Image_denoising_autoencoder_model.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Image denoising autoencoder model.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "authorship_tag": "ABX9TyMMSXYsBCDUNKIj/Jx5R+2+",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/anikch/Image-denoising-using-convolutional-autoencoder/blob/main/Image_denoising_autoencoder_model.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FZOiRmJD2Oi5"
      },
      "source": [
        "Problem Statement:\n",
        "\n",
        "The dataset is similar to MNIST but includes images of specific clothing and accessory. The objective is to add some noise to the images and then  use    an Autoencoder to denoise them. \n",
        "\n",
        "Dataset Description: \n",
        "\n",
        "MNIST Fashion Dataset\n",
        "\n",
        "•Total Images: 70,000\n",
        "\n",
        "•Train Images: 60,000\n",
        "\n",
        "•Test Images: 10,000\n",
        "\n",
        "•Image Size:28 x 28\n",
        "\n",
        "•Classes:‘T-Shirt/top’, ‘Trouser’, ‘Pullover’, ‘Dress’, ‘Coat’, ‘Sandal’, ‘Shirt’, ‘Sneaker’, ‘Bag’, ‘Ankle Boot’\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V8zm7okf1qft"
      },
      "source": [
        "# Importing required libraries\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.datasets import fashion_mnist\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UGI3lN712ruq"
      },
      "source": [
        "# Loading and checking data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8yTz6lSF17kj",
        "outputId": "4d59c267-a83b-41cb-d354-36cd7d8297af"
      },
      "source": [
        "# Loading fashion_mnist dataset\n",
        "(X_train, y_train), (X_test, y_test)= fashion_mnist.load_data()\n",
        "\n",
        "# Checking shape of train test dataset\n",
        "X_train.shape, X_test.shape, y_test.shape, y_test.shape"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((60000, 28, 28), (10000, 28, 28), (10000,), (10000,))"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "KdzvYsae1-on",
        "outputId": "69101105-e9d8-4e6b-a8d0-3777aa907b8e"
      },
      "source": [
        "# Checking arbitary image\n",
        "plt.imshow(X_train[17,:,:])"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f6dfcb65110>"
            ]
          },
          "metadata": {},
          "execution_count": 17
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "XTsD0Rkj2Bio",
        "outputId": "94a0a1f8-b479-44fa-a3e1-69b69d35f102"
      },
      "source": [
        "# Checking unique count of data in training dataset\n",
        "sns.countplot(y_train)\n",
        "plt.title('Count of different clothings')\n",
        "plt.show()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZY3TsBpM3TeD"
      },
      "source": [
        "# Data Pre-processing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "SBOqOC_L3AMe",
        "outputId": "90e58d48-8287-47e1-8766-01106ef894d8"
      },
      "source": [
        "# Normalizing data\n",
        "X_train= X_train.astype('float32') / 255.\n",
        "X_test= X_test.astype('float32') / 255.\n",
        "# Checking arbitary image\n",
        "plt.imshow(X_train[17,:,:])"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f6dfca65810>"
            ]
          },
          "metadata": {},
          "execution_count": 19
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "RMOtfg5tA0Zd",
        "outputId": "aaf44f44-f034-46f9-b97f-ab998ed9bf00"
      },
      "source": [
        "# Adding salt and pepper noise\n",
        "noise_factor= .2\n",
        "X_train_noisy= X_train + noise_factor * np.random.normal(loc= 0.0, scale= 1.0, size= X_train.shape)  \n",
        "X_test_noisy= X_test + noise_factor * np.random.normal(loc= 0.0, scale= 1.0, size= X_test.shape)\n",
        "\n",
        "# Checking arbitary image\n",
        "plt.imshow(X_train_noisy[17,:,:])"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f6dfc9ce790>"
            ]
          },
          "metadata": {},
          "execution_count": 20
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hPMMuLH4A1Fr",
        "outputId": "b5b3e25a-9cbb-4029-c5ab-7e5cd584a4c4"
      },
      "source": [
        "# Changing shape\n",
        "X_train_noisy= X_train_noisy.reshape(X_train.shape[0], 28, 28, 1)\n",
        "X_test_noisy= X_test_noisy.reshape(X_test.shape[0], 28, 28, 1)\n",
        "\n",
        "print(X_train_noisy.shape)\n",
        "print(X_test_noisy.shape)"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 28, 28, 1)\n",
            "(10000, 28, 28, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3yp4FAng4Mmk",
        "outputId": "4acd286a-5ed1-4b25-e76e-14039fe9b0c0"
      },
      "source": [
        "# Input layer\n",
        "input_encoder= tf.keras.layers.Input(shape= (28,28, 1))\n",
        "\n",
        "# Building Encoder\n",
        "encoded= tf.keras.layers.Conv2D(64, (3, 3), activation= 'relu', padding= 'same')(input_encoder)\n",
        "encoded= tf.keras.layers.MaxPool2D((2, 2), padding= 'same')(encoded)\n",
        "encoded= tf.keras.layers.Conv2D(32, (3, 3), activation= 'relu', padding= 'same')(encoded)\n",
        "encoded= tf.keras.layers.MaxPool2D((2, 2))(encoded)\n",
        "\n",
        "# Building Decoder\n",
        "decoded= tf.keras.layers.Conv2DTranspose(64, (3, 3), strides= 2, activation= 'relu', padding= 'same')(encoded)\n",
        "decoded= tf.keras.layers.Conv2DTranspose(32, (3, 3), strides= 2, activation= 'relu', padding= 'same')(decoded)\n",
        "decoded= tf.keras.layers.Conv2D(1, (3, 3), activation= 'sigmoid', padding= 'same')(decoded)\n",
        "\n",
        "# Autoencoder\n",
        "autoencoder= tf.keras.models.Model(input_encoder, decoded)\n",
        "\n",
        "# summary\n",
        "autoencoder.summary()"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 28, 28, 1)]       0         \n",
            "                                                                 \n",
            " conv2d_3 (Conv2D)           (None, 28, 28, 64)        640       \n",
            "                                                                 \n",
            " max_pooling2d_2 (MaxPooling  (None, 14, 14, 64)       0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " conv2d_4 (Conv2D)           (None, 14, 14, 32)        18464     \n",
            "                                                                 \n",
            " max_pooling2d_3 (MaxPooling  (None, 7, 7, 32)         0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " conv2d_transpose_2 (Conv2DT  (None, 14, 14, 64)       18496     \n",
            " ranspose)                                                       \n",
            "                                                                 \n",
            " conv2d_transpose_3 (Conv2DT  (None, 28, 28, 32)       18464     \n",
            " ranspose)                                                       \n",
            "                                                                 \n",
            " conv2d_5 (Conv2D)           (None, 28, 28, 1)         289       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 56,353\n",
            "Trainable params: 56,353\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zaSuumh42S1t"
      },
      "source": [
        "# Compiling model\n",
        "autoencoder.compile(optimizer = tf.keras.optimizers.Adam(lr = 0.001), loss= 'binary_crossentropy')"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nz0I4E9_378s"
      },
      "source": [
        "# creating callbacks\n",
        "earlystopping= tf.keras.callbacks.EarlyStopping(monitor= 'val_loss', min_delta= .001, patience= 10, verbose= 1, mode= 'auto')\n",
        "callbacks= [earlystopping]"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f8xh7I2P0ju_",
        "outputId": "ca0c66f2-08d9-42bf-ab78-0675246ec551"
      },
      "source": [
        "# Training autoencoder using X_train\n",
        "history= autoencoder.fit(X_train_noisy, X_train, epochs= 100, batch_size= 128, shuffle= True, validation_data= (X_test_noisy, X_test), callbacks= callbacks)"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.3082 - val_loss: 0.2796\n",
            "Epoch 2/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2735 - val_loss: 0.2731\n",
            "Epoch 3/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2689 - val_loss: 0.2696\n",
            "Epoch 4/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2668 - val_loss: 0.2684\n",
            "Epoch 5/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2657 - val_loss: 0.2673\n",
            "Epoch 6/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2649 - val_loss: 0.2666\n",
            "Epoch 7/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2643 - val_loss: 0.2663\n",
            "Epoch 8/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2638 - val_loss: 0.2656\n",
            "Epoch 9/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2634 - val_loss: 0.2654\n",
            "Epoch 10/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2631 - val_loss: 0.2650\n",
            "Epoch 11/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2628 - val_loss: 0.2648\n",
            "Epoch 12/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2626 - val_loss: 0.2646\n",
            "Epoch 13/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2624 - val_loss: 0.2644\n",
            "Epoch 14/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2622 - val_loss: 0.2643\n",
            "Epoch 15/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2621 - val_loss: 0.2641\n",
            "Epoch 16/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2619 - val_loss: 0.2640\n",
            "Epoch 17/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2618 - val_loss: 0.2640\n",
            "Epoch 18/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2617 - val_loss: 0.2638\n",
            "Epoch 19/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2616 - val_loss: 0.2637\n",
            "Epoch 20/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2616 - val_loss: 0.2636\n",
            "Epoch 21/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2614 - val_loss: 0.2637\n",
            "Epoch 22/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2614 - val_loss: 0.2635\n",
            "Epoch 23/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2613 - val_loss: 0.2635\n",
            "Epoch 24/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2613 - val_loss: 0.2633\n",
            "Epoch 25/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2612 - val_loss: 0.2633\n",
            "Epoch 26/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2611 - val_loss: 0.2633\n",
            "Epoch 27/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2610 - val_loss: 0.2633\n",
            "Epoch 28/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2610 - val_loss: 0.2632\n",
            "Epoch 29/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2610 - val_loss: 0.2631\n",
            "Epoch 30/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2609 - val_loss: 0.2631\n",
            "Epoch 31/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2609 - val_loss: 0.2630\n",
            "Epoch 32/100\n",
            "469/469 [==============================] - 4s 8ms/step - loss: 0.2608 - val_loss: 0.2630\n",
            "Epoch 00032: early stopping\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 350
        },
        "id": "1L-V8gSfL0tI",
        "outputId": "b9456bdd-1a5c-47a9-a6e7-7655c44d707b"
      },
      "source": [
        "# Plotting loss for different epochs\n",
        "plt.figure(figsize= [8,5])\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('Plot for model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Validation'], loc='upper right')\n",
        "plt.show()"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H_7pF7Tf6lCe"
      },
      "source": [
        "# Displays ten random images from each one of the supplied arrays.\n",
        "def display(array1, array2):\n",
        "    n = 10\n",
        "    indices = np.random.randint(len(array1), size=n)\n",
        "    images1 = array1[indices, :]\n",
        "    images2 = array2[indices, :]\n",
        "\n",
        "    plt.figure(figsize=(20, 4))\n",
        "    for i, (image1, image2) in enumerate(zip(images1, images2)):\n",
        "        ax = plt.subplot(2, n, i + 1)\n",
        "        plt.imshow(image1.reshape(28, 28))\n",
        "        plt.gray()\n",
        "        ax.get_xaxis().set_visible(False)\n",
        "        ax.get_yaxis().set_visible(False)\n",
        "\n",
        "        ax = plt.subplot(2, n, i + 1 + n)\n",
        "        plt.imshow(image2.reshape(28, 28))\n",
        "        plt.gray()\n",
        "        ax.get_xaxis().set_visible(False)\n",
        "        ax.get_yaxis().set_visible(False)\n",
        "    plt.show()"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "id": "afkXGgnmH9sG",
        "outputId": "cd1adb08-6780-4a92-dca0-162e7405394b"
      },
      "source": [
        "# Checking on training data\n",
        "predictions= autoencoder.predict(X_train_noisy)\n",
        "display(X_train_noisy, predictions)"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABG0AAADnCAYAAACkCqtqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydeZxdVZW2V1VqSIUEEAQEQVDsxlkwiAgSApGAYQgBEghhEGRSfzi1inaD2rYtSuOIojKDYIAwBAhjGJRBgdYWVBrUgA0OICBjhkpVUvX9852dZ71V90pxb+Amvs8/7HBO3XvOHtZe59z1rtU2ODgYxhhjjDHGGGOMMaa1aH+5L8AYY4wxxhhjjDHGDMUvbYwxxhhjjDHGGGNaEL+0McYYY4wxxhhjjGlB/NLGGGOMMcYYY4wxpgXxSxtjjDHGGGOMMcaYFsQvbYwxxhhjjDHGGGNakI6RnNzV1TXY09MTERHVfytYOnzhwoXp2DrrrFPafX19pb1kyZJ03sDAAL+r5rE111xz2O+NiHj66adLm9c4evTodN6TTz457HkREcuXLy/tjo6OYdt6TfxevUb2B/tCv2vp0qXDHlu6dGksW7asLZpAR0fHYNW3ej+jRo0q7e7u7nSMY8Xz2Af6787OznSMY/C3v/2t5nfx89kn7e35HeOyZcuiFhzTRYsWlfbYsWPTeZyPOkeee+65Ya+J7YiItrYVQ6P9wWMLFy58cnBwcL2aFz0Cenp6Bqs5pmugv7+/tDm/9PrYn729vem8tddeu7R1nfL+eX/at5xf/C6d57z+cePGpWPPP/98afO+9DN4HdofvOe11lqrtB9//PF0HuerzrVqji5cuDB6e3ubsha7u7sH11hjjSHfHZHvW+3T4sWLS7v6+4jcPxF5fag95ZrjvFfbzX6oN4Zjxowpba63iDxfeC/6GRwnXWOcg7x2jntEvhf2oX73Y4891rS1OHr06DKOOm+4Bp566ql0jHsE76+eXdZx5Fzg2tHPYF/TJqi94t/pnKSN4NzSe+Zn6vjQ3nLO1LPtOp+qv1u4cGEsXbq0KWuxq6trsJo7ei28Zp2XpJZdjMh9rvaU87neOuJnch7oOqKN47Xruexj3SeI7ouENofXrp+vezzn7aJFi5q2Fumj6hrgNTz77LPpGMeulq8TkfdF7Vt+H/08+poReQy4VnQMuJ71WC2/RecW54l+Bo/x2nXesd90bVQ2obe3N/r6+pq2Fqsx1LnHvtP5VsvGqR0jumfy+/h56tuwn+lv6JzYcMMNS1v7lefWW3+8RvVtuIfU+mz9u3rPVo8//njT1mJHR8dg1fe6FjkmugfxGO2EfgbtodpAzlPu+9rPtcZO+499prad1885qedxHHUMeL08j/cfkeehfn51bjOfF3t6egYrO6TriNdcr1/r+Rv8DLXJtH/8O12zvC72ST0fSO06fRuOoV4vP0P7n3sDz9NnK469fj55+umnh12LI3pp09PTE9tvv31ERLzpTW9Kx9iRP/3pT9Ox/fffv7T//Oc/l/a9996bzuONbrTRRjWPTZ48ubR1cc2ZM6e03/rWt5b2G9/4xnTeaaedVtpbbrllOvbMM8+U9ite8YrSftWrXpXO40MBvzci4r3vfW9p33nnnaU9c+bMmt/1hz/8IR2rXmw88MAD0Sy6urpiiy22iIiIddddNx3jpHvta1+bjt13332lTcOhTgLH6ZWvfGU69pa3vKW0zz777NL+p3/6p3QeN6Lf//73pa2bNJ0hdSY4P//7v/+7tLfbbrt03h//+MfSrvqlYv78+aVNJ4z9FPHCN4/bb7/94WgSa665ZsyaNSsihhqxP/3pT6WtD+EcLxqP3/72t+m8qVOnlvavf/3rdIwPBXRytt1223Qex58G+MEHH0zncQ1PnDgxHbvllltK+69//WtpL1iwIJ1Hw60bCNfpHnvsUdrf+c530nkbbLDBsNcbEbHppptGRMS8efOiWayxxhoxadKkiBhq72699dbSph2LiPif//mf0h4/fnxpP/bYY+k8viSorr9is802K+03vOENpX377ben87jWuf64LvU6uN4i8oult7/97aX90EMPpfM4TpxjERH333//sNeujsSb3/zm0r7xxhvTsbe97W2lfeKJJzZtLa6xxhrxvve9LyKGOve0G7Nnz07HuEf85je/Ke31118/nUd7qOP4i1/8orR33nnnYb83Iu8t3HPUPvC7+cARkW0E16KuFX6mOk6PPPJIab/jHe+o+RlPPPFEad91113pWDWHrr/++mgWo0ePjm222SYihr4kpT2t92Kax/QlBX8AuOeee9Kxf/7nfy5tzlFdR3Q0OQ90HdHG6bH/+7//K22+tNYXnNzTdH/mddDmqJ3iPOOajci+309/+tOmrcWenp6yx6+3XvZ3N99889JWO849nfudzst99tmntNmXEXnNVXtzRMRZZ52VzuOY8EFef/jbaaedSlsfaOhH02+hnxaR9xVd65tssklp0xekvxoR8frXv7609SVKtQ/cfffd0Sw4hjr32He0HxF5PDj31G+nT6Z7Jr+PfsqECRPSeZy/9DfoT0ZEHHfccaWt6/nhh1dMe9oHfRjkNapvQ5vPv9Pr4IPj6173unSM8+Jb3/pW09ZiZ2dnWff6LLDxxhuXtr7M4h70y1/+srTpn0Vk/1V9Stpf+gT6XewXrild27S3+sBPn5/Pkrr3cRy59iKynXn00UdLm/cfkX1s3WMqv/B///d/o1mMGzcuZsyYERFD1xH3SbVPfGHPua39z8/Q/XyHHXYobb68ou8Rkf0U+oP1fCB9hqN/yTHU6+VnaP/vueeepc35qM8qtMn6QocvcebMmTPsWrQ8yhhjjDHGGGOMMaYFGVGkTcSKN7389Ski/+LHX3r4NxH5zaFGR+y4446lzbdzERF33HFHaX/ta18rbY2S4S9G/HwNQ6p+GY0Y+usE34LyLR7f3kfkt24HHnhgOsZfFPmLqv7KxreB2h/VLy18A9wo7e3t5dcS7RNGrmj4Hu/1Na95TWnrLy+8Vo3CYf/xF3f9Ln4G546GGfOXPf3l/8c//nFpMwpAfznj2299M86/4y8XfBMekaN6dN7WC3tthKVLl5ZfnfirhX5nvXBdjqmG2V599dWlrSG5/KXm0EMPLW39VYC/3jEaiG/QI+pLSDhPGHWjEU28T11HvA6udf31lxEh+hnVr1HNHs/q3ikXjMiRgRolSBiRw193I3JfajQWv++yyy4rbf0VjvPid7/7XWnrry78dYfnReSovUsuuaS09ZeQetIT/pJDm6P2lL9+aL/pL8jNYvHixeWXV/0Fh78Oap/RjrCPNOKSa4C/ekdEbL311qXNX3fe9a53pfOmTZtW2vxVXG0H15X+ksR1RDvCaNSIHOGkkVD8BY6/qOov3rSj7373u9Oxau7WCy0eKdwXtU94LRoRx32HbV2L9aJa+Esc15vui9w/GX2rEY78PP2Vj33O9aa/hNO/0wgQRgNxDqvclMfUnuq/m0VHR0eJsNFIYt6TzimOHdew/iJOe3vIIYekY/xM3p9GCL/nPe8p7XoRIfRV9Jf/d77znaXNtajRRYwc4d9EZJ/m/PPPL221MZyHjICLWGHr9RfjRujo6Cg25bbbbkvHGFX0/e9/Px2rolYj8i/4KqPnfOYzR0T2YbjXMxo1IkeNM4KmitYb7rs0auimm24qbY69Rj3TT9Hnrmuvvba0OW/1uzi/9TlmypQpsTLo7u4u0W18ZojIffbqV786HWM0FSNu1Q6xz9SnpM9KG6hRTIyO5DpS34E+h0b1cH1wr1D7w+/SqH5GOHP96brnc7bO3Wr+6xxphN7e3rLXqM9Nv/3mm29Ox/h8zetRP5fPcLvssks6xr7kd2lEJ/1NRqGrjeeYqnqFz6qcI/p8y/cN6q/SH+Hc1HuuF135Qp4vHGljjDHGGGOMMcYY04L4pY0xxhhjjDHGGGNMC+KXNsYYY4wxxhhjjDEtyIhz2lRaXa38RJ2u6vqpl6V+UnODUMuvWmxq8i+//PLS1qzV1GxTY62aW+rstEoG9W7UqanWndp01R6yUgH1m/WqYmkOgUoLV6+09Uhpb28v2krV5HE8VMvHe2f2burbI3KfaJ4UHqNOV7PqMw8E54Rm8L/yyitLW6uZ7bbbbsNer+ZMoT5SK4PwM6mZpRY5IvdbvTLDzaStra3oPHV+UHeu1UOY/4D6d9VW1iuNTS3tr371q9LWyly8d+Yj0rxC1HxqiUPqVWk7NM8O17Nml2d/UJuuum+uP82XUeW70TwTjTBq1KjSF1oxiH2i1XNor7j+NJcMbZBWAKPemvm3NKcN7TDtRVVFsOK6664rbVbDicj5dJg7QzXfHDddi8wVwBwR3Bci8n6g85Z9xfxIjdLV1VV0+VptgOOoObO4PlgFpF61E80pwb2WemvN8cX9iZ+nOYFos1VfTY05z9O8NbQdmiuFc2Hu3LnD3kdEzlHFXFMRK/KMNHMt9vf3l75V34brqF7eJd63rjeuYc03VKtMva4j7l1/+ctfSlsrxfA8zSVB+8p5pvm9rrjiitJWv4SVzliZ5ec//3k6jzkE1P/SHEbNoq+vr/gM+h3MI6Hjw/sgrAgSkfNNcAwi8p7BtvYt90LOrRtuuCGdxxw06ufSDvC7dN0z94ruD1y3vK8f/vCH6TzOyb333jsd0324GXR1dZV5q33HCjOnnnpqOnbuueeWNue2VtzjuLGCU0SuIsNnmlr+QETOK6Z2jOOkuY1oCzm+ek3Md6YVwGgj2FeaW5H7uuYm0XwkzaK7u7t8r94TfUX6kBF576f/oX3LKsWaH4/jT/9Y86/Rd6e9ok8Uke2D5lThHkq7otWumG9F7Q99E+4VWsWXuQMvuOCCdKzybfUZpBE6OzuLb6o5StmvulfR76Gd2XXXXdN5tEnq83Ff4/6ha5F2nZWldb4wP5Y+83MPZh+rP8wcZ/osz32Xc0J9Za5hrWKo83M4HGljjDHGGGOMMcYY04L4pY0xxhhjjDHGGGNMCzIieRRDpbS0HUMOtYQ2w55YMlFDvfmZKkWihIYhpSqFoOSDIYxaNpvhaBqeRwkFSwBqKChDmTSEm+VlGQr4k5/8JJ3HcuAqu6nCJ7VcbyO0t7eXcGeVZDBsUstrM0yMx7T/GZqnoV4MWaRsR+U9tcpHqgSNJd30uxgazdA3DR+m5EZLzjGcjnNCS7gxbFDXhUpbmsXy5ctLSK1KMtifU6dOTccYcsjzNPyaa0Jlbgwp5XrTMq4MI+Xna/gm/04lMwytZOgyS7pHZCmHyksY3k/ZhY4j5Su6Nir7pqHFjbBs2bISUqkyBqJ2kv3FkEwNEWZ5yPnz56djDNmkDJChoRE5tJx2SCUkDH/W9cwwVUpDtDQl55LaFUrZGB6rpRspKdAS9Fq2tVmMGjWq2B+G1Efk/lS7xrBelvWmBCAiSxpVxsgxYb9QHhqR9x2uRS2LyXHkWtHr51pReQ73bi1fzvXNcH4N+2dYuMpuqutQCUAjdHR0lGvTdcQSuSobpj3geOp6ps1UORklLCz3q+HitAMaNk9oC7VfOV9oh2krIiK22mqr0lZJMeWt/Ayd3xx7+mwRQ/eUZtHT01P68+KLL675nTo+7LPdd9+9tFXiQIkLQ/Ejcj/R3up3cYwp4aS8MSL7lCrTqlXWV2VetMWU70Xke6E8U30Bymd0z6y+r5lrccmSJeUZQn2FnXbaqbRPPPHEdGz8+PGlzbmnvj9LTKv8at999y1tSiHVL+EzzowZM0pb9yP6ynwmiMhSC+6ROk70ZdVecx5wPPWeKYtUG6blp5vFM888U1IYqEyWe4vaMu5xHEddi+x3SjYj8njRbtKHjMjPdLSNaqPnzJlT2vQxIrLUhp+n8t099tijtO++++6oBf1VTR3BZ0mVrVZ+r87BRujr6ytzif5ZRL4/3asoa6PdpaQtIve5SpH22muv0ube8rOf/Sydx2O8JvU9OM/1WZ7+MGWq6s9xXWk/c/3RDqvvzWdE7rPDXddwONLGGGOMMcYYY4wxpgXxSxtjjDHGGGOMMcaYFsQvbYwxxhhjjDHGGGNakBHltOnr6yu5HTTPAnXsqj2kRpx6fZYOjsh6NC3T9+CDD5Y2NfRacpqaW56nGmrq4FT3TV0/9Ytajpj6NupTI7Iumvk33ve+96XzqHmltj1ihQ6+mWUV29raSj+rrpJ5YbRfOd48piWlmbNCtdHUjWvOHMLcDJwHWmKNeRre8Y53pGO1dMAnnXRSOo/6cp3Ts2fPLm1qZjVvDTWtmrOoXr6QRliyZEnJK6Kl0DnXVUvMfmeODy2BzP7UMre8f7bPOuusdN4555xT2sccc0xpa14hrm1dR1xz1DDrWFFjrrk4mFOAuXpUz845o/OzWuv62Y3Q0dFRxocldiNyDgNdpyzPylKLWlKapVsnTpyYjnHcWDpS5xLHg2Oh5WOZE0HzDXF9MO+DluSmnVOtL20C8zQw30hExDXXXFPamkuCJTebTXVNmh+tra2ttLVUJddVrTxRETmvA/eViIg77rijtJnr6Oqrr07nMS8Oj6m9ok3Q/Y77OEuWslRqRN5btbw15wLv+Uc/+lE6j33AfEYRK+xbM8tGszyt5ufgHsT9IiKvD85nLY/KvVXzYl144YWlzTwm2v/8LvYd51hEzo+heRW4XzNvg34Xy9Frfgfmv+G80rwrtMma849j2sx12d/fX+bFsccem44xFwJtaETuT+Y+0LxYXGOag5D5Jth/kyZNSuexz5iv6aqrrkrnca1oDrFDDjmktGmLdQ+gD6a5mviZHDvNBUEfRn3RyoZpPrtGGBgYKGtE1wpLMU+fPj0dYz/Qp9f8XlzD3/zmN9Mx5rZgnjG9Dtph5lrU/uc8nzt3bjrGvZA+kPpb9F81Vw1tL/OFaH4kzm/NOXf77bfHyqK6F80JxPFh7siI7GMxL6LuA8wRpHsm8+TQbmpeTH4X7Rx9zYj83Ko5JGvNO50LHAP1Bbhf0KfT3K7c89RXrvLfqC1vhHHjxpXndJ0nHA/1i9n/9F80ryD3Ws1Fyf2DOXM0bxfzOtFmapl55nJS/54519hmbh69DuZ2i8jrlu8h9H0I9wb1NWrlcyWOtDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYFGZE8anBwsITvaCksLa9IGB7Ekn0MsY7IoXAaBkgpA8MbNSyOJcQYrqRl2ihV0dJplFV95CMfKW2Gokfk0HyGJEfkkC1KVFSKxWNaOrUK9dLw5Ebo7+8vYdwaZkZYviwih5OxbJ5KunitGppdS67G8NuI3Ed77713unbCMLZ99tknHePcYhjfpZdems6jNETDhzmmlBHo9XIOqnSq3rpohM7OzjI/VPpXlVmMGFoul7IThqVqaU2G83NdRuQyhAwVVSkEwzy5VjTUmKG8Gj7JcESuP13PBx54YGmrBI5h8VzrWjaV4abaH1VopUoRGmHx4sUl/F7vhzKGgw8+OB1jCCWlSBpaSUmUhmhSxsNx0tKNnL8Mu9XyjCzRrVIIllilhETXCvv81ltvTcdY/pEyS5X/UWqhEohmhvCTzs7OMnc0dJo2RcN1GdrO/U7LnXPsVGLJEH6G22tpU4ZVs61h2rSxHLeIHOLOvtVwe+67KoXk9a+//vqlvcsuu6TzKIPQ/a/qx2ZKFZcvX15slMpD2K8aks75zHmp8gSGequEjudSfqJjzb7keSoR5P6k65ljys9TG0+5pkoVKY/mXFIJJvcelUDQ5jSTUaNGlZLJKuWhT6P9wjVG6a3OMY43fZOIPMb0l04//fR03mc+85nSpqSYe1hE9nPVfrPfaZe5biKy7EJ9NUqBKM3S+6I0RH26ai7UK2E8Ujo7O4tdVz+UkhMtoc1zacfOO++8dN6uu+5a2jpnuSY4R3SN0fekfeJnR+Q5oXJd7pOU32+99dbpPEr0NG0Ex4OpFyhhjMj7kNp1+g3NlEr19PSUZwz1Q3fYYYfS1pLjvAZKUFgKXFGpCmXS3E/1eZH7H/cjSvv171SyxDnJ71J/jM8h+hxI6Q6fXbQcOs9TX7TyybhfNcqiRYvK2lb/hfuA7ov03zj3VF7MMVX5G9cE3xvo+qC0kOW6dU7QN6OMKiJLs+pJCSltUt+T48Z9Qv1Q7oUqj6LEthaOtDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYFGZE8atSoUSVEmhKJiFwJSrNAM9ztoosuKu3Jkyen8xgOrGFh/D6G8p522mnpvFmzZpU2ZREaasxQUUpBIiL22GOP0r7kkktKm6FpETlDdL0qPURlF6zWonKxKoRYs5A3ShXGrRUuGMKo4WMMCWTYn4aPMYxU5QgMGWNoncq0GLJ44403lraGA1Zh0BERn/3sZ9MxjukXv/jF0ta5STT0tJaEREMb64UjMgz0gQceqHneSBkzZkyMHz8+IoZmi2f4IOdoRO5brlkN/2Wooo4jK0sxdJdVMSLy/OJnqNyBcop77rknHeO6ZwimyjoYCqnziXIH9g3DLyOyvEHDOFWC1iyq79Q+YZi7rjGON9eR3g/7VedIrVBdreTCdc8+UDkoQ2C1ahwrk1ESpaHYlGFo2DpljVz3Gl5arYmIof3BKhxa4aoR+vr6SlUKlQ8w3F4lQAyPZ1slDjymsmFKOeqF5LKSGG2ghuNyD1LZMCV7/HwNA+d60yqEtDMMOdf5yTmvcqJq/9eQ+0ZYsmRJkTJomDbXgO6ZlDywH1TixnHSvZX7O30M3atoQ7m/qayXMhGtssfQb8pZ1d5xXuicozSEIefqY/E8lYa8kCoZL4bly5cX+8NQeUWrXvLaKTPRfYbSXrXLHFdW8mGlp4iIM888s7Q5xuozct2rPIryD+67lHhEZCmMVlujhIhjpdfBcP4pU6akY1Xls2ZKMkaPHl2ko+q/UG6k65Q+Pe9bpXl8tlCJNNM+UH6lVfYoZWP1IPVfeP0q16Nve8ABB5S2+vu0P+qL8d9c22qnOL91Hqg9ahadnZ1lPnJ+ReR9QG0U9x3OUb13+nxVVeMKPo9RLqr+DecQx0qf9Wi/rr322nSMckL6MJrugv6IjjElYrTFOlaUE+lnVH6c2qVGGBwcLBJRlf7R3/j+97+fjtHXYr/qPOA61bVIeSKlnFpplXsQ/VIdQ84r7TvK/VlhTCV5559/fmlr2gLaaI6nyq05l3Tezp8/P/4ejrQxxhhjjDHGGGOMaUH80sYYY4wxxhhjjDGmBfFLG2OMMcYYY4wxxpgWZETJUsaMGVNKjGr5Zeo/VTdIjfWMGTNKW8sEUo9LnWhE1sIxL4yWi6NWjeW6L7/88nQe9W1agph5CKi9VB22ln4jG220UWkvWLCgtLXU5Pe+973SZg6KiBWaSO3rRhgzZkzJ66FadZZb1pK71B4yt4Vq95nvRHNbsHwmP1/LB8+dO7e0mb+Cmt2IrE1W/SLnHOeV5ibh56t+nXOVpcE1D8Ts2bNLe+bMmenYyiozHLHiejXHDjWeLNMbkTX0V199dWkff/zx6TyW0dM1xr7eeeedS1vzBHA+U7up85laWepkI3KpP+bA0Hwb9957b2mr/p+fT/20asx5XdTkRqzoYy1h3Ahjx44tGnrVITNnCO1HRF63LO2s48T+15w5zBVEO6xluJmHpV4eGM45zSlErTKvXceQOUpUS8zrYt6Vww47LJ3HvqI2OWKodr5ZDAwMlJwsOn+Zh0n13NzHNKcU4f6k+Us4/szHoCW/mVuGfTtnzpx0Hu1XvWui3dS9mnmFND8P+4A2RvNoUNevucYqW9LsfbHKR6BrnPOIe2REvm7mM9C5R9ul+3ytUra6dzB/E0uPqi6ec1BznGy22WalzVLOU6dOTefRNqrt4D7J69D8I1zf9cqoN5P29vZyTVqGlnnpdK+nbeB807xY7E/1W7761a+WNvMdsKx3RM6BUS//BHP4aU6Vd77znaXN+aNzgT7qQQcdlI4xtw5ze2jfHH744aWtZbYre3HKKafUuIuR8/zzzxf/kLZE/61lrZnHhTZT/XTmOtI1xnnAPFvs74jsvzK3io4nx4Y58yJy3hXaO/VzmfvzO9/5TjrG9Uwbozk2uRY0T8rKyte3dOnS4hOqj8rnI81NxDylPE+f05gbRHPJcb3QB9bx4Vzn3qTloplrROcC92TmB9Lr5X566KGHpmO0Ebwv3Rvq5X2t+ljXbyO0tbWVvq2XO415wCKy7Wf+NfUpaGvVV+D3MUeYPofTTr73ve8tbc13SNuofu43vvGN0ma/6l49bdq00tZn+Ztuuqm0uTfo3Od1MK+oovmxKhxpY4wxxhhjjDHGGNOC+KWNMcYYY4wxxhhjTAsyInlUV1dXCTX8z//8z3SMYepaDpzhiXfddVdpa/lrlm1mafCIHIbLcD7KcSJyWBxLJmoIIEPXNByaYZEMk6NUIyJLZhg2FZHDgSmz0ftiH2iJz0qGoyGpjbB48eIiQ9D7YWlVlfmwHxiWXpVmrKBMhXKoiFyqnSFt/P8RWR7FUE4Nz6OEhN8bkSUZDBWtV1pWww0ZpsowNi3NzhBV/QyWO24mzz77bCk9qN/J/tRrZTg//077luOjIeIMO2SZTQ1bZLgpw0Z1PrP0s5YVZElxyu0YmhmR54JeL8P7GUKttoPSHQ3jrNamlgpsBJZTVChvUakWQ8QZzq1yR8rYNCyYZTBZMlbnC8NSOc81dJMhwyoh4dxiaLqWpeZ6U0kYw5/33HPP0lbpASVClGxFRNx8882xMmhraytzTEtcc02obI/7JGVoui8ynFYlpwwfp/3W8HiGKFM+o/IAjoGGKzPkl+H8+l3cTylJicjrivNCZTxc31pevvq3hsQ3Qnt7e1kvug/TTqi0gOPLUHa9Zt6fhrxznXKcdE+jD7TPPvuUtpY+59rWuUQpJEPTWRI+Iq9h3eMpbaIkSq+Dc5ry6ogsr1NJTyN0dnaWNaElyCnH0+uhP8i2htHzM7VvuZ8yRYD6npQT1vNRuQa4H0fkMaBPo/ZnwoQJpT1v3rx0jDJW+nS6B3A9b7nllulYJQOstY+9GMaOHVv2PLVB/B6WzI6I+OY3v1nalAybV20AACAASURBVPKqbaGMSCW/lNJwD2KJ74hsE7gH6z5LG6fHeC8sD61SE+4TXNsREXvssUdpX3XVVaWt8h7KHXUML7nkklgZ9PT0lHQaum/xnlTGyDHgHNXnNK4BlZTRLv3sZz8rbe1bylg571VixGtS2851T1umdvO4444rbX3m4VzgnNT5z/1U99bKV1PZfyP09fXFww8/HBFDZWyUCE6aNCkd475I/0DXAJ8F6F9G5DVBm6my1L322qu0KflVm0nJVnVPFZQHc+7oOqLfq3aFc5XH+M4jIvvbtEURQ/es4XCkjTHGGGOMMcYYY0wL4pc2xhhjjDHGGGOMMS2IX9oYY4wxxhhjjDHGtCAjStDw+OOPx6mnnhoRQ3Vl1H9qiSvq8qpy0xFDS/FRB6w6R+r0mC+GGvCIrBejxl81qdTw6vUS6vFUX8jSz5rjgfpLlnfbf//903nMu3DFFVekY1XeAM0V0gj9/f0lp8Ftt9025FiF5gyhDp/5YrQkGvOdMN+C/h1zxKg2kDpv9itLZUZknSxzq0QMLUVeoX1JneYDDzxQ83qZn0XnC/McaNllomXnG6G7u7usQc3twHwBmm+Cen1qPlVzTN2u9hn13PwMvY5auQq09DHzalx55ZXp2Nlnn13au+22W2lPnDgxncf1pjpR2g7q+jUnAUv4aU6Vyh5pTqRG6OrqKuUddW5wDDVXBPXVzPlDfXtEvgfmw4jI5VK333770tZ1T50ux/OII45I59UqmRsRcd999w17nuqbOU76GcwzwjnMvBkRWfvMvSYia9a1PH0jLFu2rNyX6r6Zy0lhfiSuMWrF9Txdz+wz6vM1RxXHm3+jum/ucar7Zsla7v9aApXa8auvvjodY84hrj/Nb0Q7oOWOqz5gfqpGWbhwYdxxxx0RkddXRL5v5p+JyP3KXEua44NrUfOc0T/g56ud5DxgXop3vetd6bxrrrmmtDU/D+c955z6UdyTtYw9bQmvUfOAcZ/UPFfMM9FM6N/Qj4jI1622n33LnGh6T8yJoSV9d95559L+8pe/XNpa7pxrk3lsNEcFczqor8zcNfQ/WHY+Iudz0/xVzJ9A30ptKn0DtT/V2tCy8I2wZMmSYqt1XnJ96NxmThfmIlI/l/ZPcxbRb2T5YB1r5h3hGGrZe46h2g6uCfrRmpOPf6e2lmNIf1tz79FPUFZWye9FixaVda79R1+rXp5J+uA63szPpTaQ/c51xT0yIu9x9Dm0nDN9Cc1rVuWWjMh7hz7PfeADHyht5oSLyHvHfvvtV9qa44RjNWfOnHSseh5lefJG6erqKv2in8tnLrXnzPNDf1zzUHGear4m2muex/wz+t38XrVVnD9qr/g8+rWvfa20mVsyImL33Xcvbc0FxdyTzIWjNpn3pb7NC8mZ6UgbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWpARyaO6u7tLOKyGtp9//vmlrWH6lMww3FHDNRmuq6GElO4whIjh8BE5fJDltXl9+hl6HVWZuogcJqyhdQxj1zBLhvwxxFbDnxjCq5KP6vuaWcKtu7u7hJBpWTuG8rM8ZkQO8WIpO5VucF7ovbIf+Hkc24gc+ssyxvp5DFtXKQ5LerK8oIYZM0xR5wHlPgytO+qoo9J5HEMtM6wliZtFT09Pka5wfUXkcGaVr1EyeNJJJ5W2SpsoN1LJB++Rn6+SSYaFM0RZpS8MVddyih/96EeHvSYNW6ScUsMsKXtjaLqON9GQxmpOqmSkEfr7+2tKyFQySCg3oj2lLCUiS1903vNchiCrdIPjy3BYlu6OiJg2bVppa4lDylimTJlS2ir/oD1Vm8c1RjulshrK32gfIrKkRO+zEShz07XINaF7FUu3c7x32GGHdB7vQ2UdHEfaYpW0cM3RNj755JPpPIax63WwZDrHQ9cE7bmOD9cc56SGV/O7VS5Wjb/KCBqluja14VwDui/yOjmnKAOMiDjyyCNLW0OzOVbsB8pBIyIOP/zw0qbdUAkP5Y4afs1Ss5RM6FphGP5NN92UjtE3416tpYQpS6CNici2RKXljTA4OFj8yHo+ZLVeKxjCX0+Cwz1IP5/+JvuCkn39TJZaV0k3bYLaQ0phuIa/+tWvpvMOOOCA0lZfgBKS73znO6VN/zci+wYssxyxQnJaL8XASOnu7i52g7L8iOwDatl2+insH5XOcC7qWqf/yn7QctPsO/YP5Rn6eepv0A7Q71GfgOtIbQdlGPTn9VmFcmtKVCNqpxJoBpUPorIuzlktRc+1QwnKpz71qXQeJUw6T3iMkm+dv7QD7CPuzRErymlHDJVdcvzrPU+ccMIJpU3ZeUR+RmH6gRkzZqTzuNdybUdEnHfeedFsurq6im+t/iUlwLpXsZQ3/QZ9xuXepbIz7kH8ruuuuy6dxzXHZw6V+tMOaCoMpifgulc/au7cuaWtvh7lY9OnTy9tTc3CNDDqf1keZYwxxhhjjDHGGLOK4pc2xhhjjDHGGGOMMS3IiORRvb29JfyL2a4jcjiihvdRksBwI60swXAoDT1iOC1lHRqiykoMt99+e2kzm3/E0Ez9hOFWGrJFjj766NJWCQnDExlyrmF8GnpMqtC9ejKOkTIwMFBCQik9ishyBa38xPFgSJeGPVIKoeHrDNumfExDfxnCzetgpQ69Dg1LZZUMjg2lJRE5FE4rrjD0m22G+0XkOah9yjDIZoaBjx49usjUtDIA+7OqiFLB0FOG4usYbLbZZqW9wQYbpGPsd1bkYUhkRMSsWbNKm/NeqxVwfWj4KiUHDIdmWHBEztqvGfc5PmyrFIvVHC6++OJ0rAq11CoIjdDW1lauRyujca5wLPQaOLdVJsFQas38z2oI/Py99tornccM/AwDZmWNiCyX0jBarmFWJmFIfkTEvvvuW9pqV2h/uMY0DJwhsVqRS0Oem8XAwEAZE53bDMPVUFuuP1Yl0M/gmlD7wjHnHqn7J6uaMARXw3Npe3Wu05Zx/FUiSwmRyvIo16HN0eug7E8/o6q+0Mx9ccyYMWUv0D6m9FLlCZRBUQajkjHaJK1qQTvEcHiGyUdkv4pyJrW7vEatqshKJ5TrqT9EWR+lIBG5PyjXU3kU98zrr78+HVO5SbMYHBwskhSV4XJc1e866KCDSpv9XE+Co1WJWD2KFbx0ztBHom1UuTrXvcoi77333tLm+KikkWOl8qha1chUOsq5W2vPbKZv09fXV+y6VnLj2GhlKcrC2XeUI0Rku0HZYkSWwXB/4vyIyH4E0zpodVj6X+p70tZyn+A8isj+pko+6GNzz9R5QJus9kLncbNYvHhxmaf1UmFodVJKXLjPqJ9LH1BtNqXhnCe6LzLVA/0WzoOIbAf4XBmR/Seuddq/iGyztcoYpZDcC/W5knaaVY4iVkisVPLXCEuXLi3jUa8aXz2/mPfAOR+R0xyoRJpyYK4jlfLS/nEMVWrEftHqYLS9TI2g84DzUWWvvBf6cFqZl8/Z3Ccihvq9w+FIG2OMMcYYY4wxxpgWxC9tjDHGGGOMMcYYY1oQv7QxxhhjjDHGGGOMaUFGXPK70u8xZ0hE1rRpbhCWOGTJYeY3iMg6atXJUz922WWXlTZzo0REHHHEEaXNXByqy6ZGWPNoME8ANbX77LNPOo+a8B//+MfpGHNGUFutfUMNG3PkRKzQvlGf2ShdXV3lejS/DnWnqvnjePO8Rx99NJ3He9V+ZZ8wn4GWR6UGcL/99itt1U1Sp6s6bOYNoK6bGtaIrGnW3C1XXHHFsNeo+XN4XZpDRu+tWTzxxBPx/e9/PyLy/UXk9aZrjGVPmUeFpQojcllM1atT38s8JDqfmMODpS+nTp2azmMZPZZejci6YObP0VLCnHeah4prk6UDNY8P9eFaNrzqH9VfN8KoUaPKOtO18sgjj5Q2cyBE5LnNcdLcN9Sqq1aWdpP9paXGqRG+++67S1vLu7NfFeYKol77qKOOSuexz1kCMyLrmNkfWmp29uzZpc38ZhHNLddOli1bVsqmag4GomU+OT4s/au5dyZPnlzaqo9mjgPmLdp1113Teewn6u41vwqvX9czx/HSSy+t+Rk77bRT1ILjz9wFmleB+4/q4KscD83U7g8MDJR5pffNean7HfcC9r/mhGN+Gi1/zfvj+rvyyivTefQDeB731Yice0rzXPG6Pv7xj5c27XNEzj3EcY/I9oLXrtfBPBNa9pTldXXvboSBgYFiozX3BPe+d7/73ekY/bJ6a4B59TTfTbUfR+S5efDBB6fzmKOEx7773e+m82gHdL9jfhraB9povY5bb701HWN+GI7jtGnT0nm8Z+bSiVjhl2ses0YYM2ZMuXf6YBE5L+Jdd92VjjFPEvdCHSfaQs1ZQR/jfe97X2lrnhnaMfajPgfQXkyYMCEdO/fcc0ub+ydta0TO27bHHnukY/RZ+Uyz4YYbpvOYf0RzgtAPPPPMM6NZjB07tswPzS/CvCHqOzC/Em2IPsPRZuk9cX3T1nzxi19M5zGvar28Y/TH6vlS9L11r6bt1b2Lucy4f2o+Jo6x5i2qrlF9yUYYNWpUyc+i+aw4bpq/VH3rCs0lxhwxCxYsSMfoE9DnUx+c+yKfxfS5iO8e6vkstK061twndHw59nyO4b6jaM65efPm1Ty3wpE2xhhjjDHGGGOMMS2IX9oYY4wxxhhjjDHGtCAjkkctXbq0hDAddthh6RhDZrU8IcPCWCqP4dH6GQz5rL67gqXHtCQaS3cxfO5LX/pSOo+hpwyDjMjhjgyN1rAzluuaOHFiOsYSfgwr07Axht2x/HTEilBUlYA1AsOHtcwnQ/G0bDvlEAw9pUQsIoeoaslSSt523HHH0lZJAcMUOZ5aFpnfrRISlgdkKKaG+FGGcfbZZ0ctPvjBD5a2ht0xNFPHV8uUN4uOjo4S4qfh6yyFqCF8nJdcHxrezDWmkj5KzBiyqnOGoeUM5z/nnHPSeVz3GjbK9f2xj32stLWELOcWQ4EjchlOftfJJ5+czmN/aLnYyl5oaclG6OvrK32msiSWS9eygJS3UGqhksbzzz+/tFmOMCKHlrN/1CZzjfGYhl/XKimt18gwc52bLMepEhLKJCmBuuiii9J5DGe94IILah5rJj09PSXMVW01/61SC4Y3/8d//EdpazlQrgGVpdE+srSphpxzvBiyz5DeiCzdUFkBw5dpRzXUmKXWVXbJvYN+Astg6nepPLOSBOhcbYTOzs7yuSpppWRCx5d9zv1JZXvvec97SlulU5RBbbHFFqWtUl6ue+4zarspj1LJ73bbbVfa7Fddbyx3rLIajiFD1bWkPf0l7ukRETfeeGOsDChVVHkkw+3VjnOPU0kZ4f5J2UVE9mMoAVZJHX3b733ve6W92267pfO4H2m6ANpYSnK0zDD7WWVPlCpwLWl5ZtrpSZMmpWPV/l9PHjtSnn/++SLPU/vE+UYfMiLPP56n+yLHnv53RJYrUIql59FGU4qjsl5KfbSPaCMoT9N5xXtRCcmee+457DG1P/xulQmr7KhZtLe3l+c9lYjQRqmUl+PDfUb7hb6i+hzcT6ZMmVLa//Vf/5XO4/zi5+kaYDlqlc7T/6Y8Ucuz066oz05bWdmviKF+If24jTfeOB2rZD3NTKfR19dXc3685jWvKW3dFykxmj59emnrsx6fpXQPInwe0fcLvA72q5Yopy+r+xHHm+8o+NwXEXHhhReWNst6R+Q1x/mifif/jusgIqeB4V6TrnXY/2uMMcYYY4wxxhhjXlb80sYYY4wxxhhjjDGmBRmRPGrMmDFFaqKhagzR1QowDHFj9nMNp2V4vIZksfoA5Q9z5sxJ5zHMjGGoJ510UjrvBz/4QWkzi3tEDqunvEWrF/E+NXs5w+54XyqxYmiXhgkef/zxw35vI/T19ZVs9AwDi8jhuCrVYj8wjE0lMTNnzixtlVAwRJIhqxoqyFBOjg37MSLPg+OOOy4du/jii0ub48Ss/xFZeqJhtOwfzgkNy1WJHmHYtV5/I6y55pol87pW32DovIZQcp1y7CiziMhjrOH8vA+G7mqYPucQQ0O1ygirwWlmeFZ3UKkFYSgzJR4RecwZqq7fxeu99tpr07HKbmnYbKNU/X7nnXfWvBa1LZyXDEvVeUjJja5nlZ9WaIgqZZ8MW9YQVUo3VPbEMWSIvu4TtJk6NlzPlAqoXIjh6VrxQ0Pcm0Vvb2+Zm+wH/beuD87nY489trRVHsWwW5V3MtyYa1jnE6UXlPuorJT2SkPsKZei7dDKNpyfDPWOyOuP1657Ef+toe/33XffsNfeCL29vUVeoDI29jnlSxE5vJv7B/f1iByOrZ9PuRHvSfdP9iX3PpX6sL80VJ62hLZMZSj8Lr0XzgtKrFSWyvmue4iGrjcLShVpMyKy/6HSP9pO7p+s6hZRu7JNRN4XOY7VfK2gTaBcTSuwsM/Uz2LIPfc3XW+U7KnMjbIqyoZVNs/9Wv3XSr6hdq9ZqPSS1avUj6CckP2ldoxzXe0px4DSFN0/We2Hzyq6ZmulWoiIOOCAA0qbVYx0/6S8n7Y7Iqc04NpWORLtqUpUVlaF04GBgdL3KtOkP6hyREr1KGl517velc6jnEntCffTf/u3fytt9VsoW6W/oOk0aCt1rnOOcn/QPY1rR58FOG8oub3hhhuiFrrWq/2/VuWmF8Naa61V0ofoNfO5QyvCUnrJSoT10ohoNVFW4uWeqf3P9Uz/Rf0G+im0uxHZhvJZV6vxUb6k70C4Timz1Ap4nNNa6Vbt0XA40sYYY4wxxhhjjDGmBfFLG2OMMcYYY4wxxpgWxC9tjDHGGGOMMcYYY1qQEeW06enpKSUFVUNI/a1qN6lRZI6SfffdN513yimnlLaW/qV+jBpILWW7zz77lDZLaWpZb+oQqSeNyHo85gKYN29eOo+l/lSLRp0rtYBa9o8lsjX/SKWPVI1rI3R3d8drX/vaiBiqree4aU4JarSpS2QfR2RtoJaf5Byhlu+hhx5K51U5dyJyKUTtH2qVNb8BNcgs/80SmxE5P43q3KlvZu4C6mAjch9ozhyWHf33f//3aBZLliwpc7/e/NDSqocddlhpT5gwobSpO41YUeI6Yqiem3mAWNZdNZ7U3HKdam4P9js12hG5dGC9PEhz584d9vMiIt785jeXNvML6FrknNQcD1UeFbVLjcBcKNonnHsci4jadk1z/nzuc58rbS0tyHnBcdO+Y34VanaZWyAi67d13fMY7T/nYkS2hXq9vDeOG68pImv5tdSu5mNoFqNHjy7zkfM1IuvTNQcG+536cK6piFw6VOdJrXxNmneL+Rmo9dZcA/wuzXNVK7cOc6FF5LwB48aNS8f4d7wm3W94nuY8qPKK3H777dEsli9fXvZpzZXBNa+5IugrcK/SNcs9TXOG0Feo9uaIoaXO2ZfcM3fZZZd0HvNoaJ42+m3sV82ZQv9I1xH3WvoM2jccX90bmpl3gbS3t5d71BLatFnczyPy/sHcGZpPjOtb8wXRn6XPp2NAm0A7rCXKOcbMfxKR83rRXug4cs1qzjPuycxH9P73vz+dd8wxx5S2lkyu8uo1M+9ie3t7uR7db7k/ad5LPmdwXmoOF+7zmrOCY0MfQPdFrmfOq3p9rHOO+xHvU/1y+jqaA4PXq/OMMKcW83JE5DxNWlq+Efr7+0u+HLUNzD2ivgTHi7mbNEfJxz72sdLWXD9HH310aTMvjOZNoT286667Slvz53B/0lw1Z555ZmlzrHTf4v6g+fY4/vR1mBslIj8DaU7Yaq5pPzVCb29vyTujzxm0Geq3M3ct7ZOuRfqymuOQfh7X9gc+8IF03qmnnlratKfMFReR+1XXIkt70+ZrjkfOVS35Tdt79tln1zyP16X7ovpLw+FIG2OMMcYYY4wxxpgWxC9tjDHGGGOMMcYYY1qQEcdRVSGhDLeLyKHfKh9hGBXDo7Q0L8sTapg5w4b43Ztuumk6j6HZe+65Z2mrbIUhgRqCecYZZ8RwaMgcywJr6VmGx/LzWZYwIocJazhdFeKpoX+N0NfXV8K4NLyOoYiXX355OsaxYYnAKjy2gmXLKUuJyPdK6Uy9coCcOxqqzHmgZUkZNsoywCrdYHg7yxFH5PBufv5tt92WzqM8QCUFDN1rJoODgyUkWcMhea0ayskQWpbb0/BmhulrmURKjBgyqWPAvpg1a1Zpa2nTs846q7QPOeSQdIzrnrIClQIxtFJLEzLEUcskEsrFWA41YsW9PfzwwzX/fqR0dXUV+8XQ3IgcZqslYxn6Tbur4aC8Vw1fZVgmS3JrSWOWumSYK9dXRLaZGqJPe/Htb3+7tHW+PPjgg6WtIdwcQ9opDedmf+g17r///qVdr5Tmi6HqJ7XhtA0aBr755puXNtew2mXuhSojuv7660ubZWgpNYvIIbq8RpWocU/WdUr7y3Wp90VJnV4H93/KSzgHI/I+ouWiK4lYM8PA11prrZg8eXJE5FDsiIhtttmmtPUYbQb7WCXXvFcNM+d477DDDqWtkl/un5Rm047rdehapG3kXOJcjMh9q1JFlb9XqG2kVETl9CtLqtjb21vmnK4jzr3Pf/7z6Rj3E0oQ1V/gPaoskDaVY6Jh/5T10adcvHhxOo/+h/qe3BMuuuii0v7kJz+ZzuM+Tpl4RMTMmTNL+1vf+lZpq9yOPsNxxx2XjlVlxNXvaYTOzs6y5rVPKEuiTxqRpRGc22ozudfuvvvu6RjnDNuUz0Vkv5TrWeVL9P3Vbzz//PNLm/2newj914MPPjgdY/9QTqLfRWmOSjBYgriZjB07Nrbffvthv4P7tMrv2Ye0Zeqb8B5ZHjoiywIPOuig0r744ovTeUynwX1GfSnaTe1b+jH0TX72s5+l89gH6t/Qx6ad15L3vC/6ZrxGfU5qBKZEUZ+bY0g5VER+vuY169731a9+tbRPP/30dIz9xfO0hDavizaZtiIi23K1CbSN9WSRfObkHhyRJYC0Kyp3pJxO5ZSVD1IPR9oYY4wxxhhjjDHGtCB+aWOMMcYYY4wxxhjTgviljTHGGGOMMcYYY0wLMiJR+MKFC4v2XMspUrem+TGob2MJPC0DR+2mfgY119RxamlTlixlPo/rrrsuncfSaaorY+4M6oqp1VdUP0cNKXMDaC6FKVOmlLbmZ6juWcu8NkJ/f38p4am5h5hXQEvNMa8ONaivf/3r03ks1frII4+kY8zNwJJ/HPeInAuH2vrZs2en8z784Q+XtvbRZz7zmdKm/lg10vVK6DFXA69J9eWXXHJJaavOlPkQNM9BI3R0dBStvGpYFyxYUNpTp05Nx6j95TzXvE5cm6rrpDaU46glOPkZvHfNQ8W8Gtp/1KgyF8Bee+2VzmN53Dlz5qRj1IFvt912pa2ld9lvtcqEqs1qhOeee67kJFENNe2O5hRime8NN9ywtDW3Bftccw6w5CS19XrfHGtqfXl9ERHz5s0r7c7OznSMa4w6b9W5M3eG5mKizp02gfY+IuuMf/WrX6Vjd955Z6wMWC5aS9EzZ5bmM+OaY9+q3poll7VUJfMM8e90HLmOmHuDYx+R85VoHjjqtJmnSOcWc0bo3spxZc4lzaPBnCf0LSJW2CO1FY1SzSvmlYnI9lXztHGuM6eE5iKgP6D5q7ifMvcN13ZEHt/58+eXtq5Fjpv6UcwNxTwEmvOKeRvUx2KeogMPPLDmd/EzuU9EZH9Oyy43QldXV8lNQZsUkfvz4x//eDrGPYi5Gn7wgx+k87jXa0l29hnz2ND+ReQ5RB9Jy78yH81nP/vZdGzixImlTftA/zoi56PREuXMyThp0qRhryki36fmtajmUJXbphkMDAyUNa/56ZgLhOWCI3KfMJeT9v9WW21V2poLhWuJPiVzlUTkfuZ8UZ+Cc0738fHjx5c283kce+yx6byTTjqptNXfpm9LWH4+Iuc9Uh9Y8001C5aL1jHgNajfzT2d81fnAsdO/SfmfOL+oc819Ofol3J/i8g2W8fgoYceKm3OO92ruV8xx2NEtudEcyTR99Ncp1V+Uc0x1wjt7e1lPDhfI3IeV81nxn2ea0JzrHE/4hyNyL4J91bdg+lj8Hs1VxJzXvE5NSL7htyr6V9F5Pmj/gfnOJ8lNPfZl770pdLWHHma53c4HGljjDHGGGOMMcYY04L4pY0xxhhjjDHGGGNMCzIieVRbW1sJCbrwwgvTMYbpaRgmw6MYeqRyoKo8XMTQ8oQMmaPkREPrGHbHME+W/47I4W8azs8yc0RL7zK8XUPE+d2UTul3MURVy6hXoV4aztgI48aNK7Io7WPeN0vJRuTwWYbosUxvRA4L0zKiDJ9meT2V2lHecuihh5a29h3n2dlnn52OMSSN46ZlhhnipiGWDDNkWXINWWQov5ZH1dD1ZrFs2bISBq+hfpQp/eIXv0jHKHVi+LWWwmYIopY45Hzm/WlILkM+KdfQUNBaksaIXJ6S0hMNTWfIs8oYWTqXJQdVEsBwaA1lrsJ8tQxvI6y//vrxwQ9+MCKGloXlfNY5xPFm+LCWU2ToKUsTR0Qcf/zxpc11r3OJEiauS7XdlLdoqW3CeUapWkTEOeecU9oaRsvQZYboU3IZke26luttph0ly5cvL32vMlnKZDT8l3Ijjp2G5HJeaglt2jOue4bsR2QpDG2Ulpfl3zH8OSKH8tL2qs1jyVINM2d/0KaqzI2h71pasxp/lZE1wqJFi4o8+JhjjhlyrEL3D8okuIZVWsb+2nrrrdMxfialLvWkorRxKnPmOuW+HZH9FO5bGt7OfVfHhutbJUiE96nSKfpwKntuhMHBwSJ52G+//dIx2m6VTlJ6yzB39Qm4PlSuwTVMiZ3KVvn5XNv0BSNWyB0ihvpInPuUi+pewX+fdtpp6RjnBue4yiD4d7o/nHHGGRExdJ03Qm9vb1nj9FEisg3Vec9nEPa5riNKLXbaaad0jPaQ81xlSJRiZdZBNQAAIABJREFU0bb+6Ec/SudRSq8SQf6bY6E+NftA5xLnJ31v9Se4x+t6VjvfLJiKgXM5IvsB+lxFP5LPmWrL+FypzwbcZ+ivq0SZflatPTIi4sgjjxz22iPyPKTMWeG+qLJklq//7ne/W9oq1aWd1nuu/q6Za3FwcLDIg9V+cJ+cPn16OvbjH/+4tGn7aWcj8j6paRhOPPHE0uZ6Ux+cNo7rV0uDU0al5eNpQymV+spXvpLO47ri30TkOcdr0vQV3EN4XxFD59ZwONLGGGOMMcYYY4wxpgXxSxtjjDHGGGOMMcaYFmRE8qiBgYES6qQZnBk6zRDfiIgzzzyztBkiphUWGM7PLPEROWyIIegaCsYwZFbC0EzzDEvVUEIeYxiyVjmgTEjDfxlex7D1yZMnp/NYTUXlAlXFq2ZWrFm2bFk8/fTTERHxkY98JB1jpYRp06alYww7Y1iwSmcYysnKKRERX/ziF0ubGfK1OgX/zTBClbhxbFQawvBhftc111yTzmPIuVazoRyHkgyGMkbkOa0hi1qVqVkMDg6WMF/NQM7wQa1IwftlmJ6uRfathgYztJBrRyVFrErAtlbfYlZ3ldZQQkJJjlZpoJziJz/5STrGsFqGQ6uUhXNIqx1Ufaq2rRH6+vqKjFIrbzGEWcOqv/GNb5Q25WQaGs8+0nuljeNaueCCC9J5DJVnaLGGqFKmpbaW6+OAAw4o7VtuuSWdR/una5HzlrZI1xfnZmXnhvu7ZtLX11fGT+0Qx6SSwlUwNJg2XqUbDL9W2Q3hHrT//vunY5RJUi7AcYsYKvMlPJdSXpXB8hp1PdM+MpSclQsVlbVVNkfDkxuhs7OzhGd//vOfT8e432kVGcoTuBeqrIaoneQcrlf5g74N7YP2Hf2eyy+/PB2jvIX3pWvjrLPOKm21tdxvOG91LnH96VpXiUazaG9vL3uNfgdlEurzcS+kbWSIfkQeY5VC8P7pZ7z//e9P5zH8nvJfrTbDa1JZJKUElB7Xk/CffPLJ6RgraNFO8ZoicnVPlWlVzwEqNW6ENdZYo/j/rMwTke2azm3uhZT8qFSU46b7Yi35m+4llGfT/mnVvquuuqq01RZSHsrr0L7kXOXfROT1x3VP6WNErqSj1fjqVcVtBFY4VRvCfUD9evo7rH6q+yKlQnwWi8j7GNelysQp62G/qFSR8+6yyy5Lx/j5lIuqrJd9oHsX01Hw+YLPxBFZJq6+TyXl+fSnPx3N4qmnniqSP5XRcc/4whe+kI5R8ks7TP8vIvt16gMzhQa/W6sxUfbENAlqH7h/qq2lfJBzRKt80XarjJ3pRvj8oD4bbbfK8F5IpWhH2hhjjDHGGGOMMca0IH5pY4wxxhhjjDHGGNOC+KWNMcYYY4wxxhhjTAsyIlF4R0dH0RuqvpAasXr5RajJU80xtbNa8pW6sJkzZ5a26uB4HjV31L1FZF02c2pERGy44Yal/Za3vKW0tcww9av8m4isjWVZWs3zQk2b6ncrraSWtW2E3t7eIWUDK6iP1bKD1Pcyt4WW8mPeHy1n9q//+q+l/fWvf720r7766nQetYgsA6/5i6hRV20y7+XSSy8tbZ0HHN96uWGY80Wvg/2puVC0LGiz6O7uLjpP1ShzLur1UC/Mtail+G6++ebSplY6Iud1Ye4G1Y6zr6m1Vy0ux4C5gyIifv3rX5c2ddma54XzVY/RJrAMoJbi4/hrScCVwTPPPFOuRzXZzDGw5ZZbpmPMSUIbqvaDx7SUIOcMx5C5hyJy2WyuMS2nzTmnZVo5L84999zS1hKPPO+tb31rOkbNMUu2qv2nzdcyxvVKkTfC6NGjy3zR/E/8Ts21xNxOzOej5dS53qjzjsilNbmOqlxJFdTdc55rHqR6ZdG5Xji3NI/DzjvvXNq6njm/mItFx4aafy0zPH/+/IgYmlepEfr7+8t+ojpz5tXSdUSNO/NesPxuRMTcuXNLW/NScD5znKjBj8h7K/MeMC9eRMQhhxxS2rrGqr6LyOVkNVfXrbfeWtqaJ4X5KOgTaR4N5j3UHDKa16dZtLW1lbwA9A8i8jxiTqbq7yqYz0J9Wd6T5jhkDgvaYi09y+ugvdB5x2Oa9492lLlctMw590zNu8h74bVrXjP6e1ryvpqveu2NsGzZspJnRW0L93k9xnXE+ay5ONj/OmeZC4XPGeoPMO8P9yZdbzxP84X95je/KW3aU835x+vV/Ja035wHfOaKyPuiHtP8cc2iu7u7zBctcX3//feXtuZT5F7AY/qcwHw0zCcSkXMccu0cd9xx6Tz2BW0U51JEzhOjZbhpZ+ibqY3jc4jmu6Ed5d6hzxq0y+rbV3Za8y81wpgxY4qvpz4ZfS+d2xw33pvm8uG/9VmSfhBzounYcM2dd955pa35YTjWurdyT+NaV/9yl112Ke0bbrghHeNaZI6iww47LJ33uc99rrTV39Znl+FwpI0xxhhjjDHGGGNMC+KXNsYYY4wxxhhjjDEtyIjkUUuXLi2h1QwJjMjh3VWJsAqW2GMYvZbFZMi9hpkzfOmuu+4qbZVYMYSboeQaEsgwJD3GMCeGvn32s59N5zG8W0NPGfbFUFaGXEbk8n677rprOlaFmDWztOmmm24ap5xySkQMDb1jmc+jjz46HaOEjPetsjD2g47vEUccUdrnnHNOaR966KHpPIYu8/Mff/zxdB5DLFWGxPJ6lAro9TKMUqUCDG1l+K/O/Xr3rNKlZjE4OFhK32nZdZYI1tBXhuPxWhkCHZHXm8rz9tlnn9Km1EnP49qhxErLl7LkpIZZ15J8aDg/JTk6xgwjpbzhjDPOSOcxjFpL4Fb2oplSxTFjxpSygZzLETmsnfYjIo8Vw1I1tJLh1xqCzLB/zgkN76aUhiGlGjbKsdY1wJKoKkkjDL2vF97Ne2YJ94gcWqz2QiUazaS6Ly2Xy3B2LRHJe2TotJajZHg8xzQih/JyDag95N9xjFVSx5LWuqfRpjIUX6+JkgOVrXKuMdxa1zPnkH5GJetUCXUjdHV1lfnHEPeILG9RqWIteYKGxnMNUNKmxyhZ1VKhPI9+g5bMvfLKK0u7nmR19uzZpa2yF85jlWTwGMPFKWWNyLZSw8A5b1Wa2whdXV3F3uv84L6j9pDXOmXKlNLWz/joRz9a2pTzR+R1xX2R+3FEDqunn6vXRPul30V5HP0xnbvcW3XvovT8y1/+cmnr3sd5R9l0xAqpn0qeG6Gtra3cB9eU/lslGZQW0nbpeVybup65P/3iF78obfWPJk+eXNq0XVoGnutZ/RLK8HjtKm3lM472M/0GykXpP0Rkf07TI2yzzTalzXtulPb29rLuVcJJP09tA/c7+msq06ffq3sEn20ozdLxodSG64/jEZFTNlCqE5HHh+Ot0mA+x333u99NxzgG9Pd03hGVWw8ODkZEfYnzSBkcHCzrh3YmIq8j9W24F/AZRKV/XJu6f/DZiXNEJav8fK4BfTdAX1bnHK+Da0z9KEruaVsj8r7I+afPNPwulV1qqpbhcKSNMcYYY4wxxhhjTAvilzbGGGOMMcYYY4wxLYhf2hhjjDHGGGOMMca0ICNOllLlNXjDG96Q/v9ll11W2qrDW2uttUqb+nfNqULNbaXPq6DGnbps1aZRt0wdsJ5HbaOW2+V1USOsuklq/FX7Rg3p1KlTS1vLlbFsIfMaRKzQmWtugUZ46qmnipZdNbbMVaK5LbbeeuvSpo5W8+0ce+yxpa2lQvl948ePL23tO+Ygob5QNd/Mo6F67Vq5arRsIO9TdeMce+Yb0tLgLFWnmm9qn5tJf39/0dvrGFA/rOuU8556fS1fyjXM9RuRdfjMOaPrgzk2qPfUcrj8Ls3nwWukzljvuV75X+YOYZ4IvQ6WQNWSgNUaVLvUCD09PaX/tORxPZ18lcsoIo+NXvNWW21V2prroJaWW/XQzL9B/a2WLSYsyxoR8frXv760qQOupy/XY5yf++67b2lrLhRqwz/84Q+nY1/4whdqXnMjjBo1qoyD5v+grlrz3TB3Ae2QaqCpk9f5x/5kLjAtc8v1QZug11urnHxE7fLOWnqXeQh0H6Gdpm3Xdc+8GrrWKy295qpohOeff77oyXVfZI4FzTPD3E7cBzQnAkuDa+lU2k3uVTpfqPHnmtUcC7R3artpOzhHNIcg817odRDma9FcGbwOzbGhJYCbxfPPP1/K4jKnYUSep5pziHnqmDNB89wxr4PmEGPZYZYFZrn3iGyn6fswn1REnhe67rkWa+V9i8h2Wkug0z/bY489SlvzldGe6zr905/+FBFD7UgjtLW1FRuoObdoM5hHMyLbA47nTTfdlM7jfqT7B++D/r36hpw/9M+5viLyGGp+Ho4bff9Zs2al85i/Sm0T/835rfdF/0I/Q6+rWfT29habwHylEdnP4PNcRO5r2n4tZU27rHsV90zudzqOXHN8ZtA9hzZby8Rzz2S+1QkTJqTzfvWrX5V2veeCI488srQ1JyXntY5b5b9eeOGFNT97pIwaNar0s+ZHY45AtQu8buYI02d+2ivNQUi4rnQP5l7Itai5uZhPR/uVJdjr5WzlXqj2n/7XzjvvXNpcvxH5mUZzpvGZiSXPiSNtjDHGGGOMMcYYY1oQv7QxxhhjjDHGGGOMaUFGJI/q7Ows4aIaqsbwOw1BY1gvyztruBUlOCytFZHDnq6//vrSZjh8dY0VEydOLG0N32S4soa7MQSKoawaUktU4sPwfvaV3hfLLjKkL2JF2B3Dbhtl6dKlJdxVJQgMkdaQUoYJM1RQZSkMudWQXoa4VWG1w30GQyl5jVrijn+nYakMx2bYnYZiMsRSw+21DHOFlqzW+UNUDtcsOjo6Sjgsy65G5FBbluuMyOHsDGnUvqVkSeVg/DuGxOt4c3xY1pUypIgsrTzkkEPSsXvvvbe0KRvRcFiGqGrYKEOFKfdRCQlDdlVqVK0ZLUXdCIsXLy4lNrVMNueYzlnaMobxailElvGtJ99k2DylLRHZhtIGq1SRoaEqk2NIO+9Fw0Zpy9XmzZgxo7Rpu9WefuhDHyptllTV725mmeHu7u4SSs/w6Ig8j7q7u9Mxhu9SNqRzm3ZIpbK77757aV9zzTWlrVIIhgbz3ikViMjjqNIESj44B/UzKNnS8ta05/wM3Z8Z6q9yjcrX0BD2Rhg9enSxS1qeliHp+p21wpnVfvAzdK+iHJglh3UM2UdcY7qP099SGSMlH7RlumY5B1W6yfFQOQ7hHjJnzpx0jLZXS+M2wrrrrhsHHnhgRAyVKnO9aTg/y83Sz1OZG+9XZaD08+hDqryMc4N2VG0e+4jzIiLvFywTq/s95XH0uSJyuWv6bbrH0X6rXGBl0N/fX+Y+nwkisv3T8rjsL6ZD0JLrXH/q49GW0Q7o/sw5Qr9E/T3uT2r/Kd+bNGnSsN8bkUtd6zMIfWX67CpLpU1WqdJFF10UK4Ourq5iHzUtAfdilZlwDOhbq4STfaE+OfcPypJUMkNfgmtdJaz0i1QyybXDfZY+VkS2A7qPnHnmmaVNn0tlblzfKqOu/Gh9Nm+Ezs7Osofo/dCeXHfddekYr40+kcpkKZ2i7xGR90U+86s95f7HsWA/RuR9Rq+DtpdSKX0m5FzSlAO8Ls5VTd1Beb/6WPrv4XCkjTHGGGOMMcYYY0wL4pc2xhhjjDHGGGOMMS3IiOVRVagUw+0icnjRxz72sXSMIYPMHq0haFdeeWVpM4w3Iodo7rjjjqWt4avMaM1QIw35nDlzZmlr6C5DQBnmxRDaiByyuttuu6VjrJLFz9hpp53SeWeddVZpa/WAKut5M6tkdHZ2lrAzVrWKyGHVCkPhGC6tIdy8Vg3ppVyGYa8aNsoQN4775Zdfns5j9n1WU4jI0juGM6o0hyHtGp7HEHFeI0Pi9bs0jPa2226LlcHAwEBZVxrazvA+rejB+2d49Dve8Y50Hsdbq8Ex7HzvvfcubQ1znT9/fmkzHJ1rIyJnWtfQU64dhqazMlJEXt+8/4iI8847r7R5nxqGzGpSOq+1Cksz4Biy8kxEDttW26UyygqVmPB+VFbDkE2uZ5UbMQSdfafyD9phlX/Uqrxw0EEHpfNOPPHE0mY1k4gsRWA1CB0X9pvKRtRGNIunn3667CG6H3G9qZSEcjBWRFDpxrbbblvauna4x7EvaDcjsmSQtpfhvhFZVqAVF7k2GWauc4ao1IvriuHWalM53uonVJUfmrkv9vT0FBmC2kxKaClzisj3wP5XmSF9IP0MVpDgnqZjSDkA+1UlUNyrVLp7xRVXDPt3ajNp/zVEX2UKFRryzzB93UNWVsWahQsXluofuh/R/1D7TvtI6QYr5kXkftIxpu/JCi6VXKuCNoLrV2WGnIdqyzgnuX+qDJZ7tUqb6O9ccsklpa3VVPmZahMqSXEz1+J6660XRx11VEREfOlLX0rHuN5UtsdjlAdpGgbOPa0gSUnx4YcfXtoqv+EY1tpLI/Ja52dHZN+Ta0f3d+4bKoHmMc4DrdzKz1ebz0quavsaoa+vr0gp99xzz3SMPsLpp5+ejjFlBP2ib37zm+k8rkWV0HLt0y5r9SKOF31KTfHBFAtqvyndpgxM7SHXn/pjlKzx81UOx7mrMqjq+9Q3a4TBwcEyVup7MB2CPtfSt6a/ob7sueeeW9qUQ0XkdAWUhVGmHZGlZexHrYTKPVhl25Q2UUalUkLafEq7InJ1Us4rrWJIm6PPyPr8MxyOtDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBZkRDltFi9eXPSQmreA+sJzzjknHaOeuV7JV5bomjt3bjrGssMs4ae6Pmp4qXHVssWVLj5iqJa4KsMbkXWoqmWkPvJrX/taOsZSjtRXaklA6so1T0eFlhZrBOpMTznllHRsu+22K23tV5ayY/4eaqEjhup2CXW2tXJUROT7pSZfSx/fcsstpa05EZgLhfp81RByjqhGkTpZaqRZfi4ia1U1FxNz3FxwwQXRTKp70Tw91FurJpY6W2raVf9J/b9q96n/vPXWW0tbc+swHwDnhX4X82Np3hHmzqA2ffLkyek8rqvzzz8/HauVl0pzKbGvNL9UVT5atciNMHr06GJ71H7suuuupa25UHidzGOjeQVo83Tec85wXWqOmAkTJpQ2bYJqkzk2OudoTzl3tAQt54jmhuJncF1qSViOr+YE0XwMzWLNNdcs2nstJa72ntxxxx2lzT7TOUbt+HDfXcGS36rF5txg/hzVv/O7WGo2Iu+ZtJv1csDUywXF8dH5yb2V5VsjVuR10L9phN7e3nIf2nfMb6Blnqm1p+1X/4jrRfP80CYzF4CuMe5xzG2gdpcaevVZ6GMxD6HmXuN36xymP1flj4kYOk9pt3Tv5v7STMaOHVtyHvzwhz9Mx2hf1PYzjxRzuGgJVtpNLSm+//77lzZLympJc+Y04NxW+801wLUXkceR+Xk0v9Tuu+9e2pqbiLmy+Pk33HBDOo/rjHnw+G+9vkZ47LHH4uSTT46IoTlcpk2bVtqaY+Puu+8ubfqy9957bzqPObI0fw8/k3+neXHYD/SP1HYzL46WKGfeC96n7ouap4hwbOgPa241Pmdo3p2qVHSz6ezsLDk6NK8H7ajmMql8rYi8JvQ5rV6uJfqUzBmm59Gn4f6p53H8WVo+IvdfvWdY5gzVHDwcE/oF3Esjsp+geRCrZ6VmPi8ODAwUW63P0LNmzSptzQ1F+0rfTdcH/TfNKco8QszXxNL2EXmt016rH8p9TJ/16ANzPV988cXpPO59mgP2sMMOG/YzdKz5roRzM+KF5XpzpI0xxhhjjDHGGGNMC+KXNsYYY4wxxhhjjDEtyIjkUe3t7SW0VyUOlDFoqTOG0DJMdtKkSek8huhqeUKG61LWo+G/DDNjaBnlShE5XE+lQAyZZDiXfgb/TktmMqSaIYwaaswQN5XPVCV2mxlKPDg4WCQVlI1EZKmTfidD40466aTSVhkD54GGejFkjn2pIXMMLWM/avi1hnkS9jNDRTU0naUctZQpw84vvfTSYa8pIksAeP8ROXSvmVCqqGGjDFXU7+e64rWqvIxjTElLRJYOcUx0zXIOMfRRQ54pj6N0IyJiv/32K+0777yztDWEm9ekYYuUtjG8XcPnGSKvJVb32muviMhlzBulr6+vSAg0NJ6SE5aGjsiyBtphXW+8bw1tpR2mHdBSuFwDPKZrhfZUZR0Mc2X/6zXxM0899dR0jP3DEFUNkaeMk/tOxFDpYrNYtmxZmbeUy0TkMHDdq2g72We6f1JGpPaF8lHuW2rnKNehpE7tPGUdGmLP/Zn2WyUTlINouW7adtoBlaGwjK5Kbqv1rbKTRhgcHCxzTPdo2gItN8r1QamLSsQpKWZZ+ohsX2n/PvGJT6TzvvWtb5U2baaWiGeYP21ARLbDlIHpWuQ81tLZ3P+4nlVGyzmie4PKYJtFe3t78TN07nF8dH1QDkGJEW2XfuYJJ5yQjtEfPOOMM0qbJYEjakv/VObGvVX9y6uuuqq06V9y3UREnHbaaaVd+ZMVlFFwLqislH6WSnWqfaSZJdw7OjrKnFP5K6XPamu5L7BfDznkkHQe5Roq3+T3VWXHI4baJ/qNtKe6FumnqB9FyRL39BcqxYrI/c55qz7hlVdeWdoHH3xwOqby02axaNGi4rPp+udaUfvO+2C/qKSFEijdF3/yk5+UNtfRjjvumM6jfeQ8V7kP/SV9DuG8o9+mz4vsZ91buedzPul5RH3Uak/Q+dMIS5cuLXNdJa70r1TGVWtO6Xm8B/U9uTYpe1c5HW0SP09lbBw37SP2P/dntR1XXHHFsN8bkf1L+uU6X+66666an6GpLobDkTbGGGOMMcYYY4wxLYhf2hhjjDHGGGOMMca0IH5pY4wxxhhjjDHGGNOCjCinTU9PT9HvaU4blrVSLV+tMsOaS4Z6NJYSjsh5Tz70oQ+VNrVuEVkPSa2vls2lbldLfPIaWZJLtd3Uv2qODeYIofZS9f/XXXddaasetvoMvcdG6OnpKbo5aif135pjg/1FTaFqSdmX5557bjpGnS3HSTX+nAfUlLPUe0TOaaD5eZgjgqXf9Hr5+aovpP6VWkZqWCOyzlj1lvVK/jZCd3d3mR+qCaZ+eNNNN03Hvv71r5c256JqeNlnqv9nng6WmdQSwexr9oPqrZnLiWVTI/K9TZ8+vbS1JCN1onvvvXc6xvLEXKc6n3jPmreoymXQzHKK3d3dRXer9pTlRrVfaZ9YClFLHNIOqx6Z2mLmO9HcFrSbLOGu85xrUfuI+W5ox3XO0b5qOUvaC+aG0e+q8jwNd41ahrlZLFq0KH7+859HRM41EpFtBfeBiKy55lrRkpbsM5amjMgl2dkXOrdp57gWteQ3bbHOJ34+7aHm4OFY6Z7GMaCN1vm/3nrrlbaWVq7mru5RjUDtvu7zzLXEcrQK+/WII44Y8vkVmpeCa5N7EPN3RGRbyNwJ48ePT+dxD9Y5N3ny5NKu5mzE0PwvtD+ai4l5FjiemteEx7RUuubEahaPPvponHjiiREx1A/lPap9Z36Ut7zlLaWtfuiPfvSj0lZfjjnXWJpaSzgzbwHtq+Zz+9SnPlXaLGcdke0F83JoThv62DqfmGeJY6frrcrnFjF0f6j+Tn3JRujv7y/2Sn3fXXfdtebfMQcGx37evHnpPOaW0XK8zLv46U9/urR1vtJGMFcQ50dEXttaZp5rmP41c91EZNuhOUG4nk8//fTS1jLn3A80dwj3Ic0D1widnZ3F/mtJeO4RVVnwCpbN5vORwtyFmqPkuOOOK23aOd1nuD54bKuttkrnMd+U+tvcr2kbNV8Sc55oqWeO17bbblvaV199dTqPvrOWkK/6Uff0Rujs7Cz2Re0Y+1ztDseUNkOf0/QzCec6n0E0jyGPsY/1PPo2mpuOvsjMmTNLm7mrIrJdUZvAPYV+oJYy53NlvfcctXCkjTHGGGOMMcYYY0wL4pc2xhhjjDHGGGOMMS1Im5ajqntyW9sTEfHw3z3RNJtNBwcH1/v7p/19PIYvKx7HVR+P4eqBx3HVx2O4euBxXPXxGK4eeBxXfTyGqwfDjuOIXtoYY4wxxhhjjDHGmJcGy6OMMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQfzSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQTpGcnJbW9vgyroQU5/BwcG2ZnyOx/Bl5cnBwcH1mvFBrTKOPT09pb3OOuvUPNbWtmL6jho1Kp3X3r7i3fHy5cvTsf7+/tL+85//XNpLly59kVfcOF6LqwWr5FrkOoqI6OrqKu211167tLu7u9N5vb29w7a5viLyWhw9enQ6xrW5ePHimp8xOPjSTWuvxdWC1WItrrXWWqU9duzY0n7yySfTeX19faU9MDDwor6L/9Zjtc7j2o6IWLZs2Yivox6r21pk36255prpGG0e7aKOxRprrFHa7O+IiOeee660X0qb+XdYJdeiyaxua3FVgGu/Set52LU4opc2xvwjsRIW4cPN+JCXG/bLFltsUdr7779/Ou9tb3tbaXd2dpY2nduI/HC4cOHCdOyPf/xjaR9//PFo0iyfAAAgAElEQVSlvWDBgpFetjFklVyLHR15y954441Le9q0aaX9ute9Lp33u9/9rrTvv//+0v7rX/+azuNDxuabb56O8SHjl7/8ZWk//vjj6Ty+UG3Gw6BZ7Vkl16K+1Jw4cWJp77jjjqV9+umnp/MefnjF7fIFqq4VvmTRdc8fPviCtt5L3TFjxqRjXLeLFi0q7RZ6gbDSqffCiz7LhAkT0jH+gPTss88O+zcREdtss01pP/XUU+nY/PnzS/vl/BFKWCXXojEvN7TRfEHbgD0ddi36pY0x/x/dwLkBc+HpLybkH8Hh4a9OP/jBD0p7o402SudtsMEGpV3v4Y19xl8hIyLGjx9f2nRUZ8yYUfMzjFnVoO3hr/QREYcffnhp64tRvpzhZ+gDGqPe6q1F2jw9jw8dzz//fGnzwTMivwg6+eST07FbbrmltBmtY0wroj7Beuut+OHzhBNOSMf40oYvUn7729+m8+65557SfuSRR0qbL0Uj8ksbfm9ExKtf/erS5n7MdRmRXzbwpW5ExH333VfaH/nIR0r76aefjtWZelFK9CNoh/XF25IlS0qb9k+jiPkS/JxzzknHrrnmmhFctTHm5aZe5KL+e2XhnDbGGGOMMcYYY4wxLYhf2hhjjDHGGGOMMca0IH5pY4wxxhhjjDHGGNOCOKeN+YeGGkXqjyMipk+fXtrMtXLppZem83hsdcmtUi9B3w477FDam2yySWn/7W9/S+cxCSKTD+tnUxOumnzqRJkzhzk6Ipwfw6x6cB2cdtpppb3bbrul85gnQdcH1xyTWWr1KK4jVj5Re8XzNFcNk5XWW8+vetWrSvvrX/96OsY8IBdffHEY02pwvc2aNSsdO+aYY0pb80Yxz8kTTzxR2q94xSvSeZrAuEKT2HJd6d/QV+Faf+yxx9J5tBdcvxE5iTltzuzZs4e9vtWFF+qjjRs3rrR1bJjEmfZO81owL9ijjz76oq7DmH90VkJRmIavQ/NXvVQ40sYYY4wxxhhjjDGmBfFLG2OMMcYYY4wxxpgWxPIos8ryYkLmNFT5/e9/f2kfffTR6RhDXa+77rrSvvnmm9N5LJGp5cDrlQdvZdifGhq85ZZbljZlSRoazJBujhVlUxE5hJgh5hERzz77bGmvv/76pa1lkS2PMqsaW221VWnvtNNOpc1SvBER66yzTs3PoDSC4boaust/0ybpWqQkSm0qz6UUS7+L8iuVnH7+858v7Wuvvba0VfZlzMvF+PHjS/uTn/xkOkaJi5bGpm/BY5tuumk675lnnintv/zlL6W94YYbpvMou1l77bXTMa4rrktKtSNyOXCuy4gs8dl7771L+6KLLkrncX9eHXihZXvf/e53l7baSZZc7+hY8RilfUV/Rku/Wx5lzAujXrqGlbmO9Hv5XWpPX6r17EgbY4wxxhhjjDHGmBbEL22MMcYYY4wxxhhjWhC/tDHGGGOMMcYYY4xpQZzTxrQ0zJeg+RGYf0Fzx7Ak7Sc+8YnS/vCHP5zOY26GH/7wh+kYS9I+8sgjpa35F1j6U3WN1K+vShpmari17Cn78LnnnittzX3De+d5WnqUulHNOcTP5LEvfOEL6bxPf/rTpb1w4cIwptVQffTMmTNLm/kldH0w14XaOeqqly9fXtqaW4G5LphHQ20q/63Xy+9i7gfmdIjIeXZU9808HUcddVRpa2nwVclWmlUfzvVddtmltB944IF0HtffJptsko5xz+R56667bjrvwAMPLO03vOENpa1rkfncuLYj8ppj3hquvYhsO9Zbb7107Kmnnipt5t3R62X58tUB2ha1k7Rrv/nNb0r75z//eTpv++23L23mL9LcerS7tPFm9eTlyr2yOsK+pL1TH2hl9msrjpkjbYwxxhhjjDHGGGNaEL+0McYYY4wxxhhjjGlBLI8yLxm1Si0yZD4iYtttty3t97znPaX9xje+MZ3H8N6NNtooHdt8881Lm2GpLDkbEXH22WeX9guV1dQrA/e6170uHWO4LKVYLxe1yqSrtGnGjBmlraXQH3744dJm2OIGG2yQzqNsjOdpCDfngh5jeDFDxI844oia18v5ExGxYMGCMKs2tebtqoSWqWeIPdeAlqHlvatMgqHCPI+lZiOypPOhhx4qba7RiIiNN9645vX39PQM+3carkyZSL2ymJSHnXrqqek8vX5jViZcO29/+9tLW+Uzo0ePLm2VG/3+978vbZYGV9kTP4Pfq/4H17pKJrkWuUeqPKeeT0NZFeXktEsREXPnzq35Gasb7PPf/e53pf3tb387nbfVVluVNvtf7RbHl3I3s+qi/j9TOBDdq019tF/573HjxpW2pqegj7Gq+oYjwZE2xhhjjDHGGGOMMS2IX9oYY4wxxhhjjDHGtCCWR60kamUR1/Ct1SHs/8VAOc4xxxyTjjHE98YbbyztefPmpfNYPUrD/O+9997SpoxKQ/lfDDpO/EzKuSIi/va3v5X2X//614a/u1FqzTGVnn3qU58qbZU48N+1KmZE5PBQhnPzbyJydQpWYojI4casQPWHP/whnbfFFluU9gknnJCOHXnkkaXNUGZjXkqmTJmS/k3ZBGUM9WQSGnLNf3P96WdQInr//feXtoYas3KOShW5bvm9Kufidegxyjxe85rXlPY73/nOdN6tt94apjb/qH7DSwElgqwaGZH9FvU5uJY4z1V6TP+G+1G9cdS1SEkGv0v3Vr1GUsuubLbZZjX/ZnVD/XTaK/bJH//4x3Qex43S73qV9HQMzaoD54nKoWgvuKczjUDEUIkjsQ0fCvuEqSX4LBGRJaxcs/X6dFWu8uVIG2OMMcYYY4wxxpgWxC9tjDHGGGOMMcYYY1oQv7QxxhhjjDHGGGOMaUFWu5w2tbRq9co0N0PDVq9cWa22/ltzh6xusJ95329729vSef/yL/9S2ixPq+PUKtpD6ijvuOOOdGzDDTcs7VbIaUM4Bl/5ylfSMZYDZc6ZiFyylPpe1dMTalI19w3Lja677rrpGPPYUC+sn8HP32233dKxvfbaq7QvvfTS0m6V+bO6oXlMXqit5XysV/aaGuZ6n9cK48v1cdBBB6VjzG3BvBdcXxE5f4LuEbVK+mrpWeZaYKlizbfBdc8ym/pdXLPaz8zdoCWTeS6P6Zp1Tpuh1MqroP3P/aie3/NivldphTXWDLg+1l577dJ+8MEH03lcHzwvIpd05p6pJb+5nrkGdG1z/Wkuq1plwzfYYIN03mtf+9qoBa+Da/YVr3hFzb9ZHajnj3PfoY+htpC5CpkvjP5KRPZnmMMrIuK3v/3tSC7bvITUK+tNGxARMWHChNJef/31S/u+++5L5/35z38ubX0WoL3QPfMfFe4t9Gfq7Udcv/X2pnr+Zas/hzvSxhhjjDHGGGOMMaYF8UsbY4wxxhhjjDHGmBZklZFHvVDZE//NkFcNUa1XHlVDpyrqha3p3/Dfb3rTm2r+3YIFC0q71cOymskrX/nK0ma55ogsdVkVQgUZhrfnnnumY+uss05p33PPPS/ZNb0QGGK9zTbbpGMM36xX+pdhoypZqlWKT+e5ykEIz6WchP2qn6/XMXXq1NKmPMqMjFp2MSKv5+nTp6djV111VWn/6U9/Ku16tlvLatY6pucxjLYV7Okmm2xS2roPsEQwZUrsS0Xndq19kTY0IqK3t3fYaxo7dmw6j2VJKfeIyHKKTTfddNj/H5H3Xb3eWiHPq7skox61ynfreqPd3XHHHUubUrWIiJ/+9Kel/cwzz6RjteR0L/T6/l97Zx+zZVn+8cPccD/XGqa8SAYoAoIJ8pZGamKZ5Vvaak6UtNnS9bLI6aw/Wm211XKZ6ZzLVatpCmrQMGY2Z1KSiqAJAYWAvGWCKEblqj94fn/dZ5/z63Oe3s/D83LdD9/PX8fznNd93td1vt/Xju9x6D22+7mmy6hKskC9b84XnWPc7/bt29ftZyJyWRU/U5Nk6H3wcxwnlD5GRJx44onJrqWj5v7JuT0U6Im8j+3A9ucZIiJi79693danc5F9o/Io0yxq8myeUXUff9e73pXs8ePHJ3vChAnZdZRO0Y6IuPfee5N9zz33JFtlzjVpeNPX2J7Sm/2jNtfZ5jwDReS/Qe+///5k6xpfOwOXfqvWzrm96UN72hhjjDHGGGOMMcY0EL+0McYYY4wxxhhjjGkgfmljjDHGGGOMMcYY00AGNaZNb1NJ1rSHjF1zxhlnJPv888/Prtu4cWOy77777qyMGn1qXFWzVrt/aiAXLFiQbNUy3n777cletWpVsb6hxsyZM5OtMU00hkrTGT16dLI///nPZ2Wf+cxnBvp22oYaa9W7U0ur/VHqn9r8YGwLvY4pMxnnIyJi9erV3dZ35plnZtcxtbpSixFi6nB9pcZf091ec801yZ4xY0ZW9vzzzyebsZJqMZB0bT3yyCOTzb6eMmVKdh3jeWiq+sGIj7Vw4cJka+wXxmhiHA29T84JjVFC2FeaepaxamrtrG1GGHeHfadpw7mWaMyhUvp3TU3Mz2m646EG24H9oe166aWXJvsrX/lKsT7ON02d/tOf/jTZOv/apd0YPLWYLL397v6Ce3htbDM+jc5nwjLd07i3ss103jOGkZYxBtT//d//JZuxqyLys5XG+GKfMKaPxn+snXO7q6uT0DHLtYZ9PWvWrOw6poJnf7IvIvIYN0N9HetESr8lNa03Yz5qXLClS5cmm3ufzom5c+cm+/LLL8/Kvv71ryf7c5/7XLK5BkTk5+FFixZlZb/+9a8jonlra7v0Nm5NaQ3Vc//3v//9ZOtvTsYf4v65du3atu+xNJY0fg7HD2OfRbS3RtjTxhhjjDHGGGOMMaaB+KWNMcYYY4wxxhhjTAMZcHkUXYh6kkqS19J9U9NFn3vuuck+66yzkq2pu5gyeP369VnZ3/72t2TTfYnu7BG5+6+6WTItHNMw0k1dr3vmmWeysk51OS3BNqJ7oLYJXRNr6WP7G/Yv3fe/+tWvZtdRorFhw4as7Iknnuinuzt42Acqp+DYUxdx9gP7tJaekHNKXbg5t3We8rtq0gGuCTpv6PpYS1Fu3gzbi2vm+973vuw6rpNPPvlkVnbOOecke+LEicleuXJldh3dRjWl+0UXXZTsOXPmJJuyhoiI73znO8levnx5DDS1PU3nGKWyRx99dLI1bSzbRd1nuSfxOk0zzPlB1+CaDE0lGZy3dNtWaSWfWeVRrL+Wqpj3q+02lGGbfPKTn8zKvva1ryWbMgyurRH5XqVy7AcffDDZNakdqcleaeu4Pfvss5NNOUlExLp165LdhHMOz5GUNuk+w/mn+xjnH880em58+eWXk8221f2I87R2RqrBOrX+UtpqlRVwTOr60450qsno2OZY5Fqlfci2e/XVV5Ot8ihK0CmBNYNDTVrDPYdpvCPy/v7Tn/6UlVF2WJL/RuS/M3/0ox9lZRwb06ZNS/a1116bXcc1ViXFnUBvf/PzOl2DSvLi0047Lbtu3Lhxyd6+fXvxe7nv6rsBfrc+C886PPPecMMN2XWU0911113F+yhhTxtjjDHGGGOMMcaYBuKXNsYYY4wxxhhjjDENZEDkUeoi3UJdo2pZoShxuPHGG5Ot7r9///vfk71p06Zkq7v4cccdl+zrr78+KzvhhBO6rU/diekqpc/IZ2Gkf5WQ0LXunnvuycqGmnyD7nwnn3xysrXtrrvuumS/9NJLyWbmi4iIHTt2JLvmZscydeXnuND63/Oe9ySbrpMq4XnxxReTfeedd2Zl6tbcJJjBS12l6RauLuJ8Jo7z0jzXMpW+lLJpROTSM70PwvvQrB6c65yLtUw5nUhPXE9LaB/SHZd9oesYXYa1n2666aZkT58+PdmbN2/OrqOrusplOC4oG+D6EJGv0YORLUrbnFl+OPYicjfoT3/608nm2qifU/d79gO/W+cAxwY/U3M11rnIfq3JnLnGapYG1s8xQ1mb3uNQQ9uLewvd5invjsjPCpQjaj/Vzg2cf7///e+T3ZO5wvunlObmm2/OrqN7usqGL7vssmRrVqPBgPfK9tNMShyXeh7kWsRxr2cOSm3YltoHtT5hHbxHlYvyPvTcwn2c9zhmzJjsOp6RNBNWa+wNxlm1tt9xTuh1bNfabxC2q+6LJWmrXsf1mtmDIiJ++MMfJruWicwcHBwL+juQmSh51lHJKSWNumeWshf1ZNxRWknZpc5Z1nnSSSdFJ1A7K5QkUG9VRkplOqfY95QtRuR9eNVVVxXr4JlFs0Ixoy3fV1BmGZHv6z/+8Y+zMmePMsYYY4wxxhhjjOlQ/NLGGGOMMcYYY4wxpoH4pY0xxhhjjDHGGGNMA+l1TBvVUbcb26KmfaX+9tRTT83KLrjggmRTz/vwww8X74NaMk0NTo2iMmnSpGRTY1ZLG666ZWp/aav+jvFztN06PaZNLVYN0XZlrId58+Yle9SoUdl1TIHK2EMR5bTUP/jBD7LrPvWpTyVb9a6MecK0s6+99lp2Hfv+j3/8Y3QK1K6rdpNzkel9I8ra8Vr6TPadpr7kXNR+pOaYdeh8o/6c+uCI/Fk4t4dCTJt2Y4voel2qY/jw4VkZ2+7555/vtu6IvM01RsXUqVOTPXv27GRrimCOM71f6rypK7711luz69asWZPswUglrH3AttDx9uijjyb7scceSzZjMEVEPPTQQ8nWuch1jvNPY8nwc9zT9DrGKKmlR6V9xBFHZNexTk3X/dRTTyX7i1/8YvG6oQbb9fzzz8/KqKGnTp6phCPymEIc29r+7DfG24iI+MY3vpHsX/7yl8l++umns+v27NmTbE1tPWvWrGRfccUVyeY8j8jXcn4mIl+v20093pfo+sL4bjxf6jrHtqjtH0zJrTENOD84Z/W8pPFjSvfBvtI6uC7rHr979+5k80yjMec4hnSetsbaYKT+rsXHqO13tTML4XhmjIqIfFywL3TvYwykE088MStj/CHGZzQHD/v44osvTvaVV16ZXbd8+fJkcz/WGI/cZ3XccT+txYSrxVniuYsxW7lvROTjTn+vNBXuT5pKnc+zbdu2rKwU56k27/nbgu8MIvK21LMhv4v9yb1Z71fvgzHNuLbqmZrrq8Z/bCe2lT1tjDHGGGOMMcYYYxqIX9oYY4wxxhhjjDHGNJAey6NaLkHqpk3XI3Xvphs8XWY1fSmlQqecckrxHuhmxjS0EbnrEaUVmkqLbtovvPBCVvbRj3402XRBU7lSLV1jSZqg7k90mZw4cWJWRhlAJ6LyB6aoo0upuuhTLsW2mzNnTnYd04YuW7YsK2M6ywsvvDDZ7NuIiM9+9rPJXrJkSVbGMcP033fccUd2Hce+SnOaBt2n6fasLup026tJYVimrsEld1B159+5c2exDsokObf1Or3HEqxv5cqVbX2mydTaleNe001z/S7Nt4jcZfWJJ55ItqZk5tqoa+2KFSuSTZd/Xe+4hu7atSsro1Tk8ccfTzblUBH5s9Rc9vtLOqX18u+aWy/bjHIHrUPlD5Rh0DW4Jl9mO+u84d+1FJxcX/W7nnnmmWSrJJapVGvSBNY50DK32ripyS4ogzn77LOzsi9/+cvJ1nMP90K65avrNPuedeiYoDu2fte4ceOSTUmQwvVVpQJc/9lWKhsujU393GCgbasu7C1URs8zqkp52f88++hayTWKfafrd22Obdmypdt7HDFiRHYd12V9Zt7Xiy++mOzTTz89u45jZvv27VlZ657bSVXb19TGENu1JllVuNZQHqV1sJ1rEjSi51zK0y2POji0f6699tpkf+QjH0k2zw4R+W8/jpna3qdzsZTKu5byW+vg+YzryL59+7LreGZX6etgr6mEaw33vgULFhSv++53v5uVMfQJJauarpvvCih/o/wwIl9PFfYbz6i691H2pGOEayD3CZWxbd68Odn6HqV2jy3saWOMMcYYY4wxxhjTQPzSxhhjjDHGGGOMMaaB9EgedfjhhydXn4ULF2ZllIioqykzItDVUt2G6AKvbql0taWtbmZ0LatlKPj2t7+dbGZFiYj4whe+kOwZM2Ykm9GhI3L3KM2wQFe4mnSALqoXXXRRVtaJ8ii66Kn7dUnCou7dbCO2ubrT0xXuE5/4RLEOuipfeuml2XXa94TPQjc5ushF5C7/tYwPTaAke9K25byqZTWjS3fNRZp1qJs2Xa61/caPH99tmd4TM0tpNjK6L48cObJ4j51CyVVXs6tRoqGu/GwHjl9dx+hi+txzzxXrq0ld6HrPdZh7gaJS1EceeSTZzKak13FNHozsUTVq91O7b+5pKnfhPGDf1eSCdMHVecTvVhkBr+Vc1+9avHhxsmt7Zo3B6LvW89WyZmkZzzBXX311sj/84Q9n13FNUhdoSofY1zUpVi1LCceBzme6Y9eyfJJaFhT2vWYdYtuo9Ke2XgwEPBtG5PfHe1NpU02yRNju+qycO7oXlu5R98VSNht1tyeU40Tkz0w5qq7tlLVzD+D36ThrEtr+JTmLUsvMyN8g7E9t/5IcPeLN+7WpU8uOOXfu3Kxs/vz5yb755puTTeluRC6Bu+SSS5J97733Ztfx/F+TTtGuSaV1/eHvl1KWxoj8LKv1t+R3Tcg6zN9c/B2ov524R7z//e/Pyih7okxs+vTp2XUMXcHsVJqNiu8X9B0F/6bEUaVYDGeidXAf5293HUsPPPBAskvZ+CLKa5M9bYwxxhhjjDHGGGMaiF/aGGOMMcYYY4wxxjQQv7QxxhhjjDHGGGOMaSA9imkzbNiwOP744yMiYvbs2cXrVKdLPe7atWuTrTEwGMdGtcRMl0c9qeq+mOabmjPVbzPOjup0X3nllWRTo6oxeHiPtdSpvHeNrcNn1jSh3/ve9yKinqKwabD9v/SlL2VlbAfqLrUPqbUupcVVqGXUOu+8885k12LYKPzuWmrrRYsWJbuW8rEJcAxzflDvGZGPe9XfUt/LPq2lr2O7aFp0zoFaalim/dO5yLHB6yLysaZp6DsR6pzZXueee252HddJhXp6purVOca4M9Qc65rJPtT5wXgev/3tb5NNfXBEPpZ0jDz77LPJrsVk6aS1sqRZ1vgQjAtz3HHHZWVclzhPVf/PduK+pWloa+stv4tjUGNGrFu3Ltm9jU0zGGm+W22k5xfOMU0jSt395MmTk71169bsOtapsZy4JnFd03YtxT3SfuJ36ZpcisOicRBqqagZV6mW4pZ16PmoFntlIJg3b172N++PaxnTs0ZETJo0Kdm6vvKZSjENI/L2rO2f7Efdg7k+0ta4KdwLd+7cmZVxndm4cWOyeTaOyPdufebBjGVTWyNq50vSbppkfW6uoZy/er5gvDiNRzZx4sS2vvtQo9QnuoZwrGscJqb8Zkw9HQucc1y/9WyiqbdJbQ0kLON3ReSxG3ke0/2G67fGbmyV9Vfqb62Xz63nCMYO5TPo7yOuk7rvtt41ROT9y/NFRMTdd9+dbMbHZL9H5OPlmmuuycqYFp7tqr9H9uzZk+w1a9ZkZXyWJUuWJPvpp5/OrtN3AMQxbYwxxhhjjDHGGGM6FL+0McYYY4wxxhhjjGkgPZJH/fe//01ptG699dasjKnT5syZk5XRZZ2unOqqS1mSusLR9YjuwJrGkO5MlACoaxfTc/7hD3/Iypj+kC6ldI2KyN2oVJJRkpeoyxPr0JSZrfvX9OeDDdtS7/mWW25JtvYN3cLY9zWXObq5qjsg21hdT++///5k33777d08xVvD+2Dqd5UorF+/PtlNSzOscCzSfV37gGUqQaErNee2unKyv2qpTSkJoEtkRMQxxxyTbEq4+P+IvK/U/ZBp+1RW1YmwP5j+kKlZI3LXbJWb0t10xYoVyVYpBPuQ30WZQETEfffdl2yV2nGMlKSnEfmar+OFaz7XjqbPt76AfaDPy7Zlu+icZVszrbS6NbOdVfrAOUZbJWlN26/a4Ygjjkgu8ZdddllWxjOFtiv3P66tmtab8gqVWpTSd9fSYtfSFpfOSnr/3Fv1uTj/dG/lPOV36ZmN41bLtM6BgOeWG264ISvjfsK9UFNtc+/X8wjlxqW5ElHuA117OZ5UYjVy5MhkM7WtynNYh94H1wHuFTqf+V2alveFF16IiDfLWQeDknyw3c9E5GOEeyYlKxERTz31VLKZ3lfPURxXGv5AJT1DmZpkR9cejlOuh/pbg2c5lf6xH/jdut9RlnTUUUclm79nI3IpjM5FjiHeuz4z19SrrroqK+N91WSRXM9HjBiRlbXms+49B0tr7dbn4d8f+MAHsrKzzjor2VxDtQ6e9/X3NfuGv6853yLyvmFKcV2fOZ8ff/zxrOy9731vsrk3aV+zDt0beI9btmxJtu59lN5t2rSpWEdJempPG2OMMcYYY4wxxpgG4pc2xhhjjDHGGGOMMQ3EL22MMcYYY4wxxhhjGkiPYtocOHAg6aw0nRbjSGjaQaa2o0ZPNVvUBqq+l7pH1qGxD0qphTXOAlNEX3755cUyfq+mQqxpXqkL5j3VdOqqy1Q9XVOgJp8xbCLyttu/f39WRr0etY06DljG9lcdPMtUx3nbbbclu5YqvF2orVXNt/Z9k+F8oY5WdZfUxuscYwpixlGpxcdg+2l/U8uqMW3Yx9Qc63dRX0p9fkSuV2VMgloKw4FGdd1E10Lqh+fPn59sxjaI+F+8gYg3j1H2Kdcx1Y3zb6Yq1jamvly/i2seYwGsWrUqu47rJNM4RuS6ZWqVa+vpUIHjVGMmcN5yvunY1vnd3WcUHZOcw6xP52KtT9pJaTkYDB8+PD72sY9FRMQHP/jBrKwW46S0z2vsLP6tsfzYB7WYQvybZwPta85n3Re5h3If175m/RpXoZQWXsdmrWwwYotxvDElcETEt771rWSz/XSvZ5nGCWO7cH3V2Aq8jrGmdB7xOr0Pxs9hH+vay9gcegZmHTy/67mT5zZNfdy6djDmcn9/J9c4/S7OF+7PukxXSzEAABIcSURBVFcfe+yxyWZsIK1/qNJO+ul2z+daF8e6ri+l31z6Xb/5zW+SzXGu8VVqcfRKseT0LHXaaacle8qUKVkZv4/rre4VpXEX8b/1vC9/jxx++OHpfrheROQxZ6688sriffIZdI3jmf4vf/lLVsbf26eeemqy9+7dm13H3yCl35gREdOmTUv23Llzs7LSvqi/OVnG9xoR+bMwXhjP4Xr/Wj/TmTumjTHGGGOMMcYYY0wH4Zc2xhhjjDHGGGOMMQ2kR/55XV1dydXzr3/9a1bG9HXq5jR58uRk0+WnJovR1IV0caNrrbqellIoqosvXdeuuOKKrIyuTfwc03hFRGzfvj3Z6jbKNKF0S9U6mAJ38+bNWVmrzibIpNg3CxcuTLbKWZiOTVPSlfpGXX/pTkdbUwTTBe/hhx/OylS+cbDwu3RsNsnN/62gex/7Q/tA25pwfnBu6zhVeVzpOq4D+hm6s5akXRG5i/jatWuzsnPOOSfZfP4myKNa91BzjafbZUTuKrphw4ZkU3oUEbF169Zka/+WXL91DWJ6d/aNpnUuuXJG5M9CCeiNN96YXbdr165k63paqr+T5l5voat3Lb1zySVZP1eSLUbk46LdOlQOpWO5VEeT2L9/f9pD1K2d8kpKBCPKcm91w69JKAjPNjxDRORyb85nndt0uda1g3OR96HrPffumiyZ6LrOs5mOiTlz5iSba9hAwXUtIpdTn3feecmm631EPRU625ZtofJitl8p/XfEm+cf4Tzl3qcSqFI6Yq2fa7u6869cuTLZO3bsyMpadeoYHAhq6bp78rkSbNda+uBZs2YlW8+8bBedA6NGjUp2U2WjB0trfOtaRmrn/9J5NSIfz7wuIpfJcG3T+2B/UbJEKY1+TuVH7P8xY8YkW38ncK5rH/MeuV7odToOSWvf0jAbB8OBAwfSd3L/icjvU8/cnB8c97o+8XfbXXfdlZXxPHvBBRck+9FHH82uo2SZ/XTCCSdk17HsoYceysoee+yxbu9X04vzPYeu6/xdxOdniu+I/Dcz98GIXCanMrAW9rQxxhhjjDHGGGOMaSB+aWOMMcYYY4wxxhjTQHocvrzlrqXRuukytmTJkqyMrkdE3T9rkgC6sdF9Sd3o6ZbEz9CtSevXMrqg0abLXUTuRq3ZEOiKRRdqdZumCx0jzUe8WWY2kKirKZ+BUdCfe+657Lonn3wy2ZRxRETs3r072XRd06j6jFJOl/NaprBf/epXWVlfSMrozki3R3UT1L9J09xemT2JqMs7XU/1vuned8wxxyRbM6twLrIOvY6omz7nEevT8cnP6brCMcT69LrByETUahddW4hmKaErP8e5ri18Hs1UUcuMQeg2zzbX9mdb1jIX8TlVKspnUTdzrvNck/WZOY5r7dFktP2436krOcc23XNr0ibWofONn2N9EXm7l6SuEfme1peu2v3JG2+8kfYy3dO473C9i4iYPXt2sk8++eRkM7NGRMSECROSrX3DdmWZSlE4F0tSNS3TuU1ZO889Kj9kBlDNmsn9jrIEzbjCcavZOlasWBFNYvXq1cnmGrVgwYLsOkpaVILNc4uejwnHE+e6jgvOMe1H9gHLdK+onXN5v5zrOmf5t8qXW9/dFxk6D5a+OF+xDs4/lQNzrHO8aFZdovKomvxtKHDkkUcmydF1112XlXGf1mzDXA95XlPpNqWLKjMpyY30HMTxzPOBymJYh54jWMZ5oPIozjfdR5iJiGuq/uahZE8lYaecckpERGzatCn6iq6urnRe0N853Cf1dzj3v7FjxyZb24TSoUmTJmVlzOI4c+bMZGtIDv7N+lVexDVO1zH+vuV41DMq5V1aB9dXSv3nzZuXXadtRZjNrMTQXjWMMcYYY4wxxhhjOhS/tDHGGGOMMcYYY4xpIH5pY4wxxhhjjDHGGNNAehzTpkVNP6q6e+rAarEPqOFVDWlv0kLWNKPt6l9rcRx4j7XUqaX6InJ9pOqCW+3Y17FQSik7+Qyqv50xY0ayqZvWe6ZWU+tg7AM+Uy22EdH6HnjggWSrRr4v4lewDsa0Ua1kTc/dhDg2hJpPzlMdE9Tka9wLTWHXQvuNf7MO6nIj8rmi8W5KsVd0jeHf1NBG5Bppxm5oUt/U7kXL+AztjnON8cQxy35SDTXrr8U5KtWnZYxrVZs3Gn+hXZoWQ6o31OKJ1WJ1cR2txcdgmfYB55veB8cG55vudYyBorEBmkxpvHAtfOmll7KyZcuWdWv3hHZTFZfin+j6XEsZW5qnup5yvOgawzp7O9/afeb+Qp+X6w1Tzeo8uuSSS4pl3Fu552i78O9a7BvWr/szY9UwHoamNK6ldWfKcq4xmpa3tt+0xkmnrrU1eM5hTKqIiK1btyabsVU0zgjnkaaIZyzH2prcqfznP/9J7XTTTTdlZRz3OrZLcdV0zWDsEY0hxjMmz+vbt2/PrmM8otJZR79b9zuut5xHGquTc1OfmfUz9o3G9uNvTj1Hb9y4MSIiXn/99RgIGONGf58z/hDbZPTo0dl1tRTaXIcZP0d/IzCeLONQ1WLCso0j8v2A81TjuU2dOjXZ2ofsD7bNz372s+y63/3ud8nWuGi1eJ8t7GljjDHGGGOMMcYY00D80sYYY4wxxhhjjDGmgfRaHlWj5k5L1B2KLmLqelRzvy+h9RN1jyUlN3N1W6y5qtP9r7dSnf5wIR42bFhKLT5//vysjO6GTA3a+lwLustq6jq6d9EtLiJi+vTpyaZLm7r0st8oiVL30l/84hfJ7osU3zX4/Jr6rpNSN7JP+EzaB0RduCkV45zQdlA5Wwt1IaYLorqDltIT19J6M/VuRO7OWnNHbyr9kaq6JHGgW3/tM7Wy2nXtrml9sWZ2qsu+7n1c92ryQVLbq2p9wOu0bt5XLeW0pgo3ddodp7yO86O3e19tHNRSUXNccN+g1EDvUeE99/fe3VNqMtyaTKIkZ9L5Uaqjdh7WlOyEZ7CatEbXFZ5jKFt95ZVXsuvYHlpH0/quL2F6dz17UuZB6ZSufZSn1VJ+sz6VTJD+OAv0FwcOHHhTCvoWHKc1CWdN6sPfKGvXrs3K+nPv1zHPucn9WUN89AZdo2vP1ZqbfS2va4252m/mmoSfbbJnz57suvXr1xfrWLVqVbffpe3fn3Ni9erV2d9Lly7tt+9ql875tWmMMcYYY4wxxhhzCOGXNsYYY4wxxhhjjDENxC9tjDHGGGOMMcYYYxpIv8S0UahVazeVZLu6vJoum/EramkXlVL8nIHWk/aHLvPAgQMpbsXpp5+elbFvNB4Ntak1HTZT72kZY5LUUn6znamZ13gbGnenP2E6T9WXl1KoNxGmCuU4r6WEVtivteuoA6ceVvv7jDPO6LZurZ/zWeMn1FJrcqy99tpr3f5/qNOTlOKDdR99TU/04E1C4yJw3dT5xrFe2z85P1ifrl2luDUR5fg5ep2mOjXNpLdrAvdnpmzVMdeUNacdSvej++K+ffuSrbHteC33McY10e/iuap2vtS1jG3Nz+l1pZTGEXmMlf379ydbY6q0m758qME2WbduXVa2bNmyZC9evDjZu3btyq7jGNG4gWx/HUtDga6urnRm0znAvUTHUGke9GQ9aXpsu97GE6uVtejr36mt9utJrELeWy3eYe33Q+n7mtifA4k9bYwxxhhjjDHGGGMaiF/aGGOMMcYYY4wxxjSQAZFHkZp7F926+sIFqpTWVss6yY33YBk2bFgcd9xxEfFmCRQlLOri/q9//avb+lSyVOvfdlPG0pWSsiRNt1ZLT9zXULLw5z//OSvr6xR7fYn2waRJk5JNmZe67rKvVEJB19aaK2bJfVPro6RO5xslUTW5I8eyyjj4nFu3bi3WYTqf0piLaPY8JUzhq2jaTUoyavtdKcVqu/LiiLxtOcf0ujFjxhTrNJ1PLSV2k6iN7Xappfzetm1bVka5C+fE8OHDs+tKe6ZKSPjdWsY9rfS9Efm5jXKfiIijjz662zpUqsP76KSU0wcLz5e6r5x00knJZop0lXfv2LEj2XqGZn9QJjeUJGit8VJK/d1fNL3N2v3N2SnzrS8kbgdz7aGEPW2MMcYYY4wxxhhjGohf2hhjjDHGGGOMMcY0kAGRR9GllNIIlTHU3NdLZaWMFm9VX9Oji/cXBw4cSK616nI7atSoZKsrJ7NE0H1Wr6u1Od1s2W8qzeF9Pfvss8n+5je/Way7P+AYoTyKLq8RzXZh1ExKdOuli/U73vGO7Lo9e/YkW93MS5mgVPZUGgvqcs46NDNXyTVb3W0pE9H76ER3U9M7OFY7RQ6lcB2OyLM9aWYprpUl+VJEPu45x2pyslqWPM43lamqRMCYgeKwww5ra/zpPlAa2zo/OCf07DN27Nhkc87qHsy5yX1Mz0Hct15//fWsjGvEhAkTkq3PxXvUPZNrB6U63Pv1PoY63D/OPPPMZKtkddq0aclmRkpdd5nhlGfoiIiXX3452ZTc1yR5h1JfmM7F47T/sKeNMcYYY4wxxhhjTAPxSxtjjDHGGGOMMcaYBuKXNsYYY4wxxhhjjDENZFBTflP3G5FrOTV1YakO1c4xjgH1vX2R/nEo8O9//zs2bNgQERFXX311VnbhhRcm++Mf/3hWduKJJyabcRW0/akH7+67W7BvmG4yItde33LLLcn+xz/+Uay7P6DOnSmlaymwa88/GGgMDGr02c5MrR4RsWrVqmTPnDkzK2NKX/a/pgrl/GYb1WLf1OLnsA90LPDvESNGZGUcd7t27QozdCmltu4kjjrqqOzvt7/97cl+9dVXszLOYa6pOsf4N9tIr+P6oGsZ5zc/pzEYGG9DY4I4ppTpT972trelfUJjyXA90P2D45R7kM7Fffv2JXvNmjVZGVN787veeOON7DrGQOHc0bnI+adnJKbrZgya7du3Z9cx3pSmHuc9co3ReX8oxVQpxTHUcwPPM5MnT062/qbhWUnXyaVLlxbLyFBvc2NM+9jTxhhjjDHGGGOMMaaB+KWNMcYYY4wxxhhjTAMZcHlUjXZT87KsXXdruxj+j1abbdmyJfv/bbfdluw77rgjK2MaTcpU1AWZLr6aRpquo3T31T6kBGDz5s3JHugUvvy+66+/PtlMzxjx5tS4TYJ9FVF2/1XX7J/85CfJfvDBB7OyiRMnJpupMOkKHJG7Y/O6d77zncV7VDkX75f27t27i9dp6lQ+G1OgqhTrUFojSqmeO12+MhT6cNu2bdnfXF8oi4jIxzrXV+1HtgslVbU5oN9Vkhhr6mOmttUUuDXZszEHS1dXV5IEqeSXe4vu2Ux/zbPJ1KlTs+seeeSRZC9evDgru++++7L76M7uCTVJf6lM1/Xly5cn++KLL87KPvShDyWbZ7VDeY5yDZ0xY0ayNQ0617WRI0cmW6VwTKWu52H2VV+MF2PM0MeeNsYYY4wxxhhjjDENxC9tjDHGGGOMMcYYYxqIX9oYY4wxxhhjjDHGNJDDeqKfPOywww5abEktrsbR6E2smkOFrq6uPslZ3hd9KPVlfw9VPS71xwcxNtd0dXXN7ov7abcfNebQrFmzkj1nzpxkM5VpRMTPf/7zZGssIbYF9f/a90xjyXHC+Ehan8LPMe6Ajjum3aRWPyJPsbpo0aJka1ycdsduU+diD7+72/8P1fnbDQM+F3tQX/Y3Y3NojKpjjz022e9+97uTPXbs2Oy6448/PtlMY6xzkbEbdu7cWSzbu3dvsjXNMGPyaIyHvmYozEXTt3OxNX90LeM+o7GWOA84/0aPHp1dx9TY69aty8p4LmjKOspn5v4ZkceZYwp0xqSKOLT2RbbXeeedl+zx48dn140bNy7ZU6ZMSfY///nP7DqePTT2F+MjrVy5MtmDHFOosfuiaZ+hMBdN93PRnjbGGGOMMcYYY4wxDcQvbYwxxhhjjDHGGGMaSE/lUa9ExPa3vND0NeO6urpG9EVF7sNBxf3Y+bgPhwbux87HfTg0cD92Pu7DoYH7sfNxHw4Nuu3HHr20McYYY4wxxhhjjDEDg+VRxhhjjDHGGGOMMQ3EL22MMcYYY4wxxhhjGohf2hhjjDHGGGOMMcY0EL+0McYYY4wxxhhjjGkgfmljjDHGGGOMMcYY00D80sYYY4wxxhhjjDGmgfiljTHGGGOMMcYYY0wD8UsbY4wxxhhjjDHGmAbilzbGGGOMMcYYY4wxDeT/AUIFjY07r3cfAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1440x288 with 20 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "id": "hHW8RZiz6tI4",
        "outputId": "9e5b0004-528c-4a5f-9a8a-e0119b584961"
      },
      "source": [
        "# Checking on testing data\n",
        "predictions= autoencoder.predict(X_test_noisy)\n",
        "display(X_test_noisy, predictions)"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x288 with 20 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8_YtyQdjF0y0"
      },
      "source": [
        "# Conclusion"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jUwwJt_fF23s"
      },
      "source": [
        "- We have first Loaded the data nad vizualized it.\n",
        "- Normalized the data and added salt and pepper noise.\n",
        "- Vizualized the noisy image.\n",
        "- Created out encoder and decoder architectures and created autoencoder model.\n",
        "- Our input data shape is ( 28 x 28 x 1). I have used 2 Conv2D layers and 2 MaxPool2D in encoder. 2 Conv2DTranspose layers and 1 Conv2D layer in decoder.\n",
        "- From model summary it can be seen that output shape for input and output is same (28, 28, 1)\n",
        "- Compiled the model with 'Adam' as optimizer and binary_crossentropy as loss function.\n",
        "- Trained the model with epochs= 100 and earlystopping (monitor= 'val_loss', min_delta= .001) callback.\n",
        "- Our model trained for 24 epochs and stopped by early stopping.\n",
        "- Created a function to see random 10 noisy images and autoencoder output of those noisy images.\n",
        "- Checked random images from testing and testing dataset and denoising versions of those images from autoencoder.\n",
        "- Autoencoder has done pretty good job."
      ]
    }
  ]
}