Classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 2
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython2",
      "version": "2.7.6"
    },
    "pycharm": {
      "stem_cell": {
        "cell_type": "raw",
        "source": [],
        "metadata": {
          "collapsed": false
        }
      }
    },
    "colab": {
      "name": "Copy of main.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": false,
        "id": "WdfvvDpEI27_",
        "colab_type": "text"
      },
      "source": [
        "# Intro to Deep Learning\n",
        "#### Jordan DeKraker & Haider Al-Tahan\n",
        "\n",
        "This is the practical component of this tutorial will focus on hands-on experience with Tensorflow. Topics will include:\n",
        "\n",
        "1.   Introduction to Tensorflow 2.0 and Keras. \n",
        "2.   Train and evaluate  of dense and convolutional neural networks for real problems.\n",
        "3.   An open session for exploring and adapting existing NNs or for helping attendees build their own networks.\n",
        "\n",
        "You can use [Google Colaboratory]('http://colab.research.google.com') notebook file which is a [Jupyter notebook]('https://jupyter.org/') with free limited access to a GPU. Jupyter notebooks allows python programs to run directly in the browser which can be great way to learn and use TensorFlow.\n",
        "\n",
        "If you are using Google colab notebook and want to utilize the free GPU for training, do the following:\n",
        "\n",
        "1. At the top-left of the of the menu bar, Select `Runtime` > `Change runtime type`. \n",
        "2. Click on `Hardware accelerator` > `GPU`.\n",
        "3. At the top-right of the menu bar, select `CONNECT` to initialize your python interpreter.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%% md\n",
          "is_executing": false
        },
        "id": "0_CkgEzjI28A",
        "colab_type": "text"
      },
      "source": [
        "# TensorFlow Installation\n",
        "Note: you can also use this notebook locally on your machine. Just make sure that you have TensorFlow 2.0.\n",
        "When installing TensorFlow on your local machine, you also need to make sure that you have the right CPU or GPU version depending your machine.\n",
        "* If you have a GPU with CUDA and Cudnn or using google colab installed then you just do `pip install tensorflow-gpu==2.0.0` in your terminal.\n",
        "* If you don't have CUDA and Cudnn installed but have a GPU then follow [these instructions]('https://www.tensorflow.org/install/gpu').\n",
        "* Otherwise you can use your CPU and install TensorFlow by running `pip install tensorflow==2.0.0` in your terminal."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "pycharm": {
          "name": "#%%\n",
          "is_executing": false
        },
        "id": "w75RfYSoI28B",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "5d89624f-aca5-4336-c451-e8a602d367c4"
      },
      "source": [
        "# Without GPU\n",
        "# !pip install tensorflow==2.0.0\n",
        "# With GPU \n",
        "# !pip install tensorflow-gpu==2.0.0\n",
        "# !pip install matplotlib\n",
        "# If using colab\n",
        "%tensorflow_version 2.x"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TensorFlow 2.x selected.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": false,
        "id": "Fp7YwNbCI28E",
        "colab_type": "text"
      },
      "source": [
        "# Imports"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "pycharm": {
          "name": "#%%\n",
          "is_executing": false
        },
        "id": "VWEtvKfpI28E",
        "colab_type": "code",
        "outputId": "5ddad763-e549-4323-a6a0-0537e946dd18",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras.models import Sequential \n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "import seaborn as sns\n",
        "from matplotlib.lines import Line2D\n",
        "print(tf.__version__)\n",
        "%matplotlib inline"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.0.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DHTJ3Y-jjN9d",
        "colab_type": "text"
      },
      "source": [
        "# Neural Networks Basics"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gsIyAkw3jT_h",
        "colab_type": "text"
      },
      "source": [
        "![alt text](https://miro.medium.com/max/465/1*fX2hHozXqOOnj_bMmQpMIg.jpeg)\n",
        "[image source](https://becominghuman.ai/artificial-neuron-networks-basics-introduction-to-neural-networks-3082f1dcca8c)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gEzhhSwUjZ34",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def CtoF(c):\n",
        "    return c * 1.8 + 32\n",
        "\n",
        "xtrain = list(range(1, 1000))\n",
        "ytrain = map(CtoF, xtrain)\n",
        "\n",
        "\n",
        "model = Sequential()\n",
        "model.add(keras.layers.Dense(1, input_shape=[1]))\n",
        "model.summary()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7iyq6cUejZPj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer=tf.keras.optimizers.Adam(0.1), loss='mean_squared_error')\n",
        "history = model.fit(np.array(xtrain), np.array(list(ytrain)), epochs=10)\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.plot(history.history['loss'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DVwRnlB7jl-B",
        "colab_type": "text"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FcZSq4SMje8G",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "xtest = list(range(0,100))\n",
        "ytest = map(CtoF, xtest)\n",
        "\n",
        "y_pred = model.predict(xtest)\n",
        "plt.xlabel('C')\n",
        "plt.ylabel(\"F\")\n",
        "plt.plot(xtest, list(ytest))\n",
        "plt.plot(xtest, y_pred)\n",
        "plt.legend(['True value', 'Predicted Value'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%% md\n"
        },
        "id": "MYYrObBLI28H",
        "colab_type": "text"
      },
      "source": [
        "# Dataset\n",
        "For the first example we will use varous dataset. Keras provide us with basic image datasets that we could load with ease. Here is few that we are going to test:\n",
        "\n",
        "1.   MNIST -> this dataset consist of 60,000 (28x28) grayscale images for training and 10,000 images for testing. There are two variations of this dataset MNIST contain images in 10 classes of hand written digits. The other version of MNIST is called Fashion MNIST that consist of  60,000 (28x28) grayscale images of 10 fashion categories.\n",
        "\n",
        "2.   CIFAR -> this dataset consist of 50,000 (32x32) color images for training and 10,000 images for testing. There are two variations of this dataset CIFAR10 contain images in 10 classes (6000 images per class). While, the CIFAR100 contain images in 100 classes (600 images per class).\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "pycharm": {
          "name": "#%%\n",
          "is_executing": false
        },
        "id": "6s5FmZhyI28I",
        "colab_type": "code",
        "outputId": "cbca1727-74f5-4c8f-b59f-cd06d50404ec",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from keras.datasets import cifar10, mnist, cifar100, fashion_mnist\n",
        "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
        "x_val, x_test, y_val, y_test = train_test_split(x_test, y_test, test_size=0.5, random_state=1)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3AYSRAZXJLzq",
        "colab_type": "text"
      },
      "source": [
        "# Visualize Dataset\n",
        "\n",
        "The images are NumPy arrays, with pixel values ranging from 0 to 255. The labels are an array of integers, ranging from 0 to N. Each integer correspond to different image class.\n",
        "\n",
        "For our data variable (`x_train`), if we print the shape of our training data variable we will get a shape that corresponds to (number of samples, width of images, height of images, number of channels). In this case, 3 dimensional channel it means that we are dealing with colored images. \n",
        "\n",
        "For our labels variable (`y_train`), we same number of labels as our `x_train` data. Hence, if we print out the shape we get a matrix of shape: (number of samples, features).\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jwxj4XqoJo75",
        "colab_type": "code",
        "outputId": "18164ff6-c62c-480e-d717-e469c941b670",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        }
      },
      "source": [
        "number_of_classes = y_train.max() - y_train.min() + 1\n",
        "epochs = 10\n",
        "print(f\"Training Matrix Shape for data: {x_train.shape}\")\n",
        "print(f\"Training Matrix Shape for labels: {y_train.shape}\")\n",
        "print(f\"Number of Classes: {number_of_classes}\")\n",
        "print(f\"Validation Matrix Shape for data: {x_val.shape}\")\n",
        "print(f\"Validation Matrix Shape for labels: {y_val.shape}\")\n",
        "print(f\"Testing Matrix Shape for data: {x_test.shape}\")\n",
        "print(f\"Testing Matrix Shape for labels: {y_test.shape}\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training Matrix Shape for data: (60000, 28, 28)\n",
            "Training Matrix Shape for labels: (60000,)\n",
            "Number of Classes: 10\n",
            "Validation Matrix Shape for data: (5000, 28, 28)\n",
            "Validation Matrix Shape for labels: (5000,)\n",
            "Testing Matrix Shape for data: (5000, 28, 28)\n",
            "Testing Matrix Shape for labels: (5000,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1-pi7r-9LhBw",
        "colab_type": "code",
        "outputId": "d740029a-9cb1-4390-f0bb-cc974b67b756",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "number_of_samples = 5\n",
        "plt.figure(figsize=(number_of_classes,number_of_samples*10))\n",
        "image_shape = x_train.shape[1:]\n",
        "\n",
        "c = 1\n",
        "for i in range(number_of_classes):\n",
        "    random_samples =  np.random.choice(np.argwhere(y_train == i)[:,0], number_of_samples)\n",
        "    for j in range(number_of_samples):\n",
        "        plt.subplot(number_of_classes,number_of_samples,c)\n",
        "        plt.xticks([])\n",
        "        plt.yticks([])\n",
        "        plt.grid(False)\n",
        "        plt.imshow(x_train[random_samples[j]], cmap=plt.cm.binary)\n",
        "        c += 1\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAj8AAAopCAYAAAB+zBfWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n07OMPhxtAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x3600 with 50 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gyHGQ2yg0CfT",
        "colab_type": "text"
      },
      "source": [
        "# Model Construction\n",
        "\n",
        "We often represent neural networks as a set of layers. Each layer has its own number of units, activation function, etc... For this example, we construct a very basic model consisting of only fully connected layers.\n",
        "\n",
        "A fully connected layer (called dense layer in Keras) require a one-dimentional input. Hence, the first layer of our network only transforms the 2d image to a one-dimensional and has not learning parameters. Resulting in an array of $height \\times width \\times channels = Array Size$.\n",
        "\n",
        "We have chosen 256 nodes (or units) for our first hidden fully connected layer. Increasing the number of nodes, increases complixity and computation time. Sometime it is more advantagous to increase complixity by increasing the number of layers rather than the number of nodes.\n",
        "\n",
        "The last layer consist of 10-nodes that return an array of 10 probability scores that sums to 1. Each node contains a score that indicates the probability that the current image belongs to one of the 10 classes."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "40JPm_AcjsfE",
        "colab_type": "text"
      },
      "source": [
        "![alt text](https://miro.medium.com/max/844/1*Cc0LQodOm_vmYEm-DgTV8g.png)\n",
        "[image source](https://towardsdatascience.com/coding-up-a-neural-network-classifier-from-scratch-977d235d8a24)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ctQap6x_F1UN",
        "colab_type": "text"
      },
      "source": [
        "![alt text](https://svgshare.com/i/G7H.svg)\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OQPaDxtW0Htf",
        "colab_type": "code",
        "outputId": "d8498abd-8610-4f9b-edee-86d7946a7fee",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255
        }
      },
      "source": [
        "model = Sequential()\n",
        "model.add(keras.layers.Flatten(input_shape=image_shape))\n",
        "model.add(keras.layers.Dense(64, activation='relu'))\n",
        "model.add(keras.layers.Dense(number_of_classes, activation='softmax'))\n",
        "model.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "flatten (Flatten)            (None, 784)               0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 64)                50240     \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 50,890\n",
            "Trainable params: 50,890\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2QT-kVay0tOJ",
        "colab_type": "text"
      },
      "source": [
        "# Compile the Model\n",
        "\n",
        "Before training we need to compile our model. When compiling our model we also need to specify few parameters:\n",
        "\n",
        "1.   Optimizer: Optimizers update the weight parameters to minimize the loss function. More optimizers can be found [here]('https://keras.io/optimizers/').\n",
        "\n",
        "2.   Loss Function: acts as guides to the terrain telling optimizer if it is moving in the right direction. The less our loss is the less difference there is between the predicted output and the actual output. More Loss Functions can be found [here]('https://keras.io/losses/').\n",
        "\n",
        "3.   Metrics: Used to monitor the training and testing steps. More Loss Functions can be found [here]('https://keras.io/metrics/')."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aJ4UGk480vWA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# adam_op = keras.optimizers.Adam(learning_rate=0.001)\n",
        "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CWiajd1S02qx",
        "colab_type": "text"
      },
      "source": [
        "# Train the Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4gwwT1R802NI",
        "colab_type": "code",
        "outputId": "e3cf5908-447d-41af-933f-675d19a9f2ff",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        }
      },
      "source": [
        "model.fit(x_train, y_train, epochs=epochs, validation_data=(x_val, y_val))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 60000 samples, validate on 5000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 5s 87us/sample - loss: 2.2259 - accuracy: 0.7861 - val_loss: 0.6823 - val_accuracy: 0.8390\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 4s 71us/sample - loss: 0.4770 - accuracy: 0.8861 - val_loss: 0.4428 - val_accuracy: 0.9028\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 4s 73us/sample - loss: 0.3398 - accuracy: 0.9139 - val_loss: 0.3164 - val_accuracy: 0.9246\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 4s 73us/sample - loss: 0.2818 - accuracy: 0.9266 - val_loss: 0.3235 - val_accuracy: 0.9282\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 5s 80us/sample - loss: 0.2503 - accuracy: 0.9351 - val_loss: 0.3048 - val_accuracy: 0.9320\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 5s 78us/sample - loss: 0.2277 - accuracy: 0.9406 - val_loss: 0.2645 - val_accuracy: 0.9326\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 4s 71us/sample - loss: 0.2166 - accuracy: 0.9435 - val_loss: 0.2757 - val_accuracy: 0.9308\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 4s 74us/sample - loss: 0.2108 - accuracy: 0.9455 - val_loss: 0.2333 - val_accuracy: 0.9412\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 4s 74us/sample - loss: 0.2087 - accuracy: 0.9454 - val_loss: 0.2624 - val_accuracy: 0.9412\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 4s 73us/sample - loss: 0.2017 - accuracy: 0.9485 - val_loss: 0.2428 - val_accuracy: 0.9426\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f35bc524f98>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CWlnrij83JLh",
        "colab_type": "code",
        "outputId": "27dee9fd-6c21-4ef9-f378-6462ba1e691b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 351
        }
      },
      "source": [
        "fig, axs = plt.subplots(1,2,figsize=(15,5))\n",
        "epochs_range = range(len(model.history.history['loss']))\n",
        "axs[0].plot(epochs_range, model.history.history['loss'])\n",
        "axs[0].plot(epochs_range, model.history.history['val_loss'])\n",
        "axs[0].set_xlabel('Epoch')\n",
        "axs[0].set_ylabel('Loss')\n",
        "axs[0].legend(['loss', 'val_loss'])\n",
        "\n",
        "axs[1].plot(epochs_range, model.history.history['accuracy'])\n",
        "axs[1].plot(epochs_range, model.history.history['val_accuracy'])\n",
        "axs[1].set_xlabel('Epoch')\n",
        "axs[1].set_ylabel('Accuracy')\n",
        "axs[1].legend(['accuracy', 'val_accuracy'])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f35a6658d30>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA34AAAE9CAYAAABUerD/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeXjU1dn/8fedjYTsgUBC2FFJ2AQJ\nIFpRsSpVKkp/itQFcKG2Lm1t69Naqz4ubZ/W7rUqVQTcqUrrQrVatVQFZZFFSUSILIEEQlZCCNnO\n748ZwhgTmJCZTBI+r+uaKzNnzvnOPXFMcnPOuY855xAREREREZGuKyzUAYiIiIiIiEhwKfETERER\nERHp4pT4iYiIiIiIdHFK/ERERERERLo4JX4iIiIiIiJdnBI/ERERERGRLi4i1AEEUs+ePd3AgQND\nHYaIiATZ6tWr9zrnUkMdR2eh348iIsePln5HdqnEb+DAgaxatSrUYYiISJCZ2bZQx9CZ6PejiMjx\no6XfkVrqKSIiIiIi0sUp8RMREREREenilPiJiIiIiIh0cV1qj5+ISGdRW1tLfn4+1dXVoQ6lQ4uO\njqZv375ERkaGOpQuR5/BjkWfdREJtqAlfmbWD1gE9AYcMM8594cmfa4A/gcwYB/wbefcOu9zW71t\n9UCdcy47WLGKiLS3/Px84uPjGThwIGYW6nA6JOccxcXF5OfnM2jQoFCH0+XoM9hx6LMuIu0hmEs9\n64AfOOeGAacCN5rZsCZ9PgfOdM6NBO4F5jV5/mzn3GglfSLS1VRXV9OjRw/9wX0EZkaPHj00IxUk\n+gx2HPqsi0h7CNqMn3OuACjw3t9nZjlABrDRp8/7PkNWAH2DFY+ISEejP7iPTt+j4NL3t+PQfwsR\nCbZ2Ke5iZgOBMcAHR+h2LfBPn8cO+JeZrTazuUe49lwzW2Vmq4qKigIRrojIcSEuLi7UIYiIiEg7\nCXpxFzOLA14Avuecq2ihz9l4Er+v+DR/xTm308x6AW+YWa5zblnTsc65eXiXiGZnZ7uAvwERERHp\n1Orq6oiIUD07ETm+BXXGz8wi8SR9TznnXmyhzyjgUWCac674ULtzbqf36x5gCTA+mLE653h53S6W\nbyk+emcRkS7EOcePfvQjRowYwciRI3nuuecAKCgoYNKkSYwePZoRI0bw3//+l/r6embPnt3Y93e/\n+12Io5fO7uKLL2bs2LEMHz6cefM8W/1fe+01TjnlFE4++WTOOeccACorK5kzZw4jR45k1KhRvPDC\nC8AXZ66ff/55Zs+eDcDs2bO54YYbmDBhArfddhsffvghEydOZMyYMZx22ml8+umnANTX1/PDH/6Q\nESNGMGrUKP70pz/x1ltvcfHFFzde94033uCSSy5pj2+HiBxHnHOUV9Xyya5y/vVJIU8s3xrU1wtm\nVU8DHgNynHO/baFPf+BF4Crn3Caf9lggzLs3MBY4D7gnWLF6X5P/ey2XMf2TmTikRzBfSkSkQ3nx\nxRdZu3Yt69atY+/evYwbN45Jkybx9NNPc/755/PTn/6U+vp6qqqqWLt2LTt37uTjjz8GoKysLMTR\nS2c3f/58UlJSOHDgAOPGjWPatGlcf/31LFu2jEGDBlFSUgLAvffeS2JiIhs2bACgtLT0qNfOz8/n\n/fffJzw8nIqKCv773/8SERHBm2++ye23384LL7zAvHnz2Lp1K2vXriUiIoKSkhKSk5P5zne+Q1FR\nEampqTz++ONcc801Qf0+iEjX09Dg2Ft5kPyyA+wsPcDOZr5WHqz7wpiLx2QQHx2cY12Cue7hdOAq\nYIOZrfW23Q70B3DOPQzcCfQA/uLd1Hzo2IbewBJvWwTwtHPutSDGCkBmWgK5Bc2uRhURCZr/ffkT\nNu4K7M+eYX0SuOvrw/3q++677zJz5kzCw8Pp3bs3Z555JitXrmTcuHFcc8011NbWcvHFFzN69GgG\nDx5MXl4eN998MxdeeCHnnXdeQOOW0AjlZ/CPf/wjS5YsAWDHjh3MmzePSZMmNR5rkJKSAsCbb77J\ns88+2zguOTn5qNe+9NJLCQ8PB6C8vJxZs2bx2WefYWbU1tY2XveGG25oXAp66PWuuuoqnnzySebM\nmcPy5ctZtGiRv29dRI4TdfUNFJRXN5vQHbrV1DV8YUxCdAQZyd3pl9KdiUN60Dc5hoykGDK8X+O6\nBS89C2ZVz3fxnM93pD7XAdc1054HnByk0FqUlR7P25/uobq2nujI8PZ+eRGRDmXSpEksW7aMV199\nldmzZ3Prrbdy9dVXs27dOl5//XUefvhhFi9ezPz580MdqnRS77zzDm+++SbLly+ne/funHXWWYwe\nPZrc3Fy/r+FbDbPpcQixsbGN93/2s59x9tlns2TJErZu3cpZZ511xOvOmTOHr3/960RHR3PppZdq\nj6DIcai6tv4LyVx+adUXErzCimoamlQYSY3vRkZSDMP6JHDesN6NCd2hr8GazfOHfor5yExLoL7B\nsXlPJSMyEkMdjogcJ/ydmQuWM844g0ceeYRZs2ZRUlLCsmXL+PWvf822bdvo27cv119/PQcPHmTN\nmjVccMEFREVF8Y1vfIOhQ4dy5ZVXhjR2CYxQfQbLy8tJTk6me/fu5ObmsmLFCqqrq1m2bBmff/55\n41LPlJQUzj33XB588EF+//vfA56lnsnJyfTu3ZucnByGDh3KkiVLiI+Pb/G1MjIyAFiwYEFj+7nn\nnssjjzzC2Wef3bjUMyUlhT59+tCnTx/uu+8+3nzzzaB/L0Sk/ZUfqPVJ5KoOz9R52/ZW1nyhf3iY\nkZYQTUZyDKcO7vGlpK5PUkyHnjxS4ucjM93zyyKnoEKJn4gcNy655BKWL1/OySefjJnxq1/9irS0\nNBYuXMivf/1rIiMjiYuLY9GiRezcuZM5c+bQ0OBZuvKLX/wixNFLZzZlyhQefvhhsrKyGDp0KKee\neiqpqanMmzeP6dOn09DQQK9evXjjjTe44447uPHGGxkxYgTh4eHcddddTJ8+nV/+8pdMnTqV1NRU\nsrOzqaysbPa1brvtNmbNmsV9993HhRde2Nh+3XXXsWnTJkaNGkVkZCTXX389N910EwBXXHEFRUVF\nZGVltcv3Q6Sja2hwVFTXUrK/htKqGkr311JSVUPp/hpKqmoo83lcUV2Lc3BoUt4wfI+rPDRbb/j0\nMU+/w/cbOzfe9203b7vva9B4v/nXNoPiyhp2lh5gX5P9dd0iwhoTuaz0BDKSYuibEkNGUncykmPo\nHd+NiPB2OQ0vKMy5rnMCQnZ2tlu1atUxj69vcAy/6zWumDCAn00dFsDIRES+KCcnR39M+qm575WZ\nrfbuCRc/NPf7UZ/Bo7vpppsYM2YM1157bbu8nv6bSHtqaHDsq66jpKrGk8gdSuaqaijZX9vkcQ2l\nVbWUVdV8aWnjIVHhYSTHRpLcPYqU2CgSoiMbky3nwOF87h9up9l294U+vu18od1nbAuvQTPtSTGR\nnr11yYeTuoykGHrGRX1h+Xhn1dLvSM34+QgPM4b2jie3UAVeREREjmdjx44lNjaW3/zmN6EOReSo\nDiVxpVU1h2fg9tdQVlXb4uPSIyRxkeHWmMAld49iaFr8Fx77JniHvnaPCu8SSVNXpsSvicy0BN7I\n2Y1zTh9eERGR49Tq1atDHYIIALX1DeyuqKagvJpd3j1ou8oOUFBW7d2HdpDSqlrqW8jiIsKM5Ngo\nUrwJ24m94nweR5HcPbLxcUpsFEndI4nrFqG/g7sgJX5NZKbH89yqHRRVHqRXfHSowxERERGRLso5\nR1lV7eFkrmlyV17N7mYqRybGRNInKYa+yTGM6Z90eCbu0CycN5FLio0kXklcx1FfCwf3wcEK71ff\nm7fttFsgSP+9lPg1kZmWAEBuwT4lfiIiEjRmNgX4AxAOPOqc+2WT5wcA84FUoAS40jmX7/N8ArAR\n+Ltz7qZ2C1xE/FZdW9/sTN2ucs/jgrJqDtTWf2FMVHgY6UnR9EmM4bQhPemTFE0fb8XIjKRo0hNj\niA3iWW/SjEMJW3V5CwlbC4lcdZP2ugNHf61x10NU96C8DX1qmshM81T2zC2sYNJJqSGORkREuiIz\nCwceBM4F8oGVZvaSc26jT7cHgEXOuYVmNhn4BXCVz/P3AsvaK2YR+aKGBkdR5UF2lR1gV9nh5K6g\n/PDj4v01XxqXGt+NPonRDO0dz9lDe3mSusTDyV2P2CjCwjr5DN3BfbD1PWioDXUkX+Yc1FUfTtaa\nJmeNSZxPe1310a9r4dAtHqIToFuC535cL+gx5PDjxq/xPn2btEfGBO2tK/FrIjk2irSEaHIL9oU6\nFBER6brGA5udc3kAZvYsMA3PDN4hw4BbvfffBv5+6AkzGwv0Bl4DVN1UJEjq6hvILdzHhp3l5JdW\nscu7r66g/ACF5dXU1n9xDWZsVHhjAjciI4E+iTGNj/skRZOWGE23iI57zlubFX4Mqx6D9Yuhpvmj\nVTocC2+SgCVAXG/ocUKTBK1J0tYtHrolfjFh6+BLapX4NSMzPZ6cQiV+IiISNBnADp/H+cCEJn3W\nAdPxLAe9BIg3sx5AKfAb4Ergq8EPVeT4UVhezdodpXy0vYyPtpexfmcZ1bWec0sPHd7dJymaU/on\nk57oWXrZmNglxpAQcxzup6s7CBtfgpWPwo4VEBENw6fD6JkQkxzq6JoXEXM42YuI7vAJW6Ao8WtG\nZloC723Oo7a+gchOfEijiEigxMXFtXgw9tatW5k6dSoff/xxO0fV5f0Q+LOZzcazpHMnUA98B1jq\nnMs/0h+YZjYXmAvQv3//oAcbbEf6DIoci+raejbsLOej7aWs3eFJ9ArKPUv6IsON4X0SmTm+P6P7\nJTG6XxJ9k7sT3tmXYAZS6TZY/TiseQKq9kLKYDjvfhj9TeieEuropBlK/JqRlR5Pbb1jS1FlY7EX\nERGRANoJ9PN53Nfb1sg5twvPjB9mFgd8wzlXZmYTgTPM7DtAHBBlZpXOuR83GT8PmAeeA9yD9k6O\nM3V1dURE6M+nzsY5x9biKj7a7pnNW7ujjJyCCuq85TL7JseQPTCFMf2SGN0/iWHpCURHduElmceq\noR42v+mZ3fvsDc9M2dALYNy1MOgsCNOESUemn1zN8K3sqcRPRLqiH//4x/Tr148bb7wRgLvvvpuI\niAjefvttSktLqa2t5b777mPatGmtum51dTXf/va3WbVqFREREfz2t7/l7LPP5pNPPmHOnDnU1NTQ\n0NDACy+8QJ8+fbjsssvIz8+nvr6en/3sZ8yYMSMYb7cjWgmcaGaD8CR8lwPf9O1gZj2BEudcA/AT\nPBU+cc5d4dNnNpDdNOnrDAL5GaysrGTatGnNjlu0aBEPPPAAZsaoUaN44okn2L17NzfccAN5eXkA\nPPTQQ/Tp0+cLM9cPPPAAlZWV3H333Zx11lmMHj2ad999l5kzZ3LSSSdx3333UVNTQ48ePXjqqafo\n3bs3lZWV3HzzzaxatQoz46677qK8vJz169fz+9//HoC//vWvbNy4kd/97nfB+LaKV/mBWtZ5Z/E+\n2uGZ0Sur8hQaiY0KZ1TfJOZOGsyY/smM7pdEany3EEfcwVUWwUeLYNUCKN8OcWlw5m1wyixIzAh1\ndOInJX7NGJwaS2S4kVNYwcXowywiQfbPH0PhhsBeM20kfO2XLT49Y8YMvve97zX+0b148WJef/11\nbrnlFhISEti7dy+nnnoqF110Uav2qzz44IOYGRs2bCA3N5fzzjuPTZs28fDDD/Pd736XK664gpqa\nGurr61m6dCl9+vTh1VdfBaC8vLxt77kTcc7VmdlNwOt4jnOY75z7xMzuAVY5514CzgJ+YWYOz1LP\nG4MWUCf/DEZHR7NkyZIvjdu4cSP33Xcf77//Pj179qSkpASAW265hTPPPJMlS5ZQX19PZWUlpaWl\nR3yNmpoaVq1aBUBpaSkrVqzAzHj00Uf51a9+xW9+8xvuvfdeEhMT2bBhQ2O/yMhI7r//fn79618T\nGRnJ448/ziOPPOL3t1GOrq6+gU27K/mocW9eKVuK9gOeCakTe8Vx/rA0RvdPYkz/JE7sFa8lm/5w\nDrav8MzubfyHp0LnoElw3r2QeSGER4Y6QmklJX7NiAwP44Re8arsKSJd1pgxY9izZw+7du2iqKiI\n5ORk0tLS+P73v8+yZcsICwtj586d7N69m7S0NL+v++6773LzzTcDkJmZyYABA9i0aRMTJ07k/vvv\nJz8/n+nTp3PiiScycuRIfvCDH/A///M/TJ06lTPOOCNYb7dDcs4tBZY2abvT5/7zwPNHucYCYEEQ\nwgu6QH4GnXPcfvvtXxr31ltvcemll9KzZ08AUlI8+47eeustFi1aBEB4eDiJiYlHTfx8Z6Pz8/OZ\nMWMGBQUF1NTUMGjQIADefPNNnn322cZ+ycmewhaTJ0/mlVdeISsri9raWkaOHNnK75b42lNRzRrv\ncs2PtpeyPr+88Sy8lNgoxvRL4pIxGYzul8yofokkRCtBaZXqClj/HKyaD3s2eipXjrsOsq+B1JNC\nHZ20gRK/FmSlx/Pe5r2hDkNEjgdHmBUJpksvvZTnn3+ewsJCZsyYwVNPPUVRURGrV68mMjKSgQMH\nUl3tx9lFfvjmN7/JhAkTePXVV7ngggt45JFHmDx5MmvWrGHp0qXccccdnHPOOdx5551Hv5gEXif/\nDAbisxsREUFDQ0Pj46bjY2NjG+/ffPPN3HrrrVx00UW888473H333Ue89nXXXcfPf/5zMjMzmTNn\nTqviOt5V19bzya5y75LNMtZuL2NnmecQ7MhwY1h6AjPG9WNM/yTG9EumX0rM8VdVM1CaHsWQfjJc\n9CcY8Q2Iij36eOnwlPi1ICstgRfX7KRkfw0psVGhDkdEJOBmzJjB9ddfz969e/nPf/7D4sWL6dWr\nF5GRkbz99tts27at1dc844wzeOqpp5g8eTKbNm1i+/btDB06lLy8PAYPHswtt9zC9u3bWb9+PZmZ\nmaSkpHDllVeSlJTEo48+GoR3KR1ZoD6D5eXlzY6bPHkyl1xyCbfeeis9evSgpKSElJQUzjnnHB56\n6CG+973vNS717N27N3v27KG4uJi4uDheeeUVpkyZ0uLrZWR4toIsXLiwsf3cc8/lwQcfbNzPV1pa\nSnJyMhMmTGDHjh2sWbOG9evXt+Vb1qU559heUtW4XHPtjjI2FlQ0npWXkRTDmP5JzDl9IGP6JzO8\njwqwtFndQc8yzpWPHT6KYcQ3IPtayDjluDnm4HihxK8FmenxAOQWVnDakJ4hjkZEJPCGDx/Ovn37\nyMjIID09nSuuuIKvf/3rjBw5kuzsbDIzM1t9ze985zt8+9vfZuTIkURERLBgwQK6devG4sWLeeKJ\nJ4iMjCQtLY3bb7+dlStX8qMf/YiwsDAiIyN56KGHgvAupSML1GewpXHDhw/npz/9KWeeeSbh4eGM\nGTOGBQsW8Ic//IG5c+fy2GOPER4ezkMPPcTEiRO58847GT9+PBkZGUd87bvvvptLL72U5ORkJk+e\nzOeffw7AHXfcwY033siIESMIDw/nrrvuYvr06QBcdtllrF27tnH5p3g45/hkVwWvbijg1fUFbC+p\nAqB7VDij+iZy3RmDGd0viTH9kuiVEB3iaLuQ0q2w6nH46EnvUQxDdBTDccCc6zoVnrOzs92hjddt\nVbTvIOPuf5M7pw7jmq8MCsg1RUQOycnJISsrK9RhdArNfa/MbLVzLjtEIXU6zf1+1GewfU2dOpXv\nf//7nHPOOS32OV7+mzjn+HT3Pl5ZV8CrGwr4fO9+IsKM00/oybnDejN2QDIn9oojQmcpB1ZDvecI\nhlWP6SiGLq6l35Ga8WtBanw3esZFkVtYEepQREREpJMqKytj/PjxnHzyyUdM+o4Hn+3exyvrC3hl\n/S62FO0nzOC0IT2ZO2kwU4ankaytNcGhoxjES4nfEWSmJZBbqMqeIiIAGzZs4KqrrvpCW7du3fjg\ngw9CFJEcbzrjZzApKYlNmzaFOoyQySuq5JX1nmWcn+7ehxlMGJTCnNMHMWVEGj3jQnx+nnPQUAf1\nNd5bbQv366BbvGcZZEwyRHTwc/+cg+3LPXv3dBSDeAUt8TOzfsAioDfggHnOuT806WPAH4ALgCpg\ntnNujfe5WcAd3q73OecW0s4y0+J5YsU26hucznsRkePeyJEjWbt2bajDkOOYPoOdw7bi/d6ZvQJy\nCiowg3EDUrhn2nCmjEijV3yTvXrOQVUx7P0MKguPkHz53K87WqLW3H3fNp/bsYjsDjHeJDAmyfP1\nUFL4hVuTtsgg71PUUQxyBMGc8asDfuCcW2Nm8cBqM3vDObfRp8/XgBO9twnAQ8AEM0sB7gKy8SSN\nq83sJefckQ/ZCbDM9AQO1jWwtXg/Q1Lj2vOlReQ44JxT2fGj6Er70DsifQY7js7+Wd9RUtVYoGXD\nznIAxg5I5s6pw7hgZDppidFQUwUlm2H7Z1C8GfZu9nwt/gyqy/17obAICI/yzFiFRzVz36ctMvEo\nfb33I6KO/Hx4FFg4HKyAA6VwoAQOlHnve29Fnx5+rqGu5fgjuzeTHPrcvpQ8pviXMOooBvFD0BI/\n51wBUOC9v8/McoAMwDfxmwYscp6fdivMLMnM0oGzgDeccyUAZvYGMAV4JljxNiczzVPZM6egQomf\niARUdHQ0xcXF9OjRQ394t8A5R3FxMdHRquQXDPoMdhyd9bO+q+wASzcU8PL6AtbtKANgTN94fjk5\nga/22kfP6vVQ/CL84zMo3gLlO754gYQM6HECjPh/0PNEz/2EPhDezZN0RXT7YgIWFtnxC5A4BzX7\nvcmhT2JY5fu47PDzezcdfr6htuXrRsQ0SQyTDieH2z/QUQzil3bZ42dmA4ExQNNF+BmA70+BfG9b\nS+3t6oRecYSHGbkF+5g6qr1fXUS6sr59+5Kfn09RUVGoQ+nQoqOj6du3b6jD6JL0GexYOstnfXdF\nNa+uL2DZulwq8nMZHFbAlfEl/K5fMf3qdxJZuhXeP3h4QLcET0I34DTP18bbkK45E2UG3eI8t6T+\n/o9rTBh9ksUvJY8+9/duPtwnqb+OYhC/BD3xM7M44AXge865gJfINLO5wFyA/v1b8T+YH6Ijwxnc\nM1aVPUUk4CIjIxk0SEfFSOjoMyhHVXsASvIoz89hS+5aKnbkkFC1lUuskGusEg7VN6mJBDcIUk+A\nzPM9id2hGbzYVM08+eMLCWM//8c5p++v+C2oiZ+ZReJJ+p5yzr3YTJedgO+nu6+3bSee5Z6+7e80\n9xrOuXnAPPCcU9TmoJvITE9gzbZ23VooIiIi0j4aGqAi31NYpXiLZ79d8Wbqiz4jrCIfw5EInALs\ntRQOJA8mrO+p0HfY4dm7pAEQrkLxIaGkT1ohmFU9DXgMyHHO/baFbi8BN5nZs3iKu5Q75wrM7HXg\n52aW7O13HvCTYMV6JFnp8by8bhcV1bUkRKv0rYiIiHQyzkHlHijJg5It3gRvs+dryRaoq27sWhsR\nyw7rw4bq/uQ1jGd//EAGDR3NhHHjOaFvWgjfhIi0VTD/eeZ04Cpgg5kdqr18O9AfwDn3MLAUz1EO\nm/Ec5zDH+1yJmd0LrPSOu+dQoZf2lpWWAMCnhfsYN1DrpkVERKQDcg4qd3uSu+Ith5O8kjwo+dxT\n6fEQC4eUQdDjBA4OOJMN1an8a3c8L+d3p6A6kYE9Ypma3YepJ6cztHe8iv+IdBHBrOr5LnDEnxTe\nap43tvDcfGB+EEJrlcx0T2XP3IIKJX4iIiISOoeSu6aJXXGe52vt/sN9wyI8SzBTBsOA0z1fU4ZA\nyiAqotN589MSXllfwH/fL6K23tEvJYaLJ/XhwpHpDO+ToGRPpAvSguyjSEuIJjEmkpzCfaEORURE\nRLo652BfYZPEbotn1q6l5K7HEBh4ujexGww9BkNi/y/tu9tbeZC/Lstj0fJlHKitJyMphjmnD2Lq\nqHRGZiQq2RPp4pT4HYWZkZkWT26BKnuKiIhIADQmd1uaLM08NHNXdbhvWAQkD/QkdAO/cjixS2k+\nuWvO4YRvGwfr6pk2OoOrJg5gTL8kJXsixxElfn7ISk/gb6t20NDgCAvTD0gRERHxQ0XB4WIqjTN4\nnx8huRsCA8/wzOClDPI8Tux3zBUzm0v4bpp8AkNS4wLz/kSkU1Hi54fMtHj219STX3qA/j26hzoc\nERER6cjqDsLL34N1Tx9uC4s8PHM3aJJ3z92hmbtjT+6ao4RPRJqjxM8Pmemeyp45hRVK/ERERKRl\nVSXw3FWw7V2YeBMMmeyZwUvoG/Sz7pTwiciRKPHzw0m94zCDnIIKzh+uM2xERESkGSV58NRlULYN\npj8Koy5tl5dVwici/lDi54fuUREM7BFLboEqe4qIiEgzdnwIz1wOrgGu+runymaQKeETkdZQ4uen\nzLR4clTZU0REAsDMpgB/AMKBR51zv2zy/AA8Z9mmAiXAlc65fDMbDTwEJAD1wP3OuefaNXj5sk+W\nwJIbID4drngeep4Q1JdTwicix0KJn58y0xJ47ZNC9h+sI7abvm0iInJszCwceBA4F8gHVprZS865\njT7dHgAWOecWmtlk4BfAVUAVcLVz7jMz6wOsNrPXnXNl7fw2BDzHMrz3B3jzLug3AS5/GmJ7Bu3l\nlPCJSFsog/FTVno8zsGm3fsY0z851OGIiEjnNR7Y7JzLAzCzZ4FpgG/iNwy41Xv/beDvAM65TYc6\nOOd2mdkePLOCSvzaW30dLP0hrH4chl8CFz8MkdFBeSklfCISCEr8/JTlreyZW6jET0RE2iQD2OHz\nOB+Y0KTPOmA6nuWglwDxZtbDOVd8qIOZjQeigC3BDVe+pLoC/jYbtvwbvnIrTP4ZhIUF/GWU8IlI\nICnx81NGUgxx3SLI1T4/EREJvh8Cfzaz2cAyYCeePX0AmFk68AQwyznX0NwFzGwuMBegf//+wY73\n+FGe76ncWZQLX/8jjJ0V8JdQwiciwaDEz09hYcbQtHhyClXZU0RE2mQn0M/ncV9vWyPn3C48M36Y\nWRzwjUP7+MwsAXgV+KlzbkVLL+KcmwfMA8jOznaBfAPHrYJ18PQMqNkPVz7vOaMvgJTwiUgwKfFr\nhcy0eF5etwvnHGYW6nBERN8RjZQAACAASURBVKRzWgmcaGaD8CR8lwPf9O1gZj2BEu9s3k/wVPjE\nzKKAJXgKvzzfrlEf7z59DZ6/BmKS4ZrXofewgF1aCZ+ItAclfq2QmZ7AUx9sp6C8mj5JMaEOR0RE\nOiHnXJ2Z3QS8juc4h/nOuU/M7B5glXPuJeAs4Bdm5vAs9bzRO/wyYBLQw7sMFGC2c25te76H486H\nf4V/3gZpo+Cbz0F8WkAuq4RPRNqTEr9WyEqLByC3sEKJn4iIHDPn3FJgaZO2O33uPw98aUbPOfck\n8GTQAxSPhnr4189gxYNw0tfg/z0GUbFtvqwSPhEJBSV+rXCSN/HLKdjH5MzeIY5GREREgqZmP7w4\nF3JfgQk3wPk/h7DwNl1SCZ+IhJISv1ZIiI6kb3IMOarsKSIi0nXt2w3PzPAUc5nyf3DqDW26nBI+\nEekIlPi1UmZaArmq7CkiItI17cnxHNdQtRdmPAWZFxzzpZTwiUhHosSvlbLS43krdzfVtfVER7Zt\nyYeIiIh0IHnvwHNXQ2Q0zFkKfcYc86X+sXYnP35hgxI+EekwlPi1UmZaAg0ONu+pZERGYqjDERER\nkUD46El4+bvQ8yT45mJI6nf0MS14K3c3ty5ex9gByfxi+kglfCLSIYSFOoDOJjP9UIEX7fMTERHp\n9JyDf98L/7gRBp4B17zWpqRvzfZSvvPUGoalJzB/9jglfSLSYQRtxs/M5gNTgT3OuRHNPP8j4Aqf\nOLKAVOdciZltBfYB9UCdcy47WHG21sAesURHhmmfn4iISGdXW+1J+D5+Hk65Gi78LYRHHvPlNu/Z\nxzULVpKWEM3jc8YR100Lq0Sk4wjmjN8CYEpLTzrnfu2cG+2cGw38BPiPc67Ep8vZ3uc7TNIHEB5m\nDO0dT26hZvxEREQ6raoSeOJiT9J3zl3w9T+2KenbVXaAqx77kMjwMJ64dgI947oFMFgRkbYL2j9F\nOeeWmdlAP7vPBJ4JViyBlpmWwBs5u3HOYWahDkdERERao3gLPHUplOfD/5sPI77RpsuVVdUwa/6H\nVFbX8dy3JtIvpXuAAhURCZyQ7/Ezs+54ZgZf8Gl2wL/MbLWZzT3K+LlmtsrMVhUVFQUz1EaZ6fGU\n7K+hqPJgu7yeiIiIBMj2FfDoV+FAKcx6qc1J34Gaeq5duIptJVX8dVY2w/okBChQEZHACnniB3wd\neK/JMs+vOOdOAb4G3Ghmk1oa7Jyb55zLds5lp6amBjtWwDPjB5BboH1+IiIincbHL8DCi6B7Clz3\nJvQ/tU2Xq61v4Man17Bmeyl/vHw0pw7uEaBARUQCryMkfpfTZJmnc26n9+seYAkwPgRxtSgzzVPZ\nU/v8REREOgHnYNkD8Pw1kDEWrn0Degxp4yUdP3lxA2/l7uG+i0cwZUR6gIIVEQmOkCZ+ZpYInAn8\nw6ct1sziD90HzgM+Dk2EzUuOjSItIVozfiIiIh1dfS28dDO8dS+MvBSu/rtnxq+N/u+1T3l+dT7f\n/+pJXDFhQAACFREJrmAe5/AMcBbQ08zygbuASADn3MPebpcA/3LO7fcZ2htY4i2aEgE87Zx7LVhx\nHqvM9Hg26iw/ERGRjqu6HBZfDXnvwKTb4OzbIQBF2R79bx4P/2cLV57an1vOOaHtcYqItINgVvWc\n6UefBXiOffBtywNODk5UgZOZlsB7m/dSU9dAVERHWDErIiIijcq2w1OXQfFnMO0vMOaKo4/xw98/\n2sl9r+Zwwcg0/veiEaruLSKdhjKWY5SVHk9tvSNvb2WoQxERERFfO9d4KndW7IIrXwxY0vefTUX8\n8G/rmDi4B7+bMZrwMCV9ItJ5KPE7RqrsKSIi0gHlvgoLLoSIbnDtv2DwmQG57NodZXz7ydWc1Due\nR64eS7eI8IBcV0SkvSjxO0aDU2OJCg8jR5U9RUREOoYVD8GzV0BqJlz3b+iVGZDLbimq5JoFK+kZ\n140F14wjIToyINcVEWlPQdvj19VFhodxQq84zfiJiIiEWkM9vPYT+PARyJwK0/8KUd0DcundFdVc\n/diHhBksumY8veKjA3JdEZH2psSvDTLT43lv895QhyEiInL8OlgJL1wHm/4JE2+Cc++BsMAswyw/\nUMvVj31IWVUNz31rIgN7xgbkuiIioaClnm2QlZbA7oqDlOyvCXUoIiIixx/n4OnL4LPX4YIH4Pz7\nA5b0VdfWc/3CVXy+dz/zrs5mREZiQK4rIhIqSvzaIDM9HoBc7fMTEZFWMrMpZvapmW02sx838/wA\nM/u3ma03s3fMrK/Pc7PM7DPvbVb7Rt6B5K+Ebe/B+b+A8dcH7LJ19Q3c/MxHrNxWwm9nnMzpJ/QM\n2LVFREJFiV8bqLKniIgcCzMLBx4EvgYMA2aa2bAm3R4AFjnnRgH3AL/wjk0B7gImAOOBu8wsub1i\n71BWL4CouIAd1wDgnOOOv3/MGxt3c/fXhzN1VJ+AXVtEJJSU+LVBanw3esZFacZPRERaazyw2TmX\n55yrAZ4FpjXpMwx4y3v/bZ/nzwfecM6VOOdKgTeAKe0Qc8dSXQ4fvwgjvgHd4gN22d++sYlnV+7g\n5sknMOu0gQG7rohIqCnxa6PMtARyCzXjJyIirZIB7PB5nO9t87UOmO69fwkQb2Y9/Bzb9W34G9Qd\ngLGBW+m64L3P+dNbm5k5vh+3nntSwK4rItIRKPFro8y0eD4t3EddfUOoQxERka7lh8CZZvYRcCaw\nE6j3d7CZzTWzVWa2qqioKFgxhoZznmWeaSOhzykBueTL63bxv69s5Lxhvbl32gjMLCDXFRHpKJT4\ntVFmegIH6xrYWlwV6lBERKTz2An083nc19vWyDm3yzk33Tk3Bvipt63Mn7HevvOcc9nOuezU1NRA\nxx9auz6Cwg1wyiwIQIL27md7uXXxWsYNSOGPM8cQEa4/j0Sk69FPtjbKTFNlTxERabWVwIlmNsjM\nooDLgZd8O5hZTzM79Hv6J8B87/3XgfPMLNlb1OU8b9vxY81CiIiBUZe1+VIb8sv51hOrGJIax19n\nZRMdGZjjIEREOholfm10Qq84wsNMlT1FRMRvzrk64CY8CVsOsNg594mZ3WNmF3m7nQV8amabgN7A\n/d6xJcC9eJLHlcA93rbjw8FK2PA8jJgO0W07W+/zvfuZ/fiHJHWPYuE140mMiQxQkCIiHU9EqAPo\n7KIjwxncM1YzfiIi0irOuaXA0iZtd/rcfx54voWx8zk8A3h8+fgFqKn0LPNsgz37qrl6/gc44Ilr\nx9M7ITow8YmIdFCa8QuArPQEcjTjJyIiEnxrFkJqFvQbf8yXqKiuZdb8lRRX1vD47HEMTo0LYIAi\nIh2TEr8AyEyPZ2fZASqqa0MdioiISNdVuAF2rvYc4XCMRV2qa+uZu2gVn+3ex8NXjuXkfkkBDlJE\npGNS4hcAWWkJAHyq8/xERESCZ/VCCO8Go2Yc0/D6Bsf3n1vLirwSfnPZyUw6qYtVOxUROQIlfgGQ\nme6t7FmgfX4iIiJBUVMF6xfDsGnQPaXVw51z3PmPj/nnx4XcOXUY00Yff2fei8jxTYlfAKQlRJMY\nE0mOZvxERESCY+Pf4WC5Z5nnMfjDvz/jqQ+28+2zhnDNVwYFODgRkY5PiV8AmBmZafGa8RMREQmW\n1Quhxwkw4PRWD31yxTZ+/+ZnXJbdl9vOHxqE4EREOr6gJX5mNt/M9pjZxy08f5aZlZvZWu/tTp/n\nppjZp2a22cx+HKwYAykrPYFPC/fR0OBCHYqIiEjXsicHdqzwHOHQyqIu/9xQwM/+8THnZPbi55eM\nxI6xKIyISGcXzBm/BcCUo/T5r3NutPd2D4CZhQMPAl8DhgEzzWxYEOMMiMy0ePbX1LOjtCrUoYiI\niHQtaxZBWCSM/marhi3fUsx3n13L2P7J/PmbpxARroVOInL8CtpPQOfcMqDkGIaOBzY75/KcczXA\ns8C0gAYXBJnpnsqeOs9PREQkgGqrYd0zkHkhxPb0e9gnu8qZu2gVA3t259FZ2cREhQcxSBGRji/U\n//Q10czWmdk/zWy4ty0D2OHTJ9/b1qGd1DsOM8gt1D4/ERGRgMl5GQ6UwtjZfg/ZXlzFrPkriY+O\nYOE140nqHhW8+EREOomIEL72GmCAc67SzC4A/g6c2NqLmNlcYC5A//79AxthK3SPimBgj1hyNeMn\nIiISOGsWQtIAGHSmX92L9h3kqvkfUNfQwLNzJ5KeGBPkAEVEOoeQzfg55yqcc5Xe+0uBSDPrCewE\n+vl07etta+k685xz2c657NTU0B7EmpkWrxk/ERGRQNm7Gbb+F065GsKO/idL5cE65iz4kD0VB3l8\n9jhO6BXfDkGKiHQOIUv8zCzNvKW1zGy8N5ZiYCVwopkNMrMo4HLgpVDF2RpZ6QlsK6li/8G6UIci\nIiLS+a1ZCBYOY648ateDdfV864lV5BTs4y9XnsKY/sntEKCISOcRtKWeZvYMcBbQ08zygbuASADn\n3MPA/wO+bWZ1wAHgcuecA+rM7CbgdSAcmO+c+yRYcQZSZlo8zsGm3fv0C0dERKQt6mpg7dMw9GsQ\nn3bErg0NjlsXr+O9zcX89rKTOXtor3YKUkSk8wha4uecm3mU5/8M/LmF55YCS4MRVzBleSt75hYq\n8RMREWmTT5dC1V7P2X1H4Jzjf1/+hFfXF/DTC7KYfkrfdgpQRKRzCXVVzy4lIymGuG4R5BZon5+I\niEibrF4ACX3hhHOO2O2h/2xh4fJtzJ00mOsnDW6f2EREOiElfgEUFmYMTYsnp1CVPUVERI5Z6VbI\nextOuQrCWj5/72BdPX948zPOH96bH0/JbL/4REQ6ISV+AZaZFk9uQQWe7YoiIiLSamueAAs7alGX\nj7aXcbCugUvH9iMszNopOBGRzkmJX4BlpidQUV1HQXl1qEMRERHpfOrr4KMn4YRzIfHI+/VW5BUT\nZjBuUEo7BSci0nkp8QuwrDTPmUE52ucnIiLSep+9DpWFMPbIRV0Alm8pZnifRBJjItshMBGRzk2J\nX4Cd5E38crXPT0REpPVWL4S4NDjx/CN2q66t56PtZUwc0qOdAhMR6dyU+AVYQnQkfZNjNOMnIiLS\nWuX5sPkNz96+8COfOLVmeyk19Q2cOljLPEVE/KHELwgy0xI04yciIkdkZlPM7FMz22xmP27m+f5m\n9raZfWRm683sAm97pJktNLMNZpZjZj9p/+iD5KMnwTV4qnkexYotxYSHGeMGKvETEfGHEr8gyEqP\nJ6+okura+lCHIiIiHZCZhQMPAl8DhgEzzWxYk253AIudc2OAy4G/eNsvBbo550YCY4FvmdnA9og7\nqBrqPdU8h0yG5IFH7b48r5gRGYnER2t/n4iIP5T4BUFWegINDjbvqQx1KCIiEkRmdrOZJR/D0PHA\nZudcnnOuBngWmNakjwMSvPcTgV0+7bFmFgHEADVA599fsPnfUJEPpxy9qMuBmnrW7ijTMk8RkVZQ\n4hcEmarsKSJyvOgNrDSzxd6lm/4eJpcB7PB5nO9t83U3cKWZ5QNLgZu97c8D+4ECYDvwgHOu5Bjj\n7zjWLITuPWHoBUftunpbKbX1jomDVdhFRMRfSvyCYECPWKIjw7TPT0Ski3PO3QGcCDwGzAY+M7Of\nm9mQAFx+JrDAOdcXuAB4wszC8MwW1gN9gEHAD8xscNPBZjbXzFaZ2aqioqIAhBNE+wrh03/CmCsg\nIuqo3Zfn7dX+PhGRVlLiFwThYcbQ3vHkFmrGT0Skq3POOaDQe6sDkoHnzexXRxi2E+jn87ivt83X\ntcBi72ssB6KBnsA3gdecc7XOuT3Ae0B2M3HNc85lO+eyU1NTj+m9tZuPngRX79cyT4AVeSWM6ptI\nbLcjV/4UEZHDlPgFSWZaAjkF+/D8PSAiIl2RmX3XzFYDv8KTgI10zn0bT9GVbxxh6ErgRDMbZGZR\neIq3vNSkz3bgHO/rZOFJ/Iq87ZO97bHAqUBuwN5Ue2togDWLYOAZ0OPoE6X7D9axbkeZlnmKiLSS\nEr8gyUyPp2R/DUWVB0MdioiIBE8KMN05d75z7m/OuVoA51wDMLWlQc65OuAm4HUgB0/1zk/M7B4z\nu8jb7QfA9Wa2DngGmO2dXXwQiDOzT/AkkI8759YH6w0G3efvQNk2GDvbr+6rt5VS1+A4VYmfiEir\naI1EkGSmeQqx5Rbso1d8dIijERGRIPkn0FhYxcwSgCzn3AfOuZwjDXTOLcVTtMW37U6f+xuB05sZ\nV4nnSIeuYfVCiEmGzBbz5C9YnldMZLiRPfBYiqmKiBy/NOMXJKrsKSJyXHgI8D27p9LbJv6oLILc\nV+HkmRDp3z+SLt9SzMl9k+gepX+7FhFpDSV+QZIcG0VaQrQqe4qIdG3mfDZze5d4KiPx17qnoaHW\n76IulQfr2LCzXMs8RUSOgRK/IMpMj9eMn4hI15ZnZreYWaT39l0gL9RBdQrOeYq69DsVemX6NWTl\n1hLqGxwThyjxExFpLSV+QZSZlsCWokpq6hpCHYqIiATHDcBpeI5iyAcmAHNDGlFnse09KN7sd1EX\ngBVbiokKD+OU/trfJyLSWlqOEkRZ6fHU1jvy9lY2FnsREZGuw3uO3uWhjqNTWr0AuiXCsGl+D1mR\nV8zofknERIUHLy4RkS7Kr8TPzIYA+c65g2Z2FjAKWOScKwtmcJ1dVvrhyp5K/EREuh4zi8Zz0Ppw\nPOfsAeCcuyZkQXUGVSWw8SU45WqI6u7XkIrqWjbsLOemyScGOTgRka7J36WeLwD1ZnYCMA/oBzx9\npAFmNt/M9pjZxy08f4WZrTezDWb2vpmd7PPcVm/7WjNb5WeMHc6gnrFEhYeRU6h9fiIiXdQTQBpw\nPvAfoC+gql5Hs/45qD8IY/0r6gKw8vMSGhw6uF1E5Bj5m/g1eA+bvQT4k3PuR0D6UcYsAKYc4fnP\ngTOdcyOBe/EklL7Ods6Nds5l+xljhxMZHsYJveLILdDfACIiXdQJzrmfAfudcwuBC/Hs85OWOOdZ\n5pkxFtJG+j1sRV4xURFhjOmfFLzYRES6MH8Tv1ozmwnMAl7xtkUeaYBzbhk+h9o28/z7zrlS78MV\neP6VtMvJTI8nVzN+IiJdVa33a5mZjQASgV4hjKfj2/EhFOX6fYTDIcvzijmlfxLRkdrfJyJyLPxN\n/OYAE4H7nXOfm9kgPMtbAuVa4J8+jx3wLzNbbWZHrI5mZnPNbJWZrSoqKgpgSIGRlZbA7oqDlOyv\nCXUoIiISePPMLBm4A3gJ2Aj8X2hD6uBWL4CoOBjxDb+HlFfV8smuCiYO7hm8uEREuji/irs45zYC\ntwB4f8HFO+cC8ovNzM7Gk/h9xaf5K865nWbWC3jDzHK9M4jNxTYP7zLR7Oxs11yfUMpMjwcgt7CC\n04boF5aISFdhZmFAhXf1yjJgcIhD6vgOlMEnS+DkGdAtzu9hH24twTk4dXBKEIMTEena/JrxM7N3\nzCzBzFKANcBfzey3bX1xMxsFPApMc84VH2p3zu30ft0DLAHGt/W1QuVQNU/t8xMR6Vqccw3AbaGO\no1PZ8DeoO9D6ZZ5biukWEcZo7e8TETlm/i71THTOVQDT8RzjMAH4alte2Mz6Ay8CVznnNvm0x5pZ\n/KH7wHlAs5VBO4PU+G70jIsip0D7/EREuqA3zeyHZtbPzFIO3UIdVIfkHKxeCGmjoM+YVg1dnlfM\n2AHJdIvQ/j4RkWPl7wHuEWaWDlwG/NSfAWb2DHAW0NPM8oG78BaEcc49DNwJ9AD+YmYAdd4Knr2B\nJd62COBp59xr/r6hjigzLYHcQs34iYh0QTO8X2/0aXNo2eeX7VoDuzfAhb8Bz+94v5RV1ZBbWMGt\nXz0piMGJiHR9/iZ+9wCvA+8551aa2WDgsyMNcM7NPMrz1wHXNdOeB5z85RGdV2ZaPE+s2EZdfQMR\n4f5OsoqISEfnnBsU6hg6jdULIbI7jLy0VcNW5Hn2900covP7RETawt/iLn8D/ubzOA/wvxzXcS4z\nPYGDdQ1sLa7ihF7+b2YXEZGOzcyubq7dObeovWPp0A7ugw3Pw/DpEJ3YqqEr8oqJiQxnVF/t7xMR\naQt/i7v0NbMlZrbHe3vBzLrkuXvBkJl2uLKniIh0KeN8bmcAdwMXhTKgDunjF6B2P4xtXVEX8CR+\n2QOTiYrQihkRkbbw96fo43jOJ+rjvb3sbRM/nNg7jvAwU2VPEZEuxjl3s8/teuAUQEs7mlq9EFKz\noO+4Vg0rrjxIbuE+Th2sZZ4iIm3lb+KX6px73DlX570tAFKDGFeX0i0inCGpsZrxExHp+vYD2vfn\nq2C9p7DL2NmtKuoC8OHnJQBK/EREAsDf4i7FZnYl8Iz38Uyg+Aj9pYnMtARWbysNdRgiIhJAZvYy\nniqe4PnH1GHA4tBF1AGtWQjh3WDUZa0eujyvmO5R4Yzq27p9gSIi8mX+Jn7XAH8CfofnF9z7wOwg\nxdQlZabH89K6XVRU15IQHRnqcEREJDAe8LlfB2xzzuWHKpgOp6YK1i+G4RdD99Yfb7h8SzHjBqYQ\nqYrYIiJt5tdPUufcNufcRc65VOdcL+fcxaiqZ6tkpSUA8KnO8xMR6Uq2Ax845/7jnHsPzwqZgf4M\nNLMpZvapmW02sx8383x/M3vbzD4ys/VmdoHPc6PMbLmZfWJmG8wsOlBvKKA+WQIHK+CU1hd12Vt5\nkM/2VGqZp4hIgLTln9BuDVgUx4HMdG9lzwLt8xMR6UL+BjT4PK7H5/ijlphZOPAg8DU8y0Nnmtmw\nJt3uABY758YAlwN/8Y6NAJ4EbnDODQfOAmrb9jaCZM1C6HEiDDit1UNX5Hl2lOj8PhGRwGhL4te6\nHdrHubSEaBJjIsnRjJ+ISFcS4ZyrOfTAez/Kj3Hjgc3OuTzvmGeBaU36OCDBez8R2OW9fx6w3jm3\nzvuaxc65+ja8h+DYkwM7PvAc4dDKoi7gWeYZ1y2CEX0Sjt5ZRESOqi2Jnzt6FznEzMhMiydHM34i\nIl1JkZk1nttnZtOAvX6MywB2+DzO97b5uhu40szygaXAzd72kwBnZq+b2Rozu+1Ygw+q1QshLBJO\nnnlMw1fkFTNuYDIR2t8nIhIQR/xpamb7zKyimds+POf5SStkpSfwaeE+GhqUM4uIdBE3ALeb2XYz\n2w78D/CtAF17JrDAOdcXuAB4wszC8BRm+wpwhffrJWZ2TtPBZjbXzFaZ2aqioqIAheSn2mpY9wxk\nTYXYnq0evqeimi1F+7XMU0QkgI5Y1dM5F99egRwPMtPiqaqpZ0dpFQN6xIY6HBERaSPn3BbgVDOL\n8z6u9HPoTqCfz+O+3jZf1wJTvNdd7i3g0hPP7OAy59xeADNbiufg+H83iW0eMA8gOzu7ff/FMecl\nqC7znN13DJZ79/epsIuISOBo/UQ7ykz37FPIKdA+PxGRrsDMfm5mSc65SudcpZklm9l9fgxdCZxo\nZoPMLApP8ZaXmvTZDpzjfZ0sIBooAl4HRppZd2+hlzOBjYF6TwGxeiEkD4SBk45p+Iq8EuKjIxje\nR+f3iYgEihK/dnRS7zjMILdQ+/xERLqIrznnyg49cM6V4lmWeUTOuTrgJjxJXA6e6p2fmNk9PnsG\nfwBcb2brgGeA2c6jFPgtnuRxLbDGOfdqQN9VW+zdDNvehVOuhrBj+zNjRV4xEwalEB6mOnIiIoHi\n7wHuEgDdoyIY1COWXM34iYh0FeFm1s05dxDAzGKAbv4MdM4txVO0xbftTp/7G4HTWxj7JJ4jHTqe\nNQsgLAJGX3lMwwvLq/l8736umNA/sHGJiBznlPi1s8z0eDbu0oyfiEgX8RTwbzN7HM8xR7OBhSGN\nKJTqamDtM3DSFIjvfUyXWKH9fSIiQaGlnu0sMy2BbSVV7D9YF+pQRESkjZxz/wfcB2QBQ/Es3RwQ\n0qBC6dNXoWrvMRd1Ac/5fYkxkQxL1/l9IiKBpMSvnWWmxeMcbNqt5Z4iIl3Ebjxn214KTMazZ+/4\ntHoBJPaDIZOP+RIrPi9m/KAUwrS/T0QkoJT4tbMs779g5hYq8RMR6azM7CQzu8vMcoE/4anAac65\ns51zfw5xeKFR8jnkvQNjroKw8GO6xK6yA2wrrmKilnmKiASc9vi1s4ykGOK6RZBboH1+IiKdWC7w\nX2Cqc24z/P/27jy+rrrO//jrc2/2JmnSNm3SPZRCWtbWDC2ggGWxoAM6jmNhWF1w/A3oKDMj/H7+\n1GGc0XEWdX6DMyIiiAiDuExnRBEFN0y0hbLYpkCb7muaNE3TNNu9n98f56S5TdM2aXJzbpL38/E4\nj3vO9yz3k9PCt598NzCzj0UbUsTWPAIWg0WnNqkLBN08AS3cLiKSBmrxG2GxmHFmeRF1avETERnN\n/gjYBTxnZl8zs8sJJncZnxJdsOZbMP8qmDjjlB9TW99IaUE2Z04rGsbgREQE0pz4mdmDZrbXzH5/\nnPNmZv9qZhvM7BUzW5xy7hYzeyPcbklnnCOtqryIul0tuHvUoYiIyClw9x+4+wqgCngO+Atgqpn9\nu5ldFW10EXj9aWjdA4uHVl3X1DeypHKyxveJiKRBulv8HgKWn+D81cD8cLsd+HcAM5sEfBpYAlwA\nfNrMStMa6QiqqijmYHs3Ow+0Rx2KiIgMgbsfcvdvu/sfAjOBNcAnIg5r5L34MBRVBC1+p2hbUxvb\n9x9WN08RkTRJa+Ln7r8Emk5wyXXANz1QC5SYWQXwNuAZd29y9/3AM5w4gRxVFpQHXVg0zk9EZOxw\n9/3ufr+7Xx51LCOqeRu88Uwwti9+6lMHaP0+EZH0inqM3wxgW8rx9rDseOVjwhk9iZ/G+YmIyGi3\n5lvB56KbhvSYmvpGJk3I4YxphcMQlIiI9BV14jdkZna7ma02s9UNDQ1RhzMgxXnZzCzNp04tfiIi\nMpolE8FsnvOWQempr1vv7tRubGTpaZMw0/g+EZF0iDrx2wHMSjmeGZYdr/wYYbeaanevLisrS1ug\nw21BRbFa/EREZHTbWTM2hQAAIABJREFU8FNo2QFvGtqkLtuaDrPzQLvW7xMRSaOoE7+VwM3h7J5L\ngQPuvgt4GrjKzErDSV2uCsvGjAXlRdQ3tNLelYg6FBERkVPzwsMwoQzOuHpIj6mp3wdo/T4RkXRK\n6wLuZvYYcBkwxcy2E8zUmQ3g7v8BPAVcA2wA2oDbwnNNZva3wKrwUfe6+4kmiRl1qiqKSTps2NvK\n2TMmRh2OiIjI4LTsgtd/DBfdCVk5Q3pUzcZGphTmMq9M4/tERNIlrYmfu19/kvMO/Plxzj0IPJiO\nuDJBVTjBS92uFiV+IiIy+rz0LfAELL55SI9xd2rrmzS+T0QkzaLu6jluzZk8gbzsmMb5iYjI6JNM\nwovfhLlvgcnzhvSozY1t7G5pVzdPEZE0U+IXkXjMOHNaEet3a2ZPEREZZeqfg+at8KZbh/word8n\nIjIylPhFqKq8mLpdBwl6vIqIiIwSLz4M+ZNgwR8O+VE1GxuZWpTLaVMmDENgIiJyPEr8IlRVUUTT\noU4aDnZEHYqIiMjAtO6F9T+E866HrNwhPcrdqalv5MJ5kzW+T0QkzZT4RaiqvBiAOo3zExGR0eKl\nb0Oye8hr9wHU7ztEw8EOdfMUERkBSvwi1DOz5/pdGucnIiKjgHswqcvsC6HszCE/rmZjML5PC7eL\niKSfEr8IlU7Iobw4TzN7iojI6LD519C0ERYPvbUPoKa+kYqJecyZXDAszxMRkeNT4hexqooi6tTi\nJyIy7pjZcjN7zcw2mNnd/ZyfbWbPmdkaM3vFzK7p53yrmf3liAX9wkOQNxHOeueQH+Xu/La+kaWn\naXyfiMhIUOIXsQUVxWxsaKWzOxl1KCIiMkLMLA7cB1wNLASuN7OFfS77JPCEuy8CVgBf6XP+X4Af\npTvWI9qaoG4lnPteyM4f8uM27G1lX2ununmKiIwQJX4Rqyovoivh1O9rjToUEREZORcAG9y93t07\ngceB6/pc40BxuD8R2NlzwszeCWwC1o5ArIGXH4dE57Cs3QdBN0/Q+n0iIiNFiV/EFlQEdfr6XRrn\nJyIyjswAtqUcbw/LUn0GuNHMtgNPAXcCmFkh8Angb070BWZ2u5mtNrPVDQ0NQ4vWPejmOaMapp01\ntGeFausbmVGSz6xJQ289FBGRk1PiF7HKKRPIiceo261xfiIicpTrgYfcfSZwDfCImcUIEsIvuvsJ\nu4q4+/3uXu3u1WVlZUOLZNtvYd9rw7KEA0Ay6dTWN2l8n4jICMqKOoDxLjse4/SphWrxExEZX3YA\ns1KOZ4Zlqd4PLAdw9xozywOmAEuAPzazLwAlQNLM2t3939IXrsG8y+GsPxqWp72+9yBNhzpZetqk\nYXmeiIicnBK/DFBVUcTzG/ZFHYaIiIycVcB8M6skSPhWADf0uWYrcDnwkJktAPKABnd/S88FZvYZ\noDW9SR8wewnc9L1he1xtz/p98zS+T0RkpKirZwZYUF7MnpYOmg51Rh2KiIiMAHfvBu4AngbqCGbv\nXGtm95rZteFldwEfNLOXgceAW93do4l4eNXUNzJrUj4zS7V+n4jISFGLXwaoqigCYP2uFi46fUrE\n0YiIyEhw96cIJm1JLftUyv464OKTPOMzaQkujZJJ57ebmrhywbSoQxERGVfU4pcBqsqDmT3rdmuc\nn4iIjG3rdx+kua1L3TxFREaYEr8MUFaUy5TCHNbv0syeIiIytmn9PhGRaCjxyxBV5cWsV4ufiIiM\ncbX1jcyZXMD0Eq3fJyIykpT4ZYiq8iJe33OQ7kQy6lBERETSIpF0flvfyIVq7RMRGXFK/DLEgopi\nOrqTbG5sizoUERGRtKjb1UJLe7fG94mIRECJX4Y4MrPnbo3zExGRsalW4/tERCKT1sTPzJab2Wtm\ntsHM7u7n/BfN7KVwe93MmlPOJVLOrUxnnJng9KmFxGPG+l0a5yciImNTzcZGTpsygWnFeVGHIiIy\n7qRtHT8ziwP3AVcC24FVZrYyXJcIAHf/WMr1dwKLUh5x2N3PT1d8mSY3K868sglq8RMRkTGpO5Hk\nd5uaeMd506MORURkXEpni98FwAZ3r3f3TuBx4LoTXH898Fga48l4VeXF1KnFT0RExqB1u1o42KHx\nfSIiUUln4jcD2JZyvD0sO4aZzQEqgWdTivPMbLWZ1ZrZO9MXZuaoqihiR/NhWtq7og5FRERkWNVs\n7BnfNyniSERExqdMmdxlBfCkuydSyua4ezVwA/AlM5vX341mdnuYIK5uaGgYiVjTZkF5MYDG+YmI\nyJhTU9/IvLIJTC3S+D4RkSikM/HbAcxKOZ4ZlvVnBX26ebr7jvCzHvg5R4//S73ufnevdvfqsrKy\nocYcKc3sKSIiY1F3IsmqTU3q5ikiEqF0Jn6rgPlmVmlmOQTJ3TGzc5pZFVAK1KSUlZpZbrg/BbgY\nWNf33rGmvDiPifnZGucnIiJjyqs7DnCoM8GFp02JOhQRkXErbbN6unu3md0BPA3EgQfdfa2Z3Qus\ndveeJHAF8Li7e8rtC4CvmlmSIDn9fOpsoGOVmVFVXqQWPxERGVNqwvX7lmh8n4hIZNKW+AG4+1PA\nU33KPtXn+DP93Pcb4Jx0xpapFlQU88TqbSSTTixmUYcjIiIyZLX1TZwxrZAphblRhyIiMm5lyuQu\nEqoqL6KtM8G2/W1RhyIiIjJkXYkkqzc3ceFpGt8nIhIlJX6pDjXCUT1OR96CimBmT43zExGRseCV\n7c20dSZYqsRPRCRSSvx6dB2Gh94Oj/8pHG6OLIwzphVhppk9RURkbKitbwJgiRI/EZFIKfHrkZUH\nb7oF3ngavnoJ7FwTSRj5OXEqJ0/QWn4iIjIm1GxspKq8iEkTcqIORURkXFPi18MMln4YbvsxJBPw\n9atg1QORdP2sqtDMniIiMvp1didZvaVJ3TxFRDKAEr++Zv0BfOiXUHkp/PAu+N4HoaN1REOoKi9m\nS1Mbhzq6R/R7RUREhtPL25tp70pq4XYRkQygxK8/EybDDU/Ask/C778LX3sr7K0bsa+vKi/CHV7f\no+6eIiJjlZktN7PXzGyDmd3dz/nZZvacma0xs1fM7Jqw/Eoze8HMXg0/l4189ANTs7ERM1hSqfX7\nRESipsTveGIxuOSv4Ob/CiZ7+doyePnxEflqzewpIjK2mVkcuA+4GlgIXG9mC/tc9kngCXdfBKwA\nvhKW7wP+0N3PAW4BHhmZqAevtr6RBeXFlBRofJ+ISNSU+J1M5SXwZ7+C6Yvh+x+ClXcGM4Cm0YyS\nfApzszTOT0Rk7LoA2ODu9e7eCTwOXNfnGgeKw/2JwE4Ad1/j7jvD8rVAvpll3MroHd0JXtiyX908\nRUQyhBK/gSgqD1r+3nIXvPhN+PqV0LgxbV8XixlnlhdpZk8RkbFrBrAt5Xh7WJbqM8CNZrYdeAq4\ns5/nvBt40d070hHkUKzZ2kxHd1ITu4iIZAglfgMVz4LLPwU3fAcObIf7L4N1/5W2r6sqL6Judwse\n8YLyIiISmeuBh9x9JnAN8IiZHam3zews4B+AD/V3s5ndbmarzWx1Q0PDiAScqra+kZjBBRrfJyKS\nEZT4DdYZVwWzfk6ZD0/cDD++B7o7h/1rqiqKOdjezc4D7cP+bBERidwOYFbK8cywLNX7gScA3L0G\nyAOmAJjZTOD7wM3u3m8XFHe/392r3b26rKxsmMM/uZqNjZw1fSIT87NH/LtFRORYSvxORcnsYL2/\nJX8GtV+Bh66B5m0nv28QFpQXAbB+l8b5iYiMQauA+WZWaWY5BJO3rOxzzVbgcgAzW0CQ+DWYWQnw\nQ+Bud39+BGMesPauBGu2NrP0NLX2iYhkCiV+pyorB67+B3jPQ7B3PXz1LfDGM8P2+DN7Er/dGucn\nIjLWuHs3cAfwNFBHMHvnWjO718yuDS+7C/igmb0MPAbc6kH//zuA04FPmdlL4TY1gh/juF7cup/O\nhNbvExHJJFlRBzDqnfUumHYOfOcWePSP4S1/CZfdE4wJHIKivGxmTcqnTi1+IiJjkrs/RTBpS2rZ\np1L21wEX93PfZ4HPpj3AIajdGIzv+4O5avETEckUavEbDlNOhw/8FBbdBL/6J3jknXBwz5AfW1Ve\nrBY/EREZdWrqGzlnxkSK8jS+T0QkUyjxGy7Z+XDdv8E7/x22rw66fm761ZAeuaC8iPqGVtq7EsMU\npIiISHod7kzw0rZmlqqbp4hIRlHiN9zOvwE++DPILYZvXgu/+mdIJk/pUVUVxSQdNuxtHeYgRURE\n0uOFLfvpSjgXav0+EZGMosQvHaadBbc/F4z/+9m98Nh7oa1p0I+pCid40Tg/EREZLWrrG4nHjGqN\n7xMRyShK/NIltwje/XW45p+g/ufw1UuCLqCDMGfyBPKyY9Tt0jg/EREZHWrqGzl35kQKczV/nIhI\nJlHil05mcMEH4X1PB/sPLofa/wD3Ad0ejxlnTiti/W61+ImISOY71NHNy9ua1c1TRCQDpTXxM7Pl\nZvaamW0ws7v7OX+rmTWkrEP0gZRzt5jZG+F2SzrjTLsZi+FDv4T5V8KPPwHfuRXaB5bMVZUXU7er\nBR9gsigiIhKVF7bspzvpLFXiJyKScdKW+JlZHLgPuBpYCFxvZgv7ufQ/3f38cHsgvHcS8GlgCXAB\n8GkzK01XrCMivxRWfBuuvBfq/hvuvxR2v3rS26oqitjf1kXDwY4RCFJEROTU1dQ3khUzqueO7ipb\nRGQsSmeL3wXABnevd/dO4HHgugHe+zbgGXdvcvf9wDPA8jTFOXLM4OKPwq3/A12H4YEr4MVvnrDr\nZ1V5MQB1Ws9PREQyXM3GRs6bVUJBjsb3iYhkmnQmfjOAbSnH28Oyvt5tZq+Y2ZNmNmuQ945Ocy6C\nD/0KZi2BlXfCD/4XdLb1e+mCimBmz/Wa2VNERDJYa0c3r+44oPF9IiIZKurJXf4bmOvu5xK06j08\n2AeY2e1mttrMVjc0NAx7gGlTWAY3fR8uvRtefgweuBwaXj/mspKCHCom5rFeLX4iIpLBVm1uIpF0\nLtTC7SIiGSmdid8OYFbK8cyw7Ah3b3T3nsFrDwBvGui9Kc+4392r3b26rKxsWAIfMbE4vPUeuPG7\n0LoHvvZWePXJYy6rKi/SWn4iIpLRajc2kh03Fs/W+D4RkUyUzsRvFTDfzCrNLAdYAaxMvcDMKlIO\nrwXqwv2ngavMrDSc1OWqsGxsOv3yoOvntLPhu++HH94F3b2TuVRVFLNhbytf/cVGDrR1RRioiIhI\n/2rrG1k0q5T8nHjUoYiISD/Slvi5ezdwB0HCVgc84e5rzexeM7s2vOwjZrbWzF4GPgLcGt7bBPwt\nQfK4Crg3LBu7Js4IJn256E5Y9QA8+DbYvxmAm5bOoXpuKZ/70XqWfu5nfPIHr7Jhb2u08YqIiIRa\n2rt4dccBlqqbp4hIxrKxtD5cdXW1r169Ouowhm79D+H7HwYD3vVVOPNqANbtbOEbz2/iv17eSWd3\nkkvPKON9b67kkvlTMLNoYxYRGUFm9oK7V0cdx2iR7vrxZ3V7eP/Dq/n2B5dw0bwpafseERE5uePV\nkVFP7iL9qXo7fOgXUDoXHlsBz3wKEl0snF7MP77nPH5z9zI+fuUZrNvVwi0P/o4rv/hLvlW7hbbO\n7qgjFxGRcai2vpGcrJjG94mIZDAlfplqUiW87ydQ/T54/svw4PJg4fdEF1MKc/nI5fN5/hPL+OJ7\nzyM/O84nf/B7Lvzcs3zuR3XsaD4cdfQiIjKO1NQ3snh2CXnZGt8nIpKplPhlsuw8eMcX4d1fh5ad\n8J83whfPhmc/C81bycmK8a5FM1l5x8U8+WcXcvHpk/naL+u55AvP8efffpEXtjQxlrryiohI5jnQ\n1sXanS0s1fp9IiIZLSvqAGQAzvljWPhOeOMn8MI34Jf/FGzzr4Q33YrNfxvVcydRPXcS2/e38UjN\nFh773VZ++Mouzps5kdsuruSacyrIyVKeLyIiw+t3m5twRwu3i4hkOCV+o0U8C6quCbbmrfDiN+HF\nR+DxG6BoOiy+CRbdxMzSWdxzzQI+esV8vvviDr7x/Cb+4j9f4u+fquOmpXO4YclsJhfmRv3TiIjI\nGFGzsZHcrBjnzy6JOhQRETkBzeo5miW64fUfB62AG34GZnD6lVB9G8y/CmJxkknnl2808ODzm/nl\n6w1B99DzZ3Dbm+dSVV4c9U8gInJKNKvn4KSzfrzmy7+ipCCbb39waVqeLyIig3O8OlItfqNZPAsW\nvCPY9m8JWgHXPBLMBFo8AxbfTGzRTVx25gwuO3MqG/Ye5BvPb+Z7L+7gP1dv46J5k7nt4kqWVU0l\nHtNyECIiI8nMlgNfBuLAA+7++T7nZwMPAyXhNXe7+1PhuXuA9wMJ4CPu/vRIxt6jua2Tut0tfPyK\nM6L4ehERGQQN+horSufA5f8XPrYW/uQRKKuCn38evnQ2fHsFvP40p08p4O/edQ419yzj7qur2Lzv\nEB/85mqW/fPPefDXmzjY3hX1TyEiMi6YWRy4D7gaWAhcb2YL+1z2SeAJd18ErAC+Et67MDw+C1gO\nfCV83oirrQ/G92nhdhGRzKcWv7Emng0Lrw22/ZvhhYdhzbfg9R9B8UxYfDMli2/izy6dxwfeXMnT\na/fw4PObuPd/1vEvz7zOe6pncutFc5kzeULUP4mIyFh2AbDB3esBzOxx4DpgXco1DvT0yZ8I7Az3\nrwMed/cOYJOZbQifVzMSgaeqrW8kPzvOeTM1vk9EJNMp8RvLSufCFZ+Gy+6B156CFx6Cn/89/OIf\n4IzlZFXfxtvPXsbbz63g5W3NfOP5TTxSs4WHfrOZKxZM47aL53LhaZMxUzdQEZFhNgPYlnK8HVjS\n55rPAD8xszuBCcAVKffW9rl3RnrCPLHa+kaq55Zq1mgRkVFAid94kJUDZ70z2Jrqg1bAlx6F134I\nE2fD4ps5b9GNfGnFIu65ZgHfqt3Co7/dyjPr9lBVXsT73lzJtedN18K8IiIj63rgIXf/ZzO7EHjE\nzM4e6M1mdjtwO8Ds2bOHPbjG1g7W7z7IH543fdifLSIiw0+/ohtvJp0GV/4NfGwdvOchmFQJz30W\nvngWPP6nTNv7a+66Yj6/uXsZX3j3uQD89ZOvcPHnn+Wff/Iae1vao41fRGRs2AHMSjmeGZalej/w\nBIC71wB5wJQB3ou73+/u1e5eXVZWNoyhB363qQlAC7eLiIwSavEbr7Jy4Kx3BVvjxqAb6EuPwvr/\ngZLZ5C2+hT9ZdBPvqX4LNfWNPPjrzfzbcxv4j19s5B3nTue2i+dyrsZ0iIicqlXAfDOrJEjaVgA3\n9LlmK3A58JCZLSBI/BqAlcC3zexfgOnAfOB3IxV4j5r6Rgpy4pw7c+JIf7WIiJwCJX4Ck+fBVX8L\nyz4ZJH6rvwHP/i38/HPYmddwUfVtXHTTZWzZf5iHfrOZ76zezvfX7KB6Tim3XVzJW6vKKMjRXyUR\nkYFy924zuwN4mmCphgfdfa2Z3QusdveVwF3A18zsYwQTvdzqweK7a83sCYKJYLqBP3f3xEj/DDUb\nG6meO4nsuDoPiYiMBlrAXfq3bwO8+BCseRQONwUTxSy+BRbdyMGsUr6zejsP12xmS2MbZjB38gQW\nVBSxsKKYBeFWMTFPE8OISFpoAffBGe76cV9rB9Wf/SmfWF7Fhy+bN2zPFRGRodMC7jI4U06Hqz4L\ny/4v1P130Ar4s7+B5/6eoqq3877q27jl45fw/MYm1mxtpm5XC2t3tvDUq7uPPGJifjYLKoqOJIIL\nK4qZP62Q3KxhmiQmmYRkFyQ6IdEVbp29x2aQWwx5xZBdEByLiMiQ1dY3AnCh1u8TERk1lPjJiWXl\nwjl/HGz73ugdC7juB8QnncYli2/mkopKmNoFZ3fS3tHO3uaD7G1upaH5IPubW2nefpjGZCe1dLPG\nEkzOMyYXGJPyjJIcpzgHci2RksB1hgldn0Sub3I3mJ5NsazeJDC3GPIm9h6n7ve9JvVcVp6SRxER\ngm6ehblZnD29+OQXi4hIRlDiJwM3ZT687e/CVsCVQSvgTz9z1CV5wOxwOyIGyexcEpZFN3E6klkc\nPhCjfX+cZrJoIAuPZZOdm0tObh4FeXkU5JdQkJ9PLCsH4jlB4hYP9+PZ4ZaT8tnPNZ6EjhZoPwDt\nLeF+S2/Z/s1Hl3GSbs+x7D5JYT9J5DHniiE35Tg7bzj/REREIlFb38gfzC0lS+P7RERGDSV+MnjZ\neXDunwRb81boaA2TrT6JVyxMymJxYmbEgGwgHygB9h/qpG53C3W7DlK3q4W6XS28saeVzkQSgJys\nGGdMK2RBeTELp4djB8uLmViQPfw/UzIJna1BQnhUgtgCHQeOkzy2QOvG3v3Ogyf/nnhukBhOnAEl\nc6BkNpTOgZK5wefEWUoORSSj7W1pZ2PDId77B7NOfrGIiGQMJX4yNCWnvihw6YQcLpo3hYvmTTlS\n1pVIsrGhNUwEg4Twudf28p0Xth+5ZkZJ/jFjB2dPKiAWG0I3zFgs7PY5hG5LyQR0HDxxK2P7AWhv\nhgPbYfer8NpTQbfVVIXlYTKYmhjOCT6LZwYJtohIRGrC8X1av09EZHTRvyAlo2THY1SVF1NVXsy7\nFvWW7z3YzrqdR7cOPvdaA4lk0D1zQk6cM8t7k8EFFcWcPrWQ4ryskZtZNBaH/JJgG6hkElp3w/4t\n0Lwl/Nwa7G+thd8/GXRZ7WFxKJ5xdDJYMrt3v7A8SGJFRNKktr6Jotwszpqu9ftEREYTJX4yKkwt\nymPqmXlcdubUI2XtXQne2BO0Dq4Lt5Uv7+TR3249ck1+dpxpxbnB/cW5TCvOY2pR+BmWTyvOpTB3\nBBPEVLEYFE8PtjkXHns+0RW0DvYkgz0JYvNW2PDTIGlMFc8JuosekxjODfYLJmfWBDXuQYtn12Ho\nbk/5bIOudug+HHx6MugeW1oJ+aWZ9TOIjDO19Y1cUDmJ+FB6WYiIyIhLa+JnZsuBLxMsTvuAu3++\nz/mPAx8gWIC2AXifu28JzyWAV8NLt7r7temMVUafvOw458ycyDkze3/r7O7saD5M3a6DbNrXyt6W\nDvYe7GBPSztrd7bw7Pq9tHUeOxvokQQxJTFMTRgjSxDj2TCpMtj603UYmreFieHmoxPDnS8FazCm\nyp5wbCthapfSnKIg8ToqCTveZ5/k7LifJ3nWySbV6St3YhDrpMpgfcnSns+5QdKrrrAiabP7QDub\n9h3iT5ecejd/ERGJRtr+hWRmceA+4EpgO7DKzFa6+7qUy9YA1e7eZmYfBr4AvDc8d9jdz09XfDI2\nmRkzSwuYWVoATOv3mtaObva0tIdJYfuR/T0HO9h7kgSxICfO1KIgQextPQwSxbKi3hbFEUsQs/Oh\n7Ixg60/HwaO7j6YmhpufH9iENAMVz4Gs/GBymqy8ILaez9wiKJx6bHlWXnh9fnB8zLn83mU0DmyH\npk3BbKz7N8OetbD+qWDpjx4Wh5JZRyeDqQniUMZwZppEFxzaB617oHUvHNrbu9+6N/izz0tZluSY\nreToZUs0qZAMQK3G94mIjFrp/NX4BcAGd68HMLPHgeuAI4mfuz+Xcn0tcGMa4xEBoDA3i8KyQuaV\nFZ7wup4EcU9LOw1hq2FPgrinpZ3f7zjA7gPtHO7qP0FMTQanFeUe6WpaVpTLpAk5lOTnUFKQTV72\nMC1o35/cIig/O9j6cofD+4Mkqicx7Dx0nOSs57MgJVHLOzrRi6Xx5wCoOO/YsmQCWnb2JoP7w8Sw\naROs+69jWzzzJx2bDPYkiMXT0/8znEwyAW1NYQK3Bw41HJ3MpZa1Nfb/jJwwyc4tCt5H+wE43Hx0\ngtyfnhlnB7SVHFumxHFcqNnYSHFeFgsqxtAvUURExol0Jn4zgG0px9uBJSe4/v3Aj1KO88xsNUE3\n0M+7+w+GP0SR4xtIgujutHZ0H+lO2pMg7knpYvrq9mZ+2tLRb4IIkJcdO5IElhRkU1oQ7E/Mz6E0\nLCspyKEkP/gsLchmYkE2uVlDTFLMoGBSsM1YPLRnRSUWtvCVzILKtxx7vme9xp5ksGd/xwuw9gfg\nKX8m8Zywy+vc3oRwUkpimDPh1GLsSbCPtMrt7SeZC/cPNRw9mU+PrPwgmSucCpNOg9lLYUJ4XDit\n99yEqZBT0H8M3e0pM8seOHqW2eOVN2/pPe47+2xf8dyTtDCmJI3l5x6/lVoyWu2mRpacNlnj+0RE\nRqGMGAxjZjcC1cClKcVz3H2HmZ0GPGtmr7r7xn7uvR24HWD2bI05kJFlZhTlZVOUlz2gBHFP2L20\nua0r2A53hvud7G/r4kBbFxsbWtkflnUljj/+LT87HiaBvQniyZLFkvwccrLG0ayfeRODlsL+WgsT\n3XBg29EthT0J4rbfBUtwpJow9dhksLQS8DCRazh+S11/rW2x7N6ErXgGVJwfJnHToLCsd39CWdB6\nN5Suw2a9XWmLyk/tGV39JY7HSxrDrXlb73WpieOyT0LZX536zyOR2Nl8mC2Nbdxy4dyoQxERkVOQ\nzsRvB5C6uuvMsOwoZnYF8H+AS929o6fc3XeEn/Vm9nNgEXBM4ufu9wP3A1RXVw9ylgiRkZGaIJ4+\n9cRdTHu4O22dCZoPB0ng8ZLF5rYuDhzu5PU9rUfKu5PH/09hQk48SAp7EsSU1sbC3GzysmPkZceD\nz6w4edlxcnvKsuIp53uvGdIailGJZ6VMnPPWo88d6Qa76djWwi2/gVeeoN9JaSwWJGo9LXFlC3qT\nu56WuZ6WutE2O2l2OB6zqP+xsyeVmjgOZskTyRg1GzW+T0RkNEtn4rcKmG9mlQQJ3wrghtQLzGwR\n8FVgubvvTSkvBdrcvcPMpgAXE0z8IjJumBkTcrOYkJvFjJL8Ad/n7hzqTByVLO5v66T5cBcH+iSL\n+9u6WH+gJUwou46sizhYOfFYb3KYkjD2JIm5RyWMwfncPtfl9iSTWX0SyzDhzIpbsMViZMWN7PAz\nK2bDP5HOUd0NpPtgAAAJ+UlEQVRg33Ts+e6OoDVr/+bg2p7WuYJJ0Y8TzFRDTRwlcrX1jZQWZFNV\nXhR1KCIicgrSlvi5e7eZ3QE8TbCcw4PuvtbM7gVWu/tK4B+BQuA74T/cepZtWAB81cySQIxgjN+6\nfr9IRI5iZsH4xNwsZpYO/D53p70rSXtXgvbuRO9+V7jfnaCjq095d7L3fFeCjn7ua+3oZl9rZ3hv\n6j0JTjHPPEY8FiSA2fEY8ZiRnZIgZsWMrHjspOez40Y8FiM7FiaY4T1ZseBc1lHnY+Gz5hGPGbbP\niFkbxmFiBjEzzII/i/6OjfAzLI+lXEfK8fGu631W73WW8j095dnxGNlZQfzZsaP3R2UrrUSqpr6R\nJZWT9XdHRGSUSusYP3d/CniqT9mnUvavOM59vwHOSWdsInI0MyM/J05+zsi0WLk7XQkPk8wEHV3J\no5LMoxPKJN2JJF1JJ5FI0p0M7u0O97uTSboTQVkiGVzXnQjKes4H55yuI+VJ2rs9vC94Tt/z3cne\n84mkn7AL7WjTkwBnh0ns4PdPcC4rTDTjFiSbsbAsPD9/aiGnnWRWXcks25ra2L7/MB9483HWFBUR\nkYyXEZO7iMj4Y2bkZBk5WTGK87KjDmdA3P1IMtiTbCbccQ/OJR2c4DOZDMt7jvu7LhmUQ/DZ73Up\nn55ynYfXJY9zXXdK8tuZSB5JaHv2+x737jtd3Um6k0k6w/2uRJJDnYkj+91JpzNlv6s7eeQ5A8mN\n77ryDO68fH5a/6xkePWs33fhvCkRRyIiIqdKiZ+IyACZ9bSSQdCDXfrqaTXtCpPI7kTySELZs19W\nmBt1mDJI154/nTmTJ3DGNLXUioiMVkr8RERk2MRjRjwWTM4jY0duVpwLKidFHYaIiAzBOFrQS0RE\nREREZHxS4iciIiIiIjLGKfETEREREREZ45T4iYiIRMDMlpvZa2a2wczu7uf8F83spXB73cyaU859\nwczWmlmdmf2rhYvhioiIHI8mdxERERlhZhYH7gOuBLYDq8xspbuv67nG3T+Wcv2dwKJw/yLgYuDc\n8PSvgUuBn49I8CIiMiqpxU9ERGTkXQBscPd6d+8EHgeuO8H11wOPhfsO5AE5QC6QDexJY6wiIjIG\nKPETEREZeTOAbSnH28OyY5jZHKASeBbA3WuA54Bd4fa0u9elNVoRERn1lPiJiIhkthXAk+6eADCz\n04EFwEyCZHGZmb2l701mdruZrTaz1Q0NDSMasIiIZB4lfiIiIiNvBzAr5XhmWNafFfR28wR4F1Dr\n7q3u3gr8CLiw703ufr+7V7t7dVlZ2TCFLSIio5USPxERkZG3CphvZpVmlkOQ3K3se5GZVQGlQE1K\n8VbgUjPLMrNsgold1NVTREROyNw96hiGjZk1AFuG+JgpwL5hCGc80TsbPL2zwdH7Gryx/s7muPuo\nbsYys2uALwFx4EF3/zszuxdY7e4rw2s+A+S5+90p98WBrwCXEEz08mN3//hJvkv1YzT0zgZP72zw\n9M4Gb6y/s37ryDGV+A0HM1vt7tVRxzGa6J0Nnt7Z4Oh9DZ7emQw3/Z0aPL2zwdM7Gzy9s8Ebr+9M\nXT1FRERERETGOCV+IiIiIiIiY5wSv2PdH3UAo5De2eDpnQ2O3tfg6Z3JcNPfqcHTOxs8vbPB0zsb\nvHH5zjTGT0REREREZIxTi5+IiIiIiMgYp8QvZGbLzew1M9tgZnef/I7xzcxmmdlzZrbOzNaa2Uej\njmm0MLO4ma0xs/+JOpbRwMxKzOxJM1tvZnVmdsxC1XI0M/tY+N/l783sMTPLizomGd1URw6O6shT\no/px8FRHDs54rx+V+HFkTaT7gKuBhcD1ZrYw2qgyXjdwl7svBJYCf653NmAfRYstD8aXCdYpqwLO\nQ+/uhMxsBvARoNrdzyZYI25FtFHJaKY68pSojjw1qh8HT3XkAKl+VOLX4wJgg7vXu3sn8DhwXcQx\nZTR33+XuL4b7Bwn+RzMj2qgyn5nNBN4OPBB1LKOBmU0kWKT66wDu3unuzdFGNSpkAflmlgUUADsj\njkdGN9WRg6Q6cvBUPw6e6shTMq7rRyV+gRnAtpTj7eh/0ANmZnOBRcBvo41kVPgS8NdAMupARolK\noAH4Rtj95wEzmxB1UJnM3XcA/wRsBXYBB9z9J9FGJaOc6sghUB05YKofB0915CCoflTiJ0NkZoXA\nd4G/cPeWqOPJZGb2DmCvu78QdSyjSBawGPh3d18EHAI0vugEzKyUoDWmEpgOTDCzG6ONSmR8Uh05\nMKofT5nqyEFQ/ajEr8cOYFbK8cywTE7AzLIJKrRH3f17UcczClwMXGtmmwm6Si0zs29FG1LG2w5s\nd/ee35Q/SVDJyfFdAWxy9wZ37wK+B1wUcUwyuqmOPAWqIwdF9eOpUR05OOO+flTiF1gFzDezSjPL\nIRjouTLimDKamRlBn/I6d/+XqOMZDdz9Hnef6e5zCf6OPevu4+o3TYPl7ruBbWZ2Zlh0ObAuwpBG\ng63AUjMrCP87vRwN9pehUR05SKojB0f146lRHTlo475+zIo6gEzg7t1mdgfwNMEMPw+6+9qIw8p0\nFwM3Aa+a2Uth2f9296cijEnGpjuBR8N/cNYDt0UcT0Zz99+a2ZPAiwQzC64B7o82KhnNVEeeEtWR\nMlJURw6Q6kcwd486BhEREREREUkjdfUUEREREREZ45T4iYiIiIiIjHFK/ERERERERMY4JX4iIiIi\nIiJjnBI/ERERERGRMU6Jn0iGMLOEmb2Ust09jM+ea2a/H67niYiIjBTVjyLDQ+v4iWSOw+5+ftRB\niIiIZBjVjyLDQC1+IhnOzDab2RfM7FUz+52ZnR6WzzWzZ83sFTP7mZnNDsunmdn3zezlcLsofFTc\nzL5mZmvN7Cdmlh/ZDyUiIjJEqh9FBkeJn0jmyO/TleW9KecOuPs5wL8BXwrL/h/wsLufCzwK/GtY\n/q/AL9z9PGAxsDYsnw/c5+5nAc3Au9P884iIiAwH1Y8iw8DcPeoYRAQws1Z3L+ynfDOwzN3rzSwb\n2O3uk81sH1Dh7l1h+S53n2JmDcBMd+9IecZc4Bl3nx8efwLIdvfPpv8nExEROXWqH0WGh1r8REYH\nP87+YHSk7CfQGF8RERn9VD+KDJASP5HR4b0pnzXh/m+AFeH+nwK/Cvd/BnwYwMziZjZxpIIUEREZ\nYaofRQZIv9EQyRz5ZvZSyvGP3b1nyupSM3uF4LeS14dldwLfMLO/AhqA28LyjwL3m9n7CX5z+WFg\nV9qjFxERSQ/VjyLDQGP8RDJcOIah2t33RR2LiIhIplD9KDI46uopIiIiIiIyxqnFT0REREREZIxT\ni5+IiIiIiMgYp8RPRERERERkjFPiJyIiIiIiMsYp8RMRERERERnjlPiJiIiIiIiMcUr8RERERERE\nxrj/DymGwN8fUwh5AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1080x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ac3oQ4a108no",
        "colab_type": "text"
      },
      "source": [
        "# Quantitative Accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KoTEbgI10_yp",
        "colab_type": "code",
        "outputId": "5ce243bd-adfd-44d6-f713-35edafb84af9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "test_loss, test_acc = model.evaluate(x_test,  y_test, verbose=2)\n",
        "print('\\nTest accuracy:', test_acc)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "5000/1 - 0s - loss: 0.1398 - accuracy: 0.9442\n",
            "\n",
            "Test accuracy: 0.9442\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MTEOMGhO1QYw",
        "colab_type": "text"
      },
      "source": [
        "# Qualitative Analysis"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z-k0CY5H1TDw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot_image(i, predictions_array, true_label, img):\n",
        "  predictions_array, true_label, img = predictions_array, true_label[i], img[i]\n",
        "  plt.grid(False)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.imshow(img, cmap=plt.cm.binary)\n",
        "  predicted_label = np.argmax(predictions_array)\n",
        "  if predicted_label == true_label:\n",
        "    color = 'blue'\n",
        "  else:\n",
        "    color = 'red'\n",
        "  plt.xlabel(\"{} {:2.0f}% \".format('Highest Probability', 100*np.max(predictions_array)), color=color)\n",
        "\n",
        "def plot_value_array(i, predictions_array, true_label):\n",
        "  predictions_array, true_label = predictions_array, true_label[i]\n",
        "  plt.grid(False)\n",
        "  plt.xticks(range(10))\n",
        "  plt.yticks([])\n",
        "  thisplot = plt.bar(range(10), predictions_array, color=\"#777777\")\n",
        "  plt.ylim([0, 1])\n",
        "  predicted_label = np.argmax(predictions_array)\n",
        "\n",
        "  thisplot[predicted_label].set_color('red')\n",
        "  if not isinstance(true_label, int):\n",
        "      thisplot[true_label[0]].set_color('blue')\n",
        "  else:\n",
        "      thisplot[true_label].set_color('blue')\n",
        "  cmap = plt.cm.coolwarm\n",
        "  custom_lines = [Line2D([0], [0], color=cmap(0.), lw=4),\n",
        "                Line2D([0], [0], color=cmap(.5), lw=4),\n",
        "                Line2D([0], [0], color=cmap(1.), lw=4)]\n",
        "  plt.legend(custom_lines, ['True Label', 'Other Labels', 'Predicted Label'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0lqvx8e3K_0m",
        "colab_type": "code",
        "outputId": "3e793412-c210-4ae8-9fb8-86220a13ab98",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 228
        }
      },
      "source": [
        "predictions = model.predict(x_test)\n",
        "i = np.random.randint(0, y_test.shape[0])\n",
        "plt.figure(figsize=(6,3))\n",
        "plt.subplot(1,2,1)\n",
        "plot_image(i, predictions[i], y_test, x_test)\n",
        "plt.subplot(1,2,2)\n",
        "plot_value_array(i, predictions[i],  y_test)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[3]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAADCCAYAAAB3whgdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQuL/AxXEwKMcq9FdAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x216 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aVf003n9u-Ca",
        "colab_type": "text"
      },
      "source": [
        "# Convolution Neural Networks\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bO6Gn6ChKmCO",
        "colab_type": "text"
      },
      "source": [
        "![alt text](https://i.ibb.co/3Y79sCZ/C3-CB95-B9-D9-B4-4063-B753-22-CF51-B8-C884-png.jpg)\n",
        "![alt text](http://cs231n.github.io/assets/cnn/cnn.jpeg)\n",
        "![alt text](https://miro.medium.com/max/1170/1*Fw-ehcNBR9byHtho-Rxbtw.gif)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xZKF2qa6vCwS",
        "colab_type": "code",
        "outputId": "9b7018e2-32f0-49b7-de26-42ac7cfecf24",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 765
        }
      },
      "source": [
        "from keras.datasets import cifar10, mnist, cifar100, fashion_mnist\n",
        "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
        "x_train = x_train.reshape(x_train.shape[0], x_train.shape[1], x_train.shape[2], 1)\n",
        "x_test = x_test.reshape(x_test.shape[0], x_test.shape[1], x_test.shape[2], 1)\n",
        "x_val, x_test, y_val, y_test = train_test_split(x_test, y_test, test_size=0.5)\n",
        "\n",
        "number_of_classes = y_train.max() - y_train.min() + 1\n",
        "model = Sequential()\n",
        "model.add(keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=x_train.shape[1:]))\n",
        "model.add(keras.layers.MaxPooling2D((2, 2)))\n",
        "model.add(keras.layers.Conv2D(64, (3, 3), activation='relu'))\n",
        "model.add(keras.layers.MaxPooling2D((2, 2)))\n",
        "model.add(keras.layers.Flatten())\n",
        "model.add(keras.layers.Dense(64, activation='relu'))\n",
        "model.add(keras.layers.Dense(number_of_classes, activation='softmax'))\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n",
        "model.fit(x_train, y_train, epochs=epochs, validation_data=(x_val, y_val))\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_6 (Conv2D)            (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 13, 13, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_7 (Conv2D)            (None, 11, 11, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 5, 5, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_3 (Flatten)          (None, 1600)              0         \n",
            "_________________________________________________________________\n",
            "dense_6 (Dense)              (None, 64)                102464    \n",
            "_________________________________________________________________\n",
            "dense_7 (Dense)              (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 121,930\n",
            "Trainable params: 121,930\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train on 60000 samples, validate on 5000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 12s 193us/sample - loss: 0.3140 - accuracy: 0.9439 - val_loss: 0.0804 - val_accuracy: 0.9728\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 11s 183us/sample - loss: 0.0667 - accuracy: 0.9807 - val_loss: 0.0658 - val_accuracy: 0.9836\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 11s 183us/sample - loss: 0.0547 - accuracy: 0.9838 - val_loss: 0.0522 - val_accuracy: 0.9864\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 11s 183us/sample - loss: 0.0451 - accuracy: 0.9865 - val_loss: 0.0680 - val_accuracy: 0.9798\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 11s 183us/sample - loss: 0.0382 - accuracy: 0.9885 - val_loss: 0.0527 - val_accuracy: 0.9868\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 11s 185us/sample - loss: 0.0323 - accuracy: 0.9897 - val_loss: 0.0620 - val_accuracy: 0.9842\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 11s 184us/sample - loss: 0.0270 - accuracy: 0.9917 - val_loss: 0.0512 - val_accuracy: 0.9870\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 11s 183us/sample - loss: 0.0222 - accuracy: 0.9933 - val_loss: 0.0587 - val_accuracy: 0.9858\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 11s 188us/sample - loss: 0.0221 - accuracy: 0.9934 - val_loss: 0.0614 - val_accuracy: 0.9862\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 11s 183us/sample - loss: 0.0203 - accuracy: 0.9940 - val_loss: 0.0578 - val_accuracy: 0.9894\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fd7b7fa4be0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3J6zqWmLzf4c",
        "colab_type": "code",
        "outputId": "da9c3654-3248-45e7-da79-836dc6dd8161",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 351
        }
      },
      "source": [
        "fig, axs = plt.subplots(1,2,figsize=(15,5))\n",
        "epochs_range = range(len(model.history.history['loss']))\n",
        "axs[0].plot(epochs_range, model.history.history['loss'])\n",
        "axs[0].plot(epochs_range, model.history.history['val_loss'])\n",
        "axs[0].set_xlabel('Epoch')\n",
        "axs[0].set_ylabel('Loss')\n",
        "axs[0].legend(['loss', 'val_loss'])\n",
        "\n",
        "axs[1].plot(epochs_range, model.history.history['accuracy'])\n",
        "axs[1].plot(epochs_range, model.history.history['val_accuracy'])\n",
        "axs[1].set_xlabel('Epoch')\n",
        "axs[1].set_ylabel('Accuracy')\n",
        "axs[1].legend(['accuracy', 'val_accuracy'])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fd7b9d15be0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA34AAAE9CAYAAABUerD/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIvf/AeiaTTPGLn5VAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1080x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YxTUPne4zjF0",
        "colab_type": "code",
        "outputId": "ff8339a2-cc41-4c0f-fba6-21189b74a82c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "test_loss, test_acc = model.evaluate(x_test,  y_test, verbose=2)\n",
        "print('\\nTest accuracy:', test_acc)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "5000/1 - 0s - loss: 0.0266 - accuracy: 0.9878\n",
            "\n",
            "Test accuracy: 0.9878\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8ru2WeX5JD8E",
        "colab_type": "text"
      },
      "source": [
        "# CIFAR Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9UQeErbfJGso",
        "colab_type": "code",
        "outputId": "f84d6bdf-4852-4fb7-9053-86594ff88574",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 947
        }
      },
      "source": [
        "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
        "x_val, x_test, y_val, y_test = train_test_split(x_test, y_test, test_size=0.5, random_state=1)\n",
        "\n",
        "number_of_classes = y_train.max() - y_train.min() + 1\n",
        "image_shape = x_train.shape[1:]\n",
        "\n",
        "model = Sequential()\n",
        "model.add(keras.layers.Flatten(input_shape=image_shape))\n",
        "model.add(keras.layers.Dense(64, activation='relu'))\n",
        "model.add(keras.layers.Dense(number_of_classes, activation='softmax'))\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n",
        "model.fit(x_train, y_train, epochs=epochs, validation_data=(x_val, y_val))\n",
        "\n",
        "fig, axs = plt.subplots(1,2,figsize=(15,5))\n",
        "epochs_range = range(len(model.history.history['loss']))\n",
        "axs[0].plot(epochs_range, model.history.history['loss'])\n",
        "axs[0].plot(epochs_range, model.history.history['val_loss'])\n",
        "axs[0].set_xlabel('Epoch')\n",
        "axs[0].set_ylabel('Loss')\n",
        "axs[0].legend(['loss', 'val_loss'])\n",
        "\n",
        "axs[1].plot(epochs_range, model.history.history['accuracy'])\n",
        "axs[1].plot(epochs_range, model.history.history['val_accuracy'])\n",
        "axs[1].set_xlabel('Epoch')\n",
        "axs[1].set_ylabel('Accuracy')\n",
        "axs[1].legend(['accuracy', 'val_accuracy'])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "flatten_1 (Flatten)          (None, 3072)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 64)                196672    \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 197,322\n",
            "Trainable params: 197,322\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train on 50000 samples, validate on 5000 samples\n",
            "Epoch 1/10\n",
            "50000/50000 [==============================] - 5s 97us/sample - loss: 5.3757 - accuracy: 0.0986 - val_loss: 2.3027 - val_accuracy: 0.0940\n",
            "Epoch 2/10\n",
            "50000/50000 [==============================] - 4s 84us/sample - loss: 2.3027 - accuracy: 0.0991 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 3/10\n",
            "50000/50000 [==============================] - 4s 83us/sample - loss: 2.3028 - accuracy: 0.0989 - val_loss: 2.3025 - val_accuracy: 0.1014\n",
            "Epoch 4/10\n",
            "50000/50000 [==============================] - 4s 80us/sample - loss: 2.3027 - accuracy: 0.0970 - val_loss: 2.3028 - val_accuracy: 0.0990\n",
            "Epoch 5/10\n",
            "50000/50000 [==============================] - 4s 85us/sample - loss: 2.3028 - accuracy: 0.0974 - val_loss: 2.3026 - val_accuracy: 0.1032\n",
            "Epoch 6/10\n",
            "50000/50000 [==============================] - 4s 82us/sample - loss: 2.3027 - accuracy: 0.0996 - val_loss: 2.3027 - val_accuracy: 0.0950\n",
            "Epoch 7/10\n",
            "50000/50000 [==============================] - 4s 82us/sample - loss: 2.3028 - accuracy: 0.0985 - val_loss: 2.3029 - val_accuracy: 0.0950\n",
            "Epoch 8/10\n",
            "50000/50000 [==============================] - 4s 81us/sample - loss: 2.3028 - accuracy: 0.0970 - val_loss: 2.3026 - val_accuracy: 0.1018\n",
            "Epoch 9/10\n",
            "50000/50000 [==============================] - 4s 83us/sample - loss: 2.3028 - accuracy: 0.0999 - val_loss: 2.3026 - val_accuracy: 0.0990\n",
            "Epoch 10/10\n",
            "50000/50000 [==============================] - 4s 83us/sample - loss: 2.3028 - accuracy: 0.0976 - val_loss: 2.3025 - val_accuracy: 0.1032\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f360e450908>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAE+CAYAAADf8LIWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 1080x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M42U5NTusm0e",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predictions = model.predict(x_test)\n",
        "i = np.random.randint(0, y_test.shape[0])\n",
        "plt.figure(figsize=(6,3))\n",
        "plt.subplot(1,2,1)\n",
        "plot_image(i, predictions[i], y_test, x_test)\n",
        "plt.subplot(1,2,2)\n",
        "plot_value_array(i, predictions[i],  y_test)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rCRR1C9wJfzK",
        "colab_type": "code",
        "outputId": "730db778-d89d-4f25-dd9a-385f966ae7e6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "model = Sequential()\n",
        "model.add(keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=x_train.shape[1:]))\n",
        "model.add(keras.layers.MaxPooling2D((2, 2)))\n",
        "model.add(keras.layers.Conv2D(64, (3, 3), activation='relu'))\n",
        "model.add(keras.layers.MaxPooling2D((2, 2)))\n",
        "model.add(keras.layers.Flatten())\n",
        "model.add(keras.layers.Dense(64, activation='relu'))\n",
        "model.add(keras.layers.Dense(number_of_classes, activation='softmax'))\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n",
        "model.fit(x_train, y_train, epochs=epochs, validation_data=(x_val, y_val))\n",
        "\n",
        "fig, axs = plt.subplots(1,2,figsize=(15,5))\n",
        "epochs_range = range(len(model.history.history['loss']))\n",
        "axs[0].plot(epochs_range, model.history.history['loss'])\n",
        "axs[0].plot(epochs_range, model.history.history['val_loss'])\n",
        "axs[0].set_xlabel('Epoch')\n",
        "axs[0].set_ylabel('Loss')\n",
        "axs[0].legend(['loss', 'val_loss'])\n",
        "\n",
        "axs[1].plot(epochs_range, model.history.history['accuracy'])\n",
        "axs[1].plot(epochs_range, model.history.history['val_accuracy'])\n",
        "axs[1].set_xlabel('Epoch')\n",
        "axs[1].set_ylabel('Accuracy')\n",
        "axs[1].legend(['accuracy', 'val_accuracy'])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_2 (Conv2D)            (None, 30, 30, 32)        896       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 15, 15, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 13, 13, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 6, 6, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_3 (Flatten)          (None, 2304)              0         \n",
            "_________________________________________________________________\n",
            "dense_6 (Dense)              (None, 64)                147520    \n",
            "_________________________________________________________________\n",
            "dense_7 (Dense)              (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 167,562\n",
            "Trainable params: 167,562\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Train on 50000 samples, validate on 5000 samples\n",
            "Epoch 1/10\n",
            "50000/50000 [==============================] - 6s 122us/sample - loss: 1.9988 - accuracy: 0.3776 - val_loss: 1.5029 - val_accuracy: 0.4702\n",
            "Epoch 2/10\n",
            "50000/50000 [==============================] - 6s 110us/sample - loss: 1.2956 - accuracy: 0.5442 - val_loss: 1.3104 - val_accuracy: 0.5498\n",
            "Epoch 3/10\n",
            "50000/50000 [==============================] - 5s 106us/sample - loss: 1.1423 - accuracy: 0.6013 - val_loss: 1.1557 - val_accuracy: 0.6020\n",
            "Epoch 4/10\n",
            "50000/50000 [==============================] - 5s 106us/sample - loss: 1.0330 - accuracy: 0.6404 - val_loss: 1.1965 - val_accuracy: 0.6082\n",
            "Epoch 5/10\n",
            "50000/50000 [==============================] - 6s 111us/sample - loss: 0.9478 - accuracy: 0.6725 - val_loss: 1.0738 - val_accuracy: 0.6338\n",
            "Epoch 6/10\n",
            "50000/50000 [==============================] - 5s 105us/sample - loss: 0.8828 - accuracy: 0.6943 - val_loss: 1.1542 - val_accuracy: 0.6228\n",
            "Epoch 7/10\n",
            "50000/50000 [==============================] - 5s 104us/sample - loss: 0.8135 - accuracy: 0.7172 - val_loss: 1.0564 - val_accuracy: 0.6452\n",
            "Epoch 8/10\n",
            "50000/50000 [==============================] - 6s 110us/sample - loss: 0.7571 - accuracy: 0.7374 - val_loss: 1.1272 - val_accuracy: 0.6202\n",
            "Epoch 9/10\n",
            "50000/50000 [==============================] - 5s 107us/sample - loss: 0.7049 - accuracy: 0.7558 - val_loss: 1.1384 - val_accuracy: 0.6510\n",
            "Epoch 10/10\n",
            "50000/50000 [==============================] - 5s 105us/sample - loss: 0.6514 - accuracy: 0.7754 - val_loss: 1.1117 - val_accuracy: 0.6620\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7ff62e52d8d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAE9CAYAAABZZMC4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
            "text/plain": [
              "<Figure size 1080x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QqEUBt-asnvZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predictions = model.predict(x_test)\n",
        "i = np.random.randint(0, y_test.shape[0])\n",
        "plt.figure(figsize=(6,3))\n",
        "plt.subplot(1,2,1)\n",
        "plot_image(i, predictions[i], y_test, x_test)\n",
        "plt.subplot(1,2,2)\n",
        "plot_value_array(i, predictions[i],  y_test)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}