FoodVision.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "EkanP5cIcUpA",
        "wCjmMVftc3wz",
        "9--EkbwZd1-E",
        "8vDgpCa9d4ka"
      ],
      "gpuType": "T4",
      "authorship_tag": "ABX9TyNZwy+xSGsvyYYaf0cE6cdy",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "6547c2cb386c4ef58c97104fba7c7ac1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_17732210b64948ab83974d9f90b0a4ab",
              "IPY_MODEL_a23315fae5874c96a0e8170be0462528",
              "IPY_MODEL_66f5733c51894663aee3d7c248e87b0e"
            ],
            "layout": "IPY_MODEL_8a9d253b5e4946e7bf4dd646d5c3b636"
          }
        },
        "17732210b64948ab83974d9f90b0a4ab": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_960828c9759f49cbafbb81e8fe9e7143",
            "placeholder": "​",
            "style": "IPY_MODEL_f08bc1b45be740d983a909b4ec9ceef0",
            "value": "100%"
          }
        },
        "a23315fae5874c96a0e8170be0462528": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d1467714df704392af373f26e4004027",
            "max": 10,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_26e2000687664a29a6d29d2e2b5103ed",
            "value": 10
          }
        },
        "66f5733c51894663aee3d7c248e87b0e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_b36538665b9845c7950b2dd7fdb26c58",
            "placeholder": "​",
            "style": "IPY_MODEL_74ca3f57e25e498b84ec672faaa0611f",
            "value": " 10/10 [43:29&lt;00:00, 260.60s/it]"
          }
        },
        "8a9d253b5e4946e7bf4dd646d5c3b636": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "960828c9759f49cbafbb81e8fe9e7143": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "f08bc1b45be740d983a909b4ec9ceef0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "d1467714df704392af373f26e4004027": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "26e2000687664a29a6d29d2e2b5103ed": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "b36538665b9845c7950b2dd7fdb26c58": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "74ca3f57e25e498b84ec672faaa0611f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Bruno-Messias/FoodVision/blob/main/FoodVision.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 0. Getting setup"
      ],
      "metadata": {
        "id": "EkanP5cIcUpA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchvision\n",
        "\n",
        "print(torch.__version__)\n",
        "print(torchvision.__version__)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FzZmisUTarJP",
        "outputId": "38d4d726-7c03-41b4-fc8a-2fad9689b088"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.1.0+cu118\n",
            "0.16.0+cu118\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mnOd01KkST3m",
        "outputId": "d7768c86-4aef-4862-da44-2cdaf3e90494"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch version: 2.1.0+cu118\n",
            "torchvision version: 0.16.0+cu118\n"
          ]
        }
      ],
      "source": [
        "# For this notebook to run with updated APIs, we need torch 1.12+ and torchvision 0.13+\n",
        "try:\n",
        "    import torch\n",
        "    import torchvision\n",
        "    # assert int(torch.__version__.split(\".\")[1]) >= 12, \"torch version should be 1.12+\"\n",
        "    assert int(torchvision.__version__.split(\".\")[1]) >= 13, \"torchvision version should be 0.13+\"\n",
        "    print(f\"torch version: {torch.__version__}\")\n",
        "    print(f\"torchvision version: {torchvision.__version__}\")\n",
        "except:\n",
        "    print(f\"[INFO] torch/torchvision versions not as required, installing nightly versions.\")\n",
        "    !pip3 install -U torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113\n",
        "    import torch\n",
        "    import torchvision\n",
        "    print(f\"torch version: {torch.__version__}\")\n",
        "    print(f\"torchvision version: {torchvision.__version__}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Continue with regular imports\n",
        "import matplotlib.pyplot as plt\n",
        "import torch\n",
        "import torchvision\n",
        "\n",
        "from torch import nn\n",
        "from torchvision import transforms\n",
        "\n",
        "# Try to get torchinfo, install it if it doesn't work\n",
        "try:\n",
        "    from torchinfo import summary\n",
        "except:\n",
        "    print(\"[INFO] Couldn't find torchinfo... installing it.\")\n",
        "    !pip install -q torchinfo\n",
        "    from torchinfo import summary\n",
        "\n",
        "# Try to import the going_modular directory, download it from GitHub if it doesn't work\n",
        "try:\n",
        "    from going_modular.going_modular import data_setup, engine\n",
        "    from helper_functions import download_data, set_seeds, plot_loss_curves\n",
        "except:\n",
        "    # Get the going_modular scripts\n",
        "    print(\"[INFO] Couldn't find going_modular or helper_functions scripts... downloading them from GitHub.\")\n",
        "    !git clone https://github.com/mrdbourke/pytorch-deep-learning\n",
        "    !mv pytorch-deep-learning/going_modular .\n",
        "    !mv pytorch-deep-learning/helper_functions.py . # get the helper_functions.py script\n",
        "    !rm -rf pytorch-deep-learning\n",
        "    from going_modular.going_modular import data_setup, engine\n",
        "    from helper_functions import download_data, set_seeds, plot_loss_curves"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1B4yeG9taiLj",
        "outputId": "dd43f5d7-0090-41c7-ae6a-f6ff92ab7dc8"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[INFO] Couldn't find torchinfo... installing it.\n",
            "[INFO] Couldn't find going_modular or helper_functions scripts... downloading them from GitHub.\n",
            "Cloning into 'pytorch-deep-learning'...\n",
            "remote: Enumerating objects: 4036, done.\u001b[K\n",
            "remote: Counting objects: 100% (1224/1224), done.\u001b[K\n",
            "remote: Compressing objects: 100% (223/223), done.\u001b[K\n",
            "remote: Total 4036 (delta 1069), reused 1087 (delta 998), pack-reused 2812\u001b[K\n",
            "Receiving objects: 100% (4036/4036), 650.87 MiB | 41.96 MiB/s, done.\n",
            "Resolving deltas: 100% (2362/2362), done.\n",
            "Updating files: 100% (248/248), done.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
        "device"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "oWAHoIHGWf3y",
        "outputId": "ba72fe8e-bdc4-4e3e-fd11-75531a94a9e0"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'cuda'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 1. Get model"
      ],
      "metadata": {
        "id": "iWmJnhRecZhu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def create_model(num_classes:int=3,\n",
        "                 seed:int=42):\n",
        "    \"\"\"Creates an feature extractor model and transforms.\n",
        "\n",
        "    Args:\n",
        "        num_classes (int, optional): number of classes in the classifier head.\n",
        "            Defaults to 3.\n",
        "        seed (int, optional): random seed value. Defaults to 42.\n",
        "\n",
        "    Returns:\n",
        "        model (torch.nn.Module):  feature extractor model.\n",
        "        transforms (torchvision.transforms): image transforms.\n",
        "    \"\"\"\n",
        "    # 1. Setup pretrained model\n",
        "    weights = torchvision.models.EfficientNet_V2_S_Weights.DEFAULT\n",
        "\n",
        "    # 2. Get EffNetB2 transforms\n",
        "    transforms = weights.transforms()\n",
        "\n",
        "    # 3. Setup pretrained model\n",
        "    model = torchvision.models.efficientnet_v2_s(weights=weights).to(device) # could also use weights=\"DEFAULT\" .to(device)\n",
        "\n",
        "    # 4. Freeze the base layers in the model (this will freeze all layers to begin with)\n",
        "    for param in model.parameters():\n",
        "        param.requires_grad = False\n",
        "\n",
        "    # 5. Change classifier head with random seed for reproducibility\n",
        "    torch.manual_seed(seed)\n",
        "    model.classifier = torch.nn.Sequential(\n",
        "        torch.nn.Dropout(p=0.2, inplace=True),\n",
        "        torch.nn.Linear(in_features=1280,\n",
        "                        out_features=num_classes, # same number of output units as our number of classes\n",
        "                        bias=True))\n",
        "\n",
        "    return model, transforms"
      ],
      "metadata": {
        "id": "BM9wsJo5fN0H"
      },
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model, transforms = create_model(num_classes=101,\n",
        "                                 seed=42)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K5WR46BjfOEW",
        "outputId": "b3c66916-3ad1-4277-e30f-2b06ce6fa722"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Downloading: \"https://download.pytorch.org/models/efficientnet_v2_s-dd5fe13b.pth\" to /root/.cache/torch/hub/checkpoints/efficientnet_v2_s-dd5fe13b.pth\n",
            "100%|██████████| 82.7M/82.7M [00:00<00:00, 147MB/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# # Print a summary using torchinfo (uncomment for actual output)\n",
        "# summary(model=model,\n",
        "#         input_size=(32, 3, 224, 224), # make sure this is \"input_size\", not \"input_shape\"\n",
        "#         # col_names=[\"input_size\"], # uncomment for smaller output\n",
        "#         col_names=[\"input_size\", \"output_size\", \"num_params\", \"trainable\"],\n",
        "#         col_width=20,\n",
        "#         row_settings=[\"var_names\"]\n",
        "# )"
      ],
      "metadata": {
        "id": "ZTuGNEHbfOHT"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 2. Get Data"
      ],
      "metadata": {
        "id": "xnRpWmP8fHDD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create Food101 training data transforms (only perform data augmentation on the training images)\n",
        "food101_train_transforms = torchvision.transforms.Compose([\n",
        "    torchvision.transforms.TrivialAugmentWide(),\n",
        "    transforms,\n",
        "])"
      ],
      "metadata": {
        "id": "_jLYraPVffoF"
      },
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(f\"Training transforms:\\n{food101_train_transforms}\\n\")\n",
        "print(f\"Testing transforms:\\n{transforms}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mf-rSrNqfvLY",
        "outputId": "b8da506b-d995-416b-a6b6-9df12874b24a"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training transforms:\n",
            "Compose(\n",
            "    TrivialAugmentWide(num_magnitude_bins=31, interpolation=InterpolationMode.NEAREST, fill=None)\n",
            "    ImageClassification(\n",
            "    crop_size=[384]\n",
            "    resize_size=[384]\n",
            "    mean=[0.485, 0.456, 0.406]\n",
            "    std=[0.229, 0.224, 0.225]\n",
            "    interpolation=InterpolationMode.BILINEAR\n",
            ")\n",
            ")\n",
            "\n",
            "Testing transforms:\n",
            "ImageClassification(\n",
            "    crop_size=[384]\n",
            "    resize_size=[384]\n",
            "    mean=[0.485, 0.456, 0.406]\n",
            "    std=[0.229, 0.224, 0.225]\n",
            "    interpolation=InterpolationMode.BILINEAR\n",
            ")\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from torchvision import datasets\n",
        "\n",
        "# Setup data directory\n",
        "from pathlib import Path\n",
        "data_dir = Path(\"data\")\n",
        "\n",
        "# Get training data (~750 images x 101 food classes)\n",
        "train_data = datasets.Food101(root=data_dir, # path to download data to\n",
        "                              split=\"train\", # dataset split to get\n",
        "                              transform=food101_train_transforms, # perform data augmentation on training data\n",
        "                              download=True) # want to download?\n",
        "\n",
        "# Get testing data (~250 images x 101 food classes)\n",
        "test_data = datasets.Food101(root=data_dir,\n",
        "                             split=\"test\",\n",
        "                             transform=transforms, # perform normal EffNetB2 transforms on test data\n",
        "                             download=True)"
      ],
      "metadata": {
        "id": "bZzzIOeuWqBR"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Get Food101 class names\n",
        "food101_class_names = train_data.classes\n",
        "\n",
        "# View the first 10\n",
        "food101_class_names[:10]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6IJS05FTea7D",
        "outputId": "064c319b-1274-404b-c48a-534f9447c831"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['apple_pie',\n",
              " 'baby_back_ribs',\n",
              " 'baklava',\n",
              " 'beef_carpaccio',\n",
              " 'beef_tartare',\n",
              " 'beet_salad',\n",
              " 'beignets',\n",
              " 'bibimbap',\n",
              " 'bread_pudding',\n",
              " 'breakfast_burrito']"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 2. Creating an  feature extractor"
      ],
      "metadata": {
        "id": "rVq6MvaeXHdT"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.2 Creating DataLoaders"
      ],
      "metadata": {
        "id": "RDCgiU8gapyx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def split_dataset(dataset:torchvision.datasets, split_size:float=0.2, seed:int=42):\n",
        "    \"\"\"Randomly splits a given dataset into two proportions based on split_size and seed.\n",
        "\n",
        "    Args:\n",
        "        dataset (torchvision.datasets): A PyTorch Dataset, typically one from torchvision.datasets.\n",
        "        split_size (float, optional): How much of the dataset should be split?\n",
        "            E.g. split_size=0.2 means there will be a 20% split and an 80% split. Defaults to 0.2.\n",
        "        seed (int, optional): Seed for random generator. Defaults to 42.\n",
        "\n",
        "    Returns:\n",
        "        tuple: (random_split_1, random_split_2) where random_split_1 is of size split_size*len(dataset) and\n",
        "            random_split_2 is of size (1-split_size)*len(dataset).\n",
        "    \"\"\"\n",
        "    # Create split lengths based on original dataset length\n",
        "    length_1 = int(len(dataset) * split_size) # desired length\n",
        "    length_2 = len(dataset) - length_1 # remaining length\n",
        "\n",
        "    # Print out info\n",
        "    print(f\"[INFO] Splitting dataset of length {len(dataset)} into splits of size: {length_1} ({int(split_size*100)}%), {length_2} ({int((1-split_size)*100)}%)\")\n",
        "\n",
        "    # Create splits with given random seed\n",
        "    random_split_1, random_split_2 = torch.utils.data.random_split(dataset,\n",
        "                                                                   lengths=[length_1, length_2],\n",
        "                                                                   generator=torch.manual_seed(seed)) # set the random seed for reproducible splits\n",
        "    return random_split_1, random_split_2"
      ],
      "metadata": {
        "id": "rLUzTQUPolb2"
      },
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create training 20% split of Food101\n",
        "train_data_food101_20_percent, _ = split_dataset(dataset=train_data,\n",
        "                                                 split_size=0.2)\n",
        "\n",
        "# Create testing 20% split of Food101\n",
        "test_data_food101_20_percent, _ = split_dataset(dataset=test_data,\n",
        "                                                split_size=0.2)\n",
        "\n",
        "len(train_data_food101_20_percent), len(test_data_food101_20_percent)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "o9iIH-4xommX",
        "outputId": "bb12815d-aa19-4d88-8e0a-acbf80128af7"
      },
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[INFO] Splitting dataset of length 75750 into splits of size: 15150 (20%), 60600 (80%)\n",
            "[INFO] Splitting dataset of length 25250 into splits of size: 5050 (20%), 20200 (80%)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(15150, 5050)"
            ]
          },
          "metadata": {},
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "import torch\n",
        "\n",
        "BATCH_SIZE = 32\n",
        "NUM_WORKERS = 2 if os.cpu_count() <= 4 else 4 # this value is very experimental and will depend on the hardware you have available, Google Colab generally provides 2x CPUs\n",
        "\n",
        "# Create Food101 20 percent training DataLoader\n",
        "train_dataloader = torch.utils.data.DataLoader(train_data_food101_20_percent,\n",
        "                                               batch_size=BATCH_SIZE,\n",
        "                                               shuffle=True,\n",
        "                                               num_workers=NUM_WORKERS)\n",
        "# Create Food101 20 percent testing DataLoader\n",
        "test_dataloader = torch.utils.data.DataLoader(test_data_food101_20_percent,\n",
        "                                               batch_size=BATCH_SIZE,\n",
        "                                               shuffle=False,\n",
        "                                               num_workers=NUM_WORKERS)"
      ],
      "metadata": {
        "id": "6PYhhJmyXEwv"
      },
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.3 Training model"
      ],
      "metadata": {
        "id": "ViKGHJu7bGtz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from going_modular.going_modular import engine\n",
        "\n",
        "# Setup optimizer\n",
        "optimizer = torch.optim.Adam(params=model.parameters(),\n",
        "                             lr=1e-3)\n",
        "\n",
        "# Setup loss function\n",
        "loss_fn = torch.nn.CrossEntropyLoss(label_smoothing=0.1)\n",
        "\n",
        "# model_compiled = torch.compile(model, backend=\"aot_eager\")\n",
        "\n",
        "# torch._dynamo.reset()\n",
        "# model_compiled = torch.compile(model)\n",
        "\n",
        "# Set seeds for reproducibility and train the model\n",
        "set_seeds()\n",
        "results = engine.train(model=model,\n",
        "                       train_dataloader=train_dataloader,\n",
        "                       test_dataloader=test_dataloader,\n",
        "                       epochs=10,\n",
        "                       optimizer=optimizer,\n",
        "                       loss_fn=loss_fn,\n",
        "                       device=device)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 234,
          "referenced_widgets": [
            "6547c2cb386c4ef58c97104fba7c7ac1",
            "17732210b64948ab83974d9f90b0a4ab",
            "a23315fae5874c96a0e8170be0462528",
            "66f5733c51894663aee3d7c248e87b0e",
            "8a9d253b5e4946e7bf4dd646d5c3b636",
            "960828c9759f49cbafbb81e8fe9e7143",
            "f08bc1b45be740d983a909b4ec9ceef0",
            "d1467714df704392af373f26e4004027",
            "26e2000687664a29a6d29d2e2b5103ed",
            "b36538665b9845c7950b2dd7fdb26c58",
            "74ca3f57e25e498b84ec672faaa0611f"
          ]
        },
        "id": "3vtf1U9tbDhV",
        "outputId": "a64fbfe3-7d61-4464-a214-0ec58b4eacff"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/10 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "6547c2cb386c4ef58c97104fba7c7ac1"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 1 | train_loss: 3.5516 | train_acc: 0.3148 | test_loss: 2.6651 | test_acc: 0.5007\n",
            "Epoch: 2 | train_loss: 2.7502 | train_acc: 0.4741 | test_loss: 2.3738 | test_acc: 0.5561\n",
            "Epoch: 3 | train_loss: 2.5417 | train_acc: 0.5162 | test_loss: 2.2590 | test_acc: 0.5787\n",
            "Epoch: 4 | train_loss: 2.4365 | train_acc: 0.5413 | test_loss: 2.2052 | test_acc: 0.5951\n",
            "Epoch: 5 | train_loss: 2.3608 | train_acc: 0.5635 | test_loss: 2.1840 | test_acc: 0.5991\n",
            "Epoch: 6 | train_loss: 2.3180 | train_acc: 0.5752 | test_loss: 2.1791 | test_acc: 0.6018\n",
            "Epoch: 7 | train_loss: 2.2860 | train_acc: 0.5823 | test_loss: 2.1509 | test_acc: 0.6088\n",
            "Epoch: 8 | train_loss: 2.2678 | train_acc: 0.5850 | test_loss: 2.1514 | test_acc: 0.6078\n",
            "Epoch: 9 | train_loss: 2.2342 | train_acc: 0.5960 | test_loss: 2.1292 | test_acc: 0.6151\n",
            "Epoch: 10 | train_loss: 2.2257 | train_acc: 0.5932 | test_loss: 2.1417 | test_acc: 0.6100\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.4 Inspecting loss curves\n"
      ],
      "metadata": {
        "id": "4T7UqlrKbWQT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from helper_functions import plot_loss_curves\n",
        "\n",
        "plot_loss_curves(results)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 641
        },
        "id": "nqA0RXYCbT1y",
        "outputId": "af66239f-102c-4387-b034-019d5b266870"
      },
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x700 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABL4AAAJwCAYAAACH0KjyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAADVXklEQVR4nOzdd3hUZfrG8e+kTXpCGikEQi8JAaQoIpYFpCgL2NsP0dVd67rLqsiqFFFRWF2sWIF1BXEtKKsiiyg2pLdQpIdACkkI6WWSmfP7Y5KBSA0kOSn357rmYubMOWeeSZhkcs/7Pq/FMAwDERERERERERGRJsbN7AJERERERERERETqgoIvERERERERERFpkhR8iYiIiIiIiIhIk6TgS0REREREREREmiQFXyIiIiIiIiIi0iQp+BIRERERERERkSZJwZeIiIiIiIiIiDRJCr5ERERERERERKRJUvAlIiIiIiIiIiJNkoIvERERERERERFpkhR8iYip5s2bh8ViYd26dWaXIiIiIiKVXn/9dSwWCxdeeKHZpYiInBcFXyIiIiIiIlLN/PnziYuLY82aNezZs8fsckREzpmCLxEREREREXHZv38/K1eu5MUXXyQ8PJz58+ebXdJJFRUVmV2CiDQCCr5EpMHbuHEjw4cPJzAwEH9/fwYNGsSqVauq7VNeXs7UqVPp2LEj3t7ehIaGcskll7Bs2TLXPhkZGdxxxx20atUKq9VKVFQUo0aNIjk5uZ6fkYiIiEjDNX/+fFq0aMFVV13Fddddd9LgKzc3l7/+9a/ExcVhtVpp1aoVY8eOJTs727VPaWkpU6ZMoVOnTnh7exMVFcU111zD3r17AVixYgUWi4UVK1ZUO3dycjIWi4V58+a5to0bNw5/f3/27t3LiBEjCAgI4NZbbwXgxx9/5Prrr6d169ZYrVZiY2P561//SklJyQl1//rrr9xwww2Eh4fj4+ND586defzxxwH47rvvsFgsLFq06ITjFixYgMVi4Zdffqnx11NEzOVhdgEiIqezbds2Bg4cSGBgII8++iienp68+eabXH755Xz//feuvhNTpkxh+vTp3HXXXfTr14/8/HzWrVvHhg0bGDJkCADXXnst27Zt48EHHyQuLo7MzEyWLVtGSkoKcXFxJj5LERERkYZj/vz5XHPNNXh5eXHzzTcze/Zs1q5dS9++fQEoLCxk4MCB7NixgzvvvJMLLriA7OxsFi9ezKFDhwgLC8Nut3P11VezfPlybrrpJh566CEKCgpYtmwZW7dupX379jWuq6KigqFDh3LJJZfwj3/8A19fXwA++ugjiouLuffeewkNDWXNmjW88sorHDp0iI8++sh1/JYtWxg4cCCenp788Y9/JC4ujr179/Lf//6XZ555hssvv5zY2Fjmz5/PmDFjTviatG/fnv79+5/HV1ZETGGIiJho7ty5BmCsXbv2pPePHj3a8PLyMvbu3evalpaWZgQEBBiXXnqpa1uPHj2Mq6666pSPc/ToUQMwZs6cWXvFi4iIiDQx69atMwBj2bJlhmEYhsPhMFq1amU89NBDrn0mTZpkAMann356wvEOh8MwDMOYM2eOARgvvvjiKff57rvvDMD47rvvqt2/f/9+AzDmzp3r2nb77bcbgPHYY4+dcL7i4uITtk2fPt2wWCzGgQMHXNsuvfRSIyAgoNq24+sxDMOYOHGiYbVajdzcXNe2zMxMw8PDw5g8efIJjyMiDZ+mOopIg2W32/nf//7H6NGjadeunWt7VFQUt9xyCz/99BP5+fkABAcHs23bNnbv3n3Sc/n4+ODl5cWKFSs4evRovdQvIiIi0tjMnz+fli1bcsUVVwBgsVi48cYbWbhwIXa7HYBPPvmEHj16nDAqqmr/qn3CwsJ48MEHT7nPubj33ntP2Obj4+O6XlRURHZ2NhdffDGGYbBx40YAsrKy+OGHH7jzzjtp3br1KesZO3YsZWVlfPzxx65tH374IRUVFdx2223nXLeImEfBl4g0WFlZWRQXF9O5c+cT7uvatSsOh4ODBw8C8NRTT5Gbm0unTp3o3r07jzzyCFu2bHHtb7Vaef7551myZAktW7bk0ksvZcaMGWRkZNTb8xERERFpyOx2OwsXLuSKK65g//797Nmzhz179nDhhRdy+PBhli9fDsDevXtJSEg47bn27t1L586d8fCove46Hh4etGrV6oTtKSkpjBs3jpCQEPz9/QkPD+eyyy4DIC8vD4B9+/YBnLHuLl260Ldv32p9zebPn89FF11Ehw4dauupiEg9UvAlIk3CpZdeyt69e5kzZw4JCQm88847XHDBBbzzzjuuff7yl7+wa9cupk+fjre3N08++SRdu3Z1fRIoIiIi0px9++23pKens3DhQjp27Oi63HDDDQC1vrrjqUZ+VY0s+y2r1Yqbm9sJ+w4ZMoQvv/ySCRMm8Nlnn7Fs2TJXY3yHw1HjusaOHcv333/PoUOH2Lt3L6tWrdJoL5FGTM3tRaTBCg8Px9fXl507d55w36+//oqbmxuxsbGubSEhIdxxxx3ccccdFBYWcumllzJlyhTuuusu1z7t27fnb3/7G3/729/YvXs3PXv25IUXXuD999+vl+ckIiIi0lDNnz+fiIgIXnvttRPu+/TTT1m0aBFvvPEG7du3Z+vWrac9V/v27Vm9ejXl5eV4enqedJ8WLVoAzhUij3fgwIGzrjkpKYldu3bxr3/9i7Fjx7q2H7+yN+Bqm3GmugFuuukmxo8fzwcffEBJSQmenp7ceOONZ12TiDQsGvElIg2Wu7s7V155JZ9//jnJycmu7YcPH2bBggVccsklBAYGAnDkyJFqx/r7+9OhQwfKysoAKC4uprS0tNo+7du3JyAgwLWPiIiISHNVUlLCp59+ytVXX8111113wuWBBx6goKCAxYsXc+2117J582YWLVp0wnkMwwCcq2lnZ2fz6quvnnKfNm3a4O7uzg8//FDt/tdff/2s63Z3d692zqrrL730UrX9wsPDufTSS5kzZw4pKSknradKWFgYw4cP5/3332f+/PkMGzaMsLCws65JRBoWjfgSkQZhzpw5fP311ydsnzJlCsuWLeOSSy7hvvvuw8PDgzfffJOysjJmzJjh2q9bt25cfvnl9O7dm5CQENatW8fHH3/MAw88AMCuXbsYNGgQN9xwA926dcPDw4NFixZx+PBhbrrppnp7niIiIiIN0eLFiykoKOD3v//9Se+/6KKLCA8PZ/78+SxYsICPP/6Y66+/njvvvJPevXuTk5PD4sWLeeONN+jRowdjx47lvffeY/z48axZs4aBAwdSVFTEN998w3333ceoUaMICgri+uuv55VXXsFisdC+fXu++OILMjMzz7ruLl260L59ex5++GFSU1MJDAzkk08+OeliRi+//DKXXHIJF1xwAX/84x9p27YtycnJfPnll2zatKnavmPHjuW6664DYNq0aWf/hRSRhsfMJSVFRObOnWsAp7wcPHjQ2LBhgzF06FDD39/f8PX1Na644gpj5cqV1c7z9NNPG/369TOCg4MNHx8fo0uXLsYzzzxj2Gw2wzAMIzs727j//vuNLl26GH5+fkZQUJBx4YUXGv/5z3/MeNoiIiIiDcrIkSMNb29vo6io6JT7jBs3zvD09DSys7ONI0eOGA888IARExNjeHl5Ga1atTJuv/12Izs727V/cXGx8fjjjxtt27Y1PD09jcjISOO6664z9u7d69onKyvLuPbaaw1fX1+jRYsWxp/+9Cdj69atBmDMnTvXtd/tt99u+Pn5nbSu7du3G4MHDzb8/f2NsLAw4+677zY2b958wjkMwzC2bt1qjBkzxggODja8vb2Nzp07G08++eQJ5ywrKzNatGhhBAUFGSUlJWf5VRSRhshiGL8Z1ykiIiIiIiLSjFVUVBAdHc3IkSN59913zS5HRM6DenyJiIiIiIiIHOezzz4jKyurWsN8EWmcNOJLREREREREBFi9ejVbtmxh2rRphIWFsWHDBrNLEpHzpBFfIiIiIiIiIsDs2bO59957iYiI4L333jO7HBGpBRrxJSIiIiIiIiIiTZJGfImIiIiIiIiISJOk4EtERERERERERJokD7MLOBsOh4O0tDQCAgKwWCxmlyMiIiKNgGEYFBQUEB0djZubPutrqPQ+T0RERGqqJu/zGkXwlZaWRmxsrNlliIiISCN08OBBWrVqZXYZcgp6nyciIiLn6mze5zWK4CsgIABwPqHAwECTqxEREZHGID8/n9jYWNf7CGmY9D5PREREaqom7/MaRfBVNew9MDBQb4hERESkRjR9rmHT+zwRERE5V2fzPk8NL0REREREREREpEmqUfA1e/ZsEhMTXZ/I9e/fnyVLlpz2mNzcXO6//36ioqKwWq106tSJr7766ryKFhEREREREREROZMaTXVs1aoVzz33HB07dsQwDP71r38xatQoNm7cSHx8/An722w2hgwZQkREBB9//DExMTEcOHCA4ODg2qpfRERERERERETkpGoUfI0cObLa7WeeeYbZs2ezatWqkwZfc+bMIScnh5UrV+Lp6QlAXFzcuVcrIiJSSwzDoKKiArvdbnYpco7c3d3x8PBQD69mQK9XqW36+SEi0nycc3N7u93ORx99RFFREf379z/pPosXL6Z///7cf//9fP7554SHh3PLLbcwYcIE3N3dT3nusrIyysrKXLfz8/PPtUwREZET2Gw20tPTKS4uNrsUOU++vr5ERUXh5eVldilSR/R6lbqinx8iIs1DjYOvpKQk+vfvT2lpKf7+/ixatIhu3bqddN99+/bx7bffcuutt/LVV1+xZ88e7rvvPsrLy5k8efIpH2P69OlMnTq1pqWJiIickcPhYP/+/bi7uxMdHY2Xl5c+8W+EDMPAZrORlZXF/v376dixI25uWrOnqdHrVeqCfn6IiDQvFsMwjJocYLPZSElJIS8vj48//ph33nmH77///qThV6dOnSgtLXW9YQF48cUXmTlzJunp6ad8jJON+IqNjSUvL0/LXIuIyHmp+r3Upk0bfH19zS5HzlNxcTEHDhygbdu2eHt7V7svPz+foKAgvX9o4E73fdLrVerS6X5+iIhIw1aT93k1HvHl5eVFhw4dAOjduzdr167lpZde4s033zxh36ioKDw9PatNa+zatSsZGRnYbLZTDiu2Wq1YrdaaliYiInLW9Ol+06DvY/Og77PUBf2/EhFpHs77p73D4ag2Out4AwYMYM+ePTgcDte2Xbt2aS69iIiIiIiIiIjUuRoFXxMnTuSHH34gOTmZpKQkJk6cyIoVK7j11lsBGDt2LBMnTnTtf++995KTk8NDDz3Erl27+PLLL3n22We5//77a/dZiIiIiIiIiIiI/EaNgq/MzEzGjh1L586dGTRoEGvXrmXp0qUMGTIEgJSUlGq9u2JjY1m6dClr164lMTGRP//5zzz00EM89thjtfssREREpEbi4uKYNWtWrZxrxYoVWCwWcnNza+V8IlJdbb5eRUREmpsa9fh69913T3v/ihUrTtjWv39/Vq1aVaOiRERE5ESXX345PXv2rJU/gNeuXYufn9/5FyUiJ6XXq4iISMNQ4+b2IiIi0jAZhoHdbsfD48y/3sPDw+uhIhE5Fb1ejzndolciIiLnS0uZiIhIs2cYBsW2inq/GIZx1jWOGzeO77//npdeegmLxYLFYmHevHlYLBaWLFlC7969sVqt/PTTT+zdu5dRo0bRsmVL/P396du3L99880218/126pTFYuGdd95hzJgx+Pr60rFjRxYvXnzOX9NPPvmE+Ph4rFYrcXFxvPDCC9Xuf/311+nYsSPe3t60bNmS6667znXfxx9/TPfu3fHx8SE0NJTBgwdTVFR0zrVI02HWa7UpvV7tdjt/+MMfaNu2LT4+PnTu3JmXXnrphP3mzJnjeg1HRUXxwAMPuO7Lzc3lT3/6Ey1btsTb25uEhAS++OILAKZMmULPnj2rnWvWrFnExcVV+/qMHj2aZ555hujoaDp37gzAv//9b/r06UNAQACRkZHccsstZGZmVjvXtm3buPrqqwkMDCQgIICBAweyd+9efvjhBzw9PcnIyKi2/1/+8hcGDhx4Vl8bERFpmjTiS0REmr2ScjvdJi2t98fd/tRQfL3O7lfxSy+9xK5du0hISOCpp54CnH8AAjz22GP84x//oF27drRo0YKDBw8yYsQInnnmGaxWK++99x4jR45k586dtG7d+pSPMXXqVGbMmMHMmTN55ZVXuPXWWzlw4AAhISE1el7r16/nhhtuYMqUKdx4442sXLmS++67j9DQUMaNG8e6dev485//zL///W8uvvhicnJy+PHHHwFIT0/n5ptvZsaMGYwZM4aCggJ+/PHHGoUO0nSZ9VqFpvN6dTgctGrVio8++ojQ0FBWrlzJH//4R6KiorjhhhsAmD17NuPHj+e5555j+PDh5OXl8fPPP7uOHz58OAUFBbz//vu0b9+e7du34+7uflZfmyrLly8nMDCQZcuWubaVl5czbdo0OnfuTGZmJuPHj2fcuHF89dVXAKSmpnLppZdy+eWX8+233xIYGMjPP/9MRUUFl156Ke3atePf//43jzzyiOt88+fPZ8aMGTWqTUREmhYFXyIiIo1AUFAQXl5e+Pr6EhkZCcCvv/4KwFNPPeVaaAYgJCSEHj16uG5PmzaNRYsWsXjx4mqjNn5r3Lhx3HzzzQA8++yzvPzyy6xZs4Zhw4bVqNYXX3yRQYMG8eSTTwLQqVMntm/fzsyZMxk3bhwpKSn4+flx9dVXExAQQJs2bejVqxfgDL4qKiq45ppraNOmDQDdu3ev0eOLmK0hv149PT2ZOnWq63bbtm355Zdf+M9//uMKvp5++mn+9re/8dBDD7n269u3LwDffPMNa9asYceOHXTq1AmAdu3anfmL8ht+fn6888471aY43nnnna7r7dq14+WXX6Zv374UFhbi7+/Pa6+9RlBQEAsXLsTT0xPAVQPAH/7wB+bOnesKvv773/9SWlrqel4iItI8KfgSEZFmz8fTne1PDTXlcWtDnz59qt0uLCxkypQpfPnll64gqaSkhJSUlNOeJzEx0XXdz8+PwMDAE6YZnY0dO3YwatSoatsGDBjArFmzsNvtDBkyhDZt2tCuXTuGDRvGsGHDXFO2evTowaBBg+jevTtDhw7lyiuv5LrrrqNFixY1rkOaHrNeq1WPXRsawuv1tddeY86cOaSkpFBSUoLNZnNNT8zMzCQtLY1Bgwad9NhNmzbRqlWraoHTuejevfsJfb3Wr1/PlClT2Lx5M0ePHsXhcADOleO7devGpk2bGDhwoCv0+q1x48bxxBNPsGrVKi666CLmzZvHDTfcoIUBRESaOQVfIiLS7FkslrOewtQQ/faPuocffphly5bxj3/8gw4dOuDj48N1112HzWY77Xl++8ekxWJx/eFZmwICAtiwYQMrVqzgf//7H5MmTWLKlCmsXbuW4OBgli1bxsqVK/nf//7HK6+8wuOPP87q1atp27ZtrdcijUtjf62C+a/XhQsX8vDDD/PCCy/Qv39/AgICmDlzJqtXrwbAx8fntMef6X43N7cTpiaXl5efsN9vvw5FRUUMHTqUoUOHMn/+fMLDw0lJSWHo0KGur8WZHjsiIoKRI0cyd+5c2rZty5IlS0666ryIiDQvam4vIiLSSHh5eWG328+4388//8y4ceMYM2YM3bt3JzIykuTk5LovsFLXrl1d/YCOr6lTp06uPkAeHh4MHjyYGTNmsGXLFpKTk/n2228B5x/wAwYMYOrUqWzcuBEvLy8WLVpUb/WL1IaG+nr9+eefufjii7nvvvvo1asXHTp0YO/eva77AwICiIuLY/ny5Sc9PjExkUOHDrFr166T3h8eHk5GRka18GvTpk1nrOvXX3/lyJEjPPfccwwcOJAuXbqcMIItMTGRH3/88aRBWpW77rqLDz/8kLfeeov27dszYMCAMz62iIg0bQq+REREGom4uDhWr15NcnIy2dnZpxzd0bFjRz799FM2bdrE5s2bueWWW+pk5Nap/O1vf2P58uVMmzaNXbt28a9//YtXX32Vhx9+GIAvvviCl19+mU2bNnHgwAHee+89HA4HnTt3ZvXq1Tz77LOsW7eOlJQUPv30U7KysujatWu91S9SGxrq67Vjx46sW7eOpUuXsmvXLp588knWrl1bbZ8pU6bwwgsv8PLLL7N79242bNjAK6+8AsBll13GpZdeyrXXXsuyZcvYv38/S5Ys4euvvwbg8ssvJysrixkzZrB3715ee+01lixZcsa6WrdujZeXF6+88gr79u1j8eLFTJs2rdo+DzzwAPn5+dx0002sW7eO3bt38+9//5udO3e69hk6dCiBgYE8/fTT3HHHHef75RIRkSag2QdfK3Zm8vtXf2L8h5vMLkVEROS0Hn74Ydzd3enWrZtrGtDJvPjii7Ro0YKLL76YkSNHMnToUC644IJ6q/OCCy7gP//5DwsXLiQhIYFJkybx1FNPMW7cOACCg4P59NNP+d3vfkfXrl154403+OCDD4iPjycwMJAffviBESNG0KlTJ5544gleeOEFhg8fXm/1i9SGhvp6/dOf/sQ111zDjTfeyIUXXsiRI0e47777qu1z++23M2vWLF5//XXi4+O5+uqr2b17t+v+Tz75hL59+3LzzTfTrVs3Hn30Udfotq5du/L666/z2muv0aNHD9asWeMKvU8nPDycefPm8dFHH9GtWzeee+45/vGPf1TbJzQ0lG+//ZbCwkIuu+wyevfuzdtvv11t2qebmxvjxo3DbrczduzY8/lSiYjUrfx02PYZLH0c5gyH90bBV4/Amrdh3wrISwWtal0rLEYjWB88Pz+foKAg8vLyCAwMrNVz/7g7i/97dw1xob6seOSKWj23iIg0PKWlpezfv5+2bdvi7e1tdjlynk73/azL9w9Se073fdLrVc7FH/7wB7Kysli8ePFp99P/LxGpNxU2OJwEB9c4L4fWQt7BMx/n5Q+hHSCsU+Wlo/PfkHbg2bx/btXkfV7j7g5aC+KjgwBIPlJMQWk5Ad4nXyVGREREREQarry8PJKSkliwYMEZQy8RkTpVmFkZcq12hlxpG6GitPo+FjdoGQ+t+kGrvmDYIXsXZO92/puzH2yFkL7JefntscFtjgVhrn87gW8oWCz19UwbhWYffIX4eREd5E1aXinb0/K5sF2o2SWJiIg0KPfccw/vv//+Se+77bbbeOONN+q5IhE5leb8eh01ahRr1qzhnnvuYciQIWaXIyLNhb0CDm91BlxVYVfugRP382nhDLli+zr/jekNVv9Tn7fCBkeTK8Ow4wKx7N1QlgdH9zsvu/934uMcH4aFVv7bIg7cm2cE1Dyf9W90iw4iLa+UbQq+RERETvDUU0+dskePphCKNCzN+fW6YsUKs0sQkeag6AgcOm7KYup6KC/+zU4WiOgKsf0qw65+zimLNRmJ5eEF4Z2cl+MZhnNEWfYuOLL7uEBsF+QehJKjzvDt4Orqx7l5OqdIHj86LKwThHUA76Bz+lI0Fgq+gISYQL7ZcZitaXlmlyIiItLgREREEBERYXYZInIW9HoVEalFDjtkbj8Wch1cAzl7T9zPGgSt+jgDrtjK0Vx1FSZZLBDQ0nlpO7D6fbZiZ32/HSGWvRsqSiB7p/PyW/4tfzNlsvLfwFbg1vjXRFTwxbE+X9vT8k2uRERERERERERMUZwDh9YdG9GVugFsBSfuF9b52JTF2AudIVFDCIi8fCGyu/NyPIcD8lN/E4hVXi/MgMLDzkvyj9WP8/BxjggL/U0gFtrB+ViNhIIvnCO+AHZnFlJabsfb093kikRERERERESkzjgcztFPrpUW1zjDoN/yCoBWvY9NWWzVx9lHqzFxc4PgWOelw6Dq95XmnzhlMns3HNnrHCWWkeS8/FZQ69+MEKu87t+ywTXXV/AFRAZ6E+LnRU6RjZ0ZBfSIDTa7JBERERERERGpLaV5laO51lautrje2ST+t0LaO0dxVY3oiugKbk14cIx3oHNqZkzv6tvtFc4m/SdMm9zp7COWl+K87F1e/ThrYPVALH6Ms7eYiRR8ARaLhfjoQH7cnc3WtDwFXyIiIiIiIiKNlWE4Q5qqKYsH10DWr4BRfT9PX2fgU9WEvlVf8NOCd4BzBcjQ9s5L5+HV7ys6ctzosF1wZI/z36PJUJbvbPifut65b0xvBV8NRXx0ED/uzmab+nyJiIiIiIhIY2WvgKwdzuChMAs8rODhXfmvtfpt9+Pv83auJFjtPmuDm7Z2UmUFzud7cO2xsKs098T9WsQdm7IY2w8i4p0Bj9SMXyj49Yc2/atvryiDnH3Vp0yGdzWnxuPoO1ypqs/XtlSt7CgiInIyycnJtG3blo0bN9KzZ0+zyxERERHDgNyUYyNsUtdD2iZnb6ba4n6SUMzDG9y9Th2aVbv/bI49WSB33O3jwzfDcIYrVX25Dq6FzG1gOKrX7eEN0Rcc14S+H/hr1ds65WF1Tg2NMD/sOp6Cr0pVKzvuyCig3O7A070BrMggIiJynMsvv5yePXsya9asWjnfuHHjyM3N5bPPPquV84nIMXq9ikidKM6BtA3O1QYPrXMGXcXZJ+5nDYToXhDcGuzlUFEKdpvz34qyyn9/c9teduz68exlzktZ/TzFkzo+KLPbnP26fiso9tiUxdi+0LK7M3CTZk/BV6U2Ib74Wz0oLKtgb1YhXSIDzS5JRERERKTJsNlseHnpj1CRs1Ze6lxNL3XdsdFcOftO3M/NEyITKhuU93H+G9rBuZLfuTCMY2GZKxSrCsmqth0XklWUVQ/NThqq2c5w7EkCueP7cdltzktV+ObuBVE9j01ZbNUPAqPO7flK02c0Anl5eQZg5OXl1enjXD97pdFmwhfGx+sO1unjiIiIeUpKSozt27cbJSUlxzY6HIZRVlj/F4fjrOu+/fbbDZzvAF2X/fv3G0lJScawYcMMPz8/IyIiwrjtttuMrKws13EfffSRkZCQYHh7exshISHGoEGDjMLCQmPy5MknnO+77747bQ379+83AGPjxo2ubStWrDD69u1reHl5GZGRkcaECROM8vLyMz6+YRjGd999Z/Tt29fw9fU1goKCjIsvvthITk4+66+JYZzi+1mpvt4/yPk53ffphO+vWa/VRvh6NQzDePTRR42OHTsaPj4+Rtu2bY0nnnjCsNls1fZZvHix0adPH8NqtRqhoaHG6NGjXfeVlpYajz76qNGqVSvDy8vLaN++vfHOO+8YhmEYc+fONYKCgqqda9GiRcbxf15MnjzZ6NGjh/H2228bcXFxhsViMQzDMJYsWWIMGDDACAoKMkJCQoyrrrrK2LNnT7VzHTx40LjpppuMFi1aGL6+vkbv3r2NVatWGfv37zcsFouxdu3aavv/85//NFq3bm3Y7fYzfl2qnO7nh0i9s9sN4/AOw9jwvmH896+G8calhjE1xDAmB554eamXYXxyt2GsesMwDq41DFsT/D/scBhGeZlhlOQZRmGWYeQeNIzsPYaRsdUw0pMMo7zU7ArFZDV5n6cRX8eJjwlkTXIOW9PyuLZ3K7PLERGR+lJeDM9G1//j/j0NvPzOateXXnqJXbt2kZCQwFNPPQWAp6cn/fr146677uKf//wnJSUlTJgwgRtuuIFvv/2W9PR0br75ZmbMmMGYMWMoKCjgxx9/xDAMHn74YXbs2EF+fj5z584FICQkpEblp6amMmLECMaNG8d7773Hr7/+yt133423tzdTpkw57eNXVFQwevRo7r77bj744ANsNhtr1qzB0hga6Ip5zHqtQqN8vQYEBDBv3jyio6NJSkri7rvvJiAggEcffRSAL7/8kjFjxvD444/z3nvvYbPZ+Oqrr1zHjx07ll9++YWXX36ZHj16sH//frKzTzKl6jT27NnDJ598wqeffoq7uzsARUVFjB8/nsTERAoLC5k0aRJjxoxh06ZNuLm5UVhYyGWXXUZMTAyLFy8mMjKSDRs24HA4iIuLY/DgwcydO5c+ffq4Hmfu3LmMGzcOt3Md4SJS3/LTqvflSt0ItoIT9/MLPzaKK+YC5/RF35r9vm6ULJbKvmAaJSrnT8HXcar6fGllRxERaWiCgoLw8vLC19eXyMhIAJ5++ml69erFs88+69pvzpw5xMbGsmvXLgoLC6moqOCaa66hTZs2AHTv3t21r4+PD2VlZa7z1dTrr79ObGwsr776KhaLhS5dupCWlsaECROYNGkS6enpp3z8nJwc8vLyuPrqq2nfvj0AXbs2rEaoIueqobxen3jiCdf1uLg4Hn74YRYuXOgKvp555hluuukmpk6d6tqvR48eAOzatYv//Oc/LFu2jMGDBwPQrl3Nl6O32Wy89957hIeHu7Zde+211faZM2cO4eHhbN++nYSEBBYsWEBWVhZr1651BXwdOnRw7X/XXXdxzz338OKLL2K1WtmwYQNJSUl8/vnnNa5PpF6U5kPaxupBV0H6ift5+jqDrZgLKoOu3s6+VfpQSOS8KPg6TtXKjtvT8nE4DNzc9ANGRKRZ8PR1juYw43HPw+bNm/nuu+/w9/c/4b69e/dy5ZVXMmjQILp3787QoUO58sorue6662jRosV5PW6VHTt20L9//2qjtAYMGEBhYSGHDh2iR48ep3z8kJAQxo0bx9ChQxkyZAiDBw/mhhtuICpK/TnkNMx6rVY99nkw4/X64Ycf8vLLL7N3715XsBYYeKyP7aZNm7j77rtPeuymTZtwd3fnsssuO+fHB2jTpk210Atg9+7dTJo0idWrV5OdnY3D4VyJLSUlhYSEBDZt2kSvXr1OOapt9OjR3H///SxatIibbrqJefPmccUVVxAXF3detYrUigqbc4XB1PXHGtBn76JavyoAixtExFcPucK7gLv+RBepbXpVHad9uD9eHm4UllWQklNMXNjZDWcXEZFGzmI56ylMDUlhYSEjR47k+eefP+G+qKgo3N3dWbZsGStXruR///sfr7zyCo8//jirV6+mbdu2dV7fmR5/7ty5/PnPf+brr7/mww8/5IknnmDZsmVcdNFFdV6bNFKN9LUK9f96/eWXX7j11luZOnUqQ4cOJSgoiIULF/LCCy+49vHx8Tnl8ae7D8DNzQ3DqP6HfHl5+Qn7+fmd+P0aOXIkbdq04e233yY6OhqHw0FCQgI2m+2sHtvLy4uxY8cyd+5crrnmGhYsWMBLL7102mNE6oRhOJvNp2441oA+fYuzWftvBbc+FnDF9IaoHo3255lIY6Pg6zie7m50jQxg86E8tqblKfgSEZEGxcvLC7vd7rp9wQUX8MknnxAXF4eHx8l/pVssFgYMGMCAAQOYNGkSbdq0YdGiRYwfP/6E89VU165d+eSTTzAMwzXq6+effyYgIIBWrVqd8fEBevXqRa9evZg4cSL9+/dnwYIFCr6kSTD79bpy5UratGnD448/7tp24MCBavskJiayfPly7rjjjhOO7969Ow6Hg++//9411fF44eHhFBQUUFRU5Aq3Nm3adMa6jhw5ws6dO3n77bcZOHAgAD/99NMJdb3zzjvk5OScctTXXXfdRUJCAq+//rpriqhInSvM+k1frvVQmnvift7BxwKuVn0g+gLwDz9xPxGpFwq+fqNbdBCbD+WxLS2fqxNNap4qIiJyEnFxcaxevZrk5GT8/f25//77efvtt7n55pt59NFHCQkJYc+ePSxcuJB33nmHdevWsXz5cq688koiIiJYvXo1WVlZrl5acXFxLF26lJ07dxIaGkpQUBCenp5nXc99993HrFmzePDBB3nggQfYuXMnkydPZvz48bi5ubF69epTPv7+/ft56623+P3vf090dDQ7d+5k9+7djB07tq6+fCL1yuzXa8eOHUlJSWHhwoX07duXL7/8kkWLFlXbZ/LkyQwaNIj27dtz0003UVFRwVdffcWECROIi4vj9ttv584773Q1tz9w4ACZmZnccMMNXHjhhfj6+vL3v/+dP//5z6xevZp58+ad8evSokULQkNDeeutt4iKiiIlJYXHHnus2j4333wzzz77LKNHj2b69OlERUWxceNGoqOj6d+/P+AM3i+66CImTJjAnXfeecZRYiI1ZiuC9M3VQ67clBP3c7dCVGJl0NXHOXUxpJ36cok0JHW7wGTtqM/lyN9flWy0mfCFcds7q+r8sUREpP415uXrd+7caVx00UWGj4+PARj79+83du3aZYwZM8YIDg42fHx8jC5duhh/+ctfDIfDYWzfvt0YOnSoER4eblitVqNTp07GK6+84jpfZmamMWTIEMPf398AjO++++60j79//34DMDZu3OjatmLFCqNv376Gl5eXERkZaUyYMMEoLy83DMM47eNnZGQYo0ePNqKiogwvLy+jTZs2xqRJkwy73V6jr8npvp/1+f5Bzt3pvk96vZ7769UwDOORRx4xQkNDDX9/f+PGG280/vnPfxpBQUHV9vnkk0+Mnj17Gl5eXkZYWJhxzTXXuO4rKSkx/vrXv7pepx06dDDmzJnjun/RokVGhw4dDB8fH+Pqq6823nrrLeP4Py8mT55s9OjR44S6li1bZnTt2tWwWq1GYmKisWLFCgMwFi1a5NonOTnZuPbaa43AwEDD19fX6NOnj7F69epq53n33XcNwFizZs0ZvxYn05j/f0ktKi8zjLw0w0jdYBjr5hnG5w8axusXG8aUYMOYHPibS5BhvNrPMBbdaxhr3nYeU15m9jMQaZZq8j7PYhi/mZzfAOXn5xMUFEReXl61hpx1YdPBXEa/9jOhfl6se2KwllUXEWliSktL2b9/P23btsXb29vscuQ8ne77WZ/vHxqL1157jZkzZ5KRkUGPHj145ZVX6Nev3yn3z83N5fHHH+fTTz8lJyeHNm3aMGvWLEaMGHHO5/yt032f9HqV05k2bRofffQRW7ZsOafj9f+riXLYoTgHirKgONv5b1F25SWrcvuRY9dL8059roCo6n25onuCd1C9PRURObWavM/TVMff6BIZgLubhSNFNjLyS4kK0rBpERERafw+/PBDxo8fzxtvvMGFF17IrFmzGDp0KDt37iQiIuKE/W02G0OGDCEiIoKPP/6YmJgYDhw4QHBw8DmfU6Q2FBYWkpyczKuvvsrTTz9tdjlS1xwOZx+touzjgqwsKDpy8iCrOIcTVlA8E4sb+IZBRJfqQVegWt+INAUKvn7D29OdDuH+7DxcwLbUfAVfIiLSbDz77LM8++yzJ71v4MCBLFmypJ4rktr04osvcvfdd7samb/xxht8+eWXzJkz54QeSwBz5swhJyeHlStXunpJxcXFndc5pfY059frAw88wAcffMDo0aO58847zS5HasowoKzg2EisM4ZZ2WCcw0IsPiHgFw5+YZWXcGe4VXXd9W+4sxm9m1utP1URaRgUfJ1EfEwgOw8XsDUtj8HdWppdjoiISL245557uOGGG056nxpHN242m43169czceJE1zY3NzcGDx7ML7/8ctJjFi9eTP/+/bn//vv5/PPPCQ8P55ZbbmHChAm4u7uf0zkBysrKKCsrc93Oz8+vhWfY/DTn1+u8efPOqpG+1CNb8SmCrOOmGBYfd91uq/ljWIOqh1h+YZVB1m/CLb9wZ+jlrj91RcRJPw1OIj46iE83pLItTW/ERESk+QgJCSEkJMTsMqQOZGdnY7fbadmy+gd6LVu25Ndffz3pMfv27ePbb7/l1ltv5auvvmLPnj3cd999lJeXM3ny5HM6J8D06dOZOnXq+T+pZk6vV6kXDjsUZEDeQcg9CHkpzn8L0quHW+XFNT+3p99pgqxw8As9dt03FDystf/8RKRZUPB1EgnRzsZo21JP0+hQREQatUawtoucBX0f647D4SAiIoK33noLd3d3evfuTWpqKjNnzmTy5MnnfN6JEycyfvx41+38/HxiY2NPe4y+z1IX9P8KqCiDvEPHBVvH/5sC+WngKD+7c7lbTx5YuYKs30w59PKt2+cmIlJJwddJdKsMvtLySskpshHi52VyRSIiUluqehUVFxc3+elAzUFxsXOUQdX3VU4uLCwMd3d3Dh8+XG374cOHiYyMPOkxUVFReHp64u7u7trWtWtXMjIysNls53ROAKvVitV6diM39HqVutQsfn6UFVQPsn4bcBVmnPkcFncIjIHgWAiKdf4bGAP+EdVHalkDwGKp++ckIlJDCr5OIsDbk7hQX5KPFLMtLY+BHcPNLklERGqJu7s7wcHBZGZmAuDr64tFb9QbHcMwKC4uJjMzk+Dg4GrhjJzIy8uL3r17s3z5ckaPHg04R3QtX76cBx544KTHDBgwgAULFuBwOHCrbPq8a9cuoqKi8PJyfihY03PWlF6vUheazM8Pw3BOM6yafniyUVuluWc+j4dPZajV6liwFdT6WNAVEKV+WSLSqOkn2CnERwdVBl/5Cr5ERJqYqtEoVX9MS+MVHBx82tFFcsz48eO5/fbb6dOnD/369WPWrFkUFRW5VmQcO3YsMTExTJ8+HYB7772XV199lYceeogHH3yQ3bt38+yzz/LnP//5rM9ZG/R6lbrS4H9+2CucvbR+21/LdfsQVJSc+TzewSeGWa5/WzunIypQFpEmTMHXKcTHBPJlUjpb1edLRKTJsVgsREVFERERQXn5WfYukQbnt9Pw5PRuvPFGsrKymDRpEhkZGfTs2ZOvv/7a1Zw+JSXFNbILIDY2lqVLl/LXv/6VxMREYmJieOihh5gwYcJZn7M26PUqdaFB/PwoL4G81FOP2MpPBcN+5vP4R5480Kq6bQ2o++ciIufEVuHgpz1ZLNt+mLIKB2H+VkL9vAj1txLq70WYn/PfED8vvD31nudcWYxG0NUxPz+foKAg8vLyCAwMrJfH/H5XFrfPWUO7MD++ffjyenlMERERqT1mvH+QmtP3SZq0vFRI33zyHltFWWc+3s0TgmJODLOCqqYmttJqhyKNTLndwcq9R/hicxpLt2WQX1pxVsf5Wz0I9fdyBWNh/l6EVgZjoa7AzLmtha8nHu5uZz5pI1aT9w8a8XUK8ZUN7vdlF1FYVoG/VV8qERERERE5Czn74Id/wOaFpx+15el3+tFa/i3BTaM8RBq7CruDVfty+DIpja+3ZnC0+NgI5vAAKyMSIokK9uFIYRlHCm1kF9lc148UlVFuNygsq6CwrIIDR4rP+HgWC7Tw9ToWhvlbCfPzIqQyKAurFpZZCfT2aNI9NJXmnEKYv5XIQG8y8kvZkZ5P37gQs0sSEREREZGGLGc//PgP2PTBscCrZXdo0ebEEVvBrcGnhfpriTRRdofB6v1H+HJLOl9vzeBIkc11X5i/F8MTorgqMYq+cSG4u53654BhGOSXVnCksIycIhvZlWHYkUJnOFY9JLNxtNiGYUBOkY2cIhu7z6JFpqe7xTV6LMTP68Qpl8ePLvOz4uPVuAJ5BV+nkRATSEZ+KVtT8xR8iYiIiIjIyR09AD/MhM0fgKNy2lKHIXD5Y9Cqj7m1iUi9cTgM1h04yhdb0vgqKYPswjLXfS18PRmWEMXIxCj6tQ0566mIFouFIB9Pgnw8aXcW6+7ZHQZHi22nCMaOBWRV2wrKKii3G2Tkl5KRX3pWNfl6ubtCsN9OuQyrDM+q7gvx8zJ92qWCr9PoFh3ENzsy2ZaWb3YpIiIiIiLS0OSmOKc0bpp/LPBqPwgunwixfc2tTUTqhcNhsPHgUb7Yks5XSekczj8WdgX5eDIsPpKrEqPo3z4Uz3oIgNzdLIT5WwnztwJnXtyitNxOTpGtcoqlMwzLqfz3ZKPLbBUOim12inNKOJhz5pVl597Rlys6R9TCMzt3Cr5OI6Gyz5dWdhQREREREZfcg/DjC7DxfXBU9uppd4Uz8Gp9obm1iUidMwyDTQdz+bIy7ErLOzZSKsDbg6GVYdeA9mF4eTTsJvPenu5EB/sQHexzxn0Nw9lrzDXlsrDMNXos+zcjyY4UOQO0MD/zF+BQ8HUa8TFBAOzJLKS03K7lQ0VEREREmrO8Q87Aa8O/jwu8Lq8MvC4ytTQRqVuGYbA1NZ8vtqTxxZZ0UnOPjXbyt3owpFtLrk6M4pKOYVg9mmZ2YLFYCPD2JMDbkzahfmfc3+4waAhdDBV8nUZ0kDctfD05WlzOrsMFJLYKNrskERERERGpb3mp8NOLsOE9sFc2qG57qTPwanOxubWJSJ0xDIPt6fl8uSWdL5PSq62o6OvlzuCuLbkqMYrLOoVroMxJnK5pf31S8HUaFouF+OggftqTzba0fAVfIiIiIiLNSX4a/PRPWD/vWOAVN9DZtD7uElNLE5G6YRgGOw8XOMOuLensyy5y3eft6cagLs6RXZd3jmh0qxs2Vwq+ziA+JpCf9mSrz5eIiIiISHORn35c4FXZqLrNAOcIr7YDTS1NROrGnswC/rvZObJrT2aha7vVw40rOkdwVWIUg7pG4OulGKWx0XfsDOKjnX2+tLKjiIiIiEgTV5ABP82C9XOhorJZdeuL4YqJzpFeloYxbUdEase+rEK+3JLOF1vS2Xm4wLXdy92NSzuFM7JHFIO6tsTfquikMdN37wyqVnbckZ5Phd2BRz0sPyoiIiIiIvWo4DD8/BKse/dY4BV7kTPwanuZAi+RJuTAkSK+qJzGuD392AAXT3cLAzuGc1X3KIbEtyTQ29PEKqU2Kfg6g7hQP/y83Cmy2dmXXUSnlgFmlyQiIiIiIrWhMNMZeK19FyoqV2hr1c8ZeLW7QoGXSBNxMKeYL5OcYVfScW2MPNwsDOgQxlWJUQztFkmQr8KupkjB1xm4uVnoFh3I2uSjbE3NU/AlIiIiItLYFWbBypdgzTvHBV59nT282v9OgZdIE5CWW8JXSen8d0s6mw/mura7WeDi9mFcnRjF0PhIWvh5mVek1AsFX2chPjqItclH2ZaWzzUXmF2NiIiIiIick6LsyhFe70B5sXNbTG+4/O/QYZACL5FG7nB+qXM1xqR01h846tpuscBFbUO5KjGKYQmRhPlbTaxS6puCr7MQX9nnSys7ioiIiIg0QkVHYOXLsOZtKC9ybovu5Qy8Og5R4CXSiGUWlPL11gy+2JzO2gM5GIZzu8UCfduEcHUPZ9gVEeBtbqFiGgVfZ6FqZcftafk4HAZubvrFKCIiIiLS4BXnwMpXYM1bYCt0bovqCVf8HTpeqcBLpJHKLixzhl1b0li9/1jYBdC7TQuu6h7FiO5RRAYp7BIFX2elY0t/vNzdKCir4ODRYtqE+pldkoiIiIiInEpxDvzyKqx+87jAq4ezh1enYQq8RBowu8PgaLGNI4U2jhSVkVNkI6fIeTunyMberEJW7TuC47iwq2dsMFcnOsOu6GAf84qXBqlGwdfs2bOZPXs2ycnJAMTHxzNp0iSGDx9+xmMXLlzIzTffzKhRo/jss8/OpVbTeLq70TkygKTUPLal5Sv4EhERERFpiEqOwi+vwao3wFbg3BbZ3Rl4dR6hwEvEBOV2B0eLbBypCrCKbOQUOgOt7CIbOYVV253bckvKq43gOpXuMUGusCs2xLfun4g0WjUKvlq1asVzzz1Hx44dMQyDf/3rX4waNYqNGzcSHx9/yuOSk5N5+OGHGThw4HkXbJaEmECSUvPYmprHiO5RZpcjIiIiIiJVSnJh1euwajaU5Tu3tewOlz8GXa5S4CVSi2wVjmpBVU6RjexCGzmVt6tGZlWFXHkl5ef0OMG+noT4eRHmZyXEz4sQfy9C/byICLByaadwDUiRs1aj4GvkyJHVbj/zzDPMnj2bVatWnTL4stvt3HrrrUydOpUff/yR3NzcMz5OWVkZZWVlrtv5+fk1KbNOdIsOAg6yLc38WkREREREhMrAa3Zl4FW5EFVEfGXgdTW4uZlankhjUFpurxZUHakcjXWkcjSWc6RWmet2QVlFjR/DYoEWvs7gKsTPi1B/578hflZCj7sdWhlytfD1xMNdr1+pHefc48tut/PRRx9RVFRE//79T7nfU089RUREBH/4wx/48ccfz+rc06dPZ+rUqedaWp1IqFzZcVtaHoZhYNGnRiIiIiIi5ijNc05nXPWa8zpARDe4bAJ0/b0CL2n2DMMg+UgxydlFJwRXrimGRWXkFNoostlrfH43C9WCqqrRWM5tXoT6W13XQ/y8CPb1wl2LxIlJahx8JSUl0b9/f0pLS/H392fRokV069btpPv+9NNPvPvuu2zatKlGjzFx4kTGjx/vup2fn09sbGxNS61VXSIDcbNAdqGNzIIyWgZqdQgRERERkXpVmu9sWP/Lq1Ca69wW3hUunwBdRynwkmYrt9jGpoO5bDqYy8aUXDYfyiW3+OynGHq4WSpHYFWNvrJWG50VWjk6qyrMCvLxxE1BljQSNQ6+OnfuzKZNm8jLy+Pjjz/m9ttv5/vvvz8h/CooKOD//u//ePvttwkLC6vRY1itVqxWa01Lq1M+Xu50iPBn1+FCtqbmKfgSEREREakvZQWw+g1YeVzgFdbZGXh1G6PAS5oVW4WDXzPynUFXSi4bD+ayP7vohP28PNxoH+5PmGta4clHY4X6WQn08dCsJmmyahx8eXl50aFDBwB69+7N2rVreemll3jzzTer7bd3716Sk5Or9QVzOBzOB/XwYOfOnbRv3/58aq938dFB7DpcyLa0fAZ1bWl2OSIiIiIiTVtZAax5C1a+4lyxESCsk3NKY/wYcHM3tz6ROmYYBqm5JWxMyXWN6NqamkdZheOEfduG+dEzNth16RoViJeHQmGRc+7xVcXhcFRrRF+lS5cuJCUlVdv2xBNPUFBQwEsvvWT61MVzER8dyKKNqWxNzTO7FBERERGRpqusENa+DT+/DCU5zm2hHZ2BV8I1CrykySosq2DLQecorqqwK7vwxL+3g3w86REbTK/YYHq2DqZnq2Ba+HmZULFIw1ej4GvixIkMHz6c1q1bU1BQwIIFC1ixYgVLly4FYOzYscTExDB9+nS8vb1JSEiodnxwcDDACdsbi/joIACt7CgiIiIiUhdsRbDmbVj5MhQfcW4Lae8MvLpfp8BLmhS7w2DX4YLjpiweZXdmIYZRfT8PNwtdowLpGRtMr9bO0Vxtw/w0NVHkLNUo+MrMzGTs2LGkp6cTFBREYmIiS5cuZciQIQCkpKTg1oTn13erXNkxNbeEo0U2JeoiIiIiIrWhoszZtP7nl6A427ktpF3lCK/rwP28J6qImO5wfulxUxaPsuVQHsUnWVExJtiHnq2do7l6tQ4mPjoIb0+FviLnqka/Qd59993T3r9ixYrT3j9v3ryaPFyDE+TjSesQX1Jyitmens+ADjVr2i8iIiIiIr+RkQSf/hEytztvt2gLlz0K3W9Q4CWNVonNzta0PDamHHWN6ErLKz1hPz8vd3oc15erZ+tgIgK0kJpIbdJvkhpKiAkkJaeYral5Cr5ERERERM6Vw+4c4fXds+AoB98wGDwFetwE7p5mVydy1hwOg33ZRWw6mOsKun7NKMDuqD5n0c0CnVoGHDdlsQUdIvxxd9OURZG6pOCrhuKjg/gqKUN9vkREREREzlXOPlh0Dxxc7bzd+SoY+RL4h5tbl8hZyCmyseng0cq+XM6piwWlFSfsFxFgdY3i6hXbgu6tgvC36k9wkfqmV10NxVf2+dqappUdRURERERqxDBg/TxY+jiUF4FXAAx/HnreAmrULQ1QWYWd7Wn5lX25nCstpuQUn7Cft6cb3WOCKqcstqBX62CigrzVgF6kAVDwVUNVKzvuzy6iqKwCPyX2IiIiIiJnVnAYFj8Iu50rwtNmAIyeDS3amFuXSCXDMDiYU8LGg0ddTei3p+VjsztO2LdduB+9Ylu4mtB3jgzA073pLvQm0pgptamh8AArLQOtHM4vY0d6Pn3iQswuSURERESkYdv2GXzxVyjJAXcvGDQJLrofmvCK8NI4FNsqWLb9MF9sSWf9gaPkFNlO2KeFr2dlX64W9IwNpkerYIJ81YdOpLFQ8HUO4qODOJyfybY0BV8iIiIiIqdUkgtLHoUtHzpvR3aHMW9By26mliXNW4XdwU97svl8UxpLt2VQbLO77vN0t9AtOohelass9modTOsQX01ZFGnEFHydg4ToQL79NZOtqerzJSIiIiJyUvtWwGf3Q/4hsLjBJX+Fyx4DDy+zK5NmyDAMthzK47NNqfx3cxrZhcdGdrUJ9WVUzxgu7xxOt6hAvD3dTaxURGqbgq9z0K2yz5dWdhQRERER+Y3yEvhmKqye7bzdoi2MeRNaX2huXdIsHThSxGcb0/h8Uyr7sotc20P8vBiZGMWoXjH0ig3WiC6RJkzB1zlIiHGu7LjrcAFlFXasHvpEQERERESE1A2w6E+Qvct5u/cdcOXTYPU3ty5pVo4UlvFlUjqfbUxlQ0qua7u3pxtXdotkdK9oBnYMVzN6kWZCwdc5iAn2IcjHk7yScnYfLiQhJsjskkREREREzGOvgB9fgB9mgKMC/FvC71+FTleaXZk0EyU2O8t2HOazjan8sCuLCocBgJsFBnQIY0yvGK6Mj8Tfqj+BRZobverPgcViISEmkJ/3HGFrap6CLxERERFpvrJ3O0d5pa533u42Cq76J/iFmluXNHl2h8HKvdks2pjK0q0ZFB3XpL57TBCje8UwskcUEQHeJlYpImZT8HWO4qOD+HnPEfX5EhEREZHmyTBg7TvwvyehogSsQXDVP6D79aB+SVJHDMNgW1o+izY6m9RnFpS57mvVwocxvWIY1TOGDhGaXisiTgq+zlF8tLPP19Y0rewoIiIiIs1Mfhp8dh/s+855u+1lMPp1CGplbl3SZB3MKebzTaks2pjK3qxjTepb+HpyVWIUY3rFcEHrFmpSLyInUPB1juIrV3bckZ6P3WHg7qYfsCIiIiLSDCR9DF+Oh9I88PCGwVOh3x/BTY3CpXYdLbK5mtSvO3DUtd3q4cbgbi0Z0zOGSzuF4+Wh/3sicmoKvs5R2zA/fL3cKbbZ2ZdVSMeWAWaXJCIiIiJSd4pz4Mu/wbZPnbeje8GYtyC8k7l1SZNSWm5n+Y5MFm1M5ftdmZTbnU3qLRYY0D6MUT2jGZYQSYC3p8mVikhjoeDrHLm7WegaFcj6A0fZlpav4EtEREREmq7d38Dn90NhBljc4dJH4NKHwV3hg5w/u8Ng1b4jfLYxla+3ZlBQVuG6Lz46kNE9YxjZI5rIIDWpF5GaU/B1HhKincHX1tQ8RveKMbscEREREZHaZStyNq9f967zdmgH5yivVr3NrUsaPcMw2J6ez+eb0vh8UyqH8481qY8J9mF0r2hG94zRAAMROW8Kvs5DVZ8vrewoIiIiIk3OwbWw6E+Qs9d5u9+fYPAU8PI1tSxp3FJzS/h8UyqfbUxl1+FC1/YgH2eT+tE9Y+jTpgVu6qEsIrVEwdd5iI9xruy4LS0PwzC0goiIiIiINH4VNvhhBvz4AhgOCIiG0a9B+9+ZXZk0UnnF5c4m9ZtSWbM/x7Xdy8ONwV0jGN0zhss6h2P1cDexShFpqhR8nYeOEQF4ulvIL63g0NESYkP06ZeIiIiINGKZv8KiP0L6Zuft7tfDiJng08LcuqTRKS23892vmXy2KZXvfs3CZncAzib1F7UNZUyvGIZ1jyRQTepFpI4p+DoPXh5udGoZwLa0fLam5in4EhEREZHGyeGA1bPhm6lgLwPvYLj6n5BwjdmVSSPicBis3p/D55tS+TIpnYLSY03qu0QGMKZXDL/vGU1UkI+JVYpIc6Pg6zwlRAexLS2fbWn5DO8eZXY5IiIiIiI1k3sQPrsXkn903u4wGH7/KgTqva2cnV8z8lm0MZX/bkojLa/UtT06yJvf94xhdK9oukQGmlihiDRnCr7OU3xMIKyDrWl5ZpciIiIiInL2DAM2fwBLJkBZPnj6wpVPQ587nfPRRE4jPa+Ezzel8dnGVH7NKHBtD/D24KruUYzuFUO/uBA1qRcR0yn4Ok9a2VFEREREGp2ibPjvQ/DrF87brfrCmDchtL25dUmD9+PuLF7/bi+r9h/BMJzbvNzduKJLOGN6xXB55wi8PdWkXkQaDgVf56lrVAAWC2QVlJGZX0pEoLfZJYmIiIiInNrOr2Hxg1CUCW4ecPljMOCv4K4/DeTUCssqeObLHXywJsW17cK2IYzuFcOIhCiCfNWkXkQaJv12O0++Xh60D/dnT2Yh29LyFXyJiIiISMNUVgBL/w4b3nPeDu8C17wFUT3MrUsavF/2HuGRjzdz6GgJALf3b8MfL2tPTLCa1ItIw6fgqxbERweyJ7OQral5XNElwuxyRERERESqO/ALLPoT5B4ALND/fvjdk+CpD23l1Epsdp7/+lfmrUwGoFULH2Ze14P+7UPNLUxEpAYUfNWChOggPt+Upj5fIiIiItKwVJTBd8/Azy8DBgTFwujZ0Hag2ZVJA7f+QA4Pf7SF/dlFANxyYWv+PqIr/lb9CSkijYt+atWC+Gjn0rxa2VFEREREGoyMrc5RXoe3Om/3vBWGTQfvIHPrkgattNzOP5ft4u0f9+EwIDLQm+evS+SyTuFmlyYick4UfNWCqpUdDx0tIa+4XI0dRURERMQ8DjusfBm+fQYc5eAbCiNfgq4jza5MGrikQ3mM/88mdmcWAnDtBa2YNLIbQT76+0ZEGi8FX7UgyNeTVi18OHS0hG3peVzcPszskkRERESkOcrZD5/dCym/OG93Gg6/fxn81YdWTs1W4eDVb3fz2oq92B0GYf5Wpl/TnSHdWppdmojIeVPwVUsSooOcwVdqvoIvEREREalfhuFcrXHp38FWCF7+MOw56HUbWCxmVycN2I70fP72n81sT3f2K746MYppoxJo4edlcmUiIrVDwVctiY8O5OttGWxTny8RERERqU+FmbD4Qdj1tfN264thzGxoEWdqWdKwVdgdvPnDPmZ9s4tyu0ELX0+mjU7g6sRos0sTEalVCr5qSUKMs8/XVq3sKCIiIiL1Zfti+OIvUHwE3L3gd09A/wfAzd3syqQB25NZyN8+2szmg7kADOnWkmfHdCc8wGpuYSIidUDBVy2pWtlxX1YhxbYKfL30pRURERGROlKaB0smwOYPnLdbJsCYNyEywdy6pEGzOwzm/ryfmUt3UlbhIMDbg6m/j2dMrxgsmhIrIk2U0plaEhHoTXiAlayCMnakF9C7TQuzSxIRERGRpihlNXzyB8g7CFhgwENwxd/BQ6N15NQOHCni4Y82szb5KACXdgrn+Wu7ExXkY3JlIiJ1S8FXLYqPDmTFziy2p+Up+BIRERGR2lecAx/cBCU5zh5eo9+ANv3NrkoaMIfDYP7qAzz71a+UlNvx83Lniau7cVPfWI3yEpFmQcFXLUqIDmLFziy2pqrPl4iIiIjUge+edYZe4V3grm/AGmB2RdKApeaW8OjHm/l5zxEA+rcLZcZ1icSG+JpcmYhI/VHwVYuq+nxtS9fKjiIiIiJSyzKSYN27zuvDZyj0klMyDIOP1h3iqS+2U1hWgbenG48N68LY/nG4uWmUl4g0Lwq+alHVyo47MwqwVTjw8nAzuSIRERERaRIMA756FAwHdBsF7S4zuyJpoA7nl/LYJ1v4bmcWABe0DuaFG3rSNszP5MpERMyh4KsWtWrhQ6C3B/mlFezOLCA+OsjskkRERESkKUj6GFJWgocPXPmM2dVIA2QYBp9vSmPy4m3klZTj5e7G367sxF0D2+GuUV4i0owp+KpFFouF+Oggftl3hG2p+Qq+REREROT8lRXCsied1wf+DYJjza1HGpzswjIeX5TE0m2HAUhsFcQL1/egY0tNhxURUfBVy+KjA53BV1oeoDclIiIiInKefpgJBenOVRwvftDsaqSBWZKUzuOfbSWnyIaHm4WHBnXknsvb4+mutisiIgD6aVjLqvp8bU3Tyo4iIiLSsLz22mvExcXh7e3NhRdeyJo1a06577x587BYLNUu3t7e1fYZN27cCfsMGzasrp9G85K9G355zXl92HPg6X36/aXZyC228ecPNnLv/A3kFNnoEhnA5w8M4MFBHRV6iYgcRyO+alnVyo470vOxOwzNpxcREZEG4cMPP2T8+PG88cYbXHjhhcyaNYuhQ4eyc+dOIiIiTnpMYGAgO3fudN22WE58XzNs2DDmzp3rum21Wmu/+ObKMODrx8BRDh2GQCeFiuK0fMdhHvs0iayCMtwscO/l7fnzoI5YPdzNLk1EpMFR8FXL2oX74+3pRrHNzv7sIjpE+JtdkoiIiAgvvvgid999N3fccQcAb7zxBl9++SVz5szhscceO+kxFouFyMjI057XarWecZ/jlZWVUVZW5rqdn69R8qe0cwns+QbcPJ2jvU4SPErzkl9azrT/buej9YcAaB/uxws39KRnbLC5hYmINGAaA1vL3N0sdI1yjvpy9vkSERERMZfNZmP9+vUMHjzYtc3NzY3Bgwfzyy+/nPK4wsJC2rRpQ2xsLKNGjWLbtm0n7LNixQoiIiLo3Lkz9957L0eOHDltLdOnTycoKMh1iY1VT9STKi9xjvYCuPgBCOtgbj1iup92ZzPsnz/w0fpDWCxw1yVt+fLPAxV6iYicgYKvOpBQuZrjNvX5EhERkQYgOzsbu91Oy5Ytq21v2bIlGRkZJz2mc+fOzJkzh88//5z3338fh8PBxRdfzKFDh1z7DBs2jPfee4/ly5fz/PPP8/333zN8+HDsdvspa5k4cSJ5eXmuy8GDB2vnSTY1K1+B3AMQEA0DHza7GjFRUVkFT3yWxG3vriYtr5Q2ob58+Mf+PHF1N7w9NbVRRORMNNWxDlT1+dKILxEREWms+vfvT//+/V23L774Yrp27cqbb77JtGnTALjppptc93fv3p3ExETat2/PihUrGDRo0EnPa7Va1QfsTHJT4McXndevnAZWtc5orlbvO8IjH28hJacYgLH92/DY8C74eunPOBGRs6WfmHXAtbJjaj6GYZy0EayIiIhIfQkLC8Pd3Z3Dhw9X23748OGz7s/l6elJr1692LNnzyn3adeuHWFhYezZs+eUwZechaWPQ0UJtBkACdeaXY2YoLTczsylO5nz834MA2KCfZhxXSIDOoSZXZqISKOjqY51oGNLfzzcLOSVlJOaW2J2OSIiItLMeXl50bt3b5YvX+7a5nA4WL58ebVRXadjt9tJSkoiKirqlPscOnSII0eOnHYfOYN9K2DHYrC4wfAZamjfDG1MOcqIl3/k3Z+codeNfWL5+i8DFXqJiJwjjfiqA1YPdzq1DGB7ej5bU/Np1cLX7JJERESkmRs/fjy33347ffr0oV+/fsyaNYuioiLXKo9jx44lJiaG6dOnA/DUU09x0UUX0aFDB3Jzc5k5cyYHDhzgrrvuApyN76dOncq1115LZGQke/fu5dFHH6VDhw4MHTrUtOfZqNnL4atHndf73gWRCebWI/WqrMLOS9/s5o3v9+IwICLAynPXdud3XVqe+WARETklBV91JD46kO3p+WxPy2NYwtkv8S0iIiJSF2688UaysrKYNGkSGRkZ9OzZk6+//trV8D4lJQU3t2OTAY4ePcrdd99NRkYGLVq0oHfv3qxcuZJu3boB4O7uzpYtW/jXv/5Fbm4u0dHRXHnllUybNk09vM7V6jcheyf4hsIVfze7GqlHW1PzePijzfyaUQDA6J7RTPl9PMG+XiZXJiLS+FkMwzDMLuJM8vPzCQoKIi8vj8DAQLPLOSv/WpnM5MXb+F2XCOaM62t2OSIiIs1OY3z/0Bzp+1Sp4DC80htsBTDyZeh9u9kVST0otzt47bs9vPrtHiocBqF+XjwzJoFhCZouLCJyOjV5/6ARX3VEKzuKiIiIyFn7Zooz9Iq+AHr9n9nVSD3YdbiA8f/ZxNbUfACGJ0Ty9OgEQv01YlJEpDYp+KojXaMCsVjgcH4ZWQVlhAfoF5iIiIiInETKati8wHl9xExw0/pTTZndYfDWD/v457Jd2OwOgnw8eWpUPL/vEa3V4EVE6oCCrzriZ/WgbZgf+7KK2JaWx+WdI8wuSUREREQaGocdljzivN7zNmjVx9x6pE7tyyrkbx9tZmNKLgC/6xLB9Gu60zLQ29zCRESasBp9nDR79mwSExMJDAwkMDCQ/v37s2TJklPu//bbbzNw4EBatGhBixYtGDx4MGvWrDnvohuLhOggALal5ZtciYiIiIg0SBveg/TNYA2CwZPNrkbqiMNhMOen/Qx/6Uc2puTib/VgxnWJvHt7H4VeIiJ1rEbBV6tWrXjuuedYv34969at43e/+x2jRo1i27ZtJ91/xYoV3HzzzXz33Xf88ssvxMbGcuWVV5KamlorxTd06vMlIiIiIqdUnAPLn3Jev2Ii+GuGQFN0MKeYm99exVNfbKeswsElHcJY+tdLuaFPrKY2iojUgxpNdRw5cmS128888wyzZ89m1apVxMfHn7D//Pnzq91+5513+OSTT1i+fDljx449h3Ibl4QY54ivqoaVIiIiIiIu3z0DJTkQ3hX63mV2NVLLDMNg4dqDTPtiO8U2O75e7kwc0ZXbLmytwEtEpB6dc48vu93ORx99RFFREf379z+rY4qLiykvLyckJOS0+5WVlVFWVua6nZ/fOIOjqhFfKTnF5JWUE+TjaXJFIiIiItIgpG+BdXOc10fMAHe9T2xKisoqmPhpEos3pwHQLy6Emdcn0ibUz+TKRESanxoHX0lJSfTv35/S0lL8/f1ZtGgR3bp1O6tjJ0yYQHR0NIMHDz7tftOnT2fq1Kk1La3BCfb1IibYh9TcEran5dO/fajZJYmIiIiI2QwDljwKhgPix0DbS82uSGrRnswC7nl/A3syC3F3s/Do0M7cPbAdbm4a5SUiYoYar5XcuXNnNm3axOrVq7n33nu5/fbb2b59+xmPe+6551i4cCGLFi3C2/v0DRwnTpxIXl6e63Lw4MGaltlgqM+XiIiIiFST9BGk/AKevnDl02ZXI7Xov5vT+P2rP7Mns5CIACsf3H0Rf7qsvUIvERET1XjEl5eXFx06dACgd+/erF27lpdeeok333zzlMf84x//4LnnnuObb74hMTHxjI9htVqxWq01La1BSogJ4n/bD2tlRxERERGBsgL435PO6wP/BkGtzK1HaoWtwsGzX+1g3spkAC5qF8LLN/ciIkArNoqImO2ce3xVcTgc1fpx/daMGTN45plnWLp0KX369Dnfh2t0NOJLRERERFy+nwGFGdCiLVz8oNnVSC1Iyy3hvvkb2HQwF4D7Lm/P+CGd8HCv8eQaERGpAzUKviZOnMjw4cNp3bo1BQUFLFiwgBUrVrB06VIAxo4dS0xMDNOnTwfg+eefZ9KkSSxYsIC4uDgyMjIA8Pf3x9/fv5afSsNUtbLjnsxCSmx2fLzcTa5IREREREyRvRtWzXZeH/YceDSNGQ7N2Q+7snho4UaOFpcT6O3Bizf0ZHC3lmaXJSIix6lR8JWZmcnYsWNJT08nKCiIxMREli5dypAhQwBISUnBze3YJxuzZ8/GZrNx3XXXVTvP5MmTmTJlyvlX3whEBFgJ8/ciu9DGrxn59GrdwuySRERERKS+VTW0d5RDx6HQeZjZFcl5cDgMXvl2D7OW78IwICEmkNdv6U3rUF+zSxMRkd+oUfD17rvvnvb+FStWVLudnJxc03qaHIvFQnx0EN/vymJrmoIvERERkWbp1y9h77fg7gXDpptdjZyHnCIbf/lwEz/sygLg5n6xTB4Zj7enZnaIiDRE593jS84sPjqQ73dlsV19vkRERESan/ISWDrReb3/AxDa3tx65JxtOpjLfe+vJy2vFKuHG8+M6c51vbVAgYhIQ6bgqx5U9fnamqqVHUVERESanZ9fhtwUCIyBSx82uxo5B4Zh8P6qAzz1xXbK7QZxob7Mvq03XaMCzS5NRETOQMFXPaha2XFnRgHldgeeWuFFREREpHk4egB+etF5/cpp4OVnbj1SY0VlFfx9URKfb0oDYFh8JDOuTyTQ29PkykRE5Gwo+KoHrUN8CfD2oKC0gt2HC+kWrU+GRERERJqF/z0OFaUQNxDirzG7GqmhPZkF3Pv+BnZnFuLuZuGxYV24a2BbLBaL2aWJiMhZ0tCjemCxWOhWOQx6m/p8iYiIiDQPe7+FHf8FizsMfx4UljQq/92cxu9f/ZndmYVEBFj54O6LuPvSdgq9REQaGQVf9aSqz9e2NPX5EhEREWnyKmywZILzer+7oWW8ufXIWbNVOJiyeBsPfrCRYpudi9qF8MWfL6Ff2xCzSxMRkXOgqY71pKrPl0Z8iYiIiDQDa96E7F3gGwaXTzS7GjlLabkl3L9gAxtTcgG47/L2jB/SCQ/16BURabQUfNWTqhFf29PycTgM3Nw0RFpERESkSSrIgBXPOa8PngI+wWZWI2fpx91ZPLRwEzlFNgK9PXjxhp4M7tbS7LJEROQ8KfiqJ+3C/LB6uFFks5N8pIh24f5mlyQiIiIidWHZZLAVQkxv6Hmr2dXIGTgcBq98u4dZy3dhGM6ZGrNv7U3rUF+zSxMRkVqgMbv1xMPdja6VDe63qs+XiIiISNOUsgq2LHReHz4T3PR2uyE7WmTjjnlr+ec3ztDr5n6xfHLvxQq9RESaEP0mrkfq8yUiIiLShDns8NXDzuu9/g9a9Ta3HjmtTQdzufqVn/h+VxZWDzf+cX0Ppl+TiLenu9mliYhILdJUx3rkWtkxVSO+RERERJqc9fMgIwmsQTBostnVyCkYhsH7qw7w1BfbKbcbxIX6Mvu23q7ZGSIi0rQo+KpHx4/4MgwDi0UN7kVERESahOIc+Haa8/oVfwf/cHPrkZMqKqvg74uS+HxTGgDD4iOZcX0igd6eJlcmIiJ1RcFXPerUMgAPNwtHi8tJyyslJtjH7JJEREREpDZ8Ow1KjkJEN+h7l9nVyEnsySzk3vfXszuzEHc3C48N68JdA9vqw2gRkSZOwVc98vZ0p0OEP79mFLAtNU/Bl4iIiEhTkL4Z1s11Xh8xE9z1Fruh+WJLGhM+3kKRzU5EgJVXb7mAfm1DzC5LRETqgZrb17OqPl9a2VFERESkCTAM+OoRwID4ayDuErMrkuPYKhxMWbyNBxZspMhm56J2IXzx50sUeomINCP6OKqexUcH8vF62K6VHUVEREQavy0fwsHV4OkLVz5tdjVynLTcEu5fsIGNKbkA3Hd5e8YP6YSHuz77FxFpThR81TPXiC+t7CgiIiLSuJXmw7JJzuuXPgxBMebWIy4/7s7ioYWbyCmyEeDtwT9v6Mngbi3NLktEREyg4KuedY0KxGKBjPxSsgvLCPO3ml2SiIiIiJyLH2ZA4WEIaQf9HzC7GgEcDoNXv9vDP7/ZhWE4Z1vMvrU3rUN9zS5NRERMonG+9czf6kHbUD8AtqnPl4iIiEjjlLUTVs12Xh/2PHjow0yzHS2ycce8tby4zBl63dwvlk/uvVihl4hIM6fgywTdogMB2KY+XyIiIiKNj2HAkgngqIBOw6DTlWZX1OxtOpjL1a/8xPe7srB6uPGP63sw/ZpEvD3dzS5NRERMpuDLBFV9vrapz5eIiIhI4/PrF7DvO3D3gmHTza6mWTMMg3//ksz1b6wkNbeEuFBfPrt/ANf1bmV2aSIi0kCox5cJ4jXiS0RERKRxshXD1393Xr/4z87+XmKKYlsFf/80ic82pQEwNL4lM6/vQaC3p8mViYhIQ6LgywTx0c4RX8lHiskvLdcvZxEREZHG4ueXIC8FAlvBwPFmV9Ns7cks5N7317M7sxB3NwuPDevCXQPbYrFYzC5NREQaGE11NEGInxfRQd4A7FCDexEREZHG4Wgy/DzLeX3o0+DlZ2Y1zdYXW9IY9epP7M4sJCLAygd3X8Tdl7ZT6CUiIiel4Msk8ZV9vrYq+BIRERFpHJY+DhWlEDcQuo02u5pmx1bhYMribTywYCNFNjsXtQvhiz9fQr+2IWaXJiIiDZimOpokPjqQZdsPq8+XiIiISGOwZ7mzqb3FHUbMBI0uqlfpeSXcP38DG1JyAbj38vb8bUgnPNz1Ob6IiJyegi+TJERrZUcRERGRRqHCBksmOK9f+CeI6GpuPc3Mj7uzeGjhJnKKbAR4e/DiDT0Z0q2l2WWJiEgjoeDLJPExzpUd92QVUlpux9vT3eSKREREROSkVs+GI7vBLxwuf8zsapoNh8Pg1e/28M9vdmEYzhkTs2/tTetQX7NLExGRRkTBl0kiA70J9fPiSJGNXzMK6BkbbHZJIiIiIvJb+enw/Qzn9cFTwDvI1HKai6NFNv76n02s2JkFwM39Ypk8Ml4fFouISI1pUrxJLBYL3aKdo77U50tERESkgfpmMtgKIaYP9LjF7Gqahc0Hc7n6lZ9YsTMLq4cb/7i+B9OvSVToJSIi50QjvkyUEBPEj7uz2ao+XyIiIiINz4FfYMuHgMXZ0N5NnxnXJcMweH91CtP+ux2b3UFcqC+zb+tN16hAs0sTEZFGTMGXieIrR3xt14gvERERkYbFYYcljzivX/B/EHOBufU0cSU2OxM/3cJnm9IAGBrfkpnX9yDQ29PkykREpLFT8GWiqpUdd2QUUG534KnlmEVEREQahvVzISPJ2dNr0GSzq2nSDMPgkY8388WWdNzdLDw2rAt3DWyLxWIxuzQREWkClLSYqHWIL/5WD2wVDvZmFZpdjoiIiIgAFB2B5dOc1694AvzCzK2niZv7czJfbEnHw83Cv+7ox92XtlPoJSIitUbBl4nc3I41uFefLxEREZEG4ttpUJoLEfHQ506zq2nS1iXn8OxXOwD4+4iuXNJRIaOIiNQuBV8mi9fKjiIiIiINR9omWD/PeX3ETHBXZ5C6kllQyn3zN1DhMLg6MYo7BsSZXZKIiDRBCr5MVtXna5tGfImIiIiYy+GArx4BDEi4DuIGmF1Rk1Vhd/Dggo1kFpTRMcKf569N1PRGERGpEwq+TBYfU7myY3o+DodhcjUiIiIizdiWD+HQGvD0gyunmV1NkzZj6U5W78/Bz8ud2bf1xs+qkXUiIlI3FHyZrEO4P1YPNwrLKjiQU2x2OSIiIiLNU2k+LJvkvH7ZIxAYbW49TdhXSem89cM+AGZe34MOEf4mVyQiIk2Zgi+Tebi70SUyAFCfLxERERHTfP88FGVCSHu46D6zq2my9mQW8shHmwH446XtGNE9yuSKRESkqVPw1QDExzj7fGllRxERERETZO2E1W84rw9/Hjys5tbTRBWVVXDP++spstm5sG0Ijw7tbHZJIiLSDCj4agC0sqOIiIiISQwDljwKjgroPAI6DjG7oibJMAwmfLKFPZmFRARYeeWWXni4608RERGpe/pt0wC4VnZMy8cw1OBeREREpN7sWAz7VoC7FYY+a3Y1Tdbcn5P5Yks6Hm4WXr/1AiICvM0uSUREmgkFXw1A58gA3N0s5BTZyMgvNbscERERkebBVgxLH3deH/BnCGlrbj1N1LrkHJ79agcAfx/RlT5xISZXJCIizYmCrwbA29OdjpWr2ajPl4iIiEg9+XkW5B2EoFi4ZLzZ1TRJmQWl3Dd/AxUOg6sTo7hjQJzZJYmISDOj4KuB6KY+XyIiIiL1J2c//DTLef3Kp8HL19RymqIKu4MHF2wks6CMjhH+PH9tIhaLxeyyRESkmVHw1UBU9fnSiC8RERGRerD0cbCXQdvLoNsos6tpkmYs3cnq/Tn4ebkz+7be+Fk9zC5JRESaIQVfDUTVyo7bNeJLREREpG7t/gZ2fgluHjB8BmgUUq1bkpTOWz/sA2Dm9T3oUNnWQ0REpL4p+GogqqY6puWVklNkM7kaERERkSaqwgZfT3Be7/cniOhibj1N0J7MQh7+aDMAdw9sy4juUSZXJCIizZmCrwYiwNuTuFBnbwn1+RIRERGpI6tehyN7wC8CLp9gdjVNTlFZBfe8v54im51+bUOYMEzBooiImEvBVwMSH+Ps87UtTX2+RERERGpdfhr8MNN5fchU8A4yt54mxjAMJnyyhT2ZhUQEWHn1ll54uOvPDRERMZd+EzUgVX2+tqZqxJeIiIhIrVs2CWyF0KovJN5kdjVNztyfk/liSzoebhZev/UCIgK8zS5JREREwVdDUrWy43aN+BIRERGpXQdWQtJHgAVGzAQ3vQ2uTeuSc3j2qx0A/H1EV/rEhZhckYiIiJN+4zcgVSO+9mUXUVhWYXI1IiIiIk2EvQK+esR5vfftEN3L3HqamMyCUu6bv4EKh8HViVHcMSDO7JJERERcFHw1IKH+VqKCnEPCd6Rr1JeIiIhIrVg/Fw5vBe9g+N0ks6tpUirsDh5csJHMgjI6Rvjz/LWJWCwWs8sSERFxUfDVwKjPl4iIiNSV1157jbi4OLy9vbnwwgtZs2bNKfedN28eFoul2sXbu3rPJsMwmDRpElFRUfj4+DB48GB2795d10+jZoqOwLdPO6//7gnwCzW3niZmxtKdrN6fg5+XO7Nv642f1cPskkRERKpR8NXAxEdrZUcRERGpfR9++CHjx49n8uTJbNiwgR49ejB06FAyMzNPeUxgYCDp6emuy4EDB6rdP2PGDF5++WXeeOMNVq9ejZ+fH0OHDqW0tLSun87Z+/YpKM2Flt2hz51mV9OkLElK560f9gEw8/oedIjwN7kiERGRE9Uo+Jo9ezaJiYkEBgYSGBhI//79WbJkyWmP+eijj+jSpQve3t50796dr7766rwKbuo04ktERETqwosvvsjdd9/NHXfcQbdu3XjjjTfw9fVlzpw5pzzGYrEQGRnpurRs2dJ1n2EYzJo1iyeeeIJRo0aRmJjIe++9R1paGp999lk9PKOzkLoB1v/LeX3EDHBzN7eeJmRPZiGPfLwFgLsHtmVE9yiTKxIRETm5GgVfrVq14rnnnmP9+vWsW7eO3/3ud4waNYpt27addP+VK1dy880384c//IGNGzcyevRoRo8ezdatW2ul+KYoIcY54mtPZiGl5XaTqxEREZGmwGazsX79egYPHuza5ubmxuDBg/nll19OeVxhYSFt2rQhNjb2hPd8+/fvJyMjo9o5g4KCuPDCC097zrKyMvLz86td6o4BEV2h+w3Q5uI6fJzmpaisgnveX09hWQX92oYwYVgXs0sSERE5pRoFXyNHjmTEiBF07NiRTp068cwzz+Dv78+qVatOuv9LL73EsGHDeOSRR+jatSvTpk3jggsu4NVXX62V4puiqCBvWvh6UuEw2HW4wOxyREREpAnIzs7GbrdXG7EF0LJlSzIyMk56TOfOnZkzZw6ff/4577//Pg6Hg4svvphDhw4BuI6ryTkBpk+fTlBQkOsSGxt7Pk/t9GJ6w59+hKv+UXeP0cwYhsGET7awJ7OQiAArr97SCw93dU8REZGG65x/S9ntdhYuXEhRURH9+/c/6T6//PJLtU8BAYYOHXraTwGhvj8JbFgsFotr1Jf6fImIiIhZ+vfvz9ixY+nZsyeXXXYZn376KeHh4bz55pvndd6JEyeSl5fnuhw8eLCWKj4Fdw/wDqrbx2hG5v6czBdb0vFws/DarRcQEeB95oNERERMVOPgKykpCX9/f6xWK/fccw+LFi2iW7duJ903IyOjxp8CQj1/EtgAdVOfLxEREalFYWFhuLu7c/jw4WrbDx8+TGRk5Fmdw9PTk169erFnzx4A13E1PafVanX1i626SOOwLjmHZ7/aAcDEEV3pGxdickUiIiJnVuPgq3PnzmzatInVq1dz7733cvvtt7N9+/ZaLarePwlsYBK0sqOIiIjUIi8vL3r37s3y5ctd2xwOB8uXLz/lyP3fstvtJCUlERXlbGLetm1bIiMjq50zPz+f1atXn/U5pfHILCjlvvkbqHAYXJ0YxZ0D4swuSURE5Kx41PQALy8vOnToAEDv3r1Zu3YtL7300kmHvUdGRp7TJ4tWqxWr1VrT0pqMqpUdd6TnU2F3qG+CiIiInLfx48dz++2306dPH/r168esWbMoKirijjvuAGDs2LHExMQwffp0AJ566ikuuugiOnToQG5uLjNnzuTAgQPcddddgLM9w1/+8heefvppOnbsSNu2bXnyySeJjo5m9OjRZj1NqQMVdgcPLthIZkEZHSP8ef7aRCwWi9lliYiInJUaB1+/5XA4KCsrO+l9/fv3Z/ny5fzlL39xbVu2bJk+BTyDuFA//LzcKbLZ2ZddRKeWAWaXJCIiIo3cjTfeSFZWFpMmTSIjI4OePXvy9ddfu9pSpKSk4OZ27MO2o0ePcvfdd5ORkUGLFi3o3bs3K1eurNbi4tFHH6WoqIg//vGP5Obmcskll/D111/j7a2+T03JjKU7Wb0/Bz8vd2bf1hs/63n/CSEiIlJvLIZhGGe788SJExk+fDitW7emoKCABQsW8Pzzz7N06VKGDBlywieFK1eu5LLLLuO5557jqquuYuHChTz77LNs2LCBhISEsy4yPz+foKAg8vLymk0fiOvfWMna5KO8eEMPrrmgldnliIiINDrN8f1DY6TvU8O2JCmde+dvAOD1Wy9gRPcokysSERGp2fuHGs2hy8zMZOzYsXTu3JlBgwaxdu1aV+gFzk8K09PTXftffPHFLFiwgLfeeosePXrw8ccf89lnn9Uo9Gqu4tXnS0RERERMtDerkEc+3gLA3QPbKvQSEZFGqUbjlN99993T3r9ixYoTtl1//fVcf/31NSpKjvX50sqOIiIiIlLfisoquOff6yksq6Bf2xAmDOtidkkiIiLnRF3TG6iEGOeIr+1p+TgcZz0bVURERETkvBiGwYRPtrA7s5CIACuv3tJLiy2JiEijpd9gDVSHCH+8PNwoKKvg4NFis8sRERERkWZi7s/JfLElHQ83C6/degERAVqsQEREGi8FXw2Up7sbXSKdqzmqz5eIiIiI1Id1yTk8+9UOACaO6ErfuBCTKxIRETk/Cr4aMPX5EhEREZH6kllQyn3zN1DhMLg6MYo7B8SZXZKIiMh5U/DVgGllRxERERGpDxV2Bw8u2EhmQRkdIvx5/tpELBaL2WWJiIicNwVfDVjViK9taXkYhhrci4iIiEjdmLF0J6v35+Dn5c4bt/XGz1qjxd9FREQaLAVfDVjXqEDc3SxkF9rILCgzuxwRERERaYKWJKXz1g/7AJh5fQ86RPibXJGIiEjtUfDVgHl7utM+3A9Qny8RERERqX17swp55OMtANw9sC0jukeZXJGIiEjtUvDVwCWoz5eIiIiI1IGisgru+fd6Cssq6Nc2hAnDuphdkoiISK1T8NXAddPKjiIiIiJSywzDYMInW9idWUhEgJVXb+mFh7v+NBARkaZHv90auIQYjfgSERERkdo19+dkvtiSjoebhdduvYCIAG+zSxIREakTCr4auKoRX6m5JRwtsplcjYiIiIg0duuSc3j2qx0ATBzRlb5xISZXJCIiUncUfDVwgd6etAn1BWB7ukZ9iYiIiMi5yywo5b75G6hwGFydGMWdA+LMLklERKROKfhqBOLV50tEREREzlOF3cGDCzaSWVBGhwh/nr82EYvFYnZZIiIidUrBVyMQr5UdRUREROQ8zVi6k9X7c/DzcueN23rjZ/UwuyQREZE6p+CrEXCN+ErTiC8RERERqbklSem89cM+AGZe34MOEf4mVyQiIlI/FHw1AlUjvvZnF1FUVmFyNSIiIiLSmOzNKuSRj7cAcNclbRnRPcrkikREROqPgq9GIDzASstAK4YBO9TgXkRERETOUlFZBff8ez2FZRX0axvChOFdzC5JRESkXin4aiQS1OdLRERERGrAMAwe+zSJ3ZmFRARYefWWXni66+2/iIg0L/rN10hoZUcRERERqYl5K5P57+Y0PNwsvHbrBUQEeJtdkoiISL1T8NVIxMdoxJeIiIiInJ11yTk88+UOACaO6ErfuBCTKxIRETGHgq9GomrE167DBZRV2E2uRkREREQaqsyCUu6bv4EKh8HViVHcOSDO7JJERERMo+CrkYgJ9iHY15MKh8Huw4VmlyMiIiIiDVCF3cGDCzaSWVBGhwh/nr82EYvFYnZZIiIiplHw1UhYLBb1+RIRERGR05qxdCer9+fg5+XOG7f1xs/qYXZJIiIiplLw1YhoZUcREREROZUlSem89cM+AGZe34MOEf4mVyQiImI+BV+NSLeqEV9pGvElIiIiIsfszSrkkY+3AHDXJW0Z0T3K5IpEREQaBgVfjUhC5cqOO9LzsTsMk6sRERERkYagqKyCe/69nsKyCvrFhTBheBezSxIREWkwFHw1Im1D/fD1cqe03MG+LDW4FxEREWnuDMPgsU+T2J1ZSHiAlVdv6YWnu97ii4iIVNFvxUbEzc1CtyjndEf1+RIRERGReSuT+e/mNDzcLLx+6wVEBHqbXZKIiEiDouCrkdHKjiIiIiICsC45h2e+3AHAxBFd6RsXYnJFIiIiDY+Cr0YmXis7ioiIiDR7mQWl3Dd/AxUOg6sTo7hzQJzZJYmIiDRICr4amfiYqqmOeRiGGtyLiIiINDcVdgcPLthIZkEZHSL8ef7aRCwWi9lliYiINEgKvhqZjhEBeLpbyC+t4NDRErPLEREREZF6Nuub3azen4Oflztv3NYbP6uH2SWJiIg0WAq+GhkvDzc6RwYA6vMlIiIi0twYhsHCtSkAPDOmOx0i/E2uSEREpGFT8NUIxUepz5eIiIhIc5SWV0p2oQ0PNwvDEiLNLkdERKTBU/DVCCVU9vnamqYRXyIiIiLNyZaDuQB0ahmAt6e7ucWIiIg0Agq+GqFuWtlRREREpFnaUtnqokdskMmViIiINA4KvhqhrlEBuFkgq6CMzPxSs8sRERERkXqy5VAuAN1jgk2tQ0REpLFQ8NUI+Xp50C7c2chUo75EREREmgfDMNhyyDniK7GVRnyJiIicDQVfjVRCdGWfL63sKCIiItIsJB8ppqC0otoq3yIiInJ6Cr4aqXj1+RIRERFpVqqmOXaLCsTTXW/jRUREzoZ+YzZS8VrZUURERKRZqZrm2EPTHEVERM6agq9GKj7K+Ybn0NES8orLTa5GREREROpaUmXw1b1VsLmFiIiINCIKvhqpIF9PYkN8ANimUV8iIiIiTZrdYbhG+mvEl4iIyNlT8NWIVY36Up8vERERkaZtT2YhxTY7vl7urtW9RURE5MwUfDViCerzJSIiItIsVDW2T4gJwt3NYm4xIiIijYiCr0ZMKzuKiIiINA9Vje0TYzTNUUREpCYUfDViVSs77s0qpNhWYXI1IiIiIlJXtqRWBl+xweYWIiIi0sgo+GrEIgK8CQ+wYhiwI73A7HJEREREpA7YKhzsqBzhr8b2IiIiNaPgq5FLiHaO+tLKjiIiIiJN067DBdjsDoJ8PGkd4mt2OSIiIo2Kgq9GztXnK1V9vkRERESaos2Vje0TWwVhsaixvYiISE0o+GrktLKjiIiISNOWVNnYvrsa24uIiNSYgq9GrmrE167DBdgqHCZXIyIiIiK1bXPVio6tgs0tREREpBFS8NXItWrhQ6C3B+V2g12H1eBeREREpCkpsdld7/ES1dheRESkxhR8NXIWi8U16mt7mvp8iYiIiDQl29PzsTsMwvytRAV5m12OiIhIo6PgqwlQny8RERGRpmmLGtuLiIicFwVfTYBrZUeN+BIRERFpUpJc/b00zVFERORcKPhqAqpGfG1Pcw6FFxEREZGmYXPliK8eamwvIiJyTmoUfE2fPp2+ffsSEBBAREQEo0ePZufOnWc8btasWXTu3BkfHx9iY2P561//Smlp6TkXLdW1DfPHx9OdknI7+7OLzC5HRERERGpBQWk5+yrf23XXiC8REZFzUqPg6/vvv+f+++9n1apVLFu2jPLycq688kqKik4dtixYsIDHHnuMyZMns2PHDt59910+/PBD/v73v5938eLk7maha1QAANvU50tERESkSdiamo9hQEywD2H+VrPLERERaZQ8arLz119/Xe32vHnziIiIYP369Vx66aUnPWblypUMGDCAW265BYC4uDhuvvlmVq9efY4ly8nERwexISWXbWn5jOoZY3Y5IiIiInKeqhrbd4/RaC8REZFzdV49vvLynKOLQkJCTrnPxRdfzPr161mzZg0A+/bt46uvvmLEiBGnPKasrIz8/PxqFzk918qOqRrxJSIiItIUbKl8X5cYq+BLRETkXNVoxNfxHA4Hf/nLXxgwYAAJCQmn3O+WW24hOzubSy65BMMwqKio4J577jntVMfp06czderUcy2tWTp+ZUfDMLTctYiIiEgjVzXiKzEm2NQ6REREGrNzHvF1//33s3XrVhYuXHja/VasWMGzzz7L66+/zoYNG/j000/58ssvmTZt2imPmThxInl5ea7LwYMHz7XMZqNjS3883S3klZRz6GiJ2eWIiIiIyHk4WmTjYI7zPZ2mOoqIiJy7cxrx9cADD/DFF1/www8/0KpVq9Pu++STT/J///d/3HXXXQB0796doqIi/vjHP/L444/j5nZi9ma1WrFa1cCzJqwe7nSMCGB7ej7b0vKJDfE1uyQREREROUdV0xzjQn0J8vU0uRoREZHGq0YjvgzD4IEHHmDRokV8++23tG3b9ozHFBcXnxBuubu7u84ntaeqz5dWdhQRERFp3JKqpjm2Cja1DhERkcauRiO+7r//fhYsWMDnn39OQEAAGRkZAAQFBeHj4wPA2LFjiYmJYfr06QCMHDmSF198kV69enHhhReyZ88ennzySUaOHOkKwKR2OPt8HWJbmhYDEBEREWnMNh+qbGzfStMcRUREzkeNgq/Zs2cDcPnll1fbPnfuXMaNGwdASkpKtRFeTzzxBBaLhSeeeILU1FTCw8MZOXIkzzzzzPlVLifQyo4iIiIiTUOSK/gKNrcQERGRRq5GwdfZTE1csWJF9Qfw8GDy5MlMnjy5RoVJzXWJDMRigcyCMjILSokI8Da7JBERERGpocz8UjLyS3GzQHx0oNnliIiINGrnvKqjNDx+Vg/ahfkBaLqjiIiInOC1114jLi4Ob29vLrzwQtasWXNWxy1cuBCLxcLo0aOrbR83bhwWi6XaZdiwYXVQefNSNc2xQ4Q/ftZzWotKREREKin4amKcfb5gu4IvEREROc6HH37I+PHjmTx5Mhs2bKBHjx4MHTqUzMzM0x6XnJzMww8/zMCBA096/7Bhw0hPT3ddPvjgg7oov1lRY3sREZHao+CriVGfLxERETmZF198kbvvvps77riDbt268cYbb+Dr68ucOXNOeYzdbufWW29l6tSptGvX7qT7WK1WIiMjXZcWLVrU1VNoNtTYXkREpPYo+GpiqkZ8aaqjiIiIVLHZbKxfv57Bgwe7trm5uTF48GB++eWXUx731FNPERERwR/+8IdT7rNixQoiIiLo3Lkz9957L0eOHDltLWVlZeTn51e7yDGGYZCUqsb2IiIitUXBVxNT1QA1JaeYvJJyk6sRERGRhiA7Oxu73U7Lli2rbW/ZsiUZGRknPeann37i3Xff5e233z7leYcNG8Z7773H8uXLef755/n+++8ZPnw4drv9lMdMnz6doKAg1yU2NvbcnlQTdehoCTlFNjzcLHSJDDC7HBERkUZP3TKbmGBfL2KCfUjNLWF7Wj7924eaXZKIiIg0MgUFBfzf//0fb7/9NmFhYafc76abbnJd7969O4mJibRv354VK1YwaNCgkx4zceJExo8f77qdn5+v8Os4VaO9ukQF4O3pbnI1IiIijZ+CryYoISaQ1NwStqXlKfgSERERwsLCcHd35/Dhw9W2Hz58mMjIyBP237t3L8nJyYwcOdK1zeFwAODh4cHOnTtp3779Cce1a9eOsLAw9uzZc8rgy2q1YrVaz+fpNGmb1dheRESkVmmqYxOkPl8iIiJyPC8vL3r37s3y5ctd2xwOB8uXL6d///4n7N+lSxeSkpLYtGmT6/L73/+eK664gk2bNp1yhNahQ4c4cuQIUVFRdfZcmrqkqsb2MWpsLyIiUhs04qsJqlrZcVuaVnYUERERp/Hjx3P77bfTp08f+vXrx6xZsygqKuKOO+4AYOzYscTExDB9+nS8vb1JSEiodnxwcDCAa3thYSFTp07l2muvJTIykr179/Loo4/SoUMHhg4dWq/PralwOIxjwZdGfImIiNQKBV9NUNWIrz2ZhRTbKvD10rdZRESkubvxxhvJyspi0qRJZGRk0LNnT77++mtXw/uUlBTc3M5+MoC7uztbtmzhX//6F7m5uURHR3PllVcybdo0TWU8R/uPFFFQVoHVw42OLf3NLkdERKRJUCLSBEUEWGnVwodDR0uY8fVOpvw+3uySREREpAF44IEHeOCBB05634oVK0577Lx586rd9vHxYenSpbVUmcCxaY7x0YF4uqsjiYiISG3Qb9QmyGKxMG2UcxrCvJXJfL315MuUi4iIiEjDocb2IiIitU/BVxN1RZcI/nRpOwAe/XgzB3OKTa5IRERERE7nWH8vNbYXERGpLQq+mrCHh3amV+tg8ksrePCDjZTbHWaXJCIiIiInUWF3sDVNwZeIiEhtU/DVhHm6u/HyTb0I9PZg08Fc/rF0p9kliYiIiMhJ7MkqpLTcgb/Vg3ZhamwvIiJSWxR8NXGxIb7MvL4HAG/+sI9vfz1sckUiIiIi8ltbDjpHeyXEBOLmZjG5GhERkaZDwVczMDQ+knEXxwHwt/9sJj2vxNyCRERERKQaNbYXERGpGwq+momJI7qQEBPI0eJyHvpgExXq9yUiIiLSYCSlqr+XiIhIXVDw1UxYPdx59eYL8Ld6sCY5h5eW7za7JBEREREByirs7EjPByAxJtjcYkRERJoYBV/NSFyYH9Ov6Q7Aq9/t4afd2SZXJCIiIiI7MwootxsE+3oSG+JjdjkiIiJNioKvZmZkj2huubA1hgF/+XAjmQWlZpckIiIi0qxtPuSc5tg9JgiLRY3tRUREapOCr2Zo0tXd6BIZQHahjb8s3ITdYZhdkoiIiEizlVTZ2L6HGtuLiIjUOgVfzZC3pzuv3nIBPp7urNx7hNe/22N2SSIiIiLN1paqEV9qbC8iIlLrFHw1Ux0i/Hl6dAIA//xmF6v2HTG5IhEREZHmp8RmZ9fhAkAjvkREROqCgq9m7Nrerbj2glY4DHho4UaOFJaZXZKIiIhIs7ItLQ+HAREBViKDvM0uR0REpMlR8NXMTRsdT/twPw7nl/G3jzbjUL8vERERkXpT1dg+UdMcRURE6oSCr2bO18uD1269AKuHGyt2ZvH2j/vMLklERESk2ahqbJ+oaY4iIiJ1QsGX0CUykCm/jwdg5tKdrD9w1OSKRERERJoHNbYXERGpWwq+BICb+sYyskc0FQ6DP3+wkdxim9kliYiIiDRp+aXl7MsuAiAxRsGXiIhIXVDwJQBYLBaeHZNAXKgvqbklPPLxFgxD/b5ERERE6srWytFeMcE+hPpbTa5GRESkaVLwJS4B3p68essFeLm7sWz7YeatTDa7JBEREZEma0uqM/jqEavRXiIiInVFwZdUkxATxONXdQXg2a92sKWy4aqIiIiI1K4tamwvIiJS5xR8yQnG9m/D0PiWlNsNHliwkfzScrNLEhEREWlyqhrbq7+XiIhI3VHwJSewWCzMuLYHrVr4kJJTzMRPk9TvS0RERKQWHSks49DREgAStKKjiIhInVHwJScV5OvJKzf3wsPNwpdb0lmwJsXskkRERESajKr+Xu3C/Aj09jS5GhERkaZLwZecUq/WLZgwrAsAU/+7nR3p+SZXJCIiItI0/H97dx4eVXm+cfw7M9n3hJA9QNiXEHaRVQQEqdqi2KqlLrTVnxpUitZKVRRFUauWukGlLq0WsbWi1gWqIKugyJaw7yYEsgGZLCSTZeb3x0kCYZOw5MxM7s91nStnzpwzeYaB8HLzvs/JrFvmqNleIiIiF5WCLzmj3wxOYXjnGCqrnaTPXUeZo9rskkREREQ8Xl1j++5qbC8iInJRKfiSM7JaLTz/8x7EhQWwp6CMRz7apH5fIiIiIueprrF9D834EhERuagUfFU74Nu/wo7/mV2J24oK9uOlm3phs1qYvz6HD9buN7skEREREY+Va68gv8SB1QJdE8LMLkdERMSrKfhaPQu+eBD+9zDUaBnf6VySEsXkKzoCMPXjzezMKzG5IhERERHPVLfMsWNsKEF+PuYWIyIi4uUUfPWdAIFRULgDNr5ndjVu7a7L2jGkQzTlVTWkz11HeWWN2SWJiIiIeJwMNbYXERFpMgq+AsJhyP3G/pIZUFVhbj1uzGq18OIvehId4s+OvFKm/Xez2SWJiIiIeJyMHCP4UmN7ERGRi0/BF0C/30JYIhTnwJo5Zlfj1lqG+vOXG3tiscC8Ndl8vCHH7JJEREREPIbL5apf6qjG9iIiIhefgi8A3wAYNsXYX/4CVNjNrcfNDWofzT3DOwDwxw8z2VtYZnJFIiIiIp4h+3A5RUer8LVZ6BQXanY5IiIiXk/BV50eN0F0Ryg/At+8bHY1bu++ER3onxJFWWUN6f9cR0WV+n2JiIiI/JiMnCIAusSH4e9jM7cYERGRZkDBVx2bDwx/1Nhf9SqU5ptbj5uzWS385cZeRAX7seVgMU9/vtXskkRERETcXl1j++6JWuYoIiLSFBR8Ha/LNZDYB6qOwrI/mV2N24sLD+CFX/QA4B+rfuCLzIMmVyQiIiLi3o7194owtQ4REZHmQsHX8SwWGPm4sf/9W3B4r6nleILLO8Xwf5e1BeDB/2SQffioyRWJiIiIuCen08WmnGIAuquxvYiISJNQ8HWilKHQbjg4q+Drp82uxiM8MKoTvVtFUFJRzcT31lNZ7TS7JBERERG3s6ewjFJHNQG+VjrEhJhdjoiISLOg4OtURkw1vmb+G3Izza3FA/jarLx0Uy/CA33ZmF3Ecwu2mV2SiIiIiNupW+aYmhCOj03DcBERkaagv3FPJaEXdLsWcMGiJ82uxiMkRQbxp+vTAPjbir18tSXP5IpERERE3Et9Y3stcxQREWkyCr5O5/JHwGKDnQvhh2/MrsYjjOoWx4RBbQB44IONHCgqN7cgERERETeixvYiIiJNT8HX6US3h963GPtfTQOXy9x6PMSUMV1ISwqn6GgV9763nuoa9fsSERERqapxsvmAGtuLiIg0NQVfZ3LZH8AnALJXw46FZlfjEfx8rLx8Uy9C/X34/ocjvPjlDrNLEhERETHdzrxSHNVOQv19SGkRbHY5IiIizYaCrzMJi4f+dxr7i6aBs8bcejxE6xbBzBjXHYDXluxm6Y4CkysSERERMVd9Y/vEcKxWi7nFiIiINCMKvn7M4EkQEA75W4y7PMpZuTotgfH9WwEw+f0N5BdXmFyRiIiIiHkycozG9mnJWuYoIiLSlBR8/ZjASBg0ydj/+imodphajid59OqudI4L5VBZJffN20CNU33SREREpHmqm/GVlhhhah0iIiLNjYKvs9H/TgiJg6IsWPu22dV4jABfG6+O702Qn41Vew7xyuJdZpckIiIi0uQqqmrYnlsCQJoa24uIiDSpRgVfM2bMoF+/foSGhhITE8PYsWPZvn37j15XVFREeno68fHx+Pv707FjRz7//PNzLrrJ+QXBsD8Y+0ufA0eJufV4kHYtQ3jq2lQA/rJoB6t2HzK5IhEREZGmtS23hKoaF1HBfiRFBppdjoiISLPSqOBr6dKlpKens3r1ar788kuqqqoYNWoUZWVlp72msrKSK664gn379vHBBx+wfft25syZQ2Ji4nkX36R63QxRbeFoIayeZXY1HuXaXkn8vE8SThfcN289haVaLioiIiLNR90yx+6J4VgsamwvIiLSlHwac/KCBQsaPH777beJiYlh7dq1DB069JTXvPnmmxw+fJhvvvkGX19fANq0aXNu1ZrJ5gvDH4EPfg0rX4K+v4bgaLOr8hjTftaN9dlF7MovZfK/NvL2bf10RyMRERFpFjL2G43te2iZo4iISJM7rx5fdrvxl3hUVNRpz/nkk08YMGAA6enpxMbGkpqaytNPP01NTc1pr3E4HBQXFzfY3ELXayEuDSpLYPmLZlfjUYL8fHj1l70J8LWybEcBf122x+ySRERERJpE/YyvpAhT6xAREWmOzjn4cjqdTJo0iUGDBpGamnra8/bs2cMHH3xATU0Nn3/+OY8++igvvPAC06dPP+01M2bMIDw8vH5LTk4+1zIvLKsVRj5m7K+ZA0XZ5tbjYTrFhTLtp90AeP5/2/l+32GTKxIRERG5uMoc1ezKLwXU2F5ERMQM5xx8paens2nTJubNm3fG85xOJzExMbz++uv06dOHG264gYcffpjZs2ef9popU6Zgt9vrt+xsNwqY2o2ANkOgphKWPGN2NR7nF32T+VnPBGqcLu59bz1HyirNLklERETkotl8oBinC2LD/IkNCzC7HBERkWbnnIKviRMn8umnn/L111+TlJR0xnPj4+Pp2LEjNput/liXLl3Izc2lsvLUoYe/vz9hYWENNrdhscDIx439jXMhf5up5Xgai8XCU9d2p02LIA7YK/j9BxtxuVxmlyUiIiJyUdQtc0zTMkcRERFTNCr4crlcTJw4kfnz57N48WJSUlJ+9JpBgwaxa9cunE5n/bEdO3YQHx+Pn59f4yt2B0l9ofPV4HLC4ifNrsbjhPj78Move+Nns/LV1nzeXLnP7JJERERELoq6xvZpiVrmKCIiYoZGBV/p6em8++67zJ07l9DQUHJzc8nNzaW8vLz+nFtuuYUpU6bUP77rrrs4fPgw9913Hzt27OCzzz7j6aefJj09/cK9CzMMfxQsVtj2KWSvMbsaj5OaGM6jV3cB4JkvtrIxu8jcgkREREQugsyc2uArOcLcQkRERJqpRgVfs2bNwm63M2zYMOLj4+u3999/v/6crKwsDh48WP84OTmZhQsXsmbNGtLS0rj33nu57777eOihhy7cuzBDTGfo8Utj/6vHQcv1Gu1Xl7ZmTGocVTUuJr63Dnt5ldkliYiIiFww9vIq9haWAZrxJSIiYhafxpx8Nr2YlixZctKxAQMGsHr16sZ8K88w7CHI/Df8sAJ2L4L2I82uyKNYLBaeGZdGZo6d7MPlTPkwg1d/2RuLxWJ2aSIiIiLnLbN2mWNyVCCRwR7a4kNERMTDnfNdHQWISIZLbjf2v5oGx/Uxk7MTHujLK7/sjY/VwueZubz7bZbZJYmIiIhcEBk5RYAa24uIiJhJwdf5GjwZ/EIhNwO2zDe7Go/UMzmCh8Z0BuDJT7ew+YDd5IpEREREzl9Gthrbi4iImE3B1/kKbgGD7jX2F0+HGvWpOhe/GZzCiM4xVFY7mTh3PaWOarNLEhERETkv9Y3tNeNLRETENAq+LoRL74agaDi8B9b9w+xqPJLFYuH5n/cgPjyAvYVlPDw/86x6yomIiIi4o8JSBzlF5VgskJoYZnY5IiIizZaCrwvBPwQue9DYX/osVB41tx4PFRnsx0s39cJmtfDxhgP8+/v9ZpckIiIick7qGtu3jQ4mNMDX5GpERESaLwVfF0qf2yCiFZTmwbezza7GY/VrE8XkKzoCMPWTTezIKzG5IhEREZHG27i/CNAyRxEREbMp+LpQfPzh8keM/ZUzofyIqeV4srsua8eQDtFUVDlJ/+c6jlaq35eIiIh4loz9df291NheRETETAq+LqTu10NMN6iww4qZZlfjsaxWC3++oSctQ/3ZmV/K459sNrskERERkbPmcrmOC74izC1GRESkmVPwdSFZbTBiqrH/7WwoPmBuPR4sOsSfv9zYE4sF/vX9fuavV78vERER8QwH7RUUljqwWS10jVdjexERETMp+LrQOo6G5EuhusJodC/nbGC7aO4d3gGAh+dvYk9BqckViYiIiPy4utleHWNDCfSzmVyNiIhI86bg60KzWGDk48b+unegcJep5Xi6e0d04NK2URytrCF97noqqmrMLklERETkjDLqGtsnqr+XiIiI2RR8XQytB0CH0eCqga+nm12NR7NZLfzlxl60CPZj68Fipn+2xeySRERERM4oM6e2v1eygi8RERGzKfi6WEZMBSyweT4cWG92NR4tNiyAF2/oCcC7q7N49etd1Dhd5hYlIiLigV599VXatGlDQEAA/fv357vvvjur6+bNm4fFYmHs2LENjrtcLqZOnUp8fDyBgYGMHDmSnTt3XoTKPUeDxvaJEeYWIyIiIgq+Lpq4VEj7hbG/6Alza/ECl3VsycTL2wPwp4Xb+cVfV7GvsMzkqkRERDzH+++/z+TJk3nsscdYt24dPXr0YPTo0eTn55/xun379vHAAw8wZMiQk5577rnneOmll5g9ezbffvstwcHBjB49moqKiov1Ntxe1uGj2Mur8LNZ6RQXanY5IiIizZ6Cr4tp2BSw+sLuxbBnqdnVeLz7R3XkuevTCPH3Ye0PRxjzl+X8Y9U+nJr9JSIi8qNefPFFbr/9diZMmEDXrl2ZPXs2QUFBvPnmm6e9pqamhvHjxzNt2jTatm3b4DmXy8XMmTN55JFH+NnPfkZaWhr/+Mc/OHDgAB999NFFfjfua2PtbK8uCWH4+WioLSIiYjb9bXwxRaVA3wnG/qJp4FJAcz4sFgu/6JvMgklDGNiuBeVVNUz9eDM3v/ktOUXlZpcnIiLitiorK1m7di0jR46sP2a1Whk5ciSrVq067XVPPPEEMTEx/OY3vznpub1795Kbm9vgNcPDw+nfv/8ZX9PhcFBcXNxg8yYZ2UWAGtuLiIi4CwVfF9vQ34NvMOSsha3/Nbsar5AUGcS7v+nPEz/rRqCvjZW7DjH6z8v41/fZuBQuioiInKSwsJCamhpiY2MbHI+NjSU3N/eU16xYsYI33niDOXPmnPL5uusa85oAM2bMIDw8vH5LTk5uzFtxexl1je2TFHyJiIi4AwVfF1tIDAy429hf/CTUVJtbj5ewWi3cMqANn983hD6tIyl1VPPgBxn89u/fk1/cfPuKiIiIXAglJSXcfPPNzJkzh+jo6Av62lOmTMFut9dv2dnZF/T1zVTjdLGpPviKMLcYERERARR8NY2B90BgFBTugI3vmV2NV0mJDuZf/zeAKWM642ezsmhbPqNmLuO/Gw+YXZqIiIjbiI6OxmazkZeX1+B4Xl4ecXFxJ52/e/du9u3bxzXXXIOPjw8+Pj784x//4JNPPsHHx4fdu3fXX3e2r1nH39+fsLCwBpu32FNQytHKGgJ9bbSPCTG7HBEREUHBV9MICIch9xv7S2ZAlWYkXUg2q4X/u6wdn947mNTEMIqOVnHPe+tJ/+c6DpdVml2eiIiI6fz8/OjTpw+LFi2qP+Z0Olm0aBEDBgw46fzOnTuTmZnJhg0b6ref/vSnXH755WzYsIHk5GRSUlKIi4tr8JrFxcV8++23p3zN5qCusX1qYhg2q8XkakRERAQUfDWdfr+FsEQozoE1fzO7Gq/UMTaU+XcPYtLIDvhYLXyWeZBRf17Kl1vyfvxiERERLzd58mTmzJnD3//+d7Zu3cpdd91FWVkZEyYYN+K55ZZbmDJlCgABAQGkpqY22CIiIggNDSU1NRU/Pz8sFguTJk1i+vTpfPLJJ2RmZnLLLbeQkJDA2LFjTXyn5sncXwRomaOIiIg78TG7gGbDNwCGTYFPJsLy56H3zcZMMLmgfG1WJo3syIjOsdz/7w3syCvl9n98z7jeSUy9pivhgb5mlygiImKKG264gYKCAqZOnUpubi49e/ZkwYIF9c3ps7KysFob93+iDz74IGVlZdxxxx0UFRUxePBgFixYQEBAwMV4C26vbsaXGtuLiIi4D4vLA26DV1xcTHh4OHa73bP7QNRUw6wBRq+vob+H4Y+YXZFXq6iq4c9f7eD1ZXtwuSA+PIDnrk9jSIeWZpcmIiJNwGvGD17OWz6nqhon3R5bSGW1k68fGEZKdLDZJYmIiHitxowftNSxKdl8YPijxv6qV6E039x6vFyAr40pY7rwwZ0DaNMiiIP2Cm5+4zse+SiTMofurikiIiIXzvbcEiqrnYQG+NCmRZDZ5YiIiEgtBV9Nrcs1kNgHqo7Csj+ZXU2z0Kd1FJ/fN4RbB7QG4N3VWYz5y3K+23vY5MpERETEW2Qct8zRYlFjexEREXeh4KupWSww8nFj//u34PBeU8tpLoL8fJj2s1T++dv+JEYEknX4KDe8voqnPttCRVWN2eWJiIiIh8vMKQLU2F5ERMTdKPgyQ8pQaDccnFWwZIbZ1TQrg9pHs2DSEG7om4zLBXOW7+Xql1ewMbvI7NJERETEg23Mrp3xlajG9iIiIu5EwZdZRkw1vmb8C3I3mVtLMxMa4Muz16fx5m19aRnqz678Uq6b9Q0v/m87ldVOs8sTERERD1NRVcOOvBIA0pIjzC1GREREGlDwZZaEXtDtWsAFi54wu5pmaXjnWP43aSg/7ZFAjdPFS4t3MfbVlWzLLTa7NBEREfEgWw4WU+100SLYj4TwALPLERERkeMo+DLT8EfBYoOdC+GHb8yuplmKDPbjpZt68eovexMZ5MuWg8Vc8/IKXv16F9U1mv0lIiIiPy5Tje1FRETcloIvM7VoB71vMfa/mgYul7n1NGNXpcXzv99dxsgusVTVuPjTwu1cP3sVuwtKzS5NRERE3NzG/UUAdFdjexEREbej4Mtsl/0BfAIgezXsWGh2Nc1ay1B/5tzShxd+3oPQAB82ZBdx1UvLeXPFXpxOhZIiIiJyanUzvnokqbG9iIiIu1HwZbaweOh/p7G/aBo4a8ytp5mzWCyM65PEwklDGdIhmooqJ098uoVf/m012YePml2eiIiIuJlSRzW7ameId1fwJSIi4nYUfLmDwZMgIBzyt0Dmv82uRoCEiED+8etLmD42lSA/G6v3HObKmct477ssXFqSKiIiIrU25dhxuSA+PICYUDW2FxERcTcKvtxBYCQMmmTsf/0UVDtMLUcMFouFX13ami/uG8IlbaIoq6xhyoeZ3PbWGnLtFWaXJyIiIm7g+Mb2IiIi4n4UfLmL/ndCSBwUZcHat82uRo7TukUw791xKY9c1QU/HytLdxQw6s9L+Wh9jmZ/iYiINHN1je3T1NheRETELSn4chd+QTDsD8b+0ufAUWJuPdKAzWrht0Pa8vm9g0lLCqe4oppJ72/grnfXUViqGXoiIiLNVWaOZnyJiIi4MwVf7qTXzRDVFo4WwupZZlcjp9A+JpQP7xrI/Vd0xMdqYcHmXEb/eRkLNh00uzQRERFpYkVHK/nhkHHzm+6JCr5ERETckYIvd2LzheGPGPsrX4KyQnPrkVPysVm5Z0QHPp44iM5xoRwqq+TOd9fxu/c3YD9aZXZ5IiIi0kTqZnu1bhFERJCfydWIiIjIqSj4cjddr4W4NKgsgeUvml2NnEG3hHA+njiIu4e1w2qB+etzGDVzKUu255tdmoiIiDSBjNrG9prtJSIi4r4UfLkbqxVGPmbsr5kDRdnm1iNn5O9j48ErO/PBXQNpGx1MXrGD295aw5QPMyl1VJtdnoiIiFxEG7OLAOihxvYiIiJuS8GXO2o3AtoMgZpKWPKM2dXIWejdKpLP7h3ChEFtAHjvuyyunLmMVbsPmVuYiIiIXDRqbC8iIuL+FHy5I4sFRj5u7G+cC/nbTC1Hzk6gn43HrunGe7dfSlJkIPuPlHPTnNVM++9myitrzC5PRERELqD8kgoO2iuwWKCbljqKiIi4LQVf7iqpL3S+GlxOWPyk2dVIIwxo14IFk4Zy0yXJALy1ch9XvbScdVlHTK5MRERELpTM2v5e7VuGEOLvY3I1IiIicjoKvtzZ8EfBYoVtn0L2GrOrkUYI8fdhxnVpvDWhH7Fh/uwpLOP6Wd/w3IJtOKo1+0tERMTTbaxrbK9ljiIiIm5NwZc7i+kMPX5p7H/1OLhcppYjjXd5pxj+N+kyru2ViNMFry3Zzc9eWcnmA3azSxMREZHzkLm/CFBjexEREXen4MvdDXsIbP7wwwrYvcjsauQchAf58ucbejL7V71pEezHttwSfvbKSl5etJPqGqfZ5YmIiEgjuVwuMjTjS0RExCMo+HJ3Eclwye3G/lfTwKmgxFNdmRrPwt8NZXS3WKqdLl74cgfjZn3DrvwSs0sTERGRRjhgr+BQWSU+Vgtd48PMLkdERETOQMGXJxg8GfxCITcDtsw3uxo5D9Eh/sz+VR9m3tCTsAAfNu6385OXVvDyop2UOqrNLk9ERETOQkZ2EQAdY0MJ8LWZW4yIiIickYIvTxDcAgbda+wvng41VebWI+fFYrEwtlci//vdZVzWsSWV1U5e+HIHg55ZzF++2om9XJ+viIiIO6trbN8jWcscRURE3J2CL09x6d0QFA2H98C6f5hdjVwAceEBvD2hHzNv6Enb6GDs5VX8+asdDH5mMc8v3M6RskqzSxQREZFTyMwpAiBNje1FRETcnoIvT+EfApc9aOwvfRYqj5pbj1wQdbO/vpx8GS/d1IuOsSGUOKp55etdDHp2MTM+30pBicPsMkVERKSW03lcY/tEzfgSERFxdwq+PEmf2yCiFZTmwbezza5GLiCb1cJPeySw4L6hzP5Vb7rGh3G0soa/LtvDkOcWM+2/m8m1V5hdpoiISLP3w+GjlFRU4+djpVNcqNnliIiIyI9Q8OVJfPzh8keM/ZUzofyIqeXIhWe1WrgyNZ7P7h3MG7f2pUdyBBVVTt5auY+hz33NIx9lsv+IZvuJiIiYJWN/EQBd48PwtWkoLSIi4u70t7Wn6X49xHSDCjusmGl2NXKRWCwWRnSJ5aO7B/KPX19CvzaRVNY4eXd1FsP+tIQ/fJDBD4fKzC5TRESk2alb5tgjScscRUREPIGCL09jtcGIqcb+t7Oh+IC59chFZbFYGNqxJf/6vwG8d/ulDGzXgmqni/e/z2b4C0uZ/P4GduWXml2miIhIs1E346u7GtuLiIh4BAVfnqjjaEi+FKorjEb34vUsFgsD2rVg7u2X8p+7BnBZx5bUOF18uD6HK/68lIlz17E9t8TsMkVERLxajdPFppxiQDO+REREPIWCL09kscDIx439de9A4S5Ty5Gm1ad1FH//9SV8nD6IkV1icbng04yDjJ65jP9753s25djNLlFERMQr7covpbyqhiA/G21bhphdjoiIiJyFRgVfM2bMoF+/foSGhhITE8PYsWPZvn37WV8/b948LBYLY8eObWydcqLWA6DDaHDVwNfTza5GTNAjOYK/3dqXz+8dwk+6x2GxwMLNeVz98gp+/fYa1mfp5gciIiIX0sbaZY6pieHYrBZzixEREZGz0qjga+nSpaSnp7N69Wq+/PJLqqqqGDVqFGVlP95ke9++fTzwwAMMGTLknIuVE4yYClhg83w4sN7sasQkXRPCeG18H/43aSg/65mA1QKLt+Vz7WvfcPMb3/Ld3sNmlygiIuIVMtXYXkRExOM0KvhasGABt912G926daNHjx68/fbbZGVlsXbt2jNeV1NTw/jx45k2bRpt27Y9r4LlOHGpkPYLY3/RE+bWIqbrEBvKX27sxVeTL+P6PknYrBaW7yzkF39dxQ1/XcXKXYW4XC6zyxQREfFYamwvIiLiec6rx5fdbvyvV1RU1BnPe+KJJ4iJieE3v/nNWb2uw+GguLi4wSanMWwKWH1h92LYs9TsasQNtG0ZwvM/78GSB4Zx0yWt8LVZ+HbvYcb/7VvGzfqGr7fnKwATERFppMpqJ1sPGjeS0YwvERERz3HOwZfT6WTSpEkMGjSI1NTU0563YsUK3njjDebMmXPWrz1jxgzCw8Prt+Tk5HMt0/tFpUDfCcb+ommgQENqJUcFMeO67iz9/eXcOqA1fj5W1mUVMeGtNfzs1ZX8b3OuAjAREZGztD23hMoaJ+GBvrSKCjK7HBERETlL5xx8paens2nTJubNm3fac0pKSrj55puZM2cO0dHRZ/3aU6ZMwW6312/Z2dnnWmbzMPT34BsMOWth63/NrkbcTEJEINN+lsqKBy/nt4NTCPS1kbHfzh3vrGXMX5bzWcZBnE4FYCIiImeSkVMEQFpSOBaLGtuLiIh4Cp9zuWjixIl8+umnLFu2jKSkpNOet3v3bvbt28c111xTf8zpdBrf2MeH7du3065du5Ou8/f3x9/f/1xKa55CYmBAOix7DhY/CZ1+ArZz+mjFi8WEBfDI1V25a1g7/rZiL//4Zh/bcktIn7uO9jEhTLy8PVenxeNjO68V0CIiIl4pI9to8dE9UcscRUREPEmj/oXrcrmYOHEi8+fPZ/HixaSkpJzx/M6dO5OZmcmGDRvqt5/+9KdcfvnlbNiwQUsYL6SBEyEwCgp3wMb3zK5G3FiLEH/+cGVnVj40nHtHdCA0wIdd+aVMen8DI19cyr++z6aqxml2mSIiIm5lY21j+zQ1thcREfEojQq+0tPTeffdd5k7dy6hoaHk5uaSm5tLeXl5/Tm33HILU6ZMASAgIIDU1NQGW0REBKGhoaSmpuLn53dh301zFhAOQ+439pfMgKoKc+sRtxcR5MfkKzqy8qHhPDCqIxFBvuw7dJQHP8jg8ueX8M9vf8BRXWN2mSIiIqYrr6xhZ34pAD2SNeNLRETEkzQq+Jo1axZ2u51hw4YRHx9fv73//vv152RlZXHw4MELXqichX6/hbBEKM6BNX8zuxrxEGEBvkwc3oGVfxjOlDGdiQ7xY/+Rch6ev4lhf1rC37/ZR0WVAjAREWm+thy0U+N0ER3iT1xYgNnliIiISCNYXB5wW7fi4mLCw8Ox2+2EhYWZXY57W/cOfDIRAiPhvo3GTDCRRiivrOG977L467Ld5BU7AGgZ6s8dQ9oy/tJWBPmpf5yIeAaNHzyDJ3xOb63cy7T/bmFE5xjeuK2f2eWIiIg0e40ZP6iLtbfpcRNEd4TyI/DNy2ZXIx4o0M/GrwensPT3l/Pk2FQSIwIpKHHw1OdbGfzs17z69S5KKqrMLlNERKTJZOyvbWyfpP9QFBER8TQKvryNzQdGTDX2V70K278wtx7xWAG+Nm6+tDVfPzCMZ8d1p1VUEIfLKvnTwu0MfvZrZn61A/tRBWAiIuL9Mmob2/dQY3sRERGPo+DLG3W+GtpeDlVH4b0bYcEUqK40uyrxUH4+Vm7o14rF91/Gi7/oQduWwdjLq5j51U4GP7uY5xdu53CZfn+JiIh3KqmoYk9hGaAZXyIiIp5IwZc3sljgl/+CARONx6tfgzdHweE95tYlHs3HZuW63kl8+bvLePmmXnSKDaXEUc0rX+9i8LOLmfH5VgpKHGaXKSIickFtyinG5YLEiECiQ/zNLkdEREQaScGXt/Lxg9FPwU3zjEb3B9bDXy+DTR+aXZl4OJvVwjU9EvjiviHM/lUfuiWEcbSyhr8u28PgZxfz+CebybVXmF2miIjIBVG3zLF7omZ7iYiIeCIFX96u0xi4cwW0GgCOYvhgAvx3ElSVm12ZeDir1cKVqXF8es9g3rytLz2TI3BUO3n7m30Mfe5rHvpPBl9vy+doZbXZpYqIiJyzusb2ackKvkRERDyRj9kFSBMIT4JbP4UlM2D5C7D2Lcj+Dn7+NrTsaHZ14uEsFgvDO8dyeacYVuwq5OVFu/hu32Hmrclm3pps/GxW+qVEMqRDS4Z2aEmX+FAsFovZZYuIiJyVjJwiQI3tRUREPJWCr+bC5gMjHoU2g+DDOyB/M7x+GVz1AvT8pdnViRewWCwM6dCSIR1a8u2eQ3y04QDLdhSQU1TOyl2HWLnrEM98sY3oEH+GdohmaMeWDO4QrX4pIiLitg6XVZJ92Jgln6qljiIiIh5JwVdz02443LkSPrwd9i6Fj+6CPUuNAMw/xOzqxEv0b9uC/m1b4HK52FNYxvIdBSzbWciq3YcoLHXw4focPlyfA0C3hDCGdjRmg/VpHYmfj1Zgi4iIe8jMMZY5pkQHEx7oa3I1IiIici4UfDVHobFw83xY8SJ8/TRkzIOctfDztyCuu9nViRexWCy0axlCu5Yh3DYoBUd1DWt/OMKyHYUs31nA5gPF9dusJbsJ8rMxoG0LhnZsyZAO0aREB2tZpIiImCYjuwhQY3sRERFPpuCrubLaYOjvofUg+OA3cGgnzBkBV86Avr8GhQ1yEfj72BjYLpqB7aJ5aExnCkocrNhVUB+EFZZWsmhbPou25QOQFBlYOxssmoHtowkL0P+2i4hI08monfGVlqTgS0RExFMp+GruWg807vr48d2wYwF8NtlYAnnNSxAYYXZ14uVahvpzba8kru2VhNPpYmtucX0I9v2+I+w/Us7cb7OY+20WNquFXskRRpP8jtGkJUVgsyqgFRGRiydjfxEAaWpsLyIi4rEsLpfLZXYRP6a4uJjw8HDsdjthYWFml+OdXC5Y/Rp8+Rg4qyCitbH0MbGP2ZVJM3W0sprVew6xbEchy3YWsKegrMHz4YG+DG4fzdCO0Qzp0JKEiECTKhURd6Xxg2dw188pr7iC/k8vwmqBzMdHE+yv/y8WERFxF40ZP+hvcDFYLDAgHZIvhQ8mQNEP8MZoGPm4cVxLH6WJBfn5MLxzLMM7xwKw/8hRlu8sZNmOAlbsKsReXsVnmQf5LPMgAO1jQhhaOxusf0oLAv1sZpYvIiIeLmO/scyxfUyIQi8REREPpr/FpaGkPnDncvjkHtjyMfzvYdi3HMbOgqAos6uTZiwpMoibLmnFTZe0orrGycb9dpbtKGDZzgI2ZhexK7+UXfmlvLlyL34+Vi5pE1U/G6xzXKia5IuISKNomaOIiIh3UPAlJwsIh5//Hb5/ExZMMXp/zR4M4/5m9AQTMZmPzUqf1pH0aR3J767oiP1oFSt3G7PBlu0o4IC9ghW7ClmxqxDYRkyof31vsMHto2kR4m/2WxARETdXN+Orhxrbi4iIeDQFX3JqFgv0+w0kXwL/nmDc9fHtq2DYH2HIZOOukCJuIjzIl590j+cn3eNxuVzsLiirnw22es8h8ksc/Gfdfv6zbj8WC6QmhDOkQzRDO7akd6tI/HysZr8FERFxIy6Xq37GV3fN+BIREfFoam4vP85RCp8/ABvfMx6nXAbXzYHQWHPrEjkLjuoavt93pDYIK2TrweIGzwf72RjQrgVDO7ZkaIeWtG4RpGWRIl5C4wfP4I6fU/bhowx57mt8bRY2TRuNv4/+w09ERMSdqLm9XFj+IXDtbEgZCp/dD3uXwuxBcN3r0G642dWJnJG/j41B7aMZ1D6aKUB+cQXLdxayfGcBy3cWcqiskq+25vPV1nwAkqMCa5vkt2RAuxaEBfia+wZERKTJ1S1z7BQXqtBLRETEwyn4krPX85eQ2MdY+pi/Gd65zlj2OOyPYNNvJfEMMWEBjOuTxLg+STidLrYcLGbZTqM32NofjpB9uJx/fpvFP7/Nwma10LtVBEM7tGRIx5Z0TwzHZtVsMBERb5eRUwSosb2IiIg3UFohjdOyE9y+CBb+0Wh+v/wF+OEbo/F9eJLZ1Yk0itVqITUxnNTEcO4e1p4yRzWr9xyqXxa5t7CMNfuOsGbfEV74cgcRQb70bhVJWlI4PZIiSEsKV6N8EREvlJFtzPhKS1RjexEREU+n4EsazzcQrv4ztBkC/70PslYZd30cOws6jTG7OpFzFuzvw4gusYzoYvSvyz58tH422De7DlF0tIrF2/JZvC2//prEiEB6JIfTPTGCHknhpCaFa3mkiIgHczpdbMqpDb4040tERMTjKfiSc5d6HST0gg8mwIH18N6NcGk6jHwcfPzMrk7kvCVHBTG+f2vG929NdY2TjBw7G7OLyNhvJ2N/EbsLysgpKienqJzPM3Prr2vbMrh+RlhaUgTdEsII8FWPGBERT7D3UBkljmr8fax0iA0xuxwRERE5Twq+5PxEpcCv/wdfPQ6rXzW2rG/g+reM50S8hI/NSu9WkfRuFVl/rKSiiswce30QtjHbTk5ROXsKythTUMb89TkA2KwWOsaG0qM2CEtLCqdTXCi+NqtZb0dERE4jY38RAN0SwvRzWkRExAso+JLz5+MHVz4NKUPgo7uM2V9/HQrX/MWYFSbipUIDfBnYLpqB7aLrjxWWOsjcb2fj/qLar3YKSx1sPVjM1oPFzFuTDYC/j5WuCWENZoa1jQ7Gqub5IiKmqrujo5Y5ioiIeAcFX3LhdBoDd66AD34D2auNJZB7l8GVM4y+YCLNQHSIP5d3juHyzjEAuFwuDtorjBlhtTPDMvbbKamoZn1WEeuziuqvDfH3ITWxLgwzArGkyEAsFoVhInJhvPrqq/zpT38iNzeXHj168PLLL3PJJZec8twPP/yQp59+ml27dlFVVUWHDh24//77ufnmm+vPue222/j73//e4LrRo0ezYMGCi/o+LqZjwZca24uIiHgDBV9yYYUnwW2fwZKnYfmLsPYtyP4Ofv42tOxodnUiTc5isZAQEUhCRCBXpsYDRuPkfYfKyDhuZtimA3ZKHdWs3nOY1XsO118fFexXPyOsbqlky1DdSVJEGu/9999n8uTJzJ49m/79+zNz5kxGjx7N9u3biYmJOen8qKgoHn74YTp37oyfnx+ffvopEyZMICYmhtGjR9efd+WVV/LWW2/VP/b399yfUdU1TjYf0IwvERERb2JxuVwus4v4McXFxYSHh2O32wkLCzO7HDlbuxfDh3dAWQH4BsFVL0DPX5pdlYhbqq5xsjO/tMHMsG0HS6h2nvwjOj484LgwLILuSeGEB+pOkiIn0vihof79+9OvXz9eeeUVAJxOJ8nJydxzzz089NBDZ/UavXv35qqrruLJJ58EjBlfRUVFfPTRR+dclzt9TlsPFjPmL8sJ8fch47FRWn4uIiLiphozftCML7l42g2HO1fCh7fD3qVG/6+9y+Anz4O/7pIkcjwfm5Uu8WF0iQ/jhn7GsYqqGrYeLK5tnm+EYbsKSjlor+CgvYKFm/Pqr0+JDm4wM6xbQjiBfrqTpIgYKisrWbt2LVOmTKk/ZrVaGTlyJKtWrfrR610uF4sXL2b79u08++yzDZ5bsmQJMTExREZGMnz4cKZPn06LFi1O+1oOhwOHw1H/uLi4+Bze0cWRWbvMMTUxTKGXiIiIl1DwJRdXaCzcPB9WvAhfPw0b34P93xtLH+NSza5OxK0F+Nro1SqSXsfdSbLUUc2mHHuDmWHZh8vZW1jG3sIyPt5wAACrBTrGhjaYGdYpLhQ/H92hTKQ5KiwspKamhtjY2AbHY2Nj2bZt22mvs9vtJCYm4nA4sNlsvPbaa1xxxRX1z1955ZVcd911pKSksHv3bv74xz8yZswYVq1ahc126vB9xowZTJs27cK8sQtsY+0dHbXMUURExHso+JKLz2qDob+HVgPhP7+FQzthznCj6X3fX4Mad4uctRB/Hy5t24JL2x6bTXG4rLK+aX7dzLD8EgfbckvYllvCv77fD4CfjzGrrMdxM8PatgzBplkNInIaoaGhbNiwgdLSUhYtWsTkyZNp27Ytw4YNA+DGG2+sP7d79+6kpaXRrl07lixZwogRI075mlOmTGHy5Mn1j4uLi0lOTr6o7+NsqbG9iIiI91HwJU2nzSDjro8f3QU7F8Jnk42ljz99CQI0wBQ5V1HBfgzrFMOwTseaU+faK9i4v6hBIGYvr2JjdhEbs4uAHwAI9rPRPjaUpIhAEiIC6hvxJ9ZuEUG+uqukiBeIjo7GZrORl5fX4HheXh5xcXGnvc5qtdK+fXsAevbsydatW5kxY0Z98HWitm3bEh0dza5du04bfPn7+7tlA3xHdQ3bco1ll2mJEeYWIyIiIheMgi9pWsEt4Jfvw6pX4avHYMtHcGA9/PwtSOxjdnUiXiMuPIC48DhGdzP+Qetyufjh0NH6u0hm7LeTmWOnrLLmuDDsZIG+tvpArC4MOz4ciwsP0PJJEQ/g5+dHnz59WLRoEWPHjgWM5vaLFi1i4sSJZ/06TqezQX+uE+3fv59Dhw4RHx9/viU3uW0HS6iqcREZ5EtyVKDZ5YiIiMgFouBLmp7FAgMnQqsB8MEEKPoB3hgNV0yDS+/W0keRi8BisdAmOpg20cH8rGciADVOF7vyS9lbWMaBonIOFJWTU/+1gsJSB+VVNewuKGN3QdlpXhdahvgbQVhkbTAW3nDmmGaNibiHyZMnc+utt9K3b18uueQSZs6cSVlZGRMmTADglltuITExkRkzZgBGL66+ffvSrl07HA4Hn3/+Oe+88w6zZs0CoLS0lGnTpjFu3Dji4uLYvXs3Dz74IO3bt2f06NGmvc9zlZFjLHPsnhShn1kiIiJeRMGXmCepD/zfMvjvvbDlY1j4R2Pp49hZEBRldnUiXs9mtdApLpROcaGnfL6iqoZce0V9IFYXih0oOnbMUe0kv8RBfomDDaeZNRbkZzsuCAsgIfzYrLGkyEBiwzRrTKQp3HDDDRQUFDB16lRyc3Pp2bMnCxYsqG94n5WVhdV67M9iWVkZd999N/v37ycwMJDOnTvz7rvvcsMNNwBgs9nIyMjg73//O0VFRSQkJDBq1CiefPJJt1zK+GMyan+GpSWq/YKIiIg3sbhcLpfZRfyY4uJiwsPDsdvthIWFmV2OXGguF3z/Biz4I9Q4ICwRxr0BrQeYXZmInIHL5eJwWWWDWWIH6sOxY7PGfozFAjGh/if1F0uo7TuWGBFIeKBmjUnjafzgGdzlc7py5jK25Zbw+s19GNXt9H3PRERExHyNGT9oxpeYz2KBfr+F5P7w79vg0C54+yq4fAoMnmzcFVJE3I7FYqFFiD8tQvxJS4o45TkVVTUcPG7W2IHjZo3VzSKrrHaSV+wgr9jB+qyiU75Ow1ljtTPHjnusWWMicj6OVlazI68E4LQ/z0RERMQzKfgS9xHXHe5YCp/dDxnzYPF02LcCrn0dQmPNrk5EzkGAr42U6GBSooNP+bzL5eJQWWWDWWIHisrJOVLOAbtxrLC0kqOVNezKL2VXfukpX6du1lhi/RLKINKSwunVKoL4cDWpFpEz23KgGKfL+DkSFx5gdjkiIiJyASn4EvfiHwLX/RXaXmYEYHuWwOzBcN3r0O5ys6sTkQvMYrEQHeJPdCNmjeUcqZ01Zj82c+z4WWPrTpg1FhvmT6/kSHq2iqBXcgTdk8IJ8tNffyJyzMb9RmP7tCT19xIREfE2GvmLe+r5S0jsYyx9zN8C71wLQ+6HYVPApt+2Is1JY2aN5RwxwrE9hWVszC5iW24JecUOFmzOZcHmXKC2qX9saH0Q1qtVJG2jg7Fa1UNMpLnK2F8EaJmjiIiIN1KCIO6rZSe4fTEsmAJr34Llz8MPK2H4o9BqAFjVz0dEzjxr7GhlNZn77WzILmJ9VhHrs4+QV+xgy8FithwsZu63WQCEBvjQM/lYENYzOYLIYD8T3o2ImCGzdsZXd834EhER8ToKvsS9+QbCNTMhZQh8ch9krYK3f2Lc+bHbtdD9eojvaTT4ERE5QZCfD/3btqB/2xb1xw7ay9mQVcT67CI2ZBWRkVNESUU1y3cWsnxnYf15bVoEGWFYbRDWJT5MDfRFvJC9vIo9hWUA9NCMLxEREa+j4Es8Q+o4SOgFy16Arf+F4hxY9YqxRbUzArDUccYsMRGRM4gPDyS+eyBjuscDUFXjZHtuCeuzi1ifdYQN2UXsKShj36Gj7Dt0lI82HADAz8dKakIYvVpF0qtVBD2TI0iMCMSi4F3Eo23OMWZ7JUUGEqWZniIiIl7H4nK5XGYX8WOKi4sJDw/HbrcTFhZmdjlitqoK2PUVbPoAti+A6vJjz8V2h+7jjBAsopV5NYqIR7MfrWLD/mNB2PqsIuzlVSed1zLUv3ZWmBGEpSVFEOKv/1NyFxo/eAazP6dZS3bz7IJt/KR7HK+N79Pk319EREQarzHjB43OxfP4BkCXq43NUQrbvzBCsF1fQV6msX31OCT3h9TrodtYCIkxu2oR8SDhQb5c1rEll3VsCRgN9PcdOsr6rCOszypiQ3YRWw8WU1Di4MsteXy5JQ8AqwU6xobWB2G9WkXSvmWIGueLuLHMnCJAje1FRES8lWZ8ifc4ehi2fgKZH8C+FUDtb22LFVKGGiFYl2sgMMLMKkXES1RU1bApx14fhK3POsIBe8VJ54X6+5CWHE6vZKNXWM9WEUSH+JtQcfOj8YNnMPtzGvTMYnKKypn72/4MbB/d5N9fREREGq8x4wcFX+Kdig/C5vmw6T+Q8/2x4zY/aH+FsRyy4xjwCzKvRhHxOnnFFQ2CsIz9dsqrak46LzkqsD4I69Uqgq4JYfj72Eyo2Ltp/OAZzPycDpU66DP9KwAyHh9FWIBvk35/EREROTda6igSFg8D7ja2w3uNAGzTfyB/C2z/zNh8g6HTGKMxfrsR4KOGtiJyfmLDArgyNY4rU+MAqK5xsiOvlPXZR+rvJLkrv5Tsw+VkHy7nk421jfNtVromhNUvkezdKpKkSDXOF7nYMmob27eNDlboJSIi4qUUfIn3i0qBoQ8YW94Wox/Ypv/AkX21+x9AQAR0/amxHLLNYLBq5oWInD+f2kCra0IY4/u3BsBeXkXG/qL6IGx91hGOHK1iQ7YxU6xOi2C/Br3C0pLCCdU/zEUuqIxsI/hKSwo3uRIRERG5WBR8SfMS2xVip8LwRyFnXW3w9SGU5sK6fxhbSCx0u9YIwZL6gmZciMgFFB7oy5AOLRnS4Vjj/KzDRxsskdxysJhDZZV8tTWfr7bmA8aPorAAX0L8fQgN8CHE34eQ2q/1j/19a4/ZjttveH6wnw82NdsXAY41tu+uxvYiIiJeS8GXNE8WCyT1MbZR0+GHlUZT/C0fQ2kefDvb2CJaQ+o4YzlkbDezqxYRL2SxWGjdIpjWLYIZ2ysRMBrnbz5QXB+EbcguYv+RcuzlVdjLq877ewb72epDsZAAX0L9GwZppw7VjGOhxwVqfj7W865FxCwul4uN+40ZXz0040tERMRrKfgSsdqMuz6mDIWfPA+7FxtLIbd9BkU/wIoXja1ll9oQbBxEtTW7ahHxYgG+Nvq0jqRP60ggBYDDZZUcLnNQUlFNqaOa0opqSmq/ljqqKXM0fFz/vKPK2K+optpp3M+mrLKGssoa8nCcV51+PlYjNKsNwoL9fRo8NoKyYwHbibPP6s4N9LWpn5k0ubxiBwUlDqwW6Jag4EtERMRbKfgSOZ6PH3S60tgqj8KOBUYItvN/ULAVvp5ubAm9jVlg3a4zGumLiFxkUcF+RAWf+004XC4XjmpnfShW6qiuD9EahmZVJ4VqJwZtdXeqrKx2cqi6kkNllef13j6/dwhdE3TXRWlaG/cXAdAxNpRAP/X2FBER8VYKvkROxy8IUq8ztvIi2PapEYLtWQIH1hnbwoeNZvip10HXsRAUZXLRIiKnZrFYCPC1EeBrIzrE/7xeq7rGSZmjhtLKY2HZ8TPRjg/V6h87qimtqKoN2mooqd13uiA0QMMRaXqZ+9XYXkREpDnQSFPkbARGQK9fGVtpAWz5yOgJlr0a9i03ts9/D+2GG03xO/8E/EPNrlpE5KLwsVkJD7ISHnR+d5l0uVyUV9UQ4KPZNtL06mZ8qbG9iIiId1PwJdJYIS3hktuNrSgLNs83QrDcDGNJ5M7/gU8gdBxtLIdsfwX4BphdtYiI27FYLAT5aSgiTc/lcpGZo8b2IiIizYFGmyLnI6IVDLrP2Ap2GEshN30Ah3YZs8K2fAT+YdD5aqMpfsowsOmPnYiIiJmyD5dTdLQKX5uFTnGaoS0iIuLN9C9wkQulZUe4fAoMewgObjQCsE0fQnEObJxrbEHR0G2ssRwyuT9YrWZXLSIi0uzULXPsEh+Gv5baioiIeLVG/at7xowZ9OvXj9DQUGJiYhg7dizbt28/4zVz5sxhyJAhREZGEhkZyciRI/nuu+/Oq2gRt2axQEJPGDUdJm2CCQug32+N0OtoIaz5G7x1JczsDv97BA5sAJfL7KpFRESajbpljmpsLyIi4v0aNeNr6dKlpKen069fP6qrq/njH//IqFGj2LJlC8HBwae8ZsmSJdx0000MHDiQgIAAnn32WUaNGsXmzZtJTEy8IG9CxG1ZrdB6gLFd+SzsXQKZ/zHuEFm8H7552dgCIiAqBSJTGn6NagshcZoZJiIicgFtzC4CIC0xwtQ6RERE5OKzuFznPtWkoKCAmJgYli5dytChQ8/qmpqaGiIjI3nllVe45ZZbzuqa4uJiwsPDsdvthIWFnWu5Iu6jqgJ2fWk0xd+xAKorTn+uTwBEtjk5FItMMXqM+fg1WdkiIp5E4wfP0NSfk9PpovvjCymrrGHBpCF0jtPvDREREU/TmPHDefX4stuNaeJRUVFnfc3Ro0epqqo64zUOhwOHw1H/uLi4+NyLFHFHvgHQ5RpjqyqHw3vhyF44vOe4/b3GXSOrK6Bgm7GdyGKF8KRTh2JRKeCvhr0iIiLH21NYSlllDQG+Vtq3DDG7HBEREbnIzjn4cjqdTJo0iUGDBpGamnrW1/3hD38gISGBkSNHnvacGTNmMG3atHMtTcSz+AZCbFdjO1FNNdizjUCsLgw7su9YOFZ11AjHirJg79KTrw9uefpQLLil0Y9MRESkGcnYb/zHbWpCOD42tRIQERHxduccfKWnp7Np0yZWrFhx1tc888wzzJs3jyVLlhAQEHDa86ZMmcLkyZPrHxcXF5OcnHyupYp4LptPba+vlJOfc7mgNP+EUOy4r0cPQVmBse0/xQ0l/EKMJZSnCsXCkozvLSIi4mXqgq/uamwvIiLSLJzTv2wnTpzIp59+yrJly0hKSjqra55//nmeeeYZvvrqK9LS0s54rr+/P/7+/udSmkjzYbFAaKyxtR5w8vMV9pPDsLoZY/b9UFkKeZuM7URWH6N/2PFN9uvDsTbGLDUREREPtHF/EQA9kiJMrUNERESaRqOCL5fLxT333MP8+fNZsmQJKSmnmIVyCs899xxPPfUUCxcupG/fvudUqIg0UkA4JPQ0thNVO+DID6eeKXZkH9RU1vYb2wO7T/HaofGnuANl7X7Q2ff8ExERaUpVNU62HDB6x2rGl4iISPPQqOArPT2duXPn8vHHHxMaGkpubi4A4eHhBAYaM0BuueUWEhMTmTFjBgDPPvssU6dOZe7cubRp06b+mpCQEEJC1FBUxBQ+/tCyo7GdyFkDxQdOHYod3guOYig5aGxZ35x8fUC4EYCFJxkzw3z8jTtT+gQct+8PPoEnPD7T1wDjhgA+AWC1XfxfHxER8Uo78kpwVDsJ9fchpUWw2eWIiIhIE2hU8DVr1iwAhg0b1uD4W2+9xW233QZAVlYWVqu1wTWVlZVcf/31Da557LHHePzxxxtfsYhcXFYbRCQbW8rQhs+5XHD08GlCsT1QmmcssTy4wdguSn0+Zx+WnfV5/ieEdKf5avMHqxohi4h4qszj+ntZrbrBi4iISHPQ6KWOP2bJkiUNHu/bt68x30JE3JnFAsEtjC3pFMuWK8uO3XWyNNdYUlldcfLXqopTHz/dV2fVse/hrDb6k1WWNtnbPsZi9DuL7QZx3SE21diPaKU7ZIqIeICNamwvIiLS7Oi2bSJy4fgFG0FQbLcL+7o11VDjOC4QO0NIdtpQ7XTXlJ85eKsqB+pCfxcc3m1sWz85Vp9/WO37ToW4VONrTBfj10NERNxGZk4RoMb2IiIizYmCLxFxfzYfYzMjSHK5jFlm1RXgKIGC7bV3w9wMuZugYJvR9yxrlbHVs0CLdrWBWPfaWWKpEJ6s2WEiIiaoqKph28ESALonasaXiIhIc6HgS0TkTCwWsPkam38ohCVAu8uPPV9dCYd2GiFYXuaxQKwsHw7tMrYtHx873z/8WAhWF4rFdAG/oKZ/byIizci23BKqnS6igv1Iigw0uxwRERFpIgq+RETOh4/fccs7bzh2vDTfmBmWWzs7LG+TMVvMYTfuhtngjph1s8NSGy6XDE/S7DARkQskY38RYMz2suhnq4iISLOh4EtE5GIIiYGQ4dBu+LFj1ZVQuKN2qWRdKLYJygqOmx320bHzA8KPNdCvC8RaanaYiMi52JhtNLbvocb2IiIizYqCLxGRpuLjZ4RXcakNj5fmQ27msZlhuZugcDtU2OGHlcZWx2KFqHbHLZes7R+m2WEiImdU19i+uxrbi4iINCsKvkREzBYSA+1HGFud6koj/Mrb3DAUKysweood2nma2WGpx/qHxXQFX/WxEREpc1SzK78U0IwvERGR5kbBl4iIO/Lxg7juxtbjxmPHS/KOLZWsa6T/Y7PD6nqG1YViYYmaHSYizcrmA8U4XRAXFkBMWIDZ5YiIiEgTUvAlIuJJQmONrcHsMIfRO6yuZ1jdcsmjhcdmh22ef+z8gIiGM8NCEwAXuFxn8ZUzP3fi8VMd+7Gv9dec7bWnqMnmCxGtjZsGRKaoL5pIM1ff2F6zvURERJodBV8iIp7Ox//Y7LA6LlftnSUzG95ZsnAHVBTBDyuMrbkIS4SotsbWol3tfjuIStFyUJFmIGO/GtuLiIg0Vwq+RES8kcVy3OywkceOVzugYPtxSyUzofzIsWuwnOVXTn3sdNec9etzDtecoqbqCji8Fw7vNpaBFucY277lJ/9a1YVixwdidTPFfLUkSsQbHJvxFWFqHSIiItL0FHyJiDQnPv4Qn2ZszYHLBUcPw+E9Rgh2aPdx+3vAcaZQzGKEYi3qZofVhWPtILKNQrHz5XKBoxiOHoLwVmDTkEQuDvvRKvYdOgpAWqJmfImIiDQ3GmWKiIj3slgguIWxJfdr+Fx9KHZiIFa77yiG4v3GtnfZiS8M4UknLJ+sDceaayhWU2X8eh4tNMKsstqvddvxj+v2nVXGtZMyIaKVufWL18rMMZY5JkcFEhnsZ3I1IiIi0tQUfImISPPUIBS7pOFzLpcRzJwUiNXOFKssAXu2se1deuILQ3hy7Uyx45ZO1oViPv5N9Q7PncsFlaW1AdWpwqxCKDt0bP/oIWNJ6bnwDYKK4gtbv8hxNtYuc0zTMkcREZFmScGXiIjIiSwWCI42tlb9Gz7nchkB0OHaUKwuEDu857hQLMvY9iw54XWtx80Ua9ewr1hkG/C5SLNRaqqh/PCPh1dlx83QqnGcwzeyQFAUBLWAoGhjPzi6dr9F7X7UscdBLXTHTbnoMmsb22uZo4iISPOk4EtERKQxLBYIaWlsrS5t+JzLBWUFpwjEar9WlkJRlrGdMhRLPnnpZIt2ENH6WCjmckFl2RnCq1PM0qooOrf36hN4clh1yse1QVdgBFht5/a9RC6SDM34EhERadYUfImIiFwoFguExBjb6UKxkwKx2uWTVWVQ9IOx7fn6hNetnSnmrDHCrOqKcykOAiNPCKtanBxe1c/SagF+wef8SyHiDgpKHBywV2CxQGpimNnliIiIiAkUfImIiDSF40Ox1gMaPudyQWn+qe88ebguFMtqeI3Nv2GAdcrw6rhZWQERunOiNDuZOUUAtI0OJjTA19xiRERExBQaAYuIiJjNYoHQWGNrPbDhcy4XlObBkX1g8z0WZvkFG9eJyGkNbBfN/LsHUuqoNrsUERERMYmCLxEREXdmsUBonLGJSKME+Nro1SrS7DJERETERFazCxAREREREREREbkYFHyJiIiIiIiIiIhXUvAlIiIiIiIiIiJeScGXiIiIiIiIiIh4JQVfIiIiIiIiIiLilRR8iYiIiIiIiIiIV1LwJSIiIiIiIiIiXknBl4iIiIiIiIiIeCUFXyIiIiIiIiIi4pUUfImIiIiIiIiIiFdS8CUiIiIiIiIiIl5JwZeIiIiIiIiIiHglBV8iIiIiIiIiIuKVFHyJiIiIiIiIiIhXUvAlIiIiIiIiIiJeScGXiIiIiIiIiIh4JQVfIiIiIiIiIiLilRR8iYiIiDQTr776Km3atCEgIID+/fvz3XffnfbcDz/8kL59+xIREUFwcDA9e/bknXfeaXCOy+Vi6tSpxMfHExgYyMiRI9m5c+fFfhsiIiIiZ03Bl4iIiEgz8P777zN58mQee+wx1q1bR48ePRg9ejT5+fmnPD8qKoqHH36YVatWkZGRwYQJE5gwYQILFy6sP+e5557jpZdeYvbs2Xz77bcEBwczevRoKioqmuptiYiIiJyRxeVyucwu4scUFxcTHh6O3W4nLCzM7HJERETEA2j80FD//v3p168fr7zyCgBOp5Pk5GTuueceHnroobN6jd69e3PVVVfx5JNP4nK5SEhI4P777+eBBx4AwG63Exsby9tvv82NN954Vq+pz0lEREQaqzHjB834EhEREfFylZWVrF27lpEjR9Yfs1qtjBw5klWrVv3o9S6Xi0WLFrF9+3aGDh0KwN69e8nNzW3wmuHh4fTv3/+Mr+lwOCguLm6wiYiIiFwsPmYXcDbqJqVpYCQiIiJnq27c4AGT2y+6wsJCampqiI2NbXA8NjaWbdu2nfY6u91OYmIiDocDm83Ga6+9xhVXXAFAbm5u/Wuc+Jp1z53KjBkzmDZt2knHNc4TERGRs9WYcZ5HBF8lJSUAJCcnm1yJiIiIeJqSkhLCw8PNLsMjhYaGsmHDBkpLS1m0aBGTJ0+mbdu2DBs27Jxfc8qUKUyePLn+cU5ODl27dtU4T0RERBrtbMZ5HhF8JSQkkJ2dTWhoKBaL5YK/fnFxMcnJyWRnZ6u3hBvT5+Q59Fl5Dn1WnkOfVeO5XC5KSkpISEgwuxTTRUdHY7PZyMvLa3A8Ly+PuLi4015ntVpp3749AD179mTr1q3MmDGDYcOG1V+Xl5dHfHx8g9fs2bPnaV/T398ff3//+schISEa5wmgz8pT6HPyHPqsPIc+q8ZrzDjPI4Ivq9VKUlLSRf8+YWFh+k3mAfQ5eQ59Vp5Dn5Xn0GfVOJrpZfDz86NPnz4sWrSIsWPHAkZz+0WLFjFx4sSzfh2n04nD4QAgJSWFuLg4Fi1aVB90FRcX8+2333LXXXed9WtqnCcn0mflGfQ5eQ59Vp5Dn1XjnO04zyOCLxERERE5P5MnT+bWW2+lb9++XHLJJcycOZOysjImTJgAwC233EJiYiIzZswAjF5cffv2pV27djgcDj7//HPeeecdZs2aBYDFYmHSpElMnz6dDh06kJKSwqOPPkpCQkJ9uCYiIiJiNgVfIiIiIs3ADTfcQEFBAVOnTiU3N5eePXuyYMGC+ub0WVlZWK3HbvhdVlbG3Xffzf79+wkMDKRz5868++673HDDDfXnPPjgg5SVlXHHHXdQVFTE4MGDWbBgAQEBAU3+/kRERERORcEXRq+Jxx57rEG/CXE/+pw8hz4rz6HPynPos5ILYeLEiadd2rhkyZIGj6dPn8706dPP+HoWi4UnnniCJ5544kKVeMHpz47n0GflGfQ5eQ59Vp5Dn9XFZXHpHt8iIiIiIiIiIuKFrD9+ioiIiIiIiIiIiOdR8CUiIiIiIiIiIl5JwZeIiIiIiIiIiHglBV8iIiIiIiIiIuKVmn3w9eqrr9KmTRsCAgLo378/3333ndklyQlmzJhBv379CA0NJSYmhrFjx7J9+3azy5Kz8Mwzz2CxWJg0aZLZpcgp5OTk8Ktf/YoWLVoQGBhI9+7d+f77780uS45TU1PDo48+SkpKCoGBgbRr144nn3wS3ZdG5OxonOf+NM7zXBrnuTeN89yfxnlNp1kHX++//z6TJ0/mscceY926dfTo0YPRo0eTn59vdmlynKVLl5Kens7q1av58ssvqaqqYtSoUZSVlZldmpzBmjVr+Otf/0paWprZpcgpHDlyhEGDBuHr68sXX3zBli1beOGFF4iMjDS7NDnOs88+y6xZs3jllVfYunUrzz77LM899xwvv/yy2aWJuD2N8zyDxnmeSeM896ZxnmfQOK/pWFzNOE7s378//fr145VXXgHA6XSSnJzMPffcw0MPPWRydXI6BQUFxMTEsHTpUoYOHWp2OXIKpaWl9O7dm9dee43p06fTs2dPZs6caXZZcpyHHnqIlStXsnz5crNLkTO4+uqriY2N5Y033qg/Nm7cOAIDA3n33XdNrEzE/Wmc55k0znN/Gue5P43zPIPGeU2n2c74qqysZO3atYwcObL+mNVqZeTIkaxatcrEyuTH2O12AKKiokyuRE4nPT2dq666qsGfL3Evn3zyCX379uXnP/85MTEx9OrVizlz5phdlpxg4MCBLFq0iB07dgCwceNGVqxYwZgxY0yuTMS9aZznuTTOc38a57k/jfM8g8Z5TcfH7ALMUlhYSE1NDbGxsQ2Ox8bGsm3bNpOqkh/jdDqZNGkSgwYNIjU11exy5BTmzZvHunXrWLNmjdmlyBns2bOHWbNmMXnyZP74xz+yZs0a7r33Xvz8/Lj11lvNLk9qPfTQQxQXF9O5c2dsNhs1NTU89dRTjB8/3uzSRNyaxnmeSeM896dxnmfQOM8zaJzXdJpt8CWeKT09nU2bNrFixQqzS5FTyM7O5r777uPLL78kICDA7HLkDJxOJ3379uXpp58GoFevXmzatInZs2drQORG/vWvf/HPf/6TuXPn0q1bNzZs2MCkSZNISEjQ5yQiXkfjPPemcZ7n0DjPM2ic13SabfAVHR2NzWYjLy+vwfG8vDzi4uJMqkrOZOLEiXz66acsW7aMpKQks8uRU1i7di35+fn07t27/lhNTQ3Lli3jlVdeweFwYLPZTKxQ6sTHx9O1a9cGx7p06cJ//vMfkyqSU/n973/PQw89xI033ghA9+7d+eGHH5gxY4YGRCJnoHGe59E4z/1pnOc5NM7zDBrnNZ1m2+PLz8+PPn36sGjRovpjTqeTRYsWMWDAABMrkxO5XC4mTpzI/PnzWbx4MSkpKWaXJKcxYsQIMjMz2bBhQ/3Wt29fxo8fz4YNGzQYciODBg066XbxO3bsoHXr1iZVJKdy9OhRrNaGf1XbbDacTqdJFYl4Bo3zPIfGeZ5D4zzPoXGeZ9A4r+k02xlfAJMnT+bWW2+lb9++XHLJJcycOZOysjImTJhgdmlynPT0dObOncvHH39MaGgoubm5AISHhxMYGGhydXK80NDQk3pyBAcH06JFC/XqcDO/+93vGDhwIE8//TS/+MUv+O6773j99dd5/fXXzS5NjnPNNdfw1FNP0apVK7p168b69et58cUX+fWvf212aSJuT+M8z6BxnufQOM9zaJznGTTOazoWl8vlMrsIM73yyiv86U9/Ijc3l549e/LSSy/Rv39/s8uS41gsllMef+utt7jtttuathhptGHDhuk2127q008/ZcqUKezcuZOUlBQmT57M7bffbnZZcpySkhIeffRR5s+fT35+PgkJCdx0001MnToVPz8/s8sTcXsa57k/jfM8m8Z57kvjPPencV7TafbBl4iIiIiIiIiIeKdm2+NLRERERERERES8m4IvERERERERERHxSgq+RERERERERETEKyn4EhERERERERERr6TgS0REREREREREvJKCLxERERERERER8UoKvkRERERERERExCsp+BIREREREREREa+k4EtEmgWLxcJHH31kdhkiIiIichForCcip6PgS0Quuttuuw2LxXLSduWVV5pdmoiIiIicJ431RMSd+ZhdgIg0D1deeSVvvfVWg2P+/v4mVSMiIiIiF5LGeiLirjTjS0SahL+/P3FxcQ22yMhIwJiaPmvWLMaMGUNgYCBt27blgw8+aHB9ZmYmw4cPJzAwkBYtWnDHHXdQWlra4Jw333yTbt264e/vT3x8PBMnTmzwfGFhIddeey1BQUF06NCBTz75pP65I0eOMH78eFq2bElgYCAdOnQ4afAmIiIiIqemsZ6IuCsFXyLiFh599FHGjRvHxo0bGT9+PDfeeCNbt24FoKysjNGjRxMZGcmaNWv497//zVdffdVgsDNr1izS09O54447yMzM5JNPPqF9+/YNvse0adP4xS9+QUZGBj/5yU8YP348hw8frv/+W7Zs4YsvvmDr1q3MmjWL6OjopvsFEBEREfFiGuuJiGlcIiIX2a233uqy2Wyu4ODgBttTTz3lcrlcLsB15513Nrimf//+rrvuusvlcrlcr7/+uisyMtJVWlpa//xnn33mslqtrtzcXJfL5XIlJCS4Hn744dPWALgeeeSR+selpaUuwPXFF1+4XC6X65prrnFNmDDhwrxhERERkWZEYz0RcWfq8SUiTeLyyy9n1qxZDY5FRUXV7w8YMKDBcwMGDGDDhg0AbN26lR49ehAcHFz//KBBg3A6nWzfvh2LxcKBAwcYMWLEGWtIS0ur3w8ODiYsLIz8/HwA7rrrLsaNG8e6desYNWoUY8eOZeDAgef0XkVERESaG431RMRdKfgSkSYRHBx80nT0CyUwMPCszvP19W3w2GKx4HQ6ARgzZgw//PADn3/+OV9++SUjRowgPT2d559//oLXKyIiIuJtNNYTEXelHl8i4hZWr1590uMuXboA0KVLFzZu3EhZWVn98ytXrsRqtdKpUydCQ0Np06YNixYtOq8aWrZsya233sq7777LzJkzef3118/r9URERETEoLGeiJhFM75EpEk4HA5yc3MbHPPx8alvKvrvf/+bvn37MnjwYP75z3/y3Xff8cYbbwAwfvx4HnvsMW699VYef/xxCgoKuOeee7j55puJjY0F4PHHH+fOO+8kJiaGMWPGUFJSwsqVK7nnnnvOqr6pU6fSp08funXrhsPh4NNPP60fjImIiIjImWmsJyLuSsGXiDSJBQsWEB8f3+BYp06d2LZtG2DchWfevHncfffdxMfH895779G1a1cAgoKCWLhwIffddx/9+vUjKCiIcePG8eKLL9a/1q233kpFRQV//vOfeeCBB4iOjub6668/6/r8/PyYMmUK+/btIzAwkCFDhjBv3rwL8M5FREREvJ/GeiLiriwul8tldhEi0rxZLBbmz5/P2LFjzS5FRERERC4wjfVExEzq8SUiIiIiIiIiIl5JwZeIiIiIiIiIiHglLXUUERERERERERGvpBlfIiIiIiIiIiLilRR8iYiIiIiIiIiIV1LwJSIiIiIiIiIiXknBl4iIiIiIiIiIeCUFXyIiIiIiIiIi4pUUfImIiIiIiIiIiFdS8CUiIiIiIiIiIl5JwZeIiIiIiIiIiHil/wdRZTt3wxPHGQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.5 Saving EffNetB2 feature extractor"
      ],
      "metadata": {
        "id": "O-GS6bGIb6cd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from going_modular.going_modular import utils\n",
        "\n",
        "# Save the model\n",
        "utils.save_model(model=model,\n",
        "                 target_dir=\"models\",\n",
        "                 model_name=\"model.pth\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MIstYDs_bUH2",
        "outputId": "daa8f9d9-cfb6-4d7d-e123-e67f930bee18"
      },
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[INFO] Saving model to: models/model.pth\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.6 Checking the size of EffNetB2 feature extractor"
      ],
      "metadata": {
        "id": "XDYF_rBvcwG8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from pathlib import Path\n",
        "\n",
        "# Get the model size in bytes then convert to megabytes\n",
        "pretrained_model_size = Path(\"models/model.pth\").stat().st_size // (1024*1024) # division converts bytes to megabytes (roughly)\n",
        "print(f\"Pretrained feature extractor model size: {pretrained_model_size} MB\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "76My23AUbDkw",
        "outputId": "c627bf84-6ab3-4994-df67-726be58ac3fa"
      },
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Pretrained feature extractor model size: 78 MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.7 Collecting EffNetB2 feature extractor stats"
      ],
      "metadata": {
        "id": "lE1ahpn1crif"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Count number of parameters in EffNetB2\n",
        "model_total_params = sum(torch.numel(param) for param in model.parameters())\n",
        "model_total_params"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SNi596GYbDoa",
        "outputId": "af39948d-97f1-4a50-d87f-15e4a8b02a00"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "20306869"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a dictionary with EffNetB2 statistics\n",
        "stats = {\"test_loss\": results[\"test_loss\"][-1],\n",
        "          \"test_acc\": results[\"test_acc\"][-1],\n",
        "          \"number_of_parameters\": model_total_params,\n",
        "          \"model_size (MB)\": pretrained_model_size}\n",
        "stats"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iFjvK8ydbDsV",
        "outputId": "a16127fa-82f7-43c3-8216-05ce21205e12"
      },
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'test_loss': 2.141662242291849,\n",
              " 'test_acc': 0.6099987828627069,\n",
              " 'number_of_parameters': 20306869,\n",
              " 'model_size (MB)': 78}"
            ]
          },
          "metadata": {},
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 3. Turning our FoodVision Big model into a deployable app"
      ],
      "metadata": {
        "id": "JkXEHjlU6gwA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from pathlib import Path\n",
        "\n",
        "# Create FoodVision Big demo path\n",
        "foodvision_big_demo_path = Path(\"demos/foodvision/\")\n",
        "\n",
        "# Make FoodVision Big demo directory\n",
        "foodvision_big_demo_path.mkdir(parents=True, exist_ok=True)\n",
        "\n",
        "# Make FoodVision Big demo examples directory\n",
        "(foodvision_big_demo_path / \"examples\").mkdir(parents=True, exist_ok=True)"
      ],
      "metadata": {
        "id": "arwy5GCnbDwn"
      },
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.1 Downloading an example image and moving it to the examples directory"
      ],
      "metadata": {
        "id": "48FYRQ1T7IXm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Move trained model to FoodVision Big demo folder (will error if model is already moved)\n",
        "# !mv models/model.pth demos/foodvision"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PkhJHvEkdzR-",
        "outputId": "dd002b24-2bde-4507-ffb5-1dfa2cbf1f7a"
      },
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "mv: cannot stat 'models/model.pth': No such file or directory\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.2 Saving Food101 class names to file (class_names.txt)"
      ],
      "metadata": {
        "id": "_X1Cqmha8MqF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create path to Food101 class names\n",
        "foodvision_big_class_names_path = foodvision_big_demo_path / \"class_names.txt\"\n",
        "\n",
        "# Write Food101 class names list to file\n",
        "with open(foodvision_big_class_names_path, \"w\") as f:\n",
        "    print(f\"[INFO] Saving Food101 class names to {foodvision_big_class_names_path}\")\n",
        "    f.write(\"\\n\".join(food101_class_names)) # leave a new line between each class"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VvMQZQ5RdzVQ",
        "outputId": "c3ce5335-9ed8-4dc5-f5e4-8804e92a8785"
      },
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[INFO] Saving Food101 class names to demos/foodvision/class_names.txt\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.3 Turning our FoodVision Big model into a Python script (model.py)"
      ],
      "metadata": {
        "id": "mM82SwTe8bAN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%%writefile demos/foodvision/model.py\n",
        "import torch\n",
        "import torchvision\n",
        "\n",
        "from torch import nn\n",
        "\n",
        "def create_model(num_classes:int=3,\n",
        "                 seed:int=42):\n",
        "    \"\"\"Creates an feature extractor model and transforms.\n",
        "\n",
        "    Args:\n",
        "        num_classes (int, optional): number of classes in the classifier head.\n",
        "            Defaults to 3.\n",
        "        seed (int, optional): random seed value. Defaults to 42.\n",
        "\n",
        "    Returns:\n",
        "        model (torch.nn.Module):  feature extractor model.\n",
        "        transforms (torchvision.transforms): image transforms.\n",
        "    \"\"\"\n",
        "    # 1. Setup pretrained model\n",
        "    weights = torchvision.models.EfficientNet_V2_S_Weights.DEFAULT\n",
        "\n",
        "    # 2. Get EffNetB2 transforms\n",
        "    transforms = weights.transforms()\n",
        "\n",
        "    # 3. Setup pretrained model\n",
        "    model = torchvision.models.efficientnet_v2_s(weights=weights).to(device) # could also use weights=\"DEFAULT\" .to(device)\n",
        "\n",
        "    # 4. Freeze the base layers in the model (this will freeze all layers to begin with)\n",
        "    for param in model.parameters():\n",
        "        param.requires_grad = False\n",
        "\n",
        "    # 5. Change classifier head with random seed for reproducibility\n",
        "    torch.manual_seed(seed)\n",
        "    model.classifier = torch.nn.Sequential(\n",
        "        torch.nn.Dropout(p=0.2, inplace=True),\n",
        "        torch.nn.Linear(in_features=1280,\n",
        "                        out_features=num_classes, # same number of output units as our number of classes\n",
        "                        bias=True))\n",
        "\n",
        "    return model, transforms"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EpMIqBFIdzZK",
        "outputId": "36a6d43d-11ce-46cc-d5d0-05f4d80db37c"
      },
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Writing demos/foodvision/model.py\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.4 Turning our FoodVision Big Gradio app into a Python script (app.py)"
      ],
      "metadata": {
        "id": "luZwVseM8-_Q"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%%writefile demos/foodvision/app.py\n",
        "### 1. Imports and class names setup ###\n",
        "import gradio as gr\n",
        "import os\n",
        "import torch\n",
        "\n",
        "from model import create_model\n",
        "from timeit import default_timer as timer\n",
        "from typing import Tuple, Dict\n",
        "\n",
        "# Setup class names\n",
        "with open(\"class_names.txt\", \"r\") as f: # reading them in from class_names.txt\n",
        "    class_names = [food_name.strip() for food_name in  f.readlines()]\n",
        "\n",
        "### 2. Model and transforms preparation ###\n",
        "\n",
        "# Create model\n",
        "model, transforms = create_model(\n",
        "    num_classes=101, # could also use len(class_names)\n",
        ")\n",
        "\n",
        "# Load saved weights\n",
        "model.load_state_dict(\n",
        "    torch.load(\n",
        "        f=\"model.pth\",\n",
        "        map_location=torch.device(\"cpu\"),  # load to CPU\n",
        "    )\n",
        ")\n",
        "\n",
        "### 3. Predict function ###\n",
        "\n",
        "# Create predict function\n",
        "def predict(img) -> Tuple[Dict, float]:\n",
        "    \"\"\"Transforms and performs a prediction on img and returns prediction and time taken.\n",
        "    \"\"\"\n",
        "    # Start the timer\n",
        "    start_time = timer()\n",
        "\n",
        "    # Transform the target image and add a batch dimension\n",
        "    img = transforms(img).unsqueeze(0)\n",
        "\n",
        "    # Put model into evaluation mode and turn on inference mode\n",
        "    model.eval()\n",
        "    with torch.inference_mode():\n",
        "        # Pass the transformed image through the model and turn the prediction logits into prediction probabilities\n",
        "        pred_probs = torch.softmax(model(img), dim=1)\n",
        "\n",
        "    # Create a prediction label and prediction probability dictionary for each prediction class (this is the required format for Gradio's output parameter)\n",
        "    pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))}\n",
        "\n",
        "    # Calculate the prediction time\n",
        "    pred_time = round(timer() - start_time, 5)\n",
        "\n",
        "    # Return the prediction dictionary and prediction time\n",
        "    return pred_labels_and_probs, pred_time\n",
        "\n",
        "### 4. Gradio app ###\n",
        "\n",
        "# Create title, description and article strings\n",
        "title = \"FoodVision\"\n",
        "description = \"An Efficient Net V2 s feature extractor computer vision model to classify images of food into [101 different classes](https://github.com/mrdbourke/pytorch-deep-learning/blob/main/extras/food101_class_names.txt).\"\n",
        "article = \"Created as an study in image classification\"\n",
        "\n",
        "# Create examples list from \"examples/\" directory\n",
        "example_list = [[\"examples/\" + example] for example in os.listdir(\"examples\")]\n",
        "\n",
        "# Create Gradio interface\n",
        "demo = gr.Interface(\n",
        "    fn=predict,\n",
        "    inputs=gr.Image(type=\"pil\"),\n",
        "    outputs=[\n",
        "        gr.Label(num_top_classes=5, label=\"Predictions\"),\n",
        "        gr.Number(label=\"Prediction time (s)\"),\n",
        "    ],\n",
        "    examples=example_list,\n",
        "    title=title,\n",
        "    description=description,\n",
        "    article=article,\n",
        ")\n",
        "\n",
        "# Launch the app!\n",
        "demo.launch()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bPXQ_dTCdzcd",
        "outputId": "e1860a08-272e-4344-8ac9-f6e53294746e"
      },
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Writing demos/foodvision/app.py\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.5 Creating a requirements file for FoodVision Big (requirements.txt)"
      ],
      "metadata": {
        "id": "5HOn1xm9-7kO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%%writefile demos/foodvision/requirements.txt\n",
        "torch==2.1.0\n",
        "torchvision==0.16.0\n",
        "gradio==3.1.4"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Yy_HV7-RbD0F",
        "outputId": "cd46c143-f7a7-4eb0-89bf-2847618d3399"
      },
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Writing demos/foodvision/requirements.txt\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.6 Downloading our FoodVision Big app files"
      ],
      "metadata": {
        "id": "I4FUIIhS_Anj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Zip foodvision_big folder but exclude certain files\n",
        "!cd demos/foodvision && zip -r ../foodvision.zip * -x \"*.pyc\" \"*.ipynb\" \"*__pycache__*\" \"*ipynb_checkpoints*\"\n",
        "\n",
        "# Download the zipped FoodVision Big app (if running in Google Colab)\n",
        "try:\n",
        "    from google.colab import files\n",
        "    files.download(\"demos/foodvision.zip\")\n",
        "except:\n",
        "    print(\"Not running in Google Colab, can't use google.colab.files.download()\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 146
        },
        "id": "7vADMFNLSzy0",
        "outputId": "0803c654-d7a3-417b-b944-2277bc493077"
      },
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "  adding: app.py (deflated 55%)\n",
            "  adding: class_names.txt (deflated 48%)\n",
            "  adding: examples/ (stored 0%)\n",
            "  adding: examples/hamburger.jpg (deflated 1%)\n",
            "  adding: model.pth (deflated 7%)\n",
            "  adding: model.py (deflated 56%)\n",
            "  adding: requirements.txt (deflated 6%)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_b62c9479-e162-40e2-9dea-4b00b92b2c0b\", \"foodvision.zip\", 76275893)"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Zip foodvision_big folder but exclude certain files\n",
        "!cd going_modular/ && zip -r ../going_modular.zip * -x \"*.pyc\" \"*.ipynb\" \"*__pycache__*\" \"*ipynb_checkpoints*\"\n",
        "\n",
        "try:\n",
        "    from google.colab import files\n",
        "    files.download(\"going_modular.zip\")\n",
        "except:\n",
        "    print(\"Not running in Google Colab, can't use google.colab.files.download()\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "K03cL5BlFQ8o",
        "outputId": "0391f15f-eef9-4caa-ffd8-74f687edf54a"
      },
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "  adding: going_modular/ (stored 0%)\n",
            "  adding: going_modular/README.md (deflated 50%)\n",
            "  adding: going_modular/engine.py (deflated 74%)\n",
            "  adding: going_modular/train.py (deflated 57%)\n",
            "  adding: going_modular/data_setup.py (deflated 66%)\n",
            "  adding: going_modular/model_builder.py (deflated 67%)\n",
            "  adding: going_modular/predictions.py (deflated 62%)\n",
            "  adding: going_modular/utils.py (deflated 57%)\n",
            "  adding: models/ (stored 0%)\n",
            "  adding: models/05_going_modular_script_mode_tinyvgg_model.pth (deflated 12%)\n",
            "  adding: models/05_going_modular_cell_mode_tinyvgg_model.pth (deflated 12%)\n",
            "  adding: README.md (deflated 61%)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_7168777d-aa0f-4ac1-a098-2c3e5968ef22\", \"going_modular.zip\", 71635)"
            ]
          },
          "metadata": {}
        }
      ]
    }
  ]
}