src/mnist/04_CCGAN_MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CCGANs - Context-Conditional Generative Adversarial Networks\n",
    "\n",
    "Brief introduction to Context-Conditional Generative Adversarial Networks or CCGANs. This notebook is organized as follows:\n",
    "\n",
    "1. **Research Paper**\n",
    "2. **Background**\n",
    "3. **Definition**\n",
    "4. **Training CCGANs with MNIST dataset, Keras and TensorFlow**\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Research Paper\n",
    "\n",
    "* [Semi-Supervised Learning with Context-Conditional Generative Adversarial Networks](https://arxiv.org/pdf/1611.06430.pdf)\n",
    "\n",
    "## 2. Background\n",
    "\n",
    "**Generative adversarial nets** consists of two models: a generative model $G$ that captures the data distribution, and a discriminative model $D$ that estimates the probability that a sample came from the training data rather than $G$.\n",
    "\n",
    "The generator distribution $p_g$ over data data $x$, the generator builds a mapping function from a prior noise distribution $p_z(z)$ to data space as $G(z;\\theta_g)$.\n",
    "\n",
    "The discriminator, $D(x;\\theta_d)$, outputs a single scalar representing the probability that $x$ came form training data rather than $p_g$.\n",
    "\n",
    "The value function $V(G,D)$:\n",
    "\n",
    "$$ \\underset{G}{min} \\: \\underset{D}{max} \\; V_{GAN}(D,G) = \\mathbb{E}_{x\\sim p_{data}(x)}[log D(x)] + \\mathbb{E}_{z\\sim p_{z}(z)}[log(1 - D(G(z)))]$$\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Definition\n",
    "\n",
    "Context-Conditional Generative Adversarial Networks (CC-GANs) are conditional GANs where the generator is trained to fill in a missing image patch and the generator and discriminator are conditioned on the surrounding pixels.\n",
    "\n",
    "CC-GANs address a different task: determining if a part of an image is real or fake given the surrounding context.\n",
    "\n",
    "The generator $G$ receives as input an image with a randomly masked out patch. The generator outputs an entire image.  We fill in the missing patch from the generated output and then pass the completed image into $D$.\n",
    "\n",
    "### Network Design\n",
    "\n",
    "<img src=\"../../img/network_design_cc_gan.png\" width=\"600\"> \n",
    "\n",
    "\n",
    "### Cost Funcion\n",
    "\n",
    "$$\n",
    "\\begin{aligned}\n",
    "    \\underset{G}{min} \\: \\underset{D}{max} \\; V_{CCGAN}(D,G) =& \\mathbb{E}_{x\\sim \\mathcal{X}}[log D(x)] + \\mathbb{E}_{x\\sim \\mathcal{X}, m\\sim \\mathcal{M}}[log(1 - D(x_I))] \\\\\n",
    "    x_I =& (1 - m) \\bigodot x_G + m \\bigodot x \\\\\n",
    "    x_G =& G(m \\bigodot x, z)\n",
    "\\end{aligned}\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Training CCGANs with MINIST dataset, Keras and TensorFlow\n",
    "\n",
    "CCGANs implementation using \"U-net\" model and convolutional neural network and the [Keras](https://keras.io/) library.\n",
    "\n",
    "* **Data**\n",
    "    * Rescale the MNIST images to be between -1 and 1.\n",
    "    * Resize 32x32\n",
    "    \n",
    "* **Generator**\n",
    "    * **U-net network**.\n",
    "    * The input to the generator is the **normal distribution** $z$.\n",
    "    * The last activation is **tanh**.\n",
    "    \n",
    "* **Discriminator**\n",
    "    * **Convolutional neural network** and **LeakyReLU activation**.\n",
    "    * The last activation is **softmax**.\n",
    "    \n",
    "* **Loss**\n",
    "    * Discriminator: loss=['mse', 'categorical_crossentropy'].\n",
    "    * Adversarial: loss=['mse']\n",
    "\n",
    "* **Optimizer**\n",
    "    * Adam(lr=0.0002, beta_1=0.5)\n",
    "\n",
    "* batch_size = 64\n",
    "* epochs = 100\n",
    "\n",
    "### 1. Load data\n",
    "\n",
    "#### Load libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:03.403650Z",
     "start_time": "2018-08-06T17:31:03.048479Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-11-07 15:04:10.683353: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2 AVX AVX2 AVX_VNNI FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2023-11-07 15:04:10.753507: I tensorflow/core/util/util.cc:169] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import skimage as ski\n",
    "import tensorflow as tf\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Getting the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:05.716705Z",
     "start_time": "2018-08-06T17:31:05.133812Z"
    }
   },
   "outputs": [],
   "source": [
    "# load dataset\n",
    "(X_train, y_train), (X_test, y_test) = tf.keras.datasets.mnist.load_data()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Explore visual data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:06.159238Z",
     "start_time": "2018-08-06T17:31:05.718782Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x300 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "WIDTH_SIZE = 4\n",
    "HEIGHT_SIZE = 3\n",
    "\n",
    "fig = plt.figure(figsize=(WIDTH_SIZE, HEIGHT_SIZE))\n",
    "for i in range(10):\n",
    "    plt.subplot(2, 5, i+1)\n",
    "    x_y = X_train[y_train == i]\n",
    "    plt.imshow(x_y[0], cmap='gray', interpolation='none')\n",
    "    plt.title(\"Class %d\" % (i))\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    \n",
    "plt.tight_layout()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Reshaping and normalizing the inputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:06.583614Z",
     "start_time": "2018-08-06T17:31:06.161237Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train.shape (60000, 28, 28)\n",
      "y_train.shape (60000,)\n",
      "X_train reshape: (60000, 28, 28, 1)\n",
      "X_train reshape: (60000, 28, 28, 1)\n",
      "y_train reshape: (60000, 11)\n"
     ]
    }
   ],
   "source": [
    "print('X_train.shape', X_train.shape)\n",
    "print('y_train.shape', y_train.shape)\n",
    "\n",
    "if tf.keras.backend.image_data_format() == 'channels_first':\n",
    "    X_train = X_train.reshape(X_train.shape[0], 1, 28, 28)\n",
    "    X_test = X_test.reshape(X_test.shape[0], 1, 28, 28)\n",
    "    input_shape = (1, 28, 28)\n",
    "else:\n",
    "    X_train = X_train.reshape(X_train.shape[0], 28, 28, 1)\n",
    "    X_test = X_test.reshape(X_test.shape[0], 28, 28, 1)\n",
    "    input_shape = (28, 28, 1)\n",
    "\n",
    "# the generator is using tanh activation, for which we need to preprocess \n",
    "# the image data into the range between -1 and 1.\n",
    "\n",
    "X_train = np.float32(X_train)\n",
    "X_train = (X_train / 255 - 0.5) * 2\n",
    "X_train = np.clip(X_train, -1, 1)\n",
    "\n",
    "print('X_train reshape:', X_train.shape)\n",
    "\n",
    "# y to categorical\n",
    "num_classes = 10\n",
    "y_train = tf.keras.utils.to_categorical(y_train, num_classes=num_classes+1)\n",
    "\n",
    "print('X_train reshape:', X_train.shape)\n",
    "print('y_train reshape:', y_train.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:28.833217Z",
     "start_time": "2018-08-06T17:31:06.586070Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train reshape: (60000, 32, 32, 1)\n"
     ]
    }
   ],
   "source": [
    "X_train = ski.transform.resize(X_train, [X_train.shape[0], 32, 32, 1])\n",
    "print('X_train reshape:', X_train.shape)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Define model\n",
    "\n",
    "#### Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:29.941228Z",
     "start_time": "2018-08-06T17:31:28.835729Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-11-07 15:04:17.335806: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.353561: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.353681: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.354820: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2 AVX AVX2 AVX_VNNI FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2023-11-07 15:04:17.355943: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.356035: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.356094: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.402178: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.402281: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.402338: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:980] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2023-11-07 15:04:17.402400: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1616] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 1091 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 3080 Ti Laptop GPU, pci bus id: 0000:01:00.0, compute capability: 8.6\n"
     ]
    }
   ],
   "source": [
    "# Number of filters in first layer of generator\n",
    "gf = 32\n",
    "k = 4\n",
    "s = 2\n",
    "\n",
    "# imagem shape 28x28x1\n",
    "img_shape = X_train[0].shape\n",
    "\n",
    "# Generator input\n",
    "img_g = tf.keras.layers.Input(shape=(img_shape))\n",
    "\n",
    "# Downsampling\n",
    "d1 = tf.keras.layers.Conv2D(gf, kernel_size=k, strides=s, padding=\"same\")(\n",
    "    img_g\n",
    ")\n",
    "d1 = tf.keras.layers.LeakyReLU(alpha=0.2)(d1)\n",
    "\n",
    "d2 = tf.keras.layers.Conv2D(gf * 2, kernel_size=k, strides=s, padding=\"same\")(\n",
    "    d1\n",
    ")\n",
    "d2 = tf.keras.layers.LeakyReLU(alpha=0.2)(d2)\n",
    "d2 = tf.keras.layers.BatchNormalization(momentum=0.8)(d2)\n",
    "\n",
    "d3 = tf.keras.layers.Conv2D(gf * 4, kernel_size=k, strides=s, padding=\"same\")(\n",
    "    d2\n",
    ")\n",
    "d3 = tf.keras.layers.LeakyReLU(alpha=0.2)(d3)\n",
    "d3 = tf.keras.layers.BatchNormalization(momentum=0.8)(d3)\n",
    "\n",
    "d4 = tf.keras.layers.Conv2D(gf * 8, kernel_size=k, strides=s, padding=\"same\")(\n",
    "    d3\n",
    ")\n",
    "d4 = tf.keras.layers.LeakyReLU(alpha=0.2)(d4)\n",
    "d4 = tf.keras.layers.BatchNormalization(momentum=0.8)(d4)\n",
    "\n",
    "# Upsampling\n",
    "u1 = tf.keras.layers.UpSampling2D(size=2)(d4)\n",
    "u1 = tf.keras.layers.Conv2D(\n",
    "    gf * 4, kernel_size=k, strides=1, padding=\"same\", activation=\"relu\"\n",
    ")(u1)\n",
    "u1 = tf.keras.layers.BatchNormalization(momentum=0.8)(u1)\n",
    "\n",
    "u2 = tf.keras.layers.Concatenate()([u1, d3])\n",
    "u2 = tf.keras.layers.UpSampling2D(size=2)(u2)\n",
    "u2 = tf.keras.layers.Conv2D(\n",
    "    gf * 2, kernel_size=k, strides=1, padding=\"same\", activation=\"relu\"\n",
    ")(u2)\n",
    "u2 = tf.keras.layers.BatchNormalization(momentum=0.8)(u2)\n",
    "\n",
    "u3 = tf.keras.layers.Concatenate()([u2, d2])\n",
    "u3 = tf.keras.layers.UpSampling2D(size=2)(u3)\n",
    "u3 = tf.keras.layers.Conv2D(\n",
    "    gf, kernel_size=k, strides=1, padding=\"same\", activation=\"relu\"\n",
    ")(u3)\n",
    "u3 = tf.keras.layers.BatchNormalization(momentum=0.8)(u3)\n",
    "\n",
    "u4 = tf.keras.layers.Concatenate()([u3, d1])\n",
    "u4 = tf.keras.layers.UpSampling2D(size=2)(u4)\n",
    "u4 = tf.keras.layers.Conv2D(\n",
    "    1, kernel_size=4, strides=1, padding=\"same\", activation=\"tanh\"\n",
    ")(u4)\n",
    "\n",
    "generator = tf.keras.models.Model(img_g, u4)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Generator model visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:29.951429Z",
     "start_time": "2018-08-06T17:31:29.943419Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      " input_1 (InputLayer)           [(None, 32, 32, 1)]  0           []                               \n",
      "                                                                                                  \n",
      " conv2d (Conv2D)                (None, 16, 16, 32)   544         ['input_1[0][0]']                \n",
      "                                                                                                  \n",
      " leaky_re_lu (LeakyReLU)        (None, 16, 16, 32)   0           ['conv2d[0][0]']                 \n",
      "                                                                                                  \n",
      " conv2d_1 (Conv2D)              (None, 8, 8, 64)     32832       ['leaky_re_lu[0][0]']            \n",
      "                                                                                                  \n",
      " leaky_re_lu_1 (LeakyReLU)      (None, 8, 8, 64)     0           ['conv2d_1[0][0]']               \n",
      "                                                                                                  \n",
      " batch_normalization (BatchNorm  (None, 8, 8, 64)    256         ['leaky_re_lu_1[0][0]']          \n",
      " alization)                                                                                       \n",
      "                                                                                                  \n",
      " conv2d_2 (Conv2D)              (None, 4, 4, 128)    131200      ['batch_normalization[0][0]']    \n",
      "                                                                                                  \n",
      " leaky_re_lu_2 (LeakyReLU)      (None, 4, 4, 128)    0           ['conv2d_2[0][0]']               \n",
      "                                                                                                  \n",
      " batch_normalization_1 (BatchNo  (None, 4, 4, 128)   512         ['leaky_re_lu_2[0][0]']          \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " conv2d_3 (Conv2D)              (None, 2, 2, 256)    524544      ['batch_normalization_1[0][0]']  \n",
      "                                                                                                  \n",
      " leaky_re_lu_3 (LeakyReLU)      (None, 2, 2, 256)    0           ['conv2d_3[0][0]']               \n",
      "                                                                                                  \n",
      " batch_normalization_2 (BatchNo  (None, 2, 2, 256)   1024        ['leaky_re_lu_3[0][0]']          \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " up_sampling2d (UpSampling2D)   (None, 4, 4, 256)    0           ['batch_normalization_2[0][0]']  \n",
      "                                                                                                  \n",
      " conv2d_4 (Conv2D)              (None, 4, 4, 128)    524416      ['up_sampling2d[0][0]']          \n",
      "                                                                                                  \n",
      " batch_normalization_3 (BatchNo  (None, 4, 4, 128)   512         ['conv2d_4[0][0]']               \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " concatenate (Concatenate)      (None, 4, 4, 256)    0           ['batch_normalization_3[0][0]',  \n",
      "                                                                  'batch_normalization_1[0][0]']  \n",
      "                                                                                                  \n",
      " up_sampling2d_1 (UpSampling2D)  (None, 8, 8, 256)   0           ['concatenate[0][0]']            \n",
      "                                                                                                  \n",
      " conv2d_5 (Conv2D)              (None, 8, 8, 64)     262208      ['up_sampling2d_1[0][0]']        \n",
      "                                                                                                  \n",
      " batch_normalization_4 (BatchNo  (None, 8, 8, 64)    256         ['conv2d_5[0][0]']               \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " concatenate_1 (Concatenate)    (None, 8, 8, 128)    0           ['batch_normalization_4[0][0]',  \n",
      "                                                                  'batch_normalization[0][0]']    \n",
      "                                                                                                  \n",
      " up_sampling2d_2 (UpSampling2D)  (None, 16, 16, 128)  0          ['concatenate_1[0][0]']          \n",
      "                                                                                                  \n",
      " conv2d_6 (Conv2D)              (None, 16, 16, 32)   65568       ['up_sampling2d_2[0][0]']        \n",
      "                                                                                                  \n",
      " batch_normalization_5 (BatchNo  (None, 16, 16, 32)  128         ['conv2d_6[0][0]']               \n",
      " rmalization)                                                                                     \n",
      "                                                                                                  \n",
      " concatenate_2 (Concatenate)    (None, 16, 16, 64)   0           ['batch_normalization_5[0][0]',  \n",
      "                                                                  'leaky_re_lu[0][0]']            \n",
      "                                                                                                  \n",
      " up_sampling2d_3 (UpSampling2D)  (None, 32, 32, 64)  0           ['concatenate_2[0][0]']          \n",
      "                                                                                                  \n",
      " conv2d_7 (Conv2D)              (None, 32, 32, 1)    1025        ['up_sampling2d_3[0][0]']        \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 1,545,025\n",
      "Trainable params: 1,543,681\n",
      "Non-trainable params: 1,344\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# prints a summary representation of your model\n",
    "generator.summary()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Discriminator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:30.106450Z",
     "start_time": "2018-08-06T17:31:29.953950Z"
    }
   },
   "outputs": [],
   "source": [
    "# Discriminator network\n",
    "k = 4\n",
    "\n",
    "discriminator = tf.keras.models.Sequential()\n",
    "discriminator.add(\n",
    "    tf.keras.layers.Conv2D(\n",
    "        64, kernel_size=k, strides=2, padding=\"same\", input_shape=img_shape\n",
    "    )\n",
    ")\n",
    "discriminator.add(tf.keras.layers.LeakyReLU(alpha=0.8))\n",
    "discriminator.add(\n",
    "    tf.keras.layers.Conv2D(128, kernel_size=k, strides=2, padding=\"same\")\n",
    ")\n",
    "discriminator.add(tf.keras.layers.LeakyReLU(alpha=0.2))\n",
    "discriminator.add(\n",
    "    tf.keras.layers.Conv2D(256, kernel_size=k, strides=2, padding=\"same\")\n",
    ")\n",
    "discriminator.add(tf.keras.layers.LeakyReLU(alpha=0.2))\n",
    "\n",
    "img_d = tf.keras.layers.Input(shape=img_shape)\n",
    "features = discriminator(img_d)\n",
    "\n",
    "validity = tf.keras.layers.Conv2D(1, kernel_size=k, strides=1, padding=\"same\")(\n",
    "    features\n",
    ")\n",
    "\n",
    "label = tf.keras.layers.Flatten()(features)\n",
    "label = tf.keras.layers.Dense(num_classes + 1, activation=\"softmax\")(label)\n",
    "\n",
    "discriminator = tf.keras.models.Model(img_d, [validity, label])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Discriminator model visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:30.115318Z",
     "start_time": "2018-08-06T17:31:30.108486Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_1\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      " input_2 (InputLayer)           [(None, 32, 32, 1)]  0           []                               \n",
      "                                                                                                  \n",
      " sequential (Sequential)        (None, 4, 4, 256)    656832      ['input_2[0][0]']                \n",
      "                                                                                                  \n",
      " flatten (Flatten)              (None, 4096)         0           ['sequential[0][0]']             \n",
      "                                                                                                  \n",
      " conv2d_11 (Conv2D)             (None, 4, 4, 1)      4097        ['sequential[0][0]']             \n",
      "                                                                                                  \n",
      " dense (Dense)                  (None, 11)           45067       ['flatten[0][0]']                \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 705,996\n",
      "Trainable params: 705,996\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# prints a summary representation of your model\n",
    "discriminator.summary()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Compile model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Compile discriminator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:30.167853Z",
     "start_time": "2018-08-06T17:31:30.117566Z"
    }
   },
   "outputs": [],
   "source": [
    "# Optimizer\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.0002, beta_1=0.5)\n",
    "\n",
    "discriminator.compile(\n",
    "    optimizer=optimizer, \n",
    "    loss=['mse', 'categorical_crossentropy'],\n",
    "    loss_weights=[0.5, 0.5], \n",
    "    metrics=['accuracy']\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Combined network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:30.728960Z",
     "start_time": "2018-08-06T17:31:30.169843Z"
    }
   },
   "outputs": [],
   "source": [
    "# The generator takes noise as input and generates imgs\n",
    "masked_img = tf.keras.layers.Input(shape=(img_shape))\n",
    "gen_img = generator(masked_img)\n",
    "\n",
    "# For the combined model we will only train the generator\n",
    "discriminator.trainable = False\n",
    "\n",
    "validity, _ = discriminator(gen_img)\n",
    "\n",
    "d_g = tf.keras.models.Model(masked_img, validity)\n",
    "\n",
    "d_g.compile(optimizer=optimizer, loss='mse', metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:30.736364Z",
     "start_time": "2018-08-06T17:31:30.731183Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_3 (InputLayer)        [(None, 32, 32, 1)]       0         \n",
      "                                                                 \n",
      " model (Functional)          (None, 32, 32, 1)         1545025   \n",
      "                                                                 \n",
      " model_1 (Functional)        [(None, 4, 4, 1),         705996    \n",
      "                              (None, 11)]                        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 2,251,021\n",
      "Trainable params: 1,543,681\n",
      "Non-trainable params: 707,340\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# prints a summary representation of your model\n",
    "d_g.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:30.743590Z",
     "start_time": "2018-08-06T17:31:30.738902Z"
    }
   },
   "outputs": [],
   "source": [
    "def mask_randomly(imgs, mask_width=10, mask_height=10):\n",
    "    y1 = np.random.randint(0, imgs.shape[1] - mask_height, imgs.shape[0])\n",
    "    y2 = y1 + mask_height\n",
    "    x1 = np.random.randint(0, imgs.shape[2] - mask_width, imgs.shape[0])\n",
    "    x2 = x1 + mask_width\n",
    "\n",
    "    masked_imgs = np.empty_like(imgs)\n",
    "    for i, img in enumerate(imgs):\n",
    "        masked_img = img.copy()\n",
    "        _y1, _y2, _x1, _x2 = y1[i], y2[i], x1[i], x2[i],\n",
    "        masked_img[_y1:_y2, _x1:_x2, :] = 0\n",
    "        masked_imgs[i] = masked_img\n",
    "\n",
    "    return masked_imgs\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T17:31:31.014551Z",
     "start_time": "2018-08-06T17:31:30.746003Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaEAAAGdCAYAAAC7EMwUAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAf0ElEQVR4nO3df3TU9b3n8dcQyAiYjGUh80NCmmKQKsptQSEUIdAll9wtC9LuQb3XG9bWIwqc5UaXFt2z5vSeJRQvFPdQsbU9VKsUerai3isC6YUEXUobWFgQLRdLqLFkzCGFTAg4Icln/+gyNRLg+4EZPpnJ83HO9xzm+33nM++v35hXPvnOfMZnjDECAMCBfq4bAAD0XYQQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGf6u27gs7q6unTixAnl5OTI5/O5bgcAYMkYo9bWVkUiEfXrd/m5Tq8LoRMnTig/P991GwCAa9TQ0KDhw4dftiZlIfTcc8/pmWeeUWNjo26//XatWbNG99xzzxW/LicnR5I0WX+j/hqQqvYAACnSofN6R1sSP88vJyUhtGnTJi1ZskTPPfecvvKVr+iHP/yhysrK9N5772nEiBGX/doLf4LrrwHq7yOEACDt/P8VSb3cUknJCxNWr16tb37zm/rWt76lL37xi1qzZo3y8/O1bt26VDwdACBNJT2E2tvbtW/fPpWWlnbbX1paqt27d19UH4/HFYvFum0AgL4h6SF08uRJdXZ2KhgMdtsfDAYVjUYvqq+qqlIgEEhsvCgBAPqOlL1P6LN/CzTG9Pj3wWXLlqmlpSWxNTQ0pKolAEAvk/QXJgwdOlRZWVkXzXqampoumh1Jkt/vl9/vT3YbAIA0kPSZUHZ2tsaNG6fq6upu+6urqzVp0qRkPx0AII2l5CXaFRUVevDBBzV+/HgVFxfrRz/6kT788EMtWLAgFU8HAEhTKQmhefPmqbm5Wd/97nfV2NioMWPGaMuWLSooKEjF0wEA0pTPGGNcN/FpsVhMgUBAJZrNm1UBIA11mPOq0etqaWlRbm7uZWtZRRsA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4Ex/1w0AuLIP1kx03QKS5JYle1y30KswEwIAOEMIAQCcIYQAAM4QQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcIYQAAM6wdhxwKT6f99Ls7BQ2gkzSPxS0qu+IfpyiTnoHZkIAAGeSHkKVlZXy+XzdtlAolOynAQBkgJT8Oe7222/Xr371q8TjrKysVDwNACDNpSSE+vfvz+wHAHBFKbkndPToUUUiERUWFuq+++7TsWPHLlkbj8cVi8W6bQCAviHpITRhwgS99NJL2rZtm1544QVFo1FNmjRJzc3NPdZXVVUpEAgktvz8/GS3BADopXzGGJPKJ2hra9PIkSO1dOlSVVRUXHQ8Ho8rHo8nHsdiMeXn56tEs9XfNyCVrQGX14teon30e19K6fi4fkavqLeqT8eXaHeY86rR62ppaVFubu5la1P+PqHBgwfrjjvu0NGjR3s87vf75ff7U90GAKAXSvn7hOLxuN5//32Fw+FUPxUAIM0kPYSeeOIJ1dbWqr6+Xr/5zW/0jW98Q7FYTOXl5cl+KgBAmkv6n+M++ugj3X///Tp58qSGDRumiRMnas+ePSooKEj2U8ER3wC7+x++LO+/6/gGD7Ib+8bBnmvNoBusxu4a7P3PxGdvtusbfVf8izdb1Wel4T0hG0kPoY0bNyZ7SABAhmLtOACAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMCZlH+UA9JAvyyrct9tI63qO27yvmbb6ZF267s1j/X+cVg3FZ6yGnts3h88164f8bbV2LZG/mJBSsfH9ZPdYPd92JmiPnoLZkIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAMyzbk6ksluIxE8ZYDR35/u+t6v9L6Feea4f067Aa+wafz6LWbnmifla/o2VbjY2+q/P3x1230KswEwIAOEMIAQCcIYQAAM4QQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcIYQAAM6wdlymMl2eSwec+JPV0K0dfqv6UFan59q8rButxu4tzhvv5yhJz5/+Qoo6Qa9njOsOehVmQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcIYQAAM4QQgAAZwghAIAzhBAAwBnWjstUFutTdUabrIZ+75/HWdX/h3v+s+dan89uXa2JweOea/9npM5q7Jauc55rX2oZbTX2z575G6t6/RXrjSEzMRMCADhjHUK7du3SrFmzFIlE5PP59Nprr3U7boxRZWWlIpGIBg4cqJKSEh0+fDhZ/QIAMoh1CLW1tWns2LFau3Ztj8dXrlyp1atXa+3ataqrq1MoFNKMGTPU2tp6zc0CADKL9T2hsrIylZWV9XjMGKM1a9boqaee0ty5cyVJL774ooLBoDZs2KBHHnnk2roFAGSUpN4Tqq+vVzQaVWlpaWKf3+/X1KlTtXv37h6/Jh6PKxaLddsAAH1DUkMoGo1KkoLBYLf9wWAwceyzqqqqFAgEElt+fn4yWwIA9GIpeXWcz+fr9tgYc9G+C5YtW6aWlpbE1tDQkIqWAAC9UFLfJxQKhST9eUYUDocT+5uami6aHV3g9/vl9/uT2QYAIE0kdSZUWFioUCik6urqxL729nbV1tZq0qRJyXwqAEAGsJ4JnTlzRh988EHicX19vQ4cOKAhQ4ZoxIgRWrJkiZYvX66ioiIVFRVp+fLlGjRokB544IGkNg4ASH/WIbR3715NmzYt8biiokKSVF5erp/+9KdaunSpzp07p8cee0ynTp3ShAkTtH37duXk5CSvaySVicet6kds/tiqvvXfhnqu7Xfebnma7Xd7H/sn/8nufuMdN3iv//7OmVZjj/7nI1b1zX81yqoeSBfWIVRSUiJzmXXJfD6fKisrVVlZeS19AQD6ANaOAwA4QwgBAJwhhAAAzhBCAABnCCEAgDOEEADAGUIIAOAMIQQAcIYQAgA4QwgBAJxJ6kc5oG/o/LffW9UP/sNHFoN3Wo09LHuc59rNU75kNfbtI/7oubZfvOfPy7oU037eqh7IVMyEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGdYtgcpZ+LxlI09sPETz7WHP7jZauz2EVmea2eX/NZq7Pe+cKtVPZCpmAkBAJwhhAAAzhBCAABnCCEAgDOEEADAGUIIAOAMIQQAcIYQAgA4QwgBAJwhhAAAzhBCAABnWDsOaa1f3WHPtQU3fclq7P9R+DXPtT8p2mg1dul/nGhVP3JTm+faAY2nrMbuavZe39XaajU2cCXMhAAAzhBCAABnCCEAgDOEEADAGUIIAOAMIQQAcIYQAgA4QwgBAJwhhAAAzhBCAABnWLYHac10dHiuHfS/j1iNfTr3Ns+1H1UNtBp7TfkLVvWrSv7ac+37799sNXbBvwQ91/p3HrQa27S3WxQbq7GRGZgJAQCcIYQAAM5Yh9CuXbs0a9YsRSIR+Xw+vfbaa92Oz58/Xz6fr9s2caLdisEAgL7BOoTa2to0duxYrV279pI1M2fOVGNjY2LbsmXLNTUJAMhM1i9MKCsrU1lZ2WVr/H6/QqHQVTcFAOgbUnJPqKamRnl5eRo1apQefvhhNTU1XbI2Ho8rFot12wAAfUPSQ6isrEyvvPKKduzYoVWrVqmurk7Tp09XPB7vsb6qqkqBQCCx5efnJ7slAEAvlfT3Cc2bNy/x7zFjxmj8+PEqKCjQm2++qblz515Uv2zZMlVUVCQex2IxgggA+oiUv1k1HA6roKBAR48e7fG43++X3+9PdRsAgF4o5e8Tam5uVkNDg8LhcKqfCgCQZqxnQmfOnNEHH3yQeFxfX68DBw5oyJAhGjJkiCorK/X1r39d4XBYx48f15NPPqmhQ4fq3nvvTWrjAID05zPGbsGmmpoaTZs27aL95eXlWrdunebMmaP9+/fr9OnTCofDmjZtmv7xH//R832eWCymQCCgEs1Wf98Am9aApOp/c8Rz7Yd/+3mrsZ9fcOn32fXktgGfeK7tlN0abLPe/XvvxT8bZjX25/71mOfazo8v/SpapJcOc141el0tLS3Kzc29bK31TKikpESXy61t27bZDgkA6KNYOw4A4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwJuUf5QCkq47Gjz3XjnjFbux/+HihVX3g7z/yXLtm5C+sxn75thc91/7XBXYLEf9+WJHn2uG/sPuduKMxalWP3omZEADAGUIIAOAMIQQAcIYQAgA4QwgBAJwhhAAAzhBCAABnCCEAgDOEEADAGUIIAOAMy/YAl9LV6bm0448nrIYe+i+fWNWfPTnSc+2ceyqsxv67slrPtc8UbLYae9Xf/XvPtTs/92WrsUdUsmxPJmAmBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnGHtOMCBzuY/WdXf8Ks2z7VFx0ZYjf2Loi95rv1vE9+1Grsy9K+ea0/PGGg19qlXR3uu7Xr3qNXYNusG4towEwIAOEMIAQCcIYQAAM4QQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcYdkeIBl8Pqvy/pGwVX18VMhzbVs422rsyE1/tKq3MUDe/7vclH3OauxTts2gV2ImBABwxiqEqqqqdNdddyknJ0d5eXmaM2eOjhw50q3GGKPKykpFIhENHDhQJSUlOnz4cFKbBgBkBqsQqq2t1cKFC7Vnzx5VV1ero6NDpaWlamv7ywq/K1eu1OrVq7V27VrV1dUpFAppxowZam1tTXrzAID0ZnVPaOvWrd0er1+/Xnl5edq3b5+mTJkiY4zWrFmjp556SnPnzpUkvfjiiwoGg9qwYYMeeeSR5HUOAEh713RPqKWlRZI0ZMgQSVJ9fb2i0ahKS0sTNX6/X1OnTtXu3bt7HCMejysWi3XbAAB9w1WHkDFGFRUVmjx5ssaMGSNJikajkqRgMNitNhgMJo59VlVVlQKBQGLLz8+/2pYAAGnmqkNo0aJFOnjwoH7+859fdMz3mZerGmMu2nfBsmXL1NLSktgaGhqutiUAQJq5qvcJLV68WG+88YZ27dql4cOHJ/aHQn9+L0M0GlU4/Jf3QTQ1NV00O7rA7/fL7/dfTRsAgDRnNRMyxmjRokV69dVXtWPHDhUWFnY7XlhYqFAopOrq6sS+9vZ21dbWatKkScnpGACQMaxmQgsXLtSGDRv0+uuvKycnJ3GfJxAIaODAgfL5fFqyZImWL1+uoqIiFRUVafny5Ro0aJAeeOCBlJwAACB9WYXQunXrJEklJSXd9q9fv17z58+XJC1dulTnzp3TY489plOnTmnChAnavn27cnJyktIwACBz+IwxxnUTnxaLxRQIBFSi2ervG+C6HWQQX3+7W6BZw4Z6ro2PjliN/fH4G6zqO+72/mbvKQUfWI39zWG7PNfe7bf7f/Jwu/f14B448JDV2KF7f+e9uHf9mMt4Hea8avS6WlpalJube9la1o4DADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnLmqj3IAUqZfllV5VuDyS4J8Wtfn7ZbWOfGVgOfawV/r+UMbL+X5Uf/Lqn68v9Nzrd9yuaszXd7H/m3cami93DzFc+3Z391kNzhL8WQEZkIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZ1o6DPZ/PrjzL+3pwWTeHrcZunnyz59o/zTprNfaP737Oc+2UG6yGlu3vf+ct1kk72dlmNfam1tGea//p7ZlWY4/c6H1dui/s/LXV2MgMzIQAAM4QQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcIYQAAM4QQgAAZ1i2B9b6B/Os6uOjvS+tc+RvvS/xI0nfm/pzz7V/PShqNfaNPr9FdWp/n3v+9Bc8167ZVmY19hdej3uu/eL+I1Zjd531vlSS94WJkEmYCQEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGdYOy5d9LNbU61/fsRz7ftPeK+VpILRdmuwzQy97bn2n25812rsUQN8nmsH9RtoNfbJzjbPtc+c/IrV2Ju3FlvVR3Z1eK4d/btGq7G7Gj/2XNv5ySdWYwNXwkwIAOCMVQhVVVXprrvuUk5OjvLy8jRnzhwdOdJ9Vd358+fL5/N12yZOnJjUpgEAmcEqhGpra7Vw4ULt2bNH1dXV6ujoUGlpqdrauv/ZYubMmWpsbExsW7ZsSWrTAIDMYHVPaOvWrd0er1+/Xnl5edq3b5+mTJmS2O/3+xUKhZLTIQAgY13TPaGWlhZJ0pAhQ7rtr6mpUV5enkaNGqWHH35YTU1NlxwjHo8rFot12wAAfcNVh5AxRhUVFZo8ebLGjBmT2F9WVqZXXnlFO3bs0KpVq1RXV6fp06crHu/50xurqqoUCAQSW35+/tW2BABIM1f9Eu1Fixbp4MGDeuedd7rtnzdvXuLfY8aM0fjx41VQUKA333xTc+fOvWicZcuWqaKiIvE4FosRRADQR1xVCC1evFhvvPGGdu3apeHDh1+2NhwOq6CgQEePHu3xuN/vl9/vv5o2AABpziqEjDFavHixNm/erJqaGhUWFl7xa5qbm9XQ0KBwOHzVTQIAMpPVPaGFCxfq5Zdf1oYNG5STk6NoNKpoNKpz585Jks6cOaMnnnhCv/71r3X8+HHV1NRo1qxZGjp0qO69996UnAAAIH1ZzYTWrVsnSSopKem2f/369Zo/f76ysrJ06NAhvfTSSzp9+rTC4bCmTZumTZs2KScnJ2lNAwAyg/Wf4y5n4MCB2rZt2zU1lM6ygnlW9WeKP++59o9T7F7IeFPRnzzXrhn9M6ux78i+9EvuexLp7/2en99nd3/wvOn0XPtG2yCrsR+vm++5NvhLu75vOWT339B85H09uI6zZ63GBlxi7TgAgDOEEADAGUIIAOAMIQQAcIYQAgA4QwgBAJwhhAAAzhBCAABnCCEAgDOEEADAmav+PCFcrGOk3UrhDTO9137/qy9bjT3eH/VcO7z/jVZjH2y3+7b57013ea6tb/t3VmP/3z/e7LnW/1u78/z83nOea7N277cau/N8u1U9kKmYCQEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhDCAEAnCGEAADOEEIAAGdYOy6JsmJxq/qco4M91/5D9n227aRM1p/svm1y6r3/ruNv6bIae8TxTzzX9t/7f6zG7vrE+9jGamQAFzATAgA4QwgBAJwhhAAAzhBCAABnCCEAgDOEEADAGUIIAOAMIQQAcIYQAgA4QwgBAJxh2Z4k6nr3d1b14Xctai17wcXsFgQCcD0wEwIAOEMIAQCcIYQAAM4QQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcIYQAAM4QQgAAZwghAIAzViG0bt063XnnncrNzVVubq6Ki4v11ltvJY4bY1RZWalIJKKBAweqpKREhw8fTnrTAIDMYBVCw4cP14oVK7R3717t3btX06dP1+zZsxNBs3LlSq1evVpr165VXV2dQqGQZsyYodbW1pQ0DwBIbz5jjLmWAYYMGaJnnnlGDz30kCKRiJYsWaJvf/vbkqR4PK5gMKjvfe97euSRRzyNF4vFFAgEVKLZ6u8bcC2tAQAc6DDnVaPX1dLSotzc3MvWXvU9oc7OTm3cuFFtbW0qLi5WfX29otGoSktLEzV+v19Tp07V7t27LzlOPB5XLBbrtgEA+gbrEDp06JBuvPFG+f1+LViwQJs3b9Ztt92maDQqSQoGg93qg8Fg4lhPqqqqFAgEElt+fr5tSwCANGUdQrfeeqsOHDigPXv26NFHH1V5ebnee++9xHGfz9et3hhz0b5PW7ZsmVpaWhJbQ0ODbUsAgDTV3/YLsrOzdcstt0iSxo8fr7q6Oj377LOJ+0DRaFThcDhR39TUdNHs6NP8fr/8fr9tGwCADHDN7xMyxigej6uwsFChUEjV1dWJY+3t7aqtrdWkSZOu9WkAABnIaib05JNPqqysTPn5+WptbdXGjRtVU1OjrVu3yufzacmSJVq+fLmKiopUVFSk5cuXa9CgQXrggQdS1T8AII1ZhdDHH3+sBx98UI2NjQoEArrzzju1detWzZgxQ5K0dOlSnTt3To899phOnTqlCRMmaPv27crJyUlJ8wCA9HbN7xNKNt4nBADp7bq8TwgAgGtFCAEAnCGEAADOEEIAAGcIIQCAM4QQAMAZQggA4AwhBABwhhACADhjvYp2ql1YwKFD56VetZYDAMCLDp2X9Jef55fT60KotbVVkvSOtjjuBABwLVpbWxUIBC5b0+vWjuvq6tKJEyeUk5PT7cPwYrGY8vPz1dDQcMW1iNIZ55k5+sI5SpxnpknGeRpj1Nraqkgkon79Ln/Xp9fNhPr166fhw4df8nhubm5GfwNcwHlmjr5wjhLnmWmu9TyvNAO6gBcmAACcIYQAAM6kTQj5/X49/fTT8vv9rltJKc4zc/SFc5Q4z0xzvc+z170wAQDQd6TNTAgAkHkIIQCAM4QQAMAZQggA4EzahNBzzz2nwsJC3XDDDRo3bpzefvtt1y0lVWVlpXw+X7ctFAq5buua7Nq1S7NmzVIkEpHP59Nrr73W7bgxRpWVlYpEIho4cKBKSkp0+PBhN81egyud5/z58y+6thMnTnTT7FWqqqrSXXfdpZycHOXl5WnOnDk6cuRIt5pMuJ5ezjMTrue6det05513Jt6QWlxcrLfeeitx/Hpey7QIoU2bNmnJkiV66qmntH//ft1zzz0qKyvThx9+6Lq1pLr99tvV2NiY2A4dOuS6pWvS1tamsWPHau3atT0eX7lypVavXq21a9eqrq5OoVBIM2bMSKwfmC6udJ6SNHPmzG7XdsuW9Fobsba2VgsXLtSePXtUXV2tjo4OlZaWqq2tLVGTCdfTy3lK6X89hw8frhUrVmjv3r3au3evpk+frtmzZyeC5rpeS5MG7r77brNgwYJu+0aPHm2+853vOOoo+Z5++mkzduxY122kjCSzefPmxOOuri4TCoXMihUrEvs++eQTEwgEzPPPP++gw+T47HkaY0x5ebmZPXu2k35SpampyUgytbW1xpjMvZ6fPU9jMvN6GmPM5z73OfPjH//4ul/LXj8Tam9v1759+1RaWtptf2lpqXbv3u2oq9Q4evSoIpGICgsLdd999+nYsWOuW0qZ+vp6RaPRbtfV7/dr6tSpGXddJammpkZ5eXkaNWqUHn74YTU1Nblu6Zq0tLRIkoYMGSIpc6/nZ8/zgky6np2dndq4caPa2tpUXFx83a9lrw+hkydPqrOzU8FgsNv+YDCoaDTqqKvkmzBhgl566SVt27ZNL7zwgqLRqCZNmqTm5mbXraXEhWuX6ddVksrKyvTKK69ox44dWrVqlerq6jR9+nTF43HXrV0VY4wqKio0efJkjRkzRlJmXs+ezlPKnOt56NAh3XjjjfL7/VqwYIE2b96s22677bpfy163ivalfPpjHaQ/f4N8dl86KysrS/z7jjvuUHFxsUaOHKkXX3xRFRUVDjtLrUy/rpI0b968xL/HjBmj8ePHq6CgQG+++abmzp3rsLOrs2jRIh08eFDvvPPORccy6Xpe6jwz5XreeuutOnDggE6fPq1f/vKXKi8vV21tbeL49bqWvX4mNHToUGVlZV2UwE1NTRcldSYZPHiw7rjjDh09etR1Kylx4ZV/fe26SlI4HFZBQUFaXtvFixfrjTfe0M6dO7t95EqmXc9LnWdP0vV6Zmdn65ZbbtH48eNVVVWlsWPH6tlnn73u17LXh1B2drbGjRun6urqbvurq6s1adIkR12lXjwe1/vvv69wOOy6lZQoLCxUKBTqdl3b29tVW1ub0ddVkpqbm9XQ0JBW19YYo0WLFunVV1/Vjh07VFhY2O14plzPK51nT9LxevbEGKN4PH79r2XSX+qQAhs3bjQDBgwwP/nJT8x7771nlixZYgYPHmyOHz/uurWkefzxx01NTY05duyY2bNnj/na175mcnJy0vocW1tbzf79+83+/fuNJLN69Wqzf/9+84c//MEYY8yKFStMIBAwr776qjl06JC5//77TTgcNrFYzHHndi53nq2trebxxx83u3fvNvX19Wbnzp2muLjY3HzzzWl1no8++qgJBAKmpqbGNDY2JrazZ88majLhel7pPDPlei5btszs2rXL1NfXm4MHD5onn3zS9OvXz2zfvt0Yc32vZVqEkDHG/OAHPzAFBQUmOzvbfPnLX+72kslMMG/ePBMOh82AAQNMJBIxc+fONYcPH3bd1jXZuXOnkXTRVl5eboz588t6n376aRMKhYzf7zdTpkwxhw4dctv0VbjceZ49e9aUlpaaYcOGmQEDBpgRI0aY8vJy8+GHH7pu20pP5yfJrF+/PlGTCdfzSueZKdfzoYceSvw8HTZsmPnqV7+aCCBjru+15KMcAADO9Pp7QgCAzEUIAQCcIYQAAM4QQgAAZwghAIAzhBAAwBlCCADgDCEEAHCGEAIAOEMIAQCcIYQAAM4QQgAAZ/4f9bR5ZX1JeEkAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(mask_randomly(X_train[0:1])[0].reshape(32, 32))\n",
    "plt.show()\n",
    "plt.imshow(mask_randomly(X_train[0:1])[0].reshape(32, 32))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Fit model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T22:11:27.329963Z",
     "start_time": "2018-08-06T17:31:31.016456Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-11-07 15:04:18.068245: I tensorflow/stream_executor/cuda/cuda_dnn.cc:384] Loaded cuDNN version 8800\n",
      "2023-11-07 15:04:18.120787: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-11-07 15:04:18.121321: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-11-07 15:04:18.121337: W tensorflow/stream_executor/gpu/asm_compiler.cc:80] Couldn't get ptxas version string: INTERNAL: Couldn't invoke ptxas --version\n",
      "2023-11-07 15:04:18.121935: I tensorflow/core/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n",
      "2023-11-07 15:04:18.121995: W tensorflow/stream_executor/gpu/redzone_allocator.cc:314] INTERNAL: Failed to launch ptxas\n",
      "Relying on driver to perform ptx compilation. \n",
      "Modify $PATH to customize ptxas location.\n",
      "This message will be only logged once.\n",
      "2023-11-07 15:04:18.241195: I tensorflow/stream_executor/cuda/cuda_blas.cc:1614] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 1/100, d_loss=0.382, g_loss=0.290                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 2/100, d_loss=0.275, g_loss=0.271                                                                                                                      \n",
      "epoch = 3/100, d_loss=0.318, g_loss=0.244                                                                                                                      \n",
      "epoch = 4/100, d_loss=0.304, g_loss=0.237                                                                                                                      \n",
      "epoch = 5/100, d_loss=0.297, g_loss=0.324                                                                                                                      \n",
      "epoch = 6/100, d_loss=0.223, g_loss=0.731                                                                                                                      \n",
      "epoch = 7/100, d_loss=0.234, g_loss=0.533                                                                                                                       \n",
      "epoch = 8/100, d_loss=0.216, g_loss=0.717                                                                                                                      \n",
      "epoch = 9/100, d_loss=0.162, g_loss=0.854                                                                                                                       \n",
      "epoch = 10/100, d_loss=0.182, g_loss=0.788                                                                                                                       \n",
      "epoch = 11/100, d_loss=0.219, g_loss=6.012                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 12/100, d_loss=0.146, g_loss=0.801                                                                                                                       \n",
      "epoch = 13/100, d_loss=0.108, g_loss=0.762                                                                                                                      \n",
      "epoch = 14/100, d_loss=0.104, g_loss=0.795                                                                                                                      \n",
      "epoch = 15/100, d_loss=0.125, g_loss=0.861                                                                                                                      \n",
      "epoch = 16/100, d_loss=0.091, g_loss=0.873                                                                                                                       \n",
      "epoch = 17/100, d_loss=0.103, g_loss=0.604                                                                                                                      \n",
      "epoch = 18/100, d_loss=0.098, g_loss=1.033                                                                                                                      \n",
      "epoch = 19/100, d_loss=0.070, g_loss=0.811                                                                                                                       \n",
      "epoch = 20/100, d_loss=0.076, g_loss=0.927                                                                                                                      \n",
      "epoch = 21/100, d_loss=0.064, g_loss=0.793                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 22/100, d_loss=0.062, g_loss=0.867                                                                                                                      \n",
      "epoch = 23/100, d_loss=0.054, g_loss=0.825                                                                                                                      \n",
      "epoch = 24/100, d_loss=0.060, g_loss=0.827                                                                                                                      \n",
      "epoch = 25/100, d_loss=0.042, g_loss=0.941                                                                                                                      \n",
      "epoch = 26/100, d_loss=0.221, g_loss=0.598                                                                                                                      \n",
      "epoch = 27/100, d_loss=0.044, g_loss=0.799                                                                                                                      \n",
      "epoch = 28/100, d_loss=0.031, g_loss=0.804                                                                                                                      \n",
      "epoch = 29/100, d_loss=0.023, g_loss=0.776                                                                                                                      \n",
      "epoch = 30/100, d_loss=0.027, g_loss=0.874                                                                                                                      \n",
      "epoch = 31/100, d_loss=0.040, g_loss=0.825                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnUAAAFBCAYAAAAR9FlyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA1RklEQVR4nO3de7SVdZ3H8e9BTRSPECqXIzdFCREQU1S8pVFqOY5TXsoSxcoxsxY2zuhYFs6Us2rWxKw1XczSnHQqmUpNFBUVsERFEAEVlJsIckQUBI4gopwzf7TObz6/L2fvs/c++/rb79darvme+T3s85z9PM/ev57v9/n+Gtra2toMAAAANa1bpXcAAAAAXcekDgAAIAFM6gAAABLApA4AACABTOoAAAASwKQOAAAgAUzqAAAAEsCkDgAAIAF75rJRa2urNTc3W2NjozU0NJR6n1AkbW1t1tLSYk1NTdatW37zd455beKY1x+Oef3hmNefXI95TpO65uZmGzhwYNF2DuW1du1aGzBgQF7/hmNe2zjm9YdjXn845vWns2Oe06SusbGxaDtUTv/8z/9c6V0wM7Mf/OAHFf39hRy/Wj3m+CuOef3hmNcfjnn96ez45TSpq9VbtN27d6/0LlSFQo5frR5z/BXHvP5wzOsPx7z+dHb8cprUAeX2L//yL5XeBTMzmzx5cqV3AQCAnPD0KwAAQAKY1AEAACSASR0AAEACmNQBAAAkgEkdAABAApjUAQAAJIBJHQAAQAKY1AEAACSASR0AAEACmNQBAAAkoGqWCevW7f/nl8OGDYvG+vXrF+I333wzGluzZk2IW1paSrR3qFdXX3119PPmzZtDvGLFimhs1apVIW5ubi7lbqHGHHvssSHWzzMzs5deeinE/pxC5/bcM/4a69u3b4g//elPh3ju3LnRdvpeb9++vUR7h1LQ+YKZ2X777Rfigw46KMRtbW3RduvXrw9xqsecO3UAAAAJYFIHAACQgKpJv+61114hPuOMM6Kx008/PcSabjUze+6550Ks6S+gGP7zP/8z+llv3/t0ziOPPBLiqVOnhnjjxo3Rdj4lgOq2xx57hFjTPGZmPXv2DPHOnTtD/MYbb0TbDR8+PMRHHHFENLZ169YQk37NX/fu3aOfx4wZE+J/+7d/C/FPf/rTaLvbb789xK+++mppdg5F06NHjxAfdthh0diIESNCPGrUqBDv2rUr2m7hwoUh9qVcH//4x4uxm1124403dunfc6cOAAAgAUzqAAAAEsCkDgAAIAFVU1Onjygffvjh0dgpp5wS4sbGxmhM61heeeWVaOzuu+8u5i6iDr388svRz7179w6xr/3UWimt17jnnnui7d5///1i7iKKLFuLjKOPPjoa02M+b968EL/11lvRdrNmzQrxE088EY1t2rSp8J0to67W+hST7suHPvShaOzQQw8Nsdbb+WN33333hZiausrI1prEf9drXeqECROiMa27b2pqCnFra2u03WuvvRbid955Jxrzn9O1ijt1AAAACWBSBwAAkICqSb/mSluf+J9HjhwZjZF+rR2+LUG1uPTSS6Of//Zv/zbEn/3sZ6OxQYMGhfhLX/pSiKdPnx5tR/q18rJ9jvTv3z8a01YHX/jCF6IxTcf/5Cc/CfEHH3wQbbdu3brCdxZZ+fTrgAEDQqzpPX/MGxoaSrtj6JSu/mAWX2tadmVm9tGPfjTExx13XDSWqU2UT+8OGTIkxP74k34FAABA1WBSBwAAkICaS78iHZoO0bRmNZk/f37085YtW0Ls0zmTJk0KsS7aTpqnOuy///4hPuuss6KxcePGhXj06NEZX8OnefTJ2Ouuuy7EP/jBD6LtduzYkd/OImd6XM3MPvnJT4ZYr9Hly5dH2+m1jMq48soro58vvvjiEOtTrGa7P5WOjnGnDgAAIAFM6gAAABLApA4AACABSSWpN2zYUOldQB6027dfueHII48s9+50aNeuXdHP+kj88ccfH41t3bo1xDNmzAixb2+B8tljjz1CfMghh4R4/Pjx0XZav6Nd583MZs+eHWK/+sNVV10VYq0HGjt2bLSdtjt58sknozFqu7rG16z6Wtd22i7DLG55tWLFiuLvGDrlW1ntu+++Gcey1SbrZ+zjjz8e4j/+8Y/Rds8//3yI/TXaq1evzne4BnCnDgAAIAFM6gAAABKQVPp1/fr10c/aaVwXWDczO+2000Ks7Q38rfuNGzeG+IYbbojGbr311hD7NB06p+9ZtaY/fMfzUaNGhfjwww+PxnQRd10snBUkSkuP0YknnhiNaRd6Telv3rw52u7OO+8MsS8F0JTriBEjojFdUWLgwIEhzvZ5wGdF+WjKrqWlJRp77733yr07MLMLL7wwxNpKyMyssbEx47/T7+KnnnoqGtNVexYvXhxi38ZGr2X/Xa9zglrGnToAAIAEMKkDAABIAJM6AACABCRVU+eX4tHamDlz5kRjurTMJz7xiRD7PLu+5s6dO4uyn9jdtm3bop8nT55coT0x69bt//+3zoQJE6KxCy64IMT+fHjooYdCrHUdtDQprgMPPDD6+YwzzgjxFVdcEY1pnZsu8fXtb3872u7RRx8NsW9bossTDR48OBrT2jz97PD72LNnzw5fD+Xj66t8jR3KQ6/D7du3R2NvvPFGiP11qHV02o7GLP681dZS2epXteY+JdypAwAASACTOgAAgARUTR5AU17aBb4zmh5tbm6OxlauXBliva1rZvbcc8+FWNtP+FuyS5YsCbHe4jWLV0RA7dIu5mZxG4zzzz8/Ghs0aFCIH3zwwWjsN7/5TYg1BYCu22effUJ80kknRWNf/vKXQ3zMMcdEY88880yIdWWIxx57LNrOp3qUpk41nWtm9uEPfzjEmlZatmxZtJ22YyAdXz56THr06BGNkQavjLlz54ZYv/fNzPr27RtivWbMzBYtWhTiF154IRrTY6mfFV6/fv1C7D9H9FypZdypAwAASACTOgAAgAQwqQMAAEhARYsKtAVA//79Q+zrVrI9eqy1cvPmzYvGXnzxxRD7R6f18ehVq1aF2Nfz6fJiPsevOfi99947GtPlg7RuwD9irUtLUWtTGQMGDIh+njhxYoiPP/74aEzrKu+6665obMGCBcXfuTrSvXv3EPuWIEcccUSIP/OZz0RjRx99dIh9LdvPfvazEN97770hzrZ0m167ZmannnpqiL/whS9k/N1vv/12iG+77bZou/nz54f4nXfeyfi7UTqjR4+OfvbHGeWxZs2aDuN8ZKuDPuyww0Lsv891acdPf/rT0Zhvk1KruFMHAACQACZ1AAAACaho+lVXdTj77LNDfNppp0Xb+UfR1dKlS0Ps01+aDvE0rapxNv4ReL19P2zYsGjshBNOCLGmj30aWNssbNmyJRrT/SdlU1x6+/6UU06JxsaNGxdi/3i8pvS1LY4ZLW5y0dDQEGJtB2IWX0Oa8jSLV33p1atXNDZjxowQP//889HYzJkzQ5xrecPIkSOjny+66KIQ6+oVZmbvvvtuiPXc0BUqzGhxUw18Sv+AAw4IsS/xYfWg6qOf2ZpuNTP7zne+E2KdP2RrW6OfRWakXwEAAFBFmNQBAAAkgEkdAABAAipaU6c1NRdffHGIfXuQbLRVgMbFou1IdIkos7gO8MILL4zGjj322BBnq6nT1ip+KbPf/va3Ida6IbPdW6Ogc3pe6fG57rrrou2GDh0aYv++P/TQQyHesGFDsXcxedrG6LzzzovGLrnkkhD79hNa93r77bdHY7fcckuItUVQoXytn9Zi+dorbXOkyw2+/vrr0XZcr6Xj39v33nuvw+20htvMbOzYsSFeuHBhNLZixYri7BwK5uvh9Hh9//vfj8b8kl+58DV1kydPDrFfMkzPsfXr14f47//+76Pt/vznP4d427Ztee9TMXCnDgAAIAFM6gAAABJQ0fSrtioZM2ZMiH0X6ErS1QauueaaaEy7y/u2K/o36G1e3wn7Yx/7WIh9albTsevWrYvGVq9eHWLaJXTMv9d6+/76668Psa5mYmb22muvhfj++++PxubMmVPMXaw7Ws5wyCGHRGPa7d239/nFL34R4jvuuCMa8yu9dJWuXmFmNmLEiIzbaqpv5cqVIU6xvc2NN95Y6V3o0KZNm6Kf9fy46aabQrzffvtF22kqPVvrC5SPlmeMHz8+Gps0aVKI/Uo/Pl3aVf719HOrX79+IdbPJTOzM888M8RLliwp6j7lijt1AAAACWBSBwAAkICK3nPWtKTe/s7nVupRRx0VYv/E3KxZs/Lep0MPPTT6+aqrrgqxXwBYn6byT9Jk4rfT282NjY3R2IQJE0LsO2j/7//+b4h//vOf5/S764GmwfUJV7M45XriiSeGuLm5OdpuypQpIdYnGs1Y2SMX3bt3j37W6/KGG24IsabDzczmzZsXYn3y2yx+qsyvAFNI6sWnVK+99toQawrFLH76dfPmzdHYM888E+LHHnssxO+//37e+4TC+OOf6enXYqfoUqXfawcddFA0pqu++C4VuopKoZ+TuoKP/z7XJ1x9ujzXY6upev/Es6ZL/d997rnnhlg/3zQVaxZ/n1cKd+oAAAASwKQOAAAgAUzqAAAAElDRmjqtjdFHg7WzvFn2FSaOO+64EH/xi1+Mxt59990QP/300xlfQ+vctEbPzGzcuHEhbmpqyvjvfK3N8uXLQ6ztDY455phoO60N0MemzeK8vq8/evXVV0NczzV1BxxwQPTzqaeeGuLLLrssGjvhhBNCrC0ztIbOLG5j4lf5SLFVRbH5VRf0utHzv0+fPtF2v/71r0M8c+bMaEyvr3yOgV6jeq5cffXV0XZaL+vrabSb/KpVq6KxadOmhZjWQkjBqFGjQuxXStJaZP0OMotXc3niiScK+t1ai+prnZctWxbiwYMHR2OZPhO05tUsrs31K1CtXbs2xL4eW79/te7ef2dXg+rbIwAAAOSNSR0AAEACKpp+1RTYvffeG2JNoZnFned9akdTOL7LtC647hfpVpoC1TYiZvEj3NkeV9ZUr1l8K1pfI9fWJ55fsaJnz54FvU4K9G//xCc+EY195StfCbFvA6Ppfm2ZMX369Gg7XYyddGtudAUVXYXFLC5h0O3mzp0bbactBvwqAdloCsSndD/5yU+G+OCDDw7xGWecEW2nqVmfcl+wYEGIH3jggWiskLZJQDXTa8GXI+lnql5PZnEbE/2+9WlaTaPu2LEjGtN2ND6F29LSEuLevXtn/gPEK6+8Ev28YsWKEPvWSNqexH+GVdMqV53hTh0AAEACmNQBAAAkgEkdAABAAipaU6f5dF2iQ2tYzLIvW6L1cD4PfvbZZ4fYtxJRWpOjLVLMdl+6K5N99903+lkfex46dGhOr+GXOtH3Z82aNdHYokWLcnrNFPTt2zf6WevofA2k1lX6Wg5tmXHPPfeEeP369dF21NHl77vf/W7GMT2Pb7nllozb+WNZCN9aSNscaS2mX97n7bffDvHDDz8cjf3+978PsS5lZrZ7/R1Q64488sgQ++9UrQn338V/8zd/0+G/01YhZmYvvfRSiH1NnV5Pr732WjSmn+f+NXP9zNYl/7TFlVm8DJnWAZvFdYbbt28PsS6NZrb7508lcKcOAAAgAUzqAAAAElDR9OsHH3wQYu0erV3azeJHpX27kyFDhoTYt/nQztgal4L/3fqz3hr2XbL1EW7tpm0WP37tWyc88sgjhe9sDdhvv/1CfOaZZ0ZjV1xxRYjHjBkTja1cuTLEv/vd76IxXXlDzymk4+tf/3r0s34+rFu3LsSzZ8+OttM2C3fffXc0NmfOnBD7axRIjaYa/YoJ27ZtC7Fvs6VtRj7+8Y+H2Lfx8mVGSldi0pIss7jNUT6vmYkvtRo5cmTG19O/W1enuummm6LtqqEcgzt1AAAACWBSBwAAkAAmdQAAAAmoaE2d0lqVu+66KxqbOXNmiL/61a9GY1pv5Zct0bo2rdEyi3Pyvm4gE59n15+1PtAsXjZMH82+//77o+10OTH/ePdf/vKXEC9evDinfaxVfhkWbUFz+eWXZxzTGjqzuI7upz/9aTRGHV36/PJBWpejtbr/9V//FW3n29ogPYUu0VhvdAlFbfVjZvaxj30sxIMGDcrp9fz7vvfee4fYtwzr379/h7FZ3L6se/fu0ViuNXX6Pe2/93UOokuSmcWfIz/5yU9C7FuLVUPNLXfqAAAAEsCkDgAAIAFVk37NZsOGDSHWthRmcfuBo446Kho766yzQnzyySdHY7oChK5Yke0Wvb+1qo856z6amS1cuDDEmurRbtr+NQp5LLuW6XvtV424+uqrQzx69Oho7M033wyxpgrMzH784x+HuBTpVt3nbOeKTydn4tP29XYOFNv//M//RD/fdtttIdYWQXrdoT74a4t0bMd0RSf/ffWb3/wmxNr6JBv/Ph922GEh1tWBzHZfwUJpacURRxyR0+/2tOXIPvvsE41pyZRveTR16tQQ66oy/vO7GnCnDgAAIAFM6gAAABJQE+lXpak3M7ONGzeGWLvCm8WLtvfq1Ssa04XfdSHivfbaK+Pv9ilWXeVBn1Q1i2/L6qoRu3btyvj69UafgvJPNetiy/6Y6BOumg4wK/0Trnoe+SctNc0wfPjwEPsntfQpZxaIL65rr702+nnnzp0h5tqD0vSh746Av9LF683M1qxZE+K1a9cW9Jr6xOh9990XjWXrRNHU1BTicePGFfS7V69eHWItwTKLSzJ8twn9XtHPkWosl+FOHQAAQAKY1AEAACSASR0AAEACaq6mzuewNb/ta2a0nsav1vDoo4+GWOua8mlponl2VivIjdaxXHLJJSGeOHFitN1BBx0UYl+7oXWVvi5Ca/F05QkzsxEjRoRYV/nwrzFmzJgQ9+nTJxrT+r5s9Zc9evQIsW9voufp1q1bo7H58+eHeMaMGdHYnDlzQrxu3bqMv7ueaVsC1Af/2avXyaZNm0Lsa1v1Oh88eHA0ptch/l9ra2uXX0M///JZgWHVqlUhLrT2WOcEvn5P/zY/X6jG2rlMuFMHAACQACZ1AAAACai59Guh/G1jXbDXL96L0tEUyOGHHx7iF198MdpOO5n79Iou9HzuuedGY9omxXc815VDhgwZEmJdKNosbkej6Ruz3FeUyLVbvT8vNRXsV0jR85T0K/BXvqu/lmtMnz49xBdccEG03ebNm0PsW3eg+uhx3rJlSwX3pLpxpw4AACABTOoAAAASwKQOAAAgAXVTU4fqoLUQd999d4hr6ZHxcvG1QiwhBuzOf3ZoDe5tt90W4meffTbaTpeYfOGFF0q0d0B5cacOAAAgAUzqAAAAEkD6FWWlqRFd1QMAikFXLNCVIVglAvWAO3UAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAWpoA6LLJkydXehcAoO5xpw4AACABTOoAAAASwKQOAAAgAUzqAAAAEsCkDgAAIAE5Tera2tpKvR8ooUKOH8e8tnHM6w/HvP5wzOtPZ8cvp0ldS0tLUXYGlVHI8eOY1zaOef3hmNcfjnn96ez4NbTlMG1vbW215uZma2xstIaGhqLtHEqrra3NWlparKmpybp1yy/TzjGvTRzz+sMxrz8c8/qT6zHPaVIHAACA6saDEgAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACWBSBwAAkAAmdQAAAAlgUgcAAJAAJnUAAAAJYFIHAACQACZ1AAAACdgzl41aW1utubnZGhsbraGhodT7hCJpa2uzlpYWa2pqsm7d8pu/c8xrE8e8/nDM6w/HvP7kesxzmtQ1NzfbwIEDi7ZzKK+1a9fagAED8vo3HPPaxjGvPxzz+sMxrz+dHfOcJnWNjY0hbmhoyGt239bWVtTtkD89fqX8N6ic9muy/TrqyjHv1q1bh9d5a2trhzGqA9d5+op5naNy/GdrPvOfzo5fTpO69h1o/6AvxS3bYkzqCt2v1CaU/sIv5H3Rf9PZv0/t/atW2Y6DjrW1tXXpmGe6zrOdE3oO5DrGebO7fD7suc7rTzGv8/aY41oahX5GZnqdXK/znCZ1+os7O5G6MgPtqlzemI5+Tu2kLsXfs8cee0Q/652abHd0qpXus//bVLa/xb/PxX7ffd1Ejx49Qvz+++9HYzt37iza7921a5eZFf4/kjzqdgqX7b0r5vmWaSKvv8Ofj+3nSbXJNFHN9n7VyndAsT9bu/J3l+vcrFX5nG/ZauTyfS95+hUAACABTOoAAAASwKQOAAAgAXnV1LXzOd5qqZnJZz80h10LNWCV0l5n49+jbEWfnb1eR69RbrofH/rQhzJu52vVKllHpO/XBx980OFYMd/TfGpUc1Utx7+a5PpelvL9an/iOZ/P9koeS/3de+65Z8Yxvbb9tfvee++FuJrOxWq6RvKpn8/0ncr36+7yqbfL97OWO3UAAAAJYFIHAACQgLzTr531rypFv7lipIH8a/j0FTqWSxubfI55pdMJ7TQl8O6770Zj1ZJy8mmLakoXlTJ9UG8qfSzNqrc9SSb6nvn2Ppmu0WwlJNWkkqnsfPj91HOoWt/balXMdjvcqQMAAEgAkzoAAIAEMKkDAABIQFFamhSaD860nIt/RD3bMk7Z6gG01iLbY9U8cp1dvjUXtbB8jO7j4MGDo7EzzzwzxL169YrGVq9eHeJFixZFY8uXLw9xKWqUqFlJQy1cH7Vk7733jn4eMmRIiE8//fQQP/nkk9F2K1euDPG2bdtKs3M1IJ9zrlrqY/2yWn7O0C6fur9Urj3u1AEAACSASR0AAEACipJ+LQa9rdu9e/do7OCDDw5x7969o7F99tknxD6N+vbbb4d406ZN0Zhuu3HjxhBr2wiz+G9N5fZsMeS6okS1vmcf/vCHQ3zllVdGYxdeeGGI9fwyM3vnnXdCPG3atGjspptuCvFbb71VlP1U1fpedoZrqHrSVl1RjL+hWMdf0299+vSJxi6//PIQX3TRRSG+9dZbo+3uuOOOEGsqtlC1+DnYFaWeB/iVfvQze/jw4dHYyJEjQ6wtqvwx0e96v/+axvcrCb3++ushXrNmTYi3b98ebbd169YQV6qsizt1AAAACWBSBwAAkAAmdQAAAAkoqKYun2W8sslUl3XQQQdF202cODHEH/nIR6Kxnj17Znx9rWtasmRJNKY1dtOnTw9xc3NztJ3m1vPJkafeJqXUNULFWI7Lt8LRmoxvfetbIT7//POj7Q444IAQ+0fndWzChAnR2Jw5c0L8hz/8oYA9RqXo+bbXXntFYz169AhxtnZL+pniW9ro50GK9VXVJFMN9qGHHhpt549ztSjnUoSdKXctqNa1+VZTn/3sZ0P8+c9/PhrTNjb6me3fP18zn0lLS0v085tvvhniJ554IsSzZ8+OtnvkkUdCvGPHjpx+V7Fxpw4AACABTOoAAAASkHf6taPbwYXeLtZ/p7dMfYr1nHPOCXFTU1M0pu1PfMpTb6HqLXkzs+effz7Eixcv7vDfmMUtLPzr69/q0y20cegaTY34R9uzPfY+dOjQEP/d3/1dNKa37/v16xdi387g5ptvDrFPv2qqdtCgQdHYpZdeGuI//vGPIS7W8ec8Kh6fVtpvv/1CfPzxx0djp556aog1hW9mtnDhwhDffffdId6yZUu0XQrHLtfVgfIZy/X1Pf0s1hYWZmbr16/v8Hf7cgzfOqtaVPpcKWdrFl8+dcopp4T4m9/8ZjR2zDHHhFhLIsziY5tt/xsbG0OcbbWJfffdNxrTz4cDDzwwxGPHjo2203mFtj4xK19JFnfqAAAAEsCkDgAAIAEFPf3qFeOWrKa5fKpUn2Dyt8x9+k3pk2rDhg3L+Jr6ROOzzz4bbaeLtGs3arP4Nr9/albTtqrUXbhL9TtKzf8Nmh7VJ5vMzPr37x/iAQMGRGN6XP3qI/p04ty5c0N8/fXXR9vpcfa34VetWhXi6667Lhrr1atXiPV89ql5VIamaHShdzOz733veyHW7vRm8RN5Ph2vaVZ9an/KlCnRdr5DfWr8Z46+T7mmZrOVt2Tjn2LNVDKjKwx19Pu6yu9voU+OVvrp11L/Tr2edMUPM7NJkyaFWEtpzHZ/8lxlKnfyx1hT9f47Wq9ln9LX19T5gn5PmcXfAT79Wi7cqQMAAEgAkzoAAIAEMKkDAABIQFFq6opB61300WWzuI7O59VzrT/wuXWtxRszZkyIfcsUXSXgsccei8Y0B79t27asv6+Uqq2Grtg1lroyiJnZK6+8EmJfF6H/TmsfzMwGDhwYYq2VfOONNzK+hj7KbhafH/vvv380pquW1PqKIg0NDVV3XuXC11dpC5JvfOMbIfbtbvSY6zVvFl/bo0ePjsa0/vfb3/52iF944YVou1mzZnX4eqkq5/n//vvvRz9r7ayO+XOj1HWOxbh+UqiX9rTW+bjjjovGtEbN/+0ffPBBiP0x37p1a4hffvnlEPsVH1566aWM+6Wf5371kVGjRoVYv1f8ChXVUDvLnToAAIAEMKkDAABIQEXTr3p7VdOq++yzT7Rdtlv5ejtab8+amW3evDnEy5Yti8b0kWVtkeHTu3pbV1uYmMWPyKfctqKYq4jkStuK6GLKZvEtbn/Mdb/86iBr164Nsd429+eXlgL4FK6uLqDdyc12b4dTy9ra2rKmfsrZdb4z2l1ej4+Z2de+9rUQjxgxIsSaojMzmzZtWoinT58ejW3fvj3E48ePj8ZuuOGGEPfp0yfEn/rUp6LtnnzyyRDXQ/q1nPxnQKYWNPo577dD+ehnqm87pd+jet2Zmb3++ushfu6556Kxhx9+OMRa+uBLa/Rc8atSaGlN3759ozFtVaJpWr9Kia42UWhLm67iTh0AAEACmNQBAAAkgEkdAABAAqqmpYnmt3W5HbM4b+3z1FoPpS1GzMzmz58f4gcffDAa09fRx6r9MkBai7du3bporBoeXzarzGPvxf4d/vV8PVwhr1Po8dF2N2PHjo3GTjjhhBD7ffzLX/7S4X7Uqnz+hlL/vXqO+1rGc845J8TaVsQsXmJOayp/9atfRdvdeeedIfZtCrStjbatMYvPAf3cylbrieLy5562sdE2Jr49kW9xUkoptiYplNZIz5w5MxrTViW+XdXTTz8dYl3m0cxs9erVIdbr17/veg5onZyZ2ZFHHhnij370o9GY1mPqeePr//XzRvfXrHx199ypAwAASACTOgAAgARUNP2qqU7tJO27OevtTn/bWh97XrBgQTQ2derUEP/5z3/OuB8rVqwIsb8lu2jRohBre5OO9gW1yafcBw8eHOKJEydGY3r7/q677orGsnUrT005061mcUrtyiuvjMa++tWvhthfvxs2bAjxjTfeGOJ77rkn2m7Hjh0h9ueDfv4MGzYsGtOWDJpyfeqpp6LtfEoXxeNTYJoq09Idn37VlYrKnR6t53S8tgJ74IEHojFt/ePbDmmLMr1ezeIyLD2u2mLEzOzYY48NsU+xjhw5MsS6+pBZ/BmgaVSfUtXzrdwrmLTjTh0AAEACmNQBAAAkgEkdAABAAipaU6fLtBx11FEhHjBgQLSdLt3ll3RavHhxiH/3u99FY/fff3+IfbsTpcuP+GXC9BFr/7u19sbXSGSqA/TbaR2Oz89nWx5NUdvXNX65mC9/+cshPvroo6Mxrb/051uhbVjwV3rtaWsAM7MrrrgixJ/73OeiMW1B42vZfvSjH4X48ccfD7GvyVFak2MWnwMXX3xxNKZ1WrokETV15eOPly73pDVOfumnStbU1fNnttaWrVq1KhrLVmuoY3rNm8XthE488cQQawsqs7htidbpmsXzkWxLleoyf1rnZ2a2dOnSEOvcoZy4UwcAAJAAJnUAAAAJqGj6VW+h6qoOud76NDObPXt2iH2Xab01mi2VqWlP/9hxtlu+mrbzt3L1Nq+mfv3t2ldffTXEvoO23qKv59v1paCpvvHjx0dj559/foj9+aArEfgWJsU+RoW0Pail88T/faNHjw7xP/7jP0ZjJ510Uoh96vRPf/pTiH/+859HY8uXLw9xrumQnj17Rj+ffPLJIR4+fHg0pp8dU6ZMCbGWdJjtvsIEukbPHZ9+1Wtbt/NlFppuK3WLkVq6LsvJvy/Z3idNpftr9JprrgnxhRdeGGK/+oy+hi930s96/xnz7rvvhlhbm82aNSvaTstzKnXNc6cOAAAgAUzqAAAAEsCkDgAAIAEVranT/LbWO/hlejT37XPdr732Woh1+RGz3FuCKJ/T1330S46ceeaZIdalisziZUZ0n5ubm6Pt7r333hA/9NBD0Zjm5zWn39F+onNaN6NLxNxyyy3Rdvvuu2+Ip02bFo3NmDEjxNna5BQq10f6vVo8H7Smyczs3//930M8dOjQaEyvhZtvvjka02Pi61Jz/QzIthTYqFGjQqznhpnZW2+9FWJdNs7X79Xi8akVvnZJ6671ffe1dyrbdw4qQ797zcz69OkT4n/6p3+Kxr70pS+FWOcS/jOz0NpJPY/03Fi5cmW0nY4VMv8oBu7UAQAAJIBJHQAAQAIqmn5V2pnd3yLVW+OaJjEz6927d4h9OkdfJ9f0h78Nf/DBB4f4qquuisYmTpwYYp+azXSbd9CgQdHPxx57bIi/+c1vRmM33XRTiO+4445oTFcvqLbUTrWkCf2x7NevX4inTp0aYu1Gbmb2wgsvhPjOO++MxlavXh3iUvwt2V4z28oktUL3++qrr47Gxo4dG+Jnn302Gps8eXLGMS1vyPWY+PdPr99x48ZFY4cffniIfYsbTf2++eabee8HCqPv76ZNm6KxefPmhVhLZPz3g173PtVXqdUAqkExSkCyfVZluzb0+33w4MHR2KRJk0LsV3bRsgh9fZ8C1X3JlprNNqYtc3ybnEqlXBV36gAAABLApA4AACABFU2/6hOdL774Yog/85nPRNvpLU2fUvvIRz4S4gEDBkRjGzZsCLFfUFtv0ertVP/k2z/8wz+E+JxzzonGDjjggBBnuy2d7Zas/j29evWKxvQWs648YWb28MMPh9inCtr3pa2treJpoHL/fr0Nr0+4mpn98Ic/DHH//v1D7FeG0Kcrn3766Whs+/btRdnPeqLXxv777x/iSy65JNpu/vz5Ib7++uujsQULFoQ4n07t+ruzXWtnn312iM8777xoTEswNm7cGI1pqh7Vwa/m0c6nWDXV58fqQbdu3Tr83tLrxD8xrCsn+RWW9ElwfQo92/ef//26mtSIESOiMX0K3c8D9DtQyzF8RwxdYcZ3otASjDFjxkRjffv2DbHOF5YsWRJtVw3nEXfqAAAAEsCkDgAAIAFM6gAAABJQ0Zo6rXO75557QuxbHWju3uf4tW7q9NNPj8b0UXddecIsrvXSlgX+d48fPz7EPXv2jMZybZmSbSzTo9Jmcb3ghAkTojGtP9Lawc5+XymUu72G/j6twTAzO/XUU0N87bXXRmNao7Fu3boQ+xUlpk+fHuJSrBpRDP4YV3OLE61/OeSQQ0Ksnf/NzKZMmRJirbE1y6+OTmkbi6amphCfe+650XaXXXZZiLWGzu+nfk6ZmT3wwAMhZhWCyvDXgh4vPSa+DkvPKT9WDxoaGsJ/Sr9jteWWmdkZZ5wRYv99NXPmzBBr2yG/you+7/41tM2M1qybma1duzbE2gLNLK7b0zq3RYsWRdtp/bQ/b/SzybdO01pgbWPiv3tpaQIAAICiYFIHAACQgIqmX/XWuD5qfN9990XbaVfwxsbGaGzgwIEh9q1Q9PHrOXPmRGP6CPQVV1wRYu1qbxanXP2t4mwK6a7tbwfr7/NpRn0ffEf1cnRDr+SqEXpMND1uZnbdddeF+NBDD43G9Fb5f//3f4fYp9Q0VV8Nt9NzUenWNdloKkOvyfvvvz/abuHChSHWtgSd0XPRrxowcuTIEF966aUh9ueNpnp82xJNsf7yl7+Mxkq9wgg65993bZWV7Zho+4l6TL+2tzTxf7t+1/jWHieddFLG19PvKz0GmzdvjrbLln71KVelqdTFixdHY9pOZenSpSHWVV7879bPIrN4LuH3S/+dlo0tW7Ys2q4aPgPq70wGAABIEJM6AACABDCpAwAASEBFa+o0/6w5+B//+MfRdvr4srasMIvz/9qaxCyuvdL2IGbxY9vHHHNMiHWZKbPsy37kmj/Xuiz/b7T+zS9lpvUAviZw69atHb6+/o5qyO8Xg68nPO2000L8ta99LRrTZd780mq/+MUvQqx1dG+88Ua0Xa3U0VWrbHWjuqyWtigwi2tv8jl3teWRXxru61//eoh1mSG/j9qC4ZFHHonGHnzwwRCvWrUqGtu5c2fO+4ndFaM212+n9cbZXr/elwnbtWtXh++Pvp++7ZB+P+rSWf5nrVfT7yqzuMWJ/87LVreu7aVaWlqiMV2+UV9DW6SYxXMJPyc466yzQuyXC9XzQ7+XdflCs3geUyncqQMAAEgAkzoAAIAEVDT9qjTl9dxzz0Vj3/nOd0L8r//6r9GYpmN96rRfv34h7tOnTzSmt941feNvR+faqd+nALRdi94a9t21V65cGeKXX345GtNbu08//XQ0prefK9HJvtSpXb2Fftxxx0Vj11xzTYiPPPLIaEzTYz/84Q+jMW1NoWmFVNLU1UrPz7fffjvEfrWObC0m9DrU69XM7Oijjw6xplvN4o74mjaZNm1atN3s2bNDrO2V/H4WurIFSseXS2T7PFe9evUKsW+FUw/aP/f8+6fXq34/mcXfzSeccEI0pu/7EUccEWK/+oOW0/jrXH+3T9tqSyqfttXvQ10tyJfuDB48OMRDhw6Nxg488MAQ+88Y/Qz47W9/G2K/UlU52ol1hjt1AAAACWBSBwAAkAAmdQAAAAmompo65fPSWmv23e9+NxqbNGlSiE855ZRoTB9n9nl9ranLtW7O01osX5egjzbPnTs3xPPmzYu209oen5/XWjxfQ6CtFEpdE9bZ+1OK3z9o0KAQf+tb34rGjjrqqBA3NzdHYz/60Y9C/NBDD0Vj1NFVRqaWPv6a0foXX+OkP48bNy4au+yyy0Lsl4bTtjZTpkwJ8VNPPRVtp7WuvkaVc6V0ivHe+jpHranKtiSVfif4eux64q9D/ZzUpbnM4vZPM2bMiMa0hZh+vx5//PHRdieffHKIta7RLG41NmDAgGhM66ezLeum168/N3JtL6b1e2ZxC6z/+I//CLH/Xq4G3KkDAABIAJM6AACABFRl+tXT26IvvvhiNPaNb3wjxHr71yxuZ3DttddGY5rC0VvF2fjb1Joe9e1Ifv/734d46tSpIX7rrbei7TRNm20lA3+ruNCUcSHa2tpK/vt8akSP64knnhiNaWuKX//619HYww8/HGLfdZw0Wnnk+j77c6p///4h9u0GNIUzduzYaExTNk888UQ0dvPNN4d46dKlIfZpGc6N2uWPpabc9fPVt6nQ8ybX74Byy7Y6S1dlei0t7dm4cWM0tmnTphAvWbIkp9e75ZZbop811a1lNmZmF1xwQYjPO++8aEzbkfhWJdoOSb9L/Pun54Om6c3ikow77rgjGtOyKf3er0bcqQMAAEgAkzoAAIAE1ET6VflbvDt27OgwNjObPn16iGfNmhWNXX755SHWJ2j9a+jt52eeeSYa0xSrTwvrLdpSrPhQ7nRRKX6fPsHku5N/5StfCbG/1a4Lwft0tt569yldTW/r7/ZPUul2PiWut/P9IuD6OpoS8u9dtidAdSyVlKC+Z9lWhpgwYUKI/ZPs+l6//vrr0di9994b4j/84Q/RmE8flVIxnqJH/vx1qNe9Xof+Oh8+fHiI/YpDL730UjF3sWjlK+2vU6xzJt/XKeT3+n+jT5r7FO73vve9EOtTpmZx+tU/AX/SSSeFWM8HXWHIzGzZsmUh9qtlvPLKKyHWlW/MSr+STLbzI9/3nDt1AAAACWBSBwAAkAAmdQAAAAmouZq6fGgu2tfK3X777SGeM2dOiDXfbxbn0jdv3hyNbd26tcPtzLK3Jym2Uj72nul3FPr7tN7hkEMOCbEeD7O427s+hm4Wv9e+E/yQIUM6jM3i7uXaFmPYsGHRdvr3NDY2RmPadsN3Q8+0Skm288bX7jz22GMhfv7556MxbdXgO5m373O112dpHZ2vY7roootCrO0mzMyWL18e4vnz50djjz/+eIh9LUw5ZXvvi1FTVY7rvFDFrAnKl1+BSFc90JUBdIUhs7imTuu1zOK/p5re59Tp96b/3NfPgDVr1kRj9913X4j12PnvYf1ZW7f431eM7+9KXa/cqQMAAEgAkzoAAIAE5J1+7UoaoZK3sf3tVE2BLViwIMTZ2k9Uk2zHodiPvRdzNQlNv2qLAZ8208fLdQUJs7h9jKZazOLUrE+dakp3//337/D/bxa3RPCvof9OY7P479E0o1+cvnfv3iH27Vq0TMCnWLV9i0/3F/uR+4aGhoLPH3++6Pui77VPe2t6zJdBaLrFt0HQFiflLHvwyrnKSzXI5TOoM8X4jPLHXMsdHn300RAffPDBGbcr9yoB+ndX+ryp1rR+ttSpT7n7z4tqVK7jzJ06AACABDCpAwAASACTOgAAgATkXVNXLfn2YqpkHU6uKl134RV6HuiSaVoP9/3vfz/aTluV+HY0zc3NIV63bl00pnVofvkgrW3TOizfmkT51+jRo0eIfdsNrfPQ2jFfL6g1dX4JucWLF4fYP7avj9yX8jpsaGjoUk1dR6/XTuvr/Pun78Xq1aujMV2iT5f6MauNeppiSOGztxT1W/41tFXFrbfeGmJtCWQWf3bMmzev6PuV6+v5sXK3h0nhvKo12a6Drh4P7tQBAAAkgEkdAABAApJeUaLWFJqaqEQqrtCO65pi27JlS4inTZuW8d+U4u/zqc1MSpH21hRkNvmkbIr5HrUf83z/TS40Ra0pcLN4BY2pU6dGY7Nnzw6xb2NTLW1MKrmqQq0ox/ugnzHa/mbp0qUZ/001HZ9y7ktH52w1vRe1rFJpde7UAQAAJIBJHQAAQAKY1AEAACSAmroqUqzH3oupGuorqnUZm0JVewudTDV12eooc132SJcz02XPzMymTJkS4vnz50djulSctqzoaF9KqdpaC1VSLVyH+XymVoty1mZW63uQulK+79ypAwAASACTOgAAgASQfnWypfqKkQbM5zVyTfVk2q5Yt3grcYs+W6qv0HYqmV6j1Ar9Xfn8be2/oxrTKbpP2tJk06ZN0Xa6aoRfhUPTtpX8G3NtR+NV43EppWr8e6tpn4rxGYbaVYrvtHbcqQMAAEgAkzoAAIAEMKkDAABIADV1TqkfgS9GDV1n2ut+2traitI+o5jLhGWSz2uU8/fl016gkBrIfM6HbPWdWlNX6ZYp/m/SejhdwmnHjh3Rdi0tLQW9fjn59zbTMmHlqNnU67wY70lH13k5pda6SGU7H/I5V0pxzLOp1FJXqSv0sz2X95g7dQAAAAnI6U4dM/DyKPbTqv7/FrovHf37Up8T1XLOlfoOYrFfvxjHPNO/Lfcd2WpU6P6X4u8ux3VeTpX+/aVUqqbB5T7mKR+jUsv1vevs87ez18npTl2uaRFUh9bWVmttbQ0Hv5DjxzEvnvbb5u3/tR+f9mOU63+5vGYxjnlra2uUJkVmmY7rrl27ov9K+bvbcZ2nj2Nef/I95g1tOUwfW1tbrbm52RobG1kmp4a0tbVZS0uLNTU15d1fi2Nemzjm9YdjXn845vUn12Oe06QOAAAA1Y0HJQAAABLApA4AACABTOoAAAASwKQOAAAgAUzqAAAAEsCkDgAAIAFM6gAAABLwfxoYFCIPLUwNAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 32/100, d_loss=0.029, g_loss=0.848                                                                                                                      \n",
      "epoch = 33/100, d_loss=0.023, g_loss=0.821                                                                                                                      \n",
      "epoch = 34/100, d_loss=0.022, g_loss=0.837                                                                                                                      \n",
      "epoch = 35/100, d_loss=0.031, g_loss=0.817                                                                                                                       \n",
      "epoch = 36/100, d_loss=0.023, g_loss=0.850                                                                                                                      \n",
      "epoch = 37/100, d_loss=0.023, g_loss=0.892                                                                                                                      \n",
      "epoch = 38/100, d_loss=0.033, g_loss=0.724                                                                                                                      \n",
      "epoch = 39/100, d_loss=0.021, g_loss=0.970                                                                                                                      \n",
      "epoch = 40/100, d_loss=0.020, g_loss=0.775                                                                                                                      \n",
      "epoch = 41/100, d_loss=0.018, g_loss=0.800                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 42/100, d_loss=0.018, g_loss=0.813                                                                                                                      \n",
      "epoch = 43/100, d_loss=0.021, g_loss=0.931                                                                                                                      \n",
      "epoch = 44/100, d_loss=0.016, g_loss=0.781                                                                                                                      \n",
      "epoch = 45/100, d_loss=0.022, g_loss=0.798                                                                                                                      \n",
      "epoch = 46/100, d_loss=0.015, g_loss=0.847                                                                                                                      \n",
      "epoch = 47/100, d_loss=0.011, g_loss=0.871                                                                                                                      \n",
      "epoch = 48/100, d_loss=0.017, g_loss=0.856                                                                                                                      \n",
      "epoch = 49/100, d_loss=0.019, g_loss=0.823                                                                                                                      \n",
      "epoch = 50/100, d_loss=0.016, g_loss=0.756                                                                                                                      \n",
      "epoch = 51/100, d_loss=0.016, g_loss=0.847                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 52/100, d_loss=0.013, g_loss=0.781                                                                                                                      \n",
      "epoch = 53/100, d_loss=0.096, g_loss=0.546                                                                                                                       \n",
      "epoch = 54/100, d_loss=0.017, g_loss=0.811                                                                                                                      \n",
      "epoch = 55/100, d_loss=0.027, g_loss=0.859                                                                                                                       \n",
      "epoch = 56/100, d_loss=0.012, g_loss=0.772                                                                                                                      \n",
      "epoch = 57/100, d_loss=0.012, g_loss=0.818                                                                                                                      \n",
      "epoch = 58/100, d_loss=0.017, g_loss=0.827                                                                                                                      \n",
      "epoch = 59/100, d_loss=0.013, g_loss=0.784                                                                                                                      \n",
      "epoch = 60/100, d_loss=0.012, g_loss=0.903                                                                                                                      \n",
      "epoch = 61/100, d_loss=0.012, g_loss=0.878                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 62/100, d_loss=0.016, g_loss=0.818                                                                                                                      \n",
      "epoch = 63/100, d_loss=0.012, g_loss=0.844                                                                                                                      \n",
      "epoch = 64/100, d_loss=0.013, g_loss=0.764                                                                                                                      \n",
      "epoch = 65/100, d_loss=0.014, g_loss=0.857                                                                                                                      \n",
      "epoch = 66/100, d_loss=0.012, g_loss=0.786                                                                                                                      \n",
      "epoch = 67/100, d_loss=0.012, g_loss=0.823                                                                                                                      \n",
      "epoch = 68/100, d_loss=0.012, g_loss=0.841                                                                                                                      \n",
      "epoch = 69/100, d_loss=0.009, g_loss=0.833                                                                                                                      \n",
      "epoch = 70/100, d_loss=0.010, g_loss=0.825                                                                                                                      \n",
      "epoch = 71/100, d_loss=0.011, g_loss=0.762                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnUAAAFBCAYAAAAR9FlyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA2g0lEQVR4nO3de7BV5XnH8efgBUQOFwEFwlVAbmpR8X4jGJGoSbQ1NYnWtJ0kk6Yz6XTazNh0mthpO3U6aZ3WptGaRGLsRTCIUaKYREAqooIoCspNbnJAhIAekZvh9I8Mb37v49nr7L3Zl7Xf/f3MZObZeRd7r7PftdZ+Xe+znrelo6OjwwAAANDQutV7BwAAAHDsGNQBAAAkgEEdAABAAhjUAQAAJIBBHQAAQAIY1AEAACSAQR0AAEACGNQBAAAk4PhiNjpy5Ii1tbVZa2urtbS0VHufUCEdHR3W3t5uQ4YMsW7dShu/0+eNiT5vPvR586HPm0+xfV7UoK6trc2GDRtWsZ1DbW3dutWGDh1a0r+hzxsbfd586PPmQ583n676vKhBXWtra8V2CLVXTv/luc9vv/32eu+CmZndeeed9d6FglLrc3SNPm8+9Hnz6ar/ihrUcYu2sZXTf3nu8x49etR7F3IvtT5H1+jz5kOfN5+u+o8HJQAAABLAoA4AACABDOoAAAASwKAOAAAgAQzqAAAAEsCgDgAAIAEM6gAAABJQVJ06AKinyZMnh3jq1KlR24gRI0J88ODBqG39+vUhfuCBB0J86NChyu4gcu+zn/1s9Prss88O8UknnRS1LV++PMRPPfVU1LZ79+4q7B1QGdypAwAASACDOgAAgAQwqAMAAEhA0+TUdesWj19PPvnkELe3t3f6/5uZjR07NsSTJk2K2gYOHBjiAwcORG1bt24N8YoVK0K8ffv2aLuOjo4u9x35NGXKlOj1smXL6rQn6dM8uj/7sz+L2jSn7v3334/aFi5cGOL/+Z//CTE5dY3ljjvuqPh7/vrXvw6xP27GjRvXaVytfQEqhTt1AAAACWBQBwAAkICkp1/1MfXTTz89ajv//PNDrGUPhg0bFm13ySWXhHjGjBlR25gxY0Lsb9+vXLkyxLNmzQrx7Nmzo+3efvvtEOt0APLvi1/8YvT6nXfeCfHmzZtrvTtJ0ylWjc3Mjhw5EmKfzkB6A4DOnHbaadHrUaNGhfjw4cNRm5a4yTvu1AEAACSAQR0AAEACGNQBAAAkIOmcuo997GMh/sM//MOo7atf/WqIV69eHeL+/ftH2+3cuTPEfnmYlpaWEPfu3Ttq0yVoNJ/Pl1J49NFHQ6z5dWbk2OXdJz/5yei1lrG56667QuzzM9C1E044IXp93HHHFfXvNL/O7KPLhgFoTPUsJfOpT30qF/tRDO7UAQAAJIBBHQAAQAKSnn4dNGhQiD/xiU9EbT179gzxmWeeGeJ58+ZF2333u98NsZY+MTM744wzQuyn4qZNmxbic845J8T/9E//FG2nq1n4z96zZ4/B7Pjj83mY+vI3N9xwQ4h19YK33nor2o4yG13T1Akzs379+hX17/x065o1a0JMOgNQf7q6k7+26znK+Voe7tQBAAAkgEEdAABAAvI5r1Um/4ScTrF27949ajtw4ECI77777hDfc8890XY6deZvB+vTqhs2bIja2traQnzuued2uk9m8dTvkiVLojamX3/j8ssvr/cudMo/oalThBrrsWDGtEIx/Oot+jR5Fl3Vw8zsRz/6UYj9k+fIN3/NRhrGjx8f4gsvvDBqe/XVV0O8bNmymu1TSrhTBwAAkAAGdQAAAAlgUAcAAJCApHLqfBkEna8/5ZRTora5c+eGWPPofPmJDz/8sODnaZtfbcLn2B2lq1CYxY90+zb8xrXXXhu93rdvX532JPbwww9Hrz/96U+H+Ic//GGIb7zxxmg7zbEjv65zJ554YvRa8xf9eaIlEnzO6rhx40K8efPmEPuVJ5A/XA/TdMstt4RYV2owi8/zmTNn1mqXksKdOgAAgAQwqAMAAEhAw0+/6tTLWWedFbX5VSSUrg6xY8eOEGdNt2bx02gsJF45jzzySPR6yJAhId6/f3/UptNqvXr1itpWr14dYp3C9SVHskpfaAmar33ta1GblmDQab9LL7002u5nP/tZiN97772Cn9VsNH1i0qRJUduAAQNC7Ffk0Ne+Tc9nVvJoLEOHDq33LqAKNm3aFGK/ooReN8eOHRu1+dQodI47dQAAAAlgUAcAAJAABnUAAAAJSCqnbvjw4VGb5uXs2rUratPlSOpZVuKDDz4Icbn5fKnzy8Von/u8m2uuuSbEV155ZdS2atWqEM+fPz/EutxbpeiSdT7Xc8GCBSEmp+63+vfvH+IRI0ZEbX369Cn47zRXzudDbtu2rdPtPC2h4ssf6XnpryOonMGDB0evs3Ki0bi2b98eYp8TrWVsyKkrD3fqAAAAEsCgDgAAIAENP/2q0zQTJ06M2rS8xfPPPx+1vfDCCyGu5/TrmjVrQvzuu+/WbT/yLKvEiJ/a/NKXvlSwbdq0aSHWqbg333wz2s5PCaA2dCraT4lrCZq+fftGbXreaFqFWTzV07t37xDrsWBmduqpp4Z44MCBUdvhw4dDrFP4ZmYrVqwI8datWw2l0RUEpk6dGrXdcMMNIV66dGmN9qhr48ePD7FOCb7//vv12J2Go+eyph+Zxb/FrLZTHu7UAQAAJIBBHQAAQAIY1AEAACSg4XLqevToEb2+4oorQnzxxRdHbVrOQJdmMjPbsmVLRferZ8+e0Wv/eP5RmudnZvbGG2+EmJy6zmm5CTOzM844I8QzZsyI2iZPnhxiX8KitbU1xCNHjiz4/lm0VInmA3maD7Jx48aojSXkOqf5ST437sILLwyxz6l75513Quxzr7REwrXXXhvib3zjG9F2uvSc71c9jnx5ne985zshJqeudLo03PXXXx+1XXTRRSHOU07dV7/61RC//PLLIfb9r/mWe/fujdr870AzaW9vD7HPQ9T81bVr19Zsn1LCnToAAIAEMKgDAABIQMNNvw4aNCh6ffnll4fYV6BetGhRiJcvX17V/dKSCH6/dPpGp4rM4tvN6NyoUaOi13/xF38R4ptvvrno99HSKL/61a9CXMpKHjqFO2DAgKL+ze7du6PXPKrfOe0HP0Wd9Z1ltWlZm7/+678O8ejRo6Ptjj++uEuhTheaxcempmD4Ug34DV0Nxixe9eXss8+O2rQEzbe+9a2oTVMavv/970dtd955Z4j12utXn9Ep/S9/+csF9yvr2Dhw4ECId+zYEbX9/d//fYj1t8hvq+/RDPS6efLJJ0dtmt6S1ylqXXHGX3t0arle+8+dOgAAgAQwqAMAAEgAgzoAAIAENFxOnc+LOO2000K8c+fOqG3JkiUhXr9+fVX3y+cGaEkTzeV69NFHo+0038qX4Ghmmlvx2c9+NmrTMja+xI1+hz5fUUtmzJkzJ8S+3ECWYcOGhdgvQ1bI8OHDo9ellFBpVlqKxL/2bZqn1adPn6hNy5ho/pvvA81/8Xky+nkTJkyI2v7yL/8yxK+//nqI9dpjVlreZsr8dVKXAvPniX7vPj9Jl/b7yU9+ErXpNUDz4b797W9H2914440h7t+/f9Smx4BfNrBQvqSWSTIzu//++0N87733Rm3f/e53Q/zaa691+n6p+uQnPxlinwev5YR8/mVefOELXwixz5d+4oknQlzK70ol5fNbAwAAQEkY1AEAACSg4aZfzzvvvOi13rLX6t5mZk8++WTB97njjjsquVsf8eKLL3Yae++9915V96NRnXnmmSGeNm1a1KbTHH4qTvnp+H/9138NsVZ7r3ZJgT179kSvmYrrmk9F0Ne+TafqfXqGHjs6Feen0HRqfsOGDVHbwIEDQ3zuuedGbTrdq9N5/pxv5j7Xc/Skk06K2nRKvFevXlFbVikoXYlHyxOZxWWHtGyJrlBhFq9M4qd3dSp91qxZUdtPf/rTEOsxpdOtZvHKN+PHj4/a9DhdtWpV1JZ6Gs4FF1wQYj23PF8+6Omnnw6xpsGYmf3Jn/xJUZ+tx5hP1dBUHt8Hev7qKhj6O2Jm9vzzz4eY6VcAAACUjUEdAABAAhpi+lWn26ZOnRq16fTrL37xi6hNb9HnVbNVEy/WTTfdFOIxY8ZEbTqd42+T79u3L8T6hJxZfDwUu5KHVg83i49FffLa09v1CxcujNq06jg6558I1MXS/ROoOpX1mc98JmrTqTg9bvwC8XfffXeI/ROat912W4j98abvyVPNndPvSFeJMIufdvSpFHoOPfvss1Hbgw8+GGI/dXr99deH+Jvf/GaI/XVEn670TzHOmzcvxP7JVZ261/fs3r17tJ3+PT5tSI/hZ555puD7NyrtV30C3Sy+bmrqhKerepjFK474cy1rGldpn2Q9XZv1hL2udvP4449H27377rtF7Uc1cacOAAAgAQzqAAAAEsCgDgAAIAG5zKnz89maJzNx4sSoTR919zlUjZCv1sylDpTPR9FK4z4PR48PX/1fywPcddddUdvKlSsL/rtCrrnmmui1VsDv2bNn1KbHm+bv+ZImPgcIH+XLE23ZsqWof1dsOQhftqStrS3EV111VdTmy1GgNJpfdfXVV0dt/txWem33x4PmXPrzUPNeTz/99BD7a4yWk9IVZszMZs+eHeJdu3ZFbUOGDAmx/jadeuqp0XaaL+Zz9rTcRbHXokaVVXYqi+bDdva60vS6rCWOzMzmz58f4qeeeirEL7zwQrQdOXUAAACoCAZ1AAAACcjN9Kveoh08eHDUpgsA+9v1P/vZz0Ks1ZwbRerVw7PoY+KDBg2K2rRUja9Cr/yqEbqQulYgNyt+9Q6tNK4V6c3ilS78tEKh6ddGm165/fbbo+rq9eBXfNFSNZ5Ojz333HNRm07Z6OoSCxYsiLbTVQn8wvL+2ERp9DzRVRzM4hUZvP3794fYpzDo8eDLWVx22WVFvb+m6/gSOpoW40tr6HH0e7/3eyH2K2LoqhT6O2UWr0Rw6NChgvvYqPSa51dd0BQZf23X7z3rnPepVXq9XbduXYinT58ebafpW76/dJ99ObT/+I//6LQtjyle3KkDAABIAIM6AACABDCoAwAASEBucup02Y/Pf/7zUdtZZ50VYr/EkuZQ6Vw9GkvW4+tZeTGvvPJK9PrRRx8NcbE5dH4psOuuuy7EF1xwQdSm5RP8UmOa21VO+RSUR0tm+GWHXnzxxRBrXpZeN8zi3J5+/fpFbVnHn+YA7dixI8TNnCvr6Xfk86uylsTS5dpGjBgRtWmZGV0mzizOe81ahkr7/NJLL43adBkvLWFiZnbuueeGWHMEfQ7vk08+WbBt/fr1IU7xWNHyIJs3b47afvjDH4bY/2ZrHp2WtPGKzanzZWa03I3PqdPPW758edTmS+rkGXfqAAAAEsCgDgAAIAG5nH695ZZborb+/fuH2E+36WPpWY9AI3902kFLSpjFU2V+mlMrw/tHzxctWlTUZ+ut9yuuuCJq+8pXvhLi8847r+B7+DILWk5j8eLFIWbVkGOnU9h+Ovu0004LsS8/oZXh77777hD7cjT670aPHh216bbvv/9+1LZ69eoQ0+ed0/7yZad0as6XstJpcN+vOiXuV4rQ1I2s1QzGjRvXaWwWTx/63xVd3eTZZ58N8Q9+8INou5deeinEuoKEf/9m88gjj4TYlyDSMjaVWJ3BT9NmTXXrWELTNhoNd+oAAAASwKAOAAAgAQzqAAAAEpCbnDrlH0PX/AO/nMu2bdsKvo/mXfjHl1F/mt/Q1tYWtWk/+3wXXRZowIABUdukSZNCrPkZZnEJA82b0hw6s7gkgpbL8Pus+VpmZnPmzAnxsmXLDJWj37UvkaAlLfzyZrrMm+ZbXnzxxdF2X/jCF0Ks/W8WX3+2bt0atT300EMhXrhwYcH9b2Z6zvgyQ0899VSI/TV61KhRIfZ9MmXKlEru4kdyJXX5Qf+bM2/evBDr8l/+GpZiqZJK0zJAlaKlcHypLB1b+P7RvG7fl42EO3UAAAAJYFAHAACQgFxOv3p6a9xPf+gjy3369InaBg0aFGI/3YJ8mz9/foh1Cs0snn699dZbo7ZPfOITIX711VejttNPPz3EWlk8q+q8p1Nxvhq+L3GCytFpTj8l/rWvfS3EfnWB888/P8SzZ88Ocbdu8X/P6lSML5ly8ODBEOuqIWZxeQt0zX+399xzT4jXrFkTtWmpIX/91hUf/PGgZUy0VJYvb6Lnsl9B4P777w+xTrGaffQYQP1pqpVOzev5bxaXyfGlanbv3h3iSpRTqRfu1AEAACSAQR0AAEACGmL6VZ9g+Yd/+Ieo7dvf/naIfZVunVbT2/BmZv/8z/9cyV1EhWmldr+Ysj4Vpwtqm8ULOPsq9Ho8+Ok3lfXU2qZNm0KsC3b7fUZl6VNyDz74YNSm0+Df/OY3ozZdbSKLrlqifWxmNnfu3BD/7//+b9SmVehROp3mevzxx6O2n//85yHu3bt31KZPxE+fPj1q02vAjTfeWPA99Bqg03dm8ZPzOi2H/OvZs2eI9UlYs/g3wFfO0FUk1q5dW6W9qz7u1AEAACSAQR0AAEACGNQBAAAkIDc5dfqo+8qVK6O24cOHh9iXLcnKf/KPsKNxaHkQXanBLC5bcckllxR8j3JXEdHjZuPGjVHbf//3f4f4v/7rv6K2Dz/8sKzPQ9f0PNdq/2Zmjz32WIj9ygBXXXVViG+66aYQ+5UnNH9r5syZUZvmSvqyNYcOHepq15FB+1XzGv1rXz5o7969IV69enXUpn27YMGCEP/t3/5ttJ3+5mzYsCFq05UvWBkiTT4n+oknngixz89vJNypAwAASACDOgAAgATkZvpVq7bfddddUZs+huwXctYF3f20zLp160Lsp2V0GkWnUPSxZrO4ynl7e3vhPwAVpVMjS5Ysidp0WsZX9J8xY0bB99TpWJ1See6556Ltli1bVrBNjw9dZB6141cl0MW3dbF1s3ilgFmzZoXYryKifemn4nSqD/Xhp8N0OtZPzWr6hK4G4UtYKF3M3ayxF3Rvdno8+FUj9NqhpW/8a1/WqJFwpw4AACABDOoAAAASwKAOAAAgAbnJqdO57hUrVkRt//7v/x5iLW9iFi8hpkvOmJm9/fbbIT5w4EDUpnkzWorCz8FTpqL+fBmJxYsXh9gv4aPlJ3zZiu7du3f6/uvXr49eaxmTrVu3Rm3kVeaPnqO7du2K2vT1a6+9VrN9Qv1ovqzmyml5E6RFxw+6xJfmxHt+KTB/7WhU3KkDAABIAIM6AACABORm+lX5Ct5a0sKXt0Dz2b9/f4i1ZEVnrwEAadMxw/bt20PsV/3REmi+fNnmzZurtHe1xZ06AACABDCoAwAASACDOgAAgATkMqcOQP3ceeed9d4FADhmjzzySL13oea4UwcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJKGpQp4vlovGU03/0eWOjz5sPfd586PPm01X/FTWoa29vr8jOoD7K6T/6vLHR582HPm8+9Hnz6ar/WjqKGLYfOXLE2trarLW11VpaWiq2c6iujo4Oa29vtyFDhli3bqXNtNPnjYk+bz70efOhz5tPsX1e1KAOAAAA+caDEgAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACWBQBwAAkAAGdQAAAAlgUAcAAJAABnUAAAAJYFAHAACQAAZ1AAAACTi+mI2OHDlibW1t1traai0tLdXeJ1RIR0eHtbe325AhQ6xbt9LG7/R5Y6LPmw993nzo8+ZTbJ8XNahra2uzYcOGVWznUFtbt261oUOHlvRv6PPGRp83H/q8+dDnzaerPi9qUNfa2lqxHUJ5/H9RdXR0FP1vy+k/+ryxHUuft7S0WEtLS+Yx5tv0+DzuuOOiNv/6qBNPPDF6ffDgwRB37949atu3b1+nn2Vm9utf/7rgfjaTep7nWdcn7X/6qrLydG0vdNevlN8qdK2r/itqUMct2trI+p6PZVBXTv9Vqs+PZb/Lef8sxX62vmejXpCOpc+PDurK/Tz/bwu9V9Z25bZlDTYbtS+LlafzvFBbNc7XZpanPmdQVxtd9V9Rg7pK7YCfBy62s48cOdLp+5XyHsUq5YDP2rbQfpV79yOrze+Hfl/1kLVv+neU8/2Vu10pmv0iVM7xo9+Z//eHDx/u9N8cOHCg4PvpXTuz4s/Lal8fypX6wNPvt74u9prHzYPG1qjHbmp4+hUAACABDOoAAAASwKAOAAAgAVXPqdM8Op9Tp7k35eRgVENWvotX7dy1FHMUyJspXwq5V8Xyx4k+KXvSSSdFbT169AjxoUOHorb29vYQF8rtqwa//3rtK7fvGrXPs/LmGvVvAkpVq7xa7tQBAAAkgEEdAABAAqo+/apTlH66shFuvZdT26zc9yhXtd+/q0K0qI1a90E9p8r8Z0+ePDnEV155ZdQ2duzYELe1tUVtTz/9dIgXL14c4lqnTmQV3W3Uci3l7HfW917vv6cZVevaTl9+VK3SyrhTBwAAkAAGdQAAAAlgUAcAAJCAiufU5bVkhS+nomURNB4yZEi03YgRI0J86qmnRm29e/cOsc8V2bJlS4h/+ctfhvi9994rZbdzo1LLfZFr0RjqfR7781Vz6m688caobfz48SHev39/1KZ/x5IlS0Jcz6X0sr5b/3ercpYerKRS1txVeT3njz/+tz9/fh8LldtKTSX/tpS/p0bCnToAAIAEMKgDAABIQMWnX0tZkaHS/Gf16dMnxEOHDo3atAzCgAEDQnzRRRdF251zzjkhHjx4cNR28sknh9j/3Tt27AjxP/7jP4b4oYceirY7cOBAwfdAbehqBWZxv55yyilRm5am+OCDD6K2d999N8QHDx6s5C526eixX4ljqKWlpdNSB7Us2+NXjdBzb/jw4VGb9tdxxx0XtfXt2zfEuv95KtfSKCVNyv28vEx1a7qMmdlZZ50V4j179kRtmzdvDvH7779fpb1rbI3we+XPmawpd722N8LfVgh36gAAABLAoA4AACABDOoAAAASUPVlwmpJ58vNzEaOHBniP/3TP43arr/++hBrTo734YcfdhqbxbkiPXr0iNq0FIrm1G3atCna7tlnnw3x4cOHC+5HvZWbY1DP3ISsXCXNoxszZkzUdtlll4XYl7jp1atXiNetWxe1LViwoNO2auQUlVte4lg+o1iV+HzNhTOLc+p8blTWMlS6L1nlQiqtnt9dtdTimCuW5k76PEq9jupvgubQmZl9/etfD/H27dujth//+MchXr58edSW5z4qVWfHadY5o23V+B60L/3vubZl5cNp/59wwgnRdpqPe+jQoaht9+7dIfb50o3U59ypAwAASACDOgAAgAQkNf3as2fP6LWWLbnqqquiNl0dQm/XvvTSS9F2jz76aIg3btwYtek0kK9yr1N4p512WohnzpwZbXfLLbeE+Pnnn4/a/HRvvXR16zlPt6Z1usBPy+hUg5Yq+cpXvhJtd80114RYy92Yxbfzd+7cGbXp5+mx4m/zV0I1Vxco9B5Z022VPgZ8SoSe275f9bP9ihK7du0KcT1XCcj6PD1ms6Y4s6aWa6Geq0b4qbis6fj169eHWNMsbr311mi7q6++OsT+uNm3b1+IV69eHbX5qbnU1PI8z5oe1d9vs7gv165dG7Xt3bs3xL/61a86fT8zs3/5l38JsT/XHn744RBrKo1ZvEJU3nGnDgAAIAEM6gAAABKQ1PRrv379ote60Ldv01vof/RHfxTixx57LNou64lUvXX8zDPPRG3f+MY3On3/gQMHRtu1traG2E8x5GX6Nc/801m6EsGgQYOitjPOOCPE1157bYj9KiLar1lTXjqFbxY/Tbdo0aIQr1q1quB7VEo13rPWKxhoX1588cVR2+jRo0Psp2yUP1/rtRpAKVOV9VxxQXW2ikiWWk7/+s/SFVv8tV0rDwwbNizEPgVHp/R9H+h1JOt4a2RHV47xqn086mf638Pzzz8/xP66rH1y3nnnRW16bZ8zZ06INb3JzOzKK68MsU+L0Wnbt956K2pj+hUAAAA1xaAOAAAgAQzqAAAAElBSTt3ROfisx5yzKqlXO+/H5zjpPLvPE5g7d26ItWyJz2PL2medk9dq1GZmP//5z0Os8/rdu3ePttN99m0HDhwo+Nm11FXV8VrTkhZ9+vSJ2iZOnBjiadOmRW1aWmbp0qUh/t73vhdtp9XJ/WoTv//7v1/w/bVMype+9KUQ33HHHdF27777bojzVA7mqEK5NtVeQUBz6oYOHRq1ae6N3zc9tzXXysysvb2908+qdb5glqzPrtWqDYX6XNX6O9Lr4ZQpU6K2z33ucyGeOnVq1Ka5UXpd1vxls/i48WVK9LVfLUjP30ZWqM/1PKxGft2FF14Y4r/6q7+K2iZPnhxif21Xvg90n7dt2xbirLJWvl8nTZoUYs3tM4tLnOQlB7YQ7tQBAAAkgEEdAABAAipS0qTcxasrQR83HzduXNSmt3J1Ss0sXrA5q2xJlqxK9m1tbZ2+vy9bolMFjVTCpNbT7EpvtftH4rWMjfa/WdxHixcvDrHerjeLb69ntflSCpdcckmnsV+VQsts5LnPaz3dpv3av3//qK1Xr14F/12hxbzNzN57770Q61RMued8NWSlr9R7er6en69T8FdccUXUNn369BAPGTIkahs5cmSItZ+1JIZZfDz4qTidqtWVDFJSqG8rPb3op0AvuOCCEE+YMCFq0zJU/rdS9yurzIyuCOTLWvljQGkqlL/+1HOMUyru1AEAACSAQR0AAEACGNQBAAAkoKScuqNz8MU+gt/VtpWguQ8+d6lv374hfuedd6K2119/vaL74f9unfPXXCGf26dlS3xbXnTWh3nJo9T8GbN4KaAzzzwzalu9enWI9ZH4rPwqX+pAl4vRvEmzOHdk8ODBIfbHpS5Bk+eculrT88SX98kqs6Df4b59+6I2PU61pM2ePXui7fQ9fQ6Qvr/Pna30OVuvHLaOjo665+/5HNXrrrsuxDfffHPUpnl0Pr9K/45i88N8/pYeKxqbxedvvb+zY1Grffefo9db33d67vnfGH3tl4f0S34d5fvV/zvVu3fvEPsc3qzf8LzhTh0AAEACGNQBAAAkoOFLmgwbNizEfrpNb5m++eabUVuhSvPlOvnkkwvulz4S76eHdPo175Wq80JLHVx66aVR22WXXRZi/1i6fvf+tnwhfupAH4n3bTp10LNnzxD7lUGypgAaRbWnbrKmpf1n67Z+ZZfhw4eHWL933wdjx44t+HlvvPFGiJctWxa1bdy4McR5n5bpSmerBVWb9sOnPvWpqO3WW28Nse8fPX/91JumTGiKhF9xSKfb/G/Y2WefHeI//uM/jtq0JNGuXbs6jVGYnic+vSXr+NN/p6WKzMy2bt0a4k2bNoVY+9gs/p32aRb62l9/9Dc8T+WQOtP4vy4AAABgUAcAAJACBnUAAAAJqEhOXS35XJgLL7wwxFOmTInaNPdh0aJFUZsvTXCs/CPQmpOhj2370iqaG0B5i875ZXouuuiiEOuSMGZx3ozPmfjYxz4WYi1T4PtE+VwbzcnweXmFckv37t0bvc577lVLS0td8qs0p9TnnuprzVc0i8vT+L7U3FZdnmjSpEnRdnrc+PNQ83UWLFgQtc2aNSvEy5cvD3Ejlrqoxz6fcsopIZ42bVrUNnr06BD70hcHDx4MsS61aGa2YsWKEGufaP+bmZ111lmd7oeZ2YgRI0L8u7/7u1GbllPRnLBvfetb0XY7d+4Msf+9yVNeVq37XZdkyypbkpU766+pmtuq133/HllL8unvjL8+6PGgpbHyiDt1AAAACWBQBwAAkICSpl+PTst09v8XoxK3ef3Uy6hRo0Lsb6EvXbo0xHPnzo3aKnH7W6cEdJrHzGzq1Kkh1u/n1VdfjbbTaaW8TtmU2ueV/jvOOOOM6LWuGjFhwoSoTadEsyqZF1vSJOsWvU7nmsW3/fWz/LGW9+nXo2q9Ooy+v6ZOmMXTV1mrsvhSBxdffHGIdTrPp0tkHQ9aGmfQoEEF91lXqfHTx42gHlPuv/M7vxNiLVVkFk/T+XJPOuX6k5/8JGqbOXNmiHV61K8iouev71c9z/3qJrrShfJlVx544IEQ+9+fbdu2hbhepawK/Z5X+hjwaTADBw4MsZ9WV/570fPep2Hp+ay/F76MjZ7n/m/XNp2aNzO7/fbbQ/znf/7nIfYllPKAO3UAAAAJYFAHAACQAAZ1AAAACSgpp67QXHu18zB07vv000+P2jQnwy/9NX/+/IJtmjOhj8f7uXqd89dH2c3ichr+sfdx48aFWHMD/u7v/i7aTnMr8phTV0x5i2rst5YOue2226K2K664IsS6bJfnc680D8eXQSjE51oNHjw4xJqjZRYfK62trSGePHlytN3ixYtD7HPA8iSrHEA16Hnuy9hoXpPPw+nXr1+I9dgwi/Nr9N+VsrShHgPa/2Zmf/AHfxDiVatWhVhLnZjF15g8n+dZKrHfvu+++MUvhtjnMWUtr6hLt82ZMydq27BhQ4g1z03zK83MRo4cWXC/suj3pPli55xzTrSdXh/8b8d9990XYr+EZa1069bNWlpaMnP6qpEHr9dDn8uYdQzqNcD/O82P1O2ySqZ4+tvv8/NvuummEGvpNC1dZhb/5tTrPOdOHQAAQAIY1AEAACSgrBUlan1bUafYbrnllqjtzDPPDPH69eujNp3a0sfjzczGjBkT4r59+4b461//esH312kes3haxt/C1u9o165dIV6zZk20XZ4qi3emmCn3ahwPWqpEp9jN4tUg/OPy2g++irseH8VOv/qpAy1v0adPn6hNb9/rfvmpl7yvHHLkyJG6lLdQflpap998ioROdfv+KjTlWkoZCf13/rP1eNAUDE39MIuvAXksaXO0r6vd535aXdMs/Ll86NChEOvUtllcxmTdunVRm15Tr7766hBrmSmz+HfFX4d1lZIXXnghatNruE6/63Ho39+vdrRw4cIQb9myJWqr1fWhmue5njNawsQsLv/ly8VkyTpn9fNKSa0o5v3M4mNTxws+PScP5zZ36gAAABLAoA4AACABJU+/1mNKRp9gu+SSS6I2vd352muvRW36VJF/+lWnaXQKwK8ModMrfupA+fdfsmRJiPWJV61+j8L0CSlfaV77wR+POuWqi3mbmf3N3/xNiP2KBUpvvfupU3362k+36PSATh/qSgOd7TN+Q7+Xt99+O2rT6TD//ekUSNbKEDqttXnz5qjtrbfeCrGfLho+fHiI/dPWhdIQ9GlXs/qtGlCqaqwiolPWftUFrfjvr696Td20aVPUpk+4+j7X68Wtt94aYp8+o3+bn8KdPXt2iJ944omoTffz4x//eIj9yjd6rIwfP77gPvop/Vqp5nVIpyv99VufXPVT7llTp1mrBelrPdfK/W79NGqhYzGP13Lu1AEAACSAQR0AAEACGNQBAAAkoKySJtXm58E1b8HntOiKDFpl3Lf58hb6WufLs+bq/Xy/zt23tbVFbfrIvc/takT+e6nEY+PK51ZoKZnevXsX3BffrytXrgzxd77znaht7dq1Re2LPmbvV4OYOHFiiH0uj5Zg+OUvfxni1atXF9wur+qRK6LHlM951Jw6n+9S7GoAeqzMmzcvatu9e3eIzzvvvKhtwIABIfYlGPQaoHleH3zwQbRdHnNvakX7tVevXlGbfrf+GqBlh3xO3TvvvBNi3/+XX355iDUH1p+vejzob4WZ2Z49e0Lsc/1GjRoVYs0PyypvsXPnzqjNr5CRJ5U4VvW80HI+ZvH37nNN9be/lBVt9D11O993eoxl/Z7v3bs3avvFL34R4gcffDDEvhROHs5z7tQBAAAkgEEdAABAAnI5/aplRMzMrrzyyhDr7XqzeJrLVx3XaVVfpVvb9Da53tY3i8sZeHq71k+p6etaryBw9LZyJW8FV3q61fNTKLq4ty8dot+nlqIwi8sPPPfcc1Fbsat36JTKRRddFLXptLCnUyp6XPrVEfJQdbwr9VhRQqdG/LS6Tof47y/r2CxUJuXpp5+OttO0Dt/HWWVS9Jh65ZVXCu5jI6hWfxdaXccsPjf8dVJf+zYtO6OLuZvFK3tov/q/T/vIT9ONHj06xP434Jxzzgmxpob4KXedMp45c2bUtnTp0hDXc1WhavW5/jb68kFPPvlkiKdPnx616So9fmpWf7N9utPLL78cYi0fc/7550fbZa1gov330ksvRW333ntviPV3JY+lirhTBwAAkAAGdQAAAAlgUAcAAJCAknPqqpVro3krfp5d86t8rsqaNWtC7B9716V6fN7NhAkTQqxz9b58huZ6+b9b5+D9HL8uUVbr3KQ8PFZdKp/foDkzvk371S8Nt3DhwhAXWzZAl4wzM7vqqqtC7Mtb9O3bt+D7aFmMFStWhLgR86tqISsXzufUad5RKd+n5rzosaJ9ZWY2cuTIEJ9yyilRW9aydFpqRUvm5DHXpl70u/Df+5YtW0I8ZsyYqE3zqyZNmlTw/bVsiVmcR6XHmO87vbZr/5vFy5f5MjY9evQIseYEPvbYY9F2ixcvDrHm0JnFJVQa8XpdCl+e6L777gux/n6bxfn0vqyI/k77/EU99zQf0udbajkaf/3ZunVriDUn2szs+eefD7FfArDSjnWpPu7UAQAAJIBBHQAAQALqWtJEbzPqo8Y33XRTtJ22vfjii1Hbq6++GmKtQG4W37b0lcwvuOCCEOv0m58C0Km/AwcORG0vvPBCiP/t3/4talu2bFmn+1ENfgWOo687OjoqMvVXi+kBv586reUf+dd+8NPeuq2fNtHSKDqldvXVV0fb3XzzzSHWaXqz+Lv207vr1q0Lsa5ukvr0Srmyphn81IuWGvJTLzpNl1UlXvmyJTrNPnbs2KhNjxU/LazTRzqVSJ93zk/FaZqKlgoxi1MdLr300qhNr9/+2q6v9Xjw10mdRh0yZEjUplNsei0yi0t06LScL1ui03n+726mlAx/Lmi5EE1T6erfKX+e6+v169eH+HOf+1y03dChQwu+/+uvvx7iZ555Jmrz530tlVqijDt1AAAACWBQBwAAkAAGdQAAAAkoOaeuFrki/jFkzV1atGhR1KaPhvulZDSnasaMGVHb5z//+RDr36TLypjFuQ/+8etZs2aF2M/B+/y7StMcAl/y4+jrSuXU1YLvuyVLloRYywuYxblxEydOjNr0tc/F+/jHPx5iXW7u05/+dLSdLgukeTfezp07o9eaY+nzvhpNPZaY0/w3n6+oy8Fp7qJZXP7GL/ekeVRTpkwJsS4lZBaXUvDL0ul3oXlSZnHZCr/EYKUV+h4bKX/Pl4P46U9/GuLBgwdHbbo8pC8lpHnQ/vpX7PGm/87nXmp+9v/93/9FbQ899FCItf993pz2SyP1US2V+734f6evdYlO/zusbf73QUuirV69uqz9ylLtpTaP4k4dAABAAhjUAQAAJKCskibHWvG4M1lToDodMm3atKhNHzX203Q6LePLpGg1ceVvw+s0kD6KbRY/Ol3t6VZPvy8/dXn0b2ikW/5+mlhL10yePDlq02k0rR5vFlee9++px4MqNH1t9tHvUEtt6BSxmdkDDzxQ8N81mmqtHKPToVnlR7Zv3x61PffccyH26Rnjxo0L8aBBg6I27UtftqIQf9y8/fbbIfZ9/qMf/SjE/jwsR62maOrFX1+19JOfctcSE5/5zGeiNj3P/Yowhb7DrO/WX7/1Wv+f//mfUZuuDlGJPsex077Va4CuIGEWl0fT1UDM4tSARk6f4U4dAABAAhjUAQAAJKDk6deWlpbM29ilTNnotvrkkC7Kbmb25S9/OcT6BKOZ2WWXXRbtW7H7pbf6tWK4vyWrTzfNnj07atMpwmo/Zeqf6tPbyP4pnqNTAh0dHVVffLhSfP/oSiE6/WUWf9e9e/eO2vR1Jaay/Hf78ssvh3jevHlRm38y8lhV6jwr53MrOf2q71PseeKntfRptMcffzxq0ycjfQV5XW3CryigdL/81O/DDz8c4u9973tRm64ugNLp9UnPLbP4GvCDH/wgatOUDJ9aoyvC6OoSvv+1z3U1EDOzZ599NsQrV66M2phyzR/tW/1t9NcbvRb56gV6Llfj97yUFTKOBXfqAAAAEsCgDgAAIAEM6gAAABJQVkkT/1h6Jeh885w5c6K22267LcSaI2GWXfE/K5dHS1NorHlyZmY//vGPQ/zaa69FbeU89lxuORifx6F5gP49jn5GI5fV0FUDHnzwwahNjwGfY6nlb3wOTaG8BX88a56PL2Fx7733hviJJ56I2irxfdcrj04df/zx1tLS8pF8wnoeT1q6aMWKFVGb5uP6shjTp08Psa4U4rfTvCl//Xn66adDrFXnU1OtMjbl0mu2rvBgFq/y8Morr0Rt99xzT4hvuOGGEF933XXRdlrGZO7cuVHbI488EuL29vbid7rCqlE+7Fg+P0s9j53jj//tUKZfv34h9uWqso6prN/UWjrW3wDu1AEAACSAQR0AAEACypp+rQa9reinV77//e+H2FcW11UC/K1JnR7V6uRm8TRa9+7dQ7xq1apou7Vr13b6frWgt2H9VKKuiKGLFJtV/tZxZ7eDq317WqdEFy1aFLXptOCOHTuithkzZoR4wIABUZv+HfoevmSKlirxZUu0zIJfwLvadP+r+f2fdNJJ1tLS8pGUhWqX7cmif68/Dzds2BDi++67L2rTUhgTJkwIsZ9yf/PNN0Psj6ndu3eHuNrlLAqlUnTVVi3Ffka1rwf+e9fX/njQ/tq4cWOI58+fX/D933rrrei1lrXJ05R0rdXjmCuHXpv0/M1KFfPly/x5nxelXve5UwcAAJAABnUAAAAJYFAHAACQgJJz6mqRX6CPFpuZ3X///SFetmxZ1KZLBHmaa+GXgdESBpqv5h9zrnQeUbnLqPlcBs0X849tH/0bOjo6KvI4flf7XO3H7v1SZ0uXLg3xtm3borYnn3wyxBMnTozatIyF5uRoSRszs5deeinE/rjRMgjVkNXntcrt2b9//zGVt6h2WRb/HnqMv/HGGwX/nc+rrfR+VUPWflV6n/P6HZRKz+09e/aEWJf+ahT17hN/Lutr31Yof60Wf4P2ueZIa2ksM7NBgwZ1up3ZR38HaqmY87zY75E7dQAAAAlgUAcAAJCA3JQ0Uf42o67k4Fd1yEMF/mrJ+tt27dpVcLujt6Ir9ffn7TF2XV1AS86YxbfbH3/88YLvod9Nno6TWpUtyXL0c7OmXsotdVDPvy9P/ZxXWVP+pRwPSJ8vs1XoGKj1saFpNloOzcxszZo1IfZTsz6Vp1Fxpw4AACABDOoAAAASwKAOAAAgAbnMqfO0ZEcpj/jnJeej3P3IenxcS7n07NkzajuaG1DJvzmvOTPlPnKf178na7/KKSdQjnKWwsrr95mSape4KTV3lj4vXr3KE3Wls9JFWctTapsvpaXXoEpej0qly2b6JSafeeaZgv8uL2OEY8WdOgAAgAQwqAMAAEhAQ0y/qnJXZMiit5jLvW3sb0Xra721qreGzbJLBhTazixe6UJXztBtK3U7OS9TBZ0pZTq+0LaNUhYnb6VlUH217vNarmCRmlL6qtzyQJXsg5aWlvC/Qk444YTotf4+ZqW+5FUtj+FSptwr2efcqQMAAEgAgzoAAIAEMKgDAABIQEPk1Gn+WJZy5/Qr8fi1n/fW9zzxxBNDXMrcedZc+p49ewq2Hf2MZsyBKedvbpTvqZglmxrlb8Gxo8+rp9L5YZVesrES79etWzdraWkp+/fPlztBtqxjKqtclebIF9NX9AoAAEACirpT1yj/JVjP/cy6y5a1eHyx+1zOvzuWp2Abpc+bSSWe9C33/VEf1Xjav9C/of9/K0+/JZXa1v+bYn4fiv1d66qtGVWqH0vpL7MiB3Xt7e3F7lvTyio5sn///oq/fylV/9vb261Pnz4lfR59nm9dndj0efOhz5vPsfR5MVN5Bw8eLNh2+PDhkj4XhZUyBd5Vn7d0FDGcPHLkiLW1tVlra2tD1KLBb3R0dFh7e7sNGTKk5PwH+rwx0efNhz5vPvR58ym2z4sa1AEAACDfeFACAAAgAQzqAAAAEsCgDgAAIAEM6gAAABLAoA4AACABDOoAAAASwKAOAAAgAf8PSr116y9Ex4QAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 72/100, d_loss=0.012, g_loss=0.761                                                                                                                      \n",
      "epoch = 73/100, d_loss=0.015, g_loss=0.866                                                                                                                      \n",
      "epoch = 74/100, d_loss=0.010, g_loss=0.768                                                                                                                      \n",
      "epoch = 75/100, d_loss=0.026, g_loss=0.844                                                                                                                      \n",
      "epoch = 76/100, d_loss=0.010, g_loss=0.822                                                                                                                      \n",
      "epoch = 77/100, d_loss=0.011, g_loss=0.802                                                                                                                      \n",
      "epoch = 78/100, d_loss=2.174, g_loss=2.963                                                                                                                      \n",
      "epoch = 79/100, d_loss=0.009, g_loss=0.831                                                                                                                      \n",
      "epoch = 80/100, d_loss=0.011, g_loss=0.812                                                                                                                      \n",
      "epoch = 81/100, d_loss=0.016, g_loss=0.843                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 82/100, d_loss=0.012, g_loss=0.833                                                                                                                      \n",
      "epoch = 83/100, d_loss=0.010, g_loss=0.843                                                                                                                      \n",
      "epoch = 84/100, d_loss=0.010, g_loss=0.906                                                                                                                      \n",
      "epoch = 85/100, d_loss=0.008, g_loss=0.845                                                                                                                      \n",
      "epoch = 86/100, d_loss=0.012, g_loss=0.866                                                                                                                      \n",
      "epoch = 87/100, d_loss=0.010, g_loss=0.825                                                                                                                      \n",
      "epoch = 88/100, d_loss=0.011, g_loss=0.775                                                                                                                      \n",
      "epoch = 89/100, d_loss=0.009, g_loss=0.770                                                                                                                      \n",
      "epoch = 90/100, d_loss=0.009, g_loss=0.820                                                                                                                      \n",
      "epoch = 91/100, d_loss=0.009, g_loss=0.785                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch = 92/100, d_loss=0.009, g_loss=0.849                                                                                                                      \n",
      "epoch = 93/100, d_loss=0.009, g_loss=0.823                                                                                                                      \n",
      "epoch = 94/100, d_loss=0.009, g_loss=0.805                                                                                                                      \n",
      "epoch = 95/100, d_loss=0.007, g_loss=0.850                                                                                                                      \n",
      "epoch = 96/100, d_loss=0.010, g_loss=0.858                                                                                                                      \n",
      "epoch = 97/100, d_loss=0.009, g_loss=0.811                                                                                                                      \n",
      "epoch = 98/100, d_loss=0.010, g_loss=0.843                                                                                                                      \n",
      "epoch = 99/100, d_loss=0.010, g_loss=0.839                                                                                                                      \n",
      "epoch = 100/100, d_loss=0.016, g_loss=0.831                                                                                                                      \n",
      "epoch = 101/100, d_loss=0.010, g_loss=0.832                                                                                                                      \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "epochs = 100\n",
    "batch_size = 64\n",
    "smooth = 0.1\n",
    "\n",
    "real = tf.ones((batch_size, 4, 4, 1))\n",
    "real = real * (1 - smooth)\n",
    "fake = tf.zeros((batch_size, 4, 4, 1))\n",
    "\n",
    "fake_labels = tf.keras.utils.to_categorical(\n",
    "    np.full((batch_size, 1), num_classes), num_classes=num_classes + 1\n",
    ")\n",
    "\n",
    "d_loss = []\n",
    "d_g_loss = []\n",
    "\n",
    "for e in range(epochs + 1):\n",
    "    for i in range(len(X_train) // batch_size):\n",
    "\n",
    "        # Train Discriminator weights\n",
    "        discriminator.trainable = True\n",
    "\n",
    "        # Real samples\n",
    "        img_real = X_train[i * batch_size : (i + 1) * batch_size]\n",
    "        real_labels = y_train[i * batch_size : (i + 1) * batch_size]\n",
    "\n",
    "        d_loss_real = discriminator.train_on_batch(\n",
    "            x=img_real, y=[real, real_labels]\n",
    "        )\n",
    "\n",
    "        # Fake Samples\n",
    "        masked_imgs = mask_randomly(img_real)\n",
    "        gen_imgs = generator.predict_on_batch(masked_imgs)\n",
    "\n",
    "        d_loss_fake = discriminator.train_on_batch(\n",
    "            x=gen_imgs, y=[fake, fake_labels]\n",
    "        )\n",
    "\n",
    "        # Discriminator loss\n",
    "        d_loss_batch = 0.5 * (d_loss_real[0] + d_loss_fake[0])\n",
    "\n",
    "        # Train Generator weights\n",
    "        discriminator.trainable = False\n",
    "\n",
    "        masked_imgs = mask_randomly(img_real)\n",
    "        d_g_loss_batch = d_g.train_on_batch(x=masked_imgs, y=real)\n",
    "\n",
    "        print(\n",
    "            \"epoch = %d/%d, batch = %d/%d, d_loss=%.3f, g_loss=%.3f\"\n",
    "            % (\n",
    "                e + 1,\n",
    "                epochs,\n",
    "                i,\n",
    "                len(X_train) // batch_size,\n",
    "                d_loss_batch,\n",
    "                d_g_loss_batch[0],\n",
    "            ),\n",
    "            100 * \" \",\n",
    "            end=\"\\r\",\n",
    "        )\n",
    "\n",
    "    d_loss.append(d_loss_batch)\n",
    "    d_g_loss.append(d_g_loss_batch[0])\n",
    "\n",
    "    print(\n",
    "        \"epoch = %d/%d, d_loss=%.3f, g_loss=%.3f\"\n",
    "        % (e + 1, epochs, d_loss[-1], d_g_loss[-1]),\n",
    "        100 * \" \",\n",
    "    )\n",
    "\n",
    "    if e % 10 == 0:\n",
    "        samples = 5\n",
    "        idx = np.random.randint(0, X_train.shape[0], samples)\n",
    "        masked_imgs = mask_randomly(X_train[idx])\n",
    "        x_fake = generator.predict_on_batch(masked_imgs)\n",
    "\n",
    "        for k in range(samples):\n",
    "            # plot masked\n",
    "            plt.subplot(2, 5, k + 1)\n",
    "            plt.imshow(masked_imgs[k].reshape(32, 32), cmap=\"gray\")\n",
    "            plt.xticks([])\n",
    "            plt.yticks([])\n",
    "\n",
    "            # plot recontructed\n",
    "            plt.subplot(2, 5, k + 6)\n",
    "            plt.imshow(x_fake[k].reshape(32, 32), cmap=\"gray\")\n",
    "            plt.xticks([])\n",
    "            plt.yticks([])\n",
    "\n",
    "        plt.tight_layout()\n",
    "        plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Evaluate model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-08-06T22:11:27.495321Z",
     "start_time": "2018-08-06T22:11:27.332075Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAioAAAHFCAYAAADcytJ5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAABoKklEQVR4nO3dd3hT1f8H8PfN7m4plLbsvTeo7CEqICgKioAyXF9QwPFzL1xYt4gKOBgKKKKg4gBkKyAyK1NAZBRLKbO7aZOc3x8ns01LKWmTG96v58nT9ub25uQ2Td73nM89VxFCCBAREREFII2/G0BERERUEgYVIiIiClgMKkRERBSwGFSIiIgoYDGoEBERUcBiUCEiIqKAxaBCREREAYtBhYiIiAIWgwoREREFLAYVoivM3LlzoSgKFEXBunXrit0vhEDDhg2hKAp69erl08dWFAUvvvjiJf/e0aNHoSgK5s6d65P1iEg9GFSIrlARERGYNWtWseXr16/H4cOHERER4YdWERF5YlAhukINGzYMixcvRmZmpsfyWbNmoXPnzqhdu7afWkZE5MKgQnSFGj58OADgq6++ci7LyMjA4sWLcffdd3v9nXPnzuGBBx5AjRo1YDAYUL9+fTz77LMwm80e62VmZuK+++5DbGwswsPD0a9fPxw8eNDrNg8dOoQRI0YgLi4ORqMRzZo1w0cffeSjZylt2LAB1157LSIiIhAaGoouXbrg559/9lgnNzcXjz32GOrVqweTyYQqVaqgY8eOHvvn33//xR133IHExEQYjUZUr14d1157LZKTk33aXiJy0fm7AUTkH5GRkRg6dChmz56N//3vfwBkaNFoNBg2bBimTp3qsX5+fj569+6Nw4cP46WXXkLr1q3x+++/IykpCcnJyc4PfiEEBg8ejE2bNuGFF15Ap06dsHHjRvTv379YG/bt24cuXbqgdu3aeOeddxAfH48VK1Zg0qRJOHPmDCZPnnzZz3P9+vW47rrr0Lp1a8yaNQtGoxHTp0/HoEGD8NVXX2HYsGEAgEcffRTz5s3Dq6++inbt2iEnJwd79uzB2bNnndsaMGAArFYr3nzzTdSuXRtnzpzBpk2bcOHChctuJxGVQBDRFWXOnDkCgNi6datYu3atACD27NkjhBCiU6dOYsyYMUIIIVq0aCF69uzp/L2ZM2cKAGLRokUe23vjjTcEAPHrr78KIYRYtmyZACDef/99j/WmTJkiAIjJkyc7l91www2iZs2aIiMjw2PdCRMmCJPJJM6dOyeEEOLIkSMCgJgzZ06pz83betdcc42Ii4sTWVlZzmUWi0W0bNlS1KxZU9hsNiGEEC1bthSDBw8ucdtnzpwRAMTUqVNLbQMR+RaHfoiuYD179kSDBg0we/Zs7N69G1u3bi1x2GfNmjUICwvD0KFDPZaPGTMGALB69WoAwNq1awEAI0eO9FhvxIgRHj/n5+dj9erVuOWWWxAaGgqLxeK8DRgwAPn5+di8efNlPb+cnBz8+eefGDp0KMLDw53LtVot7rrrLpw4cQIHDhwAAFx11VVYtmwZnnrqKaxbtw55eXke26pSpQoaNGiAt956C++++y527twJm812We0jootjUCG6gimKgrFjx2L+/PmYOXMmGjdujO7du3td9+zZs4iPj4eiKB7L4+LioNPpnEMkZ8+ehU6nQ2xsrMd68fHxxbZnsVjwwQcfQK/Xe9wGDBgAADhz5sxlPb/z589DCIGEhIRi9yUmJjrbAQDTpk3Dk08+ie+//x69e/dGlSpVMHjwYBw6dAiA3FerV6/GDTfcgDfffBPt27dHtWrVMGnSJGRlZV1WO4moZAwqRFe4MWPG4MyZM5g5cybGjh1b4nqxsbE4deoUhBAey9PT02GxWFC1alXnehaLxaO2AwDS0tI8fo6JiYFWq8WYMWOwdetWrzdHYCmvmJgYaDQanDx5sth9qampAOBsd1hYGF566SX8/fffSEtLw4wZM7B582YMGjTI+Tt16tTBrFmzkJaWhgMHDuCRRx7B9OnT8fjjj19WO4moZAwqRFe4GjVq4PHHH8egQYMwevToEte79tprkZ2dje+//95j+RdffOG8HwB69+4NAFiwYIHHel9++aXHz6Ghoejduzd27tyJ1q1bo2PHjsVuRXtlLlVYWBiuvvpqLFmyxGMox2azYf78+ahZsyYaN25c7PeqV6+OMWPGYPjw4Thw4AByc3OLrdO4cWM899xzaNWqFXbs2HFZ7SSikvGsHyLC66+/ftF1Ro0ahY8++gijR4/G0aNH0apVK2zYsAGvvfYaBgwYgL59+wIArr/+evTo0QNPPPEEcnJy0LFjR2zcuBHz5s0rts33338f3bp1Q/fu3TF+/HjUrVsXWVlZ+Oeff/Djjz9izZo1l/3ckpKScN1116F379547LHHYDAYMH36dOzZswdfffWVcyjr6quvxsCBA9G6dWvExMRg//79mDdvHjp37ozQ0FDs2rULEyZMwG233YZGjRrBYDBgzZo12LVrF5566qnLbicRecegQkRlYjKZsHbtWjz77LN46623cPr0adSoUQOPPfaYx2nEGo0GS5cuxaOPPoo333wTBQUF6Nq1K3755Rc0bdrUY5vNmzfHjh078Morr+C5555Deno6oqOj0ahRo8se9nHo2bMn1qxZg8mTJ2PMmDGw2Wxo06YNli5dioEDBzrX69OnD5YuXYr33nsPubm5qFGjBkaNGoVnn30WgKyxadCgAaZPn46UlBQoioL69evjnXfewcSJE33SViIqThFFB5yJiIiIAgRrVIiIiChgMagQERFRwGJQISIiooDFoEJEREQBi0GFiIiIAhaDChEREQUsVc+jYrPZkJqaioiIiGLXHyEiIqLAJIRAVlYWEhMTodGU3mei6qCSmpqKWrVq+bsZREREVA4pKSmoWbNmqeuoOqhEREQAkE80MjLSz60hIiKissjMzEStWrWcn+OlUXVQcQz3REZGMqgQERGpTFnKNlhMS0RERAGLQYWIiIgCFoMKERERBSwGFSIiIgpYDCpEREQUsBhUiIiIKGAxqBAREVHAYlAhIiKigMWgQkRERAGLQYWIiIgClt+Dyn///Yc777wTsbGxCA0NRdu2bbF9+3Z/N4uIiIgCgF+v9XP+/Hl07doVvXv3xrJlyxAXF4fDhw8jOjran80iIiKiAOHXoPLGG2+gVq1amDNnjnNZ3bp1/dcgtSvMA3QmoAwXeSIiIlIDvw79LF26FB07dsRtt92GuLg4tGvXDp9++qk/m6RemSeBtxoC3z/g75YQERH5jF+Dyr///osZM2agUaNGWLFiBcaNG4dJkybhiy++8Lq+2WxGZmamx43sTv8NFGQD/23zd0uIiIh8xq9DPzabDR07dsRrr70GAGjXrh327t2LGTNmYNSoUcXWT0pKwksvvVTZzVQHm9X+1eLfdhAREfmQX3tUEhIS0Lx5c49lzZo1w/Hjx72u//TTTyMjI8N5S0lJqYxmqoMjoDCoEBFREPFrj0rXrl1x4MABj2UHDx5EnTp1vK5vNBphNBoro2nq4wwqVv+2g4iIyIf82qPyyCOPYPPmzXjttdfwzz//4Msvv8Qnn3yCBx980J/NUif2qBARURDya1Dp1KkTvvvuO3z11Vdo2bIlXnnlFUydOhUjR470Z7PUiUGFiIiCkF+HfgBg4MCBGDhwoL+boX4spiUioiDk9yn0yUeEI6iwRoWIiIIHg0qw4NAPEREFIQaVYMGzfoiIKAgxqAQL1qgQEVEQYlAJFs6AIgCbza9NISIi8hUGlWDh3pPCXhUiIgoSDCrBgkGFiIiCEINKsGBQISKiIMSgEizcz/ZhUCEioiDBoBIsPHpUeIoyEREFBwaVYMGhHyIiCkIMKsGCQYWIiIIQg0qwYI0KEREFIQaVYMEaFSIiCkIMKsGCQz9ERBSEGFSCBYMKEREFIQaVYMEaFSIiCkIMKsGCNSpERBSEGFSCBYd+iIgoCDGoBAsO/RARURBiUAkW7FEhIqIgxKASLNijQkREQYhBJViwmJaIiIIQg0qw4NAPEREFIQaVYOEeTgR7VIiIKDgwqAQL1qgQEVEQYlAJFqxRISKiIMSgEixYo0JEREGIQSVYMKgQEVEQYlAJFqxRISKiIMSgEizYo0JEREGIQSVYsJiWiIiCEINKsGCPChERBSEGlWDBGhUiIgpCDCrBgj0qREQUhBhUggVrVIiIKAgxqAQL9qgQEVEQYlAJFsLm+p5BhYiIggSDSrBgjwoREQUhBpVgwRoVIiIKQgwqwYI9KkREFIQYVIKBzcYaFSIiCkoMKsFAFBnqYVAhIqIgwaASDIoGEwYVIiIKEgwqwaBYUGExLRERBQe/BpUXX3wRiqJ43OLj4/3ZJHVijwoREQUpnb8b0KJFC6xatcr5s1ar9WNrVKpoDwqDChERBQm/BxWdTsdelMvFoR8iIgpSfq9ROXToEBITE1GvXj3ccccd+Pfff/3dJPVhUCEioiDl1x6Vq6++Gl988QUaN26MU6dO4dVXX0WXLl2wd+9exMbGFlvfbDbDbDY7f87MzKzM5gYu1qgQEVGQ8muPSv/+/TFkyBC0atUKffv2xc8//wwA+Pzzz72un5SUhKioKOetVq1aldncwMUaFSIiClJ+H/pxFxYWhlatWuHQoUNe73/66aeRkZHhvKWkpFRyCwMUe1SIiChI+b2Y1p3ZbMb+/fvRvXt3r/cbjUYYjcZKbpUKsEaFiIiClF97VB577DGsX78eR44cwZ9//omhQ4ciMzMTo0eP9mez1Ic9KkREFKT82qNy4sQJDB8+HGfOnEG1atVwzTXXYPPmzahTp44/m6U+DCpERBSk/BpUFi5c6M+HDx4spiUioiAVUMW0VE7FggprVIiIKDgwqAQDDv0QEVGQYlAJBgwqREQUpBhUggFrVIiIKEgxqAQDzqNCRERBikElGDiCis7k+TMREZHKMagEA2dQMXr+TEREpHIMKsGAPSpERBSkGFSCgaMmxdmjwhoVIiIKDgwqwYA9KkREFKQYVIIBa1SIiChIMagEA/aoEBFRkGJQCQbOGhV7UBFWQAj/tYeIiMhHGFSCgaMHRWtwLRM2/7SFiIjIhxhUgkHRoR/3ZURERCrGoBIMihbTui8jIiJSMQaVYFB0HhWAQYWIiIICg0ow8NqjwknfiIhI/RhUgoGwhxL3Ylr2qBARURBgUAkGjlCi0QManecyIiIiFWNQCQbOoKJlUCEioqDCoBIMHPUoGh2DChERBRUGlWDg7FHRyV4VgMW0REQUFBhUgoFHUGGPChERBQ8GlWDAGhUiIgpSDCrBgDUqREQUpBhUggFrVIiIKEgxqAQD1qgQEVGQYlAJBqxRISKiIMWgEgxYo0JEREGKQSUYeK1RYVAhIiL1Y1AJBl5rVFhMS0RE6segEgxYo0JEREGKQSUYOGtUGFSIiCi4MKgEA56eTEREQYpBJRhwwjciIgpSDCrBgMW0REQUpBhUggFrVIiIKEgxqAQDTvhGRERBikElGHDCNyIiClIMKsGANSpERBSkGFSCAU9PJiKiIMWgEgzci2kVDv0QEVHwYFAJBqxRISKiIMWgEgxYo0JEREEqYIJKUlISFEXBww8/7O+mqA9rVIiIKEgFRFDZunUrPvnkE7Ru3drfTVEnTvhGRERByu9BJTs7GyNHjsSnn36KmJgYfzdHnVijQkREQcrvQeXBBx/EjTfeiL59+/q7KerFGhUiIgpSOn8++MKFC7Fjxw5s3bq1TOubzWaYzWbnz5mZmRXVNHVhjQoREQUpv/WopKSk4KGHHsL8+fNhMpnK9DtJSUmIiopy3mrVqlXBrVQBmw2AkN8zqBARUZDxW1DZvn070tPT0aFDB+h0Ouh0Oqxfvx7Tpk2DTqeD1Vp86OLpp59GRkaG85aSkuKHlgcY90DCYloiIgoyfhv6ufbaa7F7926PZWPHjkXTpk3x5JNPQqvVFvsdo9EIo9FYWU1UB/dAomjdimlZo0JEROrnt6ASERGBli1beiwLCwtDbGxsseVUCo8eFQ79EBFRcPH7WT90mRhUiIgoiPn1rJ+i1q1b5+8mqI+wub5njQoREQUZ9qionSOQKFpAUTjhGxERBRUGFbVzn0PF/SuLaYmIKAgwqKhdiUGFPSpERKR+DCpq57wgYZGgItijQkRE6segonbOHhWt51f2qBARURBgUFE71qgQEVEQY1BRO9aoEBFREGNQUTsGFSIiCmIMKmrnLKZ11KgwqBARUfBgUFG7Yj0qvCghEREFDwYVtePQDxERBTEGFbUrsUeFQYWIiNSPQUXtWKNCRERBjEFF7VijQkREQYxBRe1Yo0JEREGMQUXtik2hz6BCRETBg0FF7VijQkREQYxBRe14rR8iIgpiDCpq5+xR4enJREQUfBhU1I7FtEREFMQYVNSOxbRERBTEGFTUrtjQj/2rsAE2m3/aRERE5CMMKmpX0oRvACBYUEtEROrGoKJ2JdWouN9HRESkUgwqaldSjYr7fURERCqlu/gqFNBKqlFxv4+I6DLZbDYUFBT4uxmkEnq9Hlqt9uIrlgGDitoVHfpR3F4YDCpE5AMFBQU4cuQIbCzQp0sQHR2N+Ph4KIpyWdthUFG7YjUqGkDR2M/64dAPEV0eIQROnjwJrVaLWrVqQaNhxQCVTgiB3NxcpKenAwASEhIua3sMKmpXNKg4vrcWMKgQ0WWzWCzIzc1FYmIiQkND/d0cUomQkBAAQHp6OuLi4i5rGIjRWO2KFtMCnPSNiHzGapVDyAaDwc8tIbVxBNvCwsLL2g6DitoVLaZ1/55BhYh85HLrDOjK46vXDIOK2nkd+nFcmJDFtEREpG4MKmpXUo2K+31ERFQiRVHw/fffV9j2x4wZg8GDB1/WNtatWwdFUXDhwgWftElNGFTUjjUqRERejRkzBoqiQFEU6PV6VK9eHddddx1mz57tcar1yZMn0b9//wprx/vvv4+5c+de1ja6dOmCkydPIioqyjeNsqvokOYLDCpqxxoVIqIS9evXDydPnsTRo0exbNky9O7dGw899BAGDhwIi0W+R8bHx8NoNPr8sa1WK2w2G6KiohAdHX1Z2zIYDD6Zk6SiXG7BbGkYVNTOEUbcJ3pjjQoREQDAaDQiPj4eNWrUQPv27fHMM8/ghx9+wLJly5y9HO69CgUFBZgwYQISEhJgMplQt25dJCUlObd34cIF3H///ahevTpMJhNatmyJn376CQAwd+5cREdH46effkLz5s1hNBpx7NixYkM/vXr1wsSJE/Hwww8jJiYG1atXxyeffIKcnByMHTsWERERaNCgAZYtW+b8naJDP47HWrFiBZo1a4bw8HBnKHPYunUrrrvuOlStWhVRUVHo2bMnduzY4by/bt26AIBbbrkFiqI4fwaAGTNmoEGDBjAYDGjSpAnmzZvnsV8VRcHMmTNx8803IywsDK+++mp5/0QXxaCidoI9KkRUeYQQyC2w+OUmhPDJc+jTpw/atGmDJUuWFLtv2rRpWLp0KRYtWoQDBw5g/vz5zg9wm82G/v37Y9OmTZg/fz727duH119/3WOOkNzcXCQlJeGzzz7D3r17ERcX57UNn3/+OapWrYotW7Zg4sSJGD9+PG677TZ06dIFO3bswA033IC77roLubm5JT6P3NxcvP3225g3bx5+++03HD9+HI899pjz/qysLIwePRq///47Nm/ejEaNGmHAgAHIysoCIIMMAMyZMwcnT550/vzdd9/hoYcewv/93/9hz549+N///oexY8di7dq1Ho8/efJk3Hzzzdi9ezfuvvvuMuz58inXhG8pKSlQFAU1a9YEAGzZsgVffvklmjdvjvvvv9+nDaSL8Faj4uhdYVAhIh/LK7Si+Qsr/PLY+16+AaEG38xT2rRpU+zatavY8uPHj6NRo0bo1q0bFEVBnTp1nPetWrUKW7Zswf79+9G4cWMAQP369T1+v7CwENOnT0ebNm1Kffw2bdrgueeeAwA8/fTTeP3111G1alXcd999AIAXXngBM2bMwK5du3DNNdd43UZhYSFmzpyJBg0aAAAmTJiAl19+2Xl/nz59PNb/+OOPERMTg/Xr12PgwIGoVq0aANdU9w5vv/02xowZgwceeAAA8Oijj2Lz5s14++230bt3b+d6I0aMqNCA4lCuHpURI0Y4k1VaWhquu+46bNmyBc8884zHTqJKwBoVIqJLJoTwWu8xZswYJCcno0mTJpg0aRJ+/fVX533JycmoWbOmM6R4YzAY0Lp164s+vvs6Wq0WsbGxaNWqlXNZ9erVAcA5Db03oaGhzpACyKnq3ddPT0/HuHHj0LhxY0RFRSEqKgrZ2dk4fvx4qW3bv38/unbt6rGsa9eu2L9/v8eyjh07lrodXylXNN2zZw+uuuoqAMCiRYvQsmVLbNy4Eb/++ivGjRuHF154waeNpFJwHhUiqkQhei32vXyD3x7bV/bv34969eoVW96+fXscOXIEy5Ytw6pVq3D77bejb9+++Pbbb53TwpfaxpCQMhW86vV6j58dZya5/wyg1AtBetuG+/DYmDFjcPr0aUydOhV16tSB0WhE586dy3QV7KLPwVuwCwsLu+h2fKFcQaWwsNBZIb1q1SrcdNNNAGRXmnshD1UCzqNCRJVIURSfDb/4y5o1a7B792488sgjXu+PjIzEsGHDMGzYMAwdOhT9+vXDuXPn0Lp1a5w4cQIHDx4stVclUPz++++YPn06BgwYAECWbZw5c8ZjHb1e77xMgkOzZs2wYcMGjBo1yrls06ZNaNasWcU32otyvdpatGiBmTNn4sYbb8TKlSvxyiuvAABSU1MRGxvr0wbSRTCoEBGVyGw2Iy0tDVarFadOncLy5cuRlJSEgQMHenwQO7z33ntISEhA27ZtodFo8M033yA+Ph7R0dHo2bMnevTogSFDhuDdd99Fw4YN8ffff0NRFPTr188Pz650DRs2xLx589CxY0dkZmbi8ccfL9YrVLduXaxevRpdu3aF0WhETEwMHn/8cdx+++1o3749rr32Wvz4449YsmQJVq1a5ZfnUa4alTfeeAMff/wxevXqheHDhzuLhpYuXeocEqJK4qxR4YRvRERFLV++HAkJCahbty769euHtWvXYtq0afjhhx+8XtE3PDwcb7zxBjp27IhOnTrh6NGj+OWXX6DRyI/LxYsXo1OnThg+fDiaN2+OJ554oliPRKCYPXs2zp8/j3bt2uGuu+7CpEmTip2F9M4772DlypWoVasW2rVrBwAYPHgw3n//fbz11lto0aIFPv74Y8yZMwe9evXyw7MAFFHO872sVisyMzMRExPjXHb06FGEhoaWeDqWr2VmZiIqKgoZGRmIjIyslMcMOLP7Acf/AG6fBzSXQ3CYOxA4+jswZBbQaqh/20dEqpafn48jR46gXr16MJlM/m4OqUhpr51L+fwuV49KXl4ezGazM6QcO3YMU6dOxYEDByotpJBdqUM/gZnyiYiIyqpcQeXmm2/GF198AUDO0nf11VfjnXfeweDBgzFjxowyb2fGjBlo3bo1IiMjERkZic6dO3vMxEdlwBoVIiIKYuUKKjt27ED37t0BAN9++y2qV6+OY8eO4YsvvsC0adPKvJ2aNWvi9ddfx7Zt27Bt2zb06dMHN998M/bu3VueZl2ZeFFCIiIKYuU66yc3NxcREREAgF9//RW33norNBoNrrnmGhw7dqzM2xk0aJDHz1OmTMGMGTOwefNmtGjRojxNu/J4nfCNM9MSEVFwKFePSsOGDfH9998jJSUFK1aswPXXXw9AzoJX3qJWq9WKhQsXIicnB507d/a6jtlsRmZmpsftilfa0I8oeaIgIiIiNShXUHnhhRfw2GOPoW7durjqqqucweLXX391nt5UVrt370Z4eDiMRiPGjRuH7777Ds2bN/e6blJSknMa4KioKNSqVas8zQ8urFEhIqIgVq6gMnToUBw/fhzbtm3DihWui1Nde+21eO+99y5pW02aNEFycjI2b96M8ePHY/To0di3b5/XdZ9++mlkZGQ4bykpKeVpfnBhUCEiNcs9B5iz/d0KCmDlngc5Pj4e8fHxOHHiBBRFQY0aNco12ZvBYEDDhg0ByAscbd26Fe+//z4+/vjjYusajUbn1P1kxwnfiEitLAXAhWOARg/Et/R3ayhAlatHxWaz4eWXX0ZUVBTq1KmD2rVrIzo6Gq+88kqpF1AqCyEEzGbzZW3jilLqRQkZVIgogDneo/heRaUoV1B59tln8eGHH+L111/Hzp07sWPHDrz22mv44IMP8Pzzz5d5O8888wx+//13HD16FLt378azzz6LdevWYeTIkeVp1pWJE74RkWoJ19fyTZLuUy+++CLatm3r72aU25gxYzB48OAyr3/06FEoioLk5OQKa5MvlGvo5/PPP8dnn33mvGoyALRp0wY1atTAAw88gClTppRpO6dOncJdd92FkydPIioqCq1bt8by5ctx3XXXladZVybOo0JEauV+ZqKwAUrxa+9crk2bNqF79+647rrrsHz5cp9vP5C8//77KOdVcQJauYLKuXPn0LRp02LLmzZtinPnzpV5O7NmzSrPw5M7x1AbgwoRqU3RoALfB5XZs2dj4sSJ+Oyzz3D8+HHUrl3b549xKQoKCmAwGHy6TavVCkVREBUV5dPtBopyDf20adMGH374YbHlH374IVq3bn3ZjaJLwBoVIlKrYkHFt3JycrBo0SKMHz8eAwcOxNy5cz3uf/3111G9enVERETgnnvuQX5+vvO+FStWwGQy4cKFCx6/M2nSJPTs2dP586ZNm9CjRw+EhISgVq1amDRpEnJycpz3161bF6+++irGjBmDqKgo3HfffSgoKMCECROQkJAAk8mEunXrIikpyfk77777Llq1aoWwsDDUqlULDzzwALKzXWdGzZ07F9HR0fjpp5/QvHlzGI1GHDt2rNjQz/Lly9GtWzdER0cjNjYWAwcOxOHDhy9zr1a+cgWVN998E7Nnz0bz5s1xzz334N5770Xz5s0xd+5cvP32275uI5WGNSpEVJmEAApyfHMzZwOFefJmzrr4+pc4rPH111+jSZMmaNKkCe68807MmTPHOTSyaNEiTJ48GVOmTMG2bduQkJCA6dOnO3+3b9++iI6OxuLFi53LrFYrFi1a5Kyj3L17N2644Qbceuut2LVrF77++mts2LABEyZM8GjHW2+9hZYtW2L79u14/vnnMW3aNCxduhSLFi3CgQMHMH/+fNStW9e5vkajwbRp07Bnzx58/vnnWLNmDZ544gmPbebm5iIpKQmfffYZ9u7d6/WCwDk5OXj00UexdetWrF69GhqNBrfccstln/RS2co19NOzZ08cPHgQH330Ef7++28IIXDrrbfi/vvvx4svvui8DhBVAs6jQkSVqTAXeC3RP4/9TCpgCCvz6rNmzcKdd94JAOjXrx+ys7OxevVq9O3bF1OnTsXdd9+Ne++9FwDw6quvYtWqVc5eFa1Wi2HDhuHLL7/EPffcAwBYvXo1zp8/j9tuuw2ADCAjRozAww8/DABo1KgRpk2bhp49e2LGjBkwmUwAgD59+uCxxx5ztuv48eNo1KgRunXrBkVRUKdOHY92O7YHAPXq1cMrr7yC8ePHewSpwsJCTJ8+HW3atCnx+Q8ZMqTY/oiLi8O+ffvQsqV6TgcvV48KACQmJmLKlClYvHgxlixZgldffRXnz5/H559/7sv2UWlsNjir5hlUiIicDhw4gC1btuCOO+4AAOh0OgwbNgyzZ88GAOzfv7/Y5VqK/jxy5EisW7cOqampAIAFCxZgwIABiImJAQBs374dc+fORXh4uPN2ww03wGaz4ciRI87tdOzY0WO7Y8aMQXJyMpo0aYJJkybh119/9bh/7dq1uO6661CjRg1ERERg1KhROHv2rMeQksFguGipxeHDhzFixAjUr18fkZGRqFevHgAZlNSk3BO+UQBwDyIexbSsUSGiCqIPlT0bvpB9CshKk9/H1AdMERd/7DKaNWsWLBYLatSo4VwmhIBer8f58+fLtI2rrroKDRo0wMKFCzF+/Hh89913mDNnjvN+m82G//3vf5g0aVKx33Uv2g0L8+wFat++PY4cOYJly5Zh1apVuP3229G3b198++23OHbsGAYMGIBx48bhlVdeQZUqVbBhwwbcc889KCwsdG4jJCQEiqKU2v5BgwahVq1a+PTTT5GYmAibzYaWLVuioKCgTM8/UDCoqJlHUGGNChFVAkW5pOGXUulMgD5Efq83+Wy7FosFX3zxBd555x3nRXMdhgwZggULFqBZs2bYvHkzRo0a5bxv8+bNxbY1YsQILFiwADVr1oRGo8GNN97ovK99+/bYu3evc3b1SxEZGYlhw4Zh2LBhGDp0KPr164dz585h27ZtsFgseOedd6DRyEGPRYsWXfL2z549i/379+Pjjz92lmNs2LDhkrcTCBhU1OyiQYU9KkQUwCrorJ+ffvoJ58+fxz333FPslN2hQ4di1qxZeOqppzB69Gh07NgR3bp1w4IFC7B3717Ur1/fY/2RI0fipZdewpQpUzB06FBn3QkAPPnkk7jmmmvw4IMP4r777kNYWBj279+PlStX4oMPPiixfe+99x4SEhLQtm1baDQafPPNN4iPj0d0dDQaNGgAi8WCDz74AIMGDcLGjRsxc+bMS94HMTExiI2NxSeffIKEhAQcP34cTz311CVvJxBcUlC59dZbS72/6GlcVMFKDCoc+iEiFaigoDJr1iz07dvX67wiQ4YMwWuvvYZGjRrhhRdewJNPPon8/HwMGTIE48eP97jQLiALZDt16oStW7di6tSpHve1bt0a69evx7PPPovu3btDCIEGDRpg2LBhpbYvPDwcb7zxBg4dOgStVotOnTrhl19+gUajQdu2bfHuu+/ijTfewNNPP40ePXogKSnJo+enLDQaDRYuXIhJkyahZcuWaNKkCaZNm4ZevXpd0nYCgSIuYRq7sWPHlmk99zG8ipSZmYmoqChkZGQgMjKyUh4zoGSfBt62dzlOviC7ZAFgy6fAL48BzW8Gbv/Cb80jIvXLz8/HkSNHUK9ePY/eBJ84fwzIs08SGlkDCC9+ii2pV2mvnUv5/L6kHpXKCiBURu6nJrsXVTl7VFijQkQBTLi9RwXh1O/kG+U+PZkCgLc5VNx/5tAPEQUy93BSATPTUnBgUFEzBhUiUrMKnkKfggODipo5hnY0RS7kxaBCRGrAoEJlwKCiZiX2qLBGhYh86xLOu7iEjTKoBDNfvWYYVNTsokM/DCpEdHm0WnngUyGzmbJGJajl5uYCAPR6/WVthxO+qRlrVIiogul0OoSGhuL06dPQ6/XO2VJ9osDiCivmQsB+QUBSNyEEcnNzkZ6ejujoaGfYLS8GFTVjjQoRVTBFUZCQkIAjR47g2LFjvt14xilXT4ouC7jAXuBgEh0djfj4+MveDoOKmjnmICixRoVBhYgun8FgQKNGjXw//DN9JGCzbzO+DTB0lm+3T36j1+svuyfFgUFFzRxBRCmpR4VHJ0TkGxqNxrcz09qsQOZh18/Z0YCvZ76loMBiWjVjjQoRqZWlSD1KIetTyDsGFTVjUCEitSrMK/1nIjsGFTVjMS0RqVWxoJLrn3ZQwGNQUTNO+EZEalU0qBQdCiKyY1BRMw79EJFaWexBxXEyQGEur6BMXjGoqBmDChGplaNHJbSKaxl7VcgLBhU1Y40KEamVI6iEVCm+jMgNg4qasUaFiNTKEUqM4YBG77mMyA2Dippx6IeI1MpRo6IPlTeAQYW8YlBRMwYVIlIrRyjRmQB9iPzewqBCxTGoqJkzqLBGhYhUxhFU9CGA3uS5jMgNg4qa2Uq6KKHjZwHYbJXaJCKiMvEIKo6hH076RsUxqKjZxYpp3dchIgokHkHFPvTD6/2QFwwqanaxGhX3dYiIAomjHkXHHhUqHYOKmpVUo6KwR4WIApx7j4qONSpUMgYVNbtojQoYVIgoMHkb+uFZP+QFg4qalalGhZO+EVEA8lpMy6BCxTGoqFlJPSqK4hr+YY8KEQUii/s8Khz6oZIxqKhZSTUqgCu8CPaoEFEAKuTMtFQ2DCpqVpagwh4VIgpE9lORj2YKZFjs71cMKuQFg4qalTT0476MNSpEFIjspyK/9usRfPvXGY9lRO4YVNSspGJawO0KyuxRIaIAZJE9KtnCgPR8rccyIncMKmpWalDh0A8RBTB770m+MCDbpvdYRuSOQUXNWKNCRGplr0fJhwH5wuCxjMgdg4qalalGhUGFiAKQvZg2HwbkweCxjMgdg4qalalGhcW0RBSA7MM8ecKIPBg9lhG582tQSUpKQqdOnRAREYG4uDgMHjwYBw4c8GeT1IU1KkSkRlYLYCsEAORDj3xw6IdK5tegsn79ejz44IPYvHkzVq5cCYvFguuvvx45OTn+bJZ6sEaFiNTI7Zo+eTC6alR4rR/ywsuheOVZvny5x89z5sxBXFwctm/fjh49evipVSrCGhUiUiO3WhQz9G5DPwwqVJxfg0pRGRkZAIAqVap4vd9sNsNsNjt/zszMrJR2BSzWqBCRGtlrUSwaIwQ0yIfj9GQGFSouYIpphRB49NFH0a1bN7Rs2dLrOklJSYiKinLeatWqVcmtDDCsUSEiNbJP7FaokT0peYLFtFSygAkqEyZMwK5du/DVV1+VuM7TTz+NjIwM5y0lJaUSWxiAGFSISI3sgaRQsQcVRzGtzQJYC/3VKgpQATH0M3HiRCxduhS//fYbatasWeJ6RqMRRqOxElsW4Jw1KiymJSIVsdeoFNiDitkRVAA5/KPV+6NVFKD82qMihMCECROwZMkSrFmzBvXq1fNnc9SHNSpEpEb2HhVXUNFDQLHfxzoV8uTXHpUHH3wQX375JX744QdEREQgLS0NABAVFYWQkBB/Nk0dOPRDRGrkmD5fcfSQK7BoTdBb83iKMhXj1x6VGTNmICMjA7169UJCQoLz9vXXX/uzWeoheHoyEamQvZjWfcjHUa/CHhUqyq89KkIIfz68+nHCNyJSI+f0+W5BRWPyuI/IIWDO+qFycNSfKN6CiqNGhUGFiAKM2wUJHQqcPSq8MCF5YlBRszLVqLCYlogCjL3XJFe4BxVe74e8Y1BRMxbTEpEa2WtUcmyuoGJWOOkbecegomY8PZmI1MjZo+KaLyXfcb0fC4d+yBODipqVOuEba1SIKEDZ61Cyre5BxTH0wx4V8sSgomasUSEiNbKHkWybK6g4zwBijQoVwaCiZqxRISI1shTvUcllUKESMKioGYMKEamRY2Zat9OT2aNCJWFQUTNelJCI1MgeRtwnfHMOA7FGhYpgUFGzMp31w6BCRAHGS4+K81RlnvVDRTCoqBmLaYlIjRxT6DtOSQaQbbO/Z3Hoh4pgUFErIVijQkTqZCk+hX62lacnk3cMKmolbK7vWaNCRGpiDyP5woAIk3yvynLWqHDohzwxqKiVewBhjQoRqYk9jOTBiJhQ2ZOSL9ijQt4xqKjVRYMKa1SIKEA5i2n1iArR27/n6cnkHYOKWpU5qLBHhYgCjMVxerIRIQYtTHqNq7DWwqBCnhhU1Mq9p4RBhYjUwlrofF/KhwEmvRYmvdZt6IdBhTwxqKiVM6gogMbLn5E1KkQUiNyCSD4MCNFrEKLXIo9DP1QCBhW1Ku3UZPflrFEhokBiDyICCszQI8Teo+Ic+mFQoSIYVNSqzEGFPSpEFEDsNSgWjRGA4hz6MbNHhUrAoKJWzqDiZQ4VgEGFiAKTPYgUamQPigwqGtd1fyx5ckJLIjsGFbUq7YKEAIMKEQUm+zwphYoJgAwqIe5DPwCv90MeGFTU6qJDP45iWtaoEFEAsU/2VqCRPSgh9qDiPp0+h3/IHYOKWrFGhYjUyB5CzHAM/Whg0mthgwZWxTGNPmenJRcGFbViUCEiNbIX05oVGVTkhG+yB9iilcNB7FEhdwwqasUaFSJSI+f0+XKox6STxbQAUKhhUKHiGFTUqqw1KoI1KkQUQBxBRdiHfgyyRgUAChXOpULFMaioFSd8IyI1soeQPMh6FJNO4xz6KXAGFdaokAuDilpdLKgonEKfiAKQvUYl1z5vSohBixCDfL9y1K3w9GRyx6CiVpzwjYjUyN6jkmuz96jotTDq5EeRWXHMTsseFXJhUFErZzEtz/ohIhWxB5Ucm9s8KvYeFUfdCmtUyB2DilpxwjciUiNnUHH1qJh08v2KV1AmbxhU1IrzqBCRGtlDSJbVEVQ0zh4V5/V+GFTIDYOKWjGoEJEaWTx7VEL0rnlUctmjQl4wqKgVJ3wjIjVynp5sn/BN75qZ1lG34ggzRACDinqJixXTskaFiAKQI6g4JnxzCyq5Np3HOkQAg4p6ceiHiNTIbQp9g1YDrUZxzkybbePpyVQcg4paMagQkRpZXEHFUZvi6FHJtjqunswJ38iFQUWtLmXCNyEqp01ERBdT6B5U5PuXo0cly+oY+mGPCrkwqKiVo/ZEKSmouC0XtopvDxFRWdh7S/KEwXlackjRYlrWqJAbBhW1KuvQj/u6RET+Zu8tyYfBOdGb0T4ElAde64eKY1BRKwYVIlIjewjJgxEme4+KUaeBosjwAoBDP+SBQUWtylqj4r4uEZE/CeHqUREGmOwXI1QUBSad1i2ocOiHXBhU1KqsFyV0X5eIyJ+sBc6auXy4alQAIMSgdc6twqBC7hhU1OqiQz8aAIrnukRE/uQWQPJhcBbRAoBJp2GPCnnl16Dy22+/YdCgQUhMTISiKPj+++/92Rx1uVhQcb+PQYWIAoE9gNigQQF0ztOTAcBk0PLqyeSVX4NKTk4O2rRpgw8//NCfzVCni9WoAJUbVC6kANvnAlaGIiIqgX2yN4vWBEDxDCo6t6EfXuuH3JRyOF7x+vfvj/79+/uzCep1sRoV9/sqI6gsfwr4+yegIAfo/GDFPx4RqY+9p6RQcVznx3WsHGLQIh/2mWltFsBaCGj1ld5ECjyqqlExm83IzMz0uF2xyjT047gwYSVM+Pbfdvk1+auKfywiUif7ZG+FGhlUPGpU9BrkO+ZRAXiKMjmpKqgkJSUhKirKeatVq5a/m+Q/gVSjkn0ayDopvz+1Gzi1t2Ifj4jUyR4+ChTXlZMdQvRaFEAHm+Njidf7ITtVBZWnn34aGRkZzltKSoq/m+Q/gRRUTu32/HnX1xX7eOQiBPDNWGDerYClwN+tISqdfbI3s1K8R8Wo1wJQYNU6TlFmjwpJqgoqRqMRkZGRHrcrViAV06bZg0pIFfl11zeVP3dL2m7g3JHKfcxAkLYL2LsEOLwa2L/U360hKp09fJjtZ/d41KjYQ4tFY7Kvy4JaklQVVMhNmYpptZ7rVhRHULnqPsAUDWSlAkd/L77e6QPAuteBfB/XFh3dAHzcA5gzoHLqcQLJvh9c32+dVf7tCAGsnAx8/4AsYiSqCPbhnHxnUPGsUQGAAkdQ4Zk/ZOfXoJKdnY3k5GQkJycDAI4cOYLk5GQcP37cn81Sh0Aa+nEElRodgRa3yO93LfJcJz8TmD8UWJcErHjGd4+ddx5Y8j8522VWKnB6v++2HeiEAPZ+7/r5+Cbg1L7ybeuvr4CNU4HkBcCOz33RuitDxgnAnOXvVqiHvUclTxQPKo4elUKlAudSOfI7cGKb77dLFcqvQWXbtm1o164d2rVrBwB49NFH0a5dO7zwwgv+bJY6XFKPSgUGlcI84MxB+X18K6DNHfL7fT8ABW5jzMufBjLsAXTnPCBl6+U/thDAz/8HZJ5wLTv+x+VvN1AU5AB/flzymVTp+4BzhwGtEWjQRy7bVo5elYwTwLInXT+vf1M+NpXuwHJgaivgzfrAvFuALZ/K+YSoZG4XJASKnvUjvy9QKmDox1oI/PIE8PlA4LNrgdUvc84nFfFrUOnVqxeEEMVuc+fO9Wez1KGia1SsFuDsYeDAMmDDVPnh5a0KP32f7M0IqwZExAO1rgai6wAF2cCBX+Q6+38CkucDUIBa18hlPz/qfUjKZgXOHJIh5GJ2LQL2LAYULdDYPh/P8c2X/lxLc/4osGYKkP63b7dbmsJ8YPMM4P02wLIngO/HASe2F1/PMezTsC/Q9WH5/V8LL+0IXwjghwcBc6bsEYuuA2SfAjZPL1/bzVnAvqXAtjkyjAZr4Mk9ByydKF/71gLg8Brgl8eAqS2BT3oDJ//ydwt9r8AHxa32HpVcm7ehH/m9o9DWZ8W0WaeAzwcBWz52Lfv9HWDeYCArrfj6OWeD93WrUn6d8I3KSAhZMFm1CRDfUi6rqB6V0weBHyfJeVGsRc4iMUYC14zzXOYY9olvBSj2awu1Hgb89qb80KzXQ24PALpOAjpPBD7sIItAt82WdS0OeReAhSOAYxvlh2af54D6vVzbdXf+qPxgAIBeTwO1OgEHlwHHfNijknsO+OJm+Vgb3gWueQDo+SRgDPdcz2YDctKB8Ore21pWFrMcgln/JpD5n1ym0QO2QuD3t4HhRXpWHEGl+c1yP8c2As4ekmdddbq3bI+59TPg33WALgS45WMgdSew5F5g4zSg4z1AaJWLb+PCcdm7cHCZrBfyeN0oQGwD+fpoNghoOgjQGcrWNps9BOhNZVvfm/xMYMN7so0D3irb8ymLn/9P/s2rNgGGzpbFzAeWASl/Aqk7gM/6Av2S5D4sz2vi3/XAP6tkAA2L9U2by0sIGcp2LQJumubqNS0Pey9JrpATuYUYihfTmp3T6Pvg9OSULcDXdwHZafL965aPZQD68SFZRzezO3Dzh7LH5chv8nZ6P6APk+91nSf47jXjTc4Z2csUVbPsv2OzAsc2yZ7Q7DQgO10eXEQkyNdLeLUKa66/MKiowdbP5IeyRg8MfA9of9cl1qiUsZj24Apg8b3y6BqQH15VG8rtpO6UH0SlBRUHR1A5vAb49m4g9yxQvSXQ+1lAZwT6PC+fz5pXgOaD5T9W1ilg/hDXqc7/bZNHPHW6Adc+D9S+xrV9q0XWpZgzZQ9N90flP7uilcNAF1KA6MucY8daCCwaJUOKPgwozAE2TQN2fwv0ew2o20M+v39WAv+sBnLPyN6IFrfIW0Ib+QHl6CE6mSy3FdccqNPV88Pn/FHZA7FzvtwOAETWAHo+IXuopneWvVOn9gLVW8j70/8GTv8tXxNN+snH6nQvsPxJWVRblg/Is4eBlfZh1r4vyr91lfrApvfl3/X3d4AbppT8+zlngVWT5VCeuyr1gZi6sl4mOw04+4+87f0OCIuTr98OY0v/G+Vnyr//6YPA4I9kGLsUQgC7vwF+fV62AZD7edQPxYPmpdqzWB44KFrglpny4CG+JdD1IfmhsXSS/F/5+f9kcBs0DTBFApkn5ezN+36QHyy3fAzUaF98+8f+ABbcBljNcv2R38qwVxJzlqy7SNkiDzAUDRCZID+4IhLk9+HxMkiHxtovWHoJtn7m+hsvnQjE1ANqX+25jhAyYG+fA9ToADToDdTvLV8L7q9De1DJscmg4q1HxXW9Hy+9GkIAJ7bK9hzbJF9PMXXttzry/+3CMRlMzx+T69oKgWpNgWHzgaqN5HYS2gCLRgPpe4Evby/+OIU58vW/5VM50/Y1D8i/oS/t/wn4bhxQkAXEtwaa3QQ0vwmo1qTk3/lnlXxNp5dQi7ZjHtDrKXkA6JjV11Igzwjc8bk8M9IQLv8HjBHyFhYne8MdrxVjFAAhewsdt9CqQLXGvn3+l0ARoix97IEpMzMTUVFRyMjICN5TlXPPAdPaAfkXXMuueRA4tQc4sh645ROgzTDvv/tZX/mPesdXQNMBJT+GEPKoc/XLAARQuzNw04fyTUajAc78I3tBNHrgiX89/2FnXS+PIofMAloNdS3/tI9rtlqtAbhvrWdv0Ce9ZK9K2zuBHv8nx/jPH5X/NLd+IkPTtlmuo/PQqq5pta0F8s3HEAGM3yDfpADZ5Z66A7j1M6D1bWXexV73x0+PyDddQzhw7yr5prfsCfkmWBYxdeWHQ9pu72+4jsBy4RhwaCUA+79hRALQZRLQ8W5XT8I3Y+SHfMsh8ugdkB8Ka6cAjW4ARtoLl/MuAO82k0eMY5cBdbqU3D5ztgyGKZuBut2BUUtdH2CHVgELhsi/28QdxQOFzQbs/AJY9aIsZgaA2l1kYGrcX34YOD6cstPlPji2SQYxR2hQNLKHZeDU4kes1kL5Qf3vWteyXs/I4FaW3om0PcAvj8viYkC+jvPOy1uDPsDwr4v36hzdKMNFZqq8ZfwH5JwG6nYD+k52BfGsU8D0q+W2ejwB9Hm2+OMLAfzxkQxxNov8YI+Itw9Lur3dGqOAUd/JD3aHs4dlDUXeeXmAYLPIcDH8a9lr6JB7TvZI7vteBlhRxrPdFK0cpm02CLj+FUAfUvr6J7YBs/vJ/7eYesD5I/L371vrel3YbMCKp4E/Zxb//ejasle0Xk/5Olv/BrBtFmZiKF7PvxWr/68nGlSTwfG7nSfwyNd/4csqn6JL7lpAZ5I9q3U6y8B++m/5QXzmQNmeq0Pzm4GbP5Ifyu4KcuX/9K6v5f9rvR7yVqebrHVb+5oMMgAQEiMPtDreXfpwu4OlQP59LPlAuzuBsKqu+2xWeVLBb295/93YRkD9nvI517pa7sP0/cCvz8meOwAwRQGJ7WX4jKgu3x/3fOsacqzaBOj9jDxA2jlfvpYvR6vbgCGfXd42iriUz28GlUD306PyAzuuBdBsoPxHd1c0ILib3U/+ww2dA7S81fs6BTnyKGnPYvlzh7FA/zeLv5F/0EEeFd8213Vmj80GJNWUH8QPbvE8EvjzE2DZ4/L7616WR5vuUrYCs/rK70OqAHnn5JvFXd/JDxZAdm3+9pb8Rys6fKUPlW8+7s9r+TPA5o9kb8LAd4s/12ObZHjKOy8/1PPOy9BTv5d8To43ky2f2oeVFGD4QvkBDMijwQ1TZaizmmUvUcNrgYbXyQ+yf9fKQHHwV89TK/Wh8oipSj35RuLtaKhBH/km2Lg/oC3SS3ZyF/Bxd/nhPmGbPLqe0VWG1ZunA+1GutZdOkkeObmHmrzzsk3/bZeFz2cOuoaWDBHA+I3yaNRBCDmmf/R3oO1IYLC9XiU/EzixBVibJHu8ALkPbnzHs8erJNZC2TO09TPZxQ7ID78Ri1xHa46ameQFsier+U1yOAyQf6ObpwOG0JIfI3WnDOg2i+wR7PEY0GWiDEuf3yRfqy1ulW+6Gq0MoL8+C+z/sZSGK0Dr2+UH1fKn5HOIbwXcu6b0YayUrcC3Y4EMtwLbmp3kkfOBX+T/pjFKvuZrdpDDAJ/1lWGgRgf59/tmjHxOOpNsc1xzWT+U/KVnDUdUbdnLUfMq+frJSpOBK+uk/D77lNy+e1BKaAPcPs/zb+8u95wcGsk8Ids8eAYw+wb5uotvBdy9QrZr6SR7DRpkbykEcHidPICxFTnVXWsErGa8ZR2OjwoHYdNTfZAYLcPSst0nMX7BDvwv/hCeLvjA1btYlC4EaDFY9sYWZMv9df6o/FtqtLJnM6aO/Brb0NW7WRKb1Xv4sNlkEFyX5DphILG97NVObFvy9o7/Cfz0sOv/XGuUB05XjweiagCL75M9sYBc1u1h4NCv8jV4eG3xfRZeXQYNYZMHi1fdL1/XRQO+zSrfK1e/JHuxPbYRD3QYLevZCvNkL1xBNpCfIQ8msk7KW+ZJuVzRyH2maAAoQON+sifZhxhUgsXJXcAnPeULdPRPQL3u8oPwu/GuD0L34FDUvFtlAteHygDSZaLs2gPsww2z5RFK3jl59Nb/TaDTPd63teJZ4I8PgTbDZXc3II/+Pmgv3zie+c/znz33nOxtqdYEuP0L728EP0xwdSlXbwXc+a088iwqO13+o2r0sjtTa5C9OkWPkPYtBRbdJUPdA5s87zt7GPiwEyBKGAZTtDKw1L5GzvUirN4DFiBDjsUsj2S8MWfL/V6YL98kqzbyfP45Z2UdzvE/5BFt25Gld+0DwILbgUMrgHZ3yXHoDzvIv9ljhzzfsByhRqOXPQGHVsrhB2/POyIBGPC2DMBFndgmj+wVjTyaOvmXnAfH8UFnCJcf3FfdXzxYlcXJv4Cv75Rd9MYo4LY5MvStTQLWvy7/HsMXAo2vB3Z8IQO7rRBIaCuXO17HRf38GLD1U9kreOunnr1B/6wGvhwmt9PxbtlTsXGaDJ2KVg5ZVm8BRCbKoTedAdj0gSvEK1q5HzV64H/rXcNwpck9J4cMw6vLD/uoGnK5OUv2Gh3/Q9ZODP8KWPWSDILRdWQvXnicPJD4Zqz828PxYWv/G8S3kh90DXrLNl+M1SI//E9sk3VjuWdlT8GQz+QHmDubDfjyNjnUUKUBcP86+T934bjsucw9I5+PopEf5opWBlr3+hVztnydO2o/0nY72/5/BeOw2NYDO56/DlXCZNhbeyAdY+dsRasaUfhxQlc5ZHp8k+yJStkiX+dtR8gQboq6+PP1FZtVvleuflkONysae1h4QrbJEYLyLsiQsG2OfJ6hsbI3JHWna1vGKMCcIQPeoGnFe8PzM2RYSdkieztP/uU6SGt2kxyivdh7Rd55YN0bwO5F8jXS8R6gSf+Au8Ajg0owEEJOYHZ8kwwit8113XfyL+CrEfJI597VQM2O3rfx33Y5Tu74R9Ea5BtJ1imZ4B1veFG1Zfio27Xk9hz5XZ7aFxorPxw1WmDPEnnEWKMDcN+aS3+OOWflEENEonyTC4m+9G24y04H3m4EQAGePCLfhB1WvgBsfF92qzboLe8LiZFHF/t+kF2k7toMl0eQl1Mc60spW4BZ18kPyfajZC9bg2uBu5YUX/ez6+QHnru45rLXplpToGpjGZ4uViT49Z3FexqiagMNeskC5rJ8OJYm54x8jON/yA+6VkNdl18YOBXoONa17tGNct28c0CTAcULiwH5P/NeS/l/MWIR0PiG4uvsWQx8ew88ehbq9QD6vQFUb+69nak75WR4R9bLn6+dLOuiLpc52x5W3EK1KQq4Z5VnPYDVInv4ts+RPze6XhZ51utR/tfnhRRZg5W6A4Ain09iexl+NTrZO/jHh/ID9d7VrmFbQAaHuQNdR/4avQyazQaV/pi554BjG2HOSEer72NRAD32v9wPIQYZ4v84fBbDP92MhnHhWPVoz/I9r4qUlSYP2PZ861qm0bneS3LPunoy2t4ph9ZCYuTw++YZ8n1GWOX/0B3z5UHMxRTmAanJ8nVR0utTpRhUgsHub4HF98jeiglbi9cJ5GfK+gb3IlZvhJBH97+94/mGCMgPuk73yjf0i427WguBtxrIxH/3CtnzsOoleTZMhzHAoPcv+SlWiGnt5dwiI76RR+OAHC9+t5k8Chy2wHsPwpl/5IfY/qXyKGjonMs726QizB3oOePvoGmyO7eof1bLo/C4pkDTgUDTGy9+FOZNZqo8NTuiuqwVqNlRHuX7ksUM/Pgw8NeXrmXd/w+41stcSmm7gZndvPckAa7eJF2IDKol1V9s+VTWsETXAq6fIj9gL/aBL4QMKheOyx6wstQplIU5WxZzHtsoP/BHfS/rYkp6/IhE3xU1FubL4uvtc0tep+jQosOOecDSCXJf3zG/eI9MKc5mm9Hh1VUAgH9fGwCNRu775JQLGPzRRtSIDsHGp/pcyjOpXIfXynmHvNXKVGkADJoqQ2RRGf/JM+ya9K/YM4lU4lI+v3nWT2U7vll2xXd9qOSuuIIcWdkNyDdtb2dHmCIvHlIA+QbcsK+8Hd0oh1pCY2XX96V8eGn1shZjz7fAweUyqHg748ffaneWQeX4H66g8vdPMqSEx8uxVm+qNgR6PSlvgarHY66gomhlCPGm4bXA0z6Y3TkyUZ5xU5F0RtmbVq2JLA5uM9xe5+BFfCt5S9stA2WHMZ73H1gmvzboU3qR6FX3yfAWWrXsp0orihwa9DVjODDyG2DTh7Jo1FtIqajH15vkAUbtzrImyGKWwww2ixz6aX2b95ACyDO3YhvIIa1LDMF5hXIY0qjTOEMK4JpC32wp41mK/tKgNzBhiyzGdRRp552TB3N1upZ8gBNVo+T9SaViUKlsPz4sz9MPiSm5HmTDe3I6+Og6sq7EV+p2LX1452Ia95NB5cByOVbqDCqtfdI8n6h9jSzsc5/4zXHE2P6u8tVTBIp6PWXPxn/b5Aeav+fX8BVFkQWF14yXwaU0LYfI192excWDykF7UGnS/+KPebnDVr5kCPNvQG5zR/nmRintrLJS5BfKM5TcT00GXPOo5BUEeFBxMITKm6PuiCoML0pYmczZ8hQ7QBZceRt1y88ENtuLVa9/NbCGHxr1lUfyp/fLgrzsNABK2YoKK0vtzvLrf9vlEeLZw/baAkXWdqiZogAD3pRnjvR8wt+t8b2LhRRAnrEDyF7JrFOu5Zmp9losxXttCgWMfHuPSkiRoOIILvkWG1RckUAVgEGlMrlVvePUbtc8I+6SF8gJgKo2Lrlr319CYlxBYP2b8mtsQ3lEGChiG8gufatZfnA5LrDX6DpZe6J2NTrIM0JKGiIIdjF1ZFATNs8rRx9cLr9WRB0N+ZQjqDiGehwcQcVqEyi0MqiQC4NKZSp6/Y+tszx/tllldTggu8EvdQbJyuCYU+TQCvk1kOpTANnr4JjT48hvwM4F8vuiwwSkXi2HyK+O04YBORwJlG3Yh/wqzxlUivaouN7v8gO9ToUqVQB+EgYxxymwjgr5vUvkKXsOB36RZ/KExACtL+N6GhWpaDFqoAUVwNXrs+lDWUQbkSBncKXg0HwwAEXOM3EhRRaf/7tO3teklBmYKSA4alCKBhWDVgNHbW2+WupUqFIwqFSm1GT5tdO9coIzS768cJ+Dozelw9jSZ9/0p6qN5Cl4DoFUSOvgCCrmDPm1ncqLaMlTZII8uwKQEyAeXiuH+qLryHliKKDlW2QxbdEaFUVRXHUqhWW8JABdERhUKktBruu8+4S2rsmsts2WRbWpyfa5FHSeVxQORO7d64HYo5LQWs7vAMhZJNVeREvFOS6dsHeJ67TkJgMCZ4I+KpGjt8Qx0Zs755k/hexRIRcGlcpyao8sAAyvLo8IW98upyE/e0iewbDZfj2VFrcE1qmT3ji61yNrlDyNvD9p9a7Zehted/lXUqbA0/xmeQZa6k45hTvA+hSVcNSfFC2mlcscPSoMKuTCoFJZHIW0jmmTjREyrADyQoN77FOhX/NA5bftUtXtKqeXHzrH3y0p2TXj5TBAr6f83RKqCGFVXROgFWTLa6iUc14Pqlwl1ajIZfIjiT0q5I5BpbI46lMS2rqWdbxbfj36u7xuRq1rgBrtK7tl5dN2hLxaa6BqeiPw4J/q2Z906Rxn/wByjp8Au+gaeVfShG/uy9ijQu4YVCpL0R4VQNZ31Ozk+rmzCnpTiAJF0xvlhTYBnu2jInklTPjmvoxBhdwxqFSGwnw5mysAJLb1vK+jfRr96NpAkxsrtVlEqhYSLS/l0OIWGVpIFUqa8A1wFdjyrB9yx3M2K8OpvfJCX6GxsgDVXethgLVA9qzwFFqiS9P5QX+3gC5RSVPoA4BRx7N+qDh+MlYGx0RvCW2Lnz6p0QAdRld2i4iI/CK/hJlpAfceFQYVcuHQT2VwBJWiwz5ERFeYkqbQBwCTjmf9UHEMKpXBWyEtEdEVqLSzfpw9KpxCn9wwqFQ0ixk4tU9+735qMhHRFai0s36cpydbWExLLgwqFS19v5wjxRQtz+whIrqCmR1BxVDyzLR57FEhNwwqJdh14gKyzZbL35B7fQqvQ0JEVzhnjYqu5JlpWUxL7hhUvFi0NQW3Tt+EZ5bshhDi8jbmnJGW9SlERM4aFV6UkMqIQcWLBnFhEACW/pWKr7akXN7GnIW0bS+3WUREqld6jwonfKPiGFS86FCnCp64oQkA4MUf92Jfamb5NmQtlJO9AexRISKC64yekFJ6VDj0Q+4YVEpwX/f66NM0DgUWGx78ckexepWMvELM23wMi7am4K+UC8gt8FLPkrYLsJrllV2r1K+klhMRBa58S8lT6LNGhbzhzLTeWC3Q2Cx457Y2uHHa7zhyJgfPLNmN9+9oC7PFhnl/HMOHa/9BRl6h81cUBahdJRQNqoUjLsKI2sYcjNx3P6IAFNa6BnoW0hLRFc5itaHQKuv+Sjs9mTUq5I5BxZsTW4EvbkJMYjssadQWk3eG4/e/MvGMUYv1B04jNSMfANAwLhzVI404kJaFM9kFOHY2F8fO5iICuVhoeAVRmuM4IapiQspQvHs6G/Wrhfv5iRER+Y/7/CheZ6bl0A95waDiTepOeaHAlD8Rn/InPtYD0AN/J9dCnPUq/BnRE7fe0AdD2teEViN7Ss5km3EwLQvH08+i++b/oUbmMWRqY/CI5kUkZ4Rh6Mw/MHdsJ7SuGe3PZ0ZE5Dfu86MYdV6unsxiWvKCQcWba8YDTfoBxzcDxzdDHN8M5cwBNNWkoKkmBShcDGxtBeTdIs/mCauGqmHVULV2FLr8+QCQuRMwRiFyzFLMiGiCMXO2YM9/mRj+yWZ8Mqojujas6u9nSERU6VwXJNRA8TIczh4V8oZBxRtFkcWvVeoDbUdAAYCcs8A/K4E9i4HDa4BTu+XNG10IMOJrIKE1qgL46r5r8L9527Hp8FmMnbMVzw9shhCDDsfO5uDo2VycOJ+L+EgTOtatgo51YtA8MRJ6LeuciSi45Jcyfb77ctaokDsGlbIKiwXa3CFvueeA/T8CB34BLqQAuWeAnDOAsAKGcGDoHKBOZ+evRpj0mDO2Ex75Ohm/7E7D8z/s9foQy/akAZD/rFfVq4Kn+jdFs4TIizYtI68QC/48hq+3pqBjnSp47daWMHqZo+BK8vmmo3jlp334dFRH9G4a5+/mEBFKvyAhAJgMrrN+hBBee13oysOgUh6hVYAOo+XNwWYD8i8AOhNgCC32K0adFh8Mb4+6sQewYm8a4qNMqBMbhrqxoagRHYpj53Kw/eh5bDt2Hhl5hVh/8DQ2HT6DiX0aYXyvBl57WP67kIfZG45g4ZbjyLGP/R47m4vUC3n4eFQHRJr0FbUHAlpegRVTVx2ExSbw1ooD6NWkGt/wiAJAaRckBFwBxiaAAqvtij/gIolBxVc0GhlgSqHVKHiiX1M80a9pievYbAKH0rPxzq8H8Ou+U3h35UGs2JuGt4a2QcO4cOw6cQGb/z2LP4+cwx+Hz8Jik6f6NakegZvaJmLGusP449+zuOPjzZh7dyfERZh8+jTVYPGOEzifK08d33cyExv/OYtujVgXRORvrhqV0od+5LoMKiQxqAQYjUZBk/gIfHxXByz9KxWTl+7F3tRM3PThBui0SrFq+M71Y/G/nvXRs7HsNejZuBrGzNmCfSczMWTGJnxx99WoVzXMT8+m8tlsArM3HAEAJESZcDIjH5/8/i+DClEAyCssebI3ANBrNdBqFFhtAvmFVkSFXJm9wuSJFZsBSlEU3Ny2BlY+0hM3tKgOi00gv9CG2DADBrSKx0s3tcDKR3rgq/uvQa8mcc6hjZY1orB4fBfUrhKKlHN5GDJjE+b9cfSKqaJf/Xc6/j2Tg0iTDrPHdIJGAX47eBr7T5bzMghE5DPOYlov0+c7OAtqC66M9yy6OPaoBLhqEUbMvLMD9p3MhF6rQaO48IvWW9SJDcPi8V0wdq48Lfr5H/bi/dX/4N7u9TDy6tqI8FK7IoTA0bO52PjPGew4dh41Y0Iw4uo6iI9S19DRp7//CwAYcXUdNEuIRP+WCfh590l89vsRvHM7r7dE5E/5pVyQ0MGk1yDb7Jpqn4hBRQUURUGLxKhL+p1qEUZ8O64LFm1Lwcfr/8V/F/Lw+rK/MX3tP+hUtwpCDFqEGrQINeiQbbbgj8Nn8d+FPI9tTF93GP1bJWBMl7poXzsaQgAnM/NxOD0bR87kwCYEIkx6RJh0iDDqUC3CiIZlCFIV5a+UC9hy5Bx0GgVjutQFANzbvR5+3n0SS//6D4/f0ER1wYtKJ4SAxSZ4Or9KOM/6KaVHxcQeFSqCQSWImfRajOpcF8Ovqo0fklMxfd0/+Pd0Dlb/ne51fb1WQbvaMbiqbhVsOXoOW46cw49/peLHv1JRMyYEZ7MLLjq/Qd3YUNzUtgZubpuIBpV8yQBHb8pNbRKdgcT9+czZdARP929WqW2iipOVX4j7vtiGg6eyMfPODriqXunF7OR/eWXqUeHstOSJQeUKoNdqMLRDTdzSrgY2/nMGqRfykFtgRV6hFbkFFmgUBR3qxOCqelUQanC9JPamZmDuxqP44a9UnDifZ9+WgtpVQlG/WjgMOg2y8i3Iyi9EVr4FJ87n4ujZXExbfQjTVh9CqxpRqB0birPZZpzLKcC5nALkF9pQMyYE9auFoX7VcNSvFobYcCMMWg2Meg0MWg1CDFrERRgRbtR57Z2x2QQKrDaPMwdOnM91zkNzb3fPK1Xf36M+thw9hy83H8eE3g29Dn2RuuQWWHD33K3YevQ8AGD07C2YPaYTOjeI9XPLqDSuGpWSe8BCODstFeH3oDJ9+nS89dZbOHnyJFq0aIGpU6eie/fu/m5WUNJqFPRoXK3M67dIjMJbt7XBU/2bYv/JLNSICUGtmBDoSuhmzzFbsHLfKfyQ/B9+O3QGu//LwO7/Moqt93daFv5Oy7ro44catKgeaUK1CCMA4FxOAc7nFOB8bgFsAmhQLQwd6sSgfe0YbDt2HlabQNeGsWie6DlJXp+mcahfLQz/ns7Bwi0puK9HfW8PRyqRX2jFvZ9vw9aj5xFh0qFZfCS2HD2HsXO3YNboTrxERQArW4+Ka9I3IsDPQeXrr7/Gww8/jOnTp6Nr1674+OOP0b9/f+zbtw+1a9f2Z9PITWy4Ed0aGS+6XphRh8HtamBwuxo4m23Gyn2nkFtgRWy4AVXC5M2o0+L4uRz8ezoH/57JwZHTOcjIK0SB1YYCiw1mixW5ZiuyzBbkFlhx5EwOjpzJ8fp4h0/n4PDpHCzadsK57L7uxUOIRqPgvu718fSS3Zjyy358/NthNKgWjkbVw9GgWjgSokyIizQhLsKIuAgTdBoF2QUWZOdbkG224FxOAQ6fzsahU9k4fDobh9OzAcii5bpVw1CvaihqVwlDlTADokL0iAzRISpEjxC9tliPUG6BBakX8pF6IQ8nM/IgBFA90oTqkSbER5kQE6ov9jvZZtlbdeJcHk6cz0V6lhnVI03ysWPDUCMmxHlxzGBntlgxbr68HEWYQYvP774KzRMiMW7+dqw7cBp3z92Kz0Z3RPdGMpAXWGz470IeCq021KsaxlqWClRoteF8bgGqhRtLrFMz24dzSjvrx8Rp9KkIRQgh/PXgV199Ndq3b48ZM2Y4lzVr1gyDBw9GUlLSRX8/MzMTUVFRyMjIQGTkxaeaJ/XIK7DiVGa+vGWZoVGAKqEGxIQZEBtmgKIo2HXiArYfO48dx8/jr5QMdKgTg3n3XOX1TTK/0Io7P/sT246dr9TnodMo0GkV6DXyAzLLbLno+hqNAghAQEAIOCf1K4leqyA2zAidVrE/ngY6jQK9VuN8bJ1WgVajQFEUKJCXswJgD4c25BdaYbbYoACIMOkQGaJHhEmPcKMW2WYrLuQW4EJuoezNsglEh8rgGR2qR0yoAREmHcKMOoQb5VcF8tIOjltmfiHMhTKImi02mAttEBAw6bUI0cvCbpNeK5874GyjVlFgMsh1THotNv97FusOnIZJr8HnY6/C1fXlUI/ZYsX4+Tuw5u90GHQatK0VjRPncnEyMx+OdziDVoPG8eFokRCF5omRiDDpIATg2LuFVhuy7UOZmfkW5Jgt0GkVhOh1CDNqEWLQwqjTQqvI8KtR5K3oy81iE8gxW+w3Obyq12oQbpL7J8Kkg0mvhc0mC4Gt9q82m4BNCFiF/B4AdFoN9FoNDFr5dy0pjyqQ7VAUBXqtAqNOC6NeA5NOC4NOgRBytlchBARQ7LGtNpvzZ5sQsNrkayO/0Ip8ixX5hTZYbTaEGXWIMOkRaZLP478L+diXmoG9qZn4Oy0LBRYbIk06NE+MRIvEKLRIjESVMANsQsBiFfjij2PY8M8ZPH5DEzzYu6HX53LfF9uwct8p3Ne9Hno1iXO2L9tsla8l+y2nwIJQg+s1F2HUwWTQymFkndxveq3ifN5WIZ+fVqPAoNXA4FxH4xwGzyuwIqdATt+v1cj/J61GA60Gzn1os/99NPbtOB7H8fdx/x/T2F8ciuL6Gwm3/23Ha9PxGnKsZxOOv4OATcjlHo+l0aDAKv8u+YVyGN9qLyp3fw9wbMdilc9f2NvveN3bbAKZ+fJ/9EKuvNmEQFSI3vm/HR2qR82YEDSMiyj1fehSXcrnt9+CSkFBAUJDQ/HNN9/glltucS5/6KGHkJycjPXr1xf7HbPZDLPZ7Pw5MzMTtWrVYlChMl8XJMdsweHT2fgnPRuH0rNx9EyOPRCZcTrLjAKrq4BPr1Wcb8p1q4ahUVw4GtpvgIKjZ3Jw9Kzs8Uk5n4dMx4dyXmGp4SLCqENidAgSok3QKArSMmQgO5tTUOLvON4sakaHolqEEWmZ+Th6JgfHzuWiwHJlFR0adBrMHt2p2CR+ZosVE77ciZX7TnksN+k10Gk0yL5ISKTK8+Kg5hjTtZ7X+x5auBM/JKdWcouoNP1bxmPGnR18us1LCSp+G/o5c+YMrFYrqlev7rG8evXqSEtL8/o7SUlJeOmllyqjeaQyZT0lOsyoQ+ua0WhdM7rYfUIIXMgthFUIRJh0F52+u0OdGK/LhRDOYmWLVaDQakOh1QabAOIijSVeg8lsseJsdgEEXD0KChSEGbUlFgDbbAKpGXm4kFuIQqsNVptAoVXAYrM5H9tiE877HEdSjuMTg04Do04Lk15+FUIgs0ivQphRh+gQPWLC9IgKMUCrUXA+11EvVIjzOQXIdvQgFFiQbZZHpJEhekTZb5EmPUINWhh1smjaqNNCgezezyu0Iq/AivxCq/3oVx5xAoDFKpxHjHmFVggB3NGplrMnxZ1Rp8X0ke2xct8pFFptqBkTilpVQlAt3AghgJTzudiXmuk6+reHUse+1mkUey+SztlDZLXZ/5YFFuTY2+jYfzYbYLXvR/fDPY0C51F+uEmHMIMWBVaBrPxC53BiXqHVfrQuj441GsWjp0Zrfz0Xuv0dC60Crv4f99eb/ejYfqRcaPXsJSu0yp4yjXxBQQGg08gZYIvdFPnV0Vtg0mtg0ms9wp6jeD4r34IqYQa0cOs9SYg24Z/0bOxNzcQ++y2nwOJ8rlqNgiphBvRvleD19QwAQ9rXxIG0LBRabc52yp4trfP1FBWiR6hRh/xCK7Ls+zQ7vxD5hTbnMHKB47krsj5PY39uVnsxvmMdi00gRK91TtkQotdCp1VkL4Sjt8t+IKSx95Io9v1daP/bFNj/x917SYSQvSHO3hP730r2tLh6WFx/Q+HstdFqZG+ixr7PbEKg0OJ6LItVwKCTfx9Hb6NGUeT/vU042+54XWvsf1v3HkDH84kw6RAdIntOokMN0CjAhbxCZNh7UC/kFvp9dnO/9aikpqaiRo0a2LRpEzp3dl1peMqUKZg3bx7+/vvvYr/DHhUiIiL1U0WPStWqVaHVaov1nqSnpxfrZXEwGo0wGi9e1ElERETBwW8l8AaDAR06dMDKlSs9lq9cuRJdunTxU6uIiIgokPj19ORHH30Ud911Fzp27IjOnTvjk08+wfHjxzFu3Dh/NouIiIgChF+DyrBhw3D27Fm8/PLLOHnyJFq2bIlffvkFderU8WeziIiIKED4dR6Vy8V5VIiIiNTnUj6/OU0jERERBSwGFSIiIgpYDCpEREQUsBhUiIiIKGAxqBAREVHAYlAhIiKigMWgQkRERAGLQYWIiIgCFoMKERERBSy/TqF/uRyT6mZmZvq5JURERFRWjs/tskyOr+qgkpWVBQCoVauWn1tCRERElyorKwtRUVGlrqPqa/3YbDakpqYiIiICiqL4dNuZmZmoVasWUlJSeB2hCsT9XDm4nysH93Pl4H6uPBW1r4UQyMrKQmJiIjSa0qtQVN2jotFoULNmzQp9jMjISP4jVALu58rB/Vw5uJ8rB/dz5amIfX2xnhQHFtMSERFRwGJQISIiooDFoFICo9GIyZMnw2g0+rspQY37uXJwP1cO7ufKwf1ceQJhX6u6mJaIiIiCG3tUiIiIKGAxqBAREVHAYlAhIiKigMWgQkRERAGLQcWL6dOno169ejCZTOjQoQN+//13fzdJ1ZKSktCpUydEREQgLi4OgwcPxoEDBzzWEULgxRdfRGJiIkJCQtCrVy/s3bvXTy0ODklJSVAUBQ8//LBzGfezb/z333+48847ERsbi9DQULRt2xbbt2933s/97BsWiwXPPfcc6tWrh5CQENSvXx8vv/wybDabcx3u60v322+/YdCgQUhMTISiKPj+++897i/LPjWbzZg4cSKqVq2KsLAw3HTTTThx4kTFNFiQh4ULFwq9Xi8+/fRTsW/fPvHQQw+JsLAwcezYMX83TbVuuOEGMWfOHLFnzx6RnJwsbrzxRlG7dm2RnZ3tXOf1118XERERYvHixWL37t1i2LBhIiEhQWRmZvqx5eq1ZcsWUbduXdG6dWvx0EMPOZdzP1++c+fOiTp16ogxY8aIP//8Uxw5ckSsWrVK/PPPP851uJ9949VXXxWxsbHip59+EkeOHBHffPONCA8PF1OnTnWuw3196X755Rfx7LPPisWLFwsA4rvvvvO4vyz7dNy4caJGjRpi5cqVYseOHaJ3796iTZs2wmKx+Ly9DCpFXHXVVWLcuHEey5o2bSqeeuopP7Uo+KSnpwsAYv369UIIIWw2m4iPjxevv/66c538/HwRFRUlZs6c6a9mqlZWVpZo1KiRWLlypejZs6czqHA/+8aTTz4punXrVuL93M++c+ONN4q7777bY9mtt94q7rzzTiEE97UvFA0qZdmnFy5cEHq9XixcuNC5zn///Sc0Go1Yvny5z9vIoR83BQUF2L59O66//nqP5ddffz02bdrkp1YFn4yMDABAlSpVAABHjhxBWlqax343Go3o2bMn93s5PPjgg7jxxhvRt29fj+Xcz76xdOlSdOzYEbfddhvi4uLQrl07fPrpp877uZ99p1u3bli9ejUOHjwIAPjrr7+wYcMGDBgwAAD3dUUoyz7dvn07CgsLPdZJTExEy5YtK2S/q/qihL525swZWK1WVK9e3WN59erVkZaW5qdWBRchBB599FF069YNLVu2BADnvvW2348dO1bpbVSzhQsXYseOHdi6dWux+7iffePff//FjBkz8Oijj+KZZ57Bli1bMGnSJBiNRowaNYr72YeefPJJZGRkoGnTptBqtbBarZgyZQqGDx8OgK/pilCWfZqWlgaDwYCYmJhi61TEZyWDiheKonj8LIQotozKZ8KECdi1axc2bNhQ7D7u98uTkpKChx56CL/++itMJlOJ63E/Xx6bzYaOHTvitddeAwC0a9cOe/fuxYwZMzBq1CjnetzPl+/rr7/G/Pnz8eWXX6JFixZITk7Gww8/jMTERIwePdq5Hve175Vnn1bUfufQj5uqVatCq9UWS4Tp6enF0iVduokTJ2Lp0qVYu3Ytatas6VweHx8PANzvl2n79u1IT09Hhw4doNPpoNPpsH79ekybNg06nc65L7mfL09CQgKaN2/usaxZs2Y4fvw4AL6efenxxx/HU089hTvuuAOtWrXCXXfdhUceeQRJSUkAuK8rQln2aXx8PAoKCnD+/PkS1/ElBhU3BoMBHTp0wMqVKz2Wr1y5El26dPFTq9RPCIEJEyZgyZIlWLNmDerVq+dxf7169RAfH++x3wsKCrB+/Xru90tw7bXXYvfu3UhOTnbeOnbsiJEjRyI5ORn169fnfvaBrl27Fju9/uDBg6hTpw4Avp59KTc3FxqN58eUVqt1np7Mfe17ZdmnHTp0gF6v91jn5MmT2LNnT8Xsd5+X56qc4/TkWbNmiX379omHH35YhIWFiaNHj/q7aao1fvx4ERUVJdatWydOnjzpvOXm5jrXef3110VUVJRYsmSJ2L17txg+fDhPMfQB97N+hOB+9oUtW7YInU4npkyZIg4dOiQWLFggQkNDxfz5853rcD/7xujRo0WNGjWcpycvWbJEVK1aVTzxxBPOdbivL11WVpbYuXOn2LlzpwAg3n33XbFz507nNBxl2afjxo0TNWvWFKtWrRI7duwQffr04enJlemjjz4SderUEQaDQbRv3955Gi2VDwCvtzlz5jjXsdlsYvLkySI+Pl4YjUbRo0cPsXv3bv81OkgUDSrcz77x448/ipYtWwqj0SiaNm0qPvnkE4/7uZ99IzMzUzz00EOidu3awmQyifr164tnn31WmM1m5zrc15du7dq1Xt+TR48eLYQo2z7Ny8sTEyZMEFWqVBEhISFi4MCB4vjx4xXSXkUIIXzfT0NERER0+VijQkRERAGLQYWIiIgCFoMKERERBSwGFSIiIgpYDCpEREQUsBhUiIiIKGAxqBAREVHAYlAhoqCiKAq+//57fzeDiHyEQYWIfGbMmDFQFKXYrV+/fv5uGhGplM7fDSCi4NKvXz/MmTPHY5nRaPRTa4hI7dijQkQ+ZTQaER8f73GLiYkBIIdlZsyYgf79+yMkJAT16tXDN9984/H7u3fvRp8+fRASEoLY2Fjcf//9yM7O9lhn9uzZaNGiBYxGIxISEjBhwgSP+8+cOYNbbrkFoaGhaNSoEZYuXVqxT5qIKgyDChFVqueffx5DhgzBX3/9hTvvvBPDhw/H/v37AQC5ubno168fYmJisHXrVnzzzTdYtWqVRxCZMWMGHnzwQdx///3YvXs3li5dioYNG3o8xksvvYTbb78du3btwoABAzBy5EicO3euUp8nEflIhVzqkIiuSKNHjxZarVaEhYV53F5++WUhhLyS9rhx4zx+5+qrrxbjx48XQgjxySefiJiYGJGdne28/+effxYajUakpaUJIYRITEwUzz77bIltACCee+4558/Z2dlCURSxbNkynz1PIqo8rFEhIp/q3bs3ZsyY4bGsSpUqzu87d+7scV/nzp2RnJwMANi/fz/atGmDsLAw5/1du3aFzWbDgQMHoCgKUlNTce2115bahtatWzu/DwsLQ0REBNLT08v7lIjIjxhUiMinwsLCig3FXIyiKAAAIYTze2/rhISElGl7er2+2O/abLZLahMRBQbWqBBRpdq8eXOxn5s2bQoAaN68OZKTk5GTk+O8f+PGjdBoNGjcuDEiIiJQt25drF69ulLbTET+wx4VIvIps9mMtLQ0j2U6nQ5Vq1YFAHzzzTfo2LEjunXrhgULFmDLli2YNWsWAGDkyJGYPHkyRo8ejRdffBGnT5/GxIkTcdddd6F69eoAgBdffBHjxo1DXFwc+vfvj6ysLGzcuBETJ06s3CdKRJWCQYWIfGr58uVISEjwWNakSRP8/fffAOQZOQsXLsQDDzyA+Ph4LFiwAM2bNwcAhIaGYsWKFXjooYfQqVMnhIaGYsiQIXj33Xed2xo9ejTy8/Px3nvv4bHHHkPVqlUxdOjQynuCRFSpFCGE8HcjiOjKoCgKvvvuOwwePNjfTSEilWCNChEREQUsBhUiIiIKWKxRIaJKw5FmIrpU7FEhIiKigMWgQkRERAGLQYWIiIgCFoMKERERBSwGFSIiIgpYDCpEREQUsBhUiIiIKGAxqBAREVHAYlAhIiKigPX/vBSb2kSDQ3AAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting the metrics\n",
    "plt.plot(d_loss)\n",
    "plt.plot(d_g_loss)\n",
    "plt.title('Model loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Discriminator', 'Adversarial'], loc='center right')\n",
    "plt.show()\n"
   ]
  }
 ],
 "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": "3.10.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}