jonpsy/capstone-project.ipynb

## capstone-project.ipynb
{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "feedforward",
      "language": "python",
      "name": "feedforward"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.9"
    },
    "colab": {
      "name": "Capstone Project.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/NanuSai/3de0b30c9e8a84aa3323a595213838be/capstone-project.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rffwkDWTz_Fo",
        "colab_type": "text"
      },
      "source": [
        "# Capstone Project\n",
        "## Image classifier for the SVHN dataset\n",
        "### Instructions\n",
        "\n",
        "In this notebook, you will create a neural network that classifies real-world images digits. You will use concepts from throughout this course in building, training, testing, validating and saving your Tensorflow classifier model.\n",
        "\n",
        "This project is peer-assessed. Within this notebook you will find instructions in each section for how to complete the project. Pay close attention to the instructions as the peer review will be carried out according to a grading rubric that checks key parts of the project instructions. Feel free to add extra cells into the notebook as required.\n",
        "\n",
        "### How to submit\n",
        "\n",
        "When you have completed the Capstone project notebook, you will submit a pdf of the notebook for peer review. First ensure that the notebook has been fully executed from beginning to end, and all of the cell outputs are visible. This is important, as the grading rubric depends on the reviewer being able to view the outputs of your notebook. Save the notebook as a pdf (you could download the notebook with File -> Download .ipynb, open the notebook locally, and then File -> Download as -> PDF via LaTeX), and then submit this pdf for review.\n",
        "\n",
        "### Let's get started!\n",
        "\n",
        "We'll start by running some imports, and loading the dataset. For this project you are free to make further imports throughout the notebook as you wish. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O3vI8jSIz_Fs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "from scipy.io import loadmat\n",
        "import numpy as np\n",
        "import pandas as pd"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8OrHY7TRz_Fx",
        "colab_type": "text"
      },
      "source": [
        "For the capstone project, you will use the [SVHN dataset](http://ufldl.stanford.edu/housenumbers/). This is an image dataset of over 600,000 digit images in all, and is a harder dataset than MNIST as the numbers appear in the context of natural scene images. SVHN is obtained from house numbers in Google Street View images.\n",
        "\n",
        "* Y. Netzer, T. Wang, A. Coates, A. Bissacco, B. Wu and A. Y. Ng. \"Reading Digits in Natural Images with Unsupervised Feature Learning\". NIPS Workshop on Deep Learning and Unsupervised Feature Learning, 2011.\n",
        "\n",
        "The train and test datasets required for this project can be downloaded from [here](http://ufldl.stanford.edu/housenumbers/train.tar.gz) and [here](http://ufldl.stanford.edu/housenumbers/test.tar.gz). Once unzipped, you will have two files: `train_32x32.mat` and `test_32x32.mat`. You should store these files in Drive for use in this Colab notebook.\n",
        "\n",
        "Your goal is to develop an end-to-end workflow for building, training, validating, evaluating and saving a neural network that classifies a real-world image into one of ten classes."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r8BHW8P_2wxw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "035f5906-abaf-41cb-ae70-b5f13a52efd1"
      },
      "source": [
        "# Run this cell to connect to your Drive folder\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YWdiz3n_z_Fy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Load the dataset from your Drive folder\n",
        "\n",
        "train = loadmat('/content/gdrive/My Drive/data/train_32x32.mat')\n",
        "test = loadmat('/content/gdrive/My Drive/data/test_32x32.mat')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Sot1IcuZz_F2",
        "colab_type": "text"
      },
      "source": [
        "Both `train` and `test` are dictionaries with keys `X` and `y` for the input images and labels respectively."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_Q1n_Ai2z_F3",
        "colab_type": "text"
      },
      "source": [
        "## 1. Inspect and preprocess the dataset\n",
        "* Extract the training and testing images and labels separately from the train and test dictionaries loaded for you.\n",
        "* Select a random sample of images and corresponding labels from the dataset (at least 10), and display them in a figure.\n",
        "* Convert the training and test images to grayscale by taking the average across all colour channels for each pixel. _Hint: retain the channel dimension, which will now have size 1._\n",
        "* Select a random sample of the grayscale images and corresponding labels from the dataset (at least 10), and display them in a figure."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-WIH5hyXz_F4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "##Preprocessing\n",
        "\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "\n",
        "train_x = np.transpose(train['X'], (3,0,1,2)) / 255.\n",
        "test_x = np.transpose(test['X'], (3,0,1,2)) / 255.\n",
        " \n",
        "encoder = OneHotEncoder(sparse=False)\n",
        "\n",
        "train_y = encoder.fit_transform(train['y'] % 10)\n",
        "test_y = encoder.fit_transform(test['y'] % 10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zmGJK3xgz_F8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Show random images\n",
        "import random\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def random_imgs(x, y):\n",
        "    num_samples = x.shape[0] \n",
        "    indexs = random.sample(range(0, num_samples + 1), 15)\n",
        "    fig, axs = plt.subplots(nrows = 3, ncols = 5, figsize=(10,10))\n",
        "\n",
        "    for idx, ax in zip(indexs, axs.flatten()):\n",
        "        #usual case\n",
        "        if(x[idx].shape == (32,32,3)):\n",
        "            ax.imshow(x[idx])\n",
        "        else: #graymap\n",
        "            ax.imshow(x[idx,...,0],cmap='gray')\n",
        "        ax.set_xlabel(\"label: \" + str(np.argmax(y[idx])))\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6SR4gYffz_F_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 544
        },
        "outputId": "cf3320e6-974a-4f96-8efd-b6faed058bca"
      },
      "source": [
        "random_imgs(train_x, train_y)"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 15 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UXYwWhHpz_GD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Take mean and reshape\n",
        "\n",
        "#Weighted average\n",
        "def rgb2gray(img):\n",
        "        return np.dot(img, [0.2989, 0.5870, 0.1140])[...,np.newaxis]\n",
        "    \n",
        "train_x_gray = rgb2gray(train_x).astype(np.float32)\n",
        "test_x_gray = rgb2gray(test_x).astype(np.float32)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "auFZ63dtz_GH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 544
        },
        "outputId": "2522e90f-f528-4344-a6d5-82c5d5691afd"
      },
      "source": [
        "random_imgs(train_x_gray,  train_y)"
      ],
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 15 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7e7iSyWXz_GN",
        "colab_type": "text"
      },
      "source": [
        "## 2. MLP neural network classifier\n",
        "* Build an MLP classifier model using the Sequential API. Your model should use only Flatten and Dense layers, with the final layer having a 10-way softmax output. \n",
        "* You should design and build the model yourself. Feel free to experiment with different MLP architectures. _Hint: to achieve a reasonable accuracy you won't need to use more than 4 or 5 layers._\n",
        "* Print out the model summary (using the summary() method)\n",
        "* Compile and train the model (we recommend a maximum of 30 epochs), making use of both training and validation sets during the training run. \n",
        "* Your model should track at least one appropriate metric, and use at least two callbacks during training, one of which should be a ModelCheckpoint callback.\n",
        "* As a guide, you should aim to achieve a final categorical cross entropy training loss of less than 1.0 (the validation loss might be higher).\n",
        "* Plot the learning curves for loss vs epoch and accuracy vs epoch for both training and validation sets.\n",
        "* Compute and display the loss and accuracy of the trained model on the test set."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l14VCBFVz_GO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Flatten\n",
        "from tensorflow.keras.regularizers import l1_l2\n",
        "\n",
        "def get_mlp_model(input_shape):\n",
        "    \n",
        "    return Sequential([\n",
        "        Flatten(name='flatten_1', input_shape = input_shape),\n",
        "        Dense(128, activation = 'relu',name='dense_1'),\n",
        "        Dense(128, activation='relu',name='dense_2'),\n",
        "        Dense(128, activation='relu', name='dense_3'),\n",
        "        Dense(128, activation='relu', name='dense_4'),\n",
        "        Dense(64, activation='relu', name='dense_5'),\n",
        "        Dense(10, activation='softmax', name='output')\n",
        "    ])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rK_QkNaYdoW_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_test_results(model):\n",
        "    test_loss, test_accuracy = model.evaluate(test_x_gray, test_y, verbose = False)\n",
        "    print(f\"Test accuracy is {test_accuracy}\")\n",
        "    print(f\"Test loss is {test_loss}\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zhQWGAwSlyYf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot_models(history):\n",
        "    \n",
        "    \n",
        "    fig, axs = plt.subplots(nrows = 2, ncols = 1, figsize=(10,10))\n",
        "    axs[0].plot(history['val_loss'], label = 'validation loss', color = 'orange')\n",
        "    axs[0].plot(history['loss'], label = 'training loss', color = 'blue')\n",
        "    axs[0].set_ylabel('val loss')\n",
        "    axs[0].set_xlabel('epoch')\n",
        "    axs[0].legend()\n",
        "    axs[1].plot(history['val_accuracy'], label = 'validation accuracy',color = 'orange')\n",
        "    axs[1].plot(history['accuracy'], label = 'training accuracy',color = 'blue')\n",
        "    axs[1].set_ylabel('val accuracy')\n",
        "    axs[1].legend()\n",
        "    axs[1].set_xlabel('epoch')\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "beEZO1kvz_GR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "shape = train_x_gray[0].shape\n",
        "model = get_mlp_model(shape)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YxJXq3xYz_GU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "outputId": "23b27c74-df9d-4491-85d1-497f28dbb2c4"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_9\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "flatten_1 (Flatten)          (None, 1024)              0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 128)               131200    \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 128)               16512     \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 128)               16512     \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 128)               16512     \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 64)                8256      \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 189,642\n",
            "Trainable params: 189,642\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pPPbzGhVz_GW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint, ReduceLROnPlateau\n",
        "\n",
        "def get_callbacks(model_type):\n",
        "    \n",
        "    early_stop = EarlyStopping(monitor = 'val_loss', patience = 10, verbose = 1)\n",
        "    plateau = ReduceLROnPlateau(monitor='val_loss', patience = 10, factor = 0.2)\n",
        "    \n",
        "    checkp_path = 'model-' + model_type + '/checkpoint-epoch: {epoch:02d}'\n",
        "    checkp = ModelCheckpoint(filepath=checkp_path, monitor='val_accuracy',save_weights_only=True,\n",
        "                                save_best_only = True, verbose = 1)\n",
        "    \n",
        "    return early_stop, plateau, checkp"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7b5_8VsCz_GZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(\n",
        "    optimizer = 'sgd',\n",
        "    loss = 'categorical_crossentropy',\n",
        "    metrics = ['accuracy']\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b0kH6VYqz_Gc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "early_stop, plateau, checkp = get_callbacks('mlp')\n",
        "history_mlp= model.fit(train_x_gray, train_y, epochs = 30, callbacks= [early_stop, plateau, checkp], validation_split=0.15, verbose = 2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HyX1eZY8dGtw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "8ed432a5-d2b9-4c59-b2d5-248bbca8f634"
      },
      "source": [
        "get_test_results(model)"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test accuracy is 0.7932160496711731\n",
            "Test loss is 0.7185637950897217\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "To8uDyIqmral",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 606
        },
        "outputId": "c1d64b3d-f9e2-4e3c-fe13-96d74270eb0c"
      },
      "source": [
        "plot_models(history_mlp.history)"
      ],
      "execution_count": 110,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ncPtDtCLz_Gg",
        "colab_type": "text"
      },
      "source": [
        "## 3. CNN neural network classifier\n",
        "* Build a CNN classifier model using the Sequential API. Your model should use the Conv2D, MaxPool2D, BatchNormalization, Flatten, Dense and Dropout layers. The final layer should again have a 10-way softmax output. \n",
        "* You should design and build the model yourself. Feel free to experiment with different CNN architectures. _Hint: to achieve a reasonable accuracy you won't need to use more than 2 or 3 convolutional layers and 2 fully connected layers.)_\n",
        "* The CNN model should use fewer trainable parameters than your MLP model.\n",
        "* Compile and train the model (we recommend a maximum of 30 epochs), making use of both training and validation sets during the training run.\n",
        "* Your model should track at least one appropriate metric, and use at least two callbacks during training, one of which should be a ModelCheckpoint callback.\n",
        "* You should aim to beat the MLP model performance with fewer parameters!\n",
        "* Plot the learning curves for loss vs epoch and accuracy vs epoch for both training and validation sets.\n",
        "* Compute and display the loss and accuracy of the trained model on the test set."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yk2mH3Npz_Gh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.layers import Conv2D, MaxPool2D, BatchNormalization, Dropout\n",
        "\n",
        "def get_cnn_model(input_shape, keep_prob):\n",
        "    \n",
        "    return Sequential([\n",
        "        Conv2D(32, kernel_size = 3, \n",
        "               input_shape=input_shape, name='conv_1'\n",
        "               ,activation='relu'),\n",
        "        Conv2D(16, kernel_size = 3 , name='conv_2', activation='relu'),\n",
        "        MaxPool2D(pool_size=2, name='pool_1'),\n",
        "        Flatten(name='flatten_1'),\n",
        "        Dense(32, activation='relu', name = 'dense_1'),\n",
        "        Dropout(keep_prob),\n",
        "        Dense(32, activation='relu', name = 'dense_3'),\n",
        "        BatchNormalization(),\n",
        "        Dense(10, activation='softmax',name='output')\n",
        "    ])\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lbgRgZ5cz_Gn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "shape = train_x_gray[0].shape\n",
        "model = get_cnn_model(shape, 0.3)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YhL_OkAYeZkA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 469
        },
        "outputId": "843be48d-afc5-445a-f856-52ed355e083f"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 79,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_10\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv_1 (Conv2D)              (None, 30, 30, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv_2 (Conv2D)              (None, 28, 28, 16)        4624      \n",
            "_________________________________________________________________\n",
            "pool_1 (MaxPooling2D)        (None, 14, 14, 16)        0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 3136)              0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 32)                100384    \n",
            "_________________________________________________________________\n",
            "dropout_7 (Dropout)          (None, 32)                0         \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 32)                1056      \n",
            "_________________________________________________________________\n",
            "batch_normalization_7 (Batch (None, 32)                128       \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 10)                330       \n",
            "=================================================================\n",
            "Total params: 106,842\n",
            "Trainable params: 106,778\n",
            "Non-trainable params: 64\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nkmS2vV2z_Gs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(\n",
        "    optimizer = 'adam',\n",
        "    loss = 'categorical_crossentropy',\n",
        "    metrics = ['accuracy']\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VytQECDVz_Gv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "64cd605d-90a0-412a-a48a-82d60e0eb1a2"
      },
      "source": [
        "early_stop, plateau, checkp = get_callbacks('cnn')\n",
        "history_cnn = model.fit(train_x_gray, train_y, epochs = 20, callbacks= [early_stop, plateau, checkp], validation_split=0.15)"
      ],
      "execution_count": 81,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "1942/1946 [============================>.] - ETA: 0s - loss: 1.5307 - accuracy: 0.4750\n",
            "Epoch 00001: val_accuracy improved from -inf to 0.59969, saving model to model-cnn/checkpoint-epoch: 01\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 1.5306 - accuracy: 0.4751 - val_loss: 1.2025 - val_accuracy: 0.5997 - lr: 0.0010\n",
            "Epoch 2/20\n",
            "1942/1946 [============================>.] - ETA: 0s - loss: 1.2866 - accuracy: 0.5612\n",
            "Epoch 00002: val_accuracy improved from 0.59969 to 0.64710, saving model to model-cnn/checkpoint-epoch: 02\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 1.2866 - accuracy: 0.5612 - val_loss: 1.0597 - val_accuracy: 0.6471 - lr: 0.0010\n",
            "Epoch 3/20\n",
            "1944/1946 [============================>.] - ETA: 0s - loss: 1.2322 - accuracy: 0.5828\n",
            "Epoch 00003: val_accuracy improved from 0.64710 to 0.66967, saving model to model-cnn/checkpoint-epoch: 03\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 1.2324 - accuracy: 0.5828 - val_loss: 1.0268 - val_accuracy: 0.6697 - lr: 0.0010\n",
            "Epoch 4/20\n",
            "1941/1946 [============================>.] - ETA: 0s - loss: 1.0480 - accuracy: 0.6571\n",
            "Epoch 00004: val_accuracy improved from 0.66967 to 0.77823, saving model to model-cnn/checkpoint-epoch: 04\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 1.0478 - accuracy: 0.6571 - val_loss: 0.7265 - val_accuracy: 0.7782 - lr: 0.0010\n",
            "Epoch 5/20\n",
            "1942/1946 [============================>.] - ETA: 0s - loss: 0.8330 - accuracy: 0.7347\n",
            "Epoch 00005: val_accuracy improved from 0.77823 to 0.81973, saving model to model-cnn/checkpoint-epoch: 05\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.8330 - accuracy: 0.7347 - val_loss: 0.5892 - val_accuracy: 0.8197 - lr: 0.0010\n",
            "Epoch 6/20\n",
            "1939/1946 [============================>.] - ETA: 0s - loss: 0.7471 - accuracy: 0.7638\n",
            "Epoch 00006: val_accuracy improved from 0.81973 to 0.82755, saving model to model-cnn/checkpoint-epoch: 06\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.7469 - accuracy: 0.7639 - val_loss: 0.5551 - val_accuracy: 0.8276 - lr: 0.0010\n",
            "Epoch 7/20\n",
            "1944/1946 [============================>.] - ETA: 0s - loss: 0.7071 - accuracy: 0.7756\n",
            "Epoch 00007: val_accuracy improved from 0.82755 to 0.83138, saving model to model-cnn/checkpoint-epoch: 07\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.7069 - accuracy: 0.7756 - val_loss: 0.5499 - val_accuracy: 0.8314 - lr: 0.0010\n",
            "Epoch 8/20\n",
            "1939/1946 [============================>.] - ETA: 0s - loss: 0.6782 - accuracy: 0.7850\n",
            "Epoch 00008: val_accuracy improved from 0.83138 to 0.83493, saving model to model-cnn/checkpoint-epoch: 08\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.6780 - accuracy: 0.7850 - val_loss: 0.5310 - val_accuracy: 0.8349 - lr: 0.0010\n",
            "Epoch 9/20\n",
            "1943/1946 [============================>.] - ETA: 0s - loss: 0.6573 - accuracy: 0.7919\n",
            "Epoch 00009: val_accuracy improved from 0.83493 to 0.83793, saving model to model-cnn/checkpoint-epoch: 09\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.6575 - accuracy: 0.7918 - val_loss: 0.5250 - val_accuracy: 0.8379 - lr: 0.0010\n",
            "Epoch 10/20\n",
            "1946/1946 [==============================] - ETA: 0s - loss: 0.6438 - accuracy: 0.7969\n",
            "Epoch 00010: val_accuracy did not improve from 0.83793\n",
            "1946/1946 [==============================] - 16s 8ms/step - loss: 0.6438 - accuracy: 0.7969 - val_loss: 0.5248 - val_accuracy: 0.8365 - lr: 0.0010\n",
            "Epoch 11/20\n",
            "1942/1946 [============================>.] - ETA: 0s - loss: 0.6323 - accuracy: 0.8009\n",
            "Epoch 00011: val_accuracy improved from 0.83793 to 0.83875, saving model to model-cnn/checkpoint-epoch: 11\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.6329 - accuracy: 0.8008 - val_loss: 0.5211 - val_accuracy: 0.8387 - lr: 0.0010\n",
            "Epoch 12/20\n",
            "1944/1946 [============================>.] - ETA: 0s - loss: 0.6215 - accuracy: 0.8057\n",
            "Epoch 00012: val_accuracy improved from 0.83875 to 0.84721, saving model to model-cnn/checkpoint-epoch: 12\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.6213 - accuracy: 0.8057 - val_loss: 0.4995 - val_accuracy: 0.8472 - lr: 0.0010\n",
            "Epoch 13/20\n",
            "1940/1946 [============================>.] - ETA: 0s - loss: 0.6103 - accuracy: 0.8068\n",
            "Epoch 00013: val_accuracy did not improve from 0.84721\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.6106 - accuracy: 0.8069 - val_loss: 0.5013 - val_accuracy: 0.8462 - lr: 0.0010\n",
            "Epoch 14/20\n",
            "1945/1946 [============================>.] - ETA: 0s - loss: 0.5983 - accuracy: 0.8120\n",
            "Epoch 00014: val_accuracy improved from 0.84721 to 0.84730, saving model to model-cnn/checkpoint-epoch: 14\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5984 - accuracy: 0.8120 - val_loss: 0.4960 - val_accuracy: 0.8473 - lr: 0.0010\n",
            "Epoch 15/20\n",
            "1940/1946 [============================>.] - ETA: 0s - loss: 0.5922 - accuracy: 0.8142\n",
            "Epoch 00015: val_accuracy did not improve from 0.84730\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5923 - accuracy: 0.8141 - val_loss: 0.4995 - val_accuracy: 0.8449 - lr: 0.0010\n",
            "Epoch 16/20\n",
            "1945/1946 [============================>.] - ETA: 0s - loss: 0.5864 - accuracy: 0.8153\n",
            "Epoch 00016: val_accuracy did not improve from 0.84730\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5864 - accuracy: 0.8153 - val_loss: 0.5135 - val_accuracy: 0.8393 - lr: 0.0010\n",
            "Epoch 17/20\n",
            "1945/1946 [============================>.] - ETA: 0s - loss: 0.5727 - accuracy: 0.8225\n",
            "Epoch 00017: val_accuracy did not improve from 0.84730\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5728 - accuracy: 0.8224 - val_loss: 0.5074 - val_accuracy: 0.8445 - lr: 0.0010\n",
            "Epoch 18/20\n",
            "1940/1946 [============================>.] - ETA: 0s - loss: 0.5710 - accuracy: 0.8211\n",
            "Epoch 00018: val_accuracy did not improve from 0.84730\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5707 - accuracy: 0.8212 - val_loss: 0.5012 - val_accuracy: 0.8458 - lr: 0.0010\n",
            "Epoch 19/20\n",
            "1943/1946 [============================>.] - ETA: 0s - loss: 0.5697 - accuracy: 0.8213\n",
            "Epoch 00019: val_accuracy did not improve from 0.84730\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5696 - accuracy: 0.8213 - val_loss: 0.4941 - val_accuracy: 0.8434 - lr: 0.0010\n",
            "Epoch 20/20\n",
            "1946/1946 [==============================] - ETA: 0s - loss: 0.5600 - accuracy: 0.8250\n",
            "Epoch 00020: val_accuracy improved from 0.84730 to 0.85722, saving model to model-cnn/checkpoint-epoch: 20\n",
            "1946/1946 [==============================] - 15s 8ms/step - loss: 0.5600 - accuracy: 0.8250 - val_loss: 0.4654 - val_accuracy: 0.8572 - lr: 0.0010\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "60mJypwQz_Gx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "d366a79f-ea62-4fa7-80b6-e45d4ff43ac3"
      },
      "source": [
        "get_test_results(model)"
      ],
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test accuracy is 0.8370466828346252\n",
            "Test loss is 0.5398384928703308\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w2v80qosz_G0",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 606
        },
        "outputId": "1c5c17ee-1466-463c-c9b5-9fb9e5a9416a"
      },
      "source": [
        "plot_models(history_cnn.history)"
      ],
      "execution_count": 111,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3iBYFPWUz_G4",
        "colab_type": "text"
      },
      "source": [
        "## 4. Get model predictions\n",
        "* Load the best weights for the MLP and CNN models that you saved during the training run.\n",
        "* Randomly select 5 images and corresponding labels from the test set and display the images with their labels.\n",
        "* Alongside the image and label, show each model’s predictive distribution as a bar chart, and the final model prediction given by the label with maximum probability."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XMYYWs0oz_G5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_best_model_mlp(model):\n",
        "      checkp = tf.train.latest_checkpoint('model-mlp')\n",
        "      model.compile(\n",
        "          optimizer = 'adam',\n",
        "          loss = 'categorical_crossentropy',\n",
        "          metrics = ['accuracy']\n",
        "      )\n",
        "      model.load_weights(checkp)\n",
        "\n",
        "      return model\n",
        "\n",
        "def get_best_model_cnn(model):\n",
        "      checkp = tf.train.latest_checkpoint('model-cnn')\n",
        "      model.compile(\n",
        "          optimizer = 'adam',\n",
        "          loss = 'categorical_crossentropy',\n",
        "          metrics = ['accuracy']\n",
        "      )\n",
        "      model.load_weights(checkp)\n",
        "\n",
        "      return model\n",
        "  \n",
        "def random_img_from_model(model):\n",
        "        num_samples = test_x.shape[0]\n",
        "        indexs = random.sample(range(0, num_samples + 1), 6)\n",
        "\n",
        "        fig, axs = plt.subplots(1,5, figsize = (8, 8))\n",
        "\n",
        "        for idx, ax in zip(indexs, axs.flatten()):\n",
        "              ax.imshow(test_x[idx])\n",
        "              prediction = model.predict(test_x_gray[idx][np.newaxis, ...])\n",
        "              ax.set_xlabel(\"prediction:\" + str(np.argmax(prediction)))\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8yUPWbFCz_G8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model_mlp = get_best_model_mlp(get_mlp_model(train_x_gray[0].shape))\n",
        "model_cnn = get_best_model_cnn(get_cnn_model(train_x_gray[0].shape, 0.3))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W48syko0z_G-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 156
        },
        "outputId": "923310bd-a7ee-4af1-8f35-4f9184d99fa5"
      },
      "source": [
        "print(\"MLP Result: \")\n",
        "get_test_results(model_mlp)\n",
        "\n",
        "print('\\n')\n",
        "print('CNN Result: ')\n",
        "get_test_results(model_cnn)"
      ],
      "execution_count": 118,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MLP Result: \n",
            "Test accuracy is 0.7904502153396606\n",
            "Test loss is 0.7076961994171143\n",
            "\n",
            "\n",
            "CNN Result: \n",
            "Test accuracy is 0.8370466828346252\n",
            "Test loss is 0.5398384928703308\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J_qfLvK-z_HA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 142
        },
        "outputId": "73a6f31b-6d4d-415e-cb3c-daac312a3cab"
      },
      "source": [
        "random_img_from_model(model_mlp)"
      ],
      "execution_count": 161,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAB9CAYAAAB6W7C8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOy9a6ht25Ye9LXex5hzrb33OfeeE+699bpVV5KKxRVBoYhBf0TUgPinRE0wQowgFEjUiEW0UBSRgPVDfPwQ8ValsEpKoqCgSEBI/QmClEWKSEhCNCmrUqWVxHrec+5ea44xem/+aM8+5lxrr73XumcfcPVzxp5zjTkevbfeenu31omZ8dye23N7bs/tuT23z7aV992B5/bcnttze27P7f+P7ZkBP7fn9tye23N7bu+hPTPg5/bcnttze27P7T20Zwb83J7bc3tuz+25vYf2zICf23N7bs/tuT2399CeGfBze27P7bk9t+f2HtqjGDAR/eNE9NeI6K8T0Y8+Vaee293tGeafbXuG92fbnuH92bdnmL+/Ru+aB0xEFcD/AeAPAvhVAD8P4I8w8195uu49t9yeYf7Ztmd4f7btGd6ffXuG+fttj9GAfx+Av87Mv8jMC4A/A+CHnqZbz+2O9gzzz7Y9w/uzbc/w/uzbM8zfY5sece93A/iV9PevAvgH7rvh5atX/PFHH1/8jQEgaeOX9XIGAbB/KP9Efsl4Uv9mZjADDAZ3vVD+BzMDzPH9vGfyNCJ5bwGKfi9Fz1Hqw65Lw5nzk/e2X/3VX/l1Zv6S/vlWMH/1wRf449/1lUtDubvR5Q7SAx/ABldmMDN6Z/TWwZDv8gyZOyqEUgqI5HOA5cUe6BzqXNqzouvkn2SIQmPfz0cwvuxv/tJfe2d4A8AXv/CSv/PLHzseRV/veS8H6jMDXRAVvXd0w9vOgru8H8M5sJjTtSwwz/f4X7YedK7A5/hPIIEnFL6U1x3d0QMMnTSc8Lt2C/fXf+c3DeZvDe/DNPFxnu+75I5uvRmj37hU713OfO+fd/9wDtM3UcX7njz2Tzr86c3tO+P4R1/8In/Xd32H94QDlYD8/UJfUi8u/z6slVjjPeFmYDKN73vTS3dNyEPQoTuvufMxfHG8w7rXX37lb/5ShvfQHsOAH9SI6IcB/DAAfPTRR/jXf+TfsF+0k1BCLcRm33llkQoNBinS11pHwoA0sV0RgwncCdyBZdnQWkfbGtalgZnRtgbujHXdsK0NvXesTX7rYDB3KM0BETBNE0opOBxmHI8zSi04Hg+otWCaKmot3n8Cw+ZVGII+pNCdTM5Gm9uP/Mi/+svvDO+Pv4x/89/+z9LDA756cVqlxvzowsLl6DKx3qqMNF3Ye9eDsZwWbNuGm9e3+OSbr9Faw+n2hNY6SimotWCeJ7z84AWmacKrVy9xdXVErYQ6lRFEXAAU9M44nTaZp62htS59gDDxWednmiumuQq4S15GrOvWsMsMQPGyf+lf+ANvBW+BRcD8K1/6In7yP/4TgmtN8GvbmoDeB1V0TITeCMxA21iO1nF7K2O7eb1gWTYs64bb2xWtM5atyTqhArEeErjIOLo+fW0blnVF5451W3xtdcjYO4tA1HpD7w3btuF0WtF7R1ulzwpVVCqYpwNKKZinCbVOKAjGXEAotg5BQkS7ruGemHpm7JTEZwL+9P/0Z94Zx4/TjB/82u++k9Yyjd8bRIxnJvSM6BfWpAlxwarZmW7I2+x0aPdmsCkM2I0/kzadFcowxLj+THTMzClfodQq/ubdE5yIFYAI/8tf+ivvDO/v/I6v4L/+r34cjRlNh9W6CY5ymJjtB5OvucATJ+lgQHCSgbY1Edp7R9NjWRa03rGuK9rWMEDfQNsJ6DKv2mvlAQYUg6uu/EKoRRSow2FKdMJvH+dAubwrAFD8HpQCdoF53Vas2woA+Ff++D9/J7wfw4D/bwBfTX9/j54bGjN/A8A3AOCrX/3e83UyaAoh3ai+g6zFiKZETmxrrUoI4BoDM6NxBzdhBm0TQJ1OC9Z1w7puWG6D2PQmxHxbN/TOwYC5OwMuhUCFcDgIs23tgN431DqBiFBraHEyji4TXWiQtATbtLO4IKHuicD5qn4jzDO8v/drv/dcRtsTgt07LhIyBloXQgGVIZgISveDUBEJoe8Ny7LgdDrh5uYWr19/C1vrON2e0FtHrQWlFhzajOlQZS7aEb3PrhFnULTe0VvDtnXc3JzQWsO6NmytgVBQqIBKQZtlfjpPAM1ipZiqCj+6GhlgGijgncB+CLz3MP+B7/8e7qr524IMCT7NgRIcY8Dr2rGtHa11FxaXZcWyNMHP1tG7EDMwoZQJpYrAUeYZRAVUq8ChN2xtUwYsjLj1jsYdnTu21uS3dcHWNtCyYmtCSXsT/IUyKSY9dmxHGAa58Js4GYTFGcYzAuMzSrET4sfA+8Pr6z27Se84byGOKfJfWAN+HQfBBrGfMyFUbjVNiIYbVbVQ7V+F1gtQCDXKOhKwi74IUTeaGG/PT9kP4sK8XBb834qmfP3rP8DGVAcuC+BcCMj90t6npedMrSvNVmFVGLAoRFtr2NY1MeBNhc+iTxYiRFRAhSCYGoK1izVp7KQCTyGh62RK0UW6sLeVxFz2gsCfDqHzsnQwTfUBJpTHMeCfB/D9RPR3QSbsnwXwzz3s1ljEbv51Bkyj5EkJMc1USYRpmjHVqtK3ahFNJq23DcyiibVmxGwRgnZacauMYDvJb9vWdOI71iaMWBgwOwMutaC1hmmawMqc57lhniu4V8zTBJ7YGbDiljABZrBL/XzGAw0WssrunbV3gHmmMCNiMZ3jyF1v5y7Em1SLpyJXZ2YpQxCGs64rlmXB7emEm9tb1YBFG6u1qLDScLUeAMg8CVyT+Ti9u20N67rh9nSLTb+3rYOooFBFKQVgxjRNIAJqLeAijN4ENNN+AySBW3eM/u3hzXDm66bjnuechMGxMLimv7etY9sEF5dlRWtqmdHzrZmkLQSllIpaRSOdDkfR/A8HlFqF4SqzXZsKm705413WFY0byqmC1gWdCWVZwWhAaWJBsr6qvutalE10IugDw1A8Z3+CsZ0Lsreu6UfB+y2bT7st0HuoJPkYkFBEOUe6jfN1/vf5EW+7TOhHfMwv2Asxe+aL8fokKF3+e2hvDfNL9Es6oprujvnu++iaqV7YmdEbK702K1pTwVwsNL13bNuK1prSnKpM1IRBBhE7PZJzaquhcKM4RLL1tFwghLu5p8yclc4NM6XMmIrQ8VIL5kIXXJpje2cGzMwbEf3LAP5nABXATzLzX37Anfq5k4uVAZmGC5yjpwOShAmXUlTGCcYmWoUQ7NY61nUTLXddsC4r1nXFuix6Tn4TAtdd4mIW85wwYNG6S+/YagUzy7tVeJjnFdPUhfAXmQWiLn0ycy7h4njs+Zd+ezqY30dgLvy9Uwp9lpLGQ6W4K0D65dPnwpSdR7rGXA1Gy03oCn9xV4Kff4eYSHXultNJGPCyYduaM6JSxA0gWni9BL17/7zU3hXHs+Yb44v3Mid4NPm+mVZvwoVaZlZlvl0JnJmda50xH46YpgnH6xeoteJwfYVpntXcLO9d26pa7ybaRNtwWk7YekOpFWWpYBBOpwWggrI2sY7r+4LxZk0inXMlMiG5EatLqu8dhO4x8C5JI40+pJcnRZdVwB/kL52zoYsqrDndpfE32CsQepb8sddFk0Lha0aeO2rHYs73/pt2aIKjra0swHs/ii9A1vM+Tza2O8jAW8NcrYOtq7mZgd4TY81MFuP0m8xjMDUNOARWY75dXTjCeFktm+72cG1XY3CIUMwSpvwB0OvI6EkAwBivzaXFnmBnyift5CCi6x9G73ycrO9hky2Enr2pPcoHzMx/FsCffcwzDIuLqJrC7MynNRrxlfGJ1ltLRS1KaJkA7uAO9MbY1obltKpva0FrDTffusHpJFrwzc2taLtq6uMONXvwBQYMN4syM2qtaJtIZdM8ASBMkxCx3kWAmCY1bZD4GWyinNVSSEaZCd/DK98d5nut2jnl3a88Z8xmXkcETu0uDD8IYEqm/DAyX5FgBenN18Mckm/Ag1QKZpxOJ3z66Wus64pPP3mNbduwJAY8zwfUWlEKgfmAaa47LST6eokn3A++t4M3Y/SFjyZoFi2BoeZkYNvk++m04vZ2xbo2vH59QmuMZREXSueCzkXdHTOIKg7HF7i+vsZ8POKDDz/ENM948cErHK+vRFhS6XtVU/SyqfC5rfj05lvYtg3f/OSbuLl9jTp9C8vSUNcVbe1g3tS/BaAUSNRhGTVfGvwPMfgsxVE6eob83TPwLjSllmGmE4NK79S+Za3F1uNes+T4GgwYCBzSz6IcldI5Y4JhLDBGINeT90kP5WKFhIGYAgETSDvDfKiA/Ma9A2aJSi6wpkhl6xUgX08ZPvv2NjBnAOumMQUqpBkD7igu7OhzhxkXuJJxqCR4x/oPbXdzBmx0QUzGFUQFVfGRdOyVjB8IjzBGnBltbq79wuZF6RR6Us7ZXPSj9sz7Bwo2yRgLWOHua+ie9m0Pwto3Y2pnLZkEIqgq5FZb+CqHp+dBpTKIpKRRt601DbrasPnfDd3M1Gliu0mjbioaAWd99qheYySFsLUNINHStlpRp4LCFYXTvckUE51Oz0dM5/0W6LdrGckyodmbeM//yudk8ZPeOPr9TIq/4/27wew13rNI3SRShoTJaLog5VjVZSAMuFYOgc3ntPvzz2TwAIeP7uzkY9te7HcqtIdHPgK3XDCxYEIYMSmu8U/TjOlwwOFwwOF4xHw44Pr6BY4vruTaSqpZr+jMmNcFy7aK+RniTzstJ7TesMyrulZYCFzpMR+u7SJ9z/N/z9j1lgBx/uPca/kuLXD8nFMaI5Rh7OdZiW7u5tD33fPS0M/AgeD15tbx9UIqu4AG6570S7QuABrMJjEk5gIgtzCkNZZNKRfW8F60Fub7RHitr3X3in8aMBg84PhuXoypJdp49xH0OccHZRMzyIQg0vgy05BlBlxmvIBpcnvMhXXX5cd07v4WUqbImfp3f8i974EBAxFQAATQKCGM2NExTFTQTiMsIambyeKkATo3Nze4eX0rGvCtnLu9OWFdhXg31XJldQC1FBAKGIzSQ/oME0JoZcyMbRN/ZWsbALH399axLiuOxxn08loiowsD1SlAfO6lMRc4npD7nrW8GIypvul95B8S9MCu2ZNi6sBrkmAvz1btSS4VIWhLGnBrquWKf731DZ1rPJtFWFqWE7518y2sy4pvvRbtbV3kWXWa0Zq4AK6uDiiV0NqMzs3NQXeMCpkZPGUj/0/hpuMh1Up6F7eToIJEzHMvwnC7asedwv1RZpQ6o9QJ19cvUacZrz78EC8+eIWr62t89PHHmI8HfOGjL+L65QtQKahTAQMSjNU7Tqsw32Vd8Dvf/CaWdcHheIVPPv0U03TA7Ul89tvWQaWgbx2NWrh6BlM0nCBmrhW8oQ/nqagW6DC/JBy9K7BJLB4u1OV+RMBS/K7EkoDcIyhD3PtzVY9UvIfTW6HrahZ1AVUdKAoqYxSmoVZPtfO3+qIhJhCTxDto5gYKo3cTMqKf8QXO5OOi9NynlOa916qtsqUVEhhVfwt64UKLnUsWPzssZkficCT2xjTg1oz5WpR4MNNSCqqZnquYoyvpOVDKPsnMNfEXH4zgYestLrF581kdcTlTj+yOscCvAll3tpbf1N4DA5aWQxIMiXIeJwigTqFB7qQTi5yToCvRSNdNwtQX9Rdu64bltLgDX8wafVisYSJSG76a7konMBdk6ctaaFkFtAhCTLUCulCvmgQEuJkZCIafJMDBL/FtZcLnhG7/nnN5Os66cKTBCg9b1+eaMrtmV5J0q3OYJF7rr/3e2oZlER++MIkN2xKCVFGfdOsbep88gM7nzRhgrLwLo70MgbdtITibtUbnlgmgAg8yc36Wx6tHN/opwWqlVkzTjDpNOB7Ez3t9fY0XL17i+sU1Xn3wAQ7HIz78wod48eolqBbUSYhiMGBJTTqdTmAAJw2Q68xY1hXHqytlZjO21gC2dCTo+pQOpyB+iLkzoOdf0nnXSooJImLio3OUfDd4qzA3WBOUD1GCazRjsBo5onSFM2NN19v8BUO1k/G7PTfzQeuXuaGEEYd7Te5NQkgXmHSQWOszE78st+/eHwDxpyb8P5und20Md61EbANHQKcDVQUdhJhuEeN2r5me7Rgtkz0934ZBLghlFx+V7P8ln4MEEoXnSJPAUBO/aQ3BvI0F54fxAM+Ih6CEFATFxeJL4d72nhjw3hwE50t7s6WLfJyUYlYNv6jPd+voW8e2NNWOJK9Rgnck4EpSOzZ3+ANwP8JUJaUpeiNm7M4sEqkVkmiBdPKcDmBDqQXLsoKIMM2TBBp1cv9fYWCgXGlCs0R2Pvanaplqvu1949fcw4tPVESXwhrjYYvGVqohKCfqyZqeY1qjE8wdUYpUENWoBrOYRh4Xvf9SuPeg/T4t3EXrFSGskBAnCVMQnUvwxw6L1jdiZNaXWPAEiXmY6oT5MGM+HHC8usKL62tcXb/A9YsXOFwdcPXiGsfroxAkzUmfuKIzo24Nh+2A+TBj6w3zafFYiNNpwdXVFQDCfDigtQ7iFVAGbNprEBp27TIz3j1uUEKabLl+MmagfTocxHzuBUyMlqjwQSqwg1PwE4DwY4999wh/hH/Qi+8k3HPfLvYpeWLhsPTFqtkaVmxmEASNgTXpeC8EdKE72FQ7JnJFrJeCju4LgfLnXRA1oeQJ4K2PC0FR4XdpGWVhtCPwxWiwMF1xMZkwnTMIAISpmKDBn+R1BMg0YLJcdEq4FaM1JU6EX/P3Gr1hMMx3bhH/Oo9MYI1q3gOPE4azWrGEDpEDm3cup0vts/cB5z/2YvAZhuSZFSzqttCZQZ3QNgm6aq1hOW1YtxWn04LbW9GUbm9OKYRdCzcQABRMk2i+QtSkmo5pwhYA1JrmCDNjXVYnkq3JwpbCEiHZTtOE1ptr78IMgE4dxOS1H/bM9tvFemNh3M+E3/T+4Xc++yLXmARaggHXWlGr5PoGASK/Pft82E2XIzMFaTDcXgvxJW2MNphv7x2lk4ug+/EJWL4dTFjMUYWk31xkITKLfygshNrPlrSATY9m5j01xxOh1Io6TThcHXA8XuHFixd4+eoVrl++wKsPXmG+OuDFqxe4fnUtvVBR3gRWKWrAWJYVDJI87WUBA1jXhhfXv41CFa+Pr8Edag4VhiU8xwCf4Me8xwAQJ1jaBwEeFOGXki7/x7GFQoTjcR7Mz02JeNkYrQFbF627wzQewMuW6NpwuRDkzNb9icpMXXlV03OtRQNCC2qxuVKfbgmfr0XnZxM0qX/E8KFvHaxxK+gdvUGZgsSoEEm0ca/sfTaFxW0tyQQ9BHcamJ8AxRlw8+q5ZZbi666JPqkCkgudmp+uaaCRCZEFbxroitEWK8TkdQN24h87bpmyQ8PzctwJu604mKrTmJ7S6M7Gm5g8JIXSqrvkgiD3tc+cATsboCCB8EUbQJeWpRgFMWfCwu6L8ACWxqFFNCm00ZSZRgBYRFuXUjFNkcJiDLi1otKzEE3qHa00135z70KBS9Jb8nWwSVM5p3mgTgk+T6gFs0n7d9G4/QIdf3zgO8aFnhdMFNXICyjODQTd4Idzkz+BnCCSEUztYk63IP07CwB3jcMqeQXuXQTCOzWLOo0AmPCDA6yaLyfp39woKXq6mxujD/AgJbSkuelFtd1aqqdmxJDNzCr+5wIehKJpkmAuCerSKle1ohQ9zC3Do3orzPfchiBzYsIygpDBtIQMYz4XwN8J2ECpJRhwD0bsOKcdNIYflCaXtoXjl/neMw7XmqqqKerWkhhwLWn8wYAlK2Ly68mK8xgcVFvq1DxupxWxX5Yibi6oEG+BRoWGDOVoRoiyydThT0+nAtvr0nfjRxmeuS+mRebALVvro1JmDw6hyObFrGhndCWZgJHN4knAOectdm4vGKY+5yu17xdhkBgBM/knez7f3e29mKDJcSEzWCNVxQdqYdyhNZDPDzcB3rZ2bGvz1BQpALFiPVne6BJSFctinSYp3HB1PGCqFYerI45XV2rSEJCEb0L8j2KGBmjVhaKO+yAn0n9LKm8UWjK4OMGXbL9dENRgRnpPjSKQ4LxxmgNjXsBu6TjBAgh1qqi9K5EXE+E0TWBN1TJGvF8EEvW8uXGHVPuYakWrpkXnz6xty7sO84yjVi2rNQWpaK8HJvAGy8C7tFKrmAzFqQcuZlIW7XbbotrV7e2CtjXcnhZ1m3QsJ7G4kNo+qTbMvaFwBwtAUCcxR0+HWf3DFQUloi93skchgCphmgoOx4PEKlxdY1k2XF/f4OrqJcAFV4cr9I2BBvRV02B6cwJkZdSJUyGCkIaUYBoz6grl0DLMr0Zj9965USEcr+bIaOiMYtWUmIAmwTTNI4tt/mMNmtDtwVLqQxftdXIcq1rUpVTy6wvBmTAp3AlyTdHgs9k1YHIG7EtdHb5tWbEtUhKXtCJUIfLgJOphnm1NTLrN56BjrLCn1OiMnjxNhEmHlTW1Z9psZqHWtGVL64m0qt7Uh9zZ8/EiBqYojrCfdx/6pEI1kaYjhQWAlWnLu7TKXHLl5HgcKPxJ8cfmdYDXTgl0FwcSI7Z3JuuLAYbZtODPmQacrVMhOAQp32tAaikYmC9cyoAvuuFI2oR8tjMtTSLptCbxVDHPk56flBk2lCIIINHQDbVU9MqgLbS3nYCVpLzRF8mX1sMl8DwxE+b04lCM3vIdLsRGOpY/H5c04ICx+cEs2G2UXnevUcksS+523iqgDfcmcTunIBQqnhfs1F57ez42080eEhX+5hY+OekbqVhp5QSlNnlOZ+uuBQ/BKAwQSQjsUNRDCYVZEwzG4qe9q/9BIsMXWdysXTUGotbq5S0t7QnUFTIhDNOwXpNWbnhG+Y3s/wXJToTwkTA3UzAcLzt6sbQegGUIKvSrVUE7His4KiiZ1mrM2NxUIsxpxG1NGhkBVZkERF4CoMxZnzUrLuYyqzZq0midrTOsuHJVjb7pPQwGcde0GjGF75YITByO1LFxne707ke1zMxyH0QLFnqzD0AdD/8BOfXQn2EaGpluYmtegq3MFzVENGMXiJf4gvzdve+M5NtHbKiTRfMBtsmX6+NKnyJo6KcUG08M+P723qKgo3FakXttyK/QTxJfFEcVltbYfWfb2tIReaNS/rA60CddTPNhwjxNOF4dcH1lwSuSzG01SddVtNdam/rPpGoRbcU7aZqKJy3smW9ndebjjsT+yxP1JHrZpWfvV859P7i2m02hPNzhwWTK/FAYtVS00sXEXyq4dteAAdZASVtE8QkA3LoWGRAidzhI1C+VitevbwFI3rUQHYuYbHqIfM7oYCrIpUxHzde+0sijn6C5vys1j5El+d67FLHfNqt41bA2qfm8tYZumixV1G2TWITWU9SmCSUR5BZUHaMAM3ZOiVowWVJXTCmTlvWUqm5eB5qN8RqRDYZsQlFNBVpKycKy5tnfmZPxOCwnAqapoHQVVEiIXyeAmggLRAyC1Gf3qtYWaFUIZVJmO0sxh2maMM+y2crhMMu5uarlzDYLIY9JKDCYSs4vKCKezXrjlhoz61v/lUgvNyecSsGyFqzLAqIorStgtPQ1cWUNdCQzAd4TGYPy0wiY8bz8je1/F9CN9xhjigpXYlExDTi77IQVhEANh6lqvVAaq2uD0yrraqbvrJvsMPtmDkjaqwXq2XyRFgSiQphY0vdKCs5z4kBhxewWU6BlZLtp9yxBweYff0AhrPfNgBNBTGKxS3EwDZj8stCIQ8pp6vfdWtTT3TZlxFrjWUxH1QlEnQrmWTTf4+GAq6uD5E+qCdp225GKTVJzWAKshLnYxguuXZLZ/LIktgsmslDEvA4UMc5C5HeXPWmjfSe0K/uzuf6pAn5M7Tm/kwqh9CA21bU0DcaqyjDN/Kcwc42JBbGLPosImOcZV9fXYBDmeRotGizmzUhdUANZSu8ZifxOziWTfemJJB4HHozsMSCaNsWuMKINw3F23bozX6nZHP2T8x1T1x3DOKKTwwT/UBeGUR54EXpjwFSKugaSMKmsKwhkHp9pZOZ3D2YMmHDWUZg0qOi8N48ux0FwZtd7R4NVZpII9GZMl3QmrFSkwqBUcs12ngVH53nyjVeOxyMmtZDNc+y25QGBPgKJFylVzlu9c/O5BwO2WBbrvuBdBUnUOQG3cxVNuBQ066d/et3/KHto1hVjSIYjO3x4KibMaQkxMjlJFET7EylHukZTASRmG0DGrdE6Nvp5y/j8xC/Mmi0BtxpAa4w+W5Cytq1KALP690msn1wYLrvrOjYlK6/fKJwTecuy+U/EeLypvccoaM6avSCTA12vUP7G+UaO32yrQSte0Bv7Yaa9s2oqCnQ3xak2nE1N8nxZDL3Lud675pvtTKcJ5/aBRUHUdypJbk9scj5vl5jtfcx3jzR75vWGR5lAUdQPPFVM8wQGY6oVfepoDeituQWjlO6mIiDSN6yIhRDFGW1rmOdZLRonmA9ZItZJtcRNq55tACzvVqU77yunebFPenKJh9UvN5rG4Au0JeGx9UjL6CrJg7oKJOHfjCpf1vVMoHbzEbQq+uTDD7/Z5SP32Sm96xxuRk24bwzYU83UB8b6O/yvELke2wjif+1SucKtMS4U1ILSRZAGQ4QBt1qpn7aSCt8SH2IMt9aCw1E2fTkcZt22zhgwoXr1JS2sAoZVX8ruF4uQdiHJKDgURsxYpzBZX4xNMwE4/enmT8Qzc1CkPyah/t4y87aN878c8+q02vpk33cMOOPwwIhtPl2hoeH7AAbA7/Fdx3Q99d49a8U04PDfRn/skeZG2KeImRJoq+GMKuq6NCvFtgqzb2tPfXozPN8TA75MyEWqTzSEAuBASF6i6qsvrQHbxthWCchaVy1qr9qvhbfXWsBVHlCqLMx5rjjMFYfDhKujbOlWqmwjVVbC1sSjvq4TmIGpFLRSUFDcFG4exEy8nLhloSH/PQ7729ze5g3Ghu/SHMfmZs+0+G1hTKoBtO2A6+srTFPFcloBIt3tR0otbrcPWAMAACAASURBVGsDGM58ANLNFciJ5OFQUKvsSfvy9UtM06R1vRcwS7lFRsdpuQUK43iasSxHjey1Eo4AFSMKXfFN5osyJ3mKluY7m6Nkxxd4INa2dixrw7I03XZQXSnN9sYWjXlSPJ6mpkSlCSHNeVkXLCiXZo2hcSLGhKGCAhWAKiCx0m7y5A6wbhhRxLYQa5SSf7+E8GrRwu4H4wJLxfLguieCNxHhME8idJUOIpZ0nU6YegEVBlDQuICY0YnVLK0+QA1MK7XgcJwwzTOOR3FJ1WnCixfXsuHFIQL7bP/YWYP8PEALUOsLBkG9uHCU/JBdo8K6akzLiqUWNItk19/cwpO0OH1IMLv0qT8lXAhm/JS0xukys5jEEUpLFtyi1rtUruv6PQTKkflaOue+bGe8l4d7t02sXnlDnXVd5TdVwkQT706v9wy4tdktmsyMPkGzAMKak5tpvrZV4rY1nE6L7C2/bG6ONrn1vvb+TNBJG7FJu+9Cc2gPLIF3k+0Eb9Q6LnG/oFuRLmDfAdHgSkrcHmlbPO9Sv/cMWdVuWVhZ6/+2a7/A2y69yxIyj49Q5mDWhP0tZmkoLlkWZDOpMTwjGt3ni1VbGq0JtjguPcsXjW6/lyVuK2Mn/c/SbfSV6PL3d26DwMj+GQQpjqZV3HLqUc/wcHE1CTd8YYYUxUyx8nEkJevubhqcbX2N68yVX47e+Dvz/UkT3vfNN0ci8x3bCnk8wFWEkmcbmnpf4F/kd457UrdzmpybiqsFXqkVZ5owTVW1ZGXAU3FT81TNB2+lE5MZNfXW5xJwhgBKpS4BJ1zDvJvmAZsLO2f3BFXK70M6J8FcjwZ5EgIAVribCTrJB4MGPMTFpDF5wOUZXTxnwPnZBg7bdrartmsBuJJxkDVgEX56ikUwjb21XEueUXrMkQkzg7CTaIwHUG62f7HssBcWpvvbe0pDilUZfQxic5fMxkxidubQJnJOJTeLgo7dZlxDs8VQoAEjEihRJzlK1VxTTTEoTTIEnA8QzoBuhJWItNIONI1s3BpLauFS1MRNproxku/b0eLp7yYFx7x49IMTFvmPc9FltvPyRyHxh/Vmmqj5cVTDs4WkmoCcTjWkM6swt4FH6RKs5Cf3WBSdwx88VkU/Zw7jUJ9uJmxcvtNLY02Ta7g5rbi5PWFZN9yeFiyLpdBZZSDDXRm7bS/YVANl/Tv+C0j1uM3PDUIrlMmqT0tShnQHF0ihm84SINY8wMQIZZRb1Nkbnivv5nTetpKxDTNyuom1x8Jc/LsxUmNUmfBnU+cIETLlPwmLe1O0+IRFM548MIvcTG20RNqurvAgrBNgDINUyFI3mjCD7kR83Vasq+zeJgFF8GAfDzN0phAU00d2F/F/SqHfpAgbi/7NiGIotrvcZhvhKGMUeCSRwXiCbz+dXStJomSAEP5kq+2/rg2rlhpe1wW2s13UrdbALDXnWD7xNEn+fOeOMhVgkxHUTQV8jVy399ne8r133N7K/gLruuLm5qRb3Ip1T+j8pW1Rx/bZMuBB2JZlbJJZiOvKxfLCHMLAsSPaSiC8iH1ISI6ELumatKxmtJQsb/Z/l1pVogNZf5JE2sO0Ii3vhxYPGJBoh1SfWcAVRsZ7QSbxX9jNsXz2i/0bRRny3YN6PNxr/vZLaUf2TLvL4HlJJPGFaEJNguGYGsAwTu6awh1BZ2ddf4JmUvOg/SqObrrH77ZtWNYt3CUa8Bd+rLF/whQTQ7F3ACP80neb4N0yUBR1vRG2l6wxYEaUcuwczN61mx28LC0KyALvjiEiFbE4f8QTtIhuDoERyos5JiV+hgFo1IIRGrAXLJHPKR2z1hEwF4cw4NQdlYAC34UoWZlCWWI0zGFYc8TFIEKbblai1j2Zl3M8kLkJddv5YuZbJjc/RC17GMjj2buTjjd2mP+Xwz/KQz8CQ7NAZJaNQYlIQBNFrHnsRzBjMQOvraV0P4uM3sBg9f0XMCYc+gEgiTo3XmABuE3fzuk5sWXi6nUnrFbEukjWjQSdvhnW78cEnZAjM4S8aJmHOYb5kOIIX4BpwaaRZh5AujSdaFsRiJTCMXTNTBWGNCa5+VaGTQoTOJUMQcL8PpKLCtduA5GC5TiT2dm3bSE+ZRu33Hvrm52m7hm2XzCuwuFm49Wjxi+EyIlQKmtn85uJaX6f67KDMBPXGNMdzM9B5nDe0UfA5o4mmi88yGpTv64sVKlTLhuGWJyClZ80OFh/JLe3uEUl78nrIiFCCx4Q/7xf+Q9nnIqhzpRDKyYq4t8jgsSlh3bry8w1zOKwL8iwj2Ic3q1kQXlsY2as66I51A1bl01ZTJO0bUitGl7vDR3sBTSyVmyWG4JBXgoDlXSekYOHui5fHafW/zZGLvnCtNv5CrDSqiHUA63ZVpu2acyKrct4mOEM2HAkMzHdgylBeAfXQVZ+PMxJQ4RzwJfPsDHfnjTPzmfRzy5sJ3OzB+/B1neizQx/jjxLaHA25xuCu0tiWO/KV2AuKwA9NvHp3KXAiipbRa2WVu3MGHAOvMrVFd2VqXEQD3Uvvh8TdPpmBQoGFszJpGGfIEdWIXBZY+iuOVjot5lHPc80Fy6o+bO6eSwICqP3TZnuhrZt6E0O21OYm5maNHUDTr6UAWfNL/piTfVNBYNqdgaVJ2fCFzTKB912oQ904Y+E42d994UVxN4O9qLl7EzYtceBL9qcmtaSj0hNkOUVOyoxsip5rpV9O5gvEOZbS41bNTXudFpxe7vg5mbB65tFzi1NAq+2HqZnZ4pWHrWCtDykCZLG2qREDKM6ixuHdFkuCmuCqxzGeFNuMJUqKFAKqPcQitK021q1VD8ZQFM6r5aIs17Q8PGY1rnjdHsbBU16x7KuGoyz6V7SEYy5WQCb4k73/rGvYfuvoHhhlxwYtbUGAtC7bkvahWFKvQAtvjFLidtSCuY5bUeo23qWosxUDcqtrViWE5blhHU9YVlXrNvqDLgZ3dPANqeNPs+Z8er3PXo/AbobjAx9cjNcsM0VpCpgBGHlaH4gmFRYIExgCReT9duE6q5BiL1FeWGrtuV9JJLtHc26kX4dfdHy8GVZdDMeofutTejcxMpBEowVfKG7BtxTbp1YeLRG9Z0VBc/be2DAWfU1kUUAFv7fzKj0kxmRDxwOcbjWGxybeb/sjaDlCj8E3/0ixPkwn7jEw4P0gyTFGSLKG0af7qipYVwbuPCd+UwTftq2W306Xk6M37oBlc4f1oOQgM/PjX9fMr37tJkgljThbA4PiRnnz0rw5YQDIZIPg96933yIwDmjeLcWQmIEW5mG1tQvZZ8ev4C99hsCRmgHe9gplmeT3l34tW+BoGf3DAJOikbPxDfHcDCzp50QWKsijgKPw5boDqng3RozxO+WAnEsCGcfObx/7WiNohAWE0KN1jatUKaTtW2LCOp9w9akeMY8i0maeQYwqQl7ckubD9/ojUYHr2rOXLdNfJdDDivcIijMl3fjyI2GubV3+Sp9AtiLkG3omOGUGJwf3RlumM7ZhXXvcsK5rJeEQB7P80HYrlTF0kqFCXZNLeq6TzEIKHp/YUkpNRrHCOGx96IR9ClNqqj1JwHOtF2rWFYrpyJDpP79z2keMJBZgQAM0MlkWwzFgS83sE+EI2NvGvlq0lCKJtXFJ1KTSWumRUyoZUIts1T9oQpwcdOOIf66bFjXzT+3taFvuqVh28C9AaQar0q0U7WKQKJUeF5ZVjZ0SANt5LRCnlwjMyDuiDcZrC9cq/0ZZYKRSo8M0KRtwrgLtWqsGK0BoGCovVu+dQRQwYQtg4trtJHj6aX9Svc3xaLH2XGJ6JsJTcb5dMzXzM5W4crMzTe3C25uTri5WXFzs6iWZsVkhLhmRmDWGdkgIQL7DH7dNYKOyh3B5B7Y2cxsNPBN1klB0braveuc9g7iBmJCkdIcAjMThrvockSMKFfghXHBCIuR488TBL211vDNTz4Jv3WPKPKmfbJPwx+TC7hBazcLHfBKS4mpSFAnYzltGunasK4ntN5we/stLMtJGPB2QinA1YsDpqni5ctrvHz5QkvcXov/WMvfSlBdBzrj5vYGp9sF3/zkm/it3/ktnE4rXt/caqxAR3MLQ4RzBTqbfufbNPnx7cuwILGIoHutfnMw+97eLSwOrgmbAsOI++yJg7A3xndANV+2whrqy4XSr2mq6Fprf6oTOjPmWfKAp7b5bkvrusrWsCspnREhB2Bs64rWCIUYRIKzVhGrMgGUKv2V2NFKmK7wnO1KTO3rJgGX6yp8402WzPenAbt2Q/5pEpXTXG2un2QNCTkYYSd1+R3BAszkZJ9kBAfjvrSutVhd3iY+B077Aw8+DDNBUM79M/8DXCP2MZsi40f00q/h9PejW5ZWFXY7jYCBAeHjJA/zkDX+UZTgEFd9PuOafSDFeUt+IWOYWRhITzN/ixPxoU9Z0j4fTqJcF9//NBA3rWWs0tZamEJNE474BcO/wJ+Q2krAj+wNKZ1FGa+vAzqH9H3YFIRcm+EvaawEWHxghUA9e0kDh+SvMOkxwqrgZ9KEGKpkDepdG7OYED2nGSF0dUTg0jjvFH3ldA5hrHThzSoaaYH/1jacTie0tuH16xssyy1aW9HaSTRfaho5XXE8zlK5qusmGhnECqdt2zTaecFpXbCsqgFrNTQfj0BVP1OpRMfvCEM7E7YfBeG7mwmv0Y+RDg+Bg9lKs3tGxm1b33t6lXHccAiAp4tC129hFehZ10LvIGiUc+/oXdw63ILZMndwAzrXWFOs7qysLupLjQHLuhCBikpRWDT0KndZtPd97b0w4FicAviy065iwMHkXLq3yTQNuGukYN9EG2bTjKOqkEjqoqsWmkA0odAEq5TUO6FtMmnbpqkAa8O6yLG4FiyRdr3FDh6WtlCLRkzmDQgsms8+LxDHGPKOSz5Zs0CJczI8BlHkzmQmnM6fdc/FjFBW9+9O2nJOIbJ7uLMru5e0VgukGEz8lDVg89hBgk17spYkCpa1rb05exj6Ixszogyq1iQ3DXhdNj+kXJ6ZvYT5CqElFQqFsIj2A2fom0diStrFvEnpyqJpFnf2CzE/YVKN4EUZvm2YEdsV9kagSYRQEIG4IXZ3Z1hxCY8iptiswDTgriUCjXCSXe/S6KMgrus9FRhRHLBzUhXJiLh0vSqZYdZUGQDrsiqACISKqW5Al4puFlTZ+4ZluUVvDafbE5Z1Qe8rWl+ldGWbtdxqEnJNi20dGzUP/FmXDa+/9Ro3N7e4PS3YNhHWWADq9/v6cCabSpomwWM/37hw/tGNJF2rA2Blbp7LYuuOI+3nEgN2pcgeSaay0NjhxJQBSR8lzTgxXK6+xEn94+yR49O2SSDbtoFIrJsSCS0JfcxRiwDWBzMvawWzaVLabuVa9XdXV9xCIeNb1+ZpSL3dKfF7++z3A3apN+UE3qsh8aDhMKy4tvl5Gpgz823Y+39k3gkWaCLF5qXqD1CkItMm126rVlNZhICKGbq5GXrTEHd/YtpVqWpgVzajnDHfrAED8DSnTIie0E8WoV7nDMi+32uuulN9ouEaAjDu/qHv0IGfRUGbHqCL1oKWczCWpxdz+B8H5ksRIEMIpr1P9vf+pC3b3mKgb9WYOWqRb81NUevaxtSj1Yq3EzoXZwymrNl4TAPKe143q8KzbUOkr4Wi5JHk0Q7anQkrmrpniGmBirVUMempW4F7lxTXLvOBnAWQ0m7E9VLSOu/+TiDm3pjw4+GtJQFZc6S1DoC7qyABTC6LWa+JVTu14v3QCkqKexpQw51083do4NSGbb1F78KA121B502OLv7FEAJ1PhXHW++gJuZQq+T2+vUNXr8WM7QE4rH3MzCXlAkb8zXYns9xhstuie7ueveW17LTFy+7yh6pHMFOpr1GR8Y+B0UI26HSqUSrbNeiPQ2RR1SwliI1H27dVmytYa0FzB21Fwmc2hidC3qX8y26oeMLmj5NukFHrahVdsyruqeAoW8I/MA0bRqDgM+zCTpjRyTJuEkj0W77zIcI4BqK7hWQkg8s3TPiYUQQGlFrboqI6im+TdwmDHddFiWikiLA3FE1ifswT0OVnFqKSk+58lP4Pgd2mBhTNPZF+1TQ9ufuzuR2lzbM4JRGe0llpLA+G2mgkIjlqtGkdP7u+EwzOPR7TFuAf79TbOP9eB9Cep4A6KrpjAFXUmJS6pTHjkadKXI7FTeHgCuV1Mw1zEjFGHr4f8UCpHutcncClRmuNQsydE3FzXoGpfOgq3zOCtLbhGah0jWZnWLrbgNOpkO/5nEsgUwjixyupB1m4XtPB5RZ677dnYFamjO5SqvsP42KWltiwEITmGV3KtknnAdYOw1jYyLw9zXIPduuXG4U9knPGuaL/Jw33n1mruAn+LEgPmuxnaa+nC2HOV9lypVtxtkDV2QrtKhqNwjn44rljEsmcCeaaZ8McSkyA1SkPxIXIXPmm+fsG0m0u8yvWH6q5YCnXPCiFk5ClKf0frFEp4usKpkKs+5L/jljwL7dMlwj0Vlxvx5k4cga18VfVFolJV7ougBWP1qPrejCj5BLiCkxYwL3gt4I2yqqQGvAqt+NqS+3K5bTgpvXt/j0m59iXTfcfPotrOuK+XDA8XDAYZ7xSgMtrq+OWi92wmGuKNWkpqLxLTpWc9+UvC+uiVLy4ZtVPzHs39QuMeG7WZwtFUnhglk0lKHY4s9ELy8uZ7pGKgetPG2Bl6V4pSWSFxvbwaUrQvN1AriTlPeS8xM38UmuUtt5aVhOG5bbTQKytBKW5/2qEJihlHPWVZ1UpqebsGvRgXWT5P9pnrAuK0opog33joKiNtagzZaZ25ixaXpU7D0cWhfMxG/aMMuuPLLXQTGgwvKCC9ncA1RskwODtLy9EGI7TifMd2PW2zWSXc4YADc3KZoJOniTCRXy1t4Za9tQmqz/Ugp6A2rdsE5i7aqlYps3CYIzgZ4bWltVWBfzMxUGFT5LpbPqeIaT29ZQSAK6bm5OWJYVN68X3L6W/HAT1kx7DyteSqRJwmr8vRMy77UoPq7JODYXDATgNIw30/OShPJBzCNTUsKSBcJg3rURmf+cdatTY96gFDNRqqwXvZtZq11tWnhjWQb65tY0ACgTqACHacI8zZgPs29DeTjMyowrJmf+o3Iom2HIiVqqVDZjpY2fJwZMQBLhaIcjhjTsg6HhRjjAg+iGuswm6TlgRgIkCzM036z9cgd6UcLCUXLMtZhNcoEtl42UKOWSddNUPAeQnDkogmStzcZxEUJZinwiBsyAqbCcFkO2DXAYkvSePDc8PkueEKrPPc3oQBDb+6/3BX3/U/3zPhPmaL1KiHH2/ASHp1GANXK2D5qqa76pYEwIh9GHkPZDiHSmRvDrPS3GKwx1jKb3GNBeOzv3zV0EjK3G3cl9P82crNLS2ZQog873jrrog+B6VxOBLMyS1udL2v++W9yhkdtd8qmblC0lEusXF0alitJtlyMCa9YFI7IuCvHZsGMKRt8og5NVJCwkfcAL6SAP9CDOjyDbMd/07RIt4XvWzIMaj2MaNH3E5zDLJFpwNzpjOHNBCRFcSgIpBx4OOGdWGl0nrv0b3eccc3KZVtgbHX9cI7dc+Eta+sh8o8/Sx1JE5Ki1YK71jdj9njZjMEar3xmuGdqEDgRbGUalAtZKI3UqmFpFmyqmWXZyqZOa8SohMgrkGdsmPrh6WvD69S1qLSLlqn+HTDrQkPfbmxss64Lbm1vc3p4kBUGlU0mul/qw11dXmA8Trq+vcDweUGfph2i9HbGx5OgrORc+cvv2aL8jE8wLli5dPv7B+bxwAzYV1WnCQPrkSiIJlKpRA1tu1+t2C4OVMHIIlW5GzClIlPoSsgJFFzIzd6JxF7kfBveoJoEYi/h8l9EHLFqnCH69iVFONvceXRWSfkSo0yRbOtbJi3BY8ZlNUyuWVTRhKoR1WzFvmwQNsZWYlBZFQZoear72utkG6CCkTYWGbmWYDEpKrCRCGvqp2jCJligMpAuDZYZVNRq2aXaR8N0bFcJ8mHWvZXlB8cjl/HRKh5X2BLix9Jc6ege2YtkPss77xjof5oPU7aG8IjO06IMGF3IRjbAD29ZBJDu09aaIDUgxltcnLKcFp9sVp9OqwTsasMZWcCMxNscv+N8DKC/BZrgzn3v3ZoFGmfmacG9pRmKmHesgG6nPgrMxNVdIBmE9+tsxliwmuhy34t5xZYLuAsS50JpdWpZaJMFW2eerf2sxDq+mylHL2gVLqiAQJtmBBuADCj5ntaAvogKrBOPUdi93R8BGp+5E2Dd6r0U2UuixB6gTeoR/sjN7UMzptGiABXS7utRDFUFvTwvWRYqhb+vmtUTB0N1PJEjlcJjEZDHPUpx9Kh54NTAaH3Mm9rz7PAfPY1s25+81HTs3mnkvaEJJqpS/4zFjoJP0250MphQNmtyFgSVG67xzRz+HIC6K+Y1HJMEiS/47Jnx3exomnIOwLFp58Ad7rnpoRTK+YHBu/i3KfC2wD8UtOK11T1fZtoaa3mMEyOcIcMYt9wXz7ZbGZMzQhZV99GoU2RC5NnzEwoQJslmDxmJYPW6c12iC/nJRw37LRiBMtaK1LmuaW1ie9k9P8EDCL25dhZWG4pxNzZydUEtX2gIl/t0lCaEvNkIzgZLPU2+MtnWwW++lxsByWiO7wmuBWzTtiP+Z6ea/7253s+bHizxA27oD0YRjwKw7hhOhHg7M0i0oo1Y6XKK45H9z6nfSOjHcI/1xgV/f4W4qBeiYDpX6UkITzruu2QYyo+huk2PONgJ073KjS/M0oaC+gea8x1KUQ7fYFiQlANGZRctMAgBQp4q5i09sniaAddedHhsri9Atb2utgUA4uYZR0LaWgKuHvn9ZTthUu1i3TRKuEzBtr8hahRnXqWCaJ12k1ncjqmaKZu+PT5wBwJFBzzx+nQztMiK8IQI63ZvN1q74mgQ4OqS85b1iLUc6kBTJ3DRIAIiFve/ruHBz+gyGp6QdfFxuvqwRpN6+EQ4Pahwujt542CwkfJN6Po3BArB8p6KkKYyF+y2AKrZhG7ZFUwbcWlOmIlhoBf7d7+ulAXfma4eP9icRFeIC213KcdcJ7R6ugi97bWb89QmgTqIBN/Z9iHwfb0lLs2htg17x+wKPIIJ/LnbRRDhqWnCiWIEdAqh0x2UQlLlmDTsHelp5XMUHliINp0WL+K/GhEMw82JDKsfwBewNkWgHjsfC8wHN6Tfvz56dHE/dM+mDInChKE6miyMtlRdk/gEtTbnfBSsLH3JecNmyMIa85WFbMRUiVahT9iRrgDmthkSBdF29ScT87H3Al5qqPfKRUYyMSod8OVX0TlphhsDc0bYNRIRlnsDMErVWC0oLDXjdRNLcVGMopeBQJ3H+c97ZSIAm20pJgfRlWQXQZFIVaYm5gnmS3VEOhwmH4zRKahmhspZtWtoZ/w0m/AC++ATt4WRwz4QBSNBF0jR8uI6tEghRS0Eb/D2Jbmdiru/h3sEW0UgZJheC1nzhkhJV+OKz4KLzUWeG/6DhP7gxROPpGvUs5uaoDd1Secqh8IP2LHwnoU35s1kZuzJf06q3tqFuVdKStk3gUHUbNeUS2yYlD8WVIhXdmqUwdU3hswL62TznuyVJtSFJCRNGTA7Lu5hvZsJ8dtlThGGVQjheXYmxmwgrbZiamHOpmcAYxFYHBoJEubpABADd/pb62tQ60JoKkXDLVq36ObHHpcmatqOAO8m2qCQpjoXE8tBbx+1pxWsLwro54XRakgk6Dk9nMvgBiHiOYMIZ6tGeWIJPzeZ10Anvet2ezsWCTTTlDe8j7GjBBZxzwT187RJQ25wRZw04a8JSfa8Ew1Yp6K6sxeH1bkJPg2QG94bWtjeO7T35gNOcqPabR7ufGJMu8kSIiYDDce7aViLQrl2FdN9aQ6Eie0kydgw4JCXZWDn8vsyMmswq8YXGSaARNQf82jHhIbUmMZKHIubTtMTM3uGVrgWDL96feOPZmPJfsj5CY80dGhi2MeLBYkAxDJ3GvWXcfttbVfKrngriYULk3Wd8j84i+p47t1MogsClABjop1+zJy42M4Hbthbs2UGY9iQ1NSIh+aqlI8GWk+n8ErUamfBubE/QiEjTAJvuQQuUuoHRlS5g8Pnt7z3rEu/G50RaAnDlXHyGCdbEERr+dk0YY/BnDr5yhjvgbWw+AyTIJqvIHQDR69OCHujOE7S7Za6LP+a/3pWsWUbLWfqR4q7vCw6j4LGjHZulxyxHbkFSl2YndIwWpUFzNlASQim+gDsEwwdbg/1uwUTbe2PAQ2MMnQ1g54sEAqUwmAumSQI7WuvY5g1gCVppvUdoe5HodAZkR6PO6LS5xrypBmyRzyFrstZ7DnMdANSSdujYUZHwiMn9QksvX+uDNrLFI2LeFbX35C0z3bd8XczPfsEliqHPlL2X4UUaPPiCJVBHnhdRpba93RmMyfZrJXdHEEXRByd2zuwQBDWe4n0c4f40xIlZA566HK71NnYtODZgUMFtZ74MYYKR89x7J5Xuw28rPT+X7LkzuHSwJoKEZcb+suhp03ybv8esQDnSdBcnH3PXGUBHp0hJ8tQT7GZwEHKeBr9Lrfjgww8xHW4xn3QDexStRXwCWeWpdYMwNcGrUqT+c+TZZsZK4K7uAJa7JL6K3a1CBCmBCABUtL58BdEE8vrycu3WZLTLIuVIb08Nt7cblmXDskg98K2ZuVrrPjOLFr2j4KIAX4adJS9ckD2fEOLxrniq0j3nQIkmnNGXu3pxhwYwCNvxuXdDCd2QNLQGM/WvWNcNyyJFT7Zt092mVo9CL4VcWFvX1eOLtm3T50rNBzZlzgRQAO5i9P5E/1uXGvBvap8LBpwlPeBcAx4nxkyYYhIrvnVa3ukoa0wRkCHEjHwjggICk5kqdr4i3VHFKroMDDL151IpuNGker5eHHmV8FJCee7FdwAAIABJREFUzrsY/Ls2Xw65H2cL+DLyR194d51p8CJ5js/Jn/Yg0zzNFw7XjM/MQjtcyPeZaXsMxrIH2eeO4V7sOiHfFRc9Ecx1LLYxQN/7l7Jp3PtzySC7g8ugwQbzvfh+fTYjTfzZC0YGDof/HepNoqsMLdCi13txxN2rLlkLnxLHicQHvPnuQcA0CeErtaKqVuMBh1lTJTrDAVsnbMPlGDd3yEYTpunpp+BnqoBnNEHvYY1Z6h1RF9xSkTQ4LucQIwmQ+7Ug793BLf3JRGfwziN8ik1H7ljld7e7SPkDLqKMv3f0ZbAwATsXTdPARw083O1IxixBvIBWVPM67XtNeNCQ9gR17JOv/8uWl9w+Fwx4oKF2TpkmINpR3pzZ1Hz9Kw4viaeLy3euKAC0KAZFXtfsGvAEw3rWT9l1I2nARJj0vlJFxu+A1OHdJKVpq03dd+zMQrYmS93MkqAyoihdmTXsp2lBCOPllDoi3+6WSkfUewijCpOQvdfMxrJtWARkZSYxpgiItpcHQWSBb7Ld2FhX2p6VGIkFFoH30tG9cHpsY5YdUVaLN7BIZWUQUa9NGUGqlmZShrHFphpwYdkDlZhUUBSmIBuARzzCNGnpPM1NB5FL7gJ/0bqneQMDqLbLku/axQ5DM9OZBh6wVY4i2zeB0JRIAsVy6bW8aqSFiLCbBZ83meYe2kopuLq+EqFaHbXL0lBqxWLlPjX6W74LPJgBtpKVzcaWCHpReFjQVrL4kMOzKE7rbkpkteUlBWlZVpQiNYcBYDnJdoMnjYDedqlHuR64oHESRpXmD4ktDkPXNu5B5KCP346WZMJ3allhGn+Qf4KOZeUmXhhBiexrzaoXmiYsG1/I0TwWiMA0eVYNd7H13B5WzL1HrBEA26WvRMe0exR0SOdMnr9+zhnwTrrLQkVmwHZxmIgQCAok7QCuBFk6UqmkjFADtLS8GBFhnmZhBgh2ZBGLtvVY3kbLKzDVqgtCKxNtRQNgNAp6EkQhY75d2dyO+dp3Y7wW4e0jfioq5YA5/34/4yGHS56LnTx4oaXr01Ze7qPXlA5oIQQwlNizRw2XIoERBjfbkWdMESDXbAAE883qwwNA6Kj2BFyYGb6VnBzCfFvraOpmEcVJBMNIdVAYGeTYgkMIrTdULmK9gTA4IviWl7UW1CnK5k2JAQOWE1ncRzpNshZKys023AwtPZVGHFQ5BvcmqTvoujVhaL6Fwq0QfEEEpwASfP08tpVCuH5xpcYkCV47nTZQKZhnMTVOU1eTNMK8rEA2rdT6RYB6LRhIcLEdEYLgRnS/ZFUYA5blvm0dwIaihX4AwnJapB6BMeBtQ9tU+00C56D9ZhQm6f/Ioy4wX7og5DyhXJ8DmOJzd9El08fZJRSXpr/PrtPfLjJfw1EVslrvWJvU6183EXjWddMg3PNPqajFHiTKXejV4big9wmH4+ZpfSEow9eqtLECF6trRxjw/bB8zxqwIbSZR5Tk22e+MpsZXFviqDLkBCP7khGEWs542lApUq9T0pAkXywTnm3b0FvVSlhNXxy9MpPGqr6CdZ1R64oy2Y5I9rxE3LLUBBvr3SvjabTh81X4NrzmkgYcbIJ2V+5fFUKRMdFsOubBysGu/VrqvQlWpnFkJl6IEhMOAjD4QNnEs3H8GM6wd/BpaFSqdKQEwUxgppF7SBLlu+BQzURNxlNCqAB2Gtj5MfikkgZhxSRqITTbRCQHMJ6Z9tkPswxhH6BivycG7tQ/MS8Xrg30TyVcEqnGPwmjnaS4TusRoElGIJ2GJEvX6HlyOuTwTzai8IkHfCMGQe9n2+O6S/Rz4dCAdUe11Stg7VwUnd2MHmlNBrhQE0SDDzxKSvIONvYlCc9PzIjNmiBv4d3bRo18H0A1fKcR9umu+4mVC609cDEL4fYI2uN4jMFMxdvWAJBkFKxC19vWpSY4FXARnOZicB+FIVu7o9Xl86wBp86XNDlSUSYz3QBq13xKKY7RfYME+2xqOjbmezhMKvUL052mCfM8o2o1KyspacW6LWpOnPYNbd2wnBaRsDYreC+LqQOor29QpwlMhK13zIcJV+WAwgWkVbvOw4lCwHjC9fDgdokNX2b2Z6E3fl5aSu3w84yh3JEhfykinHjJzgIzp7JqVb1vsO0lCwOMKn0gSAQ6FUy1gFnMrHWqKLW4+SSbsZvnueaRvFkqf2zrHViWFadFK07pjlqWf2sWZFdh03ZZRr4kDQtAE+JWavFYBKPDRQXJeZp0/1kpBDNZ9SzdOcYMwUSECgJ30Y6ZOZVRrZG6l/JdjcGybS3XG9Cb73ajsb2IDekAi0Z1ZqvM2KO1d5rTY1sphKvro+jcCsP5cAIAzNOEptqvVCU21wQ85SqC9EYNtFst9mIgD+brpufBDSL3ey3nhsRNFjADy7KgrQ23p8VrP6+b1n9u7PgR8aCCIzmy2sGIlAOchZ/cdszhKRrvj/TOs3e8gfnKPYkOUTzDxY07hNT984f+Gb9I7661gLmIsKYbmGwEMHcsi6Sa9d5RtwrmjqnKmjrOR7GaHIACqYhGU5GqWFrlw7c1BDzNcGsNy7Z8nhlwkpMM/km65OESY8CRUhB1dkMb9nwv09QIQvih/i41z83zpMx5Rt7Pl1n3ivTAKyHsdavoJHmdTDKxnaMaEQNqgq4olXSP2whocSL4AN3zadkyDwg+PvUh79jrvvEtGPNlBMuysFnH9tGCRHANdww2Mjp9TlDONDVjrJQ1x8tEXniK7vBEoWV6ROWTgN0inaMOdFhn4PiQlBlfAKFBpCpRiWm5TGPj3plCxzQ8GtQJ+1YIaj3IZT1DKziXw2w+OBhpVhftiiRz7W4PJow0FpzPz7s1icmogxWAxjrDOypulqnY8ISG52UL2jgIGq7d47ONz5ikPciKwkRVtFz7OfncewgEnOY6Q+mu77EcU8Bo3kGJLsRDvEvbKUTGEtkFhvNbLgn3zpBB5126ZL6mNJb8qxsJ7ngHYRCWainoXiNC7jGe0jTbxQKxSilRuCbV6pZ/ii9Not0y5RzP8rllwNJIxZxaiyOz+FJYy8/maFJB4s6MRR3pom2oFrw13alDtDAhNBOIgMPxgGmacJgnXF0dUUqR2s21YtJtqJi73N9FO1iXVSUn9sLpYKC1DeumlbdYSuCBZNurzlc4HA+oIFQtBc0J7x1P7lIsn8TsvH9ovPAuSfT+plieFlkWJUy4EGRkvZThNbBV2TMTqBW2VyULnhLDov2KllIGok6FQF5gXXz5FgGfbdAuoDnDQyjlO6jQhe+PbZ05KhxtsuXcsrQhEEgGNDIGE9OYxfdLAEoXWHu5SMdpkoCrefJDtF/bfas4S0hirvP6UgmFxZowp81EJivrWi0oyzonQMymPROQXFjIPFmpkdWlkKrJ+Vl20eNbISnKs84Nc5PAG/GFW+DmyDTPmQeBvBB9ZtQmQJMzRbuaUM7dKXp3h+QLZ1/6ujW1qm1eDtcjoL1YS9cN3JEY5yClDX28SwvcKy7jv08D8/0b+R7J9V7m658h/GbGPPTXOB2SgK1XSl0YjfnRfZxlUwRCrZPu4yvV4Yhk3+d6K7sntbbh5saYZfOKidN0Qmsdp6tFY3+KBO2qsobCoE4SpqLdMyZu7qdt+9wy4J1mYppNiQVDIDC1MLlwbEZuKr5V/tm2LaRL1ThM6zKJeJ6tZvOE49UBtRQcr45a07m6Bryum+5cE2XKVk1rMI2tsezv2prk6dVaMM26afdUZePtXtDZNm1jRxayB33mjd2mc5e5+V42lBeZX8q7U0qYjdsZNaDwoVFmvoAT5KEkoify7kgPWRS15IC75gc1tarVwhhvVtbOhqjKRjCqu8ztb9k4qq41qwm9NY90TdvYh/abCAsBzhy6lkDMEhy5BmtBV2o+nsKVsvdL+fug7IYoPUP2r97v+pIf4pacxIgvEXOxkHPSeq3rHH8MvXqCRgg4TF2Jb1hJhsG7hnLhIWdz74QncD1ppUajBg1Yh9gTjIQQR1CQ0I2mpsqw4Dm+Dtp27h9gUtGOeuahnX0fxpJh8Yg2BmHlPo/toq93/3f+cObrV2HPhIexJTwTg08BUXd34jSR+uOFJphmW2uVvQFaw7oCwEmZp6zOUsSlCSB4yya8RQRi1oyEEZeYrdDKaOW4r72RARPRVwH8NICv6FC/wcz/KRF9DOC/AfA1AL8E4A8z82+9+Xl50kYpyD8pzEOk4rUhq5kEjNluXXaZ2XKxe9WASim6C0/RXWWKBmvIgp1n+ZzqhKlW2NZuRNDUjmyiC1reuywqMQ8xei9Y1w3TNEV1m52kHfrOZUmWWTSc3/zN38RP/dR/iU8++cSu/LLC6Z3g7cgJETMpi+vDdemELwqK6+5A/iSYDqPLJjRoGcMhkApIWhGcemVTpxHzMLeKQGNm1CCDCcKZ6V7Eff0hDe23fuP/xU/++H+Ib/7ObwPA30NEf+JdcdxMjbJwIxBLgsJKIoQXVBcV8MjN5KOmY770yHkfTfFnj1WGGLQ3RAxCXnPkmjW5aRqw7VelWIH6urr0wfY6FQWgu5XDBRq2NxvTjh1kPn39Gj/7c7+Am9Ot9uRxOJ5NfwwzAbPq3UlwwEgQL8VhuG7lDw0ImlxChIjmJ9pFktsTYqtIzyv1LU6T+8wtNcmM68gZs+XanvZpUNYTEAY+qxN8Wlf80v/ztyTISE49jqbc23Z4uDMdZ1mIhi93N5u94Vz60+KGjPlWCDwFbyVVFFTEvYiObZtkR7HWABTUetICKJZ6Kr+R7hdgzFeYeQha0q+oaOiR2KbA7Tn0hVbu/VXaBuBHmPnrAH4/gD9ORF8H8KMAfpaZvx/Az+rf79DCH2easCzmkGAtqrWr1rvqsVilEzX3WfpH6zpdpMn4U8Wk2wceDjOOxwOOxxmH4wHHqwOOV7N/Hg6yw9E0V8+vtEALC+KQrQ0XnJYFp9OC29sFp5OYwpd1w6aa+qCJyQidXQwQ8MUuGzr/0//UP4N/99/59/An/+SPAsCXHwdv8kpLxoRhjOrCtcjXPOzx543hcVjEYhYyM45s76UEH6RlshOgdDtI203dbCJC8uM+/07JD8wE4iLj7Xkc43j2Z0ut+EN/5Ifx7//YTwDAX8UjcFysKIKX6xL5h1sKCtvZQwJoGIWQEBbYmaRvmaZaX0nMODiqPa/HM4Ypk2szM7Gj1vEoNf+WtO4yJQ286o5NyoUVv8xtNOyt3cQc+/v/3h/AH/qDfwA/9A//g8AjcZwR9X+ziZk9LiRHbgcMclrJILWlhevBYwpXj8IvpDux7SwH2gex1jVsWnPb69Cn3HBnzGzutuhdPvgCDl9Sec9PBZ59z5e/hK//nu/D3/21rz4a3g9uFyxtvm0kHiArQxkcoK4Z9sOjRRSJZWe8ijpNHmh7OB5wdXWF6+trvHz5Ai9fvsTLV6/w6oMP8PLVK7x48QrX19eYplnS9kApgCrSl9Z1U8uFCVUSetg1+KozvLjKpm4nj4C+n/++mQEz868x8y/o908gBOq7AfwQgJ/Sy34KwD/5pmftnvumK5xQhYQYvuB90fJ9UINLlCrqu6S6c8i76a3aJObcyFigREG4XGHLfRvGlvtsZzMDyOr+OOIPP/wQX/3q94IZOB6vAODmKeB9ccEOWnr08iLrTYxuF9ayG108ZS9sRBGO0X9mqvRQu7Wn4g+584kIDvPn0aIjvuRnGa11mqvA+OIXPsb3fd/3m6TU8QgcvxQgOPqkQ9i8cLP3ybDfr9+N3//cKUsBqTynu31zfPymPQNZkx6UL3tH0pYdhr42whyrPEi1sv06l7m8vrrC7/riF8Askcp4DI77XCdGm7XeC3Qm9z+PR54X2vL5PUk5KCmYzaw6Saixd5tZ0l0siTYNAjpDtW2jL2biDA0ytK5zun6JmtpI5nnCi+sjGGxVn56GpgzQuacnDhQarsrjGCgKYUS6dN0lmmX3ZN+yuas8SLGGtXMQHMtOiNWXZV6zn7PcIbagObsvB2s9QIl5Kx8wEX0NwN8P4OcAfIWZf01/+lsQE/UDmzFWc+CPPiYCed6/FzDvImGsnn4k2u66SpCD+9taA0M2zC5FTH4EuPl5PshRpyqfVYNQ5gl9a2h9AxXGNBdsWwkptxdQrbIVWalgkmIeKBUoBCoVpUyAJuPvF5ZCMCTuPfM1aSEpMb/xG78OAC/eHd4XJOf8vsucNv7133VOdguN0mfcldQFZ8Sivc51Ak3AYZox2RaSWou7rRu4NyxTxel0C+YZbZs90K1rnVci0coOhxnXV1doDbg6nKS036YmJMWTOkkVHCnIMqOiqiAm6U29iTQrE+UG8QMegePMEoTl1piNsa7mWyr6bgwa6KjYhDozmlah6DESGWeCmbhxRq/LDEi74Kbjko5sUqWEQlY8BSWeUrgoZnR/rqHWSCT3RFpw5ZNvvQYegeMdrAUtJMe2bVvaajGEFwK5FaWD4WVrM1PzPYylj0x2ncClVqBOhGnSegJziQC2qQKkmix3oLGkb+muOGIFMLeEbQqwnxV1UYT9dgTXhcb3/JbvtVedpD7xI2iK9ozGmIuLL8UO++j+rg4X2p0UeGRuPROG7P1CV9VVkphzKdWLaEwQYRgAaps0PVCCIw+HA4DwEQME1oyXcYOGjt4LumwfLdXdTLzVGA/bc1v6eSHCe9cezICJ6BWA/w7Av8bM38xSODMz3VFklIh+GMAPA8BHH3280woT4/HvIReF1BGfWfN1f+uwjVcsokGzKwQyc5GlLaiJuWr+I9TkZnuzUvLz4OywBEH5buY3uoR4diJpEaPGHLWvDcFuT7f48Z/4LwDgV94Z3h9/edCWgCAv+z6ND7l8gs5OaXatyw7nDyR7DwiVCFzCHG0acSfofsuyaUbbGlqpjg+m1cj45Kia0z1V2bSeuaFTV80hCrS01rSs6OQdcljDhCR59u3tLQD8bgB/9F1h/vL6eth7t/VY+FT4HDo6HjYR2j6UwGQhKL1v1ErfpIEMfY1PQd/Y6zSLa3cTyqTBK6MXN4LhSKzdM3wxmqpjW7eGP/dzvwA8Asc//uiVMrNxQwkv2Wn9tT5TwK/4OuSQGuTlaa3CeaPcEyVAzSxvFjSAgN40Jznhrm+mcSHgiqE+/8R0ByYcYI8wjYwn4ykAWn5UT6bHtNbxf/3qrz0K3l/60pfy+TPmexETR93iPnnifDB+hsdfE820N7s1zc6otdNVAiLJdgE8DqgmjdiK2PhLzA0RX+9vg8Uj3nlfe4gPGEQ0Q5jvzzDzf6+n/zYRfaf+/p0A/s7lPvE3mPkHmfkHX716pc87uyb5Qnj3d9ps3M16MUi/TrVkK24u98T2U65N6Drxyj3gwS0aSiun65MqoD4u09XDLET63MtSjwfK3CnJhuS2tYYf/4lv4Ad/8PcBwG+/O7y/cPll0iMHSJiKE/Lpp5+9D/neJIGrgskMjXDe/ZieHe8K1YSYQRYZvbve7gnamRfM6JLINq9ROSO0reEb//mfAoDffAyOXx1mDcJqg98zpyAJISiIPNxxTYgJNUyquUzn3s+5w1obDu5yFcgFirGJ+SqgIf5mg/OFYziPC3M3EkX5V33XmhJSSgWD8Od+7i/g93z1e4BH4PgHL03gsSyIKL6f172tP48AT0KMaVQeb0ChURX1hU8119kWP6PElEjMyFG/T7MWNiniynK6kgS9gJkGtBnhv4s27MF64fc3nWUW5vvxFz54FLw//PBumuJddKvMu7WR0Tn79MOUoySBnvXD5tp99snFaKZnt37qfBoTroNPP3doFFHdMrs12dlqbdg2K2csJYtLqbivvZEBk/TiTwP4q8z8H6Wf/kcAf0y//zEA/8ObnnXegnDY/ow56dnSjvxol329I/O1nUZYg6HY8yhlQPkwCVEJmxXPMA3aekla1N6ZrzxAaJBVQrlfb7B370ni3p/BzPiZn/lpfMd3fAX/6D/yj+VLHw/vs5W6C8zaHf5b+n3Ua+4ebwaxUaBUTnjPM8bPRNxJmTc6g7xCEA/vCcFBhabss7G0ppTaFP4Z8ut/+qf+E3zHd34VAP52GsZbw5wZznyDEfcwSxlBsHgDSsQ346L9p0FpFlE7zuHIhC8rMOc6LRnz9Y0ykn9cn0tgEXxEyvXiNP4+zivB8HonwOnbJGWsItdM/vN/4X/HRx9+iL/vB35vvvjt4Q1G7xta35wRd2XCOZ3NQFE0itwEH6Hj4QcWDdeYLyQYbQpLmRXyOdhxmHE8zjgeJxwPE2bNssixJIazDNOoWGEloebChLM08KZR5/Hv4R1w92uY8cu/9ndwdTjgyx9/9Ch439vOUe1Cby63nbiQrF/pGYPbhVzAOX+ZCVex1qhYwFzdpfAFA5Z01F1ZV9zxfghcmwZeSdClBWuJa7XWSfYNuKc9xAT9DwH4owD+EhH9RT33bwH4MQD/LRH9iwB+GcAffsCzMoxgBc5lMPpPEQTt5vcCzoDs2o0S16zhBHHNeZ0JuU2CSrSCh+fG8zt40MrdFwebhH2w1h4D314G/MVf/Bv4+Z//3/Bd3/Xd+A9+7E8BwNeJ6J/AI+BNuy+X9CLXPF2VvHDjwH4pXXRZBnd+kGDKiSAm3uwmzf21HlTjdyQezbh4Hvt3JeUNCT8AWVR/4//8y/i5//Vn8d3f8zVA4P0X8c44niw4yaqTd/PK+OM4abjlXctwsi9ZM+7nMPr/2vv2WNu2s67fN8aca597C62UYlOg9GJphUaTik0jQhTQoPYP8YWh+keb+MAHWgkGG0nwD6PhETQGIkoAQ0KDtVgjQWKL1UYhYKy1L3tbSmspBSqFmNJ77zl7rTnG5x/fY3xjrrX23ufsddbaZ3d+J/usteaaa84xvzHG9344NZYZ9uIG/uCtvnZjprGGuryiu2YQRx2n7ONtu7SljYUp6Z7RCOonfvO38Iu//DE89znPxpve+p8N59da4zL3int9jrh348Ci4sQ+DdaetC0RD+CJXadiGUon4qSxCoxUkxRM0XzsXdaNbj+RCbYd0tChksLO6/ZluCYj3Kg32D519y7+36c+jTtnKzz5kY9dE9+9wtBD0BrNEhhU94vNsT2ztWPtuuEacyZvAo7eywWqYIImEhncrUOBbremGpKlERmw3bgx3nbjUjTNbDPhfL3W9FipIyG/SXuE4gaXMmBm/llsTbnDH7ns97thpgWCgVLFd8It16qF55NIiuZsZEiDc80FjuH8hu0+TxIB2eR5lOYrab7lvoD+VAome7V7aYoTUdaKRE2yN4kWNF8hV2fEL37xl+D7vu8HnEl+89/6a+9n5p/Wr+8b3/clAtCO3X3p1fcvMNuALCWEW9Uf85EwgkZscwH360ued0WqhMJWGKWZ/e0mzoxD4B67q0Jq60YLKumaMhy/5KW/F//yh98CAPimv/TH3s/MrwiPcV84Z2atSy6VcKzYPiW9MbW0HwtsYwixaFE57EQFDNVAW2euaNquanatMfdQH9LuEMQTwJm3aoqqMUrgUqvF7Z3Aok+VizdjcL92E6NaQ3aObUeU8TNg/vnnf+7z8E1/7s96rv8PvvnN11rjTFoaVlN/Spk01Wdys3RLAYsSpuqhXs0o2ACUVlg64rgasTqTetursxHDkFzztYAs5gpQRamEzUa0qZIIZtZvooxJnE0YdkWZyOMBu3VusMV8416woCgO5zE+6/E7ePmXvdh/864nP3wtfMut4sjmNEOVk/DVRcx3F0O35d9oN5xRWhMXo9vGvK3vs3W7Mz+vWEkAJEZCahXlLK1O3SOWviR11gc3VRNFK04rFrVeSyrqer3GM08/DWZG1uDcnOQal7iAj18Jy0hCJA1GZ0y6NomuzUuIHnbgpgV3mpWSNTNT6O8BWwQm8jYxuEnPcKnftV5LIeik6agBzwsiIFx/zoh96BEh+xC1U1M9COzhs6YjdYPovo3Hd2+arXSCKNTzHqKi3zX6PXczxE26Dycz7cHWUmTQPP81qbBHFxKI+wFbSy0wyDSURmj1znPVqBFPwM0Hzj7dJFc7Dba3zshraw8Y1VHTfgHM8bvDrdOZDDok6u8B1YLlHMFjTHeSySdnvjoSv9RFmtT94xwRJ3XXs6CNnU0oN0pk0GiC+RlNA/ZqYZm8gYs1FbE/ZmESVswnKgBdt6jOehSew7lwpH1Ah1L70jXM/ruO+fp7w9DDoih7CArRXpm+DTvMjSv54VuXVYLpGaYEkZ/taz9orDDrC++gV7QdIdE041jsBt0DdOZnFYatNgVXBmd0HcYug5OUoqSw0G1BGqFFYTAKzLskdVcZOQGVqv92KkVK/mnvR0N+DHAZckbyKLdhK+LNo95YzQlWw7M0rXdTrNNS8QhdcGtmPgxjF01n7Q13LnUlPFsTQ/2rmUkeGuxlvpf9KBJ0e+XujPlvAhkPDAOQFnXUNNfEqCCvUEMlYZoK8lSQszQ5l8LaFISt2TNwCN7zsqVRmxNG6JV/SSPeZ2LHtYDZNVNP2q+iZbkpNMpnPhbVxgjaPxZOhaPZWKK7i2/8oqVYc85ar7yAM2DCX5spI1QxpqIv5+opPFosg7iKH9ifrX9OY9JzF0GarwSGpHUAW8z+uohnBqoVQJis3Z8UT5i8CIr45aIsAQrPpW9SasTb/MSSqpi8iM+4GrSufMa4WmFcjUgJGIaEWitS2iDVHFwOQUjSfyLME6xNI8I+cDQ2mc2ZLQUcc+NM+zDTXlzbfliwh55dQFS2DtP+b23flBIfWdaTxVaYxQvMqESyfpO6FEwo8zKRRsvlb7NeY7Neezc9AOCcYS5HmbPq3xmf2Ww2Wmxng/VaOubVXKR1YR2VoV+MueMz4E4YMSIRNB1ELVMc2ElV+in03rRykN7mjbmZKaxyT+dQz25GaHWEk2pdMqAaUgXcFG3VayYz+UkiJOk1JXouudmDuvZhBhwk0z3EXgnw3M8yQdB0AAAgAElEQVRwLLj/O/bMd/sqTciKz91pJGhKVWVhPmZKTF1P3YLqvZiTb4y5haFpWbFAi5lR5RilOOYmJdP1eYHcG9aUwwhD0/jgo+ZthCtTNrrKgXg1ixC7SdjMxmZ+doFDKXkK8ZXWW9nWoQU82npugUvN/Mwc7PbzqVZN0nRqrwEOwBtwREHTCJlp2zXYWA8AtbALDta/25qnlKlIF7OuMmAQGoPwa/TDGbBFPmfL923tH+Xz2GpPJ7uelESMPCgScReEVBiMgmj8wd6dFRkz8+7o6KgFH42UxDiYoHgCuxmxH9uhnc5P9c2gbqgghHj9eOUfFc0UXdT/FIN6/a+oIDs1IVRa2c4sQNzfC2hrxATeMhVMm0l+V7kVetJxXARHZ8CxFVtjumGPzzc7NQLWpPg59NpZrB7TpF7zEzQzVS0qJSnSpStJIN5eUNvmoY09mbkiNWnZ+qm2Fq9tF3gaUjdaY9U7urEcCObsZucx6g2HF19l12axDcKzI5G6xZs2FdCJEeDCz1QqUhLBJ5eCsSZYz00RmOBz1hgLYAymma3RNGxLf2K4CdKGYnN7EOyzmokt/Q3Grwje9UnIseLFLCZmPp6Py/Cp/wIBkc1f3VJghCDzoA0r9EIIxKlU1Kl0xMeYV1vvrAxW0iQyxSyAqrETupbZxqs2K45rn/25PD/fCmTY3F2TB9u1S2EvF2h/sbmHRZB75x6G+N1hj9IqJ3n0LLUa2dHUnFOroCRd3MIzxz+7eKTnVYWBqoxEFyjb2IzOGNOhgKIZI+N4rBPYOgRdD8E7oBuiPWM3BDOPcz92tHXR8NOPb05Btu7tViF0vKIJucqAU0JSa1HnlvB2s1Zxr/ifrWqb92wxQ252rqhVGjQA0HLE5zhfn6sGXFHUZ2xjuXEMGIisUqvVRiE7vPFJ5GbGaeacy+5A7VpKgFsBDyGSxKTJ+EKgigbOmMlqzpB9EaFtUktP8AbxQ9aC+dRt5j6EXaNUYe+tpsqV0rIPAG1xs9ELOxzPilrMZVujV/j9xTZrR5hIK0Kl4msAzCCGVzsDgPU0IW0SxlVGqWb66deLCp2dcOYV1LzNW50FYvUCwZzZXRcqF9dGZXxyv8oFFUULRhR05e8MYUEThhMruAbVrDJmat1gfb4GiLA+3yDnAWMFEg2KasGZ5yxquoSbz87ldVoXlE1BnaQ5PFWWtnqGICtoUKGKmzFpNcsxKRMmeHaI7VMWQmjCiXOjQwDD6+9u1gXTWt5Pm9I0FN3Hcr5aqFxjFYS7ME19jWevf+1tG6Vqnr2XVqT6PGYBi20Qbc1VuDDAFhgItK5rtWfU25wWTdrZBbs0zH3fXQM4/MVr91kg20Por3DJOb72Z4e536vx1QQx9wEr7Zq7PKq6LOu0kSC9aULVP+KKTOLvH42ua+AXGNrOUCOdmXH37jO4e/cuzu+d4+mnnwKr9ksgdVtOl+LzBEFY+yamfRPYA6KWZN92k79ncfVeKXLpkud/6E3f0Z/AQRvg2YKOUm8faT0/boNsA5X5JGfk7WEOpofNkTHDE3Uvdgrmp+25DM2O2iztJQ/7fN52BbYG9A331eaBW6F633AIcxcFg6gVx7nt5rcNwvjbnv1+DWgipr9zIhArNQHeoUoZrQ3QA0w6bQoukJbg99xsJlDKWK/XyMOglzAhMLk5rtQamkRs3EcqZtoKLn0HlyYqtrnfUctLeY/mDZMw42ZusT1l6U2ezH0QYGYRIDat4YFFhMcgL8/HpT5CI26NljdKoVKeNZ8IuaHU9m3T5oNp08yeTj/skecLvz1DwOZVHrq9Nw15znCZGsPeKTFfAzqNe+bnDB9kOc/mYOtSOyjHBWPlgLMWyBfft3uaybhjwDGnvc6qpVHoNe5z3bf4jNdxK5RnEzByFcY/qXB8IzVgZ7W2kImAopIg2YKmdl6wM7TPcWLh9Ms/d3ezLZdgiX5cVfsmuF/AEqpjMIcEX7XgHmdCymSTSkw5JQxaRcX+5pWO4saNgkgfufcQ4AEv24/niqFKc2JiJlazCgRrAKgF37GWyppqkV6cBGzMtFoYtdpGVybqts92K2G9mpunZQmlGfzcajITiIBtIvCgQFbIod2TtStRqWtMk6ZUZN3ww9BrTAQrmK/5MWJlIfM7aR30e3fP8dSnnxJrAUlA0N2759rjesC4WnUCoMU3TNOEu3fvYZoKPv3bT+Hu3bt4+qlnsHnmHNNmA2wqUgFyJTASJDSuMWEXfpQQxQpl7vtMilEWc18tBWWzVouKEbvD8ITNZoPf+MQn8czT93D37hqb8wnn9zbaEk46cSHb/tdXNUObxpRyBgFSiMEKbdxZIeeEO4/fwTBkrM4kDSkPmh8KaJ9fuEWi1iK4LTI/5+cTNpsqwVWsAaUYZEUoY55VzJTxaQDXpf5dm5SeSqIF/9BOvn8I6Nxrak64jI51Qsvs+P3sP2ekzO63NbeDrHn1AZciUem1xQltNtqbebP2zlx5SBg5u/ViGAbcuXMHKUnwnVRv08abtTHd9XqN9fka5/fOcX4uecC2yqQj2ubSZzkJAwaCgGw2BaKmADhBbf/Hz04I5pLXfFPrIvRXXY2WdmLCfiy4PXfURwf/fJHMGav7jZzpRi149sNODu8XdNQ3rg27LtFJxuQaaLv/nkuFuZmfJXRhPmOGemMusz+EOTVJVnOtU51rEno96n41e8ygeQap183TkQnPhW46EL4xe0zTyM3/WydIQfdBakNX8qL/hiMXSj0gKDJSWaubacJ6vQEoOUOVjT9pdZ+xW38mQJap4N49IT73nhHz2eZ8LebnjUqlFdpGMlR9U1MBMavLAF2Vs2ZKaBo0NODKmLBJyeSa/VxYvn+oteKZp5/B3WfWuHfP8q+LRsQCgJYVzMZ4tQGDaac+aWguJWXCOWesVqPmhw4ekGUpka1HecFm2sirCvDWNKYWq16mriYCEBpXNBy2cSB8NHx1x8Kpbu0JeOzScZzu9dd/UBDlY7cmK9pw/93OcymK8mZtIX8GMlsz4GspztX8fWyW4HX8Sei8dyNzJWvyIFtL6XO3A2UPrB2GIVTMSoh+3b63c8zJt9z4OV72w8kYsIERaErkG7hWdTTNB28T5CX0oJ1upC4nsXSl6HytFCV29YWpdGzJTqVaz04haiVExYmUJETFNN/m+23VceJr1+qQmiYMapIpYPxv1wTR7PU60Limv4ub3R0rdvasMcRcatU5insEQHuuGZVoJnxNE0vNZz6UAcMwQoSiCcwSPCHXZ081s7Qesk5TYC+IYmbdWNHJCvOXOmEzbZA2pOkogwfdOXZnWv51gUhTUlCVrjKoaocV3mjHFMI0Je3UUuHR+SmLNhFdGWo1ENSy+qEI5/fOQZRwfi5MJ+eMp86eRh5H6busWp0RgAojVozNWojQ+t65+5A3d88lzWIq4v/VIiakzIrY3DIVVKoX5JD8MK2AghrkRgl4QS3gUoAqUaKUIN3DIMLqdYErY31vg/W5Mr2NBWD1BV8QNN9ogI4WqUZfIg0BnEO6+bFo1Lyor1ORVKfqAWBaAKaI/585KR8x7V+0NPlggksbXxzPLt7ZH7PgxMako5F9N+u+DsznrFkXrgpxpPMviGaEJX6ttMSEULZIfstasTQhG5cSbFb3C1fGZtpI2tpmjVKE5ozDoLR7RE7i4z87WyGlhHGUY0a7xGqifED3qdB7xYNNK/o1tQ9Oz4BJB25I10LW3NY8OnuVPZkpDNo+LWl6SkVthbo7qVAJkEVIArDEu2naYCriNF+fi3YgdT3Vn6S9pgTpamLMpBVTMvKgr/p5iA58S8qeFd3nnczq+hrBbmgatW1QP9x9b+PYwXiBGREJBq85DzO/F8wHG+sgSwm/YRxQa8W4GkGJUIqk71BW8xuaZjFpRSPzMzLQJNtYmlENzsJ8gakmbKY1KAHrzQrjNCAl0nzN5MPtB389IAKGMWnDbmmXR6WoND6Jp0X4sRZvqCqUjMgqHSVvgUdaDKIxAzG5AXjmHqZNRcoZd5++p2kzGgCImeRtlh99b9WD6qS5yprjzrWCNwVUK6gysp6b1ApB6lulWsVlxAxwUS1MWoAC7BWxLLoUZQKXjS5wcQNZyt51gSvj/Jk11mutxVsqyiRBdxZ4R2wsSbX5qF46oY6zH+I6/EbNl12raWXy7JamWKu0nyyTBW8qE67KgLnZEzxKXBtgeGCgtWe9gBC4yEDtc5Spey/3jMwckhdjv+a7D9rtd2nRu4/LD9l9t2YGZhaXCjNLTq/tDWPxuu6r1gMQE/RGTdCTrnnGOI4AgGEYMWhFrJUy4CGPTivM+m9WOOEHyV2QLiCxVNq6CkZOzoCNettwTfrcOsV4qr+nwNhIGHKlLeZrpk2btJREGyIQLCJzKq2ggef7evk69n1hZdDmGu/c7xtNf00Tp/55TLDA/Hnj4rkubF+E4ldRpJ6VrOkjoJsFoR2K2352DwqFVaDCBgFSYYZbG8ghYxizfid+Xo86nQe7+P+8jRuK38uracetBVzf6INd81fNvEPOdSC4HxJANeDMyxEqozLzNMsxKWWo6xJa2pBJib5I90ZkiAqACalW5WskGrRv/LmQZYgKgWlq3WEtTMBc5VVTlUopXgbT83fts9eMDlHNXJXXt886+xhSAoPVn20dag7AgGER7+ausFRD+J8xvrnm2ATinmHJ3Ig5s3IFVULVPwaBCykqo3mzZUxY+dG5r9MZe5sdJKB1UFMm3FWs2lqTget22m+kLzb3+tRxHx9awJ9frzOzXXCzmXJ0ZcVjB2H0WIuZudrmPgbXxmAtQGg6tGGCNGSwCPcBRBp/QY0BA9xM0xoJX+uAcTWCreMZw+nYZXBSBtwzKCO43hgdlqCemOUvayk4bukBXBllyB4xy6HBA6CFHUrViDQ4k5XrS2qERIKKD+f83jlKLdioKZqLBKFkAlIegASMqxHjuBJn/Zk46c/ORpytVhjHUerHWvcNaztHcePANcUtxnxYDHevHZHZt7F3QDThA7rB2Tb6HklBv3LGSBDNlwir1Qq1VuQhi/l/KpimjZiZEyEPWhh9lK4yICmSQkggMz37fahZRfRP9omZqCaUQp5sn3P21oDeMu6ApgcitY5UILOs4VYFS9RfRkLlSda8NbSvgTiTGBZFCCqoJWvOekKZTGOelEiI+Rom8MGyBhRJbm4Psx8Yk2sWli85TYI7NfujVnCZXKOF5VHWIuex+nYtmogZouK3toY5E8ZBgsLykJrV6hAMuDLOzydJo1Lfr6QZQotv9MZY34akPlloEBETUEms6VrYA2CUKcHM65I6BvEnUwuyswBO8c23IiBmBofPgJjtvBMVGJbr5cICwqs9I2HfLgMRPEyu1y77ve/P/4B4virMMvx2naGvbVaaVfDq4xPyuY205pJSC4gy3aJCpEWpw7TURFip5Wg1Cv3OOWMcLYix+X9lrkWQND9wooTVaoUhZ70uZgz44ic6QSUso8w944malQVItc3SmLWZhjzCWF8rGkMwaci1C25pG+yMAyrJs2q/QnDMSW8T5QFK6gMgQIt3NyFgCObmrMyWZpqwaUbwEfaa70OLgN4Fs5V+8b13bH37/ZbayLPT9EQTxJOQvJQT0pCQuUmagJSQIzPvz9I+dt4hMF/z9wTLHFzL3Pln37fnP1RByn69Bvw6LYxrs7Z1ywwyzRIVqASmigSSMqzK3OTV1lRVgk6+Tpvm17QBgpaIjI9oplRlwGDJkwSzasfFBRlRLYtRNRdewZbuI2O3gDO7UdJOe0POKpw0DZjy9de8a8Cm4QQ5YJcVqaMrvnDatTptidTkTBAByWqnKEJtH1vRDw/mZHON7Nk7gWHONd0LZcEdx/0ae74DoBZuo7vXXONxPe8YW3uuix/CnjP017nKrYUO25qeWRO2ID7r7AYtcNb6AyfP67aywpF2Oz+BaMUANFhPzNdFTeGy8ODpa5fB8RmwErx51J5pST59SlAJhBwkzXEUgj2OG6luA0atSUsWWuCDSjlgbACXYNZry+sKq51ZEqvV5Fw2G8R2bwTpqkFEGNKARAnjaoWzlWjAjz/rDoY84LHHzrxG7JB7ZmyR0fJ0amDi+wm8PzBEn9PlJ6Nt855zX0QstszVWi+XOeGMVwABZRAtQ2r5btxFkDK0Dq9YEsbVCMqWziRsUnzwpGUCB4CBcZDlbJHDQuy7IYN9bQA1ZTUBXkQw7g+IIGYprmCWiNshFWmF5ndpUrpV12n+SRZNjAhG8SlVpEnM0ylLxHTOE1IelEiY6VnmtDFfNA2YAxpU+DANmSWXRszjJTBdZcqoxmyLqZXyXpmui/0ACMqcVUjOiaRz0Golnwfdf4RWieoaYMF6U2F3S1u9b6ak9EYiXGV85K8EalHIQViXamwFmRnThpBKUhcWmSQDscrIbywSutYqQZxqcZOUFw7aMJxAB9F7a/3sogtzTHE4SOi3swmX8Vzb8ofUgY05zUd28V4K9D0Mh2bfh5vIziByvzhpsFVlRmaWmAVITIUJtcxATeaKYLGWsuT4RpN1IvKI59XqDMM4eHekaFZgF+a1I5P+djWOKNOEO2crcJhbK+JyGZymFjTCsyml9nB6P6cFaaSUkAHkoWIwe/2YUWoC14Rpsjw0wExDlbWij5p5LSesG4ZrQ+akV6KDoMUkcgnpTGu/rlYrmaxhwGN3zpBzxp07Ky3c3vyXMSWp3feynDfBDBkhvTZQww2HYw9R4XaBKozBTXcMDOMAEKFkMXcKwRt0jixHljCOSSXT0HEq6YYMfsScE5glGK6qWRGmbbkUG7QbL/2XGsW771aM+0D92CVJQfdqhAEN7cb0CGCqEpek1aIsCtTN+0RAYRTSaOkqDEVM0gWtvrlpxK1Zhb3CHg9w7cFVRIb6dosyf2s5WBvjraZ9V7ESsJjSrdVGjzWLipYAlUSEISecnQ2zuTwQuiHV00QLlqERAGuy0XUrM6tBJ4zYRpOqSUiq9Wr60LQpyJnVFwyZjxQJcssNleCejVrbJi/O4CZx2xbcP3pceXP2xVv4ha+jyIS7rwPt6HahM+Hrwz6r2eUdly5S48P34ZhbqOaWsFkwZTLmaoU2Uis9zPpd1khmA1mTIsiena2cGUt72UYz7K7CeOVapMy71orVauyQbtkwl8HJg7CcCesnoQlmTtN6UckQTGK6JPYC6VyraqhQiUPSBIqWnqtachKQKNtu1ap52YJ0TAIgwOvC5iRRcaaJDZobuFpZ/8ihK9buwWG0LeWFJ97GA7fIyA4l1wQyIv5AMGdK2xuhnWffawk4oJnv47Ykkw4ZQMa4kgWcClCqRf+aJplck21pXHId14CHhEGtItM0+r0BtApGwSWwTfpUIu+e48FB1mHGkCXtoWbtnFXZ2JdqXXJvrtIxCVyRkgpnNk6SYCzrJyvCpZidOTMSZxBV59OkRcij/DNnwjXaZ6tpfc0EzbWZoFFVy9WoaWJ7AkZCUa2LnSGkiEXTgDUzYBwbA07uOrg+vlU+sEexWXA9F/4ajhiH6gSvpv1SJWXAjJrkJoJ+amU4gVYe1xkwYzNN3p2pTJMKBeaTJku8UAEpdKtidCNuO2pbWN7FVOdMex8cUu7ug73C1RWlV53dLWNclB7Cnp8HhQKtMQKTClAQVxageDdhqWr4o/lpbcTJ+gcnjOOAcRgBkrrgsi1rMz8zwErXLYgzJeVBWSLd3Tyu9cMvgxMz4LZUuFY3YhEEobZRk+xuMAhjSUgFOJsyiAZI9GzxQBsCxJdLEO1KFFr354qUj3AnExfZNSuRbLITjjtnouU+dudMquKMEnBlmm/O0px7GAdT3mXhWOrRZVjg1kjb/N+Hwm8jctQdvtpO3D6pbYbI16N63anaKsErfsl8J1rmjzOGwfL0Jo30bdGpaVArhLWOI7WKsDA1HhmrUvHY42cSWEdanUh9eDlLlPXQRVUHqZbbKCORvg4QEVbjADN9Ek2aFsOYigXlsLTnIw3sQQUTo2rINMFCpy1oJwFkwl0FKCFnsehQSshJzdOad97NHTeBznOy1T0To5urM15JzYAGXAnzUWKnBSQSMZCqLK0keBNXi+FSbj/mhHHMWJ0NeOyxlRc1aJbnQyx0SftpMrxRftrzt/XzgKOKWknTxEwglrRGKvKs4p1XF8bMZM0seaYSVyKBWNB0KGO41gGpMf39667fSbu/t2fgg1lwrgaNGZo6vq257oYm8Pi17BsVJOMlCEYD+r+tVEnbz27l4vZdeO2Iq9LplLKYkzWIlnWYEjtU3GIWaRyDMAZTtjH4VhacrzQdp2XALujov60JaJuZ0IKvJLK0TwWC2fdz7aKOmdrmjKa5qJHJi16fLEjEHPIZeRgk0Gpox8z02cpNUiMs4dKudOmak3ncv626EPqHBXOJOizm7rhrrr0ZPV6EjOhGG/fW2MOWMoGEgFRFumJKYPHa6fVsrhuTd9Tq+vAONSHcfxjktdaKulXTdb6Jt/FwCPJFaGuzanH2nKSLULJGBhwWAwcNTDcwkxWigWu0zQcmWmi18TJQoUVKgODnC5PgwoYJHtUZsUeNVm2V59+H9KJOiAKsl665aSzzTwT+tl5iXWVjvsKAya97XTDBwk28OgvN8BMYQ7AG6S7rNM5GH4QRp6QlUNsJOkvazyow4Fqqx5546pumiHElH2vTONo4fX+EMdqKIB3lXMTdfgJq17kEX9eFjvkivN13e9r+IM8UBfWw16/gf9ulFXc4isw5vHbD0TdGH8zVKMtA1cFuP9kAk8+LMWa2ziUdo74c2TfABG2mAQ6MgLy4RpsUMTuKDwlYrQYnAATpiMK1YpqsApBujtKioM2vw6yEEi3CliAEY9D+nmdnKwzjgNUoAVY5Zdw5ExP0YKXqkkj4QnDJCdDcIuPPGtbVRVNzeOb7YKxllzfHhaROwIvnyRfNuqgan3Je88URAE6AmDZEO66u+KnGlaE3q608nf5+wNAipFna7HlT+irViFJKOBtXmlYwau/mrJpjb9p6YCv9HGeJcOfOStdE0eAPqEtkA9LOTEUjkKmJ20ConiVHtQMREgjSjZyoOLGo2YIKTTvOogFTC/cycsfO7JsG7JUqlOESM8iCq1JLIzK+alqu1D+X6Rg0991bt+n9iICzOyPGccSdx0Y87hpw828fKgwxxobJgch0BclW07p1IDQca3AUAEYR8zMTEjMoSelPIpI1SULIKyYxfaP1mZ6K1SaelO4YA26NYDwNSamPzEE0U1ITULvhN8GhO3cLZs+964wDrfM2pB1M+PJfu2BvP9iW7xux9Ej+HYOPwsBOa+OMmMYhO29ROmIlKJvmCzAnt1r5Ncy0HMdjFc1MUOPWP/giODkDBhAGbXI9gThtbVALG2eGB95wZVT1AW80CnYYMkrRHGFY/1FZ+rEZS6ZWuMCCpoZRqiWNowRVjaMUts9almzIMQWpacBNW+M2yzu0LLTbz8woO867NifeM4D7hjiOPkXMbwO4FtwPm/vzuv1q7SBNCgU8Qobgvn9Xe/0ewffDjNVq9EjTkosXZk9aWL3VdM29NoyZBH0AVBGpD3iQJ2QGhjyAUDFl29RWC1gYb9NvqmrHVXXOoplCpn2aed7wnDzAi4i0VWDtni+uNajJzAUaZb4WWEXmiiEtoKFLOKFpupZWJAyYNOK/NSWxPUCACqoZ4yB/ZjlyX/chGLDJD01ZV13F6lib5mLEPM6zjUL1yAowSZCNpH2pQJHs5yI0FC7uF6za97kEQZ9rrCHPaGZnHYvjyJhMZESNi8hK8Anc/fx2Loctgh6z3ecDCvdzXF54ro3RhPPg/4/j7mC2IffRwznO5KczrbX7Lt6ixYdYC8rY1tCwt5UyGri9vNUPWoDFeNllNPzk7QjdROAbxIJ4jED0Yo0FOYkPkcFjAldhmquzETnbQ/ca8KR+OGuBRCBkSh1RSylhHDTi+WwlzHcccOdshZwIKw3GkvSXGOHcdv/OtgZh8nb5JnYtlusz3zlQQH5/r/u+0pX5Oof/ReuwO4ufX7VlQgtOUcbr97GD+htLXTFfuRAxrQVO5L436YySMGgnE9GAsxdHacFyreTjYYA0OBDKHNgT8oX5G4PkYEVQrCiiOjcEA64Z21aw9aK+J2QjCmJJiJHfhkPzUwHQfF2APKo5mpitXCLUtNz6++akWnCCp9gZA87qEoimau+fG3rnpmy4PiQniJdr0c7+daAplYOM50yAnUEAEsxjZnVokFurYiZ9lcVcX90MXarRm+q4Nnw3d4Lz3gs2kXBSr451yWPv2sqN2baASJrRpuvBxZrF1rgDo77IVb3FiPVArbI+SddX9PHavO6guth3qOM/ML8tA9Omlbj1UrfNHtV+pIqibVuejZkjE74YTtyOEGGwMdy7ugYcy6z1ydMqltIAgCTtQ6tiGXFlbikAFggDJkmqhxTUsILaFrU2jgMSEc7ORoyDlCWTwtzSpFkiamXzNuZtsnFbQkT9HtsldMTPkQkflvnu2izXYb6mX1zhtt1jMLq9Z8Qo6QK2VzMHhyAGZ7/OWCyvmsCJkZPm+Gk1mqIdcSSKXdaDWS9s7ZjFg5ThH8wEraYsdzUwMAw2loSi67KzThkBsN2seyJqdI0MKDXjBGSLglZCm9hN0K3RgQaeefcfdt0QqEGAlLk1g6gwWLmVFCMjP5aTMF7RgM0CJHnvdh0iqAUpqyAbGLBZfQ7EhNmJib1uT6ZpokasyQivP3tTMqmaptoK/FQNlBNmq/nrQctt/mCdLWfAtnrlfzfb7skHYn2gmUyx+7xdH8J82qKO/x8G43vwu/f0biNfLDSEAwQTQnWl1BLoZLvGvpSoC4br9/MALgC1rrHZaervj7lyaJsznqFKwVX7Xp+oEpYh9zrLoV3H3wU8uYakC9JMcu2ns4CccK2WM9iY/nYw0u5FuLUY4ucrPOMT2mYAAAiFSURBVO6hNd8+dP8yLjPbGddTlC8YlF18djhsri0W381tf7yf63aM4hzvGceWaelgQOF6MwLUHw2Y2Eb4/Lt4xtZ6sy+VYDVT9S5m17Ts/Us2In3XyLGFX9rxE5r/1tFxGHzvXqYNV/uwunU8HHCthuFpRwh4bALWHLNzZWIubO/B7wVPMof5THb5vbNLOJ3tmPF14eoxE23VAm5/NiHERjTfCDyfn27Vd1dv65/DaYRtLKH1u9hDYk34sbgSG16fLrdrHHrrsCb6i+4HOryp84KbEX0SwNMAfvNoNz08PA/HHf+LmPnzHuSHC74fCB4Y38CC8weEZY0v+D42HBPne/F9VAYMAET0DmZ+xVFvekB41Mb/qI13Do/i+B/FMUd41Mb/qI13Do/a+B+18e6Cm/IM129HssACCyywwAIL3DcsDHiBBRZYYIEFTgCnYMA/eIJ7HhIetfE/auOdw6M4/kdxzBEetfE/auOdw6M2/kdtvLvgRjzD0X3ACyywwAILLLDAYoJeYIEFFlhggZPAURkwEf1xIvogEf0SEb3+mPd+ECCiFxLRfyGi9xPR/yai1+nx5xLRzxDRh/T1c0491l2w4Pu4sOD7+LDg/Liw4PvAEKtQPcw/ABnAhwH8LgArAO8G8LJj3f8Bx/wCAF+u7z8bwC8CeBmA7wbwej3+egDfdeqxLvg++dgXfC84v9U4X/B9+L9jasCvBPBLzPwRZl4D+NcAvv6I979vYOZfZ+Z36vtPA3gSwBdAxv2jetqPAvhTpxnhhbDg+7iw4Pv4sOD8uLDg+8BwTAb8BQB+JXz+uB57JICIngDw+wD8dwDPZ+Zf168+AeD5JxrWRbDg+7iw4Pv4sOD8uLDg+8CwBGFdAYjoswD8WwB/h5l/O37HYsNYQskPCAu+jwsLvo8PC86PCzcV38dkwL8K4IXh8xfqsRsNRDRCJu4NzPxmPfx/iegF+v0LAPzGqcZ3ASz4Pi4s+D4+LDg/Liz4PjAckwH/DwAvIaIvJqIVgG8E8JNHvP99A0mrjR8G8CQz/5Pw1U8CeI2+fw2Af3/ssV0BFnwfFxZ8Hx8WnB8XFnwfGo4ckfYqSBTahwF8+ymizu5zvF8FMU28B8C79O9VAD4XwNsAfAjAfwLw3FOPdcH36f8WfC84v+04X/B92L+lEtYCCyywwAILnACWIKwFFlhggQUWOAEsDHiBBRZYYIEFTgALA15ggQUWWGCBE8DCgBdYYIEFFljgBLAw4AUWWGCBBRY4AdxqBkxEX01EP6Xv/+RF3TuI6HcQ0d8Inz+fiH7iQON4ORH9AhG9i4jeQUSvPMR1byLcIJx/g3Y/qUT0ikNc8ybCTcF3uOa3EhET0fMOed2bAjcF30T0RqUn7yKijxLRuw5x3ZsINwjn/5CI3qM4fysRff61L3rqPK0HzO3KVzzvqwH81BXPfQLA+x7SeN8K4E/o+1cBePupcfgZgPMvA/C7AbwdwCtOjb/bjm+9/gsBvAXALwN43qlxeNvxHe7zvQC+49Q4vO04B/Ds8P5vA/gX173mjdOAiegJIvoAEb2BiJ4kop8gosdVyvsuInongG8goq8jop8noncS0Zu01qf1q/yAnvdnwnVfS0Tfr++fT0T/jojerX9/EMB3AnixSjffo+N4n55/h4j+FRG9l4j+FxF9Tbjmm4noP5L0lfzuPY/FAJ6t758D4NceBu4eFG4jzpn5SWb+4MPF3IPBbcS3wj8F8G24YXWMbzG+rdLTnwfw4w8Ddw8KtxHn3NeQfhYOsc5PLQXtkWAYwFfq5x8B8HcBfBTAt+mx5wH4rwCepZ//HoDvAHAH0q3jJQAIwL+BSk4AXgvg+/X9GyFFuQHpcfkczCSn+BnAtwL4EX3/pQA+pvd6LYCP6O/vQCT/F+p5PwTVvCDa2Md0bL8K4EWnxvNtx3m45tvnx079dxvxDWnv9s/0/UdxgzTg24jvcM0/BOAdp8bxZwrOAfwjHdv7AHzedfF04zRghV9h5p/T9z8GKScGCMIB4A9Amir/HInv4zUAXgRB6v9h5g+xYOvH9lz/awH8AAAwc2HmT10ynq+yazHzByAT9FL97m3M/Clmvgfg/ToOMPNfZuZ36Dl/HcC3MPMLAXwLpDbpTYPbhvObDrcG30T0OIC/DyGeNxVuDb5n13k1bpj2G+DW4ZyZv13p+BsAfPNlCLgMhute4CHBXLW3z0/rKwH4GWZ+dTyJiF7+sAe2A87D+4LdOH0NgNfp+zdBpKqbBrcN5zcdbhO+XwzgiwG8Wyyi+EIA7ySiVzLzJ44zxEvhNuEbAEBEA8Q8+/uPMagHgFuH8wBvAPDTAP7BdW56UzXgLyKir9D3fwHAz86+/wUAX0lEXwIARPQsInopgA8AeIKIXqznvRq74W0QrRRElInoOQA+DeCz95z/3wD8RT3/pQC+CMD9+Bd/DcAf1vdfCykAftPgtuH8psOtwTczv5eZfyczP8HMT0AatX/5DWK+wC3Cd4A/CuADzPzx+/zdseBW4ZyIXhI+fr2O81pwUxnwBwH8TSJ6EsDnQM0MBsz8SYjd/seJ6D0Afh7Al6r54K8C+A/qvN/X4/F1AL6GiN4L4H8CeBkz/xbEFPI+Ivqe2fn/HEDS898I4LXMfI4LgIh+iFr6y18B8L1E9G4A/1jHeNPgVuGciP40EX0cwFfo2N5yNTQcDW4Vvh8BuI34/kbcXPMzcPtw/p163fcA+Do0q+YDw43rhkRET0Ac7r/nxEP5jIEF58eFBd/HhQXfx4cF51eDm6oBL7DAAgsssMCthhunAS+wwAILLLDAZwIsGvACCyywwAILnAAWBrzAAgsssMACJ4CFAS+wwAILLLDACWBhwAsssMACCyxwAlgY8AILLLDAAgucABYGvMACCyywwAIngP8PRP6VhpR28eEAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x576 with 5 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2Lx68VA7z_HE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NxKa37PWz_HL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "014aa316-b3b7-4ace-cb34-21b05a954663"
      },
      "source": [
        ""
      ],
      "execution_count": 150,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1, 32, 32, 1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 150
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BLORj1fNzYGN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}