conditional-generative-adversarial-network.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introduction to a Conditional Generative Adversarial Network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prerequisites\n",
    "\n",
    "### Knowledge\n",
    "- https://arxiv.org/abs/1411.1784\n",
    "\n",
    "### Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "import tensorflow as tf\n",
    "import logging\n",
    "tf.get_logger().setLevel(logging.ERROR)\n",
    "\n",
    "import tensorflow_datasets as tfds"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1) -1.0 1.0\n",
      "(60000,)\n"
     ]
    }
   ],
   "source": [
    "# defining some important konstants\n",
    "z_dim = 100\n",
    "batch_size = 50\n",
    "num_classes = 10\n",
    "\n",
    "# Load the dataset\n",
    "(X_train, y_train), (_, _) = tf.keras.datasets.mnist.load_data()\n",
    "\n",
    "# Rescale -1 to 1\n",
    "X_train = X_train / 127.5 - 1.\n",
    "X_train = np.expand_dims(X_train, axis=3)\n",
    "\n",
    "# Labels for real and fake examples\n",
    "real = np.ones((batch_size, 1))\n",
    "fake = np.zeros((batch_size, 1))\n",
    "\n",
    "print(X_train.shape, np.min(X_train), np.max(X_train))\n",
    "print(y_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CGAN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Generator(tf.keras.Model):\n",
    "    \n",
    "    def __init__(self, num_classes, z_dim):\n",
    "        super(Generator, self).__init__()\n",
    "        self.model = tf.keras.Sequential([# Reshape input into 7x7x128 tensor via a fully connected layer\n",
    "                                          tf.keras.layers.Dense(256 * 7 * 7),\n",
    "                                          tf.keras.layers.Reshape((7, 7,256)),\n",
    "                                          # Transposed convolution layer, from 7x7x256 into 14x14x128 tensor\n",
    "                                          tf.keras.layers.Conv2DTranspose(128, kernel_size=3, strides=2, padding='same'),\n",
    "                                          tf.keras.layers.BatchNormalization(fused=False),\n",
    "                                          tf.keras.layers.LeakyReLU(alpha=0.2),\n",
    "                                          # Transposed convolution layer, from 14x14x128 to 14x14x64 tensor\n",
    "                                          tf.keras.layers.Conv2DTranspose(64, kernel_size=3, strides=1, padding='same'),\n",
    "                                          tf.keras.layers.BatchNormalization(fused=False),\n",
    "                                          tf.keras.layers.LeakyReLU(alpha=0.2),\n",
    "                                          # Transposed convolution layer, from 14x14x64 to 28x28x1 tensor\n",
    "                                          tf.keras.layers.Conv2DTranspose(1, kernel_size=3, strides=2, padding='same'),\n",
    "                                          tf.keras.layers.Activation('tanh')])\n",
    "        self.label_embedding = tf.keras.layers.Embedding(num_classes, z_dim, input_length=1, dtype=tf.float64)\n",
    "        self.flatten = tf.keras.layers.Flatten()\n",
    "        \n",
    "    def call(self, x):\n",
    "        z_sample, label = x\n",
    "        label_embedding = self.label_embedding(label)\n",
    "        flatted_label_embedding = self.flatten(label_embedding)\n",
    "        joined_representation = tf.math.multiply(z_sample, flatted_label_embedding)\n",
    "        img = self.model(joined_representation)\n",
    "        return img\n",
    "\n",
    "generator = Generator(num_classes, z_dim)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discriminator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Discriminator(tf.keras.Model):\n",
    "    \n",
    "    def __init__(self, num_classes, img_shape):\n",
    "        super(Discriminator, self).__init__()\n",
    "        self.model = tf.keras.Sequential([# Convolutional layer, from 28x28x1 into 14x14x32 tensor\n",
    "                                          tf.keras.layers.Conv2D(64, kernel_size=3, strides=2, padding='same'),\n",
    "                                          tf.keras.layers.LeakyReLU(alpha=0.2),\n",
    "                                          # Convolutional layer, from 14x14x32 into 7x7x64 tensor\n",
    "                                          tf.keras.layers.Conv2D(64, kernel_size=3, strides=2, padding='same'),\n",
    "                                          tf.keras.layers.BatchNormalization(fused=False),\n",
    "                                          tf.keras.layers.LeakyReLU(alpha=0.2),\n",
    "                                          # Convolutional layer, from 7x7x64 tensor into 3x3x128 tensor\n",
    "                                          tf.keras.layers.Conv2D(128, kernel_size=3, strides=2, padding='same'),\n",
    "                                          tf.keras.layers.BatchNormalization(fused=False),\n",
    "                                          tf.keras.layers.LeakyReLU(alpha=0.2),\n",
    "                                          # Flatten the tensor and apply sigmoid activation function\n",
    "                                          tf.keras.layers.Flatten(),\n",
    "                                          tf.keras.layers.Dense(1, activation='sigmoid')])\n",
    "        self.label_embedding = tf.keras.layers.Embedding(num_classes, np.prod(img_shape), input_length=1, dtype=tf.float64)\n",
    "        self.concat = tf.keras.layers.Concatenate(axis=-1)\n",
    "        self.img_shape = img_shape\n",
    "        \n",
    "    def call(self, x):\n",
    "        img, label = x\n",
    "        label_embedding = self.label_embedding(label)\n",
    "        reshaped_label_embedding = tf.reshape(label_embedding, [-1, self.img_shape[0], self.img_shape[1], self.img_shape[2]])\n",
    "        concatenated = self.concat([img, reshaped_label_embedding])\n",
    "        prediction = self.model(concatenated)\n",
    "        return prediction\n",
    "\n",
    "discriminator = Discriminator(num_classes, (28, 28, 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator_opt = tf.keras.optimizers.Adam()\n",
    "discriminator_opt = tf.keras.optimizers.Adam()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When training the generator, you use"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "@tf.function\n",
    "def train_discriminator(img_and_label, label):\n",
    "    with tf.GradientTape() as tape:\n",
    "        # calculate loss\n",
    "        prediction = discriminator(img_and_label)\n",
    "        loss = tf.reduce_mean(tf.keras.losses.binary_crossentropy(y_true=label, y_pred=prediction))\n",
    "        accuracy = tf.reduce_mean(tf.keras.metrics.binary_accuracy(y_true=label, y_pred=prediction))\n",
    "\n",
    "        # gradient descent and only update discriminator\n",
    "        gradient = tape.gradient(loss, discriminator.trainable_variables)\n",
    "        discriminator_opt.apply_gradients(zip(gradient, discriminator.trainable_variables))\n",
    "    return loss, accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "@tf.function\n",
    "def train_generator(sample_and_label, label):\n",
    "    with tf.GradientTape() as tape:\n",
    "        image = generator(sample_and_label)\n",
    "        prediction = discriminator([image, sample_and_label[1]])\n",
    "        loss = tf.reduce_mean(tf.keras.losses.binary_crossentropy(y_true=label, y_pred=prediction))\n",
    "        # gradient descent and only update generator\n",
    "        gradients = tape.gradient(loss, generator.trainable_variables)\n",
    "        generator_opt.apply_gradients(zip(gradients, generator.trainable_variables))\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 0 d_loss 0.7184702635589544 g_loss 0.6863342644726468 accuracy 0.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 2000 d_loss 0.0006694221584597513 g_loss 8.9094727458689 accuracy 1.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 4000 d_loss 0.17648908886972342 g_loss 5.554217104309655 accuracy 0.949999988079071\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 6000 d_loss 0.21547514413880636 g_loss 3.7847938480376446 accuracy 0.9399999976158142\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 8000 d_loss 0.37768603214029506 g_loss 2.1620883897700893 accuracy 0.8299999833106995\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# init lists for losses and accuracies\n",
    "d_losses, d_accuracies, g_losses = [], [], []\n",
    "\n",
    "ITERATIONS = 10000\n",
    "for i in range(ITERATIONS):\n",
    "    # pick a random sample from the datas\n",
    "    idx = np.random.randint(0, X_train.shape[0], batch_size)\n",
    "    imgs, labels = X_train[idx], y_train[idx]    \n",
    "    \n",
    "    # Generate a batch of fake images\n",
    "    z_sample = np.random.normal(0, 1, (batch_size, z_dim))\n",
    "    gen_image = generator([z_sample, labels])\n",
    "\n",
    "    # Discriminator loss and train discriminator on fake and real images\n",
    "    d_loss_real, d_accuracy_real = train_discriminator([imgs, labels], real)\n",
    "    d_loss_fake, d_accuracy_fake = train_discriminator([gen_image, labels], fake)\n",
    "    d_loss = 0.5 * np.add(d_loss_real, d_loss_fake)\n",
    "    d_accuracy = 0.5 * np.add(d_accuracy_real, d_accuracy_fake)\n",
    "\n",
    "    # Generator loss\n",
    "    labels = np.random.randint(0, num_classes, batch_size).reshape(-1, 1) # get random labels\n",
    "    g_loss = train_generator([z_sample, labels], real)\n",
    "    g_loss = g_loss.numpy()\n",
    "\n",
    "    # visualization\n",
    "    if i % 2000 == 0:\n",
    "        print(\"Iteration\", i, \"d_loss\", d_loss, \"g_loss\", g_loss, \"accuracy\", d_accuracy)\n",
    "        image_grid_rows = 2\n",
    "        image_grid_columns = 5\n",
    "        # Sample random noise\n",
    "        z = np.random.normal(0, 1, \n",
    "                  (image_grid_rows * image_grid_columns, z_dim))\n",
    "\n",
    "        labels = np.arange(0, 10).reshape(-1, 1)\n",
    "        # Generate images from random noise\n",
    "        gen_imgs = generator([z, labels])\n",
    "        # Rescale images to 0-1\n",
    "        gen_imgs = 0.5 * gen_imgs + 0.5\n",
    "        # Set image grid\n",
    "        fig, axs = plt.subplots(image_grid_rows, image_grid_columns, \n",
    "                                        figsize=(10,4), sharey=True, sharex=True)\n",
    "        cnt = 0\n",
    "        for i in range(image_grid_rows):\n",
    "            for j in range(image_grid_columns):\n",
    "                # Output image grid\n",
    "                axs[i,j].imshow(gen_imgs[cnt, :,:,0], cmap='gray')\n",
    "                axs[i,j].axis('off')\n",
    "                axs[i,j].set_title(\"Digit: %d\" % labels[cnt])\n",
    "                cnt += 1\n",
    "        plt.show()\n",
    "        \n",
    "        d_losses.append(d_loss)\n",
    "        d_accuracies.append(d_accuracy)\n",
    "        g_losses.append(g_loss)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x1440 with 50 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "image_grid_rows = 10\n",
    "image_grid_columns = 5\n",
    "\n",
    "# Sample random noise\n",
    "z = np.random.normal(0, 1, \n",
    "          (image_grid_rows * image_grid_columns, z_dim))\n",
    "\n",
    "labels_to_generate = np.array([[i for j in range(5)] for i in range(10)])\n",
    "labels_to_generate = labels_to_generate.flatten().reshape(-1, 1)\n",
    "\n",
    "# Generate images from random noise\n",
    "gen_imgs = generator([z, labels_to_generate])\n",
    "\n",
    "# Rescale images to 0-1\n",
    "gen_imgs = 0.5 * gen_imgs + 0.5\n",
    "\n",
    "# Set image grid\n",
    "fig, axs = plt.subplots(image_grid_rows, image_grid_columns, \n",
    "                                figsize=(10, 20), sharey=True, sharex=True)\n",
    "\n",
    "cnt = 0\n",
    "for i in range(image_grid_rows):\n",
    "    for j in range(image_grid_columns):\n",
    "        # Output image grid\n",
    "        axs[i,j].imshow(gen_imgs[cnt, :,:,0], cmap='gray')\n",
    "        axs[i,j].axis('off')\n",
    "        axs[i,j].set_title(\"Digit: %d\" % labels_to_generate[cnt]) ## NEW\n",
    "        cnt += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fe6fc67af28>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot training losses for Discriminator and Generator\n",
    "d_losses = np.array(d_losses)\n",
    "g_losses = np.array(g_losses)\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.plot(d_losses, label=\"Discriminator Loss\")\n",
    "plt.plot(g_losses, label=\"Generator Loss\")\n",
    "plt.title(\"Training Losses\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fe6fc642fd0>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlYAAAE/CAYAAACEto0QAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3Xl8lOW99/HPLzshIUAS1gTCkqACshjRilVksa641J261FZrLa0Kfc7xaXs81vac87RHUBRbS1tFbevaRapY2RXFBVBcAAkBgQTQhLCEAFkmuZ4/ZoghZBlgknsy832/Xnk5c9/X3PfvmomTL9d1zT3mnENERERETlyM1wWIiIiIRAoFKxEREZEQUbASERERCREFKxEREZEQUbASERERCREFKxEREZEQUbASiXBm9riZ/UeIjznFzBYc52O/bmYbQlmPiEi4MF3HSqTjMrMtQE/AB9QC64CngTnOuToPS2sXZrYM+JNz7g9tdOwRQC/nXFWojy8ikUkjViId36XOuVSgP/D/gH8H/thWJzOzuLY6dnsyvybfA80sB/g64IDJ7VhWxDy/ItFKwUokQjjn9jnn5gHXAjeb2TAAM5trZr8M3M4ws1fMbK+Z7Taz5YfDhZllm9nfzKzUzMrMbHZg+y1m9raZPWRmZcD9gW1vHT63mTkzu9PMNprZfjP7hZkNMrMVZlZuZi+YWUKg7TgzK27w2C1m9mMz+9jM9pnZ82aWFNjXLVBvqZntCdzOCuz7L/zhZ7aZVTSo9ywzWxk41kozO6vBuZaZ2X+Z2dvAQWBgM0/nTcC7wFzg5oY7zKyTmc0ws62Bc7xlZp0C+84O9HmvmRWZ2S0NzvvdBsdo6vn7gZltBDYGts0KHKPczFab2dcbtI81s5+Y2abA87068Po9ZmYzGtU7z8zuaaafIhJiClYiEcY59z5QjD90NDY9sC8T/xTiTwBnZrHAK8BWIAfoCzzX4HFnAJsDj/mvZk79DeA04Ezg34A5wLeAbGAYcH0LZV8DXAAMAE4FbglsjwGexD8a1w84BMwO9POnwHJgqnMuxTk31cy6A68CjwDpwEzgVTNLb3CuG4HbgdRAf5tyE/DnwM83zKxng30PBvp5FtA90Nc6M+sPvAY8iv/5HQmsaaHPjV2O/3k+JXB/ZeAY3YG/AC8eDpzANPzP50VAF+BW/EHxKeD6BmE5A5gYeLyItAMFK5HItAP/H+TGaoDeQH/nXI1zbrnzL7QcA/QB/o9z7oBzrtI591bD4znnHnXO+Zxzh5o556+dc+XOubXAp8AC59xm59w+/IFjVAv1PuKc2+Gc2w38E3+gwDlX5pz7q3PuoHNuP/5Qd24Lx7kY2OiceyZQ67PAZ8ClDdrMdc6tDeyvaXwAMzsbf5B7wTm3GtgE3BDYF4M/xNzlnNvunKt1zq0IrMG6AVjknHs28NyWOeeOJVj9j3Nu9+Hn1zn3p8AxfM65GUAiMCTQ9rvAz5xzG5zfR4G27wP7gAmBdtcBy5xzXx5DHSJyAhSsRCJTX2B3E9v/FygEFpjZZjO7N7A9G9jqnPM1c7yiIM7Z8I/3oSbup7Tw2C8a3D54uK2ZJZvZ7wLTbuXAm0DXwAhbU/pw9CjUVvzPx2Gt9eVm/KFwV+D+X/hqOjADSMIfthrLbmZ7sI6oKzA9uj4w3bgXSAucv7VzPYV/pJDAf585gZpE5BhpkaRIhDGz0/EHibca7wuM+kwHpgfWYC0xs5X4/6j3M7O4ZsKVVx8fno5/lOYM59wXZjYS+BCwZuragX+0qaF+wL8a3G+2L4G1UtcAsWZ2OOwl4g9zI4BPgEpgEPBRo4cX4R/5a8oBILnB/V5NtKmvK7Ce6t/wjzytdc7Vmdkevup3UaCGT5s4zp+ATwP1ngz8o5maRKQNaMRKJEKYWRczuwT/2qg/Oec+aaLNJWY22MwM/5RRLVAHvA/sBP6fmXU2syQzG9ue9TcjFf9o197A+qn/bLT/S45cgD4fyDOzG8wszsyuxb9m6ZUgz3c5/ufkFPzTkSPxh5PlwE2BS1g8Acw0sz6BReRfM7NE/OuxJprZNYFzpweCIPjXWl0ZGIEbDHwniH77gFIgzszuw7+W6rA/AL8ws1zzO/XwOjLnXDH+9VnPAH9tYepWRNqAgpVIx/dPM9uPfxTjp/gXbH+7mba5wCKgAngH+I1zbqlzrhb/OqTBwDb8C9yvbevCg/Aw0AnYhf9Tev9qtH8WcFXgE4OPOOfKgEvwj3SV4R/1uaTBtF5rbgaedM5tc859cfgH/4L5Kea/FMKP8Y9crcQ/3forIMY5tw3/YvLpge1r8F8HC+AhoBp/EHwKfwhryeuBvhbgn8qs5MipwpnAC8ACoBz/5TU6Ndj/FDAcTQOKtDtdIFREJMKY2Tn4pwT7O73Ji7QrjViJiEQQM4sH7gL+oFAl0v4UrEREIoSZnQzsxX9JjYc9LkckKmkqUERERCRENGIlIiIiEiIKViIiIiIh4tkFQjMyMlxOTo5XpxcREREJ2urVq3c55zJba+dZsMrJyWHVqlVenV5EREQkaGbW3Je2H0FTgSIiIiIhomAlIiIiEiIKViIiIiIh4tkaKxERkeNVU1NDcXExlZWVXpciESYpKYmsrCzi4+OP6/EKViIi0uEUFxeTmppKTk4OZuZ1ORIhnHOUlZVRXFzMgAEDjusYmgoUEZEOp7KykvT0dIUqCSkzIz09/YRGQhWsRESkQ1KokrZwor9XrQYrM3vCzErM7NNm9puZPWJmhWb2sZmNPqGKREREOoDY2FhGjhzJ0KFDGTFiBDNmzKCurg6AVatW8aMf/eiEz/H444/z9NNPH9NjzjrrrOM+39y5c9mxY8dxP76hkSNHct1114XkWB1JMGus5gKzgeZe2QuB3MDPGcBvA/8VERGJWJ06dWLNmjUAlJSUcMMNN1BeXs7Pf/5z8vPzyc/PP6Hj+3w+7rjjjmN+3IoVK477nHPnzmXYsGH06dMn6MfU1tYSGxt7xLb169dTW1vL8uXLOXDgAJ07dz7umlri8/mIiwuv5eKtjlg5594EdrfQ5DLgaef3LtDVzHqHqkCRY7X7QDV//7CYZRtKWLejnF0VVdTVOa/LEpEI1qNHD+bMmcPs2bNxzrFs2TIuueQSAN544w1GjhzJyJEjGTVqFPv37wfgV7/6FcOHD2fEiBHce++9AIwbN467776b/Px8Zs2axf3338+DDz5Yv++ee+4hPz+fk08+mZUrV3LllVeSm5vLz372s/paUlJSAFi2bBnjxo3jqquu4qSTTmLKlCk4538vfOCBBzj99NMZNmwYt99+O845XnrpJVatWsWUKVMYOXIkhw4dYvHixYwaNYrhw4dz6623UlVVBfi/PeXf//3fGT16NC+++OJRz8ezzz7LjTfeyPnnn8/LL79cv72wsJCJEycyYsQIRo8ezaZNm1p8Lg5/Q8uuXbs4/DV4c+fOZfLkyYwfP54JEyZQUVHBhAkTGD16NMOHDz/ifE8//TSnnnoqI0aM4MYbb2T//v0MGDCAmpoaAMrLy4+4HwqhiHl9gaIG94sD23Y2bmhmtwO3A/Tr1y8EpxY52n/841Ne/eTIX7+4GCMjJZEeXRLpkZpIZmoSPVIP3//qdkZKIvGxWnooIsdu4MCB1NbWUlJScsT2Bx98kMcee4yxY8dSUVFBUlISr732Gi+//DLvvfceycnJ7N791fhFdXV1faC4//77jzhWQkICq1atYtasWVx22WWsXr2a7t27M2jQIO655x7S09OPaP/hhx+ydu1a+vTpw9ixY3n77bc5++yzmTp1Kvfddx8AN954I6+88gpXXXUVs2fP5sEHHyQ/P5/KykpuueUWFi9eTF5eHjfddBO//e1vufvuuwFIT0/ngw8+aPK5eP7551m4cCGfffYZjz76KDfccAMAU6ZM4d577+WKK66gsrKSurq6Fp+L5nzwwQd8/PHHdO/eHZ/Px9///ne6dOnCrl27OPPMM5k8eTLr1q3jl7/8JStWrCAjI4Pdu3eTmprKuHHjePXVV7n88st57rnnuPLKK4/70gpNadfxM+fcHGAOQH5+voYQJOTW7Sjn1U92cuvYAVw0vBcl+6soKa/0/zfwU7znEB9u20vZgeomj9G9c0IgfAVCVyCMNb7dKSG2yceLSPv6+T/Xsm5HeUiPeUqfLvznpUNDcqyxY8cybdo0pkyZwpVXXklWVhaLFi3i29/+NsnJyQB07969vv21117b7LEmT54MwPDhwxk6dCi9e/sniAYOHEhRUdFRwWrMmDFkZWUB/jVPW7Zs4eyzz2bp0qX8+te/5uDBg+zevZuhQ4dy6aWXHvHYDRs2MGDAAPLy8gC4+eabeeyxx+qDVXN1rlq1ioyMDPr160ffvn259dZb2b17N/Hx8Wzfvp0rrrgC8F8vCmjxuWjOpEmT6ts55/jJT37Cm2++SUxMDNu3b+fLL79kyZIlXH311WRkZBxx3O9+97v8+te/5vLLL+fJJ5/k97//favnOxahCFbbgewG97MC20Ta3cyFBaQmxXHXhFzSklv+F0hNbR27Kqoo3V9FSfnh4BUIYeVVlO6vpLCkgtL9VfiamEpMTYwjs2HoCox6ZTa8n5pEl05x+vSSSBTYvHkzsbGx9OjRg/Xr19dvv/fee7n44ouZP38+Y8eO5fXXX2/xOC2tR0pMTAQgJiam/vbh+z6fr9n24F9s7/P5qKys5M4772TVqlVkZ2dz//33H9flBZqr89lnn+Wzzz6rn7orLy/nr3/96zEvZI+Li6v/MEDj+hqe+89//jOlpaWsXr2a+Ph4cnJyWuzP2LFj2bJlC8uWLaO2tpZhw4YdU12t1h2CY8wDpprZc/gXre9zzh01DSjS1j4q2sui9V8yfVJeq6EKID42ht5pneid1qnFdnV1jj0Hq78a9QqMgJUGfkr2V/JR8V5Kyqs4VFN71OMT42ICYevoUa+GwSy9cwIxMQpgIscqVCNLJ6K0tJQ77riDqVOnHvUPqU2bNjF8+HCGDx/OypUr+eyzz5g0aRIPPPAAU6ZMqZ/+CmakJhQOh46MjAwqKip46aWXuOqqqwBITU2tXwM2ZMgQtmzZQmFhIYMHD+aZZ57h3HPPbfHYdXV1vPDCC3zyySf1C+CXLl3KL37xC2677TaysrL4xz/+weWXX05VVRW1tbXNPhc5OTmsXr2aMWPG8NJLLzV7zn379tGjRw/i4+NZunQpW7duBWD8+PFcccUVTJs2jfT09COe45tuuokbbriB//iP/zixJ7MJrQYrM3sWGAdkmFkx8J9APIBz7nFgPnARUAgcBL4d8ipFgjBjYQHdkuP59tnHd7Xc5sTEGOkpiaSnJHJyCx/LcM5RUeWrH/Eq2V8ZCF5fhbHC0gpWbNpFeeXR/7KMjTEyUhL8gas+iCWS2SWp/naPLklkpiSSEKd1YCJeO3ToECNHjqSmpoa4uDhuvPFGpk2bdlS7hx9+mKVLlxITE8PQoUO58MILSUxMZM2aNeTn55OQkMBFF13Ef//3f7dL3V27duW2225j2LBh9OrVi9NPP71+3y233MIdd9xBp06deOedd3jyySe5+uqr8fl8nH766a1+SnH58uX07dv3iE8VnnPOOaxbt46dO3fyzDPP8L3vfY/77ruP+Ph4XnzxRS644IImn4sf//jHXHPNNcyZM4eLL7642XNOmTKFSy+9lOHDh5Ofn89JJ50EwNChQ/npT3/KueeeS2xsLKNGjWLu3Ln1j/nZz37G9ddffwLPZNPs8CcE2lt+fr47vDhP5ESt2rKbqx5/h3svPIk7zh3kdTmtqqyprQ9dpQ2mH0uOuF1F2YEqmvpftFtyfP3oV2ajqciGwaxzYnh9DFkkVNavX8/JJ5/sdRnSQb300ku8/PLLPPPMM03ub+r3y8xWO+davYaG3nUlIsxYUEBGSiI3fa2/16UEJSk+luzuyWR3T26xna+2jrID1UeFrtKKyvrwtbn0ACX7K6mpPTqBdU6I9Y9yNTMVefh2Wqd4rQMTkajwwx/+kNdee4358+e3yfEVrKTDW1G4i3c2l3HfJaeQnBBZv9JxsTH07JJEzy5JQFqz7Zxz7D1Y89UC/EaL8UvLq/h0+z5K9pdwsProdWAJcTFkHnE5iqNHwXqk+qdDY7UOTEQ6sEcffbRNjx9Zf4Uk6jjnmLGwgF5dkrjhjOi9NpqZ0a1zAt06JzCkV2qLbQ/UrwNreBmKSkoDYezzXQd47/Pd7D149AXzYgzSU75a/9Vw1CvziNuJJMbpchQiEn0UrKRDe6OglNVb9/DLy4eRFK8/5MHonBjHgMQ4BmS0/BUTVb7aBovv/WvBShtcD6xkfyVrD1/Zvol1YGmd4o8a8coMLMBvuBg/RevA5Dg55zSFLSF3omvP9Y4mHZZzjpkLC8jq1olr8rNbf4Ack8S4WLK6JZPVreV1YLV1jrIDh8NX01OR73++m9KKKqp9dUc9PjkhtsnLT3wVxPz3uyVrHZh8JSkpibKyMtLT0/V7ISHjnKOsrKz+4qXHQ8FKOqyF677k4+J9/PqqU3X5AQ/FxlggCLX8RuSco/yQ76tF+EcEMP/U5Pod5byxv4qKqqMvRxEfa2SmNLr8RBOL8dM6xRMXY8TGmP7gRrCsrCyKi4spLS31uhSJMElJSfVXqz8eClbSIdXV+UerBmR05spRfb0uR4JgZqQlx5OWHE9uz5bXgR2s9tWHrtL9jS9DUUnR7oOs3rqH3c18LdFh8bFGXEwMcbFGfGwMcTGB/8Zao9sxrbQN7D/idgzxMUZs/WOO3Nf04/2PiQtsj2/02IbnjouNOeKYColHio+PZ8CA0F6zTiQUFKykQ5r/6U4++2I/s64bSZy+NDniJCfEkZMRR04r68Cqff6vJWq4GH9/pQ9fbR01dQ5fbR2+OkdNbR2+Woevro6aWnfk/lp3xO1KX61/W+CxvtrAY+rqGm131NTVNXmdsbYQG9NMGGwQ0IIPdi2FvKOD3VHB8ai2wYTSr7brk6USyRSspMOprXM8tLCA3B4pXHJqn9YfIBErIS6GPl070adry19L1JZq61oPYUcHO38o8zUZ8hoHuMbHPDIgNnn8wH8PVvsC+4MLmU19CKEtmNEgDLYcwo4Ijs0Ew8Nhr2+3TozM7srwvmm6OK54Rr950uG8vGY7m0oP8Jspo/UvX/FcbIwRGxMZn0itq2sY+FoOecEEu8YBrmHb5oNl0wG1ylfHgSrfEdvrawoc9/AlQmIM8nqmMjK7q/+nX1dye6Tq/ULahYKVdCg1tXXMWryRU3p34YKhvbwuRySixMQYiTGxdNTBnrKKKj4q3suabXv5sGgv8z/ZyXMriwD/txAMz0pjZHa3+sDVK+34P/kl0pwO+r+PRKu/ri5ma9lB/nBTPjH616eINJCeksj4k3oy/qSegH8EbkvZAdYU7a3/+eNbm+u//qlXl6T6ES1NIUqo6DdIOowqXy2PLilkRHZXJpzcw+tyRCTMxcQYAzNTGJiZwpWj/R+fr6ypZe2O8vqg9VHRXv619gt/+8AU4qh+XRmRpSlEOT4KVtJhvLCyiO17D/E/Vw7XR89F5LgkxcdyWv9unNa/W/22xlOIr368k2ffb3oKcVS/roHv7hRpmoKVdAiVNf7RqtNzuvH13AyvyxGRCNLUFOLnZQf4qMEU4h+Wb8ZXpylEaZ1+E6RD+NO7WynZX8Uj14/SaJWItKmYGGNQZgqDjmMK0b8wvhuDe6RoCjFKKVhJ2DtQ5eO3yzYxdnA6Zw5M97ocEYlCmkKUYClYSdh76p0tlB2oZtqkIV6XIiJSr7kpxDXb9voDV6MpxN5p/inEEYHLPZyalUZygv4MRxq9ohLWyitr+N0bmzlvSOYR/1IUEQk3DacQv3la81OIr32qKcRIpmAlYe2Jtz5n36EajVaJSIekKcToo2AlYWvvwWr+uPxzvjG0J8Oz0rwuR0QkJI53CvHwz3BNIYY1vTIStua8uZmKah/3TMrzuhQRkTYTzBTimqI9mkLsIBSsJCyVVVQxd8UWLh7em5N6dfG6HBGRdnU8U4inBq4WPyJLU4heUrCSsPT4G5uorKnl7okarRIRgZanENcU+acRf/+mphC9pmdYws6X5ZU8/c5WLh/Vl8E9UrwuR0QkLB3rFGJsjJHXM5WR2WmaQmxDClYSdn6ztJDaOsddE3K9LkVEpEM51inElMQ4hvdNq/96npHZmkI8UQpWEla27z3Es+8XcXV+Fv3TO3tdjohIh6cpxPalZ0rCyuwlGwGYOl6jVSIibeH4pxC7Mipw5XhNITZPwUrCxtayA7y4qpgpZ/Sjb9dOXpcjIhI1gptC3MGz728Djp5CHJXdlR6aQgQUrCSMzFq8kdgY4wfnDfa6FBGRqHesU4h90pLqvwcxmqcQo6/HEpYKSyr4x4fb+c7ZA/SvHhGRMHSiU4gj+3VlUGbkTyEqWElYmLV4I0nxsdxx7iCvSxERkSA1NYW4q6KKjwJfOB2NU4gKVuK5z74o558f7eDOcYNIT0n0uhwRETkBGSmJTDi5JxNODn4K8fAV4yNhCrHjVi4R46GFBaQmxnH7OQO9LkVEREIs2CnE+Z9ExhSigpV46pPifby+9kvunphL1+QEr8sREZF2cDxTiKdmpdUvjg/nKUQFK/HUzIUb6Jocz61nD/C6FBER8VBrU4hripqeQhyZ3ZXzhvQgt2eql+XXU7ASz6zeuoelG0r5twuG0CUp3utyREQkjBzLFKJhClYiMxduIL1zAjd/LcfrUkREpANobgoxLozWX8V4XYBEp3c2lfF2YRnfHzeIzonK9yIicnwyUhLDao2ugpW0O+ccMxduoGeXRL51Zn+vyxEREQmZoIKVmV1gZhvMrNDM7m1ifz8zW2pmH5rZx2Z2UehLlUixfOMuVm7Zw9TzBpMUH+t1OSIiIiHTarAys1jgMeBC4BTgejM7pVGznwEvOOdGAdcBvwl1oRIZnHPMWFhA366duOb0bK/LERERCalgRqzGAIXOuc3OuWrgOeCyRm0c0CVwOw3YEboSJZIsXl/CR0V7+eH4wSTGabRKREQiSzCrhvsCRQ3uFwNnNGpzP7DAzH4IdAYmhqQ6iSh1dY6ZCwvon55c/9FZERGRSBKqxevXA3Odc1nARcAzZnbUsc3sdjNbZWarSktLQ3Rq6SheX/sF63aWc9eEXOJj9bkJERGJPMH8ddsONFwMkxXY1tB3gBcAnHPvAElARuMDOefmOOfynXP5mZmZx1exdEi1gdGqQZmduWxkX6/LERERaRPBBKuVQK6ZDTCzBPyL0+c1arMNmABgZifjD1YakpJ6r3y8g40lFdwzKa/DfJGmiIjIsWo1WDnnfMBU4HVgPf5P/601swfMbHKg2XTgNjP7CHgWuMU559qqaOlYfLV1PLxoIyf1SuWiYb29LkdERKTNBHXJa+fcfGB+o233Nbi9Dhgb2tIkUvztw+18vusAc248jRiNVomISATTCmJpU9W+Oh5ZvJFTs9KYdEpPr8sRERFpUwpW0qZeWFVE8Z5D3DMpDzONVomISGRTsJI2U1lTy+wlhZzWvxvj8vQpUBERiXwKVtJm/vLeNr4or2S6RqtERCRKKFhJmzhUXctvlm3iawPTOWvwUZc0ExERiUgKVtImnn5nC7sqqph+fp7XpYiIiLQbBSsJuYoqH4+/sYlz8jLJz+nudTkiIiLtRsFKQu7Jtz5nz8Eapk/SaJWIiEQXBSsJqX0Ha5izfDMTT+7JiOyuXpcjIiLSrhSsJKT+8NZm9lf6mKbRKhERiUIKVhIyuw9U88Rbn3Px8N6c0qeL1+WIiIi0OwUrCZnfvbGJgzW13D0x1+tSREREPKFgJSFRsr+Sp97ZwuUj+5LbM9XrckRERDyhYCUh8Zulm6ipddw1QaNVIiISvRSs5ITt3HeIv7y3jatGZ5GT0dnrckRERDyjYCUnbPaSQhyOH04Y7HUpIiIinlKwkhNStPsgz68s4trTs8nqlux1OSIiIp5SsJIT8sjijcTEGFPP09oqERERBSs5bptLK/jbh9v51hn96ZWW5HU5IiIinlOwkuM2a/FGEmJj+P64QV6XIiIiEhYUrOS4FHy5n3kf7eDms3LITE30uhwREZGwoGAlx+WhhQV0Tojje+cM9LoUERGRsKFgJcds7Y59vPbpF9x69gC6dU7wuhwREZGwoWAlx+yhhQV0SYrjO2cP8LoUERGRsKJgJcfkw217WLS+hNvPGUhap3ivyxEREQkrClZyTGYuLKB75wRuGavRKhERkcYUrCRo73++m+Ubd3HHuQNJSYzzuhwREZGwo2AlQXHOMWPBBjJTE7nxzByvyxEREQlLClYSlBWbynjv8938YNwgOiXEel2OiIhIWFKwklY553hwwQZ6pyVx3Zh+XpcjIiISthSspFXLNpTy4ba9/HB8LknxGq0SERFpjoKVtMg5x4yFG8ju3omr87O8LkdERCSsKVhJi15f+yWfbi/nrgl5xMfq10VERKQl+kspzaqrczy0sICBGZ25fGQfr8sREREJewpW0qxXPtnJhi/3c9fEXOI0WiUiItIq/bWUJvlq63h4UQFDeqZy6akarRIREQmGgpU06eU1O9hceoB7JuUSE2NelyMiItIhKFjJUWpq65i1eCND+3ThG0N7eV2OiIhIh6FgJUd5aXUx23YfZPr5eZhptEpERCRYClZyhCpfLY8u3sjI7K6cN6SH1+WIiIh0KEEFKzO7wMw2mFmhmd3bTJtrzGydma01s7+EtkxpL8+9X8SOfZX8+PwhGq0SERE5RnGtNTCzWOAxYBJQDKw0s3nOuXUN2uQC/xcY65zbY2Ya6uiADlXXMntpIWMGdGfs4HSvyxEREelwghmxGgMUOuc2O+eqgeeAyxq1uQ14zDm3B8A5VxLaMqU9/OndrZTur2L6JK2tEhEROR7BBKu+QFGD+8WBbQ3lAXlm9raZvWtmF4SqQGkfB6p8/PaNTXw9N4MzBmq0SkRE5Hi0OhV4DMfJBcYBWcCbZjbcObe3YSMzux24HaBfv34hOrWEwtwVW9h9oJppk/K8LkVERKTDCmbEajuQ3eB+VmBbQ8U+CDvZAAAV0klEQVTAPOdcjXPuc6AAf9A6gnNujnMu3zmXn5mZebw1S4iVV9Yw583NTDipB6P6dfO6HBERkQ4rmGC1Esg1swFmlgBcB8xr1OYf+EerMLMM/FODm0NYp7ShPy7/nH2HarhHo1UiIiInpNVg5ZzzAVOB14H1wAvOubVm9oCZTQ40ex0oM7N1wFLg/zjnytqqaAmdPQeq+eNbn3PhsF4M65vmdTkiIiIdWlBrrJxz84H5jbbd1+C2A6YFfqQDmbN8MweqfRqtEhERCQFdeT2Kle6vYu7bW7j01D7k9Uz1uhwREZEOT8Eqij3+xiaqfLXcPfGozxmIiIjIcVCwilJf7KvkT+9u5crRWQzMTPG6HBERkYigYBWlHltaSG2d464JGq0SEREJFQWrKFS85yDPrdzGNadnk9092etyREREIoaCVRR6dHEhhjH1vMFelyIiIhJRFKyizJZdB3jpg2JuOKMffbp28rocERGRiKJgFWUeWbyR+FjjzvMGeV2KiIhIxFGwiiKFJfv5+5rt3PS1HHqkJnldjoiISMRRsIoiDy3aSHJ8LN87Z6DXpYiIiEQkBasosX5nOa9+vJNvjx1Aekqi1+WIiIhEJAWrKDFzYQGpSXHc9nWNVomIiLQVBaso8HHxXhau+5Lbvj6QtOR4r8sRERGJWApWUWDGggK6Jsfz7bE5XpciIiIS0RSsItzqrbt5o6CUO84dRGqSRqtERETakoJVhJuxoICMlARu+lp/r0sRERGJeApWEWzFpl2s2FTGneMGk5wQ53U5IiIiEU/BKkI555i5oIBeXZK44Yx+XpcjIiISFRSsItQbBaWs2rqHH4wfTFJ8rNfliIiIRAUFqwjknGPmwgL6du3EtfnZXpcjIiISNRSsItCi9SV8XLyPuybkkhCnl1hERKS96K9uhKmrc8xYsIGc9GSuHN3X63JERESiioJVhHnt0y/47Iv93D0xj7hYvbwiIiLtSX95I0htneOhRQXk9kjh0hF9vC5HREQk6ihYRZB5H22nsKSCeyblERtjXpcjIiISdRSsIkRNbR2zFm3k5N5duGBoL6/LERERiUoKVhHibx8Us6XsINMn5RGj0SoRERFPKFhFgGpfHY8sLmREdlcmnNzD63JERESiloJVBHh+VRHb9x5i2qQ8zDRaJSIi4hUFqw6usqaW2Us2cnpON87JzfC6HBERkaimYNXB/fm9bXxZXsW0SUM0WiUiIuIxBasO7GC1j98uK+SsQel8bVC61+WIiIhEPQWrDuypFVvZVVHN9PPzvC5FREREULDqsPZX1vC7Nzcxbkgmp/Xv7nU5IiIigoJVh/XEW1vYe7CG6ZOGeF2KiIiIBChYdUB7D1bzh+WbOf+UngzPSvO6HBEREQlQsOqAfr98M/urfNwzSWurREREwomCVQdTVlHFk29v4ZJTe3Ny7y5elyMiIiINKFh1ML97czOVNbXcPVGjVSIiIuFGwaoDKSmv5KkVW7h8VF8G90jxuhwRERFpRMGqA/nNsk346hx3Tcj1uhQRERFpQlDByswuMLMNZlZoZve20O6bZubMLD90JQrA9r2H+Mt727j6tCz6p3f2uhwRERFpQqvBysxigceAC4FTgOvN7JQm2qUCdwHvhbpIgdlLCgH4oUarREREwlYwI1ZjgELn3GbnXDXwHHBZE+1+AfwKqAxhfQJsKzvIi6uKuG5MNn27dvK6HBEREWlGMMGqL1DU4H5xYFs9MxsNZDvnXg1hbRIwa/FGYmOMH5w32OtSREREpAUnvHjdzGKAmcD0INrebmarzGxVaWnpiZ46KmwqreDvHxZz45n96dklyetyREREpAXBBKvtQHaD+1mBbYelAsOAZWa2BTgTmNfUAnbn3BznXL5zLj8zM/P4q44iDy/aSFJ8LHeMG+R1KSIiItKKYILVSiDXzAaYWQJwHTDv8E7n3D7nXIZzLsc5lwO8C0x2zq1qk4qjyIYv9vPKxzu45awcMlISvS5HREREWtFqsHLO+YCpwOvAeuAF59xaM3vAzCa3dYHR7KGFBaQkxHH7OQO9LkVERESCEBdMI+fcfGB+o233NdN23ImXJZ9u38e/1n7B3RNz6Zqc4HU5IiIiEgRdeT1MzVxYQFqneG49e4DXpYiIiEiQFKzC0Afb9rDksxJuP2cgXZLivS5HREREgqRgFYZmLiggvXMCt5yV43UpIiIicgwUrMLMu5vLeKtwF98fN4jOiUEtgRMREZEwoWAVRpxzzFxQQI/URL51Zn+vyxEREZFjpGAVRt4q3MX7W3YzdfxgkuJjvS5HREREjpGCVZhwzjFjQQF90pK49vTs1h8gIiIiYUfBKkws+ayENUV7+dGEXBLjNFolIiLSESlYhYG6OsfMhQX0657MN0/L8rocEREROU4KVmHg9bVfsHZHOXdNyCU+Vi+JiIhIR6W/4h6rrXM8tKiAQZmduXxUX6/LERERkROgYOWxVz7eQcGXFdw9MY/YGPO6HBERETkBClYe8tXW8fCijZzUK5WLh/f2uhwRERE5QQpWHvr7h9v5fNcB7pmUR4xGq0RERDo8BSuPVPvqmLV4I8P7pnH+KT29LkdERERCQMHKIy+uLqJ4zyGmnZ+HmUarREREIoGClQcqa2qZvaSQ0f26Mi4v0+tyREREJEQUrDzw7Pvb2Lmvkh+fP0SjVSIiIhFEwaqdHaqu5bGlmzhzYHfOGpzhdTkiIiISQgpW7ezpd7awq6KK6ecP8boUERERCTEFq3ZUUeXj8Tc2cU5eJqfndPe6HBEREQkxBat2NPftz9lzsIZpk/K8LkVERETagIJVO9l3qIY5b25m4sk9GZnd1etyREREpA0oWLWTPy7fTHmlT6NVIiIiEUzBqh3sPlDNE29v4aLhvTilTxevyxEREZE2omDVDn735iYOVPu4Z6JGq0RERCKZglUbK9lfyVMrtnDZiD7k9kz1uhwRERFpQwpWbey3yzZRU+u4S6NVIiIiEU/Bqg3t3HeIP7+3jW+O7suAjM5elyMiIiJtTMGqDT22tBDnHD8cn+t1KSIiItIOFKzaSNHugzy/sohrT88mu3uy1+WIiIhIO1CwaiOPLtmImTH1PI1WiYiIRAsFqzbw+a4D/PWD7XzrjP70SkvyuhwRERFpJwpWbWDWogISYmP4/rhBXpciIiIi7UjBKsQ2frmflz/awU1n9SczNdHrckRERKQdKViF2EOLCuicEMcd52i0SkREJNooWIXQ2h37mP/JF9w6NodunRO8LkdERETamYJVCD20sIAuSXF85+sDvS5FREREPKBgFSJrivayaH0Jt58zkLRO8V6XIyIiIh5QsAqRmQsL6JYczy1jB3hdioiIiHgkqGBlZheY2QYzKzSze5vYP83M1pnZx2a22Mz6h77U8LVyy27eLCjl++MGkZIY53U5IiIi4pFWg5WZxQKPARcCpwDXm9kpjZp9COQ7504FXgJ+HepCw9mMBRvITE3kxjNzvC5FREREPBTMiNUYoNA5t9k5Vw08B1zWsIFzbqlz7mDg7rtAVmjLDF8rCnfx7ubd3DluEJ0SYr0uR0RERDwUTLDqCxQ1uF8c2Nac7wCvnUhRHYVzjgcXbKB3WhLXj+nndTkiIiLisZAuXjezbwH5wP82s/92M1tlZqtKS0tDeWpPLCso5YNte5k6fjBJ8RqtEhERiXbBBKvtQHaD+1mBbUcws4nAT4HJzrmqpg7knJvjnMt3zuVnZmYeT71hwznHzAUFZHfvxNWnZbf+ABEREYl4wQSrlUCumQ0wswTgOmBewwZmNgr4Hf5QVRL6MsPPgnVf8sn2ffxofC4JcbpqhYiIiAQRrJxzPmAq8DqwHnjBObfWzB4ws8mBZv8LpAAvmtkaM5vXzOEiQl2d46GFBQzM6MwVo1pabiYiIiLRJKiLLjnn5gPzG227r8HtiSGuK6y9+slOPvtiP7OuG0lcrEarRERExE+p4Bj5aut4aFEBeT1TuPTUPl6XIyIiImFEweoYvbxmB5tLDzBtUh4xMeZ1OSIiIhJGFKyOQU1tHbMWb2Rony58Y2gvr8sRERGRMKNgdQz+urqYbbsPMm1SHmYarRIREZEjKVgFqcpXyyOLNzIyuyvjT+rhdTkiIiIShhSsgvT8yiJ27Ktk+vkarRIREZGmKVgFobKmltlLChkzoDtnD87wuhwREREJUwpWQfjTu1sp2V/FdK2tEhERkRYoWLXiQJWP3y7bxNmDMzhjYLrX5YiIiEgYU7BqxdwVWyg7UM208/O8LkVERETCnIJVC8ora5jz5mbGn9SD0f26eV2OiIiIhDkFqxb8cfnn7DtUw7RJGq0SERGR1ilYNWPvwWqeeOtzLhjai2F907wuR0RERDoABatmzHlzMxXVPu7RaJWIiIgEScGqCbsqqnjy7S1cemofhvRK9bocERER6SAUrJrw+LJNVPlquWtirteliIiISAeiYNXIl+WVPPPuVq4YlcWgzBSvyxEREZEORMGqkceWFlJb57hrgkarRERE5NgoWDVQvOcgz76/javzs+mXnux1OSIiItLBKFg1MHtJIYbxw/GDvS5FREREOiAFq4CtZQd4cXUxN5zRjz5dO3ldjoiIiHRAClYBsxZvJC7GuHPcIK9LERERkQ5KwQooLKngHx9u5+azcujRJcnrckRERKSDUrACHl5UQFJ8LN87Z6DXpYiIiEgHFvXBav3Ocl75eCe3jh1Aekqi1+WIiIhIBxb1weqhhQWkJsVx29c1WiUiIiInJqqD1SfF+1iw7ku+e/ZA0pLjvS5HREREOrioDlYzFm6ga3I8t56d43UpIiIiEgGiNlit3rqbZRtK+d45g0hN0miViIiInLioDVYzFhSQkZLAzWf197oUERERiRBRGaze2VTGik1lfH/cYJIT4rwuR0RERCJE1AUr5xwzF26gZ5dEppzRz+tyREREJIJEXbB6c+MuVm7Zw9TxuSTFx3pdjoiIiESQqApWzjlmLthA366duDY/2+tyREREJMJEVbBavL6Ej4r3cdeEXBLioqrrIiIi0g6iJl3U1TlmLCwgJz2ZK0f39bocERERiUBRE6z+tfYL1u8s566JucTFRk23RUREpB1FRcKorXPMXFjA4B4pTB6h0SoRERFpG1ERrP750Q4KSyq4Z2IesTHmdTkiIiISoSI+WPlq63h4UQEn9UrlwmG9vC5HREREIljEB6u/fbCdLWUHmX7+EGI0WiUiIiJtKKhgZWYXmNkGMys0s3ub2J9oZs8H9r9nZjmhLvR4VPvqmLV4IyOy0ph4cg+vyxEREZEI12qwMrNY4DHgQuAU4HozO6VRs+8Ae5xzg4GHgF+FutDj8fyqIrbvPcS084dgptEqERERaVvBjFiNAQqdc5udc9XAc8BljdpcBjwVuP0SMME8TjKVNbU8tqSQ/P7dOCc3w8tSREREJEoEE6z6AkUN7hcHtjXZxjnnA/YB6Y0PZGa3m9kqM1tVWlp6fBUH6ZWPd/JFeSXTzs/TaJWIiIi0i7j2PJlzbg4wByA/P9+15bm+ObovfbomcdYgjVaJiIhI+whmxGo70PAbi7MC25psY2ZxQBpQFooCj5eZKVSJiIhIuwomWK0Ecs1sgJklANcB8xq1mQfcHLh9FbDEOdemI1IiIiIi4abVqUDnnM/MpgKvA7HAE865tWb2ALDKOTcP+CPwjJkVArvxhy8RERGRqBLUGivn3HxgfqNt9zW4XQlcHdrSRERERDqWiL/yuoiIiEh7UbASERERCREFKxEREZEQUbASERERCREFKxEREZEQUbASERERCREFKxEREZEQMa8ukG5mpcDWNj5NBrCrjc8RzqK5/9Hcd4ju/qvv0Sua+x/NfYf26X9/51xma408C1btwcxWOefyva7DK9Hc/2juO0R3/9X36Ow7RHf/o7nvEF7911SgiIiISIgoWImIiIiESKQHqzleF+CxaO5/NPcdorv/6nv0iub+R3PfIYz6H9FrrERERETaU6SPWImIiIi0m4gIVmZ2gZltMLNCM7u3if2JZvZ8YP97ZpbT/lW2nSD6f4uZlZrZmsDPd72oM9TM7AkzKzGzT5vZb2b2SOB5+djMRrd3jW0piP6PM7N9DV73+9q7xrZiZtlmttTM1pnZWjO7q4k2Efn6B9n3SH7tk8zsfTP7KND/nzfRJiLf84Pse0S+3x9mZrFm9qGZvdLEvvB43Z1zHfoHiAU2AQOBBOAj4JRGbe4EHg/cvg543uu627n/twCzva61Dfp+DjAa+LSZ/RcBrwEGnAm853XN7dz/ccArXtfZRn3vDYwO3E4FCpr4vY/I1z/Ivkfya29ASuB2PPAecGajNhH5nh9k3yPy/b5B/6YBf2nq9ztcXvdIGLEaAxQ65zY756qB54DLGrW5DHgqcPslYIKZWTvW2JaC6X9Ecs69CexuocllwNPO712gq5n1bp/q2l4Q/Y9YzrmdzrkPArf3A+uBvo2aReTrH2TfI1bg9awI3I0P/DReLByR7/lB9j1imVkWcDHwh2aahMXrHgnBqi9Q1OB+MUe/ydS3cc75gH1AertU1/aC6T/ANwPTIS+ZWXb7lOa5YJ+bSPa1wLTBa2Y21Oti2kJguH8U/n+9NxTxr38LfYcIfu0D00FrgBJgoXOu2dc+0t7zg+g7RO77/cPAvwF1zewPi9c9EoKVtO6fQI5z7lRgIV8leolsH+D/CoYRwKPAPzyuJ+TMLAX4K3C3c67c63raUyt9j+jX3jlX65wbCWQBY8xsmNc1tZcg+h6R7/dmdglQ4pxb7XUtrYmEYLUdaJjIswLbmmxjZnFAGlDWLtW1vVb775wrc85VBe7+ATitnWrzWjC/GxHLOVd+eNrAOTcfiDezDI/LChkzi8cfLP7snPtbE00i9vVvre+R/tof5pzbCywFLmi0K5Lf84Hm+x7B7/djgclmtgX/kpfxZvanRm3C4nWPhGC1Esg1swFmloB/wdq8Rm3mATcHbl8FLHGB1W0RoNX+N1pXMhn/moxoMA+4KfDpsDOBfc65nV4X1V7MrNfh9QVmNgb//+8R8ccl0K8/AuudczObaRaRr38wfY/w1z7TzLoGbncCJgGfNWoWke/5wfQ9Ut/vnXP/1zmX5ZzLwf93bolz7luNmoXF6x7X3icMNeecz8ymAq/j/4TcE865tWb2ALDKOTcP/5vQM2ZWiH+x73XeVRxaQfb/R2Y2GfDh7/8tnhUcQmb2LP5PP2WYWTHwn/gXc+KcexyYj/+TYYXAQeDb3lTaNoLo/1XA983MBxwCrouEPy4BY4EbgU8C600AfgL0g4h//YPpeyS/9r2Bp8wsFn9gfME590qUvOcH0/eIfL9vTji+7rryuoiIiEiIRMJUoIiIiEhYULASERERCREFKxEREZEQUbASERERCREFKxEREZEQUbASERERCREFKxEREZEQUbASERERCZH/DwkbP/UpGVOjAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "accuracies = np.array(d_accuracies)\n",
    "\n",
    "# Plot Discriminator accuracy\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.plot(accuracies, label=\"Discriminator Accuracy\")\n",
    "plt.title(\"Discriminator Accuracy\")\n",
    "plt.legend()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}