model_gradient.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "model_gradient_updated.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ksdkamesh99/LowLightEnhancer/blob/master/model_gradient.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AGxUaxHllOkt",
        "outputId": "ba3f502a-3a14-4fc7-b775-224477409680",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a_YnKk4fWmVm",
        "outputId": "f72827a4-79d2-408c-d389-362d89224766",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "cd /content/drive/My Drive/LowLightEnhancement"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/LowLightEnhancement\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vpHqtxxCV2pg"
      },
      "source": [
        "import tensorflow as tf\n",
        "import keras\n",
        "import numpy as np\n",
        "from PIL import Image\n",
        "import matplotlib.pyplot as plt\n"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5GKNxTenWMe9"
      },
      "source": [
        "img_high=np.load(\"image_high.npy\")"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QK5SsO8AWwrF"
      },
      "source": [
        "img_low=np.load(\"image_low.npy\")"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "En1X3_txnq35"
      },
      "source": [
        "img_low=img_low/255\n",
        "img_high=img_high/255"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tM8PD-nLW2V8",
        "outputId": "ebca3d36-82ae-4ac8-8588-49682b9255ac",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "img_high.shape"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1485, 96, 96, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jp7E7ko6JKnO",
        "outputId": "b1968960-1972-44da-ef92-37918010b678",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        }
      },
      "source": [
        "plt.imshow(img_low[0])"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f9b9c39beb8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QDic1QdxbVx3"
      },
      "source": [
        "## Illumination Mask Network"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-oLa_B84wQCW"
      },
      "source": [
        "input_layer_1=keras.layers.Input(shape=(96,96,3))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fp6MQbewW6yg"
      },
      "source": [
        "top=keras.layers.Conv2D(64,kernel_size=(3,3),input_shape=(96,96,3),padding='same')(input_layer_1)\n",
        "top=keras.layers.Conv2D(64,kernel_size=(3,3),padding='same')(top)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KEkWaFosn2qK",
        "outputId": "35f0e27d-6417-4f6b-c5a8-9c3e89b78848",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "\n",
        "top.get_shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<bound method Tensor.get_shape of <tf.Tensor 'conv2d_776/BiasAdd:0' shape=(None, 96, 96, 64) dtype=float32>>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1662
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WFrTH7Voeoa1"
      },
      "source": [
        "\n",
        "bottom_inp=input_layer_1\n",
        "bottom_resize=tf.keras.layers.Lambda(\n",
        "    lambda img: tf.image.resize(img,(60,60))\n",
        ")(bottom_inp)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wFYVu3v5euM0"
      },
      "source": [
        "bottom=keras.layers.Conv2D(64,kernel_size=(3,3),input_shape=(60,60,3),padding='same')(bottom_resize)\n",
        "bottom=keras.layers.Conv2D(64,kernel_size=(3,3),padding='same')(bottom)\n",
        "bottom=keras.layers.Conv2D(64,kernel_size=(3,3),padding='same')(bottom)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UysJ2KJ6hYNz",
        "outputId": "f158c7a8-30a9-4a2c-8d5f-37f95d92ea8d",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "bottom.get_shape()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([None, 60, 60, 64])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1665
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xLJxKqb3RMqk"
      },
      "source": [
        "bottom=keras.layers.experimental.preprocessing.Resizing(96,96)(bottom)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VfAOLAA5dzpI",
        "outputId": "2c218a79-c730-471f-9904-91803d866580",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "bottom.get_shape()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([None, 96, 96, 64])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1667
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-oBklAH3iLBT",
        "outputId": "e5be4d12-9e74-47c9-a516-77e0f5ab3df8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "top.get_shape()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([None, 96, 96, 64])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1668
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xE9AeLqLiNjq"
      },
      "source": [
        "merged=keras.layers.concatenate([top,bottom])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bdXU8tmnhuCx",
        "outputId": "d2c0d1d7-2f36-473a-a0e6-55c379b4d183",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "merged"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor 'concatenate_200/concat:0' shape=(None, 96, 96, 128) dtype=float32>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1670
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8bUMJqgmi06Y"
      },
      "source": [
        "merged=keras.layers.Conv2D(32,kernel_size=(7,7),padding='same')(merged)\n",
        "merged=keras.layers.Conv2D(8,kernel_size=(1,1),padding='same')(merged)\n",
        "merged=keras.layers.Conv2D(1,kernel_size=(5,5),activation='sigmoid',padding='same')(merged)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YfLuYpLSi2Bb",
        "outputId": "79e899ea-2315-43a7-94c9-bffef5a391e2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "merged.get_shape()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([None, 96, 96, 1])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1672
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0KISCHprjQ0_"
      },
      "source": [
        "model_illumination_mask=keras.models.Model(inputs=input_layer_1,outputs=merged)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZVr8vme_jY6m",
        "outputId": "9bf8d04d-13b9-4232-8303-6c795233c6db",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model_illumination_mask.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_101\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_26 (InputLayer)           [(None, 96, 96, 3)]  0                                            \n",
            "__________________________________________________________________________________________________\n",
            "lambda_300 (Lambda)             (None, 60, 60, 3)    0           input_26[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_777 (Conv2D)             (None, 60, 60, 64)   1792        lambda_300[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_778 (Conv2D)             (None, 60, 60, 64)   36928       conv2d_777[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_775 (Conv2D)             (None, 96, 96, 64)   1792        input_26[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_779 (Conv2D)             (None, 60, 60, 64)   36928       conv2d_778[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_776 (Conv2D)             (None, 96, 96, 64)   36928       conv2d_775[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "resizing_25 (Resizing)          (None, 96, 96, 64)   0           conv2d_779[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_200 (Concatenate)   (None, 96, 96, 128)  0           conv2d_776[0][0]                 \n",
            "                                                                 resizing_25[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_780 (Conv2D)             (None, 96, 96, 32)   200736      concatenate_200[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_781 (Conv2D)             (None, 96, 96, 8)    264         conv2d_780[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_782 (Conv2D)             (None, 96, 96, 1)    201         conv2d_781[0][0]                 \n",
            "==================================================================================================\n",
            "Total params: 315,569\n",
            "Trainable params: 315,569\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vMmClQsdb1Os"
      },
      "source": [
        "# Illumination Map "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wsy6Q8eKflRV",
        "outputId": "71061ee4-6c4f-4a0a-e7d9-c3fd0e7746d4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "merged.get_shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<bound method Tensor.get_shape of <tf.Tensor 'conv2d_782/Sigmoid:0' shape=(None, 96, 96, 1) dtype=float32>>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1675
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8mc6wmEYhL2a"
      },
      "source": [
        "merged=keras.layers.Concatenate()([input_layer_1,merged])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8dk7Ho8Oh7Zv",
        "outputId": "ede02463-0fcf-47fb-acdf-7b264ec60db4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "merged"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor 'concatenate_201/concat:0' shape=(None, 96, 96, 4) dtype=float32>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1677
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ji0GvexEjmey"
      },
      "source": [
        "def ieb(input_feature):\n",
        "  ieb1=keras.layers.Conv2D(32,kernel_size=(3,3),activation='relu',padding='same')(input_feature)\n",
        "  ieb1=keras.layers.Conv2D(32,kernel_size=(3,3),activation='relu',padding='same')(ieb1)\n",
        "  max_pool=keras.layers.GlobalMaxPooling2D()(ieb1)\n",
        "  avg_pool=keras.layers.GlobalAveragePooling2D()(ieb1)\n",
        "  dense1=keras.layers.Dense(8,activation='relu')\n",
        "  dense2=keras.layers.Dense(32,activation='sigmoid')\n",
        "  max_pool=dense1(max_pool)\n",
        "  max_pool=dense2(max_pool)\n",
        "  avg_pool=dense1(avg_pool)\n",
        "  avg_pool=dense2(avg_pool)\n",
        "  '''max_pool=keras.layers.Lambda(\n",
        "    lambda image: keras.backend.expand_dims(keras.backend.expand_dims(image,axis=1),axis=1))(max_pool)\n",
        "  avg_pool=keras.layers.Lambda(\n",
        "    lambda image: keras.backend.expand_dims(keras.backend.expand_dims(image,axis=1),axis=1))(avg_pool)'''\n",
        "  channel=keras.layers.Add()([max_pool,avg_pool])\n",
        "  ieb1=keras.layers.Multiply()([ieb1,channel])\n",
        "  max_pool_s=tf.keras.layers.Lambda(\n",
        "    lambda x: keras.backend.max(x,axis=3,keepdims=True))(ieb1)\n",
        "  avg_pool_s=keras.layers.Lambda(\n",
        "    lambda x: keras.backend.mean(x,axis=3,keepdims=True))(ieb1)\n",
        "  concat_slayers=keras.layers.Concatenate(axis=3)([avg_pool_s,max_pool_s])\n",
        "  spacial=keras.layers.Conv2D(1,7,activation='sigmoid',padding='same')(concat_slayers)\n",
        "  #spacial=keras.layers.experimental.preprocessing.Resizing(92,92)(spacial)\n",
        "  ieb1=keras.layers.Multiply()([ieb1,spacial])\n",
        "  ieb1=keras.layers.BatchNormalization()(ieb1)\n",
        "  ieb1=keras.layers.Activation('relu')(ieb1)\n",
        "  #ieb1=keras.layers.experimental.preprocessing.Resizing(96,96)(ieb1)\n",
        "  return ieb1\n",
        "\n",
        "\n",
        "  \n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "60ZDXV1qn3dx"
      },
      "source": [
        "ieb_1=ieb(merged)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_om_rsynoISE"
      },
      "source": [
        "ieb_2=ieb(ieb_1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c-evBYfMoQm2"
      },
      "source": [
        "\n",
        "\n",
        "ieb_3=ieb(ieb_2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sa6_PdBUoTYt"
      },
      "source": [
        "ieb_4=ieb(ieb_3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3p78MoqyodOU"
      },
      "source": [
        "ieb_5=ieb(ieb_4)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cQcPd5NbpV8h"
      },
      "source": [
        "added_ieb=keras.layers.concatenate([ieb_1,ieb_2,ieb_3,ieb_4,ieb_5])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qxnoI6qWof9J",
        "outputId": "ac01b947-25b4-4861-cb94-5444e7d49e6a",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "added_ieb"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor 'concatenate_207/concat:0' shape=(None, 96, 96, 160) dtype=float32>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1685
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DsmqNCtzojJ3"
      },
      "source": [
        "\n",
        "impnet=keras.layers.Conv2D(32,(3,3),padding='same')(added_ieb)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JC7AyMLio50w"
      },
      "source": [
        "impnet=keras.layers.Conv2D(8,(3,3),padding='same')(impnet)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PRufKmb1qXs4"
      },
      "source": [
        "impnet=keras.layers.Conv2D(1,(3,3),padding='same')(impnet)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fo2lrmTN7p4A"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4VYnEsMH0Csg"
      },
      "source": [
        "# S/L Block "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1C1alUSK2aYH",
        "outputId": "d2bef2fb-4e76-4a7b-c68b-15ee38f3abfd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "'''impnet=keras.layers.Lambda(\n",
        "    lambda x: x+keras.backend.constant(0.001)\n",
        ")(impnet)'''"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'impnet=keras.layers.Lambda(\\n    lambda x: x+keras.backend.constant(0.001)\\n)(impnet)'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1689
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H9t5JSBf0Hm2"
      },
      "source": [
        "s_l=keras.layers.Lambda(\n",
        "    lambda input:input[0]/input[1]\n",
        ")([input_layer_1,impnet])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tOiIZy9z2GDX",
        "outputId": "a3aeeb9b-b797-46c5-ea83-3f684001a0d1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "s_l"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor 'lambda_311/truediv:0' shape=(None, 96, 96, 3) dtype=float32>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1691
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pW_1R0v4CdGe"
      },
      "source": [
        "# Correction Network"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4jPUCt4RrJ8b"
      },
      "source": [
        "def correction_network(input_feature):\n",
        "  conv1=keras.layers.Conv2D(32,kernel_size=(3,3),strides=(1,1),activation='relu',padding='same')(input_feature)\n",
        "  conv2=keras.layers.Conv2D(32,kernel_size=(3,3),strides=(1,1),activation='relu',padding='same')(conv1)\n",
        "  conv3=keras.layers.Conv2D(16,kernel_size=(3,3),strides=(1,1),activation='relu',padding='same')(conv2)\n",
        "  conv4=keras.layers.Conv2D(16,kernel_size=(3,3),strides=(1,1),activation='relu',padding='same')(conv3)\n",
        "  conv5=keras.layers.Conv2D(3,kernel_size=(3,3),strides=(1,1),activation='sigmoid',padding='same')(conv4)\n",
        "  #conv5=keras.layers.experimental.preprocessing.Resizing(96,96)(conv5)\n",
        "  #conv5=keras.layers.multiply([impnet,conv5])\n",
        "  return conv5"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NptH968rEyrm"
      },
      "source": [
        "final_output=correction_network(s_l)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cXMJRxWLVeIS"
      },
      "source": [
        "# Custom Loss Function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AhmYFUinViQl"
      },
      "source": [
        "import loss as l"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lsu1IugnpmYg"
      },
      "source": [
        "import keras.backend as K"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zF-IjANHnDTH"
      },
      "source": [
        "def enhancement_loss(x,y):\n",
        "  x=K.cast(x,dtype='float32')\n",
        "  y=K.cast(y,dtype='float32')\n",
        "  norm=tf.norm(x-y)\n",
        "  return norm"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-qSiyVGSGYT0",
        "outputId": "db55dda5-2794-4222-94be-7f15c714b069",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "\n",
        "enhancement_loss(img_low[0],img_high[0])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=float32, numpy=78.62597>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1697
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k4XaTActnVE9"
      },
      "source": [
        "\n",
        "def color_loss(x,y):\n",
        "  \n",
        "  x=K.cast(x,dtype='float32')\n",
        "  y=K.cast(y,dtype='float32')\n",
        "  cosine_loss = keras.losses.CosineSimilarity()(x,y)\n",
        "  colorloss=1-cosine_loss\n",
        "  return colorloss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TWfXJy4eGj7r",
        "outputId": "8f37bd4f-afeb-4d35-9096-9302b9850fb0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "color_loss(img_low[0],img_high[0])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=float32, numpy=1.9773979>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1699
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HFJKcdYncCoT"
      },
      "source": [
        "sobelFilter = K.variable([[[[1.,  1.]], [[0.,  2.]],[[-1.,  1.]]],\n",
        "                      [[[2.,  0.]], [[0.,  0.]],[[-2.,  0.]]],\n",
        "                      [[[1., -1.]], [[0., -2.]],[[-1., -1.]]]])\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TGhU4VG_ZwXf"
      },
      "source": [
        "def expandedSobel(inputTensor):\n",
        "    inputChannels = K.reshape(K.ones_like(inputTensor[0,0,0,:]),(1,1,-1,1))\n",
        "    return sobelFilter * inputChannels\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vx00HkqNZ68P"
      },
      "source": [
        "def squareSobelLoss(yTrue,yPred):\n",
        "  \n",
        "  yTrue=K.cast(yTrue,dtype='float32')\n",
        "  yPred=K.cast(yPred,dtype='float32')\n",
        "  filt = expandedSobel(yTrue)\n",
        "  squareSobelTrue =K.square(K.depthwise_conv2d(yTrue,filt))\n",
        "  squareSobelPred =K.square(K.depthwise_conv2d(yPred,filt))\n",
        "  newShape = K.shape(squareSobelTrue)\n",
        "  newShape = K.concatenate([newShape[:-1],\n",
        "                              newShape[-1:]//2,\n",
        "                              K.variable([2],dtype='int32')])\n",
        "  squareSobelTrue = K.sum(K.reshape(squareSobelTrue,newShape),axis=-1)\n",
        "  squareSobelPred = K.sum(K.reshape(squareSobelPred,newShape),axis=-1)\n",
        "  return K.mean(K.abs(squareSobelTrue - squareSobelPred))\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T3xENVa_ejXz"
      },
      "source": [
        "def MeanGradientError(outputs, targets):\n",
        "    outputs=tf.cast(outputs,dtype='float32')\n",
        "    targets=tf.cast(targets,dtype='float32')\n",
        "    filter_x = tf.tile(tf.expand_dims(tf.constant([[-1, -2, -2], [0, 0, 0], [1, 2, 1]], dtype = 'float32'), axis = -1), [1, 1, outputs.shape[-1]])\n",
        "    filter_x = tf.tile(tf.expand_dims(filter_x, axis = -1), [1, 1, 1, outputs.shape[-1]])\n",
        "    filter_y = tf.tile(tf.expand_dims(tf.constant([[-1, 0, 1], [-2, 0, 2], [-1, 0, 1]], dtype = 'float32'), axis = -1), [1, 1, targets.shape[-1]])\n",
        "    filter_y = tf.tile(tf.expand_dims(filter_y, axis = -1), [1, 1, 1, targets.shape[-1]])\n",
        "\n",
        "    # output gradient\n",
        "    output_gradient_x = tf.math.square(tf.nn.conv2d(outputs, filter_x, strides = 1, padding = 'SAME'))\n",
        "    output_gradient_y = tf.math.square(tf.nn.conv2d(outputs, filter_y, strides = 1, padding = 'SAME'))\n",
        "\n",
        "    #target gradient\n",
        "    target_gradient_x = tf.math.square(tf.nn.conv2d(targets, filter_x, strides = 1, padding = 'SAME'))\n",
        "    target_gradient_y = tf.math.square(tf.nn.conv2d(targets, filter_y, strides = 1, padding = 'SAME'))\n",
        "\n",
        "    # square\n",
        "    output_gradients = tf.math.sqrt(tf.math.add(output_gradient_x, output_gradient_y))\n",
        "    target_gradients = tf.math.sqrt(tf.math.add(target_gradient_x, target_gradient_y))\n",
        "\n",
        "    # compute mean gradient error\n",
        "    shape = output_gradients.shape[1:3]\n",
        "    mge = tf.math.reduce_sum(tf.math.squared_difference(output_gradients, target_gradients) / (shape[0] * shape[1]))\n",
        "    return mge"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VIeTO682IlDt"
      },
      "source": [
        "\n",
        "def max_rgb_filter(img):\n",
        "  # img=tf.keras.preprocessing.image.img_to_array(img)\n",
        "  \n",
        "  r=img[:,:,:,0]\n",
        "  g=img[:,:,:,1]\n",
        "  b=img[:,:,:,2]\n",
        "  max_c=tf.maximum(K.maximum(r,g),b)\n",
        "  '''\n",
        "  b_broadcast = K.zeros(K.shape(r), dtype=r.dtype)\n",
        "  bool_r=K.less(r,max)\n",
        "  bool_g=K.less(g,max)\n",
        "  bool_b=K.less(b,max)\n",
        "  r=K.switch(bool_r,b_broadcast,r)\n",
        "  g=K.switch(bool_g,b_broadcast,g)\n",
        "  b=K.switch(bool_b,b_broadcast,b)\n",
        "  # print(K.shape(r))\n",
        "  r=K.expand_dims(r)\n",
        "  g=K.expand_dims(g)\n",
        "  b=K.expand_dims(b)\n",
        "  img=K.concatenate([r,g,b],axis=-1)\n",
        "  # print(K.shape(img))\n",
        "  # img_rgb_filter=tf.keras.preprocessing.image.array_to_img(img)\n",
        "  return img'''\n",
        "  return tf.expand_dims(max_c,axis=-1)\n",
        "\n",
        "def light_mask_loss(input_img,pred_img,true_img):\n",
        "  pred_img=tf.cast(pred_img,tf.uint8)\n",
        "  true_img=tf.cast(true_img,tf.uint8)\n",
        "  input_img=tf.cast(input_img,tf.uint8)\n",
        "  m_i=max_rgb_filter(input_img)\n",
        "  m_t=max_rgb_filter(true_img)\n",
        "  # m_t=m_t+K.constant(0.001,shape=m_t.shape,dtype=m_t.dtype)\n",
        "  m_div_it=tf.divide(m_i,m_t)\n",
        "  m_div_it=tf.cast(m_div_it,tf.uint8)\n",
        "  light_mask=tf.subtract(pred_img,m_div_it)\n",
        "  light_mask=tf.cast(light_mask,tf.float32)\n",
        "  lightmask_loss=tf.norm(light_mask)\n",
        "  return lightmask_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5PwX_w2om0WI"
      },
      "source": [
        "a1=max_rgb_filter(tf.expand_dims(img_low[1],axis=0))\n",
        "a2=max_rgb_filter(tf.expand_dims(img_high[1],axis=0))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mK_-n1VioNDU",
        "outputId": "44fe4fea-dbd6-4701-a4e3-ce6b558fdff0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "b=a1/a2\n",
        "img_low[0]-b"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(1, 96, 96, 3), dtype=float64, numpy=\n",
              "array([[[[-0.00823529, -0.00039216, -0.01215686],\n",
              "         [-0.00028011, -0.00420168, -0.00812325],\n",
              "         [ 0.00571185, -0.00213129, -0.00997442],\n",
              "         ...,\n",
              "         [ 0.0246646 ,  0.03250774,  0.02858617],\n",
              "         [ 0.02787493,  0.03571807,  0.02787493],\n",
              "         [ 0.03443329,  0.03051172,  0.03051172]],\n",
              "\n",
              "        [[-0.00951189, -0.00166875, -0.01343346],\n",
              "         [-0.01096257, -0.00311943, -0.01096257],\n",
              "         [-0.00472189,  0.00704282, -0.00472189],\n",
              "         ...,\n",
              "         [ 0.0302521 ,  0.0302521 ,  0.0302521 ],\n",
              "         [ 0.00361991,  0.00361991,  0.01538462],\n",
              "         [ 0.03051172,  0.03443329,  0.03051172]],\n",
              "\n",
              "        [[-0.02039216, -0.02039216, -0.03215686],\n",
              "         [ 0.01176471,  0.02352941,  0.01568627],\n",
              "         [ 0.01176471,  0.02745098,  0.01568627],\n",
              "         ...,\n",
              "         [-0.02009804, -0.01617647, -0.02009804],\n",
              "         [-0.0022409 ,  0.00560224,  0.00560224],\n",
              "         [ 0.03963964,  0.03571807,  0.0317965 ]],\n",
              "\n",
              "        ...,\n",
              "\n",
              "        [[-0.03349405, -0.03349405, -0.03741562],\n",
              "         [-0.00249554,  0.00142602, -0.00249554],\n",
              "         [-0.01879648, -0.01095335, -0.01879648],\n",
              "         ...,\n",
              "         [ 0.03636364,  0.05204991,  0.04812834],\n",
              "         [ 0.02925104,  0.03709418,  0.03317261],\n",
              "         [ 0.05756065,  0.06540379,  0.06540379]],\n",
              "\n",
              "        [[-0.02663962, -0.02271805, -0.02663962],\n",
              "         [-0.04543611, -0.04543611, -0.05720081],\n",
              "         [ 0.00236646,  0.00628803,  0.00628803],\n",
              "         ...,\n",
              "         [ 0.0276348 ,  0.03547794,  0.04724265],\n",
              "         [ 0.02532948,  0.03709418,  0.04885889],\n",
              "         [ 0.04617331,  0.05401645,  0.06185958]],\n",
              "\n",
              "        [[-0.02494375, -0.01710061, -0.02494375],\n",
              "         [-0.02271805, -0.01095335, -0.01879648],\n",
              "         [-0.02002801, -0.01610644, -0.01218487],\n",
              "         ...,\n",
              "         [ 0.04374057,  0.05550528,  0.06726998],\n",
              "         [ 0.03155637,  0.03939951,  0.05508578],\n",
              "         [ 0.01101774,  0.02670401,  0.03062558]]]])>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1706
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bl3kWicqVkwl"
      },
      "source": [
        "def custom_loss_wrapper(input_tensor):\n",
        "  def custom_loss(y_true,y_pred):\n",
        "    # lm_loss=light_mask_loss(input_img=input_tensor,pred_img=y_pred,true_img=y_true)\n",
        "    # print(lm_loss)\n",
        "    e_loss=enhancement_loss(y_true,y_pred)\n",
        "    c_loss=color_loss(y_true,y_pred)\n",
        "    s_loss=squareSobelLoss(y_true,y_pred)\n",
        "    total_loss=e_loss+s_loss*0.2+0.2*c_loss\n",
        "    # total_loss=total_loss+(10*lm_loss)\n",
        "    return total_loss\n",
        "  return custom_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "15nBh2bV3Ecy"
      },
      "source": [
        "# Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W-1aYMZU2H-D"
      },
      "source": [
        "model=keras.models.Model(inputs=[input_layer_1],outputs=final_output)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oufR4Epz3R4F",
        "outputId": "406b1bf3-342e-4a6b-d0b7-32b85ecee0f1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_103\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_26 (InputLayer)           [(None, 96, 96, 3)]  0                                            \n",
            "__________________________________________________________________________________________________\n",
            "lambda_300 (Lambda)             (None, 60, 60, 3)    0           input_26[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_777 (Conv2D)             (None, 60, 60, 64)   1792        lambda_300[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_778 (Conv2D)             (None, 60, 60, 64)   36928       conv2d_777[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_775 (Conv2D)             (None, 96, 96, 64)   1792        input_26[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_779 (Conv2D)             (None, 60, 60, 64)   36928       conv2d_778[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_776 (Conv2D)             (None, 96, 96, 64)   36928       conv2d_775[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "resizing_25 (Resizing)          (None, 96, 96, 64)   0           conv2d_779[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_200 (Concatenate)   (None, 96, 96, 128)  0           conv2d_776[0][0]                 \n",
            "                                                                 resizing_25[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_780 (Conv2D)             (None, 96, 96, 32)   200736      concatenate_200[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_781 (Conv2D)             (None, 96, 96, 8)    264         conv2d_780[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_782 (Conv2D)             (None, 96, 96, 1)    201         conv2d_781[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_201 (Concatenate)   (None, 96, 96, 4)    0           input_26[0][0]                   \n",
            "                                                                 conv2d_782[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_783 (Conv2D)             (None, 96, 96, 32)   1184        concatenate_201[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_784 (Conv2D)             (None, 96, 96, 32)   9248        conv2d_783[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling2d_125 (Globa (None, 32)           0           conv2d_784[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling2d_125 (G (None, 32)           0           conv2d_784[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "dense_250 (Dense)               (None, 8)            264         global_max_pooling2d_125[0][0]   \n",
            "                                                                 global_average_pooling2d_125[0][0\n",
            "__________________________________________________________________________________________________\n",
            "dense_251 (Dense)               (None, 32)           288         dense_250[0][0]                  \n",
            "                                                                 dense_250[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "add_125 (Add)                   (None, 32)           0           dense_251[0][0]                  \n",
            "                                                                 dense_251[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "multiply_250 (Multiply)         (None, 96, 96, 32)   0           conv2d_784[0][0]                 \n",
            "                                                                 add_125[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "lambda_302 (Lambda)             (None, 96, 96, 1)    0           multiply_250[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lambda_301 (Lambda)             (None, 96, 96, 1)    0           multiply_250[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_202 (Concatenate)   (None, 96, 96, 2)    0           lambda_302[0][0]                 \n",
            "                                                                 lambda_301[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_785 (Conv2D)             (None, 96, 96, 1)    99          concatenate_202[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "multiply_251 (Multiply)         (None, 96, 96, 32)   0           multiply_250[0][0]               \n",
            "                                                                 conv2d_785[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_125 (BatchN (None, 96, 96, 32)   128         multiply_251[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_125 (Activation)     (None, 96, 96, 32)   0           batch_normalization_125[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_786 (Conv2D)             (None, 96, 96, 32)   9248        activation_125[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_787 (Conv2D)             (None, 96, 96, 32)   9248        conv2d_786[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling2d_126 (Globa (None, 32)           0           conv2d_787[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling2d_126 (G (None, 32)           0           conv2d_787[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "dense_252 (Dense)               (None, 8)            264         global_max_pooling2d_126[0][0]   \n",
            "                                                                 global_average_pooling2d_126[0][0\n",
            "__________________________________________________________________________________________________\n",
            "dense_253 (Dense)               (None, 32)           288         dense_252[0][0]                  \n",
            "                                                                 dense_252[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "add_126 (Add)                   (None, 32)           0           dense_253[0][0]                  \n",
            "                                                                 dense_253[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "multiply_252 (Multiply)         (None, 96, 96, 32)   0           conv2d_787[0][0]                 \n",
            "                                                                 add_126[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "lambda_304 (Lambda)             (None, 96, 96, 1)    0           multiply_252[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lambda_303 (Lambda)             (None, 96, 96, 1)    0           multiply_252[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_203 (Concatenate)   (None, 96, 96, 2)    0           lambda_304[0][0]                 \n",
            "                                                                 lambda_303[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_788 (Conv2D)             (None, 96, 96, 1)    99          concatenate_203[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "multiply_253 (Multiply)         (None, 96, 96, 32)   0           multiply_252[0][0]               \n",
            "                                                                 conv2d_788[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_126 (BatchN (None, 96, 96, 32)   128         multiply_253[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_126 (Activation)     (None, 96, 96, 32)   0           batch_normalization_126[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_789 (Conv2D)             (None, 96, 96, 32)   9248        activation_126[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_790 (Conv2D)             (None, 96, 96, 32)   9248        conv2d_789[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling2d_127 (Globa (None, 32)           0           conv2d_790[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling2d_127 (G (None, 32)           0           conv2d_790[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "dense_254 (Dense)               (None, 8)            264         global_max_pooling2d_127[0][0]   \n",
            "                                                                 global_average_pooling2d_127[0][0\n",
            "__________________________________________________________________________________________________\n",
            "dense_255 (Dense)               (None, 32)           288         dense_254[0][0]                  \n",
            "                                                                 dense_254[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "add_127 (Add)                   (None, 32)           0           dense_255[0][0]                  \n",
            "                                                                 dense_255[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "multiply_254 (Multiply)         (None, 96, 96, 32)   0           conv2d_790[0][0]                 \n",
            "                                                                 add_127[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "lambda_306 (Lambda)             (None, 96, 96, 1)    0           multiply_254[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lambda_305 (Lambda)             (None, 96, 96, 1)    0           multiply_254[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_204 (Concatenate)   (None, 96, 96, 2)    0           lambda_306[0][0]                 \n",
            "                                                                 lambda_305[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_791 (Conv2D)             (None, 96, 96, 1)    99          concatenate_204[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "multiply_255 (Multiply)         (None, 96, 96, 32)   0           multiply_254[0][0]               \n",
            "                                                                 conv2d_791[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_127 (BatchN (None, 96, 96, 32)   128         multiply_255[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_127 (Activation)     (None, 96, 96, 32)   0           batch_normalization_127[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_792 (Conv2D)             (None, 96, 96, 32)   9248        activation_127[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_793 (Conv2D)             (None, 96, 96, 32)   9248        conv2d_792[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling2d_128 (Globa (None, 32)           0           conv2d_793[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling2d_128 (G (None, 32)           0           conv2d_793[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "dense_256 (Dense)               (None, 8)            264         global_max_pooling2d_128[0][0]   \n",
            "                                                                 global_average_pooling2d_128[0][0\n",
            "__________________________________________________________________________________________________\n",
            "dense_257 (Dense)               (None, 32)           288         dense_256[0][0]                  \n",
            "                                                                 dense_256[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "add_128 (Add)                   (None, 32)           0           dense_257[0][0]                  \n",
            "                                                                 dense_257[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "multiply_256 (Multiply)         (None, 96, 96, 32)   0           conv2d_793[0][0]                 \n",
            "                                                                 add_128[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "lambda_308 (Lambda)             (None, 96, 96, 1)    0           multiply_256[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lambda_307 (Lambda)             (None, 96, 96, 1)    0           multiply_256[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_205 (Concatenate)   (None, 96, 96, 2)    0           lambda_308[0][0]                 \n",
            "                                                                 lambda_307[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_794 (Conv2D)             (None, 96, 96, 1)    99          concatenate_205[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "multiply_257 (Multiply)         (None, 96, 96, 32)   0           multiply_256[0][0]               \n",
            "                                                                 conv2d_794[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_128 (BatchN (None, 96, 96, 32)   128         multiply_257[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_128 (Activation)     (None, 96, 96, 32)   0           batch_normalization_128[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_795 (Conv2D)             (None, 96, 96, 32)   9248        activation_128[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_796 (Conv2D)             (None, 96, 96, 32)   9248        conv2d_795[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling2d_129 (Globa (None, 32)           0           conv2d_796[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling2d_129 (G (None, 32)           0           conv2d_796[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "dense_258 (Dense)               (None, 8)            264         global_max_pooling2d_129[0][0]   \n",
            "                                                                 global_average_pooling2d_129[0][0\n",
            "__________________________________________________________________________________________________\n",
            "dense_259 (Dense)               (None, 32)           288         dense_258[0][0]                  \n",
            "                                                                 dense_258[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "add_129 (Add)                   (None, 32)           0           dense_259[0][0]                  \n",
            "                                                                 dense_259[1][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "multiply_258 (Multiply)         (None, 96, 96, 32)   0           conv2d_796[0][0]                 \n",
            "                                                                 add_129[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "lambda_310 (Lambda)             (None, 96, 96, 1)    0           multiply_258[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "lambda_309 (Lambda)             (None, 96, 96, 1)    0           multiply_258[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_206 (Concatenate)   (None, 96, 96, 2)    0           lambda_310[0][0]                 \n",
            "                                                                 lambda_309[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_797 (Conv2D)             (None, 96, 96, 1)    99          concatenate_206[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "multiply_259 (Multiply)         (None, 96, 96, 32)   0           multiply_258[0][0]               \n",
            "                                                                 conv2d_797[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_129 (BatchN (None, 96, 96, 32)   128         multiply_259[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_129 (Activation)     (None, 96, 96, 32)   0           batch_normalization_129[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_207 (Concatenate)   (None, 96, 96, 160)  0           activation_125[0][0]             \n",
            "                                                                 activation_126[0][0]             \n",
            "                                                                 activation_127[0][0]             \n",
            "                                                                 activation_128[0][0]             \n",
            "                                                                 activation_129[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_798 (Conv2D)             (None, 96, 96, 32)   46112       concatenate_207[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_799 (Conv2D)             (None, 96, 96, 8)    2312        conv2d_798[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_800 (Conv2D)             (None, 96, 96, 1)    73          conv2d_799[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "lambda_311 (Lambda)             (None, 96, 96, 3)    0           input_26[0][0]                   \n",
            "                                                                 conv2d_800[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_801 (Conv2D)             (None, 96, 96, 32)   896         lambda_311[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_802 (Conv2D)             (None, 96, 96, 32)   9248        conv2d_801[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_803 (Conv2D)             (None, 96, 96, 16)   4624        conv2d_802[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_804 (Conv2D)             (None, 96, 96, 16)   2320        conv2d_803[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_805 (Conv2D)             (None, 96, 96, 3)    435         conv2d_804[0][0]                 \n",
            "==================================================================================================\n",
            "Total params: 469,900\n",
            "Trainable params: 469,580\n",
            "Non-trainable params: 320\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fTl_O4_A-QM-"
      },
      "source": [
        "# Plot a DL Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LTFNjv_L3ykC"
      },
      "source": [
        "# keras.utils.plot_model(model,show_shapes=True,show_layer_names=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7M1EsQJ6WME3"
      },
      "source": [
        "# Model Compile"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mZxjQMDUPItf"
      },
      "source": [
        "opt=tf.optimizers.Adam()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t_CWJZchR_Bb"
      },
      "source": [
        "EPOCHS=3\n",
        "BATCH=28"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "493B_EW-AmjN"
      },
      "source": [
        "\n",
        "\n",
        "import os"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DhEfUNb7B9PP"
      },
      "source": [
        "\n",
        "import random\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NCD2BehXNLGq",
        "outputId": "0677b9a6-77ba-4749-c4a4-deb42968ba93",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "for i in range(EPOCHS):\n",
        "    b=0\n",
        "    for j in range(0,img_high.shape[0],BATCH):\n",
        "        b=b+1\n",
        "        img_inp=img_low[j:j+BATCH]\n",
        "        img_out=img_high[j:j+BATCH]\n",
        "        with tf.GradientTape() as tape:\n",
        "          img_pred=model([img_inp])\n",
        "          lm_loss=light_mask_loss(input_img=img_inp,pred_img=img_pred,true_img=img_out)\n",
        "          e_loss=enhancement_loss(img_out,img_pred)\n",
        "          c_loss=color_loss(img_out,img_pred)\n",
        "          s_loss=MeanGradientError(img_out,img_pred)\n",
        "          total_loss=e_loss*4+s_loss*0.25+c_loss*1+lm_loss*5\n",
        "          # according to paper:- total_loss=e_loss*1+s_loss*0.2+c_loss*1+lm_loss*10\n",
        "          mse=tf.losses.mse(img_out,img_pred).numpy().sum()\n",
        "        # os.system('cls')\n",
        "          print(i,' ',b,' ',total_loss.numpy(),' ',mse)\n",
        "          grads = tape.gradient(total_loss, model.trainable_variables)\n",
        "          opt.apply_gradients(zip(grads, model.trainable_variables))\n",
        "\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0   1   6277.1177   14669.367\n",
            "0   2   850.4958   14045.191\n",
            "0   3   601.32117   7134.47\n",
            "0   4   573.4203   6543.363\n",
            "0   5   8024.948   9629.088\n",
            "0   6   971.90656   18352.64\n",
            "0   7   588.2557   6685.233\n",
            "0   8   488.02246   4591.542\n",
            "0   9   527.00024   5405.6895\n",
            "0   10   651.6235   8215.294\n",
            "0   11   6081.5444   8839.862\n",
            "0   12   610.3707   7134.1665\n",
            "0   13   406.69748   3198.3662\n",
            "0   14   13278.507   3154.375\n",
            "0   15   642.2758   8157.974\n",
            "0   16   470.00302   4339.923\n",
            "0   17   22370.38   5059.4033\n",
            "0   18   56217.805   8435.213\n",
            "0   19   76726.99   5891.633\n",
            "0   20   18899.861   6060.003\n",
            "0   21   89373.44   7048.701\n",
            "0   22   43993.81   4754.746\n",
            "0   23   50015.24   5655.733\n",
            "0   24   61292.38   5613.078\n",
            "0   25   72493.68   5454.119\n",
            "0   26   26082.367   5308.2295\n",
            "0   27   119296.78   5666.412\n",
            "0   28   43070.484   4488.017\n",
            "0   29   32570.363   4741.201\n",
            "0   30   46373.836   4348.2324\n",
            "0   31   67532.79   3757.825\n",
            "0   32   57084.906   4478.6265\n",
            "0   33   54186.145   4873.2104\n",
            "0   34   136296.84   5036.11\n",
            "0   35   85055.32   3857.1553\n",
            "0   36   188670.31   4509.6387\n",
            "0   37   80287.42   5145.626\n",
            "0   38   61183.598   3521.0273\n",
            "0   39   63802.6   3950.7441\n",
            "0   40   117816.336   3456.578\n",
            "0   41   106063.04   3636.8784\n",
            "0   42   55446.207   3383.3682\n",
            "0   43   13939.109   4976.679\n",
            "0   44   43048.418   4069.371\n",
            "0   45   13498.454   3640.9175\n",
            "0   46   22088.205   3936.009\n",
            "0   47   58081.875   3304.5073\n",
            "0   48   11072.508   4511.7437\n",
            "0   49   98138.18   4786.473\n",
            "0   50   70357.45   4665.5654\n",
            "0   51   122105.75   3536.7485\n",
            "0   52   85435.02   3457.2576\n",
            "0   53   61393.48   4389.1016\n",
            "0   54   32.716454   10.083247\n",
            "1   1   6265.563   14286.926\n",
            "1   2   843.2288   13806.883\n",
            "1   3   601.8191   7143.551\n",
            "1   4   577.3823   6631.4297\n",
            "1   5   8020.866   9513.845\n",
            "1   6   958.603   17845.7\n",
            "1   7   580.64166   6500.725\n",
            "1   8   481.48303   4459.1772\n",
            "1   9   525.4949   5367.2163\n",
            "1   10   647.28986   8099.3203\n",
            "1   11   6079.3   8776.227\n",
            "1   12   612.61334   7180.602\n",
            "1   13   409.01   3232.6235\n",
            "1   14   13274.247   3085.7627\n",
            "1   15   662.18176   8672.721\n",
            "1   16   484.37866   4606.434\n",
            "1   17   22360.982   4879.2036\n",
            "1   18   56211.55   8276.337\n",
            "1   19   76731.43   5985.5947\n",
            "1   20   18902.31   6115.458\n",
            "1   21   89375.56   7097.063\n",
            "1   22   43996.227   4800.2334\n",
            "1   23   50017.6   5703.828\n",
            "1   24   61291.59   5594.3037\n",
            "1   25   72490.91   5396.112\n",
            "1   26   26080.191   5263.5674\n",
            "1   27   119296.58   5664.5747\n",
            "1   28   43070.074   4484.042\n",
            "1   29   32569.127   4722.764\n",
            "1   30   46373.734   4349.0483\n",
            "1   31   67532.21   3748.1726\n",
            "1   32   57086.086   4505.7983\n",
            "1   33   54183.008   4814.8896\n",
            "1   34   136291.06   4921.531\n",
            "1   35   85048.836   3747.5608\n",
            "1   36   188663.48   4384.5933\n",
            "1   37   80282.82   5056.5234\n",
            "1   38   61175.164   3380.0723\n",
            "1   39   63797.543   3860.2112\n",
            "1   40   117813.69   3409.0867\n",
            "1   41   106062.03   3613.7444\n",
            "1   42   55440.12   3280.0051\n",
            "1   43   13936.101   4916.3223\n",
            "1   44   43046.152   4029.914\n",
            "1   45   13492.02   3531.0474\n",
            "1   46   22077.324   3744.7708\n",
            "1   47   58079.617   3267.641\n",
            "1   48   11072.638   4515.195\n",
            "1   49   98135.3   4732.587\n",
            "1   50   70355.89   4639.357\n",
            "1   51   122105.984   3544.8145\n",
            "1   52   85428.83   3361.3826\n",
            "1   53   61390.383   4334.3867\n",
            "1   54   34.02889   11.317509\n",
            "2   1   6258.705   14081.169\n",
            "2   2   836.36554   13603.176\n",
            "2   3   601.12225   7135.259\n",
            "2   4   579.4511   6683.0703\n",
            "2   5   8020.4487   9506.449\n",
            "2   6   952.56396   17628.023\n",
            "2   7   581.879   6528.741\n",
            "2   8   483.17462   4491.808\n",
            "2   9   522.6877   5315.4263\n",
            "2   10   642.2065   7984.391\n",
            "2   11   6075.7754   8689.546\n",
            "2   12   607.7657   7080.792\n",
            "2   13   405.46045   3180.5417\n",
            "2   14   13267.679   2988.6611\n",
            "2   15   665.88885   8774.052\n",
            "2   16   483.37418   4593.398\n",
            "2   17   22357.45   4814.495\n",
            "2   18   56207.65   8182.6904\n",
            "2   19   76731.3   5987.1177\n",
            "2   20   18899.166   6048.0493\n",
            "2   21   89375.24   7094.932\n",
            "2   22   43995.438   4783.619\n",
            "2   23   50018.26   5720.1177\n",
            "2   24   61287.047   5497.4395\n",
            "2   25   72486.07   5295.1245\n",
            "2   26   26075.424   5166.9053\n",
            "2   27   119295.09   5627.266\n",
            "2   28   43063.566   4360.5146\n",
            "2   29   32567.637   4693.433\n",
            "2   30   46371.074   4294.86\n",
            "2   31   67533.22   3761.3066\n",
            "2   32   57084.766   4472.3975\n",
            "2   33   54180.19   4759.125\n",
            "2   34   136290.72   4912.337\n",
            "2   35   85048.16   3733.166\n",
            "2   36   188662.4   4362.799\n",
            "2   37   80281.99   5038.8457\n",
            "2   38   61174.59   3372.8027\n",
            "2   39   63796.125   3838.8223\n",
            "2   40   117813.67   3412.9644\n",
            "2   41   106060.0   3585.921\n",
            "2   42   55439.168   3270.018\n",
            "2   43   13935.749   4911.435\n",
            "2   44   43043.82   3990.7979\n",
            "2   45   13487.731   3463.6477\n",
            "2   46   22077.959   3759.1685\n",
            "2   47   58078.16   3245.769\n",
            "2   48   11071.868   4507.29\n",
            "2   49   98130.62   4642.7764\n",
            "2   50   70349.0   4503.7344\n",
            "2   51   122105.23   3528.376\n",
            "2   52   85431.46   3398.6174\n",
            "2   53   61384.79   4229.2954\n",
            "2   54   32.114815   9.542242\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BQyNJedGhVEI"
      },
      "source": [
        "model.save('model_improved.h5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XBMpNtzAnzIH"
      },
      "source": [
        "# Inference"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eI-M34tELINB"
      },
      "source": [
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pimgUvGgRL72",
        "outputId": "70c0ada3-9144-467e-c928-388ae5b040c1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.engine.functional.Functional at 0x7f9bf01f9d68>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gGlTUUzqAh3L"
      },
      "source": [
        "def high_light(index):\n",
        "  img=np.expand_dims(img_low[index],axis=0)\n",
        "  a=model([img])\n",
        "  plt.imshow(img[0])\n",
        "  plt.show()\n",
        "  plt.imshow(a[0])\n",
        "  plt.show()\n",
        "  plt.imshow(img_high[index])\n",
        "  plt.show()\n",
        "\n"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "INEk0URjIIf0",
        "outputId": "09e75cfc-9d0e-4738-ca0b-cfc7beb8f11e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 770
        }
      },
      "source": [
        "high_light(1443)"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD7CAYAAACscuKmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9WbMlSZLn9VNb3P2sd4ktt8qqrLWre5qWBoTpaWEQRBAQXhBegQ8wT3wAPss88MQDHwAZEeAB5oEBYWgGeq+uysqqXCMyIu5ydndblAezc+69EfdGZlVWR5XQYSkn49xz3P24m5maqv71r2qiqrxpb9qb9v//Zn7bN/CmvWlv2utpb4T9TXvT/oG0N8L+pr1p/0DaG2F/0960fyDtjbC/aW/aP5D2RtjftDftH0j7RsIuIv+piPxERH4mIv/Nb+qm3rQ37U37zTf5dePsImKBvwP+Y+BT4F8D/6Wq/vVv7vbetDftTftNNfcNzv33gJ+p6s8BROS/B/5z4E5hbxqno87Xv/TGP6hQrmMQqZ8JyOHs64tS+cJai7Hm5tf1ZM164xwRwRiLiOCsQcQwmkxp2o6y3unhuHIBBQXrHMbaq9+Qco+/i+3rL9zKbYfun/3QBzeunVFVcs5kzQhgMOSc2G7W5BQZTWe03Zgw9Az9FtVMSgkRYTSa4nzD9UHdd7te63upz6E5o5rJOZNSZLW+IMXIaDSlaTr6YcN2uy7jaRpUldBvSTkWe9VATokUE1mVnMqPiSlTJKXyMlJeIrAf1pQgZy2f1XvNWeqUKPNi30fGKMZqmbd4UCGkSM65PqRgreBcOT6jqCopZXLKqMphLK7mvdY+l5vTfj/GInhrEWNIKZFSPtxXzpCzvjyAfDNhfxf45NrfnwL/+MWDROSfAf8MoGs9/+Tf+R5GEiKZnCFlMCqYbDBicK5FjEGM1lfpUNX9ZJYyKUQ4OpkxPZpABk2KYBDvAWHoB3JMiFGMKN63jMdTvPMcz+aMuhE//uM/4d3v/ICsmZQzxhga1yBACgHNytHJPabzY1BFs2KMwXXdQeBvE4x9u034XnX8Xee/eB0pF7px3IuvKkVXx9S/df+XQtZcryd1MTQ3XtdbP2yJcWCzWbLaLHDiGNuO3WbDT/7yz9isF/z+v/2nfOv7P+bp5x/x6c//kqHvWS0XON/yo3/rTzi9/zbWNVjrUCCpklUZQkSBUeNw1jJsN4TthiH0rHcLFosz/uzP/mcWi+f8/o//hHff/QEf/uLP+Yu//Vc0dsS9yXvkmPjkZ3/OcvEceyTYqbC8XPD8yVOGIbFaJnLOdKOM9criEhYLaKwwbw3Ggh0papTlQtluM22T6dpEiIb1ypGTkHJZ9IwxiBG6cWIyTzjp6OQtcrI8uXzOarfBZotJhvHUce9+C0bZxIGYEovzNdtNTxgMoS8KTmyZ88YriBJ7IQ1lxRGBrEpKCWMND05PGHUtZ5dLLhdrclZSVJaLdOd8+ibC/rWaqv5z4J8DzKad9tseMXUlRxBM1dKAKJkI2WBQREH3qxYvalRhu94iKNYYnPMY67AUrd36BrWZlAZSHFAbUU2AxXgPzvHlF5+yWa7o5nPGJ6c462hdhAzb7aZqBcgh1AXG4HzDxHmMk19JcF9nE15SCNe+0Ksj9saV1IWgau7D4XL1jN55rLE8+ewjfvHhnyNi8W6Ecw0Pv/0B3reoCJ/+4qcszh+zWp8jGGazY4y1PPv8p1w8/Yj773yfk4fvF+2WlRB6Ls6ekXJiOpvRtC2r8+eszp/RD2uWm2eA8uMf/bsY65kd36cbz5gfzXl4PKX1U05P3yaGzBePP0P7zDtvPeStd095/PiXMPSsNz3rzQZVYTLKjEZKjkIfwBuFJhEzLM6FlME5ZTSG2TxzdJpYLZXLpSFmGI0yxirDoIQguCzk3BAxbNOKnIUYe9DEfNxwNO4IGrlYXZJUiVnJWdn1mRAMOQvG1CEpqzg5lnmuqmCrhYHBohhJGKtgtmQZaN3AvIUQYaM3TOGX2jcR9s+Ab137+7362d1NlRhTNYWLGWhFwUoxvaSYOUIiZxCUjFYNVMwhKCYWQBgCvQHfeKy1aD1OKCu1GIPmQNSM5gyaKba5AWtZXl6wXS6Zp7ewkzHJeiQbNCubbU8Mga5taa2t2s4WgdB8QDava90Xhf83tRh83evsj6r2D4oiIi9YBrcsBVX+r1sS+/dX5qrDiNJvVzx//AkYi/EjxpMj3vr2D5jPTzl7+gWX589Yry4Yhi3etbRdhyCsLr8kpp7x0T2O9L2ipXImhsB2syCGgPWGLMp6s+Dy8jn9sGSxfkzbjnj3ez9iOj0lipAFRl3HrGtpujGzoyOGIWPaCfg186MHvPvwPVK/4/H4p6SkYHoQpWlg3MKqUbynCJpTcoDt1hADTI+g9Uo3VqbHmZiLFZIUrM/4RhkiJBWSOrI6QMipJydFNYIqXWs4nnoWu8jzyy0pK6oGVYhBSbm6TUYPQ6MIJFNHKCM2AwYpXyJWsEZRCWQizmQ6B6JCb8q439W+ibD/a+AHIvIBRcj/C+C/+qqTjN1rjCLsogZjqh+NkPTKZJeDRi9a3VgHItgqtHtTM8VMvwsYm8m6xRiDNYKR4pD5psV5jxiLirDrAykL0fR4I4z6ORISgkUaMNYwmozRrMyOTpgdHZXfN+UejCk+vKr+Tmj3/T0UH/hl019EbvjGVy7RVbv+91677zW7iJBTJGtmNrvHdz74Q1LKpFSu+PGHfwNiuHfvAaf37jM/mnP/4dtsNwuefP4hIWyJcY0SsU9+yobIbrtiefkMg2XSnOBsy7DZoFnxzYh7b73P+dnnPH78EWGX2GxXWN8ymhzRtGOm03vMjt9hND7m0aNvk5Lywx+e8eDhfT747vd4+71vsRsyp7/8EOIF3qzIklExZCNYn+naYjVmUcTD0b1c+iYrMSjLRSaq0G8t47YhO0FTz7BLiEDTgKgybMpcFRIomOpTb3aBx2nJZhhYbxRFce7auABZC0ZgLRQ4SxliJit4n3EONBmILQDGtIgoEkr/DwOEqIQrg+zO9msLu6pGEfmvgf8RsMB/q6p/9cqTBKy9mkCiAlp8RW9t0ewxk1WLllbFVKEVYzDWI0awJITynQIpKTkGxKTqVwuuKcCasxbftljrEGvJGLZ9Tx8SrUl4UWbbU2SIBcCzRdgnTYcRy/zkHvOT0xt+8nXB+F0R+EO7DjC++Pk1IPK29qLAG2MOn8UcySkxP37AyE8Iw8B2tWS1vODP/t//jeXqgn/yH/wnfOe7P8C5FteMefz5T/mrv/5fWC2fYTyIhc1j5cvlcxbnX/Dks79h2p3yj37wHzGd3GPnPDFGZkenzB6eMISezWKLsT3rzQLXNsyP7jGfzFnNHzA/eY/p9JS33/4AEUOmZ7264O13fsDDB++zW/d8Mv9L8uDx9lNCFfYkgvOBUZeJWdnGArKdHCnWwMWzzGYFwyVcLMCJZdJ1aIbVNhGTYjw0LWhU+nUuvrYri4AxDi+W9W7gbDEQk9IPZQhGYy3WRMHZyKqEVCyMzpdxilpM/a5Tug5yb0jbtoCRzqJADFtCjgwBhqjkv09hr5PjXwD/4useL3svXQRjiqCjgrGCccV00Zhvgk1F5hFNYENZIFCMQFaD0YrgFu+/rHoVbLK2gChQtJR3DsSQYiARMdV92Ow2XCzO8H3HWBXfNBwf3adtO5yrSPytCPXdJvwr+uzu/vkabsCvHSotFzxc42Xz/uXr3wAIqzVgjUXbDjGGnCJD7HHOYETZbTdcXlzgm5am7dluVwgG70bMTk5puhHd5B7N6BinSr84Z9zOabspvhlhxSIZQr9jvVpgjOHd93+AMZbZ/B7daIZzDWIMXTvl+OgtRqMZ1nkEYTo5xlpP244BQYwgrgBe3oNPkBOEQcipzEFvFdsWP3w0UkQU58u5miHFgikFjUVAjWKKz0OOBTg3rgi7tVpM8QriZVVihKxS52VBy0VgNFGsVzYrYbUolm5IN/s9RSEMoBE0R7rRhIfvvosxwuPPPma9WpTnaiwZIWfDxcVw5xz4ewfoXmxGDM6UlWwP01lvsZ1Hs6J9KOGdXP1GLaaVMYLVhJiy0hqRGnZzKKWzqFaANeC9xXtXrpkz1giTUUdW5fxiSYiR3DRE53h29pxNirTNiPnsAZPxjAenjzg+OqL1TUHiuRma+l2sA7C/Lynq5Uq77++7HmfM4V3t47qw1k/3aHMZhxJmI2ckK9Z7XNuSY8S1DdlBN/JstnBx9oxffPgh7aihm7SsFs9wpmUyafjud/+U43vv0DQtjW84f/o5U53gXMf8+C28HxVMJWc2i0suLr5kOj3iH//T/wxrLVmLJLTdBDGW4+NHjLs5YgzWNSDw4MF3yJqx1c3CZqRL2FFmPC1IewrCZgCtbmHbRabHPcYq1mc0w3rRsN064g5CL0Qgm12JDHnFOQi7IohNm2nHsWj0aj3tEfucoR8K0m5dUb0plls7fZg5vp95+pkjbC2osu1LFEVFEWDYGcJOEM1I3nH88AF//O//U5z3/Mv/6X/gYnXB0dQwnwgqDjUtj7/c3jk/XruwUwWzvEq0FiCnar7nF8NIlFVUwaLFGqioJeyxpUNA6fBJTpkoxYci62G13U9qVUgpgyZMGLC7HZrAyxJRZbNe0nVdCbV5X8KBYq+e4haBfxHU+lXb1zn/xe9eCstdv68XhPw6gWHPLTgcL4K8IPAAmtPVdao7JcagtqzYxlpG4ymTfo737cFFM8bgfctkegqAb0YY44ghEMLAbrclxogxGecbmrYjhQJwGWMwWJxrGI2nGGMJYXdYhACMdfimK/dY789aj9HMELaE0NMPm+I6OkPXdeScCUMJUVmbEFcE1+CwAo21qAXnig9d+npvOWZEr8KUV7HwMjfl2jDkrDXeXdYvsXUhkCu8RLOgSQ4+fq7nUeP9UnFUKcNU52wmxYAx5RznLI13dI0jKvT51Qro9Qv7wcQWjLWIdaSYuLxYXQk7Rfg17SefKYC9lPMaZ7E1zplyKlELUTQbkho0wW65JiXw1uCNIWdo2i0YgzUOvKXf7YhhzSgmNAkbWbN4/hzvG1JYM5vP+f6P/ohvffAjnPM0rb3xKC/Gon+d9uu4Ar/SNV4U/EouUt1H4pR8TeC5hkGoZoZhAApPwTUNYBAp8eaQE+IbvvPDPyYMO8btlNZ3NOMJ3WxOTpGHb/+AGAeWi2c8f/oJn3z2EY+ffEruN6T1Jffuv8sP//BPOTp+yPLyGf1uw6id45sRTdtWrQBtOwKurBIRQWzxX8k1PFsjCp9+/td8/vnfsF48RSUzGo354Fvfpx96Pn/ylOV6zXgSmEy2hKFls7yH9x0nk0d457gY/4Ld5BmahdjvnURFjODFYsSQ7UBuIllhswLV4hqgkKTAdX1fuRlOadty3/2g5CxcPHFsz6HvpYSdM6R6vnMFyPYm40yxLnZbw+JywV//n/8K5y0ybLl/fMTp5Jij8Zyz9RlfXPyCGOOd8+S3oNnh4L1XLaBUYkXWAyNu7/PA1Yp48PmNXDNFi0ZXzYiWSZtRhiESYwZnMd6RYypsLq3Ivim/kWImhkQYAqCkFHDWcnH2lDjsWL+3JIT+1xLEV5Fqrmvx677xbwLsu/Eb8vJ3UtXG4bcqcy2lRNZUetlUvzTHuihkwFamWcFXQhzIOTGeztE8pbMN3nqa0ZR2NAOgGx8Thh2Ly6dsN0vOz57w+Re/xOWEz4Hp/BTrLL7xGGcQC75p6MZTnHXsWXXGmBeeq1gZRdC5wRPYbC44P/+cOKwP3IjpZErjG5rmErPb4j20bSYnSKHB0GGZ44zH2QZnwdnK/7i2XhbAuGhfYzMpCjGa4t+nKvQCmUIYgxoMlpvRkGEn5N6QAUQPY3IIS0v5bWsgmnJmGAYuzr7EO4uSabzDu47GThBd0g+xyszt7bUKuyqEoEAmJiH1iZR6RPQAluWsB7OFPWpf44wlhgmQMEbpRp6mbQkh0G8D1mqZNKKEEAlDpHUdvnE4Zw+8Be/KAPjjUzKWGHt2w6aY7E2LbRq68YzReM7l+Rk/+8lfcv/BW3zrOz84hN1+k+3vA82/Eozbf0coQosqpi6QXz77JU+efoxvCtuwbTruH72Ns57l5VOGYcd0fo/Z/D7Ly6f81V/8S1IMzI/u07Qj3nr4AdPje1jnsc6S61gNGrl4/kuePvmI5dkXxPWa2b1HfOvbP2Q2nfPs+aes1s/o05aYIydtw9jew1R+w3Vyz6Hp4aFAhJwj54tn7Po1u+0aq4J1M5rpETGsSOmnxLzEmiWOJamHjU7pt4ZhuyaFHc8vAs5b+t0Ck6FrM3KqoAZNvsxLG4CM2yXSUM1xNQUgNq5o9pgK5lRoH6QM2001rqQKvgqJEnrbj1HjTPEHUiZlpU9CMFSQL5JIBCJZDZutJ8RixZLWrLcLctSrfrmlvV5hB2ICRElGGUJiGCLeC5OJPQh0rqZ8VSMHKyDn0hdBC/rZjgXnixsQa8gu5YwaQ4yRGCKq4NwVh17QChAamm6G+DHrxXM26wusdzg3RnyDb8c07Zj1asl6s8R5z3v6/Zee6UBiqAvz/kH1jl6/C9j7xgJ//bIHhf2Ka8oV/UKMQZJycfmUX376N4zGE45P7jMdzbk/f4QRw3a9YLU6w/mG+dEDtptLPv7Fn5PiwLvv/x7T2SliP6CbzK5+QovJimbWy6dcnn3Mdrkg7XraZsKj938P7yzL5TOWywStASdM9SFiCvfbvCjoe1jmhcfVnFltLlmtLxn6LRbBmBHWzxmsw2wDYrdY2WBlSw4NuzBi6COx35GislhvsA7ikBBVfANuVASawZJRArFEchLYXAA31YI9GSlCrzmTa1SpCLuSh6KtfVuiRkr16alWq4Ct/PlQODnkakWlqCiJTCbJloywGkb0vaczkUZW7Iah8v/vHvPXb8ZrIobytiC9GcEcVsF9wsU+CcA6aLwtsXNXwTxNqGZyisQQSDkiBqw1NL7BWsPgQonZ50zfB1QF1wwYI4RQdfyQwSzZbTfEEMlZMbJGY2Kz2WJsQ9N4nPekpMSUEclFOF56rutvb8ar9+b5qxNNvpkJf/jNg8l5+7VuC++Zit53zYh5d4RRYbhcshsyi9kXtM2IIaxRTeQUiMMO7xvefus7hGHAakvcZmIo/X0jxAc457n/6AOMddx/S+gHGE+P6WyDaCovyXSj+/huxqg9unPK5lTi/SH09P22goSOftjw7IsPOT9/zPLiU9arJ8RoiIMhpjWb3TkhbNgFJWHJSci50lddycEIQUqCjGTEFw69aWHYKosLS0oQhRLmigJZUK08TwWVVL37opVKxKlocmty9UgrKGcLTTalElpDKlN0D9BRMCojpmInmWwBo8Wfx6BqiCmxGiJDBtdMEFncOUdes7CXFSoMSkolLmlN0QAHYK6G3ciFP2tEaFtbTeziQPV9yfSJMTD0QkoZY8BZw6jtsNbSNwMpluyj7aYnJcU0xQfcr7pJL0mUlTtnQSSShkBsehbHK7J4xpMJI+MJUYkxImLw+9H4qqetPuRtwv4qwf+1e3dPc9WbPu5d7RCqqxp00kw5HT9gt12yev4Ubdecd56m7chaOWJxYOjXtL7lg+/8Af12y7Mn5wzrRBxKQtHhuapmc67hnW/9AacPv8Nkco/x+Jjl5TlfPv6UFLdkDRgRjiZvMz56pxCsuEK+rz9HDAMpDGw2C1bLC8Q5/HjCbrvks4/+nC8//5BNOKePCzarLRfnSyBjXHEfd4OS8fRJGPoS5tKmYENDX4XRZWgjrlNGYyUlw+VSGPrijyOGxtmSyaZK0n3WWU0wkoyKHpic1iq+hvXSYCjcEnAedABixY8o4Jp1JUblECyGpBD7jHM1UcYKjVgMjhC2XOQtmDF+dISY9Z3j/XqFXTnEzwuoBnWhuwYaC7IPEdfP9vRNzXIFYJhiN6dCosf5YqqHEMgpVXJEQe2N2ft1Ui2E8tgawyHklzN1cCzGOKS+EMPe/sh5b4bXG/9VHv06Kn5LuOz6Md8ElT9YCNfCaF83lNd2Y2bzE9q2o6l+d9fN8Y0HcSCWph1jrcc3HZPpCd5PiKEhxkTXjQ69UsDPwG6zKGmnYmiaMc63GFfOH49nhEFYR1cFpoQ3Rcrio5oJYUBEsLaMWUyBIezYbVdslmeFl7W27Polm905Q1qhOSBqKmgXQahunAUjaAJrLK1zhCzswlDmS1MYcIqAFvdwt830OyHGkiar9ioYV8Bexdnie1+x2GqsXMqcMjVuLkA8+Oh6mOuFCSoHwo3Yot01V5DaKK4B3xha2+CskLxgkuJaj2sMmBG4ER/vU75vaa8ZoFNizIX/jtb4qClgRdKD32JFSEPxh1LK9H3E2dppBrwxYA0J6EOkbT2TcQdZWS0XoNC0nsmkRTAYLYkvQzA455gfj3HOsFwt2O42pJyIMRRksxnTdmOadoprJmA9SSxRIaSMpLLC2l8HnYcr9JgrQbstpfTX7+NcF9F9PPrrAYqCcHL6iOnkqKJ3xddOsQe0CHlNTzXW49sxo8kJqpm33yvWi2+aSoApvIbtZsEnP/9/yDlx/+0fMp6eYl1DztCOpjx6t2OzuWSxeU4IPVFLvFmMYK2pKbVrjEiNtxtW2wXrzSWXTz/j7NMP6cOay+1T+rDh2cVP2aYFXkd0jAkm4GzRhN2kAxF2qUejMvIj2mbGYr3mbLlDbKSd9Lgmk/uOHDq228B26NlthGFXaNleKPnrlPwM54TRyJASDLsi9KECC3sauBWlRtdIsVBkFcBknHWYxpFSYtfHYhF0gvUQQ2aIGVph6g2TkeW0G+GtoZlnhlHm3tGck/kMsQ3ix/zd33x65xi/dp89X4MfZa+m98quTrI96WY///dhOFUt4bUK12sNcRxCV0DOqVBscXXC1m/2ZIa60OzDOdY5TNLD7+4vlPXK/7TWIWKvkFO9/pt6E6R7ob1Evrmm2a/76S++3597V/sqBt+LlNi7rlX86vKdsx7T1jwE5woJZbdGNeObks7KnmhSfWWtQNYLFy2aPUV22wUpRULoS+jTaM3bNljb4HyLc8VNQEoxjBgzg0ZSHAihL0ITA9YY8gHTyWhOhbLbbxnChpQjmeIG5lTG3BqD2JICXUBJC2po2o5JO6dPBW1Ha6jLgckOlYY+KGGIpLjP6SiJLMYUjb53mfahskMks05oodRSKH1Shr1YtHUOix40e6oJmbX7qhVZjjey9/Gv8knKglgsFjnITH6lvflahT2r0lf03dp9GmoV9nqXFafDOYPYwvTKqiW/3QAo210sfrotx/Q5I9UyMJX8EIZEDHoVs3QJ7ywpBRYXG6wR2umMo+k9tptt8XVU6PsVMQ1cPPuMfrPg3fd/wOn8EeNuRsyKSZmYE0Yhx1IQwTmHde6KXXWtHXxn2Rt/HITh1SmlX1/T6x3XE6nI7x14wf74rBUWNoIx/hDDtiLIaAKqJePwlnu6jdGnmokpMgwbLi4/Iww7shvTXZ5z/8F7nJy+hVJCUtaPePjuj0kpgnVs+hUXq0uG9UUhUNX05djvsLa8n09PsSHhorLdLaDxbPslq2HDEBzrzZZhfYaYxGg0xbqG0egUVcFKRrPj0Xs/4nvf/0M+/vhnPDn/AgSOZy2jscE1J1g34fOP15w/X+Jt4u13B8RkskZUM8uFsNns+SJUgFkquGYwStX+PUmVPhYBTuRaSaeY51Srk6xkU5RaWpWFw/my+BSRzwxp4NllsXJxczAtn19s+PjssuabGLb97s558trj7CllvLfsEwMAqMSagwNfSR0OU8IW+2WOsiiklAkhY1GcGHKEQMQYofGFi5xqXMMYLQuJMThNKDD0ASPQzec0XUeMGeeGUgIpxJLJtFsV1pTCqB3jXFPSEff3IxBjOJRdepW5fF1wDwIPN4STa5+9KIzX21102buSWr6KoXfbQnOdbGMqseWudtu97rV6iD3DsKbvN6zXlwS1zI8f7I8qI24c48kJSTO73YoQSwLN6uI5TeOZTsc467HicTbRjSc0viG2Y0bjGSrQbqYkzRjTAQ0xbtjuetrW0nTVerAjshqEFvLAaHLM6aN3uFie0TTFlWwbS9cY/LTFtSN8G8l5h3PCeBowFmIqVsNqLcRUkfYrLLK8tIKLKEgF5vb0Weq8l6LZVfSg5fcWbgp6APEse4OxFr0YItZavBHEWVZ9ZL1bIyhWlJh+i5VqXm6Fm1xsbkVcfS+H6hUFhKvc65wL8r4nJ4iAcQZvrph0RcMX9l3TNlWbFqAnx0iKAVTJteiYdxZnhDQM7JZLQh/RLFjb0k6Oa7w0MgwrVDK2belT4suzZzhrWLYNgrJZXBD6HfcfPOTe/YeHMFaxJqrPbF4WlBe1920aPtVBexWq/qJG37sr192MrIqoIncsRr/KwvJ129OnH/HhR/+GfrdilwXclCyerPZgwqlevZBSmsxajwJdO4HJEd5b2qbF+4bZ/ATnGjarC9aLZ6wWz7k8/4L15pLHTz9iu1tx9vxL1psVrR9z/9EJxhmcLyWw+qFEU+KQyDHx2ScfMsQNm/UzHjzsUCxoZLuFJ88WbPsVq8uAuICaRB8TNufqhlTqtpSQsDW2CHKqLqKt4UynqBMab5nRklLmcrMjpkwcKH2hGWVAyIwmBW8ZdiU0SIY4gPdK2ypOoDC2lc1mTdCenCKuZoJqJaTd1V67sBfzLaNqEHIxPypItz9A9mukGCAVLZ2LVhUjWGewh+OqK2AVceAaj7O2+tmGsAONsXKPE2Cw1uOsIYeBISVilLqSekazE0SEYfWMGHYlXuobhpRYXJ5jjWHUdIhmLp5+Qb8pRRXmx/cOqbsigjN7YbldYA5U4Tu0+3Vt+1XCfiU5txxTcYfb4vg3fPqvMXZfN9vv7Pwz/vYn/ysinvH4bbwfk8WjpbDSlW9a/y/1BqxxoELXjLHjUM14i/cNk+kRznkW549Znn/J5fIpzy++YL254OnZxzW99ozdrmfy8KS4CmLIYglhYHNxVinUCU2JLx9/wtPnv2QyNjx40KLqWa139LvE48/WPD/rsa5wO9QoIWaSVTO5szsAACAASURBVFpTi1QaKXiA1ApGKGikJOtUH9pJEXbbcOzHxJhY96FU6BmEXOPrYiKuKWG+AkQZ4j7mH8C7TOsVK+BFyKpsdxu2QWispbGmRAOuRWBua78Vzb7PdCv+OJDBZi29WIU+Vcpgrnx5ESHEVDrZXk28GpmoVW+UGAI5JxpqGSsjWF/SYLXG87NmIiAF3UNsS9uOMM5TkpQLM0+cIcSe1eqM3bBltbnEGMd4NC+aO2a8axGxBWCx4ORacY6voRlfFKAXhel6nP56MYn9ufsehZsCuwcxb7vmraDgS2P0Mrh3629fW6z6fkUYdsRhTeOKCzAZT3B+wmQ8pWkmOOf3xsfVtcoFi/AASbRk21mH9R3iHLthi4k9qlqy4UZzjjXTjqZgDNt+QzYj1ps1ru0YYihul0LOA9g1xg+0I4/KlCFs6Ic13ln6bdGMw5AYhlSILtkwasccHU1BeyRfoloUQ8yKMUrT1aKpCRBDN+pKCC3VFFfJhT9iMtH0RJMRUzkhvriYMVXGHFTSWAkRlyhVARtDgM2muAx9HdccC4NPBYLuQ3gZeYW4/5YSYUrb+8CQC4lGqVoIQozEnHHO4BpTJ1OJuY5GHufksBhYrWCSwq7fYaSa8XsktmtJSQlDAjKBXJH14tePJmMm86MyQGlAgaZrsdbQhzXPn33CcnHG8yefYqxnMiv51w8fPGI6niLiiakIjvWmuhe3sOzuaLcJ/Isa31r7kk99WEz2K/oela1hioL8XjPxuUKMr//u1aJ0c4H4umZ8IQ8lVstnrBZP6bdnjBqDb1uOj49pmhnj8SlNM6NpukOABK5F+aBksRlDADQNSONpRhMQYbVdkFPRnF07wbdjpscPiSly/63v0fdbmulfcrk8Y315yWa5JOdCvEIGxJ/hXGJ8NKUZzzg/i/SrM4xaNq6Y35thIMRMDALZMZuc8P5775CHBbuLNWHIfLkStkEQq0zmNQ6/Ki7gbH5UinukHaqR7Saw2wSwiZ3dFJfKJpwqvlWch7SGYVMKVRjdZxQWSzdXfGi3K9lxCmguYzz24C2EAAPgPXRN5lVD9tqFXXhRk1xpkQMuZGqmUznhYKWmpDWJoBxXaJ7lfUq5QCICuSbC7GESKBOyAP9K3sc5agJDypmcY609Fuu1Dahh2G3Yrhf02zVh6PGNoWk6um5MNxrTjSc4f6WtkipkSsilLlyq+Ss1/Ythspcz4fRuARR5Wavv0f99Z1UJU5U7Bb74z19tpt91/4VsM2I8PuH4+D2cnzCdnOL8mK6b4NyolJG+DmaZm9EBVcE3DW03xvnmUIcgDj0xDqTdhtQPiK01CTVjxWHEFg5DDhgUZ0wt8ChkLXz3TCxAli3hVGc6jCgxRwpYFBEpdd+cB0gMw4CGgulotRY9peijMaAJeicFCM6RpJkcIzknhIxvSlxeNdXwcelnqViKtaW81SHf/dAPZUaJ7C3hK7kp41qeqxSDsTVLz4OEO8fotQq7SIkLFk2gB/8WKi9YiyYWa/CApxSgTDmRkxKGsnJ1bUl2ca6UnkoxsVmXhzS1xl3fbxAxOCvYCsqNuxYxEPuivdU0qHHIboc1z8jZsOsLnbFzGWdBo2GzXBFjZOgHRt2c99//PrOjU+bHx3SjDl+FPSZlOySsEUZSKubEVDZVMFKok68y8fem+vV2JfRQhHU/7HLjvH3/HqrPUM3/PYhXr3H7WnFbBVquZtZXjqsgYpnNHzKZnjI/fodH7/xRCUHZwkIUU9ydPdlH9n6vCI0viP++cEnn7pHnc4Z+x3p5QQw9q815KTv9/EvWqwWTbsJ8PKuJ35bUb9mcf8n64gta0zLqOuKQCMmxHTLPF4GQA/NJpPWOaTOBaUOWBdv0OSIB5yLOKdOpxxpDiJd88umAI9FhEOMZzxJjC51RGpNZdhZnDCEkVptzYoRhp+SozE8Mx/eFEDObPhACxMGXUtGuuALTsXJ0osQgbFfFJYg5krJijeBsWQRESmVjTaWPohpSEiajhlHX4jz4VrFyduc4/VbN+BueW1Xf+/8KSUEOGXB7GGk/0bUCdEaklJ/co897ymIuJrtmQW0RtOIuSKFmaqkoikoJucWBlAwxlnpm+5zkMAyI3VRhK5O6aTvarhRX8G1b6I77AEPWg0bNymFhE7NPz73NfL6JfH9V+E2v/f+2828sJncI610EnsM93nLiXQvU/jrONah6xDRYN60aKtU7LVbUHkQqEQupQFcFK6VoLGM86lzRjvuOra+UEiEGQugZel+iNtkTh54cBnII2K6j9Q1OMyY6Yo5oMoXumhJqUwVRLREpFY3qxiWC4mwhCsUYGYYdWRTvSsRAyo3jXNlgom0MXWsO4cYYK0uuzkPjDJJztVDkgE+gpcadESkc+Xw1VHsZoMbtrw9jKdRUqLXF66oRIAA1vGp1fv016IBcHMvyQSmLXSNxSr/rQQoK66xFMFiT8dYxG3VFM2mq2wIVxpCzhmY2KUkvu6GE6moMT8UiVoghs1gWwkHSwoGeH48Ydw0pR9brbf2thLWeyewBTdORUIaQadoJ09kJfjznbH3Jloz4wrX23pd6d2hhgmWlD+ngcijF94pSGF3e7bV76YLrPvjVonW18wjsQ3C/GqW2WAoleWhfCvurXIlvlHl3zQo5IMOyB2OLz2nq9a0tySTlN6/hG1LMCZFSneb4/lvknDnOD0kpMpmdslxdsLz8kidnnxFjYLfdkkIg75SJPeHk+B3mJw+xWJw6Fqszel2z3lyy6RdstueEviTu2LZnOjYoriS6ZLDeMDKGlItlJkQyPTFn1s9LEUp5OKY9aelmwoMJLBeJZ0+3pCEznSmuUVQSF5eRnCDGUjZ9dlQiDrtdYrkTOiwdjhCUvmZeijosFJqtFIr5MAy1P7UuesXmT3HHeh0OoF+IvyNx9n2I5boZukfU9357rDdrTSlFRdXwzhq6Ualx1u+2JYElK5mMtZ6uLckY293A9Thz0QhSdx9JRdNSiTCUggG7QRmGQsfsmpIM49sRTTdj12/JYYeYhqabY/2IzbAjWaEPPaM4qhtUFOJs4ToJsRZqvKavD+8c9vD5XeGw2z6//u9dCPv+u/11Spz/1a7DV7VXEXduO/Ym2v4y2m+kmO+uRlmuhxOueIaC8yW9eC/8qiWvwjQt2/6SVb9g2G1ZXlygMdOoozVjRqNjpkcPaEzDyI3wbcv88yM0BZab52y2y6LhU8KSaRshq6XvtWxHZkwJrymIrXvO6UAKymbjCTvh5KRFzQTnldZnYoyolhp6Tat0Y9hsle2ugMCixYrpurK4bXthiGCCwVlPjImYCtW75MaDIRfsp9axBwrLhpJLorlYH30fMZWIl9LvCBovUuiPhzTWisZn3ZegKgNa3kVCLCCIdZSiFnFVJ3HN7q87+EmKhGxQMt6XQnzdaIL3DWEoedaiGSsFzfdtW6qgOEvUgswbLJph2wdcEmbDgLeB6ewIN3oL70Y0foJvOloaWm1obIPzvobE9jx9KqjCYUG5Lk9fURPwRl9dhRdfPunrhvUOS+uLCP7fc7tyWW5+vjfbb5CN9vK9//Ng4ewXtky/3ZBiIA49opnZ5JRvvff7bLcrnvvPySkzG89w1rPaPeWXn/wfzMYPuX/8PTbDliEnApmQS63442Ph+MixGyLnq4EwwOWFkCJMR5m2gdBnNjtBSWgudQydNZjO0g9wcV5M95QGNtuI2IQfpbKLiyg5G1L0NA1MJ/sCLCVyIXaf7ZYJaSBrxrlUy1tVIHVP4LElLKlKKbVGhZ7NFbDnPDSdYi7uHpPXK+xGaFpHCqlUf60FIUCRK0YsUFIZRRLeCU1TyDV9DoiA96ZuBVULDlpwudTL9k2pfnpyOmc0nnB5sePyYodowFCIGpPpuMbRLUETGcWIJeVMPwRsgtAPNHZg8tY7HL/1bqndPRTEs6OhUY+3RfPsBShfsyb2YNiegAFl6uY9Ev0VMneXZv9a/fzSuftff01N73hP6QtnvzrL7/oz5JTZrpclhh96UGU2PWE2O2W9XiDqyJq4d/8RTdvwF3/xc37x8f/Og3u/R9c8YNNv6DURUELOhJSZn1i+/V3Dky8zHz9JbNZw8dSRouDvR1qbGXawXJR8a9FCw3bO4pyh7yGEshPRZr0tO9u6SNNkjEvVpWtJ0eNGcHRcumK1TuRQd5Otwh5TKUDvSo5O3TFWiqA7sLHgC6lSdVXBuUKnNbUSbdNCNymf3dV+CymuRdC1xsiz7vN+qwFXgYe9T1tINC+ATVpyi8UokkshjBi11oy3GEwtILklpYS1JYxm1IIIIRS0k5AQMQUETApi8LbFuZamG9GMRqSU2axWpbSxeELKLJbP8X3DSX+PLk447F5jDd4V4d/HkosLr4fbz1oWOKNl0l9vt5nqcuX33Nru4rvf9f3XcQPuai/G7G+LCNz8/urdwVeXmxyEV7kG1xmSTVvKescUCCExxIF+2BHCwHg8LcU3xkd47xEZMQwOY8bMju9hGo/rJtjYM5kdFd5GGrh41rNellXZXJHQS1lnyq4v3bgQZzQW/KhrR1jj2A6w3e0QUSYzT0yJTR8qV6OkX+dYKNpDD4tFyZpLtf79fhvpve1lTNHkaFEQSMGYI7VwpWipUmNewEZq3D2GUlHnVTvDvN6st6zsdj37nWBSLQvkrKk7r1xVkTHWVNBBcN4cOkUVUih7WyOm4JaSkUHqVk8N1hjWqzWqS4wd0bTjIni5oMPrTV+Ta0whoNSNHhvfcjyZ03YjpkenjGdz+n5g+fnnjMZj5vMjYgxcXjzDOsfRW/dppjNiLDHgru04nh0fYtaV+nzAKYwUwGcXChNwJDW08qomXKX1chOpf+Vpv6Z/fle7mmD5BQT/OvZytRiYusjtLRlXIyLOloIiL173Vc0Yy2R+TE6p1ILfBJaLZ3x59jmjdsLbD96naUbM5/coG1CestpMsM1D3v3Oj7hcnjP67G8ZjOHRxEPesnz6mJ//1RcMUuaaq5s8ZlGSKAFwo8zRREnRMmw8gudodB9nGtZPnnF+ecHJacfDd6YMQ+Dxky1DUGL0pMEQhkSKPaulsFwK3sPR/YR1SoyGYTCl+KmWUJxrirDnWKv9Dlzt4SZ11xlXq+HkXHdLKmMQg7LbUvJO7mi/BW68HszcfY66HibFTVhHZJ9CeK04InrY0mmPcksyhFDCbMmWFMISXst1h8xaeskWkoakfbKw1NBF2ezA7ndrrYvIoSaY2DIIuaQ3mro9dEqZYegPnx/8kCroh6C47gGmK01/SHN+RZjtQAgSfUng4SZ6/7X6/lc4dv8cL56veh0ALAe9vIjI1fgd7vXKWvv6P39zUTE1AUWkhrpiJNqBXb+lMCzHWONBFeccSmC7O2O7u6DfrRn6LSMf8V5ZSSZGJQIRIUU9aNiunTCdWELuGdKWLIaspcZx1lJuu0SDilmdk5JiqQKbwp40Iwdz/PA0WkOzL7ise2VwHbSVfcSqXKxYHwiNK777UCvYGlNkQrUqlle0106qKf5OLkCD7jdmzPS7soWOrT6IXnt8ucJqKGG5UnF2CIkwRIZc9vp21pLGxVJoGlvMf5uBhPMd3ewIzRmzfE6MQwkHqdA0DW3bFvPSKFkim82KlDOj+QPmRyeQBobNBtu0PPrWD3FNR8zC2bNnjEcd467DYq80eh1Nw5WgSxX8Sim4AUJ9dedxEL4X0fG74vXX223H/tptL/DXNmy4/u/ebcxZiXUGGpFqnn4T6wKs8/i2w7kWg2W9XvH87P/GGMOjo3fpfIfGS956MGHoP+b/+jf/Hcvlmg//9icM/ZbvftsxPjYs/RLrMnGrnJ8JKRW6ddc4vvveP+Lb33qPz55+xC+++AkpZrZbasj3OdYIuyGiybJeKk8+3TEMqVwnGhpvCqEqGEiCa5R2nDF1+6kYFE1FSI0F19biF1lADVYaxFiwO3ItHZ2z4L3jeFQ2y3i22LAdAm2rNB5iFIb+1aHV10+Xrfey1wpUgkzKWrB1uT4hr613enWysbXEcLzy/WMoSGYIEc32Ggi0T4O1WN8WYXcOq4laEgPnPW3THn5GycQwYIxjJAbfdKQhEXYRqw3taIZvR+SU6bdbWueQ9hqk8ILPeg1qKAsBe0abHgT+ql/uNtPv0uS3afq7rIW7rv9KrX9A1l/m8N+FM+xR4gOXgJv/Xn+mr30fcKiQY63FWc8w7FiszhCEmZtCExESXevJecv5xXOWiw2b1TlxGMixxeAwEpCyxVrJQMvgbKHSdu2EyfgejX+K5kLJlXpvMfckTeRsEEothd02MQyZ0Jf4tzMlTrQ3sfcUXBHqtlB7Uli1eMx11QaHGnwilZhSuCnWlHRZoWyGihTXw7mqTG5Jp77eXrOwCzntSSK5pKq6uptGLPHIffjKVNtGsxJCiRdbZ7FG6LoOZy1GdhjT34zb51pSqv7idH7E6cNHhJhZbXakFBlqZlw7nuDaEa1vadsxOUX6zYqQEuvNkn7o6Y4eQI2bb0Ig2sA0bFHRurlAQHIoBQyA8WyGwVx74vJAZRG5MrVyhg0JazKdd3h30499lcBfP+6297cJ0F3n/aqavpiX5fWqezxQoWu/ZPZ7melX4xQvtrpiHlwDEU5OHjKZnvD02cc8efYTAB689S73Tt6lN2vWw5KYVgxBGRl45/iYMAysFgObTSBmpZkZZl55ZErcOg6gGvnZL/6OTz9/zGZ3yXozMBmf8ke/9wcgiSfP/ozt7jl2NmLmGoaobIaCvjetBVGOThJNl1ktLOulpdQtLa5q3xcmnzHQjkpXpqEoAXGKMRnf7hApm6jkUDahyOqwRuh1V1SFi9hWiRnWm7oe22s0vFvaa68uW9zgut+bM2W72f3GEHnvu8uVQtcKOtiSxy7G4H1D4wudUnWf9lpQ9c1mn3BQXIF2NOb43n2WyzXnl2tCGIgpFuqL97jRGN+M8M2YOOzQzZqsuSC9MZDq3lllh46IplgyqYyw2ywJ/Y5R2zLqxvimOThjAge1dkPT14B8FtCYMCJlPzpeCEVp8dXLZW4K6s3klZcz5K6f81WWwCuH6674/vXKO7f8TlkLKke//htzPrgu1+/9+nXu1Opy9aYAncJkcsRJNyHmLao7UGV+fML9R+/xxfOf0zybQZ+J/RovmZPJlOB7vlwFNkOkmyrtxNC5zLEUf3uzKorlydPHbDaP8bZQYo/GI95/9/sYSWy3f0WKEd+BGsdym1jvSo6q9xaxmfFsoJskQhS2W3uF9GcOeerWgXPFnM8RsFUhmIxrU8EnVMhqqDYvihIYyvy2CaNKGkpxVrFgm71Kub295h1him9tXA0vAHHIB6CuHlSOS4WPhi0mi1hD07qS+JITfVBSygeabSEaG7rpGBHLZNTSNI5h2PH4s0/Y7nqWl6X4oeZSlEJkSdgler9i2zSkENguF4BwdHJM0zQsF18yhL4WYTTk0PP88ScYYwtYWJ+jaTqcbdjnzSNyA3C82Q+AFrZWuf87O+wrXfqvQt0PgJq5Cn/tj3mVsN8leFpW0XqR27+/wiOufZ4rIGvvtiq+CnAUgaYbH0x5ZxuOjx7yB7//H6I50bVTwrDFSKRtSuLUpg8Mu4H1rmcYBpabzLoXxI/oRg3eBvx08/8x92Y9tmVJntfP1lp7OMeH6/feiLgRkRk5VFZ1FS2g6G6BQC1BS3wA3voNgYTUz0g80OIT9BMSry3xABIPIIEETwiEVEg0rZ6ruroyO7MysyJjuBF38vGcs4c18WBr7XPcr/uNrBaKrJ3p4X732WcPay9bZvY3s78xz5HtZkdMiccnT3h6sibmHTFviLLll5/9EZC4uNgxDA3Ra5unOSfa3iC2oVmdauw8v2Hc7AizxuhThHkwxXRX61RShJyxJtP2IMaWoqHMNM4gmVYaupWSsIhREPp6owVX3heqq9oFJpeusO+YMN86B11Iia4TXCP4KTOPB7SalD+zqFlfrIDcqM/d95qtNvtA8p5cml1nMjFnrLOsj0+1kV/f0jnH9c01L1++IITIOGjijVbGgR8Sxm6wTjBONLSz87im4fHTJ3R9x9X5czY3/5JHT5/x7Ds/Is4jL15+Tc6J0ydP6ddHCNB3a5qmK1GGuphVB+Ou37xHYxNq1dTtrQKVXwPAe8hEh33Ew2ikcTn+mwT9XyVrb4mspP33a0KRtubaFwP9K20irFYn9KsTLSxJifeefsKjf/sDUggMmyumaYeRmdUqs9kFbnYT4zhxvRuZZ8/5TWIYhb4/QuIZbTtxdNIyjhMvXmgbqI/e/4Anxx9xM3/N5Tgzjtf8+E//npZYB4vkhimBz57GGvq1pV2tefLhdxExPP9VYHOdmccAOerCMBoQowQpRsjBk1OgWUF/BBlHTmtiymwGzcx7+qhjddRjbMa5xHYbGV7DPOeFfXYZmizkYHiHYv+WzXgplLjWqMCZxN3aDllADQ2LVQ5xMbqimULxXEkos+a66kPGjJ8nUkyMOZKcZRxGxnHCGMv66EgPC75oXyDH0hFG8QFrtDSVHMjJk2JQGiHvmQYtm226FUYM67Vm6a1Wa5quwzl3WzSref3QeBTALsTE5GOJQ7+dWfbrCN5Dpnnt/Q0l7HOAd5oHFom7wn5rASqW17sWoRp9OHzO+sefJ8/+1vcPojFUDEJKpllpFaYElwNkcKaltWvW7SlOetyZZfYzIb/GDTuOjhx91+FcAAlkwgKe+eAZw8BumLnZaD/Cedb8gtZoJ9tWDI3RkGzOiRRmpt0NYIk+kKKUPJFlVQe03BlEF/kMIQrTVMcmFoBT302MWYlVs1JSq6ur877vlbHJzwbvZcFJ3rV9q8JuROh6S9MKzgm5xCgPsVpTTZJsCgmkpe80822YCrGEFrUuqzvFz48xMI0XgLC1Siq5G2eG0fPo7IwPv/MJAG9evWGeJyQraUHOmewzRgyrvqFpLMSBMEVCnIkIu2HkzeuXrFYnfPjdH7Hqj3jv7D2OVmvee/oRJ2ePlBmn8I89ZIIfzPsly243R+aQWHWOk1Wt9/5XD1EdftdahQtjUC6/anCICHLAfrPc1gNa/e4T3Cfw1WJItQPn4qPfDyb+uZ7x7ngW7CDHSJgmvB/ZbC4ZhhtSyKzsKWd9Rzw7Ask0bSbmic9e/2MutyOnbc9Je0pInjFsiWkmRvDBcLm7ZsqRV6+veP71DiOJ1umc7U+gaWG9aui7ju124vJixzR4xi9GchLGQf1w10J7pIJqTSZGYbtRXrnqvg0DbHaC8jHuEKM57q5ROvTreaLrIK+lJOmoJfzB+5bjY8PrV8LFeck2te8e029ds+s7Uo29CAb7d1mTaFKi8J3vqeliqqZhPWEJYpXaYE1V1GMiWWvXo7LT7ivA6qRUokAjSoAZk1YYVTL+lAIx1K4q6jcpt5j2LmvaTuvaO2VUMdbeCSHvxeFQ09V1YDkmVz7x2hcOvkkG/jzx9cNrVmvmFlvQXSF8ZwjuXf60LFp/Obz85/BrmkCV3+mc3HqWes6D+7513OKmivq9zinBZXfCPBlap+h14zImW1Z9j89rGjHk7Mk5kHIg5QhZK91CCIzzyDR7vM84k5FmTwFd+lmUtN99cUv0xbxOpYWV1Xz3Gl4zgLUK0Fb3unLYCUorvYTscs1ATJgAzstBEQzLvKo/yyr+ju1bR+OVsslijcVaRSQ1Tq5gm3FFENk/yFIGmyKIlHI+ZQdJNc4eVDO3rcMWhhoj4FKiidoH/uLNV4gos40Bjo9XdH3DZrPl8vIaGlg1gIFhNzENAbs+4/TsBGsUEGo77YziXMfq+ISjk0c0fV+EaS/YRm6Pf52XVQDqi86UTqK5cJKndMDg8/amfen0pdtfo3njYX18ZT/Z1yHkvTn/wGluCViNqS//lFu/714350LTdQASx5QhKOOQLQNU14JKw12vWzn0qgK4LyRprKNdH9P0a9xqpeHTcYOftnz9/CdcXX3KNG159eaKmBOr7kOOzj7h8tVnvHz9J2Q3E1uPnxMrmxAH2+2Gq6sd05RYNZamheOTgJiMjwk/iLqTXsNew6SVatFr3nvXBxpXWpk1RuvRJ6VNO15r7seNwDjqIlazSqWM21z2G6OJS2PIXG11kWicSvbFRebqKjL7KvEVtX94+0ZhF5FPgP8eeFZe2d/NOf83IvIE+B+BHwCfAn8z53zxTeerM11E01TFCJL2dEs17CaGpXHe4vvWYw4mRU21jVGXZU19PeCULz85R+Zp0JBQ0qIG1zi6vmU3DMSUsLncj9GsvgQ442j7tXJ9icO5Ri0Ca3Ftu5TL6kTUVfntuX+Y7qv/USHcP1fMh+nD7wDdvtHMfhvkqzXmFd5YYMND4TkQSNmf6NZry7ydRPOWoB+cp1ou9fpC6fCT0sIzeGifVxaWZVdezrK3Ru48u4hgXUPOmVXTkDM0TY/vjlldfo61niwj43xJjJlH3VN695iL+Dm73Tk0yiIUopJFNCYTfWKaaqKNmu+uVT64MGViMniv/duCz0vjiBDLfDVJq9Wc1gHUohgxatqLgHVZe8alfeFLfd4Uy/tyKr7hoNCrM8qFMI2lLFx0hB6O/ey3X0ezB+C/yDn/UxE5Af6JiPyfwH8K/F85578jIn8b+NvAf/muEwkosb6+NS1z9WnRMDkrwYSIaEO8Qra/aM1iE/pZ6YZDYXMhgSSdZSlG9fxdh7UWa1SLd03L6SPlhN9utRpunkdy9kzTTMqibYlLs4Kzs2d03ZpmdYTt1qz6E05OntK0PcePHtP1K45OTlmfaJfTWzpPPYv9pK/sK3fHo5gB9TAfEjejxxmhKz3p71aI5cPv/hrbIvjs/cSYbkkkzcIYU03xt7dbhThF697VtrUP2+1S3/0CpvdTxqks0iJa1aja++C57jHbH3xGhJQi15evmOeBaR6Z/cw0Z548/l26/jXbeM406lyQfAAAIABJREFUbbm4+JQwfc7lzdcMOZPmRIyBFCBIIjfQri2pEUqsBJG05LmHYIlRmPCkdI2PAesixqgwkhXkm71q5rZJOAGbG3KEcSjzNIE1QjIQjWrs45UizWNQKywGDS9bAdNQeh/o81pjcGh0K6SEMfrzru0bhT3n/BXwVfn7RkR+AnwH+I+Av1EO+++AP+AbhB0RZQEtsy6V8tRFA1Ey6RAaY5UeWvYFF+UeiD4Rqr9etKQkPSanRJbSscM2WDNjEBrXcHJ8ShaY/EyetT1RDODncFC0oM0iHj1+j+OTx0gpGD599IT3n31fUzVdi2ta1us1/Xq9n8C3Bu7tXd/wAT4m8phorOZW22wWIdznIby76u3huLuUYdf4/v5ewOV98Y/mB7ytQW8/wb4w5/De9uDeXtPkQg7y1jnyfiQaMXvz/fD+f12BF8gpsbl+w3Zzyehn5uiRmDg9/QG27ViNf0Q0npsvXnN9PqL/03LjedCkGAPgwImla9TPRpTsdB408y0Gp8KMJ6ZRqcyt+vG2NDsJSTGkrgErSVNds9XELB9LXkjhjtOIHE0jnJ4o7mPGiA+JcReJoZrzqkC819dWlUEsved0Eb4Txr6z/bl8dhH5AfBXgH8APCsLAcDXqJl/33f+FvC3ANpGG9f6WW9aqXaqSc8CVi2VQiKFVTbuwR4yMZYQhVWQr84a7aypxH7OOdq+JYZI8ImUhfPzC42v+1nDKI2jaSwxG+aQcc4uqbuz90zzpKaYtTTOcHZyjHEN2bZF6Js9SMTbInyg4JBbHtXhC1EToGrdUBatySesyQU1P/DFpFSSPSDw9xXJVGuvmvBVG5u6kKBa+ZsTdMo5ud/HX7R6tUAEnLW3nlYnuCzuiNwJXdwK+R1c5zDz7t77NIb18Rmuaen9jA+e3fUF11dfcbO74fp6x3YYCGZA+pE8zwSvYdVcqL/7Iy3NZZcxXhCXERfxE4xbzXs3AmIKA1GkaHIw4nDdCWDI044YfalsLHhMrr0Kyk9lNK3vRjIO9flziooBRM24M1bpqMkUbgclTY1RraG2UctMgeT53nkBfw5hF5Fj4H8G/vOc8/WdSZClcjHd2XLOfxf4uwBH6zZnEuOQlMhP9vFBVzq2zikXrkgV4hAy8xTLPegLn0Mk5UTTOlyrboEpiJ5qdqHpGtbrNdqO1zH7iedfPscY4eh4RdM4+r5lvV4hZtQuHwg5ar38OE6I3dG1DW1r6RrDs/eeYF3HbFqyaF+5hxTP4nYeCHwdoH049NDWV/PaZwil2Z81SpVdAbvq2iEPc8rVi90VegNL/r52xUXzAqRo+1zwjjuVbHe3e3109oDcobDWPmi3lPUBnqBA48OL1uKrZ0p8+oF7AIy1nD39CHIm+JkQPF/6n/Hp5b/g6uY1r15dMkxXJLnCHI/E68w8oC5kBNNkjh8lmkbHadyA9BmzDuw2wuUbIQRTWjcLEqqwFv5217E6eoaII4Yvyd4vrkrKiZBKxltWZZYj5LJYGAEniU5UqaWY8AFmL8yzujmuSYgYmkaffRqV6WnVWbpO22ql2qb8ge3XEnYRaVBB/x9yzv9L2f1CRD7KOX8lIh8BL7/pPDmzmO23fbiD37CgsIf17uUMy4RKmdKHOy3VQXqcaqp5mhmc+m4harfVmFTDx6j59CGkxUqwJYyiqKYQgsfPM9YabLJKYnF1QdOtaU6faqLP/fO+uBqhAg2AJgZZZ982S+Xgy0X2q5aXrFlnISYV/G9ImnjXJqLofSyqZR+2KXd0BwS9TwbvDfUtIEI5H3dz3t8WzoWB5gGQr2bhLfd2sP++lNplYUuqBLzfMc8D43jNdnfFOG0L6GkRHCY7rKmcb6KkpU7LTcVClkQiQYQ0W1Kg1HIIbat876mUqaaYwZeByAF1Yeq9FwrSrL52rBVvFLoyp6m0OUCIiupn9NhqAegA6UtRt6JYvlKsMys0jSFETcJ5F0r366DxAvy3wE9yzv/1wUf/G/CfAH+n/P5fv+lcKWd2Yyj0UeXBSlaXtaY8nPofsfRch8XAW3jYK/gVYsmkKwi6UBhBRYj5De7yCurCkdREFoFhVGoj45z6yUmwRmPpVrSpwW67YZ4nYn5KMiu+evGKafMPePT4Kf/aX/336PruVgaa3qFuIQSuLy4I3i8U0OvjNcenJ7eOW8a4SLks0BclHJcZ5oAPwqp12MbVL+h5HvDf8z1v3DqrgE5M5KAU1ZVauhJMpJJvIOwFsHLFfVPeej3W3DnP3e1Q0EXh91sLoBKSKDAlSuCHAlppWQ/vLgB675Fhd02YJy5uPuNm9zWff/VLPnv+Y3zcYhpH744JG0P0npXbwslO73VOWKeFJGLBy8yIJ+wc4WJNBtpO0fPjk4hrItMcmH2CjTBeGu0G41+Xuv1ZU7ANGqILwjhozXyI+gAnjyL9KrPbGIbREAOMg1bNZaPsS7kACcZpzTsIMarzLkawAqtVw8mxYzdGtlfzO9/Rr6PZ/zrwHwN/LCJ/WPb9V6iQ/08i8p8BvwL+5jeeqfg3UgkqFg0vi+9SNQ4HAZdDjamgT9lXEV1um82CUkOHosFrSCzlOhkzkUTwkdnGhf2ErFa/ADEGkFKkIo6YMsMw0K8ntL5574DlnKmcuDlDTJEQPMF7TfQwprScYjnm7rYIneyPSZTFkLz0tVs0ZXngdwJpec9hVwVYzN5X50DwcsUBqr/9DUDgfSGw5Zr1/Hc1+p1Q3jvPD4vP+033kMuk0KpJzXqc/UBMQXkMtL1veU9e3RixSxQAW9o3ZQ23VbcmRiEGCwZF3K2GzKxLmAQmqSWg95DIcSYti2JpzBg1PSSWn0PPpXYpzxliFmIJ22pJrM7qGmsvr2YZmwqk1opPIw+7RHX7ddD4/+cdZ/kPv+n7t84FC4V0VqIaFaiooJ0IdK1ZJF5MaQkU97F0zZIrMWOt/NOJmevEVfPaz0obWCc3JaafTTWRhe0uMEwZ5wxtYzHGKsmAMWQRXMqsVie898F36LuW46M1x0fHSA6kaUfKBrEwhonRKzCSS/5yyIlkwZVMQQ6AxLe2g9XqMC0iZ/CxVMflgI+K1K9ah8g+9vxQUsvtS+gx1hikcfeCbAqemYNwWU1q2S+09Xr35c4f/tuWKMZDAl1dtcPvH+6riT/LvRmzH77i4lWrsL7ffn1ME1eMfkcM8KMffswPvvvvcHPzgp/+6f/BdvuKy/FLYhogCDk4cqoCKdy81jj9sJ3wIWCNY7XuEQlYM5OJzGNknDLBQ4hGy0tLg9Bp9AiGkFpStmx3I7tdZBhhUhceV0Dl4GHYatsnI2gDUtsAGR80vdm1iabXVT/6Mn8Ox0wgY0nZYZxlfeQwZrp3vOE3Uc+ei6nGolcUYQ+aOdY0WihTAwlQni8d+vD7cy2av2imOnFiqY83orXjYrXfminfS1mBuDwr6YCRwuclSYk8TWkR1DSs1if0q5716TFd30EpfMBGIDJ7zziPRdDLy8jFZyiUwYcubsUlbm1yMB4HH6WyKPqliMLRL2V1B44+D2vKQ7EReCfh492EnKrpD4+6H/F/+zz3+eLLNe5R2PvPC3nJnacwB2O4YDnVGhPBNR3WJdrmiNadsjo54fTkPS7OP+f5l/+EOA8Ya5TkIQJYBctyJidDGJvSAkwbM1pjcdZhJGFFY+5bXzIeY+GYK+2+yElbN+dcOOscMVqli5pLW2dAGjXvY9RQb0kKVdfW6YKWg4aCbUnCSV7ZYwsiUge4WE5aRyJSadQf1u7fMlONqvRcgRmKEEhJjSxHVbM1poy1hq4vgzZFFeJ6nOxN97qI+KSpmM4aTFM6tSSwxtI2DmMMTaP+eQg6YWKMTLNSEmVRrjBQK+TizctCg2SwjeHR6SNWv/f7HB2d0to11hQSjagtena7LdZYHh2f0rhmqQNYfN+HtPvBdst1Kf+JxRqSkNhKwAg0zhT84xAsPPx2PUfFPr7hwsu4vt1gsmr3VMCzxdwHbL4dEntrwTg477L/lkV/d9Goz5CXGP1dfAQEWVpFHeAAYjg6eoRzDZvrN/ziT/8h29052bTY7oxpbrnZCNk0OGfVhaMhhsiYt8TkWXVnHNmG2U/M0w3OGFx7CjkTww0+ePq2pekbrr1nnCdtpZx0hJsuYK0SYXivLqApfdk7p0olJvXfNXGGAiB7ctZGEClnZFZmJyewbnU+zrG4GT6qFWJnyJrFN86ptD67f/sNNHasKlkWTXxrepZ3Wnu5ucZoAf8Uy0Qrhx1aoYuwg2Y8QddpDD0EzUSy1tA5h7EGW3IWU/J7ltCcMUnAxKXDi4mJyzcv8dOOGGd8GHny3od89NH3SabluJ9pXaP+eMz4aeLq8oK2aXly8lh57UoYTarwHCrkO5vsH+etj6uGz1nBL2uENUpVJFSAM986z8Nne/d2KLiH0RARIce0tOg6vPMatzfFDdgnw94v8Mv57l0QymKbWWr97zY/qMj/2/csrI9OWR+dcnXxFb/8+T/SjiuuxXWPGb0Ke792dK3BmA6RE5CZkLeEHDjuT1n3T7i8+YrN9hXZrViZp4VpZsTPkeNVy/G6Z9gK0zyXPqI6BirskZwF750Ws9hCVNEoqepu0AaQitIX97T2c9Mhxc/lfG1mtUrEBHPSRT8FTfYZ0bwQHyLTFG5hQ3e3b59KegGC4ECVK0Am2oAxyr4tVAwJP0dCSMW818QZMrhGmTxDSbzRU9VJysL3tgAh7DW5oBOXlDHOacy5IL/kvFBLayhnxjYNx+tj+tUjhnHGbbd061Oc60gxLia8JkQEQgz4EKhZTblp94NwKNWHY7Ps2ouIufWFvdumCTiaQJQBs9TimyW0o189+O47/Ox3bVUoK/hWF5bDJBu555xv+eXFGngXsn94DnPw6NVkX8xU2ecd6PkSw25Dip62O6Jte1brUz76+HeY5oGb6YrdsMFZTewKHvIukbMh54kYQ6l7EE1gkYmco1Y+pswm7wBoTcK1QpgSl96z3SZyUlbhGl0oTGZAxjmNLKXAAgBK1kTcDGTZ27SHJrhA6fgiIMo1F2IJNycp8fsiI6WRqWtqP/f7t29d2FPW8FgVPkV9zVLMMAwzKSaM02SbOYfCfAK2sRin4EhOma5vOT5qGMdAjNMy4CKQIwSUQ14aFfhcQzhRmz+mkhTRWMdqdaTJD8GTTcK6jqZttLHFsONs/SHPvvM7tG3P5dWWYYz03QmNNIWwX/PyY/BIzgzzqCmUSfvMdU1zF5wG7vrC97qyOt8PvpuyPv8cStZb0bSds6zakk0ldnGR9Ny/vqDfJ4i1P58xhq706wvFETXFSquhvJowcwjqGW4L/H33c9scr3+nEuHQ81kxJSdi/2ygZvD56y8Ydtc8ff8Tzh5/yJOn3+H09AO220s+/fQPubx6Qdc2iATGIeGvMmICxikXYN+vNaTVRLK5IkaPHxvGlLiMb3Am8/Fj6Bv48sXMywtPjEIKTXmTmo03TYL3YCTR95HZgw+QkybmRIGUowr6smKBlBLXqrCaVrAOSMLoNQ3X+xJNKu5siNq+zLpEu3qbDOZw+w31Z1/Uzl7JFeAtlR9F8PQfMaiPp0CXdtEgF7+8lmzuF/y3keZ8W5CWsJMc7jz4Z0axhZTB6jUzUjrHptvCIEJMCtLlnOjaDmMs3s9AZuUaGmNJOTPPyphj3cGwV7Cr/hwKweEDLCpOF61aQbZYAXlP/WQEssnU5hQPbXfR8Lv7b29qXu/H+QCAk4fRgNutr3jn/SxXutfnf1hjlYOJcSL4gXG8YbvtNTJjLDFHxnnHOO1IKSCSFcyzjULFOaJJKop9pFRAuOTL/WRsA670XjNWNbK6GKamAiyOi5QkmEwuzKIH45H2hi1SSqFLzkm6PR3LAkBpAgmmVtVVbK7MCwWiDV37FwmNFxB7p+jBWAXWfAEnagVcZZotnTusNXRri7WGo26tVNImIpK0sYQVJO+R5lqjnTJLWmIqZpFrtCEEPhBDBEMhMMiLNtJOL5Gj4zNWR8fEmHn59eccHZ/y+MlT+lVPs+oxfcewveL1+QtWqyM++uj7hBh48Up7h3//kx9ydvaUeR55+eIrur7nyZP3tSy2DMsyF2qkgSJaUgVEalOQ8nr3RS1lVuk9Fx792mpJpIBGD6H0v4bk7ZNrpAjDXvuqSfzw+SrYlHMmii481ppb0YD7wnb77+9r95fJ/8B9phzxYcPozxm+uuSrL3+C647o14+5unnNT37xD7m6fslmeI1zkbMn3+Hs6W8z7r7m6s0fEuLMbo6EmIlDJoXMNCXERvrO8ejpWvkXUL/eroXuSF3FHIpbU6oJTecRm5kHyzwoSGzKW55rS+WSrdc00OhUVErorAJdhdnYTNsIfW+02jMk7TwjqSD3meyhO17z7IOnvHj+5YPv8lvnjb/XlM17pLeucDXmXCvaFnxLROPizi4llbrSSVmd9Tr7WbEPZS2CdDhxayQgH8YDiqWR1Ew21pHRApoQtOxoMZFFiDkSo2qBru8x84z3M9M0aFOKUu/u/Yx17i3Qarnmnb/vDlWFO24dn29/rshvXvLfKabzMhoPIOX3bfelx97yqY158DwVaEvLAla41Q58//1958Usf/C+5LYVuOy+dYw+7jhtGIeRbjWTTcswbtjuLtnursgpYq2l61esj04RLhk3CfEB8WFZYGKpKTc20zTQ9wZjFVRLMZNNAQ2rz51YQsVi1FfNaHqs3v7ecoXq1mQNr1lNvKndgvabLh5i9h1bjRRSabM/V85a5ema9p0L+LffEaZ0uFCf+SCrjAI8lSoksVrhkxOFCDBpXDxmrsO2FM/oMSlROoBo+WvVzgLKTd+oBtL02gx4xBQsACWqmCdfBlXDc13nNO6ZPeO44fj0CU+ffRfXNNzsbpjCzOr4lJwzq37N97/3O6SU2W5v1FwHGmsZdhsu3ryk71e89/6zEre9J6xF1diHC85euBfh3yvzt1aEpZQ0ZnZT0DEypbtIOa+zhsa9+7XfW1X2Dg18uB26Bvtj9pmSsYTtREpmWBH8DEhZXPdZf4fo/v3XqJu1LR88+0ucPf4uv/rlP+P1q8+xTcuqX+P9mnXXkvqe9slvEY8DM5GXL/8Ea645eWLxc8c4O3LMdEeqUU+SJyRPTIF53hCjdn8JAXaDklU0DvpOKaNz0HcwekPwKtD9SgtW5rksHsVVaBpt0qiptIrKW9FqOi1g175ys8/atHFbrIil86vOFtsIphF8mHj+/CWz9w++12+311vRiHWiqqleQCbZa/hcTFdjNb4MWQsWoi4EoRT1dq2jcVa7cDq7p7cqZaKUkJBzRnPjvWKgIqEg74XBNmVCCFjncI3R1kLO4BpLJhL8iHOOR48/IEtmGK6Yw8Rud0PXdpw9fo9Hj5+y3W549eprJbNEXQo/jew2N/Rdz8nJKYtnl6upXlZ+Dn3xO9qt+vUHYalFz9125wveodrTCOD2yXtQLKM7mvWtrWoMbgvVXW381tfu7qvYyMESlrM23TRGMAc7q8AvX+Xta9ZrHAp/lhp6dJyefkhKkS8//zHTdENKnrbt6NqOzjWEpqU/ekTuHa+uP+X65kuOjgKnZxbXaCTDSKbrM02XQRKIZxwjr14H5hl2OwXJ5nmPQbRdcaWMavJxEpJXq8A2afHPAWo/dWf1Z/bKAZ+zCroWdZniWqINJKIuJILgDoxWzVTUUF5InstLf09YdL/9Brq4UihxKX6l7jdllY+imIYtAiySiZTJkNRkioVO14pmxWnRzMFkK2a9asJc2kBXq1uKGX+bFENZRhIxBARhtTqm73tm7/EhEMPE7vqNLhJaAM3Fi88ZLl+yufmIm+0GP49sN5fEqLnxNcyXYijda9ir8Xu2vYbfH3ILsCkPsSyWD/iwS85BAYCq/47I4i8/VESzf095OS7n/FaSTf0Mbi8Gd01sU3CHyU/EmOjaltY1d661N+V1kTLLIBy+09uDJff/jdJSrVeniMDm5hXTPPDoyQ9ZH3+HvnmEMz2PN9/ng+3v0jWek/WA9zMrOWccB67GTxn8a+ZJmMcVIUQaO2HajHQQjeBMZu6BlPFTJCXwXgkuJq8pr63V0JssuSGi1XWiTk2IBkym6RMxwDyWcY/qE+SC+mWNrOn3Sj07RaFpJV1JqS7Y1EPbt9wRBipzSemGXBIxwBld3UJJA3XO0jROCwGkhLYK2FNBHyNKxaMTuwI5mt9djAUVOK85iUqdbMrnSoyxzCXRmwreL5zwJyen3Gwuibsbgh/YXL7EuYbV6gQReP38l8QwcHL1A05vNkiOmKRJ/ql04kwxkKLXOvt8SL98n1d+z1pwYLYvn9V95fdbvj2ZlAuoVwolrDFvCey7BL5aGIepse9Mf63fuyWE6pPmnBmnkWmeETmmcY6qm/Pheco1EgmTzRJb35/79u+3NjUN6doVx8dPEBGurr8CaXny3u9hTMvjs2es+mN2wxW78RpJMy7uCH7LByc/Zxgv+Mln5+zefM40r7m+OMaYmbb3tDbRBSFZaFcwCYw3ie3rQIjCMDliCY/FlHFdwjQJkyr/gBKsGqMVcDEKxkW1IiZIoypBcl46AefMMgbZgK2sPsWCC0nTqHOiUNY+vH37obfinOt7LDHZ+jdq2uesyTTBaJzCNZYUE3OIxfRX5LNaPJppBYh2jlF/pjSRKBdV6ui9kCOZ2lI9CUtmXi7afbfZQM5Mfiz7ZuZpR04dXdsj1uLaFa5tSDGwuXpF6xyrvldKrEZ7uDdtjyvHK4VyneYPqOXy0X0rtMr2bWP3oSVjsWJ4t4TcJ7CZfY7/fcenpH3pN5trtWLmSQtKCmBqrStApBAKf8DV1SWzn2jNx5wcHS3deiswW1cuKSGwhazi8P5uDdkel9kX0uizr9aPePzku4zzwM1wQ86RSMAYh181OPFsN19wcf0lOWTSHJGcsKbHdY9p2se0zWPliMuenCLe75enXK2rEkazjQqn8RV8LCG1BESQVOjYqK4rUPLZc3E9Y5BlPmqR07J2LUrRGmVPFoHgw1IQpkMlSNX6D2zfvhmf6vvTSeMRnBhcMdli1DTZEZh95Oio4/h4xTx5xt1EiEnj7sUsbxrBB8EHELGYVuvMk5/IqaAZRgfLlcSdafSaICKq6WvlU7UGYow8//JznLOsjlb0q44579imSNv1rHqHsStOHj+j6Y+4eP0Fbz79Y07P3uP4k9+laTu6oxOcazk+fcx6dYRr1/iYSsMAu6zOD2kpeUjg8/6zasbfRenrZNOfUtJ6zzWWsJbsiSPr3yFqIZC906EmxEAIgfOLN/z8T3/MsN1w/vI50zgQ/EwMkfXRMcePHiO2RfpjUsrcXLwi+Inur/27fPTsQ+qClCiuTtZsNpXois7vp26tgjukm87138uA6e/3P/gRT55+jy+e/5yvfvoP8fMN8/gCIwmXfoBfn/HZF3+PX335/zKOPZubJ6zXT/g3/43/gOOjj1mvf5fTo8Rw84KYv9S2YF5VrDMoG20QUlAsbX0EYc5E7wloz7WYRV3LQZBkaayW2AavBBeusTgrTJN2hEmVCAOtqd9rbwWhG6eFLsfHPcYIu90W772a716tWdMYxPwF8tmB5UVCXb3q6g51hqeSWRPjQTbWwUqX2fvfCKXWu5j7JUkhU1dI9c8ri229cL61gubl3JIjfp6J0dB0DTk12jTCC9EY5nlUUz3rwKbomactflzj550uQt2atMQLFZXJOS+A3P7e9osMFAFeOPgecMAOQbl8EG688+FD7ttd3/qtUtN7vjhNIyEE5nlimiauL8+5ubpk2G24ub5kHgcdsxC0+21KYBtoR3LO7G4uScGz2Vyz2VwvLkKMUSsGM7T9Eda1BO8JJSkpl8VxvT7CGkuMnhwj1jmaplGLUHNRCXFeQp3W2DJ/BmIcyWkmSWIcLiFNbLev2W0vGKee7U7HbBw3OOuUm86rn+ScNhmpIHet1tQ5I4sfnnOJHuVcqMsLui6WJPs4S53Dh/M5J71W7Q9oSgg5m8ofWPEnbuMYdVd+lz7fb7+BJhGCdXprKbG0aw4+LkKf0XBYDDAwLbnnqo0FadQct41DrCPNnnkcAW0BZYzRsFyd1FkXjSmqZIkxSoQoZllfvK/MKqmEphWU8tOMbyw5GggBP0+M44BrGkL0rE9P2Vy9ZNptuIqBGHe03Zqz979PvzqhsQ3WFX55McpzxwMvR9hXxVU/7R1DubzzWyeTgnjX4x4w9A+0eD3P4SQyds9QE4LnT3/2Y158/ZybqwuuLs4JfmYaNqQYmKdRO6LESE6Jq4vXnL/5mhAzw5DIiHZMbSw/+eN/yvnrF0zTxG67wYeZzXCFsZa//Pt/nQ8++j7PP/sFX/7q50QCPk8cH5/y1/7q3+D00VO++vwXXJy/5Nmzj/neJz8kRs+wvWGeB85f/4JpvOHs8YecnL7HxfnPmIefkbOw7p6SM3zx/F+wG15zfXPJzU2ni0f6nGl4zS9/GmjsihevPuf65g20W54+C8xj5vK8xNiTMsVUn9u5TNtHrFcLM3ghGyEEYdW1rPqGYQgM16P2bgtlfueomXOR0l1GaJuSeGQ0Nu2jlsmSEylFvI9cXW0KaJfREtm6iKRSGvkwQveb0ewii0BR/faUb69YRROHEGE8qH4X5d2qDSYoqHot7ZMgSt0rTp+uxPRzMRdFDK52UikrqZqQVTCyZumWfsopRs15z5oFlaNm1xnnGHbX2Ebw80D0nilrqKZp17SrM6UgnkdiCJqgQ/Wpcvn/7SSjRUEvoJwsION+y7f272vDDzz6cs6lT/0B0PZOBP7Asqr3krPy9F1enPPi6+dcnr/i4tVLhIxdLKMCkqZEzurP76Ytfo7cXE+A8OjxKV3fc37+kpgC4zBwfX2F9xPXu3Osc3zw3d9hdfqEV69f8OUXf4ZPM1Pe8ejRU/7S7/5V2v6Ii8vjnMDlAAAgAElEQVRzXrx8Ttf1PPvgGcF7bjaXTNOWN2++ZNxd6KJphHG8IqctIg3OtiUP4pyLq88YJyHMBRvIW1KcuLn6HEPL9vqcYbehs55Vl2vwRTNfk/YXSMkUnCdjnY6adRQNr1rfNoambZh9IlOVGUu0RDKQa8UghTlHXQNBmZXAFK5FlYc4K4NSjdffnhoPCzr8hqregt/nc+s+TXipM7UuBJnS0CDuDzYiNI3FOSH4wOZGVzyMGjvGmoWPGwTbWJxzajLmgyymXKmoFeG0ZWIv3TiSxuSn2cNOu6u2jSugoLb/uXzzWrXKFPBeyD6wHSasmxjHT+m6NW1zBFEZSh4/frpXsodC94Cqv99t35va9WXnIpn3ibGPWhnXOGXwvXWm5R7yLUGv5/fzxPn5C3bbG169+JyLN1/jp5GmEWKIDNME1EqrzOwHfJiYZs9uNylBiAAYbjZbtrsJnwe20wvCHJl2avLfjFcgwj//o/+bP/vVnzBc37C7umGcRy43l1weXfNH7/19Hp094c3zX3Fz/orzF7/iZz/+RyXyEDESWbXXOOv54strvvz655wcHfHD7/37hDBz+eYr5umayV8yhSti6CH0pRXZkWbK9RcYyTzJM6erwNWY+PoL9aclgMswkxdwN3khInhxhADzpJ1bQkpEMqYdaY48vYmcJo2n73ZqIegYK6hW/fPSsp0pVssyUotjrGhaeRYNY9UQnmTR/gpG+evfsZb/ZjS79jQruQOVQ7wyvJjb2k79ugPZMJReb+BDJExpKS5Q0oV99ZUOiqXpGtVyIagVUAA+y/57YkzpKqv6UdNlRXP2yZp1Vvwx51QjbK6vGDYGpAVpCCkyzQMiI8Mm0jYdT55+yKpf8+jsTDXOAy/jQdP+HduhwC8K/s55QhH2u0AbHPro+0lyKPTez7x58xU315dcnL/k5uqiaCClCJv9TMoJa0FIDPMNk98xjoHtLqBFSx2Q2e1GTYiSwBAiORjS5AgpsptuiDmy+/kG4yy9O6Z3J2x3O16/fkXXX/OLX/wJx6fHDOev8JsbdsPAdrfFWEvXr+k7y2/9YM3JseX64mt2Y+BHP/wr/Ou/928xjpdcvfmM4K/w8YY5bsnBILFDjMW6FdZ6bHuOtZ6VsZheuBkyb15qTP2kLwkxxWvPMZGj9hMOorwJPgTNW8+FickFXA+twDpqfXoIEERpsHIWrcq0FApzSpxeFVLXJJwDg1myLpOCSjiTlwKcClrKX0Rhl2rNUk3UA/11z6xf/NNipsSUsamE50Ils9OvpRKLSIUa+rDcsgIcgiylhbJHW/QuzB44oboXSV0AXyiFE7XOOOpqm5WfzDUtj548w1pH3x7Rth1Hj57SrU9wbV+ffq/cD58168DI8sF+fxXkt6y0O4tiRa9n79ntNhgRHp+eYJpmMeVzCWlN88jl5UtC8EsB0DztFHws4NM07nj+/M/YbW+4uHzJMF7vMZCU8Em7kcx+JOfIOA/MYSZE7Z8GiZi83mhpZ52Sx89eBSWWcKRR12nYTZqTvra4dUMMGlEJfubq4hXTdEPYbInjzOw9MWhZcUoDMRgur2dCMqRksWK4unzOv/jxH5CSZ7vd4qPQ2RNO+qcEGqIYYlKmWHxkjBYh4wchTbCdM48eZ1KAeSrTpC2VccUaFWkgNTgrPDnriTHz+nxiGCPJW5gNzAkJESmKRIQlbVwBRqOWQq4lr6WOwIKzylPgnFXcaYy6aFrNkY9RLQUxBRh8h9z9ZsgrAJF9XbOwJ1tQ349yTP1O9Turpk8E0YqqMOfFQsjoApCMYJwp2VtJAb66phQBE2QfxtmvJqX6BnKhfU8xk0ztYaZE/TaWjDKpWIAnxUS7+oAPPv4tmrZnvT6hbToev/8djh89peuPbq1htwD0d2wHlvatF7nsP9DodWEYx5Gvvv4Kaw2rrqFxVgFP0QaXKUeub8752c//GcOw1Y5mOXN1/jXXV68gGXI0eD9zcfEKP094Hwgx4X1knoPSfLUdOUV2wwUhzIWH/6CLTdaadwGcU2srppk4zOQsRFVTiLWYBJurHdvNTH4CrRH8PJOjxyfPy69/peeIDZLsEmBOWRF35zLdKrIZ4WR9xtHqhBcvfsrP/vTv03VrPvrwd2iahlX7lMY0zG5kbkc2Q2C7mQkxMgcFxC4uA9td4un78OHHsNvAV79S3riVVYA5pYSPmu2XY0/fO957ryPlzPX1OdfTQBwcqevAe8w8YmJxeUx9/6IYQCh+eaG0kDIHrdXc+6Y3tOsGP0WGndcxNhlsLlVvUhhyZVF8922/kXRZjQPLXptyCELJgbOqf9zNuV6K99P+u9WUh4M2S0VD14KXlO+eV2O7dT0sU4+7VWnVKlDyhnrc/vw1UhB9YBq2unN1jBhD1/esj45p2pb9ylL6pB3mu5bfb4fR9v+WW/sOgLeyUJkSYJ+nkcuL1wiZk84yrtc0S5OJQBbPZnPJ1eVrxmkglAYaN9cXbDc3qnGiIQTPNA3E4MtCl8lZmXhyzkjQxSPEoBxqxT0To1pHFXoud63mrxbAmLKuqmmlZc95HzPPWRs+JL0vJDNPEIPQWoczRkOhMRYAUYVkGHSQuibQ9yOZgLECkgjpBomWJFuy2TH5mZvNxDDNTCGU8FpZrMg4q/MoeiXKWB1rdChliLOa4QYWEHn2mXGqVYBqfqcE41hSYb1o+WyQJQGsanCp/yzvXmW9zLlcylpn5Tlc9Wrqz1IyZgVypZv+Bu3xLafLqnasaHidBjHVvm+UTpR3gSOK+Q1kwc8lsQaKSaUaXWusVes2ZSSUhrlcKdcJWIQ6RWKKWNHqMMXLy2Ab2QdVKYBdVBCka6ufahdgxXvY3tzw8vOf069P6Po1TddxdvaE73z3e0p7pXdcHkj2ue0Hgr7khtyxx5YEGsoClvfJKCHoxG+co3WOq4tz/uUf/2P8PPL1L0/oS5ZfigmxI6bZEWJkN0zEENnsRrwPpDCX1N7SaScnQgyQE02jed7jmJi9AkdzKOj7pOxClRfdOkNjVbCbRsMhMei7cK2h6UpPvWIFuEaR8q5xeBuxKHFEjJ45zKSUmIJmkD19ekS/7thtI8NmWsK1IvD6VYNzBswOu9rg2hVP+meIeOb4KXOawLwhu4GvL+BXfwbZJKRRpDx7FbB1C6ddRnBsLxqaPvLJD2d8yPzZzxM3V4LLhgZD8pHNuMEMwjjeqPBJZn0sjD6xe6Op0jFCTgZfQDjblpbNojUelTxTFzqjC27KTCHreG8n+g4+el81/8tLYTcJqcmIzSWB6m1O3sPt24+zVzO+/KcK/KHyvgtUv3UateBum7gFWKpI+6348XIV2POl7e/l8IIqUPusrEWbVi1+6E7ngsrmvQBqwYua+8ZYXNPQNu0B8JiXqyxmzp2tjs3+6nmpEFTsQDMAtdgmMc/a3qqxFu8s280lu80l8zRwJRNj44jBE0PEuBHTbIvPaYgxM+x2zD6UNK5UohKp4BFqPrkFRK0TM5FSKGHLw6SnO89T6X8OVi8R0YQmqeBoDUruQ4pKg7X/kQJmVfxF98dl3EWkaH+Y5qjP4zpaa4FAjANZBqwJIJmYwStBDcaq8nHOasdfM+MkEArFlHFC22kV5r6AisO3Q0qZaa77Lc4aolcwWjGfOm9LpaVExO6xIoFSR6BKQHPjWXLlJSZSswe199Mm75XFWzPp9vatF8KEmulWG0AcxtgFrLUYESLaE9tUX7UI1AG2t2yaiAAsgBylrbMmLhhrtAy28MCpDodKQ5STIqkAGFP6cOk1Kz17dT3EKB8agUJxpPcpYlgfPeLjT37E+viEZx98l6OTRzjrmKaRpmkwbcMiwg+VrL01YnrQOA3sxoHL8xd88elPCWEmJc3bH6eAD4k074jTlmncMm2+RtsYrxisW4Ah5oksO4qxSsqUyr5E8p7olbaphj81R11pkSWDc/DkTPuPb3Y6g60ruQ8Rciz1CaZBqNTGGeO0XRFScJWk4KoYg1iNWSv//owZAJOZ50Dy2uzQOk1I2t5smIaSxEM4mD8wz5ri++almuRtO7JeXWHdxGq1wdnEsfkRrT3l2fuXOHfJPFi2lw1N0/Pd735C2zV89eqfc3XzGTfnwstXnrMzw6OzI4wRTo8DziQNG06BpjH0zhCjsNtpokvbtDjrcBLomoj3MJQiFyncgOsjaPvMuBUGrzhAv9KFYZMShQuFnAtffpPwSXj5WnPsd17r5HNWssuqmP7CAXSptoM50JZ1WlekUlLRsVUJyqFyOJSSCvCVQppiJWgiQl5YPaAWydQFIy8AXW2tVLXpUrJpbpvT1afKCS27LYIgYgGhaTpOHz1lfXzK8fEjVutjjBhFjAuF021FvhfmgwFaVPti9ZDxwTOOO66u3vD8i1/g/UjOQxH2yOwT8+6SeXOunPKNLmbzlJT3rgA4mZmYJ50ckknsufW07qDEdgvwEXMxcUv5ZOOEVbc3qaSsiIcYhjFm8TxrPY0VMLYgIuXd1H5wqWIe5R7mELGz1mbn0i+pWjnzNDNPs2rZyt1WhD2W7LTdDsRk2k6JSJt2wtqZ3AiSH2H5kPXKkMQzXDvizYq+PeL9p99nddRxNXzK9WCZY+J6k2gaQ/QNxgldi+IWPjLPaop3HcyekmyjCV3OOHBFsWXFLhKy9CJ0DbQd+FEFWso7y7k810GoJpe5GDNsd7pv4S3OAtnsySvfIe3ffuitaFLttyU4EaIIqXT1qHN/n/xSfe5DWiLN2Koeigp78RKKsFpjscbRdi1t32rTwUZj5b40m3B233xASjOHpnWALBOonhXKQkQpIMnVwa718QbXdjx5+gEnjx7z8SffY310wsnpI5quw97ha1sGo16imG01ziai17m8vmQcB1589We8fPEpNxevuXj5WelM69Euc2X5CjusUeGcvJ6kSToWiCK4Pkbm4DFG6FvBIlirZ0lG6wqARaNToiVNp8kbWRLbnSfGROM0ieT6JhJCxpmS4BFLsojRholiwDSCNaLCm7W7oWQDUZgndUusg35tgcw4Kh+AUnppUZS+J31XJqk1Z6zQrWs1ow6lgo2RfjVjbMSHQMJgLYzjlzTuXHvxRU/wE+boCtudgHyHHI9wODp3wtF65Ozplq4LXF2OWCP4KUIsYd+oOMQ0Z0JQkz9nmN0E4hXDiEIMhuC1F0F15cad4GfDPKobEkJW7U9WduM2QxKktIRWjMqUrEUBo4osVQII0e+9y1r89mmpymov5WVZq75YLKt3XcuWtMJF1spDyt4HT7kiuhXcqv5QYZuxjq7VHuzVZIgpMY1TQV0pZYEsdd9NoyVH3iuAxMI+ugfWYrECqrdlMBijjKWPH7/H2ZP3+Og7n7A+PtlbCQ+8hLu784FLH1Pi4vKCq+tLPv3ln/DZL/6QMA1MmystMw3KDdD1DU1jkewxEok5MfmCLjpIxmBcxJIIKTPOmcYKq4bSzVWwsAh7qvH4A3ylbQxNY5inyG7wCr5ZXbiHXWAYE+vOll59AYnqCrkWNeELNbgmowQEi2RNdgo+ELMi2P3aMg2JcVAkvWmVv28alEbMVJdBIBlojXYMMuU95gyb68Rum4gp0rRgXGKMBjGJK55jJODkGCfHiJuwq2tsN5K5IUXBSkNrTlivE48e3yAxcXWVMFlobEmdjcrlHoMyIKUkeK8vzodZIxexFFEFIXpT3BGdNeNO2zalLNQGkHEseIgtAHDIENWSDAF1r2z5bTJIxtTEUyCbd7NXfMu0VCpU1JRUSu6+qNAbEdpW012V7WWPhNwy3BeNq+dY6HzYs7IIiZQVrY41LmfUKrCaf6i5yIL6jeXeKpZkrCaBaAz9tgZWCmtD1x/TNC3rkzPWJ2d89PH3OH70mNXRsboMi6AfgIIHY3FrW6w2LS+dxi3jNPDVFz/nzeuveP3iC7Y315oFGNNi7VQCQslJu7WkSEja0x20NNIsWEYk+sQ0JqI1NKX75+BLi+iQSCEvPjtSCmXQQiEFBZf2BgtNdN/pgtFYMLamMxbqKavUSequlxJmr5rd2EyOWgQVo0Y7pICrKSvDSy7NPHOpftSmmxTXTKM33SrgHGS0SEWwdI2l6yN9H7Au41YekcQ8CjE47CrR9wOZRMot3gsvXr/EyDUvX51zdb1lTjON0ZDfOOjCZgut1DhBKDXoNXOzcRZEzXlfOsRQ0q6r5alkLOoG5rQH6IxRMx5qH7gDMNiINhwFKKyySzmsZExTQcK/SD67aI54yKgplEuoxmpdurXC0VpbGs1zYiz8dEufryUuf4DaoanzOVEmXEktlEQi4GPA+FA0t5Y9Oqe57vW01qoloAOmvehc2yFimSfN0srF2rCAayxN2/DhBx9zcvKYD7//23z0/d/m6PiU99//mK7vcYV66UGr6i18bm85zH7i5euvuL4+50/+2R/w5Wc/Yxon5nGicYZ1p7ax1uMLhozJkTlowktImTFpwkfTSklmSdgYmMbA5tpjrSldUEVJFGPGWVFT3Bqa1hTsQwV+GAIpJazV8KNGCNRsfHTSkBP4FIgpokRiCqo551ToUaaF5CN+jNhGcK3iN9Po8T5BbYVdGlIokK9/x6Cul5rxZnHzrEucPJppO4AWskFOeyS0YCawHusiR6dbMonXL9Zsh47uycTZB1v82LO5PGEa4PnznzJNgcvzid3Gc/ooc/YEhgRXV6pdE9zyjc1BinZXAFgfMn4GkyOSlbLKllJsaxNidGEtk5ocde6uelU2w6b2IVQX1hqhaXUB8CHoWHuVnb5PNB16zfEuO+3t7TdgxitIcxhNV7BoD4xl9h+K1Px5JSpclq5F2WpySi5mdaoJJhkoIZoYooJDRXPeK4BF+KzVXu1dv8LYlhQjfp4O3nCpUjKOo5Mzzp58wOmjJxwdHdP3K6yzJTRymOKwB1veNuf3O4Zhy+bminHccP76uYbQtldM4w4/ayzdmlK7b/ZYQoq1n7iGeFJWjZhsRkgKXmbl36/8AADBR9UyMZJCJiulL5oVqMfU9lIV9FStX6DQJQalSR0VZJFlJcsLb2CBAgslUAlZFqFeANLy7mqKcqoasCC4FVytiO7C7IJVauZCYSYkBcYkKYOMLVl9SInOJMSkoiUjoI0Ux3FmGALTFPC+8LNHKf3YSn/1qqGpGNPh3N3/CArkKT4lC7KiXGFKT+0oHHJRoxwV1nGNLmraFhrEQtuq+1jv4RAQrViAtn+6b3KX8z780f//m1qG+7sR0Qc0VtNbRYQ5ghRuOGelUEFrPDgOpabd3E4eOERkJ7RxREOhIcar1jLag71ycCvYViYFCgpZY+m6Ftf2PPnwB3SrE7789JcMu1EnqqQSrnO0zYrf+ku/zw9+9Jdp+5Z21WGtQwoX0dtZynuBf3vTzz775U/5+3/wv+PnLSFeEOPEm5ev8LMCYpotJvignWarSbibi5lbSnVj1viuTYnG/n/MvcuPbVuW3vUb87HW3jsizjn3ZlZWWS4sKIMMEogeHToWTbBo0QIhGkjuIiEEco8GNGgBLZAlGvQs0eI/sGhjGdFBtgrLRVW6qvJm5nlExN5rrfkYNMaYa+24dW9mWbZu5U6dPDfiROzHWnPO8Y1vfOMbK1MWtrWylMqyNBMwSUP7zUuPXgTOZ4iZ2hq3pXiHYebe7KJVI5PMW93KlSkZAqvNFGjRF66grLcNEZhCJYpt9oS1DpdbZasm0qnV0q2uylbMNrwBTeqxVmT4pnvO3KCWhJYLRJjOr8RUbJOHhS6d7uXX2qwhJ0+Nk1YkVNbaqH1F+Uxr8OVT4eVFKavNgrvdIL2YLr52OxaGM4pIIPo0mBgFCCaG0UDC3uTmyrrWO9WFP9H95p/ewXwWCyI92KGwGSKaz7ahP/9Sqc8wn+DD15GyKV9Kp3bowcm7FlhXIZ/g/Vfw6eP377+/IA+6g2E/4Kz9f+9H5LUIavmMOhTYD9C73xnYZWz6USPem2CqCRh6FAKCpOhCjoPlFwxBxBjIKTKfTsynCyllj1R2Q/cIIYHz5ZF377820ihBCBbVxzSao2433rC66OTevFF3uenzl4/87I//P2q5gjyjWtjW9c6pZ3ym/dNaZPaIPWh8k++6VFWN+rS6tkd29fy5dtcP9B31jGvYajNfgBjvhkF4tu7qruYOGSF6iVMP+W7wA6I1g+YtdAh99w40tN73nLzrkdb1e+syF/WEkXa5kOm49YK2uHOoQTo9WMeiSXAHErFSYIjdDqfgOa4qeL2+VVNmDj/C1g0e17uuSkMTdxFcPYCpd7F5Gc3WyLj9uouIDLXaDLecfYko9GrQXIEwgQY7NBGcqBS/vbZuA2JeeGrIAMwY6DcmslvO0XxQgHUN2Qx20z+PzSZi+d4cXLeuthHns+eGpaFN94NCx5P7hRmL0yCiwTRyYsqZEDNPX70nTYm6rvRSiNF6vWOENCkhVsryYm2HdbN6fMyE7F1RrbFuG8ty5Xp74d2HD7z78N6MJn3IRHTjy195PYDeG//vP/y/+ekf/j7f/PE/RvgCurIsV3q3iNc7++jp3jrSzcBgHCUDAUf/DENPrR1eXk3Z1WqjVxO5iBi8XNZGCML7r07Mp8S6KbfXm9XdPfWx/Fv94MBhqEPaXQhlh0XfTHxjnmmBUrvl+laDAlUenzIPD9lESWUQVmMju2FlEsIpmszUFWQ4atmqve7DU+SrryJ5ajzfXli7Es8FScbVdBo5KdPsh1G2DTjFTK8zJDvw1lbp9Yao8vRoXu7LYlbQEpTlapH4w48MTXz+pVk+19C9QSoQmsmmtVkakU+dmDpprpz8sFh9GdIEqvD62Qc/jMoHozoE57PdyynD+QSXs3B5gCtKacpWTQeRA1TXrZQVXn5x9Mp/1+MH18YbHGXPw7zPh1jFoInn6KMkRDeICpBzMJln7XsO9G2wvEdWRgOMRdMWPYqHyPxwZj7NrBKoRO8uEiQ0YipI6NSyoj2irTqrbI6xqp26rdTWKGWjbCshBE7nB9vkbiZ5xz6wf6UwRuoOWNy78s3P/ojf/wf/F9vtE+gNKO6B12k6iDKPpKo2nRaDlvb0B7614/FQJS6rD3n1kGmOOeqa+mbKwjkxP0xsZWVby5tEcDR2WGTzA9ZzyxBd0OQMtaqJZVDjPuo+btsMRrQr8yUjKRqp6MjiyHLZo1dIWOvnIKrasBE3+P4YhMtDREJnLYWGmUZklK6FrgUJME22WYM5RZIloi3TXABTdKE3a16a56P4K5u9TikwR3h4sEv4/IkdiahCVGG4IXkt2MqcUyPGjiaoltlZee1qeoDtZsxV9z0gwcqUCThjtzMlMx3JE+QZ0uYpWofZh38E16z0CmsF/X6/yR9eGz/q5rY4/NtdKaVbmSaOSZrddMTdoTigZSyo5ovQD4QByCWYr3aw+iUM91FxW2Bb5L3a4L7TfCbMFx6e3vHuw9colcoriHCef4sgM58/f6bUymmaebi8Q1G2vHA6X7i8e8/D+/eoCC8vL+ScufjY3zDyDXlLCA4yxaa6+ky45UbdbmzurV5rxaaR9B027h58qJUEYIeFySeLIg6Hu2nbQVkXKBGSWPmnuYvp7tuPElolFoFqFlw5w+kcHD562ci7DKfJRgm3auhlqN+0+4F5Mog6Nvi6NJvE49FLW/chCEZyhtBJ0RR8dIOzg4BrvVNdTBOcGJ0mmMSqLr0EYu5MFyVPavyPQCNa+atGljUTg3KeTRloRGelEWgaqb2gPhY5TydQYdk2oFr7aRGKjk2OT4xhJ0JNdOQXM9phWIvQe9rzka3AbfHKhs90y7mTol2K2ozjmJMdnNtVKCKsVygbLDfl5UtluTkznzxFUic31bgAayP+/u33597sYprQ/xP4qar+DRH5l4C/A/wI+HvAf6yq2696jvucWnyxDla19k5ogTwZUmy1e1/G4RM/fOFGAW7vFPIcKbhyLgShOHEfgnnOxRi8jNOppZOS8vD4yOV04se/81f4S3/lX6H2jS/rL1GFc3oiaOSnf/gH1FYJMfHw+BWgLOXK6Xzh8cNXPH342nzQvjxzOp2Y80zQgCTfkd/SwHvaSe+d27KwrDfW5ZW6XinrjWWx6bFKQ4KizgiPDWVP9/Zzp2x12Ob19d4Nhqvqbt83T7YZh0uQudPYQRRrJW4g1Zxb0znw9MFFMH46t2IlonkOPF6seyukYKaSL42uSs6BKRuyqLVStmplvdKJ0TZb93QCIiEkYuzEUEgBaJiPu6OS1pop58RmzwcJTHMgZUsT2mbl2tMJ8uy2TIBqcovlmd4fyKlzkhshVOr2SmkblUQjUXtFkyIamWazqQrXhlLRJugW2DaD7iIwzZ18VpZF2Dbf7GPkc4oogbJFzyqtdXdZ4OWL5fMmCoLTDHNW64LD5haefMz77SVQm+w2V7fXTkidVoQYMiQoW/W94dmVGAL455Wz/2fA/wO886//O+C/V9W/IyL/M/CfAv/Tr3oCy8UN9jj/whjwN+C3divLjYgxSjMHKhjQ1f7ytHH/xrCkVr0jNFxILW7tU7eVTZQ2T/ScqHVjKzdaX2nrF1DFOOOJ+XLm3Y9+zDyfaW0jpczXP/oJl4cnTvPZUENKhBhJMXqvt6lI7iP8brHcNrb1lerGEMvtldfPP0PbQpTCabaosawjFz/69oP4IMGTPedA3F7CpdXOthnZlLORN7W3nfTqXjcSvxkh2GRSO3AtXVJ39V2X5qjpjsQUt6PaDCENp6spCy0Y5LXx2mI+dVPgcgrUZAe0iLrSTf1gat4MM9K2cRh6s8wQnGDtrZbHRtIUOJ8jj0+REK01t6qyJkNtaCaHTOvCrS6UqiRthKBsNdB6ouHNVqNUCLRe7LD0Cy7BBSvdW6U5IvpOIgq7Rr82KzG2Nu6Zaf9bO8rHYt+mVkchHrx6Exv/pHaIuCZqN7FcF0yJNwhEf92IdSEcxhXfvzAS2NUAACAASURBVNv/XJtdRH4X+PeA/xb4z8Wo5H8H+A/9R/5X4L/m1232IMynRNm6ne7ecJ9SYD6ZnHDbOrXo3QIdjO6AcsfzaVdUrOwzRAtWvnFmFEAykrJ1vrkM6fbyme0mnLIpzK7XT3z+8jN6fWF9/ce2sB8bKT3x/kdf8y/O/wa35y+8/PxnvHv/Ff/qX/s3+fDVj/jw4UeICKf5xDRZTX5db6axZyaESE6JKJHrcuXT54/cXj/y8z/5fZbrF/7kD/8Bt5dPRgZuV+ZYefhghNq2WW42Uo6ARbHLKfDjrwymfnnulDrYZihr5fnLyjwHvvqQUIXPX0xnUCXsTLG57HhvQAC6Tb/tpaOlsV4bTYvlztFSpfMlMuXAtllpKmX/HsIskd7g5x8br9fO02Pk8Zw4JYgtUWtnqSbiyVOkC2ytcX01T8AQA0msWScGKJvl+oO5FhGyy3Uvl8x8Tnz19cxPfvvE7XbjZ3/6wtobZSnEpHx4euDdwxNfri98/PJzBOFLPDlxOhEko7LRsYaaPEXosKxX1oW9yhAzSLR6e72addRoY/VxbBDs51pXtluzcmAd6KntASdPWGpS1F5rsecKyTZ0qXB7ttcYw0zPl850VrYNvnwKnlYc7y1GDEVivQ/fN8VnPP68kf1/AP5L4Mm//hHwSVUH9/dHwF/+rl8Ukb8J/E3gkKd6GWPA0D1sDLjutXQjeSwy3Tm33f2/XZSjzPF2WOOfOeU8Svbm5ahmXufbtnB7/YK2V8p6RYA8v4IG0OJSyMA0Z6Ypk3MgJqHVlW19JcZHRE4ohVLMcbW2DQmB83wh55lleeH1+Rturx95ff456+0L6/Uj6+0z9A0JBQnNhSG6d+2B+Kgk4x+S2yIFNaFFjOLiEfVykkGg4RAj3+IMgtjvj2aKOO6FX7chu6zFEFca17UHl7B6rTwGa3rxCCM7wrJOuFF/nnIwIZB4e3MYHYnskt7kRhejJDv4wSHLBXHyM+zvNQQlJUvHYjyQoiGdURURTE9g6UbwtSWihKSmt2DwkcaK7689BovsC3kQq8c628eH7X/0qAzpWH93Rp9ipOb4bOrkXuiHaEe7EXZjFdt9D0RJhnidnB37aPzgva/C9z1+7WYXkb8B/ExV/56I/PVf9/Pffqjq3wb+NsA0RbWROcHhiS0uFMrW6MpuqPfj37rw7v2Z19eVz1+uJqpxJ5F928vx9+EnZ+z+OEPQhjYfk1O9bn8yWL9tleVWWJY/5uPP/hAJjZwXYgrUouTpxMtz4fpaeXh6x+/8tb9KnhJL+Snt0zd0/cT54xO/9du/R55+z2ax//KPKNvCy5eP9Nb4yV/6l3n68BP++A//Pn/w+/8Htays1xd6K4g8cz5vQAEt1jm1wXXtvL7afLF5TsxTIPnmOZ2MqU4x8PTO5361aoimFz5/ttw1J9soOYImyE5+nubA40O0Q8UxvAzfCgGid2GtVufN7qjSWoPVvOKjmAzp4jT8Tbt3ESrzBFMys4sYhHCZ6NqZ3Xt/3Rq3W7GxR4tXTx58DdTGsrRdvYdalA0CKSfynCi1stWNy7uVqq+EqfLhtxqtQa0JVWE+T8wPJ87hylM36LvdDCG+3sxv7qsPyvsng97rQJPY+V734YxCaz4fPtl+H1WUupob7CrqZKogbg4pavX4KCbdbs0akyQopwf7u6xCLbbByyiAjAK/r+WuRt7N04X3l6/YSuHjpy+0bj31ISh1M3LbiD4/7L/n8eeJ7P828O+LyL8LnLCc/X8EPohI8uj+u8BPf/3OZ9+s4q4c3Y/k1v10Guq5FDmdMlupXu/1GqKOiOXPOYCB//E0yv9hTDwxONW9OG0sv5hxZa20eqOWZ2KC+WITZOfTM6orda30rRHDA4/vHggRan2l9SvTNdHrwtP7n1BrYSsLt9sXtuXK50/f0Gvl8vRbpPmRl+ef8/njH9BroddmE0tDMekmBai2uLrsZbbujRc5sWsBUvLDLcL5bHns5kYGUx4iJBOTgEG92MQhsjG582Rtueak66+135hBIIIMoQ7CcO2VADHokVYxUMFghF0l5jdJkluFBGOPt+Jede7Hpj4yCcGdfu5bi9kj7SjFltZp3a5Vx8ql08n5iMX83UIKe6ddzvaxCva8pXVKbaj6tXKB0BDqDFrI8nJxG3OT2SKu1x+Xxa+TDQl2gnCYn6rp4aPYWhvc0kAUtfga7kfaMBb0IGAVa6oRN/dEjQPqciBk9QOx97E/vn/7/drNrqp/C/hbdu/krwP/har+RyLyvwH/AcbI/yfA//7rnmvcvD0i36nj1E9za8AQltvKL3/eWbfqpMSvxijjXwZpJAM9OdljTTKW/3U/Mm+vL9SyUutGKQsxCfMWSCnQ+iemKXKaJ776KnGaX6jLT+lduS03e43yQjud+RNWPn/5U9bbM58//dQ7oS6IJP7kp/+Qb/7kH/H5l/+A5fUL2hvazXHVGGo7cHpvoGaNdLnA7/5u9OYeJUiltUhrkZQg5W4Lphk/0bSjdC6XwO/8ZD4iuyhffYh0xQdrRHIU5iyUprxcK60ptRi0rmqbtfn0ERRuV6uFp2RRa85CnIR163x5sby+emn0w2PiR+/MUSYM8gpbsM8aqU3Js/CQO0qlfzEZMM9DJQenh8C6NHNuDcI5J0vnymaLPVdS7giJuk7EBOfTjEjjcrkBlTg9s/SNojeGZr05B3Q+Z06SWNbGH/1RI+XA6TQh0omykaKSZ2O5x4HUnZRUZW+RUEfoqtCKV4J8Q9dqaYpG9wjA/PtAWW+Wspyi8nCB0mH1kmPd2LX0qH3dG9Rl5fblF97BaeKkUgHX7Y+0w5b8PyNB9z2P/wr4OyLy3wB/H/hf/ry/OHIRXCesYkSUbQBzk1lXs/1p6pqFIVy4fwxGXmAMOtibEmRkkNadNQZBmBZcPWW4UYq5qw4EUXomJUA2pgnO50ceny6EeKOuv6DWzsuzDSs8p5XQZ27bDT7+nG175fb8DSHOvPvwV0kx8PHnf8S6vFCWP2Fbr36yec3Zy1/bZqx0isI0KSkFHh9Mk26jeTvrElhXj9TRcvI+yEibhcrpJHz91UTvjd42EDidvF7umyb4ZdNF2VaDgLXj03KMSUd1H2O9LZaH5tyJySJVj4GC8oo10tj7gceHzOMpUwqsiy3uGIWmipRgiz4LSSLLovS+ekRqO7ucT6a8UzXOYJqSsdbVjDCnrFhDoVBrJkZzj4mxE/ICoqztytZuNDfJGMikqzDPiZiFL7/c+PKxcXkITPMEdEKoxNDswJtAm6UmZcMmrd5Ff7hDp2MC8BBzNXWbbOMBzPnHru/mBN/jAzxMsGLpQS2Y8EYwG3N8tlyDtW70uu1EJcHbYHGUOt7Pr6q78U+52VX17wJ/1//7HwH/1j/N7wOoHB1oKQkpW35cxRVJ2Tersx7DhfS+7CZuMi93cEZGp5mwWyrhUfzbZ8SoHXe3ZeluCmiHjmnCUwxMk4BUalvs+9eGOlkURKh1Y1ka8XQmxor2Qqs2yfTl+Z8gknn58oX1dqW3z/RSCaGTUycEh8pitdRSALVFqwoxVgLmlku06oL1/fej7u7MXdeGYuOg5wmX19rBlrJB9hSGoaHX61Ef5xyopdpgB7ej2qGhGqlmh7MRdCsB6co8KZdZvLFDfYyWz39LEM8wrMeq69R9AYBAipHLJVOrsq7H4E4rQ7npZRdqbeRkark8CfO5k2aYpkZrz14Ss66y5dbpGuh9QjUjrZDCQsM2bGvKdK6+Zqw2Xmvn5dUmxpZqB58aKLJ1mmxNlaJ7fn1o9w8CDWBzprx5GpB8HXW1Ap/qSFOF2+a5P0LV4GPHPHXplt4O3jTEwyasG0tn3DHseWuwtr9/bnX2f/aHkxkScQuoyHkKLJglD0DMVvvtxaK9VqVu9wzkaESxOqV9UNltpWK8b5W1hpMd/Yu9h50sbd1rqEdzSasmj52mmfMpIFJY68a2BG6vkZQy7z58IKZIKQu1dE7hwikXtG7U7UbvjdvyhdaVzx+v3F43UqjkWE3p9dhNOORDEmpRnzhiZFiMndisNDlPyXrvI8g0zB7ss7SgnpaYo22K8HA2yLoVE/VEX6wpGKtfu8ltu6iV1Vqnbp1laaYjn/AD1spJpQ12vBOj0KuyLsq7R8gfnEMIltOmABFlyhDO7tW3WSfjfgC7wUTO8PRe2dZmpbahlFQ30egKmFljjJGn9xMPD5Fp7qSsEK6U/mIMu1iEfX45sZXEJA9McoFwY4qmlN+8bfX8rhBC96qH9Yd/+lRsbaXRCGS7LGZzlS3JjCrGferlbkkHI4d7h3UznYi4FZcE8ZTQmpWG/VTvyquYvNUm+ybf6O7oO3gEd7MN0RBNr7ANcOgOT+L7wtyG5Ddos4/HKP57eUOCRXkFUg77dBcNwRZZ0Tcnqf0tg4ODuw+5e8SPlxoBRc1YUXzRj5/p/l+D9MrZCLExkztYqnhnJsBepuneZNO2jZZeaW21mwy2YFsniMlBg9muoF1NPhqElE1Hv7OL4029IaiGuKjvyjcd11CH9r/v0WOUf7590+1p/XN6Oc4ISnXkdKCn+7KS3n3dgRxtM08JcrQ2ZBumydGTYGe6l8HEp5JaedAmvXok88aX0c8++tjvqRnd/2emIjEL8ykgMSEpEZNStj7gnc9lV5rbYfdmET3lbpG8KLerstxMqCIRP7B0XxfdRypZJBZv67UDcKyVQSH1JnsAkZEjcUcyiuefjib3UmU09l49hQpBmSZDU63Yet0RK575IcTJnrt6mY4R9EYA+xWPH3yzBwQ0gjPPpdsHeniKnpPkI3qrcH0tO1lSq1+0fcPzLQ/t0Ryi+/U1aB/o2qz8IeKstdAGNBYlJ3P7fPfBYNOyFG4LzKtJNGOIxDjRxSalaB9dBwr9E/22oBHSbOWavpoF1GkSTjlSt0LZVnpVrs+Wyz69PxGzySdDtkXQFcJuHG4dfoXGVgKlDsRinG1Q0xg0Nd857eLGBsZPAJYK4TEjGJyestlMLUtlWbrB1wbJdevajg2o/vutQhfl6RL5+kPk8RR5PAVSEiYftSUE2xDNuIAogZM1tiOyAY1aA6UIpVS2rbJtjW01p5phdNG77toC16Sx9ZXUAx8eLrz7MDOfTNRzuzW++dOV1uH8cGJ+iKxLY9leKaVwWwz3Pn0o5un3C3OffX2GlxfT+j884fPU7NTdlm497MW06Uf48KEOCfpmpTNLBxshwnSytVhXOwSUakBmHNLqTTnYWpsmk8SuS+U0w4efeFQvwnoVQg6EFIy8WyFk5fKVMXLPHwPb4hqF6D513zoov/34gTf7PXHmJ2O3zSvxEI+IHI0P97m4zYe7l9ccLiHfFcnuvzKRhDd16FHuGSUOO4UPh5k6+IIKLdlrpeQpwhv04N5ubFirZtzr/cD+mZrXbEzn79FtFzbbZ9uTtHFejesDHtnvIIAcKGb0D4goTVw+Ke7gM35Gx+Ggu85jRNYdMYwNPhAMx9+DWU/O5k/5KOdFh7Kqsl/b8X66DKRh5cDeO6XKGzQxojpOEsJdRPP3V2pnK6N3H1Ax/wCB3pNHvohIRNUm/dhsOhMbZSc1WxOfpupDET2KGyQWF1uJNxQJtR7CITjWp13AwyvuiOR+sTy6j+648bt2BsveKTiQ3c6XjF/3iB3CDuLGEv2zKFCPL3/V44d1qhHcGNCgTSmBbbPIloZX2uaGCq6WKm5GSD9y9JGTh2Bjm4ZQyb5rP7RX3jjgaO+KiliXkVodNgZDDatLdFOyBWw5F6CJQCKGRA7RGO3o/ECwxWV2v907yMzxM7mkzOaVFUppzkDbsMMYO/lU966l4BvGoLAttKa4y613THnEV1xElKyhdVkLW6nWIJItD8zRxk25C5QPyhBCUKJ3vuUELdt72rYONF/k47rp3h/99Bg5zZEfv8v89ofJWi+THSoNM56YYiCFSKlmedV7o5RCU8hBeJgTnz+v/PzjRiQwheCNL8f7VL8WU7QDYV0tzfjZP+nkLGwvV56eCu+/Dnz9k4QQeTjNRoC1RL0Kt9eF11u1foRiI6LWm9B7pBZ7/pSEebbP+PLFBn+laCIlDYA0k+iGZOtVTNpbi5W7jCiwAym47qCPuetdQYXi8+lCtApCBJKnsDEEIJBi4zw1ksLy0cnVYp7wRGuA6V3AJcm3ZzPhUAMNxvw39SEov2E5ewgunlDPQ5sQVIyA8rMpAD2YgGNICFHdoy5gUXiIOsYhsH//0NG3PVIdedQQTbgRiLPPdhiVVejRWOwAjIKquEtN2DkCg0/BeioZ0212QdX4HB6tLA1xEYYv8FrdV/070Im9X6GMXFGPaIGCitCD4Zxauhs22uvGoPvcr+6/q3Iwwqi9jxAG+25tsa0pLagLneyiqrpvfAycpsh5ijycoslhXZLb7kJX8B4Fi9iuIVAQyeRgQqbrrTDHZDbRXRhzzeDY7BLZBUa9K9cXK/HlWKmbIjEzP+TdgFMQ1s02RC1CKerz55q76tjUGXUiNgTTpLdixBoqNG9cinMj5NGcZZur+2eqxe/F3mg1+KPD4WZE29asvyEj5Hk49IivH7OwCoOwBurCrgewWzA8/8J+71vxQ78PWbUz92Ne+2/SZh8XQtSianSINPzUbPPJLkzYNnNosXzeNvfuh9aPemxwiG02RJbLhCCEpqbJ7iD7ghrRRN0oA2tUQFkWrx1n9ydPSgydXiptM3WdqNU7H6aZaU7UYn3XrXe26+YHmd3XkCKiJ7PF0mIR2h1s7+fRDegn0ViX1vxrP7RML3YvQrcS3E5yer23rGrjrIsZTaZkKKVWg+kpgWQjK+uYb093gihwPoGS6czO4m9e0kvkFJAopkkIDPMZbqVZzjwpU44OuU3CW7fNDFpCoxFZl43lWiFD1rAbkSDDf92QUfDOsegb8HS2XFmicS3b1lmvC32CmG1DbPWBVrO91+nMsgivr4XufQOtK/XaWde+E259R96WjqnAHIVpimyLueoqDXWvgLrJHk3V0WGezdu+uvOrdhOG3evpezOs2bzEliTYfPXR1dkPUrOr+cPjZVIRtQoEAQ0TKFRdkaYkMXRp/ght75D8rscPP/7JWUXwOq+f4hLuGWSlLlDWoUbypgUXxeB18a7mPhpU7OIN6C1WckoOx7U5KdpG7nPPrONlI7uJ62LDAmOzvvh5UuZsthabNKYcmXNCVEjBhlCg1fO/xnXZUFW3qra80oY8VnqPdgg09z53xKK+oRUYrv9ao8N23Z1MxUuJjGhcBolmm743dSipFNHdzDM6C21ySvN3Hx7wrdkqC8H60ec5gEwQnkhReXe5EUOnehohweSmo6OydXi5meKtzp15CpYrd9PTl3W1a56VLoF1LaxLQxpMMlh52wBtQFJvmBEgiplhzLP17YcYaAilNtabjZJOjwUhsNWZXhKnc2KaE0GU5bYg0qyu3o4W3TFxdowUsE3ms+RjZM6BsijLUlEvVSrQi+5oBLWGrWkOfnh6ILnPpcWO7N7McWir1XmQThTsug2puJdU03zk74Nass0uaJi8slRQaaRoxPG2Vdal/AYRdHeba+iHu1+0w1vadnytljsCO0n35oPIm7/25xpQujdzAx3lH8ThpkP4QWfsN9r12uMlrK/eIt+22UWdpnGKWl11K42wVkqpBlfbcaxWt1yKO/Go5uSSA+nR3FMm5ymM6LKJJfSwL5bRGWU5+pAW26a1A0K959wWy+6cizDsrG1DH/rprgH1lGSazA1X/P2eTpHTbHXhra30plxvlSD2b1MWopgLrDoiM5sk91mrYpN9guyWVSQ7JJpklAQ0kGIcS8Cjl+5GlCGwD5ZQ2CPV5qKYrJAU1hs8R2UqimYb6LG9LrTWaEBVYd02HwklSLID7XQy+LyuFuFRS7+DQJysqtA63BbrN4hpSGfDt0g4P+yasi5DBKQHwYYHsGj3sdQhO3ZRawdVKweq/6zxVkOkpDtZKNh7VhX3S/D+9ZD2yUet9zcNdt/1+AvwoLPST0rWWbS1u42K5+3K3qQvBLeqOuD7/UYdF3Y44Iw5bEWEpoHkk0+qmp3vsLmyJ7CLVqpNJd3LoViKgFo+FwTO58jlnIkx0Lvl0rfrZhZHzTTL3aP0GGmEuj94jARRTrP17X/4SpBgsKs3ZfZ55q0Hik8DGR75ezpslu6mpV4tXzudzAlFb8Yy52DDNsyxxxbnbbHGE4L7nHXoGpEQuJwn9KTE92I6AIdYy6psz6+mn3/pBIHf/W3l8Wx1pNq6V0Zsk29N2KqwdUE24XIWnmZTn2UJ9B5Y+pneZ0IohHCzQynZBzRnHfVxYL7Z09EQpGq2zgCXpjDBa1Fu1850grUkJCi39YXWO49b4HwR2ibUEkzdlyoxKk+PJy6nxPPzRu+FJhblYxQeHq2n/uWlsr6aR/90Mka+lNFHax5+aoQMW+2sm5E0kg6GXQSI9r3elFLMDOTx0SS+dTPHHvVAF6Iw56GAa76mDRH4RD5ag3VbAThfokugZ2KaqerGIL9ROTtAl10ifsDpg+ThbhMjby/gIOHA5K3Hw8k7h04Hs32Qc7vVMeP5df/PnTu5I/GEw/vNRDZmqGBjiEyfbKRb30nH4zFey/7EaJZOk4t2RIQW+/4xd4uh+2uBf+31cyOJbJPJXXjxT7l/rRy8xIDG4piwudWXiTbsd2IYE17Zo5eZeB1WzN0jl/1A9xLSsIE6bl3wQ6Y0m6ke1JhlY/nbfq32YZ07Obknz8fn1kMqPdj6PV8Wq6yM4YnCkIya0u36qvQq9GLkVXeJtn32Y4TVWF+GItwlyVGQVX/eIsC9J83nrulw2b27BXsZFRj2dMGvMeqQvbsbjd6v62O9A9C8i088rZFjXeDrRKQTgv05cofvfvzAQyKEKIFWx03aMfXOxO9R/A7aIXbD0s70OtQfVtQh+Iy3Y+NOKZFzphSD2a1ZfjpUZAf7bVEvxJFNHKSKdqU2ZauNflXK1pjmSEwnskaWNVJ8aIARhey5pg0qUBDzXjtfAvPTTBAl+wA+mQwFrEtnWX0BDQjon7NuNvvbFrlvOOkgwVl/W+hjQoy5uwYnfqwhpXXdF1PQzprcgrSz8wASog9eaEhXpqAkUbIPEGy98HJ1FttTkin6CGL/35wD05RZa+Wbz5aTXrwN8VYWSltZ142yNqYYx401f8B9Y1vVQn3gRFntgB22ZaVY70AGpuQWzt0mtL57CoSofPOnhY+/sBlvebb7km/B105FtbBu+3a3a92V66tdl+J5eY+2+RF15DOktIGUmyEt50JUQatf40kJsaEt0Ir1f5zP5sQ0BmzUzcpyMYzqia9FQHw4Xt+gbYpkJWQLDNlt0q1T0kp6eVLWdbO04DcmZ8c2WbfgsG/a+2T8TRT/jpLUWLT7YepR+yiB+OKT8K3Ifhdp777fRwh5k8PrESg9sle/kCGYzDS6AtARlxPk4k06R7oxPsvwZRNsGqhyJ89UO1SCN6sgsg8PsB5/i+pNjmaJcdl2AHTvpLLnwexQeBf39LvNfxeQ9lRqlBDxyBb9EOlKcaitQYl7Qfp4jjGXr3fzx0tiTrCCkZetic2C0xFhxy8e0cwuvaGK3X8QF+3IcasGLxHu8tQRPWsRblczoZRk19tKp3jTkO5cgPj7hrcowpaFy7bHvUQYcNJe/1vOwd/6LxMq2e+M8tkYedXvUO0OIHS8IfY1PpaiOJIRfycj/bEPYrLtt6vjzz5+8PFP0TGTcnehPIoJwxABtyqyfzwiud/4vQNIPbKPyapmSDiQwVYq3RnT5lAcsZpxCGJSzeLH8Xjo3Y3CNkdw2CiYJnnb2h7JUd3nhAsR6QlQcrSS1pQm96Er9GaWVU0tFGytuqBFqM2ipIrlcL0aR7HVire6+0h4ex1UTBOOQezB+qqaJiCm4KVGP+TCiGDWJx3EKhAAt6uJe4IM3oN96qh4WrBth0od7VzOgeli/7ZWZdsE0UrdAq9L5eVa7B6eTmb02Bc3v7Rd1ppyWyrq0tpW+6GqixjZFLzmLkKabGOfL6beO10Cl6dgE1onM9n8xTfWJPTpl53bDWKKnKfJXHrF0JSNaLI24iDRjpWTTYJ5fbXmq1GWs5UYTT48fOX8kCpqzTH75B1Gn4XV4ptLadFO3wwdGpdi9yFlRSeFkZYAVA8WYSgv7WtV02bsve5AiJ0QlXUtbFtnW7txn79i//3AXW/uIIu67+Z3dKUxTk37+Z2ZHiorYETMPYVkMLiB5KOSbYRRf5P/d/diHoihVquN3/e6Wyr6LVa+22bp6mYR1UirXjt9aAXwTdh9oo0PlIvB1HdC9YVup3HvPqrYBR/NW6/F04dWR1S3XH50yQURkuc2zaPw7jIzNnuwkdLjpN8v77gOzbQJRNvRZVNKbcdwQW9D3VNEtYN2+JQ3VbM2Nhd7E480ZdvMwnpd7E9PkS0ms09C6VoNauKlP6+rt3stvqpJzezk32WzIbpbT7bZZ6ezcL4ES2miW069KMtNud08FVBXEsZOpe2VGnDLcYn7Ra8B5HqUZHv3JpdgJYfuLr2j18B6AHy1OkwS/He9ZUI8ZGtXm34jcJqibfhso6i6mvLOSoB+vHQDTQOhggcqD5Lg6WKwa1irW33rr9rqfxF1dt9sDmpQVbfvOaCZLeVD8tqUvSVxrN7ej2XctbNtxcsQwSN532G63Zwx09zqvwOiGal0Bwc9iRuON+P9pBiYp2QGFy4EmVokem5sb7Iy7Kq1WlmFXiihEoK1zhoeN138spqibFk7W7E68jRb5NEy3r/YMiq2iKIvbruGdvvGMI0R/Xs3CeuoUDjVAdhbChjP0MuYx2PGk9tq6cTYEEHMocb86oxviWor8ZQCE8nLm+YvV5o5z2zFvPlFhdoqESGlCSQxTQvTvBKC0H16avfKx+U8kVNkqYWbtX7t92aeTTxjnnhW0355mX9E2wAAIABJREFUtpbgNBsPdH0tXK+mypxzslJhtomt5jh8GFbWCr25U4+jilYPIlCwQ3Hp3dNE2dWGDATGkXKq6xzuONlxxRGONGokAdrG8WCtwapiGglP4SS4U01VCPYnhsD5bJLe4q3J/U62+4bA+47HD19n3xnYkZMc9ejjx8QbVjBIK3fwXXGrad1Pz947y1bJsTNPyYmUvjeIdB0b3J7HJJSjycZrrA5bB3gYffHDaCPGwDSb3nyrndCFs9d9hyoKANnsv7wdsmxG0MUEKWMwspiC7XYz7fuymtBjmm2SjShHj6UjhGFV1IKi2SBh9nlyOlxJPRoi+Lgk+zLebXhBEE9+m9o1CU5Svq6dl1e/H+4NcLEBqd6VZ8ORgyqnFJkkUdUObdP4F0QLWzHJqqjYdBvMDSamwDR3ppNNMLRhjn0fGfV4nnm4zPzy5ZXry2ZRLrqo5mQqx5ytr77Wzu1mU37PmqgVXl83Xl46l9lUdHPu5FxRb4BRb54RFXovlGamHctyNMX0wZ5j5d+tmaVZmoKhqzbCNvuf4KlPdwS60wjfueFHVcBub8QFXd3Kv7Z2nKAbAzOkG4k8w+UhEUT4/NlSpnvCY6QT3/f4gWE8iBzjlsPevWJQB9mpDY5Y9B0kyN0paR/XQr5iDqWj/3uQPTv34RflDQE4lHtyd//ubuQ4UUzfbpHQ6sxwu5lyKk22kQ1m+qJwLWmeghtIGIRXn0jTq3Vkje6rPoihdpcrO8RT+l3FQbwUJ2gw2e24XqrOMQiEFNzD7hCDjMmrBauTj98cPuejDyEkkwPH6OYXAda1UZtymeFxDogKy2J19lqtjCQOc4XA5L4ExaF6zl616N1Leeyw3d8e61b39MpSKmt0ShGyaxFQg615gsuToY7o0X+ahbmKt4/amCqbz6d0n6AyCLHR7XYIjmxRhCEtVnZ76ANJjnUie7R1B2wLEuNH9v3n8m7uehroO9weCMoq9+IDNEZJVHY0K74o7fDsLqjxgSsqoMFKb/FI1b7r8YOX3lJMDEcSg9yR1hprMamfOUUNaaj93rBTGqeh+C4PKkcoxoQwt2XbuYEdovsBM9RmY/zviN735a4dYIS7v8VThVro3fqPVWFbbQLs5SFxfkjMc+DdlG3TFVtcl8tk9ktto9SFXpXbtdjghLXv+a7lbkpbh8GB5cPrZhDzcjZXWGu7tfdfYx2DiEHN6612K0mdrI3OuuBCsBln1QZG2GY/+rdXlyVvm13jOQqPT9YlF32Y5qdfVL58qfzl30r8+MFaSj99apRmyKR4DqvNBFOPD4lalZeraUADwpyi+c1hB3GtoKNspcqXlxvPrzc0GQeTJDDHaH0Ip5mUhet1ZV0L775Wfvf3lFbg+sn8+d69j6Rp2Fl3Lo+J02mmlE5Zr9TiLIMo66asxd6DCaqMyFUxTz8bozWitULF1xM7h0C01Go0OI3KV/ABDqPipF5TVxzJYN1vQc15t6itzWm2RTf2h/gobDRAN+eh22quOiHZSCxrhPDmrdQPo9XvePzALa5CSonWbCOEAYE8lBrTy55L76Wl/bT61TAFvIkARcV4fR35lQwJqeXfA8KP3HQQQW8uloxc6O57eizOUaoZ45TVjRBFRrQyhGFTat6SheKshNAP5nugCzG4OZjJncRq+7cIg/gRGB1ye3/4EMBwlK9GuQd/X/efqXm6owM2cbzVvTwkPnXHpcLdNQjNN8R+Xfx3Rigb97HVzsZIefw9++E9rsfOwwQgiOfFByIRkT0amzy4E6K1Hce4fzRgDMnQ3REnuM/6/iY5Gq9484nvetTvVl33cqTckWAjRQ531/JNYPXntyrKgVzHdR1fDLOWsd6P63igHmeYLOVxYDEGawj7D/3Kxw+62VOM/OjDO16uN9atOHhxImmAcSdrRpnpzQV3KDUW7R7p4bhK/oHf2CQ1ixSnUyaGwOU0EaOwVMvbppiZc7YT0/dYG+mAk4gpuPQzKCUcxhPaoRUbQjAlcyO1UpdtoJfbldcV5pNwOguxC9Mp2UgnbwQpviGTw+YUIw/n2SL78sLaGtdr43azn5ndH76D187d5cVXX6udq2/e29XKSnt92i+RIRW7kNPEmyairubvLk0Qn9ry/iHwdMqkAL98bgRvtLnjSW0xqrBVZSuFIIHznAF4XWyO2q1UU51VqJtdw+RG81vx+fHjnkuACiF0ajeybsydv740nj8VUoic5guisLy+8vljJbwXzlGoS+P5s43gvjytKDYTvWw22zxMyraKDWLsQu9jPDZ7X8ZYfM2v1e7Z3th99+YUaN5sNDaq3RtFpLne3xPULjsYVedGUoxoV8q9XTWm0NSudqglxTzmZX8/vZu4yRAav1naeGtTNfF+CA3x3SLHgO83wXtIKt88xsLS4wsFZMf33/4dh+5inuzGqme3OG50NVPAKRnLOVoL6Yp4qW4M/zvMM0b+P24gu/XQOHhG2aWYcyMxRyDtqQXeDIQKkUGKsf973GcaGYxsipWtspB91nxvgAwHXd0/eocdhtZqm/1QMVgXnYlkbNmlPARId5HYWXnxCDJls0lqrVOaDXXcnVX2qzzQmasfg2v0/f2sm7W9jvun+/Mf6Gn/vhOzVlUAddtqcQKzVqsekEyGLHJ8r1WsNi4KWyNmZY5mMrnecVoR2U1Duoz899tr9nhP4BFejkqReCVJ5fjZfU36k+0c0A4qTCDUcfI0YNWCPja75/l9lORskZl8255nkNSW8rkKcpRjvufxg2722hq//PRMinCZj66BVAqttR3yqtrPHvbG9gHi6PByKPhGKuybX922uLpf92nKnOdMTonzeWbKia8+vCOlyOfnL7zerkx54pRnunbWunpEPyaYbKVxuSQuFxu4MKmbKjTb4CkL88mcW7S7B7rPCh89ynVrvHyyWm1ZDfa/vlRqVc4XZToraERJbK3x+flGa8qX58r1ai4tyUk38xG2+Ws7hPP/H8owH4vIlIM1Ywxo2sWRSuDRI4IILp31FKTBerXIFSfzy3t4EE6TNesMk0URO4SSKtp80qmYsOk0mZHEurgbz13EBh9lNdvnUU+HkufqXdzLSJXaux8E4TCvlEDdMi9fMjFGbunG5sMgUeXLs7Ks5i331cnsudpio2H03qlxT9M8XaDvRhApHYf4m4fvp2kyDqWjFO1WfPF8vhbAHXFGHBtQ34LBcRSMVmMRYTonTwvrbiBilULdp76mu4ADdhgUd++N/im+7/GDbvbWOi8vVz68OzPN5rcr7nWd10jvQnD/r7rXxT1ies4NIF3H9fJ/v//b/qHpMd31NCWmKfN4npmmiQ/vn8g503tFtJNTZppmaquUvvnGMGWaGTF05mauMkEhO8yrDvmiO9tEZ4vVF8y+uDu0MkgfV2hVeLma0WI6wZw8mrWENmXZVmpVrtd62GyHjnbvc+/CEOKL9IPv2P/Y6588au8jpcT03SKmJAQzn2ztQC+9K1tz7UGwZZm8p3wMPOzuaCtdidUPF2+9jcmmrW5r5+VlNYOKOwXUXtrLttlL9xJgupMJOzoYpbDuufzgXWqJLDdDSrewmrjEy423RXm9AUl5jxIItG2UKQcCkrtgYeFyN4UURcMYvzR+/G3ETFmYZ6E0uJW+p1Qqh0gmRP9zjwjfQG3xun4npsiUzVK6a0Pb8IoXm+Y7lFXjGjm67Or6C95yB9/1+ME96IIopdhQqpwT0zQxpcC7x4tLRzcntKD1uvdxj4u1X3i5FzDc4ayd2TPcVHrjVgohRZvtHYXr9YaElWXdKK1Ru7KW5sRWd1miL2ZfmMa+G9uZYyAFH3OsNsnl4RSZZ2GaDOpdS/dhB07iNaWu3tyx2WJIKfvY5ELvjbZ1ylpt3JD3U5dq76nWwOa6+xyVFiB3c6NRb1YZtlNWgrNNUsT17AMiqm32IFb5EL9mIeBKMUDFZtGJTWHN2T3hxe+LHlBdXZRSi2kGttLJMZgqrxxOLEMJ2Lp6q/GIshaNxiRYRI6hHQOtIfvmP13gdILL+8jTu2gH09LsOQdB6Yq82yv84k+Vy4PwL/zlMzkHPj+/sGyNUpWtGvSvw3K6jnLbXRj29xU4ugnB0pxS7Fr4njxKun5dd8KV43tw5z7kiKp3kKa0Xu01fF7rgArBU7y910MgJmv66QyeQGm/WQo6g37rtrEW4ely4jwlck48Xk6oKrfrlVIrpVmv8GDsnQplaOpHxrzrtfeXeJu7l9Z43TYmF1oECXx5fjE3lbbReqOUjVKazTKb3HfcU4rhaY/CcmvkLDYFVWBZLC+esvB0CeRZOJ2MpdaX0WVnOHArndut0aqyXM1G6d3TmZyjTUmtlWVtvD5b7X252QLsHq22evj2peDMeHIbLu9VH1UHMy20iySoazTsuvU+6vQePXd+QfZNxQ4hDWpPkzClRvQkvvt1Vjf8qJs5v25ezkrSWYNbVFeLuDZhRndfttZscmoQ86APd6f5YKbvb+uQLJ8fOh8+wOW98O7rxLZ0Pq5tlxn34aXXlOfPcH2BH/848q//aw88PkaarujrRqlKWY0UG1r4UvE00F90j8R2LfDDCqwCszA2ov2OOU9ZB+Vg4u+fR2AvFUdveqrdPeBRQm0exIaMz9ZwiELeT2a7V2myoGfXePRx9D3mfdfjL0AuO8ikw8wwiBFSfejUMWInp+inozq80TcLwJ7w269wV8JwXGto9GCEcormbFLNl314sWkUYh/vkz2n2sUYYlrl6Dr6KRkrb+hdfRDgXUOHHtEpBuF0Sk6WHW++tWYddT5tZC/j9XtI7mlFN/FHqepR2QQWczT/9jrquYqTBYYoZHAfgGrYL5o1Ft29Vh+5v5e63NgjBOvwO0pSb0m2UYob6U9tnWU1LULKNmm2eHrQRIk+B233FkyRIG4v3Ud0FNOhq2vKZxPVzHMgJxthLWoVndEUEsIgN3WPiIN4XNaNkKLNt9sivbqi0k0MLIDI/rl8Ke1xY3xrz719ncr9z+gh2hl/bBG5Xh4OC8F4sC3df6bt5dLxfOzIdRzCMuq091vhW+ns9z1+WKcaNci0x+Q+E3xjp2xim1H/OU2ZEKJDayPv1k2doHp7BwTvDoLdSGAQKzHDHO3ampm/8PQ4A8LL9cbz64q6tDYEaIQ9AkKktUopjRSt7zrHyJQyIQjvL/A4BaZogyhtPnawhpEx9MDTgvMpcfpqYnjR19r55c+vXK/2/KUdZA2KKen0INbEP8/m/dxBYI0mHT1fLlwuE9els9wUddM9Rbndmh9IBt1j6jvRV1c9FpUjIRExCenW2JoQXzpbFs55YoqJTvXntsac2uzwKVV32etyKyzPhXlOfP3jCzEKt9tGrY3YhNQb6wJ16+QUeLyciDFwfd3Ytma2yMGbgDYlJeEnPz5xPkceL5F5DsSwofUF1CbUtmYEX0o2NcYO2ECWSIrKH3/zkfRZuH2OlHWi4iwaGMOPi7fEJqS2wcVEO8yqr6sYvbITAnFwJtG1Hdr2DXovmW3dyn2DqNvFOcH+bYzn9uN453tGm/JIYUwjYils97xdwUw07tSX3/f4C4nsBuddG906oXfPcVz5FgIhuGGiWh83jMhqz3MP5cfvondX4A205+50VPeks58faqUhYFDv7wzRRg/XKsgYFLB/BntO6/dmL4noHuF0j5QHneBqvXiQPeqfvw4o5qWfewBzzwsN8YxBWiMyZVQG9HDMHRBymB/aGXrfYXhs8mGtPC7q/Wc0MYtQw50ltxyX+Ij0x+fcibWmpP3ayk6wjtLZfbntDeQdX4vYASd2v1IeElwj2kQxuKsjug43l31BGEPtkHor3X34A9W98rpfj+G64w7Nb9cNvKkI3RN1Y+0dPydv79dx445revet7td+yGP3xqVRFXn7LHev6bhVj/v253n8wJHdWilzss60ZS38or7wcJl9RraQpwkJCWUF3ezYkugkilkbtd6dvxhSHH+ITT3Z+7795m3FJ3RqI3So7uOFNus4wlxJ7EYae/30fiLnwO0aWJfo8M0g+m0drHDfS2yT/17Ih6tMKX1ffLelsK6VlALn02TE0tZY17abQYbgc+a6OcyomgY8xkPH7eeGRxSBoHx8XvjyuoAEVw4qIbRjIwrEHJlScDivaDC/OksN7RqWahAcVabkRhTV3Fxr7WytuDlmpmG+e9XJVKvtGykXxcwTm3a++fkVEWGezd+tltG/bsu29c7z62oHazEvOuKIfAIpElMkSEJIRopVu96nnMxuvIqXn2wOevFqR5yUMFVCCpRrpoqwLI1ai2nieyBm5eGdlVFfv4CaF+Y+JrwWu/6Xs12j4UxbWodmUTh42fI02SSCdbGSqoxr7waajA2uVie3VAnSpPu/BbFxYTEK69YpVRHVN8YV93yG/bfw7UPhux4/sOHkUMHZSd9aY6mdnCPaOyHYyR1TIJbNSCiP9GAXTQm7CME1Kd86QYeuXnZyZVgGj370bhP6fFO4EOTu0BAR5ikwn9JxerqsShVvRrCWS+smc3smf01Rb/LwCGwyz07pjd4j83REv1oPUmXAu32Twh6desfaaN+8T/t73axXPuVIHsMiR5PJjiqshi0e2UXdFWd02SlUdH8v0eXDJo+16TTNVWzjfwPBDI+U7pB1OMT2br7rIs4ehzsTyTtupBRPaL2Hf5+MKsemF9f/92ZlvthMZ3DcW+4INfUc2XziEKUVk0+35n72TXxai5CzcSBvWoHvlxSWIoC53JqHgP094HoMwzlW2AiGqe/i0CCZDf3YmhQ9TDaHfFj8cEhRKMHbaORNSLtbrUf+buvhbdPYtx8/6GY3oiURfb65qt240jvruhFiMCZbIadITie/MVaGOM8TrVue2lpn3UzwIjuUtAEM4LmbO8tmV6O9fFlsMzydSCny/unCw+XM8+3G55erq8osIgZNTDKR5s5jatQaWbfo8N1nn2t30YwNqIxdiN2GVo7hDTHYgkrR7FdEhG2ru2e7+gbCy2Ojpp4GgaOHkQVidds8DaJo8BOGDMpmfd7Bo3ZXuN0atZrJwZYbcw5cTnYARedHhsHiMH4U2AmuMQ6rVROqlGojpWtTSrOIWooZYHQ3c7CH+MForhzrVmgavJRoh0L3KTxiXcnUAf+TuQI3J/FaU55fN7bNRlDlFFhXofVE6Z1l62xFDZ53cfsrpbZBWCqtbTuhFg8FCihst+AiqYMP0XHo+p91tWs9JNqDjBzr2ubCVUOJ/U4jwpFOCcPq+3jeEKxjcnQ8jvR2nxOP7AeBPZ+r5jwdlKj7AfKrBkTAX4Bcds6eRClUV2DV3lhLtdPaT6lpyqQUKaWgfbPTfbZ2Rdvsja744WAwSOTIYwabPza7KFxfV3KOnC4ZScLT49mcTD7C8+vVDBiqiWMDkRwyeS5ECSzb0aygHogiximoT6PF1W2jdRFkL80Mf7QB36vrwNEjopdmiyoEY59FnBwaN95/Lrn4cJA3Iy+3/24m7hGbUnJbmo9CUkoOyMUssQfJJJg9tcp9O6W9+xDEdQBQuyBFoZp9lXXY+aYqVnM+nISce+iHieRWodGODr9+OLiOiDrGZIGjOEchrXVer4UtNYJMFvVWYV0jVWFtVjdvzVBI38dG4f59SmsVAc6zzbtXNU94gG0Rj7b9uM4wiHR6N/gOx2YfQ0pg6BXMrBMOIc6IuF3HRja1pfkYAt1n7yWQJt7xeRwkO7uuw0b9yP0lh110o/j7+TXJ+w8vqgniWmdnNaO1oDoJ7d5zQoiRkBJRTbGGw1obaN+oPr0jOBkjEt6UycbYoJ0Mw8UZMOzN3NDC6pOCDQiI0SNLSt6Omxz6jd5keza7E2FnRntTtHbCVsz3LtpIBBsr1Nk23fXMpTYzJmh4OqPOyto1MRjMvtnvSUXjFT1qFt03yEEGuu7aN1kI4vPknBDtFi0JED3dsYGPYV+k2vte5mxd3y7chhtWsDfeWK+9Rcbe1aen2KCI89kY62kK7uDrvvOwDwtJKVtqVq0FNcXg6OnYfPYaNiihVpzEjSSxXtNAI8dOjUodM+gaLDc7OOaTlQCnbM/T1Q4IODTpXTlMLRmbmj9LgImlFkH1gP57tB9Q/O5rf17EEc0d09kqbMu4vmHXhNGH4YXATtaJt7vZwWGHctit3tAhLf7uxw/fzx4CxXOeGIQpGNzempIQco622XIi5oyEaB5uDilbs571vhUE00rHaJtznMxdbSZbHrbUsCu/ErbRowqlVpo28zgXa5OcAjYGaTJ13/VWuW39kFN79LQPZHe6dmPUY7fGDYLN5orZ4G0tnZfnyufPdhCMElXEDqTVSasQAiEO0wi7oUZkee6eBmS0RbisJrwJwdtkncCw4YYmoTOllZ3/ikXabWsW1SeD6HP21CqYa0ppwuK56ZhsY0aWttHWzTZqnu9RlL2mNivLdT9k3j0Gd7ix1KZWZbV16m44gek0mXZBKy13gk/XtXF3ZmiRY2SK1p1WCuRsVuGgTFGpofGcq4mlNp9yswnbosynwPv3E9McmERtHFaT/ZAeY5tGIB0ARUdol2OzwkA9cE8Pqw5xjJXrRNjFOgzEhFlPgxukilA32BYlJmE6DQdcr1ggLmQSf32FaKYWQ+sbghAl0cVnnP2mbHbAN4h6cHQo1w9iQVDvYPOcxL65l75ETK465USrnerlOiuNWC08qO4QfvRzG+QJ+wIuwZJbcSeU02zjOyQEUrLm6FJsTExOcZdgaoDhjxVcuSbdIqNpla28FO3tUKWi2O8Oz/MRJu+truFOiOPNKuOQGvKD0F0SD3dNGqM9WHfIKUGRfZHJDstHjqr/P3Pv8mNLlqV5/dZ+mNl5uPu9EZEZmVlU0ahU0EItWkKMEVL/Acx6ykvqWYsh/AlMe9pqCTFgAGLCjElLTHuAxAA1lCigS1WVr8iIuP46x8z2i8Fa2+z4zYjIlLp0q0zy69fdz7Fzjtne6/Gtb32rKqjWjU4LPf2RN2SRhliEBL2HYS+ZKfLeRRvYIov9NsPePdfVg+3W2qGRRfBeOxLFa4rU6b8tI5IMMI1M1q2oht/b2GMzZD2Kcb0UZxCWdL0CRet19p0jhEoICswhGsZrStijJwvbbz+PXiK7d+zSaY39g2NG9+2i388p/ZPvkeZG861tj95Ae0BuHm1bwc5va6Y2qtwM3mh87/F7bXYReQf8M+Dv2av858CfAv8D8HeAfwX8w9batz98ItR7mckspW2hVBDr6CmF2grSBrxZtCbVJnUoZ/vudOA4DXzTXsgp470nek+z4Y8Ap8PIED0vl4XlMhOaTiCtFZ6eZ0Jw3D8MHKZADIG704khjtzfv6M1+Oqb3/DN4zMPdwc+u5tY5sz1RSWLNzkbr/RG1yIBDTGt6U1rwh7SeiGxkjNcL2ochlEXT7EN1uza1NaoSQEsSbrx1tQbKxrVPIaTm4XYej1ZKGtvqe2GUYij21D42w3Ya/M6414gWJ2/J6xOZb+P46jSUEFrRsFXZGjMKzw9atkw1d1LcrM5SoHlWaMgh9bHi+z5eK4aph6PA9M0cHc4McaBdU0sa+Lx6ZWnp5noPT/+4oG704HOlahVcYONO+G0Dp+KmNeslp4Iw6Tjn1xwDIeJGAI1VIgLNEfFaWPSk7WOumZz3Lpn3p2RMwfkvGInxZRq+vqWfl/qLYaxP7enYwrCaSoqXo3vuhYDTvf0cwOf0Z1X+yAKtOxXSqbZYHjp/SLfc7jv/9Ob458A/0tr7e8Cfx/4P4H/GvjnrbU/Af65/fw7jl09Fjoi2b0cmxXbXYQx39vNF3vDv7dQSG6+vDPv7ru3vXkdO0culWx0SRENu4bgGAfP+TBxmkbU7hQ2QUzpAxjljR6c9NKQMTqaTQxxqFSx2JgYxQZuAB7YrDo3IWK9Adx6uH6LCt/+fGvFewT05rFt9wZiyFtjpwDX7f3oEzqmoiXQXZo7BHdDJunRiDEB+zipul9f+us1TL11/+rX/Nb3dcLN4TBxd3ficJgYYiT4sButqJ2LQ1TjHIK392lfRmPWAZk3m8XsV19H/Zfieomw7RLa33F8h+P+7sfdRiw3Ec7be8DG/9hO3n9vj92xF8NjbKN3h7ARzyysb/Tn/PBGh9/Ds4vIA/AfAv+pvfkVWEXkPwb+I3vYfwf8r8B/9UPn8l64Ow18/SEzz0owmUbPYfCMIVptV0PnklYWdGOu1jDRrCUzF+02c044TEHLHc0kJuzuLmtRGuqqdStV78w45whOVKjRatfFSDC+eeqaQND+9qlxuRReXp44DAP3xwOtVdZlMRZb1agjqwZ8pbeAOvIccEEIcuA4DjwFEGZabaq8Kj23k22TlYZShnGUVTeYGJ9gp1nK1piTbja+lvTFCEuoOKOwFXi3/vtSKU494TQG06DXaCEOnhAjpcIxTQiq6CI0FktD5rlwnbUS4nwguKY8826Ua8WJ14m1wIp6/d7PFdhFOhSArDy/XCml8sd/9J6f/fhLfv2rX5Guv1KKczMPvlnEPZ7ubqNaO/NxUtrtfL3wSqbP6KsN5lchL5X1mnCuak+CafyLUy3AlDQ10Y2kBKTg7bUNEOvvffP69E2+A89di673C3gHYbCUw0J0TKVHjJMAWxpOHwPWbgxra7p/zseI81CCTtIti6euwdxiQaSHGb99/D5h/L8FfAX8tyLy94H/DfgvgS9ba7+wx/wS+PK7niwi/wj4R6Blj2HwBmY0YkBldbzbAIueh9ZaaOY5cmlbCCSwDSvoE2By6UBW24Qau+cuRsXtHt1/5CWpqKpJqhRfaVW134P3DGHgab7wellwd574LtBqpSSd7FLdHnmotpqOQ8ZQVtAhEc4HvLvSTXktVlv3bL/ri1jzdN2Yvemib3S9nkbCqOzrvj9f2sbWGgbdGF0jrkcE2a6d941ooof91UNwDKOn1a7k2nA22GEu3fBWrnNR0C0EIxPpyW8di3flPOUHAAAgAElEQVRi06H65zIPtXk3uxa1siwJ7zyH8cC7h3c8Pz7SbdWbSIX9y9bWG885BC3YB7fQB4VkGxSSFkctqLQ3TQdDYh7et134s7LpEWokepsx3xopw0dkz+puDS/saDyyb+jtE/T7ueEk7NhK/0Bm1Ppn35iIAZLXlMg1R6k62aar6Hzf8fts9gD8+8A/bq39CxH5J3wUsrfWmnyPQn1r7Z8C/xTg7jS0l9fVPHJEaMwpg3jmnDU3FEHE433Ah6DId8v9XCBCjAOD5TbBCWsqtGaiE9I3jPWlW266pw7auuoqpJxZk6LJ0xRxQ+ap/RwlP5yZYuRi/YqlVq7LrECYtSAGu/OZHp5aS64xn6QJYxzxIXAYZmLQcL5zuNv2nth+2NMS3mzmamBNB/ha26mXOwjlqKIbLd6QclozuWXZF0/wwuAbMVS8qe9KRfs0m9WZm2IAtWp0ldKupBqcVixa0/o5BVyy91oa66IeNXpPcOhEFmkES4WyaF3bOcc0DRwOI87q3yF4pmlkiPMGbKWUdRDIbRhMNxzKshxiVAJTcFtQ461L7HLRcP2Ln0ROp8jr68LlVddIKjtVtbm2Acf0FKcbKPbuNYStbbVvbr9dczUQru5895L7eSxC8ILvOu8WuXy8T3t+30k+iAqeiGs0r569rpWW8psI4fuO32ez/yXwl621f2E//0/oZv+ViPy0tfYLEfkp8OvfdaJSGs+XVYGfKTKvK9clgzjmlDUEjZ7ghOAjIQRyAUhbfisiDMOgxkAgO3AukZOWtUrPzUxhtbW6W0YUGS6t4lrf7I3DqABR8a88tb9CcNz5P2aQI8EFBNVeu64zHvBNo4UukeRtjhi0TWSg949PYWKaDhyGF2L03OqPQ0dR+4YX+myzTfC/ezTdVYChtjenEVvUDcFVsY2s+Xkuxcqc/S5oqhGcEEMjevDStNevCs3GHwsKGqVcybXYiCgtU9WmGMUYo3aJ1awglq3WUhvLXAw0U764o9GkbniKSNuo09M0cjhMJhJSCd5zmCaG4WLvVzf7sibGqKSYWzeoQzwbgw+Eto8AE9HSXq2N62slROF0F/niywl+Wbm+Ks24Cz2KbUJlA+6+a8uGuJndEdgn/97QXG/zca0EdXxDH+Q9ppGg97kaj79fv9tXdaLYCdLTg8rrZdXKQzANg6yqNiGo7PYPHb9zs7fWfikifyEi/05r7U+BfwD8S/v6T4D/xr7/z7/rXKBWsFv3IUYQXRBuT4vMmutABaFtjKecMmIkDOxxPkZCVdWbYuTsSs9ve1jWf+4prBqElCvXRZtPnGg4lF61/TVJRVgZouf+dML5SkkZ8cI0TKr00jKtFbxvjENXIDGyh3fGcfb4Dib6tx699ohO2NDv2ktUmHFyO5urcbOxLRxv/ULIjrj36AK6bPYtmmw8g6ApUJ8aq0Qg9Y61VnIpe5iL4gStOUpQpppzjVy0rKjqPvY2XPe2CpxF40g0dERUB/58H+eFCnfkrJWOLay1ry11M9p09Bp2lFptKOf+WfsAkl4mdE6pwD30hsa6ZOaLlvNOp5GUCiLrVhnqrD49n13XnkNvUZXqCtiPm6HYmq+M8t2JUv3YtnKDTp3Wc9cNpARjHVbTjO+pSl83PeQHBeZu0qRS/nrEK/4x8N+LyAD8v8B/psuT/1FE/gvgz4F/+LtO0tDpl6NTIGkYBu582FRM3Za/iOXGCQHGIZBS5uX1uoWtsUZFaMcRxJkaSgGTtfJVL4APGgI56b3ICshUGq9L47poqB2dI68jl28eqK0Rz4kWL9wdz3z+8J6XyzPfPn4NMXJ6eCAEz8vliTUtjFGR/NaEUvQmjoNWBYYhMAwjcQjEwZRgO7+6ewnXPa/u+lYVhGygnndbfcY6tEageTE2lhkLaRqKqwqKLrzJOWsQ1HBw8MIURPPS7glqJrdGiIHjcWJZM8t11ud5fezUHIOFniJ6/a7rVcG3vHvpjo6LcwQXOMfRcutKalYf947kmwGyjeW64EWjJ8xQdGQ9WJluTYWwZJ3Jjme19eAtDXDO4QhqyJxWEGoxYgKahyPw+rQgLXGYjvzoR2eWeeXp+ZmUKs8v6jWDE5rT1ECvUaPZqKxk+XuX5+6ovhjpSBBKzTdVCGVlxmBev5OzvOkvurIJU3ovdu8Vs9EW5r45NJpz1qX5RtNBFMfaBDn+dTZ7a+1/B/6D7/jTP/h9nv/mBU2VBNjCRbbv6iWyhW7aIKKbpppH6Oo2W1eXPbkTLHaixdsNslv3G4CnNaNHah7eGngJWsM0wKtHBLr4vApq1EotOvI3+IhIQaR8ZMkbOgBCueSHKfL+3R0pZa7XeeOMV3vvzsnWNYa9V6Hz6uGWqKEepfc+778TbOHJ/rXVhe37XrrZy1MmWLMr0PL22vZX6DCVs9Y8jULaG6/Wz+9t02/+bJs6CfRowQRBvPem/rPHwHtZ1FmvftvksXtao4q1XUFnLz0FI11lKZRct7DZBxXv8N4MotzOd/9Y7ow3P/eQ7FZCfD/eLqyP389vPdSijI7Tbevy5prf/q4/9gba0cc0XaM/4MzfHJ9+SMTdiZJWWikaBq5ZKZvaWM7rtXPb1QsfjyP3x4PWXV3QHNdWZc2F1BRujiEQmpboStFBj6Uoahy8ji+ulmt3KNGZEkmTxJwv4Byf3yvjxUWt11YyKV1xXoxw03h+fsEJ3N+fuT+fKXmm5oU5Fea6oi2VK1SHqwuRwB//4Rf823/yM77+9gP/x5/+31znlbQU7fJykSCey5J4fl2gQTTPPARvFFaVVa61d5IZECg6d63ZIIzgnbUG68IcFFjYykCtOdbi8MBg1YBkZaR5yQxh0ftSFYXX69ZYkirqNBw+BDwQUe+zZKWphuwpDoYYOU0DgmMxo5YT2qASBV+0H/2z84APgbuHO6aD3mNlw1VaKwQvPNwfTNyjcGkrx2mgtrgxIxFNOVxzBFGDcXec4IsHnl+ufJUyPjQO7xvDCF/8pHA+V9qSaXOm5MySdYBFH4c9Nx1KEcSiTaBZ6NSxFDGOe+9lENllpW7Lcr9lfHswZQpGQsPbdew6/n20VGvNDHH35GqQRZQ67uw1FQS0tdAljb5r//11beTf53AiDDGwlkzWOEXLUNx4mWoTQWunI2rvuKLwCnClZLPOW1WBQHavXzXOoZNI1Nt1+WN7I2a29Qbo4spVBRUHVfKjeqvFNy0DCo4YIqUU1mIhrujvXCuUmjd0VS205U8tI2ROhxMPn79HpHI6jog0steZaUEGvKhgx3VWsK8P9YveG/hWDfnevV73hrvXdlt3Xae8OmchO2LKL11Ycr8vrcnWJJNL2ea+90vV6E00OtBjy6kB1xpJ9tFa/TV1yglbpULpverZddKpqgGF4IkxEmPcIr5+DdWzewPBNJTW77fefP/qRwieaQjM0ZvykRAHIY664YdROQO5Vmv6sZbUHj1tEc7NYgFdTx0dl10ldlPvN898+15uI8odUeyv8bb/vEuq9Wt5KzMlOyZLz9xujci+2r//+LSb3QmHcTDOemFeVubryngYePf+Tjfrty+slqtD09zs+VVFLdwAFV6vV5aUmabR8jUhRO1Oy7OCPdHBafR2lh56uR3YqpByJZWGv2a8rATnOQSUcjtEQ7Ma0e0eTqQxHSZdmM6RawPxuDASxXOu+lhvQNWartSXFR8bh4PnEIW/93f/Dojn4e4PiMNp67t+fHzkq69+w7ouvDw9kkuipqT4hamm1qrDGZutSBElvtwSVbRbqlcuIs47RpuGs+bCvKqUt3OaM0yjbnbvtfNNnDCMA61VllygyiaKUapKfGvBXJePiTwp0KQX3Pq3wVvJbbBBFeLU81QaxSaaprRugqNqqB3OB0qF55dZyVAG+JWi91EjnN74pJ6vGBiXUyGnghfhfJ4orXD9sLC4yrF46tGRm+bWuVWOB09KjcfXQlr30U+dkbmlM00NfKtWpSh9fnwne91w2G82uWybWwsqIqqLGLy1Nfcqi12zTrmGnTgk2z96gVNvkJJu2IVbLP+7jk/e9TaEoOCJ6bKvayLEwOnuRKuNl5crJXe+lYlUzIui2qP2aM/rymVJKqxfo83w9lpDL4Wcs6UBbvMCm502T10xIow05qUQyETfcNFGLxGQAD40vFS6zLKIUy9k3RWlh3I+EgSmIbKrrzZyXsm5sswDeT0xxMAf/cGPGMYjP/vZv8fx+J5ryqy58u03X/OLn/8V18srX/3qFyzLwny5KFJtYV+1z6dEHo0CBo+Bdr01Vd2n6gfo1JS7KXIYAy/zsnsyW5rRWHbe0h0nOma6torLuoS86Sq1rHJUbfNn9Du15e/62fUfZwMYcKbusslRW4dgq+SSjeNdN0TeOfXo13llTZlgFQ4lL5nwSdHP2DdJbwTpQhQiwjQNpJS5fNApwVfxuKunDkKNBVxjGHTAZq0qzqHErh4t3Xjm/gvH1jPv0A2nKVLdvPWb57SO5LPF4S7aZq/aMtwfaqEBPYoQt1/lmwbrbaho8FpGle0v/5oA3V/X0WjknBEb1Rx95DBUgvPM10SznDr0QWLiaDVxXVdicIyD4MVzfz5yOFR8iOA8telM71qblpOkMa+ZZI0GW5O/1Ug6wCQITXRxN1TvezpGnHOsTXXS1huE3YkjBIjDAW+1XE0BMq1khEZ0noZTDW8xGSw0PZnTquo2V4ECTx9+zjI/M0zvOQx3yLvPGIeRtK789Muf6oSaVQdI1FKppVCK6tzXWkkpqadfr9SSSCmxpkQtmZIXtk42BG8WIY6RkwwW+upG7ABmHA7EOCE2e1lTk0It2gZcqmoJpKTXes1aHno4KhD560fh8VV19JVgYrX3Bs7GLBxOI9MUWXPisi60plFeSisvr488PQWu8wulJIIXvnh3ZlkT11k3ay6JZZ11fHar5ALzdWYbzS2wrKuWDlsHZw0ALcIlVTLaw+SKqLSYUgE2ZL2h9lIrDWilo5ewy96htqWFVSsnTjT1003f+fdiJ9Qm42L6BGuyVuSNGLVzIZqNFVPR1e6sdN3GqEZwXtZdLMPKnrgbo/Idx9+A4GQiugEnniFGvHVaXS+LbXYlCPgYER+4XCqvc2GMXgcxhMBnDyPiHJe1Mqdq5A+d4TUOgRZEJ7zclKE079QcXHunTe8dbX+tNMQLx7sBJ470vFBzZc5CRYjeMUaHD45hmog+kPOiUtM5kdaVwXtO4wA0Ukva0GLZVCqF6zITS8C1SllWvml/ThhGfvrTf5fz3Y94d/dAPBy3HAx2O91r0aVklmVW/b55IefM04evuV5euFyvvL68kvPKMj9TcuF6XSmlqs66wBgqcaxbBUKNq16n8XBiGI/UspLXV2rNjEE9eapCaUJJhZoyS0o8vl6JvvFvfFaZoo6mylmjAq/KHdsC9KK76f1nZz77/MzleuWbxydSqrw8J5al8vT0NcElclJwdYzwkx/dMy+JX371aCDhyuV6YV3V4HXjB1olELCJQo1mBtqZQm0VeF4qrJVDFg65EadGPHSpJ7FN3aNAFSt1XohOf18LmwxYw0pexm70VsqsprHvvAqDbBhAVQ2F1mDuG9TicxFrxmlsQi7Oq5pOa45WFd84H7VnoVZt0upGiZs1833HJ5elGkKf4ipg/GnnHS5GA3Iy0jv+LUT2PliIWamtUJu2yQoN77T5IueMEyXgIBqSbmDVTfmpoWHkVmKibSBdsVysA22inFEw4KWitNm0rrRQaVXDaPX4Aee8uYCGuGrl0U6JUQRWFVmzapyZSfchMIw6N7kW4+aHsJeiZL9+W998rXgfKdrdwjiOHI4Lx9MdtWTW5UIthet1oZSCa6oC099DT2dAUwAnMIwH4jjRSqLkE6Vm1vlMrYW1KjJc0kpJCyUX7u4WvCu8vzsSfebH1eMPBjk7r2H91uepHPW7+4nD6ClFteS6HjqtktLKslx10EYuNv6rEnxjGv22+EGJTMPg9gjlJlWrPRRuHV6zARJNDOA0HMfgcUXUG2OsuFZ1BkAt7Ju+bd5TvOkn0HCba+9Bf381W1s9hWzo9Jz2dkNaoLmF/J3+fPv3WxZobUKxaLUvC7HY/1Zc4/uOT7rZvXPc393pzHMapRVyhThGzg/vqbXyQWBdV3JO1Jy21sfgoYj2H7cETjyOyuQqS0pcLldC8Bwe7lU5ZUnWL6y3wTvHYOqf2XLVamhsrjZuKhWeX2a8d8bUQ9HaUml4ShXWvPL4/KRg3jgQQyCGgXGcEJxxKI0RRSE0NL91kEqmUmi+EF3k5D0hDhxOR+4e7rheZp6en3A+cDyd1VMIYPRUHxw+RMbxsFUTAD7/7DNqyTuCbBug1soyzwpcLldKXtUT5qKfu6QtndEw0ltuDk3c1vSjAy5NlTU9kdMjQYRD8AiFVF6pLfNTC0vXklnyikMYVKOGVhONwnVJLCkRyFwv2kjzInruy/UZmqUfvafBFeLQ+Px9pNagohssDKMwHYJ6WRMJ7QMk29q2ikKnTm+6fdmUXcVBjDYfIDI4eH+ndflvPjzx2hVvMTA3aUg+TlYdSSBGimobtXZr9wGsP6MoUabmt8BdNwIaEejPWQOUrfFJ59A1w0IKUJhXNSidY9BZodbA+YPHJ1eqETEd7NY/vJUZTBrTOSVZ1FJo0q1Yp4dWW5ydadV7o9XvdopJw3Kgm1LOFhrf/N4wZCMnmKBCZ645jT4cYvJRbGBLrUU9BBGkqR6ZcwjOxvM4LaU1B20HnXoOrQwzj7iAOI0InOocGegmWzRhvSv7terAjZW7ECtR1nAbF9qGb1YuzOQhkM0jp5wtBO6rFbv21nyBoIKVlRp0Ko9PZjR8JPtIdI5TDApGrplcPQ4YgZgdPrePNrvQml61SmWInnEI1AoxJlzv2POocTP5bOe0eSiYRl6uxerreg1ahWKDMorc3GdbCX3Db2VI+7CqzSdbL7z+ripzTmSXYrc6l7Ttvxsbr6+H3n23w2f23fL0dgPQbX32N4/cHr4/bROtePOQXv60T7Gner9fHP9JN3tthdf1EYrKLYsTXIxaXvvlr1C+8sQ0TFydjsxJJXGZV4TGsmi+fT6sDCFstEzvPMdxAITrddaFUCuDV7RckdLGWgrOiYaETliT6Lgn8VsDyjwrw2ocBsboiU4Re5o2iejuqjQPbqr4seFaVWkqr7PNnAuEcELEMc9PpDwjriBkYjjx7v4nxDgxxfcEOUAbtvyz5nSDLDekN3WAGaTObmtYLgDirLW3J4e7d4lxIASdT1/rkZwW1mWGVvFUeo9AqU2RIQvtGwWoSFCCzXr9irS+QMt4VDL52jyVwqW+UmpS4wY2yCPTBNYuVy0aasroOYaoOnOD5zpnnHsm58bPvrzj/jx1TUXThysb4aQ2dHJurVzmxMtFx0FfizY3UVWTT6sANjE2aSwQgvXVFw39p0l4uLfJQyGpNl9OVn+3a9DY0ozOZWhF49KSKyVp6XIPw7sDMxyoNGq2zd2Zet1O3KD0OVlGYctrs783m1fs/GndSTPNvW21VZWmvyWkmkYj1xVKgKIgnDagZK6XK947zudp0w2v3gPKbhIrUlZfqUPVFj9LVAQxWiUkQ2E1dBfjoZsI4uYR/EaX7Lxxbx6kGHdUMHBH29xoJlAIWLthA9+Q0HRwYttSQLwX4jDinCeXxTYONFRCaxxPxDgR/AHvRkB53Fp6si/brIo12P8M2W7mBt6owtx4g+4P1At1ryUb+l5LQnAEeh+6zllvRf+v564WOdnvWGj1sr2fJpDQOXypJa290wk72h9ab0KTYpGTNw24kUBtgnOJ42Ehl8bdeeLh/oBrGqNVm6pTayMVBd2KTY1xCGlVeejVCa12bnq7+W7XSKwi09gadYIXhtgvkl6XPuGmR2B9zcobsk3bjIA2/+iN70EAhg9sQWWviBjAd3tsEUF7axD2Hv3bGEX/6VwCs6FqeKpRun/H/vvkdXbHqGBPbbiUcKUgAqd3R7uBibJmspmrGDynMRh4V25a/kw/zesmXVZdbE6ilkpsZ7pWNs9eiinWhEyxaSIg5AJLKnhxxBhwYj3yVELpskxs8kwSvA7lK4V5LhzGA/Hg8c4z+Kj5ICutOe4f3uH9FygDLjONB+5PnxPCwHR4wIeRtMx89atfqPqL8e/7+GFv/PL9aKZm0/fRDtrdHs3KQSI9tTHDaF15db1yff6VesHj58R4QMZhZ7FZ3t/KTKuF93HiXFZaWal54bq88M3TL8k1sbRCaVUFSFwPWxWGz1kNVzaTIa2qBqEIcQiUhrYXZx3y6DsI5hxSBTFVnqOJglTLy4+nyP3DqJ79qnTd59erNcgkLtdEbY5SHWvKfPP8Qq2FYajEUXPgdV6VJSmyUW+92+WYe3zUZ6BLBda6DWTcNnlvOZabQP7G4nYn0NDAqQNzfU/0PH7z/o2P7cL+621dW/prjMT+Pr/ziXZ84pxdN2NpibVWpFaEzDhGTucj0hrz45WaChI94lUx5TAou2o1WWOxeUU6ntepLLAZB52kIuArtJ1K2qrOTZemk0yUmqi2UOWVG8HDOCiKn3LRhdgE18TkrTVcdoPmx6Uo0j0dRsLBEcQzoO23OScawun0I46nd+YNKjFEjocjIUQO53t8GFiuFy4vz8TxwHS8V2/cmm3YnuzpN/UgnbTDlrvuueAe6n/X9Ve5JUermeX5N7RaOB0eiDHgw4ALcc8FW6GsyqQbTp+BOPLySlpeKe3XLMtfsuaV5KuqtPpKFFOo6aOaLFpZqdStEAkezxg8sWpXY/HVWp9RoYDgcE1wRq2d/IDDaRNSrZzbQKkHSq4sSyHnwocXz7KufPNhJgRBCIgMXOaVx5cruVbiJDqBxlVW08RuRuBRIMworDf7plduaNByT9z3v227u4Om7Wav2ybeet/zzd/tHL3hsW92+YFprL27sJtvC9ZsSMQPH59eSrpBWhPzshr7x1FEyFeVespFc7RB9G8tBIZRSw69D73USk1qHmvTspvcICYili9Zv7fzqoDTo7FqIonOmYiUYKU6xRWoglTtfctS9TFeGII3Q1NtVprenbQUXh4veOdZw0oIkbvTe0IYGMcTIYwMgyL2HeFOLRGTIqw4j48jPoz4EPE+7Bz0213+W5b7xgub/JYuuLchfw9nOw6ACOK0EkCr22uC0GrR+rRzbBcMjRDEeVwY8LUwTne8u/8JKV95TR/IbcVZOlBbprhVX9mUb/TdK0W1tkyQiIiQXEUmTYWudaXM+c1oqA4aHoak47xMp3aDr5wKOQQHp/PAmFU/4HQcdCOL5/kFvvoQVeElKGtuipFpGGiukMOqnzWPtCJcLy+Iywok0/nrvRauWFPtgNwWzu/9AVunmuz5vPLcZfP+Fp6aYClvfrcl7TeY25to4XYF3BiJ3yU6+ck9ewOW+crr64UpDsg46qy3mgBhrZptjaL68M573CCUknDG5tKOtkqqyrJy6KxssfonKJW0FJU98tHjWyM29QyXVQGpGDT0FNH+ehHINWsXmYsaKqGyTG4QJfoI22DI4DzOBZbLytO3F0J0jIfA6XDmpz/5E07HB3wccT7y7rPP+OJHP+by8sxf/fn/R6mFMKzUJvg4EOOBGAdiHHDe62bfkr8+8WPPQbtrUOBGN3rvJcc8a0eAmj0Ge1rnno/TEWgMwwEfJp18k7NVClRpo9nr99/5MCLiOYvHi2NNr3z77Z+xrM+kksglk/3C4h9xEgj+DjHgrjVYW2JpM5GsZTTfcGdFmR7nK2WpLC+J+SXhgiOMAR8cp3MgBs87P3L00YZp6EzvMGjEdzgfERGt0RdNw4pUvvnW8cvfjPhnIddMbZXzeOLudKSEK8twQfCM5UTLgafnleeXZd9popfUOZ0F77yQi46y6o8RdJCFSCf19AYurPTWr73c3D/LtSOoFkKPFARpOw70vcF5TxFQbkH5HbW3T5yzGzgSvCLdUa2wqr40W7BCw1n/sim2IsYN14sSvMo0Y7PDgms64cQuZCdN6H5QnjVd+cbtiiDRmkN6ft+bXbRcUvfzGEbQjBXRudPY2Drlykf1PDEQ4wEfoqLgw0iIgwJJ84WcVpxXTSPdQCaF7P2bTd7JQLemvXO/3yBw33F8XATq5v5NNxaYwm5jXWZcM4VS0U42LSH288geFdgbUMHJkdgqw3DW11pfdYItgYy3Kof2J2iMqaq2UQK+808duCBg45MRUYZfLkqrHnRzdTAqUVla1iEYVT1nsGhmsMqFdqMaRNejXosSjlPEB+F0iByPnoQjty48Us2Qt000Q0P721Fi+zXsQGXzPZUyYE2sq0+wluu+ZizSujnNdkU7UNffb8/l+99rv1x7pNduPtvvc3zyMD46x7u7I+dDBAmIBFotlLQiAkNQbnpKmXleNdR2GpKvVuo4H0/EEHh8eeH5cuEweY4HDUNL7rpuajScD4Q40GomF5342udon44j4xhZFn2tlHW0VKNRnYbYTbQ9qbZGnnvjhdbEO1Hj4d077s7v0JljkWEcOR7ODOPIw/v3HI5nXp++5uf/6v9C/MB0eqfvaxhUVWUYzDioyKbzXr93cK2Z9FMvudnRq617r8T+uJtVZOfo9CzTOpM+2KHw+Oufk4tjOp+ZTiec84QUNL+3QXG1mipjTbSq5cE4nvBh4h5HTgvX579kvX7NCAxUqgh5iEr2mJWzP40DYTjohJ1VWX24QAOmaQIEHlbIKw1Hafo35bo3XiVxYWVZVq7zzOA8D+ORwQfOIkpjrpXctH06p8qy6HVxQfijP3rg/fvJBCwaTy8jy68m1lR5vL6SMjQpHA4Bt1RavgnfBYQK1Ty9AFFwoefjipSZ3CVjhCEqD/61ts1rd6NKUyfXjKy3CYj23B2LBHra2d8HJvzZDPk3mq38jk3/N5Kza4821OaUAmjNIhq2aJ5aW6d18hEJB9N0655wD2JwmxgAACAASURBVDd33W120MquwG2A072m8yZ/5Du5olMpdy/YKbXNFo9YQVR6fV4cQxw5HE6IBJwMxDDgnKc3z/Re+5RWPJ7BB1yIeOcR7zdSTW/j3LzIFon/tmd+c0Fv//ex9/7Ox/ffGlFlXVhSJYwDtUyA8ru7eupteLGdr0dMOEI4IBLI4UDzI7RMZaCg0UOrbw2Ndx4xbKCXPnuIK2i1wHmvaHrzVjUxqi/WLVkzc1lpBEqrFOv4u/0qtc8EsIitNU21Rq8DPH3jujhNM3qTj43M1caaTl5SsQiN6tpGktq8r+j7Eytzt5t16Mzj32ReN169bWSd3TDvHl16ZHJzm9/s57aTa3qp8YeOT0uqqZXrvHA4DIxhYr7OpOuVXCpLMmqrqHJN6Vq+6EZ23nGeeqdZIaXK6XTk/HDHMi88v7zinHAcA8F5YlT7uqyF6+ViYM/bGzSvidWm8SmDrRoJpxEbULVHuiVNuqQqSeXdwz0hDByOR8IwcHf+jLvzZxwOJ95/9jm1Vl4eH7m8vlBb4+XlCRCODz8xkCtYD74CYyFEXPAEH/A+bAsNZOOxd+9+W2Lrs+ibsyko26q5MXS9TCc7Sp+Ldss1P0BtlPJCWRdqPlDrWQk64u29xi3CgEZzDlCaajF1l2E8qsBH+Skx3lHSNxzWQGozz/VbVLYra5trzeRUcU0YTTgzGa8gFdXCiyiF1QHBwtdjDErSqYnciukRFMbgGY+e6Dy1wVqa5eyZZEj9ddGuuXnO/PrXz1znmS+/OPHl5yeuIZtISqOkQlkV8O1MSu9hnIT3nymc/s3XlWVhE8WgNqVvO5hGvTfXi1Z+5iTkqqXdndxlV9KwFKUkyI7aCzpoU3TQSZ94OwS9s7l27AbYTI4e3y3mvh+fvOst50LvCV/nhVqUvaS62mIjlxy9N7iZ1XdOGFRHSWebVzgME+PxSKuND+lJBQynuBE3RCClTE5J67Yx7JZYbLJMqQTxBJOgrT206jJAVsfMVFZXFPkNA+MwMYxnYhwZxzum8czpdMf7z74gp8TL4yNpXUGElBLj4cTheL9tQjHPqBFK17dzNzVz2a7ZFprbz33D9xIctx6g54TdJ4tsz2nto/NJACmKwJdla+zRdaQwsTN2nr2glu0wQ1wbmEyVa542nhGJFMmE8oyrjWstFNJ2zVtT4FRweMIbDcC1K9EQNXJqe0qqQCogBakqCuldw3sIUUlRNWETfqp+5WpAWiGnSkqF11clOX1+f7AI0xRpm5KK+tjpZlM4+gSe48k62b7V9dCchuudYONEuwftZEbttfVUebPR9S5tl1R/sI3unHZYeifbINAuSNkapNLoVZJOF+/H3ozz3ccnpss2XueFNWcVJJSKhICTAmVFvXikiZBsHptzVfuOsw1YEENFEOZ5JpXMsix6411/HeOYmzbX4TBoTkW/FLaRTdctGfhXlNWCNGFOBcmNwyEwhoG7uzu+/NEXxBCYhokQIj/6/HNOpzOn8z2n8z0i8O3XXykZJo5MLm7eO0Zl1OnUWC2txRg1dw/do+/kmL4hay0bh2C/nW9vaKv1zX5/K0+kq6l1Sds3XxWhEiWBS0xD5HC4U+9t8+lyyUpukf7e9Ln9Xgg9/gLxXjUG2h20Lwn5hSlfcW3hyIyQeXx95GV+ZgwT5/GsaVBQMfVWtNrSXLBptb1j0TQFAC8aaVR/gsFrIxIHQvNmmITZFwrCsqx8e3nh6ZJYk85jf70qf+Lx9cqHV89aVg5HjazW4llShdcLslRWVDl3zZVvvm5G3urjxrWFtfTpvhnc0tefWFDV0wcDoEW7PkVk60/ouoE6PFLv1ZrKls93XYDb5EvLgfroUrF1/rcsZ++bvayJWir3dwceHo4KejhtAOkfSkcNFUNRNawtVVteYwyIE/I802Yts3lft7p3ayjNslRCcBymgVwr82rDJsTZAsrkrGSdNenNCFFN6JIKpVaG0eMlcn9+zx/+4R8joqVD7xxffPEF7x8eiNOJOB65Xl745qtfggjT6Z3OHbeN7UPQ0pX3xEFz+hC0zBas7ff2bvXNrlrgu/bbdrT91vfNByh+sXl8c4173Kcb/BZZb5UoGScr0xCYjveUWlhLskhMtdm3Ft7teXuEIQY6OW+xptyBG2A5MF2e8e3KkUdg5levr/zq219ynO4oZ8cQIuFwVtQ+L9pliJJ0FHfR6a9SFJMJPqjh8ZEwHnHOMYiG/FFAXCMZkj7XxLeXJ16umTVlcoLrtZIKPF9mHi+qgX88DsQBUg2MqZLqSmmrbqRV19LXX+tH6x27Yqq/pWrJrN4GWU0rAjU3at5BVSfCMATDcBREjEEYR3VgrWpUcZ1VqTgEbQyqphH4Jg3oFRcr1zmvJbwfStw/OUDnBHxU1RTvtdyhApNqmryFtDpJpNcrmiHCSgL2Thv5a81KtjFutHMeb3O7vbGieoNLzRlEWzo3/ruorI/SKvW1lGwj3N/fEbzn4e7E+XTgMB0oayIEz93pDh88JRdeXl4ZcmUwYQUXonnBHVLdp5Mo8NSnpG7TRreSYd1orruQooXcRudEOtHGau+bjbdjQ4HgtsajdGMbfunUkwoVR1WENziohbyomGZPADcppKrdamxA0A0weBuiWvivG3LAje8IfuLQdODju8MD1zwzxFGVh1wjt4SzzazBfSS6aPhC2THFBqkWGoVShVS0Gy5gU3ZNDwdEI6jgGcdAmoTxoOy4ELWKEoIaXcmVUhMFbeNVNp4KmapBe5MhYc2JW1oichtRNvPIbGW0HeBs2/rCxEObPa4Zo7Nkm+DbuFkDotf+ozTgTXB3M9brB6L4T19nj04YJxOdbMqmS6WSisnuhMAQA94PjEUbRErStsalaKlmjIHDGFmT9jBncdAiwXsdDxQCcRzxXodLpJSpssB6Vdqs9T8PQXEASTb/vWrI5nzkD3/2E+7v7zifzxwOB1wtrK+vhOORn/zBHxGHgQ/f/obHD98yjRPjMOGHkfF0VmDLPHr31m7z6M68/c5p7w67yzPRbrZUz9Vrl49WYQvQCOWNh+/XuaM9+kR6X0E1UXrxI1K1q61JYYiORqTllfnpEYkRGUeNqvo0lFqorbcWf4TR94WrgvAIHmkeJyfC/R/hc+K9ROr6jLxznE5nCiu5XimtsaYLrQqegYGR0U0cw5E1rSzri1ZmRHGCa0qshuvQPNE5pCSCOJw/KNIvzhqOBu7ujzhJPLyrLHMFY8CNx5HD6Y6cr0h4pVC4XmaeX0y8IkOrNh9PBGs8JvjeLmve2ulw0toaaZOs0osiyKYhZz1YSrgx/brWoBZV71NAzyI048/qyO6bc94ce5mubU022pX5/cffjGc37yzNJKOdcs+9U7Bq2wRgC9/+0zeFqcrchja3xaFGV7hxlCKbJyobhbZ7UzXPIUQONqQhpEqMgXEcGYeBIUaGISLFSmnOsa7a2z0vM8u6agnOBZrz+Fzs4uvnExElXSCbx9/nh9+AcG0vD93Ut374Ym41nts8fb8S/dxvwvwbL9MdQa+755wo11dcGZW45AQfeiNN++j93EQM9PPvrmX7nE50g/kR5xNjmDhxZK3CtS17WkGjtWLQn1JqtdBWNKy3eFX5T87iGX1ebY0qJrPVdKgEppvgRXPicfDQZJtwMwYFZZs1BmmrqyOEhs86M0+1F26u6kdZ1NYBt1+O7dK01tF1M/YW4fUr2dWJtczcsyP7e4OuHd15NtwY2P1e9mu9v+YPHZ9YNx4G7xi9enAzgIyucfINcZrDOfG0ZVFJ4FLJKYNglFbhuqwsq9Ie+5RMJ41GYU0rpRbGMeLF4Sl4MjVlLq+ZChyPqrjaUGWVu4d73n/2frvwThx35xMxRqZxYIyB4ThymA6sy5U/+7N/SU4r+BFc4O4MVSKSKnJNeO+ZDgeCKc6MoxqKEHoI79/k3yoNbUMZek5oob3rghw3wuF7vb0Lfii4V3sa0NQTa/hf9lUgKqaBBeSVQCNSJdDE8+HDtzxeHpnuHrj/0U8IceB4OuGNPaQLXo1kV7rtHATQn2st+t79oDm2FKpzlOGBwsBZHEc38rI88pu0kJuSlwqVK1dyy9Q8k7lSWiUFLROus16b0R85xtGMQUYEslO1WmkZl1V5qJRMywVfYBTHZ/cDtcA0RWLwfH4aOVXH0gZW7hhj5osvPKdz5ptvr7y+rlxqZbGWsn3EtBnoihmh3et6DWtYrf8sOAUfvW+EQa/TPCsTU3N3T0qFNalacPBaLc/Vpg8b87kLYdKskcZuQ7Mysv3pb1cYDz3f6W/SMkInEBp9IB+Yt6vN8hUjY9jzSlF73z9XvyACtvlvesK3HsC23RQNpXeZ6WEYOJ9Pm3cUbFKsRSA9SggxktLM5fWFZbkyHN7ho7MoA0u29MbVUqk3LJ9bks9tg0svO+2iFDcO5OYHUfe9ldLeXlDs7t/EN5tHf5NQb6/ZPUz3KA317PO14IZRdet83R4jNCMutc3LvBHS6GdshhWLs9KQ4Rcugq/4MhLdxCozAQX0skVjtEalUFom1aSbSUz22yom4t02GSi3spUZ9fLb69dKs8H0rukQ0XHQ+eyHUfGgMXgCQtbOCl2C0ROrdt9tnvhmxBWwUaU3oGyLLDUSuL1/6oSMlON1XJYTZRZ6a8/u2vewe/79XrXt/m7YQX/Im0Du9nnff3zSzV5q4/m6GhhlXVbiKLmwXLX0Fr0OBCimwomAGFAtZmVr7fVH3YgxeA6TFjlzUTbVdVnJrZGXTF4yTjxffvYORAiTw3nH8XTPOB44Hk+c7+70ptaCE2E63uFDpGQde5TWb3l5/Dm1VMbjmeFwz3S6Jw4Tp+OZk3nAcZo0X5yO6uGnAzFG1bUH24CaO3fjs+u4299F9LMKJpl10+RyEzP25phaCq3U7bfqiXZBhVtD0tVoWl5p8zMtzybD7ZDgGY9wPJ24Pz/gvPbVp7papGCNRRYVtZtE0gJKhYVptGaSV1XFv3wcwHkahYxjGD0/bjDnmV++/ppcGid/z0lUFFKHGxbWNesGS1Hn3Q+esxdy9aQSDNRTYC6g4Oi1FZKtj2mMeF9YD1oBCkNTscoYOcQTyMxYrqSS+fD1K8+vK0/PhesVchKTBm9430U6ddf1HnJ3Y4C7yhdorq8U7wy5cVl1kOU0Bh3PVRu0VUvLzlKR3qsulu4GwXt1UiXfmAEHrppn326AbF2e33d84tIbXNfMNOj8MkR121PJXFZV2xsCeNEmRlODshy3WW7HNvNaT+EIDoYYDHxT2uSaiwompErNheAD9+eDIvO+Il54d3/H+e6BEAeG8YC0itQV5xzH4xEfB15fX1nXRFkvrK+/QlwkHn6MDxPH0z3DOHGYDozjpJz48xnvA+M4KShnQxq2o3sfTI54Q9134KFndj0v3SKOnvO17pftupou2Xbc1uQbtggsxjOWUKuZlmdamlW/rTrVDxgdw6gpCyJc00KphZp1w6sB8JtD2UiaOzxtn1Gjq14bEh9w4sl5ovqGb8L9UAm8Iu1rqJnRTwSJXFsi10TNjbL2zEOx9iiOwanGAHhNFVCara4bx8JOv40uAI0QrcMuKJ/ce0/0A7llbaSplcvrwvPTyuUirMkpB97pNVcT1LURoAraGOXEaLKyRaH00LqqRn1tOggiBjifHDE6SkqUrJGJKHnzxsPrlzfVo1uAbivy3KSB34kbfMfx6XP24JjXypq1jNWH0t9NysluEug1R60vq6hEaza0vqFeQlR2OWfrvw7ZvLxeqVIrKTVVnxkGLXsFLX1NxxEfAtM0Eb2oOkothOA5nt8hIixpZV5m5suFNF+BRhjfEeLI/bsviMPEeDgS48AwDAzDuJFjvN857r993MhbdXCr/8nu2AbgcPO4HhHYY7b8+Sb50Dpt2VKFLVi3lKkTPTRqcjDeQZhwaYZaGMi0lhmCUnkrQluv1Fo2vT+oUJIucgmmm6bvoNabDX4rUyVqrLWpySm9NwxU7nB+4ItWSGWllJlaE3NayXlR0onGwkQbAlFpzFXJNyXrpJghTEZUqdSacS1zIIN4xA/M1fNYX1TjwEERuFxXvs4vzPnCh0U9+poapQrH05GTG1ivifmiarcOE7XMxoiz1K2PGle25+19ZTPeDb1srUJNSsleFh3hVDoG0CwNubmjigvo957+mu1RobO+yeUWx/n+45Prxo/R83wtzKlxHCvHA0yD5+GoN+yaFDTLaaGiIU4YtE1wWRvgOAwq8nB5eeG6LtpV5VQq+v6kUr/Pa2UtjcMwMkyjLtagDK/T/TuGYeQQvE05BVohhoF37z+n0fjFX/0Fr68vpPlKXmficGQ8fs50OPL5Fz9hGMfNiHgfjO+u7DjXc/2PiDB7nr7n6voL+9YRltb2iQH0v9Xdutu17COR+3lz1T5utr/t3qKLQFRQokwTZHqPtIJbXpCSGNoLriam6AjDaM0lakCCRSg16cRaaWGbitPb7lrNWiKsGVoCHM2N6hGdvhntQfc05yl+wlP5cnyg1sSH17/gunxAUiWnmdb6kAfH5LV/vlC5lIzkjMsZgmOMI4KwpFdSSYSaGck4iTh/xpUZyaKb3euY7OfrzOtzZilXntIzlzkzp0quwvuHE6fTHY/fPrMuqyrjVmOs5bpx3VvTrCX0MGfb7DfetqfTFaXyrhUKXOfCkrRG7/oQE7Pl3mv+WsueMqgam6i2kqjA5pvCTX+tHzg+MUDXKK2pvruDIdp4YVQDDmDNzUg2HWBTRyHNEU33vdaiut20jUdMB7dcwHnH4RAYmqL+0Udlqg3WeCLgqYSgG1aVXXtnmw09FBUjCIcjTBNhODBMdwzjRIjajtqZb86HjSyjG8BtnlA+8tC33/koHN8wme05b4EX7bZi86QdqPnYqtvW28phm3HpT1FJHsTcjdQEZUbKiquJsl65PD/q+rS5cu3m7E2ctcjqKznDFXQarL2Ddis+abGHvri2zaJ983TySmsENzHEOw4+cXILpVUWG6tFB2TNHmq5NhKC7zCvltKcxxFsOIW381uXYhOk6feAippC4FBHWnBMUUdoUxtpzmQbc1XL7tHbb90TBYz3y2NevqIVlH6fmyL52Uq/e3R3E753wo6ZbzE8IAY1eJuOYgdJe4WmG3T54f3+ibveYEmNu+PIGCNQQfQDfPNyVckma8bYKKSqRoV3nsN0pAEv11dSzgzecT5ESxH1g7s4EsbI/ek9w3jQ8l1WzzSdVLYw5iuuLByn9xzu3jFfZ66vr9RcWa/PCDB6R5gmDucHptOd9ZgrzXWcJqVpjpOF7XsTi5bBRCe6WGurfvZqmvD7Zt4Al57uGordk7Y3Gxs2g2b/3cpBfSCipj0WUm9ovCaDOVtQX1FaKw1JGcpCWx8hveLWlZYS8+PXPF3/H/CReDrhQlCP6KCKAzdQRVuTpal8mDSdZa/lVKG2oEG8SWObrKtGHMO4pyx5Jc8rrSQO8T3T8BkxHzimgUu+8FX9zVZnb1R88yoTHkaO04RHtqkzYwgMPpBzoJQR8JpNZHDN41sgFE9sjmOMHMNArp5TFF594sMdDCFzvWaeni9critpLcqPn00t16MpjK25Wis5qZhHGAzLqKJKvTcbuaFVh3nVppkuKlMbOmRCrI2WbhSNTeqFaYicDiNrKnzzdFUGqJXjbCCQ6dwJrN8fzn9azy6dvrrjOAoOVW0ptDqxeiP9cs4bwcNv6ikK3nWgxYgqEolDIA4jMaqAxDBMwEqtCsiEEHVEFMFGITXWlEhpNeZaYVl0HFAzYT9xvd+8b2rTcZebOeXS2XDdo8vbTfrR8VulMz7qYb5lqf2O0KyX1z4WmdxKbm1PD7Y0QR+hhqBq+666LmXVlZKYr6+4ODGcz9a3vldBembZzMN0cLH/3mh0ttJvXtPKK9o5p+9hE2ETjzhF04MfGf1EbpkogULdrrcXlSBTxpwam1b0uvWBEip2AttMJHFEH2lo7ds7R3SeIUSkoPPpRH8XXeNaGmnNlFwMz2y8uXS8/f8tAew7b9HN/0ttmPCuBWY9PGMz8N8D9bxx24rV7tQiPn5/33F84vFPwt0UqLVwmff2w4aRIkQJLZpC60zy8+mOL774kpwTz4/fUEtmiKLh+RDxIXA+P/D5Fz8jeE8c1MsOg3pdEag5EYNwiEpVHc73gPCbb77i+a/+UpHRdSXGgafn9wreDSPee9LzK69z4jAdON/dG6+60UQ5AE0a4vV9e+/xQ1QU3QxX/Y6NqH3p+yJxnVln7aS3fl82IbO2IfPuNjJo2gjUdfH09+1NWQx2L6RIfKHmTF5UEYZlhlVR+Qos65UPryvj6Z4f/eG/yXQ8c72+sMwz2xQeU/htooizAoJiU25QBZpe86bZSOa2GcUuE4Yf8MfPaLVaJ1jBT3ccAFknJcjUQrX20clHovMUyRRJNDxVBkXjfW9TdkjJG2A2yMiP3n1BaYVaE7TK/XjHw/jA8+WZD09P5KVyagecVD4sL7w+zzq51urgSnjpQxd3RmLwnmFSGbWUVCxVy2Z9lv3NHhXtRxeTN+ubeksMejmvT6VpSsFlzZtTrKY41HsAenKmAPbHSfzb49Nz44Mwr41cKinr3DHEyrPA3kpZDdAJTIcT6zJr5tYqfTyvD54QA+PhwP3Dg3ldI+B0oMx77W939uUdPo40HPOy8vT0gZYLLWfiMCFhIsYBcYGIo62ZXPR9lFIQ54y4s+fCBnjTlWmgewPDAb4DKdXS4Y3iCXuk0K8V7J6+b+weCehJzKC3PpG1P2s3LrckpbaVxP7/9t4u1pYtu+/6jTlnVa219sc5997utLvdLWxEBLKQIChCjsIDSoIwFiIveQggZKGgvCAlREgoFg8BiRekiJAHFGQRIYQQhhiLREbiy/jZYCcomNiOkzh0+vvevudr77Wqan4MHsaYtdY+96uN7XNPc8+82vfsvdfaVbWq5phzjP/4j/9wslIzYxSXc91q+GtjXVfiuCOlxDhOzKc7Z8f1dJSdR/AwQuQD14w4NVmbu/ZGZz1ThS3+J41mAP16QiKMO4ZW2MUdVSolmtVMcWQMkRWlsqAEF8U0v/asFXeu7g4hMMUdqlbx1rSY9zdMxHAyteEKiWiSGdWq/Tp/HRxEUzMv3XpD9Xp36xufVxPVsGKZDys7lS1tut2rB2Fav3teOuzOWaWR1Ylb3bPdAEGBi8Xn4zb378nYReTPAP+GH+v/Av514IvATwPvAL8M/Guqun7ccVThtDYjTKiRG2LE3TiLdfdeAliqte1trXB/99TaFavVu4sj3zc311xdHRiHifU0WyHNaLt5VKPY7seR/e4LtNqYl0w9rszHdym1cDzNBAZkGImjEJPVlyud2VSoqsTaEDlZqWpIHKcTMSYePX7MtNvbbhyDaYkVi9vahcLMy+51r0/vRRsP3MBupF5e1Tqp4GxfbM0I3FXe2HIX5zgbu03+ViulZCPr+G4QhgGComlAawJ3660h5UpiJYLFxRfG3Gm9vTxEo9cA5CPUBeIIyeLy5tdTPU/flRjsPoBqpWV7rS13pj0nkRp2hH3i0bijloXj829QywmRSJOAaoS6Q1WYraibUDw8zJnWimHXkqi1cTcfaWrZmSkdWNfG++sTTutM2g0MIrQXK1WtWcVhlyjV2k6JsDUjOS1WlFWdgx9QSi0bku5O2GbOG831Q6xQ/fetr9r+gIsbsi1c9liKP894RvF8QeiLwCd68Z9s7CLyg8CfAn5EVU8i8t8Cfxz4ceAvqOpPi8h/CvwJ4C993LGaYuIAHmDEiDd1CCiGaO+ngRSFeRVWMR3z+fjCtMQRCAmJeyQm9odHPH50Y22CltUKNxyoUDXwbzrsma7e4nQ6cf/+d1nmE+9955usy8y4uyINxnhLw/BA064054+rUoMjwR6bxyURh4Gdy1LVFqjV3O0Qqn/W+iBu3u6BH7Pv3+K19Q9uUpBNHUbVOq5exvBK69+cY9+X/j0vMHac1nofOd9xUEKyxpQ1WhEPrTdJqESy1RWIM7q2C7yg4LaOtPvMW+8h36NpD3qWobYsRzZj12AAlnaVl0rJs+EGyx2UBR2uaeOOlHbsx7do65H64tustUJs1KBoDUidjE/hoSBiYhyhFUspoogmWm0cTwutKft0zcCeNR855hdUKmlMtpGwboIR0xCJ0iW6xSoD1RbfXKwlVPVgudXe6go68u62s3kX8iCjseGT6PYQzx5d7W5+L6S5WCxiB3eC2VBXsfmwxeTl8b268QnYi0gGDsA3gT8E/Cv++n8B/Ht8grHbenee6J03PAwTh6sbRKDkldXjrZi6ZJO57OPhig4ACab0koaRJgYuxRgYPdaO0W7Mkiv3332PeZ55/uwppWQIiTTuScO0qcXEIbmgxGDI6kv58tDFJUS2uHpZrIlkHgbyMDCkAfa2KwfvD28ToW0xN5xd9VItHo3hAvgLYuminAGnwqpuNfDdnVft8kkvpSkfuO+XEtCcwTQnvYQ2Q13NvRZhyZl1nqkkrm8fMR6uqDmzzDMlZ1qtjqd1IOniXCIw7GxBTCMMWyM1v6aGtrD9rsfziBDiYBVgu2ukTdQwumJQZT0ttLrCeEsM40Ytri2T82LUWqqBb939bQVKpooCkdIqQxRaUEqdmddKyavxDQRTEorC7WE0zvwAb687SjEpK19CTMZM7wlrIXoqM9fKWnp/wa5RYMa6dWuxR3GOzWED1Lozd6lJsC3+fa+4XCjcU9h0Bi6MfNMs+Yjxicauql8XkT8PfBU4Af8z5rY/VVWvweFrwA9+0rHsWtsGwgXXTt8fDnzpS5+jtcbXv/FN6xaTei7beq0N08TN43eQEDkeZ2qpTLuJYZysOkgXU6XZH6zDyTAgIfL829/i3W99k2Vdub+/BxH2h2umcW/EmJRM4dWNe/Sac1NE6e2VXUnVjb23lDodjyzL4l1SB3bT/nt/kwAAIABJREFUztDklNgd9oQQXGnGQLmmlSCBlAYUJWer6BoGq8WXEIhqvPO6BWHdUMHyPp1H3cg5e4wbNkCnx3RhS8E5SAibsbdaCG1lKPdItegXgXleefbsjv3NYx7/ns+ThivyulLqC9ZlppRsOEmID8Cl3lZHxisYD5a9iBdTS50F2aoRUizfZGGAQJTR5sZuZ9deK7EUypo53T0xY9h/niSBfHxCW+8oZWFZ7qlBWQeL5xMG1NW6Usts5bEe6kyDbTNrubce5xUoEFNiiKYo9LlHjj+EAyrWMbasjaUKT1dhXgtrrdwdZ6qvyceTCZdaYsLc7jiABKVmZ31iNecKvYckl4/X7X5bDLqUXTf0LmhBnxdiG7vISzv6J+zu34sb/xbwR4EfBp4CfwX4sU/6u4u//5PAnwSYBlMP6StaU3W3qDDPs4MpkXFw+ulgSrE9/dVz1+OQqMGog9qKEQ/GHdY11eJsqu2M63KiFnMh5SI9tqHCFym0joLK+eIdNZUL9Fy2+9pVR5qY5lmOmWWZqTUh3rFEe0DVwRS53H17X3e2p/6B57Vdj5/3Ika/nDEPcAEud5DLI8r2K1uA7JtaeieYgbS7IU5XhLRDQ2SeTyArpfVe7s27mjpO0HGil87Rz7vp1V9iCF2V1VH5Vg0hJ54XrbMimzXpsMXW2k+p7kmtkMoOkUbF9pwg0VBqCdQOnnWlRg2mM0hGpXmH3uAxsZ0pJbPIXl68VmfEi2nOxQBDNEp2B9BSDKRofQUu1cDYwhR9cGf08jl0gNP/ZjP47YZ2MPAhZNPdekG3xeITA3a+Nzf+jwC/qarv+sP6WeAPAo9FJPnu/mXg6x/2x6r6U8BPATy+nvTqas+yFnKtLvVrcs5ryQwpcr2/4Wp3w/7qinG3o5TMsmbiaK2RrIDBuromaZT5jml/xdXjL1BK4cmT98k5k9c7Wl1Z59Vcd4nsDleIREvZddfcRSA7nz3KSwhpxz670qp0jXs1ModCoXgRTmGdZ2KM7K+uSGkwANG9Benxcwdb3HuQizOdT9sBHufYy9nYWzVXdiPnbHUEYVscegXVwxx7X1OM/bZWoCjraaYu96T957h9623SOBL2B0ouPH33G5Rcubp9xDjtkFBBViyXHoBg7niwc4n2z6duyKt5G37vbJG1YppSA7UtLMdntFJIo6kLGcIeqBrQMBiYmQZXAHoLbbfE9Yo07SmlEOcjqo0xVII0ljiwxsI+jdyOO0fXA1UrOawomSFOTCSKCkuxuvj9ZM05owrShKfzyrw0ihtWCpXDzkira4FcgRH0yhST5lN2xp1AtVRYLdttcoNgM8z+vDsxum4G70VQbsUixhJV9Tw9em4VJedw6pPG92LsXwV+VEQOmBv/h4FfAn4B+GMYIv8TwF/9xCN1EkowJZDzJmBaY92gTI11YBxHwAocpPe8VbYqsqDF4k+P45vCsq6s60xZTrS6UotP9CDmqn9gZ5dtt49dMkocHbkgzfhddWM67/DAps8mYvi0Yuq2ImEzOrBY7AHxtB+/x9v0WEwevOfS0DdQ5+I4Wz25LwDbDnGB0n8gK7ABbdYV1QpoAiGaeqt4Zdwynyi5sL++3gphegx+IZvgIpbn57md78Kz6aGEbBFnX6zUEX7nL3QMQARiuoiFLZxqEohxJKUJNDAEy8/3Pn/GV0iIOHffU5zbU/NDBwRpns4FRKKtYfWi+6+/PwUDFYcolCibxPgQAkOyDEz226HbU3ypcUNfAy9+lP7Nh4y+Y297z8X23r2zTYmwT88PPxTwvcXsvygiPwP8daAAfwPbqf8H4KdF5D/w3/3lTzpWa8pxLqQg7IaBYUgc9jBOEze3t6ZQU51IIAaWqQgtCEGV9cUTYgjsH73NMO0M4BMr7PjOt7/Ossy8/963yTkz+C5BiMQpIsFbEofAOAwei0cIwjSO7CbrtHpzc2Ng1bpa6k87k+/Md0+p78jdnbP0yziO7K4OpJQ4HK5MkTWamOXW6QXYUmc1o6iDc4HmhSyWQpwu8tlcAHJdjUY3+q1NhAv3+SJfb5+hmeRUzoh451pVQj2hdbbXEdrdE9rdPdMoDAehroW7J3fkGvjcD3yJm0ePWZcT83Lvk6u6V6QbSKc+/arvliHYvVI3vNYqTVdabQb6tUbaXdlCFZJBy74IxXEiTQfQhubZMhPdNVYIRIYAwzTRWuW0ZkqFXTiwm66pmrkvs9f6FydADYwSQSMZIWtl1dX8t5qIGkz8skEJjWlv7v4wBnJpzPcjoUDSSqaxS5GrMbKWyjOq6SkWJ9RgXYBb0625gz2jD7rucJHx6MZ84c6HaKulbGvnB1eIjb7wEeN7QuNV9c8Bf+6lX/894J/+Xv7+4jiU3Bh2iRStOo0A027i6nANiIFvteehA830TJCSafNiKK1gXPVgX+v6gvu7FyzzieP9C59o7vaFgRAHn5SOsHs/uW4UmycxWZeXEIyokXM29lZrF16AbO68fyg2N1msNDQNaatjt/a+bROhsJgULPDypgyu12YlqgY2Bfc4+k665ecvds1tp5aO4NrvbEG5iPFbM032Vq1AzZO/ogVa2UCgmmdaPhqoJIqujXVeWJsJZI7TjlzyVpZpLa+VqGevwgzaQEFROWcxQvCFoPrCVTcuQkgG0Kl4FZ9WtJnxp3EHrVHqgpYzecQ+cbCqyJhoIpwItAZjHEhhYK7KqR6dTVjdYwhYU1ALZSrn6sPilYXSXBxCrF1UisJhl7wDTWKNfn2q1sREhCEocw6Esq23tvVjlXq1nn26nju/jOLPj0wcC+krwlle/XJYbN9dKXq0+bHj1fd6i5HDzS2Hg+W3Y7ImhuO0t9cloE3ZRWFomRQiu2GPsCdd3wLWaeP+/kTNCyVncs6cltVYUNMNaWykcU9MhrLGZL3AQwxnhhvKmExBZhiM0GNxkjcBREjOyzcU27qMoMqajTsUXEZLevwPlJxRbeQ40FLcdONtbekP15buZhficltntVkDo+zFutUM+K5+CdC9PBR6lVntTSZ8Z1e/t53xpS04HbR6fUAmKozJjEfDiETl+iAUTQwxbohzjIFWM2U++WeqhGgda0MMHrea4pB6yy2KiU+JZyOCBGK/ryVuoVzrajip96h3xdo0ESRZmnH7bM0p0iNo5SZG8no0+nNeaFSMc9+BYWWulazKIMHahEVhP/Eg5LA1wTrBWAlqJIp5ie/cXnM1TRSsLj2XyrwY5iSD9SC4v7c+BK0ZcaiUxrwWW1D90W0NJDqTkYfg26WP37RnMM4hv5GVOBu4Xnx9xHjl/dkJ1i305tEt4zAwDcMWUzY1t0+1MUVILRNTIu52xDQw7q9oqrz37nscjyeOd085HV9Yp1UZvfLomhCEIVkL5RCtM+pl0Uq/QSlF05rrddmCCyZ40YVPupgipRTmtrhLbLthb9mUumQUbuytkdNKaxFG8UmtD2MuQL0xX/QgcuPK+85rk6JuHHt4KR6WflcvvAzHD3pMXv3ve6DaswQ1BHI1ZZ81Z0rOpGQxaQzQZEIiXO0FxTwxY82ZdloujbLc2dmDiYdYeDPYtbfqGnyu9eeSM8FrFmKwhUCbksVKSUv1aw0DIY3I5t0IIe5oUpHmBTuKVYuNE8OjdwhakbbQ5sDd86fkstjOF5JxORLQlGNZrAoyBYIkK5GO5o2sy2I7tndsjdFSmCEEK5sNwtu33to5gUbldMo8fzaz1oSMhbVUIpV5bps7nksjBQv31mbGbErHtuMXL6/esJX60GZV2bxdnz4P7HrbQtrH2vorLoRJkbceP2KapouS0Mh5rqrFsU0tXhsGwmDtjEurPP/udyi1crq7o6yruaWW1AQJ2wou4kbq3V779+M4PvB00vae4BPQe3+FwEByN/ks8JhipF2IUuymncflcTN22/nipjtnLluFFuidWroYo72o/veRXqb6YJlX3RR0OxzTi0v6Dt8JOX2yWBgQvFinWVkq1f66eeGMAsMBIZB2ixmFNWWlilDW4rtJQ6S6MIV91WJpzTRd2ecddu6uh+38Wy1/D3cc4FKU5vp0/T8eiFM6x3+b/ObZ1OYMPPpnC37P4HS8J1CNahVHxnHvC1whs6BayXVBQyM0y+YE79x6QXVi8tLbUrL1HHRUsWlBdAYVSnMcpzqwWOu2icRgi+VhskaTtap5m8EatDeFoVkYN+dKroLQCVG6UYvVhSlCDFuZdOtNTC7AOcedN1ptjNYf7qPGKzX2cRz48le+uBFUorcuDmKyQ6oVKSeaNMbDFePVzYae3z97xm/8vb/NusxcpcAoAsMVcdzbbugTLA0TIoFxHC6MPTKNE49ubxFgXTst8iw4EZMBhib3HGG0OHJdVnLOiAjjaGmg4Hn/t996i91ut5FcSvG+czGwP+xJKZGXlVoKTYRW7eHF5J/fQb6+8NlK3y4mIBtLbuPQqy0etmubO2ttgnyxwCqikocV2rkAFZtQtVCqeRWye5s4ZiYJDPlEcKZfXmdOxzsj7GgjRNCWaWWhrDNlPRHSwO72HctkbJ6RmIKLCCF1FRt79j0er3g/t45Ve3hlYX3cJnXHGswAKrWe0FaJwdx5YkIHoZbM0yfvEmg8ngLTdMWBwH7ckZrpHuS2cp/fp2ljSJ9j4ICcKjp7xaWYXsLVwXQKnj57zmkuJBfMUGm0eHJQVGhNkKwOu4iToUwpKUWYgnkz69JYVyWXyJS6YIVljZ4eM8e1UkIjFDtPc+wDbEEZB+sVWHJx/Eg3nbu+AQXEtfGF3SCc5tfE2E11w2IfeozagZ0g0Aw8kxZsB5+XDXc4ne7JeaWWTAsTLfa8d9jUOM+53A+67HCOZ/H/23nNAIMDb71mve9IIQZC66+7yxoT557qfefytlS9b3z/vaPUD6rZLkIK4fzag6EfXtjyoBbej7hRZj2mbxd/s3194PhnRLdj6OcXBEIypKpWoHdztapC1GWttLo6ShegfBjzXn42IbhXUrcqOafeoRvsfG4t9eBzmoNhoF9rBGkbftA9OjxkKWqLRhgnZF2gVss8YGIa0fPovfmCoGfV4lLRYCcLwRtEdOWwagupCVuaUGrTniZU8M4s0qAFdZEQYRgCvVdhU90aQSavxEQDRHteVU29tidhDOLptQyX6bUOxn5gKnzsePXGPkykcTI6ZclQMkJCZESSsEsj2hrvvv8+z158k1ZXWllsd1wWY0jFA20wHbqUTPQwdbcxnFlxgMd2jZxXXtzfm7vlhpbiRBrHTbhgHEb2e0Pj+65pFFprOJFcQEMc8NNWWU73bgRm7FcHc21t1XehxkE2RZsOOon02n15YJASYn+U9NJV21FMLkvVTEJ94RG802stmwsoWNPK8yKgW317L40UGqme0LpS13trjeX/SRgZb4y+nI8vaBJI+xt2V4/I63Mod0iLRLJJTE23EAe6/YprseMLKILz4pW6rpR1NgP1HDo62KNqlh1Ahj5hHHW2PuetQtWCYCk/QiINA9fpFq2F07NvcZ9nrh+9zeH2HeTJSnt+BJSr6QsgkdRMmbYSqMlaYpWy0jI8v5/NNR4rhwk0C9RAbfasRYT9NBJj4JQzSylGJooVaiM0pUnjVJTclMM0cXOYCGL960qtPH/xjDVnFu9KbJ6byVWtazE+f/MiGzUhEes/yBmdV2ja543VyIegvoB+9HjF9ewOerl7rc3leaUDVGGb9POycHf/wqql8mx/ry5rFKLtPB6rhgskm3Bur9QBry3n7NVjHZDrBS6mfnJuCBliOMe/2Ipt6ihurCnZBC7ZyRzmWgTOhB8HUs+7+KZe43DKSzzTjUJ78fNGilHbYrbd/cES7phCOzfN6DFfXyz6pVwivEKDlqFlB716qxF7i2EhDYkOlAWrIThrDQii1s75Ei7a5pu77f0zi5zRI3VZVtng5A2xPB/r5Ymrgm4SLxbnhy0MsvRmrYWSF9shJdr7qxFqoo4IiaiF4HXqjb5T2m5ei+2u0+DdU6vV25vDYp8j+mYB3rapHyOYNJWooNKMZuuZjySRMQ6UIszJXPwhCUMJtNBMs59G9S09BDFPxhd3wXEYukei5znQ43c410B8xHjFO3tgGne286oyTDvS4Wq70GVd+fo3vsbxdOTZ86ccT0cO08T1zosrXAMuDTtC8Pg6hgdtj3f7PRKEWqzl8m4cmbxJobqbfu2El91+b6CdT5zYd26EYRgtY5Qz1dsW12Z9yaLXmoutAly60kaKkQsmXpdS7mBPc3adeEVdXwC2m7S9b0tHeY7c/g5s2TwvFl1Pv7+Go/8iEN1AijaLlduKtgWpJ+LyHtRMW0+0XBhSIAWhOnZCGNi9/XnLg6fEvCxkjTDcmuFP9hxkmiyGFuexS0TEMYPusfgzT+OB4BRY/D1Ui0UljSDe6TZGtLWti+s47uiMv85pEAloLbTlSMszsjwlLC84vlc4PX+fUjLp8AM0zdRyD1ppwVpAV5ejbtpoETRaRgGUECdEkmnKJcMRwtp5/IVWGse58mKtxBQYx2jtv5KVQ+daEGlEhJYrGg3kHER5dDVS20Aa4bjC8TRzfzxRo0BMXt5azd33Xdweb9uQfbW3mCCoi6iUotxfto7+kPGKjd0Q8L5rxZQY9wfv/GE64e8/fZ+nz54a5bWs1hdueGRu8HjlOetwQXt1jbhkElW7vYEs8zxDKQzjwG43YbAHzls/MA6j6cYPA5fKJt0LCJ0Q0zzH3TXRwSvN5EIyyj7fGUk9S0zhSKm9zpk4A0hy+LvjCA7AARfudzf0chas5MJz6de7YQSX7gLndUTN3Q8108qRUE+Q76BmY5g1I8FE87lpLRNiZDrcEMY9GiJryTSC6wlEJO2sYjANEKLpzGOhSHBDrt5rrmMVMVovv77jq5prDmqeGuIYSqDJOfyIyV19F9joO1irSssLmmeknJByZFkrhXtk3BMPt1BnyvLUFrp+rc5daKKWAhRFBiXQEO8vE0jWa6DrItZKXVdqzay5Ma/KDmE/Gh5BSBSUIToF18U5kI5vBPaTcf9bENIIaGFZzGvVIMRmlXKCEtQ1HNwDaq3hSQRa68o59l4LcfrC8OHjlZNqLOywSbosK6dlYZ4Xnjx9xrqstKLshh27wXbj/bQjxNG1zcJWfhqdtpoucuhg1NmeZpvGicNhz35/cJuUzaBqqayLYQHRc/ExRgMQxdzCVnCX0ZlgIZw/Q097iWwSzcBFiqQR1HvKoxsIuG3I/j87jL9HxCjDHqP3mCyIx7ce5jSvgujprc2DcHfegCBbGKLXnm8FMSJIHIGyYRs9wrD21kqaArvrPUhA779FmxPp8Q8R0s5c/rKgJJqOlk7zHbdW59nHRErjOQzpYNIFYNiNun9uxMpNgyrnug4nGIEtLphx26HatgBXETQOhNsfQMrbkNUQbgm0qgQZOTz+ClApy3NqWWgBcqgmkLHfoxQWntI0I1UNPyAgDNbUwZku4zSQJPEoVMaheYfajqRbSDJGWy7IDS1GRV4kGJAXI4pwyivLWkEq+32kogxNqQ2CBmqBuVgjiV5kKFGQ0MlNlqOvakBhbUqoSskfbe2firH3CTofj9zd3fHsxR1f/dq30NZ4dLXjMB4Yd5NpwTnqEy529GFIpJgYh4FhGPzDmpHlYmmLm6trpnFif9izOxzoBA1tjXVeqdnLOkUYxpFpZ00qou/0yzw/kH7eMAE3pPOqukWdD/OhTVHaRgmNMVqsz4eEo04fNTzCWGWduNPPG0MiRqi10ErZFsxO+0UEqgFOtTWySzhLSqaW60ZnoccImu1+YLntIMqyWoXhdZp4dHWAujK//1Vqq6TD5wj7t5FWkXwCJpN+kl5hpwZ0lUJMoy+sHSLQTc+tu/UhmBfRsXoJEIPt3ta7rp4XWfB/BVrxjq3W/aU1Ky/WMJIefRmRSDjdEecTOa+U5UQaJ67f+gohwN13/y5r+S6rKCU00rjj6voLqGSkrBQ9UdeGlgIkaMEkzZrxFPaHiTEJ41B5tDZOWrlr1o9OHQSYUmAUrJqzFbq/oAEabuyriYJIhP1VpKGWw28wajTK8qluhVRNz/NQHUMo0oji7aqrknNjLq+TsatyfzyS18y6Llb/vVp5K0TrmTYkY7YNw+ay0nflcM5L991MnUIpXlvcc+UbqWMDAGXLc2vzPH/s6TabnM3jaeDBv5c7qEejnOWkbTfqYFH3QIwQ0Tx90uP2SwFKT/5oR97Ve9ZfUGI7AtTBF2VTyun3E8yomuvuq3sc/Ry6AYVOWnGjW7N1O23WUM+9pkDAC09aoXm5Ke5OCvjuUsnzcyRGolgvN+PfN3N73fU8E2va9nkv4h46UfRhqq0vDOc/UZdq6rl5S2dYr7foXoT08CAkSANaMqUWdIX1+NzVg0bG3WOq3FPbPYHGvN5hqhKJqDtqW515aKo31lG2IKJUjVQVcqmUXKhimSBr1ubhEI6XxEAcIkqgqjiuacY/DJEgI1WrKe10EFBNFksQhtQY3V1v/sS3kNx3GbkA7kLYuJQfOl65sXc1mu9+9wljhDEa4nnt1WJv3d4yjoMZq09qEdmALbC4P3lNe+4dS2ojSmA/7a0QJaUtD283Rc5ifZP9e9jtGcfJ4+Fsruy6boYZ44WMUp9jnBH37p7aMtVXXgMN02DpOLC8d/BFonF29YOeJ7qtCYpS2FoOq3rocAYAQc/n70i9Zx1KyZS8uifgKTznxYuIEZiaEmuhNOX+lGklkzyOTkNjCMoghXL/PgoU2aEhUTUQq3kUYRgo+Z75yXeQENjf/gBx2FNlpIYBaY1Yq6URH3gzbr39n84L8EWtswcvBR/sc/ZQp8ekzo0ISoi6GXstxe5RMkmssszM89GQ7tNTUkxcv/NF9jdfJD7/Jqlmlrby/vOvE0Lk0XhLDFcs6x05m2ZdKyeqZoocEYFVE60Kd8vC6e7EbjdxdbVnvQBRBxEz2F0kTol1hePJnrG0BRHh+npkGA48P554em/FOq0IqDCOPSXbSNYEjaqBqsqpnJtH9vm4YU3JFvOPGq+4I4xyPJ5Y5oV1WZEhkDDEc/SWSsZiSz6BL3Zz7SQaLGbvnVF7PJMgDQPDaBJRKaUHab7AWea57yydFKMiNJfK2nZKeYiSP1gxxeE87d/bLw20c8ikKRp8N/Nd9kPJLbw8sdX/RM4vy+XOdwb8+vs3NP7hJfY3cCaEsgF1VsPeN+CAxAh6DlVaWW0X1eT3+IMFOJZBc1KM+ebmPfTzOuNr++wP7u3FZ8Wrsrvr0he2i7Cof6DupXQqsx8QuLi3InYtLjdGqzR1/flquvkmDzaRC2g+2S4elaDGQIwhmNdzoe9nn7lthUUaZBMQ7YIdl89m8wijpW9RJ3/5wZTeK0FovmALbKSwIQqaAkUFbV5D7/TaTb/f52EIZswvh4iX45Ua+/F04m/+yq9tqawQBuIwMe12XD96REqJaRqJIWypsxCC79SRw/5gN8arhKbdjnEaN21zicH49GLuaO/gOgzjdg0KGwGFVsnLbPlQb+3bfBJtU8kn4DmFZPTOLTRAPG0Xtsmmqiaz5RmHTtLpEtIhpQduuHH5A72JoqX+hm23tta+ZzLOA9mTDnq1RgwR0sglb56WgYaluQJlXSin54ZeS0DiyLC7JqWBMr+wVN+aqfOdAWjDHknNOAXFyTstgwzE/RfNoHZvIcPE4NJRHdcQ6kXa0lzVOFi/PdmeBtuCq14Bklshr1ZQVDsf/rKHnpw9neYNLx4wBSUSU2Da38A7Yuj5fKS0wv3Td5mffJPx6obd21+G+xesx2oEqXLPGmC3G7je7bi/X3kxV2oLlDIhojTJtFgYx2gafbXxZDZNgJhsU9JibnsLDQ0m0jkEy3ZMaiIsL5aZ+d4YiFdjoubCslp58zB5v8IwIFPixaq8PzdUK1osvzYEGIKRbxr2/RCFJ8+7LOQHxys19lIqT58+ZzeNDIOj646ET260HV1vrVldsTd2SCmx9wKa1UG4aRrZ7fcMw8DusOeyT9rgu3YXroSL3aUFWyGdNy4Stx1ZLiQ6z/HwRQ16Hz2e797HS7t3LWdVWO1dTrWv3OHB8emYxJnDyaZNZ37t2WfjfK4tnr3YMXszg00aqvcKdrqqcQG8qMTj+NB3QAm+myutFHs+qXU/+gJHaOYNJJP0lrhD4rDdf6PEdjpn3QxTgdAS9Hb12+4sW4ZiQ/B9cdwEOkRc18KerwGi7eLYZ8+nHyykgWHaE+pgwG0JlPycuh5Jh2vidEVcCoNafnvVBUQJu8QUIot4KrQJUh1Nbw0NzQunEmXNrKtJb0UvY67iWRYURJGgxj4GUjOSTjtVlpyZUmQcEjgbEpwzFq29eUSZa9uELDq1uc+5/qytsu81itljDDy6ueJwMDJLShZbT9PEztVjmhc/BAmEITBNEwcvUOgTbZomJEb208Q0TgaEtObYjElBd6EKUag5b4uA1sp6/5xWC8O0Jw6j3bCX3cWeC5czI89+7oCTv8+9AfVOnuq7zhZmqKKlPqhsKznbxPdjSA8v5Jxqqu0cs4uki1CguVAkdOWWzfa1Gyab8KKSjMddG00zIMThYNpuWkELbbmnzM/RVnxx3RF3kwFdu0dImpDhsKH3TQMSBqbxYCmxEDwUt9qD7mp3LAF89wbQSsnqdQiWSizFRDxkW976AoovVFiO2w2tZ1U68QixcGIrxinF22EpaTB1ov3+QC0rx2VPro3dcMUw3dJ2hWkaqbnBWtBaWO4b6xyZi7V4BpBiBKhp2JHGCGKNJpsq+2YEIIqBcMlCb+sn0EwtaK2rMTXHHVEih31DXODSKhGtZ7sIHHaNMQKxoNLIszLPNr/fejSAwHFxDbxsmYvSlFMtF+DvB8crNfYQhKvDjuvrK2O1OcowDiNDtHrhNefNWEKIjOPAfrcH1IoyVBhG04sfx9GBOAitIQGS71bRAb5WzfVEjPLZWqXJBatOAAAVS0lEQVTMR2pZTXJ6NMLNpmH+IJ602mY4G7nIy6tnzx37vm9oibHs8N2nNXhQ2ebdYrbj2YQNEi12RrfebSoCcm5L3CvsLN9uLtvLraH9EkCF6nn7pr0fnCDDziS5yFBXdH5GzUeIpqEf0mAKMXGgTW9DnCAZ89HiZVuUhnHyoqYuvtYV0dR/Z1dj3szZs9JW7H7IwCWnIIRzMi4EF7L0yduc6tycYYbvnFvsLiAXHpNp3IsvskZqqSFwl0aWnGlpRxwPpOFoyLgG2mJA7VoKGbHmGTH5gmph1JCuGNKAMqNkqlZDzGug5gtMQ9SoxDVCtao1jQmZIjEOTGOFCGWt5FpoVVmzxfRRlYFGDZUWTBV/WZVpDFxfjYQoZJSyKlIaNOvXbp2WPtr+XnkhjKmH2moXnByDWGlkUKt5j0DXZRMx1ZUYI7srk4sed3vX+3bwTXyX3FxdE33oqZAN9/G4Me0OxDYRh/G8q/qE2UoHt/r0y/jY/+1gXvc9O2im3nNNseopP24H+7ZCE184zrl7P4PvZOogmnJmxvW/NTDG68b13BIanBvdPRJfrOoF4t0cDBPtrrMrvaYBGFnXlXI6kYbBqsVist5saSLsbo2WPEzo7goNcat3VxdWEK9LuATjtsm/hSU9BSlnluBliCCC1kxbZ8cz5Iw9aDMOf2u0kGhhcADMWn51oNL06YvVrGMgYfXS0HGc7F7llWfvfwepmXT7ewh5pgQgL8Rc3BNStKhp3e2tdFolUzRvqLdx7SJKI0RFxfLdrQpFhRwgyMhtEp/L5UEoFaUSYiPtA+PnbggiXF8lUhLW9Tk5r9zsE1/6/GSxv7faur1O7FU43kVOYd2yVUFek5hdxAA0VSi1MUQ1sCsIRZWIMLlEFK4TRgiUVonjwNWtgXjjMBrJ4iXk/BL0aqX4TuuADrjmmzBd3W4Gt0WNbsvV3cbBsQMuJ61PRitqoScCPEUoRkyu5xrprYTWPZje7LGnFLuCztZhFd3c0+rH6TvK5UJxDmnCA2Pp2m84+LgBjm7s6h8jeEedIMlKM8cJDcpyf8fd0ydM40TIKyElhroQhgm5/aIRg6YdMSq5Nk65olq2c8TBKMuX5bZGjsH76MXtWaG64RpbWs3xCc0rbb5HQyQk87woC1qztZgqCzpcUccbk8Pq5KtaPNavlFYITazabjsX7HcHdtPEejrx3rOn7K9veecLX0G1UNNAWY+04z26rpR5pq0nhjFxuHqMBCjtPWpdSGEghERAGDVaRidZiLUuaj3do7CGwJQGHo0Hams8vz9uijyqShJj4qX9xP6dR0iItpABPJvRfMfj64HHb92Sc+PF8yNNYf94jwyJ5+PMXYyUWlnWTAjLR9rfK8+zn0ksgSEZCy6m3vO8l4JaXbvF3eYyjqkryUQe0ERfdqr76k63TTNn24llM7weWz5IS3HmcL9chbYFxmJ/I14TDbiBcuHC+Q7a/7vYnR9cayeKvNS3R5CNIqngfcLMPVZV7xbT01NeNOOvb0BX9xj0nEXomID5Ks1LSk0ZCPW1SqFJoKUJYrDyyWoVZbEWWnF9ty1mYbtnwsU98Gu0ZwUvhxm9PYI5Y30HvqjQ66SpC0VdXK+N5tr5taCIM/n6AugeZIgGttZs4VgriDZKXmitoGUh6EpoC+QjghJ78dO60nLeRE1CisTUIDSkFVvgWoBiensa3OnAeet4bwHXAWgNTrNhDutqDMchWgOL6NPRdPd8Y2q2+QwxmYhKHNFhINCo42gCIdi8GYKyG5RVPIb/mPGpiFdMLue0308c9iY02Zppuu33VpxydX3Dbrc39LNU47u7KqzvlcSQzgKH7va2B7JO+KSzeG+jXHoE3PPNRlM9y0oBW8wcuxgF25JB80baonYluWRKMXCrZxM2BdqL0OIsaGgxfisXLpcbRAjWHZTq17i57z12rU6l7YftKSeLeTfFGl+wpFmqJnjeOcZoaqzLHafn97Q8Q1uhNWoTKok6HKg3n6eqMs8vkFoJpxmJ9+TlGev81Lq0Tjdnzv/FvRPfrUMIHio5LqGXIFwPgwLdTvM6W2oPYDigrVJd069lS49pBarJbKEnWixksxb6zIhxYB8TdTlSTveIVsa2oK1yev6MZZkZx8B+iIw5o08WZJjY334BjYmQV9blCNNIGyZSUsbpCKFS5ITqSnmutHuFscFkTRtO2UKlTKBioeguRu5PK996eu+LcyEE+PztgZvdQNVCpiBpIE3BcJb7ihTlatoRD4/JumfhChmUR2Gk1cp35xOn48JVLFzfVl7MjWWTL/vw8cqr3qyGPGw7fEqGPrfGWQvuokmEVmusF2Lcem8//DyCpZfOKn3bDnqxO+vFz5sLfuECW3EGZ9S77zIdVb+YpNtQvTiuewjd4+gomVwY5cUBZMMA2Hbxy2vqb908DD/HVoHHReDiWYqHC9b2qR9+tsvrOwMcru/mYYdEp8gqzb2T7m5bzfgKKoSxWmqqe0y+s25cf7uRtrx1wPNyl3+AgZyvlU1Jxo5p3ox7akQLPS64Bk0tTWsLSPf6Oi+he1f+DHruXh0P0ULNy8ZwDK5tF2IiIQwKMVSUFSgGdGKbhFarwEPaVntuQhMRSbJ5J8KZeCPb5zCJjyoB9Ez97vevL5ZWUJTINSDBayz0vGUFURNviReVjx8xXnnMPk470mCu+jAOW3PF3bTfXHYJVk8+DCMaK+qEii6esKGwzTTINR/R0xNrOzw9BkmbThtq4ExPYSOWexcJThBp4Hlm7e4vHuM1S/MEL4k0/wok+kRr5rvFEK0U0hl77utveMDl+GA+GHBWVgNTfFKvWlO1HL2Iq8z2SrZ44akYxbe7uSq9N5n1NjMtdF8c1FzfxgJNCdefR8pCe/r/oOXIFAU5XBODEO6eegYhInEgjabqc/ei8Pz5UxhHJGRCHDgMN6Qw+CJkxlPWEwFIvQW2QBNhSBMpnUlOVk9gwpy9uq82K1LqGnQNmLEeAof9jilGql40v2zQaBRPf0bXOAjDyO7x5w2wW0/e025Ch9kKX2SlNGE9FWSd2fFtw1GGkfFzX2IqC4/KiSWfeHpcyHXleVsoujLOwlCVmitlXtEq6JyIGnn70Q3TbuLF6cSL08z1PjCFPbVVZlew1da4n4sVPsWBNQsvWG2G5AxNSeFAGq6hqDW6bMpdhdqEOCSuolKKUGog0jik+rEG/8p39h6v9y9TYh05HPYb8g5d1NE47E1ejn1xY/IijprRfA9hgPF2Q9VbNwCf7K0DXK5mql2TXcTRczN0I58YMaQ92P17JNZ3YVxDr7doDluevNUP4SjrhaFbcLntXP0a7XsHtgDFyBzNxRZ66+atEg9L7/bDdwS7UzA3XEIb2vr5rKxChr0BQgAtE0NiHCajdJYFQiRM9h4JlhZsqqzraotKiUQqu3jYqrmaYka1Lh63GjtujXY9eI+4LcXWw53ewNCzHw/0AzBt3EqANBGGYSPdaPMCcO0pSTavKgyJNJrbm1WstfdYUU2gR1tkm7KUSmiFNN9DSoSrx8TpQMrCEDJaA2UtrNWkorM2Qm0k92BKzVAsho8SOUwDh6s9a83cL0oUYYqJWgXRTC2WUsy5uTBnoKGspRrO0IoVRElAwwSSCVh6ctEzYy5KMOyg2gJvOqavibGbEZ+BtRiTUV6H0RoMSK//tsYJWjv7yvvA+W64UVf9mBBhfAwSUNcLqz5ZmiOfEsJWvtpd795RQ0I479Y+rK0y3im2+vzpblnn7Cu9+NoAI+voKhLsISK0bLLEl+foE3IDA+nue3fbdcvTc4FB9tf7V5lPfv1CGgd3Ac0rybPJPZvufbLS2Fooyx3t+ATRStRsE6s8J4bC/vHnGa6/wHo6cv/sfaAx6Epsypgi47Tn0aN3iFrINI5iLnfNFZUFCZ2Fh5F2gOJhwzCORq2VuLk8vdijrIZdaGdAAtHlm6Tawn8I0RWBzTBCtHqKWgqLx/V+O+1YKDmvLOvit1GQENnt9qRhJLUR0QOsC5TnJmCREi0lJGe03pPriVxmlnVFciE15Xp8Cw2B/XhgGnasy5HT8SmFxikVKoWnpyPHas03hhBd567SquXPo3huvYEyIhjFuQZD5pKABGhtoRZYc2FZrZnm7cGKfpZ7k/QWrK3ZIVSmWBnCa2LsBqqFDaU2TruTY8aePjHDCrCt3lbl45RRD48fFpVEGG4AvHrMKpA617zVSpJhk4pqF7tGR+c7IWMDkQaPf9Vh6i0OFHqeXaKcd5TaEXNc0trowGU2dZOQElGGy9l4lll+ydAFtgq9hm7xoF+eexuNvCwoyn6/Jw2D1bIX46MvpyMigcPNDWkYWOYjrSh1uSO/+CZSV4ZyQqikweSiD48ec/0DP8SLJ+9xd7yHujLoiaTNJKvGHde3b7EfE6dlod3dGec/FxqFmIzBCALB+RQYTrMb9ue0XDMugvjCXcpqHom36ULYUqvB2zX2qF610lBSDAzDSAbmXndwSS5SyCUzr7MBZbsDEiLTbsekirbJKcN36PHOjCuZzl7IGW2ZWmdKW1nziuRKBK7SI8K4Y3d7w3i1Z3n6FL0vLLpyF+8oNJ4vJ+JaGIKJlJZqDDmqkmggldJcIainUAUKRgyLIxCE1qwuIufCvBaGIfFoZ3UF7x8ruRWQkRgmUmpMYzbF2o8YrxyNn9ywY4wcdjt240gQYV1OZxf7AuhqrVFqoReB9Lz0GXoDOANs/W+Du4wSbKL0mnV8NzH0Xx6IU16CY0H1DNb5WYL/0HPHnQILFjNe7rxda17BWHEiHvdfoNHu6neNObtH9rkU89d6lRPaTDjiMlft2KS1W6oPFrEODraSKdosR92KMbR80YsxIjQaM1Ubp+MJ3n+XPL9gNzZqbhxfrCCVaV4Y1pl8uqPcP2EtFc3ZdOVCMw9IPEToXADOAGMrGVETY7DOP8HPb+i5cQi80QSdBixIdALTRZvqDga2kqFVUrKqMPV70h0EY1fuQMSkAjeXQtAIGsSYmPsr+3XaGa9jPUHONBqNBHHHePXYXf5CLndIULSu5PmEivFF9rJDUcYwESVSXU6sVKMIo5C8B+EwBYJWA+fUQOjk07nmRiuenBRvCunS1UtxABOv4gwNCRlp1rzzIYL8cLxiumzg+uqaw9WBaRq5ORy43R9Y8srz+2eowjgeDPAS8eo3c2F6IwHTheu5YwfPWqOWQhBhTMPmHguYnFWSre1Rp6vWWgnjSNeN3+Jnz13ji8sZZ/fhoYFgkymlZGQLpyoarqaUeQZM6D9O44Oiji4vlfO6iWX0z7bVwHvrj43/3aq1OFJBomCKpLYotFI897pl4y1kULWqPtSQPwzAGYadCS55hdzp9D65LOT3n/Li6ZFpgpsbZZ4L3/76iTUL+y+8YDi8YH72Hean36BqpLFDQuBwsOaHM42MCWfIYFTY1uPpxaSqmwRvuZxM0jtYx9pLNLqURvUFYRg6pXZGtDl+YEKTeV2BxjR4ykp7iGSoypgGhsMObabC03wBV4w6TTDt9ii+eQS7H0u9pyxHW1jjSBh2HHZvU8vK3bf/PvN8Rzm9sGsTIAhJIm+Fa0SE0efv8+OR5+vMuirzYqh5nPZWcjBkJBbyYiFXcAmsprrl5IsqUSM98V61cXdSn6fBU62VGAtlbaxzOdd4fMh45Tt7r8zqVNGeSuu79aV7rtsE7juufmDl6hMEPrioPSRyyMO/ubymhwe8+P4jPsjFrv/gXB05pINiL9NoLq/5glLqbrqofMib7XxyeV2cK6TOb/vIi6WnBi/vhudBnbBjxmEFPcVi5Z4yar1BYWfFWZrO2QL4fuqOxEN/6wPXp+fl6MHN7fn57tFc3tuPTSfZMbs35BDCdv4zUHn5N+eCmf6eDRSUcxbl4f3sVOC4gamtVlrfKOL5uEHODUU2rIXLfy/P6ynBy+vT7ZFvIVu/v9Dnbj8GiFzMoY/b1gFR/fg3/E4OEXkXuAfee2Un/Z0Zn+P775rh+/O631zzb2/8Q6r6+Q974ZUaO4CI/JKq/v5XetLf5vh+vGb4/rzuN9f8uzfCJ7/lzXgz3oz/P4w3xv5mvBmfkfFpGPtPfQrn/O2O78drhu/P635zzb9L45XH7G/Gm/FmfDrjjRv/ZrwZn5HxxtjfjDfjMzJembGLyI+JyK+LyN8RkT/7qs77Wx0i8hUR+QUR+Vsi8n+LyJ/2378tIv+LiPyG//vWp32tLw8RiSLyN0Tk5/znHxaRX/R7/t+IyPhJx3iVQ0Qei8jPiMivicivisgf+D65z3/G58aviMh/LSK71/1ewysydrFm3f8J8C8APwL8yyLyI6/i3P8fRgH+bVX9EeBHgX/Tr/XPAj+vqr8X+Hn/+XUbfxr41Yuf/0PgL6jqPwI8Af7Ep3JVHz3+IvA/quo/BvwT2LW/1vdZRH4Q+FPA71fVfxzjz/1xXv97/bBk8nfrC/gDwP908fNPAj/5Ks79O3DtfxX454BfB77ov/si8Ouf9rW9dJ1fxozjDwE/h3Eq3wPShz2DT/sLeAT8Jg4SX/z+db/PPwj8A+BtjG7+c8A//zrf6/71qtz4foP6+Jr/7rUeIvJDwO8DfhH4gqp+01/6FvCFT+myPmr8x8C/w7nR5zvAU1XtQnev2z3/YeBd4D/30OM/E5ErXvP7rKpfB/488FXgm8Az4Jd5ve818Aag+8ghItfAfwf8W6r6/PI1teX7tclZisi/CHxHVX/5076W38JIwD8F/CVV/X1YzcQDl/11u88AjiH8UWyx+hJwBfzYp3pR3+N4Vcb+deArFz9/2X/3Wg4RGTBD/69U9Wf9198WkS/6618EvvNpXd+HjD8I/Esi8veBn8Zc+b8IPBaRXtn4ut3zrwFfU9Vf9J9/BjP+1/k+A/wR4DdV9V1VzcDPYvf/db7XwKsz9v8D+L2OWI4YoPHXXtG5f0tDrC72LwO/qqr/0cVLfw34Cf/+J7BY/rUYqvqTqvplVf0h7N7+b6r6rwK/APwxf9vrds3fAv6BiPyj/qs/DPwtXuP77OOrwI+KyMHnSr/u1/Zeb+MVAhs/Dvxt4O8C/+6nDVZ8zHX+M5jr+DeB/9O/fhyLgX8e+A3gfwXe/rSv9SOu/58Ffs6//4eB/x34O8BfAaZP+/peutZ/Evglv9f/PfDW98N9Bv594NeAXwH+S2B63e+1qr6hy74Zb8ZnZbwB6N6MN+MzMt4Y+5vxZnxGxhtjfzPejM/IeGPsb8ab8RkZb4z9zXgzPiPjjbG/GW/GZ2S8MfY34834jIz/F64Awe4ovydIAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lbB_TvbrDzuR"
      },
      "source": [
        "from tensorflow.keras.models import load_model\n",
        "import tensorflow as tf\n",
        "import keras"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7CDw37dCnTp3",
        "outputId": "024b19a3-5e2a-4023-ed17-750a16cb66c4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model=load_model('model_improved.h5')"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zuBJSf8vna09",
        "outputId": "b084aeb5-6821-4be7-8504-8179d39f63da",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.engine.functional.Functional at 0x7f9bf01f9d68>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q41gcCCmnk_A"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}