3_ImageGeneration&GAN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "3_ImageGeneration&GAN.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "c3b7af621fb54e328e1c4611d5046ee3": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_b1f38a6c441646b2ac8a5b94486c5e39",
              "IPY_MODEL_3175d53b00264c94807a56fdb85c2899",
              "IPY_MODEL_dd880baf474d4fb4bf0670bfe15a62fd"
            ],
            "layout": "IPY_MODEL_a9ce0e0f689b4f3fa639dfa89a910ede"
          }
        },
        "b1f38a6c441646b2ac8a5b94486c5e39": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_e401d1656fd042c1b675cd063395221c",
            "placeholder": "​",
            "style": "IPY_MODEL_98b413b979f64386ad8890af3ea4213d",
            "value": ""
          }
        },
        "3175d53b00264c94807a56fdb85c2899": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_21b0151a800c43ee83faa3cf400ecd89",
            "max": 9912422,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_fd9fe9e04ca74559954e9048913fdafe",
            "value": 9912422
          }
        },
        "dd880baf474d4fb4bf0670bfe15a62fd": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_7629b3e849ad43c89542e46fab302263",
            "placeholder": "​",
            "style": "IPY_MODEL_60761f97352b4d45ab7032e7fd424c9d",
            "value": " 9913344/? [00:00&lt;00:00, 52554872.21it/s]"
          }
        },
        "a9ce0e0f689b4f3fa639dfa89a910ede": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "e401d1656fd042c1b675cd063395221c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "98b413b979f64386ad8890af3ea4213d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "21b0151a800c43ee83faa3cf400ecd89": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "fd9fe9e04ca74559954e9048913fdafe": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "7629b3e849ad43c89542e46fab302263": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "60761f97352b4d45ab7032e7fd424c9d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "fcc34341aa62417589e94fef7b8a7496": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_5710298fba0e42c79aab9529ce75037f",
              "IPY_MODEL_5ddb7f399fa840cf93decc83dc6166f7",
              "IPY_MODEL_a9807ef780e1404cbb46965e419d14b4"
            ],
            "layout": "IPY_MODEL_2625777c7b6d4431bcb2ee81b00534df"
          }
        },
        "5710298fba0e42c79aab9529ce75037f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_19e0b3755a814c53bb0381b93236f944",
            "placeholder": "​",
            "style": "IPY_MODEL_6568165fc73440debde28ccfe3cb14f7",
            "value": ""
          }
        },
        "5ddb7f399fa840cf93decc83dc6166f7": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_55beea1889994baa8c51a59057923c0d",
            "max": 28881,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_4f3045f15d8a44589c1650625691987e",
            "value": 28881
          }
        },
        "a9807ef780e1404cbb46965e419d14b4": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d1c20a4c46d347788c43f18f61d9eb05",
            "placeholder": "​",
            "style": "IPY_MODEL_2151f780e5844e4aba14cdaecbfd1807",
            "value": " 29696/? [00:00&lt;00:00, 800573.67it/s]"
          }
        },
        "2625777c7b6d4431bcb2ee81b00534df": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "19e0b3755a814c53bb0381b93236f944": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6568165fc73440debde28ccfe3cb14f7": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "55beea1889994baa8c51a59057923c0d": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "4f3045f15d8a44589c1650625691987e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "d1c20a4c46d347788c43f18f61d9eb05": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "2151f780e5844e4aba14cdaecbfd1807": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "9067f574a4504e739c3e82fae73f20c1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_0542090dbb9d442e80f776c40ec45580",
              "IPY_MODEL_5daa32dbcec84be4857804b98bbc7562",
              "IPY_MODEL_ad2bedcadbb64e459e471ee3cb02dc29"
            ],
            "layout": "IPY_MODEL_2e1cc6b3052c4804a40f8a21c5220683"
          }
        },
        "0542090dbb9d442e80f776c40ec45580": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_42a025ff6f784fc3ab27144c9b8cf4d5",
            "placeholder": "​",
            "style": "IPY_MODEL_aa28441f1beb4c1ebf12c79eaa4930ec",
            "value": ""
          }
        },
        "5daa32dbcec84be4857804b98bbc7562": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a6b59d262da64000956a5873f74a015e",
            "max": 1648877,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_cd461de843204453add8ac3960f7dd75",
            "value": 1648877
          }
        },
        "ad2bedcadbb64e459e471ee3cb02dc29": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_967ad85461a746a88bf9de34fb8325de",
            "placeholder": "​",
            "style": "IPY_MODEL_6bb87ff1cd5b4c6fa970e5c90e57a204",
            "value": " 1649664/? [00:00&lt;00:00, 18142752.71it/s]"
          }
        },
        "2e1cc6b3052c4804a40f8a21c5220683": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "42a025ff6f784fc3ab27144c9b8cf4d5": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "aa28441f1beb4c1ebf12c79eaa4930ec": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a6b59d262da64000956a5873f74a015e": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "cd461de843204453add8ac3960f7dd75": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "967ad85461a746a88bf9de34fb8325de": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6bb87ff1cd5b4c6fa970e5c90e57a204": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "978eb2fade4d45cc8566d0b1f537820d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_78b49df3ed3b44549f48263a588906ad",
              "IPY_MODEL_c7791e18a75b4703961ac251503565fe",
              "IPY_MODEL_951ba503007a4de09d4dece9dce8922a"
            ],
            "layout": "IPY_MODEL_f4ef679f0c624879a224ea013c2dd3ac"
          }
        },
        "78b49df3ed3b44549f48263a588906ad": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_fe0faca4836549b7af0aa287f52f8252",
            "placeholder": "​",
            "style": "IPY_MODEL_51e119155fc044a0b03f94bf7b3d60e2",
            "value": ""
          }
        },
        "c7791e18a75b4703961ac251503565fe": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_0fcee55c7eca4bbfb7f0ecdf9c514ae2",
            "max": 4542,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_90cfa1a70aa7427d8f2ffe3ff02afb93",
            "value": 4542
          }
        },
        "951ba503007a4de09d4dece9dce8922a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_3820bdccdd9b4f95b072f9abece96ec8",
            "placeholder": "​",
            "style": "IPY_MODEL_c80271a45f954d9aa4bebb520cf17ce8",
            "value": " 5120/? [00:00&lt;00:00, 114720.30it/s]"
          }
        },
        "f4ef679f0c624879a224ea013c2dd3ac": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "fe0faca4836549b7af0aa287f52f8252": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "51e119155fc044a0b03f94bf7b3d60e2": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "0fcee55c7eca4bbfb7f0ecdf9c514ae2": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "90cfa1a70aa7427d8f2ffe3ff02afb93": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "3820bdccdd9b4f95b072f9abece96ec8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "c80271a45f954d9aa4bebb520cf17ce8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Tikquuss/GflowNets_Tutorial/blob/main/3_ImageGeneration%26GAN.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# x -----  GFN -------- x\n",
        "\n"
      ],
      "metadata": {
        "id": "6t0OFIVXteoJ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this section, we will train a GflowNet whose reward is in the form  $R(y \\ | \\ x) = e^{-a*distance(x,y)}$ where $x$ is a source image and $y$ is an image constructed using GflowNet (trained with Trajectory Balance objective), and $a > 0$ : we start with a black image (only 0's), and at each step GflwoNet  has to choose a pixel and color it white (1) or choose to return the image (stop action).\n",
        "\n",
        "In [4], the *initial state* is a completely empty image ($\\emptyset$ on all pixels) and the action space of the forward policy is $I \\times C$, where :\n",
        "- $I = [1, N] \\times [1, M] ∼ [1, N*M]$ (for an image of size $M \\times N$)  : at each step, we have to choose the pixel to color (among the pixels not yet colored).\n",
        "- $C = \\{0, 1\\}$ : each pixel can be colored in black (0) or in white (1)\n",
        "\n",
        "The process continues until all pixels are colored (*final state*).\n",
        "\n",
        "The backward actions (actions of the backward policy) are simple, because they are dual to the one of the forward policy : if the forward policy has chosen a cell $(i, j) \\in I$ and colored it with the color $k \\in C$, the backward policy must choose the same cell $(i, j)$ and decolorize it. The action space of the backward policy is then $I \\times \\{ \\emptyset \\}  ∼ I$."
      ],
      "metadata": {
        "id": "2ol502fBHqd8"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![4.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "RAs9fEIcHrwO"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Image from [4]\n",
        "\n",
        "Note that this description is valid just because during training, we choose to build our objects (i.e. we move from the initial state to the final state during training). But, since we have access to the modes in the reconstruction problem (the modes are the neighbors of $x$ in the metric space generated on $I$ by the distance we use in the reward function), we can decide to train the backward policy to sample the initial state from these modes (finals states with high reward), ie samples complete trajectories leading from the initial state to these modes, and force the forward policy to follow the opposite actions of the backward policy (what I call being its dual). Then, during inference, use this forward policy to sample the final states from the initial state. This process can extremely speed up the training.\n",
        "\n",
        "In our case, to make things simple, the initial state is a completely black image (0), and at each step the model just has to choose the pixel to make white (1) or return the image as final state (stop action). The forward action space is reduced to $I \\cup \\{stop\\_action \\}$. If the forward policy has chosen a cell $(i, j) \\in I$ and colored it (white, $1$), the backward policy must choose the same cell $(i, j)$ and decolorize it (black, $0$). The backward action space is then $I$.\n",
        "\n",
        "We will use Euclidean distance, and we will choose $a$ depending on $M$ and $N$.\n",
        "\n",
        "For small images (4x4, 5x5, ..., 10x10) it works well (the model returns the mode when we sample from it), but for 11x11 for example the model ''fails'' (it returns some modes, but not $x$). There is only one \"true\" mode in this case (x), but we define the modes by taking the images that have distance below a threshold."
      ],
      "metadata": {
        "id": "Z8QWMFblFYvg"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from torchvision.datasets import MNIST\n",
        "import torchvision.transforms as transforms"
      ],
      "metadata": {
        "id": "RdHmc4d7uPWL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#M,N = 28, 28\n",
        "#M,N = 10, 10\n",
        "M,N = 6, 6\n",
        "transform = transforms.Compose([\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Resize(size=(M,N)),\n",
        "        lambda x : 1*(x!=0) + 0*(x==0),\n",
        "])\n",
        "data = MNIST(root='/content', download=True, train=True, transform=transform)"
      ],
      "metadata": {
        "id": "SVTf0TB6uXkC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423,
          "referenced_widgets": [
            "c3b7af621fb54e328e1c4611d5046ee3",
            "b1f38a6c441646b2ac8a5b94486c5e39",
            "3175d53b00264c94807a56fdb85c2899",
            "dd880baf474d4fb4bf0670bfe15a62fd",
            "a9ce0e0f689b4f3fa639dfa89a910ede",
            "e401d1656fd042c1b675cd063395221c",
            "98b413b979f64386ad8890af3ea4213d",
            "21b0151a800c43ee83faa3cf400ecd89",
            "fd9fe9e04ca74559954e9048913fdafe",
            "7629b3e849ad43c89542e46fab302263",
            "60761f97352b4d45ab7032e7fd424c9d",
            "fcc34341aa62417589e94fef7b8a7496",
            "5710298fba0e42c79aab9529ce75037f",
            "5ddb7f399fa840cf93decc83dc6166f7",
            "a9807ef780e1404cbb46965e419d14b4",
            "2625777c7b6d4431bcb2ee81b00534df",
            "19e0b3755a814c53bb0381b93236f944",
            "6568165fc73440debde28ccfe3cb14f7",
            "55beea1889994baa8c51a59057923c0d",
            "4f3045f15d8a44589c1650625691987e",
            "d1c20a4c46d347788c43f18f61d9eb05",
            "2151f780e5844e4aba14cdaecbfd1807",
            "9067f574a4504e739c3e82fae73f20c1",
            "0542090dbb9d442e80f776c40ec45580",
            "5daa32dbcec84be4857804b98bbc7562",
            "ad2bedcadbb64e459e471ee3cb02dc29",
            "2e1cc6b3052c4804a40f8a21c5220683",
            "42a025ff6f784fc3ab27144c9b8cf4d5",
            "aa28441f1beb4c1ebf12c79eaa4930ec",
            "a6b59d262da64000956a5873f74a015e",
            "cd461de843204453add8ac3960f7dd75",
            "967ad85461a746a88bf9de34fb8325de",
            "6bb87ff1cd5b4c6fa970e5c90e57a204",
            "978eb2fade4d45cc8566d0b1f537820d",
            "78b49df3ed3b44549f48263a588906ad",
            "c7791e18a75b4703961ac251503565fe",
            "951ba503007a4de09d4dece9dce8922a",
            "f4ef679f0c624879a224ea013c2dd3ac",
            "fe0faca4836549b7af0aa287f52f8252",
            "51e119155fc044a0b03f94bf7b3d60e2",
            "0fcee55c7eca4bbfb7f0ecdf9c514ae2",
            "90cfa1a70aa7427d8f2ffe3ff02afb93",
            "3820bdccdd9b4f95b072f9abece96ec8",
            "c80271a45f954d9aa4bebb520cf17ce8"
          ]
        },
        "outputId": "08c9d31a-28ec-4e18-9e00-0b8b2fd50bc8"
      },
      "execution_count": null,
      "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 /content/MNIST/raw/train-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/9912422 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "c3b7af621fb54e328e1c4611d5046ee3"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting /content/MNIST/raw/train-images-idx3-ubyte.gz to /content/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 /content/MNIST/raw/train-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/28881 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "fcc34341aa62417589e94fef7b8a7496"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting /content/MNIST/raw/train-labels-idx1-ubyte.gz to /content/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 /content/MNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/1648877 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "9067f574a4504e739c3e82fae73f20c1"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting /content/MNIST/raw/t10k-images-idx3-ubyte.gz to /content/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 /content/MNIST/raw/t10k-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/4542 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "978eb2fade4d45cc8566d0b1f537820d"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting /content/MNIST/raw/t10k-labels-idx1-ubyte.gz to /content/MNIST/raw\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "x = data[10][0]\n",
        "print(x.shape)\n",
        "\n",
        "plt.imshow(x.squeeze(0)),plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "lXGd_mGquwg8",
        "outputId": "70b455af-d6aa-4b56-b042-fa79fc68b909"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([1, 6, 6])\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<matplotlib.image.AxesImage at 0x7f10d8b71810>, None)"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x = x.view(x.shape[0], -1) # (1, 784)"
      ],
      "metadata": {
        "id": "zXB9FacCyT-e"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import torch \n",
        "plt.imshow(torch.randint(low=0, high=2, size=(1, M,N)).squeeze(0))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "9rqe7LQG0aoF",
        "outputId": "e853f8e3-2a68-4d0e-8190-fc2eb1f397e1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f10d859b0d0>"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def make_mlp(l, act=torch.nn.LeakyReLU(), tail=[]):\n",
        "    return torch.nn.Sequential(*(sum(\n",
        "        [[torch.nn.Linear(i, o)] + ([act] if n < len(l)-2 else [])\n",
        "         for n, (i, o) in enumerate(zip(l, l[1:]))], []) + tail))"
      ],
      "metadata": {
        "id": "6tVO5UzWx9yL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
        "x = x.to(device)"
      ],
      "metadata": {
        "id": "WZ0hM9WwzYIl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "n_hid = 256\n",
        "n_layers = 2\n",
        "ndim = M*N\n",
        "input_dim = ndim # embedding dim\n",
        "output_dim = 2*ndim+1 # ndim + 1 for P_F (+1 for stop action) and ndim for P_B \n",
        "\n",
        "#For GAN, it is better to learn a Z for each image and not a global Z as there.\n",
        "independent_Z = False\n",
        "if independent_Z:\n",
        "    output_dim+=1\n",
        "    logZ_TB = 0\n",
        "    model_TB = make_mlp([input_dim] + [n_hid] * n_layers + [output_dim]).to(device)\n",
        "    optimizer = torch.optim.Adam([ {'params':model_TB.parameters(), 'lr':0.001}])\n",
        "else :\n",
        "    logZ_TB = torch.zeros((1,)).to(device) # log (initial state flow), Z = 1\n",
        "    model_TB = make_mlp([input_dim] + [n_hid] * n_layers + [output_dim]).to(device)\n",
        "    optimizer = torch.optim.Adam([ {'params':model_TB.parameters(), 'lr':0.001}, {'params':[logZ_TB], 'lr':0.1} ])\n",
        "    logZ_TB.requires_grad_()\n",
        "\n",
        "model_TB"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5nQEqsWWyGJb",
        "outputId": "99ea04ee-65d2-49f9-c67f-7d26f0b047e5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Sequential(\n",
              "  (0): Linear(in_features=36, out_features=256, bias=True)\n",
              "  (1): LeakyReLU(negative_slope=0.01)\n",
              "  (2): Linear(in_features=256, out_features=256, bias=True)\n",
              "  (3): LeakyReLU(negative_slope=0.01)\n",
              "  (4): Linear(in_features=256, out_features=73, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tqdm\n",
        "import numpy as np"
      ],
      "metadata": {
        "id": "c_9hc3Wqz2bL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "losses_TB = []\n",
        "rewards_TB = []\n",
        "logZ_TB_list = []\n",
        "all_visited_TB = []"
      ],
      "metadata": {
        "id": "_ZnnfhKb3JTV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "n_train_steps = 10000\n",
        "batch_size = 1\n",
        "uniform_PB = False\n",
        "minus_inf = -1e8\n",
        "\n",
        "a = 100"
      ],
      "metadata": {
        "id": "QHUDwU-mI-FI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for it in tqdm.trange(n_train_steps):    \n",
        "    # TB loss for each trajectory\n",
        "    loss_TB = torch.zeros((batch_size,)).to(device)\n",
        "    loss_TB += 0 if independent_Z else logZ_TB # see the equation above\n",
        "    # finished trajectories\n",
        "    dones = torch.full((batch_size,), False, dtype=torch.bool).to(device)\n",
        "    # s_0\n",
        "    states = torch.zeros(size=(batch_size, ndim)).to(device) # (batch_size, ndim)\n",
        "    # actions chosen at each step \n",
        "    actions = None # (current_batch_size,)\n",
        "\n",
        "    max_steps = 1e8 #ndim+0\n",
        "    i = 0\n",
        "    while torch.any(~dones) and i <= max_steps :\n",
        "        ### Forward pass ### \n",
        "        current_batch_size = (~dones).sum()\n",
        "        non_terminal_states = states[~dones] # (current_batch_size, ndim)\n",
        "        logits = model_TB(non_terminal_states) # (current_batch_size, output_dim)\n",
        "\n",
        "        ### Backward Policy ### \n",
        "        PB_logits = logits[...,ndim+1:2*ndim+1] # (current_batch_size, ndim)\n",
        "        PB_logits = PB_logits * (0 if uniform_PB else 1) # (current_batch_size, ndim)\n",
        "        # Cells that are still black (0) are excluded from the action space of the backward policy\n",
        "        PB_mask = (non_terminal_states == 0.).float() # (current_batch_size, ndim)\n",
        "        logPB = (PB_logits + minus_inf*PB_mask).log_softmax(1) # (current_batch_size, ndim)\n",
        "        if actions is not None: \n",
        "            loss_TB[~dones] -= logPB.gather(1, actions[actions!=ndim].unsqueeze(1)).squeeze(1)\n",
        "        elif independent_Z :\n",
        "            logZ_TB = logits[...,-1]\n",
        "            loss_TB += logZ_TB + 0 \n",
        "            logZ_TB = logZ_TB.mean()\n",
        "\n",
        "        ### Forward Policy ### \n",
        "        PF_logits = logits[...,:ndim+1] # (current_batch_size, ndim+1) \n",
        "        # Cells that are already white (1) are excluded from the action space of the forward policy\n",
        "        edge_mask = (non_terminal_states == 1.).float() # (current_batch_size, ndim)\n",
        "        stop_action_mask = torch.zeros((current_batch_size, 1), device=device) # (current_batch_size, 1)\n",
        "        PF_mask = torch.cat([edge_mask, stop_action_mask], 1) # (current_batch_size, ndim+1)\n",
        "        logPF = (PF_logits + minus_inf*PF_mask).log_softmax(1) # (current_batch_size, ndim+1)\n",
        "        sample_temperature = 1\n",
        "        sample_ins_probs = (logPF/sample_temperature).softmax(1) # (current_batch_size, ndim+1)\n",
        "        actions = sample_ins_probs.multinomial(1) # (current_batch_size,)\n",
        "        loss_TB[~dones] += logPF.gather(1, actions).squeeze(1)\n",
        "\n",
        "        ### select terminal states ### \n",
        "        terminates = (actions==ndim).squeeze(1)\n",
        "        for state in non_terminal_states[terminates]: \n",
        "            all_visited_TB.append(state)\n",
        "       \n",
        "       # Update dones\n",
        "        dones[~dones] |= terminates\n",
        "\n",
        "        # Update non completed trajectories\n",
        "        with torch.no_grad():\n",
        "            non_terminates = actions[~terminates].squeeze()\n",
        "            tmp = states[~dones]\n",
        "            tmp[torch.arange((~dones).sum()), non_terminates] = 1.\n",
        "            states[~dones] = tmp\n",
        "        \n",
        "        i+=1\n",
        "        \n",
        "    dist = torch.nn.functional.mse_loss(input=states.view(M,N), target=x.view(M,N))\n",
        "    R = (-a*dist).exp()\n",
        "    loss_TB -= R.log()\n",
        "    loss = (loss_TB**2).sum()/batch_size\n",
        "\n",
        "    optimizer.zero_grad()\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    losses_TB.append(loss.item())\n",
        "    rewards_TB.append(R.mean().cpu())\n",
        "    logZ_TB_list.append(logZ_TB.item())\n",
        "\n",
        "    if it%100==0: \n",
        "        print('\\nloss =', np.array(losses_TB[-1:]).mean(), 'logZ =', logZ_TB.item(), \"R =\", np.array(rewards_TB[-1:]).mean())\n",
        "        plt.imshow(states.view(M,N).cpu()), plt.show()"
      ],
      "metadata": {
        "id": "EMKy-WBXz1qg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "38beec6a-f8d9-4266-d5a8-35fd00c99dca"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r  0%|          | 0/10000 [00:00<?, ?it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 512.6591186523438 logZ = -0.10000000149011612 R = 4.9890988e-20\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  1%|          | 99/10000 [00:09<12:31, 13.18it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 467.8556213378906 logZ = -7.15767765045166 R = 7.4563803e-25\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  2%|▏         | 197/10000 [00:15<10:41, 15.27it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 80.70758056640625 logZ = -12.277861595153809 R = 5.368969e-14\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  3%|▎         | 298/10000 [00:22<08:17, 19.51it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 2.67977237701416 logZ = -14.20073127746582 R = 1.2905354e-17\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  4%|▍         | 399/10000 [00:26<07:15, 22.06it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.3996076583862305 logZ = -14.165170669555664 R = 5.368969e-14\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  5%|▌         | 500/10000 [00:32<05:16, 30.03it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 2.0482215881347656 logZ = -12.582322120666504 R = 5.7777417e-08\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  6%|▌         | 599/10000 [00:35<05:25, 28.85it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 100.78427124023438 logZ = -12.082448959350586 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  7%|▋         | 698/10000 [00:39<07:01, 22.07it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 4.439117908477783 logZ = -11.341734886169434 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  8%|▊         | 799/10000 [00:44<07:07, 21.50it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 26.922605514526367 logZ = -10.089076042175293 R = 5.7777417e-08\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "  9%|▉         | 900/10000 [00:49<07:12, 21.02it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.18853692710399628 logZ = -8.045300483703613 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 10%|▉         | 999/10000 [00:53<06:57, 21.54it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 71.15153503417969 logZ = -6.860934734344482 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 11%|█         | 1098/10000 [00:58<06:45, 21.95it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 3.086512804031372 logZ = -5.577122688293457 R = 5.7777417e-08\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 12%|█▏        | 1198/10000 [01:03<07:21, 19.92it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.04760092496871948 logZ = -4.529385089874268 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAJpElEQVR4nO3dzYtdhR3G8efpdJoYXxBrKnmjcSGCCI1lyEYpNGATX9AuFXQlzKZCpAXRpf+AuOkmqLRFaxBUEGubhhqRgCZOYkxNoiJiMVGIrYim0Gji08XclKnEzLk358w5/vh+YHBeLtcHzTfn3jMz9ziJANTxvb4HAGgXUQPFEDVQDFEDxRA1UMz3u7jTyy+byvp1013c9UTePbii7wlAq/6jf+vLnPTZvtZJ1OvXTWvvjnVd3PVENq/e0PcEoFV78rdv/RoPv4FiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiGkVte4vtd2y/Z/uBrkcBmNyiUduekvRbSTdJukbSnbav6XoYgMk0OVJvlPRekveTfClpu6Tbu50FYFJNol4j6cMFHx8dfe7/2J61PWd77pN/nW5rH4AxtXaiLMm2JDNJZlb+cKqtuwUwpiZRH5O08LWJ1o4+B2CAmkT9uqSrbF9p+weS7pD0fLezAExq0RceTHLK9r2SdkiakvR4kkOdLwMwkUavJprkRUkvdrwFQAv4iTKgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKafQLHeN69+AKbV69oYu7LmPHRwf6njBo/PmZHEdqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqCYRaO2/bjt47bfWopBAM5PkyP17yRt6XgHgJYsGnWSVyR9ugRbALSgtZczsj0raVaSlmtFW3cLYEytnShLsi3JTJKZaS1r624BjImz30AxRA0U0+RbWk9JelXS1baP2r6n+1kAJrXoibIkdy7FEADt4OE3UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVBMkwvkrbO9y/Zh24dsb12KYQAms+gF8iSdkvSbJPttXyxpn+2dSQ53vA3ABBY9Uif5OMn+0ftfSDoiaU3XwwBMpsmR+n9sr5d0naQ9Z/narKRZSVquFS1MAzCJxifKbF8k6RlJ9yX5/JtfT7ItyUySmWkta3MjgDE0itr2tOaDfjLJs91OAnA+mpz9tqTHJB1J8nD3kwCcjyZH6usl3S1pk+0Do7ebO94FYEKLnihLsluSl2ALgBbwE2VAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFNrnq53PZe22/aPmT7oaUYBmAyi14gT9JJSZuSnBhdp3q37T8nea3jbQAm0OSql5F0YvTh9OgtXY4CMLlGz6ltT9k+IOm4pJ1J9pzlNrO252zPfaWTbe8E0FCjqJOcTrJB0lpJG21fe5bbbEsyk2RmWsva3gmgobHOfif5TNIuSVu6mQPgfDU5+73S9qWj9y+QdKOkt7seBmAyTc5+r5L0e9tTmv9L4OkkL3Q7C8Ckmpz9PijpuiXYAqAF/EQZUAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxTT5La3vvB0fHeh7AsY0xP9nm1dv6HtCIxypgWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiimcdSjC8+/YZuL4wEDNs6RequkI10NAdCORlHbXivpFkmPdjsHwPlqeqR+RNL9kr7+thvYnrU9Z3vuK51sZRyA8S0ate1bJR1Psu9ct0uyLclMkplpLWttIIDxNDlSXy/pNtsfSNouaZPtJzpdBWBii0ad5MEka5Osl3SHpJeS3NX5MgAT4fvUQDFjvURwkpclvdzJEgCt4EgNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFDPWb2l9V21evaHvCcCS4UgNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDGNfvVydG3qLySdlnQqyUyXowBMbpzfp/55kn92tgRAK3j4DRTTNOpI+qvtfbZnz3YD27O252zPfaWT7S0EMJamD79vSHLM9o8k7bT9dpJXFt4gyTZJ2yTpEl+WlncCaKjRkTrJsdE/j0t6TtLGLkcBmNyiUdu+0PbFZ96X9AtJb3U9DMBkmjz8vkLSc7bP3P6PSf7S6SoAE1s06iTvS/rJEmwB0AK+pQUUQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxTtp/PQPbn0j6Rwt3dbmkIb0uGnvObWh7pOFtamvPj5OsPNsXOom6LbbnhvTKpew5t6HtkYa3aSn28PAbKIaogWKGHvW2vgd8A3vObWh7pOFt6nzPoJ9TAxjf0I/UAMZE1EAxg4za9hbb79h+z/YDA9jzuO3jtgfx0si219neZfuw7UO2t/a8Z7ntvbbfHO15qM89Z9iesv2G7Rf63iLNX2jS9t9tH7A919m/Z2jPqW1PSXpX0o2Sjkp6XdKdSQ73uOlnkk5I+kOSa/vasWDPKkmrkuwfvSb7Pkm/7Ou/kedfP/rCJCdsT0vaLWlrktf62LNg168lzUi6JMmtfW4Z7flA0kzXF5oc4pF6o6T3kryf5EtJ2yXd3ueg0SWGPu1zw0JJPk6yf/T+F5KOSFrT454kOTH6cHr01uvRwvZaSbdIerTPHX0YYtRrJH244OOj6vEP7NDZXi/pOkl7et4xZfuApOOSdibpdY+kRyTdL+nrnncstOiFJtswxKjRkO2LJD0j6b4kn/e5JcnpJBskrZW00XZvT1Ns3yrpeJJ9fW34Fjck+amkmyT9avS0rnVDjPqYpHULPl47+hwWGD13fUbSk0me7XvPGUk+k7RL0pYeZ1wv6bbRc9jtkjbZfqLHPZKW7kKTQ4z6dUlX2b7S9g8k3SHp+Z43DcroxNRjko4keXgAe1bavnT0/gWaP8n5dl97kjyYZG2S9Zr/8/NSkrv62iMt7YUmBxd1klOS7pW0Q/MngJ5OcqjPTbafkvSqpKttH7V9T597NH8kulvzR6ADo7ebe9yzStIu2wc1/5fyziSD+DbSgFwhabftNyXtlfSnri40ObhvaQE4P4M7UgM4P0QNFEPUQDFEDRRD1EAxRA0UQ9RAMf8Fom9IVnxPiHcAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 13%|█▎        | 1300/10000 [01:09<07:15, 20.00it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.08809570223093033 logZ = -4.015066146850586 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 14%|█▍        | 1399/10000 [01:14<06:57, 20.61it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 24.32484245300293 logZ = -3.6338181495666504 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 15%|█▌        | 1500/10000 [01:20<06:54, 20.51it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 6.3147196769714355 logZ = -2.863300323486328 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 16%|█▌        | 1600/10000 [01:25<06:57, 20.13it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 2.0999598503112793 logZ = -1.9609133005142212 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 17%|█▋        | 1700/10000 [01:30<07:21, 18.79it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 3.051518201828003 logZ = -1.0194141864776611 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 18%|█▊        | 1799/10000 [01:36<07:34, 18.06it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 6.432326316833496 logZ = -0.3228550851345062 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 19%|█▉        | 1899/10000 [01:42<06:56, 19.45it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.018489688634872437 logZ = 0.7707247734069824 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 20%|█▉        | 1999/10000 [01:49<16:38,  8.01it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.5152246952056885 logZ = 1.4689204692840576 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 21%|██        | 2099/10000 [01:57<13:54,  9.47it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.6809870600700378 logZ = 1.5550636053085327 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 22%|██▏       | 2199/10000 [02:05<07:42, 16.86it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.36317571997642517 logZ = 1.6540247201919556 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 23%|██▎       | 2299/10000 [02:11<07:14, 17.74it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.7421892881393433 logZ = 1.956602692604065 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 24%|██▍       | 2399/10000 [02:17<07:35, 16.70it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.49145209789276123 logZ = 1.855776309967041 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 25%|██▌       | 2500/10000 [02:24<12:20, 10.13it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.41346263885498047 logZ = 1.8781675100326538 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 26%|██▌       | 2599/10000 [02:32<08:46, 14.07it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.4666852056980133 logZ = 1.8828063011169434 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 27%|██▋       | 2699/10000 [02:38<12:08, 10.02it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.9387232661247253 logZ = 1.9532827138900757 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 28%|██▊       | 2798/10000 [02:45<07:10, 16.72it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.943284034729004 logZ = 1.8831965923309326 R = 1.494533e-05\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 29%|██▉       | 2899/10000 [02:50<07:10, 16.51it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.7061401009559631 logZ = 1.9377720355987549 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 30%|███       | 3000/10000 [02:57<06:51, 17.00it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.05804408714175224 logZ = 1.993329405784607 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 31%|███       | 3099/10000 [03:02<06:57, 16.52it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.0305393934249878 logZ = 2.057600736618042 R = 5.7777417e-08\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 32%|███▏      | 3199/10000 [03:08<06:21, 17.81it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.5999446511268616 logZ = 1.6631559133529663 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 33%|███▎      | 3300/10000 [03:15<07:28, 14.95it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.5750818252563477 logZ = 1.8955271244049072 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 34%|███▍      | 3400/10000 [03:21<05:53, 18.67it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.09065576642751694 logZ = 2.0919559001922607 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 35%|███▌      | 3500/10000 [03:27<06:20, 17.10it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.5144419074058533 logZ = 1.971221923828125 R = 1.494533e-05\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 36%|███▌      | 3600/10000 [03:33<06:18, 16.90it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.01736217364668846 logZ = 1.980860710144043 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 37%|███▋      | 3700/10000 [03:39<06:06, 17.17it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.353585422039032 logZ = 1.971509337425232 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 38%|███▊      | 3799/10000 [03:45<06:07, 16.85it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.018329495564103127 logZ = 2.122718572616577 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAJk0lEQVR4nO3dzYtdhR3G8edpOiYaFSm1kmRC40IEERrLkC6UQlNs4gvapYKuhNkoRFoQXfoPiJtuBpW2aA2CCmJt01AjEtDoJEZrEi1BLCYKaRHRFBoTfbqYmzKVmDn35pw5Jz++Hxicl8v1IeSbc++ZmXucRADq+E7fAwC0i6iBYogaKIaogWKIGijmu13c6QVemVVa3cVdA5D0H/1bX+aEz/S1TqJepdX6iX/exV0DkLQnf/3Wr/HwGyiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiimUdS2t9p+3/Zh2w92PQrA5JaM2vYKSb+RdJOkayTdafuarocBmEyTI/UmSYeTfJDkS0nbJd3e7SwAk2oS9TpJHy36+Mjoc//H9qztedvzJ3WirX0AxtTaibIkc0lmksxMaWVbdwtgTE2iPipp/aKPp0efAzBATaJ+U9JVtq+0fYGkOyS90O0sAJNa8oUHk5yyfZ+kHZJWSHoiyYHOlwGYSKNXE03ykqSXOt4CoAX8RBlQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFNPqFDrRvx8f7+54waFvWbux7wnmLIzVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UMySUdt+wvYx2+8uxyAA56bJkfq3krZ2vANAS5aMOsmrkj5dhi0AWtDayxnZnpU0K0mrdFFbdwtgTK2dKEsyl2QmycyUVrZ1twDGxNlvoBiiBopp8i2tpyW9Julq20ds39P9LACTWvJEWZI7l2MIgHbw8BsohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKKbJBfLW295l+6DtA7a3LccwAJNZ8gJ5kk5J+nWSfbYvkbTX9s4kBzveBmACSx6pk3ySZN/o/S8kHZK0ruthACbT5Ej9P7Y3SLpO0p4zfG1W0qwkrdJFLUwDMInGJ8psXyzpWUn3J/n8m19PMpdkJsnMlFa2uRHAGBpFbXtKC0E/leS5bicBOBdNzn5b0uOSDiV5pPtJAM5FkyP19ZLulrTZ9v7R280d7wIwoSVPlCXZLcnLsAVAC/iJMqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBooZ65VPUNeWtRv7noCWcKQGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoJgmV71cZfsN22/bPmD74eUYBmAyTX6f+oSkzUmOj65Tvdv2n5K83vE2ABNoctXLSDo++nBq9JYuRwGYXKPn1LZX2N4v6ZiknUn2nOE2s7bnbc+f1Im2dwJoqFHUSb5KslHStKRNtq89w23mkswkmZnSyrZ3AmhorLPfST6TtEvS1m7mADhXTc5+X277stH7F0q6UdJ7XQ8DMJkmZ7/XSPqd7RVa+EfgmSQvdjsLwKSanP1+R9J1y7AFQAv4iTKgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKafJbWue9HR/v73vC4A3tz2jL2o19TzhvcaQGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoJjGUY8uPP+WbS6OBwzYOEfqbZIOdTUEQDsaRW17WtItkh7rdg6Ac9X0SP2opAckff1tN7A9a3ve9vxJnWhlHIDxLRm17VslHUuy92y3SzKXZCbJzJRWtjYQwHiaHKmvl3Sb7Q8lbZe02faTna4CMLElo07yUJLpJBsk3SHp5SR3db4MwET4PjVQzFgvEZzkFUmvdLIEQCs4UgPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFjPVbWuerLWs39j0BWDYcqYFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBooptGvXo6uTf2FpK8knUoy0+UoAJMb5/epf5bkX50tAdAKHn4DxTSNOpL+Ynuv7dkz3cD2rO152/MndaK9hQDG0vTh9w1Jjtr+gaSdtt9L8uriGySZkzQnSZf6e2l5J4CGGh2pkxwd/feYpOclbepyFIDJLRm17dW2Lzn9vqRfSHq362EAJtPk4fcVkp63ffr2f0jy505XAZjYklEn+UDSj5ZhC4AW8C0toBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGinHS/usZ2P6npH+0cFfflzSk10Vjz9kNbY80vE1t7flhksvP9IVOom6L7fkhvXIpe85uaHuk4W1ajj08/AaKIWqgmKFHPdf3gG9gz9kNbY80vE2d7xn0c2oA4xv6kRrAmIgaKGaQUdveavt924dtPziAPU/YPmZ7EC+NbHu97V22D9o+YHtbz3tW2X7D9tujPQ/3uec02ytsv2X7xb63SAsXmrT9N9v7bc939v8Z2nNq2ysk/V3SjZKOSHpT0p1JDva46aeSjkv6fZJr+9qxaM8aSWuS7Bu9JvteSb/s68/IC68fvTrJcdtTknZL2pbk9T72LNr1K0kzki5NcmufW0Z7PpQ00/WFJod4pN4k6XCSD5J8KWm7pNv7HDS6xNCnfW5YLMknSfaN3v9C0iFJ63rckyTHRx9Ojd56PVrYnpZ0i6TH+tzRhyFGvU7SR4s+PqIe/8IOne0Nkq6TtKfnHSts75d0TNLOJL3ukfSopAckfd3zjsWWvNBkG4YYNRqyfbGkZyXdn+TzPrck+SrJRknTkjbZ7u1piu1bJR1LsrevDd/ihiQ/lnSTpHtHT+taN8Soj0pav+jj6dHnsMjoueuzkp5K8lzfe05L8pmkXZK29jjjekm3jZ7Dbpe02faTPe6RtHwXmhxi1G9Kusr2lbYvkHSHpBd63jQooxNTj0s6lOSRAey53PZlo/cv1MJJzvf62pPkoSTTSTZo4e/Py0nu6muPtLwXmhxc1ElOSbpP0g4tnAB6JsmBPjfZflrSa5Kutn3E9j197tHCkehuLRyB9o/ebu5xzxpJu2y/o4V/lHcmGcS3kQbkCkm7bb8t6Q1Jf+zqQpOD+5YWgHMzuCM1gHND1EAxRA0UQ9RAMUQNFEPUQDFEDRTzXzG2Qb0oPKINAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 39%|███▉      | 3899/10000 [03:51<06:16, 16.20it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.2346440851688385 logZ = 1.9927643537521362 R = 9.292488e-07\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 40%|████      | 4000/10000 [03:57<06:09, 16.24it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.7721426486968994 logZ = 1.9002652168273926 R = 1.494533e-05\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 41%|████      | 4100/10000 [04:03<05:17, 18.60it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.0011555254459381104 logZ = 1.9683557748794556 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 42%|████▏     | 4200/10000 [04:09<05:33, 17.39it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.2977183163166046 logZ = 2.0399985313415527 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 43%|████▎     | 4299/10000 [04:15<05:31, 17.21it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.24043473601341248 logZ = 1.9770630598068237 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 44%|████▍     | 4399/10000 [04:22<05:51, 15.94it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.04254543408751488 logZ = 1.9164940118789673 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 45%|████▍     | 4499/10000 [04:28<05:10, 17.72it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.00010994883632520214 logZ = 2.156998634338379 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 46%|████▌     | 4599/10000 [04:34<04:51, 18.55it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.012078847736120224 logZ = 2.038860321044922 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 47%|████▋     | 4700/10000 [04:40<05:09, 17.12it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.9808588027954102 logZ = 1.7169541120529175 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 48%|████▊     | 4800/10000 [04:46<04:42, 18.41it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.3925711214542389 logZ = 1.864484190940857 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 49%|████▉     | 4899/10000 [04:52<04:30, 18.84it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.0991688966751099 logZ = 1.957149863243103 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 50%|█████     | 5000/10000 [04:58<04:35, 18.16it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.002156264381483197 logZ = 2.01462459564209 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 51%|█████     | 5100/10000 [05:04<04:43, 17.29it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.39482223987579346 logZ = 1.9922047853469849 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 52%|█████▏    | 5200/10000 [05:10<04:25, 18.06it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 2.2336696929414757e-05 logZ = 1.9165253639221191 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 53%|█████▎    | 5299/10000 [05:16<04:54, 15.98it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.07862699031829834 logZ = 2.1438496112823486 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 54%|█████▍    | 5399/10000 [05:22<04:22, 17.50it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.7122266292572021 logZ = 2.019627094268799 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 55%|█████▍    | 5499/10000 [05:28<04:44, 15.80it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.6549355387687683 logZ = 1.7839223146438599 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 56%|█████▌    | 5599/10000 [05:34<04:28, 16.42it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.24828775227069855 logZ = 1.88702392578125 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 57%|█████▋    | 5700/10000 [05:40<04:18, 16.64it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.13230225443840027 logZ = 1.7966036796569824 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAJqklEQVR4nO3dzYtdhR3G8edpOiYaLVJqJcmExoUVRNpYhnShFJpiE1/QLhV0JcymQqQF0aX/gLjpZlBpi9YgqCDWNg01IgGNTmK0JlEJYjFRSIuIpqUx0aeLuYFRYubcm3PmHH98PzA4L5frQ8g3594zM/c4iQDU8a2+BwBoF1EDxRA1UAxRA8UQNVDMt7u40/O8Mqu0uou7Rkd++KP/9j3hS95544K+Jwza//QffZYTPtPXOol6lVbrp/5FF3eNjuzYsb/vCV+yZe3GvicM2p78/Wu/xsNvoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoJhGUdveavtt24dt39v1KACTWzJq2ysk/U7S9ZKulHSb7Su7HgZgMk2O1JskHU7ybpLPJG2XdEu3swBMqknU6yS9v+jjI6PPfYntWdvztudP6kRb+wCMqbUTZUnmkswkmZnSyrbuFsCYmkR9VNL6RR9Pjz4HYICaRP2qpMttX2b7PEm3Snqm21kAJrXkCw8mOWX7Lkk7JK2Q9EiSA50vAzCRRq8mmuQ5Sc91vAVAC/iJMqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBopp9AsdaN+OD/b3PWHQhvjns2Xtxr4nNMKRGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBooZsmobT9i+5jtN5djEIBz0+RI/XtJWzveAaAlS0ad5EVJHy3DFgAtaO3ljGzPSpqVpFW6oK27BTCm1k6UJZlLMpNkZkor27pbAGPi7DdQDFEDxTT5ltbjkl6SdIXtI7bv7H4WgEkteaIsyW3LMQRAO3j4DRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDGtvfIJ0KYtazf2PeEbiyM1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8U0uUDeetu7bB+0fcD2tuUYBmAyTX6f+pSk3ybZZ/siSXtt70xysONtACaw5JE6yYdJ9o3e/1TSIUnruh4GYDJjvfKJ7Q2Srpa05wxfm5U0K0mrdEEL0wBMovGJMtsXSnpS0t1JPvnq15PMJZlJMjOllW1uBDCGRlHbntJC0I8learbSQDORZOz35b0sKRDSR7ofhKAc9HkSH2NpDskbba9f/R2Q8e7AExoyRNlSXZL8jJsAdACfqIMKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiimyVUvV9l+xfbrtg/Yvn85hgGYzJIXyJN0QtLmJMdH16nebfsvSV7ueBuACTS56mUkHR99ODV6S5ejAEyu0XNq2yts75d0TNLOJHvOcJtZ2/O250/qRNs7ATTUKOoknyfZKGla0ibbV53hNnNJZpLMTGll2zsBNDTW2e8kH0vaJWlrN3MAnKsmZ78vsX3x6P3zJV0n6a2uhwGYTJOz32sk/cH2Ci38I/BEkme7nQVgUk3Ofr8h6epl2AKgBfxEGVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8U0+S2tb7wdH+zvewKwbDhSA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVBM46hHF55/zTYXxwMGbJwj9TZJh7oaAqAdjaK2PS3pRkkPdTsHwLlqeqR+UNI9kr74uhvYnrU9b3v+pE60Mg7A+JaM2vZNko4l2Xu22yWZSzKTZGZKK1sbCGA8TY7U10i62fZ7krZL2mz70U5XAZjYklEnuS/JdJINkm6V9HyS2ztfBmAifJ8aKGaslwhO8oKkFzpZAqAVHKmBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYsb6La1vqi1rN/Y9AVg2HKmBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKKbRr16Ork39qaTPJZ1KMtPlKACTG+f3qX+e5N+dLQHQCh5+A8U0jTqS/mZ7r+3ZM93A9qztedvzJ3WivYUAxtL04fe1SY7a/r6knbbfSvLi4hskmZM0J0nf8XfT8k4ADTU6Uic5OvrvMUlPS9rU5SgAk1syaturbV90+n1Jv5T0ZtfDAEymycPvSyU9bfv07f+U5K+drgIwsSWjTvKupB8vwxYALeBbWkAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRTjpP3XM7D9L0n/bOGuvidpSK+Lxp6zG9oeaXib2trzgySXnOkLnUTdFtvzQ3rlUvac3dD2SMPbtBx7ePgNFEPUQDFDj3qu7wFfwZ6zG9oeaXibOt8z6OfUAMY39CM1gDERNVDMIKO2vdX227YP2753AHsesX3M9iBeGtn2etu7bB+0fcD2tp73rLL9iu3XR3vu73PPabZX2H7N9rN9b5EWLjRp+x+299ue7+z/M7Tn1LZXSHpH0nWSjkh6VdJtSQ72uOlnko5L+mOSq/rasWjPGklrkuwbvSb7Xkm/6uvPyAuvH706yXHbU5J2S9qW5OU+9iza9RtJM5K+k+SmPreM9rwnaabrC00O8Ui9SdLhJO8m+UzSdkm39DlodImhj/rcsFiSD5PsG73/qaRDktb1uCdJjo8+nBq99Xq0sD0t6UZJD/W5ow9DjHqdpPcXfXxEPf6FHTrbGyRdLWlPzztW2N4v6ZiknUl63SPpQUn3SPqi5x2LLXmhyTYMMWo0ZPtCSU9KujvJJ31uSfJ5ko2SpiVtst3b0xTbN0k6lmRvXxu+xrVJfiLpekm/Hj2ta90Qoz4qaf2ij6dHn8Mio+euT0p6LMlTfe85LcnHknZJ2trjjGsk3Tx6Drtd0mbbj/a4R9LyXWhyiFG/Kuly25fZPk/SrZKe6XnToIxOTD0s6VCSBwaw5xLbF4/eP18LJznf6mtPkvuSTCfZoIW/P88nub2vPdLyXmhycFEnOSXpLkk7tHAC6IkkB/rcZPtxSS9JusL2Edt39rlHC0eiO7RwBNo/eruhxz1rJO2y/YYW/lHemWQQ30YakEsl7bb9uqRXJP25qwtNDu5bWgDOzeCO1ADODVEDxRA1UAxRA8UQNVAMUQPFEDVQzP8BX8VG/3312fcAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 58%|█████▊    | 5800/10000 [05:46<04:16, 16.39it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.019952476024627686 logZ = 1.8560327291488647 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAJkElEQVR4nO3dzYtdhR3G8edpOiYaFSm1EjOhcSGCCI1lSBdKoSnW+IJ2qaArYTYVIi2ILv0HxE03g0pbtAZBBbG2aagRCWh0EkdrEi0hWEwU0iKiKTQm+nQxN2WUmDn35pw5xx/fDwzOy+XmQfLNuffMzD1OIgB1fKfvAQDaRdRAMUQNFEPUQDFEDRTz3S7u9Dyvzhqt7eKuAUj6r/6jz3PCZ/paJ1Gv0Vr9xD/v4q4BSNqTv33j13j4DRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRTTKGrbW22/Z/uQ7Qe6HgVgcstGbXuVpN9KuknS1ZLutH1118MATKbJkXqzpENJDif5XNJ2Sbd3OwvApJpEvV7SB0s+PjL63FfYnrU9b3v+pE60tQ/AmFo7UZZkLslMkpkprW7rbgGMqUnURyVtWPLx9OhzAAaoSdRvSLrS9hW2z5N0h6Tnu50FYFLLvvBgklO275W0Q9IqSY8n2d/5MgATafRqoklelPRix1sAtICfKAOKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqCYRr/Qgfbt+HCh7wmDduPlm/qe8K3FkRoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKGbZqG0/bvuY7XdWYhCAc9PkSP07SVs73gGgJctGneQVSR+vwBYALWjt5Yxsz0qalaQ1uqCtuwUwptZOlCWZSzKTZGZKq9u6WwBj4uw3UAxRA8U0+ZbWU5JelXSV7SO27+l+FoBJLXuiLMmdKzEEQDt4+A0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFNPkAnkbbO+yfcD2ftvbVmIYgMkse4E8Sack/SbJPtsXSdpre2eSAx1vAzCBZY/UST5Ksm/0/meSDkpa3/UwAJNpcqT+P9sbJV0rac8ZvjYraVaS1uiCFqYBmETjE2W2L5T0jKT7knz69a8nmUsyk2RmSqvb3AhgDI2itj2lxaCfTPJst5MAnIsmZ78t6TFJB5M83P0kAOeiyZH6Okl3S9pie2H0dnPHuwBMaNkTZUl2S/IKbAHQAn6iDCiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWLGeuUTtOfGyzf1PeErdny40PcEtIQjNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFNLnq5Rrbr9t+y/Z+2w+txDAAk2ny+9QnJG1Jcnx0nerdtv+c5LWOtwGYQJOrXkbS8dGHU6O3dDkKwOQaPae2vcr2gqRjknYm2XOG28zanrc9f1In2t4JoKFGUSf5IskmSdOSNtu+5gy3mUsyk2RmSqvb3gmgobHOfif5RNIuSVu7mQPgXDU5+32p7UtG758v6QZJ73Y9DMBkmpz9Xifp97ZXafEfgaeTvNDtLACTanL2+21J167AFgAt4CfKgGKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKKbJb2l96+34cKHvCcCK4UgNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDGNox5deP5N21wcDxiwcY7U2yQd7GoIgHY0itr2tKRbJD3a7RwA56rpkfoRSfdL+vKbbmB71va87fmTOtHKOADjWzZq27dKOpZk79lul2QuyUySmSmtbm0ggPE0OVJfJ+k22+9L2i5pi+0nOl0FYGLLRp3kwSTTSTZKukPSS0nu6nwZgInwfWqgmLFeIjjJy5Je7mQJgFZwpAaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKGeu3tL6tbrx8U98TgBXDkRoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYhr96uXo2tSfSfpC0qkkM12OAjC5cX6f+mdJ/t3ZEgCt4OE3UEzTqCPpr7b32p490w1sz9qetz1/UifaWwhgLE0ffl+f5KjtH0jaafvdJK8svUGSOUlzknSxv5eWdwJoqNGROsnR0X+PSXpO0uYuRwGY3LJR215r+6LT70v6haR3uh4GYDJNHn5fJuk526dv/8ckf+l0FYCJLRt1ksOSfrQCWwC0gG9pAcUQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UIyT9l/PwPa/JP2zhbv6vqQhvS4ae85uaHuk4W1qa88Pk1x6pi90EnVbbM8P6ZVL2XN2Q9sjDW/TSuzh4TdQDFEDxQw96rm+B3wNe85uaHuk4W3qfM+gn1MDGN/Qj9QAxkTUQDGDjNr2Vtvv2T5k+4EB7Hnc9jHbg3hpZNsbbO+yfcD2ftvbet6zxvbrtt8a7Xmozz2n2V5l+03bL/S9RVq80KTtv9tesD3f2Z8ztOfUtldJ+oekGyQdkfSGpDuTHOhx008lHZf0hyTX9LVjyZ51ktYl2Td6Tfa9kn7Z1/8jL75+9Nokx21PSdotaVuS1/rYs2TXryXNSLo4ya19bhnteV/STNcXmhzikXqzpENJDif5XNJ2Sbf3OWh0iaGP+9ywVJKPkuwbvf+ZpIOS1ve4J0mOjz6cGr31erSwPS3pFkmP9rmjD0OMer2kD5Z8fEQ9/oUdOtsbJV0raU/PO1bZXpB0TNLOJL3ukfSIpPslfdnzjqWWvdBkG4YYNRqyfaGkZyTdl+TTPrck+SLJJknTkjbb7u1piu1bJR1LsrevDd/g+iQ/lnSTpF+Nnta1bohRH5W0YcnH06PPYYnRc9dnJD2Z5Nm+95yW5BNJuyRt7XHGdZJuGz2H3S5pi+0netwjaeUuNDnEqN+QdKXtK2yfJ+kOSc/3vGlQRiemHpN0MMnDA9hzqe1LRu+fr8WTnO/2tSfJg0mmk2zU4t+fl5Lc1dceaWUvNDm4qJOcknSvpB1aPAH0dJL9fW6y/ZSkVyVdZfuI7Xv63KPFI9HdWjwCLYzebu5xzzpJu2y/rcV/lHcmGcS3kQbkMkm7bb8l6XVJf+rqQpOD+5YWgHMzuCM1gHND1EAxRA0UQ9RAMUQNFEPUQDFEDRTzPzMhQSw04JlaAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 59%|█████▉    | 5900/10000 [05:52<03:56, 17.34it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.14374490082263947 logZ = 1.8825818300247192 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 60%|█████▉    | 5999/10000 [05:58<03:21, 19.83it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.10560039430856705 logZ = 1.7303539514541626 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6100/10000 [06:05<03:55, 16.53it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.13774384558200836 logZ = 2.0305707454681396 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6200/10000 [06:11<03:32, 17.85it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.03627481684088707 logZ = 2.1307806968688965 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6298/10000 [06:18<03:30, 17.55it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.05710337683558464 logZ = 1.4079080820083618 R = 1.494533e-05\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6399/10000 [06:24<03:44, 16.04it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.30454009771347046 logZ = 1.6067792177200317 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▌   | 6500/10000 [06:31<03:34, 16.31it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.05676795542240143 logZ = 1.9894821643829346 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6599/10000 [06:37<03:26, 16.49it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.07968252152204514 logZ = 1.7924302816390991 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6700/10000 [06:43<03:27, 15.94it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.062090277671814 logZ = 2.128617286682129 R = 1.494533e-05\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6800/10000 [06:49<03:08, 16.96it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.010331026278436184 logZ = 1.7289063930511475 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6900/10000 [06:56<05:30,  9.37it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.17409968376159668 logZ = 1.6783418655395508 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|███████   | 7000/10000 [07:03<02:55, 17.08it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.2784833312034607 logZ = 2.1689364910125732 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7100/10000 [07:09<02:49, 17.14it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 3.379459857940674 logZ = 1.6436216831207275 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7199/10000 [07:15<02:38, 17.62it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.1463485211133957 logZ = 1.8045070171356201 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7300/10000 [07:23<02:47, 16.13it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.10468752682209015 logZ = 1.6259342432022095 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7400/10000 [07:29<02:30, 17.29it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.19831766188144684 logZ = 2.025891065597534 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▌  | 7500/10000 [07:35<02:29, 16.77it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.19503064453601837 logZ = 2.200392961502075 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7599/10000 [07:42<02:17, 17.46it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.17179274559021 logZ = 2.0357346534729004 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAJpUlEQVR4nO3dzYtdhR3G8efpdEw0KtI2lSQTGhdWEGljGdKFUmiKTXxBu1TQlTCbCpEWRJf+A+Kmm0GlLVqDoIJY2zTUiAQ0OomjNS9KEIuJQmxFNC2NJj5dzE2ZSsyce3POnOuP7wcG5+Vy8yD55tx7ZuYeJxGAOr7R9wAA7SJqoBiiBoohaqAYogaK+WYXd3qeV2SlVnVx1wAk/Uf/0mc54TN9rZOoV2qVfuyfdXHXACTtyV+/8ms8/AaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKaRS17a2237J92Pa9XY8CMLolo7Y9Iek3kq6XdKWk22xf2fUwAKNpcqTeJOlwkneSfCZpu6Rbup0FYFRNol4n6b1FHx8ZfO7/2J6xPWd77nOdaGsfgCG1dqIsyWyS6STTk1rR1t0CGFKTqI9KWr/o46nB5wCMoSZRvyrpctuX2T5P0q2Snul2FoBRLfnCg0lO2r5L0g5JE5IeSbK/82UARtLo1USTPCfpuY63AGgBP1EGFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMY1+oQPt2/H+fN8TxtqWtRv7nvC1xZEaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGihmyahtP2L7mO03l2MQgHPT5Ej9W0lbO94BoCVLRp3kRUkfLcMWAC1o7eWMbM9ImpGklbqgrbsFMKTWTpQlmU0ynWR6UivaulsAQ+LsN1AMUQPFNPmW1uOSXpJ0he0jtu/sfhaAUS15oizJbcsxBEA7ePgNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRTT5AJ5623vsn3A9n7b25ZjGIDRLHmBPEknJf06yT7bF0naa3tnkgMdbwMwgiWP1Ek+SLJv8P6nkg5KWtf1MACjaXKk/h/bGyRdLWnPGb42I2lGklbqghamARhF4xNlti+U9KSku5N88uWvJ5lNMp1kelIr2twIYAiNorY9qYWgH0vyVLeTAJyLJme/LelhSQeTPND9JADnosmR+hpJd0jabHt+8HZDx7sAjGjJE2VJdkvyMmwB0AJ+ogwohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFihnrlk6a+/4N/a8eO+S7ueiRb1m7sewKwbDhSA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVBMk6terrT9iu3Xbe+3ff9yDAMwmia/T31C0uYkxwfXqd5t+09JXu54G4ARNLnqZSQdH3w4OXhLl6MAjK7Rc2rbE7bnJR2TtDPJnjPcZsb2nO25D/95qu2dABpqFHWSU0k2SpqStMn2VWe4zWyS6STTq7890fZOAA0NdfY7yceSdkna2s0cAOeqydnv1bYvGbx/vqTrJB3qehiA0TQ5+71G0u9sT2jhH4Enkjzb7SwAo2py9vsNSVcvwxYALeAnyoBiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiimyW9pDe3tNy7QlrUbu7jrkex4f77vCcCy4UgNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDFEDRRD1EAxRA0UQ9RAMUQNFEPUQDGNox5ceP4121wcDxhjwxypt0k62NUQAO1oFLXtKUk3Snqo2zkAzlXTI/WDku6R9MVX3cD2jO0523Of60Qr4wAMb8mobd8k6ViSvWe7XZLZJNNJpie1orWBAIbT5Eh9jaSbbb8rabukzbYf7XQVgJEtGXWS+5JMJdkg6VZJzye5vfNlAEbC96mBYoZ6ieAkL0h6oZMlAFrBkRoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBooZqjf0vq62rJ2Y98TgGXDkRoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYhr96uXg2tSfSjol6WSS6S5HARjdML9P/dMk/+hsCYBW8PAbKKZp1JH0F9t7bc+c6Qa2Z2zP2Z77XCfaWwhgKE0ffl+b5Kjt70raaftQkhcX3yDJrKRZSbrY30rLOwE01OhIneTo4L/HJD0taVOXowCMbsmoba+yfdHp9yX9XNKbXQ8DMJomD78vlfS07dO3/0OSP3e6CsDIlow6yTuSfrgMWwC0gG9pAcUQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UIyT9l/PwPaHkv7ewl19R9I4vS4ae85u3PZI47eprT3fS7L6TF/oJOq22J4bp1cuZc/Zjdseafw2LcceHn4DxRA1UMy4Rz3b94AvYc/Zjdseafw2db5nrJ9TAxjeuB+pAQyJqIFixjJq21ttv2X7sO17x2DPI7aP2R6Ll0a2vd72LtsHbO+3va3nPSttv2L79cGe+/vcc5rtCduv2X627y3SwoUmbf/N9rztuc7+nHF7Tm17QtLbkq6TdETSq5JuS3Kgx00/kXRc0u+TXNXXjkV71khak2Tf4DXZ90r6RV//j7zw+tGrkhy3PSlpt6RtSV7uY8+iXb+SNC3p4iQ39bllsOddSdNdX2hyHI/UmyQdTvJOks8kbZd0S5+DBpcY+qjPDYsl+SDJvsH7n0o6KGldj3uS5Pjgw8nBW69HC9tTkm6U9FCfO/owjlGvk/Teoo+PqMe/sOPO9gZJV0va0/OOCdvzko5J2pmk1z2SHpR0j6Qvet6x2JIXmmzDOEaNhmxfKOlJSXcn+aTPLUlOJdkoaUrSJtu9PU2xfZOkY0n29rXhK1yb5EeSrpf0y8HTutaNY9RHJa1f9PHU4HNYZPDc9UlJjyV5qu89pyX5WNIuSVt7nHGNpJsHz2G3S9ps+9Ee90havgtNjmPUr0q63PZlts+TdKukZ3reNFYGJ6YelnQwyQNjsGe17UsG75+vhZOch/rak+S+JFNJNmjh78/zSW7va4+0vBeaHLuok5yUdJekHVo4AfREkv19brL9uKSXJF1h+4jtO/vco4Uj0R1aOALND95u6HHPGkm7bL+hhX+UdyYZi28jjZFLJe22/bqkVyT9sasLTY7dt7QAnJuxO1IDODdEDRRD1EAxRA0UQ9RAMUQNFEPUQDH/BbKQRZCOC/mOAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7700/10000 [07:48<02:19, 16.53it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.14211702346801758 logZ = 1.7746753692626953 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7800/10000 [07:54<02:07, 17.21it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.018170539289712906 logZ = 2.140127658843994 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7899/10000 [08:00<02:08, 16.29it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.30930331349372864 logZ = 2.543339967727661 R = 2.2336291e-10\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|████████  | 8000/10000 [08:08<01:56, 17.18it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.2233339548110962 logZ = 1.9777226448059082 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8099/10000 [08:14<01:51, 17.06it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.06820723414421082 logZ = 2.096052646636963 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8200/10000 [08:21<01:56, 15.46it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 1.1995435953140259 logZ = 1.9258085489273071 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8299/10000 [08:29<01:39, 17.02it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.08450424671173096 logZ = 1.368445634841919 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8399/10000 [08:36<02:28, 10.81it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.02852088212966919 logZ = 1.6641992330551147 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▍ | 8499/10000 [08:44<01:28, 16.91it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.4378869831562042 logZ = 1.834989070892334 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8600/10000 [08:53<02:01, 11.54it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.0833466500043869 logZ = 1.8649271726608276 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8699/10000 [09:01<02:11,  9.89it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.006436355412006378 logZ = 1.8378021717071533 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8799/10000 [09:08<01:13, 16.41it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 5.784285068511963 logZ = 1.905197024345398 R = 5.7777417e-08\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8899/10000 [09:14<01:08, 16.03it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.09091394394636154 logZ = 2.0152015686035156 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|█████████ | 9000/10000 [09:20<00:58, 16.97it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.004173632245510817 logZ = 1.9762988090515137 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9100/10000 [09:27<00:49, 18.17it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.05510343611240387 logZ = 1.8547438383102417 R = 0.003865919\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9200/10000 [09:33<00:49, 16.02it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.12681694328784943 logZ = 2.0469653606414795 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9300/10000 [09:39<00:39, 17.71it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.48829758167266846 logZ = 2.1457130908966064 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9400/10000 [09:45<00:36, 16.47it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.30066978931427 logZ = 2.1759448051452637 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▍| 9499/10000 [09:51<00:31, 15.76it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.0024971975944936275 logZ = 1.929572343826294 R = 1.0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPUAAAD4CAYAAAA0L6C7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAJgElEQVR4nO3dzYtdhR3G8edpOiYaLVJqJWZC40IEERrLkC6UQlOs8QXtUkFXwmwqRFoQXfoPiJtuBpW2aA2CCmJt01AjEtDoJEZrEi1BLCYK0yKiKTQm+nQxN2UqMXPunXPmHH98PzA4L5ebB8k3594zM/c4iQDU8a2+BwBoF1EDxRA1UAxRA8UQNVDMt7u40/O8Nuu0vou7BiDpP/q3Ps9Jn+1rnUS9Tuv1Y/+si7sGIGlf/vq1X+PhN1AMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVBMo6htb7f9ru2jtu/vehSAyS0bte01kn4j6UZJV0m6w/ZVXQ8DMJkmR+qtko4meS/J55J2Srqt21kAJtUk6o2SPljy8bHR5/6P7Vnb87bnT+lkW/sAjKm1E2VJ5pLMJJmZ0tq27hbAmJpEfVzSpiUfT48+B2CAmkT9uqQrbF9u+zxJt0t6rttZACa17AsPJjlt+x5JuyStkfRYkkOdLwMwkUavJprkBUkvdLwFQAv4iTKgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKafQLHWjfrg8P9j1h0G64bEvfE76xOFIDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UsG7Xtx2wv2H57NQYBWJkmR+rfStre8Q4ALVk26iQvS/p4FbYAaEFrL2dke1bSrCSt0wVt3S2AMbV2oizJXJKZJDNTWtvW3QIYE2e/gWKIGiimybe0npT0iqQrbR+zfXf3swBMatkTZUnuWI0hANrBw2+gGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoJgmF8jbZHuP7cO2D9nesRrDAExm2QvkSTot6ddJDti+SNJ+27uTHO54G4AJLHukTvJRkgOj9z+TdETSxq6HAZhMkyP1/9jeLOkaSfvO8rVZSbOStE4XtDANwCQanyizfaGkpyXdm+TTr349yVySmSQzU1rb5kYAY2gUte0pLQb9RJJnup0EYCWanP22pEclHUnyUPeTAKxEkyP1tZLukrTN9sHR200d7wIwoWVPlCXZK8mrsAVAC/iJMqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgmCZXvVxn+zXbb9o+ZPvB1RgGYDLLXiBP0klJ25KcGF2neq/tPyV5teNtACbQ5KqXkXRi9OHU6C1djgIwuUbPqW2vsX1Q0oKk3Un2neU2s7bnbc+f0sm2dwJoqFHUSb5IskXStKSttq8+y23mkswkmZnS2rZ3AmhorLPfST6RtEfS9m7mAFipJme/L7F98ej98yVdL+mdrocBmEyTs98bJP3O9hot/iPwVJLnu50FYFJNzn6/JemaVdgCoAX8RBlQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFNPktrW+8XR8e7HsCsGo4UgPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQDFEDxRA1UAxRA8UQNVAMUQPFEDVQTOOoRxeef8M2F8cDBmycI/UOSUe6GgKgHY2itj0t6WZJj3Q7B8BKNT1SPyzpPklfft0NbM/anrc9f0onWxkHYHzLRm37FkkLSfaf63ZJ5pLMJJmZ0trWBgIYT5Mj9bWSbrX9vqSdkrbZfrzTVQAmtmzUSR5IMp1ks6TbJb2Y5M7OlwGYCN+nBooZ6yWCk7wk6aVOlgBoBUdqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoJixfkvrm+qGy7b0PQFYNRypgWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiiGqIFiiBoohqiBYogaKIaogWKIGiim0a9ejq5N/ZmkLySdTjLT5SgAkxvn96l/muRfnS0B0AoefgPFNI06kv5ie7/t2bPdwPas7Xnb86d0sr2FAMbS9OH3dUmO2/6+pN2230ny8tIbJJmTNCdJ3/F30/JOAA01OlInOT7674KkZyVt7XIUgMktG7Xt9bYvOvO+pJ9LervrYQAm0+Th96WSnrV95vZ/SPLnTlcBmNiyUSd5T9IPV2ELgBbwLS2gGKIGiiFqoBiiBoohaqAYogaKIWqgGKIGiiFqoBiiBoohaqAYogaKcdL+6xnY/qekf7RwV9+TNKTXRWPPuQ1tjzS8TW3t+UGSS872hU6ibovt+SG9cil7zm1oe6ThbVqNPTz8BoohaqCYoUc91/eAr2DPuQ1tjzS8TZ3vGfRzagDjG/qRGsCYiBooZpBR295u+13bR23fP4A9j9lesD2Il0a2vcn2HtuHbR+yvaPnPetsv2b7zdGeB/vcc4btNbbfsP1831ukxQtN2v6b7YO25zv7c4b2nNr2Gkl/l3S9pGOSXpd0R5LDPW76iaQTkn6f5Oq+dizZs0HShiQHRq/Jvl/SL/r6f+TF149en+SE7SlJeyXtSPJqH3uW7PqVpBlJ30lyS59bRnvelzTT9YUmh3ik3irpaJL3knwuaaek2/ocNLrE0Md9blgqyUdJDoze/0zSEUkbe9yTJCdGH06N3no9WtielnSzpEf63NGHIUa9UdIHSz4+ph7/wg6d7c2SrpG0r+cda2wflLQgaXeSXvdIeljSfZK+7HnHUsteaLINQ4waDdm+UNLTku5N8mmfW5J8kWSLpGlJW2339jTF9i2SFpLs72vD17guyY8k3Sjpl6Onda0bYtTHJW1a8vH06HNYYvTc9WlJTyR5pu89ZyT5RNIeSdt7nHGtpFtHz2F3Stpm+/Ee90havQtNDjHq1yVdYfty2+dJul3Scz1vGpTRialHJR1J8tAA9lxi++LR++dr8STnO33tSfJAkukkm7X49+fFJHf2tUda3QtNDi7qJKcl3SNplxZPAD2V5FCfm2w/KekVSVfaPmb77j73aPFIdJcWj0AHR2839bhng6Q9tt/S4j/Ku5MM4ttIA3KppL2235T0mqQ/dnWhycF9SwvAygzuSA1gZYgaKIaogWKIGiiGqIFiiBoohqiBYv4Lr7c+KoP9UcAAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9599/10000 [09:58<00:26, 15.32it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.12214137613773346 logZ = 1.9268975257873535 R = 0.00024036932\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9699/10000 [10:04<00:19, 15.71it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.0274082999676466 logZ = 2.1169180870056152 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9800/10000 [10:11<00:12, 16.58it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.061547596007585526 logZ = 1.7290458679199219 R = 1.494533e-05\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9900/10000 [10:17<00:05, 16.68it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "loss = 0.036688029766082764 logZ = 1.827315092086792 R = 0.062176514\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 10000/10000 [10:23<00:00, 16.03it/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "dist = torch.nn.functional.mse_loss(input=states.view(M,N), target=x.view(M,N))\n",
        "R = (-a*dist).exp()\n",
        "print(dist, R)\n",
        "\n",
        "plt.imshow(x.view(M,N).cpu()), plt.show()\n",
        "plt.imshow(states.view(M,N).cpu()), plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 548
        },
        "id": "kw3VQutfCJd6",
        "outputId": "3c0ee614-11c0-47fc-aeaf-d985e7fddab9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(0.1389, device='cuda:0') tensor(9.2925e-07, device='cuda:0')\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<matplotlib.image.AxesImage at 0x7f10d4a3c490>, None)"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Samples"
      ],
      "metadata": {
        "id": "thy5hnUbc2IL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "all_images = []\n",
        "all_R = []\n",
        "all_dist = []\n",
        "for it in tqdm.trange(100):    \n",
        "    # finished trajectories\n",
        "    dones = torch.full((batch_size,), False, dtype=torch.bool).to(device)\n",
        "    # s_0\n",
        "    states = torch.zeros(size=(batch_size, ndim)).to(device) # (batch_size, ndim)\n",
        "    # actions chosen at each step : we can choose the coordinate to increment (0 ... ndim-1), or choose to return the current state as terminal (ndim)\n",
        "    actions = None # (current_batch_size,)\n",
        "\n",
        "    max_steps = 1e8 #ndim+0\n",
        "    i = 0\n",
        "    while torch.any(~dones) and i <= max_steps :\n",
        "        ### Forward pass ### \n",
        "        current_batch_size = (~dones).sum()\n",
        "        non_terminal_states = states[~dones] # (current_batch_size, ndim)\n",
        "        logits = model_TB(non_terminal_states) # (current_batch_size, output_dim)\n",
        "        ### Forward Policy ### \n",
        "        PF_logits = logits[...,:ndim+1] # (current_batch_size, ndim+1) \n",
        "        edge_mask = (non_terminal_states == 1.).float() # (current_batch_size, ndim)\n",
        "        stop_action_mask = torch.zeros((current_batch_size, 1), device=device) # (current_batch_size, 1)\n",
        "        PF_mask = torch.cat([edge_mask, stop_action_mask], 1) # (current_batch_size, ndim+1)\n",
        "        logPF = (PF_logits + minus_inf*PF_mask).log_softmax(1) # (current_batch_size, ndim+1)\n",
        "        sample_temperature = 1\n",
        "        sample_ins_probs = (logPF/sample_temperature).softmax(1) # (current_batch_size, ndim+1)\n",
        "        actions = sample_ins_probs.multinomial(1) # (current_batch_size,)\n",
        "        \n",
        "        ### select terminal states ### \n",
        "        terminates = (actions==ndim).squeeze(1)\n",
        "        for state in non_terminal_states[terminates]: \n",
        "            all_visited_TB.append(state)\n",
        "       \n",
        "       # Update dones\n",
        "        dones[~dones] |= terminates\n",
        "\n",
        "        # Update non completed trajectories$\n",
        "        with torch.no_grad():\n",
        "            non_terminates = actions[~terminates].squeeze()\n",
        "            tmp = states[~dones]\n",
        "            tmp[torch.arange((~dones).sum()), non_terminates] = 1.\n",
        "            states[~dones] = tmp\n",
        "        \n",
        "        i+=1\n",
        "        \n",
        "    dist = torch.nn.functional.mse_loss(input=states.view(M,N), target=x.view(M,N))\n",
        "    R = (-a*dist).exp()\n",
        "    \n",
        "    all_images.append(states.view(M,N).cpu())\n",
        "    all_R.append(R.item())\n",
        "    all_dist.append(dist.item())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hUrx_Izdc3Xa",
        "outputId": "f0d3ba35-aa86-476f-d0b4-07e719e7c4e8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 100/100 [00:02<00:00, 37.58it/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tp = torch.tensor(all_R).topk(k=10)\n",
        "values = tp.values\n",
        "indices = tp.indices"
      ],
      "metadata": {
        "id": "tkhIb1JYc6IW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for i in indices:\n",
        "    fig, axarr = plt.subplots(1,2)\n",
        "    fig.suptitle(\"d=%s, R=%s\"%(all_dist[i], all_R[i]), x=0.5, y=0.1, fontsize=15)\n",
        "    axarr[0].imshow(x.view(M,N).cpu())\n",
        "    axarr[0].set_title('x')\n",
        "    axarr[1].imshow(all_images[i].view(M,N).cpu())\n",
        "    axarr[1].set_title('generated')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "peRf8ZBqc744",
        "outputId": "3e017f02-2a1e-4c14-c1bf-0d76996fc895"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWoAAADtCAYAAACWGj83AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAARtElEQVR4nO3debBkZXnH8e/PYRhkEVA0yqJETTSEGLAmuFYELBVxIbFi1MSFRDPRBAt3jcZEjTFa7kYrOG6YckEtMeWCoCiWISo6KCqbigZlEySALOKwPfmjz2hzvcPtO9Pd5z2X76fq1Nx7tn5O1zO/+963T99OVSFJatdt+i5AknTLDGpJapxBLUmNM6glqXEGtSQ1zqCWpMYZ1JJWjCSV5J591zFtBrWkJiQ5PMnJfdfRIoNa0swl2abvGobMoO5BknskuSzJfbvvd0/ysyQH9lyaVqgk903yrSRXJflYko8keXW37dFJTktyRZKvJLnP2HHnJnlBku8k+Xl33HZj25c69sVJvgNck2SbJC9J8sOujjOT/Gm37+8BRwEPSHJ1kiu69WuSvCHJT5JcnOSoJLcde4wXJrkoyYVJ/nrmT2RfqsqlhwX4G+BMYHvgBOANfdfksjIXYFvgx8CRwGrgccB1wKuB/YFLgPsBq4CnAecCa7pjzwW+DuwO3B44C3hmt22SY08D9gJu2617fHeu2wBPAK4B7tJtOxw4eUHtbwY+2T32TsCngH/rth0CXAzsC+wAfAgo4J59P+fTXhxR96Sq3gWcA5wC3AV4Wb8VaQW7P7AN8Laqur6qjmUUvgDrgHdW1SlVdWNVvR/Y2B2zyduq6sKquoxRUO63zGPPq6prAarqY925bqqqjwA/AA5YrOgk6R7juVV1WVVdBbwGeGK3y58D76uq06vqGuAVW/wMNc55o369i9FoYV1Vbey7GK1YuwMXVDcM7ZzX/Xs34GlJnj22bdvumE1+Ovb1L8a2TXLseWNfk+SpwPOAvbtVOwK7babuOzL6jfPUUWaPTsFo9L7puk4d2//HmznP4Dmi7kmSHYG3AO8BXpHk9j2XpJXrImCPjKUdo+kIGAXpv1bVLmPL9lX14QnOO8mxv/rhkORujAYnRwB3qKpdgNMZhe/N9u1cClwL/P7Y+Xeuqh3Hrmuvsf3vOkHNg2RQ9+etwIaqegbwGUYvpEiz8FXgRuCI7gW9w/j1dMO7gGcmuV9GdkjyqCQ7TXDe5R67A6Mw/hlAkr9iNL+8ycXAnkm2Baiqm7rHeHOSO3XH7JHkEd3+HwUOT7JPku2Bf57s6Rgeg7oH3X+UQ4BndaueB9w3yV/2V5VWqqq6jtELiE8HrgCeDHwa2FhVGxi9sP124HJGr5scPuF5l3VsVZ0JvJHRD46LgT8A/mdsly8CZwA/TXJpt+7F3Xm/luRK4ETgXt35Psvot9Ivdvt8cZK6hyg3n7aSdGuQ5BTgqKp6X9+1aGmOqKVbgSQPSXLnburjacB9gOP7rkuT8a4P6dbhXozmdHcAfgT8WVVd1G9JmpRTH5LUOKc+JKlxBrUkNW4mc9TbZk1txw6zOLXEL7mG62pjlt5zuuxrzdIt9fVMgno7duB+eegsTi1xSn2hl8e1rzVLt9TXTn1IUuMMaklqnEEtSY0zqCWpcQa1JDXOoJakxhnUktS4iYI6ySFJvpfknCQvmXVR0rzY2xqCJYM6ySrgHcAjgX2AJyXZZ9aFSbNmb2soJhlRHwCcU1U/6j4p4hjgsNmWJc2Fva1BmCSo9+DmnyR8frfuZpKsS7IhyYbr8QO1NQhL9rZ9rRZM7cXEqlpfVWurau1q1kzrtFKv7Gu1YJKgvoCbfyT7nt06aejsbQ3CJEH9DeB3kvx29zHuTwQ+OduypLmwtzUIS/6Z06q6IckRwAnAKuC9VXXGzCuTZsze1lBM9Peoq+o44LgZ1yLNnb2tIfCdiZLUOINakhpnUEtS4wxqSWqcQS1JjTOoJalxE92ed2t2woWn9V3C1Dxi9/36LkGNsK+HxRG1JDXOoJakxhnUktQ4g1qSGmdQS1LjDGpJapxBLUmNM6glqXEGtSQ1bsmgTvLeJJckOX0eBUnzYm9rKCYZUR8NHDLjOqQ+HI29rQFYMqir6svAZXOoRZore1tD4Ry1JDVuan89L8k6YB3Admw/rdNKvbKv1YKpjairan1Vra2qtatZM63TSr2yr9UCpz4kqXGT3J73YeCrwL2SnJ/k6bMvS5o9e1tDseQcdVU9aR6FSPNmb2sonPqQpMYZ1JLUOINakhpnUEtS4wxqSWqcQS1JjTOoJalxBrUkNc6glqTGGdSS1DiDWpIaZ1BLUuMMaklqnEEtSY0zqCWpcQa1JDXOoJakxk3yUVx7JTkpyZlJzkhy5DwKk2bN3tZQLPlRXMANwPOr6ptJdgJOTfL5qjpzxrVJs2ZvaxCWHFFX1UVV9c3u66uAs4A9Zl2YNGv2toZikhH1ryTZG9gfOGWRbeuAdQDbsf0USpPmZ3O9bV+rBRO/mJhkR+DjwHOq6sqF26tqfVWtraq1q1kzzRqlmbql3rav1YKJgjrJakaN/MGqOna2JUnzY29rCCa56yPAe4CzqupNsy9Jmg97W0MxyYj6QcBTgIOTnNYth864Lmke7G0NwpIvJlbVyUDmUIs0V/a2hsJ3JkpS4wxqSWqcQS1JjTOoJalxBrUkNc6glqTGGdSS1DiDWpIaZ1BLUuMMaklqnEEtSY0zqCWpcQa1JDXOoJakxhnUktQ4g1qSGjfJR3Ftl+TrSb6d5Iwkr5xHYdKs2dsaiiU/4QXYCBxcVVd3HwR6cpLPVtXXZlybNGv2tgZhko/iKuDq7tvV3VKzLEqaB3tbQzHRHHWSVUlOAy4BPl9Vp8y2LGk+7G0NwURBXVU3VtV+wJ7AAUn2XbhPknVJNiTZcD0bp12nNBNL9bZ9rRYs666PqroCOAk4ZJFt66tqbVWtXc2aadUnzcXmetu+Vgsmuevjjkl26b6+LfAw4OxZFybNmr2toZjkro+7AO9PsopRsH+0qj4927KkubC3NQiT3PXxHWD/OdQizZW9raHwnYmS1DiDWpIaZ1BLUuMMaklqnEEtSY0zqCWpcQa1JDVukje8NOmEC0/ruwRp6uxrLcYRtSQ1zqCWpMYZ1JLUOINakhpnUEtS4wxqSWqcQS1JjTOoJalxEwd192nN30riJ2BoxbCvNQTLGVEfCZw1q0KkntjXat5EQZ1kT+BRwLtnW440P/a1hmLSEfVbgBcBN82wFmne7GsNwpJBneTRwCVVdeoS+61LsiHJhuvZOLUCpVmwrzUkk4yoHwQ8Nsm5wDHAwUk+sHCnqlpfVWurau1q1ky5TGnq7GsNxpJBXVX/UFV7VtXewBOBL1bVk2demTRD9rWGxPuoJalxy/rggKr6EvClmVQi9cS+VuscUUtS4wxqSWqcQS1JjTOoJalxBrUkNc6glqTGGdSS1Lhl3Ufdkkfsvl/fJUhTZ19rMY6oJalxBrUkNc6glqTGGdSS1DiDWpIaZ1BLUuMMaklqnEEtSY0zqCWpcRO9M7H7ANCrgBuBG6pq7SyLkubF3tYQLOct5AdV1aUzq0Tqj72tpjn1IUmNmzSoC/hcklOTrFtshyTrkmxIsuF6Nk6vQmm2brG37Wu1YNKpjwdX1QVJ7gR8PsnZVfXl8R2qaj2wHuB2uX1NuU5pVm6xt+1rtWCiEXVVXdD9ewnwCeCAWRYlzYu9rSFYMqiT7JBkp01fAw8HTp91YdKs2dsaikmmPn4L+ESSTft/qKqOn2lV0nzY2xqEJYO6qn4E/OEcapHmyt7WUHh7niQ1zqCWpMYZ1JLUOINakhpnUEtS4wxqSWqcQS1JjUvV9P98QZKfAT9exiG7ASvlz0x6LbN3t6q647wfdAv6Gtp9DrfESrmWVq9js309k6BeriQbVsofbPdaNG4lPYcr5VqGeB1OfUhS4wxqSWpcK0G9vu8Cpshr0biV9ByulGsZ3HU0MUctSdq8VkbUkqTN6D2okxyS5HtJzknykr7r2VJJ9kpyUpIzk5yR5Mi+a9oaSVYl+VaST/ddyxDZ1+0aYm/3GtRJVgHvAB4J7AM8Kck+fda0FW4Anl9V+wD3B/5+wNcCcCRwVt9FDJF93bzB9XbfI+oDgHOq6kdVdR1wDHBYzzVtkaq6qKq+2X19FaNG2KPfqrZMkj2BRwHv7ruWgbKvGzXU3u47qPcAzhv7/nwG3ASbJNkb2B84pd9KtthbgBcBN/VdyEDZ1+0aZG/3HdQrTpIdgY8Dz6mqK/uuZ7mSPBq4pKpO7bsWtWPofQ3D7u2+g/oCYK+x7/fs1g1SktWMmvmDVXVs3/VsoQcBj01yLqNf2Q9O8oF+Sxoc+7pNg+3tXu+jTrIN8H3goYwa+RvAX1TVGb0VtYUy+ijr9wOXVdVz+q5nGpIcCLygqh7ddy1DYl+3b2i93euIuqpuAI4ATmD0IsVHh9jMnQcBT2H0U/q0bjm076I0f/a1ps13JkpS4/qeo5YkLcGglqTGGdSS1DiDWpIaZ1BLUuMMaklqnEEtSY0zqCWpcQa1JDXOoJakxhnUktQ4g1qSGmdQS1LjDGpJapxBLUmNM6glqXEGtSQ1zqCWpMYZ1JLUOINakhpnUEtS4wxqSWqcQS1JjTOoJalxBrWWlGTfJJXkwK08z85J3pfk8iQ/T/LBJHeY4LgkeWmS85Jcm+TLSfbbwhpqbLk2yVlJXpxkmy0534Jz/12SzyT5v+U+X0kOS/LdJL9McmaSJ2xtPVo5DGrN00eBA4FnAIcDfwT81wTHvQR4OfA64DHA1cCJSe68hXW8EXgAcChwHPDa7jG21lOB2wMnLOegJA8GPg6cBDwS+Azw4SQPn0JNWgFSVX3XoMYl2Rf4LnBQVX1pC8/xAOArwEOq6svdugOAU4CHVdWJmzluO+Bi4I1V9apu3Q7AucA7q+ofl1lHAc+uqrePrTseuHtV/e6yL+zm575NVd203OcryQnA6qo6eGzdccDtqurBW1OTVgZH1PoN3a/w5yW5JsmngLtM4bSPBC7eFNIAVfV14H+7bZvzQOB2jEbjm467BvjUEsctx7eBvbb2JFV103KPSbIGOIix6+scAzwgyc5bW5eGz6DWzSQ5DHgH8GngcYxGhu9dZL9VSbZZYhnvr3sDZy/ykGd12zbn3sCNwA+Wedxy3JXRD4xfSXKbCa5v1RQe+x7Aan7zuTmL0f/PrRrla2UwqLXQy4Djq+pZVXVCVb0UOH6R/b4AXL/EMh7wuwJXLHKey7ttm7MrcHVV3bjIcdsn2XbpS/oNm0J4pyRPYvQD6VUL9vknlr6+H27BYy+06doXPjeXL9iuW7GtfqVbK0d358N9gSMWbDqW0QuA4/4W2GmJU146pdKm7a3dssmbq+qYBfusZ/RbxS3ZONWqpM0wqDVuN2AVcMmC9Qu/BzgHyBLnG5+zvRy44yL77MqvR4+LuRzYMcmqBaPqXYFfVNV1S9SwmNczmhPeGXgO8NwkJ1bVcWP7/JTFr3vcNF6J33TtC+eid12wXbdiTn1o3KWM5oPvtGD9wu9h+VMfZ7P4nPLm5q7Hj1sF3HOZx92Sn1TVhqr6AqNpj+8Br08y/oNnXlMfP+zOtfC5uTejH3Tfn8JjaOAcUetXquqGJN8CDgOOGtv0uEV2X+7Ux2eBlyd5cFWdDJBkLXD3btvmfAW4Eng88OruuO0Z3U+9fonHX1JVXZ/k5YxG2I8BPtltmsvUR1VtTHISo+t759imJwBfraqfb+1jaPgMai30GuDYJP8BfAJ4CHDIwp2q6nvLOWlVfTXJ54D/TPICRqPF1wEnj99DneQ9jO61vmd33C+TvJZRyF/OaBT9PEa/Df772HEHMnrDyJbc6/3x7rwvpAvqqroQuHA5J+l+8OzNr2/1e0iS3YBzq2pDt89TGf2mcY+q+nG3378AX0ryFkZvADq0W37jedetVFW5uNxsYfRi4vnALxi9c+/hjOZjD9zK8+4CvI/RHQ5XAh8Cdluwz9GMgm18XRjdjXI+cC3w38D+C/Y5tKtxnyVqKOCIRdY/tdt2/624vqO7cyxcjh7b5/Bu3d4Ljv0T4HRGo/SzgSf23Qcu7Sy+M1ErQpJXAn9cVQf1XYs0bb6YqJXigcCb+i5CmgVH1JLUOEfUktQ4g1qSGmdQS1LjDGpJapxBLUmNM6glqXEGtSQ17v8BEftjXs2BDIYAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "figsize=(6*4,4*4)\n",
        "fig, (ax1, ax2)  = plt.subplots(2, 1, sharex=False, sharey=False, figsize = figsize)\n",
        "fig.suptitle(\"Flownet agent\")\n",
        "\n",
        "## loss\n",
        "xs_TB = list(range(len(losses_TB)))\n",
        "ax1.plot(xs_TB, losses_TB, label=\"TB\")\n",
        "ax1.set(xlabel='episodes', ylabel='loss')\n",
        "ax1.set_title('loss per episodes : do not compare')\n",
        "ax1.legend()\n",
        "\n",
        "## rewards\n",
        "ax2.plot(xs_TB, rewards_TB, label=\"TB\")\n",
        "ax2.set(xlabel='episodes', ylabel='reward')\n",
        "ax2.set_title('reward per episodes')\n",
        "ax2.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "gdj6z6bSc-7M",
        "outputId": "33f148d1-57c9-46e4-e9eb-e7d70375534e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f10d46236d0>"
            ]
          },
          "metadata": {},
          "execution_count": 19
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1728x1152 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# GAN (Goodfellow et al., 2014) \n",
        "\n",
        "I took this GAN base code here : https://github.com/ikr7/wandb-pytorch-gan-mnist-demo\n"
      ],
      "metadata": {
        "id": "k40Y35GbKVRh"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "dataset : MNIST,  28x28"
      ],
      "metadata": {
        "id": "LX1k2_4PLctJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from torchvision.datasets import MNIST\n",
        "import torchvision.transforms as transforms\n",
        "\n",
        "def bin_transform(x) :\n",
        "    return 1.*(x!=0) + 0.*(x==0)\n",
        "\n",
        "M,N = 28, 28\n",
        "def get_mnist(bin = False):\n",
        "    transform = transforms.Compose([\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Resize(size=(M,N)),\n",
        "        transforms.Normalize(mean=(0.5,), std=(0.5,)) if not bin else bin_transform,\n",
        "    ])\n",
        "    return MNIST(root='/content', download=True, train=True, transform=transform)"
      ],
      "metadata": {
        "id": "6QSy6HNILd28"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "models"
      ],
      "metadata": {
        "id": "OrMQXuwNLjB_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch.nn as nn\n",
        "\n",
        "class Flatten(nn.Module):\n",
        "    def forward(self, x):\n",
        "        return x.view(x.shape[0], -1)\n",
        "\n",
        "def block (n_in_channels, n_out_channels, batch_norm=True):\n",
        "    layers = []\n",
        "    layers.append(nn.Linear(n_in_channels, n_out_channels))\n",
        "    if batch_norm:\n",
        "        layers.append(nn.BatchNorm1d(n_out_channels, 0.8))\n",
        "    layers.append(nn.LeakyReLU(0.2, inplace=True))\n",
        "    return layers\n",
        "\n",
        "class Generator(nn.Module):\n",
        "    def __init__(self, nz, bin = False):\n",
        "\n",
        "        super(Generator, self).__init__()\n",
        "\n",
        "        self.main = nn.Sequential(\n",
        "            *block(nz, 256, batch_norm=False),\n",
        "            *block(256, 512),\n",
        "            *block(512, 1024),\n",
        "            nn.Linear(1024, M*N),\n",
        "            nn.Sigmoid() if bin else nn.Tanh()\n",
        "        )\n",
        "\n",
        "    def forward(self, z):\n",
        "        fake_image = self.main(z)\n",
        "        return fake_image.view(fake_image.size(0), 1, M, N)\n",
        "\n",
        "    def predict(self, z):\n",
        "        fake_image = self.main(z)\n",
        "        return fake_image.view(fake_image.size(0), 1, M, N)\n",
        "\n",
        "class Discriminator(nn.Module):\n",
        "    def __init__(self):\n",
        "\n",
        "        super(Discriminator, self).__init__()\n",
        "\n",
        "        self.main = nn.Sequential(\n",
        "            Flatten(),\n",
        "            nn.Linear(M*N, 1024),\n",
        "            nn.LeakyReLU(0.2, inplace=True),\n",
        "            nn.Dropout(0.3),\n",
        "            nn.Linear(1024, 512),\n",
        "            nn.LeakyReLU(0.2, inplace=True),\n",
        "            nn.Dropout(0.3),\n",
        "            nn.Linear(512, 256),\n",
        "            nn.LeakyReLU(0.2, inplace=True),\n",
        "            nn.Dropout(0.3),\n",
        "            nn.Linear(256, 1),\n",
        "            nn.Sigmoid(),\n",
        "        )\n",
        "\n",
        "    def forward(self, x):\n",
        "        return self.main(x)"
      ],
      "metadata": {
        "id": "yxNKPKGEKXZ4"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "train"
      ],
      "metadata": {
        "id": "U8MyMvFXLlKz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "## Uncomment if you want to use wandb ##\n",
        "#!pip install wandb"
      ],
      "metadata": {
        "id": "7BNkrZWTMJHm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "## Uncomment if you want to use wandb ##\n",
        "# #wandb.login()\n",
        "# %env wand_db_key=\"PASTE YOURS\"\n",
        "# !wandb login $wand_db_key\n",
        "\n",
        "# import wandb"
      ],
      "metadata": {
        "id": "OIKRMg5mM13P"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import argparse\n",
        "from pathlib import Path\n",
        "\n",
        "import torch.nn as nn\n",
        "from torch.utils.data import DataLoader\n",
        "from tqdm import tqdm\n",
        "import yaml\n",
        "\n",
        "class AttrDict(dict):\n",
        "    def __init__(self, *args, **kwargs):\n",
        "        super(AttrDict, self).__init__(*args, **kwargs)\n",
        "        self.__dict__ = self"
      ],
      "metadata": {
        "id": "KT-Pqsextc3J"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Uncomment this part to train a GAN, or go to the next section for GAN-GflowNets"
      ],
      "metadata": {
        "id": "F8LTOqa_g4ln"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "params = AttrDict({\n",
        "    # number of epochs \n",
        "    \"num_epochs\": 250,\n",
        "    # batch size\n",
        "    \"batch_size\": 1024,\n",
        "    # dimensionality of the latent space\n",
        "    \"nz\": 100,\n",
        "    # learning rate for discriminator\n",
        "    \"d_lr\": 0.0002,\n",
        "    # learning rate for generator\n",
        "    \"g_lr\": 0.0002, \n",
        "    # use a GPU or not\n",
        "    \"cuda\": torch.cuda.is_available(),\n",
        "    # where to save network snapshots\n",
        "    \"log_dir\": '/content/',\n",
        "    # binary image\n",
        "    \"bin\" : True,\n",
        "})\n",
        "\n",
        "dataset = get_mnist(bin=params[\"bin\"])\n",
        "dataloader = DataLoader(\n",
        "    dataset,\n",
        "    batch_size=params['batch_size'],\n",
        "    shuffle=True,\n",
        "    num_workers=4,\n",
        "    drop_last=True\n",
        ")\n",
        "\n",
        "device = torch.device('cuda:0' if params['cuda'] else 'cpu')"
      ],
      "metadata": {
        "id": "L12TnhMlMduV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "## Uncomment if you want to use wandb ##\n",
        "# wandb.init(project='gan-mnist', config=params)\n",
        "# run_id = wandb.run.id\n",
        "# run_log_dir = Path(params['log_dir']) / run_id\n",
        "# run_log_dir.mkdir(exist_ok=True, parents=True)"
      ],
      "metadata": {
        "id": "LOJ4nopYa2pv"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "d_net = Discriminator().to(device)\n",
        "g_net = Generator(nz=params['nz'], bin=params[\"bin\"]).to(device)\n",
        "\n",
        "# Uncomment if you want to use wandb ##\n",
        "# wandb.watch(d_net)\n",
        "# wandb.watch(g_net)\n",
        "\n",
        "bce = nn.BCELoss(reduction = 'mean')\n",
        "\n",
        "d_opt = torch.optim.Adam(d_net.parameters(), lr=params['d_lr'], betas=(0.5, 0.999))\n",
        "g_opt = torch.optim.Adam(g_net.parameters(), lr=params['g_lr'], betas=(0.5, 0.999))\n",
        "\n",
        "fixed_noise = torch.randn(64, params['nz']).to(device)"
      ],
      "metadata": {
        "id": "sE8I2CXAmMSC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for epoch in tqdm(range(1, params['num_epochs'] + 1), desc='Training'):\n",
        "\n",
        "    d_net.train()\n",
        "    g_net.train()\n",
        "\n",
        "    total_g_loss_train = 0\n",
        "    total_d_real_loss_train = 0\n",
        "    total_d_fake_loss_train = 0\n",
        "    total_d_loss_train = 0\n",
        "\n",
        "    for i, (real_image, _) in enumerate(tqdm(dataloader, desc=f'Epoch {epoch :4d}', leave=False), 1):\n",
        "\n",
        "        noise = torch.randn(params['batch_size'], params['nz']).to(device)\n",
        "\n",
        "        real_image = real_image.to(device)\n",
        "        fake_image = g_net(noise)\n",
        "\n",
        "        real_label = torch.ones(params['batch_size'], 1).to(device)\n",
        "        fake_label = torch.zeros(params['batch_size'], 1).to(device)\n",
        "\n",
        "        g_net.zero_grad()\n",
        "        g_loss = bce(d_net(fake_image), real_label)\n",
        "        total_g_loss_train += g_loss.item()\n",
        "        g_loss.backward()\n",
        "        g_opt.step()\n",
        "\n",
        "        d_net.zero_grad()\n",
        "        d_real_loss = bce(d_net(real_image), real_label)\n",
        "        d_fake_loss = bce(d_net(fake_image.detach()), fake_label)\n",
        "        d_loss = d_real_loss + d_fake_loss\n",
        "        total_d_real_loss_train += d_real_loss.item()\n",
        "        total_d_fake_loss_train += d_fake_loss.item()\n",
        "        total_d_loss_train += d_loss.item()\n",
        "        d_loss.backward()\n",
        "        d_opt.step()\n",
        "    \n",
        "    g_loss_train = total_g_loss_train / len(dataloader)\n",
        "    d_real_loss_train = total_d_real_loss_train / len(dataloader)\n",
        "    d_fake_loss_train = total_d_fake_loss_train / len(dataloader)\n",
        "    d_loss_train = total_d_loss_train / len(dataloader)\n",
        "\n",
        "    g_net.eval()\n",
        "    with torch.no_grad():\n",
        "        fixed_fake_image = g_net.predict(fixed_noise).detach()\n",
        "\n",
        "    data = {\n",
        "         'g_loss_train': g_loss_train,\n",
        "         'd_real_loss_train': d_real_loss_train,\n",
        "         'd_fake_loss_train': d_fake_loss_train,\n",
        "         'd_loss_train': d_loss_train\n",
        "    }\n",
        "\n",
        "    ## Uncomment if you want to use wandb ##\n",
        "    #data['examples'] =  [wandb.Image(i) for i in fixed_fake_image]\n",
        "    #data[\"reals\"] = [wandb.Image(i) for i in real_image]\n",
        "    #wandb.log(data)\n",
        "\n",
        "    #torch.save(g_net.state_dict(), run_log_dir / f'generator-{epoch :04d}.pth')\n",
        "    #torch.save(d_net.state_dict(), run_log_dir / f'discriminator-{epoch :04d}.pth')\n",
        "    \n",
        "    run_log_dir = \"/content\"\n",
        "    torch.save(g_net.state_dict(), run_log_dir+\"/\"+ f'generator-{epoch :04d}.pth')\n",
        "    torch.save(d_net.state_dict(), run_log_dir+\"/\"+ f'discriminator-{epoch :04d}.pth')\n",
        "\n",
        "\n",
        "    if epoch%1 == 0:\n",
        "        plt.imshow(fake_image[0][0].detach().cpu()), plt.show()\n",
        "        for k, v in data.items():\n",
        "            if k != \"examples\": print(k, v)"
      ],
      "metadata": {
        "id": "g9gWxsdbKrxP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "# GAN GflowNet\n",
        "The Generator generates the images as I described above, and the reward is calculated that way : $R(x = Generator()) = e^{- \\big[ binary\\_cross\\_entropy\\_loss \\big( Discriminator(x), \\ real\\_label \\big) \\big] }$\n",
        "\n",
        "With good hyperparameters and on small images the results are acceptable."
      ],
      "metadata": {
        "id": "29foZGErlu4j"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class GeneratorFlowNet(nn.Module):\n",
        "    def __init__(self, model_TB, logZ_TB, independent_Z, uniform_PB = False):\n",
        "\n",
        "        super(GeneratorFlowNet, self).__init__()\n",
        "\n",
        "        self.model_TB = model_TB\n",
        "        self.logZ_TB = 0 if independent_Z else logZ_TB\n",
        "        self.independent_Z = independent_Z\n",
        "        self.uniform_PB = uniform_PB\n",
        "    \n",
        "    def forward(self, z):\n",
        "        batch_size = z.size(0)\n",
        "        device = z.device\n",
        "        ndim = M*N\n",
        "        minus_inf = -1e8\n",
        "\n",
        "        # TB loss for each trajectory\n",
        "        loss_TB = torch.zeros((batch_size,)).to(device)\n",
        "        loss_TB += self.logZ_TB # see the equation above\n",
        "        # finished trajectories\n",
        "        dones = torch.full((batch_size,), False, dtype=torch.bool).to(device)\n",
        "        # s_0\n",
        "        states = torch.zeros(size=(batch_size, ndim)).to(device) # (batch_size, ndim)\n",
        "        # actions chosen at each step \n",
        "        actions = None # (current_batch_size,)\n",
        "\n",
        "        max_steps = ndim*1e8\n",
        "        i = 0\n",
        "        while torch.any(~dones) and i <= max_steps :\n",
        "            ### Forward pass ### \n",
        "            current_batch_size = (~dones).sum()\n",
        "            non_terminal_states = states[~dones] # (current_batch_size, ndim)\n",
        "            logits = self.model_TB(non_terminal_states) # (current_batch_size, output_dim)\n",
        "\n",
        "            ### Backward Policy ### \n",
        "            PB_logits = logits[...,ndim+1:2*ndim+1] # (current_batch_size, ndim)\n",
        "            PB_logits = PB_logits * (0 if self.uniform_PB else 1) # (current_batch_size, ndim)\n",
        "            PB_mask = (non_terminal_states == 0.).float() # (current_batch_size, ndim)\n",
        "            logPB = (PB_logits + minus_inf*PB_mask).log_softmax(1) # (current_batch_size, ndim)\n",
        "\n",
        "            is_initial = actions is None\n",
        "            if not is_initial : \n",
        "                loss_TB[~dones] -= logPB.gather(1, actions[actions!=ndim].unsqueeze(1)).squeeze(1)\n",
        "            else :\n",
        "                logZ_TB = logits[...,-1]\n",
        "                loss_TB += logZ_TB # see the equation above\n",
        "\n",
        "            ### Forward Policy ### \n",
        "            PF_logits = logits[...,:ndim+1] # (current_batch_size, ndim+1) \n",
        "            edge_mask = (non_terminal_states == 1.).float() # (current_batch_size, ndim)\n",
        "            stop_action_mask = torch.zeros((current_batch_size, 1), device=device) # (current_batch_size, 1)\n",
        "            #stop_action_mask = torch.ones((current_batch_size, 1), device=device) # (current_batch_size, 1)\n",
        "            PF_mask = torch.cat([edge_mask, stop_action_mask], 1) # (current_batch_size, ndim+1)\n",
        "            logPF = (PF_logits + minus_inf*PF_mask).log_softmax(1) # (current_batch_size, ndim+1)\n",
        "            sample_temperature = 1 # 0.5\n",
        "            sample_ins_probs = (logPF/sample_temperature).softmax(1) # (current_batch_size, ndim+1)\n",
        "            actions = sample_ins_probs.multinomial(1) # (current_batch_size,)\n",
        "            loss_TB[~dones] += logPF.gather(1, actions).squeeze(1)\n",
        "\n",
        "            ### select terminal states ### \n",
        "            terminates = (actions==ndim).squeeze(1)\n",
        "          \n",
        "            # Update dones\n",
        "            dones[~dones] |= terminates\n",
        "\n",
        "            # Update non completed trajectories$\n",
        "            with torch.no_grad():\n",
        "                non_terminates = actions[~terminates].squeeze()\n",
        "                tmp = states[~dones]\n",
        "                tmp[torch.arange((~dones).sum()), non_terminates] = 1.\n",
        "                states[~dones] = tmp\n",
        "            \n",
        "            i+=1\n",
        "            \n",
        "        return states.view(batch_size, 1, M,N), loss_TB, logZ_TB if self.independent_Z else self.logZ_TB\n",
        "\n",
        "    def predict(self, z):\n",
        "        with torch.no_grad():\n",
        "            fake_image, loss_TB, logZ_TB = self.forward(z)\n",
        "        return fake_image"
      ],
      "metadata": {
        "id": "DoPKgMu_lvu2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "params = AttrDict({\n",
        "    # number of epochs \n",
        "    \"num_epochs\": 250,\n",
        "    # batch size\n",
        "    \"batch_size\": 1024,\n",
        "    # dimensionality of the latent space\n",
        "    \"nz\": 100,\n",
        "    # learning rate for discriminator\n",
        "    \"d_lr\": 0.0002,\n",
        "    # learning rate for generator\n",
        "    \"g_lr\": 0.0002, \n",
        "    # use a GPU or not\n",
        "    \"cuda\": torch.cuda.is_available(),\n",
        "    # where to save network snapshots\n",
        "    \"log_dir\": '/content/',\n",
        "    # binary image\n",
        "    \"bin\" : True,\n",
        "})\n",
        "\n",
        "def bin_transform(x) :\n",
        "    return 1.*(x!=0) + 0.*(x==0)\n",
        "\n",
        "#M,N=28,28\n",
        "M,N=10,10\n",
        "def get_mnist(bin = False):\n",
        "    transform = transforms.Compose([\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Resize(size=(M,N)),\n",
        "        transforms.Normalize(mean=(0.5,), std=(0.5,)) if not bin else bin_transform,\n",
        "    ])\n",
        "    return MNIST(root='/content', download=True, train=True, transform=transform)\n",
        "\n",
        "dataset = get_mnist(bin=params[\"bin\"])\n",
        "dataloader = DataLoader(\n",
        "    dataset,\n",
        "    batch_size=params['batch_size'],\n",
        "    shuffle=True,\n",
        "    num_workers=4,\n",
        "    drop_last=True\n",
        ")\n",
        "\n",
        "device = torch.device('cuda:0' if params['cuda'] else 'cpu')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I1BgWH25torY",
        "outputId": "b8bbdae8-23cf-4a45-e6fb-21ecc624c24d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py:481: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  cpuset_checked))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(dataset)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "duH376eH1gcy",
        "outputId": "e3306944-ad90-4233-ea4b-1a049a8dfbaf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "60000"
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def make_mlp(l, act=torch.nn.LeakyReLU(), tail=[]):\n",
        "    return torch.nn.Sequential(*(sum(\n",
        "        [[torch.nn.Linear(i, o)] + ([act] if n < len(l)-2 else [])\n",
        "         for n, (i, o) in enumerate(zip(l, l[1:]))], []) + tail))"
      ],
      "metadata": {
        "id": "qOYYlEhGwnf4"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "n_hid = 256\n",
        "n_layers = 2\n",
        "ndim = M*N\n",
        "input_dim = ndim # embedding dim\n",
        "output_dim = 2*ndim+1 # ndim + 1 for P_F (+1 for stop action) and ndim for P_B \n",
        "independent_Z = True\n",
        "if independent_Z:\n",
        "    output_dim+=1\n",
        "    logZ_TB = None\n",
        "else :\n",
        "    logZ_TB = torch.zeros((1,)).to(device) # log (initial state flow), Z = 1\n",
        "\n",
        "model_TB = make_mlp([input_dim] + [n_hid] * n_layers + [output_dim]).to(device)\n",
        "\n",
        "model_TB"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ewpiVLf7mFih",
        "outputId": "c931012c-551f-48b5-85e8-c2cf902513c6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Sequential(\n",
              "  (0): Linear(in_features=100, out_features=256, bias=True)\n",
              "  (1): LeakyReLU(negative_slope=0.01)\n",
              "  (2): Linear(in_features=256, out_features=256, bias=True)\n",
              "  (3): LeakyReLU(negative_slope=0.01)\n",
              "  (4): Linear(in_features=256, out_features=202, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# wandb.init(project='gan-mnist',  config=params) \n",
        "# run_id = wandb.run.id\n",
        "# run_log_dir = Path(params['log_dir']) / run_id\n",
        "# run_log_dir.mkdir(exist_ok=True, parents=True)\n",
        "\n",
        "d_net = Discriminator().to(device)\n",
        "g_net = GeneratorFlowNet(model_TB, logZ_TB, independent_Z, uniform_PB = False)\n",
        "\n",
        "# wandb.watch(d_net)\n",
        "# wandb.watch(g_net)\n",
        "\n",
        "bce_g = nn.BCELoss(reduction = 'none')\n",
        "bce_d = nn.BCELoss(reduction = 'mean')\n",
        "\n",
        "d_opt = torch.optim.Adam(d_net.parameters(), lr=params['d_lr'], betas=(0.5, 0.999))\n",
        "#g_opt = torch.optim.Adam(g_net.parameters(), lr=params['g_lr'], betas=(0.5, 0.999))\n",
        "if independent_Z :\n",
        "    g_opt = torch.optim.Adam([ {'params':g_net.model_TB.parameters(), 'lr':params['g_lr'], \"betas\":(0.5, 0.999)}])\n",
        "else :\n",
        "    g_opt = torch.optim.Adam([ {'params':g_net.model_TB.parameters(), 'lr':params['g_lr'], \"betas\":(0.5, 0.999)}, {'params':[g_net.logZ_TB], 'lr':00.1} ])\n",
        "    logZ_TB.requires_grad_()\n",
        "\n",
        "fixed_noise = torch.randn(64, params['nz']).to(device)"
      ],
      "metadata": {
        "id": "Gakp6wbPwE49"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "params['batch_size'] = 512\n",
        "dataloader = DataLoader(\n",
        "    dataset,\n",
        "    batch_size=params['batch_size'],\n",
        "    shuffle=True, # because we will update our generator 2-5 times more than the discriminator.\n",
        "    num_workers=4,\n",
        "    drop_last=True\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mmQbtopwzDE-",
        "outputId": "c9dabed4-11ee-4dab-f426-8b3b1cd7811d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py:481: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  cpuset_checked))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for epoch in tqdm(range(1, params['num_epochs'] + 1), desc='Training'):\n",
        "\n",
        "    d_net.train()\n",
        "    g_net.train()\n",
        "\n",
        "    total_g_loss_train = 0\n",
        "    total_g_loss_train_TB = 0\n",
        "    total_d_real_loss_train = 0\n",
        "    total_d_fake_loss_train = 0\n",
        "    total_d_loss_train = 0\n",
        "    total_reward = 0\n",
        "    total_log_Z = 0\n",
        "\n",
        "    k = 1 # update the generator k-times than the discriminator\n",
        "    if epoch % k == 0:\n",
        "        # unfreeze\n",
        "        for param in d_net.parameters():\n",
        "          param.requires_grad = True\n",
        "    else :\n",
        "        for param in d_net.parameters():\n",
        "          param.requires_grad = False\n",
        "\n",
        "    for i, (real_image, _) in enumerate(tqdm(dataloader, desc=f'Epoch {epoch :4d}', leave=False), 1):\n",
        "\n",
        "        noise = torch.randn(params['batch_size'], params['nz']).to(device)\n",
        "\n",
        "        real_image = real_image.to(device)\n",
        "        fake_image, loss_TB, logZ_TB = g_net(noise) #\n",
        "\n",
        "        real_label = torch.ones(params['batch_size'], 1).to(device)\n",
        "        fake_label = torch.zeros(params['batch_size'], 1).to(device)\n",
        "\n",
        "        g_net.zero_grad()\n",
        "        g_loss1 = bce_g(d_net(fake_image), real_label).squeeze(1) # (bs,)\n",
        "        \n",
        "        ##############\n",
        "        # try to train the discriminator first\n",
        "        # train the generator 2-5 times than discriminator\n",
        "        # one step generation : from noise to image (problem : high variance)\n",
        "        R = (-1*g_loss1).exp()\n",
        "        loss_TB -= R.log()\n",
        "        g_loss = (loss_TB**2).sum()/params['batch_size']\n",
        "\n",
        "        ##############\n",
        "        total_log_Z += logZ_TB.mean().item()\n",
        "        total_reward += R.mean().cpu()\n",
        "        total_g_loss_train += g_loss1.mean().item()\n",
        "        total_g_loss_train_TB += g_loss.item()\n",
        "        g_loss.backward()\n",
        "        g_opt.step()\n",
        "\n",
        "        d_net.zero_grad()\n",
        "        d_real_loss = bce_d(d_net(real_image), real_label)\n",
        "        d_fake_loss = bce_d(d_net(fake_image.detach()), fake_label)\n",
        "        d_loss = d_real_loss + d_fake_loss\n",
        "        total_d_real_loss_train += d_real_loss.item()\n",
        "        total_d_fake_loss_train += d_fake_loss.item()\n",
        "        total_d_loss_train += d_loss.item()\n",
        "        if epoch % k == 0:\n",
        "            d_loss.backward()\n",
        "            d_opt.step()\n",
        "    \n",
        "    L = len(dataloader)\n",
        "    g_loss_train = total_g_loss_train / L\n",
        "    g_loss_train_TB = total_g_loss_train_TB / L\n",
        "    d_real_loss_train = total_d_real_loss_train / L\n",
        "    d_fake_loss_train = total_d_fake_loss_train / L\n",
        "    d_loss_train = total_d_loss_train / L\n",
        "    total_reward /= L\n",
        "    total_log_Z /= L \n",
        "\n",
        "    g_net.eval()\n",
        "    with torch.no_grad():\n",
        "        fixed_fake_image = g_net.predict(fixed_noise).detach()\n",
        "\n",
        "\n",
        "    data = {\n",
        "        'g_loss_train': g_loss_train,\n",
        "        'g_loss_train_TB' : g_loss_train_TB,\n",
        "        'd_real_loss_train': d_real_loss_train,\n",
        "        'd_fake_loss_train': d_fake_loss_train,\n",
        "        'd_loss_train': d_loss_train,\n",
        "        'total_reward' : total_reward,\n",
        "        'total_log_Z' : total_log_Z  \n",
        "    }\n",
        "\n",
        "    # If using wandb\n",
        "    # data['examples'] = [wandb.Image(i) for i in fixed_fake_image]\n",
        "    # data[\"reals\"] = [wandb.Image(i) for i in real_image]\n",
        "\n",
        "    # wandb.log(data)\n",
        "\n",
        "    # torch.save(g_net.state_dict(), run_log_dir / f'generator-{epoch :04d}.pth')\n",
        "    # torch.save(d_net.state_dict(), run_log_dir / f'discriminator-{epoch :04d}.pth')\n",
        "\n",
        "    run_log_dir = \"/content\"\n",
        "    torch.save(g_net.state_dict(), run_log_dir+\"/\"+ f'generator_gfn-{epoch :04d}.pth')\n",
        "    torch.save(d_net.state_dict(), run_log_dir+\"/\"+ f'discriminator_gfn-{epoch :04d}.pth')\n",
        "\n",
        "    if epoch%1 == 0:\n",
        "        plt.imshow(fake_image[0][0].cpu()), plt.show()\n",
        "        for k, v in data.items():\n",
        "            if k != \"examples\": print(k, v)"
      ],
      "metadata": {
        "id": "HPkp4OOYmcqI"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}