: 0.00014\n",
            "[13,  6860] loss: 0.00002\n",
            "[13,  6880] loss: 0.00004\n",
            "[13,  6900] loss: 0.00004\n",
            "[13,  6920] loss: 0.00006\n",
            "[13,  6940] loss: 0.00005\n",
            "[13,  6960] loss: 0.00015\n",
            "[13,  6980] loss: 0.00010\n",
            "[13,  7000] loss: 0.00005\n",
            "[13,  7020] loss: 0.00018\n",
            "[13,  7040] loss: 0.00001\n",
            "[13,  7060] loss: 0.00005\n",
            "[13,  7080] loss: 0.00003\n",
            "[13,  7100] loss: 0.00009\n",
            "[13,  7120] loss: 0.00022\n",
            "[13,  7140] loss: 0.00023\n",
            "[13,  7160] loss: 0.00009\n",
            "[13,  7180] loss: 0.00010\n",
            "[13,  7200] loss: 0.00001\n",
            "[13,  7220] loss: 0.00051\n",
            "[13,  7240] loss: 0.00059\n",
            "[13,  7260] loss: 0.00004\n",
            "[13,  7280] loss: 0.00007\n",
            "[13,  7300] loss: 0.00037\n",
            "[13,  7320] loss: 0.00013\n",
            "[13,  7340] loss: 0.00023\n",
            "[13,  7360] loss: 0.00017\n",
            "[13,  7380] loss: 0.00009\n",
            "[13,  7400] loss: 0.00018\n",
            "[13,  7420] loss: 0.00014\n",
            "[13,  7440] loss: 0.00001\n",
            "[13,  7460] loss: 0.00006\n",
            "[13,  7480] loss: 0.00042\n",
            "[13,  7500] loss: 0.00017\n",
            "Finished training for epoch 12 time taken = 819.6900115013123\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 2th\n",
            "[14,    20] loss: 0.00001\n",
            "[14,    40] loss: 0.00015\n",
            "[14,    60] loss: 0.00004\n",
            "[14,    80] loss: 0.00008\n",
            "[14,   100] loss: 0.00009\n",
            "[14,   120] loss: 0.00001\n",
            "[14,   140] loss: 0.00008\n",
            "[14,   160] loss: 0.00001\n",
            "[14,   180] loss: 0.00009\n",
            "[14,   200] loss: 0.00006\n",
            "[14,   220] loss: 0.00001\n",
            "[14,   240] loss: 0.00029\n",
            "[14,   260] loss: 0.00010\n",
            "[14,   280] loss: 0.00001\n",
            "[14,   300] loss: 0.00006\n",
            "[14,   320] loss: 0.00006\n",
            "[14,   340] loss: 0.00000\n",
            "[14,   360] loss: 0.00001\n",
            "[14,   380] loss: 0.00001\n",
            "[14,   400] loss: 0.00003\n",
            "[14,   420] loss: 0.00001\n",
            "[14,   440] loss: 0.00002\n",
            "[14,   460] loss: 0.00051\n",
            "[14,   480] loss: 0.00024\n",
            "[14,   500] loss: 0.00006\n",
            "[14,   520] loss: 0.00004\n",
            "[14,   540] loss: 0.00084\n",
            "[14,   560] loss: 0.00018\n",
            "[14,   580] loss: 0.00009\n",
            "[14,   600] loss: 0.00016\n",
            "[14,   620] loss: 0.00012\n",
            "[14,   640] loss: 0.00002\n",
            "[14,   660] loss: 0.00001\n",
            "[14,   680] loss: 0.00015\n",
            "[14,   700] loss: 0.00005\n",
            "[14,   720] loss: 0.00001\n",
            "[14,   740] loss: 0.00004\n",
            "[14,   760] loss: 0.00006\n",
            "[14,   780] loss: 0.00011\n",
            "[14,   800] loss: 0.00002\n",
            "[14,   820] loss: 0.00001\n",
            "[14,   840] loss: 0.00010\n",
            "[14,   860] loss: 0.00004\n",
            "[14,   880] loss: 0.00004\n",
            "[14,   900] loss: 0.00004\n",
            "[14,   920] loss: 0.00018\n",
            "[14,   940] loss: 0.00001\n",
            "[14,   960] loss: 0.00014\n",
            "[14,   980] loss: 0.00018\n",
            "[14,  1000] loss: 0.00004\n",
            "[14,  1020] loss: 0.00005\n",
            "[14,  1040] loss: 0.00013\n",
            "[14,  1060] loss: 0.00013\n",
            "[14,  1080] loss: 0.00001\n",
            "[14,  1100] loss: 0.00034\n",
            "[14,  1120] loss: 0.00018\n",
            "[14,  1140] loss: 0.00015\n",
            "[14,  1160] loss: 0.00038\n",
            "[14,  1180] loss: 0.00001\n",
            "[14,  1200] loss: 0.00014\n",
            "[14,  1220] loss: 0.00001\n",
            "[14,  1240] loss: 0.00010\n",
            "[14,  1260] loss: 0.00011\n",
            "[14,  1280] loss: 0.00003\n",
            "[14,  1300] loss: 0.00011\n",
            "[14,  1320] loss: 0.00002\n",
            "[14,  1340] loss: 0.00007\n",
            "[14,  1360] loss: 0.00006\n",
            "[14,  1380] loss: 0.00001\n",
            "[14,  1400] loss: 0.00034\n",
            "[14,  1420] loss: 0.00001\n",
            "[14,  1440] loss: 0.00002\n",
            "[14,  1460] loss: 0.00008\n",
            "[14,  1480] loss: 0.00002\n",
            "[14,  1500] loss: 0.00005\n",
            "[14,  1520] loss: 0.00002\n",
            "[14,  1540] loss: 0.00005\n",
            "[14,  1560] loss: 0.00002\n",
            "[14,  1580] loss: 0.00002\n",
            "[14,  1600] loss: 0.00004\n",
            "[14,  1620] loss: 0.00003\n",
            "[14,  1640] loss: 0.00001\n",
            "[14,  1660] loss: 0.00001\n",
            "[14,  1680] loss: 0.00010\n",
            "[14,  1700] loss: 0.00002\n",
            "[14,  1720] loss: 0.00006\n",
            "[14,  1740] loss: 0.00003\n",
            "[14,  1760] loss: 0.00004\n",
            "[14,  1780] loss: 0.00026\n",
            "[14,  1800] loss: 0.00007\n",
            "[14,  1820] loss: 0.00028\n",
            "[14,  1840] loss: 0.00007\n",
            "[14,  1860] loss: 0.00003\n",
            "[14,  1880] loss: 0.00005\n",
            "[14,  1900] loss: 0.00001\n",
            "[14,  1920] loss: 0.00000\n",
            "[14,  1940] loss: 0.00016\n",
            "[14,  1960] loss: 0.00008\n",
            "[14,  1980] loss: 0.00013\n",
            "[14,  2000] loss: 0.00007\n",
            "[14,  2020] loss: 0.00003\n",
            "[14,  2040] loss: 0.00005\n",
            "[14,  2060] loss: 0.00008\n",
            "[14,  2080] loss: 0.00003\n",
            "[14,  2100] loss: 0.00001\n",
            "[14,  2120] loss: 0.00001\n",
            "[14,  2140] loss: 0.00003\n",
            "[14,  2160] loss: 0.00002\n",
            "[14,  2180] loss: 0.00001\n",
            "[14,  2200] loss: 0.00014\n",
            "[14,  2220] loss: 0.00002\n",
            "[14,  2240] loss: 0.00000\n",
            "[14,  2260] loss: 0.00031\n",
            "[14,  2280] loss: 0.00000\n",
            "[14,  2300] loss: 0.00004\n",
            "[14,  2320] loss: 0.00001\n",
            "[14,  2340] loss: 0.00003\n",
            "[14,  2360] loss: 0.00006\n",
            "[14,  2380] loss: 0.00042\n",
            "[14,  2400] loss: 0.00008\n",
            "[14,  2420] loss: 0.00001\n",
            "[14,  2440] loss: 0.00008\n",
            "[14,  2460] loss: 0.00002\n",
            "[14,  2480] loss: 0.00010\n",
            "[14,  2500] loss: 0.00019\n",
            "[14,  2520] loss: 0.00018\n",
            "[14,  2540] loss: 0.00011\n",
            "[14,  2560] loss: 0.00002\n",
            "[14,  2580] loss: 0.00004\n",
            "[14,  2600] loss: 0.00019\n",
            "[14,  2620] loss: 0.00003\n",
            "[14,  2640] loss: 0.00006\n",
            "[14,  2660] loss: 0.00005\n",
            "[14,  2680] loss: 0.00001\n",
            "[14,  2700] loss: 0.00003\n",
            "[14,  2720] loss: 0.00003\n",
            "[14,  2740] loss: 0.00004\n",
            "[14,  2760] loss: 0.00004\n",
            "[14,  2780] loss: 0.00009\n",
            "[14,  2800] loss: 0.00001\n",
            "[14,  2820] loss: 0.00007\n",
            "[14,  2840] loss: 0.00001\n",
            "[14,  2860] loss: 0.00003\n",
            "[14,  2880] loss: 0.00033\n",
            "[14,  2900] loss: 0.00002\n",
            "[14,  2920] loss: 0.00020\n",
            "[14,  2940] loss: 0.00007\n",
            "[14,  2960] loss: 0.00024\n",
            "[14,  2980] loss: 0.00007\n",
            "[14,  3000] loss: 0.00043\n",
            "[14,  3020] loss: 0.00000\n",
            "[14,  3040] loss: 0.00015\n",
            "[14,  3060] loss: 0.00001\n",
            "[14,  3080] loss: 0.00015\n",
            "[14,  3100] loss: 0.00005\n",
            "[14,  3120] loss: 0.00007\n",
            "[14,  3140] loss: 0.00007\n",
            "[14,  3160] loss: 0.00014\n",
            "[14,  3180] loss: 0.00004\n",
            "[14,  3200] loss: 0.00003\n",
            "[14,  3220] loss: 0.00014\n",
            "[14,  3240] loss: 0.00005\n",
            "[14,  3260] loss: 0.00001\n",
            "[14,  3280] loss: 0.00031\n",
            "[14,  3300] loss: 0.00006\n",
            "[14,  3320] loss: 0.00004\n",
            "[14,  3340] loss: 0.00011\n",
            "[14,  3360] loss: 0.00001\n",
            "[14,  3380] loss: 0.00002\n",
            "[14,  3400] loss: 0.00033\n",
            "[14,  3420] loss: 0.00013\n",
            "[14,  3440] loss: 0.00019\n",
            "[14,  3460] loss: 0.00020\n",
            "[14,  3480] loss: 0.00012\n",
            "[14,  3500] loss: 0.00001\n",
            "[14,  3520] loss: 0.00006\n",
            "[14,  3540] loss: 0.00004\n",
            "[14,  3560] loss: 0.00004\n",
            "[14,  3580] loss: 0.00002\n",
            "[14,  3600] loss: 0.00013\n",
            "[14,  3620] loss: 0.00000\n",
            "[14,  3640] loss: 0.00016\n",
            "[14,  3660] loss: 0.00002\n",
            "[14,  3680] loss: 0.00007\n",
            "[14,  3700] loss: 0.00001\n",
            "[14,  3720] loss: 0.00004\n",
            "[14,  3740] loss: 0.00000\n",
            "[14,  3760] loss: 0.00018\n",
            "[14,  3780] loss: 0.00001\n",
            "[14,  3800] loss: 0.00002\n",
            "[14,  3820] loss: 0.00030\n",
            "[14,  3840] loss: 0.00007\n",
            "[14,  3860] loss: 0.00006\n",
            "[14,  3880] loss: 0.00002\n",
            "[14,  3900] loss: 0.00001\n",
            "[14,  3920] loss: 0.00002\n",
            "[14,  3940] loss: 0.00015\n",
            "[14,  3960] loss: 0.00002\n",
            "[14,  3980] loss: 0.00003\n",
            "[14,  4000] loss: 0.00003\n",
            "[14,  4020] loss: 0.00008\n",
            "[14,  4040] loss: 0.00012\n",
            "[14,  4060] loss: 0.00002\n",
            "[14,  4080] loss: 0.00017\n",
            "[14,  4100] loss: 0.00006\n",
            "[14,  4120] loss: 0.00003\n",
            "[14,  4140] loss: 0.00030\n",
            "[14,  4160] loss: 0.00038\n",
            "[14,  4180] loss: 0.00021\n",
            "[14,  4200] loss: 0.00011\n",
            "[14,  4220] loss: 0.00008\n",
            "[14,  4240] loss: 0.00037\n",
            "[14,  4260] loss: 0.00010\n",
            "[14,  4280] loss: 0.00002\n",
            "[14,  4300] loss: 0.00025\n",
            "[14,  4320] loss: 0.00005\n",
            "[14,  4340] loss: 0.00020\n",
            "[14,  4360] loss: 0.00010\n",
            "[14,  4380] loss: 0.00003\n",
            "[14,  4400] loss: 0.00002\n",
            "[14,  4420] loss: 0.00042\n",
            "[14,  4440] loss: 0.00007\n",
            "[14,  4460] loss: 0.00003\n",
            "[14,  4480] loss: 0.00007\n",
            "[14,  4500] loss: 0.00001\n",
            "[14,  4520] loss: 0.00000\n",
            "[14,  4540] loss: 0.00016\n",
            "[14,  4560] loss: 0.00009\n",
            "[14,  4580] loss: 0.00007\n",
            "[14,  4600] loss: 0.00000\n",
            "[14,  4620] loss: 0.00028\n",
            "[14,  4640] loss: 0.00004\n",
            "[14,  4660] loss: 0.00017\n",
            "[14,  4680] loss: 0.00015\n",
            "[14,  4700] loss: 0.00041\n",
            "[14,  4720] loss: 0.00021\n",
            "[14,  4740] loss: 0.00010\n",
            "[14,  4760] loss: 0.00005\n",
            "[14,  4780] loss: 0.00029\n",
            "[14,  4800] loss: 0.00003\n",
            "[14,  4820] loss: 0.00021\n",
            "[14,  4840] loss: 0.00004\n",
            "[14,  4860] loss: 0.00000\n",
            "[14,  4880] loss: 0.00001\n",
            "[14,  4900] loss: 0.00025\n",
            "[14,  4920] loss: 0.00006\n",
            "[14,  4940] loss: 0.00009\n",
            "[14,  4960] loss: 0.00021\n",
            "[14,  4980] loss: 0.00026\n",
            "[14,  5000] loss: 0.00050\n",
            "[14,  5020] loss: 0.00002\n",
            "[14,  5040] loss: 0.00010\n",
            "[14,  5060] loss: 0.00015\n",
            "[14,  5080] loss: 0.00014\n",
            "[14,  5100] loss: 0.00019\n",
            "[14,  5120] loss: 0.00001\n",
            "[14,  5140] loss: 0.00002\n",
            "[14,  5160] loss: 0.00002\n",
            "[14,  5180] loss: 0.00007\n",
            "[14,  5200] loss: 0.00005\n",
            "[14,  5220] loss: 0.00005\n",
            "[14,  5240] loss: 0.00002\n",
            "[14,  5260] loss: 0.00003\n",
            "[14,  5280] loss: 0.00006\n",
            "[14,  5300] loss: 0.00002\n",
            "[14,  5320] loss: 0.00004\n",
            "[14,  5340] loss: 0.00001\n",
            "[14,  5360] loss: 0.00002\n",
            "[14,  5380] loss: 0.00003\n",
            "[14,  5400] loss: 0.00001\n",
            "[14,  5420] loss: 0.00039\n",
            "[14,  5440] loss: 0.00009\n",
            "[14,  5460] loss: 0.00010\n",
            "[14,  5480] loss: 0.00012\n",
            "[14,  5500] loss: 0.00012\n",
            "[14,  5520] loss: 0.00001\n",
            "[14,  5540] loss: 0.00001\n",
            "[14,  5560] loss: 0.00011\n",
            "[14,  5580] loss: 0.00005\n",
            "[14,  5600] loss: 0.00004\n",
            "[14,  5620] loss: 0.00001\n",
            "[14,  5640] loss: 0.00001\n",
            "[14,  5660] loss: 0.00005\n",
            "[14,  5680] loss: 0.00015\n",
            "[14,  5700] loss: 0.00013\n",
            "[14,  5720] loss: 0.00001\n",
            "[14,  5740] loss: 0.00012\n",
            "[14,  5760] loss: 0.00009\n",
            "[14,  5780] loss: 0.00018\n",
            "[14,  5800] loss: 0.00006\n",
            "[14,  5820] loss: 0.00022\n",
            "[14,  5840] loss: 0.00013\n",
            "[14,  5860] loss: 0.00003\n",
            "[14,  5880] loss: 0.00001\n",
            "[14,  5900] loss: 0.00000\n",
            "[14,  5920] loss: 0.00007\n",
            "[14,  5940] loss: 0.00021\n",
            "[14,  5960] loss: 0.00006\n",
            "[14,  5980] loss: 0.00005\n",
            "[14,  6000] loss: 0.00002\n",
            "[14,  6020] loss: 0.00001\n",
            "[14,  6040] loss: 0.00037\n",
            "[14,  6060] loss: 0.00037\n",
            "[14,  6080] loss: 0.00007\n",
            "[14,  6100] loss: 0.00004\n",
            "[14,  6120] loss: 0.00027\n",
            "[14,  6140] loss: 0.00006\n",
            "[14,  6160] loss: 0.00008\n",
            "[14,  6180] loss: 0.00001\n",
            "[14,  6200] loss: 0.00033\n",
            "[14,  6220] loss: 0.00020\n",
            "[14,  6240] loss: 0.00001\n",
            "[14,  6260] loss: 0.00013\n",
            "[14,  6280] loss: 0.00005\n",
            "[14,  6300] loss: 0.00024\n",
            "[14,  6320] loss: 0.00027\n",
            "[14,  6340] loss: 0.00009\n",
            "[14,  6360] loss: 0.00001\n",
            "[14,  6380] loss: 0.00017\n",
            "[14,  6400] loss: 0.00001\n",
            "[14,  6420] loss: 0.00010\n",
            "[14,  6440] loss: 0.00009\n",
            "[14,  6460] loss: 0.00002\n",
            "[14,  6480] loss: 0.00028\n",
            "[14,  6500] loss: 0.00004\n",
            "[14,  6520] loss: 0.00006\n",
            "[14,  6540] loss: 0.00012\n",
            "[14,  6560] loss: 0.00012\n",
            "[14,  6580] loss: 0.00013\n",
            "[14,  6600] loss: 0.00002\n",
            "[14,  6620] loss: 0.00009\n",
            "[14,  6640] loss: 0.00000\n",
            "[14,  6660] loss: 0.00002\n",
            "[14,  6680] loss: 0.00006\n",
            "[14,  6700] loss: 0.00010\n",
            "[14,  6720] loss: 0.00007\n",
            "[14,  6740] loss: 0.00001\n",
            "[14,  6760] loss: 0.00009\n",
            "[14,  6780] loss: 0.00007\n",
            "[14,  6800] loss: 0.00001\n",
            "[14,  6820] loss: 0.00002\n",
            "[14,  6840] loss: 0.00002\n",
            "[14,  6860] loss: 0.00003\n",
            "[14,  6880] loss: 0.00004\n",
            "[14,  6900] loss: 0.00002\n",
            "[14,  6920] loss: 0.00001\n",
            "[14,  6940] loss: 0.00019\n",
            "[14,  6960] loss: 0.00001\n",
            "[14,  6980] loss: 0.00005\n",
            "[14,  7000] loss: 0.00006\n",
            "[14,  7020] loss: 0.00001\n",
            "[14,  7040] loss: 0.00002\n",
            "[14,  7060] loss: 0.00001\n",
            "[14,  7080] loss: 0.00002\n",
            "[14,  7100] loss: 0.00001\n",
            "[14,  7120] loss: 0.00000\n",
            "[14,  7140] loss: 0.00001\n",
            "[14,  7160] loss: 0.00000\n",
            "[14,  7180] loss: 0.00004\n",
            "[14,  7200] loss: 0.00004\n",
            "[14,  7220] loss: 0.00001\n",
            "[14,  7240] loss: 0.00002\n",
            "[14,  7260] loss: 0.00000\n",
            "[14,  7280] loss: 0.00000\n",
            "[14,  7300] loss: 0.00004\n",
            "[14,  7320] loss: 0.00001\n",
            "[14,  7340] loss: 0.00003\n",
            "[14,  7360] loss: 0.00001\n",
            "[14,  7380] loss: 0.00006\n",
            "[14,  7400] loss: 0.00000\n",
            "[14,  7420] loss: 0.00004\n",
            "[14,  7440] loss: 0.00006\n",
            "[14,  7460] loss: 0.00003\n",
            "[14,  7480] loss: 0.00034\n",
            "[14,  7500] loss: 0.00004\n",
            "Finished training for epoch 13 time taken = 882.2553853988647\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 3th\n",
            "[15,    20] loss: 0.00001\n",
            "[15,    40] loss: 0.00018\n",
            "[15,    60] loss: 0.00001\n",
            "[15,    80] loss: 0.00006\n",
            "[15,   100] loss: 0.00001\n",
            "[15,   120] loss: 0.00003\n",
            "[15,   140] loss: 0.00004\n",
            "[15,   160] loss: 0.00012\n",
            "[15,   180] loss: 0.00000\n",
            "[15,   200] loss: 0.00030\n",
            "[15,   220] loss: 0.00007\n",
            "[15,   240] loss: 0.00002\n",
            "[15,   260] loss: 0.00005\n",
            "[15,   280] loss: 0.00002\n",
            "[15,   300] loss: 0.00019\n",
            "[15,   320] loss: 0.00017\n",
            "[15,   340] loss: 0.00001\n",
            "[15,   360] loss: 0.00004\n",
            "[15,   380] loss: 0.00004\n",
            "[15,   400] loss: 0.00008\n",
            "[15,   420] loss: 0.00022\n",
            "[15,   440] loss: 0.00005\n",
            "[15,   460] loss: 0.00054\n",
            "[15,   480] loss: 0.00003\n",
            "[15,   500] loss: 0.00000\n",
            "[15,   520] loss: 0.00000\n",
            "[15,   540] loss: 0.00003\n",
            "[15,   560] loss: 0.00001\n",
            "[15,   580] loss: 0.00027\n",
            "[15,   600] loss: 0.00005\n",
            "[15,   620] loss: 0.00001\n",
            "[15,   640] loss: 0.00003\n",
            "[15,   660] loss: 0.00006\n",
            "[15,   680] loss: 0.00001\n",
            "[15,   700] loss: 0.00007\n",
            "[15,   720] loss: 0.00004\n",
            "[15,   740] loss: 0.00001\n",
            "[15,   760] loss: 0.00001\n",
            "[15,   780] loss: 0.00002\n",
            "[15,   800] loss: 0.00001\n",
            "[15,   820] loss: 0.00001\n",
            "[15,   840] loss: 0.00000\n",
            "[15,   860] loss: 0.00001\n",
            "[15,   880] loss: 0.00009\n",
            "[15,   900] loss: 0.00009\n",
            "[15,   920] loss: 0.00004\n",
            "[15,   940] loss: 0.00002\n",
            "[15,   960] loss: 0.00014\n",
            "[15,   980] loss: 0.00020\n",
            "[15,  1000] loss: 0.00001\n",
            "[15,  1020] loss: 0.00001\n",
            "[15,  1040] loss: 0.00017\n",
            "[15,  1060] loss: 0.00001\n",
            "[15,  1080] loss: 0.00021\n",
            "[15,  1100] loss: 0.00003\n",
            "[15,  1120] loss: 0.00006\n",
            "[15,  1140] loss: 0.00002\n",
            "[15,  1160] loss: 0.00005\n",
            "[15,  1180] loss: 0.00003\n",
            "[15,  1200] loss: 0.00000\n",
            "[15,  1220] loss: 0.00001\n",
            "[15,  1240] loss: 0.00009\n",
            "[15,  1260] loss: 0.00009\n",
            "[15,  1280] loss: 0.00003\n",
            "[15,  1300] loss: 0.00002\n",
            "[15,  1320] loss: 0.00004\n",
            "[15,  1340] loss: 0.00006\n",
            "[15,  1360] loss: 0.00001\n",
            "[15,  1380] loss: 0.00001\n",
            "[15,  1400] loss: 0.00003\n",
            "[15,  1420] loss: 0.00001\n",
            "[15,  1440] loss: 0.00005\n",
            "[15,  1460] loss: 0.00000\n",
            "[15,  1480] loss: 0.00008\n",
            "[15,  1500] loss: 0.00002\n",
            "[15,  1520] loss: 0.00000\n",
            "[15,  1540] loss: 0.00001\n",
            "[15,  1560] loss: 0.00010\n",
            "[15,  1580] loss: 0.00000\n",
            "[15,  1600] loss: 0.00003\n",
            "[15,  1620] loss: 0.00020\n",
            "[15,  1640] loss: 0.00021\n",
            "[15,  1660] loss: 0.00005\n",
            "[15,  1680] loss: 0.00004\n",
            "[15,  1700] loss: 0.00000\n",
            "[15,  1720] loss: 0.00001\n",
            "[15,  1740] loss: 0.00001\n",
            "[15,  1760] loss: 0.00017\n",
            "[15,  1780] loss: 0.00001\n",
            "[15,  1800] loss: 0.00007\n",
            "[15,  1820] loss: 0.00001\n",
            "[15,  1840] loss: 0.00004\n",
            "[15,  1860] loss: 0.00005\n",
            "[15,  1880] loss: 0.00006\n",
            "[15,  1900] loss: 0.00001\n",
            "[15,  1920] loss: 0.00004\n",
            "[15,  1940] loss: 0.00002\n",
            "[15,  1960] loss: 0.00004\n",
            "[15,  1980] loss: 0.00002\n",
            "[15,  2000] loss: 0.00024\n",
            "[15,  2020] loss: 0.00002\n",
            "[15,  2040] loss: 0.00003\n",
            "[15,  2060] loss: 0.00003\n",
            "[15,  2080] loss: 0.00003\n",
            "[15,  2100] loss: 0.00002\n",
            "[15,  2120] loss: 0.00001\n",
            "[15,  2140] loss: 0.00004\n",
            "[15,  2160] loss: 0.00022\n",
            "[15,  2180] loss: 0.00010\n",
            "[15,  2200] loss: 0.00001\n",
            "[15,  2220] loss: 0.00000\n",
            "[15,  2240] loss: 0.00011\n",
            "[15,  2260] loss: 0.00006\n",
            "[15,  2280] loss: 0.00004\n",
            "[15,  2300] loss: 0.00000\n",
            "[15,  2320] loss: 0.00002\n",
            "[15,  2340] loss: 0.00001\n",
            "[15,  2360] loss: 0.00026\n",
            "[15,  2380] loss: 0.00016\n",
            "[15,  2400] loss: 0.00002\n",
            "[15,  2420] loss: 0.00001\n",
            "[15,  2440] loss: 0.00010\n",
            "[15,  2460] loss: 0.00005\n",
            "[15,  2480] loss: 0.00001\n",
            "[15,  2500] loss: 0.00008\n",
            "[15,  2520] loss: 0.00004\n",
            "[15,  2540] loss: 0.00014\n",
            "[15,  2560] loss: 0.00001\n",
            "[15,  2580] loss: 0.00001\n",
            "[15,  2600] loss: 0.00001\n",
            "[15,  2620] loss: 0.00001\n",
            "[15,  2640] loss: 0.00005\n",
            "[15,  2660] loss: 0.00010\n",
            "[15,  2680] loss: 0.00004\n",
            "[15,  2700] loss: 0.00002\n",
            "[15,  2720] loss: 0.00001\n",
            "[15,  2740] loss: 0.00000\n",
            "[15,  2760] loss: 0.00017\n",
            "[15,  2780] loss: 0.00006\n",
            "[15,  2800] loss: 0.00002\n",
            "[15,  2820] loss: 0.00001\n",
            "[15,  2840] loss: 0.00002\n",
            "[15,  2860] loss: 0.00020\n",
            "[15,  2880] loss: 0.00005\n",
            "[15,  2900] loss: 0.00009\n",
            "[15,  2920] loss: 0.00012\n",
            "[15,  2940] loss: 0.00012\n",
            "[15,  2960] loss: 0.00019\n",
            "[15,  2980] loss: 0.00008\n",
            "[15,  3000] loss: 0.00002\n",
            "[15,  3020] loss: 0.00012\n",
            "[15,  3040] loss: 0.00004\n",
            "[15,  3060] loss: 0.00006\n",
            "[15,  3080] loss: 0.00007\n",
            "[15,  3100] loss: 0.00000\n",
            "[15,  3120] loss: 0.00001\n",
            "[15,  3140] loss: 0.00002\n",
            "[15,  3160] loss: 0.00008\n",
            "[15,  3180] loss: 0.00001\n",
            "[15,  3200] loss: 0.00004\n",
            "[15,  3220] loss: 0.00004\n",
            "[15,  3240] loss: 0.00003\n",
            "[15,  3260] loss: 0.00004\n",
            "[15,  3280] loss: 0.00001\n",
            "[15,  3300] loss: 0.00001\n",
            "[15,  3320] loss: 0.00001\n",
            "[15,  3340] loss: 0.00001\n",
            "[15,  3360] loss: 0.00001\n",
            "[15,  3380] loss: 0.00000\n",
            "[15,  3400] loss: 0.00005\n",
            "[15,  3420] loss: 0.00002\n",
            "[15,  3440] loss: 0.00002\n",
            "[15,  3460] loss: 0.00001\n",
            "[15,  3480] loss: 0.00003\n",
            "[15,  3500] loss: 0.00001\n",
            "[15,  3520] loss: 0.00009\n",
            "[15,  3540] loss: 0.00002\n",
            "[15,  3560] loss: 0.00019\n",
            "[15,  3580] loss: 0.00001\n",
            "[15,  3600] loss: 0.00000\n",
            "[15,  3620] loss: 0.00008\n",
            "[15,  3640] loss: 0.00005\n",
            "[15,  3660] loss: 0.00002\n",
            "[15,  3680] loss: 0.00001\n",
            "[15,  3700] loss: 0.00032\n",
            "[15,  3720] loss: 0.00009\n",
            "[15,  3740] loss: 0.00008\n",
            "[15,  3760] loss: 0.00001\n",
            "[15,  3780] loss: 0.00001\n",
            "[15,  3800] loss: 0.00002\n",
            "[15,  3820] loss: 0.00013\n",
            "[15,  3840] loss: 0.00004\n",
            "[15,  3860] loss: 0.00001\n",
            "[15,  3880] loss: 0.00001\n",
            "[15,  3900] loss: 0.00013\n",
            "[15,  3920] loss: 0.00000\n",
            "[15,  3940] loss: 0.00022\n",
            "[15,  3960] loss: 0.00006\n",
            "[15,  3980] loss: 0.00011\n",
            "[15,  4000] loss: 0.00005\n",
            "[15,  4020] loss: 0.00001\n",
            "[15,  4040] loss: 0.00005\n",
            "[15,  4060] loss: 0.00005\n",
            "[15,  4080] loss: 0.00021\n",
            "[15,  4100] loss: 0.00009\n",
            "[15,  4120] loss: 0.00007\n",
            "[15,  4140] loss: 0.00008\n",
            "[15,  4160] loss: 0.00005\n",
            "[15,  4180] loss: 0.00005\n",
            "[15,  4200] loss: 0.00012\n",
            "[15,  4220] loss: 0.00018\n",
            "[15,  4240] loss: 0.00020\n",
            "[15,  4260] loss: 0.00002\n",
            "[15,  4280] loss: 0.00002\n",
            "[15,  4300] loss: 0.00013\n",
            "[15,  4320] loss: 0.00010\n",
            "[15,  4340] loss: 0.00007\n",
            "[15,  4360] loss: 0.00001\n",
            "[15,  4380] loss: 0.00000\n",
            "[15,  4400] loss: 0.00020\n",
            "[15,  4420] loss: 0.00003\n",
            "[15,  4440] loss: 0.00003\n",
            "[15,  4460] loss: 0.00014\n",
            "[15,  4480] loss: 0.00010\n",
            "[15,  4500] loss: 0.00006\n",
            "[15,  4520] loss: 0.00004\n",
            "[15,  4540] loss: 0.00083\n",
            "[15,  4560] loss: 0.00002\n",
            "[15,  4580] loss: 0.00053\n",
            "[15,  4600] loss: 0.00010\n",
            "[15,  4620] loss: 0.00073\n",
            "[15,  4640] loss: 0.00000\n",
            "[15,  4660] loss: 0.00020\n",
            "[15,  4680] loss: 0.00011\n",
            "[15,  4700] loss: 0.00005\n",
            "[15,  4720] loss: 0.00025\n",
            "[15,  4740] loss: 0.00028\n",
            "[15,  4760] loss: 0.00019\n",
            "[15,  4780] loss: 0.00001\n",
            "[15,  4800] loss: 0.00086\n",
            "[15,  4820] loss: 0.00041\n",
            "[15,  4840] loss: 0.00009\n",
            "[15,  4860] loss: 0.00011\n",
            "[15,  4880] loss: 0.00009\n",
            "[15,  4900] loss: 0.00018\n",
            "[15,  4920] loss: 0.00001\n",
            "[15,  4940] loss: 0.00013\n",
            "[15,  4960] loss: 0.00003\n",
            "[15,  4980] loss: 0.00013\n",
            "[15,  5000] loss: 0.00003\n",
            "[15,  5020] loss: 0.00030\n",
            "[15,  5040] loss: 0.00005\n",
            "[15,  5060] loss: 0.00001\n",
            "[15,  5080] loss: 0.00008\n",
            "[15,  5100] loss: 0.00004\n",
            "[15,  5120] loss: 0.00015\n",
            "[15,  5140] loss: 0.00026\n",
            "[15,  5160] loss: 0.00002\n",
            "[15,  5180] loss: 0.00005\n",
            "[15,  5200] loss: 0.00014\n",
            "[15,  5220] loss: 0.00014\n",
            "[15,  5240] loss: 0.00010\n",
            "[15,  5260] loss: 0.00025\n",
            "[15,  5280] loss: 0.00009\n",
            "[15,  5300] loss: 0.00006\n",
            "[15,  5320] loss: 0.00004\n",
            "[15,  5340] loss: 0.00007\n",
            "[15,  5360] loss: 0.00011\n",
            "[15,  5380] loss: 0.00001\n",
            "[15,  5400] loss: 0.00003\n",
            "[15,  5420] loss: 0.00001\n",
            "[15,  5440] loss: 0.00006\n",
            "[15,  5460] loss: 0.00010\n",
            "[15,  5480] loss: 0.00001\n",
            "[15,  5500] loss: 0.00006\n",
            "[15,  5520] loss: 0.00002\n",
            "[15,  5540] loss: 0.00005\n",
            "[15,  5560] loss: 0.00003\n",
            "[15,  5580] loss: 0.00005\n",
            "[15,  5600] loss: 0.00001\n",
            "[15,  5620] loss: 0.00001\n",
            "[15,  5640] loss: 0.00003\n",
            "[15,  5660] loss: 0.00003\n",
            "[15,  5680] loss: 0.00001\n",
            "[15,  5700] loss: 0.00002\n",
            "[15,  5720] loss: 0.00004\n",
            "[15,  5740] loss: 0.00007\n",
            "[15,  5760] loss: 0.00001\n",
            "[15,  5780] loss: 0.00003\n",
            "[15,  5800] loss: 0.00018\n",
            "[15,  5820] loss: 0.00001\n",
            "[15,  5840] loss: 0.00012\n",
            "[15,  5860] loss: 0.00003\n",
            "[15,  5880] loss: 0.00051\n",
            "[15,  5900] loss: 0.00024\n",
            "[15,  5920] loss: 0.00017\n",
            "[15,  5940] loss: 0.00004\n",
            "[15,  5960] loss: 0.00007\n",
            "[15,  5980] loss: 0.00001\n",
            "[15,  6000] loss: 0.00001\n",
            "[15,  6020] loss: 0.00016\n",
            "[15,  6040] loss: 0.00006\n",
            "[15,  6060] loss: 0.00004\n",
            "[15,  6080] loss: 0.00005\n",
            "[15,  6100] loss: 0.00044\n",
            "[15,  6120] loss: 0.00004\n",
            "[15,  6140] loss: 0.00009\n",
            "[15,  6160] loss: 0.00003\n",
            "[15,  6180] loss: 0.00039\n",
            "[15,  6200] loss: 0.00001\n",
            "[15,  6220] loss: 0.00006\n",
            "[15,  6240] loss: 0.00001\n",
            "[15,  6260] loss: 0.00003\n",
            "[15,  6280] loss: 0.00009\n",
            "[15,  6300] loss: 0.00041\n",
            "[15,  6320] loss: 0.00010\n",
            "[15,  6340] loss: 0.00026\n",
            "[15,  6360] loss: 0.00003\n",
            "[15,  6380] loss: 0.00038\n",
            "[15,  6400] loss: 0.00006\n",
            "[15,  6420] loss: 0.00048\n",
            "[15,  6440] loss: 0.00041\n",
            "[15,  6460] loss: 0.00007\n",
            "[15,  6480] loss: 0.00004\n",
            "[15,  6500] loss: 0.00014\n",
            "[15,  6520] loss: 0.00001\n",
            "[15,  6540] loss: 0.00003\n",
            "[15,  6560] loss: 0.00020\n",
            "[15,  6580] loss: 0.00004\n",
            "[15,  6600] loss: 0.00010\n",
            "[15,  6620] loss: 0.00002\n",
            "[15,  6640] loss: 0.00002\n",
            "[15,  6660] loss: 0.00002\n",
            "[15,  6680] loss: 0.00009\n",
            "[15,  6700] loss: 0.00005\n",
            "[15,  6720] loss: 0.00005\n",
            "[15,  6740] loss: 0.00011\n",
            "[15,  6760] loss: 0.00002\n",
            "[15,  6780] loss: 0.00035\n",
            "[15,  6800] loss: 0.00002\n",
            "[15,  6820] loss: 0.00002\n",
            "[15,  6840] loss: 0.00001\n",
            "[15,  6860] loss: 0.00003\n",
            "[15,  6880] loss: 0.00011\n",
            "[15,  6900] loss: 0.00002\n",
            "[15,  6920] loss: 0.00004\n",
            "[15,  6940] loss: 0.00001\n",
            "[15,  6960] loss: 0.00004\n",
            "[15,  6980] loss: 0.00000\n",
            "[15,  7000] loss: 0.00004\n",
            "[15,  7020] loss: 0.00001\n",
            "[15,  7040] loss: 0.00000\n",
            "[15,  7060] loss: 0.00001\n",
            "[15,  7080] loss: 0.00005\n",
            "[15,  7100] loss: 0.00009\n",
            "[15,  7120] loss: 0.00010\n",
            "[15,  7140] loss: 0.00009\n",
            "[15,  7160] loss: 0.00003\n",
            "[15,  7180] loss: 0.00002\n",
            "[15,  7200] loss: 0.00002\n",
            "[15,  7220] loss: 0.00006\n",
            "[15,  7240] loss: 0.00008\n",
            "[15,  7260] loss: 0.00000\n",
            "[15,  7280] loss: 0.00003\n",
            "[15,  7300] loss: 0.00001\n",
            "[15,  7320] loss: 0.00004\n",
            "[15,  7340] loss: 0.00002\n",
            "[15,  7360] loss: 0.00037\n",
            "[15,  7380] loss: 0.00008\n",
            "[15,  7400] loss: 0.00013\n",
            "[15,  7420] loss: 0.00005\n",
            "[15,  7440] loss: 0.00033\n",
            "[15,  7460] loss: 0.00001\n",
            "[15,  7480] loss: 0.00011\n",
            "[15,  7500] loss: 0.00002\n",
            "Finished training for epoch 14 time taken = 946.6912240982056\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 4th\n",
            "[16,    20] loss: 0.00034\n",
            "[16,    40] loss: 0.00017\n",
            "[16,    60] loss: 0.00005\n",
            "[16,    80] loss: 0.00005\n",
            "[16,   100] loss: 0.00012\n",
            "[16,   120] loss: 0.00019\n",
            "[16,   140] loss: 0.00002\n",
            "[16,   160] loss: 0.00001\n",
            "[16,   180] loss: 0.00011\n",
            "[16,   200] loss: 0.00003\n",
            "[16,   220] loss: 0.00016\n",
            "[16,   240] loss: 0.00011\n",
            "[16,   260] loss: 0.00002\n",
            "[16,   280] loss: 0.00001\n",
            "[16,   300] loss: 0.00004\n",
            "[16,   320] loss: 0.00003\n",
            "[16,   340] loss: 0.00005\n",
            "[16,   360] loss: 0.00016\n",
            "[16,   380] loss: 0.00021\n",
            "[16,   400] loss: 0.00001\n",
            "[16,   420] loss: 0.00009\n",
            "[16,   440] loss: 0.00009\n",
            "[16,   460] loss: 0.00002\n",
            "[16,   480] loss: 0.00014\n",
            "[16,   500] loss: 0.00003\n",
            "[16,   520] loss: 0.00004\n",
            "[16,   540] loss: 0.00020\n",
            "[16,   560] loss: 0.00003\n",
            "[16,   580] loss: 0.00002\n",
            "[16,   600] loss: 0.00005\n",
            "[16,   620] loss: 0.00000\n",
            "[16,   640] loss: 0.00004\n",
            "[16,   660] loss: 0.00001\n",
            "[16,   680] loss: 0.00004\n",
            "[16,   700] loss: 0.00001\n",
            "[16,   720] loss: 0.00001\n",
            "[16,   740] loss: 0.00002\n",
            "[16,   760] loss: 0.00002\n",
            "[16,   780] loss: 0.00013\n",
            "[16,   800] loss: 0.00002\n",
            "[16,   820] loss: 0.00001\n",
            "[16,   840] loss: 0.00001\n",
            "[16,   860] loss: 0.00003\n",
            "[16,   880] loss: 0.00001\n",
            "[16,   900] loss: 0.00001\n",
            "[16,   920] loss: 0.00002\n",
            "[16,   940] loss: 0.00025\n",
            "[16,   960] loss: 0.00006\n",
            "[16,   980] loss: 0.00009\n",
            "[16,  1000] loss: 0.00003\n",
            "[16,  1020] loss: 0.00001\n",
            "[16,  1040] loss: 0.00011\n",
            "[16,  1060] loss: 0.00000\n",
            "[16,  1080] loss: 0.00001\n",
            "[16,  1100] loss: 0.00004\n",
            "[16,  1120] loss: 0.00002\n",
            "[16,  1140] loss: 0.00001\n",
            "[16,  1160] loss: 0.00004\n",
            "[16,  1180] loss: 0.00003\n",
            "[16,  1200] loss: 0.00001\n",
            "[16,  1220] loss: 0.00001\n",
            "[16,  1240] loss: 0.00002\n",
            "[16,  1260] loss: 0.00002\n",
            "[16,  1280] loss: 0.00024\n",
            "[16,  1300] loss: 0.00019\n",
            "[16,  1320] loss: 0.00006\n",
            "[16,  1340] loss: 0.00002\n",
            "[16,  1360] loss: 0.00003\n",
            "[16,  1380] loss: 0.00028\n",
            "[16,  1400] loss: 0.00006\n",
            "[16,  1420] loss: 0.00035\n",
            "[16,  1440] loss: 0.00021\n",
            "[16,  1460] loss: 0.00015\n",
            "[16,  1480] loss: 0.00003\n",
            "[16,  1500] loss: 0.00004\n",
            "[16,  1520] loss: 0.00023\n",
            "[16,  1540] loss: 0.00016\n",
            "[16,  1560] loss: 0.00038\n",
            "[16,  1580] loss: 0.00003\n",
            "[16,  1600] loss: 0.00011\n",
            "[16,  1620] loss: 0.00042\n",
            "[16,  1640] loss: 0.00013\n",
            "[16,  1660] loss: 0.00001\n",
            "[16,  1680] loss: 0.00004\n",
            "[16,  1700] loss: 0.00009\n",
            "[16,  1720] loss: 0.00011\n",
            "[16,  1740] loss: 0.00017\n",
            "[16,  1760] loss: 0.00005\n",
            "[16,  1780] loss: 0.00007\n",
            "[16,  1800] loss: 0.00002\n",
            "[16,  1820] loss: 0.00007\n",
            "[16,  1840] loss: 0.00009\n",
            "[16,  1860] loss: 0.00006\n",
            "[16,  1880] loss: 0.00011\n",
            "[16,  1900] loss: 0.00004\n",
            "[16,  1920] loss: 0.00004\n",
            "[16,  1940] loss: 0.00004\n",
            "[16,  1960] loss: 0.00012\n",
            "[16,  1980] loss: 0.00005\n",
            "[16,  2000] loss: 0.00000\n",
            "[16,  2020] loss: 0.00003\n",
            "[16,  2040] loss: 0.00005\n",
            "[16,  2060] loss: 0.00009\n",
            "[16,  2080] loss: 0.00003\n",
            "[16,  2100] loss: 0.00005\n",
            "[16,  2120] loss: 0.00005\n",
            "[16,  2140] loss: 0.00023\n",
            "[16,  2160] loss: 0.00001\n",
            "[16,  2180] loss: 0.00001\n",
            "[16,  2200] loss: 0.00002\n",
            "[16,  2220] loss: 0.00003\n",
            "[16,  2240] loss: 0.00001\n",
            "[16,  2260] loss: 0.00007\n",
            "[16,  2280] loss: 0.00006\n",
            "[16,  2300] loss: 0.00004\n",
            "[16,  2320] loss: 0.00003\n",
            "[16,  2340] loss: 0.00006\n",
            "[16,  2360] loss: 0.00001\n",
            "[16,  2380] loss: 0.00004\n",
            "[16,  2400] loss: 0.00004\n",
            "[16,  2420] loss: 0.00003\n",
            "[16,  2440] loss: 0.00001\n",
            "[16,  2460] loss: 0.00003\n",
            "[16,  2480] loss: 0.00002\n",
            "[16,  2500] loss: 0.00016\n",
            "[16,  2520] loss: 0.00003\n",
            "[16,  2540] loss: 0.00001\n",
            "[16,  2560] loss: 0.00008\n",
            "[16,  2580] loss: 0.00000\n",
            "[16,  2600] loss: 0.00001\n",
            "[16,  2620] loss: 0.00003\n",
            "[16,  2640] loss: 0.00012\n",
            "[16,  2660] loss: 0.00001\n",
            "[16,  2680] loss: 0.00006\n",
            "[16,  2700] loss: 0.00001\n",
            "[16,  2720] loss: 0.00002\n",
            "[16,  2740] loss: 0.00001\n",
            "[16,  2760] loss: 0.00002\n",
            "[16,  2780] loss: 0.00002\n",
            "[16,  2800] loss: 0.00001\n",
            "[16,  2820] loss: 0.00002\n",
            "[16,  2840] loss: 0.00000\n",
            "[16,  2860] loss: 0.00001\n",
            "[16,  2880] loss: 0.00001\n",
            "[16,  2900] loss: 0.00000\n",
            "[16,  2920] loss: 0.00001\n",
            "[16,  2940] loss: 0.00002\n",
            "[16,  2960] loss: 0.00006\n",
            "[16,  2980] loss: 0.00016\n",
            "[16,  3000] loss: 0.00001\n",
            "[16,  3020] loss: 0.00017\n",
            "[16,  3040] loss: 0.00011\n",
            "[16,  3060] loss: 0.00001\n",
            "[16,  3080] loss: 0.00002\n",
            "[16,  3100] loss: 0.00001\n",
            "[16,  3120] loss: 0.00000\n",
            "[16,  3140] loss: 0.00005\n",
            "[16,  3160] loss: 0.00004\n",
            "[16,  3180] loss: 0.00000\n",
            "[16,  3200] loss: 0.00008\n",
            "[16,  3220] loss: 0.00002\n",
            "[16,  3240] loss: 0.00001\n",
            "[16,  3260] loss: 0.00007\n",
            "[16,  3280] loss: 0.00002\n",
            "[16,  3300] loss: 0.00000\n",
            "[16,  3320] loss: 0.00003\n",
            "[16,  3340] loss: 0.00007\n",
            "[16,  3360] loss: 0.00000\n",
            "[16,  3380] loss: 0.00005\n",
            "[16,  3400] loss: 0.00002\n",
            "[16,  3420] loss: 0.00003\n",
            "[16,  3440] loss: 0.00015\n",
            "[16,  3460] loss: 0.00021\n",
            "[16,  3480] loss: 0.00007\n",
            "[16,  3500] loss: 0.00001\n",
            "[16,  3520] loss: 0.00006\n",
            "[16,  3540] loss: 0.00011\n",
            "[16,  3560] loss: 0.00002\n",
            "[16,  3580] loss: 0.00004\n",
            "[16,  3600] loss: 0.00001\n",
            "[16,  3620] loss: 0.00008\n",
            "[16,  3640] loss: 0.00001\n",
            "[16,  3660] loss: 0.00004\n",
            "[16,  3680] loss: 0.00002\n",
            "[16,  3700] loss: 0.00009\n",
            "[16,  3720] loss: 0.00002\n",
            "[16,  3740] loss: 0.00029\n",
            "[16,  3760] loss: 0.00016\n",
            "[16,  3780] loss: 0.00004\n",
            "[16,  3800] loss: 0.00004\n",
            "[16,  3820] loss: 0.00002\n",
            "[16,  3840] loss: 0.00000\n",
            "[16,  3860] loss: 0.00019\n",
            "[16,  3880] loss: 0.00004\n",
            "[16,  3900] loss: 0.00001\n",
            "[16,  3920] loss: 0.00049\n",
            "[16,  3940] loss: 0.00003\n",
            "[16,  3960] loss: 0.00017\n",
            "[16,  3980] loss: 0.00001\n",
            "[16,  4000] loss: 0.00001\n",
            "[16,  4020] loss: 0.00013\n",
            "[16,  4040] loss: 0.00001\n",
            "[16,  4060] loss: 0.00002\n",
            "[16,  4080] loss: 0.00019\n",
            "[16,  4100] loss: 0.00002\n",
            "[16,  4120] loss: 0.00002\n",
            "[16,  4140] loss: 0.00003\n",
            "[16,  4160] loss: 0.00001\n",
            "[16,  4180] loss: 0.00001\n",
            "[16,  4200] loss: 0.00001\n",
            "[16,  4220] loss: 0.00004\n",
            "[16,  4240] loss: 0.00001\n",
            "[16,  4260] loss: 0.00025\n",
            "[16,  4280] loss: 0.00008\n",
            "[16,  4300] loss: 0.00001\n",
            "[16,  4320] loss: 0.00014\n",
            "[16,  4340] loss: 0.00025\n",
            "[16,  4360] loss: 0.00009\n",
            "[16,  4380] loss: 0.00000\n",
            "[16,  4400] loss: 0.00007\n",
            "[16,  4420] loss: 0.00000\n",
            "[16,  4440] loss: 0.00001\n",
            "[16,  4460] loss: 0.00000\n",
            "[16,  4480] loss: 0.00002\n",
            "[16,  4500] loss: 0.00000\n",
            "[16,  4520] loss: 0.00000\n",
            "[16,  4540] loss: 0.00041\n",
            "[16,  4560] loss: 0.00008\n",
            "[16,  4580] loss: 0.00024\n",
            "[16,  4600] loss: 0.00015\n",
            "[16,  4620] loss: 0.00004\n",
            "[16,  4640] loss: 0.00022\n",
            "[16,  4660] loss: 0.00002\n",
            "[16,  4680] loss: 0.00001\n",
            "[16,  4700] loss: 0.00001\n",
            "[16,  4720] loss: 0.00016\n",
            "[16,  4740] loss: 0.00000\n",
            "[16,  4760] loss: 0.00030\n",
            "[16,  4780] loss: 0.00001\n",
            "[16,  4800] loss: 0.00015\n",
            "[16,  4820] loss: 0.00005\n",
            "[16,  4840] loss: 0.00010\n",
            "[16,  4860] loss: 0.00001\n",
            "[16,  4880] loss: 0.00005\n",
            "[16,  4900] loss: 0.00002\n",
            "[16,  4920] loss: 0.00001\n",
            "[16,  4940] loss: 0.00003\n",
            "[16,  4960] loss: 0.00022\n",
            "[16,  4980] loss: 0.00013\n",
            "[16,  5000] loss: 0.00000\n",
            "[16,  5020] loss: 0.00001\n",
            "[16,  5040] loss: 0.00017\n",
            "[16,  5060] loss: 0.00004\n",
            "[16,  5080] loss: 0.00004\n",
            "[16,  5100] loss: 0.00005\n",
            "[16,  5120] loss: 0.00006\n",
            "[16,  5140] loss: 0.00001\n",
            "[16,  5160] loss: 0.00005\n",
            "[16,  5180] loss: 0.00029\n",
            "[16,  5200] loss: 0.00007\n",
            "[16,  5220] loss: 0.00014\n",
            "[16,  5240] loss: 0.00012\n",
            "[16,  5260] loss: 0.00002\n",
            "[16,  5280] loss: 0.00002\n",
            "[16,  5300] loss: 0.00001\n",
            "[16,  5320] loss: 0.00002\n",
            "[16,  5340] loss: 0.00002\n",
            "[16,  5360] loss: 0.00001\n",
            "[16,  5380] loss: 0.00001\n",
            "[16,  5400] loss: 0.00014\n",
            "[16,  5420] loss: 0.00006\n",
            "[16,  5440] loss: 0.00001\n",
            "[16,  5460] loss: 0.00001\n",
            "[16,  5480] loss: 0.00012\n",
            "[16,  5500] loss: 0.00002\n",
            "[16,  5520] loss: 0.00001\n",
            "[16,  5540] loss: 0.00000\n",
            "[16,  5560] loss: 0.00002\n",
            "[16,  5580] loss: 0.00008\n",
            "[16,  5600] loss: 0.00009\n",
            "[16,  5620] loss: 0.00002\n",
            "[16,  5640] loss: 0.00001\n",
            "[16,  5660] loss: 0.00009\n",
            "[16,  5680] loss: 0.00002\n",
            "[16,  5700] loss: 0.00007\n",
            "[16,  5720] loss: 0.00002\n",
            "[16,  5740] loss: 0.00002\n",
            "[16,  5760] loss: 0.00023\n",
            "[16,  5780] loss: 0.00022\n",
            "[16,  5800] loss: 0.00001\n",
            "[16,  5820] loss: 0.00023\n",
            "[16,  5840] loss: 0.00033\n",
            "[16,  5860] loss: 0.00010\n",
            "[16,  5880] loss: 0.00027\n",
            "[16,  5900] loss: 0.00033\n",
            "[16,  5920] loss: 0.00008\n",
            "[16,  5940] loss: 0.00025\n",
            "[16,  5960] loss: 0.00001\n",
            "[16,  5980] loss: 0.00011\n",
            "[16,  6000] loss: 0.00029\n",
            "[16,  6020] loss: 0.00026\n",
            "[16,  6040] loss: 0.00021\n",
            "[16,  6060] loss: 0.00071\n",
            "[16,  6080] loss: 0.00002\n",
            "[16,  6100] loss: 0.00005\n",
            "[16,  6120] loss: 0.00008\n",
            "[16,  6140] loss: 0.00029\n",
            "[16,  6160] loss: 0.00051\n",
            "[16,  6180] loss: 0.00001\n",
            "[16,  6200] loss: 0.00118\n",
            "[16,  6220] loss: 0.00023\n",
            "[16,  6240] loss: 0.00010\n",
            "[16,  6260] loss: 0.00005\n",
            "[16,  6280] loss: 0.00018\n",
            "[16,  6300] loss: 0.00002\n",
            "[16,  6320] loss: 0.00029\n",
            "[16,  6340] loss: 0.00040\n",
            "[16,  6360] loss: 0.00014\n",
            "[16,  6380] loss: 0.00004\n",
            "[16,  6400] loss: 0.00018\n",
            "[16,  6420] loss: 0.00023\n",
            "[16,  6440] loss: 0.00003\n",
            "[16,  6460] loss: 0.00007\n",
            "[16,  6480] loss: 0.00003\n",
            "[16,  6500] loss: 0.00006\n",
            "[16,  6520] loss: 0.00002\n",
            "[16,  6540] loss: 0.00007\n",
            "[16,  6560] loss: 0.00016\n",
            "[16,  6580] loss: 0.00008\n",
            "[16,  6600] loss: 0.00011\n",
            "[16,  6620] loss: 0.00001\n",
            "[16,  6640] loss: 0.00017\n",
            "[16,  6660] loss: 0.00014\n",
            "[16,  6680] loss: 0.00002\n",
            "[16,  6700] loss: 0.00002\n",
            "[16,  6720] loss: 0.00009\n",
            "[16,  6740] loss: 0.00008\n",
            "[16,  6760] loss: 0.00013\n",
            "[16,  6780] loss: 0.00008\n",
            "[16,  6800] loss: 0.00002\n",
            "[16,  6820] loss: 0.00015\n",
            "[16,  6840] loss: 0.00007\n",
            "[16,  6860] loss: 0.00012\n",
            "[16,  6880] loss: 0.00004\n",
            "[16,  6900] loss: 0.00002\n",
            "[16,  6920] loss: 0.00002\n",
            "[16,  6940] loss: 0.00001\n",
            "[16,  6960] loss: 0.00000\n",
            "[16,  6980] loss: 0.00005\n",
            "[16,  7000] loss: 0.00004\n",
            "[16,  7020] loss: 0.00001\n",
            "[16,  7040] loss: 0.00004\n",
            "[16,  7060] loss: 0.00024\n",
            "[16,  7080] loss: 0.00003\n",
            "[16,  7100] loss: 0.00000\n",
            "[16,  7120] loss: 0.00001\n",
            "[16,  7140] loss: 0.00006\n",
            "[16,  7160] loss: 0.00002\n",
            "[16,  7180] loss: 0.00006\n",
            "[16,  7200] loss: 0.00018\n",
            "[16,  7220] loss: 0.00089\n",
            "[16,  7240] loss: 0.00007\n",
            "[16,  7260] loss: 0.00004\n",
            "[16,  7280] loss: 0.00004\n",
            "[16,  7300] loss: 0.00002\n",
            "[16,  7320] loss: 0.00001\n",
            "[16,  7340] loss: 0.00001\n",
            "[16,  7360] loss: 0.00002\n",
            "[16,  7380] loss: 0.00010\n",
            "[16,  7400] loss: 0.00004\n",
            "[16,  7420] loss: 0.00005\n",
            "[16,  7440] loss: 0.00001\n",
            "[16,  7460] loss: 0.00004\n",
            "[16,  7480] loss: 0.00004\n",
            "[16,  7500] loss: 0.00001\n",
            "Finished training for epoch 15 time taken = 1010.3970143795013\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 5th\n",
            "[17,    20] loss: 0.00002\n",
            "[17,    40] loss: 0.00000\n",
            "[17,    60] loss: 0.00002\n",
            "[17,    80] loss: 0.00002\n",
            "[17,   100] loss: 0.00007\n",
            "[17,   120] loss: 0.00004\n",
            "[17,   140] loss: 0.00001\n",
            "[17,   160] loss: 0.00002\n",
            "[17,   180] loss: 0.00005\n",
            "[17,   200] loss: 0.00002\n",
            "[17,   220] loss: 0.00001\n",
            "[17,   240] loss: 0.00007\n",
            "[17,   260] loss: 0.00006\n",
            "[17,   280] loss: 0.00002\n",
            "[17,   300] loss: 0.00004\n",
            "[17,   320] loss: 0.00004\n",
            "[17,   340] loss: 0.00001\n",
            "[17,   360] loss: 0.00010\n",
            "[17,   380] loss: 0.00002\n",
            "[17,   400] loss: 0.00009\n",
            "[17,   420] loss: 0.00000\n",
            "[17,   440] loss: 0.00001\n",
            "[17,   460] loss: 0.00001\n",
            "[17,   480] loss: 0.00000\n",
            "[17,   500] loss: 0.00000\n",
            "[17,   520] loss: 0.00002\n",
            "[17,   540] loss: 0.00018\n",
            "[17,   560] loss: 0.00001\n",
            "[17,   580] loss: 0.00003\n",
            "[17,   600] loss: 0.00002\n",
            "[17,   620] loss: 0.00004\n",
            "[17,   640] loss: 0.00001\n",
            "[17,   660] loss: 0.00003\n",
            "[17,   680] loss: 0.00006\n",
            "[17,   700] loss: 0.00003\n",
            "[17,   720] loss: 0.00004\n",
            "[17,   740] loss: 0.00001\n",
            "[17,   760] loss: 0.00003\n",
            "[17,   780] loss: 0.00009\n",
            "[17,   800] loss: 0.00000\n",
            "[17,   820] loss: 0.00001\n",
            "[17,   840] loss: 0.00010\n",
            "[17,   860] loss: 0.00012\n",
            "[17,   880] loss: 0.00002\n",
            "[17,   900] loss: 0.00001\n",
            "[17,   920] loss: 0.00004\n",
            "[17,   940] loss: 0.00025\n",
            "[17,   960] loss: 0.00000\n",
            "[17,   980] loss: 0.00002\n",
            "[17,  1000] loss: 0.00029\n",
            "[17,  1020] loss: 0.00004\n",
            "[17,  1040] loss: 0.00000\n",
            "[17,  1060] loss: 0.00012\n",
            "[17,  1080] loss: 0.00001\n",
            "[17,  1100] loss: 0.00020\n",
            "[17,  1120] loss: 0.00002\n",
            "[17,  1140] loss: 0.00017\n",
            "[17,  1160] loss: 0.00031\n",
            "[17,  1180] loss: 0.00006\n",
            "[17,  1200] loss: 0.00001\n",
            "[17,  1220] loss: 0.00006\n",
            "[17,  1240] loss: 0.00007\n",
            "[17,  1260] loss: 0.00001\n",
            "[17,  1280] loss: 0.00003\n",
            "[17,  1300] loss: 0.00002\n",
            "[17,  1320] loss: 0.00018\n",
            "[17,  1340] loss: 0.00001\n",
            "[17,  1360] loss: 0.00008\n",
            "[17,  1380] loss: 0.00017\n",
            "[17,  1400] loss: 0.00000\n",
            "[17,  1420] loss: 0.00004\n",
            "[17,  1440] loss: 0.00001\n",
            "[17,  1460] loss: 0.00001\n",
            "[17,  1480] loss: 0.00003\n",
            "[17,  1500] loss: 0.00021\n",
            "[17,  1520] loss: 0.00002\n",
            "[17,  1540] loss: 0.00002\n",
            "[17,  1560] loss: 0.00005\n",
            "[17,  1580] loss: 0.00011\n",
            "[17,  1600] loss: 0.00002\n",
            "[17,  1620] loss: 0.00005\n",
            "[17,  1640] loss: 0.00005\n",
            "[17,  1660] loss: 0.00001\n",
            "[17,  1680] loss: 0.00000\n",
            "[17,  1700] loss: 0.00001\n",
            "[17,  1720] loss: 0.00001\n",
            "[17,  1740] loss: 0.00001\n",
            "[17,  1760] loss: 0.00002\n",
            "[17,  1780] loss: 0.00001\n",
            "[17,  1800] loss: 0.00003\n",
            "[17,  1820] loss: 0.00000\n",
            "[17,  1840] loss: 0.00008\n",
            "[17,  1860] loss: 0.00001\n",
            "[17,  1880] loss: 0.00001\n",
            "[17,  1900] loss: 0.00004\n",
            "[17,  1920] loss: 0.00000\n",
            "[17,  1940] loss: 0.00002\n",
            "[17,  1960] loss: 0.00002\n",
            "[17,  1980] loss: 0.00001\n",
            "[17,  2000] loss: 0.00002\n",
            "[17,  2020] loss: 0.00002\n",
            "[17,  2040] loss: 0.00001\n",
            "[17,  2060] loss: 0.00001\n",
            "[17,  2080] loss: 0.00001\n",
            "[17,  2100] loss: 0.00001\n",
            "[17,  2120] loss: 0.00000\n",
            "[17,  2140] loss: 0.00001\n",
            "[17,  2160] loss: 0.00001\n",
            "[17,  2180] loss: 0.00001\n",
            "[17,  2200] loss: 0.00000\n",
            "[17,  2220] loss: 0.00001\n",
            "[17,  2240] loss: 0.00001\n",
            "[17,  2260] loss: 0.00002\n",
            "[17,  2280] loss: 0.00000\n",
            "[17,  2300] loss: 0.00002\n",
            "[17,  2320] loss: 0.00001\n",
            "[17,  2340] loss: 0.00000\n",
            "[17,  2360] loss: 0.00000\n",
            "[17,  2380] loss: 0.00001\n",
            "[17,  2400] loss: 0.00002\n",
            "[17,  2420] loss: 0.00001\n",
            "[17,  2440] loss: 0.00001\n",
            "[17,  2460] loss: 0.00005\n",
            "[17,  2480] loss: 0.00009\n",
            "[17,  2500] loss: 0.00008\n",
            "[17,  2520] loss: 0.00000\n",
            "[17,  2540] loss: 0.00005\n",
            "[17,  2560] loss: 0.00005\n",
            "[17,  2580] loss: 0.00001\n",
            "[17,  2600] loss: 0.00003\n",
            "[17,  2620] loss: 0.00001\n",
            "[17,  2640] loss: 0.00000\n",
            "[17,  2660] loss: 0.00001\n",
            "[17,  2680] loss: 0.00001\n",
            "[17,  2700] loss: 0.00007\n",
            "[17,  2720] loss: 0.00002\n",
            "[17,  2740] loss: 0.00001\n",
            "[17,  2760] loss: 0.00002\n",
            "[17,  2780] loss: 0.00000\n",
            "[17,  2800] loss: 0.00002\n",
            "[17,  2820] loss: 0.00000\n",
            "[17,  2840] loss: 0.00000\n",
            "[17,  2860] loss: 0.00001\n",
            "[17,  2880] loss: 0.00000\n",
            "[17,  2900] loss: 0.00002\n",
            "[17,  2920] loss: 0.00001\n",
            "[17,  2940] loss: 0.00001\n",
            "[17,  2960] loss: 0.00007\n",
            "[17,  2980] loss: 0.00001\n",
            "[17,  3000] loss: 0.00000\n",
            "[17,  3020] loss: 0.00001\n",
            "[17,  3040] loss: 0.00009\n",
            "[17,  3060] loss: 0.00001\n",
            "[17,  3080] loss: 0.00005\n",
            "[17,  3100] loss: 0.00010\n",
            "[17,  3120] loss: 0.00003\n",
            "[17,  3140] loss: 0.00005\n",
            "[17,  3160] loss: 0.00001\n",
            "[17,  3180] loss: 0.00000\n",
            "[17,  3200] loss: 0.00001\n",
            "[17,  3220] loss: 0.00001\n",
            "[17,  3240] loss: 0.00001\n",
            "[17,  3260] loss: 0.00005\n",
            "[17,  3280] loss: 0.00000\n",
            "[17,  3300] loss: 0.00002\n",
            "[17,  3320] loss: 0.00005\n",
            "[17,  3340] loss: 0.00001\n",
            "[17,  3360] loss: 0.00041\n",
            "[17,  3380] loss: 0.00001\n",
            "[17,  3400] loss: 0.00001\n",
            "[17,  3420] loss: 0.00000\n",
            "[17,  3440] loss: 0.00000\n",
            "[17,  3460] loss: 0.00008\n",
            "[17,  3480] loss: 0.00010\n",
            "[17,  3500] loss: 0.00001\n",
            "[17,  3520] loss: 0.00014\n",
            "[17,  3540] loss: 0.00004\n",
            "[17,  3560] loss: 0.00000\n",
            "[17,  3580] loss: 0.00005\n",
            "[17,  3600] loss: 0.00009\n",
            "[17,  3620] loss: 0.00001\n",
            "[17,  3640] loss: 0.00001\n",
            "[17,  3660] loss: 0.00008\n",
            "[17,  3680] loss: 0.00001\n",
            "[17,  3700] loss: 0.00002\n",
            "[17,  3720] loss: 0.00000\n",
            "[17,  3740] loss: 0.00003\n",
            "[17,  3760] loss: 0.00002\n",
            "[17,  3780] loss: 0.00000\n",
            "[17,  3800] loss: 0.00009\n",
            "[17,  3820] loss: 0.00002\n",
            "[17,  3840] loss: 0.00006\n",
            "[17,  3860] loss: 0.00000\n",
            "[17,  3880] loss: 0.00001\n",
            "[17,  3900] loss: 0.00001\n",
            "[17,  3920] loss: 0.00017\n",
            "[17,  3940] loss: 0.00000\n",
            "[17,  3960] loss: 0.00002\n",
            "[17,  3980] loss: 0.00024\n",
            "[17,  4000] loss: 0.00011\n",
            "[17,  4020] loss: 0.00001\n",
            "[17,  4040] loss: 0.00004\n",
            "[17,  4060] loss: 0.00000\n",
            "[17,  4080] loss: 0.00015\n",
            "[17,  4100] loss: 0.00021\n",
            "[17,  4120] loss: 0.00001\n",
            "[17,  4140] loss: 0.00006\n",
            "[17,  4160] loss: 0.00002\n",
            "[17,  4180] loss: 0.00007\n",
            "[17,  4200] loss: 0.00001\n",
            "[17,  4220] loss: 0.00000\n",
            "[17,  4240] loss: 0.00020\n",
            "[17,  4260] loss: 0.00018\n",
            "[17,  4280] loss: 0.00002\n",
            "[17,  4300] loss: 0.00004\n",
            "[17,  4320] loss: 0.00007\n",
            "[17,  4340] loss: 0.00001\n",
            "[17,  4360] loss: 0.00002\n",
            "[17,  4380] loss: 0.00000\n",
            "[17,  4400] loss: 0.00005\n",
            "[17,  4420] loss: 0.00002\n",
            "[17,  4440] loss: 0.00001\n",
            "[17,  4460] loss: 0.00004\n",
            "[17,  4480] loss: 0.00001\n",
            "[17,  4500] loss: 0.00002\n",
            "[17,  4520] loss: 0.00009\n",
            "[17,  4540] loss: 0.00003\n",
            "[17,  4560] loss: 0.00013\n",
            "[17,  4580] loss: 0.00001\n",
            "[17,  4600] loss: 0.00002\n",
            "[17,  4620] loss: 0.00002\n",
            "[17,  4640] loss: 0.00003\n",
            "[17,  4660] loss: 0.00000\n",
            "[17,  4680] loss: 0.00030\n",
            "[17,  4700] loss: 0.00010\n",
            "[17,  4720] loss: 0.00014\n",
            "[17,  4740] loss: 0.00003\n",
            "[17,  4760] loss: 0.00004\n",
            "[17,  4780] loss: 0.00002\n",
            "[17,  4800] loss: 0.00027\n",
            "[17,  4820] loss: 0.00000\n",
            "[17,  4840] loss: 0.00001\n",
            "[17,  4860] loss: 0.00002\n",
            "[17,  4880] loss: 0.00002\n",
            "[17,  4900] loss: 0.00001\n",
            "[17,  4920] loss: 0.00001\n",
            "[17,  4940] loss: 0.00024\n",
            "[17,  4960] loss: 0.00025\n",
            "[17,  4980] loss: 0.00001\n",
            "[17,  5000] loss: 0.00001\n",
            "[17,  5020] loss: 0.00021\n",
            "[17,  5040] loss: 0.00001\n",
            "[17,  5060] loss: 0.00009\n",
            "[17,  5080] loss: 0.00001\n",
            "[17,  5100] loss: 0.00000\n",
            "[17,  5120] loss: 0.00001\n",
            "[17,  5140] loss: 0.00005\n",
            "[17,  5160] loss: 0.00001\n",
            "[17,  5180] loss: 0.00001\n",
            "[17,  5200] loss: 0.00003\n",
            "[17,  5220] loss: 0.00006\n",
            "[17,  5240] loss: 0.00014\n",
            "[17,  5260] loss: 0.00000\n",
            "[17,  5280] loss: 0.00002\n",
            "[17,  5300] loss: 0.00006\n",
            "[17,  5320] loss: 0.00017\n",
            "[17,  5340] loss: 0.00018\n",
            "[17,  5360] loss: 0.00003\n",
            "[17,  5380] loss: 0.00018\n",
            "[17,  5400] loss: 0.00000\n",
            "[17,  5420] loss: 0.00004\n",
            "[17,  5440] loss: 0.00003\n",
            "[17,  5460] loss: 0.00015\n",
            "[17,  5480] loss: 0.00012\n",
            "[17,  5500] loss: 0.00034\n",
            "[17,  5520] loss: 0.00007\n",
            "[17,  5540] loss: 0.00008\n",
            "[17,  5560] loss: 0.00003\n",
            "[17,  5580] loss: 0.00002\n",
            "[17,  5600] loss: 0.00002\n",
            "[17,  5620] loss: 0.00006\n",
            "[17,  5640] loss: 0.00002\n",
            "[17,  5660] loss: 0.00002\n",
            "[17,  5680] loss: 0.00001\n",
            "[17,  5700] loss: 0.00004\n",
            "[17,  5720] loss: 0.00010\n",
            "[17,  5740] loss: 0.00003\n",
            "[17,  5760] loss: 0.00002\n",
            "[17,  5780] loss: 0.00000\n",
            "[17,  5800] loss: 0.00004\n",
            "[17,  5820] loss: 0.00001\n",
            "[17,  5840] loss: 0.00018\n",
            "[17,  5860] loss: 0.00002\n",
            "[17,  5880] loss: 0.00000\n",
            "[17,  5900] loss: 0.00024\n",
            "[17,  5920] loss: 0.00027\n",
            "[17,  5940] loss: 0.00002\n",
            "[17,  5960] loss: 0.00005\n",
            "[17,  5980] loss: 0.00023\n",
            "[17,  6000] loss: 0.00000\n",
            "[17,  6020] loss: 0.00000\n",
            "[17,  6040] loss: 0.00002\n",
            "[17,  6060] loss: 0.00016\n",
            "[17,  6080] loss: 0.00001\n",
            "[17,  6100] loss: 0.00023\n",
            "[17,  6120] loss: 0.00002\n",
            "[17,  6140] loss: 0.00001\n",
            "[17,  6160] loss: 0.00005\n",
            "[17,  6180] loss: 0.00003\n",
            "[17,  6200] loss: 0.00007\n",
            "[17,  6220] loss: 0.00003\n",
            "[17,  6240] loss: 0.00004\n",
            "[17,  6260] loss: 0.00002\n",
            "[17,  6280] loss: 0.00003\n",
            "[17,  6300] loss: 0.00005\n",
            "[17,  6320] loss: 0.00002\n",
            "[17,  6340] loss: 0.00002\n",
            "[17,  6360] loss: 0.00002\n",
            "[17,  6380] loss: 0.00002\n",
            "[17,  6400] loss: 0.00052\n",
            "[17,  6420] loss: 0.00015\n",
            "[17,  6440] loss: 0.00009\n",
            "[17,  6460] loss: 0.00002\n",
            "[17,  6480] loss: 0.00003\n",
            "[17,  6500] loss: 0.00003\n",
            "[17,  6520] loss: 0.00008\n",
            "[17,  6540] loss: 0.00006\n",
            "[17,  6560] loss: 0.00001\n",
            "[17,  6580] loss: 0.00008\n",
            "[17,  6600] loss: 0.00001\n",
            "[17,  6620] loss: 0.00001\n",
            "[17,  6640] loss: 0.00000\n",
            "[17,  6660] loss: 0.00017\n",
            "[17,  6680] loss: 0.00006\n",
            "[17,  6700] loss: 0.00001\n",
            "[17,  6720] loss: 0.00036\n",
            "[17,  6740] loss: 0.00016\n",
            "[17,  6760] loss: 0.00002\n",
            "[17,  6780] loss: 0.00044\n",
            "[17,  6800] loss: 0.00019\n",
            "[17,  6820] loss: 0.00017\n",
            "[17,  6840] loss: 0.00005\n",
            "[17,  6860] loss: 0.00028\n",
            "[17,  6880] loss: 0.00002\n",
            "[17,  6900] loss: 0.00008\n",
            "[17,  6920] loss: 0.00001\n",
            "[17,  6940] loss: 0.00001\n",
            "[17,  6960] loss: 0.00068\n",
            "[17,  6980] loss: 0.00007\n",
            "[17,  7000] loss: 0.00003\n",
            "[17,  7020] loss: 0.00002\n",
            "[17,  7040] loss: 0.00001\n",
            "[17,  7060] loss: 0.00002\n",
            "[17,  7080] loss: 0.00037\n",
            "[17,  7100] loss: 0.00004\n",
            "[17,  7120] loss: 0.00003\n",
            "[17,  7140] loss: 0.00005\n",
            "[17,  7160] loss: 0.00004\n",
            "[17,  7180] loss: 0.00003\n",
            "[17,  7200] loss: 0.00051\n",
            "[17,  7220] loss: 0.00004\n",
            "[17,  7240] loss: 0.00010\n",
            "[17,  7260] loss: 0.00018\n",
            "[17,  7280] loss: 0.00000\n",
            "[17,  7300] loss: 0.00001\n",
            "[17,  7320] loss: 0.00006\n",
            "[17,  7340] loss: 0.00007\n",
            "[17,  7360] loss: 0.00004\n",
            "[17,  7380] loss: 0.00001\n",
            "[17,  7400] loss: 0.00005\n",
            "[17,  7420] loss: 0.00005\n",
            "[17,  7440] loss: 0.00015\n",
            "[17,  7460] loss: 0.00017\n",
            "[17,  7480] loss: 0.00038\n",
            "[17,  7500] loss: 0.00007\n",
            "Finished training for epoch 16 time taken = 1073.987805366516\n",
            "################################evaluation###########################\n",
            "\n",
            "no improvement with prev best model 6th\n",
            "stop training\n",
            "Finished Training................................................\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yeNAfjn0bt16"
      },
      "source": [
        "## LOAD MODEL"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mCZe3irla9B_",
        "outputId": "88a3586b-af40-4ae0-c5a7-a12a3a372572"
      },
      "source": [
        "# loading one of the models\n",
        "!gdown --id 1yCzEwzbqVmVQ9O2G8GMolzzO0g5dGLns\n",
        "PATH = 'xyz_epoch_9.pt'\n",
        "model = NNmodel()\n",
        "model = model.to('cuda')\n",
        "model.load_state_dict(torch.load(PATH))\n",
        "model.eval()"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1yCzEwzbqVmVQ9O2G8GMolzzO0g5dGLns\n",
            "To: /content/xyz_epoch_9.pt\n",
            "\r  0% 0.00/250k [00:00<?, ?B/s]\r100% 250k/250k [00:00<00:00, 84.2MB/s]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "NNmodel(\n",
              "  (conv1): Conv2d(1, 6, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
              "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
              "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ERXYpna0b_-C"
      },
      "source": [
        "## TESTING"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5IPxAMK6cB1t"
      },
      "source": [
        "# visualization of the testing set\n",
        "dataiter = iter(testloader)\n",
        "images, labels = dataiter.next()"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 112
        },
        "id": "2FAHbKAjcE47",
        "outputId": "0c388237-38e1-4cc6-c179-7050979b3f3b"
      },
      "source": [
        "# show images\n",
        "imshow(torchvision.utils.make_grid(images))\n",
        "\n",
        "# print labels\n",
        "print('Ground Truth: ', ' '.join('%5s' % classes[labels[j]] for j in range(batch)))"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Ground Truth:  7 - seven 2 - two 1 - one 0 - zero 4 - four 1 - one 4 - four 9 - nine\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4ksEdd_RcOc_",
        "outputId": "e18772c1-978b-4e4d-f0c2-7c67f7f45e5e"
      },
      "source": [
        "# testing against one of the models after training\n",
        "outputs = model(images.to('cuda'))\n",
        "_, predicted = torch.max(outputs, 1)\n",
        "\n",
        "# printing prediction\n",
        "print('Predicted: ', ' '.join('%5s' % classes[predicted[j]] for j in range(batch)))"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predicted:  7 - seven 2 - two 1 - one 0 - zero 4 - four 1 - one 4 - four 9 - nine\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vu2DHsc8cYLX"
      },
      "source": [
        "## ACCURACY"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7UpcQMIocZ71",
        "outputId": "88af7f5a-4bba-4905-b58d-252e739f96d9"
      },
      "source": [
        "# checking the accuracy after the testing\n",
        "correct = 0\n",
        "total = 0\n",
        "with torch.no_grad():\n",
        "    for data in testloader:\n",
        "        images, labels = data[0].to('cuda'), data[1].to('cuda')\n",
        "        outputs = model(images)\n",
        "        _, predicted = torch.max(outputs.data, 1)\n",
        "        total += labels.size(0)\n",
        "        correct += (predicted == labels).sum().item()\n",
        "\n",
        "print('Accuracy of the network on the 3776 test images: %d %%' % (100 * correct / total))"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy of the network on the 3776 test images: 98 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D1pCSYV5coR8",
        "outputId": "7cac4571-92e3-4474-993f-b185e0abc94c"
      },
      "source": [
        "# testing the accuracy of the individual classes\n",
        "sftmx = nn.Softmax()\n",
        "class_correct = list(0. for i in range(10))\n",
        "class_total = list(0. for i in range(10))\n",
        "with torch.no_grad():\n",
        "    for data in testloader:\n",
        "        images, labels = data[0].to(device), data[1].to(device)\n",
        "        #print(labels)\n",
        "        outputs = model(images)\n",
        "        _, predicted = torch.max(outputs, 1)\n",
        "\n",
        "        #print(sftmx(outputs))\n",
        "        c = (predicted == labels).squeeze()\n",
        "        for i in range(batch):\n",
        "            label = labels[i]\n",
        "            class_correct[label] += c[i].item()\n",
        "            class_total[label] += 1\n",
        "\n",
        "\n",
        "for i in range(10):\n",
        "    print('Accuracy of %5s : %2d %%' % (\n",
        "        classes[i], 100 * class_correct[i] / class_total[i]))"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy of 0 - zero : 99 %\n",
            "Accuracy of 1 - one : 99 %\n",
            "Accuracy of 2 - two : 99 %\n",
            "Accuracy of 3 - three : 99 %\n",
            "Accuracy of 4 - four : 99 %\n",
            "Accuracy of 5 - five : 98 %\n",
            "Accuracy of 6 - six : 98 %\n",
            "Accuracy of 7 - seven : 99 %\n",
            "Accuracy of 8 - eight : 97 %\n",
            "Accuracy of 9 - nine : 98 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1qG9hEsPh8zE"
      },
      "source": [
        "## TESTING ON A RANDOM IMAGE\n",
        "\n",
        "Let's test on a random image."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lxvKxzUQiEGD",
        "outputId": "f17d1718-1e03-40cc-aba8-c67c1f411def"
      },
      "source": [
        "# uploading the image\n",
        "!gdown --id 16UDdyu2qbcp0b0dL326d7VIT4qHCdaeA"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=16UDdyu2qbcp0b0dL326d7VIT4qHCdaeA\n",
            "To: /content/img2.png\n",
            "\r  0% 0.00/593 [00:00<?, ?B/s]\r100% 593/593 [00:00<00:00, 1.14MB/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "KIjfQ-0YifXq",
        "outputId": "0a9c9d57-d8fe-4811-d032-7d90461c67da"
      },
      "source": [
        "# plotting the image\n",
        "plt.imshow(np.asarray(Image.open('img2.png')), \"gray\")"
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7ffa07a14dd0>"
            ]
          },
          "metadata": {},
          "execution_count": 49
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6keK3teGyDlE"
      },
      "source": [
        "# defining the transformations\n",
        "testing_transforms = transforms.Compose([transforms.Resize((32,32)), transforms.ToTensor(), transforms.Normalize((0.5), (0.5))])"
      ],
      "execution_count": 50,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IH-vvQbNyHDr"
      },
      "source": [
        "# giving the image input\n",
        "input = testing_transforms(Image.open('img2.png'))"
      ],
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "udWlDfFpyTLU",
        "outputId": "88c4e8cf-d61d-4d49-e658-1e197ae65de9"
      },
      "source": [
        "# testing against the model and finding the output\n",
        "output = model(input.unsqueeze(0).to(device))\n",
        "_, predicted = torch.max(output, 1)\n",
        "print(classes[predicted])"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "5 - five\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "BFeto-h1FEyk"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}