Futurama Super-resolution.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "Futurama Super-resolution.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.5"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "e86f6f17529a43fd9e3c2ddfd51d12a2": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_39a776452218465b905042eb9351ec32",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_083ef8fbd4cf47e18149955da842e3ae",
              "IPY_MODEL_6c918e1371bc48ecb09522bee231be17"
            ]
          }
        },
        "39a776452218465b905042eb9351ec32": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "083ef8fbd4cf47e18149955da842e3ae": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_d4c316f47f07469684c152462be36e5c",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 31152,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 31152,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_7492b3d0978a476fa8bddf0baf073821"
          }
        },
        "6c918e1371bc48ecb09522bee231be17": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_649838e3d84a47c8917e1094e6c7a828",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 31152/31152 [05:25&lt;00:00, 95.84it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_39e138860d234b1288c7bf6b36dc2812"
          }
        },
        "d4c316f47f07469684c152462be36e5c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "7492b3d0978a476fa8bddf0baf073821": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "649838e3d84a47c8917e1094e6c7a828": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "39e138860d234b1288c7bf6b36dc2812": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "c64de03b6e8d4ef8a62324d55c9ca3ab": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_5e405b6157334f8985de00565c8e4724",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_86f14f9d3e0149b9abb382f19e7394b1",
              "IPY_MODEL_e28ec762b2174d75a38f3f6ce4697222"
            ]
          }
        },
        "5e405b6157334f8985de00565c8e4724": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "86f14f9d3e0149b9abb382f19e7394b1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_1f014c4d267249b9aa5a98710b0e63b9",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 244,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 244,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_f961d5c692be43a5bb9ecadb1cdb444a"
          }
        },
        "e28ec762b2174d75a38f3f6ce4697222": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_1852e8f8338942b786acd7e4512d384e",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 244/244 [01:36&lt;00:00,  2.54it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_a15bfcca2edd4146a772d48e5f0dd0c1"
          }
        },
        "1f014c4d267249b9aa5a98710b0e63b9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "f961d5c692be43a5bb9ecadb1cdb444a": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "1852e8f8338942b786acd7e4512d384e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "a15bfcca2edd4146a772d48e5f0dd0c1": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "2e6d5a4c8be945e5aa2d83885f13c400": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_ca41a502284842e497deae777f197e96",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_c6325c0e129c44b4a279bcf07a4649d5",
              "IPY_MODEL_a2ae75bb715c493f8970e3d3dd5fe281"
            ]
          }
        },
        "ca41a502284842e497deae777f197e96": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "c6325c0e129c44b4a279bcf07a4649d5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_c5fb444e0cfd4025b7962709cd4d124a",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 244,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 244,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_340915558c81434885f812e55c58a9ec"
          }
        },
        "a2ae75bb715c493f8970e3d3dd5fe281": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_489f113ddd254f81acf09a52d63174f0",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 244/244 [01:35&lt;00:00,  2.55it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_f812199881394f989cf52dc42f154474"
          }
        },
        "c5fb444e0cfd4025b7962709cd4d124a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "340915558c81434885f812e55c58a9ec": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "489f113ddd254f81acf09a52d63174f0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "f812199881394f989cf52dc42f154474": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YO1SOtFPZwHG"
      },
      "source": [
        "# Futurama Super-resolution\n",
        "This project will aim to create higher resolution versions of early Futurama episodes. Luckily, the later seasons are in 1080p, and will therefore be used to generate ground truth and from them downscaled inputs can be generated.\n",
        "\n",
        "First, we need to generate the training data. This will be done by converting futurama episodes into a set of images. Initially, let's start by attempting to do superresolution on 64x64 -> 128x128 squares of the image. Since the old episodes are in the format 640x480, I think a resolution of 1280x960 would be good enough. So the order of increase in resolution is 4 times as many pixels."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bfly8XV2ZwHK"
      },
      "source": [
        "import cv2, random, os\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "from PIL import Image\n",
        "from torch.utils.data import Dataset\n",
        "import numpy as np\n",
        "from torch.utils.data import DataLoader\n",
        "from torchvision.transforms import Compose, Normalize, Grayscale\n",
        "from tqdm.notebook import tqdm\n",
        "from torchvision.utils import make_grid\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "import math\n",
        "\n",
        "low_resolution = (64, 64)\n",
        "high_resolution = (128, 128)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ibk37sk8ZwHL"
      },
      "source": [
        "## Generating training data\n",
        "Now we need to create the data. This is done by grabbing random crops of a Futurama episode.\n",
        "![Futurma shot crop](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/futurama_shot.png?raw=1)\n",
        "\n",
        "The cropped image is then downscaled to a lower resolution, which will be the input and then the high resolution is the target for the model.\n",
        "\n",
        "High resolution                                          | Low resolution (downscaled)\n",
        ":-------------------------------------------------------:|:----------------------------------------------------:\n",
        "![Shot crop high resolution](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/highres_crop.png?raw=1) | ![Shot crop low resolution](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/lowres_crop.png?raw=1)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7dU_y-f1uGWe",
        "outputId": "1f13468c-79ec-4f9c-b9ac-72594dec0df1"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 66,
          "referenced_widgets": [
            "e86f6f17529a43fd9e3c2ddfd51d12a2",
            "39a776452218465b905042eb9351ec32",
            "083ef8fbd4cf47e18149955da842e3ae",
            "6c918e1371bc48ecb09522bee231be17",
            "d4c316f47f07469684c152462be36e5c",
            "7492b3d0978a476fa8bddf0baf073821",
            "649838e3d84a47c8917e1094e6c7a828",
            "39e138860d234b1288c7bf6b36dc2812"
          ]
        },
        "id": "RRnFP4R-ZwHL",
        "outputId": "639184e9-0f1a-4baf-ba0d-6a2096919596"
      },
      "source": [
        "def img_to_tensor(img):\n",
        "    return torch.from_numpy(np.asarray(img)).permute(2,0,1)[[2,1,0], :, :].float() / 256.0\n",
        "\n",
        "if not os.path.exists('highres'):\n",
        "    os.makedirs('highres')\n",
        "    \n",
        "if not os.path.exists('lowres'):\n",
        "    os.makedirs('lowres')\n",
        "\n",
        "cap = cv2.VideoCapture(\"/content/drive/MyDrive/Colab Notebooks/video/sample0.mkv\")\n",
        "\n",
        "s, frame0 = cap.read()\n",
        "\n",
        "# Skip the first 100 frames\n",
        "#cap.set(cv2.CAP_PROP_POS_FRAMES, 1000)\n",
        "\n",
        "cw, ch = high_resolution\n",
        "\n",
        "max_start_crop_x, max_start_crop_y = (frame0.shape[1] - cw, frame0.shape[0] - ch)\n",
        "\n",
        "frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT) - 5)\n",
        "\n",
        "lowres_data = torch.zeros((frames,3,64,64))\n",
        "highres_data = torch.zeros((frames,3,128,128))\n",
        "\n",
        "i=0\n",
        "for _ in tqdm(range(frames)):\n",
        "    if not cap.isOpened():\n",
        "        break\n",
        "    \n",
        "    ret, frame = cap.read()\n",
        "\n",
        "    if frame is None:\n",
        "      continue\n",
        "    \n",
        "    cx = random.randint(0, max_start_crop_x)\n",
        "    cy = random.randint(0, max_start_crop_y)\n",
        "    \n",
        "    crop = frame[cy:(cy+ch), cx:(cx+cw)]\n",
        "    lowres_crop = cv2.resize(crop, low_resolution, interpolation=cv2.INTER_AREA)\n",
        "    #lowres_upscaled = cv2.resize(lowres_crop, high_resolution, interpolation=cv2.INTER_CUBIC)\n",
        "\n",
        "    lowres_data[i] = img_to_tensor(lowres_crop)\n",
        "    highres_data[i] = img_to_tensor(crop)\n",
        "    i += 1\n",
        "\n",
        "cap.release()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "e86f6f17529a43fd9e3c2ddfd51d12a2",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=31152.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hWKk2XIQZwHL"
      },
      "source": [
        "### Loading data to tensors\n",
        "Next up, we need to load this data into the tensor format for use in pytorch. We define our own dataset `ImageInputTargetDataset`, which will be able to generate input-target datapoints from images in two different folders."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ot85nPq8vojJ",
        "outputId": "6781502b-91aa-477a-878c-b69fd368f81b"
      },
      "source": [
        "if torch.cuda.is_available():\n",
        "    device = torch.device('cuda')\n",
        "else:\n",
        "    device = torch.device('cpu')\n",
        "\n",
        "def to_device(data, device):\n",
        "    \"\"\"Move tensor(s) to chosen device\"\"\"\n",
        "    if isinstance(data, (list,tuple)):\n",
        "        return [to_device(x, device) for x in data]\n",
        "    return data.to(device, non_blocking=True)\n",
        "\n",
        "class DeviceDataLoader():\n",
        "    \"\"\"Wrap a dataloader to move data to a device\"\"\"\n",
        "    def __init__(self, dl, device):\n",
        "        self.dl = dl\n",
        "        self.device = device\n",
        "        \n",
        "    def __iter__(self):\n",
        "        \"\"\"Yield a batch of data after moving it to device\"\"\"\n",
        "        for b in self.dl: \n",
        "            yield to_device(b, self.device)\n",
        "\n",
        "    def __len__(self):\n",
        "        \"\"\"Number of batches\"\"\"\n",
        "        return len(self.dl)\n",
        "\n",
        "device"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "device(type='cuda')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q4IwYMZtZwHM"
      },
      "source": [
        "def imgpath_to_tensor(img_path):\n",
        "    return img_to_tensor(cv2.imread(img_path))\n",
        "    \n",
        "class ImageInputTargetDataset(Dataset):\n",
        "\n",
        "    def __init__(self, input_tensor, target_tensor, transform=None):\n",
        "        self.input = input_tensor\n",
        "        self.target = target_tensor\n",
        "        self.transform = transform\n",
        "    \n",
        "    def __len__(self):\n",
        "        return len(self.input)\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        input = self.input[idx]\n",
        "        target = self.target[idx]\n",
        "\n",
        "        if self.transform:\n",
        "            input = self.transform(input)\n",
        "            target = self.transform(target)\n",
        "\n",
        "        return [input, target]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dkM0XwF3ZwHM"
      },
      "source": [
        "With this dataset format defined, we can create a `DataLoader` for it. This class can load the datapoints in batches."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "3vOku6ngZwHM",
        "outputId": "0110929c-9659-4f96-e5e2-2b74a046dfa3"
      },
      "source": [
        "batch_size = 128\n",
        "\n",
        "dataset = ImageInputTargetDataset(lowres_data, highres_data,\n",
        "                                  transform=Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)))\n",
        "\n",
        "dataloader = DeviceDataLoader(DataLoader(dataset, batch_size=batch_size, shuffle=True), device)\n",
        "\n",
        "loss_fun = nn.MSELoss()\n",
        "to_device(loss_fun, device)\n",
        ";"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "''"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pSHk4udBZwHM"
      },
      "source": [
        "def denorm(img_tensors):\n",
        "    return img_tensors * 0.5 + 0.5\n",
        "\n",
        "\n",
        "upscale_factor = high_resolution[0]/low_resolution[0]\n",
        "upsample = nn.Upsample(scale_factor=upscale_factor, mode='bilinear', align_corners=False)\n",
        "\n",
        "def display_imgs(imgs):\n",
        "    _, axs = plt.subplots(1,len(imgs), figsize=(4*len(imgs),4))\n",
        "    for i in range(len(imgs)):\n",
        "        axs[i].imshow(denorm(imgs[i].cpu()).permute(1,2,0), cmap='gray')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "id": "aQkkA-2IZwHM",
        "outputId": "611025ae-f23b-4ad9-c826-036e83a78e97"
      },
      "source": [
        "for lr, hr in dataloader:\n",
        "    print(\"Bilinear loss: \", loss_fun(upsample(lr)[0], hr[0]).item())\n",
        "    display_imgs([lr[0].cpu(), hr[0].cpu()])\n",
        "    break"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Bilinear loss:  0.0004935362376272678\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3MbHlWxKZwHN"
      },
      "source": [
        "## Creating the model \n",
        "The model that we will use is a simple transpose convolutional neural network."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fOfntOQzZwHN"
      },
      "source": [
        "class ConvSubPixel(nn.Module):\n",
        "    \"\"\"This is the first working attempt model, it uses convolutional \n",
        "    layers and then a subpixel convolution.\n",
        "    \"\"\"\n",
        "    \n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "\n",
        "        self.conv1 = nn.Sequential(\n",
        "            nn.Conv2d(3, 24, kernel_size=9, stride=1, padding=4),\n",
        "            nn.BatchNorm2d(24),\n",
        "            nn.PReLU()\n",
        "        )\n",
        "\n",
        "        self.resconv = nn.Sequential(\n",
        "            nn.Conv2d(24, 24, kernel_size=5, stride=1, padding=2),\n",
        "            nn.BatchNorm2d(24),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Conv2d(24, 24, kernel_size=3, stride=1, padding=1),\n",
        "            nn.BatchNorm2d(24),\n",
        "            nn.PReLU()\n",
        "        )\n",
        "        \n",
        "        # 3072x4x4\n",
        "        self.sub_pixel = nn.Sequential(\n",
        "            nn.Conv2d(24, 96, kernel_size=3, stride=1, padding=1),\n",
        "            nn.BatchNorm2d(96),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.PixelShuffle(2)\n",
        "        )\n",
        "\n",
        "        self.convf = nn.Sequential(\n",
        "            nn.Conv2d(24, 3, kernel_size=3, stride=1, padding=1)\n",
        "        )\n",
        "        \n",
        "        self.final = nn.Tanh()\n",
        "\n",
        "        self.upsample = upsample\n",
        "        \n",
        "\n",
        "    def forward(self, xb):\n",
        "        upsampled = torch.atan(self.upsample(xb))\n",
        "\n",
        "        out = self.conv1(xb)\n",
        "        out = self.resconv(out) + out\n",
        "        out = self.sub_pixel(out)\n",
        "        out = self.convf(out)\n",
        "        out = self.final(out + upsampled)\n",
        "\n",
        "        #out = torch.clamp(out, -1.0, 1.0)\n",
        "        \n",
        "        return out"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gsYCqbPTck_G"
      },
      "source": [
        "class Discriminator(nn.Module):\n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "        self.conv = nn.Sequential(\n",
        "            nn.Conv2d(3, 16, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(16),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Conv2d(16, 16, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(16),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Conv2d(16, 16, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(16),\n",
        "            nn.PReLU(),\n",
        "            \n",
        "            nn.Conv2d(16, 16, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(16),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Conv2d(16, 32, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(32),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Conv2d(32, 64, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(64),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Conv2d(64, 128, kernel_size=4, stride=2, padding=1),\n",
        "            nn.BatchNorm2d(128),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Flatten()\n",
        "        )\n",
        "\n",
        "        self.fully_connected = nn.Sequential(\n",
        "            nn.Linear(128, 32),\n",
        "            nn.PReLU(),\n",
        "\n",
        "            nn.Linear(32, 1),\n",
        "            nn.Sigmoid()\n",
        "        )\n",
        "    \n",
        "    def forward(self, xb):\n",
        "        out = self.conv(xb)\n",
        "        out = self.fully_connected(out)\n",
        "\n",
        "        return out\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SGGm9h3oZwHO",
        "outputId": "2f35b671-3023-4d52-8769-10d7b553a121"
      },
      "source": [
        "generator = ConvSubPixel()\n",
        "discriminator = Discriminator()\n",
        "to_device(generator, device)\n",
        "to_device(discriminator, device)\n",
        "\n",
        "def count_parameters(model):\n",
        "    return sum([p.numel() for p in model.parameters() if p.requires_grad])\n",
        "\n",
        "print(\"Learnable parameters - Generator:\", count_parameters(generator), \" Discriminator: \", count_parameters(discriminator))\n",
        "\n",
        "def upscale_image(img):\n",
        "    out = generator(img)\n",
        "    return out"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Learnable parameters - Generator: 47311  Discriminator:  190121\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 262
        },
        "id": "EleIaJjIZwHP",
        "outputId": "7f03f709-f56f-4d25-d253-d1008e11312b"
      },
      "source": [
        "for lr, hr in dataloader:\n",
        "    print(upscale_image(lr).detach()[0].shape)\n",
        "    bilinear = upsample(lr)[0]\n",
        "    display_imgs([lr[0], bilinear, upscale_image(lr).detach()[0], hr[0]])\n",
        "    break"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "torch.Size([3, 128, 128])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6QAAADkCAYAAACPD+ksAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9SY8lTZfn9TvHzN3vEGM+wztU1VtVXfU2dCOBREm0xAYWIHWzoCVWwAfoFRJbvgYLNr1gDRuW3UJCQmLbq14UokulVhdU9Ts8Q2bGcO/1weywMHO/fm/ciLyREZEZ+Zb/nyFOWJi722x27PztmJgZEyZMmDBhwoQJEyZMmDBhwqeGfu4ETJgwYcKECRMmTJgwYcKEv52YFNIJEyZMmDBhwoQJEyZMmPBZMCmkEyZMmDBhwoQJEyZMmDDhs2BSSCdMmDBhwoQJEyZMmDBhwmfBpJBOmDBhwoQJEyZMmDBhwoTPgkkhnTBhwoQJEyZMmDBhwoQJnwWTQjphwu8wROQfisi/EpG/FJH/4XOnZ8KECbuY+uiECa8bUx+dMOHl8SSFdOqkEya8XoiIA/4n4B8Bfx/4b0Tk73/eVE2YMKHH1EcnTHjdmProhAmfBh+tkE6ddMKEV4//CPhLM/vXZtYA/wvwjz9zmiZMmLDF1EcnTHjdmProhAmfAP4Jzw6dFEBE+k76f9/3wExntnSnT/jkhAlfPm7DNZu4kU/wqd8D/r/R738N/IP9SCLyT4B/AuDxf3bmLz5B0iZMeL141X1U/J+dVqdgEVCEPpkGBphhkn6FFAdALGIyliXL9gFZRvGzjIA8LINgB2UwUcSszxyYIZDjWI6zKzOO3+c1pxWzB2QFDMFS+vqyQvJ/vdyXl+SQ/v+CjMP78n1EWeyWS5///bKwVD+W03qsjOU0jeVj6/YRdf6kdgHcyXOSo2guXzC2chTJbXwc30AcKsqqu2XT1a+zj07z6IQJj55Hn6KQPrqTLvSEf3T5X915kcjjxhTrJ6bPjGPS/ZS0jt8/fs993z0mzkN4LeX6sbivvO6L81C8Y75xHz70zn/+9n971DdfGmb2T4F/CvBV8Y3t9FHpf3zcvG9YL3xaPDLdB9N5xKPj99vo4Ye+28eTYz+y/+SX2k3lcLnslL0cKhc7KsuPaaPWKzT34DX30a+Xb+w//zv/KatNx9y+ppBIoQ1iHgkRbTtqcYgFiq6h1goxmFvDRksAZrFmIyU2yAVRlHnYUGtJEGURamot6ERZxppGPK14FnFDm+VllhspWMYNQRy1FCxjTRRloyXLUGMibLRkERpESHLsUKD2JbMQcEDtPWWIOIOmcPgQcdGoC8UFw8VI7RUX2YZHw0dj4xU1w4fIxjvEjCJEau8QjDJENk4xgzJ01M4Dwix2bNRjQBUaNuIJIsxDQyOeIMq8zz9ZxtGKYxY3dOJoxLMIGzpRGilYxA1BlFpLFrHGSPmfxwYDNloxjzUAa62YxQbFWA1yZKUVVWxxvWwd3gIrrSiso7COlc7wFiit3ZHXWqEWqWKWMWaxYaUVgjGPDeuhLaR2YSIsYk2tFRFhERtq9QSUWdzQSGoLi1iP6r+mFUcrnnmst2URa4Iom1FbWGvJMpfFrVQs4gYDbl3FnA0A1zKnYgMWeC8FajXQ0VkJrIGWTs9ZVCX/19/8H5+q+x2FB+fRjEevdT8wTn1KHJv2j15H5vnhTp735o2D7/9AnD7tr6k8Pwr3zKHwadfvQ3k+81r3KQrpUdjvpI98dpDHneEYxeO14ClK5VPy9trL5aVxX7nfN1B9DL6AMv4b4A9Gv/9+DvtkGCaYwWLzOtEP8iYfVirHCtQ4/qNgZCvLI59/xWX4sdhvI325bBXTj90OSTiozr6ecnx0H7UoWG3M4hrvNggeixWiZwQ2bOwdXTxFYkcX3qNyiRFYxR9BLgBjHX4Adw4om/hbTM/APOv4W+AUkZJ1/A2wRGXGOvwWdInKgk34DegClRPW4Tegc1TO2cTfgFSoXqb3SInwVXqPeIxvWNtvwRSTn7Cy73OGfk7kRyAQ7ec0/IhIQ7Cf43iH6IaOn6HyHtFbAj9B5BbVGyLfAmtErjD7BmwDvMfi15g1bOw9xK/AOurwDrjEzGjDD4hdAsJtfAuaymIVvgc5Q8SzCd+DniFSsgnfYXqCyizlX1JZ1OFXmMxRPWUTfw0yQ/WcOpeF45I6/hbwqHxFE78DJMvfA4bKV7TxByBk+S3Q4ORrOntHZxtUvqKLV3S2RviKYNeEeIvwhmgrNvEK4ZJoGzbxPXBJtIZNfAtyQbSOdfwR5DzVf/wek3NA2MTvMDlDzLEJ3+X6L1I4Z4hU1PHXmJwgMmcTfp3agp7kOl+ieprzPMeN86/n1PZbxApKvqKN32EopfuKxn6LAYV8k+pDIkv/DXX4kUDDmVxwE3+gsYiffUXX/JrYXXOGQ7uTT9mHnzyPSrao34f7NjNlu0P35StTH0I/n35MHj8wF5u9/jXIUyGZOXJQX//Ide59be6l1r5PUUg/+2J3woQJD+JfAL8UkT8m9c3/GvhvP3UiBoXjuMiPxzOOjfKYWWs/rXZP+MFI7DIFj8GhZD30/OeYfO8rkyMee4nk3tvuxszNz4uP6KMGYgQrcWE+sip0EA2iR2gRIoJHrSEtVDxiTc66R6wFBLMiPSsxyXQIRrQCsYBIjZkHi1kusrzB8Ik+KmvMHABqa+Igr4iWaMLKCotpga12S0wsWNRuCTFRM9EbgoVUL3pDZ22m3d4QYpvZuLfE2BCyjLVYFIRbsJDlFUbEoqCyBjOiKRrrnDeH9vmPDiXJMXpEOyDm8kplkfIfEGqieZAI1ATzJJrsJsXBUNsQzeWtlL5cBLE1Zj1leo2ZbMMRMM0yYA6xlG4zh9hmK5NlfJZDSis1sC8XiDU5P0WucxvJjOSA4VHrUhszD9a3BQ8SgFTnYjEp/7kshJR/EUPYEHGZhrxJaRZQcv5FUFsTLFGgnd3m7SeHY41DiRQ4rcEcnVWUtqYVpdMFqM/995N13ifPo8cO9Q9tvQny8Qrb58AD1rwedxSep+TtQ89+KeX2BDx3Fj91m3uKQvoqFrsTJkw4DDPrROS/A/53wAH/s5n9+eNewpFK08MRh4Gtj3o3wsfj2bWZhxNzR8F+yUn0/kQcF+dTTsKH0nR0vR5K7DMdD3votZ951/xj+qhIBIWWOT6cIBJR2UDcQACJFU42iIFSILZKbdZK1Fb5wyVqa0QgWpUUGDMiVVZgI5EyKbZWEynRLAdK1AJi1wSbIQTUrglWIUTUrghWAYbjPZ1VYIbq2yyD6o90NgMTxL4nWIWZ4u07ulgkOX5HFx0WlUJ+S4iOEJVCviNGR4iOQr4nRkeMHs+PmCkxFjjeQZZV3mMmWCwRrlOVxwphlWQrEVsjxFxGG4RItAqlQQkEqpR/1gTKpKjJBqFX4K+JFCn/XOeyi6hdE60EMdSuiCSarDKS7YpoBWYOtSssy24IL3K4z/I1Zi7LNxgOo0TtJuWZErXbrUwvVyip/o0qKYiQ67wGsZznOp0TtRKhQYipbq0FaqKViHSo1antWKr/SJLFrnKeI2JXmJW5LaQ8ixnKe9Q8iOHtHdgCQRF7z4w5UGJ6TUUJsaAM72jxNO6cH10A+XQ3Fj7XPGrYw1THI4+IHL25+5lxDK/lS8rPq8Ax7eiZ8Snr6KMV0mfppNt3DfIx1NzH0nefQst86L33xx99e9QpdSTHeyr4KDryvenZI94/AS99PvaxeCzF+b74T20Lx+A1UcrN7J8B/+xpL9nSYe6fZB4x+ewrAU88pwpPoPqME7B9272/yfBzdCZlPz8GPETlfWqTeER5fbLdzYfOthygON+b9mdoCw/iwPs/t9XhsX00Rod0c0404vUdMCfKGbS3dKLcFhW+/h7DE905ZXwPKK0/w4V36R3+AhfeI2aE4gIXroFA8BdovIbYZfkWjQ3Bv0HjCombLG/QuKLzb5BY48ItnXuDWIsL17TuDWIdvruicW8QC/juHY1eAkbZ/UidKcOz5ntqd4KJZ9b8hlrPCJTM61/R6AWdzJjXv6bVM4IsqOpfEXRJkCVFc02UOUHnlO01JhVBZxThFsMTdIYLa0CJWqGxAQRcicQ2tQJXQuxS4fpZsjJbwPwCsQ0SO8yfIXGDxIbozyCukbChzeEabgnuErEGF24J7gKhQ+M10V+CBVy8JmiizLpwRdAzQND4LsuKC+8I7hTwuPiWqCcYRZJliWmV5TnmZrjwligzol/gw1uilES/xIW3IAXBn6Q6F0fwZ/jwDkQJ/hwX3wMQ/BkaryBGgr/AxWvEOoK/xMVbJDZ0/hKNt0isCf4SiWs0boj+ErEaDash/+Oy0HBFyG2h6K5o3SVihm/f0eo5JqDdDapnKA7frFOexePqOQFPFKUrv8HbDS6umNcrgjqILz+X93iWeZT7KZB34mS82Dj4zHiKrwiRA+dGd1/+8UdjPvjx0brEjgj/VPjAmdiBggwHldNPsc59CTzpDOlzddIJEyZ8GUh+Kj/EE/2ApfTACP8sE6/s//rIdw66ZX9u524693OXLCzHvfvuLx+f5y9zutnioXb0peft5WFEIp1JcmQkkhQoSJbI2GLZs67GNpezoLEZFioSe+puDu/lntIrglqLYiCaww0Rh7MWkYiQaK9CRNRlC2JA1OHYhjtrEQwVhyPRQfv3gCCiqLUYAVAkf9dwmTJap/xYB5booGYRI1FjI4bFDSEvZ8zqJIti1ERJlNlInSxrAtAg2lvZ6tEZvwaR1MuNGsheZK3ONF3FLFkNTTQrtRHEIdbnX7NlOSLiMu03pnwScv4VIaRSF0WtS+OOKmohW0EUsRSH3hOttVs5bmWNTaozDGKDZI+74zpX25Nz7lMd9nWe6nkrx736FyQ7XUrpazOte19uUtp7GcOkp0kbqAIt+VMILVhMnnVpkkIsEaNNI0WEGDssGoEie/b9Ai1reQJ5aOTbUYQORBwcSA//233Fx46fB9+3982Dz1n68H6U8abtx5Kihjh7L7CdwI/M8UjpS2d17249fxaa9BFUZ2DoO58F97Szh9rQMXhxp0YTJkz4HcDRY3+/UHgCffcjsK/ofiiZNhLuTqSPSJbdo2Lf+4Lnmtm+HLVtpzzvaUcTbet4iERaadmECuKcQsDJmhhBQqDskhdUsUgRrlBfIhaTFdQnL7OEK8S5tKLprhDns3wNQ/h1UhxUh3BRRbprTB04pehuQARzbpC5Rzbn8OEGAHMeF24BQ1xBGW4xM9R5irgimiFS4CxZYk3KbN3cYFKBtemspswxOiS+J+oJkCijpqckKvEN6BIjIvEGcyeAYfEa1SUgEG9AZwgK8RrTKilS3RVohYiDcIVqgahHuhvQpDxKd53iquDCdVIuncN1t0l5cw4Lt4Ag3mHdbapD53Eh06edw0I6L4r3kGXxHgubUXgNtk7hsYG4yXILoU6y9bLDrIPQIN6lDYvuCvE+jVch1zlg4SrVORC71C5EJMVRBypYbguiguuuc7gObURUcaP2ot1N6uNO8SGVEb5A2z7c4bvUFtQ5NN6CKNEVFGGNYGwKgRBRM8q2o4mRJhqdn6HavpzV7IUho//fH+chmuT2DftxnrKxe+h928/tWusGq1xvpd6Lo7ZNSyfdR0+H9+wLg+yruY/P9yEV+r54n3J+OroO98p8jI9i5n3AKrsb9f4tlaeU1atTSF+avvsxOM78/eHOsZvucfjoyZ1k79HKBmzPT/R3ed2N8+UsWF8C99X/x7SLx7arz03TfVG8BH33mfAcVtZh53WkU48VqvEEuQ2XAwOx7EmHhumHFfffaRyiYfV7FPeuQu7iGJrwnW/Il60Ah6jIRllajekV0Sp8qFiHWywqWEXbXSOmIHNivcrWuRNifQWAyikxXGNEnJxj4WaQY7jFrEXlgtjdAmtULrCwxmyFk0uMDWa3iFyA1Rg3KY61WT4H64h2jdNTMCNwjWcBQNteU1ABQtNeUVIgIqzDihKPR1ixpsThUdZyRYmjRFjZhkI0hdsVDsXjaGIvF3QxKYeOkjbcZEvinBBuAEFkQeiSQiiyBLsl3UF6inS3QEeUc+hWCCtULrOiuErlEtZEu0W4GMrCyUWy6LU3OLkEOqxN5QKB2F6ncsGI7RUqZ6k+2/eonAJK7N4jcoLgid07RJYIJbF7i8gCYZHlOcIyy7Mk1+8QSkR6uUDkJMsO1VNifYWIoHJGDH1bOCN0V0BE5Cy3hYDTC2K4Scq/XhDDCrNm1BZucXqJhZT/FN4Q7Ta1KVpie4PqRVKIm/eIXAIB6ndEPU+W4PYtopcIgq+/Y61vMHEU67+mdZcEmWO6ShbnGPHtd0S3/KSU3eeE3XOkY38sG6is9/z98LttR7HYZ6J86H3HrnWSBff+9ESz4f7Z7N9swpHY3pb8gfp+SSrzB/BStPJXp5BOmDDhleKAVeu4+zZfy8LhGDPvYbWxl8YqpO3Hlv7vwlYaT+x9eeS5/Elzye+YMvuRZfFBCvm4ssa/H4r3heiofVKDCRo8mMNMkKgQHTEqGl2y+JnDmcsKmSZHMgDiUHOAZtlDpp6qObCeYukAj4lDLKl7vSw4TNI31LKMZlm2cvamKqaYZBnNyy5JsmgufwUEE2H4R0As/0RQERTFoTiJ6ac5HBFFk3qaKcXOCowWTAGfvwsS3bDsS95tNZdqLyuGoqZDPns5ovndDsty8jab/MMmOYVLDgdwg2w4S+XYhydZR7LLsj8gF3vh2/iIyw6Relm33w0OxeWNir7O+3pOXpRNHMkzs2CWyrL37pvkkOL3eTZF8EM+ZZSO5NQptU8B1PxAn9ZYELNXXmdF8lAsqb3oUPYFGhPF2pHSk/St8hXNKR+JA2ONyT1j2WOZqXubqR/9vuEd+4pyfo3sz4bjd+bZX3rziQyx+qfEdqmwgzg21iC7efnQZv/+eP/F4sjKful83lee99i+nrrROymkEyZMOIz7xsSjdKHHDkyfUsEaTfxJY7wbY2/ncTzQDhvDZnubxNm+mhfTeSscxO7kLLFt0uL8YNY/UHy/E+rosU3koXjHFsIx8b4QpVQlgousQ8WyPQFRTCNlrAgR2qDMSB5soymVzDETgileZgA0QSh0gWA0QShljojRBKWQGSJGHYRCZigFdRS8VHg8myB4KfGibILipKCQJZsgqPitjKPQJXVI/aHUBU0QTKCUGV1WVAuZYZbsk4XMACMCPitv0Qyv1VCFXjwOQQ1K9YgJDqVUlxSkWFBqVhRjiXcxe9ktMC3Tt6JDNXn/DTHlRzBCIMlSsAlQSIVKwSYYhZR4KdhEcBQp/1FwUlKIUgcdyQ4VpZQT6qCoCIWc0AQHAqWc0IS0/Cr0lDZ4DCj1JMua5OhTHeoJXfQES+EhKsFkT14So9LdkRdYFNosg9AEKHWe2wIUuf6HOpfULrzMUCmznMqiCYKTCi/FkGcvbk92qAhelCYrwoWe04REGS7lnE0sEGCmF9QxtctKLyhC2jTAf4UzwAKnsWBtRo0QiiWmzUEr44R9PH6WOLSRPGyqSmKrG5aKP9eBme1sDkaNeTNJUJPB8hpGu7g6ksc1aW77W8wbVABKOEKHvp+PdCefr5TJ9po2W44pz+csx8+mkD6X19SPoVA+loJ5nMfZcfxx+IhSO6IuxHwHGEAYdcAuOzoAMNk+K6PB11kxyOXopYWV2/fE8bOPpzK/ts76XB7E9p99bD6f8u3XVqYPQu4fGMfK2jFW0KMmh/z/+99zn2OEZxi8ZV8P6b8VB/qojUKHn305yK7lNOmgms6QIUhiUaHZ4pM+KAM1p1dqH/Tcu5u8V0k7fRaL+GPL4t7XfFwaPrfH3WMQo1K0yow1cdFQBKXsjEYiXo05LWsTnBlzAutsc1qyYZUb26WrWZEUwgu3YZNse1y4FevseufSrdkgtAYXsqYm3W564Va0CDXKZZY3KBduRYewRrl0NQFhhXLhWiKwEse5JqdGtyinkhaYa5RKDMVYi1CSqryW5L5fgTZbWnpP9Sq9gx1IjnciISanQU4DjSVrpXcNjaWnCndLl/tO6Tc0WZ5JQ2tp9Fm4ltZSWZy7jo0JAeVMm95lE+euoTFogHNX05EuR7nI8gbh3NWEnLdTVxOzfOJqDFihLF0NwC3KwjUocJPT4IAblLm0+BxeSUsxyB0lHdcIpQQqOm4QvAROaLlB8RI4zeEqxiUbbnPPuHQdq2wtvnQbVqkUuXQ16yxfuJoNQpflVP9JbkZ5bhDqnOd2JHfAOssxt4Vzt8GAWxwX5PyL55wWFeGGkiXgiawMCknnlm+LiiKuOY0167AmmEMeoIy+OhyYTx/rwfUx4/1YObzvirJ739criaN5frvusUH5PDTK7hyVyH9M9F3LC2NFpF8LychBkw0KbGfJ47Vqr4z2iq0hYjseZseK8jg1kp1efUlLrh6HWD8f9ET8Atg/Djn+/r362BPTOFlIJ0yYsIuPmufvsdk9ZMq7h/ZxKOID+3Mfevjg58a/DQpmnuis/8disoJixH4nGMtXNo2dKownbkiOjhIVUS3RCyWrpIO8p2SNc3CUE4Vjs/2pJrBnNNnuLCyOKYvnxitf50riu6XtDNMUoNnECKBgUbbtmPSnSCSSNjCDRIIlhSRKJFhSQgIxxTUhSCCYI4oQCAQcgSR32V7REWhROoNOIh2S/pOY/m5KkGTx7GDoRwEZNmICikna/DHcQPMTSQpoovxFyAq0EQmW3tcR0gLZIm2mG4gEGkuWOJWWzgoEo5BAZx4VoRKjs0ThTfn2KEopSTY0ee/dKTslIkSJRARDtuEGQZKf4IgQiLmsjEDEDsqGYOmduaYijpBzGXJ4yOExjz19nLAjx704DOEBh41kGepfRm3B5bym/IQhDzrKz4FwMWJueikNkp/vZcnptByeyjTk9CGpvWzbhSNKLhPzRNW8+BVMNP9XfBnHR2Us3k3woz24PmYYHI3Hw1Sz//z+77L38+CvMlIm75nv+4fSvismWyNL1Kwkm9H1A5YI0WJSKkWIMYCAc4nx0A8UXhMrIsaISppLTZLSaQfmny9QF00w7qdvPwaPZF09B546X08K6YQJEx6FQ4PO/nxg+398AobJ9Anj81Z1HFs3826vbZfu0WL+PRItLaX6f3plNNpWJZX8nl7HFNKkqrikkorDpVNuePF4imzp2ff08BGZ+2Jn3GfAl7AgfSGoRKyMXLcVi3pOcJHObyAmxWhjhpnQiXEjRk/SaYW8wQLv6NU74V1mAQxy7nB9eB8/7X5HfpT00xA2vSwkmbTYbDLrwIBOJSkeUbHszbWxIunSQEMBGnECrXicyEBn7/8LtGnRaUZrLW2MtDGCbahjxzq2mAVaa9nQYNFlxXkDViBiiDQ4FjhRChcpbInDUzrF25xCSqIWOEv9c4WQCs+4NhtU0OsRO+E6lykivM9lJ9ggg5FcB9kd+Xo0Kl3TMwGMGzgod3vyapCNdZbbPXnTp49kccYSK6MZ1Xkzqudmv/7zs2N5nLf3o354Jzxn70pyws14P7oz8T1pMyWivJek4EQ8KwwVaLWioUGlo4pJCd5IycafEnX16r3sHqNQ7Myl9jgr6GNTcsy772VFjU2N5kbKaLwbRyCqkbsxnXZEi0SLBJ+3W8xYa5PuoVUl0g7p02ioCM55XFDEIITI0i1weNq2Zc4CL2mO7WIg5Hn70J2cXySeoS0ctMgf+80DFv1Pgc+mkD7WU+5TGtixNMun0DE7t+2YNbeDvG5vBvl6fTXIN83brdxtw1exHuRqRNn1o7QtZstBPi1/Psh/VP3dQU73kA2/HZWH14xPfdHvS3ls/iKQlb+HvL0dogPdGfCOLbZhJ/eBOj4mzoPP9haIrHpKSIqlRIJtJ8wudkQLRAt01hItEKwjWNrzD5Ynv/yPSu+axUZOVxSnLimjWuClxImnlDkzneMpKFLk4Q63ZzUxfg58ygns3rbw8f30OTdQXhrRFGmVU6thcYWYg06ImQIrCGi+79IM0zSPSO4ByVZvY4PqcDwk3ReZZYlD/HRnZrJqCiE7J2qBkC2bQnKKBP0yO5k622Esda5DJdFnK5revRAFHckLL2DQihHEaBE6STa5xvo7KVtWIRCsJlrDJrQEawlW01qks47O2mSBy787NKXHIqXeYCghQpQbFEcTHTOdE6XE6SmFzSkoKKQkkyTwkuyHlunP1vfdgTkR852tqSzGsrCN80FZsiyj8A/Iqa7saDltmm3H6z7cDsaXgzLZ0qvSH2dINOqUH2F7l2vIZw4h2Xz7fOZyEUEkJOZIfzerpCNJs/CeoCURR+NrCA3Elnl4S5vPDn8J+JDjteekY44Vsg8dodhZ4+Tv32eZG8eNeVMDkyG+qm43d82IMab/LNLEdQqzyDpuCDEQQuDa1qlsRLA8Xlmm96oIhSuYaYVXn96PUWhBWZbEaISQLK+WLat3dYovk7IL3JlPP2YN+qQjNHttYb99vhRrabKQTpgw4V70y4pHK0uPHK8enLCPVNjuDpI27MYOyqhke6d1GEkpjZJIaJFIkJaQSWmdNZme1xEkZDtpIFpvFd2zjkqfwnTuLIoSrchEuQLB4a1AxTGcOh1ZEn538Sky+UxT5BdUF2mhazshQF7kp1+Th9r0t6Qr9gsL28Y58Obd/2DsQEaGOFmZJS9ehi6aFzJJk0XyuTEkItlJSSKaJq+pji7ReSNEGhoCJv3mUQcSCNGI0hBp2AQjWo1Rs7GYr5ppCWIgEWddKgMiGgPe5UWzQOUMkUT5FWvAlNYk5yNSUKLicbikhGWa4FAMfTmOhyPhTj3Ic8jyPPLY/0SvVwxtYZC3ediNvz0fP5b7+PvPbuu/T0N/at7yZkYuOov0DTARezWfM9zSzMl07CiGWlZYesWJ199VB0X0AUWvh8AzeF7feyHcaaf3TaPD9z8AIzksst5JkTSDZS1Ym+oLow0NXQiE0NHEVd7wNTZxlSyaIbCyOm9rJNaH5HZk2UJaeY/5GaoeEGqbU7iK0+o0+dgW3dKH+/HmY9Yqrw2fYXN3Bzb6KXth+3F4IM5HYFJIJ0yYsIu9yeyhIf4OGehDg+JubnIAACAASURBVNLByfCYwffwyDkYGNkuZGxkwex3ZpMqGYh0SamUFiM5Q0ECZGtQ1HS6SSzipMNJsqKicVhgm+ZUi+C1n1DTdRSYEYKxaVtCgK5rCdahVpJ8c6abFXd3L/sSHv98Bhw7p32Cld3Oguc5LAF3MndPudnoz/fhta9s96AaiWVg1ZQs1wtEDfMNEhIrxszSScFsNXB9/zAbPGSGrIQhEAel1JKiRvpDZ1ulxrKnzFTKOuylONIVLVGgzO03iKSFReoYKCUi4LQjkrzoOrmljhXRIi7ecBMcdQxoeMc7a9lYoIgNjbR0BEoTgnSEbE2NuS9LtrSpRNR7vAqleKJ3KTwaRVXgnVABs1mFAeumYVVD3RlN03FrLd6E0jV4maezjqJD2YXQuySTnqGLsrUWSS5HSI6Yeuvzx8i9svUYOZIWc4fkcfzeKVSEnLeUB5drNtBfvMO9cjrNmxT8YMNFPTl8HCeNizFvgLj8HgGcpLaWTLX5mp2epm3pfGrrKjozorXM2kAwoxPPxn+N6Xpnk+RVwkYzY5bvZ3odT6s9BocYTP2Z1b5/H37irjVsDFOjLZokS2Ql1zRNTds1rO1qmG9X7RVt19K1HU28SSwjAkadzzb3X03tp7LUZr2ms+H9GfLGVwSDm9satQvK4oRvz77lJ/4P85VUIFGR4VBxauFfMqvtuZTR4yjj2//3GJ9rPsrp64G29rH45ArpYzjeL+ENF3YLzsm2CCx794J00L9Ho1uPuNfh/SDf1teD/Hb9bwc5Fu8G+Y9+73yQf7HYesH99ttt+C//zt8b5G++uRjkZbk9Z1b5apD/x//1nw/yv/mrH7bf3fHEO4gv1jmP8z78sgPDp6DWHuPh+SleoF8bhgHm6Al/NJk+pE/JOPbj05Ow3T/vLTPD+c/BhtnTbjPFNnYE62hjTaTNCmmDuoiqUZWKc4JXKNRQBVXBF4pzDl8os1mB80pROFzhcCo4JzhNCzMniaobQmS1afj1D++5XtW8v25om0AMHV1s86ScztyNi9cOSE/CI8r603iWPbKNPPkLY9jOrv+DXqO/oG4ao+Iax6m1yHyNRAdBEt3TUj4jBpaslsNMZjIoTWLbWXAsaz4fDeDYOhLSQaXZlmGygrl87Upy4mX5faS9GUIElS4pRwLCLZY4B0TeEqxjExvquKK1hnXc0Niajo41BtalTSN1qBhOUr68pDG3VKFwjrLwLMuC+azkdDHDNPXXqnIUzuO8o5w5Zn5J1xnv313z17+54v3thqZZYURUWgJJ6Y0SMYUuCslJmQ2KVSIA26DIG2Olfqt0HSdLthL1iuLWqphJtCP57rOB7W2ux8h9WxC2CvQ2vLcU2/DlnsB7V2bYvNhaK++Pk9La7x3a9gIPS4qNZrV265Hc0DaCKFGEznmIG1xsWIRbOnXHmfS+MLyEN9UPUYY/dNRmnJ4QApuwoWkamq7mffyBNrZ01tKwojdfJ0p9xCziXOqD3jtOzjyz+Yz5YsFiWVAWBbOqYubyuVEVxBxeHfNZxW/fX/HdDz/yL//ln9PSIuQjANEGJ4RyN7VfJj5gGX1152MFdOdo4MObGcdgspBOmDDhk+FYZfTwJLpVPiFT+iy7GxLLimiko8uW0LhdYNIRJCmkJh24FvXgHJRzT+EV7wTvkqLpnVCWnqJwlKVnuawGuawKnCreK1631lEVpe0C72/WrNs1wTpu1tCFlM7BcntMEXzqOeeFlMTP9JnP8rVPi97Cmam3yPFZHXa/+19sFL7f77YKUm9b6a1s6YcbLKdI9mxrIXtejXQWsvpp2XtsjRFINpYVwTrWMWLcEqymJoBscHSYKoUzCoXCK4WSFVFFJW0azQtHUXiqsuBiMWM+r1iczCi0whUOP3PMxKNeYe5Z6JK2CZReub6uaeuOwaZnI9VqZzPO9n7eV9ayG6dfnMlYHr1q5wzf3p/2vr4Ne6BNGwMt9s6n74w5B9K6983H9J6dNO9nYBTLrLeh3n1z2heSnfY8xLb8uyWfx73v5VeN+yr0Od/9AIbNv+GRfltiu6E77v/DBq+k+SoQhnOg0RJ9OljHTbxiU29o25oVbzE1cJFFGfDO4dQhFIgKqkJRQVkUVFXJm6/mLBZzTk9OWJ6UlGXBbDajcom9IQhOPF49VVXxF3/za/rbSF0+Z+zE0VP/t8kfZfQlhvzPPWfLPfIx3+y725EdWrhnr2c8XPB8ltwxJoV0woQJByE7M+qx8e3J1Mx+frHRKGp5sRgtn/fsabjW9Sc+CZLOhQY6TMJAtRUfUIlU2qHeUAdFVVCVQlkop8uSqlAK75LHTacUXimrgrJwVGXByXJGWaaFb1UWuByvt5CqCCLKpmn5zY9X/Hj9jlWzQV1E1CEqiVq0vXhtp+R2M/88OHa6GDw93tX/XwC5lRyyvttYeLgRfWjX/3cZqtnLblOxWC0RNXDNoCimc5nZcse+nKjnbSRfBmJ0lmRI17E4eiU3v1EYLFgmki4uypqCUaKSnP50MkOtpbCGd1HpYgPhPd9bQxtbsDVr1nlxGzGpB3ualy45xXHK0kHlHOWi5LL0nFUFJ+enLH1k5kCKBd4ZhYvMT0/QUnEzx0+/voTK0VbKt/6UIMoVxtcqBIG3GJdWsNm02GnH+u0P2Mb4qytog+JQvDkKURzJAuclImI0IS2BVYQu5gWyQJepzw4SxVn2ZIGQLTlJTnXoJF97QlqEhVxHLlsSE+1WRuH5+plsHT5GThZKG6i5IYeTv9fzrwI2yB3bu1+7/KxkWUd5U0lxgrGTZ5W+XMbxk5MjB3R5W8MLdFkBTW6iNPdoRTI93HxJsIBZYBY6gilBZqzdDLTmtVN2x0yjQ85hDrGqjrV+PWjxHHm7tVH83vJoFrDYZv8JRsh0uqiRumjoQkdrDde8o65r2tCwDte0XUvoatru+0z/N05nHaenJ5ydLvl7P7/g4vyMk5MTFgSqsmQ2m3F5fspisWC5WPD12ZKyLCh8kS9cSuXg8y5KutJFIHtd9rKiuX3LUgA9pXSXLPwppc1woSCEMLSZF8MHLJc9nmybtdE79r55n3zMN4fztcaRbKDDue3n7PFfY4wHYn48PptCegzV8hjvu0fBdnfT4oiC28pmkG/bLf31t+tfD/Lb+leDfHKxpfX+0S+29Nr/8s/+vUH+j//Df3eQf+9yNsgLt03HuCL9yHW2jmmgo/Cbdpvm3373/SD/ZP5ng+zC6CqJV0Jn+ZRU1mO+dex3j2mTx8T/3cBj2tLHUjN3FZDx7m2E5IwoW1u62CQKrrW0scmeNRuitCARXMB5wykUhTKrHEWhLJdzFvOCWeU5O6uYzzyz0nM695TeUXih0KRkJkquy9ZSR1V6fFZCvdeBXqR7uuStRK68gAVi19F1AbUZSkmpFV6K5DAFeVVeIg9NQS9JZT3okn5HV//wh0d+P4/65n78L5XcFWPysntiDTrfoNEhQYnSJSuSSFKCSCTKOOQ6u0Ey2y4GSUpqb0lLymhvSyHHkUxTTQs/l88FJhrpmmBCK4KyoiVwS0tjtzTWsGbNxm5o6UiE9Q1RkkMb0YBiOFFK7yi9cloVXJ4uOV3M+Pas4vziguXpkvOTGWXhKHxSFp0DdeC0QAvQEhaLBagQ1ZgV6YjLzIxSPdEiag2+K4hqnHdGHYSbANatKfFJEZUiOTXKV0pEcxDT4hvLZyZzWUVjsNNFEuUwyQzjQmRrCL0jc78se3K/RdN7rCXL8QjZDskclt2opWhPH93Lz315OxQnsLU2D/FtROUVy5TdvpVF+nHfhQZDiKK0voDYIbFhEWpae2El5BNgf/wR5EUou/2705GS5CMhXf+z3extmoY61tzYLXVX01jNNe9oY/KBsLEbJArOwcXFBW/evOHi/JQ//v1zLi8vOD894Y/fzJjP58yqkqUyWGJnpU+WU+dYFJlNpIrkBi7kNmIMyiikNdW7t2/58e1b2ghV6SiKAu890sn2mpdXNIwP1n346Hp87I0Sh7wl9/KwBLPD6enn2qM3d3ul9oUwWUgnTJhwBO4u3x+jDBz/FRj5xKWn524dErW0NHS0SSGlJtDS0YB2iEZcEfGF4L0wrzzLuWM2c1yczTg7mbFYVLy5WLKcF8zLguUsLYYL1XT5dm/xHMlba2g+5zJMpDndZnQx0jnBiWHZk2DoDDVFxafrX3DpioPRcn8n868Nj9pUeIbPHKK37X9/FGesvD7YHvMzB5Xd11juH4BYOrEpvQvYkPMtML7rYDe/1v+bPKbmYO2v9BCSR1ny4pi+nSe1VfvFluhAQQ/WELKDGmVFIFJbJNg7Gmu4tY7ANVECQQSVNjkCU0fhBCdK5R2LsmReFnx9MuPy8oKz0wW/uFgwf3PJ7PSUy6VHvUedozJAwRy4oMlhgo94X5E2sALOlYBRxIBqhRHQGIiWNpOKGGhDpA6BGFucarbuuS0lcLhKYthi2xYxd+34LytvNwm2Jzx3R+UPvWfblWUkc4+8pWnfLYHj070fv/9INhrvKNv7W1SD9yhI18RIarcSO5TiwBe+MOwPQyOvxE9jGY0UomEzwBAJdDGxiEySR9zOAk1oWLdrNt2G63BFGxsCDcHdIEq6s9dB5QtmVcHPvlny05/+hK+/uuTv/vG3XJyfc3qy4KdLpSgKCu9YqqT5L4R07juj2J4e3vH4LKPGYZCvdel49/6Gt+9vCIDXktKlTd2+nD6yaI4vV3nECmekjCbp7n3tL42db9peXzyUlHHHH4LGI8ADz7wAJoV0woQJH4TdEdieVzqAQ3qM3LNS2XX0H/JCt1dA81UrtINCGrTB6Iiuw7mA18jcG0Wp+MIxq5T5zFEWynJWcLqYMZ+VvDk/4fxkwXJR8eb8hMWsZFY6ZkW/OE7UIdlNHGCjBdNIzh420/2lkXXTIRYIXUvbtNR1S9tGZuIppKLSWbr2xdywuz8oYY/Fh5S2Lw6jFnNMXg7ESe3xQCOzvZ9fOFQjsQrcZsqu1wiuRWJ/FyDDssRsu0ue5BTeO5hJaygdSs2Jy4thQaVACelye1ngaXHWcSNzNN5AvOG7UBNsA3HFSm5Tv7UOkXU6QwpU2jJzwrzwvKlmzEvP4qTivHTMi4KTkyUnJwXLZcXPf/I1/nJBuZzzs7NzGu+I3vFtpWzEaIBLUWqLbCxyiqOJHevYciIFnbWsY8fcFoQYWceWmdPk3zMYRduh1rCxW+r6iqa+ZW3Qdg7TgqryePE4PC4kr6RopA2WPX8m63DysiuDpW97Mi973x1kGRwS7Ybvxtl/Nuvcg+K7K+dF+yBblhMeluWDcfp2cVdOS9RdJ0nZy/i9+TwQbskJXJ/PZJd3gBBRPMla2hUVwTqwjlnXERA6ZqyLGaabF7XUPCcGuu7RLKrRYZlHs1SU/nytSTrGAtBpTSsNKDSsWfGe1WrFprvlNrzjdpPOcMOaWaUsK8cfnJVcXFxwvpjzx5c/4ac//SlvLs74g598xcnJknlVcT5Lx1dUBOs2ENdII/jxvaSSrnfa8TJsNjpiMPLmDaxF2bQt79694//5q+/4y7/+kU5g7s4502+YdxeUocSHVFZx1H+OLdljFcV7HeEdUNpe8gjJYA3+yGcfHWd/Cv1EDMBPrpD2GXush9bd+Fsqq4zouOPxKeo2zq3c7rz3u/VfD/JVs/WOW12sBvnf+ZNvB/m/+M/+4SD/B3/6i0H+euQF10kYZAlbWq8b0YOl2yYwjrxTjeva7W5VDPjzv/jXg/ybH+pB/sU3Px1k323f2T1x9fUUivSnpK8+llo74XnxMG1yu/12Z5dOtr9s7aGR1jqCJa99rdVES5ShKA1Ih7kO5wPqYOaF+cJTVSWnpyUnJxWzynOyKFjMCiqvLArPvPKUhWM5r5hVBVXhWc4KyiI5MipdOiulYsPVBiCjfjm2wDF4gtxmwwhdYFM3rFYbbm7XrNYNdRPoOgHnUVcmC6n55JyB7e7lo3vLsLM83gn/mBfdh0dc4zPhk6On7C6sw802SNTkk6dXWQQskjdMxpa99Hdh72hI3hxJ9DehdxiSlAmHGURWrAc3YTe0tqGxFbW9p7WOlpaOdfJOK6DZW3WpwsnshEXpuTwp+ebbrzk9mXN5UnK6OKUqC5YVzGYzyrLg5KTCzwpc6XGuYi4RiZGudTg6ZhJorMToKKWlsRlGpKSjjWkzq8gbWmZGIUYMbfL2aZHNTcvt2zU3v9pwtQrctsnSWwpUKmAVWAni03U2plgEzWpVUvZTZ4vjRTXbMaFX3rbhvRJo98aBreLXy7tWy7FVdBx/qxAO9Np75F2leRx+11Byn8UzsvUybCN56ynYdvJ2Nzx9JeaXbv/St7ntnOC7RNkN4mmcYLYBa5l1Gzr3JVB2d0tVBov7h545Ym08aks74WbDgjIpg0boOjZhwyqu6NqOtd1ww1s23YbOahq7pfAFFydLfvKzBT/9ybe8OV/yh1+fcHFxwel8xs9nBVVVURae87mn8OkIi0jEYiQARX8kzQyLiZavkq6PSqUgQ/oMBor7ztE8QFRomoZ/+6tf8cMP33Nzc4NzSlkUFMWntYzvlvNDyufdev3YqfPD/hFs9P/dNH7p0/VkIZ0wYcIOjnYYk+fOu1Po7rC4a7XqH8re/ID+htDklbOhs4bWGlpbE60j0mDSIi6gvsNVQlEo88pzfuZZLmZ89WbJ5fmSxazk7GTGsioovTL3OjgqKgtHkam3pU9Xt2j23Kn5frbtppbtLOD6pMsoJ9szOOlMeF03rDc1q01SRps2EgOIOtQ8Tnym66aJ+7mOaN5HDfrQ0uaDu8TPOrvtfu1vq0OiZ0MUnAVweSM0bNvqIQUjtefthUn99R4p8uCzEpF8/Ub+x4zkNMw2NEQaOrD3bGhYW0Ow97REagUnDU5BNDkIK5wy88r56Qlni4rfu5jx1R/8jLOzBd+cFMyWl5SF58R3qK/SFUtecJ50xkwdjoASiF2R7hqlpYsFqgGVjmCGSMRpTAtfS+5xYtLIE30+RMzS3cJt3bC5rbl9V7OuA5uQrsopVCnFoZSo+UEpT68cWaws0fgP6xV3d4XG1s3dGrE7ccbPb5XNuwrv/vh8nApzP/rxo3/vrjK7G+eh57f5u2fgsP7vKQ+aKTb9Rsr+F5JHXQcixHxWUIg4awnx9VN2B99C48p5aOMwzwfH7KXvXBk2alLJ53za3g3W0MaGpq1ZdytW3TVd11FzS61XRA2oBpYqXJ6c8ubNKX/6J1/ziz/8BV9fnPGLr085OztjURZ85RMFF4sUdMM8KDEODgfT2JHndItZuVTG1Ou0ccao0R5QpQyatuW7777n+vqWum7xvqRwJYUWiPVHXp5pDhm/Zr9uRnP0ffXy7Gd+jR1HWPtpvDP2GE+keL8eTArphAkT7uBRSikHBtDR/t3+LJx21CMhJv+4rbXJMRHpTGi6mqVFihbvIr4wyplQFJ5yVrCYe2alYzkreXO+5HQ546s3p7w5X7KYVZwtZsyLgsKlewodZJfy24VPT7dF0kQjeZYcW3J3B/ndstjha0SjaTve36x4e3XL2/e3rDaBthWIHqXEJ7cp+Qxpcmi0u895D+7ZiH2qQjc+Z/Lg954Ftl2E9CF32stzfemFsvCKoBqxKnDTVsxXS9I9nQ25WRMHK0RanPbl0TuaEdK9of291SoOyf09SkVBi2KspYRwhcUbfhvXNPGWNt6y4e2wgTR3Dc4rZ97zzaJiXpXMFzMuFyWLWcXpyZLzrxecnS346TdvkK8WzGYlX89m1C55nb4wYa0dEWPRFjSuJtBQNTOCdkRp8aEkZgqi752b0VGIESTS0lFKOkrbSqQkUUUbIoVzWIg0XU3c3FKvr/n1+j239YbQRSotKTilsDNOKHP/NGJIFlY0UneWvewawXqPu1vnPImO2ityW2Wy946bwvfpqzYKB7CtN91RHKP3sruV+0MOvbIb87NJGUiMj14enA2Z4QbHS7bjkKnPQxiFJ1lGynFPU+6v6CBvCPQed7dyzIdEd2TJMiOPw5ZKILnZSo6z+u+3viJai1jDrAuJsisVt8UZpvUXQ9l9GvrWQOq11jsl2/ohiM4wNbrQsXFrNloTQ+Cq+4H3zfes12uC3WByxdnFGScz5edL4/LyDW+Wc37/bMmf/umf8uZsyU/OZpydn1MVngojxhUEoxgqS8Bl8rZAJ1uv+s3IUafmneoYDdORs00dLo3aubs3UzrAjE2M/Ph+xb/6y3/D1bVhdsbJ/IKl/4o557jO4YIM591N+p7xYRyipo6dSR3P7rN7NqVeBsOmz4Fv7rAB5UAePwZ2xIb1C+ALVUjHBbXtBBttBvn/vfmrQf6bzV/sPP3LP5oP8n//j/+TQf4H//6fDPLliI5b2Pa93kadLm7puHGUDhs1kCDbIvYjGnGMI/P/iL4bR4Ps6FP8n/9im4fzxe9v0xa2Xnx15xb44xrTfdTcYyi791OqH/etY/DYb70Ufre96T43dpVSG/3TWUcbk2OiTVzTWp0oulqjLuJ9ZLl0LBeJxndxPmM+9ywWntNFSVU45oXjZFExLz2nyxnLeUlVeGaFUrpk+fRkD5lsFdL0y669zmSk5OUF1FaJGver7Q5/n7VoRtMFrlc17282XN02tC2YeZyWOAoUl3Z2LSmjD22Uj5KVf4y+ub/bntNw53157fCo61FkS6mDvNB4RHN/LVexfKnec49Fouw6ZtbhqhqJkjxWAmlTJfYEU8RiHi/TYkZEctMXyO0wKT79OdKGhkCiz//Ixm6o7ZZVfEttNQ0NLZt01YeCm804mxW8WVT87Kffcnoy4/ys4vLknFnlmc+Vk8WCalayPK0oinTHIFGpLJ1PDXhcF1AiEUG7DiFgNIh1KAFoSZfSBMw6IKASCKEFDXg1YkxxvNhgtfFiYAFiQDvj6rrl7fua66sNbYigQulK5q5ioTOcFmAOQ3GSqIgxbm+9TNeUZNkYhY8YE4z77L4Fchy/l7f9Zp+yK3dk2fnCQZqu7MrcCb9L/e3TpKNvHaIj9953+y/fJ8uOfDfPSRlJlrO0OdUzVCKIIgYuNIAQxdE6waxGrKUKNeF3wMvuc8DMiCHQxUDTNNzailtuWId1Ohsar3DmmM9nLM9KfvnLX/LNmyU/+3rOxcUFZ2XBt7OCy8vLNJ+6jsL1mwLZyZKkLYP+HGj/c7B29shXtYgky6lkj7nRti3tiFUit6tb3r2/4je/+ZG2UVQqqqrCqUdFsTgmmz8OD3kx/tB0tzu/pXw+dGTs2S2nfwvwhSqkEyZMeM3YtYf2e+xku0r2lWs1TazZxFsa29BajWhD6UALYXEy5+Ks4KvLBT/55ozlouLspOJ0UVJ6R+U0OSTympRQ7/CqiZYr/ZlQy2c+7a5CN6RQRou+tEja6q4yTH67+yDbBVc0ow2R23XDzarmdt3SBQHzeC3/f/beLFazLMvv+q29z/BNdx5ijozMrKqsquyq7i5jubEBm7YQMjayAGFhm8HCknkAISQksHjh1U9ISDy1BAIkhEACCV4REsL4gQdPbblLNt3V7u6qyqzOjOHe+01n2HvxsPcZvu9+X8SNiBuREZl3heLedc93vnP2Gfaw1vqv/8JIEmNSZmVKvoq9t2p/Cv1F7aU91zHAL2OUyroSW3dVo/RVDODrlk3h0a/jgkAFvJCKg6RGnEFcU0tWkV7Jl86kiAaNNsvCzrBq/lZVlJIST601nnNmfsFcFxT6hEIdhTgMNUkSYO/j0ZCjyYC7+yPuPbjN3u6Qw/2cvZ0jstSQZTUDO8QkBptBZgK0T72QRDPYicH4GvB4K4gGgxQ6XWm2hVGkuTbnPSKhLmjQG6KxcG2BPEeb0DGzRcX5rOJiXlHHRMbEpmRJTmpyjFg8JhjrAqrS5kQq0QiV5l5tMuZW+2jfkOznU3bf7fZfN2f92vEvH/PyPrDaZTfrTe5nf/tlM9esfFd6+2xvU/MeNeUvtH/Numn//rsqK45C4x0iCWBw0ry7jsSXKBnfPGmcqc3zCtD00peUWrEsF0zdORf+nJmbUpkpzizYGe5zcJhzejrg00++xZ3TPR7c2mWyM2FkDbuiJEmCVUdSz+MLER1araO2c9aGPNDGQWradwSJDi+R0A+bEi6t0bYK7O5dUfc2CswXS84uZnz5ZEZVT7AmIcsG2EgG6L3ndcoavjL7rXI1ZE/PkfzKztF4f79pXCgvNEhF5L8B/hzwh6r6S3HbIfA/AY+AfwL8BVV9+uaaeSM3ciPbREQeAP89cIswxv+Gqv6Xb6Sfrhkol4dL7f0O0FzVELFxWuOocVpTUlBSUJsSnyxIbE2a1AxHhskwYWeccvtkl6O9CSeHu9w+3mMyzEMkdJC2JVoCK25TVj169rVb7Mhae5o/GybR3lQYL60fdpQuQirdIkrb44VJo1ZlWdU8nS748nzBk/MlZWVAE1IzDFBdTSKMqbtjunbE7Tf8Ktu2yPMMxes21tp7eznH7a3JO2qAXmcfNdbBQJnWA4bzCUYc3hRo3UUGGxi6lQCFFCARAxLBl5KTaqjBvZQh1l2gFJypp3QX1Dpnro9ZaljswgyxwsgYTkYZB+MBhzsjPrp7n5PjA27dOmJ8L5RS2hvklHmo6zn2sEgcojCoUyo7B/WkLkOpUBxWDWoLoMaogUhgFpYnDsTFiEPIv5Y0a3ozaTrAS4UaJWEQUgFEsGLxCLVYRAY4A3NxTIsFT4olv+vgvBbUWw6SDG92qcwuzoZFvnrwFYh4EqNUsW5jC8eNtVobI6uB3Qadnt6H6W5iotUVaG4wdjtIbZ+VN1xzZ0z63vYVCO4aTHeT7tCWQLFj0NWV47wIsrsVpit9yK6sbe+uJ/BuhbvaVCJVDEl4gynTIc5XQMWgdsElIQPm6d4bgey+1Xn06q0CImMtPjicEIq0pDI1RVHwZf0FZ+4Z58U5S31GyZRxjxoMagAAIABJREFUptzbHfDwcMSv/uq3+PjOMd+5d8LJyQm5FTJ1GLMMc3Md+p2IILbJFRUa8yBA3+PsKmELjcOhsVMBVRsZqIWE3iylposkSo/1u5uoW0ZeUH52UfKTJxW/8xSc3WGQ7TPOD7GLFOpmnu/m9rdqsjWG+Ns63TcMkXeVCOl/C/xXhI7ayF8H/k9V/Rsi8tfj3//py5y4f6P7A4v0tls62Kz33XZnOkbbX5Sft/rn899u9eP73f7/yb/0J1fO/S/8kU9b/TDvXmfrO3Zc6eu9dtCdeoUp1xjT26V3bT3YbT+2YXqeD79y/d0Jlr0v/63f/qzVD9Lvd23utae+xp75Jph1r6tzfdM66RWkBv5jVf07IrID/G0R+T+Av8Kr9NOeEbPxw17wrP2pnWHXeLmbXNG6MUJ9QeVj3VAp8KYCWzMcKoNRynCUc7g/YG+cszcZcOtwwu54wN5kyMHOgEGWMswMuRWsgJUul6lhvm19udGQVIiF2XswtV4/6V/hKo2RrPxavTX96wuLstJ5pkXFRVExLRzOZYiG2qOGxhjdfN6Xlc1RSN1+0BWjdMtu2/Z5ZQfvtkjuN1aurY96NVALA60xeYF4jcmMTXR0NSIhItFJo+G70cQoVQJrrj9nqTMqLVj4KcuIVihY4sSB8VibM8pDGaWHt0853h9xvD/mg1sP2Nkds7M/ZDzISBMLYsidx0g4n61d7FkO45t8uAokkKQIAYKLeCQaoIiLuM7Yy5yG2qkSHFxIqMGq6ttojFqP4LFxQBBCXqUzHu89fqE8npV8OVtQFMsY7UxRSckMZOJx5CHvrTXepMes23STrof04fTbCIjo6e1ifGWfnq7NGS5HWqGLZjbnXo84AStIjq06q+y73fV08rKQXdmoa3ve1bG2sWkaB0oPshtNbuMCy26A7IKqA63IXUkdUx+uWa53Ht0kVxhPV8b3laoU3ZMoq5KFX7JYLLjwF1z4C5a6JBsY9sYTvvPxIz462eOjkz0+/vhjTncGHE4yBlmCVY9xDUw6OF77JViMMWE6cc/1PPdeKG2/20T0mt/t+7wh0qe962uIrVSVZ2fnPDu/oAKsTUiTdLUB/SGud/53Rp67frqR58kLDVJV/b9F5NHa5j8P/Kmo/3fA/8WrdtAbuZEbeS1R1c+Az6J+ISI/Bu7xOv20vwpZ2RhlxShd/VxVW7Y/R02lJbWvKPyC0i+oKVFbYBOPzZTxTsr+Xsbe3oC7J7vsT4Yc7Iw43hsxylJGecIoT0gjW24i2rLiNnnTwRCV3mTVWwiqtJ7lNui51uxLJl6MMOnKcq277vClUPrBqad0jllZM11WzIsa5wagCYmkkcbI9E+3fjdfTV7GwG0nyed84Sr7XLVd6/f5Gz43X2sfVUG8kFDjsxIqCQFF+kZp1zdD/Uzie94wPDtKVWp1qJ4xiwZo5R+zoKaSmMNpILHCMB9xNE442sn54OFdTo7GnB6OuXN4j2xoSYbCGAMm2MZpXYSanZJgfcgVVSMYbeC2FWgDwXVIJG+RGAVqFqeNYRnslGDcqNeGfBUvwZgxArU0tDiNkzdE+GrxePW4hfJ4VvBkvqQsSiAJ90VSUglwYq8WH0vGiCg+5sklsaWR9zUMM7oOa71sbHa1P3UlEtkZtuvG7KY8075xuh0SvGqovkjv2t2PxK6y7Pav7QWQXW3sk5eD7Aqg0rwT0joz0RAg8JKiGJxRxCmCI/UFbwKy+0bmUVh1oPcnmW2btTNmmp8igbnW41GvLOo5s3oWaonKlNosSHM43B9yejTgRz/8Jb51+4iPT/c5PDpiLI4RJdaY4MDqtcsY6RwIqphIQtSLpVyS1sCUOL9u+LyLeq5+p82tXHshlBBZPzs/5+wiGKRDm5DYlH66S/uV1tGx1c36UnKdU9RXFzR5vyfaV80hvRU7L8DnBIjDjdzIjXzFEp1Hvwr8v1yxn4rIXwP+GsDITLoPeiuOdU/8Jn+nxgnTqaOmxKmj1CWlhqioswUmrxhYz2BkGI1ThqOE06MRR/tjDvZG3D3eC1HR8ZDdUU5qhLQHze3XUmSlLcIKgisy8qg0cYQNEcWesSkrP1ePvKrGBbLG/J3asShLpsuCp7MFT2dLzuY1KYZEUxLJA1xXzeq0ueoAf45s9AysLGKvLFeZq65rPnvJ43yTvMmv20fHyRCXOaZ1xuBiiBGHMQXqQykNK7Y1WkQMJpbP8CYj0VBNdEFCUT+l8FPO9DFLv6SmBGZgDdYaTodJ6IfjIQ+PTzk5PuDoeJeDjw6YTIZMRgMkF1I15F6o7RJRS1ZneCmj8WXALIKuFpEiOrIsITLqMSaJrbUYtdHYSICMJv886AbFY8hRNXgPiU/xAo4K4w2KUmMjgY/gfFjAF2XJl4un/OLJM/7w6YxnlWI1ZWwGHCYjckZYHZE5g5dgPNa1x0rIUW3YdC0BARXurbRMoaGVHVvtdpjuBj0ep9FbY3dFb9bw2padWTVMO70Pu10xCLfs02fWfVnIrm8gu/K6kN3GVJc1yG4NVAyqCNklZ5a8eZbda51H2x1AjbbP0aLtBJCQtBHfKiLjRIQ6cXhxqFc+959zVp9xcXFBqV/gmZFkCfcOUu4cjvjBD3/It+4c8sHpHkdHRwwsDKwi1ROMSGRYDkZ/OFVj/EcnAoLRAIsFsD3PZIMGFAFjbbvd9ZxfpjmmagQSSoy+di5Y3/llSZy2L0SRJFTOsVgW/NZPn/Djz55xDuzpPjvumGGxR1anpC60upZuFSDoVTk8L4s2PeP9nX++TojB1yY1UlUV2f46vLCT3siN3Mi1iIhMgP8F+I9U9XyN1XhrP1XV3wB+A+AoPXnO6LbqiVx19oZqok4dlZYUfkGtFaUuqChQU5OmntE4YTRKOTzI2dsZsDvJODkcsTvO2RnnHO0OGeUZo9ySR6ZcK90iaf287bb1VkcPfUtSFFvZfbwJ7tZ5pWXjQZsvhBWh90pVVSyLknmsPzovKorKYY0gYtvao6ZdcPXv2aVbekk6M3nTbtu2vx/yTYT1XkcfPRwcKk7ItYosu2FhF4wkosFHhOpqy6ALNZVWOK1YaslMz1gyY65TKiq8OKwxTPKEySDl3skxJ4djDvfGPDi5y87+kMlezu5wHNlyBeMdVj14wUhYLYrUiPhojjqIBEVGHE3+WQMtFgAfmTmbPqvtNSPaLBab/T0qHsRjWsiuxxrFGx+NxmjoxahLVTqqi4r553POpgXTssT7ikyaHNsM0YxAsyQ4L3htojmCapO12UU2m9JRzejRd2+tRky3QXl7EVJtooM9x98lvYsOdQvQ/uDR6Zc5dC9v3wbTXdU3tbuBZcbrXNPlkh5exvVhbvXONeQ52u4UILs1Cnix1IYA2cWR+nlwCrwhlt03No82gcGegdb/LJ67dWJ476mqklorirLkzJ9x7s9Z+iVpBqPhiAcPHvCdu3s8PNnj29/+Frd3Mo5HCXmekYonwSEm0hI1RF3dU1q/8vicup9EbWXmamG2zbdeLP3nr/EYTdpaw8JbVhVPnz3jy8dPODu/ACAxCYlt6nf3DiAvd/6+bLz293US/ZrJqxqkvxCRO6r6mYjcAf5w245XW+x2r5S5NGwFOXNdDvlPnv1Wq7vhk1b/d//SP9Pq/+qf/FGr7yZdeRYAY7q/jevO3avEsrJUMr0mrRY57pVr6QEc+gNYvyP3IRD97f1xNe3lq/7m7/9eq/+jn33Z6n/05LDVk17iaGV6571iB7tKSZf3Ud739r+siEhKmET/B1X9X+PmK/fTK51j7XeQcJ8DdLWm1prSL1m6eWDR1Xko5SKeQZ6wu5dxsDfk/u09jvfH7O8MOdkfMcwsg9QwHiSkiSG1QipdxGHdaNw2DV3e2l8wvXDn9hTb3p4mGqIK6j1FGQzS2XLJrChZFBVF6RhkAqYxSC0tnDguuq/kkV2beNttPGf7+yLPu7avqVxXH1UVpBZyKSFfopWFYuVEcccwD7jmffMFC62odEHlnzLVKUtdUOgsQFyNkGcZe+OU48mQh/fucef2hOOjCXeP75GOlGSgjHUQIj2qpFUd88cNoi7kaxsw0sS8ghHa6CK+WZWH6VNBnUFs1+ZV6lhFIvIi1JkIxqgYD9EgbYxTH8uqJRpJgiIbqC8c1UXJ7OdzLmYls7JCfUWSQCIWTwqkGFKUEPXxnkAWpaHH2xgXDaRCsf/3m3lJb8ytdXjtlu3SRFmbMUbaqGQzXsSho81pld4xL5ma8aDrxEjrZWWab3Qw3e21UrdDdtdgupf0OAbr5e+KsALZlRXIbtVCdisLxiviKzJXUJLzJuSNz6MrY1z3FPogbokh8MpVLHVJUS+ZzqdccMacGRhlPB5wdDjkl3/wS3z6wTEf3j7g9PQWQ12Q+2UbgTVImxNK60zYsD7qeUDC41pzPvc5T/rEKFe0Bjeav8YER1qEyJdFwZdffsmXT55wfn6BAay1WJvE87++A+Kbti583+RVDdL/Hfh3gL8Rf/9v19aiG7mRG3kpkTBa/9fAj1X1v+h99Pr9tOe57ZYRITrgiblZ+ODF9UscFaUs8ekCYypGmScbpOS54eRwyN2TXY72xzy8dcDx3pi98ZD9SU4iGuoaxpwwkR48t89M26zKVhq4tszpN3VtKmz/kudMb/3DbdgmBBhaVTnOZ3OeXcx4ej5juihZlJ7aB4ZdQ4KVFKNmBbL7UlPitp2/DvPq1+EarijX2UeN8fihclGPGc72sVJjkwKnHvUGT4rVJYihTlK0usBpyZKSqTun9DMqfUxJDeIYp45xnjEZDPjo5JQHd25xcrzP3Y9P2NsfMx4PMCND5j2pKpVdkvgE4y1qCgSLIUNlibTcrRGy29Q91fhbIjTRC8Q8QKMG1Sz0rdrgTQaiSG1Rm6KSYJzFG0UNGC94J3hRrHq8KLUhRtSEWgSjYZ+igvrpnOmTKb9/PmM6W1AVFahFfE4qQ/ZkSIoEQkEnWGL+qIu56jGXNLDJBshqszxuSrlsheNu04UVmK4ndIcVBt3IuKv0WXallw+8vs9luK/TjkAx6OGcW9l3t+nREF+B6fZ12QLZvaTTuxfEAd3QUexIm69bJkOcRshu7XAaWHZn6S5qyjfBsvvm5tEoVm2bphF4oIPURmkYMy/MGQu34KKY8qX+PnN/TkHBkILDoeWT737Er33nPt++d8KHH37IQW6YpEpaPwatYyQZVIKTxjtt61+3ng96REMGVub28HHfHEWMxPqfuhJAedkgdRv0CQngiAjWGpyBx7Mpf//HP+YX50umPiHLRojkiKaopkicR7tWta14uUbcyDsrVyn78j8SErqPReSnwH9O6Jj/s4j8VeD3gL/wJht5IzdyI8+VPwH8W8A/EJG/F7f9Z1xnP21DjCHsEsxRR6kVtQbioqXO8VRoUpLkjjQzjCdDdnczJqOM06Mxp/tjDnYGnO6N2B1njCM018RcE7PmCL285tgyA7aLtBgnWEE7cDka9xqiBNbvyjnmi4LpfMl0XlDVgQHVRKiulQSLwcQSEbQRgvc5Y+VGXlGurY9qJDXKcJi8RDyIk15tSR+ioupQX7PUBZUuKXTKQmfUFHjjsQbyJGFnMOJkb8jhzohv3XvI6e199vYnHO3skA9S0sREaG6A0BrvwgIXQFyM0DWERE0mYKTziXBcAIl1SkIEbzXvS9qIeZcZre3COSyCmyCJGhCj2JDQGQwiA/iARDBNDUQHUnvmi5rprORitqD04d6IWDKbkdkMJEPFtnHG5l+bh7nWb4OR3Y9HruI21vU+lHcVpks7LoU91ktQrevrY5+0EZ/VeFYbEFtpXT/nvH+kq7Hy9o77UpBdVnyE/fZ139S4j/btIYzWoQSOGJwoSgPZXVLLG4Hsvvl5tPfGSP+JqMepp65rpjpj4efM6ikFSzTxHOzt8+BwyK3DCZ9+/xM+ubPP/YPg0B2KIxUXK0P4mFES49ntTdfup8rKs13dY5No90Jdaf/Lx11//kBAN0TDuChLLi6mfPb5FxRlBWLI8gFplpEkaRM7vuIZb+R9lauw7P7FLR/96etqhHG98i62g9P+3vx3W/0L/Umr//qvP2r1f/vP/eut/vF+2uopy94ZVruC74Xt+wxh0oPg9mdL7YFtt42BL7vA3BR8AXA9NMTf/NsdNHkwPG31se1ycaXu+7Fu5Jsoqvr/sP3xX0M/bSaDmF2iwdNfq6PwRZszumQGtiZJHJNJymSccno85tbxHgc7Q24f7bA/ypjkCXujlDw1ZImQio/wLW0uaIsB2Vu99VoGYWHTTxXV9R1eQ3Rd0WiQVo7pfMn5bMHFvKByiqrFmBQraYDrikHUtIuPG0P0mynX2UdVDVIZBqaE0RKpLMwtoj72z1jSRR3qZlzolNLPKf0TCkpUPMYK4yxhnKfc2j/i0d0Jtw53+fjRx4wPUwbjhD0zQhOP4kjLUAbNC1jvGoshQHM1/peOU7YFJzhBLCEXVJtpVXv/Y9+NnVZjRBK0NTyRQIYiNhiezismAZMoNcEmsgK1TULtVXU4YxBxGOeYzkvOpkvOL2YUTnEiJMYySHKyZICS4sUCBiNVZAcF20QuFYx0mawhWrmJ7XaD4dnus2F/Wf2utPuHO9garWuGap9Iqdu+Zpz2DLtWl77e21+2QXa311C9dC9eCNntGeHtNfNcyG7iGshuQmUF48sWsssbgOy++Xk0vhfNc+jNdw5P5SrmizlTd8Hcz5jrHE0cwzzn0YMH/Oi793h074RPPvmE22nNjvUYI1hXoN6h3mNMfJ4+xtylexZtJqjQY7xduwdb2v0moK7ex+qz1rJYLHh2dsHPP/uSskwwZsBgMCBPc1KbrjD09h0rymsQGt3IOyevTWp0IzdyI193UYgeV4fDUeK0otSCuc5wUuHSgiyvSHPY3c24dTzhYHfIvdM97hztsb8z4mR3xDA15JYYFW2IWJqFymrkUNqobPN7E6+sXtauYoSuGbbPOWy7qGt296rU3rOsap5czHl8Nufp+ZK6sggJuR2SmiyWfLE9wvpNM+c3YDZdv8/fgEt+k2KMi5DdEfnZHtYoNqnBVeANajJc/SWlnzJ3Tzl3F1S6oGaGtYZRlnBvb8KjW8cc7+/y8P49Tm8fMd4fMzhMGFlIjeLNAussxoHaBSIpIhZYBp0UoY6ERAZRC96EfmsymvI0SBp0J6BhySG1ARPgi+JBg9Ua4LiNrkFXkUA+5AMDrkXwDqoSEufwBmqTYOpgxNbWQJlQFzWzqmL5xZecP37GzwqldpBpwiDJGZhDcjkgtyawgPsavI8QVMX5UOPUSAfNDQZhZ0I2r/LVYbqNLhtyO6Xbro3R2IswaVNGJkhf77Pvhu26so+yHaa7oveM5k2Q3f4+bxqyWyRDvNYoFYO6DpBd3hxk962ID4aYdw4/LEOdblfzmX7O1E85L85x+RckY2V3b49f+/ghH94+4tNPP+XB/oi9QUaazkl9SVI7kiQJsFyjeO9wSLinxqycVjComJhX3cVmu8zg+I69hAO37+x40X7Nb9MLuCg2xLyd8ouLis/Oa76Yg/UTxsmEvfyEkR8xcAlaE2vUKj0GtF6E+ertvpF3V24M0hu5kRt5gYRJK+SL1hRaUGsRSrowB1NjcsfObsJ4nHB8OOTu8S4Hu0PuHO1ytDtiZ5gzziyZhSTmirYwLbbNJ6vG6Oo2Xfv7lS7rytKf+BTFeaWsHbNlxWxZMV86vDMYTUhNhiXAdVto43W0932UTZd7Eyp+LVEEvJDgMFnI1VSvkT6owvuCuU4p9YKlTilZ4KXCGGEyMOyPc+6eHHHv3jHHh/vcuXXM5HBCPkpJjSdBMaqor5G4/iOWP+miWE3NT42BmLjIVQlEuhJ1Jf5e779Bl5XNXQSniyLR8a30hgExgjFCW2O4jfqEXFUVH8asCp4sK54UJfO6DqRBYknMkCxJyYxFsXgNBEkdp2+HuLjs8pK1q1jt3drus1YBtOdg67/+fQLFle0t0mO1LIX2duhzz3RR0dUxdaVtvchp2/6VPtq7Alndui7a7vRqkN1O03afpoECgeyGcH+8GBQH6kg05Etvck++D9LkVjv11K5kWSw51zNm9YyCJflA2Nkf8uFHD/jed+7y6PSAOycH7GfC0EoouRPh842Rti0JpLtHm+bQTX+97LW8xv4xZO+859nZBc/OZswLEElIk4wszTFlgP03aS/hnWy8NOtrga+5fAMu950wSAumrf6Tix+3ejn+rNX/g3/jn2/1f+WP/bDVJ9St7n1HNehWvGfrHbA/0vZuwcqD9ps3X0Ve8gtJz2u0dN31/M1/+PNWvzv6sNXzuoMmI6532k0J319vWaNk/wpb8nWVaIwSjNFKC5ZuFhh0mVPbOTbxDEaW4+MxB/sj7t/a4/7RPvuTASd7Y8ZZQp5YBom0pEXNQqRbOvUnzHZ9Gheyq591k9Hbe97tPBjhfM6HGqQXi5LzecV0UeN9jiXDmgGppFgN9RB7K8i31t53SdYhVjfyeqLeYGohlwLZmaFlip9blgrOL6ncGRd6RqUzar2gNjVGIEszjndTbu3v8OGD+3zwrWMOD3e4tXOK5h6xnkGtqITebqgRo6gx0RiNEDuJRigxguYBZzA2GlFeIoOuBLbaOC157ZHqCC1JvffBuEQiH2+k1fZIoLgXAqmQJbDpKhgb/jsJZC0WcDY4gKyDKnUoDr+Eny0rPisqllURiH1MgrVjBtEgresE1ZrA4BsiekpYHDVkQ7aJJcVrUF2FDyo9UqE4E6/Cetcgu6ptitDqcbQ13tu+ot3I0RiVsklv9+lDcJvjrAGKG4OfhgAptGIVjrsdvrtyL2I7gv5ykN3gw2ggu/20CyV1JUggtSp7kN28XlLawXs6nMZosTGUrqYoC6bTKc/0KQUFViw7OxPu3D7mV375h/zw23e4szfGWktSL0AdxsTcbA3vqle/JVp8eRUocnnbVyXGCN4pzrnArvv4GfMF2CwlTTPyLEcqifsajNGG96lzGn1jVrqsjhFfU3knDNIbuZEbeYckWl+hdIIPtUV9gdOSioLKzsFWZGnNzjhhOLIcHgx4dO+Ao70x90/2ubW3w84wZ2+YkUpYrNg1I1RXT7gi3RJqg0Haekjf1A24LBKNUVSpnaesa+ZFxbPZkmfTgvNZjboxVnNSMyQhREmNdsC+rhj5Cxr+vI+vay7ado5tfryXPa+s/Optjsvit+Xt3XYNr3NtX7EY65GRYVbvkZ6NseIxtqR0T1nU5yyqxyz8HCixtuAkz9gfDXl4dMCHH9zl6HiPux8ccnSyx2CYY7OKpBKMVzQrMJrE51SApAgWkSrWSARRAw1rp6aIT0AFdSmBvdagdYDpGm/RGtDIplvFnq0GjcmQRg2RGBSDwbswNhgRvDOoKBaDrwSHkAhhIWsUUzlUhMpabC2ogDOKm2UsZwVPywtmT54wOztnWoP3GbnJOZCcxE8QHZFpjcPh8ahXkhCew0Vjrw+7FZE2hzMYaf19opHe23+l7Er8LtIQUMXFtpho4AaypH5pFt9/N9thL2z0GnJn9ZKunU6T1dsx90I0ZBvnQNyn8S30WYObbrJJF+nBd2UNmttc//r2xghu7stWyK60kN0+y25NzjLdwb+PkF2Bwl5Qu5KiKPjM/gEzP+VCLxiNnnFrb8ijR4/49R98wIPTQ+7ePeTWEAZ+gTpPQoiOinNUgJeGKdey6rKgN769W/fI+q5plVEKrZkXC/7xZ8/47S8veAIcyRETPWJU7pNXAxKXILVAxIC0PUsl5Jy/4JwrQYu+d+d9Eemi6o28l9dxBbkxSG/kRm5kTTQuZjTERbWi8IuQOyoFklYkmScfJxwcZuxMUk4OR9w92mV/Z8jhZMDOIGWUGjIbohfN8qvx6De+TV2fSNfUzvWv2/d5w9Jf96gqznkq5yjqmkVRsSgrisqRECC7CWlg192QkPPCOWTlEjdEFq8jMPzce7d+8IZNFa48B66siTadrHcRbzLQfdka3r7f+za5e8VoiTM5nhp0yZIZhc4DM6dUJBZG6YDbhyNO9nb48M4d7t4NpVz2dybkWUpiBfE1RkMMEPGYxiRqyMYIxodEZtwWnKDx7Vi3iMKXQRrTKUQ/m9qYwBrkrvtKC9ONuzQN0BhZElobJkKJpV2wdSE3jbl5jmpRcVaWTKsS52sMORZLKhmJGKwE80tjLcQ2jtl3JHXNoz8OdU6z/udNdPJyNCNcz3pMp+ss6zbEtldS+/ts6EbrOhv07o/L8OPu8FfvGM2zWv9G+ww3be//1Vz/CmTXt0aqj++SqGK06l6U90lUqVxBUS2ZVTPmzKhNRb6Tcf/OKbdP9/nkk2/xrYcn3NobszPOyQgkXaoaooTt7Mla/1l/q97NQa0ZT5oovHOOZVnw9OyCi9kcB6Q2I7UZxlukcZWsPO7nzB0vcHS248e7d2teSoRgjL/v17Eub90gbWA6576D6f7O+d9t9e/9yk6r/4f/5l9p9U+Ohq2e+o5Bt8+Mq+aqkNXeU/zKvGxdC02v0PAfPjtr9R//QVd/+dODP9bqSd19160QA/evevN1PQ+y8XqQ197EvMJQ3In3HbzY9pPu+3Umr3KmLdfwtuG728733sOIJRpe0RgttWDmLvBSQlIyHMJonHB4mPPo/hGHe0NuH0043h0yyRN2hznDxJJawajvSp+hlyeRbc9y4x9v+152jQ3GtMZyLzVlFYzR6bJktqxYFjXjxGAkIZEswHUxbTkLVo7UO/aGS9pkyF0PXKdb9Qqb+uhqLlJH29JbNF9xDnwRlKoPK3wj0jZ5g1H/nO3vg6gaTOnI/YzFBKqqplwuueCcysyopCA1jmGWcbSzy7c/2OHuyREfPvw2k+MBwzxlx+ZhUV95El+FkJVAU76lse3CAlJiqUIBH6C5KKgDk8bFpSdkv0gkyYnYTy9gbCRgQKLOAAAgAElEQVTn0Wb+D39LJNdXJ2DDuXzDwiOKesAoYjzeK8aG/04DxNdIqDkqCgmKFwMebB3Y+mtfUp1X/KKs+NJVeFeQ2jGJTTAyILOQisc7i/c1XoVEwvEVwdLllK7CVCWW1ei2+/4+dL13lXiI1ahksMBWjL8m4qjxr03DRHDqNZFY6dUHDbsHYqQmjrSmr5EqNc93Lds17g+Nib4dstvpTa3qTqcH2dUVYrjmHrUlfdgE2S1AMpxJe5DdkkGE7L43AdLeMyyKglk542x5RqkF2Tjj+OiIP/Kj23zw4Bbf/d73eJSVjIwP0NyiCGy0JhpmDZmPmF6CVifv23rDOcd8PufJkyecn58jQJqkpElIS2vez7400cJNc8jz5snmXZQtn7/rsv5s3/gc+hXITYT0Rm7kRlpRlEornDoKXVJTUpsCGZRkqSMfGI5PBuzt5tw6mfDR/SMOdoec7o3YyRPyxJBbgxXiQuky+URTX7D7ubkll5cv659v/lZfXt+H3i0Pg0HqKIqC+XLJbLlgVpTMi4qiUkZiMDYhkRSDbR1ETZtW2qb9Y78N2Xy/lbAgbAhdnLq2TSaaIgGqGSyWZkLfYku/ZIuucULd8KC3xKIu7dx6zd8TEXEskpppkVN8WeD1HKePqaoFqanZGxju7k043d/l0Z07fPv7D9k9GDM5HDBKDIkBMSXGJYEUxSzbchFIGVmvm1JFoaao+AR8MPiUFPEWUYOv0ljWyKJVWD4aIkwXwajB15EMSZvtYS+NJctEBHXSMtmqk2ikCOrD/2BMgdaQ4PFWcQZMofgGsusFr54aj3sKs8cFf/D0GRfnU5bzJSWGzOcYHTC24RpUDeIdRhREcbUi4iO8NkJT6cNUu+2GdeOtbwQGsS1Mt7893NnODd0fJwxIgybpLa8jlDeQL9Vtr0mkg/7aNic1wJqVJm83GrsNEy8Ns274nmeVQbfP4rt+/dC/F6t6e/2yBtnt67HhDWQ3XGNggA2OioCoQQxlMsTjUK0Y1o4ag2PAIp3gTfXOQ3bFhns9yxYsiwWz+Ywv6t+h8AsqU/LJLcMHDw/54Q8/4UcfHHK4M2QnWzLwBdb7zrC3BDj5iqN/WwhwPWralPfRWJZpVcJHwVNyKZyxxcn+0tLyQCjgIw+D4mzCrKj5/Isn/OFZwbO5o0ZIZUjOmEwHcdYJg0aoRdvz6MR+JJEgsa29utbm584xzTjTm9AuO1PWjcHWxbRyTwNhWtDrqkSMwZiAwgj3k5Du0Nv/vRC5PJf2HViXtr9Gt7wxSG/kRm5kRYJBWrP08zDxm5JsoAwGlsmO5fR4zMHukFtHE452h+yNcyaDlGFiSIyEBS/0DJdXGaGeNwFezRhttl2XrREMUk9ZVxRlxbIsqeqa2nucNvESEyKjbJjgn9/8F7b51Y23bcZo2BgAjoHIptKqPY/FYMRg1CCSxFhME3qMbKbvgodWVn6x6a/t28L2VdfBuy2KUvglhS9ZqkE15ItmqWOcWg6GA+6d7nHr8IAHd445PNpnOMnIMwn53ITaoqYxqVYYdD0SF7ANYZE0C2ONkc2GNbeNhHXvBNqD6dJWl6Td1F3Epcch68ra77Y+af8wsQRE6+NSUK/UlaNYlswuFhRVRe0cqIa6jRIMaFqDUrnEU9oj4OkacbX3YxWy27W/2aqXbkCf7mjlTrRH84BTHwxurWidZLGwFNKMOV1Bj/4xuvY017DZPbNNv6qs36XLZ1t7rMKGEzVXoNGpJ+E5a7PNh3fy2kb2NyMaHQmLas6snHNRXVDIAsmV3d0JH36wy4cf3OGjh7e4fZAzzhJScYH4z4fvIg02/Wpj7UvfkbfiE+0/cY0GTih/UxQFz87OmBcFlfPB+Sk2OnQ7w481rYXxrp+it60lPbru64vd6dL7J+0HMbc14hy8bty/5Tt8V6adjQuEbbtuftNaKPEryls1SGut+bL6AoDfWf79dvuf/7Pfa/W/+mf/6VY/HnYXZrXHJtu7FzXd9qvL1qXiC2UbNPN1pA9x/cf/5HdbfVZ32zNGvW90vpUeehfbr/G0pWnXC+noTmJ6en/gXJp5q3++7FiDd9JJqx8np91xfPdK+nfEh3SV53wVGPH7IB4fPLhaMnVniNQkieNgf8D+Xs7p4ZiP7h9yuDvkZH/MwSRjkFqGqSE1LfKPVSKfV+9vr/Q97S2DruP2Cy2pUe0ci2XBfLlgtlhSVDWV83gvoKEWo2zoDesB0a3rgDinbV18vPTtk7Wf/XsTS2PgqaUOTghd4GMpgcQkJCRYsdHMbgyTfmrEVwwbam2da3rQ74HUCtPyKWX5OUU+RlQwDo5HcDQecm//mO99fMzxwQF3T++R7qekVhhUDpVogHoX6G+NYkwD0zXYEAIFAiO2esAJJo3RLddFflQFMQI+RPtall2nPTgune5pU3ZW9vHasumqU7CBids7RRrdKyaJzLoaIvUJUOWCUcGo4o0ilccWjsLXLJYLlo+nzOuKAofBk1klMVBrihUQ8Xjx1D60PZWGWTeYrA1o19DUBNWWTddv0VdMzQ1suqtZpl0kMoyaMUapnUGjQKUeFx2FC12C1giOzGSkkqCkoAOsGBIxuDaKS6dLE6GUNlIKq7DeTfBdXdtnG2TXxEjQemSp2e7RlnXc09U4VXykNSJGxsOdSOuCSlJqk1IkBuMqxFeM6nOK94Bl1ztP7WuezZ9xUU2Z1VOSHcPR8T4PHz7kj//xD7h365Dbt28z9lNSrfHek4hgjEHV49+PIemlREQQY6jKiouLCz7/7HNmizmVq8lsihETxpUryno0t//B+tx0af37gn22rZelWRTEgyi9OUgDKzDa1Zw1EqKliQRHWNNu4fL5v0rZmMoS79GL1rIdJPrVX9qbCOmN3MiNtOLVsdApamrSvCIfwniccv/uhOP9EbePdnh0e5+9cc7+eMAgFRIrpKZbSFwejl7S6bP291c3J69NVCjOOxZFwXxZsIheXeeJEQoTc3u6aEUr0k1Anb94y1R0DfPT5ShJPLAG760S2JMddSjh40sqrVjU89aEHjAgNSkZOYmkW2MvN/J2RX3FxfIJzxYVtqg5HFtu71m+++ABtw73eHDrhDt3jxmOcwZjw8ASSrhoidE0REBtERYPAkY8gb/Wgzfhc1W0gemq4OsEvEWkYc0N73ujGwxaR45WtahzsQfE7U3OVwvZ7ekRsqtremDZDSOKAXylaK0ht9MotRFMEfwjtTWYyuCcozAOzmYU53N+URRUlQuGtFjQEaJDcgNGHaggXoPxLYpz2jrVOrZbaeF4Qi9Xkx5kdU3vZ4YGB0/oNb63dO1HkoODSFZztBVKFWpVKj9n6WsKdZT+nFB1VhloydDkjAzkNonHFVITjESnq+y7TR1Hr7Twe8c2yO7la1vZLpd1aCI/0Xht95HWqIcmDzfu3CZ2GFQbY1kokwEOj2pFXju8Co6ceTrGm+qd9h/VUvNZ/TPmyxmP+W0YKqM84U//8j0ePbjLt771Md89GjPKEtLynESqdu702lX53RhAft+kB2vtE4c5K8zrkl88e0zhHCoWm+QYSTGaBKI17VA5Ib0kvmPavaeX0A1vRLqzmdbtRNdbxLNM51R1TVmXfF7+nKqu8N4zMSNGoxGT0Zid8iCsD0Qwddo5feJ1faU5wNpEduM8/xUEU24M0hu5kRtpxeOpZIG1ynBsmEws+7sZp0djjvdGnByM2B/nTAYZo8yS2LDoMCGECKxGCF5WNg3HL3+86x9IG8+/V6VyLk48Dqdh8RSgVdEQjbURW5f/WzbbNhujQTTm8DgJhFW11iz8gtIXlFoyd3OI3lATwyhWbIzcbHIVdCQsN/J2xKundIE1djJw7I1STg+G3L+1z+nhAacnB+zsjklzQ5IqVhVRRcWFCL6EeJVIMFBENDLoxqWIDwu+YBlEa6ZZGCpxobxqSHWQnA2oBNmgv0TSbgOyaE+z7WVrmuqVqnYUVcWiriKcHgSDFRMhu81CV9cP8aKWtPJyY1UPmbC2V2PQdXuFJXatjlqVUh2FX7DQisKXVDpDNWCprQ5I1eBIozH58j1x0+O5LtEVbdvRO2P30hr4TXns3rB4dVxUF8yrKTpaMtkdcXx0wHc/usPDu7d4eHufg0RIUdTXYDv3RIM+CPLuX+vLSQukBRFq75jN5/hYaNgmIUJqYlkXoO/B7fX/t3lfOm9yiIxGXgztmqiqlK5kWS1ZVAueVI+p6iqgP6xDPKSSMG6+2xsmRV8P5nqt8jqLt2uQt2qQFn7O/7f4OwD8e3/519rtf+lP/Uqrj13V6r53Z1zvBTSvecdeB3Z7fV6Dnteod8yffvG01dN83Om9R2WkD5nrUyOsDv9vQvr3Xnp4ktrUrf7MP2v1f/jk77X6F/WXrf6D/V9q9cPkpNXfRQ/J67wvbwLi/SbFa00h54zylFunOxwfjLh1PObb94/ZHw842hmymyek1pCZzm+4Kl1O1Ctd8cqEs2GBu0XauWo1MPmaou2BFHBeKcqaZVVTVHUkLDEYY9uJtB8hDQv+tphErz2rC5Drkuaut+fpreBVumzRSquYh1hwVj9j6RYUfsnCLRARrLFtjosVi4uwuq94vrossUGv+ca9V1IrlN4xsUse3c25f7LPh7fv8p1v32JvvMfO6AByjxVPWvgQIpOYe0eN4IP/hDCeWwGtAwLN2Mh066OuEY5pwz4KSALqQZ22MN3Aghshm2FtGRZsXlf1jZBdApuuEOqRtjBdEKsQdZNAEll2G2bdamAwGtbzPlGMV2wN53gufE21WOI0EB9ZDMNEyROoNCUREyCkJqAcFCGNM6pCD7LLRshuvy9s0oPZ3xEetQTCEuYFg+mtdcM3nGiAauKZa0mpBZUveOafUvoptc5x2sTPYKATHIoTS+nHAcpsoPKhlRalyfxpiJqQHtQ2Ri43QXb7rMGvz7LbESmp9kmVYiuanDsNdz5AdpcRsptQJIK4EuMLhvWCyrzbkN26rjmbnVFScLC7y0cff8APfulT/uivPGB3lJMmCWY5a2tbN2gFCHPMO31x1yTWGpz3XEwvcB6MSUnTDGtt+860zt0ozRvzVUiDeVpNU4lEat4xXU7D/3LKU57i8RgMIzckL0uqql5te7y2dwmy+7qicXx5Vdjuu7j+v5EbuZGvSIwV9g4SdicZ92/vcHo45s7xDnePJuwMMnaHGZmE4ImNlP2N9IfV6zMLXs4Eul5zZHWi8KrU3rMoK+ZFzaJweB8iLdakIddyPUraa9frx4+fL5eM0Xg+T8OgW1NpRU3N0i9Y+GCEnlfPKPySUksqrUkkaQ8YFkrSm4xXj/3G5SoRk7itKR/xsvd3hZnxPTBos8RzZ9+A2+N7gx9w73CHB492OLl3xCBLSamwjmCAmhIjgaxIqWhyGEPOXoDVqjMtUZF3NkZDLb4CYk50YNAN77bWITQgSMegi8E71y3YalZ0Gt1t0ukx64J30i5qtAfZ1TrklRrnUQu1MZhSUSPUVpA6ofY1hamR83Oq6ZwnDryXUHvUpIhOAmRXQpVlVBCnAeLaMxqJevPO90uk9Lfr2j5NtK9htgh6904GX5vEvNLmiAZHMH5rdVRaU2nNzF2w1AWFLpm6p9R+Qa0lgUVXyDDhnonBqiFLHAaHV9uD7AZjL0B2uwhkA6MVVuG4fXqll4Xs6prOmh5yWBujll4p0QDVbc5m410skzwQOVGRtZDdlCIZ4k39Tvueagp08phb+3v8mX/qIx49vMeHHx5xL3Okfo4WGoDKAmINrmeDasMktmHEfRkJ79oVxmi5POoJq4GS1xFV384joWZU6CuPz+c8PltwNgPnE8TkJOmQRCwWwfhY9zgep8dtG8Y05I1BXJue2epKy+Jb2Yb113FuzymKgkU156fl77J0S0pKSqqwHjCGcTJhmIxJJce6SNhEmKtUtHVKvCtM7689Hyor0N+XkRuD9EZu5EZasVY42B+wN8k5ORhxsj/iaHfIzjBlnIWyLlY77MJ1R/jeVWnWC84rReUpSkdReXwsW2Ak6UVH+/LV3qFmme00RkW1oNKSuZ+x8HOWfsnCz6h8gO82k6LpRX2tWC7lxH5V8oLb+c5FcN+AWDEcjQdkYrh355Dbp2OODgYMhzmJEax3WBczHCVCcxuILt0yt4XpNlhYTyA6WlkCNkunnktCm5heD5PzmovXFhQRD9MihpvD0/XBJgrXu5JgzMQAm9cOxbB0PuZ8GqxJouMo6dquq69UO67J2lq+A0pcatPGa1n52V2ctCeReM8awp+Q011qSakVpVYsdcbSzyl0SekXOC0IhTEsYOMiPyxwjdg2J1Re1MA3IN27oFe6R80H/Wiy0DdSw07td7VbJseKpe92ZEmUyX7KnfuHfO/RHe7fPuLW/pCRCbnaPs6h4ZkFB4/2vgsx37d9z19BrmKLvuJnr9aQ8M43f82XBbNFxXwJ6i0mSUiSDONtmEWbDq+rh3lb43t/jRMM45C44vF476m05MyfsagWLKo5T+snVFrFyGiKWCGxCXmakyZZmEe1SXfoX5Cu/PmVy3VEFl5xIn6rBunJ0YR//1/7EwD8xX/2++32XItW97bbv0tb/jouMrqn7noQ3N/8yZNW38k7KGtGd2NW35felLdyk95QFKZ3ktJ0z+335j9p9d+a/WarTzlv9Ufjb7f6ndG9Vrd1/9rebq/sw2iv4m27CgR323Heh6LVw0HK9z4+ZW+U850HRxxMBhxOBoGS3oLFrbxafS9qmD83kRq939I8NadK5ZTpouJiUTFdOJyzCJbUZFhJImxX1iadtecu2nbWN/lGBLIET601FRULP2Pu5iz9IsB0fYi+LP0ieDRFGNgRA5sxtENGyYihGZGbAYkGg1vephtXuqhv/5q2SpwErwrf7bMIttvelVye54iVjG/v3+Phfskn//KAnWzATjXCVhKior5GJQn1NBsGXdFga7oId7WgRTBETSqo11DvMwHvokGYgLoIu0wkwHQ9mKh7JUBwY/RgK0x3k95AdiWcTyLLrveNrjgfWHaN0QDpSyC1UHsJpCbqqbMA2U0V6swhtWIWhp/WwueVZ1ku8aKAIZGUUQa5EUqfYTSQ+2ICDFgVEgmEP14byG7PONYusgh9dtzwCsVbQQfr1cuMs6bj2W2gvJVCrUsqKqaNAapLLvQZpZ9T+RBz0QgmThCsyUnMkEzGJGaItTnOD4JxaoTKh7WBkUBshETIbhOtbNrHa0B2W73PMiy9e9SRJ7V6vF8+LvDDPYgsu9KMleElSesFSEotCUUqiK8wvmBUnVHawdsdi15S0iTl+9/7Pj/6lV/mh987YJwnWGvwxTkWJbE25E1CzAe27Xv17o9ALy9h7eO7urgizKZTzs8XnJ0DGBKTkOc5SWUx3rT3Z9OxdMO0+sbaHk4a3mMJ518sF8yKGT+vfk5BQcGSOfPWKZ2alEE2YJgPGeeTaJSmSC1ttl7nEHt3nStXg+DqxkD8q6x3byKkN3IjN9JKnloe3dlnMsg43RuzM0zZGaSkpl+FUnueztVBp88f+XWSEB2FqvbMiorZsma+dKi3iKYkkrXlUaQhgmH9TvTdhq/oQlx3rm5oqRKIb5q6ooUuqXzJzF2w8HMKXVDIlFIKailRqTFisWIZJgOGdsTIjBjIgEwyUtJ4XaZ/mpXzvZps+eaLHGtr92DTIcNadTtsaJMx+rwmvUuyMxY+/eVdhumQydMPGOwodqcg8Yp4h1BGgiKPqEOMCwt+T8wjFdRZgglloh4ibL4WGmiurxtzRToGXQxaB0CqaDBMJS5Y1mG623QaPeJaDQFW29eVmJ/oBO/Bovgaah9Ku6gRvDEYF5xilREoMypXMTclcnGGn8+YecF5SyIZA5Nj3BijQ3JRXONC88EQ1SZvlfC/ZdmVjlm3iQQ2+/THwFU2hy6Kq03YT0IGb7jDylKJEN0lC11QaMHMneMI0NzCzXBa4gOVMQkhOj4wOUMzZGTGHCRDMsnJfEZmQtTNeYuNb7iLzLqN0dgahCuG9StCdumM8sYwhcZ4jfdIur7Wjzq3EW8N3/ZNioM2cXmhsjl1ADKTuQDZ9WQskiHOuF4o+N2Tg0nOv/jpLT66P2BfliS1IPF998Q8WrPK+tFG5LRBIsDXIhQjDXqhccOE6/v52ZxfTJfMAMOQVMbkdkJSJliVNse66VfaOizWU4XexD2SUNZKAnqhzAtQpdSSL9znXNQXzOopT/k5DocjMpWLJ5GUU3PM0IwY2RH7fo+0TjHeYr1pxxCv/af/hmXFqNxsQG4VfXuO2huD9EZu5EZaSRPL6f6YYZawO0wZZgl5YuNCFqJLj3eQ3uYNSM+4jhNq7T3L0rEsa8rKoT7FkCCShujEFpqnTfLSQ7xs+lN6TW2Wx4rDUWtN4QuWfkGpRYT/zSl1iTclSCC4MapYEVKx5DZnYHJyOyCVjIQ05rz0oEbaX4a/urzo+5uMyWv1Jr+769nnSpYabp/uIDJmIGNSKbG2wHgf41OeBtTYxKyaZVu7pmtZLAVVE8k1+ovgqMdoRkxWDIaqrr2KzVCw5m9pckKbBU2j05yi/6X+9pV9+seNhDk+RFCVUIO0uUo0GK/OO4qioCgrKq+gBiOWxKRYif21G81aU6opidK95l3e9KbmbZJ2v95O0vuwiTJ6oKSm1prSz1noPKIVpniWeKqeMRpyfUONUUtucoZmwMgMGZichIxEbcjpVyKpkPZaJGsaK5+9qptsBbL9nD22PdrN/qQeGqKxpIks0PGbXiwq/gXn/mpllKd85+4+p7spA1Mi6sOzkaRtt++ji5pwWfjrazWzNqR+MdgWRiOvPJvOOVsUlICRhMRkWJMhBBRHOya1or3/l13h1y6NQ0mgkoraBe6FJ/Vjpu6Chc4pZYaYkPaQaijnkoiwYyeMkjFDOyTXHOstpoHrauO2f0tx0UsRTlkdoK7SiLfU2d6qQXq0M+Av/3OfApA37lGInsog/euWtb++TiLS3fqy6hhqf/zTX7T6UdYx0VrXXX+93kcvq6/aqFY1PbSE6RUpnuq81f/B2d9t9af6u61+70HHDlwsbrX6p9mnrT6uJt0JVq6hd+IrQGJfFnL7PKbb14Hvfl0kTy0fnu6RGGGSJSRW1koJvGgREvbpfr7PEhfsBOhd7ZSyckyXBdNlyWxZo36IJSMxAxKSOOms9sU2L0pfo4/KqtLkF0lvq5fwTjochRYUvuCiPmPuppS6YKkXOFOitibJlEQElYTaJaTkpAzYT/YYypiBjBgwINE0wHXX4HHPvY6eR3Xz0qo3Ea8fSF789rTGhDznhrZ2zpYd3uOuKww5lPuM8jnyg6fY2mMvaiiJBCLBIDGi2ASoI+NqZMpFwWTga0AVyQhwXAWTCr4hG0p62xM6dlTTbNdVyG4SdX8F3Wlk0AXv4nFihDKw7PqwPQnnc2owsd5xlYT1VOI9LrWIQqpKmdXIwmOmwk9qy88cVHWBkZzEWBIZk2eWVISySvC+DtBfgVpDRCRBA8SVDoK7AjWliyz27fCezRz1uOiMzh60IWcKFGNLVUp/RsGCqc5Y6gW1FjgpqdWHiIsJ5SJUgksoNQNyGTIxe4ztDmO7A+xgxZIZofIh/pSIp4rs2k0dUAjctU0d1A6Cq5eimxr/rcN0VyG7DYNw2L/RoSOGavZZge+297HxScRyRE3NTYl3XpWsXlBKipOEIjUYXyERslu945Dd4XDI3Xv3EPUk4kPZJZT/n703jbFt2+67fmPOtfbetXc1p6pO39zmvXvfs/2Mn5/t5JE4NgFjFEEEQiIWYFkgRfLXECzhgEDiCygIicAnpEiRyAcQQgIJPiBBBLEQjZK4wTG2Mca4fe/ec09XVbtbzZxz8GHO1exd/Tl1uvtqHFWdUWuvPddca81ujPGf/+HCu1vn1yXWWELw+BBiFinvqZ3n8ePHvDh4gQJZnpPnOdbamF4thLbdtA4VPWM8f42iGiirkunRlGl9xGM+JxBAAqPRiMlkzGAwoKwrvPdIMGwPthnmIwb5ALM00Rml0fnb3MKbar6nOXbb+3uHJsPrCOm1XMu1tGKNsD3KMSIRppucaX1/+4lu9JMiAl8GSfcVguJ9oPYhEqZUnqoOiFoMEbJrtImQSv+rUW8jT0la7/9lpGeqrUSpEqMmviVGWYYFZSgomOHsApWSgfWoJVEkx5IiY6ch12iQbpgNRowYypBMcyxdcvLVPU7rYawTnlsyCI/z/nLOvV9RCzrJn/nuzL0vLdlAGd4TfBgx+M4mYks0qxBrYv5QDV2EzIGkBGodTFdQl9hzaWC0CabrtX3fHUx3FWrbwVoj7LaNIl4ashsHFyFCfyFGGiLjbtyLrT7uMbUSDTkXFKnibXhjMCFCIL0IWuZUNcwoMIspUixxCGiOYciGGSB+gJKRhS6vrg8pHysdsy70mHWTo6ch9lrp16kU1dVW2/aOBqYrisWz1ECljtLPmOoBpS5ZsAQpmpfFhonvr1SLhgABApaR5IxMznY2ZiRDBmoZGYcoBM3SzsuYFkgktPfQSGccrkNw+92jcSSdDNPtn2965x/Xe++5KVOacbAXIdX4DW0InlQSW0aE7CZ+8BXIbpmN8Ca80xONQRmHCiFu9ej1ku6c0Bkl7zD6+FIS3+maA1sjGkPEtM4bHzwvjiqOFo4a2LF7TNhjVG+Th4wsbQ+J5N/91mQ5LpdsCJ3nI6a6WouBGGPweOa2oK5rFn7GH/s/YB7mVCwRc8RokDMYZDHfc55jjKUImziNzpUddhnokMxnZJpHh9QqY1ebRi0y7J6BrFJeDjKbIqNNhPp9mPuuDdJruZZracWIMMrioG/Q04f6/tiq7/Ta4KWlv2gKIZKsOBeoXPTwOq9kYjFkZJJjNLHRarfQ0NWSeiWfcPgsOckJ0PsoMunGtC5FKBJMd4mjQG2FtR0z1iQAACAASURBVJ58ZBALKhJ3vYS4X8+YnFwHDBimPaNx36jVPlT3ZFjGeZHS9TpfnTf2kuW8B5PxRUUs2An40sI8Q4YVanVlkdX9Lx1bZ9qvh4CmdEWIoGlfU7dZsjEOaPeHttGtGPajfbGprUvfwdJDA8gJensOpIUqaS+hrrLbpuPaDDCpHOMTKLmBBUg0SoMzeK9U3uGqEu8cQQVDhiEjlxzRDMEiyZLuyHk6dlhpr33em+jvxeqbcvEZxWca6xYJfDxlC6OfMtdpJDCiJLc11ii5yRgZgxGFlBs1RhotQ8kYScbIDBmQk9HsFdX2XUFDMNS+nmQgngXNbWK6jbGpJ5xz2hM4ucxGX3/lx/VubNHe50CE5qaQvtHO3PWSE6R+pyI76yKq5Bp3DDdR0UiK09VZ+PIYoq2ccD/a7IlNbz2EgPOB+aJmWTgCMDITRrJJ7odYtS0ix60MKyc4cS8zh658qdOaSHt8PbHvOFEKKVj4BVN3yGP3OSUlYhw3RspkYhmNBmxPBvEeA6gMyE3AYNlgg1wH8V7I2tQuK51CmzHnAjfwUvbo6n2+D03tjRqkokruKgBcz9PRdw586TroKWJcd6NHs6LVv3jRQWI/vXG71fuQ3bo/qPWIRi4LOT0mfU9RD6Z76A5b/dcOf6XVb9w5aPW/9BM/3ur/49//9VafaMemu+l3evXumt6rbJh+VdjslxF2+0qiINrSeawOYrI2sPW83V86aVZHqmgyROvaU9aOonKUlaeulUGWkcmAXIZka3tIV1uW9H73Pr+sUbpSpKCJvMipZ+mXFKHgyB1SMCOYEvIlG2PY2LDc3N1CjOLV8+xoSlkqdQXKiJwxA90g1xE5Ayx5WrivLhgvLSd5d1+ioBMn7e/Rrhsqgz62jGSG//QFwXnsQsHV0fDJLFJHQ89mgi7j3BBhuilOlsWIZoTpxigkCpKnaKVqhOmGZOBEFGWC3Uo0lloIrl4eshu0Y9ZN0Ny4WFXIImTX+4DNIhtvUIMYITNQWzAhYAOETBCvZN5TZyXqHDyD7zrL0wCqNUOjDIxFGDPIYozFq8H5AAoDlJDSbqww6yZDrzO0unh/E3Fo4n9NvkQl5k5scplqMGTicXim6in9U5Y654ApVZihUjO0Sp5Zhpnl9ngEJsOrwKLEJbbhECxDO2JkNhCTY8kZyIDSg8WQm8jWKyiZKJ4mntRBc20CDLNyvIPjaqtHOemcqJOM5D5kt4P+ruttVFZPiMquQHY1LQAtKAxbyG5O0UJ2iwTZHb7TkF2IbTzG3L934j7Seq1WJTp2Yyep65rZdMbh0SHL5QIBsiwjy7KYn/eirzU5y5BLkvScI3UV06MtwoIn8ycc+gOOOGJkR2yNN/n0q3vs7e0xyDOePP4TlsslZekoljAabbAxHCN1jIh6H+HabX2lG0m627jINqiXl2aN+z6sdb93esq1XMuXXETEAr8MfEdV/6KIfAz8l8A+8CvAz6lqdXYp2g6g7fC1ltrlpEDH65O3t+gQNFHMKz4EnPc456ldwPnEyKkWI1nMMSYmsZueZyv14wiXl2aBrAQcDt9EXXRBSYEzBTZ3DHKYbA7Z2rKMRxn7NyaR8KWuOZrPqQmICpZhu4c0khhliciol3vyLVp/x4zRd39ePVWuoo+aLCDbjtpZ7OMxkpWIraJnVz34DrKrzkSINkJwkU0XQH0yPySlXWnIgby2nhJtjNQULYV0nuvBVz1tK2nhuyorOr733b4eUlS0uVaT7D40+y3TftaQILsa9yGakhjlN4IJ8ZgHKAxVFZhpgakKjHMoFtUhwoChidDBhhjJSsrS29xbEyDm9Mge0EYkYcU8TXUGEcWTYQhk4lioUmlJ6Y+YhgMKLahYkueBoc3Y3hiyM8qYDDP2xxMq71jWjrKqmIeAU8WSRQSGWgbkWDJUDZkAKnht3IeNMRqlT/1zGrPuafDdy0B2hYZZN57Xwp1pzHguAdmNnzo7IKii4si9R1UI5JTZ8LVBdq9mDm19mZ3TkWSQy5fUgXuGGO+iE0oVnw+pvXI0W7CsofKGgCU3Q3IZYIPltW+uVI3ICuKlmlSTXjyVKVjUC6b1Ed9xf8wiLMAs+WjseHh/hzs39/mRrz2kriqO5kv+pKipSocrlazaZZhtMsomWB1G5/QKMVPj7NEOhnEJ6aeP+rLKuZSQIvJIRP6uiPyWiPymiPyVdHxPRP6OiPxu+n/39Vf3Wq7lWs6QvwL8du/v/wD4G6r6CfAC+MsvVeoJ418XN3sN8rZnbO1VQUFDIPhIyuBCwIeYFzGmqoh5xyx2Lf9o+/XXUsHoe485RiutKENBqUtqCjSrGIyUyabh5t4Gd/Y2ubO/yc3tDXbGQybDHNumWTBk5JGpUwbpPnpQ3WMkYP07eoMvSns/77e8eh81iowCLjfILEeWaUXVLHKSxyIaiZKggpIWYbHNaqt3x9u9nNrXY9EatJd/oTtOaIyMJmraO97k21s7zoqeekyTQoZUnkboOyEZukFTXwRc/F7QdE1VvCpag6sDS1+hrgbv02IwT5BdgwRpQ6BCzH0ZGn3tMXf7vqV3y1005nhT7KB/QSSxDMfUS1UoKPWIRZiz1AWeimEmbI1y7u5s8Wh/mw/3d3i4u8PtyYgbw6xNtQUkBm+LxZInhxFI7MdCew+NidjEYfp7SPuOnVUDmxP/Ov2c48f7cZ9OP04AtfqkGk2SwdKVpwhebLIbFKshMURbnAwjnPf1DAZXNocey5XZv8HvIZHgMcFjVUEMziuLZUHlNJKJiZBJRkaGhEgAdJl3e+lHqtAmG1ZQC2rAGc9clrzwhzxzz3lSf8bCHGE3Kr56a4Mf+sotfvTr9/lTnzzkoxsb7GaesqgoK4erldxvMdAdhrqD0QxRG++HeI/asA3TDKE9hvDLyNteH71GuUiE1AG/oKq/KiJbwK+IyN8B/lXgf1LVvy4ifw34a8AvnlWQ0qO6fotW/llMq29KDB3L8MFs2upVD4I7ZHipMl/Vc9J/KofawXR/5eD/aPUf/lbnd/j3f/5fbvV/+Bv/sNX/i2fLVv/G5Cutbn0Ppmu7+wyh7tXh4mkz4Apgyl8SEZGHwD8D/HvAvy7xwfwTQPOS/jbw7wL/6Xll6bGA2Jt7ru2l16Kyb0Na0JkGau+onaNyLhql2sRHTPvvuDm6Jto4f3Xl2OXqRErsEfeMLnVB4Zcc1Qc4u8Tkno1NuHNzzP7umK9/eJMbkwG5FepqyZMXU3xdE5zia/C1ITdjhjJhyBhLNEqjMWrSSrJZHvQXVBd8Kw2sqr1dfdsB17cmV9VHQ20JTwYMdUr45AgJAVlarI8RGJNbZBEtUjMwaBkNUpMLcajVDqYLLZsukBh0E1Nqn003i+HDTo9TuAySMXFRmG6ynaMe8wkGrxGmSzQsI5Q34L0myK7i6VIq1BMwQck1LiTFKZnzFLYkVDU8DRwEYn5D9YwyR27A6YDMRoCcD8S9fWrIxaeIlrZw1L5hBT0CklaX5LgKbVs2eBSDDxahpCaw0MAiPGWhM56HIzxzjHi2csOtrS3u7m3zj37/h2xMhmTWQFHzxWNL7cHrAoMnFyUX2LAZG3ZAMErKHEvt43OzJuUcTfWLr1MSBLkZqTqCoT6zrm2PSxv1VFYhu6vw3bSWYxWyu8KmK/1IbNIvAtltluoKuVtSJ5bdIheMd4gWTOpDqtcA2b3KOTSWR2tsfA8vTTA9GK4gOOeYLxZUdU0IAWMMxliMMYk8bHXOOFUuabg2Iv3fRjDG4pyjrEqeL55zUB0wD0cEAh/c/4CvPrrFP/+PfcqjR4/Y2Zyw5Qs+++wzpkdHvHjxAuc9Rofsb04YjUZkeYa45PAVsCbrkFZt+rzebaheyiYR5GyG+fdYzjVIVfUz4LOkT0Xkt4EHwD8H/Pl02t8GfolzDNJruZZreW3yHwP/BrCV/t4HDlQbXkv+hNhvj4mI/Dzw8wAP99M+377R8IYtwrd46V4FkgmmStCATwZp7RwuxL1uMR1iXAI2+cVOKurkP9aOXWhyiQs/rw6Ho6am0oJaCkJWkY8Cww3D7u6Au7c2ubU74cHNHSYDi+B5flDEfHg+xAiZGkSzmLaGlKNRO8M6OgbWwkEvATV6I3Kp53iBcl6PXEkfvXdzB7Md8HVG9myEsRXYMi79VSPNamP4OcCk9+k6l0ncM5oWy757aC2kdk2ngfhqoyfyLp/gu8oKNPc0vTF8V/QUjW1guq2usQ4hgJGAGiUYkCpGH71IZCpVxROgDNR1zVRLtHbg425GrwPQjIFEY40QDR6bkrKq9hmktbv//vPvGRYNc6wKaDLnhIBKjqjHUFGqo9aSKiyYhwMKlngKRjmMBwNubY754P5N7t68wb37d5CBQ0NN9UVBFSrmrqL2nlpjFCljAAwwMsCQgUbGUpucRSExBQOdQairzLqdkX2c4GjV/I53eB5kV449r6QLjcneHpfmuPbaV3vBBrJLj2UXvMlTBl3fQXY1p7I5/iSy1VeXl+6fcHweVZOesyotVf2bDIYI3Zz0FofsIKG9fac1daipXYUPcUezsTlWDFbjj0hihu3vvaTXeHrHXioIIdKOj05qKp1R1AVH9QFf+D9moXOygePTGzk//P13+NoH9/ihu3tsjQw2VBzO50ynU6azOcoGNhNyGTGWCUMdkfsc0zVuAp4mlVIfRN9sh7moMdpHHDSEaWfd/ypC4f2wXi8VjhKRj4BvAX8PuJOMVYDPgTunfOfnReSXReSXn00XJ51yLddyLa8gIvIXgS9U9VfOPfkEUdW/qao/pqo/tr817hXMW5vI3uKle9IsWNP+Ue9wCbrrtYHuRU9os3f0pWt9zteaKETQCNOttaYMBRUF3lSYgWdjImxt5dzaG3N3f5N7e1vcvrHJ7uaIyTDHaCC4uA82GqSRAdAmYzQjayO9vdXiyu8rvq1T77X/16Uu8iqN5jU2uCvtoztjZKyEUYZMB8jSdvu+m32aiUQ3QnYBZA2m22g9mK5KguZG50oL022Mx9CU0+msHV+H6UpPB9Ygu7FOkvR2X1fa52nSd4NTxMe6hABSAz5BdrUxSBWtPHVdM/clwTk0BFQMqhmQkUncHy8xAEvMgKStndCsCxvjaeX502+J3XNUbIw+E3l04zqxptaSMiwowgFLnVLqHKRiPDTsToY83L/BR/f2+eD+TfZu7bG5M2I0jsQ9patY1BUuOJxGAiYhx8gAIRmkqY9G7sEmXU0ar+j8RtpUd/WTtXvrG490v/tAjhMQM6c5DvuG7kn66jnpOWp67z3D2JssockDNiTILpbajPFiX4kMcV1etX/CWh/dnkQjW+jq+YYZO9t2LG9mLj3t7iKeJ5Lv+eBwvqJyJbX3cU+4tYkOUBK7blfb0+rdRlBf5sZ6w2BNzSwccVg/50X9lAN9zNIeMhhXfP3BJt/69DY//NU7fHJjk9tGGddLjo6OODw6YjZbgIyw2YTBYJuRjBiEAXnIopNK40/Q0P5ER1b6uWy9LxsRTmOjvkdQ8QuTGonIJvBfA/+aqh6tQSVV5OTepqp/E/ibAD/8lQetG2/19Le/9HzTotLt7vj82RetPq86+GoWTn49ckk440W9SAc9mO7/dfgPWv2nf/qjVv+Fv/SPt/pW3bED/1KPWXd33J0/CZOurr16BFx3vMfoe1kYzmU9ZC/jUbsqWPBrhIr/OPDPisg/DYyAbeA/AW6ISJY8vA+B71ygkt1qbMUjeSEQzZdOAiESAVUlRVWyLAtKV1N7jw9p35uAtBZA801ZSTV63lrkPChrzMYX8DgWuqT0C2Z+CsOCfKTs7mTcuz3h5u6Yrz7Y5/7eJjvjIXsbAxaLBYvlkmfPD3h2MOXgqCD4nEyGWDNiQD/naO8t95Mr9g5L7x4vKisT6VnPIq1cL3x+uy7p9dGXgTOdUM4Vy5X10QjZzRnoFP14igaPFBbjQswbPLCwiJ53k4PWsU+bXE+A6eoaTDcaoqr0oLmnQXaTrpwK0w1nQHaxmkiVFLIQ9cS+K6I4DzZTrIWQCJhyoB7HiGyuSjCK1IHMOw7VUSxL6qcF8+Ao8Yg6hrYit45Kc/IEH1SrLWQ3o9sX2kJ2k7G3EtFLC0LT66qGkMh2MnwoWaqj0sAyPGahCw61wFCSmcA4Nzzc3+f+rV2++fWH3Lyzw8bmBrJlMAcVfrbgu08XPDlYcDiL214GYsBYbmSGcWYZGpsMUcUQqFxsrzFPa2y/Vnpw3F6kyfaAx33IbgdT1hWWXdPkL1zTW2hur5zQP67dXLdOpGQkGp9NxFUSZNe0VxbApowMS2rJIsvuwGCCQ0LJxB1RmyuH7F7dHHqCnMY++0ZEG6fL67u+tteRteWDttH7Jh9mURQ8e/aM6XzKsnQMsgE+hBa6ulLL3t8n1b+JFMbrXW7Q995T1AUvyhe8qF6wlBlihW98/Rt8+sFt/qk/9REfPHrE3sYQCXOqsuTo6Ijf/u3f5g//4I/44otniNxkOByykY/IygwRIYSeEdiLYkrPO9BPpdbm6m3GmnXyIlk9552Pdrb3d3m5UIRURHKiMfqfq+p/kw4/FpF76fN7wBenff9aruVaXp+o6r+pqg9V9SPgXwT+Z1X9WeDvAv9COu1fAf7bSxXcjJ9vyMP6rklcmEYPZ+UcVV1TORehRqEBxDRxJnpTzOWDdqd5fJvIqFNHHSrKUCaYbolmFflIGW9a9nc3uHtzk3s3t7i7t8n+9pjtjQFWFO8dZVUzW1YsCkdRBtAMwyAapW3Cmv40qfT/W7mbl1hcXersVwiQrhy8zM9rlqvso2IVmQT8yCIHA2RmIcQlvkZcXDT8Euluu2BraQs6Bl200yMEN+lA2ijd06WD7KbzSSy4q7qcqrfR0kRIFCO1EqOlXlbOaSG7DsRpIhCWNkLqUxlBFUeAooqLRlfhXEgGt6Cag2bkRGitEKOtRpLRSeeH034PWDF4mvijEKSJioJIhhHB4vAUeJ1T6QGLcBSjoiwZD2BvMuDDmzt8/PA2Hzy6w+17d9ncHTOaGAZa4VzJsiqYlgumdc3MB7wKRnIGdkRmNrAyiGy7RDZan6DMze7OprZhrc7SsBfTNPeTILusnbMeOV2Povb11XyKnSNula6lbwSvXjtmgVVS1Dud4U0eOUfEk4eQomcZtZngbXal/fZ1zKGrw0sXVX7z0nMyv74r0LaUJirXvu00U4rBq1LUNdPFgjooHkElknRFFmnTOn4uXYcLLlRqcSzskufuBc/cU575x5T5nM1t+OGv7vFn/5GHfPv77vO1/Qm3c8cWJU6V6XLB89mMx7OKWchw2ZiRbDNimyFbKX9qhB239WnnVG1b/3p+0OZYA1Xu30/Td7tzLv1YLi7SXfNlGurLfq8vF2HZFeBvAb+tqv9R76P/jthB4WUWu9dyLdfyuuUXieQM/y9xP8zfusyX39Ba/Z0WTVGeujFI6xrv+3nPpP135tO6wIOUNUXbfx2bbpHYdJ0pIasZjWFrK+PW/pi7N7e4u7/FnRub7E6GTIYROlTXNUVZMS8qFqWjrBQJedw72i5ybcpp3K+o9n76NXyZieeSX+hf9pJlv4x3tls6vHG5dB8VGzCbgbCRDNK5BR9NBg0xLQtWwcQcuq2xlYxC0XVdz4bpngHZbc5v94GuwXTX9c4IJdY15dfWBNkV0vW8tJBd72IUFA9BBaml1QnRCHcoWlRUZcWhrwg+xL2iYgjJIM0kLQoTbNeiLWS36cdNs+v2XjVLyaZvJJhuitgiNhJJSY3XJbVOqcNzFjql0jkiJZOh5ebWBh/fvsHHj27xwcM77N+5xcb2kMEIMl9S1wWLqmBWLZi5mrmPi3UrOQMzIpdIOGYSm3cTfTatMRijmPTqL/RALtKgMFbcTe13+4ZnIyf5hdaNq/4e0u7atJ+sn9OW0xitGtsd/eecTvI2J67tA1kIGDVATm0neJOv1fC1yUvPoc2YItL0wbcT2XpTDuXjxlJnUDURUheUsq6ZLRY4VVQsmAyTjNI4DvHSr/YiY7gTx8zMeV6/4Fn9lOf6BX5Ysref82Nfv8Of/cGH/Oind/hoO+emlIzDkgrlaLHg2XTG07ljLgN8NmFEMkh1szVGTdoHG+vTtGx6nfHkPnbRO3wTcum5UF7ye2tyEcjujwM/B/yGiPyf6di/Bfx14L8Skb8M/CHwMxe5YLukWYFBXrS6XybpfAF/8N3HrZ7nG61u5bTXc3KDXmkMvYe6AnVYe9gvfAfT/dWDv9fqP/2Tj1r9F/+lf7LVb5mOHfh/+90/avXffRF653TfHdQ9Zt2Vup5apUvJm2bZfRcYms8SVf0lIsEYqvr/AX/6ZcqR3u/vZVFVnA8sijL9VPgUHTUmwzT5R3t+zPTNroz0/+mDdT92kf6WGPnx6nF4FmHBMsF0NS8YDmBn2/Lw3iZ39rf45IObfHB7m53xkN3xAKMB7+LE/2K+5Nl8yYulY1ZC4SxGN7CMsTIi0yxOpL276Isox/maXqKrtbCj/sr0Ncm7DDB/1T7aQXZL9MMZGgJSZhGya8BkFlIqGDMATbDOyKwb25rJkiFKx6yrELdPrMB0+/rpkN1wCkw39GC6YQWyC2JjW1APkkUrNQRayK73YDISZDe+0QzFjSPhUa5CsCAeshoOvWNalpRHS2o8XhSjwjj3DDKl1BybDEs1vmXKziRGGCMEtZvHTDqGgsW3BpTgUjTa4imotKbUiqn/goUumOoCpMZaGA4y7t+8yYPbe3z/J/e5/8Ee4+0xsg3Z0qFFxeyw5OjxnINnC57MAssiQK1MTEYmQ3LZSBHSEUKO83HFbkRxwSAkZl0FkESzFHXTg+zGyCqdjp6iRyNy/Tg0e1VP0LWZG/vH1yG7adUTm1IsU0AldIt2Ce0dDOvEsmsGlFlAwjJCdusjKjM4YVC6GrmSOTRxDLTLhYaoJ5z5rS+VREeNICbBtL2jqiLD7uHhIcEr1liGwyHDwZBMMpx3tBvgLyGXWfuVRcGL8jlPi2cUMsVYw/d93/fxja/e4Sd/7CG3b99mKxOMn2JECAqurnny5Anf/c5jnj1/xnKxwPmAtRZrLVkWWYIj6XYAPT3Wt5LDuBcNviis+9w1r6zOf6c9m5VyTpiQG4h189F5a98OM/by/fIiLLv/K6c/pZ966Stfy7Vcyzspr3spf9Lw+E6aD81+p6BUzlPWnqr27cKv9f628QU509jqs96tXWj1LI1xUa+OOhEYVVpSS0mwMcfoeNOye2PI7b0Jt/cm3LoxZndzg8kwY2QNVeWoa8d0ueRwvuRwXjIvPWUNzhuM5ljNIrtupJNoJ8hzX8arOJCar78Bo/TLKmIU2QgEZzDTHKSOXoNkcKhPqVOQHkyXFsLa6ulFNHqMYOpxXRtdescbnd45Z+jEiGybJWU9+B6avJ20+RtjFDUZyprWeCKIi8s5LxIhfgiOgJYldZXgrgFETXIWDRAyslS4EiOkRnQlNeTxJtkZc5oSqMSKZYh6FI/XgloLyrBgEY4oKAlUjHPDZJSxt7nBo/v73Luzz637t5jcGDAYWXIctXpqX1MsFjyfL/hituTZomLpAg4h0yG5GTKUIVa6/d0mGe5B17vqavST9CwbR9lqxFNP0fslHR8K5DRdTjqux85vyjcr9emu1pEaCd5kNE8/CyGRHllqs4G/XHa4tyPSU84Z61ZshjMimm3fOeOasnbuG/Wh96mUUx1MIvtSVZZVYF7UTJclaEwJY8W0c147OTSNpcdjcDyg0Xyg3d8nRDpEhFocJRVFUfCifhZzjLJgY+h4sJPzzY/3+drDPR5sjpiYwJA4fgRV6gCzAM8WJU/mBUeVZREG1CpsMGGkGwz8CBNM+9BPyi26/h7kpINXIWkO7xuTb0QusnY4Ry5ManQt13It1/Kqctr4eAVj2RVLG8/DB6WqojFa1Z6Q0lKImLhAvoRP8JhR2sy3PaSS0rDp1lShotSSiiXelGBqRmPD9nbO7f0J925ucXt3k9s7E3bHQ4aZIRelKgJVVXM4W/BiuuDFbMmscBS14JxhqDmWxKyrtjNGu2qtPo32wKu+pd6y9NoofSkRG5Atjysso+cjZKDoxKMyiFFM75EsLka0BmOJugNJC/kQOo+3Nvs4m2hl8q2s6C6dL5KOJz00+8OkJUwy63podFo9khmlZaSBEKKBYowkll3BCvgQHUK5V0IOwQi2EoIVQiZYb/AKtXhYLqnLiqMQCAn+a41FdYhoTo7iJLQdzSa2Sx+khVM2Aay+ERej+hajEVSqYhF1CBWOGbXOKXXKIhzhxIEJTIYTbm2O+fD2Dl/58Da37uxz68Et7EaBFWXoHYU6Sl+znM/4Yjrnu9MFTxcllVNUDVaH5Gwwkg0sWRyRgraR26CQJaO6iSs2aGrbj1BKE4npcFnhDL3BavgUJRVijlOT1vtn6SJNZLmJrHZRUumXn+yWmHKnSfuS3n363NsBXh2oJw8eh+DIqeyAYMorZdm9clkh50n97MzqHje3Lnbu+kfHLdLT3KCvQ9aNrOjQMa3Ldl56pkvH4bxGQ4axBmssIQSCRGIjJDqv2i0rJz44WZlGmp++jytVACOGWmqmMuN58Zzn7gu+CH+CtZZbkwE/eH/Cn/7kJh/ducH9kcWEsk3nFryn8spBrTyeFzyelRzWOUXYALHs6jZDPyRngG0MUiGOM21N5Zh2mlwVyq91pK0QKp1/LW3Gi7ckb80gfR3oyouEpt8VEekSaf3BZ/NWz7JBq5u+D1E6l6D23d596cEErHblaw+G9Lx6tvKVX5v+cqv/1J//oNX/7Z/7C62+Jx0jbum6Ov1n//3/0uqfPenq/c2N3e4eerAa15tATtrYfZZcpKO+afjuZdvbm6jTuyy6+mtVZJVo492QjjxkWXoWhWdRenzaV2fExMWYdIss6d2EqnaospVIhR5PE6Pxqqc5jAAAIABJREFUk8jq63DqWPg5i7Bg7meEQcFgBLtbGY/uxb2iX32wz0d3dlqY7sgKFoUAi7LmxWzBH37+lN///ICnhwXPDyuqcgjesiE5GQMyHWAwGG320KX9sdKr2yXlNEKK5n47YLJ2q4jXKP36XGxp9m730w6yOyc8miIhIFWOrEB2o7PE5Br3aQrIgDZi2ofsiiEaskhi302R0XxVJ5GVyECOwXeDng/TPRGySwqC2Aayq2AjHLWF7JrIiyQkyO5GTA+RASFXTC3Y0vKFszyvA3VZxtx/yVm0MfBkRimbPKoKIgGX0oxY8e1C1jazpQpWfbuUNqGpgcGFgiqRix36Jyx0zkLnBHHYDPJhxod3bvHw9i6ffHiHDx/dYmNnjI6UzAVMCNTBEGYV5Yslf/T5IX/0x8/57PmMsgiEEHMD7+QZAzNAZIj6QYy6iFD5gGg0TH3qt038lqQ3w2wH2U17cltDUXpGY7P470N2aaG/DTtus5I4UdfV48dhvWuGrzbjZmTZbXEmQnvWILHsBskpMgMJsjuuD6lMftUsu29E1g2E+J564805bPYC0Xg/Zy3RMbS+/c0LqooYgxGhKArm85LpNOBCwOaQZdEwFZXWeXIZ6dpOBzFvjgOUZcnUTXlWP+OoPqKSksFgwKeffso3P77DX/jWB3z66adsDwxZfdiy5Erqb8vlgt/6nd/jd37nd/jDz57xZLaFsZaRmTAY5BhjWiKnk57263AJHGPiPee8q5TT5tOuTWtz4qXlOkJ6LddyLa8k6zCvly7kbc+cfUmGkmpcbNdeqX3AeaXJl3fWJNMN2n3YzPr5/b8TfZF6aq2pQxXZdIkw3WykjCcZuzsDbu9NuLM34fbuhL2tBNPNDJYIRQxBqeqaRVlxMFtyOFtytKgoa0G9wYQMMRY5nVv3FeRdM+bW63N2Q3vnKfUhQnaHSvCCWWREDGu8rw6yC6ttT/v0qyuQ3RaJCi1jrkZLpPt+C7vlZMhuX+9C/Z1OjHqsV6n5X7W3yA4p0KO0e+4kbjqMxz3J0ZOiGICXAHVFcI4qwVkhOo1UbXQgaZNsJJUpTZ5A+n6h9hwV096biI0xQw1ARaCg0jmFzqh0SaBimBvGo5ydrSGPHtzk/p097jy8xWR3RD6KkOHO9Au4ULOsS15Ml7xY1hyUnkUNAxFyY0GHGFI6June04qjms6n0ydlaj7U5h33j7fnr76Gdd/QSeP6qbp03+hfbb2nrZSfKtEQRzWlNpoXmz4L8d0pKAZvhgRzeh9+9+W0ul/MbDnbh6cdO/RbNkYbp0hjdKuAC57SBZZV7PMiBmvytGkkGjWXDXzH8pstNPGIqlJLTU3Fwi04qJ/ywj2mkorJZuDR/ibf/OQOP/DoFh/e2mbLOoYa69kMOsFYau+YLgt+77uP+ZOnc744qKhqYWM8IrNjjGYYNXEebfuAHK/gu95c28D6BSp63oLvFabQa4P0Wq7lWl5aTlvkXOwb77bExZ0SAjgfklHakYesJFxrc1/Kyas9msOrT6mDGfXYdENJGQpKlnhbgHFsTAw3buTcvbXJg9vb3N3d4u7uhL3NEUNryAmoxnQ0znuWVc1sWfLsaM7z6ZLDmaOqRthgsRpZDW3PID1p4fgyc+gbe7vNmuucWr4/re3iIjYg2y5Cdp9uIIMlOlkkyG5AXUDyaDSGWjE2ts3gI4ERrEF2WwhupwtC8JrOl5atF5EE30261zZVQAfNPQ2+S5cHVSJMt+lG6gUVSSQiEZmQGXAh1jWrwWdEmG4NwQrOCtYZgkJtPXm1gLqiAILGJa4RQ/AZYMhMoMYkY1fJRAmi1L0Ihra5OoUgGYJHFMTkMdE9FUKJZ0athyzDEU5qVAKT4QZ3tid8cHuHT756n9t397n3cB87dFiEgVcqEwgEJChlKJnVS54ezXlS1DxzgboS8txgJSOEMUYG0TiVxKKsihUlqEaYbrPPHWgwUTFamfQmuqlJhy7qSe8cYl/qw3RX4LsXgOn6FPVsyaDoIrGSDI42zizx/GbMbKK/msjVVMDZAaGB7HrXQnYLO0DfdcjuKXJqxOpKjUd968ZoU43GOAypDVYaKLyyqIi5hcmiQdo4WuQl5p00XmgwkAiyVJVSCmZmyvP6Oc/c5zwJ32EwGPDB/g5/5mu3+fYP3eeD/R0+ujFA3VHnPAsBNRYVw7J2PJ/O+c0//Izfe7Lg4BCMDLCjLYayQxYyTIjjjDSd6z2cdS7SWo5FY7XnwJWLulPOlnfaIO0/gHcRdvsq4nqw21ndua6t6aCv3tetrn3WrlPHtN5Ct9d25mHW6r929PdXvvNTP/VJq/87P/vTrX4r72C6wXV1/YPHXbrZX/29jh34wdafa/Xc5efW9SLv9rLv/yohsZdtb9+LMF1d/cW7Cbt9OWmTemuTB1RxIe5p65ZvKeIjcc+noYmonPwEGg9kWsv3IkQJqoujCEtmfsrcT/GDJYOxsLuZ8/D+NvdubvHxvV0+vnuDG+MBu5MhGxayNAnX3lM7z6KqOVosOZgteD5dcDivWRSKpD1pA7NBpnlKIXEGG2Ab5bj4G70oJPZYlOwyckFj9PjXLhgZfce7bKgt4WnOIMwJD2eIJshu4TFGMKMGshuZdVuYbgPBZZVlVzISBJcVmK4ZNJHUjk1XVZH8uB40wXH7OpeH7PqgSAPZDYKxIBb8IMbKMlXcMBo3OYrPPVIpWWn5QkYcYsHXSAsDFfKBA/Es0bgPVAGjOKIxZwnt+8+ImUoBTHCk2Ach+EhCpJ5D/5RZOGSuU2oqjIXhwPLx3bs8urvLVz66xVfu7jHaGRMyy7B2GFU8HgkOcR63NCw+rzj8vOSzo5LlskCriqEZYk2OtSMm+RDLAEeOOtNGRp2Pu0UNGsmbRFaNwGPMuj09jdAtrBdaXVmF6bYQ/qbMng5nwHfp+lpIBi4ch+yaZt+uBAwhmqw9lt3cLXGS4SWjyAT8EtGCcX1IbQbvJWT3VNHeCPSqxmSC9F5Zea8oxpjkvIh7RINCvXaOhtQHX6Ki0YEWW1YkR4rXWS6XHNQHPK+fU0jBYDDgR37kR/jWJ3f5yR98wJ07d9jJFCiw1saByAXExLQtwRgODw958vQpT548YV5VOLFM8gGj0YjRaISk3Xavss57GVTOlSF5TjAkj0FypXd/ZxieV0Gi9E4bpNdyLdfy/so7vq6/kCgp/UpQvGr06gMNjX970kvO+HH6DHh1OGpqLamJUF2beUYbOdvbQ27tjrm9O+HWjQl7mx1MN5MUDdE4qXvvKaqKZVmxrGqWlaeqQyQysjk27R012HZiOXFht3JPZ9/gyxiuF5IzinttS6z3odGKImlvqJSWLrFolhBnHcvuMQhDo67DcRtHSv/8/jna/4nR1x5K83Rdj5dxkjMiwgy1/a42dWluLXS304IRUvxOiZBdCTGa5pE2zhejg6aF87XZMZUuL2TblJrvRV0kmaeqiHiCltS6YBmmFDqn0gJjlPEgY3sy5P6dfe7d3+fOg1tMbozIRjblCW2SsEajLmggOEdRFSyKgvmypnKBEMCKZSADhjLCMIzGmUbnAo1zQLpn0B+Cjj/WdWbd7mZPaRatLr3/9QR9XeRU/aQ6RE17N3C8XEGlMcE1wa2jY14lj3Dq7wG5/HB0fFx83cboeptZuXCKosV5NCJ4gnYzRpxHDYGY4iw6upoY/SkisEreFF0oAEE8npraO+Z+yqGLOYFHG4HbO5t84+N7fPrgFg9ubLGZGUbikbCa7kgAj1IHx8HsiBdHRyzLgA+RIGmSbTKSMbkO29yp0lTj1Gd0xj7S9e9d9IW9zrnqpJd6XhT0CupzbZBey7Vcy5XLiWPT+7DYb0RoU0IEwAXFhzgpNdFRSXT27Wx00v2trRil97uZqmN6l4oqVBRhSUWBs0uGG8Lm1ga3b455cHuH+/tb3NvdYn8rwXRFsa3VEOtXOce8KJgtC6bLgkXpqB2EYLB2QC6JEVBtTOC9MsGs3UDPpjltGnotey7PmZCvxBjtra5fhYThbYjJArLjqAvL6Mk6ZFfR2ifILhGym8gfQw3GpmhVw6CbooRNMw3aENhEqK1J4fzgEkzzLMhuwmOuMuueBN9tIL7SrEnbfdnGCN7Hl5NJw7ILthRCHtl1bR3Zdp0RTLB4oMprNuo5uauoEUIwGGMjy27IQC25UZyYxO4KxsSITO2bThxhukZ9ur8hogUGhxiPhjkuvGDmn1FQUlMzzg374wEf7G/ztU8ecvfRTe59sI/NK6wKuVNqE0mTDAbnY393vmBazzgs5xwtSioHXjMyhmwwZksmGDawwWBVCBJiHLeBGtNBc6ORS4TvEqMZp8F0O/bd8xh3G/huA8eN7z/CbpvjDRw39snj+hpMN+mkunaQXRNLbCG7sZ3VdkTQGtQxdI4acOQs7RbY4r2E7J4nx5BWr4iweiPIwh6B0sq10zGPRxU8gSo5diPGyESUjrEEUVzweO8ZiD19jE9GkTRGKdHtIskg9bailJJZMeOpf8xj/yfU1Nzb2eZbH9/iJ37wQz66MeaDkYFQJhdUgwxQXJoEvHpmTvn8+VM+e/6cZW0JQcgkYy+/yUQnjPyILOQpNVWE/3c3vv6MklHaIyM67d1cCD57VZOV9so6x+B8E3H2d9og/VLDIG3n4Xs27Vh2h9mw1Y3tmHL7hBR9xt1+u+w/rYJFq//Gwa+3+gdf6ZUJ/MLPdKlk90wH0/W+K7jfEP/33/id7rjZavVb2Z1Wz8oOshtWK85JclXQ7Iuw7L5M+Zet35tm+32rIu2vVnT11/sraYEXSY08lfPU3repCoxJe0ektwuzvWVZIfqgN4HGo82uUU+pBQs3ZxkWLOSIYAsGA8/+3oRHd7f56qNbfPrgJruTIbvjnKFARuiiJOkyznmWZcWLoznPp3OOZgW1AzTDypChjOIPQzLNEhjRrFZ7Te+M0quMgvaiVethF+mf8ZokTbzNgrZJMfC+SAvZ1TnhQYLs1jlSJpbdoYWiD9mNDzlCcOMLNTmJGRLEaAvfNUY6yG6e1pqqmExa/VTIbk6nnwTf9YmJl3i9PmQXExBRfMiQTKOuIBYyAz6PkcoMjflH0ZjGxXhMGeeczxjxHIv6GmMC1sS90oOBYARKb0g2KBjFp20weVwyA5CpT31C0FBTq1KqZ+mfc+RfMA2HLFnixYGFva0dHt7f5dOP9/nw/g02tsc4NQzLyNDrjcGkiDJBsF5xZWD6wvH8jxc8/2zBwnlUlVwMO/mQzE7AbJHLRsyVilJ7h9EIzXca2YINMdrUGH6d3mfW7QzFfu7PDr57ut4s0hs4bp+JN77FVVbek/TGwG0M4ma6aHOoCugpkN1hYtn1MmhZdo2WTNyMOryfLLvniSbv0FXd2VWX9zLSsNB6hcV8zrIoqYit0xobWXa9wbRIALnU0kGERF6m1FXN3M95sXjBPMxxOO7cvMPXPvmAP/UjX+XB/ftsZwH8bKUM5z2gMVqrAecc00XB4y++4OmTJ3gfyDNLTs5oNCLLbDfvnwUdWJOrmmfa6zZlnlDs+lp13Qg+9j0FFV0pe7WsBl/y+lrTO22QXsu1XMu7KyfZEt3f788C/zxR4t62kH6AFmrUxEpjmKcbqLtncbK7NAJ1Q4qO1lRaUWmJ2ppsqAzGGbvbI/Z2Nri5M2Z3c8TWMGeYGWw7XySQUaqS84GydsyWBfNlxaJ0+CCAjWydkmOJZEZyijF66gOQE/94JTltHn9Ty6dmj9x7J6JIliC7Li33m0bQeuLje9K1BUe/5bS33v//NAea9s7rt4ETPBnrfphji54Tiol/dzDgtn7NrYXeJZpblWhmqUiMTgSHhpAwAylWJ40pZjFiTrm/xgyTaAFr4n0VxWtNrQULf8gyzCh0QcCTW2EwsNzd2+Hu3X3uPrzD5o0h+chgxCORNSqac809ihKCx9WOYlFytHBMC0/tA4JgxZLLBrkMsWRtpDLWJ76b486+432l36Zfpbd2LqhuJLtobznVwXVMuva4Xn4Hy+2opuL7tundX8tJcoqP762JiKHZWVwUFWXt2n3GBmkdu8YYJKxS7EU/bjfWxGFNurECCBIdu847pm7KQf2CI39IMAsmueej+3t85cE+H9/ZZycXhqINDTdptEwpW9Lco4LzcLioOZhXHC0d3lusHZGbMQMziMSAbQ/tG3Dn9Lh3aLpZ85FHOaP6J1ZdzjvhcnJtkF7LtVzLS8v62HWZRcuZBb0jEtfoEQ7rQsCHZipNBmmKkLZyygPoo8siTNdThxqnNUVYULCgkiV26NjcttzYGfHwzg4Pb21zf2+LW1sbjKxhYDQZpL3oUsqBVjrHoqh4MV1wMF0yXVR4bzHEXIY5Q3JyLHnHrpu8okBv2XnKg4CVrTuvLi/dWk6Rl19+9yOm77qYLCA3EmT38wkybCC7OSEopgfZ1cSyq9Feo1nja027rzKEHvtuTcusG1zD1tpBdpHmuw0EN3rdWzhue3xdl7Mhuz6aHB1kVyJklwgvtqUhZNqx7BqDtwZxOR4oBxWbbsbIF3gMIQhqMqy1qB8gkjPAUEskPCFIZB8Ggtg2ohjMCKsFRh3GWMqwpAoHzPxjFrqk0goxga3hgJubG3zjK/f4+GsP+fjrDxlvCrlC5ktqfISfasCLoiZGoqtlTVEVzA+nPClrnrqAqwMZllyGZOwwYsSGWKBGQkzCmpkQF94hYEUSZFcSZLcHtdaGcbfbE3cyZPc4y+5xnXaR3ofsngXHPVuPMGAUrLAK2RUTaZo0xUcF6myYWHYdQ+9wGGoGLLNt1Hw5ILuxe1395PdOkYBKk/83MJ2VLIoaD1gsVixWDJnJyDQjSFgl/BPTG9HTDJWM0eb118ZRUzMrZ3xWf8bj+jOWLLmRL3i4Ffhz33rIj358jx+6t40NsxiNl5BKlGSQxvKMAmooa/jOYc3nU8+zBdRuwMbGBuNshzFjxEvc8pJSWYkIAder58vvcb4Iqu6ieUjPKeXYbBnzkHdlHrtO/wuyWsJVOHffe4P0nep4l5DQQ7IWVcc5lpuNVpdTG3V3z6bXeGqqVv/N6W+2+uhOBwn+G3/1Z1dK+rC7HD2ULka7Ch4slq3+P/yDDrK7O/xKq4/d5MSark4a57+r0zrYZTrpy8pp0NzLQnC/zOzQLy+nPId39PG0/tPEChhC6CKkTUqJNkFC59WV9REbkLY9NLlGlTrULPyCMhQsmFKbJZqV7OzmPLy7zcO7O3zfBze5vTPh9taIcRYX6F10NBqlqoGgivOBZVlztCj44mDG08OY6iX4AZYxmdlIe0czLL29oycEus5/Nqf00bNN2jM/uRq5+OR40ufvQKKEC0moLeFZTq4Lwv0ZQkBcjhQh7vMcWijj3NFBdhuW3ViG5D2Yru0ChyaTlll3XW+ilSvQ3NMgu8f02HfOh+zqMciuNRCyCFe1qoQcRAKZKF5qTAH5csATGXNADuqwxpCJImoZDuNCp3Ym7nlFQUJKDSMMCHhJuYU1EILBqVD4J0z9CxZ6yFQX1NSoDexubPHR3Qlfub/NNz59xPatPawdMiiXiAiemBM4igXicwxO0Zlj8aLk9x/PeXYwZzYvUGMwMiKXCdv5EGuG+DAgECHGRpQ6CEZjRMlrii4Jq3DcPny3ixET2tji6j7Pfoy10c2K3kB2m33F6wy6DXxXT4Dm9s+h1Ztokm+ivqI9yC401ogAWV3gxOIlp8gFCQUmVEzcnNp8OSG776Wct8bR6Ejx3hNCTBLc7BM2xkSnkca5UdL+9YuKojjnKHzB4fKQaZhSUrI92uYbX/+Yb/7Abb797W/zcJyRZyGW30Do19dyEtfcy+WcFy8O+f3f/wO+ePKEo6Mp6BZ5lpMPBq0hfCkDbN14e81buU5ad4pcMSKo58y+qmLfe4P0Wq7lWt51udKw2psT7SKkQWn33B2XVR+u9I6tIFq0AeoGao2MuhUl3lSYPJAPDNvbQ/ZvjLm7v8Xt3U12xyMmg4xcJJHKsDL4xzypAec9RV2zLGumi4J5UVOUHg3DCNXVQYIBJriudlC5tVs+95mcCtM5x7h9U0mBjqVwOQmW9D6LJHe+soq5Smus1TaXHsFlH/tVvqY1/Oapj7+pe/+dae9nvYK98zrPfj/dRVqYSUPB02d87fVZsa2RZhFqAk4rlv4wMuqGOY4aY5RBZrm1vc392zd49GiPvVvbjDaHDKxGx1PzsHt1EwloAPWeqiiZLwqeHS2ZFjWFi5FUawbkZkhuRhiyaFb20A+iJz6E9nn2H9d5r645Z3VfeIOSOK1zn13eq8qJw4p0f0nvpWqbAOdajskJY/CbcwOefoYiiV23+dd8M05q7X52aB0rJ43bQjOWRRIh7z1LXTDzMw79ATVzrK24d2ebrz66zTc+esD9nQnbxmN9mYzQBNNt6q5NoqcI/S/qiulyyePnhxzOS4pKGZoJQ7PJQMZYzdp99o0Rrehx+OubkH7Hv6j0PVGnlbde5tq649RzXlGuDdJruZZreQ3SG8FWg3Eniq79v17SW5O0DoxGqa55NptlbKc3x7X5bm9RFQg4ranVUeiSggWlFMigZrwpTLZGPLq7zQd3b/DhnRs82NtmnFs2rJA3UEukM0aSMVp7R1FVzBYFh/MFz48WHM5KFoVBwoCcDXKzgdUce0Lu0ZXF30UmttPOOee763Df0721Zy2rL2jUfolXrCYLmL0I2R1+1kF2g2xggqKlQwYGDQI1KRIphFpayG6oU0oUGrbW2H4b1lxE0Lqvx4WgCGhNs45Ea6WF46bjx/TTILtOunJWILugngjZ1QTZLQw+VzQD68BLIgwKg8iyO6y4EaaMQ0nA4DUjkGGs4Oqo5ypUIi2jr4iiYijZIAsVmSjDLKcKhxGm6x6zCHMKLVDj2Brk7I43+OGP7/PpDzziK9/3gO29AQOBXCvq4EEzTBBcYo011uNDHfeLe8Ph/IgnBwf88dPnPCtqZg7UZ4ztiC2zSSa7DDAxois+hRyVgfUx9RQkyK7iVVLO0eP5PpvcoisQ3AS7PVNHW/xEA9mNUdl1lt3mnCZC20FzY1SWY8f7LLtGSE6+GNPtQ3aj0SFUNkJ2VWs2vMchOIYUXyLI7nki0swlsZ9dZC5sz2/+fsukRohJe9qb/d6Co7ENBQ0S0UdeWwIkVVrCQEUJPmCDpj2nQmU9DkdRFzwOn/MsPOUJTxiZObsbgT/z7Uf8xCf3+PaHtxj7I4wL2OARE9uZaxMie7JEG+4QCjEcVku+mM74fz474rszxdUb7NoPmTBhomNGfozzjhACapJ5LZGLqb3lEzy96/PdVURJ+8ms1ss+D513XqtonXwpuntef7uKVvZeGqSXhUGe9eLfJKSyf63Qw+yGhhcf8D2973IJfWbdHtw3TlFRfmf6f7e62XvW6v/hX/2ZVv/+nY7FF0C0gws7OnZc6cF/f/13f7fVf+sPn7b6N2/9eKvnPWZdL12dVrxGr3n+uCycdl1epS186dl0LyvS/jpXdEVZM/hWgw1vTBpzuvGnRkY6bQdp7UUsdO2b2iuh7/V1wbEMBVUoWDCjkgWaVWztWO7d3uLOrU2+9miP+3ub3N0es5UbBjYuzJslnq49Ce8dVVUxnS84mM14MZ1zMFsyXzqqasCGHZLLBkNGZJr2jradUleMUe2q/drk2NR5ctCH9VhN39h/iYumdtS9mfdZgjOE5xl5WKL35nEO8DmmDIgRZGjRMpon0kB2hQTTjWX04buNUQOKNCy7gOR9nRa+Kxld5C6tIM6E7CaYbgiRcbfpQ8cgu2jLyismGqNiEmTXBposoSEHQ8CI4qXCqJIXOc9lk5nMEPVkBgw+pmoYegyeurb45p5RgsY91AMcJKProHrOkX/GPBxwGI6oqQnGsz0c8/DWmEe3t/mBTz9m7/YeWT5iUDqMiQaykZDGCCGzMYotwSAeqjJQHjoOvrPg6edzHi//f/beLtaSZLvz+q2IyMz9cb6rqquqP2/f7r6+nplr+/pLsgZbls2IGVkIkBDSvDCgkRAvIMELvPCAhIR4QjwNshghPyCZkQUCIQGaGZgHJDSDwQNibHNnbM/96Nvf3VXnY++dmRGxeIjIj73PPlXnVJ1TXX37rNKuEzt3ZmRkZEZGrLX+67+WrHxNVCjdBGunWDujtA6JlhjBa52gjQaaaBEStDUk3STFj+oYgjuG7HbXOWLQlUuU8+gY503chOya0XbZUh6/rzdhvd200LHsSobspmsjE9hYIFL4FV4MQRy1FYgrRBum/pTWuK8NZFfg0iRO3Tje3JgMMNfdsstJP18KhBDRGDOPdTb0hoBkUiMAY2z+XdAYUQFjTUII5bpCSBwMxyfHHHPMqZ4ykQnvvn2fd9454rvf/S6v7zsmFckwhMFY6CfkDrq73lJijPz4gw/4/g9+zAcffEC9WmHZoaoqyqLE2QJ8alvHPPtlyGWgt1v3eZbmKkkZzcfepCHoK6mQ3sqt3MrLK9te00+dC7dNpOlN+II1iKdPMt2LfhQamhcMHQxorDxpD9UNeNrYUGuNlxqcp6iU3R3H0cGU+3d2eHC0y9HOhJ2qoLQpbtRkKvbzrVr3kC5WDcu6YdUG2qDEKBhb4CQTGalBtIMublFGL98FzyabloWLIEDb2jRa8D7vub/ypiOV/GE96O+pIue/Ks9wv3XDXPCUg8f1S7cQ3b5fdgSNtkmvxIzABh3Wbq0J59455x6XdctLV42IpFQPsWUVT1iGE5bxlDoTGBVWOJrPeXj3gDdeO+Leg31muxNKl0iLRDtD0WCk6uzH3QJafcSvGk5Pao5PG86aFh8iqMGaAmdKnCSDEdrlE922qHwRL0MdDZfLwSE2h/YTH6mnPmvdzdSRMWncnqGfvy7y/Hf9yzDrAgiqA+81jMudsTeut0xAu5CSLteodJHRybRRxxWLsOCxf0zwBBnxAAAgAElEQVQtp4ituXf3Du+8cZ/vfPMBb9494LAMFKbJyvAwZ29ZZSAkI0kbIl88esxnX3zBydmCGBVnLIUtKIzDdvmsWLe5b30ah4u8frlMndd5Xr2gfM1yq5Deyq3cyrXL1ae+Jy18XuBEmszRdIuebqlphOR9Gq2Kx9EwQ0sHuG43eUYNBPU0sWalC2pdEMua2S7Mdkpee7DLWw/3eevBIW/c22e3tMwLQ2UGn8PmwlSyxdLHlOrlZLnkZLnidNWwapUQLKoOR0mRPyl+dLMvL+j3m1RKn2Wf63gEfkLWsKYImDst7dJSvZ8gu2TIrgRFQsSUCjFBZ40jeStbBphu2CzDeZjuerlTCqMfFDn1yY2ZYLpJm9wO383bPYAgRoihW2wmL64iWAPBp1vVQXYDgl3ZBNktIrYVopGcj7QgCrSThntyxgfSgHF4XxBwiIO2NRQYnEpivJXEPSskqF60jro5ow2nLMOHLNrPWcRTWl2xbwsOpxU/99YDfuovfIO33nuNO/enVAYKUyd0kzqMGryASMRIgpeKJHh12yptaImrBR8vV3y8TEyjMYLgKHSHGTOmMkkw65g8rVYMMQZUI9bERK5G59HM75vs3Va0Z9Z9LshuLsuobGAdsrtRlq7MNrhv2j6UAfQ8ZDfTGmn28CrQuAlBW9CWifcEyJDd/a8NZPeqksKXt3n/voy2pHdH8D578hWhRXpUX0YRdBaqHDvaxBQznAiIUiy1RsGbCBJQlI/9x3zafpJguu4xh3vKb/zmL/Nr777Oz79xj2m5YmKgNBB8QBWCpHESx/HY0s3TShMij1aeH3z0CT/6+DOOly2zyRTnpsyZMtEJRUzJmTtvvh9NKpuP43j2vpXLya1C+iWJGT2989lAdasjuGsc7aNqRvsMmN0fL9/vy6flD/vyf/xv/kt9+ecfDgy4OmLPBTJQJokdYYHVDNv/6P0Bpnvn8BtDu8NsqGhsrJYLBukV2Wpv5Va+bElzqmCN6T9iSPDFzuoqg810rJQqkaCRNja02rCKCxoWeLtiNlNeuTvn3t0Z7755h9fu7HJ/f8peZZlYwZlNo2RevOdv3ZegShMCZ3XLognUHowpKWwBTHCUOffoCKqrbCzmtnkfZQ2mcysvj0RviF9kyO6DRZozgsPUihgQa9Amx+KtMesyQHBNV5YRek0RJ/27uYfsKogjQ++kZ+VNsNt0cMzHwhimexXIblqwRqWH7EbtILtKtDFR2YigRWLhTWlhGixK2RScun1qu8CKUliPlYAGxVYegycGg/rOQeqTJ1IhUFOzYsmCR+ERZyxoxDOxFfcPp7x+tMu33nmbe/fuUk0nlK0gRgmWDNNNHWydggoSLZXNsapRsCoELzxeRBYnNc2iBgFrHAUlO0WBMwXgsvM79WnQRN8jpHhRyYpc966R3HldLuRxzOca+24+4lKQ3S1lhXXI7lo532fGcN8x4+4my25qe+ygxiPIrvQ1pFdr4VeIGAKWxgnoCokNU7+gNfZrA9m9imyF7H5JolnJHMNjq7LCOZefjeQZNbaLG01GLzO6t+NlomrK41vXK86aU87CGQbDvXt3efONXb7zne/w6uGE6bTIsacJHqx5vnvyklOIMVKvVixXK4IP7M532JnvMHe7OHWZRyK+kIwPX1e5VUhv5VZu5Va2iNDFipCSd5tEUz+AZ7vlWl5Yw0Z8ZprAEqNuQ0tDNC3GBaYzx8H+hHtHO9w/3OHO7pS9acHECYWRIeejdi3ZLiEqrY8sG5+guh7AYU2BaIkV28eNbqIcx5C4J3bCy7G+uZp8Vdt9GVGBkOIfcfn5i6ML7rQWkXWLoNDH9Pdws0x+0a+xLoKZbfRn/3W8/3ifzXL3V54M2c065xp8Vzr3mmbPnYy2M8SWirGIsRjpkAyZiEwiSvKsMI6Fyv+r1nhd0eqSVVwQMqPu3qTi/uEerz844pUHR+zuT6nKnNJFO3dl7PtYRPt3Q6d4qSYFP/jIauVZ1p66TaqZE0cpBZWtcOIwYvo3Sqds9he95SaMPZHb5GlD+3rkooHWBQZctabh3SoMnjKV9LwLChq4OCXerbxskt41KeayKksKl9SODq477NQ9s+mfZu+7pokVT2KmP23OqMOCqDXzacFrrxzw7ut3eePeHQ4myRiV8oWn2rXzGo/qH78EOqRBUFi2gVUb8FEo3YSqnFHZKaYZco6iFz3xV+0Uhou+FeBWIb2VW7mVL0n0XOHlkX4NLYnV0lqDsxZrTQ8bUo0jlXRsyk1QuUig0YZVWNKwxLsV1RTKWcHD+3PeerjPG/cP+Mb9ffYmBTulY+IyGc1ap8io1J0teZMaH1jULV+cLnl8WnO6CmgssDLBSCIysliM2n6iPyfbT/X0ztk89qpy2Rn8iufol8I3VP+XLaYImLstzcoyeX+OlIllN8gUE4FVwFQmpRppBWPTQio2gpGk2Y2ZdWPwmVAkwW7NGmQXOpZdshKoflRukxvzHEw354hfg+mKpGNJbL/bILvGJEiwsgHZrS2xUIILPWTXGzBaEsXgJy333Skf2AYxBW0UMJbGBpZNxAnsYfEkkpRIhZcGpKGUGt98SuMf4XXBxCg7ZcGff+Ue3/np93j73de4/809pgWUrkVVMWohWoJRjFGMKD4m8Kk1LSHrrIpQB8+qXrI6fsQndeAzL2iwTN2Eqdmh4i6VVJRi8YSePMrJkG7KmUTC1Hkie+9jz7K7DtMd4LtdWa8M2Q19ufO4ruc4XffEDsRGsrbPFpZdBpbdZCExIBmyq4buDdi6KUE9qGfqAx6DlwlLt4+a+hayu0VeRsiu+DblHC0s+3u7zGbT/IwkYqKmrVFnUGOIMeJaQaJBY4QiIiZxJTzmESd6zKfNp/jiQ3amNe+99x6/9Qvv8EvvvcrDWaCkQWKLU0XF0opBTXpyjVg0BAyamHVJcHIvgnOWWiwfrQyftgUnYYILexRxj1L2KNsSa2wyGq1Z9QZUYf84yjClbPOmbnpPbyG9g7zkCulVV0qjI1+CAfkkcSML3+58gOzW7cCOG8rhYXd+uP7PzeO+/I+X/29f/g/+nd/qy7/+zVeGk8WBSTd0dNd9O4ZHQHRg1n3c+r78P/8f3+vLs/hOXy4o2SZr2d5GtyFeYuCNB+vz3MPLQiaueo7rZHj+Osu6Mvpy9dFY/UsJ6AVnkjJqTGfHj0MqmKz8dOkhE6thoqVvtGYZz2hlhStbDo4qDg8r3n79gNfv7fPgYM7BpGRSmJzaZfC2di3o4XnZNBvJTIMoqzZwumr47GTBFycrTs4C3jtEHVYLjNhEZjSeJRn7NQaPxPjK+0lyi6es8/4Az54Yu5/TnzyerqRYrlV/eW1Uh874Skj0Bn3kKHSF3l+ieIw6zColfpfKQJMXgx1kV0AKyfFTIGIGmK6YjDpVTJEhu5r2J4JqzMy62Wu5Vk6L3wTZTe1T1Z7UZygrugbZBTE6oABGkF3soLCIBSsQTUREsUIP2S2MELXF1krpC5b2gNYusZJiLgWIbYGUKzAFnioTewESsdoSYs2JnnDCGUupscayPy+4uzvjvbfe5cHDB+wd7jKNFvERHxWb00SIEazLOTKjUCZyWDRKUiBRfFTiInJ6GvjTRw2PlktWbZ28vKbA2opJ4UCLlKomgs2jL2iGuwoZsts5ijsvbAevHuIuYVBYh3L69iRoblfreZju2Kt0cfnJLLtDufPMd2Sn6fXWxRdKhu+mNUoRGgRDkILGCqo1og0Tv8TfQnZ7uQmYrnYVP0fKmG7tY016ojVEiqLAWjuYVVXxISQo7MacY4whcTAoPnjO6jNO2hNWrDg82OOVBzN+9Vd/lW+9NufO0YSOBb9TxjV/NyZ5N7twBUjruNhRPQNN23J8fMKPfvQjPv30EcfHK6LupPepGK7eC5dXM7vUNs96CyUPpHNrTRm/I56l3he/1r3FPdzKrfwEiIgciMjvicgfi8gficiviMiRiPxtEfnH+e/hzZx9oPfpPpc87Cp7v2AZ8uGJSA/XtdmVkGJe4nDl/cVnVt1MZOQzVNdLgy0ju7sVdw/nPLizy939GQfzillpqWyu+ykv9YxETCpxVBofWDae00XD6bJlVQeiN5joMGPPqI69o3quzn7qGf90YVOuZyF4mUn+5jPovbhF7bWNURXUGyQoWgUotQ/mEyGlbkGS98lIT8rb51oYlxltk/PbeoPD+AEZd9m2Yy4odx/ZPG9XZ2ekyO1T6a4HsEk7k06BMSRiL8mQXTWoKcG4DNlNsZ0hBBSPEkBiymMokjkcAlFbVmFBozVBPKWzHM6n3D/c4+HDVzg42mO6M6EgoXM1KIQhbtSY5EXuutTksWaycUpVCU1kufR8cpbYdWufaH6MFFhTYK1DxCWeCO05vNe6eH0s6pbSxTDCy7yTn/b7eByuly/ef9tea4+OjFXY9d8FMoNx9tCK6fvFqkc2uDCuS17UPHpds97ms/LSSQ4JELKn31rsiJ+kiwvVqCOFRhFNMHwVJRBofMOyOWPZnhHFc7A/4/UHd/n2u29z/94d5tMqxVvrxvOuYz7w7MHvX0Ad07zQti2nZws+/PgTHh0vWSw9Ri1WErNuf8x5nW/t7+i0F96Wa/WI9lYdzr9nx9vH+192yvsS7D1PVUhFZCIi/0BE/m8R+Uci8h/m7W+LyN8XkX8iIv+1iGx3l93KrdzKi5D/DPifVPXbwM8CfwT8+8DfVdX3gL+bv1+bDAro+VXplRTTS5xorOxea91bztUVukWmFaGwhsIKzqYk38lyG7MVNkmvLGqkiQ1NrGllhRYtbhrZ3Xc8fGWHN1895BsPD3l4uMOdnYp5kUhQnAx1dC0Y9+14Yo0xWZZXjeds1fDF2Yrjs4azZUSDw1DiOmZd3W7hTTVfNKVeTmRzMlyv/MLtV1E0LwAaX6+8mMn3WsZoD9ndtZgf7sDHBbFpCcYQVAi1QqVoCaFNahIq+BYIFoIhtArRQBRCG9LCTSE2EYkgEbSJSMhW9kYh5nLLWllDUsK0VdSD0by9K3tQn/dJCMy0QA3pu6iAT0q2If0eG7BBiF4IjWDPLFoLPoCpDbGBtgVpClQNsfTcL445dDXGFoTgaKKwEs9ZE2hapdKCQpRSYGIqAkpNjecRpS7ZlcCD+Q4/9+br/Mq33+Gtnzni3htT9g4NshNwZaQUknKcPyEEoMXYhhAcMRgkRtoo+CDExnDcLHi0OObks885rj2LIMRQUlEx0SnqdzDR4UiGr45xtMieQyXlYoUND6WOc4KOPZHDIryD+AojIiEYlQfY7Xh752Xtt4/eh+vl/Ewia+/Bzt9lzpUZPLGaIoBRS0ffFNUmgwtC46Z4KyieSWhxUYhUnBV3aYrJTUF2b3we1c5wubY1jj6XrKev7OWUZMjN/AVqMBgmhWPiLBVplaAaUZ8MRzGk+bQIUHpwbaC1K87sCR+vPuRj/wO+kB9Rzht+9u19fv2nX+Gnj5S7ZkFVH7NHTUEEtfmTaLI0RCQoLhpKU2BJ0OBEomShmPDobMUPP/mc//N73+dHxzVnbcG+ucss7lOFnX4eTZ3e3af0zuyMT5cW7Qzaz7mS2ZhHZePf2naRtX0uXfcLVkovA9mtgd9Q1VMRKYD/TUT+R+DfBf5TVf1dEfnPgb8O/I1na8bopozcxHEEyYhSDI3WAYJ6kTwvZPMmWLLWzjWCD9ydDdfWhFVfNiNm3XFz/mz5p3351/6Z9/ryP/czb46OrYdTrcF0N1+DwwuwGFmuPj8+68sfPF725W+U94ZzxGH/OGL+HRMOrDHuXgKOO95+0f6X2X5ZeRHnuGkRkX3g14B/DUBVG6ARkX8B+PW82+8Afw/4967jnJt3L30fPa8X7jmSzrr3xHdyfmmv7SP9YTcwSs/VK0BhEtmQM5BweSl3IeTJaORBihpoNDF3BttSTQzVrOTu4Yx7hzPu7U85nFfMKktpxxlGNyeA8x0jkpwzISo+RBrvqX2gbgONV0KAApeYdaVIymgKLMrXtd2PokOx/3MRlHXIQjlMcn3Pjdr/tO1Xk5dv3F1FrnOMqjfosaOILXpvhYhHsGuQXfXJM26Kjl1SEGtz3KYgxo1YWW2/SJatkF1FCvqh2EF26Rh0swLVleMag26C6YIOgY75EDGdfyet6CQzYYpNz3lURUz6dJBdJ0osIsZk7iZpMERcsLRmj2jOKIiI8SnWyyt20mKMJ0iqPGpkFU9Z6SMaPcbHBi2Esqi4f3iHo/uvsHf/LjuuwkYhNAEXWjTljUCKzutjKCwpMDKAsxEVQcVgSakmYlTCibI6jXziA03OaWONwUiJMQXOAV4IUQjqE2kSHUxXekURBo9nN4q64Xkesqvn95HhmxnNY4bz27tX2madulHuzjB+BY7L8aJy3km6e98pxTLw9drQAIYolsYoqh7RlsqvCNZcO2T3hc2j+eW62fpBwcw9+5TLSz2/tZZnkv78axCG9e1X6fFu3RTj8MQWZUFhXY53TkbdqJ2ZYjirGMGKJYRIExpWfoWIMJ/M+OY3v8nb33iT1197JSGXcrtiiHSQiu7Z1eyh7dqjMc/ZxiCqYIQownK5ZLVaZUU1t9akmHtj5AJY7fMbA/o5+QXYFbbFrm5/Cr+8ufapHlJNcpq/FvmjwG8Av5e3/w7wL95IC2/lVm7lafI28AnwX4rIH4jIfyEic+C+qn6Q9/kQuL/tYBH5N0Tk90Xk9z8bGSLWZbtvsoujjJpIMILm/HJPabBsfrnos/Xl+AJe4Eof52REsEZw+dOrWXkhPm5lih4NiVmXhmg91dQwn5cc7E852JlwMK/YnRRMnMWZ0bLvgnngvPI/xN60IdL6QONDileLgqjFkCG7MsJmig7wzdEJB2V0vBCR8S4XtOnim3DR0uV5fZ265Rn8isi1jdHPjxdobTBe0ZmHiSImqSQi5FQs+R5aGVgmjSF2eTiNpBR/CQKQn42sPHa3v2O3IS0QeyNQVx6NUxUdPTbaQ4Lz2q7fZ2DH7c7R1a95/+RyEJP37xRPl5VTIcN3E5EQEhAiNhrUTBApcAJIBAIaA0JAJKImx2YSWemSOp7SxDO8tlgrTCYldw8O2LtzwPzOPpUrMArqA/iQ8oNCzwIsCFay+TWQfHxJq+pjyWOM+EVgtQyc+IDv8oqKxUiC6hqjSZGFHJOe+lBzB8ugrg2dqjA2ZeVe7793dWwa1gZ1Z317Z5Ra3/70sXq+nm6MPmG7Dm0b3kBjHtQkRkNulSEYk9sXcFpj4o1Adm90Hh3fn7FusH716V6n52G7+XD9nOufJykTesFna71d3U85/9Ok9wTmydRZm9BGeawrkTi+l0pKeyQCRmhjSx1q6lhTOMPefMZbb7zGqw/vc+/oENv1nEg/drrv47Z319rFqvbeW0me+bppqOuGGHMdInmMpk/XJ3ngj95tw/dza5a1bds6mH6dcb6Ojf2eVNezip4fq1+2w/1SMaQiYkXkHwIfA38b+BPgkap2zDc/Al674NhLLHZv5VZu5TnEAT8P/A1V/S5wxgasSJ+AD1HV31bVX1TVX7yzNz//O5sTWCrFGGl8go0uVg2ny4bTVc1Z3dD4kKnXL5Yn6aEbOtO2Vvd66WUm2CtJrx8q3Xp9UlgmhaUqDMYodJDdrByqQtCAV08TG5ZxQa0LKBr290vu35vz+v19HhzNubs7Ya9yTAqTFNxzFzxctGampHGMboyR1ntWTcNyVbNYNayalraNhABGC6ykT5/bT7b0Tb/q2NQStsjWyVA2vp2HC61vvx5J/bAxka710DNWerNyfWP0aIq511DvOuT7O+hHBaFp8cbgMYSaBNktBO9Lojcp7UhdE+uGuKppVwtoAwQlhAZMQGwk1AFiWuRpE3tobsyQXSOZIy9ID8elh+zSQ3ATlDcrTePtF0B21RvUZ4idT4y9NqZtvgV7OoLsrgxaQ+MVmhJVQyhbXqkec1AtcYWDkDyOtcCiVZogOFskkrFwytI/4tgfcxzOWJqGO2XF2zt7vPfNV3j1jXscPjzA3A0UU09pPVIqUlqkKEhEKR7rGkI0yQiE0npD9IKESBsE34CeBT5fnfFoeYY/XRICRC0gTplIxVQKYmuRoIhGnMnvVlWsRLrEGKZTWCGzlyYo7JgwqDM9rW8fwXFJvsfEjsvadoF+O6N6rlZOsu5BHcF0YSCaWYPsGpTEshvVZHZiqN0Ub0BpmLQNNgpRJ5y5V6jdjUB2b2we7RAIaEifbjCZzIAshigONQ7vprR2xkoma3niLyPpVS7nPkDKybn5GaO/YBTXe33S2aiIEacNM+e5swMiifivZoUzBaWUWLV4o6yM5zSu+KD5kPeb9/mcz3ntYcnPf/suv/Er7/LnXt3l1Ymy355gCXhXEIoCtXYwpBlBbLJoqVVCEfAuEExI1056Nn2EZe1Z1IGzBkzhKKsJM3NIxS4uTocrEQhG00cUFYOKHX3Sc5z6cfRvZIEQ4dz2Tbjthds7tVjWZ9QBAnyxdPus7afntydPclrfdbG942M2y5c592XlUiy7qhqAnxORA+C/Bb592ROo6m8Dvw3wc998bWur46hj7WgR+4f/9IO+/MHJAGX9Sz8zML0+r1wEwbxpKO/4NfONhzt9uQ4D1Hb82vuw/aQvfxb+aV/+q3/5X+3LBzIw4/oNNt1B1ts/htqOFYgfvP9+Xz45GSC7k7tDW9cezFG16310CSvrJfr0on2ui5V3U65a15fMpvsj4Eeq+vfz998jTaQfichDVf1ARB6SDEpXkvXpatiiUQkhsmoaWh+pfaCN6dfCWvamJTiLMxc9h1cQ2aZN5XXB2nbpFa/rmlRFwBqhKgxVaSgLM2IHjdmiq8SODVBbaq1pdEmgpnTK7k7BnYMZ9w7m7E1LZoVN8aLSXck2ZXD0XAMJ1gaoEmLEh0DTNKzqmqZpaFtPCBGNINYk72h+y4z7o+8uObflCZ0wsg4/lzzP8eepIJ6fHOKFEu5f2xjVIOiJw8UVeqdGbEDEIh1ktzQpPhTB2ICPLaFtCMsz2pWAGIqZw8xTOiMxKZWBKgmym9OOSE4Xoz0cNz9LdlTusKKat5PL3QTXw3czFHgcCzmC7Ip0Y4r18WVS6hGtkofUiqJlRCSlhVFtMCbicGixg3UnTCwYEyEoElNeQmsDKkJrPEtZ8jh+xlk8I9IwE8ve0ZSjV3a4e3DApLBIDOiyRWMgDarURhFNzLoIGgzWgebutjnETKNBtEVDyhHcHDfUZw1LDUQiRiKFJREZYXNaC81jO3la+7QoSXXPDibp+478vWfK7bazzrgrz1Du788zl8eMuxvsu/lkMbuFNiG7wytCcaHBI0SxtBZUayS2VDFtvwGW3RubR7tOjlF7pbHjH+hgpT4EFosFj88+B2PZ39/HOe3jh8eSnoH1t9fT11FXWetuQoYv19ddu9YVsHTxglCWJVVlmVTAQnvIbmeoyAegUWmalqVfUscE1z04OOD+K/e4e+eI6aTGSrvG4XB5Ga5NoxI0cHx8wvHxMWdnC7z3lAImD+htZ7j4rMKTAkqvoj9cZs272ZbtMNzrmenOn+f65UrmF1V9BPyvwK8AById0TuvA+9feOCt3Mqt3Jio6ofAD0Xkp/Km3wT+EPjvgb+Wt/014L+7nhOmSSzEwKpuOVvVnCxWPD5bcrxYcbqqqVv/VA/p0+TcfDj+bH0dbiqozyedxdgaoSwsZZEUUpNNvv0/TRNrqz5BdbXGa0MwDbaI7OwUHOxNONqbsTMpmDiDFc0ejXwd26A//bV2C/fO0BzxwdM0bfq0ntYHQtC8dpZMZDSQGa1xID+5Y5/QG9cj+kz/+oO33+NL3vcL67xhuc4xqkFgabBthL0WZjF5AsiLXWdQtahK8nrGlrZZsnr8mLPHxyxPzvCtJ6pHTUB6WC9Jq5L8vHTwXXSA6XbKZsfvMXAmpf27spG8j44YfbuY0HwhwrACMUkRVTKDrtWUGkUUY0ipXrLzQ21EbNqO8UlJxCLFDFeUTKxkeG9ENGAlYE1ERWnFU8uKk/iYpS7x+GRAO5pydH+Xg/0dqsIg0UPToj6iofMkZYXUJLImQlL+RNI1J8bdBLcUjWj0tN7TnjY0i5ZaI6rJ55X4GkzS4k2CS0fJXkbtPIuDl2RNYdShPGzf2Ge0/Wrl/Iz1C05dW3xultlaZktZz20fIKHDoB7T4ZnoU5xfD9lVRAJFXGKj57rlJufRNGUNnmPtDDSxe7cLdev57PMv+OGPf8yPP/yIuvU9T8GWxl65BZvw3idDctdN0ZeG7z7BkSMiFEVBVRVMqs543Hni4uhhTkriql2xahY0vsZaODzY5d6dI/Z3dygLt1XxurT0caUR7wOnpyccn5ywWCwTUZmmdDVpHz336dq6VZl/gVbOcVu6cz+pvVeSbUsDHf7ehA/mqR5SEbkHtKr6SESmwF8C/hOSYvovA7/LdS52b+VWbuVZ5N8C/qvMdv2nwL9OWu79LRH568D3gX/l+U+TJtUQI6u65dNHJ5ytGk5XDWodrnDsTifsTStCF4/xXGc7LyPHwc3IqG4rKWn2fFIyn1bMpyXOCiKRGENK7yKBljax6mpNLSvMJFBOhYODkvtHMx4ezbl/MGO/skwKQ15eZWTBU3pIupd/WsqEDrJb16zqmrpuCF6Tm0YNQo59WcvzkRbNV7kX55epzy/rcWTPVdGXd+5nl2sZo6aMyL2GVeOo/mwHnSwJO0uinWLUorXFVGlp3/qC5Q8+47PPHvHHH3zIblWxtzfhrQND6XdQUxBlibEOMUI8q1PZGrQJPXlPbLpUKwmym0O8UpkEKY1tetaMCJ2+IGLQNqSyEbRND6EIxNDtk9h0IaERok8LdCtCVCEEwbWGWBiCU+zKEW3E24itS4geLQN3dxfc3Q0czCuKTx21BloRTtuSeWGYTGpYPKatj1k2NSqBsrK8cs1xrpIAACAASURBVDTn3Tde551v3OfwG4bStDhRjI2gBagDaTBGMRIJrcGYiLEeHw14wXqlzWzWJiqroPjaw3HNp/WKL9oGbXz2XxbEOKcURynJwJSYjRUxMXlKNYUKJFhr7t/sDTL9du29oamc+nCTcfdq5c5OIAS0f4N0ZbOlLE8pd7Bew8DlmO4tWXk3KW6QdL87zqy6mBDVgzZM2pDg6FScuT2iaW6KZfdG51GxZR48BudbJCsMn3r4s/c/4+/8nb/HJyvPnQcP2H33p5kRqLYw714VqXdRfOkwr1yPbGuX5IEuwO6k4u7+Dm88vEP56TGr7An1lSeY9L450RMWesYHzQec8ENisWT38A7ffm3Kz762wz05ZkaL04QUiqZ7fi4nUSNtTGiSJkLtA598/hmff/45q2XNtJxSuQkTN0vGoRG56OZ1Xoacc3P79aAs9WbjPWXdUDXefpNowMtAdh8CvyMi2f7J31LV/0FE/hD4XRH5j4A/AP7mszZiDOqLo5v/N/+b/6Uv/5V//p991uqfSW6aTdWMqv/W22/1ZW3+9758Zk/68p8sv9eXf+svf6cvf/f1o74c4wCtxT4fVPKzLx735UkxwHTLUXafq/bR87DVXoYB9+ssqvoPgV/c8tNvPk+95/S+/CUGpW48Xzw+4/FixWndMt2ZMzUGFfN0h9u1yKbZ7ooa1yVqFhGcMVRVyaRMH2tStJVqJBJIkTAtK10RtCHalunMMt113DmYcbAzYW9eMS8dpRVs16miV9cSSc98DGky9j5mFkNBxCRWQEyf6mEwcD5Lx1z3DXwJxuqX2ITrGqMaBD2zuNjCYQMuItZi60T0Y13yKqUYzcin7YoPTh/z4w8/4N7BEaY0RFtBCVIlQiSNCb6GS/BdQoq/6tl3R9NJ5+FUGMF3dYDBqa5BdhlBeXuWXXonBZ3nMZU05xvVvHgjk/5ohu/GHrJrRVDnERQXLVLuUlYL9qsK5wzGe0wMWLNAmbL0ngUNtanBtDgDk6Lg7vyA/f0DdnZ3KWKBSMpbqj4ipHOJIfEkiWBtQiGgCS2RILs5ZQsp8txkCHStEd94QhuGGEuB0kYQk9Y7Qh8rHjV7XdG1OM9unhMGz8QA2ZW1N+G6h3R9+9iLelGZtVp1az1DedyKoS7ZUu/42POQ3a7nsgtewQWPhx6ymywYniJ6AtwEZPfG5lHIfZIzGajqWt99+ulnvP/BR3z/h59QHB1QlBXT2RyrZ1wlFcwzNeoG+hHyve/YeTPPQlWWzGcz9vf3sPYM9Yr3Hi07T7zQNDWrZoXHM5mUTHYr3nvvHV599SFHhwcJkRAiIcZnmtc6QqPOKua9x3tPjIo1FjF5Ls33yqhcuObcqoBnOPb6xovn4GfRNV6CmfRG5KkKqar+P8B3t2z/U+CXb6JRt3Irt/LyyObrUhFCTArp58enfHG65LRuuVuWuLIEY/Kq9WYmun6hI+MtF7f3ec5jJMXZVWXBpCqpygKbiYhUAyErpK021Loi0GBsy87csbfnODqYcbBTsTctmRaOwsS0iF1bIl7dOBOj4n3Ah0DICqkZsQIKKd3Lsy3atoH/nldeLET2J1k0GDizOBuJDxpEI9JaZJUUUlMo6i0E0DbwUb3g/dPHvP/xh1hbMt2fo24ClSBVRIIhhqw7OgtNUlBNIYl/RUEcWWkF49K+GtNQ78q9ctqXUzyqZOurhpEyO3Lj9WXRYR8D6hPE1xiINiksghKKpMBYBe88RhXrLVLsUZWn7E8qCmewVonRY2WJsmTRtiykpjY1YhqcK5iVJfd2Dtjb32W6O8cFUIlESbkLjYkZQpwUIImCLZNRIAbBihKMoNZgjaZYXDGJnTcqTYzEJhDbQMiLUitQ2aRuRmzv5VSSBzFZ/oWgg2I3kA1t5A1lMDh1I+xJ5fGxOqqzA7FvEhtt7tOVx28wGJTTznPTxQ3q5v5ZAVK0N8prl9ORoUJBsbElSoFi8CazGMdAGVc0VJcdLi+VmKysxN5TluIlP/70U37844/54INHfOPuXWazKbPpFFsvU1qhG5ILHKfXI73hZJCyLJnNZuzt7mHNR2kO9T7HkUaiwKqtqdsaj2c2nXB0NONb773Lqw8ecLC/k4w1MbHzJsXyas0SkSG9kULrPcEHVBVnXGLXNWYgFroGhf1ZFOeLvKk/yU6YS5Ea3cqt3MrXWyQvNLqwl8YHTlcNH39xzGfHC5ZtYPfwkF0xlGU5WBpvrD3bpVvYPW2/p0taDYpJ5AaTsmRSlUyrCmcMRpRIoI4rFKHGcxaPwXimBewd7PDKnRmv3tvj7t6Ug2nBrDA5LC/2bb2s9KQfpAnJh0DTtrRtIPjIJlTX0Cmj2bp7yZONI72u2nfrx8r57S9gHn1eiPhXQUzhM2S3oPyTHZgsCXtLop1hqIj1BFcJTbvg85NP+N7vf5/vvf8x/9fnLTpvmK8i5bxCZIIGiP4s5dwD4qLG2CLFofqUOFREMmRXM2RX1yG7khla1+C7qa0imYn3XJnEviu5nGG6xqQ6NXtAoxcigvVCLAzRGczKEm2ktQGzqlI8u4HZrGb/0HD3/gG73/+IRRNYBMMXTYmKZ2/+A2g+hnpJHS33qymvH+3y7b9wh3sPZ+wcFhS2RWsBL6jL/LCalGljNcXktpbksI34YJEM2fVFyo0pARoNBN9iTxYce89ZVGwQRCyGihB2KZyhJODVYhJvUoqjDQnmamU9lUVSWM/DdIdykvNQ3qHceUAD9BHmYyU16KAortWZt4/3OQfH7cpywXY6yK72sON0bRYVIWJGkF2hKaaE6IGWiQ8ENQRJkF019U1Bdq9NNuGwIoL1NWRvYVtM8EGoQ+CPP/icP/r4ER+18AuH+zw83MP6GrkohvQp59zm0buKJ673bo7a/qxirEXyc+uAqXMczOfMnWVJi/gVj+QTmqKlqio+WrzPmZyyZMl7u5F3Hkz55T//Gq/vlxyYCKHNqAlDuPzUNro4zaSMAd8GfBCWjadtwTJlNtlh6nag1f4ZHUtvbNHx93H1umX/pzSpH5vr+U5VtfeujrffuOj6fP6i5FYhvZVbuZWniA6rh8wE6UOkbj1nq4ZFXbMKaRJ0zlEWBdaanAPzhbZy67Zn9RH2HgEBYwzOWKwx2Jw7NGqgiavspfDUusQQmTnHbOrYmZfszydMS0fpUtxo583oG3XVWCDoKdm9j4OHVFOs30Bk9CyJVp5nAtKNbzfDwnc5OX/Xv4zJ9aZEo0GXGbK730KRSI2sjxgbMSWoFPhWOHtU80Gz5KN2xco3lK5lUkXspMIUBiMRfCLDUiVTxeaPUchcI9sYdJVxeQzZZQTZXS/3XtFcHoMcRHSoUzoPAclz6jTlJ0WTopiVYy08xIiNEE2BcxWzyRRXFcjSQONpdcEyer6oW461oTYB62B3p2R/d8rOdJfCuVR3G4dAxziCvXe5Xb3BOLL3V1IsePaQ9h4/I5iYwhqWqpn9OmZ0foqNLaxmJax3CWYCEumNT9uIgsZey83t/T25YH8ZHTF4LWV9+2jYjEfQwAi+Xt52jqe1FTYUJ+lc5F3mzRQNa6JPBlCxWfEIQKCIixuD7N60DPcp3akQI4vlko8//ZTPv3hEAPb399nf318L67q8vHzvuRhC9sRK7/VzzuGsw5jELlz7BmtWqESWfskqrAgEptMJe7tz9nZ3KIuYng+NabwIdJDgZ5EErSUZd+uGpmlRTblJJcPGO0PNlqNzWMH61ueZZ74OxtSnyUuhkNrOnAp8ejZYhP7gT3/Ql//t114dHfHlDbrniYMcSxilaNmbzvpy6Yb6/+j4H/XlExlS4PzmL/3FvjzVoe/0OeNGzejaPvtiiF+1ZjrsE56uZDwPpOAyx36ZkIWfZLjEhaLQcThGVUJQ2pBSvZzVDWd1QxvTeCitZVIWOGt6WMyLb+9oSfSsk3o/2WV4oDFYY7J3NE1GnYe01eRJanRJIYq4CbOpY2+nGhRSmxava4u5Z+kfJUN2hxjSEDr1LymkaPaOPuXyt06el5wVLzy2k03L9YsYNjo+71Pa9xUWDQZOHIWLhLdWoIrUDhMjxkVspcTW0TaGxaOGD8KKT1gRYs2saplPFTOZYlwmVakTBFwVnHVom/KY2DISM/2u2ATfVdUM2U1KrLFsgezqxeVOIQ1JwUTysVkJjbme5DU9D9mFSKgiEhWrii9aJERsKzSmwLkJ88kUW5UYl9yt3hyzCCWfLFvOYgsmMnGwvzfhcH/OfLKbGDWjoquISPaMhhRnJpIoGbQVCCb1b0wePZc2EcRQoOTwW0xIIY8LTQtyjTGlhlHBGihd8ksGNYNCKCQyIzrI7hiaez6Vyno+0U1obuyV6cFDOcBxk+cy1ba+XTb2H+rsPDRjD9F5aC6MoYaDypv3755hNhXknPpltIMLHWQ3xZCaqEhsKUP9lYXs9vGFuTN8CJydLfjo44/5/IsvADg4OODw4CDBRn8CVJQQA11uzTWF1CVobPQNTVMjcYkPnmWzoMnpD+fzGXu7u+zs7OD0DLTN8bf0Rpw0R1+9n0x+dr331Dl9WoiZI2Vtbr6o7nWTy6XWrpxv61r6xAvmrZ8kg+qT5KVQSG/lVm7l5RWRBCfrJtIQFR8iTRtY1i3LuiVkWnfnLFXhcDbFl32VJeYFJkYonKUsHFVZUBYGZwUksIoLoAGxiG2ZVI7ZjuVwf8Ld/Rl39mbsVI7KJg8pPIcymo/uYkhbH/Aheazp4boWI5nSqPMgbJ3j9MLfLtGCpx970W+bl/208z9Zo778eX+CxBQBeaVh1RYU/2QXnayIe8cEO8FEQzxTbLlC5YRGPyd88YjweEGtDs8+avZxhUNsDaFFo8dEA0FQX2NsiRQ2eTYyeU8H001MuYBktteuzADZFcn7bCuvwXezP8x0jLuSoPB5fyNmgOyGBNlVazGnjmgD3nlkVaB4vHgKaqoiMt2tODSRzzXSxMQWvJIly2ZJKcKsKLg32+Vbb77Bu2894N47U3ZmgdIGzCyibSKEErNMbhi1hDp5Rk2htG2C2FoNtGIQLzivtKVFokW8obGBqA2TszOWUVmpwUSLswVWKoKfUzihFKUhw3VVKXNO0qjn2XTXlcM0Djchu0M5s6EzMO527LhPKg/wXdkK5RUhxz8O0NzOFNYz9IpcDOXN47NXiNMLtlNHUQbIbu2mBN2E7E44K/bQm2PZvREREcQYYmgRFGOhVYcPlqZVHi8CJ02AUtidTNgtSqRpk7XmKyo9s28JZBOJASZlxdH+AZOJwRaKNson+iniHxObSBs+wznY29vjz33zPt/5xgOOnKdsPSYHtYu1iLEpRcsVJaXbCViXwhG8h1UdaNqYPKSmwJqCQl1SIC/wxHfzoI4tLVdpRzbw3MogtwrprdzKrQzSvyA3bdljd5vmWFIlaCRoJHYwKxJhwDjm8cWLrP15FlmLh8yKduEspbMU1uJsWjy32qaFoUSqAsrKMJ8W7E5Ldqcl87KgdGZLcvNNQNtlGpW91KSFXczoQu3gusYgMXtJM2FGN+dt2GTX/my5+HzANYOItlX1pG5Ygw8OX/p78wxd+JMgGgVqiw0t7PoE2TUG8cnTKM6g0RG8o1kZaoQ25/mclpFppRhjkTiQFmn20CUWUElw1ZG7Srq+Vh1gtzAqr3u3ngbZ7cr9LRw9bzKC4w2PsWRPiKI2eU6BBN9VTcqrgi1LJtM5tpoixWkiTdGsCAWlKB2mMMzmuxwe7XF4uMukKDESUCL4iESbNTvpc9ybIkFENQimy7caDYaQILsuJVlCSO2LQozCkhQHSvYMGVIogzOgYhJkN1EZA7mdKj18uesa7f/K2rZNlt3hN+3L4+1PK4/9lmse2TVkx6i8tj9bj4Xx/pvlfO3d9NKzCSuiIUN2hSCdBzjgYk2QZ2NY/fKkC6ToDDtmyBGJ4GPs06RZm9A4si2A8aso2ZMZM9TWOcdsNqMsU2hPBBrfoBJSCiQNuKLg6OiAe3eOODzYH4W7aO9pjc/VN+uezRCUGMEYmz8pHvwiguNzyuRl26KgnUf3J+HeXrO8WIU0v2yBtbwnxgxQ0+CHJ2DhBzhqZYf9ZfSUDLY9YGwxu6GbfV3pYNpRAx/cGVK3HM2HtCrf++L7ffkvfufNvvxLbz/sy0abvhy0GLXz6pa1MGrTWTOUJ+W8L1s7PDLaPnsnXxf0eSzPkg7monN/LaG5naylIslLH+kWZ+vkRjEzBvYLWmEj++VYblaDWF/cDtueVTpPppgEMSpcQVkUlC4rpEbxbYtKxIhjp6qYzgp2dybszyv2ZhU7k4LKCs7kGnsWFxhW+VdrU8fS2C20lS7lS2bYlc340S2An6edVrtX9bPBoS6SrYrlE/cf/t+s4+sCY9oU9RZ9XFDYBfGbq5S/clVgVoo4wUwsuiiIdUl9WrGwlro0mBDZnTbsTlqMtdCR98TEc4oBh0O9AzVIFVPqGBJkNQZ6ll1yzKl0kF0dM+4+AbK7WSbDdLu40YR+T940nwm7bccmm/x/cRKQqBgV2son+G40RByuKpjNSmS+B+Uxor7PiSlRiM5hpgXzgzvcuXfInaM9KioiLSF6TL3KiqMkaHR+F5oCtBXUG1ylBB1YdqODoBYXAypKsIr1QoiGE2NzWqYEVzZicCKURYJDezXpvZpfCUENNsNtOwKjrtyN6nXI7jbG3XVW3m6cXATHXS/3KmLvZdXxPpqZYvM8YOSi/YftT4TsCqhEtkF2i9CgUqA4WiuY2CCxpQorGiZfnQW9rM+nXdqRGGIPYdXMGpumhSHlyE+KxC6WWpTCOXZ2dqiqCuccCmkeJRKJFESqquLBg/s8uH+fo8PDXEt+Ro3BaySG+Hxrx2wQUNI7SNXgnMNai7UWEwwxw/c1jmDoz7g27I/rDIC3ck5uPaS3ciu3cl6UlLR8ZJvvf8qKqGZoWWdnkk4hzTGWfUUj9bTzdHQ/XbeN+zpVp/QnQa0Ka5iUjllVsDet2JuUzCvLKgYCinOGg72Ke4dzHtzd487ejMNZzj1KxOYFl3Yrz1GXXnZO7Xo0KviY4njbHrKbFVKTCRn6lC+6fvBVpF8oryul/SL3Oe0L48Xyk/a6lXUxpce80rBqS4rv7cFkSdw7IZoqKWqnNdbV2PIR09kxe03D4zry2JR4PaSJ+0TvMUWDmBo5i5joIArRN1hHv2CWmDwFMWs7xtDDdDt2XMgKS+jW3k+A7G6B7xqBGLLpIdcZSUgLDdkrGQV1NnEtLQvUeWLhscuKIIFgPS4oZVwxM6fcb4/5uF0lJTrEhCowwl4UjsqKX3zzLq+++Rq7D/YxzlPWLeIbYtFmJd0AIZE8CYTGYYzBluBbg8QU0+gLg4TMsls6knbqqItIlIbZ8owmQqMWjQVFUeJkAu2cwjicCG3WICUqpY3EkJQ3m52nEc1w1069GyC73Vs10imErJXH5GJxVF5j09VBEexht6P9TacQ59f4OmR32H9QfMfbR8quXAWyC7WbEtWjPWRXCExYFLtE077ckF0F6YM00n0MClbKxJLeKs5ajEJbNzS+wWubrbmJvAkJGGLPtKuZox3o89YO58qybTLJBtCLFKmtuTRJ8/91iA/ZrGKgEjBlgdnb4fXdkrOJYwV8whKfnUsHVeDhnQm/8DPv8M6dHV4rwfkFNjSgAdWQIQs2xZ93Md+XFAVaYwjAMignTcA7B4VDFobSVjhT4kNeA6lge1PL5WUcRz3M9RsG2Wd8hK9aTw+f3iIvk/PlViG9lVu5lXNyDnCbJzTtJrZsWex0TckxpP0nL5tk9K7cForRW8pfMkltThqXkIhInLFU1jErC+ZVwawqOG5qDFA6YW8+4WB3yuHujHlVMHEW13uMn6C9XXo+SHXE7Bn1IWao0bBQ7YBh42pvon+vSym9lWcQFbQxmNCiM48UeRTFvEgViMGgwUKwiezOGWwIpHWXpljQALSg3cAUesiuqsEQzz08WSca1r35b+/w77+sgyzWABdsO3a03Ns8VkbtkwTZ7eG+mQDJGMAI4kpMscNkMqUqy5QzOLsLjRHmlbA3cxzszqlKgzOKiZ6OZliipP7QPGY7RdwNj3qnmPUswSJoTnGloqhhgOxG8DESNWYlLkF2TQ7ITAYq7a9P+5fq+qAa3kdD163vIVtK2/cE1vWWC14Q49t50X27SNaOfUq506hk9INAguyS3sBRelo9rKbYyq8CZHez97XTI3XwtgXvk+FABGNNnwNzY4hd6jW7rUekO93WCi7uw+vo3c5L3s2AqGKMUJUFR/t77M6PsxqazRYi7OzMOTrc58H9e8ynFaUxKdfysJLIf/JFbb5vLtGm7m8k8WFE7Tz4kjgYjDmf8uWK81xvyL3h+fFyU/DLP1bgS1FIu4E4bNEwwEvbdmCfdeXQvMKN4Kg67DNeOI8tZi9796sM17ZbDfCMX/2pt/ryD//BH/flv/pXfqUvH9oRHHe8yn9OS0ccwZ8/frzoy6UZYMRxdD5zAyPtWWC3t3K9sqmMphCnzhXaeUa1Z7uTbC6XvNgaT6YXnOClHqBd0oUumE1Ec6oGw6Rw7M8m7M8mHMxKHq8aPEJVGe4czHjlaJf7R7vszyqmpaWQXmcfjVVlHRZ9tbZ1HtLGR9oQc9qXTiXtgvpk9PdmxtGtUvrlSGwt+kVBac8I75yhHmThsD6m2MrSEE8q/HKCr6e0VUGoLGVdM580TCct1gp6Zom1RdV0jiqccdC6RO5RxaSnRcW65LnUQEorcw6yq5eH7AIaxuV0bhEd2HpF+30k1ysSEQJ+EhBNxi4/iYgqLgreWWAKsWK6f4fpzgkTZ2liAu06aznaNdw5KNjZ26U0AeOXmLZBUaIKxgsmd4aISYzDKonsKCYYryszXD4aLBmy6wwuBKIRvI3Y1tC2hk9wNMETY9t7j40RbKH4IERNjE5ptAo+SkZTZC8mIw9lvv/d67Mbf52Hchvj7uVgutshu109HXnSkyC7nVcW1kmVnsqyK5pj6jrIrgIphrcINSolKpamh+w2TPyKxn6FILsjiSH0RltVJcSI9yH1q0k5vBNs9Hohu53h5EuR3uAiiYRIhKIouX//Fd7//AzD+/Q30wh3797h4cOHvPbaa0yrBUY6Zl0ZXceg6iqkeNurXt8o7Yv3vm+byUz6z73+VK7sVf26y62H9FZu5VYulLSI2LBM0tk7s6skT7CbCta56eElV0S3SpoBEU1Mu/NJycOj/bwIE0xZ4YFqOuHtB0e8eveAh0d77E1LpoVZM9royBkytuxeRvrwk7yIadrAsvEsG0/dRmIc8pB+lTq5t97DBQvMizvp6cc+Y4NechkguxX2/9vHTFewd4x3FaJgFjWmWOJmx5Q7K+ZNYFrDiZSs/D6rdg/fCmYakSLAcUCCgwDRtz1kV0OEDrLb5KFuSWy6eeivQXbjSCHqYL2b5Y5ld1wW0AzZ7WJHIUN2fVJ2LBCNIYrF1A6sR51gVyXRRILx2NZSsYTyU372vqE6m7H6aI/3Fw1aCrt3Sn7l22/yzQf3+dZbcw5mnsokaKQ0BgmKakAJ6Z0nIJWAGEKbPZsCvhUIYEKG7EbBBqV1DtRBLFkWHs+Syeljlj6yCIbGA84gagltRYVgiQSTlG+URHYUkyLXQ3YzRPb/Z+/dYmxpsvyu34qIzNy7ruec+q79dX/dPZ6e6R48uEcaG1tmbDB+sMcjzANICCwhAUKCB5B5wOLJz0hIXB5ADDcZy1hICAEyEhIIPAgkLDw0np6Znp7unp7p+S7nfN+51mVfMiNi8RCRl71r7zq7qnadU6e7Vqmq1s4dGRmZGZd1+cdabQCc3v6zCMddjrjbHm/NgwvQ3HU8vQFtMYJun2qj27cqff2LEN9eEL8MZJeslrYRRWo3JmhA8VQ+kMwRI86KffS2Q3bXkNBkEINBdZwMjNnICG2sghHOjBAtgNlrbnHua1nhu9L0OIgRY9G0pccafv7DLxBfHPP5HswnnobATgF/4isP+Ie+ep+v7npGoQH1KUVLYVG1hM51mSLsyiUn/9Zwq0GZTuc8evKUx8eB0wmYKBg1GLXYDNRtwe/boIVgifQy001H3F0XCbhLx3MD17/K/t47hfSO7uiOXkoy/O2gue03rRe1XbzS4TdPXFimXvQTAWuEqnTcP9jFKzRRCNbhgbIqeffeHkf7OxyMS6rC4pYCU+jg79Xco8nL0KbdmTeeugk0PqJqkzeXQbqXbdCNGxEG7tUVntaLL3/xuVdqyptAmlKmSFAoI7isRrROApHcHyxd9GUrGNW83yom3KAnJc3suqTkYIPCMr5+GY67+GX+p4vHZBW/TCuFpMHl23Nbw5iQvWr5c/asimjKa6qC4Djc3+O9B4d87d0jxidTGFnuvbfHh++8w7sP7rO3W+HEpIBQmvx73WNou1T27LYoiU4Lax/RMJh1bk/6iWhQmgCnPlLHgI/pGp3K3gqiC3YpXbzxcx1aFjhd+m55Zjn3qmQDfu3hiwfYy8bpstJw7tMKw5IM8zNmbxZEjPouyNObRl0XkqzgicU51wUxCiGgA8TLtuiiqfHVPMU8btoAiKq8dXjAF997m6999QOefz6lBg7GBV/+4F3ef+s+lVVs7MeGIqkfkOq5VrtViDFSz2uOT06Y1oEmSpcyLRciN3V7tDRHvlJaP52kETrcX3VduuK93QqFdIhOKMshPLQ/Xg+i79IbXH4sSLTPpfRX/8pf7vg/80t/vOP/9M9/teN1EFnXDx6GoYcyX+UhNb4//3c/ftLxTgdRfS+f9uml9Doj7q679k1Ah7d1b6+SROiiQEoOVmRMHymwNZsrdHsb4xbntdtASSBWCmvYrUo+ePsB+3t7HB4e8ODoFK+KdY4P3zvi3t6Yo70xOy5F1m3TMgxEq1zn0BexGSmpL/oQmHvPZN4wmXvmNcTgMCRv1pLcfH1SOm/ETQiBZ8ul+wAAIABJREFUa4MbdYaN9cLn1iLutkrVGyDkxsaijwtKc4r/meOUz/OswPmIlBYZWZqTHeLcI/UuflQQRoKdBQp7RmEnGBPRKWgtaZ+pATHgTAGNI2qKJquSPHbGyWrIruu9eKaF714Bsis2KSXDKLs9ZFdaBw0igehSagghEqqAEHAx0JQQQ0Vs9tk7Ur4kFZWzfPXxMWZc8eBL7/LW/Xsc7O1QjcdIiGgIiNaoATCYYEANoiZF1q1Tm+wop7pCcE4JNgWpcTHx0Qk2eKJRglPsVJjODH8QHRPvibHGiEu+GRGKMhB9QQxCjNqpw16Th1AkwYLT1tjVkN2h57L3UKYePITvxhxzt4X+IptH2Y20AZOgDcvWwXF1AN/lfFAlls6N9OmDOgX2QsjuDJWCII7aRkysMbFh7E+obfVGLjJpfyh5/UzpxEajEWVZIALT6ZS6rvEttHdb81FnKV5BNyiXyFL9mrf8KIG33n4btQWucPC9HzFX4f79+/zcz32d944Ou2ckMUF9VdataHLpW1DS3t3JdMLjx084rQO1CmXeJtjm+k55SK9697eblvvWZkEGN6r4yv32Viikd3RHd3SLaeiIyp7RPnBR6yHNqUhizks6SAXz40JGDGVZcP9gj/E4sLe7w8H+DkEVMYa37u2zUxXsl45CYmcSalW5Lp/fIIjJpi7IVhmNaOchrX3yjoYoWYi2GLJXTNvF+xoexBXNukg5vA5dqJRCn7vtjjClR96pmTUj7HfvI+NpguwWozQu5wFXneB2T7EHc0bzQDEVTtVxOtvldLZDU4MbCVKATiImOIgQmgZrU57SBchuk0GV6yC7IkTfK0G6xHMRL2TIblJI1OdRIYCHqBFrNHlHxGG8oNagFmxjU55KA3ZWgAswPuX+WwWjvUOKe8KXZ19EbUGzV3G0M6YqLc7VuCZ5SKMGRIs0ZoxibMyNsoh1iLWEOu/9FMV7AQ/Ggy9T0CQblcY4FAehZFp61Ew5mJ2gUWnUEmOJE4elINRjnBqMKkH6/ZVOtDOAnofsts8u0TBqbs9r2koAWTm8PmR3qOAOobktZJeF8rIE2e2nHcPmkF1lFWTXEKiYFbuo8W/m2tJ5wE3ONZIMK64osK5gXnsmzYyJnzGWXUbb0rql+7PUnmxwH8Bx2/53FQO6ZsV34dwYu/5Q5E3HqsqOzHn/oOT+17/MVz98n4hQFAVH+2MqA2Y+xeBT/lJAu93VrfOldVKZ/LsZJYyI4kXxIjRS4En7uaWpoCmR6LAKKWI92ZxyNUv7unQxC+vdtV7zAHGxSRuvoTC+tCXXNG7cKaR3dEd3tBlJ/69d3xbWnayUxpgiTb6RAkNHAw2uU8bBilAVCWJlbIJChpiilu5VBZWzFEawnIeptXVe1yDdejCigmqOCopBxCLaQ44G/n46WNymC98b/e5+zKlN2B4VXHaj6eLbTu7GbDAy2V0elRgjMQbaHHxtZ9IBEmLVy184vKZvbLvLtN11wW6z0H8H6Jb2HlBEPcZZylHJHjuMxhVqDXVlUuRrC0YiEkHaja8MnpVKfhYymOiGn1maBBOktA0p3sYRiyo0OswJCoLN+YIztF7SGBZ9ycBc+fXyE1/GYSx9t+4FbfDiOmXl5UU3omXRfOUFVjyT1pP6xjqthljtgevNGJOQAIAPHh/8Vu9y3Su+wZh3Ky7WKqrpzizKqDCUboednV2UZPQpJO+BjiF50BcqGd6Jnj+0KZnWfCNdMM9kTOmNuu2xG3tAN1DtJl7O22zYvRUK6RANbugxoXYguR3PBpu7q9v7QDelYacYQpPfPhx1/F/8Yz/dl48DrOzg9kXsgB9Adq/Q2f2gXpXe4jSyBx1vh5aowYR6WVjrOkjsXWTd20irBNR2s38vMIYYU9TA2IdRfxM1m6FstLggpVxqhQjWJrhVVWbMogils1gRnLbQs1450NaKuSQBbPJ0dPBfcxWJTxYBEYsRhxWLIUN3WwF7cCcbCzid4WEwRgfK+UIwkbvh+sopNpb4WUFlTvDfeJGCBp0W2CZiSoNUluZ0hzDzcFbRjB1hR7AnDVZeYOUYTIOeRWgUvAGrqBWcKdDGERuDGyka6CG7AdoUgEPILjqE7OoVILuZFzJ8Nyl3MeR4KDlFjYgiJqV9EQ2IemIZEA2Y0OBHQvQJiuxNgYhhbwTROYxR9kYBXzdoEIqohCa111aGGBNm2YpBZwaNBjtuunzLtozZ42hwBYTkCMXGFJ03OsFFT5QExi/qAu9HfG53SEGEI9ZEBIcRx6iCWAshpPQw7SreQXZRvF4E2ZXOy9h7Lhe9mHABNFc2h+xCL+i2xy+G7Cpm4CftvbhsDNkVVYowS94rccwLg4keE2vG/pjGvJlRdiW7ukOMXaD1lGZEOpmoaTxN0/SBAl/zfab3tNk6PowrsXAs32yMsfvWmARUF1WwfY7VJOsurZFt4MRuRb06nFkkweaHVAoEI1hnsTbn834DZZfXTW3wpDvI7h3d0R1tnTT/7ZQh6OB6JnsM0yJLByOtfSDYlDtuhZ9uWPGtpmHLdXDMmBTP1lootTfQmKx0XhT1r12w5RLwn+Fzaz2jQZWgkdAqBtjkfTmnjN4A3fS7W9H0rd/Nq7jGDZEpPPJOw7wZY373EBnN0L0esmvmHlse48Yn2L2a0dTjJpF5tDydVDw7K2hmSlmBFIJMFaLpo+xaEGtQnyG7CNpkQcPQQ3OFBciu5ui4l4fs5pyo5L7uW6VJE6+a+NwACaDGokYw3hBFUbGYugQJqIsYdcmRbA1WHCIJuWFNlZS9eooURVKmfTI0ASl3qzM5anABWEQMoTYYk9rkRcALpoFQJm+qjdAYR9QC1ZLZqAE35b4/RRSCOqJWOGOxYvC1wwawGsHEdP+a42nkqWEBsgsLcxAsQ3N749lqKO9FMN1h+d75u6qe4fE+yu7wuHQKa/c+235Lb3zvjuetBe3+QNVWORdqNyJoBAJVCEnxp2LqdogmvDkDdkCxFdatSQaWrIQ7DdgYwMNk5pnUSjQVMHlpnbefDC38ImSDbEoVlQa9quJynl4RgxJXIiKW1NzcZ8hIkM2pBZg0MckqdcjvxVoqt4uVCkORRtcbIKdclvqxefsG0J1Cekd3dEcXky78A3oro5Eu8yUx7230IRKiJdo10LE3hgYr4vLCmAWLBUPrEEHUQvc6i+41SXuDQESTUhqTkJ1Qh60ouIVFZujWeNUk6z5usUGyir19i/NaEqCN+roy703fEcUm7wuk1BJNSEJYDAFtAwkxRAVk33u3r1G7vnyjT2hR2zp/PDeyQzoOf/N5otIZy1oYqBjBdEiNHB18YSKjG+ad76UzpOXfoWHtwofQnxOVvN+73U8/HKPbzTHZ0zbm1s0Hf3psMuC3Tdm4gS616LKJPm4PLSPA2ujtzgrWJE1pXnvmdYO+wfd5ntIbjHnCaTE7ibJi3k5p7R7WFf1QFpirz0gqySjjY2ReNwRN4bqscf0YXTcnvWk0XCK2fS9brvvVKqTtQrpE9WCQVmU/We8WVcd//OS4479++FbHW+1hqmEhVP3yqnPbqG9THPAyeD4S/cryQ97oIPqwbLDQLT0KMzh9GED38+PeMvfA9DBiif0Jg1MvTa8LmvsqIt2uu8YbAUdeM0YjSRjU7CUxJi2k1hisMeCVxgfmTWDWeHxhCDZvJM2CbbtPbTGjxG2f9ZeEggXZdPDNUBk9V0PrDVCWVtQ1Zyydv3SZbiH1gdpHmqAUbWAHMXlh7yX2K3ml9fVBc7ekwl9U+QWK++3fnxZrR3zkEmT3555BA5wU2CYgpSBVSZiMiTXQ7OP3RsSpxT1tMPE5xAN8qHFnAVHFREt0IFaxpkAbS/QGNyJDdiPWcR6yG3UJspvV2ahpKdJLQHYdaaoIeRkTIQbJCnX2yprsibUKJKhudMmDJtETRg2qARMDwTZJkVGDxuR9sYVF52lfmi0NcR4RVWwpxHmG5haK1opGMONICGmvrauSWhQxWKcEFwkjwTXpnqMFp54gKZNOMbXMpyV/qGNqjQiR0jSoGhSDKzwBIQbwXihUOw9imjUFn48t5AGll/2GkXW7IEFt+WE03ZaXJT739GVYbztznD++Cprbe2j73KerouxuAtmNee1MlszSz/BS4k3B3CbIrsSacfPijY2y60NIS4gI4gRrHWVZMh6PKcuKGKecnZ5xdnZGjJE38iZXkeY/CfBA2wPaAIlkuVK19dG/GqrrmufPnzOpA5IvnSIhD9qgb8KqsJ6Gkei36hVtvf2tbDeoW7v9wpejOw/pHd3RHV1A/VTcbn80khRRZw2lsxTWUcdAiEoTkqLko+b0LynxAICo9lsoW8lk8Pc2Qki6Zeiipm2gt7UgNum8NetNim2R1mK8TDH2uUh9jClYo6zwkF53Dd3WGrzRa30FBsSXKqNvhvhnCo9521M3Y8z3D6Gawc4xoRglKP28wRaRcnfC3n3PW1F51gifYXk6K3g6cdTzQLknybI0A4mSgh6FGmtKxCqxCYjm8FxNck2KyTDa7DjsIbv0kF05D9kd8qzkc52Q4bspFYqG9Ju8oklBMiooNglBIXvpJKVsUSJBJEUNNgE1NTACFEKNsYlXX+d9YqBNSHvKrCTIrrGIkRxTJQU4CbVgBIzRZLj1BuOFUCSPq4nQiCXiQAumRUO0U/bqE1RTlF1ilQx5aIIJB8HEiDMpwJJqq8ilWbeD7LIKsrscWXfpuGzAn6vnPGSXAS8LvHSKKZyPpvuyKLu9kS59k+CXZgGy27hRUqDxVCEQVIiUTN2IYPyiwey20TnDbtbCjUXz+wphDtpQFMJOWTByBUbhyfEJn5+cURtHQ9kp+k6bnDMXommD2aW6+3XqCts1ZOmMmzDaZ4+mtJni2ivmhquYtP988NXyftS0hobBeRcbdRfq0tb5kqN1W8nvwtCE5IwRhdAIMdr8TJpBE9c5u15ON6HMblLnued37r3ehMt0WPvl7/tOIb2jO7qji2lBGhKMSblIrTUU1uKsxcYkzIQY8SHl0ItZiEzU2t2HnxfpZqfHa9Dw/geHelrn5ZWNPy3XtKoJybqsGbLbw3Vjp+AvCyO3QGCTIfsaFcENlNGObsFju5AExCkaBRoDLnfQvpMkSL2DYuQYWUdlLEaS8FX7iPeBlGieHo6b3Juk8LNJG2qNKDktffJikoxL6x7lqvzqG43tttBwOA29aTooRzZh5b6vGYqbVCWli/w1xPHkfWrtVNTufV+AbKh0Y0nbKMVIjw7RDCDV9Ji6CLraz3Co5P3dAWLI++XatqX6E/S4BWW+oplvefo9d8nzs9rqee7qlx1eqRf0ZalQr2hofrAyOEelVWBvO/V32BpddZD2RGNK+VJYYVxWjIoCgLPZnLPZnCiGiEW7kFcNJve4qHbpoXaa3aW60qqiN/Fkz73aFaTD3EYMu0TfaTdZQ1deX7NxpKu0H9cp32i7NpjBHBBz3xO0NQssTEQb0gbFL+293KgJF9d3rVlHlj+saNAVpovXoJCusDeYHmq6N+6/+dI773T8ZydnHR94r+ML7aPvBu1vRwcL0TAS7W2kBaTt0NuyJhLtYscdLrgv717LfcQMDvjQ1zWrm463Oz10mji83uqIu6+S1kXrfZ3tGNIbAdNdIumWz4WDgHSpTwpnqArLuKoYVw0NHkXwUZc8pO1Yz8JXPqDL8eY7SfF2CRrdUnhRs7JgO7QIblMxTGtgficKmveOhoVoxiblTOtykLbPcoWE/4ppeaFdazm9qde/gTI6jCJ82ynWlviwYGSe4X/uGdIoclJgmogUBikNcbZLbCrUFHD0ADebsjd5yp47ozJn1L6mObFYK4g6tPAJsusL8NlTVRmi1wzZFTRIgrJ2kN2kGJP7oOQou1FbyO6aiLuwOspuy7skVEcviE2pIDTk/Z9GiHkwGlGikBRNQvaYJXis0iTeSBb8FWsM6jNfCLFJAqcpQGdAVKQi5UFVMFVMwqoqrkgzWMTgbIbsjsE2qWwUKAh4ICAUpwXNtOJTKhrNGTTNnBgNUS22jHiFGJQQDIX2kXI3h+wuwW6HMN3h8Za/ALLbqsXD/KUtTDeeK9NCdntobi/Q68LxdZDdzqMrikqg9bemQD9pDSp8jrJrCmZWkFgjcca4mVHbUZ7n3gxK99oajpLCIypYa6mqEQ+Ojjh4ckbkE84mZ0wmkz7K7k8Itc+ok5lUM/Jni9cQQSStlSFEfAh47zGqOGOoqiohPDSbtlZZ126AXjU67KqQ2paW27sM170qbby7XkSsiHxLRP5O/vxVEfl7IvJ9EflvRKS8civu6I7u6FokIn9VRH5LRH5TRP62iIyuO0YHU8yCbVIkRZp11lAVjlFRUDqH5r2NdUjw3QTPyjZOJStVFygjt5hk6XeZlmWjG5GVBjpTu5+rDwDR+kdloSzdtxfQ8s2tu8mXnbvq+EW0/M5fk/z1qpTRbY1RKQJy1FDvVfDDffisBK2JpRANxDpiignlzpSdwzlfrSIfFoZxUdGEXWbNCO8VSoOMTaovAh5iaMAExEViHdpkt0SfMKUiSWlMcrUm2G3MHsLQ8wkikY8nLQ3RDNP1Pa8tn/enCpn3ecaJef9oVoDVgwQgSFJYm8QTDOJBgiDBYHyC8EoQrFqMWghg1CBqiI2knL1q0cZAB9nNI0mE6C0SBaOSUsR4RYISoqBzgzkzyT4bwQTwwRCDA19QlzXi5hyGGUkfdzRaJWSJJMiuiYJVcKbPrdkKZC1kNz2C3nsytNmci4grA747ziK/pgwswmvbMqv5Yfk+tYsO7mGRH5QfGCn7JcHQ7SJVk+c3oXEVwQiKpwweE0Epmbl7NLa8Ecju1tbRnOe3/02HY/4sQGGhKiw7o5IHB/vc39th1+R7R/BizvkEf6LoBrzg6V1EQvB431A3DdOmxmfl19kCcm5vleQZfe3eeKEzxnXpgTZokq75gfUOlOtRX/9V6TLh3v4N4DuDz/8O8O+p6k8Dz4B/6VotuaM7uqMrkYh8APzrwC+q6h8FLPDPsoUxuozMaBUfI+BM3kPqLM6mqSRGJYQ+suTK6UlZ980bTsuewBuqu5Xq0MGavUoRvVy1l/puw/JrLb86+H8Z5fkNpW2OURGQQonOwNQhdVYZ2sCQMSImYFygHMGDynG/LBg7CzhitMSQ8iCqkaEWlIKKiCbPVU422iqGrUFJh7FWtHPed+9SVAa2pxY9sL58x8fBsVZbWSqT6pV8PP+PAioJAhsBTSiB5V+0NdlIvlbPS05fpS3kcQjn7e5HE2S3bWvIXkNt71FQNRAN0ShIoCDSQf7UZu/n4FmRvIFttx8qngvvfAO+o+HJHUZx2WLGwvGh93VVNRde7wJaV14WuKU7apU3MdmwFxFNoR9VhSDuRpSEm1xHOxp0aUOPNtoZj9gdVZQ5Mn2y5/x4zoWb0mXsopeh1kgQY4LVNzEMzC3DcIB5S8NlWnFZg+6m58mSwXmT+l4xXeRv2JQ2UkhF5IvAXwL+s/xZgD8H/Le5yN8A/qmrNqLbY6EJQ9z+fvmtg+73yfFp9/smUbvQLf9ep54ba9/gxzntfm003e+16r/kPawrv616tklDa+hrIgeMRcQBO8CnbGGMDgXJ9oOQIFnOCOPCMi4doyL5AXwI1N4nOGl8bc/ikiR0+SK6vBGX6yvS/unzTmxtTeha0wqtkqDTJl/KtDK0XmLBGta35uelNyGsLn+FxbM1dNwIdQ7QLayY16OtjNEOsvtihnzjGfLFGcYWuCbi8BRVgGYE832kPiTe/wLu6IijAu6VNTu2xtcN/nkgPFO0KVIgn8JgihKiRb0gowI1JkV+LU3mBVMkJbaF74pNz1YcYCGKYjKvokhbpuWLzLsBb5d4l4RBLEghRCPg0rXVAlYQR+YVsTFF37WKsRExKQCTNSAZxptyfCat1zqAiGpM14rJQ2vyrh9FkSLVGY1iS6AQgkuQXRkr8SDVKRaiEwobsM6jRUM1LdB6xBMzIopSiGds59kbaLGFgo1EAz5a2lyKrWicILj0PACSeck8C8db/d3IEk9bz0AZWji+CMcl8/bccRnAcVtPZwvU7flIP/4Xgy0NPbTtdJkUd8k+UiMpCjIoVTPFBkUomLmCYBTDnN3mMaWf3RRkd/vraJ6vM5uU0rzdIsYEGd0Zj9nZGVMWEEJD0zSEEN7ILT+3mUQSTLooivxeBGMMzoKRSN3UnaHoSvUPfy4hby6so7L6u0tUtnZN39YamwJQbb9vbqph/PvAv0WP1jgCnqt2OVc+Aj7Yctvu6I7uaANS1Y+Bfxf4EWkBfQH8OhuOURH5V0Tk74vI339yfLaqxOB/mtKMJOtuVThGzlJYQ/AhQWDmNY0PxGFKoltAQ+fLuW90ucDlJ9uXei6uSwPPUaf70hsIhkbWm7IuX0RXWuzkHHOj9LrEu22O0efTY8wDT71bwh/uoI8dGjzRKDEKfiaIzhE7Q8qGIz3jXqhRKTmeFjw7M0xmgViAjACryQsYhagBtRGcJsiuJuUueu36neaIZSnKrkLI+61aaC4swXeX+AzfTV7GntcW1hvJkXXzcd+f20KEu/KZJ8NzU/0mxWUKJMU6pl/1JD4I2ggSTYYO57EjkiP8ku9HcruFmKHGpoXs1oKZJAWdfC0fLTEYJFiaskFczaHWWISIo9Eyp66JKaWNphyp1qSH03rNICvEeUgMlbrFMu1xPQ/ZXVL8+vLnFcV2Q4ZAJ2R29dDXA336FwbldYnvIbvnowNvAtltPdONLYkGesiuoBTM7AHellufMra6jp5MWGlmkzaliEkplxScGHYLy54zjIDpdMbJ6YTTsyk+hFWX2jq1DqGXlmO98X1bzpehUXfb1DteDEYFK0LphMJCaQ1jV+EwWBWMKlYvByPtrnHN/ZkLsVEGgsn2jarLdb9eeumzFpFfAT5T1V+/ygVeLuze0R3d0XVIRO4Dfxn4KvAFYBf4C5uer6q/qqq/qKq/eHSwu9k1IU/mhtIZnBFC8NR1w3xe4324VR7S5ZasVEovUFk3pZtSBDX/7ULjt0Jh69Sl91CsatPWaU2lV4EVXd9qu/n7etVKenfdLY7RB4c7UEViYeC4gIlBNUNwNSttNIh4xEb2JLBLxKhhMhdOp8pkUhM0ohbE9JqPakyR7oyiIWNoBTRHKBMyZBfS8VaQbbULTcqjZkVRSOdqzEpe3gsKnOM7rSdqxyd4bJv2JdezPFTznq/22p3y24YQJv+P0p2jGeZLqxjnQaQZypvaJBl1kLYiEBVp98YHAd97IkVTvD9VgSAEk2DTI0LvXVTbBe1pryOavYTpip0c3iub3aOGJZGxVTC759T1lYWeM1A2F+en4fFhnSwdXx6fq8q/jF99P30r+kjE/bfR2BS0CsWq5kjGhmAqotitC9DbXkdX61VZGc1RodMn2HGOncJSGpjPak7PJpycnhFekUJ63XUv0SqT6OVn3MuedbmW55UynySAM4Iz6X9hHJYW8k/3+1po8BCG8+xaaOwlH5yiC+jUrbz+a9ImUXb/NPBPisgvkxJ6HQD/AXBPRFy2HH0R+HjVyar6q8CvAnzzpz5YecsqfTNM6KO7/uwXjjr+f/rWb3f8v/pL39yg2a+WFqLgDmahOIxKO6CbhI9ehobt9oPIuoXtbRVuEAZ4ODiHd7CuL28LcnLZ57Vp+dfVvi3Tnwd+qKqf57b8d6Rxu9EY3YRUkkUXSJZdhMIKu2XBpCqoJnNOpzNmc48Tw+zoEL87uv6dbYG6Nzx81wPI2mKh9vuFf6+VFGgjlmqMreSbInC2ERsleVvOqYNZElwWPPuKWR8c5CYXYrmcIvpS4fOir7em+F6LtjZGY+2ID0tG5jnxG09SMJ9JiTSKsYoZg84LNFikMczuf5H4InLff5djI0xngUcf7/LuyDGuDxgfjVHTgESsOjQYghrsSNB58ia6sUGDEKNgi+xNDWBcUvJiTDBdRXs+HxdHV6blu4i7Qh9lV0B9TGXa4wYQIYbs9TOa+fSdxiQxGtHMJ/RGzJlrjBGiz+Vt7wE1NuU+VRJMN4bULlPkcQaYIqAhRRxOUXZTBF1nlWAjcSTYHJE3ChTq8QoNUJ054qzisRkTeI4Tz46ZQUzQZ1do6ywmeIPLwnHIHhmT+VZh6XN/ruJbCG4aBHZFBN2e1434SA/ZjYPjeo4nl+kDL63mkyLfjsDlKLtmABxuW1w2UxopCKZg5gSJDSbO2W1e3FSU3RtfR5fdkFGV4D0Hhwcc3jtgPIJHx8fYh5/zgx/8gK/+9PvcPxhf+XKvihTQuEZ53nKk3PMX1+46Ly+a8nYH1Q4WDWmekChEjdeWCbcmU9JH+X3p+rfBerpN+PcqebeFG99olF1V/bdV9Yuq+hXSBu//TVX/eeB/B/7pXOxfAP6HK7fiju7ojq5DPwL+pIjs5P3d/wTw22x9jPbCTfKQpii7VeEorME3ntm8ZjKbMW88TUjiiOpQ+ZOl/7eAtP0z+NXBV7eYFoA9srRwtQLgy25i1fcbKHivim4aSiTnmBuhrY1RcREOPPW4JH66hz4vUQ05D7AQ6/zMTECKhnGY4MKciSl5OlUenTT8wYtTjusp81ijJkWYRExSZCyII0XWzXs9o+8hejEheXPE3cyT+D6ariYor8p5yG4u30N26SG4DCLudse15z2LfCTBcUP2fmZeogzKJ0+oBjqvqHrJHsqkaLf7r1M9ku8h5SSUfM8EMBFiFGgEmeU0KpEcfdcQo0GCoSk84mr2tc4pVix1dOl5tpDdKFgVrNHsJW2zHfYw2PYxLSp4qcxiKpghvHbRK8qKMhfxdPXQ8csw3Z5fLrMCptuWl0VvqdJ6ZnrIrqrpWuMzZBcCRQwYTZDd2h7gbbF+gFydbnwdVbHJuys2GxtSupGDUcX98YjDUUWMwmQeOD6d4UNKwHPj0XZvECZbtTpQAAAgAElEQVR7+yj1QtWYkR8GjYPAZZJQC0OnYefvfaldVFfym7XpPLXw3XMw6Eu+qlcFx32VUXaX6a8B/6aIfJ+0p/Q/v1ZL7uiO7uhKpKp/jxR04f8Fvk0a17/KlcbokmK29A3ZOt56H0rnKHOU3RDSHtLZLO0h9SF2uUhhaFU7r5ReDdxzVVq+0mU1shumFQ9ijQOX8+9Lz5XYiNa/9guaebU3tr3QCtuim2/NNseomAjjQFNY9NkIPbUpaiTJcxm8ZAhvAOcp4wwXG+bWcVIrTycNn56ccjKfUId5RqjmPXxCChRkErqnDSzU7QnNSmhn7BhAT9souQJdJN6Ojy1UtYXd9vDdFsrbRfFp6+qOt8+wv8awTLuntNtzGnP/ipL5ViHtYXjtdSU/M3JVwwjCHWRXkzeYyAJk1zTSZsVJkN28D1fUEG3A2MBYQ/ZkCl4tOQxwV86Q0sB071b6R9DrB9r9XVDm6HvuslOi3VHeljmvnF7My+CMTc6VwedWYU6KriwdXy6fW5efo2YeEmQ3GUoiLkaMCuBozDgrddudo7e7jq6oHzrDj2bUWRtpd68sOKhKDqoSxTD3kdOzGT4sq0Q3QzdZ++2a61vqk6CoSgIdRej3Aywuhx189+XV9vtrL9E9dXDeJgEypf2zwcNd2ZYbeCnbUHo3gez2F1T9u8DfzfzvAX/iMufnJS/XNagX2/HDQChfff/djv/DX/uNjn9W9yd/oejPHerXw6OvIrTKOsjmrYDmXtCEIeqlLPruoL5/arPxZHD8rb7a4Tvc0m1eazP4cCP4Ggj1Va69CQRh3fXWtWnbpKp/HfjrS4cvPUYXK23/yOKxfBvWCOOqZKeqGBUFIQTmtcdYy7RumPtIHROEzgyV0S4K7ToxZbvUCVTL0ZrXHF8u80ppzQWT1TZFBDRtYIa8iEWNBBMIBCzJCGAHlQlZaJZLrZFr27cddfK8CLvZWWvG6EUJzPMlWk/Qqut1gvMNJkLf1hgNc0v8Q8uOeYb/2ZAUrTOLNIoYwZSGeNagYvHliNnhIdw74Z1Rw2d+zlTPePLojKfP9jk+dLwbv4CReYpKGwo0Kp6IrQq0Tt5ROzZoSEqqq5K3MQYwZRLklmG6puj5dkdODAN+ANmNHWQ31dvzS8czllVDHg+GDOvVjscoZhUvijEDyK5recW4pCSrDiG7shqyKz1kN4wEF9LaFw0UJMhuiIbxvELrHZ6YHYI+x9JQ2TN8CDRGGRUheb4C+GDTHknWQXalg762nsskiwrLkN2+fH/upjDdIa8dnzyuPb8OstuPodXlWfDoDqPsqgSkKxVo5biymeKlxJuSWWGQOE2QXf+C2pQ3EmX3RtbRNdSuPyF4xIypqoqjoyPMo5pZXfPo0SOmky8Swi7W2pdV91opKUhL7yN783TNtrWtXfuKMqO1lhgjZ2dnNE2gML0B5fbToKWS94K2C9dL1mhpDZCDta59hpvKqOfKDa8pV5d1L6WQ3tEd3dFPGi16MzVPfmTlqCwLqrKgLByimpXShmnjmTaeuQ9YmwRDO4CatIpBC+a9OVW9v54ufV51fPmc100LbRPBtEqpMVhjcrJsiIT8Ezth8Ta0/+V0fWV0oaptWIJv+YMzLiKHkcaP4PMDsA1qp2BcCgITFS2SUC/MGe14dg8Mh/cPcNNjfN3wwnuens14fjYjEMAYDJYoiUeFGCLGJqEy+ghqc8TZrE4seEulD9QDg2i1JB2jFVgiQA/B7cq0vJI2ZC6Xad9rWMVnhRVoobmStwm0XlFty7Tez9xucpm2HYlvYbrCELIrkvdzSvojQYiGLv9pEAvqMOrQMZSjwH0b+JERGoQ6GqIElEDIcN2U6kS7dCDnI+hexJ+H7K4qw4a8nDvevp/FrRrrILusKG+WyqzyqDKApKoKNn8ZbEHqRgEXYzJw4GjMLuE62L5XQZqcL0KbvicL/2K65xe0ddAIaMRZy8H+HoWzTIPn+YunPD6b8HTm2dsfYWONzXUKbc7cZJZo4zvcTCacZVpnuM/3OSh10+v6uTYMPwzt6NI7wRK6y1AWBbvjinv7I46fH9Oo54xT1KWtBmM13bx1E5DXm0jheOkyK5RXyZPRJjrltu/htg/rO7qjO3rF1EFVuvCtkhOR59/WW2ctZVFQlSVlUSTBLcYE260bJrVn6gM+pgAe7eR33k/JhV7KbdHgDtYeX1fmVdHa6+ZcdiKCzcqotQZrBESJBCKeSFiwlP440CaRe4fW2cHBq9GrlqIuSeIicqjUO2P0ySGcjJMHShxRLFGFWFi0AGMaRjuRvQPL4YNDCjciquM4Bp5NkkLq1Se4rzjUkFyPxhJVUSOoE0JYtYdUiCELxuT9l8twXJW0T6tNB9PBdFteV8J3W54VPEiG47Y8HXS446FPCYMs8O05PWSXTpJfgOwG0+0tbfeQSkzQXLxgatNFqJQIMRpULQYHI6WolHsu4CRhwxo1BMnjNO8hNaQ9pJ1qJr2Brofv6jmlkCV+vaLaz2cv41eVZ2N+UQllcJzueM93aI1BJOR2H2n72ZsiQcgl4GLIRgZLY3cJ1m0dsrttEo0pmFZ2v0u7ZzH/Rk25SNvdioUzHB7sUxYWjZ7nx894cjLhybRmbkqiOASDMRaRhKtP6UsMFsFuME9e+56gj+w+VPgGf7tj2v55xbRmAe+7WjLsVmXJzrji/uEYYwxePad6ytzOaKzPo86uruwVUYqEezF8t5XJLvp64f+Kry7zzWWKXJZeqYdUVfCaLmns4NKDZ+2lhyZ87Y98peObs+OO/z+/3wc5+2d+/u2Ot6GvSG8DVPaW0MK8vfRYhmCKUVl1/Pvv9NDch48f9cftl/uq/PYnm8u6+q8Llb6sVemyUOA3LrG1psASZFhou78MekhYxKQE9IVjZ3fMPMLuzhnOCsTAbB55djZl73TCk7MZbtdhjWXksqiiS93wbqz2pOsfh4jBGHDOUZUF41FFWc7xUQi+oY5zRB1RYi/sbVu7Ui4lCL5KCNQC7PaS1++U+DdguMa6QB+O2JUXNN94gQQwZ3uIFhgjmErQiScaQxgVlPUOu7v7vPNTnqMnL5gw4eyk4cWLpzx9WjJ5MWdvd4QtDBILoigqEVuO0CYQomJ3KrSJhBCwI4tGCDFgKpJwHSOmBBRiiEgbfTcoUuTjl4Lv9rws8DKA6co5yK4YBvDdNTwgrueNS4oxSIbvroDslnQQVGeVaCNhFBJkVyFapcATRGiw7MRdLIeclEeIeUEhDVamNKGhNp7dsiZIkVLEeItFQRSv0vkL26i0Q8gurIHs0kfWXXf8Ir5XJi+C7y7y8QJ+ZZRdMrRYU3TZtLQkbV5aJbqTrZWqmXRRdqeFQWKDxFmKsntDkN3tUVIiDGAGcFuRIm+z0D6ysAhGDPv7+3ztZ36GvW/9gE+PT3n48CEPHz3kvXsHvP/+F9JWDTPIeHCL1k3hvMx9e1rXk2qaqyLgioLd3V2Ojo4Q+wm+DkwmE3b3dsEohSvSFkK9Gb1aVXuobOtBf0l5Vng04eJ1TvpB9VK6qE+dz416QTuvSHeQ3Tu6ozvqKMTIybROE5+YrB319nNBEQ3dxO4VoghibFosRQg+cjab82Iy5fnZlP1yzMgJKjYro4PpTBbn4aG1/ieWlh5CJ2aKYkSwYnDWUjhH4SzWkgJIZdBuG65BRTtj1A2ophvRMLDJbaVXFYFwaySB2nnmoaJ4WKAOoouY2qFOwCl4C0aQqFiJVM5yMLrH23bEqRQ8io751DCbGEIjnXcPaXonovrkLJWEfBBRsEKIcRCVtleCeu+iLAQ4Ygjf7WC60gUx6o7nHJ2tJtYfX8ErXaAkciTf7niO1ksbAXeBz9R6bFnykA4guzqA7GrXblIQpCiYmCC7tB5SMahYDBYtlLKMPCgChRFmYlIUXgJBPT4qVjOk1aTzRSV5SJcADtq16Pz8+DJ+lfdzFc9KXjfm5dzx9FfOlVlu63B96dRSQAjG5W4UsDHkfXKW2owJZni3t48Ug5cqGW+NAW1zPvbvMZheIUWTh/T+wT7vHR3y9PSMpy8m/P7DJ+wcfM67X4vsGChN7h+qCzBdfwsexy1oQj8HsNiX28/dsag4MVSuYG9njDMGjYHGT4mxJkiDp+luSmRBatkavc718eLrDu51hRK8NpbDNZ/RnUJ6R3d0Rx2FqDw9neagIbYPM54piQx5A5YmmJpXUp5Sk369ek6mc0anE56eTjjacexV7e6XHFZdlT7jXCId/L8Ni9trpeV5PZmgEUnBpJwxlM5SOIO1yaMVNWTobuylv04K3KJKeplqXuOLvNSl3ySdVDwzN2fqx9z7/RI5iPCWx80tJihqFakFsYo0ASM1lVUOykPetmOOpcLqnHpWMJ04fKMEb4mFS55NkmARY4O1FsUQ64CRFFArNp4EEjTJGyqSvJW+9/Zo8CQjinTKnDEZ+kqS0dtYJ6bXTVP9yb3fedN0JZ8+i5FO2F/k6TwbKaBpGxgp319MvECXsxSkC/6VvLKmm/+CT+0UgRBBGsF4gx8pJoIJ4I1FM3hSy0hZBd4uPKU1IBavhigJVh9y+hqjpD/5/q2kYEbpPumilA+fUTuSIykcZNLNFSur+VZ3W+DRDlobc6vPH2eBb1XHzXhdfbx9t3TbiQHTbQlpU4SJkNK+aINqoAgeD3gcta1QU99yyK5QU2bDjcl7GCPSBfgRopgE6QVEY4bs7vGldx7w9PiMzx+f8b2PPyeM9vjps4a39iK7hqyMBmxWSBsxxOw57frUTygNI+EqnZ0pfW6N39kQZBGqomBvd4fCGEQbGj8lxDnR1DQ03fh38YaCSun2DKJDj+u6Y6sC+60qs2mTto0AfKUK6WcvTvgP/8dfA2C00yf7Hbm+Gcb6jv/SO1/s+Pc//ErH/xd/53/p+L/wzX+x4/f9acc3A+jvsr/9NkEdboyGtzg0diz1n+GAHXaGD+7vdvxHj6cdH0z/fgq5vfaM1wmVfeNgugOazGt+4/c/7vaIGpOEMjtcPCUJodYaSldRN4GTJoIrkaIizgOfvThh6gPOGe7tOMajggdisERMHKiiKgNkaStxnY+/+xNN+fm0Qp0zQukM49IxKiwzF5hKSvOhGlCJaBvrUuXWP8itW5437DxvEkx3SNNJ5JNv/T5VfcaLf/inGM/m7P7OKfX4fUrm7MZnzN3buLJgZ2aZW0szD4yaRzy7t8uL2ZTDJx9xPC359EXFo49+j5H5I+wW97G7B+gM1EcMJ2hj8BhsaYg+En3EVqMEewsBW1k0KCF4zMhAhBh9imKrgi7Ad+MamC4XwnfTXtUWjtvyL4PssgDfFWEBpislKyC7YGwP2ZUioDFBdm3ZOVlxTolOCRpwEVSUYJVCPAHBi6FqSozf5dQ+wMrnlKZB7JwYAtF4KGb4KMRoIUiXFaAZQHaDthF0W4hrqzSm7t1DebMinzvyMq+reHreDgyD6yG7vRJ8Ho67ng/neBbuIWkQKcquIaM6sppdNhMacQRxTAvBRI+JM3b9Mc0th+w23vPw04eEGCiqEucc1jkIHiEZcCzzhDoQYafcwxjH3t4eX//G15nIiG9/7xM++uhjGt/wO7/zO3z4tXsc3R9hLvIO61D+kDd2N0wSB/qJ+aZkdmst49GI3UqYlMIsG6G6TAAit7qfLVML600f2n+DBa7VNWVx7Vt13kuudCMw5turUdzRHd3RKycflcdn86xwth5SOoVRcgQPk78fFZ4mKC8mM6Y+Uof0G6YzgipPj8c8P5tyb3fEWT2mFMVly/4CWmvI3mmjq0myBVwEK0JhDYU1OBNBNCuhbZTdN0zL2ibJOWbD8m8GNd7zvU+fUE/OUHYw8ynm+AXz4gyjc8pwDOUxZeE42DHYcoz3c85OH/LDT4755Okxn08DVZzT2DP+wUdPGe19gao8oNwbJxddCIgT1GbPlQQwiriYFA5Jyl+MMQ1jAxra3YPSKXhJs2wHtHTCTgeta+WgddDcczDdIT+A7MZ1ZTK8bAjfRfsyLEJ2cxSdjpfsVumiBgvEHFDJDIxpogmym65gwCjORQ5cguwaMbkJEa8BH5RKM6R+YHBKkN30YDr5UHuI3PLUuA1+EaWi53gZlBn6UVbz6663rky/COhSqWhsFp4jTmP2eFmCjDqP4G2lk+mM//Xbv0uMkTIrpK4osINnKsZ3CuP93ftUZcVoNOLhacOxVxqB2geeT2d89PljXnx4wFSqtFUjzhCa7C0lBe2CJefL9hfSzm78Su2cF19pYaUTljv0iiqSVmYECmvYKwsqERyKmjZAYESsZOW+Ncm8ZpJ+HriQlpf+VaLApsc2vMS26E4hvaM7uqOOQlSeTJrsIU1SmKoSfUA1ojEpPcYI1ljGZU2Iytms5mTWMGkC0yYQas+0btgZlTw+PmN/Z8TR/g47VigNVNZ0kDnTabxtrMY7jfQiMjnSbmkNhRWsFUQi6adVTNuf276Dc0DbWOVk4d+PJTW+4Tc+/ZyPns0ZfU+YhTPOwjNmzcf42ODjlN1ij3FpuLejPBgfEgk89c/4g88mvJjMeTqLlNMZT2uL++Fj3rl/wsH4Pgf3HdoIoh6xKRplFCGqx7ikNfl5xIrBWiH4BOW1xhLmDUYMYhzRJ6+kERkoc+2eUzkH3+3ygKbtdkkEvC5kl95TCrKCz4pQzpUMGZqbz41+ANkNKbevCAQEaQwSDL5KEEkToDYWcBi1xCJQlp63i5rKGowYQoRAJGikbmBXwIqmNDK5xzpRgigxsgjZpduW2/MXwHRfCtk9B9/tc5yugummrKk9L7ncJnwfPKkPothBkDV9M4Ts5s6yBNkNBAxeSua2AjO/1ZDdJycT/uv/61sAlGVWSJ2jNMnXrKrgeq3uC/feYm9nl8PDQz55dswnLybUBpoIMm/44aef8fn0K3zAmLEbYXzABp/gljH1IzgPU70Zuvn1OQ2JS1wjF9VB00QHaZR08bkIEZEkhxyOSkYiOCBKwOOJ4jFGCCG8ige6Ea1fyW/GWzmoPkkSr8Dd/koVUm9KHu9/AEBd193x3/jWdzp+Mp93vDS/3fEfP/qsL1P3ENL/5H/+vzv+L/7Cz3T8lx4cdPyY/lo/KTScq4d99VyXGg7SQYd7a7eHVM/Ds44PWUkBcDqARb+mMbsu6u26yLjXpSvlenqDKIphWuwAaRIPIdA0DacnM5qmoW4aYuxzIziT9pHN64bHz4+ZzOecTWcEjRTOYoxwdLiLjwnm+2CnYm9UcjguKV1SqAqVvHDoj7UicR3KcS86Id8ZQ2kthbU4k/b0pjQCoUsl0Fl3OyHvNTRc6QTHdQtql+dt2bp9Wepuc7MbfYNU9QWaBcfT6ZiCM8zhC5pZw4sTiOEZdQhMQ2QaJlhveRRH7E5fgBHmruBYZ8ykRghEPWNaz/jRx1P+nyPPVH+Pe/YrlPvv4Hb2MK4gNBUEixsVxEYJTcRYj2oSlm2Gvvqg2FGJqhJ8isRLhBAippQM2Q2YvEc1+mX4rmbvIwmm2x2nj7J7IWRXzkN2hQ6muzFk1/ROJmNDDvYk2KJz6KYouy4SVXAZIpkguw0RwWMonxeI3+HUPMCah5RGKazP+GSPK2d4HKiktB15m0IdDSanCwmd17SH7LZKo5DWmStDdlfAd7XjOXe8hfW2yuswsi4v4YdRdlvbY2yVBQGVyBCyS4bsVn5GgyOYknkRkDjDxJq95oTa3m7IboiBk5NTOuWT1K8mkzqlOiIFImr71HfNDymMUBrDSVDqEGgC1ECczvjud7/L9z+8zz3r+eIHH1BoTaER5xwqJr2bmN+6wM1qKG8wSWv0TguqKxx7e7sYY1KQxui7NCve+2QAuw3eUYaIheV+P4Rmb1c5vby8mw2CV5SB7zykd3RHd9SRcY7de0dpoYwR7wO2bqjVQd0Q65oQU5TdbuYLAcFR7kSiKwmmoPY1RgSvcDyteXIy5eGzE+q6YX9UUDdjxmVBVTh2Skdh0t7IwtAJXnd+0p4WjEqShFRrBGskezpyNrv2zzJEacHytIVGXER3L+1GqawcH/7slwjzI4rxiEk9473JKS8mT6jnNfWkZkZK/l45QygUa4S3jKUYR85OZ5x9PqVWJYpwMp/zg0dPaWJgVBveOoocHtznnXd2KShwpmKn8Cm6LIoY3w19zdhXMQmOmqCIkqC8CohJXpwclbdzfyJdf8rdt4O+noPpDrtzB9Md8kmT7Y+zxOfrroDspnmOvm9rz6v2+8cWouyadE2J2daTy6lJId8SrxijjE3EtYGeOkVOU2AkGQIBW0VRF58LPXakpeHw2oQffn5Z+UW0ow748/VsMh3I4Gzp2Xz+YisHpYAU9Kf1+1ltBW1DMNWtT+u3u7vHH/9T/2j22icHzHQ24+HDx9R1QwyBJsZsOBBsDBiNqR/PamgaYu1RjfgQeX5yym/+/kcEjTyu4f2x4X5l2d3dZeSU0hqMKbCa67nFpN2fNbTcGeV8fz5XfEV9w6xnXUAjUr8byhdWwJre8KOktDzaWoA3XdBWDaJt0kYC0XWtun01Fz9z4VxAmu7Eq1//TiG9ozu6o45cUfL2Bx8m+S1GvPfM65pi94R53TCva0LIXrgY0RCIMXlRi/0Z87pmZzLhbHJGDB5xwvHcY16cIcDznYr9UcHR/g57o4rdUcm93TG7ZVJMTWFTNEulg5Rtg5anyJdVvWaqvbHrXUhLa4xAhkybrJAOd5it2z3aW1evvL/0ZQvuS29yiyaG5Wp+ghwCuwclf/LP/zHqeofKHiDxBYTP+P7kU5rjU/j8mB/GMSYEHoQJD3dHVAY+bOZ892TMk2dnPPFPeDwLzJrAaT3nN3/0nO99OuXb35vx9fc8Xz4645tfe5u3xhX3dnYox4IRixED5izLaSn/rRHBOkNoAgaDtY44rxFS4DP1MXkZjU2ReCHDdwF0AN/V8/DdlAYwefSG/JYgu0JyWraBYlZCdtvjWUjzCOIFEwzNSDAhQ3adRTT9RhdxZcP9Yk5hBRFDjJZok+ez9hZjDU4gCHlPqmJNJGRPcQfZzXNhB2WmH+m9F1M77+Zy9N1FmG72UNNH1u3huJL5JU/sAi+EXKchB15qeVbBdGUlZLeNJkxWMoeQ3dy1aGxFVA8aKEPAY/CUzF1FlNm1BN+bpgdvvc1f+Zf/tRTMyFqOj495/Pgx3/6t3+b09IzpdLqAEmyaOd576rrm2ZNnHB8fc/bJJ+DnhOh5fnrGr337d/mtjx7xs0+mfO39I95/cMA775S8O3Y8GMHe3g7jMGcU5he07BbQZTVShYvgu8PIunF4Oj04SJeP6xAJECkMWJPGUYgxK6TJkCS0e0lZv3xJbzS60fgNmup/NSi8l11j+214pQrph1/5Cv/Rf/pfAouIAh/7D03TdPx8Puv4Rw8fdfzf+lt/s+P/5n/1Nzr+P/7b/33H/6U/8yc7/q/9c7+80I63XA87LQcNCWYYmfdyVqZ10NFtld/auRd8HibhKIredmsHdlwxblB+cO6WBsg6eO1ln+l1aR3k9zrtexNof3+fX/rH/vE08WmanEOIzOc1PgS8D8lDGnOC6RiStbdpmE6nTGcznj97zqPPHnJyfMyzp4+Zn53w2cmMR0+eURIpLNzbGfHgYI97+7t8+M4R790/4K3DPd4+GFMgWOnt52agxyxa1Td7312pQfTBi4LP6gKTLy6bq1O6+IfrRA1eeYfSVZuEQSNZcM4/ObhKwvkOTsiuputCVXUpQl/bpsvUu3LRvqQHVgb3tRDef6miNxWau46q4pA/+o0/R7UnmMN9jAaMn/ELM0v0Dd7P8H5E0AYfz4h1kdIFFYE/ddJwPJ3y8PknfOf/+DYf/+gjvvP4u8yfn+LnDZ+cPubFx8/57ceO73405mvvv8OX337APyJvs3vwPtX+24z2Dwi1EmYRayR5cSIY61AFH8BWJRqV4BVTWkQh+IgpkuIRQ8TkyGYxaIbvynn4rmn5DMft+HRcQ+zKa7wEZLegh+kW2RuikuC77bC1EaISVbCuc7gm3ikBxWmCmQanFIQOsusagzYVL3QPweBEsTYkjwyRqmiIovjsXU1DVWiiRVQxohmmSwfXB8nKqA54lvgEr10H2aUrY7rjLRxXWYTmroug2+75j12d62C6LPH97J3gyGk+Ugm0qJg0j6VWlH6KF0eQgpkzGbI7Z6c5oTYFt2Vv3yra39/nz/7ZX4KszLRr5S//yq/gfaBpPE3e/hJjpK5neJ+OvXj2gufPn/ODH/yA/+8f/Do/+tEf8J3vfIeT0ymT0ykPP3vGb1aWg9JxdLTLNz54i596/21+4Zvf5N3dilH1khQlea3Q1+RJbfv62u+5WftijyISjDE45zBGzkUvtm2WARXwWcaTl8t6q1KvLBZYXJOuIreuVXovW9WKdXux7v7TUOFepXx39Sg5Wvblx+edh/SO7uiOOiqKgvffT/u8270UMSqND0RNymmMEdXYLaatQjqbzZjOZjx79ozx7pjnz55RlgXPnhScnRxz1jTU9QyZB+omMA8wqQNFUaYUM9YyKh0jZ6hsCo3feiWW05cM0DxXoCTQvVxJ1MF/WThy9Wtvh1pldDFHbP8XkVvrQLiWMpppcbHLIszaTrFFz+wtIOcs9x48oNwxFPsjBEVCxARLjEqIARMdQT3zOMPUyaAYbOT/Z+9NgyRJsvu+33OPiMyss6v6nOmZnmvn2MUudrHchaAVRRMI6KIIgNSSkEgKoNFogiSKRlL6Qpq+kOAX0qRvMJjJBJnJQJooGgGTKJLgYSQIihRIAMIQ2AHAnd2d3dm5uqenz6rqOjIzwv3pg7vHkZVZXWd312y9tux6Genh4XF4uL/3/v5/K+c82+WYC8MVFjYNz16+wOKtPrfefZ/1u/e5fe8ew3LEaLwDW9uMvOfezjbWVly9PODyhUWuLC6BM8GqMOUvYV8AACAASURBVEPAxE7qo6NH8BKMQExrwiVhcp4ibZ27khw4yTJKSwLirW2X6+4z5e8syO6MfWsizbR9Qhe6DvR6e3rkIhxZ40cAJDARt8lmmqMLqqZ2IDU9N7IXtxxn3YNO1jNdpj3tD5lCT9Q/4WyeaMPe0jUnptezW9IKWW191xZkN10TxaCSPfGQXWstCwuLoIqPz5AYYfkctZPXR6SR956qLPHeUTnH9uYWW1tbXL58icWled5//z1WV1e5deNDNtbXuX//PneGQ9bw3B9uUY5L7m2OYfkpXn1qBb2wxKA/IMOReUe0ozDth1gkGoWTzrt9isa3eEqPkjZPqafd/x8mD3MST/u9cwb7OIF6JK/bZcIrTAhkbWJiEEZq47UxtvQAFylWL902PtRQ22/9Rx3f9+FEbvvV9yraqecIXfPMID2TMzmTWvKi4Kmnr7a2aA0VSyyXwSDVGCV1DbR3NGI0GrG2tsa5c+e4f/8ei4uLfLiwyNr9e9yyGZsb64x2thkOR2y7bdaHJQ5DGdPFiMDSoGCxl0MvJzcSBwqYZhQe2jCNs7Z9mypTDnQ0o3j/h+xumYhOtibNSaQeHqSlz6rzMclhGrIrMvqQup/sOeuhJStg+UIP5+ZYjEBJZy293g6qFufmyMwY9QZXZciiQnQmmWXB6xyj8Tk+feE8O8MHXL/7Wf71r/wK33rrLX7tK7/FrVubPHjgeHs05N2Ne+Tvb/OV90d83wsLfOHFcyxdXKCQHOMyyDcQa8AYvAYor7E2kBYhZFlk3yWkiWrDd32C75oI39UmuomCsWF7gO/KbviuRodMdJxJxITWkF1t6Qha0jDrujQ/FnwlNUy3A9ktGyZwVwXIqREoAdOB7AqmEsa5ATWIt1TWYbIxS2Ynrq81eGdQAx6hdDmZNeQiVEJYD6tKZj0uttuIRJZdxbZ1BE+M1tKK3Lb0Gr7bZtylgekGciKm6NQOg91Q3gam24bsBl3qSKqJ7ZNUfkKHNhw5pskRCSuQI5RSBMqsh/cVUNFzLuR4JWfHLqBm+ESz7NY3AqmvrUZHSWYMmTGQz5h+X7wIwKc//V18/vPfw9raGm+//Tb//J//c7761a/y+uuvc//+XbZ3trl1v+Sd9bv0P9jia+NFfu9nX+JLxRxPLy6x5HZYcJuIGHIN0XwILNHhPiQjqxUpfVj0L5bx0WFUm6O1Rdo1PsO8YT9ulIccv17POb1RB3oSRDHia9I/JzY40oxQGEvPZGTG4gnpq4xCHp1aKvuIaO4y9HaPWHvVsZ/I4nHAgqcdp4MGnDhGylXaXc++u46jIJIem0HafrbyTFp6Uetzg0ZfOdew5v7kT/5krf+5P/ff1fpP/dRP1fpP//RP1/q7NxuGXoC/+Cf+UK1/5uJ880O7Yx4QFjsL4nkUqOmsffdV50NL7BZDc/7jcaOrbyDO7WO3ucfcPmAHTzqs9aDtOwrU+jRJOs0EtTGRqEQtqBpUPVmekRcZg0Gffr/HwsI8Ozs7XLt2jY31dTY3N7n54Yfc/PAG9+7e5cMbH/BgfZ217S12PrzD/c0tbtxd4/5GgO9eWVnm6ZVFBrlhYGPElAk/futy7gXBrffruCpTeEP2SUK7+5hpIJ5q1LYM6OORBuQkrci195GAwSfSD62PLRPXpLFhdz+HTzysdb/G6GHLnyJxail9xiAbMSwKrHgKP2IsfYx6MjOi1B4iDmt2qLQPKNaO8C7Q1vZ7noxz9HuLLKyscmn1ee6urfOpb/0Wb/zLb/D219/jt771G7jxCHXb3Fgf8avXd7hRvoUUL3Ht0hWeuniF+cU5VC1+bLFZSAtVejAxMWLpwOZ5zb5r8gAndGWC7xLguxGyqw7EAjLBsrsLshu7c6SiFYm5T41Go7alR/KjNkw36ADNulWIjLsRviuZjxGuAOVFwzrJLAPNwKkPuQvF47KUK1Nw1pCXBjfOWavm8E4x6hBbgXoERy8b42SeCgn5TuN7pHIGiRN9p601mfWkPsFxhe76zN26MMG4y24G3Vl6m003rU+FWZDdLqxXmWTcDduSYQo0UcOYRDPFlxFFQpZqinIYIbsZo9wgfgfxYwbVA0qTPdEsu8Cu1/9hhoTFxUXm5gZcvHiB1157jXv37vH666/zq7/2q3zj69/gzTffZGtri+EwMPFm69e5881lvvSlL/HCyjzPLfcZDAYPP9ABRaAVpW4zu+4xEB+6/mbbUSUQrMW+FBeae+dxLjidgu7wJqROS0thPlYyJTK6HwP3YfPd40gydxYhPZMzOZOHSvdd1Eyggl1nMSYweXobSByyLGN+fp75+XnOnTvHzvYOc3Nz9Ho95uYXKF3wuVcKo+1NNoYlyhZzvRyRsK5jvlfgezmmlwXynilG1n5jnI1RCt13775jpFNk+r67DnGII+zLdREjJmhKzrDXXsmgng45Og0ZSx9N606UkuLYxAMVAIJRYubZVkQiOo9Uk+VG/bgKYLwGwhUyellOcWmRxeXzuB7YrT6r80vo3APufHSL9ftrbG5tc3dzHSdDvvKtggc7yk6Z8dL80+QmR1wONpKpCCAuOImUJpIl6TmLeutRrHvS7lACPPTZpvv7NAgFTdMeWs20/dvVJ19WsFtrvbtPgiZrexMqDRv27ndZ+xizznc/T6e2TE9p/X9UOWibwpuw25JZ5abo0joLba7jqUgOJnsEGw/Q/CyzZJllMOizsLDI+fPng1ElwvnV81hruX79Onfu3GFra4vrt0Yw3mLx8vuUT53HuBWeemqROaOY+MboOHWP8LZ72Ont6hfHVP9xiaSx0zu8i85dQvq0tLTAiHQCL6fgyTu8HPPAd5iR9MwgPZMzOZOOTBqfu37bw9WrqvR6BXNzcyEq4nzNHriyusrS8jLnb90mK3r05xfof3SL2zdvMBrvMNoaobrGsPKMSo9BOL8wh1+cwxpDbmOqk5TM7qDnlfY63O4HPtbD5HBNSFMy7UxoQ95RHz8h0pB+mdqWKRbzk2+UHhDBcIgjnAZj1ERayY0q49wIvAg7CEVeoSqULsNkLkT0qiJEIlXR0mCLEPXTsScfRKjtKGO+MMz3Cxb6n+KVC5fY/N2f5EtvP8v/+8u/zld+601+56tv82DLsba+w8999B7PX6x49SnPjw3OcWGpz8JggBbbSOyjzu8Ech5j8aVD1JDlOX48AoUsK9CqarHvxujbJHw3RS6N1CiAGr6rBAiub+W+U23YZDVGRBL7bmTWBTrwXV8mll2JbLqhY/nS1CmWqipAYi1CqYFV105AdsvcIGoxzlJZj83GLNttMIrD4Jzg49LbYZWxmkEuSmUEqQBVcuuoHOG6hFMLOTujrnSJhxqYrtaRyAayq7tYdm0rojkdspuisg2bborW7obpdll2U5RUOvV0IbtVSw8B6vBtN8uuME4su1T0XEWFUFGwky0++ZDdYxJrTUTDhO/9fp/v/u7v5sUXX2RtbY3PfOYz/PIv/zJvvPEGv/Ebv8EH6yNuPBjzfvnbfOb5q7z/8jX+nfOf4Erfc9nsAGDUkWnZGuhP6J1/VIv0BESigWk15m11Dq3KsDTAK048pVY4KqwE+L74Zjit5w+nWaIj+yiIxWk5UWuExBGQgKfUIG0uwur5lVr/S3/pL9b6j/zID9f6H/0vfqyz90/81f+11n/mz/9Xtf7Zyw18V6M36aTkpOGr+34o2qQorV3m+o1faKQNjbhqC76rLTa39ukc8HncT1uPwiz8OOW0wnSPQySx2OUZYoSVlXNkmWV19TwrK+d47rnnuHf3Lm+//U0+vHGdOx/d5M7GGjvje9xd22Rza5tnL65QXr5AlmXMFZZBJtgE35WjRDf3aDdp4Il1T7uFMlXdt2hHmfGMTDl+mLaFVC/GxAHTBJBd+heAchqjS+0Bo8VGe0ijvPZ6Ttl3duLu0yA6cU+eXHFeqJzSNyNckWGBLBpqCbsuxkb4ow+0sMYj4vAm0MXKYIwTCxik8Dg8eCgKA/MrZMU8n1lc5OLFT/K933ub//uf/T3efvPbXP/2dTYfrPP+/Q94UN7j4m9s8JnnnuVzL77A8vIykONKwdAH9cFgMRY0sO+aPMeoRsbdsKLRlQGyG1LB+Bqyq04n4Lu7GXc15vOUaLAKjfEqtLYLcY1pfEY7LLst3aZ3doDsqg/EOtYGmLxDsZkGll31ZKJ4UXymEbLrcKYidwZ1BVssYFXIcIh1cfysyLMRlZnHEC1OQsi0cjZCdjUy0Tb5EZv8zKF9kzDdpv+10q5MsOwmQ7FNINSF7LYMXEmQXWYz67b0h0F2kw7NWlVqyK6ZcLIpRTWkFIuXjFEm4HcQHTGoNqlOA2T3GMS3l3dG7jDvYW5+QK+X8wM/8Ht55ZVXeO+99/j5n/953vzqv+Gdd77N3bt3+Z2tDe6/9w5V5fjuaxeZu3ae+bm5mCosLrs5ycZHx3U9HpyU3bsfw3eS+TZ+rLX0ej2KXLAWfOVqssbEjmw8dBeonaw8+Y7h6XIcT9MpNUjP5EzO5DSISMrlBf1BHzFCv9/HWstgMGB5eYmqKqkqF4iOtrYCG2+5zVyRUeQ5S/PzrCzOAQWZCFnMvRnkZIbUjr22x9hwpGHjoQPpdIM7MewakXpiHmOkrUEhTk6P8fLsZYx2m3oyjoKTk9NjjCaR6JAIa/AkslCHvJlhth+hDCamARJQtYT8KopaC1rTrcQIo2LQmAYhIx9k9LNVzi1f5r216/RcTjbyvPnNDbbHO5QbI958/zq9LGN1cZFXV5bIrUW8Rcw4TvoUietJNUZ2tTaHElNnc17p+4GeHmkrJ3EDdao6vVxjFIZ3X7PMIBgAPkICfd1bpbP/4dv4pE9k9/dWmPVyOSUd89ilvT6T0NcFMpORWculy5fpDwasrq7y3nvvAcq4LHnvvfdYe7BFtbXJV7/1beal4umFgquDwMWSRSrqwBR9jOlfpjlqT/yR3M+zMXsstdYEn51Nq6dbY6lO9ip9ePDlqMOfcui0KaddzgzSMzmTMzlG6b5EU2oSMYFxc9Dv472yuLjI+fOrbG5uYrMMm+eYPGc0Ltm4f4/1jTU+uLsBYimKgkGv4MLiHEYGZLYgC0ztHcfnw4alg77en6ThIJoXpDiEEbApQiqBWEVJE10/xWvcRC86m/Y5z9vTGG0KhT8HNShSW/b8+eTuxhOILNtTrFHyHoylT8+FvuUNEY5pwAYgpwIqJrCsqsVpH2vCJKtyFmtDiNENS2wvrgvfLsl7OeSWatyjtyAsL87zR37oy7x55Tl+58oVfurWddbWHvBgo+TXvnabj+7nXL874CeWnuXC8hy9Yh7Nd4KB68EzQtDAXFk6RAWbFbhyjKhiswLvHKJgavguHfhul3GXGrIboqHSwHdVGmjuHjDd/UN2Q45fV0bjEqjKANO1XlqQXahyC2oxLqPKlCwvWSm2MRa8GLw3OAuVwk5puYBSGKWM/UUUjK2onITcp9KG7EoLshvubpcwSJvtSgemO1VnFuNuA6ntQHZJLLstwiS6EN8us+50yG4yfRIEeTZkF8ZZyKWLVvSrigqopMcwW0LN6DsCsts2RjVeYDFxuwan1LlzSywtLfLjP/7jvPCJT/DMCy/xN/7G32Bj7R43tjf5p298nXsPhgyd5Qcuv8plKxTVBnmegzi8Hx6xlbKfjC5PlIQxNLbbVGSZYqzHUYJ1iAmdzaMhiN8y2nePqrvlyMg4Tc68J/jCamteIG2ExOHP/VQapO17NMtZ8fnPf0+t/+2/83c6+3/5DzRw3p/82f+r1n/6zzTQ3iu9g13UTj7AJ+Ah2m8b2mfZTpzd6zV65cet8s0e0tIfSSc9pDzq+/GdDNNNEpdzBb31f/otyzIGgwHWGp555hlslrG8ssKg3+f9d9/h+vvvMdy4z+3NbczNu+TGsHN+mdwa+nlOLxMyCZMz2dcLsGHTrRvR+fUxSIQ0HeT9nSZrEhN5G4kGKdpEXZROaDTYnQ8ZJjQ1pYnwpM2dMichNaLrUd+F00FgNCleBVcaiszje4LBYLDxPWdAJbJDhvWlKaG7sQ4xFpXA+hrxqZieQ20W7kPf4KJxYCITrXpYmOvz0qc+xdzyKj94/QO+9rVv8a2332N7uMmdB3f4+vWSr3x9nk88+zSfeO4ZsiwLRrB3iMsR9TgNRp2K4ByBcVcVX3pMFo3AKkB2U2qXDmTXtN4rNWRXA5uuUFO3pvQvHR1FcjqQXXyEo86C7FoPaoPhl1GnRrU5aBaqbUN2bYqAGkfmBXzBti6QYchFEeNALYKLkF1PmfKOSjAwnQ+Q3WAQNv1dtb0+s3FQtdlxG5hu01W7+iSbLi19ErJLF7Lbmmy2obnt90TSfV1eG2juZJkUaNoDsptXQxCDF8soA3yF6JhBtU1p7HcEZLczjkozbCVUjPfhHhsRFhcX+NznPsfqhYvs7Ozwm6//Or/9xm8yGo25fuMGX/FDnnv+ebLLK1xZ7eOcJ7gLDj/XfSLkAO2pc6urYozBWkueZUCTVsk5h7MuRkhP53zuwBkjIizl0Ocao7pHlUcHjD6TMzmTI4mI/G8icktEfqe1bVVE/omIvBX/rsTtIiI/JSLfFJHfEpHP7/c4usfnYO2dPVYYYyiKnH6/z7lz57h8+TLXnr3Gs9eucfnKUyyvrCB5j5FT7m1uc2v9AXc3tljfHrFdVowqj/N6qHUwk016HMOrTH6Z9plaOEachdYaUkgmZ5dt94BXZgKZeKjh5aEXc3qtjwOepLuUo8uj6aPRYGzU2M8aQpjU+SRaNFIbpiGiWj84YuMaU4uKRU2IVgX7zNehyNxYzq2s8szzz/PZ7/4sn3rlZZ5/6ilsbtkZD7m9do9vfHCdd27e5O7GOq7SwPDrs9jSmNgjhtw0Qo0Tk2rniW0//23mrri9fbt26TJNn+443YUWaP/d1Y70aSJ42r720TFE/AjpWtuw3ltajiENWUQT425bagbfiWbMekRn9abp+nHIZIbFWb/sLjO7m02/K0Irz2WtKaJun07I6fKoxtHjkSb6WHfr5qfmcomQ5wWXLl3klZdf5otf/CKf+tSnuHr1KmIMG1vbvHfzI95+/zo37txls6wYqlJO3LPDXtXHabLNGj6nDaWQostavyuNMfWYGvpnSKPGaYPOwPQTflQyOWk4RDv2ZZCKyDsi8tsi8hUReT1um9qBz+RMzuTE5GeB/2hi218A/qmqvgz80/gd4D8GXo6fnwD+50fUxoeKiGCtIc9z+v0Bq6srXL16lZdeepHXPvlJXnntNV56+VXOX3kKO1hgbVjx3p013rl9j3fv3OfD+w+4u7nDdlmFJPI8zICSXdqswWo/cmwG+0M+u0sHMSJYY+qUANaYhv9IdaI9u1v10EjprBPb58i/1zU9ZcP7YeRnOeE+ao2SD6DqZYhmiFiw4AXURAivgIogxoTUJEII/wnBMJS8jpxWmqPRaHQVETapuK0d0BEmK9Htkr4YLpxb5A/8gf+UP/HlL/Onfvj3cfH8Cl4ybq2N+fuvf5u/96/e5hf/5XvcuzOi2hYyVxAyyytSCN6UeDPGZB7vSnxVIZlFK4cvqwjTDVFTIcJ0y/hMOdAyJKonbg+6hDIq9XZplRGVYFsnXZvyooIvQXzQtdZBSwM+ptUpQSqwPkB2dSTYHcE5QcugV96glcWMLc4qWVGxMhiSFYJYi3iDIlQYdsoM6yBXrfMtGgLLboqWNvDaFK2cZNPdzZSbjF5DiKdO01NENHVZ39GJbelCc7uESdTbm/ItIiWaNESpjNJEX5UmcqsqqFoibzheTe1NGGcDnBGUkn5VknnB02c7v8A46x8lIvOznJJxNC15MRPrUwJyodkuBP/SoF9w8fwKf/CHf4Q/+kf/GH/sx/4451bOs6WGb6zt8Pd+7Tf5h7/5VX7t/Y9431nWJG/qBDigg1ciqRWnKIpY8y9EV5lBGRRKL09Io+YTHAKPy8Kj05ZOu2ZcbuFweVPTeR+nHKYdB4Hsfr+q3ml9Tx34r4rIX4jf//xDa2l7MLub9i2ddWMzrmF7+2svv9T57W/9rZ+r9R/6of+k1n/yrzXw3f/xT/7BWl/MGsgqNMyypgUZOS7w16yH4lF3iour52pd2+fZjqIcU5MOem6PAkJxlGO0z+c426qq/0JEnp/Y/CPAvxf1vwb8P4R++CPAX9fQgF8VkXMi8pSqfnhsDTqSNH5zayy9okAELl+6hK8qekVBZoQPb3zAjfffY21nC3N/g9waLMrl5XkyWabI+hRisDKZe2+3EXoc0nWaTrzF4iTrJHtq8upaY8izjDzLyDKLEdMqMaXN03/a90GPHsE8GXjsUWqtOUv3CkEdtD2PoI96D34MWebwmQPNEJ8jxqEKHhvMiTjZN5KgZyliGtsaDVIxxP8UyQwqcYzrV3ixiBokC6uo1Cnzi3NcefkT2ME8v+etf8PXvvUOb71znc1qm/fv3+A33xnz4lV49vIFnrpyEWMzFEvlPTYaHM4Lxlow4J3WjLu+DJBdEYP3HjFhHZV6ImRXavLoBr4bI2kxsWkN341QXm3BfdMAFridotGUNe9p2QXZDaabiflVnIaAMgRDy4qiRvEZWOJabuPIdgxCj1IW6UtO35iQa0I9oo7cjnEGSpFoqIW3l/O2jt4kWKtpTg1BOpDdxjhMsNv0VAdps+nO1rs+p9SrurDb9Mskg+6k3qR/afal3jfpDWQXJiG7aXPuRlQYXM2y6zA6pl/tUB0BsnvaxtFmfXSNOG/9JtFA0RoFYYzS7xe8/PIn6Pd7vP3223z1t9/ga//mt7l58yZf1xHn3QZiDGZlgYtLdvqBj9ru1Pj9lIsSnokTxMuI1IO4aiB2yrOMubk+vX7RtOZYxrxphz/gXJfZxufHUY6yhnRWBz6TMzmTRyeXW4PjTeBy1K8C77fKfRC37RpIReQnCN5frl27dnItnSYS4LtZliEiLC8v4aoQKXmwsU5Zjrl37z4bO9s8GJbcWt9kedAjM8Lq4hxLcz2sCdM5GyNB9StfOdYhRado3V9P2GkUZ44Sr5m14WNMmm23vH210cHRmHaPcEozjcVjG2B1qnowOUZrdLYcax99+uozYZKqidrG15PSOj6e7n/8ms6wMTzSQ9LAehUClJcQyQvkSLbOiRnWKENRWJbOr2KKPp/91CdRp2xubHP9zk3Wdh7wzu2Sd69fJrfCudVF5op+ZCKCtumkKUzmqKOEJKMk3ZYU9qNublhPWevxZ0MgMEpMvsmwqbuEdtfjxfLEfeswXv04aD13Ve3MYxu4bqu4imK8huimpEiFBVOQGYuNkEDvIaXj0YlDKtGOZvdT2X27NPdummtf6iu8l56OqVPKdM9tmky2bVIm6999Drv3ju6h+rvRYKgiBo+PZ+yxWsVvxyonMo7Oujb7lU4AYI+bkZ7RUL+QZRkXLlxgMOjzxS98gXK0w4fXP+DevXvcvH2Hr8mIZ198mTkLzy4sB9gqgWV777Gsu33PN+c+Xqu7jNYTDrx0XdXhWhmTUr8UwDi2SSYLNu39mMixpGlr73oMF2e/BqkC/1hEFPhfVPVnmN2BO/JYJ7tncibfQaKqGvvoQff7GeBnAL7whS/okd5PrX33ExwWQuoSYyyZtawsn6Pf68V0MA6bZYzKinfVU+1sc3tzSHFnDafK8sIcc/2CBZ8zXwhipfbSh7qb/2OLDn9iT4DU3mMJhEaZDXBdG9Pq1JAewmDTjow+6jPfFyvvxB4HtX61/u8YJM2gT9incBx99HPf81m1PRhLRl8jXFscDsGIYiXm/xQwxjfGk4GY5yFgJlPkVCxpJafzWVheiuIqg7Ghf/rtbSQ3mNygQ6FnLb3VJX7/f/6jfNerb/DZS5f56X/0D7i3vsU3Pxryi29c58a6IL2LvPpSj7leRu4takeoOASDukCba4sCHVeoD4y76ipUFWN64CI7rI0GpGqdZ7Rh3BW0iueXGHfji0hdNF41QIFrlt0xDcvuOLHsRmhuXKznShPeTYArifA+KEswPnyqviCVYCtDWYB4g3UGZwSbO5YHQ6QIxr3xhkqESg3bZYG1UNiQ8TxFQgvrIstuhOlqguxSk67YGCVVjU44wvZ2hHISyrtbb0VZZzHx7mLZTaRKs+C+u3WD7M24qwLYsG6ZwLJrCA6KUTbAa4nqmH4Vnu9K+mzn58AMT4xl99jGUdj17j3o66UdMZ61s9Q/NlaTFRj0c/q9c/zwD/0Q5y9cwPYH/MIv/AIfbqxz+61bjC+8za0tx8WlZZaW5ukbz7yut5ybtqnPm6Si4gNSIZw0tVNrZrv2kEfwzt0l0QGHMYgoloyl+WUWBkOEdbyXuBQoQtsVXMvBexRpI+ZOAvWoCSWy34uqeziOHyYTx5ms5zDowP0apL9bVa+LyCXgn4jI1yYOPLMDT+2kcGzzxH3d04n6P/1d31Xrf/Nv/Xyt/7E//KO1/gtvvFXrX/58A/kdSJv+ub0E93hejo8amtthBzbN+ZhWairnq1r3rfP0Jz95e6LqOerxTujefpQgRCLyFHArbr8OPNsq90zc9gRKNB1FQ6LqogcIly9doqpKFMit4e6tj7jz4Q3Wd0r6mzvc2thmca6HU481BdbYmvShHguP8dbX1dXhkl2ncPzSjg5FbKnECYPUgbBkiHYyqNWfRy0HN0ZT0f3lXjtO8G+nroMM5AeTY+2jXsGPlcxW+KwCtRjNMcaBCk4SsBO8DxC+th6ifmkSmUKGsQ+aZKgJkpmw7hSBIkONRTHBOJRgIS0sLfP0K6+hZo7PvP0W77x/gxu37nF7uM71+5YbHxQ8df5VzLlFBoMBohloIE4yYmLU0CBZFiK+pSI2OF208g2zrk8GNZFZlybiKcnoDL93oLkiDeNuK0SaILsgLciuTIHs2mC/W8CHax8DxygRspsp3io2Ehp5o2QjwGeMqj6ZN2SAlFv4SQAAIABJREFUF0W9QyRCdkUp6a6t9D5EkoUG1ppOJznYamOy9RpqRzbbrLntV+Cknr4ZaWKTRmT69uZNMzXiudf2Jttts72G8gpMg+yCkrsRJYKXjLEF1CFa0ncjKjXHzbJ7/OOoTOiHeG21pwyzpw+tH7RxBCdW6NXVFT7zmU9j8ozbt2/zzje+wQfffIt33nmXVVvx7qUBr7z8Mr2BjfulvvOQthHe2Ed6Zdbj2qOUsM6+nqeJUFYVlasIaZN8Dec9SUnruE+gYvYFizrqMNo6ztSqDnFy+yI1UtXr8e8t4G8D30vswAATHfhMzuRMHp38XeCPR/2PA3+ntf3HI0vg9wHrT8760dlSkx31epxbOcfly1d4/vnnufrsNVYvXML25tiplAfDinvbQ9a2RzzYGUfWXTrrnVs+42MTaSvtz+Tvx3msNNVrHas1b2spumfY8KSM092m8AEPpm314f+m7XfYdh9HPfuQ4+2jGtZUGu9BXYTrSkgfgsY+EJ6FBG1Mk9PwhZrNte4u0cJJzg4iIZIECxXNbDRILSoJ2KcU/T4rV57i2nd9mtdeeZkXrj7N6vw8m9WIO5trXL95g431DYY7Q7woaIaQ1cdAQlQswiSihz9uR6OxGc8nNi04hcK6UE1tN812JrenfUzLSVPrOhUWDC1YcNwe+JA0oJpDytEAETZBN/F4YU2rIGrwmmHrqF+6CR4RF9LFtG8r4FTqZiTZbUg2vzzs0ZWJMu0+1HZYdct0t2tnb6Z+29/2ae/k5gjS2SOy6caL78XEZ9WT+TFGPccsH6txtO7jCnNzA5599hl+1+/6PJ//nu/hxZdeYm5hgXv317hx8xbv3fiQB8MRY6d4lbguVepbk56/ac9ah/X3EJIcK53PEerb1zFr51TTBl9VOBdT4NSERkGOdXiYmDOk66o6/XPYg8+qr1P3MUinnRM37kRIjURkHjCq+iDq/wHwl2k68F+l24HP5EzO5ARERP4mYd32BRH5APiLhP73cyLyJ4F3gRTm/wfA7wO+CWwDf+LRtzf83fPdN/GjMaZeU7q6skqv12dxaQnnPHlesLU95N6HHzASx4drmxSZMBzPMygy+nkWcrKZVkoMDu2cnn1ex1jXYY+fmAKNNHEJjWsKu+bbAVs7pfheA8u+jLq9mjBtRrzfembVuy/ncKvd9XNy9Dv7KPqoMUrWE0rp0RMTIaYOF5/7DE8VJ3aWkCdTJADwQqqSwFQrEmCvYpspp0qAsaGK07j2EcWVgmQRIjvcQXJBCouOYZDnDC6v8Id//Md469ff4PW//4v8wpu/wfv3x9xdu8fTq/fQYY9e7wL9gWKNYF2ON1tgHVZMhOwSILvVCFWHtb3AsqseYyzqfJggW8ALvgrGoSpx/Wg4wQDflTCxTDp0YLr7heyaOElOkF1BKMeB3deqUPZMgOyWQtULl9FUEiDTtmK+GOELxdvA0DtGKNWwVWaYTMmtpyLVrRS2wjnBp5yt2kB2gx4gu8mYNrXODGhuYuVNxENTysyA5joa5t6km6jLnrp0ILtd9t34DEsk6JkC2Q3vt8iyqyVo2YHsbuXLqBntLwo0RR7lOPo4xoq0xtkkL6bC4vw883Nz/Df/5X/Nr373r7F07jy/9Eu/xDsb2/yDr76Ff/YlXvYX+NzyAGstGZ6eHwKgSA1ZPe0ixiCqiA/QfytgceGdU40RHKIVaIUjLDKX+gk/huPPeCI0LaU4rSLx3IQjIRL3A9m9DPztCDfMgP9DVf+RiPw60zvw3pIG/0d58XdZ7o183xe/WOt/+a/8lUb/83+21r/w8p+q9U8tNZcsPbBTD3JKpOMJ8s35zA+K1vaGZbjr023V044Szbi3JwFZnVXnYTrFftp30HqP85xV9Y/M+OkHppRV4L89toMfQfZ7BdrGk4iSZRmDfh+ASxcvMhoO2XywSYZjuLnO2tY6vXsb4DwXlxdYGvSxxpLFwbgNXTuNMhmZkJbHVBCsCWtJw2QyRF4exsq359WovdPH2E/3MPa07X4/Qj1T6zxA+47znB9FH/UKrlSMKXFZicFitUAISyt8dFKIKpFnByWurdTkkZfaTV7DYWsRMNFIS0y8WYiYIgYKg5qIW22lo1g9f55nX32FnW3H6ze/ze07dxmXQ966dwM7r1y8skRRzGFsFtajlgH36kURG+C76qKO4KORKVYamC6QOoNkjadfEu4VIhx3UtcpevBGSBuyW+f0FcQ0bw8T61fVANmVaBwmyK7xARqtimYeKyBYnPboqSXHxCUuHtGK3ATIbkXMfxidAzoLspva0Yp47sWU29V113bT+tZAc6UDrzWtvfeE3UZ9Nqy3tT3CpmsIskAbstv0RyVzYd7hxFJaCAzFJT03pqpzGR1cHvU4+liMUuhcn9RPVlaXefXVl/nBH/xBvvGNb3D/o+t8cP02X3/rmxTjbT716jX6/T5imjft6R1Bd0vKMRqcMFIz1ltrsbbNxdDM3R5HjuyZIrvbc1gm3ofNTQ9rWB5lzvtQg1RV3wY+O2X7XaZ04DM5kzM5k31LGAGm/mStpSgKxBhWVlYYjUZsbW0z3HrAHTxr9+9yx43JjLCxU7JTeYpM6aeJ9MdhJE0zQE1f4uRSgsFgY95J2hitmQArmDnLkF3KscmjHNDbjKEH2es0SYoIWvWoDbErBUy893UkLD4OyUHY1mNWlNrAaxs5oZA00DaR2kgUBDILYkKsztQWLvMLC1x89hkqM+DyL/5dRtub3Fl7wLtrd+nPZ7y6fom55R62KEJ9ZUhx4pEG+loFgxSJ60mz8Kx7ot1rCFFPCUZobbTWeqwnkRlN0YnlA8tuhP62kmVqPB8xWnepUCZ8z1LOVwIfsYqiVjEu6iZl1TR4ycnFkhHzwWqA6xpT4VGcQE7TY52GtFa1E4pknO428rqG52w2XT+lDK19U9qZafvWxC4T9Tfta5u8Tb0t/0DrlSMTdbZbwa7vxleIZAgGZ8D4MaKe3I9QaRzmp0Ie81gUhg9hbm7A1Weu8gUVrl27xnhznZvfvs47777PSi5sPXeZrNcni0/wRA2nXxITeb00oGH5tzZiCdTPDLzU8jBH6iMcUg415j2hQ95R0r6cyZmcycdQOjaLdrcd/4tsdkTLWoOxhizPOH/+fIAS5TnleAgCH978kPXtDbLNIR9tbLE010O9p5jrYXPTGGo0k6vTK9FrG9lTrRGyzJJZE4zSOnTkW4NTNwL5yLkjHiKdPKCTMmV8PU7D9onyeh9QrAHbF0opImQXEI+LEO4MpYrRqEzBR1IjA2iM+lnVWg+RqnjB6/WUihcbEsCo4lxg3zWi+LEPEVOryPYO5IL0LLayrCwssvhCn+/6/u9H//VvceeXfpU3393g9n3Y3ljiD//7K1y9NIfJ+2BHYAUpBdUKxZP18g7jLurxFRixAb5baYTs0oHsepdgrRLhuzESV8N3ma5zEMhuiOMFyK60ILsGEyG74sFUYS2oZMp8v0LmBNMT+iqMxFCSBZbdXCnUB6irChbFtiC7VgLxSsh3KpFltzH2PBGGzd5sugkpMquMC66FqfsKXQiu0GbQ7UJ2U/7RUJ7W9ubRChgs3ROyG5ZbCON8Dq8laMWgcjgMFX228qUjQXYfucx0/j0CiceS6GdSD/Pzczz/wjN8+Q/9CL/8L1b4uQ/e5c13PqKqlGfOr/D5z38PlwZzDMbxzqlHdFhXqIl999RK6A3jssQpGCssLs0zPz9A2MD5EZUb4bSMjp+4aDyKn7imu2RKFPM4ZK/Io4jsO6K5L/bjxySPzyBtX7uTvjYTT41vz4BaUKUf/dEv1/r//td/ttb/5j/7lVr/H37o99R6vwXZdXL6bfv2A720vNhsb0F5MdMf+idtaDhOKO9xHfu0yeMe7zW+OAUoioL5hXnOu/NcuHiRBxvrLC6dY6saU6pwf2uHuw+2sKos5ZZ+lmORGOxJFhnpbXxKJRoQkXDGWovJGqhR4z2I8ZY6+jwJ8YnFHlGr95apAKSHe6C/g8Ur+FJDlC0rEbVYCdFKiQZMIgcJkN1ghCX4rmgYA+tu4QM8FWhBKVtEWkhIBZOipZmJhEcChQT4rmaIEYxVbG555doz7Ny/x7vfvMitj26ws/OA9bUb3L1/lcU5y4VVG1lSDZp5xBGMTC+IsWF9qwtGY7AtNUJ2pQlu2NaznCKbUa+7wn4gu7YV75sB2ZVoWamCtcl4CwY6mUdNTMFiQrTUGhAxeNOjF6OkzgCVYsSRmRIvBCcCgkqAV3sNE5IEHYSuUdeNijaR1bYTcVJvRzxp7ZukCwNu9uvW047Opu2NkTq7TCPN9jB5rp8v8UhaGdtqWOZKKgQvlsoIqmNEKwpfRnD6k/EGmymP4x02GdRsOZTT8JdZS7/X47XXXuP+nTu8+NJLfPDBB9y+c5e3vvlNXnjxRZZyS+VdvL9NKqBHcTot8tsTq9/70JmNWLLM0u/3KYqiUyg4xSz2oOtHdw+5xyJtKPw0ac8795rnnlDzjkWOZ6XumZzJmZzJMUrndSqQFRlzc3OsrJzj/PnzrKyusrh8jqzXx4llfXvI/QfbrG1tM6wclQ8TvBAOeTzncJIiIhhrsJEEqp0uIRSYvt+TfylOc+zy0YhqjA5WDq9VgJh5qf0QdZRJUxQqbaeJlk9sbxh4J5lboxGaoOG1QRpBo3lix7WIWMRkGJPz4tUrvPT801x54Qo2E8rxDg8e3OLe2l3WN9eBEvEgatAsGpsS2IOxBjKL9/H4tmHcDYGKYLhIvT0apNHaDuy7WofoNEXhJ3WJk7y4JhRSPeEYEjHNSjR+TSCIMhbIQLNAMIUFn4OIBkPahgm8EYuaHgUZuVi8CfFF0WCQKk1eTiWs/a18aEAo2b3ntV7fmVnrSXfruqfevttto1J2badVftIIZWJ7OkYbXpyMUO2UDp/gREnGs2J9FWHllsqENbiCI3fbWFedhpfZY5DmeiY+5brbR++BMYHJ/sUXX+SVV1/l5VdeARHur63xrW9/m/tr62zvDHHO4bzHHYDR+Ki3RDvncEISjU0I7xybZRRFQZbltQ9XVfHeR5LFg0eE65zg7c9Rz0nZfx17DaIHqecQ1R9FTn9Y70zO5Ew+1qJAnmVk1tLrFVy8dJGtrU2eefZZfDlktPmAtZ1NPlp7gKjn6soSy4MeOQ17ZZDOtOhE2tmW4z1KMw2UaCCEwVJantFm4NuVIuVUyfGEFp5kT/BRxBrI+oZKcgoxkZzD402ItmWiVMRoiCbIrmA9xPz2EbILECGaEmOEaWGpKHVYBMUT2CkNiitteO6swsghFiS36GiMsUI2MDzz3HP0reVCqfxP73/A9Q9v87Xbjk+/e5+BnePChadYnBMyY5Edg5oSNQ7btzByqFNMXgR0TqkIBTiNkN2A9/SVYiLhUYDsRmqvNmS31unqbfbdBNNlCmQ3Eim58SRkN9jGZREgu4FlVxAvmEqo1KCmopdX+DlFe0rfwzYBsrtV9jCFJ4/UiKLhnhhb4Z3gVMgiy64Soq9e2yy7zVphZS/IbmLi3Ruym5h1964n6bWtPwOyOxvWG5ftTkB2Q17alHIkpMgRxvkApw4o6Vc+sOzSY5gvomYUHBBnMiGNMZdg0JOS7sW5pSVefe2T/P4/9J/x9Xfe5+Z7b/P1j9Z48+Y9evOLPPf0Aopi1JPXBhxTl1gE329jTh4GJdau4yTF2IzCZjjnqBRcVVFVFd57RDKc81TGh8UvbajAAeRxo+T2Yrw9HsiunEgE+yxCeiZnciZPtNQQQglG2KDfZ2lpiYuXL7Fy/gJziwuUTtkcjtkajhk5pdSQg9FP1nNCom1FJ7YdWqT1aW/bffQWWIfdJuk+WqK7lGOTx2sYf/wmrV7BO8W4CqXCq6fOQ5qYdSFCc0mhwIbUSOOylfibxqhqKgMpgprChuG/huAoRkwT2ZFJVq6gMVqa9eaZW7nIhRde4crqOc7PF5jC89Fwg5uba5TbD/Deo8YgvVhXis6aGHX1MfRpTRPhsdQo9BqyW0dIG51dujZMT7QirRCjpWHyGaKl0Ti3SiJzFVtfxmCkZuCzaLNnihbU601TTlJB8FisNxhvqFJEFoc1JQ4Cy2681J6w9jTBplN8qx0tfRizbtsJo7Trmf0W6ZIUdevslulunzxW+jatDW09+griY9V6KAkR5rRDHSHVFCFVhIrcb2N91WnTmbREHv5JY+nC4gIvvPACz7/wAhcuXmI4ctxf22DjwYO6nEhNZ7bnIY9bgjPmmCVaUcFxS32O3vtokEoHMZK2f5zk8Nf05MfSxxch3c8TfFLvmxn1FllzOf70n/3Ttf6T//2fqfX1398QC/dk1KqyBUJpuw4697B94IPd3Lan42jeF5n41vre4v5Xda0yzbHHWrX27h3oyLNw7Uc5n+O7Lqfr2I9EjidQdSySjFKAXq/H4uIily9fZuP+XXw54u5HH7I1GrM5zBlWnspDlaIMtcHWnnLNul+H76PRxTulnsOITFz+7tSv0bXWm2ldMEoOTGJU79M97yNN/OKlftj6l5OSZJB/nCavqmENaWYczlYhYqZS55JUlYbsx0Nc8thh1q3XlqJ4H9OaxH2ltkolGmfRG55YKW2aoArkNsbEBLUhU6X3FpMP6C2dZ+n5nKsXVtm+d5vRaMjN0QOWNvsMN9dxy/Nk1iK5IlsGrUxMQRPibFqCycIaT+8CFFeMoGUwLsVI2B4NT+81wmajDjOYdaOBmVh2LTXLrphkxCeW3ZSrNbh3vNfAsmuicSoh8ugyxVagJqSxsRKeuUoNtgoGaSmC4jHqyGLaF4eQSdjHC1TekCUjVFuGrW+v19TaQPAdPdkbXUZc39LT9un5Qbt5TXcz7tLRU5m2MUvcvjsFTNfgrY3Tuox25hkCWFeiklMZS2VBnIY1pG7IWIsnZmyaKY+rfbOOK8k307gI5+fmee7aNV566SU2793h1ltvsra2xtr6OsrTTTQtOaX2eI8ehFhnRvOgU8de4/RhjxF6iotrSJOj23uPqtbkZglp5NVHzpmDwXanXgc54lh6XKKxzx1wzqqt+Y20DPuw4XjO7QyyeyZnciaz5Ql4fyZJE5hBv8+5c8s8c/UqD9bvMx4OecdkbA1HbAxLNoYlm2NHlhkyq4iRejipJ1bxvBQeKXnT5KGmX96JWMakoZvyTIrE1BxxiqegkWVXJf49qCl4fPb0jMr3U/H0G7LfM+nQ4D9mg/gkxLQhu8YGuOUMyG4i2oHQBxrILg3jLo3zwqAt4jqtO51K7Q/CuUA6ZAwwJhiHVpHhCDKL9MBUnkEOxbmC57/3i2xl83z4j1/nxkdDqNb4yvJtvm/xIr18Hq0yMEPIKkwlqAvrYm2vQCuHjhVDEVh2ncdIhrrAoDsVsqtpwqRdxl3XMqCc7gOyKxEODW4sXcgugYangewaXOFBCZBdb1AqetZRDSp84Rh4zxZChQ2Q3Z6SUYWULl5qyK46g6+daeF8JiG7wX7WCNkNqXMOCtntGrKNHiDcxFyNsXwkIRJp8qO24biNYdll5U3wXery8TmL9agG54OPUTg05oIUGsiulvRKFyG7BdvZAmrGpwey+0hfO+luRKN+WpG4hlKsocgt58/N8blPv8Jw/TZv/H//im+8d4PCZnzrpRdYWVlhLhfyKrHsKtJaU+pNE8SQ2O/CcNUCX8rsCKPG9tRFpfWiOSp8dwreNjjpFHWOLAvJbbwr6fdz+v0MaxzgEFPirMNlHts+l0PKpPF36PFI2f9zP80mPubAyXHXdwbZPZMzOZOjixzys0d1078L1lp6vR4LiwssLiwyNzeHtTlOhXHl2R5XbI8rhmNH5btxxFpacKBjm9Ykb/KME5t2nL2OPStNTdcYbQ9rKT56fBO1Qw2a0vyR1r/9zsyOo/XSbsSx1vwEiCbIrkPVhftds+ZqA4dM0LM2HDd+9iI70tb2lJOzcYxQRyoD062p9UgtGyejBjEZYns8dfEiV66cZ+58j9JVbG1usXHvPjujEaV3GBsMMvGTBEcCEiDBCjESKi29cS7NguzWM5ykJ0ioleak25DdWEbRuk5t1a+AsYSoqo0su1bRPEZq4+mnvqkYrLcYNVQ1WsNjpaoNSYmRZ0XwajqRROK5Nn6iNuKjjld0vk/q031Msmf5Gjas3X3rZ4tuVLRd72SdtPTaeE2+DgnXu3k71AfAeBdz6xpchOyCI9MdjJ4yyO5+mnrYMbQznjZf2m/deJnjOyI9qOFPllkuX77E5SuXWVxcZHt7h3v317hz7z6l8yAmBhMlsGs/5HSOcld22aC692h2eFiv1H9FqPkY6jG0fdyTeMwO0ug0zrfg03Ude32OKOm9NF0efoDDzEMefYR02tvxpOcJE9i1mVZ4qx3/7pf+7VqvbL/WX3/7Zq3/hy+fq3Xb9hq1vAbd+VDr2wG9e/vxROwLQjpxLdqtaP/SK5pHw/txU94cD57+KJ6Vk0rpsp/rN4tae9a+jyLNzMdVdt2BeCmzLKPf67G0sMDi4iJz8/PYomCMYeSUzVHF5qiklxkW8owi5vRLVqhIl8EyVS2tY4QGCPt9OdUlpbttl2jrIDP76ETFE1/TRM6I1OvWkiEq7RHpsBGE2pY7Qh/dY999DVS7JibESXuajD+8bZ1IaV3ldM/0aSKASmlfMlPhbBnn8DlGNKwbpQXZDfZSvZ9p6YHNVXEJsqtEyK7URm5teGgzOTEmMaIKZCbaeILmIV+fehsmsShqMq5duMDGU6ssXu1z7xslO1slD+7dYXu4zdiXzBWCjgV1gi88xgerTiuDZMEY9MNoBFsJUUwTjEJfJShvgCdLZNn1Lho50yC7SfeATkB2rZKWjIltGDIlC4RKXiNkV8LHoPgsQnZdvOaiZJXijODEUrgM6y2lCXBEQ0VuxziBCkMuTX2lN+Sa1o2GaKjQsPGm90ywr9v5Qbsw3S5kN5zQJEw3PfHt7Ql2qaS8rg10OBEsTcJ6G+O0MVVTmZQHddKA9SQ4ssZziJ96PSlkboxKjhpLaUC8IjqmV40Y296T71+aZZk/ZqnnJxpgm9Zann76aa5evcrq6iq3bt/m3r173PzoI56+ehWZ74X3rtjgEDrRNZVT3Bwpz9OUkqT31IGGqi4hT1hPG9AQ4XD6sZ23HdeyuT3hu0eQswjpmZzJmTxGmeG8mdzSetdl1tIreixEg3RhYYnB3AImL6gwrG+PWNsasrE9Yuw8ruNF3ful2XUwHtBpNHEO0x2W0+usy016fHXiL3EAFQkpX2yT9kVaNR2Tk/TJknhSB7k3k8bnKYqpzJQastvrkZk+1mSIDZBdNYK1igs2HdYQoLm1Hj7GENZImhDxS7nfxUartf6E30IdIYrovMH7QP7hSnAuQDr9yKOVB3HgRlgd0ctGrFy5wtVLz/Lp/mUKU7A+Vr76YcXGnTHl+phRBS4DzQV2MnzlUB1jehX4Cj8qowHp8aULxpEHP46TIR+gtsYDTsJ2H7ZrGYxp8bJb99H4LokpaMCXEnQPWgriBKOCH5vAoKsBsqsjsMMIzR0Z7KbBOQKMeBxZdjH0MkUXPLbvWVTFiMFpzk7Vw6qSU4X1oT5c7p6tQAKDrm0ZdfU64JZDJcFxm+3Uhmxj+DXQ3Da8dpee1tJpcnZFp13SNemJKE46+UsTdBi0kzc1QXa1bhvR2A3rnj0CalEMHsFpE+oe5wOcBZWSflWSeUHps51fZJz1Twdk90lvogIYrj37HK+99mk+87v+LaSY49bGNm98+zrvr29zb+ypTBbW8opppTI5+uFDUFc6H2lvn9Lc2mBsrTU9iDGUIqJocIIYMWTGYsTgvaeqHM55UB8dJScjHerBo17LOop6LE17+OFaEdvjlDOD9EzO5Ewes9Rxv4l/7PoAiAQjLLMZeVFQFD2KXh+TFagYhmXF9mjMzrikdD4S/HRox6a2YlpQ7sgy3Sptnc304+x5bEleXYOt035MeonbA/aTLd07Xm/cUw5yZvuBDe9+8p5g0RgB9P7QkN1Uz+7vEr9Lbfzv2q/VIRPbLgQ4X8RgphklKpaiN09/8Rz9y5cxWY56z7AasjXeYWc8hLIEF7H1mW/2VRrrKOqJBROCQZ1O9qGQXbQ722kz7krr9OK+7Vyl6VjpsqSIrJrIsmtAbYshNkF2jYCxZGQYsbgYhhbxGHHR2JT6kkFkP6aJhNa6Nvo091Y7ECedEk2p9lOdyrQjm8TL3WxvlZfmyLuDftI5QvvYu+qpC7T2iBsFOtfCeBcJeE1wtggIHqtjjPonv58meZJfwxIe2F7RZ35+ntXzF7BZzrisuHv/Pg+2dxiOy3ptaHdImX1ijXNVHwqrnTbGtx6R8L39/pp1TJ3i0J1xxGatazhmlmVYa1vn6fHpnXRSMnVecAiRdh+enAcco7T8BiclZ6RGbWld6aWlBo77/MufrPXXv/52rf/el79Q65aGlXZ3nCRpfmqJg8pBw+4dOOke9bZ/mx/M1bptJwY+YQrs44JKHBXWu59rvB/v0MeSffeRS3eKlYwxJCPPexT9PoO5efJeH8oh2+OSzeGI+dxQujl8az3I5ARp9/Rt91TuUHdwckRtH29ahXX52UdrhhzBiIkRUhvWv0jXmNrb3D26dEiDDinTJpRtCOAeB49GRLfU5Jl/XKUD2TVj0II2ZNchIRqqsyG7LrLZQkwT44nstIS1VIldNlG3JhgvWjtAwhrSxsz3uUWSNWgMiqCakRWL5OcuYF96nuxXCoyvcGyzVj5gbbTJU+MeQjRE5ypkCFIavBOskQDN3RZMRtBLDSRKFnyVjL/pkF0TYb0IkVk3GkIdxl2tt5spkF2v4XgexXtPnrUgu5FMqio8WRWeSS9K5sEbwZucvvbItWAUx1GDo2fGBPoUQ2EEbzxOlcrb6ZBdTRP2AKkVSVDbZvtklBRiTtBkzErDYJoinclBxOHsAAAgAElEQVRm98mwp4HjNmRIKQKaIMeS4jqNYavRONfQjkSGpNqwefqO3kpzi0ewmFadoOQtyO7YCsaViFb03Q5j7T/Zht4TLok8KL1Ds8wyGAxYWVkhz3O2q4o7d+7wYGODnZ1lpBdTL+1j/tQpp4Bo7VA6mMwel7vER9L12iiozHZXpOUG6n2dESPPc/KiwNqMsnJ4hcpVob884UPKbhRQd6nKfmQ/8+JH4QA6M0jP5EzO5HSJSIiSKhRFwWAwx9K5FR6srTHeEoZlydZwzFZhKL2PycFbL9N6JvX4RprpLqv97WdIkF1LZgw2wo1MhLDVy7Ce8IH0SDLloh14KdEpFWPA9oRKehTGYDAdll0rGtYcSsOmm4yTRBgZmFSDWVHrEll24wMkcXtN8hOjZ21jQl2Y/ogAY48aRTJByyqsT8uEoidcmLd8finjX8732bAD1jf7bG9UDLdHDHOlT2QE3S5QHYE4TG7R0qNOMXk0kitFjA0wXR+MTzyBZdeAesG32HT9JMtuivQ56iXivrXd+yZK56uGRMSV4a9RE1l2A4y2yg20WHZFBeuEsQaDPBfFzzvswHMOz6YYKs3YcT0sPrDsOkXcbpbdBNlNxp5P61lFggGJ1hEdr937E07fB7bl+NwkplyAqi4vtfEaLkyagCtVrQtefd3lXJ0khpgiKBmZTbbUtk5Hb0GFteHlhQTnFWx8JodZL7DsUlKUFR7w9NjOFvCmfPI7+5PePgj9wjvECL3BHFeevkZvsICXnAelZ0cNYywORb1HVMnSOIOgNQPtRKhvD2bdQ0nbmNUmut42PDX28/q8YrMm1ziqBqbg8K4UMmMY9PrMFT2KLMNXFYW19LIiON++Ex0fnajrPnc5huDLGWT3TM7kTJ58qdeYdDZiM0tRFMzNzdMfDMh7PbwqpXO1p1Pr/WdWflKNPmDRfZaPg0WaLNcRKybwNCc4kE5nsT0eOVqzvwNmD3HuJ6qotlfoxR8moG71PhP7N9O6VoBeO0W6u00GKaDulwH514LvSvObMRlZMWBw7iJzvT793GKMZ+TG7JSjsGYUoqXdWpfUqkO1tS0FRRrIQG2MTXYBHqbH6GptJUn3t7r4hJ4gwiIx7GxbuY6laTpGsASHkUrDUJqigSlwlJDSvoZMd6V7HyZucPsepxs5NeIx8RBEa7cmKNGIOtAEs2yXiXXW6wcfrtfW9ES70vpD7bQ1ri5VH40JDV6DWDaluQGP0bKTeuSJlWnP1JMoMbpY5DnnL1xgbn6BLM8ZjUtG45LSuV3j7+7TkSna9N8P28RZNWirzLRSs0aE9q2xxjRIo3Z09SSHkyf9uXgM8ugjpKfmBjRP4pd+9++p9V//h/9nrUuWN8XLhol2VuRFjsnVclAWV9nnJNW0fuvZ5txy24LsnvB87yQYak+KMW0/sN59MR+fyaFERAPjbr/P8vIya0vLUFVsbm8yrhzjykUoWnvo2Wt4gkl36GHuWPDcxi+61zFb5TvK3mXTZDdESk1k251BMHDCRqnGCfhxHec4oMAz62T2pOU0iVeoxkpuxjgzRikQn2PjeTqkZs31CtbHCJZPeSsDZLfOSdmG7zZLONG4XaOdkIw+I0IE8GFt80z6LCM8FQa1YaGlSha6X38Vd/U1lhYWWekZyEZs6SYbo3l0awyLc5BbTL9EtgXKECU0NjBg+i2DyQnQ2SqSMlm6jLu+MRa9i7DWqANdyK4N9g4KYiVulwDxncaymyu+UrxX8lxQEbwQotHW4wpPVmkgWxLIVFEVnDH0XYF1PXZMhhKW9/TsCK9CpZYipp3xAqULkN0EozXR9ePifUnox3iLoiSiI53Y3u2anbWiLb2JSrT7STrQFI9Ga/1mm/VaWg1Lv7b8B90AWruN4gLZFBqZTgPevKhGjE1OJZaxzTB+jPEj+m7I2J4CyO6T3j7inEQs6ip6vR7PPvss58+f5/b1Adv37rO9vcNoOMKYQfM47MMXcNS5Tu3s2kNCe5J3Kh233rtxYk20qXm2u7+JMQ2aQhXnXGScPpkb2e5DR5EEK67rO6n57uR1k+57pA3PP6ycQXbP5EzO5NRI1/YRMptR9HrMLywwmJtnuLVJ5T3DsTIcVziv9RrSemL0/7f37rGWLXl93+dXVWvt8+ju23fuXMbzAA8gQkSiAKORZcsOtoyUYELAUaIICylgO0KJnMiWFTkg/oj/sRTHihVHSkDEYJMIbONJsEeWHEGIHUuRGBvI8JwZuAwzc2fmPuZxbz/O2XuvVfX75Y+qWmvtfR59Tnefc3b3rG9r96m99nr81qr1q6rf+4wzPqmY8iTTgEwbRTo9W3SWwULqimY3a3gVq5NqPefGivCpyo6PjfOe84ZwPsUFid7u420B18rKeGMi3Xijdl9YdQJhISRZ0Ljsvi6iqOQyQB4rmVNt4rJbYgJL2zN9FhRtpI3vDFloY1KGw0q72DOzG2lxjxUBYqqakuxaK4Y0irjEXuh4jz9G9m6xDHdZv3XMvbfh4TsSfTDM5/Po2mUPT9HsFhwVjYpbOMwUkuHEZ+NZYnTZra62ykmX3dIeXXZtdNmlCLVlUav9qL/SmK1CDkhd8UZAiP1Jl13XO2Kj2ViaoFOPqhAU0kHCHyTe4Yz7ztFbwzrt4wUaSTluV3N23+qyiwq+mm4NGlG0xGeG0lNW+xEpWXltq12jM4uiwsZsvckEX9bxCYentuv+ltulRmg0waE4jEiDw3BiRHM4jICyJocNBIyuPPWA0RfVlceI5eUK5fxmUgTinFBGTIoQLsT2ACzhLdGmRMKRZMGquYW6+KzrlXYIWZDY398v5V/ewxuf+wyffeNV3j465t7xEpPD7AFhcrZAOnXffQLjy4ZANXXJ3RK0suL37JegZoGtZYysKqfHIQ2xsTyxg5KxPh/nq7btiiE1lvcmJmaZKJZPvf5kdrwAfdNyU4+DWSCdMWPGJjZX9DsNKbGUTdMSQkCcJ2px2U2JpHpG5r2nO9PYxb5cADJIZRuW1hOnG9Ou148TQTfcd8sJZGLsEKjJQK8Wp4v/+aeJdHxhnLbzo57t07e27hYmzrrVEcA23xnjZNtO+3Wru6YKnLOVOYwXPfW3+p8g4nHNgiYEgncco6z6juV6TYqZR3M23cHetqmlmWgpanPbEjg0znWCsEm73sKm6e7Uw7fPK5UGG7P7lp+yK+9w8uLF4MriOi/V3AbBk77Z4M2TWqRNS+h0z1EJtdmuC85KU+7JPEbYhKayvdxIFb5rfVAnrizca1vLfVSB0mWPjTLAOKvnB1f7VCiW0AnNwiB0DP/L9j6Oml2rWlJnPB3YJHWtc462bdnb26dtF/RJWa07Vut1iVm+aIzfU1R7bo1np13p3N+mcaVnHWcTXtmaU8c0gTuMrbHwwo9fHi04bjhJXAQn3Tcuhd0QSHdwASwT1nvpnS8P7c9+8a2h3cUxs+7iil/bq8o+O7UOTBVfzo+vxt5iMW5343PRyancNfbbWW6wMh25Jk0/+dK7TfWe0/F+Nq0mp6sBL9IPZ+3zvBZbvkk45/De0zYtPjSIc3QxsdbIuu+JKWevtNOl0qeLS4/e52BYzG+eb1g+lN+lLnidxzkDKRbS8vv2GvpaXsFhEWEnJ7zzZJfpzxMh+lTB47RTWT3Ozt5psJSevs+zUNtQDeK6uOwucibSQBgSEuVMuXnfpCVZEJJddkuSjqiWs+lS3HQdYCUhUDXMJYYyKOgoaG2EXA4uu4ZZKFdy0DhEfHbd1UB0+zzcf5lD77nje44PjjiOLQ+OG+KDNXr7FuYFHwyiYElITvFNtgDrsceHhAuQYs6eK15Gl11XkhyJjW66RZbZcNONAHaK+24R7j2YZndbCTXLruAaQ1Nx2W3zs1ApCaSCkhaJ0JX9BYIp6h3JYBFbfGo58i0mRwRJ7PkecCQLBC+os5LZM9BipQ6oFSEPYvnrxKgOswLULqq1P6XsN9YBFVJJJOQlW0aRbEHVwk2+1D1FxthWRYbthhAs76+Sky8lBMWV8zgiECQf2yOlLcRyHkOylbXwVyp0ipRsxTDccxVQQ+zo3YLkPKvG41RxqWcRH9D5xfmSyoyLw6xkxVb6vqdpGkJoWEXj6PiY4+Nj+j6XfvFnDeBXR9zTm1If5U5aPY58jvke62xezXv21Nb0W/RdNMvuucKocaFz1J23b+Vx3Xd3QyCdMWPGs41tyefKMVpNRMbyJ4gjqRFRYnXXnRjKdkXfeVFF4rZIOq7BRi3ukFRm64R1Qplat7Y1w1eGiXD46GduQ6yOFYtTjf3NaWDGZDAbOM86N+yzRdNp7W26dxxOapbddnDZNYzkiiupsyG7anA2cdktArdkJV0Vvt1Qw7PEmBbWkuLHaeXgab6dQTGSqmXBQeyy+5kTLBrOZRdPkURLx914j855HtqCBw8Cx8fCqhNSaICAJF9KlChIwsWAxeKmu+dARxdcSznuc3DT1bLgNE666da2Vn6SXDqmME110xUE6xn4KWfZLfJ4P7qTVpddJ5CCYMnhek8KlrPsqrEyh6rgo9DtR1hE7hL5sggdgVVqQRTnekwXiGYhUXwELa6sRamTy79kXtLCT0Zd9LmBb7JaoFibC2+kkk03d5WVWOHSxoFAnLSH7YxZb0FIUhLEGSSJRSDNWYTBYeZIMlpZk40jnNYCM1Iz+pbtxRysuPwsye2qRls1iyI2RxZa60k3rJoXSS7y3KU/vcY5dJqkCi1ziPP4pgXxqDgSsOx7jtcdKSpt2xKcQIyb56kW7Yl7/y7DBFLlCREUT7t3yP7+IYf7B6yWqyyQ2chZadIf1fjypF20kWvkPMHbTipKXXWProJjmWtrmahMc54ToIQEDC7VU+V0VmJepNu2sxWfRtcm2Zd/F2aBdMaMGWdj17wXJhazQcAUBsGsbjerA+coij16ArmmG5yQNaXw9B0LhMGdz4TBTfBEB9Usl8P5H+OezifqMc5zxgmr0FonflMMzfUepfZYcRUc7HLbJ9jEaWRPra478Q4/bZhtLM5PMwo/sju3dyjP6nKPa1NtYIU3BzdcqbZUwdSy90KalATZiu8aiN8m4iL9uLWPbf09/QDOf1BnXHc6HtV+GBVHNjyWzVNPss2eONcpmpPJRTZJmJa3GLl9aiOxrf0GJUM933BuG34bb7UI7VbKzmCopWwtLb9VLlWrLr/1Lsb+3rzD6T3JiWtaoXak26g231EcfwZY+bx3aWeIL/NimVSkuJbXJ177vSp+z0ycN4zz1wSzjfjRs8Z9MTt1h6li10SyQtuHrNQu2ctV9SS/cEN6kNPGHTvFElrXCJoVkbX2r7GZdGyDD08bX8+57qPoehLsdtkXm3yuG5kTQW2DDHNu+HgXhs84GZ+2aHrymzjh237Kp2YEPM8VYLpPNavXT3I2fMS54dPFfvjEFIdPrgtWXIVs/FwE23Sc9rn0eUTGz0QBeL//8vB5yzY/TO9imtpexoX/RZ79dUBEflJE3hSR35xs+xsi8nER+XUR+TkRuTv57YdF5BUR+YSI/LvXQuR1YagbsLlCq/2fWXHidnOm682jl6oXwTDG12ud9pkuuc57v08MGeNi2abDzOS9U7Py0WIV1jGZA6dc7gITjV3g3yOHtmGXrXF0sPLk35JFovV0tqazjs7W9NaTiIzu82cLXWO0z8nPdUYCXQePqkFaK6FbE3VFsoglwamCGimRa1qaERVcGd6SjkaMlKCaPFWtmMOyyy4GomSX3fLoTfO2Gn9c3US9y5l2gzNc4xHvc1xh8Ehw2RXQPFH2uL94kdbgkJ72sAPXYxoJqQenWAAXivIheTQoNIL4gC4DqMP5POaLN1xb52RwDZjkRZjUtie3nYHL7Uq4ayVbfUvbXB7zJQAuv60uWNluuCa3kxihEaSF1BreK7JIpFsJ7xPijBSEIIp4JbZK2zVI33JP2uI+G2ncKmfZVYeWOTcK9OoQc4OrbX1/o9aSMK5YKnMvpKEtm23NfYZJWcZkC4lasSabH5Y4qBvHD3WoGkkhaSApRM10daqsUs9xSixTx1pXLFNinZRelaSOqI6YhFSSOiUz1DxaLMamDiu/Ydnsnku4ZP505MzoyYxFv0QSqDX0DpIoRs9e/yVCv3rsIXueR8uapibAEymDShbC6torAc57gg94769znXP2OmtKw7BeO+NEk3WkkO/31OtVNVDJrNulSBfjmGX3qrLWTte6l7xEVeIOtE/bZvSxJ8a8Tj+RKdg495oynTMv0eUnk05d/n25kEAqIndF5EOFYT8mIn9ERN4hIr8gIr9b/r546avPmDHjMvi7wHdsbfsF4N80s38L+B3ghwFE5JuA7wX+jXLM/ywinovAzvjsIKolNGkqk6kWd7pSU6yKKddA/1ki78YOZ+wxCo5VcLPJY7eNPTduZeOimzaF07Snl+rLs96DJznPILRXuo1kiV7zAnelS1a6zAIpHZFYXP6evANHpcHW5+ni73LFPOoEfOuI7YIgC7wE8EZyLgtYRbhRye67WsqK+Jy8FZMsSJpkIazmErISf2kCWuIyzRWZVPJ5VDZUeFlwSUIyQaNiSXPpk1gkmZRddoOuuHX8ZZIl1jjismHfLzjc38Md3EFooAftE2YRXER6gd7QlHB7CXO64bKrXek+JbvgWm0zZAPVnuHd22znBZlYPTaPExrzOca2lH3ydgfE3rDecL2R1GGdwx07kvrsapuMKFn48p1D24Rrs8uuF4daoNOmvNsdlhJWpMMgiUSi16wRiJroNCEWSRrpNYHWdgSNqEa6sn1oW0Qt0VkEi5hFOs1/zfKxVj69RjQlNCX6lBfimhJdSqQU0dizjJEuHtPHezyIxxzHh6zjfVZpSUpLXDqmTwnTnqArYupJ2uNSl9sp4qwjaiRpj7eePtX7UfqkdEkxTSSNRE2sXXYVNuuQpDgVHA3r9kVS0/IE8tGV8yjwZGPmdWGYHLUo/pSkEdWEB5oQaNp2sBxqrYn0lDDMeY9lkBjdBobjzryOZR4r7YTmpMBiGJGUemJcs14fE1VJZV6yOhg88X1uKXKfECoQBZLPyl1MMIW1W/Ew3OON1Wu83n2OL8TXeLi4R98sc/pvF0FSzpJdXkh1lDniyWjKIVPj98e5z4taSP8W8H+a2b8OfDPwMeCHgF80s28AfrF8nzFjxhXBzP4F8OWtbT9vZjWo45eA95X29wB/38zWZvb7wCvAH7o2Yq8aRVtqZTJSVbQUVZ9mnd1NbEpC5w7bp/1YDt/IBCjbB+3ayodTJ/a8FFKiRTrtWOu6WEc7ksXiwGsnjn28nj3DUvoUX5Nr49HqCXDClfm09+rk89vc8yzY+KbK+HfzDGVDDTC1yU5W/qt/NQ4WfEFogqcNDS6E3C/DorJYaarUa5NTThaq0zXrxRwAzthqm4snO2XvEyohmyi5SvsE11VztMuCfi2zIhRL5nCvxTunqpGK6zqmxdOhPARTjGq2rtsVqzVwbLONpXLuai5NQ9tsbNcaOjK08+9mRWi1iFpHtBW9Len1iN6O6fWYpEvU1pj1UD/0UATh6UcsFeE4DdfMv1VaK70JiKhYud+ULUDFQGbis7bkMTHPoxNsKecMI6WYFQhA8J7gfXlnzxAUr2iaPYv/8jXPsN5NvYKmx9nw36aSV6zwhhY33TQZC6zE3D+FufQpKyWq9152gigCKRCJrGzFg3ifB/0DHsYH9G5dYq4Vq24ulZDJsG2T4fxcnLbP4MS1IZFeGo+MIRWRF4BvA34AwMw6oBOR7wH+RNntp4B/DvzXj7ziWQusXcPEvP/W2+PYdfdgf9wnRW4CZ2aZPWMBvjGInEiHNWlPDg9hvP9FGBWCu5yFcnr/ypgB+RPHnxjat92djWPe2YyGfTceMkm+ABuLvCd59lePPwf8g9J+L3lirfhs2XYCIvKDwA8CfM3XfM1V0vfkEDbeWTNImrMDxtijKeGd0HhHcJKTqgzuvdeIyQRYpUXhHC3ksH+5QZMycWwekJNkVhembN1yrpRfcLIxqeyiWLqZyTqPJ1EjS12yTMesWeHM4fHZqofDpFReFCn9eLEsghehIdPxbPHoe977PtJaaVxPv+jxFvAscCWWVBHq8J0UQrnfZODLrcbSlmnboDcb2yql3iRELTUpiwLEkZN7iBuN3hpy6Eou/RGyi5z3WDLWtHxp7wUsKm3qODhIhEWDawKNdogEzOXyHqYOkmABfHBIEnTtcUFwIZGS4XzJphtBXM4YXGuRjll2rWTZBTDEy2Y75ofh2uzGamZIyOtTtew+bJoXfa4xrGTZDS1QFoPeRTR4kgihS5g5zBm+U9RBOoCFObwGHro9kAcEiSz8Cm+CqEcbxaGgQh+hKfU+NXNBttaWOqC11qi3Wms0Z7qt2x1jPVIB2km7mbQDOZMuQs6gW8aonEYovzzBqqXICHbMQ3vI0o6J9gDM4yRwW9YYB3QccEtyXuBonoV0RdnkaMu4lttjLGhb1xLiaLOGDRWfx2wUp2tMGnCe1HjEIl4je/GItVxplt3nfx6dYmCaLHAeHx+zPD7OAmkIhBCyEsm2Z5QnD1eSev0JxqQ5kzjR0j6x/yRvBJPjxo2cW6u0Lg2kZOsPIdAguS6vyMZ5dglDPL7lTNw1oVvsI8u05H66jyZw0RPj3YHzHzXXbczPZ6yZhJIQ7wqey0WSGn0t8AXg74jINwO/AvxF4F1m9lrZ53XgXacd/Mwy6YwZzxBE5EeACPz0ZY81sx8Hfhzggx/84A4Ov5wx+AmqSkyR9XpF33WkFAlOaIKjDR7vamThbtzW6GS0jXPos3HtJVuC7mgNroLqCZl9dzC1dBVLkJLorWeta1a6ZM2aIIGGltE8Nk6UNaXJefdoW8dcCNfw0J4Wj37zt3yzuVbopaEpcdKGoS4XCXGSy3II+b2oCaLcZA3hGRdqOeNufq6+iELZlbfG5Oc8xyqOWmNyyMSrNtidLea2lhh+zEp8YML3Sw6/9BopdXQI1i245Q+4tThE9w4xcZAU64vVTAzRkF10FXyTF8sp5my3ZjljqwjZfVezGGXULLsZY7sIrKWd1IrYJaS+rmslu+mWESPFyf598USw7LLrJEdBx8ZBdEj0RA9iOQZURTCV7LLbgF8Y7wjKF5wnOUdiD4IhQckhrj7LBb7PJTdNGMX7XG7DzBWjhmEDZQztDQYrbbNx7LOSrZp8+sE9bjMPdnm+5LI2qkoiEenprGdlkUQcxhx8rrHqxSMScHUxX5aWntGSlXl3pHtIZ1XeW5soOxBHFIeI4kTxIgRxONewat6BEq+k1txzMY+eYWQ48Xsd8xRwkJKyWq1ymZduzX4jHCwC+20glN5y1zGzbMeJTto22efCo7vZKLgWAa66lwqCH+qv7nF4eMi9+w/wKrgQcmjtKSUAL5wh94rg6lgmgomiTkkp8UDv86X0ZR5yROta9kLAN4JLdfxyJTwh0+8m8aX1dRjm2jKvnHZvp8WXnhB2H0NXcRG/hwB8APhRM/tW4Igt91yzE6qT6W8/bmYfNLMPvvzyy6ftMmPGjCeAiPwA8F3A99mo1voc8NWT3d5Xtj3j2BzltCYi6Dr6PscqZQupJwSPc3J2PqNdwbmTmY3GU9twNioTUrVQVWFUNo7dOdRkCpbz6aolovX01tNpR69djt2ZJDoZ72l6b5spjM642NXcw2PgKni0LuChCoiTRZJsLi62be3ZRZbqtVUUHaPYb1vtelQVchi6UYbPWBdmTFg17KcJXa+IKZLUEDx7oWG/aZEQQMjuqakm47LiSVfcBF0mtD64qdNDVm6MHgFmI91TY8l0n5H+E2vejXblOBtPnJOGaf7RLAuelibnmpw7m1FBvNC4HNueS6zkhE8yDE4l29KQuKW2876Zv8txNVxhYp42GRPUTNtITkzI9rFsnqcmg1NxozugCElyMqGeWD49PUaPZUda8cWF1mMuYM6X85XtzkH5PV+/tjNt5nLbBhoZf6u0uOqumxPTqCvW9KeM52IevcyQt7WvpsRyuWS1WhH7njY49tqGRRuGMkfXkQX10eP6Ji52y6fYBmvCI5HsmhwCTdNkJbZMjnvEBW5iaVHHbFc8qUyUpD3LtORhfEhHhzrNHiY1qqNmOy9Kp/FTvcieAmGTczxOUqOLWEg/C3zWzD5Svn+ILJC+ISLvNrPXROTdwJsXuuIzszAcCf34xz82tN//3ncO7WYa2j5x97xqXMRN4kxz+9axOtlv2p7uNx37zzb5T3SsZ+xy1e6+EkdN1ls6ulm/0b0+tN/TbHrbOJ2ywHgP3vS0zTu0xM0Qke8A/grwx83sePLTh4GfEZG/CbwH+AbgX94AiVeKFCPdes3Dhw85Pj6i79Y0wbG/8OwvGhrvJlbS6xl+xrFfNradwNTaedpJBp1l/uTd7ZTdZDKLFwljXLpv7302PZzH308OGUy9ORtwJNJbxzqtWaUlS12SJBJocOIILuAJ5JyjWbNbaTfsNI/mR1pPbwJPm0eTQlopje/pFhFHIFiba1hKFsS8y++MGniFurByRWjM7rv54aXpds3WU0RKW4YknKKTV81ybTukxpkK6v0gOKAecYIFBythScunFy9wvOqQ9ZJbB3DnVsOdw8B+6HAlhrBvItJ7pM8Cky+x4DHmbL7SQLSYE5Y5ISXN1/SOVOqSbmx3QkqZJ5yT7MoLOO9yRmEM52t24ezKqyWzvittNZu47CZCUwU2y1lhF4buGa7LNUBVHKxyOE/cV1xyaGy4Z/tE7iMSaV2HxBa1fVLjifhs6U1CY9V6q4gFBI/Sk7PSOky60vYgPTld8GltAQtAbXuQXKFWzGMSAUMsoNLnl8tCETcV08BSe9aWOEqRY8teDIrQihKI9HIblX162UdZYKYkTexJIEeARtriLJwk4S2736v0eGvymOB6QrEEi0Ry9KLS6hG9a+mtofHQAc4S7UmuMTUAABzFSURBVOoLrGQ/K0GeEr6i59Gy1uv6ji996Uvcu3eP5WpJ0zTcuXOHO3fu4IMv04oh1xiqJsVz41yp8KzfZLQETs9hlucfyN4jiazocG6i0CuZaheLRbam7hisKurIChszY7lccdQdcRSPiMSsBGuacsAFEkVtKbQvG25WraZPEqb2SIHUzF4XkVdF5BvN7BPAtwO/XT7fD/y35e8/fmwqZsyY8UiIyN8jx22/U0Q+C/w35GyAC+AXihLhl8zsPzOz3xKRnyXzaQT+gpldo9rkarDtQlKto0dHR6yWx6R+TeuF/ba4Gk2E0bPFlKcvvlxsqfQ415UTzWILGmuuwbDIOCGPPu6tPhXDa6XJBstoZx1rW9PTk0iD4OCdx0vA4QFXQkdHgRZsouCaOvKeJPc6cR086pwhjdE7T4MHgeQUGSxgOlg4TXSM95GSLEcEmRSxzzRloU2cDBpFcROFiIwvT/HwG5KcZOEU1JTBvVSyFdF1ghFxy3scfvITxOVDOlWavmWvPWRv/zYp3MZzjOgKlpKvEAzIpUmygF0cuBUkObRcNLvH5vsTczkFlma34rx/cTEGotqg1Ii23c4L1JRqW4joYDmOMfOXAL0azhSHEIPkdJdRsospgtNccsYi+GXOXeqcclsijXhUArDAt1mpsKC4XXvJChl1OK2u0pm/tSpgJm6LAztcuC1jHB7VEpudZBUPJWY7WeY38y0urQgaiVbiWMURcDSuZeH3eMkHGmcEWRNsHzDUZ6USGEqiRIhikhDLC2SV8mylZIYuFmIjCzvZOtpmZZRYcWt2CA1p727Wuz0mkz/X8+gFx/iNqaK4tHbrjjfffJ2HDx+gqeel2wtu7wf2WpcTTtV9bxRPOrLLhrUwZ6g1Yt/TrdesV8eY5Qq7rnodnTCkZI+QYV1xyiO5cuNL+U8Q1tKzYs39eI8jfciKJd459mSf29xh0e0TUoszPxytg4xt+GFelo2w7MvfwaY6+3GU2xexkAL8l8BPi0gLfBL4s+Sx8mdF5M8Dnwb+40tffcaMGReGmf2ZUzb/xDn7/zXgrz3xhesgdVNz0VkW9+Ku2/c9q+Ux3WpF6nv2vWPRBNomZJfdyUlGEWa3rGgXgWz+twmb7DP8neg7zzaJnv/7WYRc8uFtn94wokX66qZLwkRxzuGcL/WdPa7EzRWbKNMMgaMIOsbUbd/TheXop/QyXBuPOkhOaE1K+Zbq4mzUlK9VfDSzkQfKsyz5qQFG4dSKFbvui42LsXLq7E47Hj24qFbhtB5dhWDNQrCul8QvvUHXrehVWZhnr1mw1+5hvsV0ma+cADHM5cWvFWWEE4bruVIQ3rQm88nePb5QVbP41naFDpZHSGb4SduRBZykVspFWbYclzcrW5TzG5eKxVnMcm1QFSSCBimCOlgV5tMYCe3RwpEuLwMdOG85Xq0Khy4nlXIOEJct11Dsh7kfapkHqW2227K1vUTLDvtUa4Yb2lbaHlfSJGUh1bsifIvigYBnIZ6F22PhDjhwLU2xZOe7T8OCf1PwLSqSIQlNzRJd3Yez6K3jUSTnh3ZVDCAO823O2PuYuDYevTGt2AX32xIuuy5bSFerJaaJ24d77O0F2iClZMo4h97IrV0mbvTMc1AUUeN7ZmTFdoyRru8wtZLnaTt7/eQkVcdX3ks58fMVry6kuuvmzLprVhylI1a2ItLT+gMWfsG+O6TRFq8B0az4ytmry9iPDIq56bz+ONSf0FU8xkkuJJCa2UeBD57y07df+oqPWgDdFBNvoev7of0bv/qvhvaf/ePfMtlpdF2wMNXO79ZSd9NNd5O2KSv5SXmtjW6I4+AfDsZXRtJ4rjqh5S+XNPVPA8Q3jt0275TWxuknlE5cLj758JWhffvO7aF9d290uQbo1uO9Ve3sNhptJ9ceL+4n95wmtE7p26034flBLfXSdR3L42Puvf0Wxw/u4VPHwd1Dbu+13F60NJJ179nnsBwsm5NRbd88tiSoyaRRv7ryfSj3UnebBMi5muioLgBtcq/DqvYSb6ZsjhO2ccIzDzn5rRymAlEjK11xpEc5c6dEJAhNaGlDSxsWNL7FWQB1pGmA3iB927C4yBPr1tg2WFFGjcp2f9dFxbPEp5og3Y/sNYn+QJEU8LFFpcO5BnEtSkd+AxaYxCwOmAdJ5P53gyBq6oYEMdk4WITaRK7VhyFqgzIg1OxAUjK6ClgASfl9U2+5k4HUCN09441O+OeN582uQ9Oal24fcne/5XYTCHIf65bE2NEvFFkHpPegDd4LzkNnMmSU7nx12fWs+5xhV7zQpZIF2OU2kjNP9ykvoZ0X+jLcj23DOUcslVdcyOVTjXytZCUxUCiCqIELuYpfNHIMaSBbdCOYCCkI1uX4y/W+YOZYaeBztmDJF7OA6g5JqaVzB3RhPz/4aGh3i0ZiTlJmTekQQ/WA7FyoGAfkHLkJtX1EbNieRTpFi7VSJKG2h2CIKFoslI6EsZfbohg5a61YQvUANejsiPtJOFJY6tul+Ion+ls4fQcLXuSBOwRd5L5yx2ANaIvzXRGEb+Ekkt+UO/jiMmy8kNti4G4hls+utEW5kmj6nuSanKzLJQKJQESW91G3z66M2s8mrCYsLy72xtHDh7zyyis8fPgQ5xwvf9VXcevwkKZpUF0PCqfmmj1Y8xB+0b4ehciqdLFSO7WuM0UE5zymWpJzuQ3ltmoeO7z3p4THCVeZZfaimJZ86/ue43jMg/SAnh4nLrtat7e5FW7l/fRkpxmbXlV5CbGl2Z7ufw3W8YtaSGfMmDHjYriCcWt6SlUlqRJjzq67Wi1ZHh8R1yu8KAeN59ai5aBtaZw7IxHDqOe9uWXNlkg8UVTC5j0LdUK1jZBRtv6O9g/Z2jbF5MyPuvkL/V7Vqo9+kkqJH7VIp2vWumKta5Lrc6mm4laqFona5b62HtGpUFkFdZCSJMYV286Q5mc7k9VpAvSj5eqdhHMge0LnAiFml8feJ0IpEhIlETQ/B3WJptxgEs2esBgqWix+2d3XlcVmcpbbarldFi3JGU5zu3M5LtUr9KK4BC4JGhRRwSdBfa6XIutEjEfY/Tc5/MRv03YraBruLG5xeOc2izu3UFvgpANR6FzWurTFkig5uQ7qUUnFFbjJLp+iEDwq2ZJrPscnmujQ1hPtEp253a4WOkk5GQ+S4yotV4zPrs4ewZFIuFqyxUdQnz9SnJk1Fx0Vy+VfRFe4dEwTjzBNJHNENcwvcd7TlPzAJkb0qbin5vfXuZItyaVcGqbUIi0OyTjRLGxS3LGLQCriy/Y0lEBzkrJrLiCkQQGbHZ1DXoRLItoas4TaW3R2n5Udcaz3AIeII1jCXHb7VjvGky2kcVLDUbXPChEJCAnB4yUrdkUcjiZnKhZBCIMlPiGIJLLk7wlOweWzhqJkS4tbWUDZ4RJ0wCSN8Nb2y0w4V3WLo4Ecc3B8tObtt9/m9Vc/RVw+ZF+M99065J1N4AVxZTxxxfqnE+XmVKCxJ7Ji5qHp8Y0Z0zk002KFX7L3SCYv7+CMslVyEsQQaJuWPd+y8isa8TjJ4SI5K+2ELmcDnZcNYz7L+CLlWdr4YBladakyuZhJqc5rxlE64n56myMeYhgLt8fB3iHBL7IXgnP0mohhjfc+95+dNL44PK01JC313If+LB4chfY0pXvyWPQpvKyzQDpjxoyL4SKD71WvEWySWbfvWa3XrFYr1qslqe+QRthvPAdtw0HbjDGkg/y56S2wMzr2LWH09B2mzlKj6Ll53CiMXo+ovSXcbvV/fe7FFkdOdJKIRDrr6HRNVxbAiA3CaNKe3taI9dnVyOoEXj958eHE48jxpr6kRsnbRjfFStwGaTu+lj0PIiABenE0fYOSSL7Hm0cl12D2muv2JZRQvANUtDxDQSULoZCT89Qsi+p1qNtrrmaTNVKxjnozUpEFfBFaUZBoJG94NYjFLSwZsk6suxXxwT38Z14l9D0SGu7uHXJ465D2YB+1psQKeywGpIngs9dNdUTOi+H6BgUg5oWmywlwsnbBFYFIy2rbiiA5bWeBbNrWYR8hu5zWyM26andAJJcm8cVN1+Esu7aiDmKAJubYMpXisitISqARsZ6ga0S1ZCRWxK1xriHYgoFC1+Etu6qrBDwpC95a3IclK3SklJ2p7VysR8uz0pK0JQusQ7sK5VWALW6vWSDNhYK85MVqsg7lLSL36e2Yzo4R8aUyahba1YHSZAurKCnl529iRNKgKMqqBQ8sSnkYj6PFDWWKUo5/Lj3gSmK2KA7vFHG5NmpOsOWIIff/M4PpsF1xwxPPYOUjk3e0XPH2vXt88fXPo92KvcbxB27d4sWm4ZYTfF+FEoaatfk84zlyY1Iv9JqxnTbBrIYcTFWYkxnUauZgR+MDrQ80oaVxHr+RSG/ifSNZ8BpiRJ9kHtmYL23yd/L0aibg6mZcdtPSVFOWuuRhyrGjToTGNbRhgcMTLYcAJBIqEe98iZHvx/CLYm1taGispebzHB3/S/z2aQLpE9z+aXg2BdIrWkxMlTO/9ltjZt0HX/j80P6j3/inh3YziWPQja65+dXO2Zl4t9zaJm6nMkksO3VHdRNX3mni7w3tyPTgp/SWykm+PHH66T5f7N8a2p/vPjO09/1I8++88RGmsDSeIKVuaPtmdNN96WDMzPvOvXcP7YXujXTY+IycTWxyu67FfQaw/QRjH1mu1zx4+JC3377H/Xv3WB49xOKatlnw4sEeL93a58WDBXveE2R7Krp5jArPLW3ogBoHt3lMXqvVhArTf25oMfxKkd+qADdaY59KjItw6qi3IQ4PgnZeMPeW644udcVSjzlORyztCClxfC4qLiWSrFjzAKkxpFoEWhvdr0QE7xq8a2jcgtbtEVygkZYgDQ6HmR+ficgYi7Z9IzswZl8UmsA9jNzyxmq/Q0xoVwK6xrlA8EKy7LrVuJZosVgCGqJll0lHIEosVuKQE0phSAyl7p5BPwowzgJmSo8RNItPPYaL+V2KXqEPJMlCbb/Oi6Xoej776TUffxX+3/VLHNkX+QML+NavvstLi0MWusgC1NpB58F7tBdkbWANEhTxRsIPZSeiuqx0MEh9QJzhXCIln4Uzx4k2GN7XNnhvp7QF5wwt2WydD6hmodO53EY9Im22Q1qup5rXiop0bXmzjBgXxBTpSCS7S2JF9J4mNDROeLHpaWgwGtaN5sRGDtwa2mAEMZKsc+1NM3rtqcvknh5ffAK6Sbsn4nAEcXQWESCU7bXdFctIQIilz7MNM2EIAcc69dk7oQOVDpUVjRi4CC4hjefIH7MMPbdsj2hKb5EDl7k8WhqSbakzWhY48YhzHLgDvDSIdOzLPk48SRIti6x0IOYkSCYElpg7xFwDzRFYfv6H/UNWruV6ipA8nzAMSzmSJZrxuc99nk9/6lP8/u//Ps45XnrpJb7+67+eF154gbZdIP2DS5zcNhdml8UlBdrtta6Uc2z+XuZE57JAVmLep27L4lyJ5WbMxCvbs8UFstU+RYjIqVNTjZGPsWe5WnHcHaNFbYPB8fExxEAXFe88mkBj9UjItZ77vke8x4eGdtGy5/YIEnDicJJD+GqGYUtW5uCrxbMpkM6YMeMrFlML6VB/NPagihdj4T2LkDWevpa2Y/dEjilNZ03AVajb0GvY1m/D5HvyLKef9wp9VU+zkk42Wak9miwSyydZpJaEUY1E6fBQ6qPlCbkmzTGmAqkrcWUJRAZrjCcUa1GN9TltgSMDRc8ejB7NsZHrFlCS9PgiNJquCMmTnNGxIsRcjqULa0LMwlkXOkKfOSOGjhCz4N83uS0m9KHHJ8GpsA4dIeUC68vQEZIjRMe6ibgIoRO6RURUCFFYSyLFRFon+nSMcMwLssbdbnlxf4/F3XfDYkFyRjruoFdIQm9SlC4GXnEkRA1zuZyIaELFZ8ugJVQCYglJCcNny2MsbrdDu7jmntsutVdTfpcMkBizy5sYKebteYGYYEiJVC2eDilWRvCo9sXd3IGu8HTstYH9tsVSQ2hfJIWG6IS2JkcyIXlPLy6Xjqk0iRBdKDZHIRljO9tQceJymLxkKmLhhZSXnwhZaR7LWJGjNItVXPI5ASKJtcslVpLvsRJv5yRka7RzJO/A5zqhHT2JbK9eOy1Wm8J3wvD8nORkUWsHQcCL0JWSPlIi4x2CFqE7L5zbbGkWI1nI6jYBaxakWrN0l7Hj9JlAUqWLkU+/+hleffUzLB/c4+7t27z7nXd538svcOgVr6tsQR+8CE4ij/FP44av+6EZpjqEAfWa6LXEag/Ky7Pv+8KQQUV8xu/1CY77bMu9goLLvLHya9ba8fbqbe6ltzjmfuHF7Otwb71mHd9imXIJNVVDkxGCz3OERfrireHU02pD4/bY87dpFwtaWXAgd3K2e3xOilQTr1meQzKRrj7Fp9J1s0A6Y8aMZwo1oVGMOTNejD0pJpzleKrGOxrvCN7lOKUdFjoedww/+7jzPCNks3XNUnp1uFSMZEqyRCILp86yW5Kay8IqxU1qWyCtWusikJZ1PM4CiWwVVXKpE6Nmfp1m5N3xVeJFIDl2MybYWwfUdyTf00iT60Baj0v7qCmd6zmI+yDGyvccxLyAWPmO/bhAzFj7HoktzoQu9Ehq8UlYh55FbJAkrH0PKdD2gVXo2UsNvnOsQqTpHf440DURFwV37OlaJXWK3leSLvFuyV1Z09wO3Dm8hb/9VeiiJYrhVz3SW87rU7xwxWWBVFVBDXGKaSrmYQNVsIQEzUlLNLdRxVJpW22X2NKUkFBcdrfbPlv0LCUGLZamSgymCXGCuLKPCCIeUFCPmUPcGiELTqoJVcW0WK7p2Wsce01Lcgtc+wLJB6KDhWaB1BDUFYGULKylMnrV97eWbsmvwSSbLmCShVAR2QhfrFl2I1PPCBkWkYnq4WVAohOhE4i+z4qeUn7JXAMukPzo4txJKg7A0Jezm6vnLs9OpMSZOjoRVCQnmysCqWO8H6pAiiDSFKFWUfPUuFFt2g23wZ3Frg41k3FfTelj5POvvcbrr7/G+vgh733Xy7x89w7vuntI6xI+RTYDYs8QSp9cZrv2Z2YlBEg11xiOmoiqOCkZus9R9l4Gj7L5Vg6YzlSnxb/U7L/R9yw55q31l3nIPY55CIBKzHNqd0SXjFXJP2flHlvfoKY5gZMVoVehTQHv9nDNbW61t9jjNpHs/hsI7MlBURI5Gsuu+VISvmXans5iQq7Z/PwF4Aj44rVddDfwTuZ7ft5xmfv9g2b28lUS87gQkQfAJ26ajlOwq+/TTNfl8KzQNfPo5fGs9O2uYKbrcph59MnxrPTtrmCm63J4Ih69Vgupmb0sIr9sZqeVkHluMd/z84/n6H4/sYv3savPd6brcpjpeiqYefQSmOm6HGa6ngpmHr0EZrouh+eVrjkqfMaMGTNmzJgxY8aMGTNm3AhmgXTGjBkzZsyYMWPGjBkzZtwIbkIg/fEbuOZNY77n5x/Py/3u6n3MdF0OM12Xw67SdRp2ldaZrsthputy2FW6TsOu0jrTdTnMdF0OT0TXtSY1mjFjxowZM2bMmDFjxowZMypml90ZM2bMmDFjxowZM2bMmHEjmAXSGTNmzJgxY8aMGTNmzJhxI7hWgVREvkNEPiEir4jID13nta8DIvLVIvLPROS3ReS3ROQvlu3vEJFfEJHfLX9fvGlanzZExIvI/yci/6R8/1oR+Ujp638gIu1N0/g0ISJ3ReRDIvJxEfmYiPyRZ72fd4U/z+GjvyoinxORj5bPd94AbZ8Skd8o1//lsu1G+11EvnHyTD4qIvdF5C/dxPMSkZ8UkTdF5Dcn2059PpLxP5b37ddF5APXTNffKPz76yLycyJyt2x/v4gsJ8/tx66Krsti5tEL0Tbz6Pm07ByPPi/8CTOPXpC2neLRXeLPQs9XJo+a2bV8AA/8HvB1QAv8GvBN13X9a7rHdwMfKO3bwO8A3wT8d8APle0/BPz1m6b1Cu79LwM/A/yT8v1nge8t7R8D/vObpvEp3+9PAf9pabfA3We5n3eJP8/ho78K/Fc3/Jw+Bbxza9vO9Hvpx9eBP3gTzwv4NuADwG8+6vkA3wn8U0CAPwx85Jrp+neAUNp/fULX+6f77cpn5tEL0zbz6PnX3zkefR74c9K3M48+mrad5dGb5s9Cw1ckj16nhfQPAa+Y2SfNrAP+PvA913j9K4eZvWZmv1raD4CPAe8l3+dPld1+CvjTN0Ph1UBE3gf8e8DfLt8F+JPAh8ouz9U9i8gLZOb8CQAz68zsbZ7tft4Z/jyHj3YVu9Tv3w78npl9+iYubmb/Avjy1uazns/3AP+rZfwScFdE3n1ddJnZz5tZLF9/CXjfVVz7KWLm0cfHzKMFu8ijzwl/wsyjT4Jd4dEb5U/4yuXR6xRI3wu8Ovn+WXabOZ4IIvJ+4FuBjwDvMrPXyk+vA++6IbKuCv8D8FcALd9fAt6evKjPW19/LfAF4O9IdlP+2yJyyLPdzzvJn1t8BPBfFPeQn7xut7sCA35eRH5FRH6wbNulfv9e4O9Nvt/084Kzn88uvXN/jqxlrvjawtv/j4j82zdE0zZ26XkNmHn00ph59PJ4FvgTdud5bWDm0UthF/kTvgJ4dE5qdAUQkVvA/w78JTO7P/3Nsj37uam1IyLfBbxpZr9y07RcIwLZdeFHzexbgSOyC8WA562fbwKn8NGPAl8PfAvwGvDf3wBZf8zMPgD8KeAviMi3TX+8yX6XHKf93cA/LJt24XltYBf5QkR+BIjAT5dNrwFfU3j7LwM/IyJ3boq+XcbMo5fDzKOXx8yfT4aZRy+OZ4E/4fnl0esUSD8HfPXk+/vKtucKItKQmf+nzez/KJvfqCb08vfNm6LvCvBHge8WkU+R3VP+JPC3yG4DoezzvPX1Z4HPmlnVNn6ILKA+y/28U/x5Gh+Z2RtmlsxMgf+F7B51rTCzz5W/bwI/V2jYlX7/U8CvmtkbhcYbf14FZz2fG3/nROQHgO8Cvq9M8pjZ2sy+VNq/Qo4J+9euk64zcOPPa4qZRx8LM49eAs8Yf8LMoxfCDvPorvInfAXw6HUKpP8K+AbJ2Vdbsln8w9d4/StHiZ38CeBjZvY3Jz99GPj+0v5+4B9fN21XBTP7YTN7n5m9n9yn/7eZfR/wz4D/qOz2vN3z68CrIvKNZdO3A7/Ns93PO8OfZ/HRVlzEfwD85vaxV0zXoYjcrm1yQP9vsjv9/meYuBrd9POa4Kzn82HgP5GMPwzcm7gkXTlE5DvIoQbfbWbHk+0vi4gv7a8DvgH45HXRdQ5mHn00XTOPPh52jkefQf6EmUcvQtcu8+iu8id8JfCoXW/mqO8kZ/r6PeBHrvPa13R/f4xsRv914KPl853kmMpfBH4X+L+Ad9w0rVd0/3+CMcvu1wH/EniF7P6wuGn6nvK9fgvwy6Wv/xHw4rPez7vCn+fw0f8G/EbZ/mHg3ddM19eRsyb+GvBb9RntQr8Dh8CXgBcm2679eZEn89eAnuxJ8OfPej7krID/U3nffgP44DXT9Qo59qa+Yz9W9v0PS/9+FPhV4N+/7v485z5mHj2frplHH03HzvHo88Kfhb6ZR8+nayd5dFf4s1z3K5JHpRw8Y8aMGTNmzJgxY8aMGTNmXCvmpEYzZsyYMWPGjBkzZsyYMeNGMAukM2bMmDFjxowZM2bMmDHjRjALpDNmzJgxY8aMGTNmzJgx40YwC6QzZsyYMWPGjBkzZsyYMeNGMAukM2bMmDFjxowZM2bMmDHjRjALpDNmzJgxY8aMGTNmzJgx40YwC6QzZsyYMWPGjBkzZsyYMeNG8P8DbO4J2BGDdMIAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 1152x288 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "udfxOyGyZwHP"
      },
      "source": [
        "train_history = []\n",
        "def train(num_epochs, lr, nbatches=-1, gen_headstart=0):\n",
        "    opt_generator = torch.optim.Adam(generator.parameters(), lr=lr)\n",
        "    opt_discriminator = torch.optim.Adam(discriminator.parameters(), lr=lr)\n",
        "\n",
        "    discriminator_loss_fun = nn.BCELoss()\n",
        "    content_loss_fun = loss_fun\n",
        "    \n",
        "    generator.train()\n",
        "    \n",
        "    for epoch in range(num_epochs): \n",
        "        d_fake_loss = []\n",
        "        d_real_loss = []\n",
        "        g_gan_loss = []\n",
        "        g_cont_loss = []\n",
        "\n",
        "        i = 0\n",
        "        for lowres, highres in tqdm(dataloader):\n",
        "            if nbatches != -1 and i > nbatches:\n",
        "                break\n",
        "            \n",
        "            # Reset gradients\n",
        "            opt_generator.zero_grad()\n",
        "            opt_discriminator.zero_grad()\n",
        "\n",
        "            ## Train discriminator\n",
        "            # Train on HR images\n",
        "            real_predictions = discriminator(highres)\n",
        "            real_loss = discriminator_loss_fun(real_predictions, torch.ones(highres.size(0), 1, device=device))\n",
        "            d_real_loss.append(torch.mean(real_loss).item())\n",
        "            \n",
        "            # Train on SR images\n",
        "            fakes = generator(lowres)\n",
        "            fake_predictions = discriminator(fakes)\n",
        "            fake_loss = discriminator_loss_fun(fake_predictions, torch.zeros(highres.size(0), 1, device=device))\n",
        "            d_fake_loss.append(torch.mean(fake_loss).item())\n",
        "\n",
        "            d_loss = real_loss + fake_loss\n",
        "            d_loss.backward()\n",
        "\n",
        "            if i < gen_headstart:\n",
        "                opt_discriminator.step()\n",
        "\n",
        "            ## Train Generator\n",
        "            fakes = generator(lowres)\n",
        "            fake_predictions = discriminator(fakes)\n",
        "            gen_loss = discriminator_loss_fun(fake_predictions, torch.ones(highres.size(0), 1, device=device))\n",
        "            cont_loss = content_loss_fun(fakes, highres)\n",
        "\n",
        "            g_loss = cont_loss # + gen_loss\n",
        "            g_gan_loss.append(torch.mean(gen_loss).item())\n",
        "            g_cont_loss.append(torch.mean(cont_loss).item())\n",
        "            \n",
        "            g_loss.backward()\n",
        "            \n",
        "            opt_generator.step()\n",
        "            i += 1\n",
        "        \n",
        "        avg_d_fake_loss = sum(d_fake_loss) / len(d_fake_loss)\n",
        "        avg_d_real_loss = sum(d_real_loss) / len(d_real_loss)\n",
        "        avg_g_gan_loss = sum(g_gan_loss) / len(g_gan_loss)\n",
        "        avg_g_cont_loss = sum(g_cont_loss) / len(g_cont_loss)\n",
        "        train_history.append([avg_d_fake_loss, avg_d_real_loss, avg_g_gen_loss, avg_g_cont_loss])\n",
        "        print('Epoch [{}/{}], d_fake: {:.4f}, d_real: {:.4f}, g_gan: {:.4f}, g_cont: {:.4f}'\n",
        "            .format(epoch+1, num_epochs, avg_d_fake_loss, avg_d_real_loss, avg_g_gan_loss, avg_g_cont_loss))\n",
        "        torch.save(discriminator.state_dict(), '/content/drive/MyDrive/Colab Notebooks/discriminator.ckpt')\n",
        "        torch.save(generator.state_dict(), '/content/drive/MyDrive/Colab Notebooks/generator.ckpt')\n",
        "        \n",
        "    if len(train_history) > 1:\n",
        "        plt.plot(train_history)\n",
        "        plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rjbwUquOZwHP",
        "outputId": "cbbb3c79-ee1a-452a-a100-441072c40621"
      },
      "source": [
        "discriminator.load_state_dict(torch.load('/content/drive/MyDrive/Colab Notebooks/discriminator.ckpt'))\n",
        "generator.load_state_dict(torch.load('/content/drive/MyDrive/Colab Notebooks/generator.ckpt'))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 331,
          "referenced_widgets": [
            "c64de03b6e8d4ef8a62324d55c9ca3ab",
            "5e405b6157334f8985de00565c8e4724",
            "86f14f9d3e0149b9abb382f19e7394b1",
            "e28ec762b2174d75a38f3f6ce4697222",
            "1f014c4d267249b9aa5a98710b0e63b9",
            "f961d5c692be43a5bb9ecadb1cdb444a",
            "1852e8f8338942b786acd7e4512d384e",
            "a15bfcca2edd4146a772d48e5f0dd0c1"
          ]
        },
        "id": "D-FA91azfLp_",
        "outputId": "2ced2639-954f-4870-aa73-386dfd20256e"
      },
      "source": [
        "train(10, 5e-4, gen_headstart=5)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "c64de03b6e8d4ef8a62324d55c9ca3ab",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=244.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "error",
          "ename": "NameError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-18-72e138efd477>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m5e-4\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgen_headstart\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m<ipython-input-17-ffb65e33e5a3>\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(num_epochs, lr, nbatches, gen_headstart)\u001b[0m\n\u001b[1;32m     61\u001b[0m         \u001b[0mavg_g_gan_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_gan_loss\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_gan_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m         \u001b[0mavg_g_cont_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_cont_loss\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_cont_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m         \u001b[0mtrain_history\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mavg_d_fake_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mavg_d_real_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mavg_g_gen_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mavg_g_cont_loss\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m         print('Epoch [{}/{}], d_fake: {:.4f}, d_real: {:.4f}, g_gan: {:.4f}, g_cont: {:.4f}'\n\u001b[1;32m     65\u001b[0m             .format(epoch+1, num_epochs, avg_d_fake_loss, avg_d_real_loss, avg_g_gan_loss, avg_g_cont_loss))\n",
            "\u001b[0;31mNameError\u001b[0m: name 'avg_g_gen_loss' is not defined"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 331,
          "referenced_widgets": [
            "2e6d5a4c8be945e5aa2d83885f13c400",
            "ca41a502284842e497deae777f197e96",
            "c6325c0e129c44b4a279bcf07a4649d5",
            "a2ae75bb715c493f8970e3d3dd5fe281",
            "c5fb444e0cfd4025b7962709cd4d124a",
            "340915558c81434885f812e55c58a9ec",
            "489f113ddd254f81acf09a52d63174f0",
            "f812199881394f989cf52dc42f154474"
          ]
        },
        "id": "hAORQTMSZwHP",
        "scrolled": true,
        "outputId": "4f5a2fcd-26ba-4e6b-ccad-0a5fa6231c86"
      },
      "source": [
        "train(10, 1e-3)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "2e6d5a4c8be945e5aa2d83885f13c400",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=244.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "error",
          "ename": "NameError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-16-7adf5d0ee538>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1e-3\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m<ipython-input-15-5672ddbc96f6>\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(num_epochs, lr, nbatches, gen_headstart)\u001b[0m\n\u001b[1;32m     61\u001b[0m         \u001b[0mavg_g_gan_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_gan_loss\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_gan_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m         \u001b[0mavg_g_cont_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_cont_loss\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_cont_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m         \u001b[0mtrain_history\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mavg_d_fake_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mavg_d_real_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mavg_g_tot_loss\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m         print('Epoch [{}/{}], d_fake: {:.4f}, d_real: {:.4f}, g_gan: {:.4f}, g_cont: {:.4f}'\n\u001b[1;32m     65\u001b[0m             .format(epoch+1, num_epochs, avg_d_fake_loss, avg_d_real_loss, avg_g_gan_loss, avg_g_cont_loss))\n",
            "\u001b[0;31mNameError\u001b[0m: name 'avg_g_tot_loss' is not defined"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "11YzTzGHZwHQ"
      },
      "source": [
        "## Things that did not work\n",
        "\n",
        "A U-net of convolutional and deconvolutional layers did not work. At best this achieved the identity function, which is of course useless. Using res-net seems to help, but again, this makes it much easier to achieve the identity function, which means it might not be a sign that it works after all.\n",
        "\n",
        "The first thing that was working was 3 convolutions from $3\\times64\\times64\\rightarrow768\\times8\\times8$ and then a sub-pixel convolution $768\\times8\\times8\\rightarrow3\\times128\\times128$.\n",
        "\n",
        "![First working attemp model](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/fwa.png?raw=1)\n",
        "\n",
        "The result of a single subframe is shown here:\n",
        "\n",
        "![First working attempt, subframe example](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/fwa_subframe.png?raw=1)\n",
        "\n",
        "When observing a single subframe, this looks very promising. You can notice that it looks a bit ragged, and there is also a small amount of color noise. When considering an entire frame, as shown below, the problems become more apparent.\n",
        "\n",
        "![First working attempt, full frame original](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/fwa_original.png?raw=1)\n",
        "![First working attempt, full frame upscaled](https://github.com/Ragusaen/FuturamaSuperResolution/blob/main/resources/fwa_upscaled.png?raw=1)\n",
        "\n",
        "It is definitely not bad, but it is also not great. This model works really well in dealing with "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vWlIgEm0ZwHQ",
        "outputId": "fccfde33-9776-4b4e-9999-ab7220ad51b0"
      },
      "source": [
        "# Switch model to evaluation mode to not keep track of gradients\n",
        "model.eval()\n",
        "\n",
        "# Get video\n",
        "cap = cv2.VideoCapture(\"/content/drive/MyDrive/Colab Notebooks/video/lowres_sample0.mkv\")\n",
        "\n",
        "video_resolution = (640, 480)\n",
        "\n",
        "cap.set(cv2.CAP_PROP_POS_FRAMES, 12000)\n",
        "\n",
        "# Number of subframes per axis per frame\n",
        "n_subframes = (math.floor(video_resolution[0] / low_resolution[0]),\n",
        "               math.floor(video_resolution[1] / low_resolution[1]))\n",
        "\n",
        "target_resolution = (n_subframes[0] * high_resolution[0],\n",
        "                     n_subframes[1] * high_resolution[1])\n",
        "\n",
        "_, frame = cap.read()\n",
        "\n",
        "# Resize to actual resolution\n",
        "frame = cv2.resize(frame, video_resolution, interpolation=cv2.INTER_CUBIC)\n",
        "\n",
        "cv2.imwrite('resources/original_frame.png', frame)\n",
        "\n",
        "# Create a new empty image to insert the upscaled into\n",
        "upscaled_image = np.zeros((target_resolution[1], target_resolution[0], 3), np.uint8)\n",
        "\n",
        "normalize = Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))\n",
        "\n",
        "# Create normalized tensor\n",
        "norm = torch.unsqueeze(normalize(img_to_tensor(frame)), dim=0)\n",
        "\n",
        "# Upscale it using SR model\n",
        "up_tensor = model(norm.cuda()).detach()[0]\n",
        "\n",
        "# Denormalise it and turn it into a opencv2 mat\n",
        "denormed = denorm(up_tensor).cpu()\n",
        "up_mat = denormed[[2,1,0], :, :].permute(1,2,0).numpy() * 256.0\n",
        "\n",
        "cv2.imwrite('resources/upscaled_frame.png', up_mat)\n",
        "        "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e-B1AGc5ZwHQ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 231
        },
        "outputId": "6a8e397c-813f-4a44-fbeb-c3a7b2abd9ca"
      },
      "source": [
        "def img_to_tensor(img):\n",
        "    return torch.from_numpy(np.asarray(img)).permute(2,0,1)[[2,1,0], :, :].float() / 256.0\n",
        "\n",
        "if not os.path.exists('highres'):\n",
        "    os.makedirs('highres')\n",
        "    \n",
        "if not os.path.exists('lowres'):\n",
        "    os.makedirs('lowres')\n",
        "\n",
        "cap = cv2.VideoCapture(\"/content/drive/MyDrive/Colab Notebooks/video/sample0.mkv\")\n",
        "\n",
        "s, frame0 = cap.read()\n",
        "\n",
        "# Skip the first 100 frames\n",
        "#cap.set(cv2.CAP_PROP_POS_FRAMES, 1000)\n",
        "\n",
        "cw, ch = high_resolution\n",
        "\n",
        "max_start_crop_x, max_start_crop_y = (frame0.shape[1] - cw, frame0.shape[0] - ch)\n",
        "\n",
        "frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT) - 5)\n",
        "\n",
        "lowres_data = torch.zeros((frames,3,64,64))\n",
        "highres_data = torch.zeros((frames,3,128,128))\n",
        "\n",
        "i=0\n",
        "for _ in tqdm(range(frames)):\n",
        "    if not cap.isOpened():\n",
        "        break\n",
        "    \n",
        "    ret, frame = cap.read()\n",
        "\n",
        "    if frame is None:\n",
        "      continue\n",
        "    \n",
        "    cx = random.randint(0, max_start_crop_x)\n",
        "    cy = random.randint(0, max_start_crop_y)\n",
        "    \n",
        "    crop = frame[cy:(cy+ch), cx:(cx+cw)]\n",
        "    lowres_crop = cv2.resize(crop, low_resolution, interpolation=cv2.INTER_AREA)\n",
        "    #lowres_upscaled = cv2.resize(lowres_crop, high_resolution, interpolation=cv2.INTER_CUBIC)\n",
        "\n",
        "    lowres_data[i] = img_to_tensor(lowres_crop)\n",
        "    highres_data[i] = img_to_tensor(crop)\n",
        "    i += 1\n",
        "\n",
        "cap.release()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "NameError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-57-3294527831bc>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mkmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mGreyUpscaleRecolor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mopt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mAdam\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparameters\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlr\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1e-3\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mloss_fun\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mMSELoss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreduction\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'sum'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mlr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhr\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdataloader\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mNameError\u001b[0m: name 'GreyUpscaleRecolor' is not defined"
          ]
        }
      ]
    }
  ]
}