dlnd_face_generation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Face Generation\n",
    "\n",
    "In this project, you'll define and train a DCGAN on a dataset of faces. Your goal is to get a generator network to generate *new* images of faces that look as realistic as possible!\n",
    "\n",
    "The project will be broken down into a series of tasks from **loading in data to defining and training adversarial networks**. At the end of the notebook, you'll be able to visualize the results of your trained Generator to see how it performs; your generated samples should look like fairly realistic faces with small amounts of noise.\n",
    "\n",
    "### Get the Data\n",
    "\n",
    "You'll be using the [CelebFaces Attributes Dataset (CelebA)](http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html) to train your adversarial networks.\n",
    "\n",
    "This dataset is more complex than the number datasets (like MNIST or SVHN) you've been working with, and so, you should prepare to define deeper networks and train them for a longer time to get good results. It is suggested that you utilize a GPU for training.\n",
    "\n",
    "### Pre-processed Data\n",
    "\n",
    "Since the project's main focus is on building the GANs, we've done *some* of the pre-processing for you. Each of the CelebA images has been cropped to remove parts of the image that don't include a face, then resized down to 64x64x3 NumPy images. Some sample data is show below.\n",
    "\n",
    "<img src='assets/processed_face_data.png' width=60% />\n",
    "\n",
    "> If you are working locally, you can download this data [by clicking here](https://s3.amazonaws.com/video.udacity-data.com/topher/2018/November/5be7eb6f_processed-celeba-small/processed-celeba-small.zip)\n",
    "\n",
    "This is a zip file that you'll need to extract in the home directory of this notebook for further loading and processing. After extracting the data, you should be left with a directory of data `processed_celeba_small/`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# can comment out after executing\n",
    "#!unzip processed_celeba_small.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dir = 'processed_celeba_small/'\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL\n",
    "\"\"\"\n",
    "import pickle as pkl\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import problem_unittests as tests\n",
    "#import helper\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize the CelebA Data\n",
    "\n",
    "The [CelebA](http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations, you'll only need the images. Note that these are color images with [3 color channels (RGB)](https://en.wikipedia.org/wiki/Channel_(digital_image)#RGB_Images) each.\n",
    "\n",
    "### Pre-process and Load the Data\n",
    "\n",
    "Since the project's main focus is on building the GANs, we've done *some* of the pre-processing for you. Each of the CelebA images has been cropped to remove parts of the image that don't include a face, then resized down to 64x64x3 NumPy images. This *pre-processed* dataset is a smaller subset of the very large CelebA data.\n",
    "\n",
    "> There are a few other steps that you'll need to **transform** this data and create a **DataLoader**.\n",
    "\n",
    "#### Exercise: Complete the following `get_dataloader` function, such that it satisfies these requirements:\n",
    "\n",
    "* Your images should be square, Tensor images of size `image_size x image_size` in the x and y dimension.\n",
    "* Your function should return a DataLoader that shuffles and batches these Tensor images.\n",
    "\n",
    "#### ImageFolder\n",
    "\n",
    "To create a dataset given a directory of images, it's recommended that you use PyTorch's [ImageFolder](https://pytorch.org/docs/stable/torchvision/datasets.html#imagefolder) wrapper, with a root directory `processed_celeba_small/` and data transformation passed in."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# necessary imports\n",
    "import torch\n",
    "from torchvision import datasets\n",
    "from torchvision import transforms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_dataloader(batch_size, image_size, data_dir='processed_celeba_small/'):\n",
    "    \"\"\"\n",
    "    Batch the neural network data using DataLoader\n",
    "    :param batch_size: The size of each batch; the number of images in a batch\n",
    "    :param img_size: The square size of the image data (x, y)\n",
    "    :param data_dir: Directory where image data is located\n",
    "    :return: DataLoader with batched data\n",
    "    \"\"\"\n",
    "    \n",
    "    # TODO: Implement function and return a dataloader\n",
    "    \n",
    "    transform = transforms.Compose([transforms.Resize(image_size),\n",
    "                                    transforms.ToTensor()])\n",
    "    \n",
    "    dataset = datasets.ImageFolder(data_dir, transform)\n",
    "    \n",
    "    dataloader = torch.utils.data.DataLoader(dataset = dataset, batch_size = batch_size, shuffle = True, num_workers = 0)\n",
    "    \n",
    "    return dataloader\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a DataLoader\n",
    "\n",
    "#### Exercise: Create a DataLoader `celeba_train_loader` with appropriate hyperparameters.\n",
    "\n",
    "Call the above function and create a dataloader to view images. \n",
    "* You can decide on any reasonable `batch_size` parameter\n",
    "* Your `image_size` **must be** `32`. Resizing the data to a smaller size will make for faster training, while still creating convincing images of faces!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define function hyperparameters\n",
    "batch_size = 20\n",
    "img_size = 32\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "# Call your function and get a dataloader\n",
    "celeba_train_loader = get_dataloader(batch_size, img_size)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, you can view some images! You should seen square images of somewhat-centered faces.\n",
    "\n",
    "Note: You'll need to convert the Tensor images into a NumPy type and transpose the dimensions to correctly display an image, suggested `imshow` code is below, but it may not be perfect."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f69cbec4cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# helper display function\n",
    "def imshow(img):\n",
    "    npimg = img.numpy()\n",
    "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "# obtain one batch of training images\n",
    "dataiter = iter(celeba_train_loader)\n",
    "images, _ = dataiter.next() # _ for no labels\n",
    "\n",
    "# plot the images in the batch, along with the corresponding labels\n",
    "fig = plt.figure(figsize=(20, 4))\n",
    "plot_size=20\n",
    "for idx in np.arange(plot_size):\n",
    "    ax = fig.add_subplot(2, plot_size/2, idx+1, xticks=[], yticks=[])\n",
    "    imshow(images[idx])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Exercise: Pre-process your image data and scale it to a pixel range of -1 to 1\n",
    "\n",
    "You need to do a bit of pre-processing; you know that the output of a `tanh` activated generator will contain pixel values in a range from -1 to 1, and so, we need to rescale our training images to a range of -1 to 1. (Right now, they are in a range from 0-1.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# TODO: Complete the scale function\n",
    "def scale(x, feature_range=(-1, 1)):\n",
    "    ''' Scale takes in an image x and returns that image, scaled\n",
    "       with a feature_range of pixel values from -1 to 1. \n",
    "       This function assumes that the input x is already scaled from 0-1.'''\n",
    "    # assume x is scaled to (0, 1)\n",
    "    # scale to feature_range and return scaled x\n",
    "    \n",
    "    x = (feature_range[1]-feature_range[0])*x+feature_range[0]\n",
    "    \n",
    "    return x\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min:  tensor(-0.9765)\n",
      "Max:  tensor(0.9373)\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "# check scaled range\n",
    "# should be close to -1 to 1\n",
    "img = images[0]\n",
    "scaled_img = scale(img)\n",
    "\n",
    "print('Min: ', scaled_img.min())\n",
    "print('Max: ', scaled_img.max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# Define the Model\n",
    "\n",
    "A GAN is comprised of two adversarial networks, a discriminator and a generator.\n",
    "\n",
    "## Discriminator\n",
    "\n",
    "Your first task will be to define the discriminator. This is a convolutional classifier like you've built before, only without any maxpooling layers. To deal with this complex data, it's suggested you use a deep network with **normalization**. You are also allowed to create any helper functions that may be useful.\n",
    "\n",
    "#### Exercise: Complete the Discriminator class\n",
    "* The inputs to the discriminator are 32x32x3 tensor images\n",
    "* The output should be a single value that will indicate whether a given image is real or fake\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "def conv(in_channels, out_channels, kernel_size = 4, stride = 2, padding = 1, batch_norm = True):\n",
    "    \n",
    "    disc_conv_layers = []\n",
    "    \n",
    "    conv_layer = nn.Conv2d(in_channels, out_channels, kernel_size, stride, padding, bias = False)\n",
    "    \n",
    "    disc_conv_layers.append(conv_layer)\n",
    "    \n",
    "    if batch_norm:\n",
    "        batch_norm_layer = nn.BatchNorm2d(out_channels)\n",
    "        disc_conv_layers.append(batch_norm_layer)\n",
    "        \n",
    "    return nn.Sequential(*disc_conv_layers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tests Passed\n"
     ]
    }
   ],
   "source": [
    "class Discriminator(nn.Module):\n",
    "\n",
    "    def __init__(self, conv_dim):\n",
    "        \"\"\"\n",
    "        Initialize the Discriminator Module\n",
    "        :param conv_dim: The depth of the first convolutional layer\n",
    "        \"\"\"\n",
    "        super(Discriminator, self).__init__()\n",
    "\n",
    "        # complete init function\n",
    "        \n",
    "        self.conv_dim = conv_dim\n",
    "        \n",
    "        self.conv1 = conv(3, conv_dim, batch_norm = False)\n",
    "        self.conv2 = conv(conv_dim, 2*conv_dim)\n",
    "        self.conv3 = conv(2*conv_dim, 4*conv_dim)\n",
    "        \n",
    "        self.fc = nn.Linear(4*4*4*conv_dim, 1)\n",
    "        \n",
    "        self.dropout = nn.Dropout(p=0.5)\n",
    "        \n",
    "\n",
    "    def forward(self, x):\n",
    "        \"\"\"\n",
    "        Forward propagation of the neural network\n",
    "        :param x: The input to the neural network     \n",
    "        :return: Discriminator logits; the output of the neural network\n",
    "        \"\"\"\n",
    "        # define feedforward behavior\n",
    "        \n",
    "        x = F.leaky_relu(self.conv1(x), 0.2)\n",
    "        x = self.dropout(x)\n",
    "        x = F.leaky_relu(self.conv2(x), 0.2)\n",
    "        x = self.dropout(x)\n",
    "        x = F.leaky_relu(self.conv3(x), 0.2)\n",
    "        x = self.dropout(x)\n",
    "\n",
    "        \n",
    "        x = x.view(-1,4*4*4*self.conv_dim).squeeze()\n",
    "        \n",
    "        x = self.fc(x)\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        return x\n",
    "\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "tests.test_discriminator(Discriminator)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Generator\n",
    "\n",
    "The generator should upsample an input and generate a *new* image of the same size as our training data `32x32x3`. This should be mostly transpose convolutional layers with normalization applied to the outputs.\n",
    "\n",
    "#### Exercise: Complete the Generator class\n",
    "* The inputs to the generator are vectors of some length `z_size`\n",
    "* The output should be a image of shape `32x32x3`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def deconv(in_channels, out_channels, kernel_size = 4, stride = 2, padding = 1, batch_norm = True):\n",
    "    \n",
    "    gen_deconv_layers = []\n",
    "    \n",
    "    deconv_layer = nn.ConvTranspose2d(in_channels, out_channels, kernel_size, stride, padding, bias = False)\n",
    "    \n",
    "    gen_deconv_layers.append(deconv_layer)\n",
    "    \n",
    "    if batch_norm:\n",
    "        batch_norm_layer = nn.BatchNorm2d(out_channels)\n",
    "        gen_deconv_layers.append(batch_norm_layer)\n",
    "        \n",
    "    return nn.Sequential(*gen_deconv_layers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tests Passed\n"
     ]
    }
   ],
   "source": [
    "class Generator(nn.Module):\n",
    "    \n",
    "    def __init__(self, z_size, conv_dim):\n",
    "        \"\"\"\n",
    "        Initialize the Generator Module\n",
    "        :param z_size: The length of the input latent vector, z\n",
    "        :param conv_dim: The depth of the inputs to the *last* transpose convolutional layer\n",
    "        \"\"\"\n",
    "        super(Generator, self).__init__()\n",
    "\n",
    "        # complete init function\n",
    "        \n",
    "        self.conv_dim = conv_dim\n",
    "        \n",
    "        self.fc = nn.Linear(z_size, 4*4*4*conv_dim)\n",
    "        \n",
    "        self.deconv1 = deconv(4*conv_dim, 2*conv_dim)\n",
    "        self.deconv2 = deconv(2*conv_dim, conv_dim)\n",
    "        self.deconv3 = deconv(conv_dim, 3, batch_norm = False)\n",
    "        \n",
    "\n",
    "    def forward(self, x):\n",
    "        \"\"\"\n",
    "        Forward propagation of the neural network\n",
    "        :param x: The input to the neural network     \n",
    "        :return: A 32x32x3 Tensor image as output\n",
    "        \"\"\"\n",
    "        # define feedforward behavior\n",
    "        \n",
    "        x = self.fc(x)\n",
    "        \n",
    "        x = x.view(-1, 4*self.conv_dim, 4, 4)\n",
    "        \n",
    "        x = F.relu(self.deconv1(x))\n",
    "        x = F.relu(self.deconv2(x))\n",
    "        x = F.tanh(self.deconv3(x))\n",
    "        \n",
    "        \n",
    "        \n",
    "        return x\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "tests.test_generator(Generator)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Initialize the weights of your networks\n",
    "\n",
    "To help your models converge, you should initialize the weights of the convolutional and linear layers in your model. From reading the [original DCGAN paper](https://arxiv.org/pdf/1511.06434.pdf), they say:\n",
    "> All weights were initialized from a zero-centered Normal distribution with standard deviation 0.02.\n",
    "\n",
    "So, your next task will be to define a weight initialization function that does just this!\n",
    "\n",
    "You can refer back to the lesson on weight initialization or even consult existing model code, such as that from [the `networks.py` file in CycleGAN Github repository](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/models/networks.py) to help you complete this function.\n",
    "\n",
    "#### Exercise: Complete the weight initialization function\n",
    "\n",
    "* This should initialize only **convolutional** and **linear** layers\n",
    "* Initialize the weights to a normal distribution, centered around 0, with a standard deviation of 0.02.\n",
    "* The bias terms, if they exist, may be left alone or set to 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def weights_init_normal(m):\n",
    "    \"\"\"\n",
    "    Applies initial weights to certain layers in a model .\n",
    "    The weights are taken from a normal distribution \n",
    "    with mean = 0, std dev = 0.02.\n",
    "    :param m: A module or layer in a network    \n",
    "    \"\"\"\n",
    "    # classname will be something like:\n",
    "    # `Conv`, `BatchNorm2d`, `Linear`, etc.\n",
    "    classname = m.__class__.__name__\n",
    "    \n",
    "    # TODO: Apply initial weights to convolutional and linear layers\n",
    "    \n",
    "    if ((classname.find('Conv') != -1) or \n",
    "        (classname.find('Linear') != -1)):\n",
    "        m.weight.data.normal_(0, 0.02) \n",
    "        \n",
    "        # Initializing the weights of convolutional (resp. transpose conv.) and\n",
    "        # linear layers as taken from a normal distribution.\n",
    "        \n",
    "        \n",
    "        \n",
    "        if classname.find('Linear') != -1:\n",
    "            m.bias.data.fill_(0) \n",
    "            # The only layers with biases are the linear layers. Those biases are \n",
    "            # initialized with values equal to zero.\n",
    "                            \n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Build complete network\n",
    "\n",
    "Define your models' hyperparameters and instantiate the discriminator and generator from the classes defined above. Make sure you've passed in the correct input arguments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "def build_network(d_conv_dim, g_conv_dim, z_size):\n",
    "    # define discriminator and generator\n",
    "    D = Discriminator(d_conv_dim)\n",
    "    G = Generator(z_size=z_size, conv_dim=g_conv_dim)\n",
    "\n",
    "    # initialize model weights\n",
    "    D.apply(weights_init_normal)\n",
    "    G.apply(weights_init_normal)\n",
    "\n",
    "    print(D)\n",
    "    print()\n",
    "    print(G)\n",
    "    \n",
    "    return D, G\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Exercise: Define model hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Discriminator(\n",
      "  (conv1): Sequential(\n",
      "    (0): Conv2d(3, 32, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "  )\n",
      "  (conv2): Sequential(\n",
      "    (0): Conv2d(32, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (conv3): Sequential(\n",
      "    (0): Conv2d(64, 128, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (fc): Linear(in_features=2048, out_features=1, bias=True)\n",
      "  (dropout): Dropout(p=0.5)\n",
      ")\n",
      "\n",
      "Generator(\n",
      "  (fc): Linear(in_features=100, out_features=4096, bias=True)\n",
      "  (deconv1): Sequential(\n",
      "    (0): ConvTranspose2d(256, 128, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (deconv2): Sequential(\n",
      "    (0): ConvTranspose2d(128, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (deconv3): Sequential(\n",
      "    (0): ConvTranspose2d(64, 3, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "# Define model hyperparams\n",
    "d_conv_dim = 32\n",
    "g_conv_dim = 64\n",
    "z_size = 100\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "D, G = build_network(d_conv_dim, g_conv_dim, z_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training on GPU\n",
    "\n",
    "Check if you can train on GPU. Here, we'll set this as a boolean variable `train_on_gpu`. Later, you'll be responsible for making sure that \n",
    ">* Models,\n",
    "* Model inputs, and\n",
    "* Loss function arguments\n",
    "\n",
    "Are moved to GPU, where appropriate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training on GPU!\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL\n",
    "\"\"\"\n",
    "import torch\n",
    "\n",
    "# Check for a GPU\n",
    "train_on_gpu = torch.cuda.is_available()\n",
    "if not train_on_gpu:\n",
    "    print('No GPU found. Please use a GPU to train your neural network.')\n",
    "else:\n",
    "    print('Training on GPU!')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## Discriminator and Generator Losses\n",
    "\n",
    "Now we need to calculate the losses for both types of adversarial networks.\n",
    "\n",
    "### Discriminator Losses\n",
    "\n",
    "> * For the discriminator, the total loss is the sum of the losses for real and fake images, `d_loss = d_real_loss + d_fake_loss`. \n",
    "* Remember that we want the discriminator to output 1 for real images and 0 for fake images, so we need to set up the losses to reflect that.\n",
    "\n",
    "\n",
    "### Generator Loss\n",
    "\n",
    "The generator loss will look similar only with flipped labels. The generator's goal is to get the discriminator to *think* its generated images are *real*.\n",
    "\n",
    "#### Exercise: Complete real and fake loss functions\n",
    "\n",
    "**You may choose to use either cross entropy or a least squares error loss to complete the following `real_loss` and `fake_loss` functions.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def real_loss(D_out):\n",
    "    '''Calculates how close discriminator outputs are to being real.\n",
    "       param, D_out: discriminator logits\n",
    "       return: real loss'''\n",
    "    \n",
    "    ones_tensor = torch.ones(D_out.size())\n",
    "    if train_on_gpu:\n",
    "        ones_tensor = ones_tensor.cuda()\n",
    "    \n",
    "    loss = nn.MSELoss()(D_out, ones_tensor)\n",
    "    \n",
    "    \n",
    "    \n",
    "    return loss\n",
    "\n",
    "def fake_loss(D_out):\n",
    "    '''Calculates how close discriminator outputs are to being fake.\n",
    "       param, D_out: discriminator logits\n",
    "       return: fake loss'''\n",
    "    \n",
    "    zeros_tensor = torch.zeros(D_out.size())\n",
    "    if train_on_gpu:\n",
    "        zeros_tensor = zeros_tensor.cuda()\n",
    "    \n",
    "    loss = nn.MSELoss()(D_out, zeros_tensor)\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Optimizers\n",
    "\n",
    "#### Exercise: Define optimizers for your Discriminator (D) and Generator (G)\n",
    "\n",
    "Define optimizers for your models with appropriate hyperparameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.optim as optim\n",
    "\n",
    "# Create optimizers for the discriminator D and generator G\n",
    "d_optimizer = optim.Adam(D.parameters(), 0.002, [0.5, 0.999])\n",
    "g_optimizer = optim.Adam(G.parameters(), 0.002, [0.5, 0.999])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## Training\n",
    "\n",
    "Training will involve alternating between training the discriminator and the generator. You'll use your functions `real_loss` and `fake_loss` to help you calculate the discriminator losses.\n",
    "\n",
    "* You should train the discriminator by alternating on real and fake images\n",
    "* Then the generator, which tries to trick the discriminator and should have an opposing loss function\n",
    "\n",
    "\n",
    "#### Saving Samples\n",
    "\n",
    "You've been given some code to print out some loss statistics and save some generated \"fake\" samples."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Exercise: Complete the training function\n",
    "\n",
    "Keep in mind that, if you've moved your models to GPU, you'll also have to move any model inputs to GPU."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(D, G, n_epochs, print_every=50):\n",
    "    '''Trains adversarial networks for some number of epochs\n",
    "       param, D: the discriminator network\n",
    "       param, G: the generator network\n",
    "       param, n_epochs: number of epochs to train for\n",
    "       param, print_every: when to print and record the models' losses\n",
    "       return: D and G losses'''\n",
    "    \n",
    "    # move models to GPU\n",
    "    if train_on_gpu:\n",
    "        D.cuda()\n",
    "        G.cuda()\n",
    "\n",
    "    # keep track of loss and generated, \"fake\" samples\n",
    "    samples = []\n",
    "    losses = []\n",
    "\n",
    "    # Get some fixed data for sampling. These are images that are held\n",
    "    # constant throughout training, and allow us to inspect the model's performance\n",
    "    sample_size=16\n",
    "    fixed_z = np.random.uniform(-1, 1, size=(sample_size, z_size))\n",
    "    fixed_z = torch.from_numpy(fixed_z).float()\n",
    "    # move z to GPU if available\n",
    "    if train_on_gpu:\n",
    "        fixed_z = fixed_z.cuda()\n",
    "\n",
    "    # epoch training loop\n",
    "    for epoch in range(n_epochs):\n",
    "\n",
    "        # batch training loop\n",
    "        for batch_i, (real_images, _) in enumerate(celeba_train_loader):\n",
    "\n",
    "            batch_size = real_images.size(0)\n",
    "            real_images = scale(real_images)\n",
    "            \n",
    "            \n",
    "\n",
    "            # ===============================================\n",
    "            #         YOUR CODE HERE: TRAIN THE NETWORKS\n",
    "            # ===============================================\n",
    "            \n",
    "            d_optimizer.zero_grad()\n",
    "            \n",
    "            # 1. Train the discriminator on real and fake images\n",
    "            \n",
    "            if train_on_gpu:\n",
    "                real_images = real_images.cuda()\n",
    "            \n",
    "            d_real_loss = real_loss(D(real_images))\n",
    "            \n",
    "            \n",
    "            z = np.random.uniform(-1, 1, size=(batch_size, z_size))\n",
    "            z = torch.from_numpy(z).float()\n",
    "            if train_on_gpu:\n",
    "                z = z.cuda()\n",
    "            fake_images = G(z)\n",
    "            \n",
    "            d_fake_loss = fake_loss(D(fake_images))\n",
    "            \n",
    "            \n",
    "            d_loss = d_real_loss + d_fake_loss\n",
    "            d_loss.backward()\n",
    "            d_optimizer.step()\n",
    "\n",
    "            # 2. Train the generator with an adversarial loss\n",
    "            \n",
    "            g_optimizer.zero_grad()\n",
    "            \n",
    "            z = np.random.uniform(-1, 1, size=(batch_size, z_size))\n",
    "            z = torch.from_numpy(z).float()\n",
    "            if train_on_gpu:\n",
    "                z = z.cuda()\n",
    "            fake_images = G(z)\n",
    "            \n",
    "            \n",
    "            g_loss = real_loss(D(fake_images))\n",
    "            g_loss.backward()\n",
    "            g_optimizer.step()\n",
    "            \n",
    "            \n",
    "            # ===============================================\n",
    "            #              END OF YOUR CODE\n",
    "            # ===============================================\n",
    "\n",
    "            # Print some loss stats\n",
    "            if batch_i % print_every == 0:\n",
    "                # append discriminator loss and generator loss\n",
    "                losses.append((d_loss.item(), g_loss.item()))\n",
    "                # print discriminator and generator loss\n",
    "                print('Epoch [{:5d}/{:5d}] | d_loss: {:6.4f} | g_loss: {:6.4f}'.format(\n",
    "                        epoch+1, n_epochs, d_loss.item(), g_loss.item()))\n",
    "\n",
    "\n",
    "        ## AFTER EACH EPOCH##    \n",
    "        # this code assumes your generator is named G, feel free to change the name\n",
    "        # generate and save sample, fake images\n",
    "        G.eval() # for generating samples\n",
    "        samples_z = G(fixed_z)\n",
    "        samples.append(samples_z)\n",
    "        G.train() # back to training mode\n",
    "\n",
    "    # Save training generator samples\n",
    "    with open('train_samples.pkl', 'wb') as f:\n",
    "        pkl.dump(samples, f)\n",
    "    \n",
    "    # finally return losses\n",
    "    return losses"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set your number of training epochs and train your GAN!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [    1/   50] | d_loss: 0.7287 | g_loss: 0.2848\n",
      "Epoch [    1/   50] | d_loss: 0.5277 | g_loss: 0.3103\n",
      "Epoch [    1/   50] | d_loss: 0.5184 | g_loss: 0.5030\n",
      "Epoch [    1/   50] | d_loss: 0.4411 | g_loss: 0.4586\n",
      "Epoch [    1/   50] | d_loss: 0.4215 | g_loss: 0.3905\n",
      "Epoch [    1/   50] | d_loss: 0.4623 | g_loss: 0.3498\n",
      "Epoch [    1/   50] | d_loss: 0.6278 | g_loss: 0.4591\n",
      "Epoch [    1/   50] | d_loss: 0.7411 | g_loss: 0.4730\n",
      "Epoch [    1/   50] | d_loss: 0.6300 | g_loss: 0.4461\n",
      "Epoch [    1/   50] | d_loss: 0.4103 | g_loss: 0.3670\n",
      "Epoch [    1/   50] | d_loss: 0.4692 | g_loss: 0.4701\n",
      "Epoch [    1/   50] | d_loss: 0.4173 | g_loss: 0.3790\n",
      "Epoch [    1/   50] | d_loss: 0.5789 | g_loss: 0.2904\n",
      "Epoch [    1/   50] | d_loss: 0.4200 | g_loss: 0.3823\n",
      "Epoch [    1/   50] | d_loss: 0.5097 | g_loss: 0.3526\n",
      "Epoch [    1/   50] | d_loss: 0.5060 | g_loss: 0.3650\n",
      "Epoch [    1/   50] | d_loss: 0.4934 | g_loss: 0.2825\n",
      "Epoch [    1/   50] | d_loss: 0.4829 | g_loss: 0.2890\n",
      "Epoch [    1/   50] | d_loss: 0.4763 | g_loss: 0.3567\n",
      "Epoch [    1/   50] | d_loss: 0.4939 | g_loss: 0.4259\n",
      "Epoch [    1/   50] | d_loss: 0.4272 | g_loss: 0.3104\n",
      "Epoch [    1/   50] | d_loss: 0.6584 | g_loss: 0.2977\n",
      "Epoch [    1/   50] | d_loss: 0.3989 | g_loss: 0.2458\n",
      "Epoch [    1/   50] | d_loss: 0.4311 | g_loss: 0.2931\n",
      "Epoch [    1/   50] | d_loss: 0.4627 | g_loss: 0.3515\n",
      "Epoch [    1/   50] | d_loss: 0.5702 | g_loss: 0.2880\n",
      "Epoch [    1/   50] | d_loss: 0.4529 | g_loss: 0.2835\n",
      "Epoch [    1/   50] | d_loss: 0.5405 | g_loss: 0.3004\n",
      "Epoch [    1/   50] | d_loss: 0.4967 | g_loss: 0.3124\n",
      "Epoch [    1/   50] | d_loss: 0.4981 | g_loss: 0.2521\n",
      "Epoch [    1/   50] | d_loss: 0.4962 | g_loss: 0.2617\n",
      "Epoch [    1/   50] | d_loss: 0.5063 | g_loss: 0.2458\n",
      "Epoch [    1/   50] | d_loss: 0.4853 | g_loss: 0.3248\n",
      "Epoch [    1/   50] | d_loss: 0.5623 | g_loss: 0.2308\n",
      "Epoch [    1/   50] | d_loss: 0.4658 | g_loss: 0.3152\n",
      "Epoch [    1/   50] | d_loss: 0.3977 | g_loss: 0.2953\n",
      "Epoch [    1/   50] | d_loss: 0.5500 | g_loss: 0.2485\n",
      "Epoch [    1/   50] | d_loss: 0.5418 | g_loss: 0.2556\n",
      "Epoch [    1/   50] | d_loss: 0.4901 | g_loss: 0.2614\n",
      "Epoch [    1/   50] | d_loss: 0.5298 | g_loss: 0.2620\n",
      "Epoch [    1/   50] | d_loss: 0.5389 | g_loss: 0.2613\n",
      "Epoch [    1/   50] | d_loss: 0.5049 | g_loss: 0.2613\n",
      "Epoch [    1/   50] | d_loss: 0.5105 | g_loss: 0.2860\n",
      "Epoch [    1/   50] | d_loss: 0.5199 | g_loss: 0.2616\n",
      "Epoch [    1/   50] | d_loss: 0.4683 | g_loss: 0.2886\n",
      "Epoch [    1/   50] | d_loss: 0.5232 | g_loss: 0.2511\n",
      "Epoch [    1/   50] | d_loss: 0.5111 | g_loss: 0.2485\n",
      "Epoch [    1/   50] | d_loss: 0.4507 | g_loss: 0.2831\n",
      "Epoch [    1/   50] | d_loss: 0.5025 | g_loss: 0.3039\n",
      "Epoch [    1/   50] | d_loss: 0.4896 | g_loss: 0.2784\n",
      "Epoch [    1/   50] | d_loss: 0.4909 | g_loss: 0.2811\n",
      "Epoch [    1/   50] | d_loss: 0.5057 | g_loss: 0.2397\n",
      "Epoch [    1/   50] | d_loss: 0.4732 | g_loss: 0.3070\n",
      "Epoch [    1/   50] | d_loss: 0.4785 | g_loss: 0.2539\n",
      "Epoch [    1/   50] | d_loss: 0.4896 | g_loss: 0.2643\n",
      "Epoch [    1/   50] | d_loss: 0.5488 | g_loss: 0.2653\n",
      "Epoch [    1/   50] | d_loss: 0.5075 | g_loss: 0.2575\n",
      "Epoch [    1/   50] | d_loss: 0.5089 | g_loss: 0.2706\n",
      "Epoch [    1/   50] | d_loss: 0.5146 | g_loss: 0.2578\n",
      "Epoch [    1/   50] | d_loss: 0.5270 | g_loss: 0.2435\n",
      "Epoch [    1/   50] | d_loss: 0.4869 | g_loss: 0.2801\n",
      "Epoch [    1/   50] | d_loss: 0.5035 | g_loss: 0.2561\n",
      "Epoch [    1/   50] | d_loss: 0.5179 | g_loss: 0.2505\n",
      "Epoch [    1/   50] | d_loss: 0.5263 | g_loss: 0.2511\n",
      "Epoch [    1/   50] | d_loss: 0.5283 | g_loss: 0.2498\n",
      "Epoch [    1/   50] | d_loss: 0.4610 | g_loss: 0.2731\n",
      "Epoch [    1/   50] | d_loss: 0.4893 | g_loss: 0.2593\n",
      "Epoch [    1/   50] | d_loss: 0.5022 | g_loss: 0.2941\n",
      "Epoch [    1/   50] | d_loss: 0.5193 | g_loss: 0.2733\n",
      "Epoch [    1/   50] | d_loss: 0.4992 | g_loss: 0.2569\n",
      "Epoch [    1/   50] | d_loss: 0.4906 | g_loss: 0.2812\n",
      "Epoch [    1/   50] | d_loss: 0.4755 | g_loss: 0.2270\n",
      "Epoch [    1/   50] | d_loss: 0.4830 | g_loss: 0.2362\n",
      "Epoch [    1/   50] | d_loss: 0.4957 | g_loss: 0.2553\n",
      "Epoch [    1/   50] | d_loss: 0.4935 | g_loss: 0.3015\n",
      "Epoch [    1/   50] | d_loss: 0.5132 | g_loss: 0.2598\n",
      "Epoch [    1/   50] | d_loss: 0.4678 | g_loss: 0.2483\n",
      "Epoch [    1/   50] | d_loss: 0.5106 | g_loss: 0.2632\n",
      "Epoch [    1/   50] | d_loss: 0.5277 | g_loss: 0.2499\n",
      "Epoch [    1/   50] | d_loss: 0.4973 | g_loss: 0.2999\n",
      "Epoch [    1/   50] | d_loss: 0.5089 | g_loss: 0.2532\n",
      "Epoch [    1/   50] | d_loss: 0.5251 | g_loss: 0.2595\n",
      "Epoch [    1/   50] | d_loss: 0.4788 | g_loss: 0.2815\n",
      "Epoch [    1/   50] | d_loss: 0.4803 | g_loss: 0.2727\n",
      "Epoch [    1/   50] | d_loss: 0.5004 | g_loss: 0.2839\n",
      "Epoch [    1/   50] | d_loss: 0.5063 | g_loss: 0.2504\n",
      "Epoch [    1/   50] | d_loss: 0.5014 | g_loss: 0.2526\n",
      "Epoch [    1/   50] | d_loss: 0.5458 | g_loss: 0.2638\n",
      "Epoch [    1/   50] | d_loss: 0.4836 | g_loss: 0.2649\n",
      "Epoch [    1/   50] | d_loss: 0.5063 | g_loss: 0.2730\n",
      "Epoch [    2/   50] | d_loss: 0.4748 | g_loss: 0.2705\n",
      "Epoch [    2/   50] | d_loss: 0.4876 | g_loss: 0.2669\n",
      "Epoch [    2/   50] | d_loss: 0.5292 | g_loss: 0.2287\n",
      "Epoch [    2/   50] | d_loss: 0.5179 | g_loss: 0.2305\n",
      "Epoch [    2/   50] | d_loss: 0.5067 | g_loss: 0.2465\n",
      "Epoch [    2/   50] | d_loss: 0.5037 | g_loss: 0.2610\n",
      "Epoch [    2/   50] | d_loss: 0.5236 | g_loss: 0.2699\n",
      "Epoch [    2/   50] | d_loss: 0.4899 | g_loss: 0.2451\n",
      "Epoch [    2/   50] | d_loss: 0.5187 | g_loss: 0.2506\n",
      "Epoch [    2/   50] | d_loss: 0.4873 | g_loss: 0.2321\n",
      "Epoch [    2/   50] | d_loss: 0.4768 | g_loss: 0.2653\n",
      "Epoch [    2/   50] | d_loss: 0.5136 | g_loss: 0.2652\n",
      "Epoch [    2/   50] | d_loss: 0.5082 | g_loss: 0.2387\n",
      "Epoch [    2/   50] | d_loss: 0.5287 | g_loss: 0.2442\n",
      "Epoch [    2/   50] | d_loss: 0.4942 | g_loss: 0.2708\n",
      "Epoch [    2/   50] | d_loss: 0.4958 | g_loss: 0.2605\n",
      "Epoch [    2/   50] | d_loss: 0.4813 | g_loss: 0.2577\n",
      "Epoch [    2/   50] | d_loss: 0.5326 | g_loss: 0.2715\n",
      "Epoch [    2/   50] | d_loss: 0.5135 | g_loss: 0.2475\n",
      "Epoch [    2/   50] | d_loss: 0.5029 | g_loss: 0.2650\n",
      "Epoch [    2/   50] | d_loss: 0.5174 | g_loss: 0.2520\n",
      "Epoch [    2/   50] | d_loss: 0.5011 | g_loss: 0.2738\n",
      "Epoch [    2/   50] | d_loss: 0.5075 | g_loss: 0.2526\n",
      "Epoch [    2/   50] | d_loss: 0.4828 | g_loss: 0.2599\n",
      "Epoch [    2/   50] | d_loss: 0.5128 | g_loss: 0.2558\n",
      "Epoch [    2/   50] | d_loss: 0.4986 | g_loss: 0.2580\n",
      "Epoch [    2/   50] | d_loss: 0.5160 | g_loss: 0.2738\n",
      "Epoch [    2/   50] | d_loss: 0.5175 | g_loss: 0.2525\n",
      "Epoch [    2/   50] | d_loss: 0.4633 | g_loss: 0.2464\n",
      "Epoch [    2/   50] | d_loss: 0.5121 | g_loss: 0.2524\n",
      "Epoch [    2/   50] | d_loss: 0.5080 | g_loss: 0.2470\n",
      "Epoch [    2/   50] | d_loss: 0.5122 | g_loss: 0.2492\n",
      "Epoch [    2/   50] | d_loss: 0.5313 | g_loss: 0.2376\n",
      "Epoch [    2/   50] | d_loss: 0.5173 | g_loss: 0.2591\n",
      "Epoch [    2/   50] | d_loss: 0.4979 | g_loss: 0.2717\n",
      "Epoch [    2/   50] | d_loss: 0.5280 | g_loss: 0.2578\n",
      "Epoch [    2/   50] | d_loss: 0.5412 | g_loss: 0.2760\n",
      "Epoch [    2/   50] | d_loss: 0.5036 | g_loss: 0.2563\n",
      "Epoch [    2/   50] | d_loss: 0.5003 | g_loss: 0.2444\n",
      "Epoch [    2/   50] | d_loss: 0.5372 | g_loss: 0.2613\n",
      "Epoch [    2/   50] | d_loss: 0.5001 | g_loss: 0.2397\n",
      "Epoch [    2/   50] | d_loss: 0.5397 | g_loss: 0.2645\n",
      "Epoch [    2/   50] | d_loss: 0.4845 | g_loss: 0.2345\n",
      "Epoch [    2/   50] | d_loss: 0.4952 | g_loss: 0.2423\n",
      "Epoch [    2/   50] | d_loss: 0.4899 | g_loss: 0.2756\n",
      "Epoch [    2/   50] | d_loss: 0.4970 | g_loss: 0.2520\n",
      "Epoch [    2/   50] | d_loss: 0.4918 | g_loss: 0.2575\n",
      "Epoch [    2/   50] | d_loss: 0.5048 | g_loss: 0.2265\n",
      "Epoch [    2/   50] | d_loss: 0.5149 | g_loss: 0.2562\n",
      "Epoch [    2/   50] | d_loss: 0.4927 | g_loss: 0.2772\n",
      "Epoch [    2/   50] | d_loss: 0.5309 | g_loss: 0.2584\n",
      "Epoch [    2/   50] | d_loss: 0.4727 | g_loss: 0.2466\n",
      "Epoch [    2/   50] | d_loss: 0.5137 | g_loss: 0.2650\n",
      "Epoch [    2/   50] | d_loss: 0.4842 | g_loss: 0.2411\n",
      "Epoch [    2/   50] | d_loss: 0.5247 | g_loss: 0.2453\n",
      "Epoch [    2/   50] | d_loss: 0.5012 | g_loss: 0.2510\n",
      "Epoch [    2/   50] | d_loss: 0.5262 | g_loss: 0.2909\n",
      "Epoch [    2/   50] | d_loss: 0.4953 | g_loss: 0.2441\n",
      "Epoch [    2/   50] | d_loss: 0.5141 | g_loss: 0.2661\n",
      "Epoch [    2/   50] | d_loss: 0.4791 | g_loss: 0.2787\n",
      "Epoch [    2/   50] | d_loss: 0.5379 | g_loss: 0.2426\n",
      "Epoch [    2/   50] | d_loss: 0.5072 | g_loss: 0.2724\n",
      "Epoch [    2/   50] | d_loss: 0.5013 | g_loss: 0.2531\n",
      "Epoch [    2/   50] | d_loss: 0.5237 | g_loss: 0.2451\n",
      "Epoch [    2/   50] | d_loss: 0.4774 | g_loss: 0.2707\n",
      "Epoch [    2/   50] | d_loss: 0.4892 | g_loss: 0.2139\n",
      "Epoch [    2/   50] | d_loss: 0.4977 | g_loss: 0.2917\n",
      "Epoch [    2/   50] | d_loss: 0.5292 | g_loss: 0.2619\n",
      "Epoch [    2/   50] | d_loss: 0.4903 | g_loss: 0.2736\n",
      "Epoch [    2/   50] | d_loss: 0.4747 | g_loss: 0.2564\n",
      "Epoch [    2/   50] | d_loss: 0.4919 | g_loss: 0.2759\n",
      "Epoch [    2/   50] | d_loss: 0.4952 | g_loss: 0.2572\n",
      "Epoch [    2/   50] | d_loss: 0.4893 | g_loss: 0.2534\n",
      "Epoch [    2/   50] | d_loss: 0.5348 | g_loss: 0.2906\n",
      "Epoch [    2/   50] | d_loss: 0.4998 | g_loss: 0.2484\n",
      "Epoch [    2/   50] | d_loss: 0.4716 | g_loss: 0.2758\n",
      "Epoch [    2/   50] | d_loss: 0.5278 | g_loss: 0.2199\n",
      "Epoch [    2/   50] | d_loss: 0.5176 | g_loss: 0.2739\n",
      "Epoch [    2/   50] | d_loss: 0.4788 | g_loss: 0.2483\n",
      "Epoch [    2/   50] | d_loss: 0.5069 | g_loss: 0.2387\n",
      "Epoch [    2/   50] | d_loss: 0.4867 | g_loss: 0.2967\n",
      "Epoch [    2/   50] | d_loss: 0.5116 | g_loss: 0.2608\n",
      "Epoch [    2/   50] | d_loss: 0.4594 | g_loss: 0.2892\n",
      "Epoch [    2/   50] | d_loss: 0.5470 | g_loss: 0.2671\n",
      "Epoch [    2/   50] | d_loss: 0.5494 | g_loss: 0.2642\n",
      "Epoch [    2/   50] | d_loss: 0.4875 | g_loss: 0.2536\n",
      "Epoch [    2/   50] | d_loss: 0.5021 | g_loss: 0.2647\n",
      "Epoch [    2/   50] | d_loss: 0.4733 | g_loss: 0.2684\n",
      "Epoch [    2/   50] | d_loss: 0.4860 | g_loss: 0.2706\n",
      "Epoch [    2/   50] | d_loss: 0.5018 | g_loss: 0.2427\n",
      "Epoch [    3/   50] | d_loss: 0.5254 | g_loss: 0.2537\n",
      "Epoch [    3/   50] | d_loss: 0.4782 | g_loss: 0.2774\n",
      "Epoch [    3/   50] | d_loss: 0.5254 | g_loss: 0.2778\n",
      "Epoch [    3/   50] | d_loss: 0.5482 | g_loss: 0.2750\n",
      "Epoch [    3/   50] | d_loss: 0.5596 | g_loss: 0.2360\n",
      "Epoch [    3/   50] | d_loss: 0.5036 | g_loss: 0.2787\n",
      "Epoch [    3/   50] | d_loss: 0.5535 | g_loss: 0.2525\n",
      "Epoch [    3/   50] | d_loss: 0.4766 | g_loss: 0.2625\n",
      "Epoch [    3/   50] | d_loss: 0.5109 | g_loss: 0.2878\n",
      "Epoch [    3/   50] | d_loss: 0.4810 | g_loss: 0.2426\n",
      "Epoch [    3/   50] | d_loss: 0.4642 | g_loss: 0.2696\n",
      "Epoch [    3/   50] | d_loss: 0.5044 | g_loss: 0.2691\n",
      "Epoch [    3/   50] | d_loss: 0.4773 | g_loss: 0.2290\n",
      "Epoch [    3/   50] | d_loss: 0.5091 | g_loss: 0.2718\n",
      "Epoch [    3/   50] | d_loss: 0.4970 | g_loss: 0.2120\n",
      "Epoch [    3/   50] | d_loss: 0.4970 | g_loss: 0.2514\n",
      "Epoch [    3/   50] | d_loss: 0.5130 | g_loss: 0.2742\n",
      "Epoch [    3/   50] | d_loss: 0.5050 | g_loss: 0.2405\n",
      "Epoch [    3/   50] | d_loss: 0.4749 | g_loss: 0.2456\n",
      "Epoch [    3/   50] | d_loss: 0.5030 | g_loss: 0.2546\n",
      "Epoch [    3/   50] | d_loss: 0.4808 | g_loss: 0.2854\n",
      "Epoch [    3/   50] | d_loss: 0.4885 | g_loss: 0.2539\n",
      "Epoch [    3/   50] | d_loss: 0.4961 | g_loss: 0.2472\n",
      "Epoch [    3/   50] | d_loss: 0.5020 | g_loss: 0.2604\n",
      "Epoch [    3/   50] | d_loss: 0.5115 | g_loss: 0.2287\n",
      "Epoch [    3/   50] | d_loss: 0.4868 | g_loss: 0.2504\n",
      "Epoch [    3/   50] | d_loss: 0.4794 | g_loss: 0.2890\n",
      "Epoch [    3/   50] | d_loss: 0.4875 | g_loss: 0.2918\n",
      "Epoch [    3/   50] | d_loss: 0.5176 | g_loss: 0.2561\n",
      "Epoch [    3/   50] | d_loss: 0.5082 | g_loss: 0.2776\n",
      "Epoch [    3/   50] | d_loss: 0.4614 | g_loss: 0.2279\n",
      "Epoch [    3/   50] | d_loss: 0.5098 | g_loss: 0.2441\n",
      "Epoch [    3/   50] | d_loss: 0.5074 | g_loss: 0.2834\n",
      "Epoch [    3/   50] | d_loss: 0.4942 | g_loss: 0.3351\n",
      "Epoch [    3/   50] | d_loss: 0.5424 | g_loss: 0.2675\n",
      "Epoch [    3/   50] | d_loss: 0.5038 | g_loss: 0.2537\n",
      "Epoch [    3/   50] | d_loss: 0.5232 | g_loss: 0.2659\n",
      "Epoch [    3/   50] | d_loss: 0.4953 | g_loss: 0.2759\n",
      "Epoch [    3/   50] | d_loss: 0.5037 | g_loss: 0.2733\n",
      "Epoch [    3/   50] | d_loss: 0.4725 | g_loss: 0.2754\n",
      "Epoch [    3/   50] | d_loss: 0.4879 | g_loss: 0.2653\n",
      "Epoch [    3/   50] | d_loss: 0.5175 | g_loss: 0.2669\n",
      "Epoch [    3/   50] | d_loss: 0.4512 | g_loss: 0.2738\n",
      "Epoch [    3/   50] | d_loss: 0.4828 | g_loss: 0.2615\n",
      "Epoch [    3/   50] | d_loss: 0.5244 | g_loss: 0.2871\n",
      "Epoch [    3/   50] | d_loss: 0.5161 | g_loss: 0.2462\n",
      "Epoch [    3/   50] | d_loss: 0.5098 | g_loss: 0.2554\n",
      "Epoch [    3/   50] | d_loss: 0.5200 | g_loss: 0.2322\n",
      "Epoch [    3/   50] | d_loss: 0.5008 | g_loss: 0.2659\n",
      "Epoch [    3/   50] | d_loss: 0.5239 | g_loss: 0.2588\n",
      "Epoch [    3/   50] | d_loss: 0.4985 | g_loss: 0.2778\n",
      "Epoch [    3/   50] | d_loss: 0.5236 | g_loss: 0.2826\n",
      "Epoch [    3/   50] | d_loss: 0.4932 | g_loss: 0.2431\n",
      "Epoch [    3/   50] | d_loss: 0.5267 | g_loss: 0.2778\n",
      "Epoch [    3/   50] | d_loss: 0.4920 | g_loss: 0.2588\n",
      "Epoch [    3/   50] | d_loss: 0.4935 | g_loss: 0.2515\n",
      "Epoch [    3/   50] | d_loss: 0.5472 | g_loss: 0.2502\n",
      "Epoch [    3/   50] | d_loss: 0.5011 | g_loss: 0.2540\n",
      "Epoch [    3/   50] | d_loss: 0.5232 | g_loss: 0.2538\n",
      "Epoch [    3/   50] | d_loss: 0.4946 | g_loss: 0.2318\n",
      "Epoch [    3/   50] | d_loss: 0.5320 | g_loss: 0.2527\n",
      "Epoch [    3/   50] | d_loss: 0.5013 | g_loss: 0.2383\n",
      "Epoch [    3/   50] | d_loss: 0.4385 | g_loss: 0.2698\n",
      "Epoch [    3/   50] | d_loss: 0.5217 | g_loss: 0.2423\n",
      "Epoch [    3/   50] | d_loss: 0.5316 | g_loss: 0.2476\n",
      "Epoch [    3/   50] | d_loss: 0.4798 | g_loss: 0.2469\n",
      "Epoch [    3/   50] | d_loss: 0.4810 | g_loss: 0.2756\n",
      "Epoch [    3/   50] | d_loss: 0.4683 | g_loss: 0.2547\n",
      "Epoch [    3/   50] | d_loss: 0.4802 | g_loss: 0.2487\n",
      "Epoch [    3/   50] | d_loss: 0.5273 | g_loss: 0.2673\n",
      "Epoch [    3/   50] | d_loss: 0.4289 | g_loss: 0.2767\n",
      "Epoch [    3/   50] | d_loss: 0.4750 | g_loss: 0.3006\n",
      "Epoch [    3/   50] | d_loss: 0.5115 | g_loss: 0.2613\n",
      "Epoch [    3/   50] | d_loss: 0.4841 | g_loss: 0.3055\n",
      "Epoch [    3/   50] | d_loss: 0.5419 | g_loss: 0.2467\n",
      "Epoch [    3/   50] | d_loss: 0.5183 | g_loss: 0.2338\n",
      "Epoch [    3/   50] | d_loss: 0.5270 | g_loss: 0.2572\n",
      "Epoch [    3/   50] | d_loss: 0.4865 | g_loss: 0.2331\n",
      "Epoch [    3/   50] | d_loss: 0.5221 | g_loss: 0.2870\n",
      "Epoch [    3/   50] | d_loss: 0.4597 | g_loss: 0.2612\n",
      "Epoch [    3/   50] | d_loss: 0.5097 | g_loss: 0.2895\n",
      "Epoch [    3/   50] | d_loss: 0.5469 | g_loss: 0.2545\n",
      "Epoch [    3/   50] | d_loss: 0.5669 | g_loss: 0.3049\n",
      "Epoch [    3/   50] | d_loss: 0.4643 | g_loss: 0.2647\n",
      "Epoch [    3/   50] | d_loss: 0.4898 | g_loss: 0.2631\n",
      "Epoch [    3/   50] | d_loss: 0.4751 | g_loss: 0.2613\n",
      "Epoch [    3/   50] | d_loss: 0.5147 | g_loss: 0.2711\n",
      "Epoch [    3/   50] | d_loss: 0.5194 | g_loss: 0.2696\n",
      "Epoch [    3/   50] | d_loss: 0.4829 | g_loss: 0.2754\n",
      "Epoch [    3/   50] | d_loss: 0.4949 | g_loss: 0.3220\n",
      "Epoch [    4/   50] | d_loss: 0.4840 | g_loss: 0.2448\n",
      "Epoch [    4/   50] | d_loss: 0.5039 | g_loss: 0.2420\n",
      "Epoch [    4/   50] | d_loss: 0.4712 | g_loss: 0.2775\n",
      "Epoch [    4/   50] | d_loss: 0.4688 | g_loss: 0.2530\n",
      "Epoch [    4/   50] | d_loss: 0.4746 | g_loss: 0.2686\n",
      "Epoch [    4/   50] | d_loss: 0.5083 | g_loss: 0.2730\n",
      "Epoch [    4/   50] | d_loss: 0.5274 | g_loss: 0.2368\n",
      "Epoch [    4/   50] | d_loss: 0.5027 | g_loss: 0.2775\n",
      "Epoch [    4/   50] | d_loss: 0.5351 | g_loss: 0.2683\n",
      "Epoch [    4/   50] | d_loss: 0.5263 | g_loss: 0.3039\n",
      "Epoch [    4/   50] | d_loss: 0.4752 | g_loss: 0.3028\n",
      "Epoch [    4/   50] | d_loss: 0.5306 | g_loss: 0.2507\n",
      "Epoch [    4/   50] | d_loss: 0.4911 | g_loss: 0.2467\n",
      "Epoch [    4/   50] | d_loss: 0.4480 | g_loss: 0.2497\n",
      "Epoch [    4/   50] | d_loss: 0.4510 | g_loss: 0.2434\n",
      "Epoch [    4/   50] | d_loss: 0.4693 | g_loss: 0.2761\n",
      "Epoch [    4/   50] | d_loss: 0.5313 | g_loss: 0.2325\n",
      "Epoch [    4/   50] | d_loss: 0.4774 | g_loss: 0.2850\n",
      "Epoch [    4/   50] | d_loss: 0.4736 | g_loss: 0.2759\n",
      "Epoch [    4/   50] | d_loss: 0.4828 | g_loss: 0.2525\n",
      "Epoch [    4/   50] | d_loss: 0.4891 | g_loss: 0.2841\n",
      "Epoch [    4/   50] | d_loss: 0.4543 | g_loss: 0.2887\n",
      "Epoch [    4/   50] | d_loss: 0.4541 | g_loss: 0.2762\n",
      "Epoch [    4/   50] | d_loss: 0.5113 | g_loss: 0.2863\n",
      "Epoch [    4/   50] | d_loss: 0.4676 | g_loss: 0.2524\n",
      "Epoch [    4/   50] | d_loss: 0.5324 | g_loss: 0.2408\n",
      "Epoch [    4/   50] | d_loss: 0.4838 | g_loss: 0.2818\n",
      "Epoch [    4/   50] | d_loss: 0.4949 | g_loss: 0.2811\n",
      "Epoch [    4/   50] | d_loss: 0.5472 | g_loss: 0.2722\n",
      "Epoch [    4/   50] | d_loss: 0.4881 | g_loss: 0.2614\n",
      "Epoch [    4/   50] | d_loss: 0.4986 | g_loss: 0.2753\n",
      "Epoch [    4/   50] | d_loss: 0.5109 | g_loss: 0.2891\n",
      "Epoch [    4/   50] | d_loss: 0.4973 | g_loss: 0.2699\n",
      "Epoch [    4/   50] | d_loss: 0.4753 | g_loss: 0.2422\n",
      "Epoch [    4/   50] | d_loss: 0.5355 | g_loss: 0.2849\n",
      "Epoch [    4/   50] | d_loss: 0.4819 | g_loss: 0.2684\n",
      "Epoch [    4/   50] | d_loss: 0.5062 | g_loss: 0.2601\n",
      "Epoch [    4/   50] | d_loss: 0.4708 | g_loss: 0.2687\n",
      "Epoch [    4/   50] | d_loss: 0.4835 | g_loss: 0.2635\n",
      "Epoch [    4/   50] | d_loss: 0.5019 | g_loss: 0.2250\n",
      "Epoch [    4/   50] | d_loss: 0.4593 | g_loss: 0.2615\n",
      "Epoch [    4/   50] | d_loss: 0.5443 | g_loss: 0.3050\n",
      "Epoch [    4/   50] | d_loss: 0.5088 | g_loss: 0.2614\n",
      "Epoch [    4/   50] | d_loss: 0.4779 | g_loss: 0.2789\n",
      "Epoch [    4/   50] | d_loss: 0.4983 | g_loss: 0.2437\n",
      "Epoch [    4/   50] | d_loss: 0.4726 | g_loss: 0.2378\n",
      "Epoch [    4/   50] | d_loss: 0.5310 | g_loss: 0.2365\n",
      "Epoch [    4/   50] | d_loss: 0.4687 | g_loss: 0.2776\n",
      "Epoch [    4/   50] | d_loss: 0.4928 | g_loss: 0.2631\n",
      "Epoch [    4/   50] | d_loss: 0.5331 | g_loss: 0.2914\n",
      "Epoch [    4/   50] | d_loss: 0.5050 | g_loss: 0.2658\n",
      "Epoch [    4/   50] | d_loss: 0.5531 | g_loss: 0.2739\n",
      "Epoch [    4/   50] | d_loss: 0.4531 | g_loss: 0.2740\n",
      "Epoch [    4/   50] | d_loss: 0.4776 | g_loss: 0.2354\n",
      "Epoch [    4/   50] | d_loss: 0.5091 | g_loss: 0.2548\n",
      "Epoch [    4/   50] | d_loss: 0.4591 | g_loss: 0.2940\n",
      "Epoch [    4/   50] | d_loss: 0.4516 | g_loss: 0.2514\n",
      "Epoch [    4/   50] | d_loss: 0.5292 | g_loss: 0.2379\n",
      "Epoch [    4/   50] | d_loss: 0.4872 | g_loss: 0.3492\n",
      "Epoch [    4/   50] | d_loss: 0.5263 | g_loss: 0.2591\n",
      "Epoch [    4/   50] | d_loss: 0.4809 | g_loss: 0.2831\n",
      "Epoch [    4/   50] | d_loss: 0.4863 | g_loss: 0.2474\n",
      "Epoch [    4/   50] | d_loss: 0.5244 | g_loss: 0.2599\n",
      "Epoch [    4/   50] | d_loss: 0.4824 | g_loss: 0.2527\n",
      "Epoch [    4/   50] | d_loss: 0.5085 | g_loss: 0.2575\n",
      "Epoch [    4/   50] | d_loss: 0.5035 | g_loss: 0.2804\n",
      "Epoch [    4/   50] | d_loss: 0.4575 | g_loss: 0.2818\n",
      "Epoch [    4/   50] | d_loss: 0.5190 | g_loss: 0.3427\n",
      "Epoch [    4/   50] | d_loss: 0.4645 | g_loss: 0.2750\n",
      "Epoch [    4/   50] | d_loss: 0.4596 | g_loss: 0.2850\n",
      "Epoch [    4/   50] | d_loss: 0.4930 | g_loss: 0.2394\n",
      "Epoch [    4/   50] | d_loss: 0.5433 | g_loss: 0.2459\n",
      "Epoch [    4/   50] | d_loss: 0.4953 | g_loss: 0.2925\n",
      "Epoch [    4/   50] | d_loss: 0.4971 | g_loss: 0.3169\n",
      "Epoch [    4/   50] | d_loss: 0.5476 | g_loss: 0.2782\n",
      "Epoch [    4/   50] | d_loss: 0.5031 | g_loss: 0.3084\n",
      "Epoch [    4/   50] | d_loss: 0.4864 | g_loss: 0.2696\n",
      "Epoch [    4/   50] | d_loss: 0.4731 | g_loss: 0.2901\n",
      "Epoch [    4/   50] | d_loss: 0.5266 | g_loss: 0.2625\n",
      "Epoch [    4/   50] | d_loss: 0.5473 | g_loss: 0.2667\n",
      "Epoch [    4/   50] | d_loss: 0.4848 | g_loss: 0.2809\n",
      "Epoch [    4/   50] | d_loss: 0.4559 | g_loss: 0.2479\n",
      "Epoch [    4/   50] | d_loss: 0.5136 | g_loss: 0.2549\n",
      "Epoch [    4/   50] | d_loss: 0.4575 | g_loss: 0.2640\n",
      "Epoch [    4/   50] | d_loss: 0.5499 | g_loss: 0.2698\n",
      "Epoch [    4/   50] | d_loss: 0.5434 | g_loss: 0.2809\n",
      "Epoch [    4/   50] | d_loss: 0.4850 | g_loss: 0.3127\n",
      "Epoch [    4/   50] | d_loss: 0.4481 | g_loss: 0.3041\n",
      "Epoch [    4/   50] | d_loss: 0.4201 | g_loss: 0.2811\n",
      "Epoch [    4/   50] | d_loss: 0.4739 | g_loss: 0.3194\n",
      "Epoch [    5/   50] | d_loss: 0.4723 | g_loss: 0.3165\n",
      "Epoch [    5/   50] | d_loss: 0.5196 | g_loss: 0.2704\n",
      "Epoch [    5/   50] | d_loss: 0.5153 | g_loss: 0.2620\n",
      "Epoch [    5/   50] | d_loss: 0.4729 | g_loss: 0.2709\n",
      "Epoch [    5/   50] | d_loss: 0.4872 | g_loss: 0.2918\n",
      "Epoch [    5/   50] | d_loss: 0.4638 | g_loss: 0.2580\n",
      "Epoch [    5/   50] | d_loss: 0.4704 | g_loss: 0.2366\n",
      "Epoch [    5/   50] | d_loss: 0.5821 | g_loss: 0.2541\n",
      "Epoch [    5/   50] | d_loss: 0.4660 | g_loss: 0.2947\n",
      "Epoch [    5/   50] | d_loss: 0.4786 | g_loss: 0.2497\n",
      "Epoch [    5/   50] | d_loss: 0.5162 | g_loss: 0.3212\n",
      "Epoch [    5/   50] | d_loss: 0.4996 | g_loss: 0.2652\n",
      "Epoch [    5/   50] | d_loss: 0.4935 | g_loss: 0.3148\n",
      "Epoch [    5/   50] | d_loss: 0.4745 | g_loss: 0.2952\n",
      "Epoch [    5/   50] | d_loss: 0.4653 | g_loss: 0.2920\n",
      "Epoch [    5/   50] | d_loss: 0.5313 | g_loss: 0.2612\n",
      "Epoch [    5/   50] | d_loss: 0.4256 | g_loss: 0.2674\n",
      "Epoch [    5/   50] | d_loss: 0.5423 | g_loss: 0.2557\n",
      "Epoch [    5/   50] | d_loss: 0.5092 | g_loss: 0.3055\n",
      "Epoch [    5/   50] | d_loss: 0.4706 | g_loss: 0.2763\n",
      "Epoch [    5/   50] | d_loss: 0.5340 | g_loss: 0.3018\n",
      "Epoch [    5/   50] | d_loss: 0.4940 | g_loss: 0.2741\n",
      "Epoch [    5/   50] | d_loss: 0.4638 | g_loss: 0.2703\n",
      "Epoch [    5/   50] | d_loss: 0.4888 | g_loss: 0.3162\n",
      "Epoch [    5/   50] | d_loss: 0.4445 | g_loss: 0.3006\n",
      "Epoch [    5/   50] | d_loss: 0.4961 | g_loss: 0.2835\n",
      "Epoch [    5/   50] | d_loss: 0.5050 | g_loss: 0.2509\n",
      "Epoch [    5/   50] | d_loss: 0.4942 | g_loss: 0.2889\n",
      "Epoch [    5/   50] | d_loss: 0.4701 | g_loss: 0.2433\n",
      "Epoch [    5/   50] | d_loss: 0.4825 | g_loss: 0.2511\n",
      "Epoch [    5/   50] | d_loss: 0.5388 | g_loss: 0.2911\n",
      "Epoch [    5/   50] | d_loss: 0.4766 | g_loss: 0.3197\n",
      "Epoch [    5/   50] | d_loss: 0.4817 | g_loss: 0.2745\n",
      "Epoch [    5/   50] | d_loss: 0.4751 | g_loss: 0.3108\n",
      "Epoch [    5/   50] | d_loss: 0.5025 | g_loss: 0.2862\n",
      "Epoch [    5/   50] | d_loss: 0.5264 | g_loss: 0.2355\n",
      "Epoch [    5/   50] | d_loss: 0.4966 | g_loss: 0.2789\n",
      "Epoch [    5/   50] | d_loss: 0.4657 | g_loss: 0.2717\n",
      "Epoch [    5/   50] | d_loss: 0.5022 | g_loss: 0.3507\n",
      "Epoch [    5/   50] | d_loss: 0.5143 | g_loss: 0.2741\n",
      "Epoch [    5/   50] | d_loss: 0.4954 | g_loss: 0.2816\n",
      "Epoch [    5/   50] | d_loss: 0.5234 | g_loss: 0.2738\n",
      "Epoch [    5/   50] | d_loss: 0.4723 | g_loss: 0.2999\n",
      "Epoch [    5/   50] | d_loss: 0.5314 | g_loss: 0.3178\n",
      "Epoch [    5/   50] | d_loss: 0.4334 | g_loss: 0.2665\n",
      "Epoch [    5/   50] | d_loss: 0.5336 | g_loss: 0.2602\n",
      "Epoch [    5/   50] | d_loss: 0.5479 | g_loss: 0.2644\n",
      "Epoch [    5/   50] | d_loss: 0.4539 | g_loss: 0.2963\n",
      "Epoch [    5/   50] | d_loss: 0.5382 | g_loss: 0.2873\n",
      "Epoch [    5/   50] | d_loss: 0.5620 | g_loss: 0.2869\n",
      "Epoch [    5/   50] | d_loss: 0.4668 | g_loss: 0.3112\n",
      "Epoch [    5/   50] | d_loss: 0.4604 | g_loss: 0.3140\n",
      "Epoch [    5/   50] | d_loss: 0.4873 | g_loss: 0.3124\n",
      "Epoch [    5/   50] | d_loss: 0.4795 | g_loss: 0.2623\n",
      "Epoch [    5/   50] | d_loss: 0.4866 | g_loss: 0.2694\n",
      "Epoch [    5/   50] | d_loss: 0.4690 | g_loss: 0.3254\n",
      "Epoch [    5/   50] | d_loss: 0.4525 | g_loss: 0.2803\n",
      "Epoch [    5/   50] | d_loss: 0.4931 | g_loss: 0.2821\n",
      "Epoch [    5/   50] | d_loss: 0.4579 | g_loss: 0.3247\n",
      "Epoch [    5/   50] | d_loss: 0.4941 | g_loss: 0.2594\n",
      "Epoch [    5/   50] | d_loss: 0.4959 | g_loss: 0.2924\n",
      "Epoch [    5/   50] | d_loss: 0.5094 | g_loss: 0.2539\n",
      "Epoch [    5/   50] | d_loss: 0.5347 | g_loss: 0.2595\n",
      "Epoch [    5/   50] | d_loss: 0.5188 | g_loss: 0.2703\n",
      "Epoch [    5/   50] | d_loss: 0.4404 | g_loss: 0.2349\n",
      "Epoch [    5/   50] | d_loss: 0.4639 | g_loss: 0.2855\n",
      "Epoch [    5/   50] | d_loss: 0.5016 | g_loss: 0.2891\n",
      "Epoch [    5/   50] | d_loss: 0.4942 | g_loss: 0.2659\n",
      "Epoch [    5/   50] | d_loss: 0.4693 | g_loss: 0.2813\n",
      "Epoch [    5/   50] | d_loss: 0.5153 | g_loss: 0.2534\n",
      "Epoch [    5/   50] | d_loss: 0.4884 | g_loss: 0.3012\n",
      "Epoch [    5/   50] | d_loss: 0.5061 | g_loss: 0.3004\n",
      "Epoch [    5/   50] | d_loss: 0.4579 | g_loss: 0.3238\n",
      "Epoch [    5/   50] | d_loss: 0.4570 | g_loss: 0.3101\n",
      "Epoch [    5/   50] | d_loss: 0.5317 | g_loss: 0.2883\n",
      "Epoch [    5/   50] | d_loss: 0.5234 | g_loss: 0.2901\n",
      "Epoch [    5/   50] | d_loss: 0.4999 | g_loss: 0.2362\n",
      "Epoch [    5/   50] | d_loss: 0.5600 | g_loss: 0.2514\n",
      "Epoch [    5/   50] | d_loss: 0.4957 | g_loss: 0.2719\n",
      "Epoch [    5/   50] | d_loss: 0.4938 | g_loss: 0.2529\n",
      "Epoch [    5/   50] | d_loss: 0.5444 | g_loss: 0.2662\n",
      "Epoch [    5/   50] | d_loss: 0.4600 | g_loss: 0.2574\n",
      "Epoch [    5/   50] | d_loss: 0.5025 | g_loss: 0.2891\n",
      "Epoch [    5/   50] | d_loss: 0.5257 | g_loss: 0.2581\n",
      "Epoch [    5/   50] | d_loss: 0.4887 | g_loss: 0.3091\n",
      "Epoch [    5/   50] | d_loss: 0.4816 | g_loss: 0.2832\n",
      "Epoch [    5/   50] | d_loss: 0.4249 | g_loss: 0.3068\n",
      "Epoch [    5/   50] | d_loss: 0.5549 | g_loss: 0.3032\n",
      "Epoch [    5/   50] | d_loss: 0.5051 | g_loss: 0.2784\n",
      "Epoch [    5/   50] | d_loss: 0.4608 | g_loss: 0.2706\n",
      "Epoch [    6/   50] | d_loss: 0.4580 | g_loss: 0.2988\n",
      "Epoch [    6/   50] | d_loss: 0.4805 | g_loss: 0.2801\n",
      "Epoch [    6/   50] | d_loss: 0.5062 | g_loss: 0.2511\n",
      "Epoch [    6/   50] | d_loss: 0.4935 | g_loss: 0.2872\n",
      "Epoch [    6/   50] | d_loss: 0.4425 | g_loss: 0.2653\n",
      "Epoch [    6/   50] | d_loss: 0.5029 | g_loss: 0.2854\n",
      "Epoch [    6/   50] | d_loss: 0.4556 | g_loss: 0.2713\n",
      "Epoch [    6/   50] | d_loss: 0.5603 | g_loss: 0.2842\n",
      "Epoch [    6/   50] | d_loss: 0.4601 | g_loss: 0.3419\n",
      "Epoch [    6/   50] | d_loss: 0.4624 | g_loss: 0.2591\n",
      "Epoch [    6/   50] | d_loss: 0.5173 | g_loss: 0.2629\n",
      "Epoch [    6/   50] | d_loss: 0.5015 | g_loss: 0.3039\n",
      "Epoch [    6/   50] | d_loss: 0.4839 | g_loss: 0.2412\n",
      "Epoch [    6/   50] | d_loss: 0.5185 | g_loss: 0.2760\n",
      "Epoch [    6/   50] | d_loss: 0.5211 | g_loss: 0.3016\n",
      "Epoch [    6/   50] | d_loss: 0.4505 | g_loss: 0.2656\n",
      "Epoch [    6/   50] | d_loss: 0.4671 | g_loss: 0.2618\n",
      "Epoch [    6/   50] | d_loss: 0.4856 | g_loss: 0.2589\n",
      "Epoch [    6/   50] | d_loss: 0.4933 | g_loss: 0.2915\n",
      "Epoch [    6/   50] | d_loss: 0.5307 | g_loss: 0.2772\n",
      "Epoch [    6/   50] | d_loss: 0.4838 | g_loss: 0.2972\n",
      "Epoch [    6/   50] | d_loss: 0.5381 | g_loss: 0.2479\n",
      "Epoch [    6/   50] | d_loss: 0.4756 | g_loss: 0.2925\n",
      "Epoch [    6/   50] | d_loss: 0.5188 | g_loss: 0.2542\n",
      "Epoch [    6/   50] | d_loss: 0.5251 | g_loss: 0.3131\n",
      "Epoch [    6/   50] | d_loss: 0.5226 | g_loss: 0.2639\n",
      "Epoch [    6/   50] | d_loss: 0.4523 | g_loss: 0.2690\n",
      "Epoch [    6/   50] | d_loss: 0.5349 | g_loss: 0.2954\n",
      "Epoch [    6/   50] | d_loss: 0.4657 | g_loss: 0.2276\n",
      "Epoch [    6/   50] | d_loss: 0.5111 | g_loss: 0.2989\n",
      "Epoch [    6/   50] | d_loss: 0.5044 | g_loss: 0.2725\n",
      "Epoch [    6/   50] | d_loss: 0.5049 | g_loss: 0.2827\n",
      "Epoch [    6/   50] | d_loss: 0.4876 | g_loss: 0.2774\n",
      "Epoch [    6/   50] | d_loss: 0.5087 | g_loss: 0.2755\n",
      "Epoch [    6/   50] | d_loss: 0.5263 | g_loss: 0.2535\n",
      "Epoch [    6/   50] | d_loss: 0.5365 | g_loss: 0.2835\n",
      "Epoch [    6/   50] | d_loss: 0.5018 | g_loss: 0.2327\n",
      "Epoch [    6/   50] | d_loss: 0.5102 | g_loss: 0.2626\n",
      "Epoch [    6/   50] | d_loss: 0.5369 | g_loss: 0.2931\n",
      "Epoch [    6/   50] | d_loss: 0.4891 | g_loss: 0.2585\n",
      "Epoch [    6/   50] | d_loss: 0.5063 | g_loss: 0.2393\n",
      "Epoch [    6/   50] | d_loss: 0.4885 | g_loss: 0.2662\n",
      "Epoch [    6/   50] | d_loss: 0.4867 | g_loss: 0.2756\n",
      "Epoch [    6/   50] | d_loss: 0.4421 | g_loss: 0.2818\n",
      "Epoch [    6/   50] | d_loss: 0.5087 | g_loss: 0.3038\n",
      "Epoch [    6/   50] | d_loss: 0.5304 | g_loss: 0.2796\n",
      "Epoch [    6/   50] | d_loss: 0.4974 | g_loss: 0.3115\n",
      "Epoch [    6/   50] | d_loss: 0.4749 | g_loss: 0.2698\n",
      "Epoch [    6/   50] | d_loss: 0.5163 | g_loss: 0.2862\n",
      "Epoch [    6/   50] | d_loss: 0.4722 | g_loss: 0.2787\n",
      "Epoch [    6/   50] | d_loss: 0.5601 | g_loss: 0.2393\n",
      "Epoch [    6/   50] | d_loss: 0.5145 | g_loss: 0.2755\n",
      "Epoch [    6/   50] | d_loss: 0.5147 | g_loss: 0.2929\n",
      "Epoch [    6/   50] | d_loss: 0.5021 | g_loss: 0.2501\n",
      "Epoch [    6/   50] | d_loss: 0.5551 | g_loss: 0.2439\n",
      "Epoch [    6/   50] | d_loss: 0.5162 | g_loss: 0.2983\n",
      "Epoch [    6/   50] | d_loss: 0.5154 | g_loss: 0.2608\n",
      "Epoch [    6/   50] | d_loss: 0.4822 | g_loss: 0.2549\n",
      "Epoch [    6/   50] | d_loss: 0.4665 | g_loss: 0.3021\n",
      "Epoch [    6/   50] | d_loss: 0.5139 | g_loss: 0.2339\n",
      "Epoch [    6/   50] | d_loss: 0.5280 | g_loss: 0.2578\n",
      "Epoch [    6/   50] | d_loss: 0.5164 | g_loss: 0.2855\n",
      "Epoch [    6/   50] | d_loss: 0.5076 | g_loss: 0.2550\n",
      "Epoch [    6/   50] | d_loss: 0.5660 | g_loss: 0.2909\n",
      "Epoch [    6/   50] | d_loss: 0.4948 | g_loss: 0.2855\n",
      "Epoch [    6/   50] | d_loss: 0.4736 | g_loss: 0.2803\n",
      "Epoch [    6/   50] | d_loss: 0.4801 | g_loss: 0.2757\n",
      "Epoch [    6/   50] | d_loss: 0.4575 | g_loss: 0.2982\n",
      "Epoch [    6/   50] | d_loss: 0.4915 | g_loss: 0.2871\n",
      "Epoch [    6/   50] | d_loss: 0.4828 | g_loss: 0.2770\n",
      "Epoch [    6/   50] | d_loss: 0.4960 | g_loss: 0.2903\n",
      "Epoch [    6/   50] | d_loss: 0.5205 | g_loss: 0.2263\n",
      "Epoch [    6/   50] | d_loss: 0.5541 | g_loss: 0.2306\n",
      "Epoch [    6/   50] | d_loss: 0.5226 | g_loss: 0.2819\n",
      "Epoch [    6/   50] | d_loss: 0.4604 | g_loss: 0.2883\n",
      "Epoch [    6/   50] | d_loss: 0.4562 | g_loss: 0.2691\n",
      "Epoch [    6/   50] | d_loss: 0.4949 | g_loss: 0.2360\n",
      "Epoch [    6/   50] | d_loss: 0.4341 | g_loss: 0.2776\n",
      "Epoch [    6/   50] | d_loss: 0.4847 | g_loss: 0.2570\n",
      "Epoch [    6/   50] | d_loss: 0.4852 | g_loss: 0.2565\n",
      "Epoch [    6/   50] | d_loss: 0.5188 | g_loss: 0.2547\n",
      "Epoch [    6/   50] | d_loss: 0.5097 | g_loss: 0.2265\n",
      "Epoch [    6/   50] | d_loss: 0.5167 | g_loss: 0.2397\n",
      "Epoch [    6/   50] | d_loss: 0.4923 | g_loss: 0.2756\n",
      "Epoch [    6/   50] | d_loss: 0.4521 | g_loss: 0.2670\n",
      "Epoch [    6/   50] | d_loss: 0.5015 | g_loss: 0.2865\n",
      "Epoch [    6/   50] | d_loss: 0.4537 | g_loss: 0.3002\n",
      "Epoch [    6/   50] | d_loss: 0.4749 | g_loss: 0.3033\n",
      "Epoch [    6/   50] | d_loss: 0.4813 | g_loss: 0.2511\n",
      "Epoch [    6/   50] | d_loss: 0.4828 | g_loss: 0.2752\n",
      "Epoch [    7/   50] | d_loss: 0.4696 | g_loss: 0.2789\n",
      "Epoch [    7/   50] | d_loss: 0.5115 | g_loss: 0.2600\n",
      "Epoch [    7/   50] | d_loss: 0.5069 | g_loss: 0.2546\n",
      "Epoch [    7/   50] | d_loss: 0.5016 | g_loss: 0.2925\n",
      "Epoch [    7/   50] | d_loss: 0.4981 | g_loss: 0.2441\n",
      "Epoch [    7/   50] | d_loss: 0.4840 | g_loss: 0.2888\n",
      "Epoch [    7/   50] | d_loss: 0.4741 | g_loss: 0.2982\n",
      "Epoch [    7/   50] | d_loss: 0.4921 | g_loss: 0.2595\n",
      "Epoch [    7/   50] | d_loss: 0.4813 | g_loss: 0.3231\n",
      "Epoch [    7/   50] | d_loss: 0.5142 | g_loss: 0.2480\n",
      "Epoch [    7/   50] | d_loss: 0.4865 | g_loss: 0.2566\n",
      "Epoch [    7/   50] | d_loss: 0.4445 | g_loss: 0.2839\n",
      "Epoch [    7/   50] | d_loss: 0.4667 | g_loss: 0.2421\n",
      "Epoch [    7/   50] | d_loss: 0.4635 | g_loss: 0.2829\n",
      "Epoch [    7/   50] | d_loss: 0.5353 | g_loss: 0.2552\n",
      "Epoch [    7/   50] | d_loss: 0.4641 | g_loss: 0.2865\n",
      "Epoch [    7/   50] | d_loss: 0.5297 | g_loss: 0.2922\n",
      "Epoch [    7/   50] | d_loss: 0.5020 | g_loss: 0.2523\n",
      "Epoch [    7/   50] | d_loss: 0.4574 | g_loss: 0.2300\n",
      "Epoch [    7/   50] | d_loss: 0.5386 | g_loss: 0.2549\n",
      "Epoch [    7/   50] | d_loss: 0.4690 | g_loss: 0.2583\n",
      "Epoch [    7/   50] | d_loss: 0.4828 | g_loss: 0.2761\n",
      "Epoch [    7/   50] | d_loss: 0.4598 | g_loss: 0.2738\n",
      "Epoch [    7/   50] | d_loss: 0.4679 | g_loss: 0.2599\n",
      "Epoch [    7/   50] | d_loss: 0.5183 | g_loss: 0.2422\n",
      "Epoch [    7/   50] | d_loss: 0.5076 | g_loss: 0.2679\n",
      "Epoch [    7/   50] | d_loss: 0.5032 | g_loss: 0.2532\n",
      "Epoch [    7/   50] | d_loss: 0.5323 | g_loss: 0.2783\n",
      "Epoch [    7/   50] | d_loss: 0.4850 | g_loss: 0.2408\n",
      "Epoch [    7/   50] | d_loss: 0.4878 | g_loss: 0.2911\n",
      "Epoch [    7/   50] | d_loss: 0.4422 | g_loss: 0.2665\n",
      "Epoch [    7/   50] | d_loss: 0.5115 | g_loss: 0.2365\n",
      "Epoch [    7/   50] | d_loss: 0.4830 | g_loss: 0.2381\n",
      "Epoch [    7/   50] | d_loss: 0.5177 | g_loss: 0.2464\n",
      "Epoch [    7/   50] | d_loss: 0.4723 | g_loss: 0.2623\n",
      "Epoch [    7/   50] | d_loss: 0.5111 | g_loss: 0.3312\n",
      "Epoch [    7/   50] | d_loss: 0.5075 | g_loss: 0.2646\n",
      "Epoch [    7/   50] | d_loss: 0.4764 | g_loss: 0.2580\n",
      "Epoch [    7/   50] | d_loss: 0.4310 | g_loss: 0.2789\n",
      "Epoch [    7/   50] | d_loss: 0.5319 | g_loss: 0.2383\n",
      "Epoch [    7/   50] | d_loss: 0.5268 | g_loss: 0.3107\n",
      "Epoch [    7/   50] | d_loss: 0.5017 | g_loss: 0.2494\n",
      "Epoch [    7/   50] | d_loss: 0.5138 | g_loss: 0.2724\n",
      "Epoch [    7/   50] | d_loss: 0.5390 | g_loss: 0.2933\n",
      "Epoch [    7/   50] | d_loss: 0.5087 | g_loss: 0.3041\n",
      "Epoch [    7/   50] | d_loss: 0.4860 | g_loss: 0.2403\n",
      "Epoch [    7/   50] | d_loss: 0.5328 | g_loss: 0.2982\n",
      "Epoch [    7/   50] | d_loss: 0.4596 | g_loss: 0.3075\n",
      "Epoch [    7/   50] | d_loss: 0.5287 | g_loss: 0.2871\n",
      "Epoch [    7/   50] | d_loss: 0.4438 | g_loss: 0.2947\n",
      "Epoch [    7/   50] | d_loss: 0.5350 | g_loss: 0.2431\n",
      "Epoch [    7/   50] | d_loss: 0.4857 | g_loss: 0.2921\n",
      "Epoch [    7/   50] | d_loss: 0.5571 | g_loss: 0.2744\n",
      "Epoch [    7/   50] | d_loss: 0.4515 | g_loss: 0.2539\n",
      "Epoch [    7/   50] | d_loss: 0.4465 | g_loss: 0.3213\n",
      "Epoch [    7/   50] | d_loss: 0.6338 | g_loss: 0.2559\n",
      "Epoch [    7/   50] | d_loss: 0.5142 | g_loss: 0.2859\n",
      "Epoch [    7/   50] | d_loss: 0.4963 | g_loss: 0.2703\n",
      "Epoch [    7/   50] | d_loss: 0.5013 | g_loss: 0.2663\n",
      "Epoch [    7/   50] | d_loss: 0.4866 | g_loss: 0.2472\n",
      "Epoch [    7/   50] | d_loss: 0.4889 | g_loss: 0.2785\n",
      "Epoch [    7/   50] | d_loss: 0.4909 | g_loss: 0.2662\n",
      "Epoch [    7/   50] | d_loss: 0.4750 | g_loss: 0.3009\n",
      "Epoch [    7/   50] | d_loss: 0.4931 | g_loss: 0.2608\n",
      "Epoch [    7/   50] | d_loss: 0.5356 | g_loss: 0.2814\n",
      "Epoch [    7/   50] | d_loss: 0.5380 | g_loss: 0.2653\n",
      "Epoch [    7/   50] | d_loss: 0.4409 | g_loss: 0.2644\n",
      "Epoch [    7/   50] | d_loss: 0.5189 | g_loss: 0.2944\n",
      "Epoch [    7/   50] | d_loss: 0.4826 | g_loss: 0.2805\n",
      "Epoch [    7/   50] | d_loss: 0.4931 | g_loss: 0.2553\n",
      "Epoch [    7/   50] | d_loss: 0.5211 | g_loss: 0.2794\n",
      "Epoch [    7/   50] | d_loss: 0.4978 | g_loss: 0.2646\n",
      "Epoch [    7/   50] | d_loss: 0.4773 | g_loss: 0.2492\n",
      "Epoch [    7/   50] | d_loss: 0.4899 | g_loss: 0.2970\n",
      "Epoch [    7/   50] | d_loss: 0.5086 | g_loss: 0.2541\n",
      "Epoch [    7/   50] | d_loss: 0.4872 | g_loss: 0.3155\n",
      "Epoch [    7/   50] | d_loss: 0.4772 | g_loss: 0.3244\n",
      "Epoch [    7/   50] | d_loss: 0.5029 | g_loss: 0.2981\n",
      "Epoch [    7/   50] | d_loss: 0.4927 | g_loss: 0.2819\n",
      "Epoch [    7/   50] | d_loss: 0.4794 | g_loss: 0.3182\n",
      "Epoch [    7/   50] | d_loss: 0.4917 | g_loss: 0.3067\n",
      "Epoch [    7/   50] | d_loss: 0.5435 | g_loss: 0.2660\n",
      "Epoch [    7/   50] | d_loss: 0.4899 | g_loss: 0.2959\n",
      "Epoch [    7/   50] | d_loss: 0.5054 | g_loss: 0.3392\n",
      "Epoch [    7/   50] | d_loss: 0.5291 | g_loss: 0.2537\n",
      "Epoch [    7/   50] | d_loss: 0.4565 | g_loss: 0.2428\n",
      "Epoch [    7/   50] | d_loss: 0.4725 | g_loss: 0.2271\n",
      "Epoch [    7/   50] | d_loss: 0.4875 | g_loss: 0.2835\n",
      "Epoch [    7/   50] | d_loss: 0.5415 | g_loss: 0.2899\n",
      "Epoch [    7/   50] | d_loss: 0.4970 | g_loss: 0.2371\n",
      "Epoch [    8/   50] | d_loss: 0.4954 | g_loss: 0.2532\n",
      "Epoch [    8/   50] | d_loss: 0.5201 | g_loss: 0.3015\n",
      "Epoch [    8/   50] | d_loss: 0.4646 | g_loss: 0.2967\n",
      "Epoch [    8/   50] | d_loss: 0.4959 | g_loss: 0.2459\n",
      "Epoch [    8/   50] | d_loss: 0.4970 | g_loss: 0.2765\n",
      "Epoch [    8/   50] | d_loss: 0.5387 | g_loss: 0.2528\n",
      "Epoch [    8/   50] | d_loss: 0.5132 | g_loss: 0.2324\n",
      "Epoch [    8/   50] | d_loss: 0.4873 | g_loss: 0.2466\n",
      "Epoch [    8/   50] | d_loss: 0.4833 | g_loss: 0.2719\n",
      "Epoch [    8/   50] | d_loss: 0.5492 | g_loss: 0.2872\n",
      "Epoch [    8/   50] | d_loss: 0.5353 | g_loss: 0.2730\n",
      "Epoch [    8/   50] | d_loss: 0.4429 | g_loss: 0.2620\n",
      "Epoch [    8/   50] | d_loss: 0.4907 | g_loss: 0.2552\n",
      "Epoch [    8/   50] | d_loss: 0.5427 | g_loss: 0.2545\n",
      "Epoch [    8/   50] | d_loss: 0.4842 | g_loss: 0.2828\n",
      "Epoch [    8/   50] | d_loss: 0.4966 | g_loss: 0.2640\n",
      "Epoch [    8/   50] | d_loss: 0.4893 | g_loss: 0.2732\n",
      "Epoch [    8/   50] | d_loss: 0.4720 | g_loss: 0.2280\n",
      "Epoch [    8/   50] | d_loss: 0.5059 | g_loss: 0.3348\n",
      "Epoch [    8/   50] | d_loss: 0.4641 | g_loss: 0.2894\n",
      "Epoch [    8/   50] | d_loss: 0.4394 | g_loss: 0.2265\n",
      "Epoch [    8/   50] | d_loss: 0.4911 | g_loss: 0.2442\n",
      "Epoch [    8/   50] | d_loss: 0.4749 | g_loss: 0.2616\n",
      "Epoch [    8/   50] | d_loss: 0.4766 | g_loss: 0.2458\n",
      "Epoch [    8/   50] | d_loss: 0.4917 | g_loss: 0.2784\n",
      "Epoch [    8/   50] | d_loss: 0.4653 | g_loss: 0.2553\n",
      "Epoch [    8/   50] | d_loss: 0.4744 | g_loss: 0.2921\n",
      "Epoch [    8/   50] | d_loss: 0.4825 | g_loss: 0.2800\n",
      "Epoch [    8/   50] | d_loss: 0.4768 | g_loss: 0.2796\n",
      "Epoch [    8/   50] | d_loss: 0.4830 | g_loss: 0.2823\n",
      "Epoch [    8/   50] | d_loss: 0.4832 | g_loss: 0.2805\n",
      "Epoch [    8/   50] | d_loss: 0.4671 | g_loss: 0.2724\n",
      "Epoch [    8/   50] | d_loss: 0.4784 | g_loss: 0.2573\n",
      "Epoch [    8/   50] | d_loss: 0.5046 | g_loss: 0.2572\n",
      "Epoch [    8/   50] | d_loss: 0.5425 | g_loss: 0.2712\n",
      "Epoch [    8/   50] | d_loss: 0.4979 | g_loss: 0.2939\n",
      "Epoch [    8/   50] | d_loss: 0.4696 | g_loss: 0.2591\n",
      "Epoch [    8/   50] | d_loss: 0.5163 | g_loss: 0.2777\n",
      "Epoch [    8/   50] | d_loss: 0.5157 | g_loss: 0.3010\n",
      "Epoch [    8/   50] | d_loss: 0.4921 | g_loss: 0.2886\n",
      "Epoch [    8/   50] | d_loss: 0.4581 | g_loss: 0.2719\n",
      "Epoch [    8/   50] | d_loss: 0.4917 | g_loss: 0.2503\n",
      "Epoch [    8/   50] | d_loss: 0.5127 | g_loss: 0.2767\n",
      "Epoch [    8/   50] | d_loss: 0.4988 | g_loss: 0.2767\n",
      "Epoch [    8/   50] | d_loss: 0.4992 | g_loss: 0.2678\n",
      "Epoch [    8/   50] | d_loss: 0.4920 | g_loss: 0.2668\n",
      "Epoch [    8/   50] | d_loss: 0.5033 | g_loss: 0.2680\n",
      "Epoch [    8/   50] | d_loss: 0.4841 | g_loss: 0.2848\n",
      "Epoch [    8/   50] | d_loss: 0.5095 | g_loss: 0.2915\n",
      "Epoch [    8/   50] | d_loss: 0.4832 | g_loss: 0.2209\n",
      "Epoch [    8/   50] | d_loss: 0.4834 | g_loss: 0.2641\n",
      "Epoch [    8/   50] | d_loss: 0.5086 | g_loss: 0.3009\n",
      "Epoch [    8/   50] | d_loss: 0.4884 | g_loss: 0.2292\n",
      "Epoch [    8/   50] | d_loss: 0.4677 | g_loss: 0.2308\n",
      "Epoch [    8/   50] | d_loss: 0.5072 | g_loss: 0.3002\n",
      "Epoch [    8/   50] | d_loss: 0.5175 | g_loss: 0.2550\n",
      "Epoch [    8/   50] | d_loss: 0.4921 | g_loss: 0.2388\n",
      "Epoch [    8/   50] | d_loss: 0.5244 | g_loss: 0.2916\n",
      "Epoch [    8/   50] | d_loss: 0.4988 | g_loss: 0.2851\n",
      "Epoch [    8/   50] | d_loss: 0.5282 | g_loss: 0.2428\n",
      "Epoch [    8/   50] | d_loss: 0.4947 | g_loss: 0.2693\n",
      "Epoch [    8/   50] | d_loss: 0.5660 | g_loss: 0.2545\n",
      "Epoch [    8/   50] | d_loss: 0.5199 | g_loss: 0.2765\n",
      "Epoch [    8/   50] | d_loss: 0.4867 | g_loss: 0.2496\n",
      "Epoch [    8/   50] | d_loss: 0.5214 | g_loss: 0.2590\n",
      "Epoch [    8/   50] | d_loss: 0.4585 | g_loss: 0.2368\n",
      "Epoch [    8/   50] | d_loss: 0.4941 | g_loss: 0.2941\n",
      "Epoch [    8/   50] | d_loss: 0.4818 | g_loss: 0.2738\n",
      "Epoch [    8/   50] | d_loss: 0.4760 | g_loss: 0.2541\n",
      "Epoch [    8/   50] | d_loss: 0.5210 | g_loss: 0.2409\n",
      "Epoch [    8/   50] | d_loss: 0.4829 | g_loss: 0.2664\n",
      "Epoch [    8/   50] | d_loss: 0.4754 | g_loss: 0.2464\n",
      "Epoch [    8/   50] | d_loss: 0.5149 | g_loss: 0.3087\n",
      "Epoch [    8/   50] | d_loss: 0.5288 | g_loss: 0.2289\n",
      "Epoch [    8/   50] | d_loss: 0.4576 | g_loss: 0.2896\n",
      "Epoch [    8/   50] | d_loss: 0.4548 | g_loss: 0.2515\n",
      "Epoch [    8/   50] | d_loss: 0.5194 | g_loss: 0.2740\n",
      "Epoch [    8/   50] | d_loss: 0.5248 | g_loss: 0.2791\n",
      "Epoch [    8/   50] | d_loss: 0.4544 | g_loss: 0.2958\n",
      "Epoch [    8/   50] | d_loss: 0.5169 | g_loss: 0.2825\n",
      "Epoch [    8/   50] | d_loss: 0.4986 | g_loss: 0.2604\n",
      "Epoch [    8/   50] | d_loss: 0.5573 | g_loss: 0.2639\n",
      "Epoch [    8/   50] | d_loss: 0.5088 | g_loss: 0.2773\n",
      "Epoch [    8/   50] | d_loss: 0.5069 | g_loss: 0.2841\n",
      "Epoch [    8/   50] | d_loss: 0.5192 | g_loss: 0.2858\n",
      "Epoch [    8/   50] | d_loss: 0.4772 | g_loss: 0.2626\n",
      "Epoch [    8/   50] | d_loss: 0.4378 | g_loss: 0.2898\n",
      "Epoch [    8/   50] | d_loss: 0.4777 | g_loss: 0.2388\n",
      "Epoch [    8/   50] | d_loss: 0.5344 | g_loss: 0.2817\n",
      "Epoch [    8/   50] | d_loss: 0.4439 | g_loss: 0.2558\n",
      "Epoch [    9/   50] | d_loss: 0.5290 | g_loss: 0.2441\n",
      "Epoch [    9/   50] | d_loss: 0.5096 | g_loss: 0.2415\n",
      "Epoch [    9/   50] | d_loss: 0.4929 | g_loss: 0.2327\n",
      "Epoch [    9/   50] | d_loss: 0.5278 | g_loss: 0.2576\n",
      "Epoch [    9/   50] | d_loss: 0.4639 | g_loss: 0.2541\n",
      "Epoch [    9/   50] | d_loss: 0.5194 | g_loss: 0.2821\n",
      "Epoch [    9/   50] | d_loss: 0.5190 | g_loss: 0.2415\n",
      "Epoch [    9/   50] | d_loss: 0.4573 | g_loss: 0.3084\n",
      "Epoch [    9/   50] | d_loss: 0.4792 | g_loss: 0.2650\n",
      "Epoch [    9/   50] | d_loss: 0.4927 | g_loss: 0.2649\n",
      "Epoch [    9/   50] | d_loss: 0.4832 | g_loss: 0.2621\n",
      "Epoch [    9/   50] | d_loss: 0.5186 | g_loss: 0.2356\n",
      "Epoch [    9/   50] | d_loss: 0.4608 | g_loss: 0.2950\n",
      "Epoch [    9/   50] | d_loss: 0.4777 | g_loss: 0.2575\n",
      "Epoch [    9/   50] | d_loss: 0.4850 | g_loss: 0.3031\n",
      "Epoch [    9/   50] | d_loss: 0.4938 | g_loss: 0.2358\n",
      "Epoch [    9/   50] | d_loss: 0.5281 | g_loss: 0.3198\n",
      "Epoch [    9/   50] | d_loss: 0.5090 | g_loss: 0.2643\n",
      "Epoch [    9/   50] | d_loss: 0.4617 | g_loss: 0.2671\n",
      "Epoch [    9/   50] | d_loss: 0.4886 | g_loss: 0.2858\n",
      "Epoch [    9/   50] | d_loss: 0.4749 | g_loss: 0.2917\n",
      "Epoch [    9/   50] | d_loss: 0.5409 | g_loss: 0.2543\n",
      "Epoch [    9/   50] | d_loss: 0.5046 | g_loss: 0.2704\n",
      "Epoch [    9/   50] | d_loss: 0.5182 | g_loss: 0.2696\n",
      "Epoch [    9/   50] | d_loss: 0.4658 | g_loss: 0.2423\n",
      "Epoch [    9/   50] | d_loss: 0.5306 | g_loss: 0.2600\n",
      "Epoch [    9/   50] | d_loss: 0.4632 | g_loss: 0.2224\n",
      "Epoch [    9/   50] | d_loss: 0.4886 | g_loss: 0.2345\n",
      "Epoch [    9/   50] | d_loss: 0.4392 | g_loss: 0.2939\n",
      "Epoch [    9/   50] | d_loss: 0.5563 | g_loss: 0.3074\n",
      "Epoch [    9/   50] | d_loss: 0.5028 | g_loss: 0.2460\n",
      "Epoch [    9/   50] | d_loss: 0.5680 | g_loss: 0.2452\n",
      "Epoch [    9/   50] | d_loss: 0.5176 | g_loss: 0.2693\n",
      "Epoch [    9/   50] | d_loss: 0.4970 | g_loss: 0.2583\n",
      "Epoch [    9/   50] | d_loss: 0.5387 | g_loss: 0.2464\n",
      "Epoch [    9/   50] | d_loss: 0.4728 | g_loss: 0.2197\n",
      "Epoch [    9/   50] | d_loss: 0.5218 | g_loss: 0.2273\n",
      "Epoch [    9/   50] | d_loss: 0.4730 | g_loss: 0.2532\n",
      "Epoch [    9/   50] | d_loss: 0.5199 | g_loss: 0.2702\n",
      "Epoch [    9/   50] | d_loss: 0.5380 | g_loss: 0.2530\n",
      "Epoch [    9/   50] | d_loss: 0.5209 | g_loss: 0.2451\n",
      "Epoch [    9/   50] | d_loss: 0.4815 | g_loss: 0.2244\n",
      "Epoch [    9/   50] | d_loss: 0.5144 | g_loss: 0.2530\n",
      "Epoch [    9/   50] | d_loss: 0.4621 | g_loss: 0.2558\n",
      "Epoch [    9/   50] | d_loss: 0.4959 | g_loss: 0.2872\n",
      "Epoch [    9/   50] | d_loss: 0.5123 | g_loss: 0.2894\n",
      "Epoch [    9/   50] | d_loss: 0.5700 | g_loss: 0.2632\n",
      "Epoch [    9/   50] | d_loss: 0.4673 | g_loss: 0.2564\n",
      "Epoch [    9/   50] | d_loss: 0.4852 | g_loss: 0.3032\n",
      "Epoch [    9/   50] | d_loss: 0.4924 | g_loss: 0.2580\n",
      "Epoch [    9/   50] | d_loss: 0.5176 | g_loss: 0.2774\n",
      "Epoch [    9/   50] | d_loss: 0.4835 | g_loss: 0.2615\n",
      "Epoch [    9/   50] | d_loss: 0.5254 | g_loss: 0.2671\n",
      "Epoch [    9/   50] | d_loss: 0.4738 | g_loss: 0.3098\n",
      "Epoch [    9/   50] | d_loss: 0.4779 | g_loss: 0.2806\n",
      "Epoch [    9/   50] | d_loss: 0.5062 | g_loss: 0.2637\n",
      "Epoch [    9/   50] | d_loss: 0.4378 | g_loss: 0.3007\n",
      "Epoch [    9/   50] | d_loss: 0.4480 | g_loss: 0.2668\n",
      "Epoch [    9/   50] | d_loss: 0.5007 | g_loss: 0.2481\n",
      "Epoch [    9/   50] | d_loss: 0.4935 | g_loss: 0.3202\n",
      "Epoch [    9/   50] | d_loss: 0.5113 | g_loss: 0.2595\n",
      "Epoch [    9/   50] | d_loss: 0.4945 | g_loss: 0.2285\n",
      "Epoch [    9/   50] | d_loss: 0.4853 | g_loss: 0.2807\n",
      "Epoch [    9/   50] | d_loss: 0.4994 | g_loss: 0.2408\n",
      "Epoch [    9/   50] | d_loss: 0.4756 | g_loss: 0.2441\n",
      "Epoch [    9/   50] | d_loss: 0.5022 | g_loss: 0.2400\n",
      "Epoch [    9/   50] | d_loss: 0.4748 | g_loss: 0.2523\n",
      "Epoch [    9/   50] | d_loss: 0.5690 | g_loss: 0.2384\n",
      "Epoch [    9/   50] | d_loss: 0.5235 | g_loss: 0.2794\n",
      "Epoch [    9/   50] | d_loss: 0.5241 | g_loss: 0.3097\n",
      "Epoch [    9/   50] | d_loss: 0.5063 | g_loss: 0.2715\n",
      "Epoch [    9/   50] | d_loss: 0.4974 | g_loss: 0.2797\n",
      "Epoch [    9/   50] | d_loss: 0.4551 | g_loss: 0.2946\n",
      "Epoch [    9/   50] | d_loss: 0.5083 | g_loss: 0.2767\n",
      "Epoch [    9/   50] | d_loss: 0.4672 | g_loss: 0.2619\n",
      "Epoch [    9/   50] | d_loss: 0.5010 | g_loss: 0.2648\n",
      "Epoch [    9/   50] | d_loss: 0.5343 | g_loss: 0.2820\n",
      "Epoch [    9/   50] | d_loss: 0.4610 | g_loss: 0.2618\n",
      "Epoch [    9/   50] | d_loss: 0.5070 | g_loss: 0.2717\n",
      "Epoch [    9/   50] | d_loss: 0.4885 | g_loss: 0.2951\n",
      "Epoch [    9/   50] | d_loss: 0.4983 | g_loss: 0.2916\n",
      "Epoch [    9/   50] | d_loss: 0.5174 | g_loss: 0.2556\n",
      "Epoch [    9/   50] | d_loss: 0.4715 | g_loss: 0.2973\n",
      "Epoch [    9/   50] | d_loss: 0.5186 | g_loss: 0.2391\n",
      "Epoch [    9/   50] | d_loss: 0.4983 | g_loss: 0.2343\n",
      "Epoch [    9/   50] | d_loss: 0.5348 | g_loss: 0.2645\n",
      "Epoch [    9/   50] | d_loss: 0.5171 | g_loss: 0.2880\n",
      "Epoch [    9/   50] | d_loss: 0.5050 | g_loss: 0.2907\n",
      "Epoch [    9/   50] | d_loss: 0.4900 | g_loss: 0.3136\n",
      "Epoch [    9/   50] | d_loss: 0.4814 | g_loss: 0.2781\n",
      "Epoch [   10/   50] | d_loss: 0.5140 | g_loss: 0.2528\n",
      "Epoch [   10/   50] | d_loss: 0.4836 | g_loss: 0.2248\n",
      "Epoch [   10/   50] | d_loss: 0.4640 | g_loss: 0.2707\n",
      "Epoch [   10/   50] | d_loss: 0.5151 | g_loss: 0.2864\n",
      "Epoch [   10/   50] | d_loss: 0.5157 | g_loss: 0.2755\n",
      "Epoch [   10/   50] | d_loss: 0.4740 | g_loss: 0.2277\n",
      "Epoch [   10/   50] | d_loss: 0.4362 | g_loss: 0.2723\n",
      "Epoch [   10/   50] | d_loss: 0.5056 | g_loss: 0.2987\n",
      "Epoch [   10/   50] | d_loss: 0.4754 | g_loss: 0.2809\n",
      "Epoch [   10/   50] | d_loss: 0.5413 | g_loss: 0.2529\n",
      "Epoch [   10/   50] | d_loss: 0.4995 | g_loss: 0.2695\n",
      "Epoch [   10/   50] | d_loss: 0.4706 | g_loss: 0.2877\n",
      "Epoch [   10/   50] | d_loss: 0.5047 | g_loss: 0.2824\n",
      "Epoch [   10/   50] | d_loss: 0.5501 | g_loss: 0.3030\n",
      "Epoch [   10/   50] | d_loss: 0.4986 | g_loss: 0.2861\n",
      "Epoch [   10/   50] | d_loss: 0.5408 | g_loss: 0.2699\n",
      "Epoch [   10/   50] | d_loss: 0.4650 | g_loss: 0.2900\n",
      "Epoch [   10/   50] | d_loss: 0.4647 | g_loss: 0.2706\n",
      "Epoch [   10/   50] | d_loss: 0.5073 | g_loss: 0.2850\n",
      "Epoch [   10/   50] | d_loss: 0.4446 | g_loss: 0.2749\n",
      "Epoch [   10/   50] | d_loss: 0.4364 | g_loss: 0.2801\n",
      "Epoch [   10/   50] | d_loss: 0.5188 | g_loss: 0.2326\n",
      "Epoch [   10/   50] | d_loss: 0.5317 | g_loss: 0.2696\n",
      "Epoch [   10/   50] | d_loss: 0.5654 | g_loss: 0.2628\n",
      "Epoch [   10/   50] | d_loss: 0.4925 | g_loss: 0.2421\n",
      "Epoch [   10/   50] | d_loss: 0.4848 | g_loss: 0.2489\n",
      "Epoch [   10/   50] | d_loss: 0.5318 | g_loss: 0.2510\n",
      "Epoch [   10/   50] | d_loss: 0.5026 | g_loss: 0.2230\n",
      "Epoch [   10/   50] | d_loss: 0.4936 | g_loss: 0.2781\n",
      "Epoch [   10/   50] | d_loss: 0.4981 | g_loss: 0.2890\n",
      "Epoch [   10/   50] | d_loss: 0.5311 | g_loss: 0.2645\n",
      "Epoch [   10/   50] | d_loss: 0.4803 | g_loss: 0.2673\n",
      "Epoch [   10/   50] | d_loss: 0.4579 | g_loss: 0.2826\n",
      "Epoch [   10/   50] | d_loss: 0.4545 | g_loss: 0.2671\n",
      "Epoch [   10/   50] | d_loss: 0.5174 | g_loss: 0.2437\n",
      "Epoch [   10/   50] | d_loss: 0.5341 | g_loss: 0.2567\n",
      "Epoch [   10/   50] | d_loss: 0.5355 | g_loss: 0.2646\n",
      "Epoch [   10/   50] | d_loss: 0.4765 | g_loss: 0.2795\n",
      "Epoch [   10/   50] | d_loss: 0.5072 | g_loss: 0.2690\n",
      "Epoch [   10/   50] | d_loss: 0.5191 | g_loss: 0.2568\n",
      "Epoch [   10/   50] | d_loss: 0.4952 | g_loss: 0.2232\n",
      "Epoch [   10/   50] | d_loss: 0.4706 | g_loss: 0.2621\n",
      "Epoch [   10/   50] | d_loss: 0.5004 | g_loss: 0.2524\n",
      "Epoch [   10/   50] | d_loss: 0.5468 | g_loss: 0.2045\n",
      "Epoch [   10/   50] | d_loss: 0.4943 | g_loss: 0.2385\n",
      "Epoch [   10/   50] | d_loss: 0.4853 | g_loss: 0.2949\n",
      "Epoch [   10/   50] | d_loss: 0.4880 | g_loss: 0.2592\n",
      "Epoch [   10/   50] | d_loss: 0.5269 | g_loss: 0.2545\n",
      "Epoch [   10/   50] | d_loss: 0.5490 | g_loss: 0.2742\n",
      "Epoch [   10/   50] | d_loss: 0.5146 | g_loss: 0.2572\n",
      "Epoch [   10/   50] | d_loss: 0.4658 | g_loss: 0.2982\n",
      "Epoch [   10/   50] | d_loss: 0.5254 | g_loss: 0.2856\n",
      "Epoch [   10/   50] | d_loss: 0.5284 | g_loss: 0.2846\n",
      "Epoch [   10/   50] | d_loss: 0.5003 | g_loss: 0.2520\n",
      "Epoch [   10/   50] | d_loss: 0.4712 | g_loss: 0.2404\n",
      "Epoch [   10/   50] | d_loss: 0.5126 | g_loss: 0.2618\n",
      "Epoch [   10/   50] | d_loss: 0.4922 | g_loss: 0.2679\n",
      "Epoch [   10/   50] | d_loss: 0.5051 | g_loss: 0.2863\n",
      "Epoch [   10/   50] | d_loss: 0.5536 | g_loss: 0.2753\n",
      "Epoch [   10/   50] | d_loss: 0.4759 | g_loss: 0.2702\n",
      "Epoch [   10/   50] | d_loss: 0.5006 | g_loss: 0.2651\n",
      "Epoch [   10/   50] | d_loss: 0.5110 | g_loss: 0.2583\n",
      "Epoch [   10/   50] | d_loss: 0.4820 | g_loss: 0.2423\n",
      "Epoch [   10/   50] | d_loss: 0.4918 | g_loss: 0.2963\n",
      "Epoch [   10/   50] | d_loss: 0.4955 | g_loss: 0.2632\n",
      "Epoch [   10/   50] | d_loss: 0.5147 | g_loss: 0.2431\n",
      "Epoch [   10/   50] | d_loss: 0.5316 | g_loss: 0.2553\n",
      "Epoch [   10/   50] | d_loss: 0.5025 | g_loss: 0.2588\n",
      "Epoch [   10/   50] | d_loss: 0.5057 | g_loss: 0.2764\n",
      "Epoch [   10/   50] | d_loss: 0.4942 | g_loss: 0.2570\n",
      "Epoch [   10/   50] | d_loss: 0.4943 | g_loss: 0.2692\n",
      "Epoch [   10/   50] | d_loss: 0.4960 | g_loss: 0.2412\n",
      "Epoch [   10/   50] | d_loss: 0.5076 | g_loss: 0.2471\n",
      "Epoch [   10/   50] | d_loss: 0.4993 | g_loss: 0.2795\n",
      "Epoch [   10/   50] | d_loss: 0.5393 | g_loss: 0.2340\n",
      "Epoch [   10/   50] | d_loss: 0.5245 | g_loss: 0.2657\n",
      "Epoch [   10/   50] | d_loss: 0.4910 | g_loss: 0.2558\n",
      "Epoch [   10/   50] | d_loss: 0.5238 | g_loss: 0.2398\n",
      "Epoch [   10/   50] | d_loss: 0.4867 | g_loss: 0.2822\n",
      "Epoch [   10/   50] | d_loss: 0.4958 | g_loss: 0.2680\n",
      "Epoch [   10/   50] | d_loss: 0.4800 | g_loss: 0.2778\n",
      "Epoch [   10/   50] | d_loss: 0.5183 | g_loss: 0.2512\n",
      "Epoch [   10/   50] | d_loss: 0.4709 | g_loss: 0.2615\n",
      "Epoch [   10/   50] | d_loss: 0.5114 | g_loss: 0.2571\n",
      "Epoch [   10/   50] | d_loss: 0.4895 | g_loss: 0.2590\n",
      "Epoch [   10/   50] | d_loss: 0.5419 | g_loss: 0.2477\n",
      "Epoch [   10/   50] | d_loss: 0.5355 | g_loss: 0.2434\n",
      "Epoch [   10/   50] | d_loss: 0.4659 | g_loss: 0.2695\n",
      "Epoch [   10/   50] | d_loss: 0.5033 | g_loss: 0.2384\n",
      "Epoch [   10/   50] | d_loss: 0.5135 | g_loss: 0.2848\n",
      "Epoch [   11/   50] | d_loss: 0.5150 | g_loss: 0.2597\n",
      "Epoch [   11/   50] | d_loss: 0.5310 | g_loss: 0.2350\n",
      "Epoch [   11/   50] | d_loss: 0.5033 | g_loss: 0.2350\n",
      "Epoch [   11/   50] | d_loss: 0.5065 | g_loss: 0.2729\n",
      "Epoch [   11/   50] | d_loss: 0.4844 | g_loss: 0.2366\n",
      "Epoch [   11/   50] | d_loss: 0.5037 | g_loss: 0.2688\n",
      "Epoch [   11/   50] | d_loss: 0.5299 | g_loss: 0.2462\n",
      "Epoch [   11/   50] | d_loss: 0.4695 | g_loss: 0.2879\n",
      "Epoch [   11/   50] | d_loss: 0.5145 | g_loss: 0.2444\n",
      "Epoch [   11/   50] | d_loss: 0.5312 | g_loss: 0.2787\n",
      "Epoch [   11/   50] | d_loss: 0.4579 | g_loss: 0.2521\n",
      "Epoch [   11/   50] | d_loss: 0.4964 | g_loss: 0.3380\n",
      "Epoch [   11/   50] | d_loss: 0.4986 | g_loss: 0.2393\n",
      "Epoch [   11/   50] | d_loss: 0.5447 | g_loss: 0.2600\n",
      "Epoch [   11/   50] | d_loss: 0.5397 | g_loss: 0.2924\n",
      "Epoch [   11/   50] | d_loss: 0.4756 | g_loss: 0.2769\n",
      "Epoch [   11/   50] | d_loss: 0.5253 | g_loss: 0.2664\n",
      "Epoch [   11/   50] | d_loss: 0.5249 | g_loss: 0.2796\n",
      "Epoch [   11/   50] | d_loss: 0.4687 | g_loss: 0.2407\n",
      "Epoch [   11/   50] | d_loss: 0.5283 | g_loss: 0.2758\n",
      "Epoch [   11/   50] | d_loss: 0.4685 | g_loss: 0.2565\n",
      "Epoch [   11/   50] | d_loss: 0.5170 | g_loss: 0.2781\n",
      "Epoch [   11/   50] | d_loss: 0.5301 | g_loss: 0.2446\n",
      "Epoch [   11/   50] | d_loss: 0.5245 | g_loss: 0.2467\n",
      "Epoch [   11/   50] | d_loss: 0.5347 | g_loss: 0.2283\n",
      "Epoch [   11/   50] | d_loss: 0.4643 | g_loss: 0.2910\n",
      "Epoch [   11/   50] | d_loss: 0.5067 | g_loss: 0.2333\n",
      "Epoch [   11/   50] | d_loss: 0.5083 | g_loss: 0.2454\n",
      "Epoch [   11/   50] | d_loss: 0.5199 | g_loss: 0.2871\n",
      "Epoch [   11/   50] | d_loss: 0.4761 | g_loss: 0.2808\n",
      "Epoch [   11/   50] | d_loss: 0.5345 | g_loss: 0.2422\n",
      "Epoch [   11/   50] | d_loss: 0.5061 | g_loss: 0.2729\n",
      "Epoch [   11/   50] | d_loss: 0.4840 | g_loss: 0.2640\n",
      "Epoch [   11/   50] | d_loss: 0.5211 | g_loss: 0.2430\n",
      "Epoch [   11/   50] | d_loss: 0.4991 | g_loss: 0.2725\n",
      "Epoch [   11/   50] | d_loss: 0.4912 | g_loss: 0.2579\n",
      "Epoch [   11/   50] | d_loss: 0.4835 | g_loss: 0.2860\n",
      "Epoch [   11/   50] | d_loss: 0.4936 | g_loss: 0.2738\n",
      "Epoch [   11/   50] | d_loss: 0.4905 | g_loss: 0.2973\n",
      "Epoch [   11/   50] | d_loss: 0.4807 | g_loss: 0.2578\n",
      "Epoch [   11/   50] | d_loss: 0.5427 | g_loss: 0.2848\n",
      "Epoch [   11/   50] | d_loss: 0.4762 | g_loss: 0.2504\n",
      "Epoch [   11/   50] | d_loss: 0.5369 | g_loss: 0.2380\n",
      "Epoch [   11/   50] | d_loss: 0.5020 | g_loss: 0.2431\n",
      "Epoch [   11/   50] | d_loss: 0.5314 | g_loss: 0.2738\n",
      "Epoch [   11/   50] | d_loss: 0.4837 | g_loss: 0.2690\n",
      "Epoch [   11/   50] | d_loss: 0.4867 | g_loss: 0.2776\n",
      "Epoch [   11/   50] | d_loss: 0.5437 | g_loss: 0.2837\n",
      "Epoch [   11/   50] | d_loss: 0.5131 | g_loss: 0.2614\n",
      "Epoch [   11/   50] | d_loss: 0.5355 | g_loss: 0.2720\n",
      "Epoch [   11/   50] | d_loss: 0.4706 | g_loss: 0.2701\n",
      "Epoch [   11/   50] | d_loss: 0.5503 | g_loss: 0.2605\n",
      "Epoch [   11/   50] | d_loss: 0.4884 | g_loss: 0.2608\n",
      "Epoch [   11/   50] | d_loss: 0.5402 | g_loss: 0.2730\n",
      "Epoch [   11/   50] | d_loss: 0.4916 | g_loss: 0.2364\n",
      "Epoch [   11/   50] | d_loss: 0.5289 | g_loss: 0.2563\n",
      "Epoch [   11/   50] | d_loss: 0.5213 | g_loss: 0.2901\n",
      "Epoch [   11/   50] | d_loss: 0.5075 | g_loss: 0.2658\n",
      "Epoch [   11/   50] | d_loss: 0.4618 | g_loss: 0.2522\n",
      "Epoch [   11/   50] | d_loss: 0.5062 | g_loss: 0.2791\n",
      "Epoch [   11/   50] | d_loss: 0.4659 | g_loss: 0.2755\n",
      "Epoch [   11/   50] | d_loss: 0.4808 | g_loss: 0.2799\n",
      "Epoch [   11/   50] | d_loss: 0.5322 | g_loss: 0.2778\n",
      "Epoch [   11/   50] | d_loss: 0.4551 | g_loss: 0.2761\n",
      "Epoch [   11/   50] | d_loss: 0.4972 | g_loss: 0.2486\n",
      "Epoch [   11/   50] | d_loss: 0.5345 | g_loss: 0.2455\n",
      "Epoch [   11/   50] | d_loss: 0.4872 | g_loss: 0.2512\n",
      "Epoch [   11/   50] | d_loss: 0.5128 | g_loss: 0.2476\n",
      "Epoch [   11/   50] | d_loss: 0.4991 | g_loss: 0.2566\n",
      "Epoch [   11/   50] | d_loss: 0.5057 | g_loss: 0.2453\n",
      "Epoch [   11/   50] | d_loss: 0.5361 | g_loss: 0.2191\n",
      "Epoch [   11/   50] | d_loss: 0.5517 | g_loss: 0.2847\n",
      "Epoch [   11/   50] | d_loss: 0.5340 | g_loss: 0.2476\n",
      "Epoch [   11/   50] | d_loss: 0.5025 | g_loss: 0.2527\n",
      "Epoch [   11/   50] | d_loss: 0.4900 | g_loss: 0.2463\n",
      "Epoch [   11/   50] | d_loss: 0.4891 | g_loss: 0.2648\n",
      "Epoch [   11/   50] | d_loss: 0.5414 | g_loss: 0.2482\n",
      "Epoch [   11/   50] | d_loss: 0.4747 | g_loss: 0.2439\n",
      "Epoch [   11/   50] | d_loss: 0.5065 | g_loss: 0.2934\n",
      "Epoch [   11/   50] | d_loss: 0.4850 | g_loss: 0.2843\n",
      "Epoch [   11/   50] | d_loss: 0.5127 | g_loss: 0.2623\n",
      "Epoch [   11/   50] | d_loss: 0.4940 | g_loss: 0.2678\n",
      "Epoch [   11/   50] | d_loss: 0.5100 | g_loss: 0.2426\n",
      "Epoch [   11/   50] | d_loss: 0.5072 | g_loss: 0.2506\n",
      "Epoch [   11/   50] | d_loss: 0.5018 | g_loss: 0.2616\n",
      "Epoch [   11/   50] | d_loss: 0.5325 | g_loss: 0.2546\n",
      "Epoch [   11/   50] | d_loss: 0.5318 | g_loss: 0.2230\n",
      "Epoch [   11/   50] | d_loss: 0.5039 | g_loss: 0.2414\n",
      "Epoch [   11/   50] | d_loss: 0.4879 | g_loss: 0.2741\n",
      "Epoch [   11/   50] | d_loss: 0.5540 | g_loss: 0.2631\n",
      "Epoch [   12/   50] | d_loss: 0.5055 | g_loss: 0.2615\n",
      "Epoch [   12/   50] | d_loss: 0.5008 | g_loss: 0.2543\n",
      "Epoch [   12/   50] | d_loss: 0.5046 | g_loss: 0.2433\n",
      "Epoch [   12/   50] | d_loss: 0.5421 | g_loss: 0.2661\n",
      "Epoch [   12/   50] | d_loss: 0.4754 | g_loss: 0.2533\n",
      "Epoch [   12/   50] | d_loss: 0.5044 | g_loss: 0.2384\n",
      "Epoch [   12/   50] | d_loss: 0.5423 | g_loss: 0.2596\n",
      "Epoch [   12/   50] | d_loss: 0.5352 | g_loss: 0.2543\n",
      "Epoch [   12/   50] | d_loss: 0.4929 | g_loss: 0.2598\n",
      "Epoch [   12/   50] | d_loss: 0.5338 | g_loss: 0.3012\n",
      "Epoch [   12/   50] | d_loss: 0.4911 | g_loss: 0.2843\n",
      "Epoch [   12/   50] | d_loss: 0.5202 | g_loss: 0.2620\n",
      "Epoch [   12/   50] | d_loss: 0.5138 | g_loss: 0.2760\n",
      "Epoch [   12/   50] | d_loss: 0.5030 | g_loss: 0.2808\n",
      "Epoch [   12/   50] | d_loss: 0.4867 | g_loss: 0.2531\n",
      "Epoch [   12/   50] | d_loss: 0.4951 | g_loss: 0.2892\n",
      "Epoch [   12/   50] | d_loss: 0.4637 | g_loss: 0.2450\n",
      "Epoch [   12/   50] | d_loss: 0.5203 | g_loss: 0.2574\n",
      "Epoch [   12/   50] | d_loss: 0.4688 | g_loss: 0.2433\n",
      "Epoch [   12/   50] | d_loss: 0.4993 | g_loss: 0.2399\n",
      "Epoch [   12/   50] | d_loss: 0.5101 | g_loss: 0.2541\n",
      "Epoch [   12/   50] | d_loss: 0.5064 | g_loss: 0.2640\n",
      "Epoch [   12/   50] | d_loss: 0.5023 | g_loss: 0.2634\n",
      "Epoch [   12/   50] | d_loss: 0.5218 | g_loss: 0.2400\n",
      "Epoch [   12/   50] | d_loss: 0.4932 | g_loss: 0.2540\n",
      "Epoch [   12/   50] | d_loss: 0.4748 | g_loss: 0.2388\n",
      "Epoch [   12/   50] | d_loss: 0.5052 | g_loss: 0.2681\n",
      "Epoch [   12/   50] | d_loss: 0.4799 | g_loss: 0.3053\n",
      "Epoch [   12/   50] | d_loss: 0.4859 | g_loss: 0.2498\n",
      "Epoch [   12/   50] | d_loss: 0.4899 | g_loss: 0.2636\n",
      "Epoch [   12/   50] | d_loss: 0.4925 | g_loss: 0.2962\n",
      "Epoch [   12/   50] | d_loss: 0.4786 | g_loss: 0.2438\n",
      "Epoch [   12/   50] | d_loss: 0.5441 | g_loss: 0.3064\n",
      "Epoch [   12/   50] | d_loss: 0.5471 | g_loss: 0.2709\n",
      "Epoch [   12/   50] | d_loss: 0.4814 | g_loss: 0.2287\n",
      "Epoch [   12/   50] | d_loss: 0.5177 | g_loss: 0.2775\n",
      "Epoch [   12/   50] | d_loss: 0.5508 | g_loss: 0.2886\n",
      "Epoch [   12/   50] | d_loss: 0.5045 | g_loss: 0.2755\n",
      "Epoch [   12/   50] | d_loss: 0.4751 | g_loss: 0.2679\n",
      "Epoch [   12/   50] | d_loss: 0.4843 | g_loss: 0.2434\n",
      "Epoch [   12/   50] | d_loss: 0.5228 | g_loss: 0.2548\n",
      "Epoch [   12/   50] | d_loss: 0.4826 | g_loss: 0.2317\n",
      "Epoch [   12/   50] | d_loss: 0.5046 | g_loss: 0.2450\n",
      "Epoch [   12/   50] | d_loss: 0.5073 | g_loss: 0.2700\n",
      "Epoch [   12/   50] | d_loss: 0.4843 | g_loss: 0.2641\n",
      "Epoch [   12/   50] | d_loss: 0.5351 | g_loss: 0.2696\n",
      "Epoch [   12/   50] | d_loss: 0.4912 | g_loss: 0.2471\n",
      "Epoch [   12/   50] | d_loss: 0.4979 | g_loss: 0.3206\n",
      "Epoch [   12/   50] | d_loss: 0.5295 | g_loss: 0.2481\n",
      "Epoch [   12/   50] | d_loss: 0.4923 | g_loss: 0.2560\n",
      "Epoch [   12/   50] | d_loss: 0.5166 | g_loss: 0.2440\n",
      "Epoch [   12/   50] | d_loss: 0.4964 | g_loss: 0.2440\n",
      "Epoch [   12/   50] | d_loss: 0.5037 | g_loss: 0.2436\n",
      "Epoch [   12/   50] | d_loss: 0.5271 | g_loss: 0.2656\n",
      "Epoch [   12/   50] | d_loss: 0.5012 | g_loss: 0.2531\n",
      "Epoch [   12/   50] | d_loss: 0.5358 | g_loss: 0.2623\n",
      "Epoch [   12/   50] | d_loss: 0.5102 | g_loss: 0.2918\n",
      "Epoch [   12/   50] | d_loss: 0.5045 | g_loss: 0.2581\n",
      "Epoch [   12/   50] | d_loss: 0.4771 | g_loss: 0.2294\n",
      "Epoch [   12/   50] | d_loss: 0.4680 | g_loss: 0.2887\n",
      "Epoch [   12/   50] | d_loss: 0.5068 | g_loss: 0.2381\n",
      "Epoch [   12/   50] | d_loss: 0.4996 | g_loss: 0.2546\n",
      "Epoch [   12/   50] | d_loss: 0.5030 | g_loss: 0.2526\n",
      "Epoch [   12/   50] | d_loss: 0.5147 | g_loss: 0.2589\n",
      "Epoch [   12/   50] | d_loss: 0.5067 | g_loss: 0.2698\n",
      "Epoch [   12/   50] | d_loss: 0.4681 | g_loss: 0.2381\n",
      "Epoch [   12/   50] | d_loss: 0.4874 | g_loss: 0.2627\n",
      "Epoch [   12/   50] | d_loss: 0.4637 | g_loss: 0.2799\n",
      "Epoch [   12/   50] | d_loss: 0.5734 | g_loss: 0.2904\n",
      "Epoch [   12/   50] | d_loss: 0.4924 | g_loss: 0.2566\n",
      "Epoch [   12/   50] | d_loss: 0.4993 | g_loss: 0.2458\n",
      "Epoch [   12/   50] | d_loss: 0.4932 | g_loss: 0.2598\n",
      "Epoch [   12/   50] | d_loss: 0.4735 | g_loss: 0.2403\n",
      "Epoch [   12/   50] | d_loss: 0.4971 | g_loss: 0.2908\n",
      "Epoch [   12/   50] | d_loss: 0.5213 | g_loss: 0.2420\n",
      "Epoch [   12/   50] | d_loss: 0.4935 | g_loss: 0.2703\n",
      "Epoch [   12/   50] | d_loss: 0.4727 | g_loss: 0.2512\n",
      "Epoch [   12/   50] | d_loss: 0.4789 | g_loss: 0.2422\n",
      "Epoch [   12/   50] | d_loss: 0.5074 | g_loss: 0.2369\n",
      "Epoch [   12/   50] | d_loss: 0.5018 | g_loss: 0.2522\n",
      "Epoch [   12/   50] | d_loss: 0.5146 | g_loss: 0.2544\n",
      "Epoch [   12/   50] | d_loss: 0.5004 | g_loss: 0.2473\n",
      "Epoch [   12/   50] | d_loss: 0.4981 | g_loss: 0.2418\n",
      "Epoch [   12/   50] | d_loss: 0.5440 | g_loss: 0.3141\n",
      "Epoch [   12/   50] | d_loss: 0.4954 | g_loss: 0.2687\n",
      "Epoch [   12/   50] | d_loss: 0.5607 | g_loss: 0.2688\n",
      "Epoch [   12/   50] | d_loss: 0.4557 | g_loss: 0.2734\n",
      "Epoch [   12/   50] | d_loss: 0.4742 | g_loss: 0.2398\n",
      "Epoch [   12/   50] | d_loss: 0.4963 | g_loss: 0.2070\n",
      "Epoch [   12/   50] | d_loss: 0.5416 | g_loss: 0.2404\n",
      "Epoch [   13/   50] | d_loss: 0.4875 | g_loss: 0.2262\n",
      "Epoch [   13/   50] | d_loss: 0.5136 | g_loss: 0.2517\n",
      "Epoch [   13/   50] | d_loss: 0.5459 | g_loss: 0.2363\n",
      "Epoch [   13/   50] | d_loss: 0.4715 | g_loss: 0.2904\n",
      "Epoch [   13/   50] | d_loss: 0.5318 | g_loss: 0.2815\n",
      "Epoch [   13/   50] | d_loss: 0.5563 | g_loss: 0.2477\n",
      "Epoch [   13/   50] | d_loss: 0.4857 | g_loss: 0.2635\n",
      "Epoch [   13/   50] | d_loss: 0.4990 | g_loss: 0.2524\n",
      "Epoch [   13/   50] | d_loss: 0.5183 | g_loss: 0.2541\n",
      "Epoch [   13/   50] | d_loss: 0.5264 | g_loss: 0.2479\n",
      "Epoch [   13/   50] | d_loss: 0.5120 | g_loss: 0.2195\n",
      "Epoch [   13/   50] | d_loss: 0.5041 | g_loss: 0.2823\n",
      "Epoch [   13/   50] | d_loss: 0.5239 | g_loss: 0.2447\n",
      "Epoch [   13/   50] | d_loss: 0.4861 | g_loss: 0.2538\n",
      "Epoch [   13/   50] | d_loss: 0.4971 | g_loss: 0.2566\n",
      "Epoch [   13/   50] | d_loss: 0.5154 | g_loss: 0.2537\n",
      "Epoch [   13/   50] | d_loss: 0.4655 | g_loss: 0.2606\n",
      "Epoch [   13/   50] | d_loss: 0.4810 | g_loss: 0.2871\n",
      "Epoch [   13/   50] | d_loss: 0.4519 | g_loss: 0.2868\n",
      "Epoch [   13/   50] | d_loss: 0.4987 | g_loss: 0.2815\n",
      "Epoch [   13/   50] | d_loss: 0.4519 | g_loss: 0.2639\n",
      "Epoch [   13/   50] | d_loss: 0.4805 | g_loss: 0.2863\n",
      "Epoch [   13/   50] | d_loss: 0.4940 | g_loss: 0.2370\n",
      "Epoch [   13/   50] | d_loss: 0.5106 | g_loss: 0.2414\n",
      "Epoch [   13/   50] | d_loss: 0.4762 | g_loss: 0.2649\n",
      "Epoch [   13/   50] | d_loss: 0.4818 | g_loss: 0.2804\n",
      "Epoch [   13/   50] | d_loss: 0.5601 | g_loss: 0.2523\n",
      "Epoch [   13/   50] | d_loss: 0.5155 | g_loss: 0.2432\n",
      "Epoch [   13/   50] | d_loss: 0.4435 | g_loss: 0.2839\n",
      "Epoch [   13/   50] | d_loss: 0.4849 | g_loss: 0.2482\n",
      "Epoch [   13/   50] | d_loss: 0.5100 | g_loss: 0.2864\n",
      "Epoch [   13/   50] | d_loss: 0.4758 | g_loss: 0.2642\n",
      "Epoch [   13/   50] | d_loss: 0.5270 | g_loss: 0.2615\n",
      "Epoch [   13/   50] | d_loss: 0.4836 | g_loss: 0.2212\n",
      "Epoch [   13/   50] | d_loss: 0.5558 | g_loss: 0.2606\n",
      "Epoch [   13/   50] | d_loss: 0.4907 | g_loss: 0.2728\n",
      "Epoch [   13/   50] | d_loss: 0.4792 | g_loss: 0.2661\n",
      "Epoch [   13/   50] | d_loss: 0.5348 | g_loss: 0.2831\n",
      "Epoch [   13/   50] | d_loss: 0.4605 | g_loss: 0.2708\n",
      "Epoch [   13/   50] | d_loss: 0.4828 | g_loss: 0.2427\n",
      "Epoch [   13/   50] | d_loss: 0.5028 | g_loss: 0.2720\n",
      "Epoch [   13/   50] | d_loss: 0.4435 | g_loss: 0.2769\n",
      "Epoch [   13/   50] | d_loss: 0.5444 | g_loss: 0.2536\n",
      "Epoch [   13/   50] | d_loss: 0.4969 | g_loss: 0.2842\n",
      "Epoch [   13/   50] | d_loss: 0.4598 | g_loss: 0.2445\n",
      "Epoch [   13/   50] | d_loss: 0.5103 | g_loss: 0.2563\n",
      "Epoch [   13/   50] | d_loss: 0.5045 | g_loss: 0.2411\n",
      "Epoch [   13/   50] | d_loss: 0.4604 | g_loss: 0.2490\n",
      "Epoch [   13/   50] | d_loss: 0.5044 | g_loss: 0.2518\n",
      "Epoch [   13/   50] | d_loss: 0.5043 | g_loss: 0.2426\n",
      "Epoch [   13/   50] | d_loss: 0.5013 | g_loss: 0.2555\n",
      "Epoch [   13/   50] | d_loss: 0.5233 | g_loss: 0.2559\n",
      "Epoch [   13/   50] | d_loss: 0.4979 | g_loss: 0.2482\n",
      "Epoch [   13/   50] | d_loss: 0.5171 | g_loss: 0.2663\n",
      "Epoch [   13/   50] | d_loss: 0.5274 | g_loss: 0.2927\n",
      "Epoch [   13/   50] | d_loss: 0.5356 | g_loss: 0.2499\n",
      "Epoch [   13/   50] | d_loss: 0.4772 | g_loss: 0.2454\n",
      "Epoch [   13/   50] | d_loss: 0.5184 | g_loss: 0.2472\n",
      "Epoch [   13/   50] | d_loss: 0.4897 | g_loss: 0.2688\n",
      "Epoch [   13/   50] | d_loss: 0.5627 | g_loss: 0.2772\n",
      "Epoch [   13/   50] | d_loss: 0.5096 | g_loss: 0.2540\n",
      "Epoch [   13/   50] | d_loss: 0.5398 | g_loss: 0.2933\n",
      "Epoch [   13/   50] | d_loss: 0.4351 | g_loss: 0.3161\n",
      "Epoch [   13/   50] | d_loss: 0.4870 | g_loss: 0.2903\n",
      "Epoch [   13/   50] | d_loss: 0.5038 | g_loss: 0.2676\n",
      "Epoch [   13/   50] | d_loss: 0.5064 | g_loss: 0.2679\n",
      "Epoch [   13/   50] | d_loss: 0.5023 | g_loss: 0.2515\n",
      "Epoch [   13/   50] | d_loss: 0.4651 | g_loss: 0.2569\n",
      "Epoch [   13/   50] | d_loss: 0.5292 | g_loss: 0.2578\n",
      "Epoch [   13/   50] | d_loss: 0.5077 | g_loss: 0.2423\n",
      "Epoch [   13/   50] | d_loss: 0.4777 | g_loss: 0.2679\n",
      "Epoch [   13/   50] | d_loss: 0.5091 | g_loss: 0.2639\n",
      "Epoch [   13/   50] | d_loss: 0.4910 | g_loss: 0.2630\n",
      "Epoch [   13/   50] | d_loss: 0.5142 | g_loss: 0.2683\n",
      "Epoch [   13/   50] | d_loss: 0.4868 | g_loss: 0.2552\n",
      "Epoch [   13/   50] | d_loss: 0.5229 | g_loss: 0.2618\n",
      "Epoch [   13/   50] | d_loss: 0.5086 | g_loss: 0.2525\n",
      "Epoch [   13/   50] | d_loss: 0.5198 | g_loss: 0.2547\n",
      "Epoch [   13/   50] | d_loss: 0.5049 | g_loss: 0.2652\n",
      "Epoch [   13/   50] | d_loss: 0.5115 | g_loss: 0.2860\n",
      "Epoch [   13/   50] | d_loss: 0.4931 | g_loss: 0.2672\n",
      "Epoch [   13/   50] | d_loss: 0.4964 | g_loss: 0.2416\n",
      "Epoch [   13/   50] | d_loss: 0.5314 | g_loss: 0.2573\n",
      "Epoch [   13/   50] | d_loss: 0.4816 | g_loss: 0.2748\n",
      "Epoch [   13/   50] | d_loss: 0.4721 | g_loss: 0.2631\n",
      "Epoch [   13/   50] | d_loss: 0.5538 | g_loss: 0.2510\n",
      "Epoch [   13/   50] | d_loss: 0.5107 | g_loss: 0.2818\n",
      "Epoch [   13/   50] | d_loss: 0.4778 | g_loss: 0.2980\n",
      "Epoch [   13/   50] | d_loss: 0.4570 | g_loss: 0.2511\n",
      "Epoch [   13/   50] | d_loss: 0.5089 | g_loss: 0.2462\n",
      "Epoch [   14/   50] | d_loss: 0.4595 | g_loss: 0.2623\n",
      "Epoch [   14/   50] | d_loss: 0.5227 | g_loss: 0.2851\n",
      "Epoch [   14/   50] | d_loss: 0.5230 | g_loss: 0.2585\n",
      "Epoch [   14/   50] | d_loss: 0.4758 | g_loss: 0.2853\n",
      "Epoch [   14/   50] | d_loss: 0.5041 | g_loss: 0.2780\n",
      "Epoch [   14/   50] | d_loss: 0.5381 | g_loss: 0.2413\n",
      "Epoch [   14/   50] | d_loss: 0.5838 | g_loss: 0.2433\n",
      "Epoch [   14/   50] | d_loss: 0.4611 | g_loss: 0.2770\n",
      "Epoch [   14/   50] | d_loss: 0.4960 | g_loss: 0.2389\n",
      "Epoch [   14/   50] | d_loss: 0.4825 | g_loss: 0.2554\n",
      "Epoch [   14/   50] | d_loss: 0.5039 | g_loss: 0.2654\n",
      "Epoch [   14/   50] | d_loss: 0.5015 | g_loss: 0.2403\n",
      "Epoch [   14/   50] | d_loss: 0.5157 | g_loss: 0.2530\n",
      "Epoch [   14/   50] | d_loss: 0.4951 | g_loss: 0.2503\n",
      "Epoch [   14/   50] | d_loss: 0.4865 | g_loss: 0.2773\n",
      "Epoch [   14/   50] | d_loss: 0.5094 | g_loss: 0.2574\n",
      "Epoch [   14/   50] | d_loss: 0.4862 | g_loss: 0.2420\n",
      "Epoch [   14/   50] | d_loss: 0.5181 | g_loss: 0.2713\n",
      "Epoch [   14/   50] | d_loss: 0.5029 | g_loss: 0.2509\n",
      "Epoch [   14/   50] | d_loss: 0.5101 | g_loss: 0.2619\n",
      "Epoch [   14/   50] | d_loss: 0.5026 | g_loss: 0.2409\n",
      "Epoch [   14/   50] | d_loss: 0.5241 | g_loss: 0.2699\n",
      "Epoch [   14/   50] | d_loss: 0.4883 | g_loss: 0.2443\n",
      "Epoch [   14/   50] | d_loss: 0.4891 | g_loss: 0.2717\n",
      "Epoch [   14/   50] | d_loss: 0.4831 | g_loss: 0.2564\n",
      "Epoch [   14/   50] | d_loss: 0.4865 | g_loss: 0.2570\n",
      "Epoch [   14/   50] | d_loss: 0.4682 | g_loss: 0.2959\n",
      "Epoch [   14/   50] | d_loss: 0.4782 | g_loss: 0.3019\n",
      "Epoch [   14/   50] | d_loss: 0.4568 | g_loss: 0.2841\n",
      "Epoch [   14/   50] | d_loss: 0.5136 | g_loss: 0.2342\n",
      "Epoch [   14/   50] | d_loss: 0.4963 | g_loss: 0.2573\n",
      "Epoch [   14/   50] | d_loss: 0.4811 | g_loss: 0.2411\n",
      "Epoch [   14/   50] | d_loss: 0.4579 | g_loss: 0.2522\n",
      "Epoch [   14/   50] | d_loss: 0.5048 | g_loss: 0.2226\n",
      "Epoch [   14/   50] | d_loss: 0.4673 | g_loss: 0.2862\n",
      "Epoch [   14/   50] | d_loss: 0.5482 | g_loss: 0.2404\n",
      "Epoch [   14/   50] | d_loss: 0.5295 | g_loss: 0.2795\n",
      "Epoch [   14/   50] | d_loss: 0.5174 | g_loss: 0.2741\n",
      "Epoch [   14/   50] | d_loss: 0.4819 | g_loss: 0.2524\n",
      "Epoch [   14/   50] | d_loss: 0.4878 | g_loss: 0.2557\n",
      "Epoch [   14/   50] | d_loss: 0.5275 | g_loss: 0.2293\n",
      "Epoch [   14/   50] | d_loss: 0.4685 | g_loss: 0.2851\n",
      "Epoch [   14/   50] | d_loss: 0.4785 | g_loss: 0.2554\n",
      "Epoch [   14/   50] | d_loss: 0.5353 | g_loss: 0.2384\n",
      "Epoch [   14/   50] | d_loss: 0.4767 | g_loss: 0.2850\n",
      "Epoch [   14/   50] | d_loss: 0.4777 | g_loss: 0.2577\n",
      "Epoch [   14/   50] | d_loss: 0.5682 | g_loss: 0.2748\n",
      "Epoch [   14/   50] | d_loss: 0.5182 | g_loss: 0.2725\n",
      "Epoch [   14/   50] | d_loss: 0.4686 | g_loss: 0.2816\n",
      "Epoch [   14/   50] | d_loss: 0.5277 | g_loss: 0.2801\n",
      "Epoch [   14/   50] | d_loss: 0.5613 | g_loss: 0.2615\n",
      "Epoch [   14/   50] | d_loss: 0.5212 | g_loss: 0.2543\n",
      "Epoch [   14/   50] | d_loss: 0.5428 | g_loss: 0.2586\n",
      "Epoch [   14/   50] | d_loss: 0.4983 | g_loss: 0.2504\n",
      "Epoch [   14/   50] | d_loss: 0.4839 | g_loss: 0.2517\n",
      "Epoch [   14/   50] | d_loss: 0.5080 | g_loss: 0.2496\n",
      "Epoch [   14/   50] | d_loss: 0.5276 | g_loss: 0.2531\n",
      "Epoch [   14/   50] | d_loss: 0.4993 | g_loss: 0.2726\n",
      "Epoch [   14/   50] | d_loss: 0.5113 | g_loss: 0.2590\n",
      "Epoch [   14/   50] | d_loss: 0.5250 | g_loss: 0.3059\n",
      "Epoch [   14/   50] | d_loss: 0.5086 | g_loss: 0.2501\n",
      "Epoch [   14/   50] | d_loss: 0.5073 | g_loss: 0.2470\n",
      "Epoch [   14/   50] | d_loss: 0.4822 | g_loss: 0.2590\n",
      "Epoch [   14/   50] | d_loss: 0.5170 | g_loss: 0.2610\n",
      "Epoch [   14/   50] | d_loss: 0.5064 | g_loss: 0.2659\n",
      "Epoch [   14/   50] | d_loss: 0.4929 | g_loss: 0.2656\n",
      "Epoch [   14/   50] | d_loss: 0.4856 | g_loss: 0.2473\n",
      "Epoch [   14/   50] | d_loss: 0.5128 | g_loss: 0.2352\n",
      "Epoch [   14/   50] | d_loss: 0.5216 | g_loss: 0.2656\n",
      "Epoch [   14/   50] | d_loss: 0.4926 | g_loss: 0.2685\n",
      "Epoch [   14/   50] | d_loss: 0.5212 | g_loss: 0.2585\n",
      "Epoch [   14/   50] | d_loss: 0.5139 | g_loss: 0.2747\n",
      "Epoch [   14/   50] | d_loss: 0.4764 | g_loss: 0.2699\n",
      "Epoch [   14/   50] | d_loss: 0.5048 | g_loss: 0.2775\n",
      "Epoch [   14/   50] | d_loss: 0.4916 | g_loss: 0.2632\n",
      "Epoch [   14/   50] | d_loss: 0.4667 | g_loss: 0.2566\n",
      "Epoch [   14/   50] | d_loss: 0.5045 | g_loss: 0.2614\n",
      "Epoch [   14/   50] | d_loss: 0.5040 | g_loss: 0.2689\n",
      "Epoch [   14/   50] | d_loss: 0.5225 | g_loss: 0.2805\n",
      "Epoch [   14/   50] | d_loss: 0.5246 | g_loss: 0.2607\n",
      "Epoch [   14/   50] | d_loss: 0.4734 | g_loss: 0.2597\n",
      "Epoch [   14/   50] | d_loss: 0.5141 | g_loss: 0.2439\n",
      "Epoch [   14/   50] | d_loss: 0.4624 | g_loss: 0.2677\n",
      "Epoch [   14/   50] | d_loss: 0.5310 | g_loss: 0.2480\n",
      "Epoch [   14/   50] | d_loss: 0.5262 | g_loss: 0.2615\n",
      "Epoch [   14/   50] | d_loss: 0.5092 | g_loss: 0.2479\n",
      "Epoch [   14/   50] | d_loss: 0.4658 | g_loss: 0.2641\n",
      "Epoch [   14/   50] | d_loss: 0.5335 | g_loss: 0.2748\n",
      "Epoch [   14/   50] | d_loss: 0.5017 | g_loss: 0.2516\n",
      "Epoch [   14/   50] | d_loss: 0.4827 | g_loss: 0.2690\n",
      "Epoch [   15/   50] | d_loss: 0.5078 | g_loss: 0.2570\n",
      "Epoch [   15/   50] | d_loss: 0.4638 | g_loss: 0.2709\n",
      "Epoch [   15/   50] | d_loss: 0.5131 | g_loss: 0.2581\n",
      "Epoch [   15/   50] | d_loss: 0.4937 | g_loss: 0.2519\n",
      "Epoch [   15/   50] | d_loss: 0.5143 | g_loss: 0.2587\n",
      "Epoch [   15/   50] | d_loss: 0.4832 | g_loss: 0.2582\n",
      "Epoch [   15/   50] | d_loss: 0.5075 | g_loss: 0.2572\n",
      "Epoch [   15/   50] | d_loss: 0.5185 | g_loss: 0.2654\n",
      "Epoch [   15/   50] | d_loss: 0.5117 | g_loss: 0.2467\n",
      "Epoch [   15/   50] | d_loss: 0.4993 | g_loss: 0.2324\n",
      "Epoch [   15/   50] | d_loss: 0.4810 | g_loss: 0.2507\n",
      "Epoch [   15/   50] | d_loss: 0.5049 | g_loss: 0.2201\n",
      "Epoch [   15/   50] | d_loss: 0.5166 | g_loss: 0.3236\n",
      "Epoch [   15/   50] | d_loss: 0.4339 | g_loss: 0.3151\n",
      "Epoch [   15/   50] | d_loss: 0.5137 | g_loss: 0.2651\n",
      "Epoch [   15/   50] | d_loss: 0.5219 | g_loss: 0.2595\n",
      "Epoch [   15/   50] | d_loss: 0.4844 | g_loss: 0.2408\n",
      "Epoch [   15/   50] | d_loss: 0.4888 | g_loss: 0.2408\n",
      "Epoch [   15/   50] | d_loss: 0.5215 | g_loss: 0.2735\n",
      "Epoch [   15/   50] | d_loss: 0.5296 | g_loss: 0.2414\n",
      "Epoch [   15/   50] | d_loss: 0.4914 | g_loss: 0.2805\n",
      "Epoch [   15/   50] | d_loss: 0.4984 | g_loss: 0.2776\n",
      "Epoch [   15/   50] | d_loss: 0.4830 | g_loss: 0.2334\n",
      "Epoch [   15/   50] | d_loss: 0.5432 | g_loss: 0.2410\n",
      "Epoch [   15/   50] | d_loss: 0.5067 | g_loss: 0.2550\n",
      "Epoch [   15/   50] | d_loss: 0.4977 | g_loss: 0.2506\n",
      "Epoch [   15/   50] | d_loss: 0.5040 | g_loss: 0.2567\n",
      "Epoch [   15/   50] | d_loss: 0.5003 | g_loss: 0.2351\n",
      "Epoch [   15/   50] | d_loss: 0.4772 | g_loss: 0.2496\n",
      "Epoch [   15/   50] | d_loss: 0.4760 | g_loss: 0.2937\n",
      "Epoch [   15/   50] | d_loss: 0.5088 | g_loss: 0.2708\n",
      "Epoch [   15/   50] | d_loss: 0.4661 | g_loss: 0.2492\n",
      "Epoch [   15/   50] | d_loss: 0.4932 | g_loss: 0.2969\n",
      "Epoch [   15/   50] | d_loss: 0.5122 | g_loss: 0.2403\n",
      "Epoch [   15/   50] | d_loss: 0.5091 | g_loss: 0.2576\n",
      "Epoch [   15/   50] | d_loss: 0.4840 | g_loss: 0.2499\n",
      "Epoch [   15/   50] | d_loss: 0.4861 | g_loss: 0.2646\n",
      "Epoch [   15/   50] | d_loss: 0.5169 | g_loss: 0.2562\n",
      "Epoch [   15/   50] | d_loss: 0.4884 | g_loss: 0.2638\n",
      "Epoch [   15/   50] | d_loss: 0.4896 | g_loss: 0.2525\n",
      "Epoch [   15/   50] | d_loss: 0.4923 | g_loss: 0.2563\n",
      "Epoch [   15/   50] | d_loss: 0.5153 | g_loss: 0.2669\n",
      "Epoch [   15/   50] | d_loss: 0.4521 | g_loss: 0.2907\n",
      "Epoch [   15/   50] | d_loss: 0.5494 | g_loss: 0.2491\n",
      "Epoch [   15/   50] | d_loss: 0.4714 | g_loss: 0.2910\n",
      "Epoch [   15/   50] | d_loss: 0.4939 | g_loss: 0.2489\n",
      "Epoch [   15/   50] | d_loss: 0.4380 | g_loss: 0.2676\n",
      "Epoch [   15/   50] | d_loss: 0.4707 | g_loss: 0.2776\n",
      "Epoch [   15/   50] | d_loss: 0.4446 | g_loss: 0.2604\n",
      "Epoch [   15/   50] | d_loss: 0.5052 | g_loss: 0.2592\n",
      "Epoch [   15/   50] | d_loss: 0.4880 | g_loss: 0.2471\n",
      "Epoch [   15/   50] | d_loss: 0.5161 | g_loss: 0.2635\n",
      "Epoch [   15/   50] | d_loss: 0.4896 | g_loss: 0.2894\n",
      "Epoch [   15/   50] | d_loss: 0.5175 | g_loss: 0.2886\n",
      "Epoch [   15/   50] | d_loss: 0.5173 | g_loss: 0.2424\n",
      "Epoch [   15/   50] | d_loss: 0.4717 | g_loss: 0.2816\n",
      "Epoch [   15/   50] | d_loss: 0.4934 | g_loss: 0.2451\n",
      "Epoch [   15/   50] | d_loss: 0.5405 | g_loss: 0.2772\n",
      "Epoch [   15/   50] | d_loss: 0.5428 | g_loss: 0.2623\n",
      "Epoch [   15/   50] | d_loss: 0.4991 | g_loss: 0.2700\n",
      "Epoch [   15/   50] | d_loss: 0.5207 | g_loss: 0.2634\n",
      "Epoch [   15/   50] | d_loss: 0.4806 | g_loss: 0.2607\n",
      "Epoch [   15/   50] | d_loss: 0.5091 | g_loss: 0.2888\n",
      "Epoch [   15/   50] | d_loss: 0.4711 | g_loss: 0.2671\n",
      "Epoch [   15/   50] | d_loss: 0.5122 | g_loss: 0.2442\n",
      "Epoch [   15/   50] | d_loss: 0.5113 | g_loss: 0.2664\n",
      "Epoch [   15/   50] | d_loss: 0.5291 | g_loss: 0.2787\n",
      "Epoch [   15/   50] | d_loss: 0.4956 | g_loss: 0.2609\n",
      "Epoch [   15/   50] | d_loss: 0.5235 | g_loss: 0.2476\n",
      "Epoch [   15/   50] | d_loss: 0.4716 | g_loss: 0.2475\n",
      "Epoch [   15/   50] | d_loss: 0.5093 | g_loss: 0.2356\n",
      "Epoch [   15/   50] | d_loss: 0.4709 | g_loss: 0.2663\n",
      "Epoch [   15/   50] | d_loss: 0.5194 | g_loss: 0.2488\n",
      "Epoch [   15/   50] | d_loss: 0.4768 | g_loss: 0.2642\n",
      "Epoch [   15/   50] | d_loss: 0.4617 | g_loss: 0.2813\n",
      "Epoch [   15/   50] | d_loss: 0.4809 | g_loss: 0.2485\n",
      "Epoch [   15/   50] | d_loss: 0.5162 | g_loss: 0.2647\n",
      "Epoch [   15/   50] | d_loss: 0.4907 | g_loss: 0.2406\n",
      "Epoch [   15/   50] | d_loss: 0.5133 | g_loss: 0.2556\n",
      "Epoch [   15/   50] | d_loss: 0.5087 | g_loss: 0.2742\n",
      "Epoch [   15/   50] | d_loss: 0.5272 | g_loss: 0.2468\n",
      "Epoch [   15/   50] | d_loss: 0.4937 | g_loss: 0.2709\n",
      "Epoch [   15/   50] | d_loss: 0.4694 | g_loss: 0.2715\n",
      "Epoch [   15/   50] | d_loss: 0.4908 | g_loss: 0.2626\n",
      "Epoch [   15/   50] | d_loss: 0.5407 | g_loss: 0.2386\n",
      "Epoch [   15/   50] | d_loss: 0.5663 | g_loss: 0.2662\n",
      "Epoch [   15/   50] | d_loss: 0.4710 | g_loss: 0.2518\n",
      "Epoch [   15/   50] | d_loss: 0.5332 | g_loss: 0.2551\n",
      "Epoch [   15/   50] | d_loss: 0.5084 | g_loss: 0.2646\n",
      "Epoch [   15/   50] | d_loss: 0.5298 | g_loss: 0.2841\n",
      "Epoch [   16/   50] | d_loss: 0.4909 | g_loss: 0.2925\n",
      "Epoch [   16/   50] | d_loss: 0.5001 | g_loss: 0.2652\n",
      "Epoch [   16/   50] | d_loss: 0.5176 | g_loss: 0.2137\n",
      "Epoch [   16/   50] | d_loss: 0.5122 | g_loss: 0.2508\n",
      "Epoch [   16/   50] | d_loss: 0.5343 | g_loss: 0.2388\n",
      "Epoch [   16/   50] | d_loss: 0.4893 | g_loss: 0.2617\n",
      "Epoch [   16/   50] | d_loss: 0.5127 | g_loss: 0.2563\n",
      "Epoch [   16/   50] | d_loss: 0.4731 | g_loss: 0.2626\n",
      "Epoch [   16/   50] | d_loss: 0.5154 | g_loss: 0.2854\n",
      "Epoch [   16/   50] | d_loss: 0.4500 | g_loss: 0.2538\n",
      "Epoch [   16/   50] | d_loss: 0.4760 | g_loss: 0.2482\n",
      "Epoch [   16/   50] | d_loss: 0.4699 | g_loss: 0.2346\n",
      "Epoch [   16/   50] | d_loss: 0.4890 | g_loss: 0.2500\n",
      "Epoch [   16/   50] | d_loss: 0.4977 | g_loss: 0.2764\n",
      "Epoch [   16/   50] | d_loss: 0.5130 | g_loss: 0.2731\n",
      "Epoch [   16/   50] | d_loss: 0.4958 | g_loss: 0.2541\n",
      "Epoch [   16/   50] | d_loss: 0.4622 | g_loss: 0.2450\n",
      "Epoch [   16/   50] | d_loss: 0.4906 | g_loss: 0.2661\n",
      "Epoch [   16/   50] | d_loss: 0.5331 | g_loss: 0.2488\n",
      "Epoch [   16/   50] | d_loss: 0.5214 | g_loss: 0.2935\n",
      "Epoch [   16/   50] | d_loss: 0.5003 | g_loss: 0.2934\n",
      "Epoch [   16/   50] | d_loss: 0.4628 | g_loss: 0.2716\n",
      "Epoch [   16/   50] | d_loss: 0.4985 | g_loss: 0.2604\n",
      "Epoch [   16/   50] | d_loss: 0.4946 | g_loss: 0.2820\n",
      "Epoch [   16/   50] | d_loss: 0.4544 | g_loss: 0.2963\n",
      "Epoch [   16/   50] | d_loss: 0.5148 | g_loss: 0.2682\n",
      "Epoch [   16/   50] | d_loss: 0.5297 | g_loss: 0.2642\n",
      "Epoch [   16/   50] | d_loss: 0.4831 | g_loss: 0.2506\n",
      "Epoch [   16/   50] | d_loss: 0.5036 | g_loss: 0.2705\n",
      "Epoch [   16/   50] | d_loss: 0.5156 | g_loss: 0.2822\n",
      "Epoch [   16/   50] | d_loss: 0.5306 | g_loss: 0.2675\n",
      "Epoch [   16/   50] | d_loss: 0.4640 | g_loss: 0.2622\n",
      "Epoch [   16/   50] | d_loss: 0.4571 | g_loss: 0.2880\n",
      "Epoch [   16/   50] | d_loss: 0.5222 | g_loss: 0.2464\n",
      "Epoch [   16/   50] | d_loss: 0.4939 | g_loss: 0.2773\n",
      "Epoch [   16/   50] | d_loss: 0.4737 | g_loss: 0.2425\n",
      "Epoch [   16/   50] | d_loss: 0.4667 | g_loss: 0.2919\n",
      "Epoch [   16/   50] | d_loss: 0.4724 | g_loss: 0.2844\n",
      "Epoch [   16/   50] | d_loss: 0.5018 | g_loss: 0.2429\n",
      "Epoch [   16/   50] | d_loss: 0.5152 | g_loss: 0.2610\n",
      "Epoch [   16/   50] | d_loss: 0.4841 | g_loss: 0.2970\n",
      "Epoch [   16/   50] | d_loss: 0.4974 | g_loss: 0.2461\n",
      "Epoch [   16/   50] | d_loss: 0.4798 | g_loss: 0.2690\n",
      "Epoch [   16/   50] | d_loss: 0.5140 | g_loss: 0.2762\n",
      "Epoch [   16/   50] | d_loss: 0.5005 | g_loss: 0.2797\n",
      "Epoch [   16/   50] | d_loss: 0.5174 | g_loss: 0.2275\n",
      "Epoch [   16/   50] | d_loss: 0.4905 | g_loss: 0.2503\n",
      "Epoch [   16/   50] | d_loss: 0.4957 | g_loss: 0.2749\n",
      "Epoch [   16/   50] | d_loss: 0.4809 | g_loss: 0.2591\n",
      "Epoch [   16/   50] | d_loss: 0.4424 | g_loss: 0.2810\n",
      "Epoch [   16/   50] | d_loss: 0.5195 | g_loss: 0.2530\n",
      "Epoch [   16/   50] | d_loss: 0.4987 | g_loss: 0.2726\n",
      "Epoch [   16/   50] | d_loss: 0.4414 | g_loss: 0.2926\n",
      "Epoch [   16/   50] | d_loss: 0.5040 | g_loss: 0.2713\n",
      "Epoch [   16/   50] | d_loss: 0.4981 | g_loss: 0.2522\n",
      "Epoch [   16/   50] | d_loss: 0.4690 | g_loss: 0.2531\n",
      "Epoch [   16/   50] | d_loss: 0.4821 | g_loss: 0.2812\n",
      "Epoch [   16/   50] | d_loss: 0.4846 | g_loss: 0.2605\n",
      "Epoch [   16/   50] | d_loss: 0.5069 | g_loss: 0.2652\n",
      "Epoch [   16/   50] | d_loss: 0.4574 | g_loss: 0.2846\n",
      "Epoch [   16/   50] | d_loss: 0.5046 | g_loss: 0.2627\n",
      "Epoch [   16/   50] | d_loss: 0.4889 | g_loss: 0.3032\n",
      "Epoch [   16/   50] | d_loss: 0.4959 | g_loss: 0.2565\n",
      "Epoch [   16/   50] | d_loss: 0.5174 | g_loss: 0.2716\n",
      "Epoch [   16/   50] | d_loss: 0.4779 | g_loss: 0.2589\n",
      "Epoch [   16/   50] | d_loss: 0.5335 | g_loss: 0.2694\n",
      "Epoch [   16/   50] | d_loss: 0.4572 | g_loss: 0.2586\n",
      "Epoch [   16/   50] | d_loss: 0.5078 | g_loss: 0.2538\n",
      "Epoch [   16/   50] | d_loss: 0.4879 | g_loss: 0.2520\n",
      "Epoch [   16/   50] | d_loss: 0.5303 | g_loss: 0.2877\n",
      "Epoch [   16/   50] | d_loss: 0.5156 | g_loss: 0.2646\n",
      "Epoch [   16/   50] | d_loss: 0.5269 | g_loss: 0.2384\n",
      "Epoch [   16/   50] | d_loss: 0.5110 | g_loss: 0.2801\n",
      "Epoch [   16/   50] | d_loss: 0.5066 | g_loss: 0.2405\n",
      "Epoch [   16/   50] | d_loss: 0.4873 | g_loss: 0.2642\n",
      "Epoch [   16/   50] | d_loss: 0.5169 | g_loss: 0.2342\n",
      "Epoch [   16/   50] | d_loss: 0.4588 | g_loss: 0.2682\n",
      "Epoch [   16/   50] | d_loss: 0.5097 | g_loss: 0.2670\n",
      "Epoch [   16/   50] | d_loss: 0.4649 | g_loss: 0.2295\n",
      "Epoch [   16/   50] | d_loss: 0.5170 | g_loss: 0.2624\n",
      "Epoch [   16/   50] | d_loss: 0.5223 | g_loss: 0.2764\n",
      "Epoch [   16/   50] | d_loss: 0.5178 | g_loss: 0.2520\n",
      "Epoch [   16/   50] | d_loss: 0.4930 | g_loss: 0.2773\n",
      "Epoch [   16/   50] | d_loss: 0.5126 | g_loss: 0.2871\n",
      "Epoch [   16/   50] | d_loss: 0.4909 | g_loss: 0.2800\n",
      "Epoch [   16/   50] | d_loss: 0.4795 | g_loss: 0.2866\n",
      "Epoch [   16/   50] | d_loss: 0.5111 | g_loss: 0.2550\n",
      "Epoch [   16/   50] | d_loss: 0.4580 | g_loss: 0.2881\n",
      "Epoch [   16/   50] | d_loss: 0.4840 | g_loss: 0.2849\n",
      "Epoch [   16/   50] | d_loss: 0.5187 | g_loss: 0.2280\n",
      "Epoch [   17/   50] | d_loss: 0.4680 | g_loss: 0.2398\n",
      "Epoch [   17/   50] | d_loss: 0.4945 | g_loss: 0.2292\n",
      "Epoch [   17/   50] | d_loss: 0.5305 | g_loss: 0.2327\n",
      "Epoch [   17/   50] | d_loss: 0.5253 | g_loss: 0.2519\n",
      "Epoch [   17/   50] | d_loss: 0.4944 | g_loss: 0.2733\n",
      "Epoch [   17/   50] | d_loss: 0.5324 | g_loss: 0.2779\n",
      "Epoch [   17/   50] | d_loss: 0.4671 | g_loss: 0.2912\n",
      "Epoch [   17/   50] | d_loss: 0.4969 | g_loss: 0.2378\n",
      "Epoch [   17/   50] | d_loss: 0.5036 | g_loss: 0.2932\n",
      "Epoch [   17/   50] | d_loss: 0.5173 | g_loss: 0.2735\n",
      "Epoch [   17/   50] | d_loss: 0.4734 | g_loss: 0.2625\n",
      "Epoch [   17/   50] | d_loss: 0.4929 | g_loss: 0.2814\n",
      "Epoch [   17/   50] | d_loss: 0.4582 | g_loss: 0.2448\n",
      "Epoch [   17/   50] | d_loss: 0.4952 | g_loss: 0.2211\n",
      "Epoch [   17/   50] | d_loss: 0.4916 | g_loss: 0.2520\n",
      "Epoch [   17/   50] | d_loss: 0.4853 | g_loss: 0.2671\n",
      "Epoch [   17/   50] | d_loss: 0.5097 | g_loss: 0.2497\n",
      "Epoch [   17/   50] | d_loss: 0.4690 | g_loss: 0.2351\n",
      "Epoch [   17/   50] | d_loss: 0.4828 | g_loss: 0.2548\n",
      "Epoch [   17/   50] | d_loss: 0.5105 | g_loss: 0.2149\n",
      "Epoch [   17/   50] | d_loss: 0.5347 | g_loss: 0.2165\n",
      "Epoch [   17/   50] | d_loss: 0.5219 | g_loss: 0.2444\n",
      "Epoch [   17/   50] | d_loss: 0.4847 | g_loss: 0.2676\n",
      "Epoch [   17/   50] | d_loss: 0.5239 | g_loss: 0.2660\n",
      "Epoch [   17/   50] | d_loss: 0.5366 | g_loss: 0.2462\n",
      "Epoch [   17/   50] | d_loss: 0.5168 | g_loss: 0.2448\n",
      "Epoch [   17/   50] | d_loss: 0.5258 | g_loss: 0.2366\n",
      "Epoch [   17/   50] | d_loss: 0.5152 | g_loss: 0.2443\n",
      "Epoch [   17/   50] | d_loss: 0.5147 | g_loss: 0.2733\n",
      "Epoch [   17/   50] | d_loss: 0.4927 | g_loss: 0.2411\n",
      "Epoch [   17/   50] | d_loss: 0.5008 | g_loss: 0.2597\n",
      "Epoch [   17/   50] | d_loss: 0.4976 | g_loss: 0.2970\n",
      "Epoch [   17/   50] | d_loss: 0.5204 | g_loss: 0.2400\n",
      "Epoch [   17/   50] | d_loss: 0.4744 | g_loss: 0.2318\n",
      "Epoch [   17/   50] | d_loss: 0.5196 | g_loss: 0.2353\n",
      "Epoch [   17/   50] | d_loss: 0.4921 | g_loss: 0.2666\n",
      "Epoch [   17/   50] | d_loss: 0.4938 | g_loss: 0.2480\n",
      "Epoch [   17/   50] | d_loss: 0.4713 | g_loss: 0.2585\n",
      "Epoch [   17/   50] | d_loss: 0.4729 | g_loss: 0.2688\n",
      "Epoch [   17/   50] | d_loss: 0.4457 | g_loss: 0.2867\n",
      "Epoch [   17/   50] | d_loss: 0.4840 | g_loss: 0.2264\n",
      "Epoch [   17/   50] | d_loss: 0.5061 | g_loss: 0.2552\n",
      "Epoch [   17/   50] | d_loss: 0.4861 | g_loss: 0.2752\n",
      "Epoch [   17/   50] | d_loss: 0.5138 | g_loss: 0.2391\n",
      "Epoch [   17/   50] | d_loss: 0.4930 | g_loss: 0.3008\n",
      "Epoch [   17/   50] | d_loss: 0.4633 | g_loss: 0.2351\n",
      "Epoch [   17/   50] | d_loss: 0.5075 | g_loss: 0.2736\n",
      "Epoch [   17/   50] | d_loss: 0.4957 | g_loss: 0.2467\n",
      "Epoch [   17/   50] | d_loss: 0.4711 | g_loss: 0.2573\n",
      "Epoch [   17/   50] | d_loss: 0.4794 | g_loss: 0.2541\n",
      "Epoch [   17/   50] | d_loss: 0.4797 | g_loss: 0.2749\n",
      "Epoch [   17/   50] | d_loss: 0.5422 | g_loss: 0.2461\n",
      "Epoch [   17/   50] | d_loss: 0.4943 | g_loss: 0.2435\n",
      "Epoch [   17/   50] | d_loss: 0.5362 | g_loss: 0.2442\n",
      "Epoch [   17/   50] | d_loss: 0.5020 | g_loss: 0.2678\n",
      "Epoch [   17/   50] | d_loss: 0.4607 | g_loss: 0.2248\n",
      "Epoch [   17/   50] | d_loss: 0.4975 | g_loss: 0.2467\n",
      "Epoch [   17/   50] | d_loss: 0.4862 | g_loss: 0.2556\n",
      "Epoch [   17/   50] | d_loss: 0.4980 | g_loss: 0.2667\n",
      "Epoch [   17/   50] | d_loss: 0.5269 | g_loss: 0.2635\n",
      "Epoch [   17/   50] | d_loss: 0.4978 | g_loss: 0.2478\n",
      "Epoch [   17/   50] | d_loss: 0.4912 | g_loss: 0.2542\n",
      "Epoch [   17/   50] | d_loss: 0.4986 | g_loss: 0.2574\n",
      "Epoch [   17/   50] | d_loss: 0.4840 | g_loss: 0.2668\n",
      "Epoch [   17/   50] | d_loss: 0.5169 | g_loss: 0.2385\n",
      "Epoch [   17/   50] | d_loss: 0.4829 | g_loss: 0.2896\n",
      "Epoch [   17/   50] | d_loss: 0.4946 | g_loss: 0.2872\n",
      "Epoch [   17/   50] | d_loss: 0.4745 | g_loss: 0.2530\n",
      "Epoch [   17/   50] | d_loss: 0.5463 | g_loss: 0.2689\n",
      "Epoch [   17/   50] | d_loss: 0.4697 | g_loss: 0.3137\n",
      "Epoch [   17/   50] | d_loss: 0.4999 | g_loss: 0.2803\n",
      "Epoch [   17/   50] | d_loss: 0.4527 | g_loss: 0.3016\n",
      "Epoch [   17/   50] | d_loss: 0.5277 | g_loss: 0.2653\n",
      "Epoch [   17/   50] | d_loss: 0.4490 | g_loss: 0.2913\n",
      "Epoch [   17/   50] | d_loss: 0.5169 | g_loss: 0.2725\n",
      "Epoch [   17/   50] | d_loss: 0.4733 | g_loss: 0.2483\n",
      "Epoch [   17/   50] | d_loss: 0.4890 | g_loss: 0.2576\n",
      "Epoch [   17/   50] | d_loss: 0.4797 | g_loss: 0.2667\n",
      "Epoch [   17/   50] | d_loss: 0.4826 | g_loss: 0.2489\n",
      "Epoch [   17/   50] | d_loss: 0.4845 | g_loss: 0.2806\n",
      "Epoch [   17/   50] | d_loss: 0.4711 | g_loss: 0.3031\n",
      "Epoch [   17/   50] | d_loss: 0.4721 | g_loss: 0.2582\n",
      "Epoch [   17/   50] | d_loss: 0.4998 | g_loss: 0.2946\n",
      "Epoch [   17/   50] | d_loss: 0.4785 | g_loss: 0.2993\n",
      "Epoch [   17/   50] | d_loss: 0.4871 | g_loss: 0.2484\n",
      "Epoch [   17/   50] | d_loss: 0.5105 | g_loss: 0.2785\n",
      "Epoch [   17/   50] | d_loss: 0.4815 | g_loss: 0.2636\n",
      "Epoch [   17/   50] | d_loss: 0.5060 | g_loss: 0.2480\n",
      "Epoch [   17/   50] | d_loss: 0.5075 | g_loss: 0.2661\n",
      "Epoch [   17/   50] | d_loss: 0.5064 | g_loss: 0.2587\n",
      "Epoch [   18/   50] | d_loss: 0.4616 | g_loss: 0.3032\n",
      "Epoch [   18/   50] | d_loss: 0.4966 | g_loss: 0.2631\n",
      "Epoch [   18/   50] | d_loss: 0.5300 | g_loss: 0.3010\n",
      "Epoch [   18/   50] | d_loss: 0.4929 | g_loss: 0.2694\n",
      "Epoch [   18/   50] | d_loss: 0.5190 | g_loss: 0.2698\n",
      "Epoch [   18/   50] | d_loss: 0.5107 | g_loss: 0.2565\n",
      "Epoch [   18/   50] | d_loss: 0.5209 | g_loss: 0.2611\n",
      "Epoch [   18/   50] | d_loss: 0.5106 | g_loss: 0.2781\n",
      "Epoch [   18/   50] | d_loss: 0.4922 | g_loss: 0.2537\n",
      "Epoch [   18/   50] | d_loss: 0.4909 | g_loss: 0.2541\n",
      "Epoch [   18/   50] | d_loss: 0.4956 | g_loss: 0.2582\n",
      "Epoch [   18/   50] | d_loss: 0.5046 | g_loss: 0.2688\n",
      "Epoch [   18/   50] | d_loss: 0.4970 | g_loss: 0.2719\n",
      "Epoch [   18/   50] | d_loss: 0.4831 | g_loss: 0.2466\n",
      "Epoch [   18/   50] | d_loss: 0.4889 | g_loss: 0.2433\n",
      "Epoch [   18/   50] | d_loss: 0.4915 | g_loss: 0.2692\n",
      "Epoch [   18/   50] | d_loss: 0.5011 | g_loss: 0.2875\n",
      "Epoch [   18/   50] | d_loss: 0.4849 | g_loss: 0.3219\n",
      "Epoch [   18/   50] | d_loss: 0.4923 | g_loss: 0.2446\n",
      "Epoch [   18/   50] | d_loss: 0.4979 | g_loss: 0.2511\n",
      "Epoch [   18/   50] | d_loss: 0.5499 | g_loss: 0.2547\n",
      "Epoch [   18/   50] | d_loss: 0.4835 | g_loss: 0.2460\n",
      "Epoch [   18/   50] | d_loss: 0.4735 | g_loss: 0.2489\n",
      "Epoch [   18/   50] | d_loss: 0.5041 | g_loss: 0.2834\n",
      "Epoch [   18/   50] | d_loss: 0.5026 | g_loss: 0.2701\n",
      "Epoch [   18/   50] | d_loss: 0.4939 | g_loss: 0.2472\n",
      "Epoch [   18/   50] | d_loss: 0.5269 | g_loss: 0.2761\n",
      "Epoch [   18/   50] | d_loss: 0.5373 | g_loss: 0.2538\n",
      "Epoch [   18/   50] | d_loss: 0.5161 | g_loss: 0.2473\n",
      "Epoch [   18/   50] | d_loss: 0.4932 | g_loss: 0.2400\n",
      "Epoch [   18/   50] | d_loss: 0.5005 | g_loss: 0.2748\n",
      "Epoch [   18/   50] | d_loss: 0.4540 | g_loss: 0.2686\n",
      "Epoch [   18/   50] | d_loss: 0.5167 | g_loss: 0.2760\n",
      "Epoch [   18/   50] | d_loss: 0.4464 | g_loss: 0.2647\n",
      "Epoch [   18/   50] | d_loss: 0.4799 | g_loss: 0.2461\n",
      "Epoch [   18/   50] | d_loss: 0.5607 | g_loss: 0.2667\n",
      "Epoch [   18/   50] | d_loss: 0.4922 | g_loss: 0.2741\n",
      "Epoch [   18/   50] | d_loss: 0.5351 | g_loss: 0.2521\n",
      "Epoch [   18/   50] | d_loss: 0.5153 | g_loss: 0.3150\n",
      "Epoch [   18/   50] | d_loss: 0.4856 | g_loss: 0.2526\n",
      "Epoch [   18/   50] | d_loss: 0.4826 | g_loss: 0.2869\n",
      "Epoch [   18/   50] | d_loss: 0.5047 | g_loss: 0.2565\n",
      "Epoch [   18/   50] | d_loss: 0.4794 | g_loss: 0.2412\n",
      "Epoch [   18/   50] | d_loss: 0.5036 | g_loss: 0.2337\n",
      "Epoch [   18/   50] | d_loss: 0.4852 | g_loss: 0.2686\n",
      "Epoch [   18/   50] | d_loss: 0.5178 | g_loss: 0.2613\n",
      "Epoch [   18/   50] | d_loss: 0.5335 | g_loss: 0.2546\n",
      "Epoch [   18/   50] | d_loss: 0.4952 | g_loss: 0.2560\n",
      "Epoch [   18/   50] | d_loss: 0.4907 | g_loss: 0.2420\n",
      "Epoch [   18/   50] | d_loss: 0.5268 | g_loss: 0.2434\n",
      "Epoch [   18/   50] | d_loss: 0.5047 | g_loss: 0.2596\n",
      "Epoch [   18/   50] | d_loss: 0.4828 | g_loss: 0.2768\n",
      "Epoch [   18/   50] | d_loss: 0.4952 | g_loss: 0.2463\n",
      "Epoch [   18/   50] | d_loss: 0.4852 | g_loss: 0.2473\n",
      "Epoch [   18/   50] | d_loss: 0.4859 | g_loss: 0.2473\n",
      "Epoch [   18/   50] | d_loss: 0.5351 | g_loss: 0.2726\n",
      "Epoch [   18/   50] | d_loss: 0.4597 | g_loss: 0.2423\n",
      "Epoch [   18/   50] | d_loss: 0.4842 | g_loss: 0.2515\n",
      "Epoch [   18/   50] | d_loss: 0.5044 | g_loss: 0.2453\n",
      "Epoch [   18/   50] | d_loss: 0.5131 | g_loss: 0.2562\n",
      "Epoch [   18/   50] | d_loss: 0.4867 | g_loss: 0.2410\n",
      "Epoch [   18/   50] | d_loss: 0.5092 | g_loss: 0.2603\n",
      "Epoch [   18/   50] | d_loss: 0.5051 | g_loss: 0.2910\n",
      "Epoch [   18/   50] | d_loss: 0.5105 | g_loss: 0.2663\n",
      "Epoch [   18/   50] | d_loss: 0.4985 | g_loss: 0.2859\n",
      "Epoch [   18/   50] | d_loss: 0.4958 | g_loss: 0.2496\n",
      "Epoch [   18/   50] | d_loss: 0.5378 | g_loss: 0.2448\n",
      "Epoch [   18/   50] | d_loss: 0.5253 | g_loss: 0.2490\n",
      "Epoch [   18/   50] | d_loss: 0.5306 | g_loss: 0.2647\n",
      "Epoch [   18/   50] | d_loss: 0.5057 | g_loss: 0.2865\n",
      "Epoch [   18/   50] | d_loss: 0.4884 | g_loss: 0.2701\n",
      "Epoch [   18/   50] | d_loss: 0.5005 | g_loss: 0.2815\n",
      "Epoch [   18/   50] | d_loss: 0.4883 | g_loss: 0.2676\n",
      "Epoch [   18/   50] | d_loss: 0.4757 | g_loss: 0.2726\n",
      "Epoch [   18/   50] | d_loss: 0.4962 | g_loss: 0.2445\n",
      "Epoch [   18/   50] | d_loss: 0.5241 | g_loss: 0.2763\n",
      "Epoch [   18/   50] | d_loss: 0.5283 | g_loss: 0.2648\n",
      "Epoch [   18/   50] | d_loss: 0.5112 | g_loss: 0.2632\n",
      "Epoch [   18/   50] | d_loss: 0.5000 | g_loss: 0.2801\n",
      "Epoch [   18/   50] | d_loss: 0.4879 | g_loss: 0.2591\n",
      "Epoch [   18/   50] | d_loss: 0.5230 | g_loss: 0.2495\n",
      "Epoch [   18/   50] | d_loss: 0.4898 | g_loss: 0.2418\n",
      "Epoch [   18/   50] | d_loss: 0.5386 | g_loss: 0.2948\n",
      "Epoch [   18/   50] | d_loss: 0.5218 | g_loss: 0.2685\n",
      "Epoch [   18/   50] | d_loss: 0.5519 | g_loss: 0.2664\n",
      "Epoch [   18/   50] | d_loss: 0.5075 | g_loss: 0.2698\n",
      "Epoch [   18/   50] | d_loss: 0.4974 | g_loss: 0.2652\n",
      "Epoch [   18/   50] | d_loss: 0.4706 | g_loss: 0.2622\n",
      "Epoch [   18/   50] | d_loss: 0.5366 | g_loss: 0.2642\n",
      "Epoch [   18/   50] | d_loss: 0.4729 | g_loss: 0.2615\n",
      "Epoch [   19/   50] | d_loss: 0.5422 | g_loss: 0.2756\n",
      "Epoch [   19/   50] | d_loss: 0.5020 | g_loss: 0.2577\n",
      "Epoch [   19/   50] | d_loss: 0.4636 | g_loss: 0.2661\n",
      "Epoch [   19/   50] | d_loss: 0.5280 | g_loss: 0.2717\n",
      "Epoch [   19/   50] | d_loss: 0.5010 | g_loss: 0.2314\n",
      "Epoch [   19/   50] | d_loss: 0.4927 | g_loss: 0.2591\n",
      "Epoch [   19/   50] | d_loss: 0.5098 | g_loss: 0.2248\n",
      "Epoch [   19/   50] | d_loss: 0.5364 | g_loss: 0.2608\n",
      "Epoch [   19/   50] | d_loss: 0.4671 | g_loss: 0.2832\n",
      "Epoch [   19/   50] | d_loss: 0.4850 | g_loss: 0.2583\n",
      "Epoch [   19/   50] | d_loss: 0.5491 | g_loss: 0.2223\n",
      "Epoch [   19/   50] | d_loss: 0.5331 | g_loss: 0.2541\n",
      "Epoch [   19/   50] | d_loss: 0.4972 | g_loss: 0.2294\n",
      "Epoch [   19/   50] | d_loss: 0.4583 | g_loss: 0.2519\n",
      "Epoch [   19/   50] | d_loss: 0.4712 | g_loss: 0.2916\n",
      "Epoch [   19/   50] | d_loss: 0.4761 | g_loss: 0.3036\n",
      "Epoch [   19/   50] | d_loss: 0.4822 | g_loss: 0.2868\n",
      "Epoch [   19/   50] | d_loss: 0.5424 | g_loss: 0.2482\n",
      "Epoch [   19/   50] | d_loss: 0.5081 | g_loss: 0.2704\n",
      "Epoch [   19/   50] | d_loss: 0.5144 | g_loss: 0.2833\n",
      "Epoch [   19/   50] | d_loss: 0.5267 | g_loss: 0.2436\n",
      "Epoch [   19/   50] | d_loss: 0.5219 | g_loss: 0.2484\n",
      "Epoch [   19/   50] | d_loss: 0.4939 | g_loss: 0.2616\n",
      "Epoch [   19/   50] | d_loss: 0.5374 | g_loss: 0.2384\n",
      "Epoch [   19/   50] | d_loss: 0.4781 | g_loss: 0.2694\n",
      "Epoch [   19/   50] | d_loss: 0.4938 | g_loss: 0.2520\n",
      "Epoch [   19/   50] | d_loss: 0.5007 | g_loss: 0.2512\n",
      "Epoch [   19/   50] | d_loss: 0.5252 | g_loss: 0.2680\n",
      "Epoch [   19/   50] | d_loss: 0.5179 | g_loss: 0.3057\n",
      "Epoch [   19/   50] | d_loss: 0.4856 | g_loss: 0.2364\n",
      "Epoch [   19/   50] | d_loss: 0.5305 | g_loss: 0.3005\n",
      "Epoch [   19/   50] | d_loss: 0.4789 | g_loss: 0.2937\n",
      "Epoch [   19/   50] | d_loss: 0.5069 | g_loss: 0.2404\n",
      "Epoch [   19/   50] | d_loss: 0.4799 | g_loss: 0.2866\n",
      "Epoch [   19/   50] | d_loss: 0.4887 | g_loss: 0.2375\n",
      "Epoch [   19/   50] | d_loss: 0.5388 | g_loss: 0.2592\n",
      "Epoch [   19/   50] | d_loss: 0.5060 | g_loss: 0.2842\n",
      "Epoch [   19/   50] | d_loss: 0.4791 | g_loss: 0.2825\n",
      "Epoch [   19/   50] | d_loss: 0.4799 | g_loss: 0.2479\n",
      "Epoch [   19/   50] | d_loss: 0.4795 | g_loss: 0.2657\n",
      "Epoch [   19/   50] | d_loss: 0.5229 | g_loss: 0.2822\n",
      "Epoch [   19/   50] | d_loss: 0.5162 | g_loss: 0.2731\n",
      "Epoch [   19/   50] | d_loss: 0.5438 | g_loss: 0.2619\n",
      "Epoch [   19/   50] | d_loss: 0.5052 | g_loss: 0.2535\n",
      "Epoch [   19/   50] | d_loss: 0.4656 | g_loss: 0.3009\n",
      "Epoch [   19/   50] | d_loss: 0.5222 | g_loss: 0.2620\n",
      "Epoch [   19/   50] | d_loss: 0.5175 | g_loss: 0.2781\n",
      "Epoch [   19/   50] | d_loss: 0.4739 | g_loss: 0.2496\n",
      "Epoch [   19/   50] | d_loss: 0.5465 | g_loss: 0.2379\n",
      "Epoch [   19/   50] | d_loss: 0.4933 | g_loss: 0.2545\n",
      "Epoch [   19/   50] | d_loss: 0.4957 | g_loss: 0.2241\n",
      "Epoch [   19/   50] | d_loss: 0.5127 | g_loss: 0.2608\n",
      "Epoch [   19/   50] | d_loss: 0.5224 | g_loss: 0.2290\n",
      "Epoch [   19/   50] | d_loss: 0.4775 | g_loss: 0.2626\n",
      "Epoch [   19/   50] | d_loss: 0.5102 | g_loss: 0.2952\n",
      "Epoch [   19/   50] | d_loss: 0.5172 | g_loss: 0.2693\n",
      "Epoch [   19/   50] | d_loss: 0.4736 | g_loss: 0.2753\n",
      "Epoch [   19/   50] | d_loss: 0.5399 | g_loss: 0.2673\n",
      "Epoch [   19/   50] | d_loss: 0.4864 | g_loss: 0.2897\n",
      "Epoch [   19/   50] | d_loss: 0.5182 | g_loss: 0.2776\n",
      "Epoch [   19/   50] | d_loss: 0.4817 | g_loss: 0.2362\n",
      "Epoch [   19/   50] | d_loss: 0.4700 | g_loss: 0.3092\n",
      "Epoch [   19/   50] | d_loss: 0.4647 | g_loss: 0.2532\n",
      "Epoch [   19/   50] | d_loss: 0.5169 | g_loss: 0.2677\n",
      "Epoch [   19/   50] | d_loss: 0.5023 | g_loss: 0.2570\n",
      "Epoch [   19/   50] | d_loss: 0.5129 | g_loss: 0.2468\n",
      "Epoch [   19/   50] | d_loss: 0.5284 | g_loss: 0.2318\n",
      "Epoch [   19/   50] | d_loss: 0.5267 | g_loss: 0.2858\n",
      "Epoch [   19/   50] | d_loss: 0.4881 | g_loss: 0.2411\n",
      "Epoch [   19/   50] | d_loss: 0.5487 | g_loss: 0.2498\n",
      "Epoch [   19/   50] | d_loss: 0.4845 | g_loss: 0.2692\n",
      "Epoch [   19/   50] | d_loss: 0.5088 | g_loss: 0.2642\n",
      "Epoch [   19/   50] | d_loss: 0.5035 | g_loss: 0.2796\n",
      "Epoch [   19/   50] | d_loss: 0.5389 | g_loss: 0.2519\n",
      "Epoch [   19/   50] | d_loss: 0.5015 | g_loss: 0.2543\n",
      "Epoch [   19/   50] | d_loss: 0.4799 | g_loss: 0.2258\n",
      "Epoch [   19/   50] | d_loss: 0.4845 | g_loss: 0.2149\n",
      "Epoch [   19/   50] | d_loss: 0.5276 | g_loss: 0.2540\n",
      "Epoch [   19/   50] | d_loss: 0.4921 | g_loss: 0.2875\n",
      "Epoch [   19/   50] | d_loss: 0.4902 | g_loss: 0.2718\n",
      "Epoch [   19/   50] | d_loss: 0.5283 | g_loss: 0.2535\n",
      "Epoch [   19/   50] | d_loss: 0.4937 | g_loss: 0.2506\n",
      "Epoch [   19/   50] | d_loss: 0.5203 | g_loss: 0.2640\n",
      "Epoch [   19/   50] | d_loss: 0.4995 | g_loss: 0.2702\n",
      "Epoch [   19/   50] | d_loss: 0.5217 | g_loss: 0.2722\n",
      "Epoch [   19/   50] | d_loss: 0.5161 | g_loss: 0.2659\n",
      "Epoch [   19/   50] | d_loss: 0.4881 | g_loss: 0.2900\n",
      "Epoch [   19/   50] | d_loss: 0.5232 | g_loss: 0.2411\n",
      "Epoch [   19/   50] | d_loss: 0.5065 | g_loss: 0.2307\n",
      "Epoch [   19/   50] | d_loss: 0.5299 | g_loss: 0.2424\n",
      "Epoch [   20/   50] | d_loss: 0.5365 | g_loss: 0.2789\n",
      "Epoch [   20/   50] | d_loss: 0.4788 | g_loss: 0.2459\n",
      "Epoch [   20/   50] | d_loss: 0.5142 | g_loss: 0.2214\n",
      "Epoch [   20/   50] | d_loss: 0.5278 | g_loss: 0.2533\n",
      "Epoch [   20/   50] | d_loss: 0.5145 | g_loss: 0.2636\n",
      "Epoch [   20/   50] | d_loss: 0.5016 | g_loss: 0.2576\n",
      "Epoch [   20/   50] | d_loss: 0.4779 | g_loss: 0.2984\n",
      "Epoch [   20/   50] | d_loss: 0.5263 | g_loss: 0.2286\n",
      "Epoch [   20/   50] | d_loss: 0.5239 | g_loss: 0.2661\n",
      "Epoch [   20/   50] | d_loss: 0.5227 | g_loss: 0.2828\n",
      "Epoch [   20/   50] | d_loss: 0.5145 | g_loss: 0.2445\n",
      "Epoch [   20/   50] | d_loss: 0.5371 | g_loss: 0.2972\n",
      "Epoch [   20/   50] | d_loss: 0.5127 | g_loss: 0.2285\n",
      "Epoch [   20/   50] | d_loss: 0.4954 | g_loss: 0.2742\n",
      "Epoch [   20/   50] | d_loss: 0.5158 | g_loss: 0.2638\n",
      "Epoch [   20/   50] | d_loss: 0.5058 | g_loss: 0.2515\n",
      "Epoch [   20/   50] | d_loss: 0.5244 | g_loss: 0.2398\n",
      "Epoch [   20/   50] | d_loss: 0.5437 | g_loss: 0.2495\n",
      "Epoch [   20/   50] | d_loss: 0.4722 | g_loss: 0.2989\n",
      "Epoch [   20/   50] | d_loss: 0.5204 | g_loss: 0.2387\n",
      "Epoch [   20/   50] | d_loss: 0.4870 | g_loss: 0.2407\n",
      "Epoch [   20/   50] | d_loss: 0.5388 | g_loss: 0.2478\n",
      "Epoch [   20/   50] | d_loss: 0.4985 | g_loss: 0.2699\n",
      "Epoch [   20/   50] | d_loss: 0.5098 | g_loss: 0.2487\n",
      "Epoch [   20/   50] | d_loss: 0.5156 | g_loss: 0.2536\n",
      "Epoch [   20/   50] | d_loss: 0.5203 | g_loss: 0.2297\n",
      "Epoch [   20/   50] | d_loss: 0.4867 | g_loss: 0.2509\n",
      "Epoch [   20/   50] | d_loss: 0.5054 | g_loss: 0.2400\n",
      "Epoch [   20/   50] | d_loss: 0.5140 | g_loss: 0.2806\n",
      "Epoch [   20/   50] | d_loss: 0.4685 | g_loss: 0.2542\n",
      "Epoch [   20/   50] | d_loss: 0.5353 | g_loss: 0.2712\n",
      "Epoch [   20/   50] | d_loss: 0.4972 | g_loss: 0.2465\n",
      "Epoch [   20/   50] | d_loss: 0.4981 | g_loss: 0.2636\n",
      "Epoch [   20/   50] | d_loss: 0.5256 | g_loss: 0.2680\n",
      "Epoch [   20/   50] | d_loss: 0.4911 | g_loss: 0.2926\n",
      "Epoch [   20/   50] | d_loss: 0.5322 | g_loss: 0.2813\n",
      "Epoch [   20/   50] | d_loss: 0.5200 | g_loss: 0.2556\n",
      "Epoch [   20/   50] | d_loss: 0.4707 | g_loss: 0.2427\n",
      "Epoch [   20/   50] | d_loss: 0.4831 | g_loss: 0.2818\n",
      "Epoch [   20/   50] | d_loss: 0.5285 | g_loss: 0.2698\n",
      "Epoch [   20/   50] | d_loss: 0.5057 | g_loss: 0.2753\n",
      "Epoch [   20/   50] | d_loss: 0.5134 | g_loss: 0.2548\n",
      "Epoch [   20/   50] | d_loss: 0.4868 | g_loss: 0.2425\n",
      "Epoch [   20/   50] | d_loss: 0.5250 | g_loss: 0.2535\n",
      "Epoch [   20/   50] | d_loss: 0.4877 | g_loss: 0.2372\n",
      "Epoch [   20/   50] | d_loss: 0.5166 | g_loss: 0.2609\n",
      "Epoch [   20/   50] | d_loss: 0.5270 | g_loss: 0.2626\n",
      "Epoch [   20/   50] | d_loss: 0.4960 | g_loss: 0.2426\n",
      "Epoch [   20/   50] | d_loss: 0.4938 | g_loss: 0.2415\n",
      "Epoch [   20/   50] | d_loss: 0.4993 | g_loss: 0.2517\n",
      "Epoch [   20/   50] | d_loss: 0.5085 | g_loss: 0.2716\n",
      "Epoch [   20/   50] | d_loss: 0.5167 | g_loss: 0.2730\n",
      "Epoch [   20/   50] | d_loss: 0.4812 | g_loss: 0.2623\n",
      "Epoch [   20/   50] | d_loss: 0.5098 | g_loss: 0.2241\n",
      "Epoch [   20/   50] | d_loss: 0.5081 | g_loss: 0.2731\n",
      "Epoch [   20/   50] | d_loss: 0.5025 | g_loss: 0.2769\n",
      "Epoch [   20/   50] | d_loss: 0.5298 | g_loss: 0.2691\n",
      "Epoch [   20/   50] | d_loss: 0.4963 | g_loss: 0.2654\n",
      "Epoch [   20/   50] | d_loss: 0.4982 | g_loss: 0.2473\n",
      "Epoch [   20/   50] | d_loss: 0.4861 | g_loss: 0.2553\n",
      "Epoch [   20/   50] | d_loss: 0.5087 | g_loss: 0.2519\n",
      "Epoch [   20/   50] | d_loss: 0.5076 | g_loss: 0.2522\n",
      "Epoch [   20/   50] | d_loss: 0.5210 | g_loss: 0.2777\n",
      "Epoch [   20/   50] | d_loss: 0.5060 | g_loss: 0.2517\n",
      "Epoch [   20/   50] | d_loss: 0.4824 | g_loss: 0.2577\n",
      "Epoch [   20/   50] | d_loss: 0.5118 | g_loss: 0.2665\n",
      "Epoch [   20/   50] | d_loss: 0.5127 | g_loss: 0.2562\n",
      "Epoch [   20/   50] | d_loss: 0.5142 | g_loss: 0.2823\n",
      "Epoch [   20/   50] | d_loss: 0.4940 | g_loss: 0.2431\n",
      "Epoch [   20/   50] | d_loss: 0.4958 | g_loss: 0.2736\n",
      "Epoch [   20/   50] | d_loss: 0.4586 | g_loss: 0.2734\n",
      "Epoch [   20/   50] | d_loss: 0.5197 | g_loss: 0.2854\n",
      "Epoch [   20/   50] | d_loss: 0.4375 | g_loss: 0.2657\n",
      "Epoch [   20/   50] | d_loss: 0.5140 | g_loss: 0.2591\n",
      "Epoch [   20/   50] | d_loss: 0.4937 | g_loss: 0.2677\n",
      "Epoch [   20/   50] | d_loss: 0.5108 | g_loss: 0.2629\n",
      "Epoch [   20/   50] | d_loss: 0.4702 | g_loss: 0.2583\n",
      "Epoch [   20/   50] | d_loss: 0.5182 | g_loss: 0.2492\n",
      "Epoch [   20/   50] | d_loss: 0.5518 | g_loss: 0.2254\n",
      "Epoch [   20/   50] | d_loss: 0.5034 | g_loss: 0.3039\n",
      "Epoch [   20/   50] | d_loss: 0.4942 | g_loss: 0.2678\n",
      "Epoch [   20/   50] | d_loss: 0.4710 | g_loss: 0.2437\n",
      "Epoch [   20/   50] | d_loss: 0.5064 | g_loss: 0.2293\n",
      "Epoch [   20/   50] | d_loss: 0.4941 | g_loss: 0.2687\n",
      "Epoch [   20/   50] | d_loss: 0.5264 | g_loss: 0.2359\n",
      "Epoch [   20/   50] | d_loss: 0.5308 | g_loss: 0.2362\n",
      "Epoch [   20/   50] | d_loss: 0.5291 | g_loss: 0.2790\n",
      "Epoch [   20/   50] | d_loss: 0.4666 | g_loss: 0.2733\n",
      "Epoch [   20/   50] | d_loss: 0.5258 | g_loss: 0.2374\n",
      "Epoch [   20/   50] | d_loss: 0.5191 | g_loss: 0.2427\n",
      "Epoch [   21/   50] | d_loss: 0.5424 | g_loss: 0.2532\n",
      "Epoch [   21/   50] | d_loss: 0.5253 | g_loss: 0.2678\n",
      "Epoch [   21/   50] | d_loss: 0.4869 | g_loss: 0.2505\n",
      "Epoch [   21/   50] | d_loss: 0.4736 | g_loss: 0.2832\n",
      "Epoch [   21/   50] | d_loss: 0.4717 | g_loss: 0.2630\n",
      "Epoch [   21/   50] | d_loss: 0.4971 | g_loss: 0.2819\n",
      "Epoch [   21/   50] | d_loss: 0.5078 | g_loss: 0.2450\n",
      "Epoch [   21/   50] | d_loss: 0.4812 | g_loss: 0.2465\n",
      "Epoch [   21/   50] | d_loss: 0.5562 | g_loss: 0.2339\n",
      "Epoch [   21/   50] | d_loss: 0.5213 | g_loss: 0.2716\n",
      "Epoch [   21/   50] | d_loss: 0.5029 | g_loss: 0.2569\n",
      "Epoch [   21/   50] | d_loss: 0.5163 | g_loss: 0.2591\n",
      "Epoch [   21/   50] | d_loss: 0.5095 | g_loss: 0.2690\n",
      "Epoch [   21/   50] | d_loss: 0.4909 | g_loss: 0.2439\n",
      "Epoch [   21/   50] | d_loss: 0.4850 | g_loss: 0.2886\n",
      "Epoch [   21/   50] | d_loss: 0.5019 | g_loss: 0.2423\n",
      "Epoch [   21/   50] | d_loss: 0.4835 | g_loss: 0.2703\n",
      "Epoch [   21/   50] | d_loss: 0.4681 | g_loss: 0.2747\n",
      "Epoch [   21/   50] | d_loss: 0.5026 | g_loss: 0.2578\n",
      "Epoch [   21/   50] | d_loss: 0.4738 | g_loss: 0.2612\n",
      "Epoch [   21/   50] | d_loss: 0.5092 | g_loss: 0.2672\n",
      "Epoch [   21/   50] | d_loss: 0.5086 | g_loss: 0.2765\n",
      "Epoch [   21/   50] | d_loss: 0.4939 | g_loss: 0.2547\n",
      "Epoch [   21/   50] | d_loss: 0.4781 | g_loss: 0.2813\n",
      "Epoch [   21/   50] | d_loss: 0.4999 | g_loss: 0.2748\n",
      "Epoch [   21/   50] | d_loss: 0.5353 | g_loss: 0.2894\n",
      "Epoch [   21/   50] | d_loss: 0.4896 | g_loss: 0.2377\n",
      "Epoch [   21/   50] | d_loss: 0.4531 | g_loss: 0.2560\n",
      "Epoch [   21/   50] | d_loss: 0.5073 | g_loss: 0.2457\n",
      "Epoch [   21/   50] | d_loss: 0.4881 | g_loss: 0.2861\n",
      "Epoch [   21/   50] | d_loss: 0.5414 | g_loss: 0.2419\n",
      "Epoch [   21/   50] | d_loss: 0.5115 | g_loss: 0.2689\n",
      "Epoch [   21/   50] | d_loss: 0.5016 | g_loss: 0.2773\n",
      "Epoch [   21/   50] | d_loss: 0.4982 | g_loss: 0.2553\n",
      "Epoch [   21/   50] | d_loss: 0.5062 | g_loss: 0.2720\n",
      "Epoch [   21/   50] | d_loss: 0.5123 | g_loss: 0.2302\n",
      "Epoch [   21/   50] | d_loss: 0.5155 | g_loss: 0.2541\n",
      "Epoch [   21/   50] | d_loss: 0.5122 | g_loss: 0.2396\n",
      "Epoch [   21/   50] | d_loss: 0.4791 | g_loss: 0.2584\n",
      "Epoch [   21/   50] | d_loss: 0.5215 | g_loss: 0.2470\n",
      "Epoch [   21/   50] | d_loss: 0.5371 | g_loss: 0.2404\n",
      "Epoch [   21/   50] | d_loss: 0.5045 | g_loss: 0.2704\n",
      "Epoch [   21/   50] | d_loss: 0.5453 | g_loss: 0.2632\n",
      "Epoch [   21/   50] | d_loss: 0.5079 | g_loss: 0.2675\n",
      "Epoch [   21/   50] | d_loss: 0.5068 | g_loss: 0.2448\n",
      "Epoch [   21/   50] | d_loss: 0.5162 | g_loss: 0.2980\n",
      "Epoch [   21/   50] | d_loss: 0.4719 | g_loss: 0.2805\n",
      "Epoch [   21/   50] | d_loss: 0.5500 | g_loss: 0.2683\n",
      "Epoch [   21/   50] | d_loss: 0.4834 | g_loss: 0.2764\n",
      "Epoch [   21/   50] | d_loss: 0.5311 | g_loss: 0.2493\n",
      "Epoch [   21/   50] | d_loss: 0.4894 | g_loss: 0.2616\n",
      "Epoch [   21/   50] | d_loss: 0.5125 | g_loss: 0.2464\n",
      "Epoch [   21/   50] | d_loss: 0.4947 | g_loss: 0.2565\n",
      "Epoch [   21/   50] | d_loss: 0.5030 | g_loss: 0.2755\n",
      "Epoch [   21/   50] | d_loss: 0.5016 | g_loss: 0.2585\n",
      "Epoch [   21/   50] | d_loss: 0.5004 | g_loss: 0.2472\n",
      "Epoch [   21/   50] | d_loss: 0.5047 | g_loss: 0.2691\n",
      "Epoch [   21/   50] | d_loss: 0.5279 | g_loss: 0.2821\n",
      "Epoch [   21/   50] | d_loss: 0.5330 | g_loss: 0.2525\n",
      "Epoch [   21/   50] | d_loss: 0.5117 | g_loss: 0.2627\n",
      "Epoch [   21/   50] | d_loss: 0.5034 | g_loss: 0.2345\n",
      "Epoch [   21/   50] | d_loss: 0.5039 | g_loss: 0.2480\n",
      "Epoch [   21/   50] | d_loss: 0.5082 | g_loss: 0.2641\n",
      "Epoch [   21/   50] | d_loss: 0.5116 | g_loss: 0.2355\n",
      "Epoch [   21/   50] | d_loss: 0.5121 | g_loss: 0.2452\n",
      "Epoch [   21/   50] | d_loss: 0.4828 | g_loss: 0.2698\n",
      "Epoch [   21/   50] | d_loss: 0.5133 | g_loss: 0.2773\n",
      "Epoch [   21/   50] | d_loss: 0.4721 | g_loss: 0.2541\n",
      "Epoch [   21/   50] | d_loss: 0.5107 | g_loss: 0.2503\n",
      "Epoch [   21/   50] | d_loss: 0.5061 | g_loss: 0.2774\n",
      "Epoch [   21/   50] | d_loss: 0.5053 | g_loss: 0.2757\n",
      "Epoch [   21/   50] | d_loss: 0.4707 | g_loss: 0.2432\n",
      "Epoch [   21/   50] | d_loss: 0.5019 | g_loss: 0.2469\n",
      "Epoch [   21/   50] | d_loss: 0.4759 | g_loss: 0.2783\n",
      "Epoch [   21/   50] | d_loss: 0.4992 | g_loss: 0.2570\n",
      "Epoch [   21/   50] | d_loss: 0.4873 | g_loss: 0.2356\n",
      "Epoch [   21/   50] | d_loss: 0.5245 | g_loss: 0.2533\n",
      "Epoch [   21/   50] | d_loss: 0.5068 | g_loss: 0.2708\n",
      "Epoch [   21/   50] | d_loss: 0.5248 | g_loss: 0.2784\n",
      "Epoch [   21/   50] | d_loss: 0.5108 | g_loss: 0.2759\n",
      "Epoch [   21/   50] | d_loss: 0.4857 | g_loss: 0.2362\n",
      "Epoch [   21/   50] | d_loss: 0.4959 | g_loss: 0.2527\n",
      "Epoch [   21/   50] | d_loss: 0.4998 | g_loss: 0.2591\n",
      "Epoch [   21/   50] | d_loss: 0.4994 | g_loss: 0.2829\n",
      "Epoch [   21/   50] | d_loss: 0.5081 | g_loss: 0.2497\n",
      "Epoch [   21/   50] | d_loss: 0.4908 | g_loss: 0.2727\n",
      "Epoch [   21/   50] | d_loss: 0.5278 | g_loss: 0.2755\n",
      "Epoch [   21/   50] | d_loss: 0.5105 | g_loss: 0.2579\n",
      "Epoch [   21/   50] | d_loss: 0.4699 | g_loss: 0.2848\n",
      "Epoch [   21/   50] | d_loss: 0.4665 | g_loss: 0.2447\n",
      "Epoch [   22/   50] | d_loss: 0.5073 | g_loss: 0.2774\n",
      "Epoch [   22/   50] | d_loss: 0.4966 | g_loss: 0.2423\n",
      "Epoch [   22/   50] | d_loss: 0.4885 | g_loss: 0.2562\n",
      "Epoch [   22/   50] | d_loss: 0.4691 | g_loss: 0.2342\n",
      "Epoch [   22/   50] | d_loss: 0.5429 | g_loss: 0.2485\n",
      "Epoch [   22/   50] | d_loss: 0.5270 | g_loss: 0.2768\n",
      "Epoch [   22/   50] | d_loss: 0.5101 | g_loss: 0.2399\n",
      "Epoch [   22/   50] | d_loss: 0.4875 | g_loss: 0.2602\n",
      "Epoch [   22/   50] | d_loss: 0.4740 | g_loss: 0.3001\n",
      "Epoch [   22/   50] | d_loss: 0.4951 | g_loss: 0.2749\n",
      "Epoch [   22/   50] | d_loss: 0.4966 | g_loss: 0.2714\n",
      "Epoch [   22/   50] | d_loss: 0.5090 | g_loss: 0.2900\n",
      "Epoch [   22/   50] | d_loss: 0.4591 | g_loss: 0.2589\n",
      "Epoch [   22/   50] | d_loss: 0.4955 | g_loss: 0.2404\n",
      "Epoch [   22/   50] | d_loss: 0.5098 | g_loss: 0.2791\n",
      "Epoch [   22/   50] | d_loss: 0.4955 | g_loss: 0.2725\n",
      "Epoch [   22/   50] | d_loss: 0.4657 | g_loss: 0.2522\n",
      "Epoch [   22/   50] | d_loss: 0.5000 | g_loss: 0.2573\n",
      "Epoch [   22/   50] | d_loss: 0.4843 | g_loss: 0.2495\n",
      "Epoch [   22/   50] | d_loss: 0.5156 | g_loss: 0.2525\n",
      "Epoch [   22/   50] | d_loss: 0.5332 | g_loss: 0.2625\n",
      "Epoch [   22/   50] | d_loss: 0.4991 | g_loss: 0.2827\n",
      "Epoch [   22/   50] | d_loss: 0.4758 | g_loss: 0.2457\n",
      "Epoch [   22/   50] | d_loss: 0.5146 | g_loss: 0.2752\n",
      "Epoch [   22/   50] | d_loss: 0.4815 | g_loss: 0.2422\n",
      "Epoch [   22/   50] | d_loss: 0.4843 | g_loss: 0.2614\n",
      "Epoch [   22/   50] | d_loss: 0.4988 | g_loss: 0.2508\n",
      "Epoch [   22/   50] | d_loss: 0.5181 | g_loss: 0.2587\n",
      "Epoch [   22/   50] | d_loss: 0.4974 | g_loss: 0.2723\n",
      "Epoch [   22/   50] | d_loss: 0.5071 | g_loss: 0.2583\n",
      "Epoch [   22/   50] | d_loss: 0.5036 | g_loss: 0.2725\n",
      "Epoch [   22/   50] | d_loss: 0.5017 | g_loss: 0.2510\n",
      "Epoch [   22/   50] | d_loss: 0.5125 | g_loss: 0.2551\n",
      "Epoch [   22/   50] | d_loss: 0.4666 | g_loss: 0.2769\n",
      "Epoch [   22/   50] | d_loss: 0.5094 | g_loss: 0.2640\n",
      "Epoch [   22/   50] | d_loss: 0.4573 | g_loss: 0.2464\n",
      "Epoch [   22/   50] | d_loss: 0.4807 | g_loss: 0.2742\n",
      "Epoch [   22/   50] | d_loss: 0.5386 | g_loss: 0.2768\n",
      "Epoch [   22/   50] | d_loss: 0.5013 | g_loss: 0.2442\n",
      "Epoch [   22/   50] | d_loss: 0.4902 | g_loss: 0.2593\n",
      "Epoch [   22/   50] | d_loss: 0.4627 | g_loss: 0.2543\n",
      "Epoch [   22/   50] | d_loss: 0.5159 | g_loss: 0.2327\n",
      "Epoch [   22/   50] | d_loss: 0.4738 | g_loss: 0.2578\n",
      "Epoch [   22/   50] | d_loss: 0.5142 | g_loss: 0.2554\n",
      "Epoch [   22/   50] | d_loss: 0.4786 | g_loss: 0.2686\n",
      "Epoch [   22/   50] | d_loss: 0.4707 | g_loss: 0.2546\n",
      "Epoch [   22/   50] | d_loss: 0.5218 | g_loss: 0.2790\n",
      "Epoch [   22/   50] | d_loss: 0.4865 | g_loss: 0.2471\n",
      "Epoch [   22/   50] | d_loss: 0.5547 | g_loss: 0.2556\n",
      "Epoch [   22/   50] | d_loss: 0.5109 | g_loss: 0.2577\n",
      "Epoch [   22/   50] | d_loss: 0.4902 | g_loss: 0.2597\n",
      "Epoch [   22/   50] | d_loss: 0.4743 | g_loss: 0.2522\n",
      "Epoch [   22/   50] | d_loss: 0.5103 | g_loss: 0.2686\n",
      "Epoch [   22/   50] | d_loss: 0.5137 | g_loss: 0.2766\n",
      "Epoch [   22/   50] | d_loss: 0.4856 | g_loss: 0.2618\n",
      "Epoch [   22/   50] | d_loss: 0.4810 | g_loss: 0.2420\n",
      "Epoch [   22/   50] | d_loss: 0.4969 | g_loss: 0.2792\n",
      "Epoch [   22/   50] | d_loss: 0.5551 | g_loss: 0.2583\n",
      "Epoch [   22/   50] | d_loss: 0.5076 | g_loss: 0.2686\n",
      "Epoch [   22/   50] | d_loss: 0.4959 | g_loss: 0.2481\n",
      "Epoch [   22/   50] | d_loss: 0.4932 | g_loss: 0.2394\n",
      "Epoch [   22/   50] | d_loss: 0.4981 | g_loss: 0.2887\n",
      "Epoch [   22/   50] | d_loss: 0.4788 | g_loss: 0.2763\n",
      "Epoch [   22/   50] | d_loss: 0.4624 | g_loss: 0.2483\n",
      "Epoch [   22/   50] | d_loss: 0.4748 | g_loss: 0.2589\n",
      "Epoch [   22/   50] | d_loss: 0.5503 | g_loss: 0.2854\n",
      "Epoch [   22/   50] | d_loss: 0.4878 | g_loss: 0.2987\n",
      "Epoch [   22/   50] | d_loss: 0.5069 | g_loss: 0.2859\n",
      "Epoch [   22/   50] | d_loss: 0.4641 | g_loss: 0.2608\n",
      "Epoch [   22/   50] | d_loss: 0.5409 | g_loss: 0.2571\n",
      "Epoch [   22/   50] | d_loss: 0.4676 | g_loss: 0.2669\n",
      "Epoch [   22/   50] | d_loss: 0.5179 | g_loss: 0.2662\n",
      "Epoch [   22/   50] | d_loss: 0.4541 | g_loss: 0.2453\n",
      "Epoch [   22/   50] | d_loss: 0.5237 | g_loss: 0.2896\n",
      "Epoch [   22/   50] | d_loss: 0.4935 | g_loss: 0.2555\n",
      "Epoch [   22/   50] | d_loss: 0.4917 | g_loss: 0.2865\n",
      "Epoch [   22/   50] | d_loss: 0.5389 | g_loss: 0.2526\n",
      "Epoch [   22/   50] | d_loss: 0.4706 | g_loss: 0.2737\n",
      "Epoch [   22/   50] | d_loss: 0.4831 | g_loss: 0.2609\n",
      "Epoch [   22/   50] | d_loss: 0.4956 | g_loss: 0.2685\n",
      "Epoch [   22/   50] | d_loss: 0.4922 | g_loss: 0.2909\n",
      "Epoch [   22/   50] | d_loss: 0.5618 | g_loss: 0.2649\n",
      "Epoch [   22/   50] | d_loss: 0.4834 | g_loss: 0.2636\n",
      "Epoch [   22/   50] | d_loss: 0.4909 | g_loss: 0.2967\n",
      "Epoch [   22/   50] | d_loss: 0.5284 | g_loss: 0.2785\n",
      "Epoch [   22/   50] | d_loss: 0.5235 | g_loss: 0.2703\n",
      "Epoch [   22/   50] | d_loss: 0.5192 | g_loss: 0.2640\n",
      "Epoch [   22/   50] | d_loss: 0.4660 | g_loss: 0.2494\n",
      "Epoch [   22/   50] | d_loss: 0.4859 | g_loss: 0.2846\n",
      "Epoch [   22/   50] | d_loss: 0.4862 | g_loss: 0.2715\n",
      "Epoch [   23/   50] | d_loss: 0.4790 | g_loss: 0.2467\n",
      "Epoch [   23/   50] | d_loss: 0.5227 | g_loss: 0.2528\n",
      "Epoch [   23/   50] | d_loss: 0.5220 | g_loss: 0.3044\n",
      "Epoch [   23/   50] | d_loss: 0.5504 | g_loss: 0.2484\n",
      "Epoch [   23/   50] | d_loss: 0.5135 | g_loss: 0.2880\n",
      "Epoch [   23/   50] | d_loss: 0.5004 | g_loss: 0.2554\n",
      "Epoch [   23/   50] | d_loss: 0.4855 | g_loss: 0.2275\n",
      "Epoch [   23/   50] | d_loss: 0.4780 | g_loss: 0.2444\n",
      "Epoch [   23/   50] | d_loss: 0.5113 | g_loss: 0.2656\n",
      "Epoch [   23/   50] | d_loss: 0.5473 | g_loss: 0.2440\n",
      "Epoch [   23/   50] | d_loss: 0.5395 | g_loss: 0.2649\n",
      "Epoch [   23/   50] | d_loss: 0.4560 | g_loss: 0.2536\n",
      "Epoch [   23/   50] | d_loss: 0.5019 | g_loss: 0.2346\n",
      "Epoch [   23/   50] | d_loss: 0.5079 | g_loss: 0.2557\n",
      "Epoch [   23/   50] | d_loss: 0.5152 | g_loss: 0.2827\n",
      "Epoch [   23/   50] | d_loss: 0.4550 | g_loss: 0.2398\n",
      "Epoch [   23/   50] | d_loss: 0.5055 | g_loss: 0.2583\n",
      "Epoch [   23/   50] | d_loss: 0.4849 | g_loss: 0.2616\n",
      "Epoch [   23/   50] | d_loss: 0.4828 | g_loss: 0.2440\n",
      "Epoch [   23/   50] | d_loss: 0.5089 | g_loss: 0.2523\n",
      "Epoch [   23/   50] | d_loss: 0.5042 | g_loss: 0.2521\n",
      "Epoch [   23/   50] | d_loss: 0.4817 | g_loss: 0.2626\n",
      "Epoch [   23/   50] | d_loss: 0.5390 | g_loss: 0.2396\n",
      "Epoch [   23/   50] | d_loss: 0.5034 | g_loss: 0.2786\n",
      "Epoch [   23/   50] | d_loss: 0.4994 | g_loss: 0.2236\n",
      "Epoch [   23/   50] | d_loss: 0.5121 | g_loss: 0.2632\n",
      "Epoch [   23/   50] | d_loss: 0.5222 | g_loss: 0.2932\n",
      "Epoch [   23/   50] | d_loss: 0.4781 | g_loss: 0.2505\n",
      "Epoch [   23/   50] | d_loss: 0.4827 | g_loss: 0.2900\n",
      "Epoch [   23/   50] | d_loss: 0.5227 | g_loss: 0.2693\n",
      "Epoch [   23/   50] | d_loss: 0.5009 | g_loss: 0.2901\n",
      "Epoch [   23/   50] | d_loss: 0.4871 | g_loss: 0.2600\n",
      "Epoch [   23/   50] | d_loss: 0.5239 | g_loss: 0.2847\n",
      "Epoch [   23/   50] | d_loss: 0.5012 | g_loss: 0.2766\n",
      "Epoch [   23/   50] | d_loss: 0.5196 | g_loss: 0.2432\n",
      "Epoch [   23/   50] | d_loss: 0.5074 | g_loss: 0.2920\n",
      "Epoch [   23/   50] | d_loss: 0.5412 | g_loss: 0.2709\n",
      "Epoch [   23/   50] | d_loss: 0.4837 | g_loss: 0.2146\n",
      "Epoch [   23/   50] | d_loss: 0.5234 | g_loss: 0.2837\n",
      "Epoch [   23/   50] | d_loss: 0.5360 | g_loss: 0.2661\n",
      "Epoch [   23/   50] | d_loss: 0.4897 | g_loss: 0.2754\n",
      "Epoch [   23/   50] | d_loss: 0.5098 | g_loss: 0.2693\n",
      "Epoch [   23/   50] | d_loss: 0.5154 | g_loss: 0.2501\n",
      "Epoch [   23/   50] | d_loss: 0.4901 | g_loss: 0.2496\n",
      "Epoch [   23/   50] | d_loss: 0.4945 | g_loss: 0.2486\n",
      "Epoch [   23/   50] | d_loss: 0.4666 | g_loss: 0.2684\n",
      "Epoch [   23/   50] | d_loss: 0.5602 | g_loss: 0.2776\n",
      "Epoch [   23/   50] | d_loss: 0.4964 | g_loss: 0.2535\n",
      "Epoch [   23/   50] | d_loss: 0.5076 | g_loss: 0.2457\n",
      "Epoch [   23/   50] | d_loss: 0.4791 | g_loss: 0.2395\n",
      "Epoch [   23/   50] | d_loss: 0.4996 | g_loss: 0.2231\n",
      "Epoch [   23/   50] | d_loss: 0.5119 | g_loss: 0.2532\n",
      "Epoch [   23/   50] | d_loss: 0.5029 | g_loss: 0.2959\n",
      "Epoch [   23/   50] | d_loss: 0.5064 | g_loss: 0.2770\n",
      "Epoch [   23/   50] | d_loss: 0.4684 | g_loss: 0.2708\n",
      "Epoch [   23/   50] | d_loss: 0.5510 | g_loss: 0.2509\n",
      "Epoch [   23/   50] | d_loss: 0.4811 | g_loss: 0.2733\n",
      "Epoch [   23/   50] | d_loss: 0.5378 | g_loss: 0.2620\n",
      "Epoch [   23/   50] | d_loss: 0.5294 | g_loss: 0.2715\n",
      "Epoch [   23/   50] | d_loss: 0.4941 | g_loss: 0.2500\n",
      "Epoch [   23/   50] | d_loss: 0.4791 | g_loss: 0.2444\n",
      "Epoch [   23/   50] | d_loss: 0.4595 | g_loss: 0.2504\n",
      "Epoch [   23/   50] | d_loss: 0.5017 | g_loss: 0.2371\n",
      "Epoch [   23/   50] | d_loss: 0.5265 | g_loss: 0.2640\n",
      "Epoch [   23/   50] | d_loss: 0.5247 | g_loss: 0.2829\n",
      "Epoch [   23/   50] | d_loss: 0.5220 | g_loss: 0.2551\n",
      "Epoch [   23/   50] | d_loss: 0.4946 | g_loss: 0.2652\n",
      "Epoch [   23/   50] | d_loss: 0.4901 | g_loss: 0.2565\n",
      "Epoch [   23/   50] | d_loss: 0.4838 | g_loss: 0.2698\n",
      "Epoch [   23/   50] | d_loss: 0.5146 | g_loss: 0.2883\n",
      "Epoch [   23/   50] | d_loss: 0.5086 | g_loss: 0.2367\n",
      "Epoch [   23/   50] | d_loss: 0.4757 | g_loss: 0.2477\n",
      "Epoch [   23/   50] | d_loss: 0.5091 | g_loss: 0.2476\n",
      "Epoch [   23/   50] | d_loss: 0.5164 | g_loss: 0.2330\n",
      "Epoch [   23/   50] | d_loss: 0.5044 | g_loss: 0.2910\n",
      "Epoch [   23/   50] | d_loss: 0.4892 | g_loss: 0.2510\n",
      "Epoch [   23/   50] | d_loss: 0.4987 | g_loss: 0.2388\n",
      "Epoch [   23/   50] | d_loss: 0.5031 | g_loss: 0.2539\n",
      "Epoch [   23/   50] | d_loss: 0.4435 | g_loss: 0.2657\n",
      "Epoch [   23/   50] | d_loss: 0.5285 | g_loss: 0.2804\n",
      "Epoch [   23/   50] | d_loss: 0.4659 | g_loss: 0.2639\n",
      "Epoch [   23/   50] | d_loss: 0.4899 | g_loss: 0.2767\n",
      "Epoch [   23/   50] | d_loss: 0.4996 | g_loss: 0.2412\n",
      "Epoch [   23/   50] | d_loss: 0.4704 | g_loss: 0.2876\n",
      "Epoch [   23/   50] | d_loss: 0.5077 | g_loss: 0.2442\n",
      "Epoch [   23/   50] | d_loss: 0.4777 | g_loss: 0.2525\n",
      "Epoch [   23/   50] | d_loss: 0.5339 | g_loss: 0.2647\n",
      "Epoch [   23/   50] | d_loss: 0.4992 | g_loss: 0.2426\n",
      "Epoch [   23/   50] | d_loss: 0.5040 | g_loss: 0.2629\n",
      "Epoch [   23/   50] | d_loss: 0.5276 | g_loss: 0.2467\n",
      "Epoch [   24/   50] | d_loss: 0.4932 | g_loss: 0.2292\n",
      "Epoch [   24/   50] | d_loss: 0.5136 | g_loss: 0.2555\n",
      "Epoch [   24/   50] | d_loss: 0.5075 | g_loss: 0.2662\n",
      "Epoch [   24/   50] | d_loss: 0.4943 | g_loss: 0.2467\n",
      "Epoch [   24/   50] | d_loss: 0.4984 | g_loss: 0.2758\n",
      "Epoch [   24/   50] | d_loss: 0.5104 | g_loss: 0.2390\n",
      "Epoch [   24/   50] | d_loss: 0.5353 | g_loss: 0.2592\n",
      "Epoch [   24/   50] | d_loss: 0.4791 | g_loss: 0.2402\n",
      "Epoch [   24/   50] | d_loss: 0.5439 | g_loss: 0.2575\n",
      "Epoch [   24/   50] | d_loss: 0.4872 | g_loss: 0.2628\n",
      "Epoch [   24/   50] | d_loss: 0.5307 | g_loss: 0.2762\n",
      "Epoch [   24/   50] | d_loss: 0.4287 | g_loss: 0.2548\n",
      "Epoch [   24/   50] | d_loss: 0.5090 | g_loss: 0.2515\n",
      "Epoch [   24/   50] | d_loss: 0.5374 | g_loss: 0.2927\n",
      "Epoch [   24/   50] | d_loss: 0.5122 | g_loss: 0.2360\n",
      "Epoch [   24/   50] | d_loss: 0.5004 | g_loss: 0.2493\n",
      "Epoch [   24/   50] | d_loss: 0.5182 | g_loss: 0.2752\n",
      "Epoch [   24/   50] | d_loss: 0.4736 | g_loss: 0.2443\n",
      "Epoch [   24/   50] | d_loss: 0.4865 | g_loss: 0.3030\n",
      "Epoch [   24/   50] | d_loss: 0.4994 | g_loss: 0.2554\n",
      "Epoch [   24/   50] | d_loss: 0.4878 | g_loss: 0.2601\n",
      "Epoch [   24/   50] | d_loss: 0.5303 | g_loss: 0.2661\n",
      "Epoch [   24/   50] | d_loss: 0.5369 | g_loss: 0.2485\n",
      "Epoch [   24/   50] | d_loss: 0.5177 | g_loss: 0.2716\n",
      "Epoch [   24/   50] | d_loss: 0.4517 | g_loss: 0.2535\n",
      "Epoch [   24/   50] | d_loss: 0.4728 | g_loss: 0.2901\n",
      "Epoch [   24/   50] | d_loss: 0.4999 | g_loss: 0.2425\n",
      "Epoch [   24/   50] | d_loss: 0.4811 | g_loss: 0.2399\n",
      "Epoch [   24/   50] | d_loss: 0.5161 | g_loss: 0.2250\n",
      "Epoch [   24/   50] | d_loss: 0.5172 | g_loss: 0.2738\n",
      "Epoch [   24/   50] | d_loss: 0.5115 | g_loss: 0.2560\n",
      "Epoch [   24/   50] | d_loss: 0.4932 | g_loss: 0.2591\n",
      "Epoch [   24/   50] | d_loss: 0.5009 | g_loss: 0.2474\n",
      "Epoch [   24/   50] | d_loss: 0.4744 | g_loss: 0.2756\n",
      "Epoch [   24/   50] | d_loss: 0.4940 | g_loss: 0.2519\n",
      "Epoch [   24/   50] | d_loss: 0.4842 | g_loss: 0.2822\n",
      "Epoch [   24/   50] | d_loss: 0.4274 | g_loss: 0.2691\n",
      "Epoch [   24/   50] | d_loss: 0.4543 | g_loss: 0.2657\n",
      "Epoch [   24/   50] | d_loss: 0.4761 | g_loss: 0.2765\n",
      "Epoch [   24/   50] | d_loss: 0.4976 | g_loss: 0.2819\n",
      "Epoch [   24/   50] | d_loss: 0.4990 | g_loss: 0.2632\n",
      "Epoch [   24/   50] | d_loss: 0.5095 | g_loss: 0.2785\n",
      "Epoch [   24/   50] | d_loss: 0.5187 | g_loss: 0.2502\n",
      "Epoch [   24/   50] | d_loss: 0.5124 | g_loss: 0.2324\n",
      "Epoch [   24/   50] | d_loss: 0.5237 | g_loss: 0.2269\n",
      "Epoch [   24/   50] | d_loss: 0.4911 | g_loss: 0.2541\n",
      "Epoch [   24/   50] | d_loss: 0.4733 | g_loss: 0.2453\n",
      "Epoch [   24/   50] | d_loss: 0.5394 | g_loss: 0.2712\n",
      "Epoch [   24/   50] | d_loss: 0.4992 | g_loss: 0.2562\n",
      "Epoch [   24/   50] | d_loss: 0.4587 | g_loss: 0.2621\n",
      "Epoch [   24/   50] | d_loss: 0.5244 | g_loss: 0.2527\n",
      "Epoch [   24/   50] | d_loss: 0.4788 | g_loss: 0.2976\n",
      "Epoch [   24/   50] | d_loss: 0.5306 | g_loss: 0.2681\n",
      "Epoch [   24/   50] | d_loss: 0.4933 | g_loss: 0.2471\n",
      "Epoch [   24/   50] | d_loss: 0.4784 | g_loss: 0.2779\n",
      "Epoch [   24/   50] | d_loss: 0.5005 | g_loss: 0.2527\n",
      "Epoch [   24/   50] | d_loss: 0.4864 | g_loss: 0.2634\n",
      "Epoch [   24/   50] | d_loss: 0.4761 | g_loss: 0.2504\n",
      "Epoch [   24/   50] | d_loss: 0.5455 | g_loss: 0.2435\n",
      "Epoch [   24/   50] | d_loss: 0.4856 | g_loss: 0.2480\n",
      "Epoch [   24/   50] | d_loss: 0.4812 | g_loss: 0.2369\n",
      "Epoch [   24/   50] | d_loss: 0.4947 | g_loss: 0.2626\n",
      "Epoch [   24/   50] | d_loss: 0.5433 | g_loss: 0.2226\n",
      "Epoch [   24/   50] | d_loss: 0.4979 | g_loss: 0.2701\n",
      "Epoch [   24/   50] | d_loss: 0.5276 | g_loss: 0.2300\n",
      "Epoch [   24/   50] | d_loss: 0.5085 | g_loss: 0.2805\n",
      "Epoch [   24/   50] | d_loss: 0.4958 | g_loss: 0.2584\n",
      "Epoch [   24/   50] | d_loss: 0.5299 | g_loss: 0.2848\n",
      "Epoch [   24/   50] | d_loss: 0.5161 | g_loss: 0.2652\n",
      "Epoch [   24/   50] | d_loss: 0.4603 | g_loss: 0.2454\n",
      "Epoch [   24/   50] | d_loss: 0.5164 | g_loss: 0.2387\n",
      "Epoch [   24/   50] | d_loss: 0.5049 | g_loss: 0.2515\n",
      "Epoch [   24/   50] | d_loss: 0.5079 | g_loss: 0.2603\n",
      "Epoch [   24/   50] | d_loss: 0.4638 | g_loss: 0.2703\n",
      "Epoch [   24/   50] | d_loss: 0.5193 | g_loss: 0.2865\n",
      "Epoch [   24/   50] | d_loss: 0.5006 | g_loss: 0.2355\n",
      "Epoch [   24/   50] | d_loss: 0.4863 | g_loss: 0.2601\n",
      "Epoch [   24/   50] | d_loss: 0.4997 | g_loss: 0.2444\n",
      "Epoch [   24/   50] | d_loss: 0.4834 | g_loss: 0.2494\n",
      "Epoch [   24/   50] | d_loss: 0.4834 | g_loss: 0.2424\n",
      "Epoch [   24/   50] | d_loss: 0.5015 | g_loss: 0.2341\n",
      "Epoch [   24/   50] | d_loss: 0.5060 | g_loss: 0.2440\n",
      "Epoch [   24/   50] | d_loss: 0.5042 | g_loss: 0.2630\n",
      "Epoch [   24/   50] | d_loss: 0.5303 | g_loss: 0.2671\n",
      "Epoch [   24/   50] | d_loss: 0.5188 | g_loss: 0.2880\n",
      "Epoch [   24/   50] | d_loss: 0.5022 | g_loss: 0.2612\n",
      "Epoch [   24/   50] | d_loss: 0.5029 | g_loss: 0.2429\n",
      "Epoch [   24/   50] | d_loss: 0.4609 | g_loss: 0.2860\n",
      "Epoch [   24/   50] | d_loss: 0.5024 | g_loss: 0.2415\n",
      "Epoch [   24/   50] | d_loss: 0.4989 | g_loss: 0.2660\n",
      "Epoch [   25/   50] | d_loss: 0.4972 | g_loss: 0.2490\n",
      "Epoch [   25/   50] | d_loss: 0.5271 | g_loss: 0.2559\n",
      "Epoch [   25/   50] | d_loss: 0.4971 | g_loss: 0.2324\n",
      "Epoch [   25/   50] | d_loss: 0.5085 | g_loss: 0.2683\n",
      "Epoch [   25/   50] | d_loss: 0.5111 | g_loss: 0.2733\n",
      "Epoch [   25/   50] | d_loss: 0.4968 | g_loss: 0.2375\n",
      "Epoch [   25/   50] | d_loss: 0.5235 | g_loss: 0.2587\n",
      "Epoch [   25/   50] | d_loss: 0.5051 | g_loss: 0.2387\n",
      "Epoch [   25/   50] | d_loss: 0.4921 | g_loss: 0.2798\n",
      "Epoch [   25/   50] | d_loss: 0.4804 | g_loss: 0.2754\n",
      "Epoch [   25/   50] | d_loss: 0.5021 | g_loss: 0.2849\n",
      "Epoch [   25/   50] | d_loss: 0.4721 | g_loss: 0.2598\n",
      "Epoch [   25/   50] | d_loss: 0.5191 | g_loss: 0.2490\n",
      "Epoch [   25/   50] | d_loss: 0.4837 | g_loss: 0.2442\n",
      "Epoch [   25/   50] | d_loss: 0.5205 | g_loss: 0.2904\n",
      "Epoch [   25/   50] | d_loss: 0.5206 | g_loss: 0.2405\n",
      "Epoch [   25/   50] | d_loss: 0.5204 | g_loss: 0.2695\n",
      "Epoch [   25/   50] | d_loss: 0.5146 | g_loss: 0.2702\n",
      "Epoch [   25/   50] | d_loss: 0.5016 | g_loss: 0.2390\n",
      "Epoch [   25/   50] | d_loss: 0.4949 | g_loss: 0.2764\n",
      "Epoch [   25/   50] | d_loss: 0.4991 | g_loss: 0.2456\n",
      "Epoch [   25/   50] | d_loss: 0.4938 | g_loss: 0.2437\n",
      "Epoch [   25/   50] | d_loss: 0.4866 | g_loss: 0.2546\n",
      "Epoch [   25/   50] | d_loss: 0.5160 | g_loss: 0.2312\n",
      "Epoch [   25/   50] | d_loss: 0.5142 | g_loss: 0.2814\n",
      "Epoch [   25/   50] | d_loss: 0.4875 | g_loss: 0.2583\n",
      "Epoch [   25/   50] | d_loss: 0.5180 | g_loss: 0.2484\n",
      "Epoch [   25/   50] | d_loss: 0.4617 | g_loss: 0.2449\n",
      "Epoch [   25/   50] | d_loss: 0.4830 | g_loss: 0.2694\n",
      "Epoch [   25/   50] | d_loss: 0.4862 | g_loss: 0.2734\n",
      "Epoch [   25/   50] | d_loss: 0.5297 | g_loss: 0.2612\n",
      "Epoch [   25/   50] | d_loss: 0.5037 | g_loss: 0.2451\n",
      "Epoch [   25/   50] | d_loss: 0.5259 | g_loss: 0.2530\n",
      "Epoch [   25/   50] | d_loss: 0.5495 | g_loss: 0.2812\n",
      "Epoch [   25/   50] | d_loss: 0.4772 | g_loss: 0.2869\n",
      "Epoch [   25/   50] | d_loss: 0.4519 | g_loss: 0.2562\n",
      "Epoch [   25/   50] | d_loss: 0.5197 | g_loss: 0.2422\n",
      "Epoch [   25/   50] | d_loss: 0.5009 | g_loss: 0.2378\n",
      "Epoch [   25/   50] | d_loss: 0.4717 | g_loss: 0.2448\n",
      "Epoch [   25/   50] | d_loss: 0.5004 | g_loss: 0.2581\n",
      "Epoch [   25/   50] | d_loss: 0.5078 | g_loss: 0.2612\n",
      "Epoch [   25/   50] | d_loss: 0.5231 | g_loss: 0.2736\n",
      "Epoch [   25/   50] | d_loss: 0.4886 | g_loss: 0.2896\n",
      "Epoch [   25/   50] | d_loss: 0.5278 | g_loss: 0.2456\n",
      "Epoch [   25/   50] | d_loss: 0.4791 | g_loss: 0.2630\n",
      "Epoch [   25/   50] | d_loss: 0.5128 | g_loss: 0.2558\n",
      "Epoch [   25/   50] | d_loss: 0.4893 | g_loss: 0.2702\n",
      "Epoch [   25/   50] | d_loss: 0.4996 | g_loss: 0.2797\n",
      "Epoch [   25/   50] | d_loss: 0.5108 | g_loss: 0.2336\n",
      "Epoch [   25/   50] | d_loss: 0.4708 | g_loss: 0.2561\n",
      "Epoch [   25/   50] | d_loss: 0.4827 | g_loss: 0.2855\n",
      "Epoch [   25/   50] | d_loss: 0.4968 | g_loss: 0.2649\n",
      "Epoch [   25/   50] | d_loss: 0.5251 | g_loss: 0.2312\n",
      "Epoch [   25/   50] | d_loss: 0.4555 | g_loss: 0.2596\n",
      "Epoch [   25/   50] | d_loss: 0.5088 | g_loss: 0.2701\n",
      "Epoch [   25/   50] | d_loss: 0.4673 | g_loss: 0.2813\n",
      "Epoch [   25/   50] | d_loss: 0.5297 | g_loss: 0.2739\n",
      "Epoch [   25/   50] | d_loss: 0.5169 | g_loss: 0.2593\n",
      "Epoch [   25/   50] | d_loss: 0.5210 | g_loss: 0.2608\n",
      "Epoch [   25/   50] | d_loss: 0.5224 | g_loss: 0.2568\n",
      "Epoch [   25/   50] | d_loss: 0.4725 | g_loss: 0.2644\n",
      "Epoch [   25/   50] | d_loss: 0.5179 | g_loss: 0.3033\n",
      "Epoch [   25/   50] | d_loss: 0.4857 | g_loss: 0.2747\n",
      "Epoch [   25/   50] | d_loss: 0.4667 | g_loss: 0.2580\n",
      "Epoch [   25/   50] | d_loss: 0.5008 | g_loss: 0.2834\n",
      "Epoch [   25/   50] | d_loss: 0.4833 | g_loss: 0.2659\n",
      "Epoch [   25/   50] | d_loss: 0.5089 | g_loss: 0.2620\n",
      "Epoch [   25/   50] | d_loss: 0.5439 | g_loss: 0.2752\n",
      "Epoch [   25/   50] | d_loss: 0.5237 | g_loss: 0.2993\n",
      "Epoch [   25/   50] | d_loss: 0.4854 | g_loss: 0.2679\n",
      "Epoch [   25/   50] | d_loss: 0.5107 | g_loss: 0.2590\n",
      "Epoch [   25/   50] | d_loss: 0.4997 | g_loss: 0.2666\n",
      "Epoch [   25/   50] | d_loss: 0.4640 | g_loss: 0.2704\n",
      "Epoch [   25/   50] | d_loss: 0.5161 | g_loss: 0.2664\n",
      "Epoch [   25/   50] | d_loss: 0.5030 | g_loss: 0.2384\n",
      "Epoch [   25/   50] | d_loss: 0.4849 | g_loss: 0.2665\n",
      "Epoch [   25/   50] | d_loss: 0.5238 | g_loss: 0.2751\n",
      "Epoch [   25/   50] | d_loss: 0.5165 | g_loss: 0.2574\n",
      "Epoch [   25/   50] | d_loss: 0.5261 | g_loss: 0.2399\n",
      "Epoch [   25/   50] | d_loss: 0.4934 | g_loss: 0.2635\n",
      "Epoch [   25/   50] | d_loss: 0.5080 | g_loss: 0.2709\n",
      "Epoch [   25/   50] | d_loss: 0.5336 | g_loss: 0.2499\n",
      "Epoch [   25/   50] | d_loss: 0.4907 | g_loss: 0.2330\n",
      "Epoch [   25/   50] | d_loss: 0.5400 | g_loss: 0.2452\n",
      "Epoch [   25/   50] | d_loss: 0.4991 | g_loss: 0.2647\n",
      "Epoch [   25/   50] | d_loss: 0.5176 | g_loss: 0.2726\n",
      "Epoch [   25/   50] | d_loss: 0.5312 | g_loss: 0.2699\n",
      "Epoch [   25/   50] | d_loss: 0.4912 | g_loss: 0.2459\n",
      "Epoch [   25/   50] | d_loss: 0.5151 | g_loss: 0.2702\n",
      "Epoch [   25/   50] | d_loss: 0.5179 | g_loss: 0.2751\n",
      "Epoch [   26/   50] | d_loss: 0.4734 | g_loss: 0.2693\n",
      "Epoch [   26/   50] | d_loss: 0.5443 | g_loss: 0.2493\n",
      "Epoch [   26/   50] | d_loss: 0.4819 | g_loss: 0.2724\n",
      "Epoch [   26/   50] | d_loss: 0.4697 | g_loss: 0.2568\n",
      "Epoch [   26/   50] | d_loss: 0.4945 | g_loss: 0.2516\n",
      "Epoch [   26/   50] | d_loss: 0.5168 | g_loss: 0.2966\n",
      "Epoch [   26/   50] | d_loss: 0.4956 | g_loss: 0.2582\n",
      "Epoch [   26/   50] | d_loss: 0.5153 | g_loss: 0.2575\n",
      "Epoch [   26/   50] | d_loss: 0.5028 | g_loss: 0.2779\n",
      "Epoch [   26/   50] | d_loss: 0.4721 | g_loss: 0.2625\n",
      "Epoch [   26/   50] | d_loss: 0.5020 | g_loss: 0.2671\n",
      "Epoch [   26/   50] | d_loss: 0.4911 | g_loss: 0.2632\n",
      "Epoch [   26/   50] | d_loss: 0.4856 | g_loss: 0.2698\n",
      "Epoch [   26/   50] | d_loss: 0.4803 | g_loss: 0.2604\n",
      "Epoch [   26/   50] | d_loss: 0.4753 | g_loss: 0.2363\n",
      "Epoch [   26/   50] | d_loss: 0.5142 | g_loss: 0.2756\n",
      "Epoch [   26/   50] | d_loss: 0.5003 | g_loss: 0.2246\n",
      "Epoch [   26/   50] | d_loss: 0.5151 | g_loss: 0.2614\n",
      "Epoch [   26/   50] | d_loss: 0.4976 | g_loss: 0.2631\n",
      "Epoch [   26/   50] | d_loss: 0.4609 | g_loss: 0.2727\n",
      "Epoch [   26/   50] | d_loss: 0.4889 | g_loss: 0.2604\n",
      "Epoch [   26/   50] | d_loss: 0.5380 | g_loss: 0.2717\n",
      "Epoch [   26/   50] | d_loss: 0.5284 | g_loss: 0.2397\n",
      "Epoch [   26/   50] | d_loss: 0.5292 | g_loss: 0.2502\n",
      "Epoch [   26/   50] | d_loss: 0.5553 | g_loss: 0.2687\n",
      "Epoch [   26/   50] | d_loss: 0.5167 | g_loss: 0.2417\n",
      "Epoch [   26/   50] | d_loss: 0.4906 | g_loss: 0.2426\n",
      "Epoch [   26/   50] | d_loss: 0.4977 | g_loss: 0.2455\n",
      "Epoch [   26/   50] | d_loss: 0.4593 | g_loss: 0.2458\n",
      "Epoch [   26/   50] | d_loss: 0.5112 | g_loss: 0.2663\n",
      "Epoch [   26/   50] | d_loss: 0.4887 | g_loss: 0.2403\n",
      "Epoch [   26/   50] | d_loss: 0.4678 | g_loss: 0.2359\n",
      "Epoch [   26/   50] | d_loss: 0.4873 | g_loss: 0.2519\n",
      "Epoch [   26/   50] | d_loss: 0.4920 | g_loss: 0.2506\n",
      "Epoch [   26/   50] | d_loss: 0.5057 | g_loss: 0.2482\n",
      "Epoch [   26/   50] | d_loss: 0.5206 | g_loss: 0.2887\n",
      "Epoch [   26/   50] | d_loss: 0.4841 | g_loss: 0.2645\n",
      "Epoch [   26/   50] | d_loss: 0.4861 | g_loss: 0.2355\n",
      "Epoch [   26/   50] | d_loss: 0.4916 | g_loss: 0.2568\n",
      "Epoch [   26/   50] | d_loss: 0.5111 | g_loss: 0.2660\n",
      "Epoch [   26/   50] | d_loss: 0.4800 | g_loss: 0.2622\n",
      "Epoch [   26/   50] | d_loss: 0.4920 | g_loss: 0.2583\n",
      "Epoch [   26/   50] | d_loss: 0.5114 | g_loss: 0.2518\n",
      "Epoch [   26/   50] | d_loss: 0.4936 | g_loss: 0.2390\n",
      "Epoch [   26/   50] | d_loss: 0.5079 | g_loss: 0.2597\n",
      "Epoch [   26/   50] | d_loss: 0.5030 | g_loss: 0.2500\n",
      "Epoch [   26/   50] | d_loss: 0.4976 | g_loss: 0.2558\n",
      "Epoch [   26/   50] | d_loss: 0.5499 | g_loss: 0.2668\n",
      "Epoch [   26/   50] | d_loss: 0.4744 | g_loss: 0.2381\n",
      "Epoch [   26/   50] | d_loss: 0.4946 | g_loss: 0.2485\n",
      "Epoch [   26/   50] | d_loss: 0.5240 | g_loss: 0.2567\n",
      "Epoch [   26/   50] | d_loss: 0.5132 | g_loss: 0.2676\n",
      "Epoch [   26/   50] | d_loss: 0.5081 | g_loss: 0.2391\n",
      "Epoch [   26/   50] | d_loss: 0.5006 | g_loss: 0.2957\n",
      "Epoch [   26/   50] | d_loss: 0.4993 | g_loss: 0.2621\n",
      "Epoch [   26/   50] | d_loss: 0.4984 | g_loss: 0.2731\n",
      "Epoch [   26/   50] | d_loss: 0.4936 | g_loss: 0.2729\n",
      "Epoch [   26/   50] | d_loss: 0.5074 | g_loss: 0.2628\n",
      "Epoch [   26/   50] | d_loss: 0.4592 | g_loss: 0.3030\n",
      "Epoch [   26/   50] | d_loss: 0.5122 | g_loss: 0.2658\n",
      "Epoch [   26/   50] | d_loss: 0.4850 | g_loss: 0.2264\n",
      "Epoch [   26/   50] | d_loss: 0.5414 | g_loss: 0.2697\n",
      "Epoch [   26/   50] | d_loss: 0.5052 | g_loss: 0.2867\n",
      "Epoch [   26/   50] | d_loss: 0.5421 | g_loss: 0.2283\n",
      "Epoch [   26/   50] | d_loss: 0.5307 | g_loss: 0.2507\n",
      "Epoch [   26/   50] | d_loss: 0.5443 | g_loss: 0.2475\n",
      "Epoch [   26/   50] | d_loss: 0.4773 | g_loss: 0.2666\n",
      "Epoch [   26/   50] | d_loss: 0.5273 | g_loss: 0.2855\n",
      "Epoch [   26/   50] | d_loss: 0.4855 | g_loss: 0.2570\n",
      "Epoch [   26/   50] | d_loss: 0.4926 | g_loss: 0.2951\n",
      "Epoch [   26/   50] | d_loss: 0.4773 | g_loss: 0.2684\n",
      "Epoch [   26/   50] | d_loss: 0.4719 | g_loss: 0.2487\n",
      "Epoch [   26/   50] | d_loss: 0.5164 | g_loss: 0.2660\n",
      "Epoch [   26/   50] | d_loss: 0.4857 | g_loss: 0.2568\n",
      "Epoch [   26/   50] | d_loss: 0.5134 | g_loss: 0.2343\n",
      "Epoch [   26/   50] | d_loss: 0.4928 | g_loss: 0.2591\n",
      "Epoch [   26/   50] | d_loss: 0.5384 | g_loss: 0.2468\n",
      "Epoch [   26/   50] | d_loss: 0.5242 | g_loss: 0.2454\n",
      "Epoch [   26/   50] | d_loss: 0.5112 | g_loss: 0.2773\n",
      "Epoch [   26/   50] | d_loss: 0.4814 | g_loss: 0.2603\n",
      "Epoch [   26/   50] | d_loss: 0.4990 | g_loss: 0.2647\n",
      "Epoch [   26/   50] | d_loss: 0.4746 | g_loss: 0.2570\n",
      "Epoch [   26/   50] | d_loss: 0.4867 | g_loss: 0.2580\n",
      "Epoch [   26/   50] | d_loss: 0.5169 | g_loss: 0.2630\n",
      "Epoch [   26/   50] | d_loss: 0.4972 | g_loss: 0.2744\n",
      "Epoch [   26/   50] | d_loss: 0.5492 | g_loss: 0.2267\n",
      "Epoch [   26/   50] | d_loss: 0.4886 | g_loss: 0.2407\n",
      "Epoch [   26/   50] | d_loss: 0.5307 | g_loss: 0.2432\n",
      "Epoch [   26/   50] | d_loss: 0.5050 | g_loss: 0.2534\n",
      "Epoch [   26/   50] | d_loss: 0.4973 | g_loss: 0.2407\n",
      "Epoch [   27/   50] | d_loss: 0.5237 | g_loss: 0.2358\n",
      "Epoch [   27/   50] | d_loss: 0.5141 | g_loss: 0.2721\n",
      "Epoch [   27/   50] | d_loss: 0.4714 | g_loss: 0.2525\n",
      "Epoch [   27/   50] | d_loss: 0.4960 | g_loss: 0.2668\n",
      "Epoch [   27/   50] | d_loss: 0.4937 | g_loss: 0.2687\n",
      "Epoch [   27/   50] | d_loss: 0.4680 | g_loss: 0.2845\n",
      "Epoch [   27/   50] | d_loss: 0.5085 | g_loss: 0.2735\n",
      "Epoch [   27/   50] | d_loss: 0.4822 | g_loss: 0.2562\n",
      "Epoch [   27/   50] | d_loss: 0.4833 | g_loss: 0.2446\n",
      "Epoch [   27/   50] | d_loss: 0.5089 | g_loss: 0.2306\n",
      "Epoch [   27/   50] | d_loss: 0.4936 | g_loss: 0.2575\n",
      "Epoch [   27/   50] | d_loss: 0.4820 | g_loss: 0.2634\n",
      "Epoch [   27/   50] | d_loss: 0.5024 | g_loss: 0.2779\n",
      "Epoch [   27/   50] | d_loss: 0.4992 | g_loss: 0.2340\n",
      "Epoch [   27/   50] | d_loss: 0.5350 | g_loss: 0.2920\n",
      "Epoch [   27/   50] | d_loss: 0.5044 | g_loss: 0.2350\n",
      "Epoch [   27/   50] | d_loss: 0.4932 | g_loss: 0.2814\n",
      "Epoch [   27/   50] | d_loss: 0.4971 | g_loss: 0.2510\n",
      "Epoch [   27/   50] | d_loss: 0.5140 | g_loss: 0.2581\n",
      "Epoch [   27/   50] | d_loss: 0.4975 | g_loss: 0.3131\n",
      "Epoch [   27/   50] | d_loss: 0.5109 | g_loss: 0.2844\n",
      "Epoch [   27/   50] | d_loss: 0.4537 | g_loss: 0.2472\n",
      "Epoch [   27/   50] | d_loss: 0.5267 | g_loss: 0.2763\n",
      "Epoch [   27/   50] | d_loss: 0.4763 | g_loss: 0.2570\n",
      "Epoch [   27/   50] | d_loss: 0.5103 | g_loss: 0.2545\n",
      "Epoch [   27/   50] | d_loss: 0.4854 | g_loss: 0.2274\n",
      "Epoch [   27/   50] | d_loss: 0.4525 | g_loss: 0.2341\n",
      "Epoch [   27/   50] | d_loss: 0.4651 | g_loss: 0.2257\n",
      "Epoch [   27/   50] | d_loss: 0.4843 | g_loss: 0.2569\n",
      "Epoch [   27/   50] | d_loss: 0.4855 | g_loss: 0.2647\n",
      "Epoch [   27/   50] | d_loss: 0.5109 | g_loss: 0.2803\n",
      "Epoch [   27/   50] | d_loss: 0.4652 | g_loss: 0.2553\n",
      "Epoch [   27/   50] | d_loss: 0.5314 | g_loss: 0.2447\n",
      "Epoch [   27/   50] | d_loss: 0.4721 | g_loss: 0.2292\n",
      "Epoch [   27/   50] | d_loss: 0.4578 | g_loss: 0.2660\n",
      "Epoch [   27/   50] | d_loss: 0.5063 | g_loss: 0.2665\n",
      "Epoch [   27/   50] | d_loss: 0.5473 | g_loss: 0.2695\n",
      "Epoch [   27/   50] | d_loss: 0.5185 | g_loss: 0.2685\n",
      "Epoch [   27/   50] | d_loss: 0.4866 | g_loss: 0.2686\n",
      "Epoch [   27/   50] | d_loss: 0.5076 | g_loss: 0.2729\n",
      "Epoch [   27/   50] | d_loss: 0.4854 | g_loss: 0.2807\n",
      "Epoch [   27/   50] | d_loss: 0.4733 | g_loss: 0.2929\n",
      "Epoch [   27/   50] | d_loss: 0.5049 | g_loss: 0.2605\n",
      "Epoch [   27/   50] | d_loss: 0.4873 | g_loss: 0.2855\n",
      "Epoch [   27/   50] | d_loss: 0.5138 | g_loss: 0.2483\n",
      "Epoch [   27/   50] | d_loss: 0.4728 | g_loss: 0.2925\n",
      "Epoch [   27/   50] | d_loss: 0.4845 | g_loss: 0.2549\n",
      "Epoch [   27/   50] | d_loss: 0.4711 | g_loss: 0.3009\n",
      "Epoch [   27/   50] | d_loss: 0.5238 | g_loss: 0.2617\n",
      "Epoch [   27/   50] | d_loss: 0.4986 | g_loss: 0.2178\n",
      "Epoch [   27/   50] | d_loss: 0.5141 | g_loss: 0.2546\n",
      "Epoch [   27/   50] | d_loss: 0.5139 | g_loss: 0.2603\n",
      "Epoch [   27/   50] | d_loss: 0.5329 | g_loss: 0.2432\n",
      "Epoch [   27/   50] | d_loss: 0.4940 | g_loss: 0.2965\n",
      "Epoch [   27/   50] | d_loss: 0.4979 | g_loss: 0.2732\n",
      "Epoch [   27/   50] | d_loss: 0.5379 | g_loss: 0.2542\n",
      "Epoch [   27/   50] | d_loss: 0.5390 | g_loss: 0.2503\n",
      "Epoch [   27/   50] | d_loss: 0.4973 | g_loss: 0.2382\n",
      "Epoch [   27/   50] | d_loss: 0.5112 | g_loss: 0.2491\n",
      "Epoch [   27/   50] | d_loss: 0.4996 | g_loss: 0.2820\n",
      "Epoch [   27/   50] | d_loss: 0.4750 | g_loss: 0.2419\n",
      "Epoch [   27/   50] | d_loss: 0.4814 | g_loss: 0.2359\n",
      "Epoch [   27/   50] | d_loss: 0.5069 | g_loss: 0.2674\n",
      "Epoch [   27/   50] | d_loss: 0.4937 | g_loss: 0.2568\n",
      "Epoch [   27/   50] | d_loss: 0.4691 | g_loss: 0.2484\n",
      "Epoch [   27/   50] | d_loss: 0.5034 | g_loss: 0.2529\n",
      "Epoch [   27/   50] | d_loss: 0.4668 | g_loss: 0.2285\n",
      "Epoch [   27/   50] | d_loss: 0.4816 | g_loss: 0.2645\n",
      "Epoch [   27/   50] | d_loss: 0.4935 | g_loss: 0.2829\n",
      "Epoch [   27/   50] | d_loss: 0.5581 | g_loss: 0.2728\n",
      "Epoch [   27/   50] | d_loss: 0.5436 | g_loss: 0.2131\n",
      "Epoch [   27/   50] | d_loss: 0.5597 | g_loss: 0.2389\n",
      "Epoch [   27/   50] | d_loss: 0.5138 | g_loss: 0.2440\n",
      "Epoch [   27/   50] | d_loss: 0.4902 | g_loss: 0.2457\n",
      "Epoch [   27/   50] | d_loss: 0.4883 | g_loss: 0.2831\n",
      "Epoch [   27/   50] | d_loss: 0.5184 | g_loss: 0.2365\n",
      "Epoch [   27/   50] | d_loss: 0.5082 | g_loss: 0.2617\n",
      "Epoch [   27/   50] | d_loss: 0.5041 | g_loss: 0.2715\n",
      "Epoch [   27/   50] | d_loss: 0.5142 | g_loss: 0.2599\n",
      "Epoch [   27/   50] | d_loss: 0.5219 | g_loss: 0.2388\n",
      "Epoch [   27/   50] | d_loss: 0.5299 | g_loss: 0.2725\n",
      "Epoch [   27/   50] | d_loss: 0.5151 | g_loss: 0.2380\n",
      "Epoch [   27/   50] | d_loss: 0.5172 | g_loss: 0.2625\n",
      "Epoch [   27/   50] | d_loss: 0.5325 | g_loss: 0.2744\n",
      "Epoch [   27/   50] | d_loss: 0.5104 | g_loss: 0.2420\n",
      "Epoch [   27/   50] | d_loss: 0.5281 | g_loss: 0.2806\n",
      "Epoch [   27/   50] | d_loss: 0.5230 | g_loss: 0.2950\n",
      "Epoch [   27/   50] | d_loss: 0.5262 | g_loss: 0.2942\n",
      "Epoch [   27/   50] | d_loss: 0.4810 | g_loss: 0.2508\n",
      "Epoch [   27/   50] | d_loss: 0.5272 | g_loss: 0.2291\n",
      "Epoch [   28/   50] | d_loss: 0.4737 | g_loss: 0.2426\n",
      "Epoch [   28/   50] | d_loss: 0.4977 | g_loss: 0.2408\n",
      "Epoch [   28/   50] | d_loss: 0.5202 | g_loss: 0.3085\n",
      "Epoch [   28/   50] | d_loss: 0.5142 | g_loss: 0.2222\n",
      "Epoch [   28/   50] | d_loss: 0.4976 | g_loss: 0.2720\n",
      "Epoch [   28/   50] | d_loss: 0.5261 | g_loss: 0.2746\n",
      "Epoch [   28/   50] | d_loss: 0.5320 | g_loss: 0.2504\n",
      "Epoch [   28/   50] | d_loss: 0.5072 | g_loss: 0.2523\n",
      "Epoch [   28/   50] | d_loss: 0.4838 | g_loss: 0.2939\n",
      "Epoch [   28/   50] | d_loss: 0.5313 | g_loss: 0.2585\n",
      "Epoch [   28/   50] | d_loss: 0.4737 | g_loss: 0.2505\n",
      "Epoch [   28/   50] | d_loss: 0.5516 | g_loss: 0.2361\n",
      "Epoch [   28/   50] | d_loss: 0.5103 | g_loss: 0.2714\n",
      "Epoch [   28/   50] | d_loss: 0.4768 | g_loss: 0.2413\n",
      "Epoch [   28/   50] | d_loss: 0.5061 | g_loss: 0.2631\n",
      "Epoch [   28/   50] | d_loss: 0.4976 | g_loss: 0.2678\n",
      "Epoch [   28/   50] | d_loss: 0.5000 | g_loss: 0.2468\n",
      "Epoch [   28/   50] | d_loss: 0.5184 | g_loss: 0.2478\n",
      "Epoch [   28/   50] | d_loss: 0.4947 | g_loss: 0.2906\n",
      "Epoch [   28/   50] | d_loss: 0.4834 | g_loss: 0.2512\n",
      "Epoch [   28/   50] | d_loss: 0.4772 | g_loss: 0.2481\n",
      "Epoch [   28/   50] | d_loss: 0.4685 | g_loss: 0.2250\n",
      "Epoch [   28/   50] | d_loss: 0.5319 | g_loss: 0.2244\n",
      "Epoch [   28/   50] | d_loss: 0.4893 | g_loss: 0.2358\n",
      "Epoch [   28/   50] | d_loss: 0.5331 | g_loss: 0.2556\n",
      "Epoch [   28/   50] | d_loss: 0.4963 | g_loss: 0.2479\n",
      "Epoch [   28/   50] | d_loss: 0.4914 | g_loss: 0.2426\n",
      "Epoch [   28/   50] | d_loss: 0.4849 | g_loss: 0.2587\n",
      "Epoch [   28/   50] | d_loss: 0.5016 | g_loss: 0.2693\n",
      "Epoch [   28/   50] | d_loss: 0.4660 | g_loss: 0.2617\n",
      "Epoch [   28/   50] | d_loss: 0.5251 | g_loss: 0.2794\n",
      "Epoch [   28/   50] | d_loss: 0.4804 | g_loss: 0.2770\n",
      "Epoch [   28/   50] | d_loss: 0.5115 | g_loss: 0.2307\n",
      "Epoch [   28/   50] | d_loss: 0.5626 | g_loss: 0.2839\n",
      "Epoch [   28/   50] | d_loss: 0.5090 | g_loss: 0.2486\n",
      "Epoch [   28/   50] | d_loss: 0.5134 | g_loss: 0.2609\n",
      "Epoch [   28/   50] | d_loss: 0.4878 | g_loss: 0.2948\n",
      "Epoch [   28/   50] | d_loss: 0.4582 | g_loss: 0.2756\n",
      "Epoch [   28/   50] | d_loss: 0.5322 | g_loss: 0.2411\n",
      "Epoch [   28/   50] | d_loss: 0.4842 | g_loss: 0.2746\n",
      "Epoch [   28/   50] | d_loss: 0.5023 | g_loss: 0.2895\n",
      "Epoch [   28/   50] | d_loss: 0.4721 | g_loss: 0.2441\n",
      "Epoch [   28/   50] | d_loss: 0.5200 | g_loss: 0.2485\n",
      "Epoch [   28/   50] | d_loss: 0.5134 | g_loss: 0.2874\n",
      "Epoch [   28/   50] | d_loss: 0.4980 | g_loss: 0.2582\n",
      "Epoch [   28/   50] | d_loss: 0.5069 | g_loss: 0.2592\n",
      "Epoch [   28/   50] | d_loss: 0.5305 | g_loss: 0.2848\n",
      "Epoch [   28/   50] | d_loss: 0.4868 | g_loss: 0.2573\n",
      "Epoch [   28/   50] | d_loss: 0.5069 | g_loss: 0.2502\n",
      "Epoch [   28/   50] | d_loss: 0.4863 | g_loss: 0.2546\n",
      "Epoch [   28/   50] | d_loss: 0.5130 | g_loss: 0.2430\n",
      "Epoch [   28/   50] | d_loss: 0.5143 | g_loss: 0.2641\n",
      "Epoch [   28/   50] | d_loss: 0.5031 | g_loss: 0.2653\n",
      "Epoch [   28/   50] | d_loss: 0.4836 | g_loss: 0.2757\n",
      "Epoch [   28/   50] | d_loss: 0.4924 | g_loss: 0.2178\n",
      "Epoch [   28/   50] | d_loss: 0.5443 | g_loss: 0.2620\n",
      "Epoch [   28/   50] | d_loss: 0.5152 | g_loss: 0.2784\n",
      "Epoch [   28/   50] | d_loss: 0.5296 | g_loss: 0.2629\n",
      "Epoch [   28/   50] | d_loss: 0.5208 | g_loss: 0.2286\n",
      "Epoch [   28/   50] | d_loss: 0.5217 | g_loss: 0.2574\n",
      "Epoch [   28/   50] | d_loss: 0.5004 | g_loss: 0.2398\n",
      "Epoch [   28/   50] | d_loss: 0.5054 | g_loss: 0.2485\n",
      "Epoch [   28/   50] | d_loss: 0.5457 | g_loss: 0.2241\n",
      "Epoch [   28/   50] | d_loss: 0.5112 | g_loss: 0.2436\n",
      "Epoch [   28/   50] | d_loss: 0.5599 | g_loss: 0.2379\n",
      "Epoch [   28/   50] | d_loss: 0.4655 | g_loss: 0.2423\n",
      "Epoch [   28/   50] | d_loss: 0.4482 | g_loss: 0.2594\n",
      "Epoch [   28/   50] | d_loss: 0.5150 | g_loss: 0.2264\n",
      "Epoch [   28/   50] | d_loss: 0.5215 | g_loss: 0.2525\n",
      "Epoch [   28/   50] | d_loss: 0.5030 | g_loss: 0.2652\n",
      "Epoch [   28/   50] | d_loss: 0.4821 | g_loss: 0.2512\n",
      "Epoch [   28/   50] | d_loss: 0.4868 | g_loss: 0.2697\n",
      "Epoch [   28/   50] | d_loss: 0.4979 | g_loss: 0.2469\n",
      "Epoch [   28/   50] | d_loss: 0.5125 | g_loss: 0.2616\n",
      "Epoch [   28/   50] | d_loss: 0.4815 | g_loss: 0.2395\n",
      "Epoch [   28/   50] | d_loss: 0.4895 | g_loss: 0.2492\n",
      "Epoch [   28/   50] | d_loss: 0.4978 | g_loss: 0.2470\n",
      "Epoch [   28/   50] | d_loss: 0.5181 | g_loss: 0.2730\n",
      "Epoch [   28/   50] | d_loss: 0.4782 | g_loss: 0.2444\n",
      "Epoch [   28/   50] | d_loss: 0.5361 | g_loss: 0.2683\n",
      "Epoch [   28/   50] | d_loss: 0.5054 | g_loss: 0.2776\n",
      "Epoch [   28/   50] | d_loss: 0.5032 | g_loss: 0.2531\n",
      "Epoch [   28/   50] | d_loss: 0.5155 | g_loss: 0.2774\n",
      "Epoch [   28/   50] | d_loss: 0.4724 | g_loss: 0.2565\n",
      "Epoch [   28/   50] | d_loss: 0.4904 | g_loss: 0.2648\n",
      "Epoch [   28/   50] | d_loss: 0.4900 | g_loss: 0.2681\n",
      "Epoch [   28/   50] | d_loss: 0.4865 | g_loss: 0.2676\n",
      "Epoch [   28/   50] | d_loss: 0.4900 | g_loss: 0.2518\n",
      "Epoch [   28/   50] | d_loss: 0.4614 | g_loss: 0.2829\n",
      "Epoch [   28/   50] | d_loss: 0.5248 | g_loss: 0.2339\n",
      "Epoch [   29/   50] | d_loss: 0.5033 | g_loss: 0.2438\n",
      "Epoch [   29/   50] | d_loss: 0.5139 | g_loss: 0.2578\n",
      "Epoch [   29/   50] | d_loss: 0.5116 | g_loss: 0.2750\n",
      "Epoch [   29/   50] | d_loss: 0.5157 | g_loss: 0.2472\n",
      "Epoch [   29/   50] | d_loss: 0.5015 | g_loss: 0.2622\n",
      "Epoch [   29/   50] | d_loss: 0.4584 | g_loss: 0.2538\n",
      "Epoch [   29/   50] | d_loss: 0.4803 | g_loss: 0.2652\n",
      "Epoch [   29/   50] | d_loss: 0.5100 | g_loss: 0.2864\n",
      "Epoch [   29/   50] | d_loss: 0.4918 | g_loss: 0.2911\n",
      "Epoch [   29/   50] | d_loss: 0.5154 | g_loss: 0.2749\n",
      "Epoch [   29/   50] | d_loss: 0.5110 | g_loss: 0.2536\n",
      "Epoch [   29/   50] | d_loss: 0.5378 | g_loss: 0.2271\n",
      "Epoch [   29/   50] | d_loss: 0.4813 | g_loss: 0.2610\n",
      "Epoch [   29/   50] | d_loss: 0.5304 | g_loss: 0.2549\n",
      "Epoch [   29/   50] | d_loss: 0.5385 | g_loss: 0.2912\n",
      "Epoch [   29/   50] | d_loss: 0.5143 | g_loss: 0.2736\n",
      "Epoch [   29/   50] | d_loss: 0.4898 | g_loss: 0.2874\n",
      "Epoch [   29/   50] | d_loss: 0.5101 | g_loss: 0.2504\n",
      "Epoch [   29/   50] | d_loss: 0.5189 | g_loss: 0.2652\n",
      "Epoch [   29/   50] | d_loss: 0.4913 | g_loss: 0.2699\n",
      "Epoch [   29/   50] | d_loss: 0.4830 | g_loss: 0.2477\n",
      "Epoch [   29/   50] | d_loss: 0.5191 | g_loss: 0.2495\n",
      "Epoch [   29/   50] | d_loss: 0.5051 | g_loss: 0.2987\n",
      "Epoch [   29/   50] | d_loss: 0.4995 | g_loss: 0.2954\n",
      "Epoch [   29/   50] | d_loss: 0.4900 | g_loss: 0.2611\n",
      "Epoch [   29/   50] | d_loss: 0.4910 | g_loss: 0.2557\n",
      "Epoch [   29/   50] | d_loss: 0.5031 | g_loss: 0.2677\n",
      "Epoch [   29/   50] | d_loss: 0.5007 | g_loss: 0.2456\n",
      "Epoch [   29/   50] | d_loss: 0.4667 | g_loss: 0.2999\n",
      "Epoch [   29/   50] | d_loss: 0.5289 | g_loss: 0.2517\n",
      "Epoch [   29/   50] | d_loss: 0.5190 | g_loss: 0.2348\n",
      "Epoch [   29/   50] | d_loss: 0.5005 | g_loss: 0.2413\n",
      "Epoch [   29/   50] | d_loss: 0.4770 | g_loss: 0.2756\n",
      "Epoch [   29/   50] | d_loss: 0.5199 | g_loss: 0.2782\n",
      "Epoch [   29/   50] | d_loss: 0.4890 | g_loss: 0.2636\n",
      "Epoch [   29/   50] | d_loss: 0.5352 | g_loss: 0.2573\n",
      "Epoch [   29/   50] | d_loss: 0.5234 | g_loss: 0.2336\n",
      "Epoch [   29/   50] | d_loss: 0.4913 | g_loss: 0.2685\n",
      "Epoch [   29/   50] | d_loss: 0.5304 | g_loss: 0.2624\n",
      "Epoch [   29/   50] | d_loss: 0.5043 | g_loss: 0.2615\n",
      "Epoch [   29/   50] | d_loss: 0.4974 | g_loss: 0.2704\n",
      "Epoch [   29/   50] | d_loss: 0.5012 | g_loss: 0.2748\n",
      "Epoch [   29/   50] | d_loss: 0.5413 | g_loss: 0.2777\n",
      "Epoch [   29/   50] | d_loss: 0.5027 | g_loss: 0.2460\n",
      "Epoch [   29/   50] | d_loss: 0.5062 | g_loss: 0.2800\n",
      "Epoch [   29/   50] | d_loss: 0.5177 | g_loss: 0.2788\n",
      "Epoch [   29/   50] | d_loss: 0.5164 | g_loss: 0.2686\n",
      "Epoch [   29/   50] | d_loss: 0.4610 | g_loss: 0.2794\n",
      "Epoch [   29/   50] | d_loss: 0.5058 | g_loss: 0.2586\n",
      "Epoch [   29/   50] | d_loss: 0.4856 | g_loss: 0.2699\n",
      "Epoch [   29/   50] | d_loss: 0.4974 | g_loss: 0.2730\n",
      "Epoch [   29/   50] | d_loss: 0.4713 | g_loss: 0.2553\n",
      "Epoch [   29/   50] | d_loss: 0.5149 | g_loss: 0.2500\n",
      "Epoch [   29/   50] | d_loss: 0.5027 | g_loss: 0.2432\n",
      "Epoch [   29/   50] | d_loss: 0.5267 | g_loss: 0.2475\n",
      "Epoch [   29/   50] | d_loss: 0.5092 | g_loss: 0.2643\n",
      "Epoch [   29/   50] | d_loss: 0.4888 | g_loss: 0.2337\n",
      "Epoch [   29/   50] | d_loss: 0.4879 | g_loss: 0.2638\n",
      "Epoch [   29/   50] | d_loss: 0.4825 | g_loss: 0.2544\n",
      "Epoch [   29/   50] | d_loss: 0.5288 | g_loss: 0.2630\n",
      "Epoch [   29/   50] | d_loss: 0.5176 | g_loss: 0.2634\n",
      "Epoch [   29/   50] | d_loss: 0.4967 | g_loss: 0.2271\n",
      "Epoch [   29/   50] | d_loss: 0.5259 | g_loss: 0.2494\n",
      "Epoch [   29/   50] | d_loss: 0.5043 | g_loss: 0.2401\n",
      "Epoch [   29/   50] | d_loss: 0.4838 | g_loss: 0.2716\n",
      "Epoch [   29/   50] | d_loss: 0.5062 | g_loss: 0.2541\n",
      "Epoch [   29/   50] | d_loss: 0.5060 | g_loss: 0.2662\n",
      "Epoch [   29/   50] | d_loss: 0.5098 | g_loss: 0.2403\n",
      "Epoch [   29/   50] | d_loss: 0.5137 | g_loss: 0.2637\n",
      "Epoch [   29/   50] | d_loss: 0.4905 | g_loss: 0.2494\n",
      "Epoch [   29/   50] | d_loss: 0.5169 | g_loss: 0.2702\n",
      "Epoch [   29/   50] | d_loss: 0.5076 | g_loss: 0.2484\n",
      "Epoch [   29/   50] | d_loss: 0.5014 | g_loss: 0.2471\n",
      "Epoch [   29/   50] | d_loss: 0.4833 | g_loss: 0.2527\n",
      "Epoch [   29/   50] | d_loss: 0.5005 | g_loss: 0.2807\n",
      "Epoch [   29/   50] | d_loss: 0.5061 | g_loss: 0.2666\n",
      "Epoch [   29/   50] | d_loss: 0.4693 | g_loss: 0.2380\n",
      "Epoch [   29/   50] | d_loss: 0.4874 | g_loss: 0.2633\n",
      "Epoch [   29/   50] | d_loss: 0.4990 | g_loss: 0.2385\n",
      "Epoch [   29/   50] | d_loss: 0.5093 | g_loss: 0.2396\n",
      "Epoch [   29/   50] | d_loss: 0.4977 | g_loss: 0.2516\n",
      "Epoch [   29/   50] | d_loss: 0.5222 | g_loss: 0.2669\n",
      "Epoch [   29/   50] | d_loss: 0.4985 | g_loss: 0.2454\n",
      "Epoch [   29/   50] | d_loss: 0.4972 | g_loss: 0.2377\n",
      "Epoch [   29/   50] | d_loss: 0.5349 | g_loss: 0.2726\n",
      "Epoch [   29/   50] | d_loss: 0.4765 | g_loss: 0.2690\n",
      "Epoch [   29/   50] | d_loss: 0.4890 | g_loss: 0.2692\n",
      "Epoch [   29/   50] | d_loss: 0.4598 | g_loss: 0.2763\n",
      "Epoch [   29/   50] | d_loss: 0.4918 | g_loss: 0.2513\n",
      "Epoch [   29/   50] | d_loss: 0.5098 | g_loss: 0.2334\n",
      "Epoch [   30/   50] | d_loss: 0.5101 | g_loss: 0.2399\n",
      "Epoch [   30/   50] | d_loss: 0.4860 | g_loss: 0.2533\n",
      "Epoch [   30/   50] | d_loss: 0.5073 | g_loss: 0.2875\n",
      "Epoch [   30/   50] | d_loss: 0.4804 | g_loss: 0.2587\n",
      "Epoch [   30/   50] | d_loss: 0.5469 | g_loss: 0.2741\n",
      "Epoch [   30/   50] | d_loss: 0.4978 | g_loss: 0.2466\n",
      "Epoch [   30/   50] | d_loss: 0.4991 | g_loss: 0.2631\n",
      "Epoch [   30/   50] | d_loss: 0.4753 | g_loss: 0.2419\n",
      "Epoch [   30/   50] | d_loss: 0.5133 | g_loss: 0.2376\n",
      "Epoch [   30/   50] | d_loss: 0.4859 | g_loss: 0.2458\n",
      "Epoch [   30/   50] | d_loss: 0.4860 | g_loss: 0.2501\n",
      "Epoch [   30/   50] | d_loss: 0.4935 | g_loss: 0.2597\n",
      "Epoch [   30/   50] | d_loss: 0.5009 | g_loss: 0.2416\n",
      "Epoch [   30/   50] | d_loss: 0.4961 | g_loss: 0.2649\n",
      "Epoch [   30/   50] | d_loss: 0.5442 | g_loss: 0.2773\n",
      "Epoch [   30/   50] | d_loss: 0.4896 | g_loss: 0.2702\n",
      "Epoch [   30/   50] | d_loss: 0.5037 | g_loss: 0.2686\n",
      "Epoch [   30/   50] | d_loss: 0.5361 | g_loss: 0.2439\n",
      "Epoch [   30/   50] | d_loss: 0.4872 | g_loss: 0.2949\n",
      "Epoch [   30/   50] | d_loss: 0.5131 | g_loss: 0.2376\n",
      "Epoch [   30/   50] | d_loss: 0.4834 | g_loss: 0.2545\n",
      "Epoch [   30/   50] | d_loss: 0.5190 | g_loss: 0.2362\n",
      "Epoch [   30/   50] | d_loss: 0.4945 | g_loss: 0.2627\n",
      "Epoch [   30/   50] | d_loss: 0.4985 | g_loss: 0.2398\n",
      "Epoch [   30/   50] | d_loss: 0.5044 | g_loss: 0.2605\n",
      "Epoch [   30/   50] | d_loss: 0.5408 | g_loss: 0.2561\n",
      "Epoch [   30/   50] | d_loss: 0.4807 | g_loss: 0.2412\n",
      "Epoch [   30/   50] | d_loss: 0.5018 | g_loss: 0.2281\n",
      "Epoch [   30/   50] | d_loss: 0.5099 | g_loss: 0.2430\n",
      "Epoch [   30/   50] | d_loss: 0.4783 | g_loss: 0.2443\n",
      "Epoch [   30/   50] | d_loss: 0.4964 | g_loss: 0.2488\n",
      "Epoch [   30/   50] | d_loss: 0.4862 | g_loss: 0.2598\n",
      "Epoch [   30/   50] | d_loss: 0.5501 | g_loss: 0.2504\n",
      "Epoch [   30/   50] | d_loss: 0.5061 | g_loss: 0.2660\n",
      "Epoch [   30/   50] | d_loss: 0.5144 | g_loss: 0.2730\n",
      "Epoch [   30/   50] | d_loss: 0.5077 | g_loss: 0.2582\n",
      "Epoch [   30/   50] | d_loss: 0.5079 | g_loss: 0.2435\n",
      "Epoch [   30/   50] | d_loss: 0.5224 | g_loss: 0.2417\n",
      "Epoch [   30/   50] | d_loss: 0.4568 | g_loss: 0.2446\n",
      "Epoch [   30/   50] | d_loss: 0.5214 | g_loss: 0.2355\n",
      "Epoch [   30/   50] | d_loss: 0.4938 | g_loss: 0.2784\n",
      "Epoch [   30/   50] | d_loss: 0.4910 | g_loss: 0.2852\n",
      "Epoch [   30/   50] | d_loss: 0.4764 | g_loss: 0.2356\n",
      "Epoch [   30/   50] | d_loss: 0.5212 | g_loss: 0.2704\n",
      "Epoch [   30/   50] | d_loss: 0.4991 | g_loss: 0.2375\n",
      "Epoch [   30/   50] | d_loss: 0.4814 | g_loss: 0.2817\n",
      "Epoch [   30/   50] | d_loss: 0.5076 | g_loss: 0.2690\n",
      "Epoch [   30/   50] | d_loss: 0.4947 | g_loss: 0.2618\n",
      "Epoch [   30/   50] | d_loss: 0.5394 | g_loss: 0.2959\n",
      "Epoch [   30/   50] | d_loss: 0.4994 | g_loss: 0.2387\n",
      "Epoch [   30/   50] | d_loss: 0.4982 | g_loss: 0.2475\n",
      "Epoch [   30/   50] | d_loss: 0.5047 | g_loss: 0.2434\n",
      "Epoch [   30/   50] | d_loss: 0.5058 | g_loss: 0.2637\n",
      "Epoch [   30/   50] | d_loss: 0.4990 | g_loss: 0.2640\n",
      "Epoch [   30/   50] | d_loss: 0.5256 | g_loss: 0.2551\n",
      "Epoch [   30/   50] | d_loss: 0.5395 | g_loss: 0.2435\n",
      "Epoch [   30/   50] | d_loss: 0.5146 | g_loss: 0.2590\n",
      "Epoch [   30/   50] | d_loss: 0.4925 | g_loss: 0.2750\n",
      "Epoch [   30/   50] | d_loss: 0.5118 | g_loss: 0.2625\n",
      "Epoch [   30/   50] | d_loss: 0.4883 | g_loss: 0.2494\n",
      "Epoch [   30/   50] | d_loss: 0.4799 | g_loss: 0.2531\n",
      "Epoch [   30/   50] | d_loss: 0.4744 | g_loss: 0.2839\n",
      "Epoch [   30/   50] | d_loss: 0.4749 | g_loss: 0.2454\n",
      "Epoch [   30/   50] | d_loss: 0.4729 | g_loss: 0.2580\n",
      "Epoch [   30/   50] | d_loss: 0.4983 | g_loss: 0.2617\n",
      "Epoch [   30/   50] | d_loss: 0.4971 | g_loss: 0.2460\n",
      "Epoch [   30/   50] | d_loss: 0.4956 | g_loss: 0.2602\n",
      "Epoch [   30/   50] | d_loss: 0.5076 | g_loss: 0.2617\n",
      "Epoch [   30/   50] | d_loss: 0.5148 | g_loss: 0.2622\n",
      "Epoch [   30/   50] | d_loss: 0.4939 | g_loss: 0.2747\n",
      "Epoch [   30/   50] | d_loss: 0.4926 | g_loss: 0.2529\n",
      "Epoch [   30/   50] | d_loss: 0.4752 | g_loss: 0.2408\n",
      "Epoch [   30/   50] | d_loss: 0.5046 | g_loss: 0.2861\n",
      "Epoch [   30/   50] | d_loss: 0.5385 | g_loss: 0.2588\n",
      "Epoch [   30/   50] | d_loss: 0.4635 | g_loss: 0.2746\n",
      "Epoch [   30/   50] | d_loss: 0.4995 | g_loss: 0.2975\n",
      "Epoch [   30/   50] | d_loss: 0.4347 | g_loss: 0.2810\n",
      "Epoch [   30/   50] | d_loss: 0.4954 | g_loss: 0.2517\n",
      "Epoch [   30/   50] | d_loss: 0.5524 | g_loss: 0.2611\n",
      "Epoch [   30/   50] | d_loss: 0.5218 | g_loss: 0.2506\n",
      "Epoch [   30/   50] | d_loss: 0.5120 | g_loss: 0.2332\n",
      "Epoch [   30/   50] | d_loss: 0.5339 | g_loss: 0.2604\n",
      "Epoch [   30/   50] | d_loss: 0.5308 | g_loss: 0.2448\n",
      "Epoch [   30/   50] | d_loss: 0.4831 | g_loss: 0.2595\n",
      "Epoch [   30/   50] | d_loss: 0.5304 | g_loss: 0.2468\n",
      "Epoch [   30/   50] | d_loss: 0.5242 | g_loss: 0.2614\n",
      "Epoch [   30/   50] | d_loss: 0.4987 | g_loss: 0.2662\n",
      "Epoch [   30/   50] | d_loss: 0.4959 | g_loss: 0.2932\n",
      "Epoch [   30/   50] | d_loss: 0.4724 | g_loss: 0.2794\n",
      "Epoch [   30/   50] | d_loss: 0.5116 | g_loss: 0.2637\n",
      "Epoch [   31/   50] | d_loss: 0.4992 | g_loss: 0.2393\n",
      "Epoch [   31/   50] | d_loss: 0.5304 | g_loss: 0.2658\n",
      "Epoch [   31/   50] | d_loss: 0.4654 | g_loss: 0.2851\n",
      "Epoch [   31/   50] | d_loss: 0.5207 | g_loss: 0.2358\n",
      "Epoch [   31/   50] | d_loss: 0.4452 | g_loss: 0.2591\n",
      "Epoch [   31/   50] | d_loss: 0.5557 | g_loss: 0.2881\n",
      "Epoch [   31/   50] | d_loss: 0.5055 | g_loss: 0.2321\n",
      "Epoch [   31/   50] | d_loss: 0.5295 | g_loss: 0.2543\n",
      "Epoch [   31/   50] | d_loss: 0.4790 | g_loss: 0.2637\n",
      "Epoch [   31/   50] | d_loss: 0.5163 | g_loss: 0.2330\n",
      "Epoch [   31/   50] | d_loss: 0.5059 | g_loss: 0.2919\n",
      "Epoch [   31/   50] | d_loss: 0.5007 | g_loss: 0.2484\n",
      "Epoch [   31/   50] | d_loss: 0.5003 | g_loss: 0.2529\n",
      "Epoch [   31/   50] | d_loss: 0.4942 | g_loss: 0.2941\n",
      "Epoch [   31/   50] | d_loss: 0.5001 | g_loss: 0.2626\n",
      "Epoch [   31/   50] | d_loss: 0.5087 | g_loss: 0.2771\n",
      "Epoch [   31/   50] | d_loss: 0.4888 | g_loss: 0.2499\n",
      "Epoch [   31/   50] | d_loss: 0.5353 | g_loss: 0.2561\n",
      "Epoch [   31/   50] | d_loss: 0.4829 | g_loss: 0.2906\n",
      "Epoch [   31/   50] | d_loss: 0.5034 | g_loss: 0.2576\n",
      "Epoch [   31/   50] | d_loss: 0.5352 | g_loss: 0.2327\n",
      "Epoch [   31/   50] | d_loss: 0.5031 | g_loss: 0.2841\n",
      "Epoch [   31/   50] | d_loss: 0.5054 | g_loss: 0.2379\n",
      "Epoch [   31/   50] | d_loss: 0.5359 | g_loss: 0.2626\n",
      "Epoch [   31/   50] | d_loss: 0.4973 | g_loss: 0.2367\n",
      "Epoch [   31/   50] | d_loss: 0.4915 | g_loss: 0.2786\n",
      "Epoch [   31/   50] | d_loss: 0.5031 | g_loss: 0.2600\n",
      "Epoch [   31/   50] | d_loss: 0.5038 | g_loss: 0.2458\n",
      "Epoch [   31/   50] | d_loss: 0.4948 | g_loss: 0.2284\n",
      "Epoch [   31/   50] | d_loss: 0.4927 | g_loss: 0.2815\n",
      "Epoch [   31/   50] | d_loss: 0.5391 | g_loss: 0.2611\n",
      "Epoch [   31/   50] | d_loss: 0.5013 | g_loss: 0.2344\n",
      "Epoch [   31/   50] | d_loss: 0.5154 | g_loss: 0.2379\n",
      "Epoch [   31/   50] | d_loss: 0.5230 | g_loss: 0.2784\n",
      "Epoch [   31/   50] | d_loss: 0.5485 | g_loss: 0.2785\n",
      "Epoch [   31/   50] | d_loss: 0.4912 | g_loss: 0.2479\n",
      "Epoch [   31/   50] | d_loss: 0.4767 | g_loss: 0.2633\n",
      "Epoch [   31/   50] | d_loss: 0.4587 | g_loss: 0.2578\n",
      "Epoch [   31/   50] | d_loss: 0.4688 | g_loss: 0.2624\n",
      "Epoch [   31/   50] | d_loss: 0.5503 | g_loss: 0.2730\n",
      "Epoch [   31/   50] | d_loss: 0.4670 | g_loss: 0.2585\n",
      "Epoch [   31/   50] | d_loss: 0.4920 | g_loss: 0.2536\n",
      "Epoch [   31/   50] | d_loss: 0.4934 | g_loss: 0.2695\n",
      "Epoch [   31/   50] | d_loss: 0.5020 | g_loss: 0.2427\n",
      "Epoch [   31/   50] | d_loss: 0.4891 | g_loss: 0.2532\n",
      "Epoch [   31/   50] | d_loss: 0.5125 | g_loss: 0.2599\n",
      "Epoch [   31/   50] | d_loss: 0.5318 | g_loss: 0.2624\n",
      "Epoch [   31/   50] | d_loss: 0.4970 | g_loss: 0.2765\n",
      "Epoch [   31/   50] | d_loss: 0.4843 | g_loss: 0.2373\n",
      "Epoch [   31/   50] | d_loss: 0.4881 | g_loss: 0.2341\n",
      "Epoch [   31/   50] | d_loss: 0.5334 | g_loss: 0.2602\n",
      "Epoch [   31/   50] | d_loss: 0.4883 | g_loss: 0.2694\n",
      "Epoch [   31/   50] | d_loss: 0.4669 | g_loss: 0.2484\n",
      "Epoch [   31/   50] | d_loss: 0.4891 | g_loss: 0.2312\n",
      "Epoch [   31/   50] | d_loss: 0.4990 | g_loss: 0.2603\n",
      "Epoch [   31/   50] | d_loss: 0.4910 | g_loss: 0.2528\n",
      "Epoch [   31/   50] | d_loss: 0.4782 | g_loss: 0.2804\n",
      "Epoch [   31/   50] | d_loss: 0.5176 | g_loss: 0.2985\n",
      "Epoch [   31/   50] | d_loss: 0.5076 | g_loss: 0.2520\n",
      "Epoch [   31/   50] | d_loss: 0.5117 | g_loss: 0.2890\n",
      "Epoch [   31/   50] | d_loss: 0.5112 | g_loss: 0.2497\n",
      "Epoch [   31/   50] | d_loss: 0.5135 | g_loss: 0.2486\n",
      "Epoch [   31/   50] | d_loss: 0.4508 | g_loss: 0.2582\n",
      "Epoch [   31/   50] | d_loss: 0.5137 | g_loss: 0.2580\n",
      "Epoch [   31/   50] | d_loss: 0.5099 | g_loss: 0.2899\n",
      "Epoch [   31/   50] | d_loss: 0.4961 | g_loss: 0.2527\n",
      "Epoch [   31/   50] | d_loss: 0.4970 | g_loss: 0.2680\n",
      "Epoch [   31/   50] | d_loss: 0.5253 | g_loss: 0.2626\n",
      "Epoch [   31/   50] | d_loss: 0.4862 | g_loss: 0.2423\n",
      "Epoch [   31/   50] | d_loss: 0.4889 | g_loss: 0.2695\n",
      "Epoch [   31/   50] | d_loss: 0.4877 | g_loss: 0.2736\n",
      "Epoch [   31/   50] | d_loss: 0.5114 | g_loss: 0.2905\n",
      "Epoch [   31/   50] | d_loss: 0.4919 | g_loss: 0.2475\n",
      "Epoch [   31/   50] | d_loss: 0.5398 | g_loss: 0.2558\n",
      "Epoch [   31/   50] | d_loss: 0.5353 | g_loss: 0.2636\n",
      "Epoch [   31/   50] | d_loss: 0.4990 | g_loss: 0.2561\n",
      "Epoch [   31/   50] | d_loss: 0.4753 | g_loss: 0.2740\n",
      "Epoch [   31/   50] | d_loss: 0.4856 | g_loss: 0.2772\n",
      "Epoch [   31/   50] | d_loss: 0.5038 | g_loss: 0.2256\n",
      "Epoch [   31/   50] | d_loss: 0.4969 | g_loss: 0.2611\n",
      "Epoch [   31/   50] | d_loss: 0.5282 | g_loss: 0.2930\n",
      "Epoch [   31/   50] | d_loss: 0.4443 | g_loss: 0.2526\n",
      "Epoch [   31/   50] | d_loss: 0.4860 | g_loss: 0.2969\n",
      "Epoch [   31/   50] | d_loss: 0.5567 | g_loss: 0.2742\n",
      "Epoch [   31/   50] | d_loss: 0.4868 | g_loss: 0.2460\n",
      "Epoch [   31/   50] | d_loss: 0.5064 | g_loss: 0.2482\n",
      "Epoch [   31/   50] | d_loss: 0.5127 | g_loss: 0.2770\n",
      "Epoch [   31/   50] | d_loss: 0.4934 | g_loss: 0.2662\n",
      "Epoch [   31/   50] | d_loss: 0.4767 | g_loss: 0.2803\n",
      "Epoch [   31/   50] | d_loss: 0.4674 | g_loss: 0.2671\n",
      "Epoch [   32/   50] | d_loss: 0.5616 | g_loss: 0.2346\n",
      "Epoch [   32/   50] | d_loss: 0.4846 | g_loss: 0.2558\n",
      "Epoch [   32/   50] | d_loss: 0.4696 | g_loss: 0.2705\n",
      "Epoch [   32/   50] | d_loss: 0.5203 | g_loss: 0.2371\n",
      "Epoch [   32/   50] | d_loss: 0.5158 | g_loss: 0.2519\n",
      "Epoch [   32/   50] | d_loss: 0.4828 | g_loss: 0.2714\n",
      "Epoch [   32/   50] | d_loss: 0.5374 | g_loss: 0.2539\n",
      "Epoch [   32/   50] | d_loss: 0.4831 | g_loss: 0.2751\n",
      "Epoch [   32/   50] | d_loss: 0.5015 | g_loss: 0.2835\n",
      "Epoch [   32/   50] | d_loss: 0.5208 | g_loss: 0.2534\n",
      "Epoch [   32/   50] | d_loss: 0.5228 | g_loss: 0.2575\n",
      "Epoch [   32/   50] | d_loss: 0.4920 | g_loss: 0.2570\n",
      "Epoch [   32/   50] | d_loss: 0.5122 | g_loss: 0.2671\n",
      "Epoch [   32/   50] | d_loss: 0.4733 | g_loss: 0.2699\n",
      "Epoch [   32/   50] | d_loss: 0.5520 | g_loss: 0.2699\n",
      "Epoch [   32/   50] | d_loss: 0.4877 | g_loss: 0.2719\n",
      "Epoch [   32/   50] | d_loss: 0.5145 | g_loss: 0.2807\n",
      "Epoch [   32/   50] | d_loss: 0.4770 | g_loss: 0.2455\n",
      "Epoch [   32/   50] | d_loss: 0.5140 | g_loss: 0.2641\n",
      "Epoch [   32/   50] | d_loss: 0.5315 | g_loss: 0.2675\n",
      "Epoch [   32/   50] | d_loss: 0.5190 | g_loss: 0.2951\n",
      "Epoch [   32/   50] | d_loss: 0.4843 | g_loss: 0.2289\n",
      "Epoch [   32/   50] | d_loss: 0.4993 | g_loss: 0.2604\n",
      "Epoch [   32/   50] | d_loss: 0.4760 | g_loss: 0.2586\n",
      "Epoch [   32/   50] | d_loss: 0.5305 | g_loss: 0.2623\n",
      "Epoch [   32/   50] | d_loss: 0.4954 | g_loss: 0.2758\n",
      "Epoch [   32/   50] | d_loss: 0.5002 | g_loss: 0.2589\n",
      "Epoch [   32/   50] | d_loss: 0.5255 | g_loss: 0.2923\n",
      "Epoch [   32/   50] | d_loss: 0.5479 | g_loss: 0.2544\n",
      "Epoch [   32/   50] | d_loss: 0.4997 | g_loss: 0.2739\n",
      "Epoch [   32/   50] | d_loss: 0.5099 | g_loss: 0.2417\n",
      "Epoch [   32/   50] | d_loss: 0.4786 | g_loss: 0.2262\n",
      "Epoch [   32/   50] | d_loss: 0.5002 | g_loss: 0.2632\n",
      "Epoch [   32/   50] | d_loss: 0.4927 | g_loss: 0.2761\n",
      "Epoch [   32/   50] | d_loss: 0.5106 | g_loss: 0.2576\n",
      "Epoch [   32/   50] | d_loss: 0.5137 | g_loss: 0.2775\n",
      "Epoch [   32/   50] | d_loss: 0.5025 | g_loss: 0.2580\n",
      "Epoch [   32/   50] | d_loss: 0.5311 | g_loss: 0.3107\n",
      "Epoch [   32/   50] | d_loss: 0.5086 | g_loss: 0.2445\n",
      "Epoch [   32/   50] | d_loss: 0.4716 | g_loss: 0.2423\n",
      "Epoch [   32/   50] | d_loss: 0.4794 | g_loss: 0.2765\n",
      "Epoch [   32/   50] | d_loss: 0.5076 | g_loss: 0.2900\n",
      "Epoch [   32/   50] | d_loss: 0.4921 | g_loss: 0.2616\n",
      "Epoch [   32/   50] | d_loss: 0.4824 | g_loss: 0.2459\n",
      "Epoch [   32/   50] | d_loss: 0.5517 | g_loss: 0.2448\n",
      "Epoch [   32/   50] | d_loss: 0.5279 | g_loss: 0.2520\n",
      "Epoch [   32/   50] | d_loss: 0.5063 | g_loss: 0.2854\n",
      "Epoch [   32/   50] | d_loss: 0.4740 | g_loss: 0.2492\n",
      "Epoch [   32/   50] | d_loss: 0.4847 | g_loss: 0.2672\n",
      "Epoch [   32/   50] | d_loss: 0.4852 | g_loss: 0.2670\n",
      "Epoch [   32/   50] | d_loss: 0.4750 | g_loss: 0.2396\n",
      "Epoch [   32/   50] | d_loss: 0.5139 | g_loss: 0.2576\n",
      "Epoch [   32/   50] | d_loss: 0.4856 | g_loss: 0.2754\n",
      "Epoch [   32/   50] | d_loss: 0.4614 | g_loss: 0.2470\n",
      "Epoch [   32/   50] | d_loss: 0.5449 | g_loss: 0.2401\n",
      "Epoch [   32/   50] | d_loss: 0.5040 | g_loss: 0.2551\n",
      "Epoch [   32/   50] | d_loss: 0.4896 | g_loss: 0.2546\n",
      "Epoch [   32/   50] | d_loss: 0.4720 | g_loss: 0.2849\n",
      "Epoch [   32/   50] | d_loss: 0.5289 | g_loss: 0.2654\n",
      "Epoch [   32/   50] | d_loss: 0.5283 | g_loss: 0.2702\n",
      "Epoch [   32/   50] | d_loss: 0.5491 | g_loss: 0.2623\n",
      "Epoch [   32/   50] | d_loss: 0.4987 | g_loss: 0.2328\n",
      "Epoch [   32/   50] | d_loss: 0.4500 | g_loss: 0.2643\n",
      "Epoch [   32/   50] | d_loss: 0.5262 | g_loss: 0.2490\n",
      "Epoch [   32/   50] | d_loss: 0.4869 | g_loss: 0.2471\n",
      "Epoch [   32/   50] | d_loss: 0.5208 | g_loss: 0.2697\n",
      "Epoch [   32/   50] | d_loss: 0.5046 | g_loss: 0.2537\n",
      "Epoch [   32/   50] | d_loss: 0.4836 | g_loss: 0.2632\n",
      "Epoch [   32/   50] | d_loss: 0.5523 | g_loss: 0.2521\n",
      "Epoch [   32/   50] | d_loss: 0.4860 | g_loss: 0.2548\n",
      "Epoch [   32/   50] | d_loss: 0.4685 | g_loss: 0.2855\n",
      "Epoch [   32/   50] | d_loss: 0.4923 | g_loss: 0.2852\n",
      "Epoch [   32/   50] | d_loss: 0.5081 | g_loss: 0.2435\n",
      "Epoch [   32/   50] | d_loss: 0.5516 | g_loss: 0.2411\n",
      "Epoch [   32/   50] | d_loss: 0.4902 | g_loss: 0.2526\n",
      "Epoch [   32/   50] | d_loss: 0.5000 | g_loss: 0.2594\n",
      "Epoch [   32/   50] | d_loss: 0.5418 | g_loss: 0.2450\n",
      "Epoch [   32/   50] | d_loss: 0.5021 | g_loss: 0.2607\n",
      "Epoch [   32/   50] | d_loss: 0.5323 | g_loss: 0.2886\n",
      "Epoch [   32/   50] | d_loss: 0.5146 | g_loss: 0.2588\n",
      "Epoch [   32/   50] | d_loss: 0.5098 | g_loss: 0.2455\n",
      "Epoch [   32/   50] | d_loss: 0.5502 | g_loss: 0.2632\n",
      "Epoch [   32/   50] | d_loss: 0.4963 | g_loss: 0.2569\n",
      "Epoch [   32/   50] | d_loss: 0.4837 | g_loss: 0.2608\n",
      "Epoch [   32/   50] | d_loss: 0.4887 | g_loss: 0.2561\n",
      "Epoch [   32/   50] | d_loss: 0.5222 | g_loss: 0.2723\n",
      "Epoch [   32/   50] | d_loss: 0.5287 | g_loss: 0.2629\n",
      "Epoch [   32/   50] | d_loss: 0.5390 | g_loss: 0.2405\n",
      "Epoch [   32/   50] | d_loss: 0.5200 | g_loss: 0.2433\n",
      "Epoch [   32/   50] | d_loss: 0.5215 | g_loss: 0.2792\n",
      "Epoch [   33/   50] | d_loss: 0.5594 | g_loss: 0.2323\n",
      "Epoch [   33/   50] | d_loss: 0.5109 | g_loss: 0.2462\n",
      "Epoch [   33/   50] | d_loss: 0.5010 | g_loss: 0.2539\n",
      "Epoch [   33/   50] | d_loss: 0.4939 | g_loss: 0.2297\n",
      "Epoch [   33/   50] | d_loss: 0.4864 | g_loss: 0.2842\n",
      "Epoch [   33/   50] | d_loss: 0.5082 | g_loss: 0.2634\n",
      "Epoch [   33/   50] | d_loss: 0.4841 | g_loss: 0.2367\n",
      "Epoch [   33/   50] | d_loss: 0.4450 | g_loss: 0.2258\n",
      "Epoch [   33/   50] | d_loss: 0.5044 | g_loss: 0.2667\n",
      "Epoch [   33/   50] | d_loss: 0.5065 | g_loss: 0.2638\n",
      "Epoch [   33/   50] | d_loss: 0.5637 | g_loss: 0.2040\n",
      "Epoch [   33/   50] | d_loss: 0.4835 | g_loss: 0.2957\n",
      "Epoch [   33/   50] | d_loss: 0.5111 | g_loss: 0.2734\n",
      "Epoch [   33/   50] | d_loss: 0.5095 | g_loss: 0.2221\n",
      "Epoch [   33/   50] | d_loss: 0.5080 | g_loss: 0.2983\n",
      "Epoch [   33/   50] | d_loss: 0.4824 | g_loss: 0.2343\n",
      "Epoch [   33/   50] | d_loss: 0.4907 | g_loss: 0.2589\n",
      "Epoch [   33/   50] | d_loss: 0.5060 | g_loss: 0.2565\n",
      "Epoch [   33/   50] | d_loss: 0.5020 | g_loss: 0.2677\n",
      "Epoch [   33/   50] | d_loss: 0.5256 | g_loss: 0.2569\n",
      "Epoch [   33/   50] | d_loss: 0.5425 | g_loss: 0.2491\n",
      "Epoch [   33/   50] | d_loss: 0.4896 | g_loss: 0.2648\n",
      "Epoch [   33/   50] | d_loss: 0.5273 | g_loss: 0.2814\n",
      "Epoch [   33/   50] | d_loss: 0.4700 | g_loss: 0.2767\n",
      "Epoch [   33/   50] | d_loss: 0.5277 | g_loss: 0.2637\n",
      "Epoch [   33/   50] | d_loss: 0.4744 | g_loss: 0.2631\n",
      "Epoch [   33/   50] | d_loss: 0.5523 | g_loss: 0.2504\n",
      "Epoch [   33/   50] | d_loss: 0.5239 | g_loss: 0.2609\n",
      "Epoch [   33/   50] | d_loss: 0.5404 | g_loss: 0.2252\n",
      "Epoch [   33/   50] | d_loss: 0.5247 | g_loss: 0.2467\n",
      "Epoch [   33/   50] | d_loss: 0.5404 | g_loss: 0.2909\n",
      "Epoch [   33/   50] | d_loss: 0.5101 | g_loss: 0.2299\n",
      "Epoch [   33/   50] | d_loss: 0.5253 | g_loss: 0.2372\n",
      "Epoch [   33/   50] | d_loss: 0.4967 | g_loss: 0.2700\n",
      "Epoch [   33/   50] | d_loss: 0.4941 | g_loss: 0.2540\n",
      "Epoch [   33/   50] | d_loss: 0.5393 | g_loss: 0.2530\n",
      "Epoch [   33/   50] | d_loss: 0.4994 | g_loss: 0.2606\n",
      "Epoch [   33/   50] | d_loss: 0.4575 | g_loss: 0.2320\n",
      "Epoch [   33/   50] | d_loss: 0.5061 | g_loss: 0.2631\n",
      "Epoch [   33/   50] | d_loss: 0.4668 | g_loss: 0.2592\n",
      "Epoch [   33/   50] | d_loss: 0.4982 | g_loss: 0.2816\n",
      "Epoch [   33/   50] | d_loss: 0.5017 | g_loss: 0.2219\n",
      "Epoch [   33/   50] | d_loss: 0.5244 | g_loss: 0.2506\n",
      "Epoch [   33/   50] | d_loss: 0.5121 | g_loss: 0.2951\n",
      "Epoch [   33/   50] | d_loss: 0.5124 | g_loss: 0.2663\n",
      "Epoch [   33/   50] | d_loss: 0.5102 | g_loss: 0.2439\n",
      "Epoch [   33/   50] | d_loss: 0.5383 | g_loss: 0.2247\n",
      "Epoch [   33/   50] | d_loss: 0.5160 | g_loss: 0.2718\n",
      "Epoch [   33/   50] | d_loss: 0.4804 | g_loss: 0.2746\n",
      "Epoch [   33/   50] | d_loss: 0.5047 | g_loss: 0.2531\n",
      "Epoch [   33/   50] | d_loss: 0.5308 | g_loss: 0.2606\n",
      "Epoch [   33/   50] | d_loss: 0.5248 | g_loss: 0.2766\n",
      "Epoch [   33/   50] | d_loss: 0.5092 | g_loss: 0.2339\n",
      "Epoch [   33/   50] | d_loss: 0.4976 | g_loss: 0.2471\n",
      "Epoch [   33/   50] | d_loss: 0.5219 | g_loss: 0.2568\n",
      "Epoch [   33/   50] | d_loss: 0.4952 | g_loss: 0.2345\n",
      "Epoch [   33/   50] | d_loss: 0.4857 | g_loss: 0.2720\n",
      "Epoch [   33/   50] | d_loss: 0.5255 | g_loss: 0.2534\n",
      "Epoch [   33/   50] | d_loss: 0.5106 | g_loss: 0.2549\n",
      "Epoch [   33/   50] | d_loss: 0.5173 | g_loss: 0.2682\n",
      "Epoch [   33/   50] | d_loss: 0.5225 | g_loss: 0.2257\n",
      "Epoch [   33/   50] | d_loss: 0.4934 | g_loss: 0.2708\n",
      "Epoch [   33/   50] | d_loss: 0.5060 | g_loss: 0.2555\n",
      "Epoch [   33/   50] | d_loss: 0.5379 | g_loss: 0.2444\n",
      "Epoch [   33/   50] | d_loss: 0.4815 | g_loss: 0.2510\n",
      "Epoch [   33/   50] | d_loss: 0.4783 | g_loss: 0.2478\n",
      "Epoch [   33/   50] | d_loss: 0.4855 | g_loss: 0.2581\n",
      "Epoch [   33/   50] | d_loss: 0.4976 | g_loss: 0.2574\n",
      "Epoch [   33/   50] | d_loss: 0.4736 | g_loss: 0.2512\n",
      "Epoch [   33/   50] | d_loss: 0.4780 | g_loss: 0.2917\n",
      "Epoch [   33/   50] | d_loss: 0.4837 | g_loss: 0.2922\n",
      "Epoch [   33/   50] | d_loss: 0.5032 | g_loss: 0.2569\n",
      "Epoch [   33/   50] | d_loss: 0.4815 | g_loss: 0.2474\n",
      "Epoch [   33/   50] | d_loss: 0.4729 | g_loss: 0.2457\n",
      "Epoch [   33/   50] | d_loss: 0.5068 | g_loss: 0.2422\n",
      "Epoch [   33/   50] | d_loss: 0.4606 | g_loss: 0.2583\n",
      "Epoch [   33/   50] | d_loss: 0.5201 | g_loss: 0.2466\n",
      "Epoch [   33/   50] | d_loss: 0.5371 | g_loss: 0.2513\n",
      "Epoch [   33/   50] | d_loss: 0.5151 | g_loss: 0.2530\n",
      "Epoch [   33/   50] | d_loss: 0.5038 | g_loss: 0.2772\n",
      "Epoch [   33/   50] | d_loss: 0.4891 | g_loss: 0.2622\n",
      "Epoch [   33/   50] | d_loss: 0.5185 | g_loss: 0.2512\n",
      "Epoch [   33/   50] | d_loss: 0.5081 | g_loss: 0.2421\n",
      "Epoch [   33/   50] | d_loss: 0.4735 | g_loss: 0.2375\n",
      "Epoch [   33/   50] | d_loss: 0.5221 | g_loss: 0.2709\n",
      "Epoch [   33/   50] | d_loss: 0.5012 | g_loss: 0.2880\n",
      "Epoch [   33/   50] | d_loss: 0.4767 | g_loss: 0.2773\n",
      "Epoch [   33/   50] | d_loss: 0.4821 | g_loss: 0.2662\n",
      "Epoch [   33/   50] | d_loss: 0.4877 | g_loss: 0.2627\n",
      "Epoch [   33/   50] | d_loss: 0.4962 | g_loss: 0.2350\n",
      "Epoch [   34/   50] | d_loss: 0.5223 | g_loss: 0.2340\n",
      "Epoch [   34/   50] | d_loss: 0.5245 | g_loss: 0.2777\n",
      "Epoch [   34/   50] | d_loss: 0.4558 | g_loss: 0.2827\n",
      "Epoch [   34/   50] | d_loss: 0.5052 | g_loss: 0.2549\n",
      "Epoch [   34/   50] | d_loss: 0.4935 | g_loss: 0.2726\n",
      "Epoch [   34/   50] | d_loss: 0.4346 | g_loss: 0.2671\n",
      "Epoch [   34/   50] | d_loss: 0.4961 | g_loss: 0.2701\n",
      "Epoch [   34/   50] | d_loss: 0.4930 | g_loss: 0.2317\n",
      "Epoch [   34/   50] | d_loss: 0.4637 | g_loss: 0.2871\n",
      "Epoch [   34/   50] | d_loss: 0.4636 | g_loss: 0.2706\n",
      "Epoch [   34/   50] | d_loss: 0.5086 | g_loss: 0.2459\n",
      "Epoch [   34/   50] | d_loss: 0.5199 | g_loss: 0.2491\n",
      "Epoch [   34/   50] | d_loss: 0.5181 | g_loss: 0.2879\n",
      "Epoch [   34/   50] | d_loss: 0.5323 | g_loss: 0.2599\n",
      "Epoch [   34/   50] | d_loss: 0.4601 | g_loss: 0.2341\n",
      "Epoch [   34/   50] | d_loss: 0.5175 | g_loss: 0.2879\n",
      "Epoch [   34/   50] | d_loss: 0.4980 | g_loss: 0.2554\n",
      "Epoch [   34/   50] | d_loss: 0.5122 | g_loss: 0.2725\n",
      "Epoch [   34/   50] | d_loss: 0.5109 | g_loss: 0.2579\n",
      "Epoch [   34/   50] | d_loss: 0.4861 | g_loss: 0.2470\n",
      "Epoch [   34/   50] | d_loss: 0.4808 | g_loss: 0.2757\n",
      "Epoch [   34/   50] | d_loss: 0.5202 | g_loss: 0.2383\n",
      "Epoch [   34/   50] | d_loss: 0.4635 | g_loss: 0.2398\n",
      "Epoch [   34/   50] | d_loss: 0.4870 | g_loss: 0.2774\n",
      "Epoch [   34/   50] | d_loss: 0.4684 | g_loss: 0.2797\n",
      "Epoch [   34/   50] | d_loss: 0.5370 | g_loss: 0.2414\n",
      "Epoch [   34/   50] | d_loss: 0.4939 | g_loss: 0.2763\n",
      "Epoch [   34/   50] | d_loss: 0.5139 | g_loss: 0.2655\n",
      "Epoch [   34/   50] | d_loss: 0.5461 | g_loss: 0.2614\n",
      "Epoch [   34/   50] | d_loss: 0.4945 | g_loss: 0.2652\n",
      "Epoch [   34/   50] | d_loss: 0.5075 | g_loss: 0.2543\n",
      "Epoch [   34/   50] | d_loss: 0.5159 | g_loss: 0.2628\n",
      "Epoch [   34/   50] | d_loss: 0.4643 | g_loss: 0.2662\n",
      "Epoch [   34/   50] | d_loss: 0.5084 | g_loss: 0.3361\n",
      "Epoch [   34/   50] | d_loss: 0.4975 | g_loss: 0.2650\n",
      "Epoch [   34/   50] | d_loss: 0.5212 | g_loss: 0.2381\n",
      "Epoch [   34/   50] | d_loss: 0.4774 | g_loss: 0.2667\n",
      "Epoch [   34/   50] | d_loss: 0.4824 | g_loss: 0.2802\n",
      "Epoch [   34/   50] | d_loss: 0.5252 | g_loss: 0.2568\n",
      "Epoch [   34/   50] | d_loss: 0.4762 | g_loss: 0.2678\n",
      "Epoch [   34/   50] | d_loss: 0.5299 | g_loss: 0.2376\n",
      "Epoch [   34/   50] | d_loss: 0.5270 | g_loss: 0.2579\n",
      "Epoch [   34/   50] | d_loss: 0.4963 | g_loss: 0.2440\n",
      "Epoch [   34/   50] | d_loss: 0.5100 | g_loss: 0.2581\n",
      "Epoch [   34/   50] | d_loss: 0.4853 | g_loss: 0.2844\n",
      "Epoch [   34/   50] | d_loss: 0.4857 | g_loss: 0.2589\n",
      "Epoch [   34/   50] | d_loss: 0.4877 | g_loss: 0.2494\n",
      "Epoch [   34/   50] | d_loss: 0.5040 | g_loss: 0.2532\n",
      "Epoch [   34/   50] | d_loss: 0.5085 | g_loss: 0.2716\n",
      "Epoch [   34/   50] | d_loss: 0.5137 | g_loss: 0.2598\n",
      "Epoch [   34/   50] | d_loss: 0.5187 | g_loss: 0.2485\n",
      "Epoch [   34/   50] | d_loss: 0.5211 | g_loss: 0.2340\n",
      "Epoch [   34/   50] | d_loss: 0.5301 | g_loss: 0.2936\n",
      "Epoch [   34/   50] | d_loss: 0.5251 | g_loss: 0.2678\n",
      "Epoch [   34/   50] | d_loss: 0.4755 | g_loss: 0.2606\n",
      "Epoch [   34/   50] | d_loss: 0.4992 | g_loss: 0.2648\n",
      "Epoch [   34/   50] | d_loss: 0.4649 | g_loss: 0.2472\n",
      "Epoch [   34/   50] | d_loss: 0.5433 | g_loss: 0.2774\n",
      "Epoch [   34/   50] | d_loss: 0.5323 | g_loss: 0.2762\n",
      "Epoch [   34/   50] | d_loss: 0.5111 | g_loss: 0.2660\n",
      "Epoch [   34/   50] | d_loss: 0.4765 | g_loss: 0.2212\n",
      "Epoch [   34/   50] | d_loss: 0.4998 | g_loss: 0.2573\n",
      "Epoch [   34/   50] | d_loss: 0.4796 | g_loss: 0.2514\n",
      "Epoch [   34/   50] | d_loss: 0.5051 | g_loss: 0.2385\n",
      "Epoch [   34/   50] | d_loss: 0.4889 | g_loss: 0.2627\n",
      "Epoch [   34/   50] | d_loss: 0.4938 | g_loss: 0.2573\n",
      "Epoch [   34/   50] | d_loss: 0.4951 | g_loss: 0.2698\n",
      "Epoch [   34/   50] | d_loss: 0.5027 | g_loss: 0.2617\n",
      "Epoch [   34/   50] | d_loss: 0.5127 | g_loss: 0.2624\n",
      "Epoch [   34/   50] | d_loss: 0.5529 | g_loss: 0.2846\n",
      "Epoch [   34/   50] | d_loss: 0.4690 | g_loss: 0.2853\n",
      "Epoch [   34/   50] | d_loss: 0.4918 | g_loss: 0.2506\n",
      "Epoch [   34/   50] | d_loss: 0.4835 | g_loss: 0.2726\n",
      "Epoch [   34/   50] | d_loss: 0.5234 | g_loss: 0.2575\n",
      "Epoch [   34/   50] | d_loss: 0.5338 | g_loss: 0.2737\n",
      "Epoch [   34/   50] | d_loss: 0.4859 | g_loss: 0.2957\n",
      "Epoch [   34/   50] | d_loss: 0.4654 | g_loss: 0.2422\n",
      "Epoch [   34/   50] | d_loss: 0.5067 | g_loss: 0.2508\n",
      "Epoch [   34/   50] | d_loss: 0.5013 | g_loss: 0.2505\n",
      "Epoch [   34/   50] | d_loss: 0.4957 | g_loss: 0.2565\n",
      "Epoch [   34/   50] | d_loss: 0.5118 | g_loss: 0.2872\n",
      "Epoch [   34/   50] | d_loss: 0.4975 | g_loss: 0.2456\n",
      "Epoch [   34/   50] | d_loss: 0.5143 | g_loss: 0.2411\n",
      "Epoch [   34/   50] | d_loss: 0.5144 | g_loss: 0.2668\n",
      "Epoch [   34/   50] | d_loss: 0.4611 | g_loss: 0.2532\n",
      "Epoch [   34/   50] | d_loss: 0.5248 | g_loss: 0.2133\n",
      "Epoch [   34/   50] | d_loss: 0.4925 | g_loss: 0.2668\n",
      "Epoch [   34/   50] | d_loss: 0.4828 | g_loss: 0.2616\n",
      "Epoch [   34/   50] | d_loss: 0.4909 | g_loss: 0.2376\n",
      "Epoch [   34/   50] | d_loss: 0.4828 | g_loss: 0.2617\n",
      "Epoch [   35/   50] | d_loss: 0.4964 | g_loss: 0.2620\n",
      "Epoch [   35/   50] | d_loss: 0.5261 | g_loss: 0.2477\n",
      "Epoch [   35/   50] | d_loss: 0.5250 | g_loss: 0.2709\n",
      "Epoch [   35/   50] | d_loss: 0.5283 | g_loss: 0.2401\n",
      "Epoch [   35/   50] | d_loss: 0.5054 | g_loss: 0.2796\n",
      "Epoch [   35/   50] | d_loss: 0.5394 | g_loss: 0.2580\n",
      "Epoch [   35/   50] | d_loss: 0.5339 | g_loss: 0.2581\n",
      "Epoch [   35/   50] | d_loss: 0.5157 | g_loss: 0.2632\n",
      "Epoch [   35/   50] | d_loss: 0.5096 | g_loss: 0.2370\n",
      "Epoch [   35/   50] | d_loss: 0.4957 | g_loss: 0.2928\n",
      "Epoch [   35/   50] | d_loss: 0.4864 | g_loss: 0.2624\n",
      "Epoch [   35/   50] | d_loss: 0.5020 | g_loss: 0.2608\n",
      "Epoch [   35/   50] | d_loss: 0.5128 | g_loss: 0.2394\n",
      "Epoch [   35/   50] | d_loss: 0.4998 | g_loss: 0.2604\n",
      "Epoch [   35/   50] | d_loss: 0.5275 | g_loss: 0.2593\n",
      "Epoch [   35/   50] | d_loss: 0.5144 | g_loss: 0.2337\n",
      "Epoch [   35/   50] | d_loss: 0.4465 | g_loss: 0.2777\n",
      "Epoch [   35/   50] | d_loss: 0.5035 | g_loss: 0.2385\n",
      "Epoch [   35/   50] | d_loss: 0.5062 | g_loss: 0.2509\n",
      "Epoch [   35/   50] | d_loss: 0.5210 | g_loss: 0.2659\n",
      "Epoch [   35/   50] | d_loss: 0.5063 | g_loss: 0.3010\n",
      "Epoch [   35/   50] | d_loss: 0.5160 | g_loss: 0.2386\n",
      "Epoch [   35/   50] | d_loss: 0.4967 | g_loss: 0.2544\n",
      "Epoch [   35/   50] | d_loss: 0.4757 | g_loss: 0.2429\n",
      "Epoch [   35/   50] | d_loss: 0.5057 | g_loss: 0.2650\n",
      "Epoch [   35/   50] | d_loss: 0.5378 | g_loss: 0.2438\n",
      "Epoch [   35/   50] | d_loss: 0.5069 | g_loss: 0.2488\n",
      "Epoch [   35/   50] | d_loss: 0.5196 | g_loss: 0.2771\n",
      "Epoch [   35/   50] | d_loss: 0.5042 | g_loss: 0.2673\n",
      "Epoch [   35/   50] | d_loss: 0.4822 | g_loss: 0.2586\n",
      "Epoch [   35/   50] | d_loss: 0.4815 | g_loss: 0.2444\n",
      "Epoch [   35/   50] | d_loss: 0.5009 | g_loss: 0.2268\n",
      "Epoch [   35/   50] | d_loss: 0.4922 | g_loss: 0.2391\n",
      "Epoch [   35/   50] | d_loss: 0.4971 | g_loss: 0.2770\n",
      "Epoch [   35/   50] | d_loss: 0.5004 | g_loss: 0.2600\n",
      "Epoch [   35/   50] | d_loss: 0.5154 | g_loss: 0.2522\n",
      "Epoch [   35/   50] | d_loss: 0.5195 | g_loss: 0.2593\n",
      "Epoch [   35/   50] | d_loss: 0.5169 | g_loss: 0.2753\n",
      "Epoch [   35/   50] | d_loss: 0.4868 | g_loss: 0.2637\n",
      "Epoch [   35/   50] | d_loss: 0.4952 | g_loss: 0.2415\n",
      "Epoch [   35/   50] | d_loss: 0.4835 | g_loss: 0.2319\n",
      "Epoch [   35/   50] | d_loss: 0.4771 | g_loss: 0.2752\n",
      "Epoch [   35/   50] | d_loss: 0.4959 | g_loss: 0.2617\n",
      "Epoch [   35/   50] | d_loss: 0.5077 | g_loss: 0.2413\n",
      "Epoch [   35/   50] | d_loss: 0.4949 | g_loss: 0.2351\n",
      "Epoch [   35/   50] | d_loss: 0.5028 | g_loss: 0.2438\n",
      "Epoch [   35/   50] | d_loss: 0.5178 | g_loss: 0.2399\n",
      "Epoch [   35/   50] | d_loss: 0.4792 | g_loss: 0.2586\n",
      "Epoch [   35/   50] | d_loss: 0.5515 | g_loss: 0.2341\n",
      "Epoch [   35/   50] | d_loss: 0.5122 | g_loss: 0.2385\n",
      "Epoch [   35/   50] | d_loss: 0.4562 | g_loss: 0.2404\n",
      "Epoch [   35/   50] | d_loss: 0.5065 | g_loss: 0.2346\n",
      "Epoch [   35/   50] | d_loss: 0.5686 | g_loss: 0.2533\n",
      "Epoch [   35/   50] | d_loss: 0.5455 | g_loss: 0.2721\n",
      "Epoch [   35/   50] | d_loss: 0.5243 | g_loss: 0.2643\n",
      "Epoch [   35/   50] | d_loss: 0.4899 | g_loss: 0.2627\n",
      "Epoch [   35/   50] | d_loss: 0.4581 | g_loss: 0.2502\n",
      "Epoch [   35/   50] | d_loss: 0.5053 | g_loss: 0.2425\n",
      "Epoch [   35/   50] | d_loss: 0.5174 | g_loss: 0.2616\n",
      "Epoch [   35/   50] | d_loss: 0.4781 | g_loss: 0.2384\n",
      "Epoch [   35/   50] | d_loss: 0.4897 | g_loss: 0.2535\n",
      "Epoch [   35/   50] | d_loss: 0.5023 | g_loss: 0.2727\n",
      "Epoch [   35/   50] | d_loss: 0.4616 | g_loss: 0.2610\n",
      "Epoch [   35/   50] | d_loss: 0.5090 | g_loss: 0.2448\n",
      "Epoch [   35/   50] | d_loss: 0.5155 | g_loss: 0.2271\n",
      "Epoch [   35/   50] | d_loss: 0.4977 | g_loss: 0.2453\n",
      "Epoch [   35/   50] | d_loss: 0.5220 | g_loss: 0.2458\n",
      "Epoch [   35/   50] | d_loss: 0.4975 | g_loss: 0.2488\n",
      "Epoch [   35/   50] | d_loss: 0.5196 | g_loss: 0.2788\n",
      "Epoch [   35/   50] | d_loss: 0.5042 | g_loss: 0.2642\n",
      "Epoch [   35/   50] | d_loss: 0.5292 | g_loss: 0.2563\n",
      "Epoch [   35/   50] | d_loss: 0.5086 | g_loss: 0.2432\n",
      "Epoch [   35/   50] | d_loss: 0.5262 | g_loss: 0.2496\n",
      "Epoch [   35/   50] | d_loss: 0.5054 | g_loss: 0.2897\n",
      "Epoch [   35/   50] | d_loss: 0.4868 | g_loss: 0.2604\n",
      "Epoch [   35/   50] | d_loss: 0.5274 | g_loss: 0.2798\n",
      "Epoch [   35/   50] | d_loss: 0.4946 | g_loss: 0.2489\n",
      "Epoch [   35/   50] | d_loss: 0.5015 | g_loss: 0.2425\n",
      "Epoch [   35/   50] | d_loss: 0.5054 | g_loss: 0.2382\n",
      "Epoch [   35/   50] | d_loss: 0.5084 | g_loss: 0.2328\n",
      "Epoch [   35/   50] | d_loss: 0.4940 | g_loss: 0.2521\n",
      "Epoch [   35/   50] | d_loss: 0.4899 | g_loss: 0.2745\n",
      "Epoch [   35/   50] | d_loss: 0.5073 | g_loss: 0.2747\n",
      "Epoch [   35/   50] | d_loss: 0.4711 | g_loss: 0.2613\n",
      "Epoch [   35/   50] | d_loss: 0.4969 | g_loss: 0.2579\n",
      "Epoch [   35/   50] | d_loss: 0.5192 | g_loss: 0.2671\n",
      "Epoch [   35/   50] | d_loss: 0.5184 | g_loss: 0.2715\n",
      "Epoch [   35/   50] | d_loss: 0.5191 | g_loss: 0.2553\n",
      "Epoch [   35/   50] | d_loss: 0.5073 | g_loss: 0.2454\n",
      "Epoch [   35/   50] | d_loss: 0.4789 | g_loss: 0.2666\n",
      "Epoch [   36/   50] | d_loss: 0.5074 | g_loss: 0.2834\n",
      "Epoch [   36/   50] | d_loss: 0.4556 | g_loss: 0.2615\n",
      "Epoch [   36/   50] | d_loss: 0.5260 | g_loss: 0.2686\n",
      "Epoch [   36/   50] | d_loss: 0.4724 | g_loss: 0.2608\n",
      "Epoch [   36/   50] | d_loss: 0.5258 | g_loss: 0.2360\n",
      "Epoch [   36/   50] | d_loss: 0.4815 | g_loss: 0.2951\n",
      "Epoch [   36/   50] | d_loss: 0.5055 | g_loss: 0.2550\n",
      "Epoch [   36/   50] | d_loss: 0.4849 | g_loss: 0.2655\n",
      "Epoch [   36/   50] | d_loss: 0.5016 | g_loss: 0.2175\n",
      "Epoch [   36/   50] | d_loss: 0.5164 | g_loss: 0.2612\n",
      "Epoch [   36/   50] | d_loss: 0.4840 | g_loss: 0.2441\n",
      "Epoch [   36/   50] | d_loss: 0.5058 | g_loss: 0.2510\n",
      "Epoch [   36/   50] | d_loss: 0.4613 | g_loss: 0.2475\n",
      "Epoch [   36/   50] | d_loss: 0.5046 | g_loss: 0.2762\n",
      "Epoch [   36/   50] | d_loss: 0.4927 | g_loss: 0.2717\n",
      "Epoch [   36/   50] | d_loss: 0.4699 | g_loss: 0.2631\n",
      "Epoch [   36/   50] | d_loss: 0.4911 | g_loss: 0.2585\n",
      "Epoch [   36/   50] | d_loss: 0.5190 | g_loss: 0.2588\n",
      "Epoch [   36/   50] | d_loss: 0.5001 | g_loss: 0.2852\n",
      "Epoch [   36/   50] | d_loss: 0.4605 | g_loss: 0.2813\n",
      "Epoch [   36/   50] | d_loss: 0.4985 | g_loss: 0.2397\n",
      "Epoch [   36/   50] | d_loss: 0.5363 | g_loss: 0.2614\n",
      "Epoch [   36/   50] | d_loss: 0.5065 | g_loss: 0.2578\n",
      "Epoch [   36/   50] | d_loss: 0.5060 | g_loss: 0.2656\n",
      "Epoch [   36/   50] | d_loss: 0.4982 | g_loss: 0.2737\n",
      "Epoch [   36/   50] | d_loss: 0.5016 | g_loss: 0.2398\n",
      "Epoch [   36/   50] | d_loss: 0.5142 | g_loss: 0.2501\n",
      "Epoch [   36/   50] | d_loss: 0.4879 | g_loss: 0.2621\n",
      "Epoch [   36/   50] | d_loss: 0.4910 | g_loss: 0.2354\n",
      "Epoch [   36/   50] | d_loss: 0.5115 | g_loss: 0.2679\n",
      "Epoch [   36/   50] | d_loss: 0.4903 | g_loss: 0.2648\n",
      "Epoch [   36/   50] | d_loss: 0.4792 | g_loss: 0.2551\n",
      "Epoch [   36/   50] | d_loss: 0.5289 | g_loss: 0.2730\n",
      "Epoch [   36/   50] | d_loss: 0.5141 | g_loss: 0.2205\n",
      "Epoch [   36/   50] | d_loss: 0.5262 | g_loss: 0.2665\n",
      "Epoch [   36/   50] | d_loss: 0.5025 | g_loss: 0.2630\n",
      "Epoch [   36/   50] | d_loss: 0.5468 | g_loss: 0.2784\n",
      "Epoch [   36/   50] | d_loss: 0.5020 | g_loss: 0.2680\n",
      "Epoch [   36/   50] | d_loss: 0.5298 | g_loss: 0.2210\n",
      "Epoch [   36/   50] | d_loss: 0.5040 | g_loss: 0.2544\n",
      "Epoch [   36/   50] | d_loss: 0.5132 | g_loss: 0.2668\n",
      "Epoch [   36/   50] | d_loss: 0.5465 | g_loss: 0.2468\n",
      "Epoch [   36/   50] | d_loss: 0.4893 | g_loss: 0.2731\n",
      "Epoch [   36/   50] | d_loss: 0.4998 | g_loss: 0.2618\n",
      "Epoch [   36/   50] | d_loss: 0.4804 | g_loss: 0.2554\n",
      "Epoch [   36/   50] | d_loss: 0.5014 | g_loss: 0.2503\n",
      "Epoch [   36/   50] | d_loss: 0.5139 | g_loss: 0.2319\n",
      "Epoch [   36/   50] | d_loss: 0.5140 | g_loss: 0.2426\n",
      "Epoch [   36/   50] | d_loss: 0.5089 | g_loss: 0.2705\n",
      "Epoch [   36/   50] | d_loss: 0.4996 | g_loss: 0.2652\n",
      "Epoch [   36/   50] | d_loss: 0.5029 | g_loss: 0.2569\n",
      "Epoch [   36/   50] | d_loss: 0.5239 | g_loss: 0.2410\n",
      "Epoch [   36/   50] | d_loss: 0.4951 | g_loss: 0.2391\n",
      "Epoch [   36/   50] | d_loss: 0.5458 | g_loss: 0.2457\n",
      "Epoch [   36/   50] | d_loss: 0.4843 | g_loss: 0.2671\n",
      "Epoch [   36/   50] | d_loss: 0.5330 | g_loss: 0.2509\n",
      "Epoch [   36/   50] | d_loss: 0.4756 | g_loss: 0.2695\n",
      "Epoch [   36/   50] | d_loss: 0.5077 | g_loss: 0.2573\n",
      "Epoch [   36/   50] | d_loss: 0.4886 | g_loss: 0.2392\n",
      "Epoch [   36/   50] | d_loss: 0.5408 | g_loss: 0.2204\n",
      "Epoch [   36/   50] | d_loss: 0.5003 | g_loss: 0.2586\n",
      "Epoch [   36/   50] | d_loss: 0.4804 | g_loss: 0.2651\n",
      "Epoch [   36/   50] | d_loss: 0.4979 | g_loss: 0.2309\n",
      "Epoch [   36/   50] | d_loss: 0.5478 | g_loss: 0.2670\n",
      "Epoch [   36/   50] | d_loss: 0.4809 | g_loss: 0.2427\n",
      "Epoch [   36/   50] | d_loss: 0.5068 | g_loss: 0.2612\n",
      "Epoch [   36/   50] | d_loss: 0.4958 | g_loss: 0.2496\n",
      "Epoch [   36/   50] | d_loss: 0.5258 | g_loss: 0.2635\n",
      "Epoch [   36/   50] | d_loss: 0.4818 | g_loss: 0.2940\n",
      "Epoch [   36/   50] | d_loss: 0.5239 | g_loss: 0.2469\n",
      "Epoch [   36/   50] | d_loss: 0.5147 | g_loss: 0.2516\n",
      "Epoch [   36/   50] | d_loss: 0.4867 | g_loss: 0.2540\n",
      "Epoch [   36/   50] | d_loss: 0.4757 | g_loss: 0.2271\n",
      "Epoch [   36/   50] | d_loss: 0.5108 | g_loss: 0.2737\n",
      "Epoch [   36/   50] | d_loss: 0.4828 | g_loss: 0.2845\n",
      "Epoch [   36/   50] | d_loss: 0.4775 | g_loss: 0.2578\n",
      "Epoch [   36/   50] | d_loss: 0.5441 | g_loss: 0.2688\n",
      "Epoch [   36/   50] | d_loss: 0.5290 | g_loss: 0.2758\n",
      "Epoch [   36/   50] | d_loss: 0.5275 | g_loss: 0.2414\n",
      "Epoch [   36/   50] | d_loss: 0.5051 | g_loss: 0.2701\n",
      "Epoch [   36/   50] | d_loss: 0.5192 | g_loss: 0.2681\n",
      "Epoch [   36/   50] | d_loss: 0.5149 | g_loss: 0.2416\n",
      "Epoch [   36/   50] | d_loss: 0.5267 | g_loss: 0.2693\n",
      "Epoch [   36/   50] | d_loss: 0.4974 | g_loss: 0.3255\n",
      "Epoch [   36/   50] | d_loss: 0.5093 | g_loss: 0.2656\n",
      "Epoch [   36/   50] | d_loss: 0.4833 | g_loss: 0.2762\n",
      "Epoch [   36/   50] | d_loss: 0.4858 | g_loss: 0.3026\n",
      "Epoch [   36/   50] | d_loss: 0.5049 | g_loss: 0.2427\n",
      "Epoch [   36/   50] | d_loss: 0.4998 | g_loss: 0.2636\n",
      "Epoch [   36/   50] | d_loss: 0.4599 | g_loss: 0.2866\n",
      "Epoch [   37/   50] | d_loss: 0.4898 | g_loss: 0.2548\n",
      "Epoch [   37/   50] | d_loss: 0.5280 | g_loss: 0.2798\n",
      "Epoch [   37/   50] | d_loss: 0.5425 | g_loss: 0.2877\n",
      "Epoch [   37/   50] | d_loss: 0.4827 | g_loss: 0.3006\n",
      "Epoch [   37/   50] | d_loss: 0.5218 | g_loss: 0.2736\n",
      "Epoch [   37/   50] | d_loss: 0.4855 | g_loss: 0.2623\n",
      "Epoch [   37/   50] | d_loss: 0.4426 | g_loss: 0.2522\n",
      "Epoch [   37/   50] | d_loss: 0.4860 | g_loss: 0.2507\n",
      "Epoch [   37/   50] | d_loss: 0.4923 | g_loss: 0.2511\n",
      "Epoch [   37/   50] | d_loss: 0.4987 | g_loss: 0.2639\n",
      "Epoch [   37/   50] | d_loss: 0.5406 | g_loss: 0.2679\n",
      "Epoch [   37/   50] | d_loss: 0.4708 | g_loss: 0.2385\n",
      "Epoch [   37/   50] | d_loss: 0.4874 | g_loss: 0.2509\n",
      "Epoch [   37/   50] | d_loss: 0.5159 | g_loss: 0.2599\n",
      "Epoch [   37/   50] | d_loss: 0.5646 | g_loss: 0.2408\n",
      "Epoch [   37/   50] | d_loss: 0.4900 | g_loss: 0.2706\n",
      "Epoch [   37/   50] | d_loss: 0.4843 | g_loss: 0.2771\n",
      "Epoch [   37/   50] | d_loss: 0.5201 | g_loss: 0.2328\n",
      "Epoch [   37/   50] | d_loss: 0.5036 | g_loss: 0.2734\n",
      "Epoch [   37/   50] | d_loss: 0.5252 | g_loss: 0.2856\n",
      "Epoch [   37/   50] | d_loss: 0.4966 | g_loss: 0.2734\n",
      "Epoch [   37/   50] | d_loss: 0.4929 | g_loss: 0.2738\n",
      "Epoch [   37/   50] | d_loss: 0.5329 | g_loss: 0.2500\n",
      "Epoch [   37/   50] | d_loss: 0.5186 | g_loss: 0.2822\n",
      "Epoch [   37/   50] | d_loss: 0.5068 | g_loss: 0.2283\n",
      "Epoch [   37/   50] | d_loss: 0.5093 | g_loss: 0.2700\n",
      "Epoch [   37/   50] | d_loss: 0.4788 | g_loss: 0.2685\n",
      "Epoch [   37/   50] | d_loss: 0.5096 | g_loss: 0.2707\n",
      "Epoch [   37/   50] | d_loss: 0.4875 | g_loss: 0.2822\n",
      "Epoch [   37/   50] | d_loss: 0.5174 | g_loss: 0.2147\n",
      "Epoch [   37/   50] | d_loss: 0.4780 | g_loss: 0.2637\n",
      "Epoch [   37/   50] | d_loss: 0.4909 | g_loss: 0.2625\n",
      "Epoch [   37/   50] | d_loss: 0.4713 | g_loss: 0.2541\n",
      "Epoch [   37/   50] | d_loss: 0.4987 | g_loss: 0.2640\n",
      "Epoch [   37/   50] | d_loss: 0.4980 | g_loss: 0.2482\n",
      "Epoch [   37/   50] | d_loss: 0.5211 | g_loss: 0.2377\n",
      "Epoch [   37/   50] | d_loss: 0.5048 | g_loss: 0.2730\n",
      "Epoch [   37/   50] | d_loss: 0.5208 | g_loss: 0.2707\n",
      "Epoch [   37/   50] | d_loss: 0.5073 | g_loss: 0.2458\n",
      "Epoch [   37/   50] | d_loss: 0.5018 | g_loss: 0.2708\n",
      "Epoch [   37/   50] | d_loss: 0.5092 | g_loss: 0.2832\n",
      "Epoch [   37/   50] | d_loss: 0.5126 | g_loss: 0.2533\n",
      "Epoch [   37/   50] | d_loss: 0.4978 | g_loss: 0.2742\n",
      "Epoch [   37/   50] | d_loss: 0.5305 | g_loss: 0.2478\n",
      "Epoch [   37/   50] | d_loss: 0.5186 | g_loss: 0.2618\n",
      "Epoch [   37/   50] | d_loss: 0.5490 | g_loss: 0.2634\n",
      "Epoch [   37/   50] | d_loss: 0.4798 | g_loss: 0.2621\n",
      "Epoch [   37/   50] | d_loss: 0.4674 | g_loss: 0.2728\n",
      "Epoch [   37/   50] | d_loss: 0.5350 | g_loss: 0.2756\n",
      "Epoch [   37/   50] | d_loss: 0.4862 | g_loss: 0.2406\n",
      "Epoch [   37/   50] | d_loss: 0.5047 | g_loss: 0.2508\n",
      "Epoch [   37/   50] | d_loss: 0.4705 | g_loss: 0.2467\n",
      "Epoch [   37/   50] | d_loss: 0.4987 | g_loss: 0.2632\n",
      "Epoch [   37/   50] | d_loss: 0.5205 | g_loss: 0.2409\n",
      "Epoch [   37/   50] | d_loss: 0.5099 | g_loss: 0.2542\n",
      "Epoch [   37/   50] | d_loss: 0.5389 | g_loss: 0.2833\n",
      "Epoch [   37/   50] | d_loss: 0.4950 | g_loss: 0.2755\n",
      "Epoch [   37/   50] | d_loss: 0.4857 | g_loss: 0.2585\n",
      "Epoch [   37/   50] | d_loss: 0.5230 | g_loss: 0.2751\n",
      "Epoch [   37/   50] | d_loss: 0.4637 | g_loss: 0.2698\n",
      "Epoch [   37/   50] | d_loss: 0.4920 | g_loss: 0.2288\n",
      "Epoch [   37/   50] | d_loss: 0.4943 | g_loss: 0.2392\n",
      "Epoch [   37/   50] | d_loss: 0.5064 | g_loss: 0.2413\n",
      "Epoch [   37/   50] | d_loss: 0.5068 | g_loss: 0.2602\n",
      "Epoch [   37/   50] | d_loss: 0.5280 | g_loss: 0.2506\n",
      "Epoch [   37/   50] | d_loss: 0.5503 | g_loss: 0.2489\n",
      "Epoch [   37/   50] | d_loss: 0.5140 | g_loss: 0.2664\n",
      "Epoch [   37/   50] | d_loss: 0.4753 | g_loss: 0.2556\n",
      "Epoch [   37/   50] | d_loss: 0.4835 | g_loss: 0.2698\n",
      "Epoch [   37/   50] | d_loss: 0.4976 | g_loss: 0.2577\n",
      "Epoch [   37/   50] | d_loss: 0.5168 | g_loss: 0.2399\n",
      "Epoch [   37/   50] | d_loss: 0.4897 | g_loss: 0.2881\n",
      "Epoch [   37/   50] | d_loss: 0.5024 | g_loss: 0.2411\n",
      "Epoch [   37/   50] | d_loss: 0.4547 | g_loss: 0.2680\n",
      "Epoch [   37/   50] | d_loss: 0.5062 | g_loss: 0.2416\n",
      "Epoch [   37/   50] | d_loss: 0.5476 | g_loss: 0.2677\n",
      "Epoch [   37/   50] | d_loss: 0.5604 | g_loss: 0.2707\n",
      "Epoch [   37/   50] | d_loss: 0.5503 | g_loss: 0.2516\n",
      "Epoch [   37/   50] | d_loss: 0.5126 | g_loss: 0.2548\n",
      "Epoch [   37/   50] | d_loss: 0.5012 | g_loss: 0.2533\n",
      "Epoch [   37/   50] | d_loss: 0.4862 | g_loss: 0.2610\n",
      "Epoch [   37/   50] | d_loss: 0.5062 | g_loss: 0.2651\n",
      "Epoch [   37/   50] | d_loss: 0.4617 | g_loss: 0.2448\n",
      "Epoch [   37/   50] | d_loss: 0.5013 | g_loss: 0.2788\n",
      "Epoch [   37/   50] | d_loss: 0.4713 | g_loss: 0.2681\n",
      "Epoch [   37/   50] | d_loss: 0.4875 | g_loss: 0.2702\n",
      "Epoch [   37/   50] | d_loss: 0.4942 | g_loss: 0.2607\n",
      "Epoch [   37/   50] | d_loss: 0.5290 | g_loss: 0.2499\n",
      "Epoch [   37/   50] | d_loss: 0.5117 | g_loss: 0.2697\n",
      "Epoch [   37/   50] | d_loss: 0.5087 | g_loss: 0.2597\n",
      "Epoch [   38/   50] | d_loss: 0.4880 | g_loss: 0.2905\n",
      "Epoch [   38/   50] | d_loss: 0.5051 | g_loss: 0.2604\n",
      "Epoch [   38/   50] | d_loss: 0.5001 | g_loss: 0.2865\n",
      "Epoch [   38/   50] | d_loss: 0.5291 | g_loss: 0.2660\n",
      "Epoch [   38/   50] | d_loss: 0.5178 | g_loss: 0.2472\n",
      "Epoch [   38/   50] | d_loss: 0.4896 | g_loss: 0.2353\n",
      "Epoch [   38/   50] | d_loss: 0.5331 | g_loss: 0.2238\n",
      "Epoch [   38/   50] | d_loss: 0.5024 | g_loss: 0.2856\n",
      "Epoch [   38/   50] | d_loss: 0.5449 | g_loss: 0.2795\n",
      "Epoch [   38/   50] | d_loss: 0.5084 | g_loss: 0.2519\n",
      "Epoch [   38/   50] | d_loss: 0.5291 | g_loss: 0.2725\n",
      "Epoch [   38/   50] | d_loss: 0.4985 | g_loss: 0.2389\n",
      "Epoch [   38/   50] | d_loss: 0.4786 | g_loss: 0.2852\n",
      "Epoch [   38/   50] | d_loss: 0.5126 | g_loss: 0.2620\n",
      "Epoch [   38/   50] | d_loss: 0.5219 | g_loss: 0.2926\n",
      "Epoch [   38/   50] | d_loss: 0.5221 | g_loss: 0.2737\n",
      "Epoch [   38/   50] | d_loss: 0.4918 | g_loss: 0.3021\n",
      "Epoch [   38/   50] | d_loss: 0.5234 | g_loss: 0.2786\n",
      "Epoch [   38/   50] | d_loss: 0.5198 | g_loss: 0.2826\n",
      "Epoch [   38/   50] | d_loss: 0.4860 | g_loss: 0.2434\n",
      "Epoch [   38/   50] | d_loss: 0.5091 | g_loss: 0.2980\n",
      "Epoch [   38/   50] | d_loss: 0.5425 | g_loss: 0.2934\n",
      "Epoch [   38/   50] | d_loss: 0.4984 | g_loss: 0.2472\n",
      "Epoch [   38/   50] | d_loss: 0.5070 | g_loss: 0.2601\n",
      "Epoch [   38/   50] | d_loss: 0.5071 | g_loss: 0.2467\n",
      "Epoch [   38/   50] | d_loss: 0.5128 | g_loss: 0.2555\n",
      "Epoch [   38/   50] | d_loss: 0.5044 | g_loss: 0.2429\n",
      "Epoch [   38/   50] | d_loss: 0.5231 | g_loss: 0.2383\n",
      "Epoch [   38/   50] | d_loss: 0.4770 | g_loss: 0.2538\n",
      "Epoch [   38/   50] | d_loss: 0.4798 | g_loss: 0.2542\n",
      "Epoch [   38/   50] | d_loss: 0.4916 | g_loss: 0.2528\n",
      "Epoch [   38/   50] | d_loss: 0.4804 | g_loss: 0.2401\n",
      "Epoch [   38/   50] | d_loss: 0.5236 | g_loss: 0.2427\n",
      "Epoch [   38/   50] | d_loss: 0.4735 | g_loss: 0.2695\n",
      "Epoch [   38/   50] | d_loss: 0.5281 | g_loss: 0.2593\n",
      "Epoch [   38/   50] | d_loss: 0.4844 | g_loss: 0.2437\n",
      "Epoch [   38/   50] | d_loss: 0.5126 | g_loss: 0.2406\n",
      "Epoch [   38/   50] | d_loss: 0.5048 | g_loss: 0.2527\n",
      "Epoch [   38/   50] | d_loss: 0.4982 | g_loss: 0.2723\n",
      "Epoch [   38/   50] | d_loss: 0.4718 | g_loss: 0.2678\n",
      "Epoch [   38/   50] | d_loss: 0.5242 | g_loss: 0.2775\n",
      "Epoch [   38/   50] | d_loss: 0.5152 | g_loss: 0.2681\n",
      "Epoch [   38/   50] | d_loss: 0.4758 | g_loss: 0.2375\n",
      "Epoch [   38/   50] | d_loss: 0.5325 | g_loss: 0.2649\n",
      "Epoch [   38/   50] | d_loss: 0.4952 | g_loss: 0.2524\n",
      "Epoch [   38/   50] | d_loss: 0.4760 | g_loss: 0.2462\n",
      "Epoch [   38/   50] | d_loss: 0.5080 | g_loss: 0.2646\n",
      "Epoch [   38/   50] | d_loss: 0.5108 | g_loss: 0.2530\n",
      "Epoch [   38/   50] | d_loss: 0.4897 | g_loss: 0.2872\n",
      "Epoch [   38/   50] | d_loss: 0.5113 | g_loss: 0.2382\n",
      "Epoch [   38/   50] | d_loss: 0.5233 | g_loss: 0.2456\n",
      "Epoch [   38/   50] | d_loss: 0.5021 | g_loss: 0.2769\n",
      "Epoch [   38/   50] | d_loss: 0.4911 | g_loss: 0.2747\n",
      "Epoch [   38/   50] | d_loss: 0.5006 | g_loss: 0.2362\n",
      "Epoch [   38/   50] | d_loss: 0.4921 | g_loss: 0.2522\n",
      "Epoch [   38/   50] | d_loss: 0.5323 | g_loss: 0.2410\n",
      "Epoch [   38/   50] | d_loss: 0.5008 | g_loss: 0.2646\n",
      "Epoch [   38/   50] | d_loss: 0.5249 | g_loss: 0.2561\n",
      "Epoch [   38/   50] | d_loss: 0.5035 | g_loss: 0.2567\n",
      "Epoch [   38/   50] | d_loss: 0.5169 | g_loss: 0.2670\n",
      "Epoch [   38/   50] | d_loss: 0.5119 | g_loss: 0.2323\n",
      "Epoch [   38/   50] | d_loss: 0.5106 | g_loss: 0.2785\n",
      "Epoch [   38/   50] | d_loss: 0.5069 | g_loss: 0.2666\n",
      "Epoch [   38/   50] | d_loss: 0.5043 | g_loss: 0.2558\n",
      "Epoch [   38/   50] | d_loss: 0.4933 | g_loss: 0.2425\n",
      "Epoch [   38/   50] | d_loss: 0.5289 | g_loss: 0.2404\n",
      "Epoch [   38/   50] | d_loss: 0.5238 | g_loss: 0.2713\n",
      "Epoch [   38/   50] | d_loss: 0.5058 | g_loss: 0.2616\n",
      "Epoch [   38/   50] | d_loss: 0.4592 | g_loss: 0.2455\n",
      "Epoch [   38/   50] | d_loss: 0.5143 | g_loss: 0.2860\n",
      "Epoch [   38/   50] | d_loss: 0.5224 | g_loss: 0.2573\n",
      "Epoch [   38/   50] | d_loss: 0.4600 | g_loss: 0.2685\n",
      "Epoch [   38/   50] | d_loss: 0.5324 | g_loss: 0.2708\n",
      "Epoch [   38/   50] | d_loss: 0.5246 | g_loss: 0.2478\n",
      "Epoch [   38/   50] | d_loss: 0.4939 | g_loss: 0.2428\n",
      "Epoch [   38/   50] | d_loss: 0.4752 | g_loss: 0.2372\n",
      "Epoch [   38/   50] | d_loss: 0.4661 | g_loss: 0.2609\n",
      "Epoch [   38/   50] | d_loss: 0.5394 | g_loss: 0.2577\n",
      "Epoch [   38/   50] | d_loss: 0.4819 | g_loss: 0.2342\n",
      "Epoch [   38/   50] | d_loss: 0.5313 | g_loss: 0.2843\n",
      "Epoch [   38/   50] | d_loss: 0.4739 | g_loss: 0.2559\n",
      "Epoch [   38/   50] | d_loss: 0.5095 | g_loss: 0.2670\n",
      "Epoch [   38/   50] | d_loss: 0.4936 | g_loss: 0.2529\n",
      "Epoch [   38/   50] | d_loss: 0.5074 | g_loss: 0.2555\n",
      "Epoch [   38/   50] | d_loss: 0.4893 | g_loss: 0.2432\n",
      "Epoch [   38/   50] | d_loss: 0.4880 | g_loss: 0.2518\n",
      "Epoch [   38/   50] | d_loss: 0.4958 | g_loss: 0.2654\n",
      "Epoch [   38/   50] | d_loss: 0.5395 | g_loss: 0.2708\n",
      "Epoch [   38/   50] | d_loss: 0.4670 | g_loss: 0.2613\n",
      "Epoch [   38/   50] | d_loss: 0.4940 | g_loss: 0.2905\n",
      "Epoch [   39/   50] | d_loss: 0.4747 | g_loss: 0.2738\n",
      "Epoch [   39/   50] | d_loss: 0.5250 | g_loss: 0.2575\n",
      "Epoch [   39/   50] | d_loss: 0.5137 | g_loss: 0.2652\n",
      "Epoch [   39/   50] | d_loss: 0.5188 | g_loss: 0.2508\n",
      "Epoch [   39/   50] | d_loss: 0.4873 | g_loss: 0.2575\n",
      "Epoch [   39/   50] | d_loss: 0.4949 | g_loss: 0.2636\n",
      "Epoch [   39/   50] | d_loss: 0.5406 | g_loss: 0.2523\n",
      "Epoch [   39/   50] | d_loss: 0.5045 | g_loss: 0.2609\n",
      "Epoch [   39/   50] | d_loss: 0.5161 | g_loss: 0.2844\n",
      "Epoch [   39/   50] | d_loss: 0.5018 | g_loss: 0.2726\n",
      "Epoch [   39/   50] | d_loss: 0.4785 | g_loss: 0.2712\n",
      "Epoch [   39/   50] | d_loss: 0.5047 | g_loss: 0.2489\n",
      "Epoch [   39/   50] | d_loss: 0.4440 | g_loss: 0.2657\n",
      "Epoch [   39/   50] | d_loss: 0.5044 | g_loss: 0.2531\n",
      "Epoch [   39/   50] | d_loss: 0.5263 | g_loss: 0.2557\n",
      "Epoch [   39/   50] | d_loss: 0.4982 | g_loss: 0.2660\n",
      "Epoch [   39/   50] | d_loss: 0.5258 | g_loss: 0.2658\n",
      "Epoch [   39/   50] | d_loss: 0.5109 | g_loss: 0.2626\n",
      "Epoch [   39/   50] | d_loss: 0.4851 | g_loss: 0.2504\n",
      "Epoch [   39/   50] | d_loss: 0.4929 | g_loss: 0.2389\n",
      "Epoch [   39/   50] | d_loss: 0.5106 | g_loss: 0.2633\n",
      "Epoch [   39/   50] | d_loss: 0.5217 | g_loss: 0.2571\n",
      "Epoch [   39/   50] | d_loss: 0.4941 | g_loss: 0.2389\n",
      "Epoch [   39/   50] | d_loss: 0.5143 | g_loss: 0.2656\n",
      "Epoch [   39/   50] | d_loss: 0.4889 | g_loss: 0.2655\n",
      "Epoch [   39/   50] | d_loss: 0.4729 | g_loss: 0.2563\n",
      "Epoch [   39/   50] | d_loss: 0.4844 | g_loss: 0.2949\n",
      "Epoch [   39/   50] | d_loss: 0.4949 | g_loss: 0.2354\n",
      "Epoch [   39/   50] | d_loss: 0.5171 | g_loss: 0.2800\n",
      "Epoch [   39/   50] | d_loss: 0.4879 | g_loss: 0.2609\n",
      "Epoch [   39/   50] | d_loss: 0.5057 | g_loss: 0.2768\n",
      "Epoch [   39/   50] | d_loss: 0.4995 | g_loss: 0.2800\n",
      "Epoch [   39/   50] | d_loss: 0.4947 | g_loss: 0.2338\n",
      "Epoch [   39/   50] | d_loss: 0.4825 | g_loss: 0.2173\n",
      "Epoch [   39/   50] | d_loss: 0.4612 | g_loss: 0.2659\n",
      "Epoch [   39/   50] | d_loss: 0.4757 | g_loss: 0.2657\n",
      "Epoch [   39/   50] | d_loss: 0.4625 | g_loss: 0.2813\n",
      "Epoch [   39/   50] | d_loss: 0.4772 | g_loss: 0.2459\n",
      "Epoch [   39/   50] | d_loss: 0.5414 | g_loss: 0.2416\n",
      "Epoch [   39/   50] | d_loss: 0.4642 | g_loss: 0.2329\n",
      "Epoch [   39/   50] | d_loss: 0.5248 | g_loss: 0.2681\n",
      "Epoch [   39/   50] | d_loss: 0.5106 | g_loss: 0.2636\n",
      "Epoch [   39/   50] | d_loss: 0.5012 | g_loss: 0.2578\n",
      "Epoch [   39/   50] | d_loss: 0.4765 | g_loss: 0.2443\n",
      "Epoch [   39/   50] | d_loss: 0.4982 | g_loss: 0.2580\n",
      "Epoch [   39/   50] | d_loss: 0.4849 | g_loss: 0.2761\n",
      "Epoch [   39/   50] | d_loss: 0.4761 | g_loss: 0.2743\n",
      "Epoch [   39/   50] | d_loss: 0.4942 | g_loss: 0.2790\n",
      "Epoch [   39/   50] | d_loss: 0.5190 | g_loss: 0.2402\n",
      "Epoch [   39/   50] | d_loss: 0.4956 | g_loss: 0.2657\n",
      "Epoch [   39/   50] | d_loss: 0.5122 | g_loss: 0.2553\n",
      "Epoch [   39/   50] | d_loss: 0.4973 | g_loss: 0.2665\n",
      "Epoch [   39/   50] | d_loss: 0.5080 | g_loss: 0.2664\n",
      "Epoch [   39/   50] | d_loss: 0.4647 | g_loss: 0.2688\n",
      "Epoch [   39/   50] | d_loss: 0.4805 | g_loss: 0.2552\n",
      "Epoch [   39/   50] | d_loss: 0.5437 | g_loss: 0.2623\n",
      "Epoch [   39/   50] | d_loss: 0.5432 | g_loss: 0.2602\n",
      "Epoch [   39/   50] | d_loss: 0.4748 | g_loss: 0.2638\n",
      "Epoch [   39/   50] | d_loss: 0.4860 | g_loss: 0.2842\n",
      "Epoch [   39/   50] | d_loss: 0.4851 | g_loss: 0.2469\n",
      "Epoch [   39/   50] | d_loss: 0.5485 | g_loss: 0.2456\n",
      "Epoch [   39/   50] | d_loss: 0.5084 | g_loss: 0.2874\n",
      "Epoch [   39/   50] | d_loss: 0.5031 | g_loss: 0.2367\n",
      "Epoch [   39/   50] | d_loss: 0.5220 | g_loss: 0.2446\n",
      "Epoch [   39/   50] | d_loss: 0.4841 | g_loss: 0.2579\n",
      "Epoch [   39/   50] | d_loss: 0.4892 | g_loss: 0.2660\n",
      "Epoch [   39/   50] | d_loss: 0.4896 | g_loss: 0.2515\n",
      "Epoch [   39/   50] | d_loss: 0.4893 | g_loss: 0.2451\n",
      "Epoch [   39/   50] | d_loss: 0.4886 | g_loss: 0.2435\n",
      "Epoch [   39/   50] | d_loss: 0.5075 | g_loss: 0.2683\n",
      "Epoch [   39/   50] | d_loss: 0.4850 | g_loss: 0.2490\n",
      "Epoch [   39/   50] | d_loss: 0.5363 | g_loss: 0.2626\n",
      "Epoch [   39/   50] | d_loss: 0.5185 | g_loss: 0.2527\n",
      "Epoch [   39/   50] | d_loss: 0.5031 | g_loss: 0.2582\n",
      "Epoch [   39/   50] | d_loss: 0.4834 | g_loss: 0.2515\n",
      "Epoch [   39/   50] | d_loss: 0.5006 | g_loss: 0.2708\n",
      "Epoch [   39/   50] | d_loss: 0.5292 | g_loss: 0.2619\n",
      "Epoch [   39/   50] | d_loss: 0.4977 | g_loss: 0.2461\n",
      "Epoch [   39/   50] | d_loss: 0.5279 | g_loss: 0.2791\n",
      "Epoch [   39/   50] | d_loss: 0.4888 | g_loss: 0.2284\n",
      "Epoch [   39/   50] | d_loss: 0.4978 | g_loss: 0.2791\n",
      "Epoch [   39/   50] | d_loss: 0.4972 | g_loss: 0.2134\n",
      "Epoch [   39/   50] | d_loss: 0.4934 | g_loss: 0.2761\n",
      "Epoch [   39/   50] | d_loss: 0.4652 | g_loss: 0.2779\n",
      "Epoch [   39/   50] | d_loss: 0.5159 | g_loss: 0.2484\n",
      "Epoch [   39/   50] | d_loss: 0.5059 | g_loss: 0.2562\n",
      "Epoch [   39/   50] | d_loss: 0.4939 | g_loss: 0.2922\n",
      "Epoch [   39/   50] | d_loss: 0.5463 | g_loss: 0.2522\n",
      "Epoch [   39/   50] | d_loss: 0.5055 | g_loss: 0.2659\n",
      "Epoch [   39/   50] | d_loss: 0.4987 | g_loss: 0.2331\n",
      "Epoch [   40/   50] | d_loss: 0.5089 | g_loss: 0.2731\n",
      "Epoch [   40/   50] | d_loss: 0.4796 | g_loss: 0.2860\n",
      "Epoch [   40/   50] | d_loss: 0.4991 | g_loss: 0.2852\n",
      "Epoch [   40/   50] | d_loss: 0.4894 | g_loss: 0.2512\n",
      "Epoch [   40/   50] | d_loss: 0.5538 | g_loss: 0.2672\n",
      "Epoch [   40/   50] | d_loss: 0.5077 | g_loss: 0.2380\n",
      "Epoch [   40/   50] | d_loss: 0.4954 | g_loss: 0.2566\n",
      "Epoch [   40/   50] | d_loss: 0.5074 | g_loss: 0.2364\n",
      "Epoch [   40/   50] | d_loss: 0.4905 | g_loss: 0.2739\n",
      "Epoch [   40/   50] | d_loss: 0.5048 | g_loss: 0.2512\n",
      "Epoch [   40/   50] | d_loss: 0.5387 | g_loss: 0.2050\n",
      "Epoch [   40/   50] | d_loss: 0.5012 | g_loss: 0.2431\n",
      "Epoch [   40/   50] | d_loss: 0.5320 | g_loss: 0.2714\n",
      "Epoch [   40/   50] | d_loss: 0.5025 | g_loss: 0.2561\n",
      "Epoch [   40/   50] | d_loss: 0.5082 | g_loss: 0.2306\n",
      "Epoch [   40/   50] | d_loss: 0.4951 | g_loss: 0.2644\n",
      "Epoch [   40/   50] | d_loss: 0.5212 | g_loss: 0.2293\n",
      "Epoch [   40/   50] | d_loss: 0.4869 | g_loss: 0.2642\n",
      "Epoch [   40/   50] | d_loss: 0.5101 | g_loss: 0.2741\n",
      "Epoch [   40/   50] | d_loss: 0.4994 | g_loss: 0.2421\n",
      "Epoch [   40/   50] | d_loss: 0.5399 | g_loss: 0.2536\n",
      "Epoch [   40/   50] | d_loss: 0.4695 | g_loss: 0.3074\n",
      "Epoch [   40/   50] | d_loss: 0.5144 | g_loss: 0.2346\n",
      "Epoch [   40/   50] | d_loss: 0.4980 | g_loss: 0.2908\n",
      "Epoch [   40/   50] | d_loss: 0.4641 | g_loss: 0.2357\n",
      "Epoch [   40/   50] | d_loss: 0.4823 | g_loss: 0.2623\n",
      "Epoch [   40/   50] | d_loss: 0.5254 | g_loss: 0.2484\n",
      "Epoch [   40/   50] | d_loss: 0.5271 | g_loss: 0.2678\n",
      "Epoch [   40/   50] | d_loss: 0.4946 | g_loss: 0.2479\n",
      "Epoch [   40/   50] | d_loss: 0.4911 | g_loss: 0.2542\n",
      "Epoch [   40/   50] | d_loss: 0.4538 | g_loss: 0.2491\n",
      "Epoch [   40/   50] | d_loss: 0.5533 | g_loss: 0.2953\n",
      "Epoch [   40/   50] | d_loss: 0.5022 | g_loss: 0.2592\n",
      "Epoch [   40/   50] | d_loss: 0.5019 | g_loss: 0.2474\n",
      "Epoch [   40/   50] | d_loss: 0.4924 | g_loss: 0.2819\n",
      "Epoch [   40/   50] | d_loss: 0.5252 | g_loss: 0.2681\n",
      "Epoch [   40/   50] | d_loss: 0.4955 | g_loss: 0.2511\n",
      "Epoch [   40/   50] | d_loss: 0.5058 | g_loss: 0.2679\n",
      "Epoch [   40/   50] | d_loss: 0.4975 | g_loss: 0.2360\n",
      "Epoch [   40/   50] | d_loss: 0.4774 | g_loss: 0.2764\n",
      "Epoch [   40/   50] | d_loss: 0.4746 | g_loss: 0.2575\n",
      "Epoch [   40/   50] | d_loss: 0.4937 | g_loss: 0.2489\n",
      "Epoch [   40/   50] | d_loss: 0.4996 | g_loss: 0.2915\n",
      "Epoch [   40/   50] | d_loss: 0.5374 | g_loss: 0.2480\n",
      "Epoch [   40/   50] | d_loss: 0.4828 | g_loss: 0.2655\n",
      "Epoch [   40/   50] | d_loss: 0.5288 | g_loss: 0.2395\n",
      "Epoch [   40/   50] | d_loss: 0.4916 | g_loss: 0.2965\n",
      "Epoch [   40/   50] | d_loss: 0.4918 | g_loss: 0.2383\n",
      "Epoch [   40/   50] | d_loss: 0.4829 | g_loss: 0.2454\n",
      "Epoch [   40/   50] | d_loss: 0.4914 | g_loss: 0.2574\n",
      "Epoch [   40/   50] | d_loss: 0.5122 | g_loss: 0.2606\n",
      "Epoch [   40/   50] | d_loss: 0.5429 | g_loss: 0.2837\n",
      "Epoch [   40/   50] | d_loss: 0.4716 | g_loss: 0.2732\n",
      "Epoch [   40/   50] | d_loss: 0.5046 | g_loss: 0.2685\n",
      "Epoch [   40/   50] | d_loss: 0.4847 | g_loss: 0.2485\n",
      "Epoch [   40/   50] | d_loss: 0.5267 | g_loss: 0.2580\n",
      "Epoch [   40/   50] | d_loss: 0.4819 | g_loss: 0.2473\n",
      "Epoch [   40/   50] | d_loss: 0.5047 | g_loss: 0.2748\n",
      "Epoch [   40/   50] | d_loss: 0.5383 | g_loss: 0.2718\n",
      "Epoch [   40/   50] | d_loss: 0.4779 | g_loss: 0.2572\n",
      "Epoch [   40/   50] | d_loss: 0.4995 | g_loss: 0.2427\n",
      "Epoch [   40/   50] | d_loss: 0.5004 | g_loss: 0.2473\n",
      "Epoch [   40/   50] | d_loss: 0.4842 | g_loss: 0.2601\n",
      "Epoch [   40/   50] | d_loss: 0.4693 | g_loss: 0.2647\n",
      "Epoch [   40/   50] | d_loss: 0.5058 | g_loss: 0.2519\n",
      "Epoch [   40/   50] | d_loss: 0.4681 | g_loss: 0.2688\n",
      "Epoch [   40/   50] | d_loss: 0.5169 | g_loss: 0.2661\n",
      "Epoch [   40/   50] | d_loss: 0.4668 | g_loss: 0.2845\n",
      "Epoch [   40/   50] | d_loss: 0.5276 | g_loss: 0.2534\n",
      "Epoch [   40/   50] | d_loss: 0.5125 | g_loss: 0.2592\n",
      "Epoch [   40/   50] | d_loss: 0.5055 | g_loss: 0.2364\n",
      "Epoch [   40/   50] | d_loss: 0.4987 | g_loss: 0.2743\n",
      "Epoch [   40/   50] | d_loss: 0.5239 | g_loss: 0.2537\n",
      "Epoch [   40/   50] | d_loss: 0.5079 | g_loss: 0.2560\n",
      "Epoch [   40/   50] | d_loss: 0.4842 | g_loss: 0.2431\n",
      "Epoch [   40/   50] | d_loss: 0.5103 | g_loss: 0.2704\n",
      "Epoch [   40/   50] | d_loss: 0.5034 | g_loss: 0.2539\n",
      "Epoch [   40/   50] | d_loss: 0.5235 | g_loss: 0.2897\n",
      "Epoch [   40/   50] | d_loss: 0.5220 | g_loss: 0.2468\n",
      "Epoch [   40/   50] | d_loss: 0.4975 | g_loss: 0.2430\n",
      "Epoch [   40/   50] | d_loss: 0.4748 | g_loss: 0.2525\n",
      "Epoch [   40/   50] | d_loss: 0.4909 | g_loss: 0.2366\n",
      "Epoch [   40/   50] | d_loss: 0.5339 | g_loss: 0.2639\n",
      "Epoch [   40/   50] | d_loss: 0.5073 | g_loss: 0.2593\n",
      "Epoch [   40/   50] | d_loss: 0.4892 | g_loss: 0.2743\n",
      "Epoch [   40/   50] | d_loss: 0.5203 | g_loss: 0.2502\n",
      "Epoch [   40/   50] | d_loss: 0.5039 | g_loss: 0.2528\n",
      "Epoch [   40/   50] | d_loss: 0.5063 | g_loss: 0.2609\n",
      "Epoch [   40/   50] | d_loss: 0.4901 | g_loss: 0.2632\n",
      "Epoch [   40/   50] | d_loss: 0.4714 | g_loss: 0.2211\n",
      "Epoch [   41/   50] | d_loss: 0.4938 | g_loss: 0.2704\n",
      "Epoch [   41/   50] | d_loss: 0.5022 | g_loss: 0.2698\n",
      "Epoch [   41/   50] | d_loss: 0.5080 | g_loss: 0.2454\n",
      "Epoch [   41/   50] | d_loss: 0.5190 | g_loss: 0.2393\n",
      "Epoch [   41/   50] | d_loss: 0.5030 | g_loss: 0.2502\n",
      "Epoch [   41/   50] | d_loss: 0.4711 | g_loss: 0.2653\n",
      "Epoch [   41/   50] | d_loss: 0.5154 | g_loss: 0.2502\n",
      "Epoch [   41/   50] | d_loss: 0.5003 | g_loss: 0.2520\n",
      "Epoch [   41/   50] | d_loss: 0.5148 | g_loss: 0.2559\n",
      "Epoch [   41/   50] | d_loss: 0.5004 | g_loss: 0.2357\n",
      "Epoch [   41/   50] | d_loss: 0.5049 | g_loss: 0.2605\n",
      "Epoch [   41/   50] | d_loss: 0.4948 | g_loss: 0.2668\n",
      "Epoch [   41/   50] | d_loss: 0.5089 | g_loss: 0.2460\n",
      "Epoch [   41/   50] | d_loss: 0.5130 | g_loss: 0.2571\n",
      "Epoch [   41/   50] | d_loss: 0.4722 | g_loss: 0.2681\n",
      "Epoch [   41/   50] | d_loss: 0.4540 | g_loss: 0.2713\n",
      "Epoch [   41/   50] | d_loss: 0.5218 | g_loss: 0.2902\n",
      "Epoch [   41/   50] | d_loss: 0.5033 | g_loss: 0.2671\n",
      "Epoch [   41/   50] | d_loss: 0.4698 | g_loss: 0.2675\n",
      "Epoch [   41/   50] | d_loss: 0.4826 | g_loss: 0.2639\n",
      "Epoch [   41/   50] | d_loss: 0.5271 | g_loss: 0.2536\n",
      "Epoch [   41/   50] | d_loss: 0.4999 | g_loss: 0.2524\n",
      "Epoch [   41/   50] | d_loss: 0.4947 | g_loss: 0.2983\n",
      "Epoch [   41/   50] | d_loss: 0.4864 | g_loss: 0.2442\n",
      "Epoch [   41/   50] | d_loss: 0.4537 | g_loss: 0.2564\n",
      "Epoch [   41/   50] | d_loss: 0.4949 | g_loss: 0.2597\n",
      "Epoch [   41/   50] | d_loss: 0.5218 | g_loss: 0.2620\n",
      "Epoch [   41/   50] | d_loss: 0.4820 | g_loss: 0.2810\n",
      "Epoch [   41/   50] | d_loss: 0.5170 | g_loss: 0.2849\n",
      "Epoch [   41/   50] | d_loss: 0.5007 | g_loss: 0.2545\n",
      "Epoch [   41/   50] | d_loss: 0.5308 | g_loss: 0.2538\n",
      "Epoch [   41/   50] | d_loss: 0.5013 | g_loss: 0.2665\n",
      "Epoch [   41/   50] | d_loss: 0.5280 | g_loss: 0.2687\n",
      "Epoch [   41/   50] | d_loss: 0.5332 | g_loss: 0.2556\n",
      "Epoch [   41/   50] | d_loss: 0.5173 | g_loss: 0.2458\n",
      "Epoch [   41/   50] | d_loss: 0.5369 | g_loss: 0.2474\n",
      "Epoch [   41/   50] | d_loss: 0.4943 | g_loss: 0.2390\n",
      "Epoch [   41/   50] | d_loss: 0.5055 | g_loss: 0.2398\n",
      "Epoch [   41/   50] | d_loss: 0.5056 | g_loss: 0.2511\n",
      "Epoch [   41/   50] | d_loss: 0.5203 | g_loss: 0.2520\n",
      "Epoch [   41/   50] | d_loss: 0.4724 | g_loss: 0.2358\n",
      "Epoch [   41/   50] | d_loss: 0.4910 | g_loss: 0.2646\n",
      "Epoch [   41/   50] | d_loss: 0.5148 | g_loss: 0.2664\n",
      "Epoch [   41/   50] | d_loss: 0.4755 | g_loss: 0.2345\n",
      "Epoch [   41/   50] | d_loss: 0.4921 | g_loss: 0.2507\n",
      "Epoch [   41/   50] | d_loss: 0.5281 | g_loss: 0.2599\n",
      "Epoch [   41/   50] | d_loss: 0.4882 | g_loss: 0.2421\n",
      "Epoch [   41/   50] | d_loss: 0.5496 | g_loss: 0.2633\n",
      "Epoch [   41/   50] | d_loss: 0.5324 | g_loss: 0.2627\n",
      "Epoch [   41/   50] | d_loss: 0.4964 | g_loss: 0.2535\n",
      "Epoch [   41/   50] | d_loss: 0.4829 | g_loss: 0.2513\n",
      "Epoch [   41/   50] | d_loss: 0.5095 | g_loss: 0.2350\n",
      "Epoch [   41/   50] | d_loss: 0.4972 | g_loss: 0.2669\n",
      "Epoch [   41/   50] | d_loss: 0.5091 | g_loss: 0.2550\n",
      "Epoch [   41/   50] | d_loss: 0.5053 | g_loss: 0.2409\n",
      "Epoch [   41/   50] | d_loss: 0.5284 | g_loss: 0.2787\n",
      "Epoch [   41/   50] | d_loss: 0.5214 | g_loss: 0.2603\n",
      "Epoch [   41/   50] | d_loss: 0.4870 | g_loss: 0.2532\n",
      "Epoch [   41/   50] | d_loss: 0.5117 | g_loss: 0.2635\n",
      "Epoch [   41/   50] | d_loss: 0.5058 | g_loss: 0.2302\n",
      "Epoch [   41/   50] | d_loss: 0.4809 | g_loss: 0.2627\n",
      "Epoch [   41/   50] | d_loss: 0.4572 | g_loss: 0.2749\n",
      "Epoch [   41/   50] | d_loss: 0.4836 | g_loss: 0.3016\n",
      "Epoch [   41/   50] | d_loss: 0.4839 | g_loss: 0.2441\n",
      "Epoch [   41/   50] | d_loss: 0.5021 | g_loss: 0.2502\n",
      "Epoch [   41/   50] | d_loss: 0.4915 | g_loss: 0.2546\n",
      "Epoch [   41/   50] | d_loss: 0.4961 | g_loss: 0.2781\n",
      "Epoch [   41/   50] | d_loss: 0.4829 | g_loss: 0.2476\n",
      "Epoch [   41/   50] | d_loss: 0.4705 | g_loss: 0.2646\n",
      "Epoch [   41/   50] | d_loss: 0.4940 | g_loss: 0.2422\n",
      "Epoch [   41/   50] | d_loss: 0.4547 | g_loss: 0.2877\n",
      "Epoch [   41/   50] | d_loss: 0.5100 | g_loss: 0.2497\n",
      "Epoch [   41/   50] | d_loss: 0.5364 | g_loss: 0.2769\n",
      "Epoch [   41/   50] | d_loss: 0.4781 | g_loss: 0.2405\n",
      "Epoch [   41/   50] | d_loss: 0.4904 | g_loss: 0.2483\n",
      "Epoch [   41/   50] | d_loss: 0.5045 | g_loss: 0.2687\n",
      "Epoch [   41/   50] | d_loss: 0.5039 | g_loss: 0.2469\n",
      "Epoch [   41/   50] | d_loss: 0.4942 | g_loss: 0.2598\n",
      "Epoch [   41/   50] | d_loss: 0.5327 | g_loss: 0.2442\n",
      "Epoch [   41/   50] | d_loss: 0.5080 | g_loss: 0.2756\n",
      "Epoch [   41/   50] | d_loss: 0.5344 | g_loss: 0.2606\n",
      "Epoch [   41/   50] | d_loss: 0.4878 | g_loss: 0.2541\n",
      "Epoch [   41/   50] | d_loss: 0.4971 | g_loss: 0.2846\n",
      "Epoch [   41/   50] | d_loss: 0.4923 | g_loss: 0.2705\n",
      "Epoch [   41/   50] | d_loss: 0.5194 | g_loss: 0.2266\n",
      "Epoch [   41/   50] | d_loss: 0.5104 | g_loss: 0.2510\n",
      "Epoch [   41/   50] | d_loss: 0.5423 | g_loss: 0.2633\n",
      "Epoch [   41/   50] | d_loss: 0.5039 | g_loss: 0.2528\n",
      "Epoch [   41/   50] | d_loss: 0.5174 | g_loss: 0.2712\n",
      "Epoch [   41/   50] | d_loss: 0.4970 | g_loss: 0.2513\n",
      "Epoch [   42/   50] | d_loss: 0.5050 | g_loss: 0.2632\n",
      "Epoch [   42/   50] | d_loss: 0.4785 | g_loss: 0.2852\n",
      "Epoch [   42/   50] | d_loss: 0.5022 | g_loss: 0.3045\n",
      "Epoch [   42/   50] | d_loss: 0.5142 | g_loss: 0.2340\n",
      "Epoch [   42/   50] | d_loss: 0.4940 | g_loss: 0.2559\n",
      "Epoch [   42/   50] | d_loss: 0.4932 | g_loss: 0.2823\n",
      "Epoch [   42/   50] | d_loss: 0.4611 | g_loss: 0.2441\n",
      "Epoch [   42/   50] | d_loss: 0.4838 | g_loss: 0.2817\n",
      "Epoch [   42/   50] | d_loss: 0.5031 | g_loss: 0.2593\n",
      "Epoch [   42/   50] | d_loss: 0.4849 | g_loss: 0.2478\n",
      "Epoch [   42/   50] | d_loss: 0.5104 | g_loss: 0.2666\n",
      "Epoch [   42/   50] | d_loss: 0.5002 | g_loss: 0.2519\n",
      "Epoch [   42/   50] | d_loss: 0.4776 | g_loss: 0.2522\n",
      "Epoch [   42/   50] | d_loss: 0.5213 | g_loss: 0.2697\n",
      "Epoch [   42/   50] | d_loss: 0.4902 | g_loss: 0.2457\n",
      "Epoch [   42/   50] | d_loss: 0.5017 | g_loss: 0.2938\n",
      "Epoch [   42/   50] | d_loss: 0.4742 | g_loss: 0.2654\n",
      "Epoch [   42/   50] | d_loss: 0.5284 | g_loss: 0.2686\n",
      "Epoch [   42/   50] | d_loss: 0.5181 | g_loss: 0.2755\n",
      "Epoch [   42/   50] | d_loss: 0.5237 | g_loss: 0.2409\n",
      "Epoch [   42/   50] | d_loss: 0.4867 | g_loss: 0.2509\n",
      "Epoch [   42/   50] | d_loss: 0.4981 | g_loss: 0.2644\n",
      "Epoch [   42/   50] | d_loss: 0.4798 | g_loss: 0.2498\n",
      "Epoch [   42/   50] | d_loss: 0.5135 | g_loss: 0.2542\n",
      "Epoch [   42/   50] | d_loss: 0.4837 | g_loss: 0.2587\n",
      "Epoch [   42/   50] | d_loss: 0.4807 | g_loss: 0.2412\n",
      "Epoch [   42/   50] | d_loss: 0.4993 | g_loss: 0.2362\n",
      "Epoch [   42/   50] | d_loss: 0.5257 | g_loss: 0.2127\n",
      "Epoch [   42/   50] | d_loss: 0.4746 | g_loss: 0.2619\n",
      "Epoch [   42/   50] | d_loss: 0.5305 | g_loss: 0.2752\n",
      "Epoch [   42/   50] | d_loss: 0.4976 | g_loss: 0.2724\n",
      "Epoch [   42/   50] | d_loss: 0.4488 | g_loss: 0.2597\n",
      "Epoch [   42/   50] | d_loss: 0.5139 | g_loss: 0.2433\n",
      "Epoch [   42/   50] | d_loss: 0.4920 | g_loss: 0.2781\n",
      "Epoch [   42/   50] | d_loss: 0.4881 | g_loss: 0.2667\n",
      "Epoch [   42/   50] | d_loss: 0.5339 | g_loss: 0.2599\n",
      "Epoch [   42/   50] | d_loss: 0.4864 | g_loss: 0.2493\n",
      "Epoch [   42/   50] | d_loss: 0.4880 | g_loss: 0.2357\n",
      "Epoch [   42/   50] | d_loss: 0.4822 | g_loss: 0.2711\n",
      "Epoch [   42/   50] | d_loss: 0.5002 | g_loss: 0.2539\n",
      "Epoch [   42/   50] | d_loss: 0.5095 | g_loss: 0.2755\n",
      "Epoch [   42/   50] | d_loss: 0.5123 | g_loss: 0.2370\n",
      "Epoch [   42/   50] | d_loss: 0.5100 | g_loss: 0.2512\n",
      "Epoch [   42/   50] | d_loss: 0.4801 | g_loss: 0.2591\n",
      "Epoch [   42/   50] | d_loss: 0.4882 | g_loss: 0.2476\n",
      "Epoch [   42/   50] | d_loss: 0.4998 | g_loss: 0.2523\n",
      "Epoch [   42/   50] | d_loss: 0.4971 | g_loss: 0.2671\n",
      "Epoch [   42/   50] | d_loss: 0.5105 | g_loss: 0.2590\n",
      "Epoch [   42/   50] | d_loss: 0.4903 | g_loss: 0.2655\n",
      "Epoch [   42/   50] | d_loss: 0.5371 | g_loss: 0.2989\n",
      "Epoch [   42/   50] | d_loss: 0.5148 | g_loss: 0.2990\n",
      "Epoch [   42/   50] | d_loss: 0.4627 | g_loss: 0.2683\n",
      "Epoch [   42/   50] | d_loss: 0.5074 | g_loss: 0.2627\n",
      "Epoch [   42/   50] | d_loss: 0.5016 | g_loss: 0.2406\n",
      "Epoch [   42/   50] | d_loss: 0.4976 | g_loss: 0.2711\n",
      "Epoch [   42/   50] | d_loss: 0.4547 | g_loss: 0.2516\n",
      "Epoch [   42/   50] | d_loss: 0.5024 | g_loss: 0.2594\n",
      "Epoch [   42/   50] | d_loss: 0.5210 | g_loss: 0.2562\n",
      "Epoch [   42/   50] | d_loss: 0.5050 | g_loss: 0.3289\n",
      "Epoch [   42/   50] | d_loss: 0.4759 | g_loss: 0.2337\n",
      "Epoch [   42/   50] | d_loss: 0.5025 | g_loss: 0.2648\n",
      "Epoch [   42/   50] | d_loss: 0.5069 | g_loss: 0.2697\n",
      "Epoch [   42/   50] | d_loss: 0.4901 | g_loss: 0.2742\n",
      "Epoch [   42/   50] | d_loss: 0.4900 | g_loss: 0.2534\n",
      "Epoch [   42/   50] | d_loss: 0.4861 | g_loss: 0.2671\n",
      "Epoch [   42/   50] | d_loss: 0.4473 | g_loss: 0.2819\n",
      "Epoch [   42/   50] | d_loss: 0.5150 | g_loss: 0.2681\n",
      "Epoch [   42/   50] | d_loss: 0.4867 | g_loss: 0.2514\n",
      "Epoch [   42/   50] | d_loss: 0.4503 | g_loss: 0.2799\n",
      "Epoch [   42/   50] | d_loss: 0.4957 | g_loss: 0.2596\n",
      "Epoch [   42/   50] | d_loss: 0.5213 | g_loss: 0.2477\n",
      "Epoch [   42/   50] | d_loss: 0.5177 | g_loss: 0.2732\n",
      "Epoch [   42/   50] | d_loss: 0.4960 | g_loss: 0.2560\n",
      "Epoch [   42/   50] | d_loss: 0.4879 | g_loss: 0.2485\n",
      "Epoch [   42/   50] | d_loss: 0.5074 | g_loss: 0.2671\n",
      "Epoch [   42/   50] | d_loss: 0.4920 | g_loss: 0.2497\n",
      "Epoch [   42/   50] | d_loss: 0.4785 | g_loss: 0.2701\n",
      "Epoch [   42/   50] | d_loss: 0.5116 | g_loss: 0.2402\n",
      "Epoch [   42/   50] | d_loss: 0.5185 | g_loss: 0.2624\n",
      "Epoch [   42/   50] | d_loss: 0.5235 | g_loss: 0.2592\n",
      "Epoch [   42/   50] | d_loss: 0.5071 | g_loss: 0.2587\n",
      "Epoch [   42/   50] | d_loss: 0.4774 | g_loss: 0.2761\n",
      "Epoch [   42/   50] | d_loss: 0.5054 | g_loss: 0.2793\n",
      "Epoch [   42/   50] | d_loss: 0.5330 | g_loss: 0.2348\n",
      "Epoch [   42/   50] | d_loss: 0.5371 | g_loss: 0.2686\n",
      "Epoch [   42/   50] | d_loss: 0.4867 | g_loss: 0.2479\n",
      "Epoch [   42/   50] | d_loss: 0.4805 | g_loss: 0.2266\n",
      "Epoch [   42/   50] | d_loss: 0.4649 | g_loss: 0.2525\n",
      "Epoch [   42/   50] | d_loss: 0.4950 | g_loss: 0.2715\n",
      "Epoch [   42/   50] | d_loss: 0.4904 | g_loss: 0.2426\n",
      "Epoch [   43/   50] | d_loss: 0.5026 | g_loss: 0.2549\n",
      "Epoch [   43/   50] | d_loss: 0.5266 | g_loss: 0.2693\n",
      "Epoch [   43/   50] | d_loss: 0.5255 | g_loss: 0.2451\n",
      "Epoch [   43/   50] | d_loss: 0.5489 | g_loss: 0.2327\n",
      "Epoch [   43/   50] | d_loss: 0.4994 | g_loss: 0.2580\n",
      "Epoch [   43/   50] | d_loss: 0.4953 | g_loss: 0.2593\n",
      "Epoch [   43/   50] | d_loss: 0.4934 | g_loss: 0.2498\n",
      "Epoch [   43/   50] | d_loss: 0.5580 | g_loss: 0.2413\n",
      "Epoch [   43/   50] | d_loss: 0.4982 | g_loss: 0.3083\n",
      "Epoch [   43/   50] | d_loss: 0.4690 | g_loss: 0.2671\n",
      "Epoch [   43/   50] | d_loss: 0.5436 | g_loss: 0.2214\n",
      "Epoch [   43/   50] | d_loss: 0.5362 | g_loss: 0.2408\n",
      "Epoch [   43/   50] | d_loss: 0.4669 | g_loss: 0.2598\n",
      "Epoch [   43/   50] | d_loss: 0.4811 | g_loss: 0.2605\n",
      "Epoch [   43/   50] | d_loss: 0.5159 | g_loss: 0.2587\n",
      "Epoch [   43/   50] | d_loss: 0.4933 | g_loss: 0.2671\n",
      "Epoch [   43/   50] | d_loss: 0.5140 | g_loss: 0.2543\n",
      "Epoch [   43/   50] | d_loss: 0.5225 | g_loss: 0.3051\n",
      "Epoch [   43/   50] | d_loss: 0.4769 | g_loss: 0.2550\n",
      "Epoch [   43/   50] | d_loss: 0.4571 | g_loss: 0.2501\n",
      "Epoch [   43/   50] | d_loss: 0.5035 | g_loss: 0.2881\n",
      "Epoch [   43/   50] | d_loss: 0.4365 | g_loss: 0.2902\n",
      "Epoch [   43/   50] | d_loss: 0.4673 | g_loss: 0.2531\n",
      "Epoch [   43/   50] | d_loss: 0.5086 | g_loss: 0.2696\n",
      "Epoch [   43/   50] | d_loss: 0.4761 | g_loss: 0.2534\n",
      "Epoch [   43/   50] | d_loss: 0.4908 | g_loss: 0.2326\n",
      "Epoch [   43/   50] | d_loss: 0.4969 | g_loss: 0.2865\n",
      "Epoch [   43/   50] | d_loss: 0.5000 | g_loss: 0.3130\n",
      "Epoch [   43/   50] | d_loss: 0.5070 | g_loss: 0.2377\n",
      "Epoch [   43/   50] | d_loss: 0.4887 | g_loss: 0.2525\n",
      "Epoch [   43/   50] | d_loss: 0.4709 | g_loss: 0.2745\n",
      "Epoch [   43/   50] | d_loss: 0.4461 | g_loss: 0.2564\n",
      "Epoch [   43/   50] | d_loss: 0.5194 | g_loss: 0.2642\n",
      "Epoch [   43/   50] | d_loss: 0.5046 | g_loss: 0.2410\n",
      "Epoch [   43/   50] | d_loss: 0.5120 | g_loss: 0.2987\n",
      "Epoch [   43/   50] | d_loss: 0.5032 | g_loss: 0.2956\n",
      "Epoch [   43/   50] | d_loss: 0.5007 | g_loss: 0.2567\n",
      "Epoch [   43/   50] | d_loss: 0.5258 | g_loss: 0.2410\n",
      "Epoch [   43/   50] | d_loss: 0.5172 | g_loss: 0.3072\n",
      "Epoch [   43/   50] | d_loss: 0.5125 | g_loss: 0.2550\n",
      "Epoch [   43/   50] | d_loss: 0.4980 | g_loss: 0.2443\n",
      "Epoch [   43/   50] | d_loss: 0.4890 | g_loss: 0.2636\n",
      "Epoch [   43/   50] | d_loss: 0.5019 | g_loss: 0.2566\n",
      "Epoch [   43/   50] | d_loss: 0.5017 | g_loss: 0.2551\n",
      "Epoch [   43/   50] | d_loss: 0.4678 | g_loss: 0.2850\n",
      "Epoch [   43/   50] | d_loss: 0.4901 | g_loss: 0.2456\n",
      "Epoch [   43/   50] | d_loss: 0.4775 | g_loss: 0.2747\n",
      "Epoch [   43/   50] | d_loss: 0.4959 | g_loss: 0.2350\n",
      "Epoch [   43/   50] | d_loss: 0.4874 | g_loss: 0.2490\n",
      "Epoch [   43/   50] | d_loss: 0.5042 | g_loss: 0.2416\n",
      "Epoch [   43/   50] | d_loss: 0.5330 | g_loss: 0.2854\n",
      "Epoch [   43/   50] | d_loss: 0.5082 | g_loss: 0.2414\n",
      "Epoch [   43/   50] | d_loss: 0.4571 | g_loss: 0.2494\n",
      "Epoch [   43/   50] | d_loss: 0.4933 | g_loss: 0.2708\n",
      "Epoch [   43/   50] | d_loss: 0.5353 | g_loss: 0.2736\n",
      "Epoch [   43/   50] | d_loss: 0.5599 | g_loss: 0.2643\n",
      "Epoch [   43/   50] | d_loss: 0.4991 | g_loss: 0.2454\n",
      "Epoch [   43/   50] | d_loss: 0.5130 | g_loss: 0.2557\n",
      "Epoch [   43/   50] | d_loss: 0.5129 | g_loss: 0.2783\n",
      "Epoch [   43/   50] | d_loss: 0.5262 | g_loss: 0.2600\n",
      "Epoch [   43/   50] | d_loss: 0.5065 | g_loss: 0.2717\n",
      "Epoch [   43/   50] | d_loss: 0.5008 | g_loss: 0.2475\n",
      "Epoch [   43/   50] | d_loss: 0.4854 | g_loss: 0.2789\n",
      "Epoch [   43/   50] | d_loss: 0.5081 | g_loss: 0.2541\n",
      "Epoch [   43/   50] | d_loss: 0.5021 | g_loss: 0.2529\n",
      "Epoch [   43/   50] | d_loss: 0.4704 | g_loss: 0.2335\n",
      "Epoch [   43/   50] | d_loss: 0.5048 | g_loss: 0.2489\n",
      "Epoch [   43/   50] | d_loss: 0.4921 | g_loss: 0.2446\n",
      "Epoch [   43/   50] | d_loss: 0.5200 | g_loss: 0.2376\n",
      "Epoch [   43/   50] | d_loss: 0.4983 | g_loss: 0.2407\n",
      "Epoch [   43/   50] | d_loss: 0.5127 | g_loss: 0.2641\n",
      "Epoch [   43/   50] | d_loss: 0.4957 | g_loss: 0.2520\n",
      "Epoch [   43/   50] | d_loss: 0.5423 | g_loss: 0.2244\n",
      "Epoch [   43/   50] | d_loss: 0.4805 | g_loss: 0.2512\n",
      "Epoch [   43/   50] | d_loss: 0.4976 | g_loss: 0.2618\n",
      "Epoch [   43/   50] | d_loss: 0.5019 | g_loss: 0.2504\n",
      "Epoch [   43/   50] | d_loss: 0.4503 | g_loss: 0.2702\n",
      "Epoch [   43/   50] | d_loss: 0.5173 | g_loss: 0.2343\n",
      "Epoch [   43/   50] | d_loss: 0.4874 | g_loss: 0.2570\n",
      "Epoch [   43/   50] | d_loss: 0.4859 | g_loss: 0.2567\n",
      "Epoch [   43/   50] | d_loss: 0.4681 | g_loss: 0.2625\n",
      "Epoch [   43/   50] | d_loss: 0.5139 | g_loss: 0.2980\n",
      "Epoch [   43/   50] | d_loss: 0.4739 | g_loss: 0.2309\n",
      "Epoch [   43/   50] | d_loss: 0.4932 | g_loss: 0.2415\n",
      "Epoch [   43/   50] | d_loss: 0.5090 | g_loss: 0.2989\n",
      "Epoch [   43/   50] | d_loss: 0.4978 | g_loss: 0.2648\n",
      "Epoch [   43/   50] | d_loss: 0.5100 | g_loss: 0.2465\n",
      "Epoch [   43/   50] | d_loss: 0.5326 | g_loss: 0.2720\n",
      "Epoch [   43/   50] | d_loss: 0.5158 | g_loss: 0.2291\n",
      "Epoch [   43/   50] | d_loss: 0.5071 | g_loss: 0.2858\n",
      "Epoch [   44/   50] | d_loss: 0.4806 | g_loss: 0.2463\n",
      "Epoch [   44/   50] | d_loss: 0.5132 | g_loss: 0.2661\n",
      "Epoch [   44/   50] | d_loss: 0.5033 | g_loss: 0.2412\n",
      "Epoch [   44/   50] | d_loss: 0.5379 | g_loss: 0.2565\n",
      "Epoch [   44/   50] | d_loss: 0.5155 | g_loss: 0.2237\n",
      "Epoch [   44/   50] | d_loss: 0.4909 | g_loss: 0.2392\n",
      "Epoch [   44/   50] | d_loss: 0.5121 | g_loss: 0.2397\n",
      "Epoch [   44/   50] | d_loss: 0.5286 | g_loss: 0.2715\n",
      "Epoch [   44/   50] | d_loss: 0.4656 | g_loss: 0.2780\n",
      "Epoch [   44/   50] | d_loss: 0.5070 | g_loss: 0.2381\n",
      "Epoch [   44/   50] | d_loss: 0.4937 | g_loss: 0.2581\n",
      "Epoch [   44/   50] | d_loss: 0.4833 | g_loss: 0.2546\n",
      "Epoch [   44/   50] | d_loss: 0.5323 | g_loss: 0.2394\n",
      "Epoch [   44/   50] | d_loss: 0.5203 | g_loss: 0.2934\n",
      "Epoch [   44/   50] | d_loss: 0.5127 | g_loss: 0.2591\n",
      "Epoch [   44/   50] | d_loss: 0.5168 | g_loss: 0.2729\n",
      "Epoch [   44/   50] | d_loss: 0.5056 | g_loss: 0.2745\n",
      "Epoch [   44/   50] | d_loss: 0.5144 | g_loss: 0.2126\n",
      "Epoch [   44/   50] | d_loss: 0.4990 | g_loss: 0.2417\n",
      "Epoch [   44/   50] | d_loss: 0.5061 | g_loss: 0.2873\n",
      "Epoch [   44/   50] | d_loss: 0.4735 | g_loss: 0.2449\n",
      "Epoch [   44/   50] | d_loss: 0.4930 | g_loss: 0.2518\n",
      "Epoch [   44/   50] | d_loss: 0.5091 | g_loss: 0.2987\n",
      "Epoch [   44/   50] | d_loss: 0.4842 | g_loss: 0.2590\n",
      "Epoch [   44/   50] | d_loss: 0.4934 | g_loss: 0.2726\n",
      "Epoch [   44/   50] | d_loss: 0.4977 | g_loss: 0.2406\n",
      "Epoch [   44/   50] | d_loss: 0.4789 | g_loss: 0.2551\n",
      "Epoch [   44/   50] | d_loss: 0.4604 | g_loss: 0.2668\n",
      "Epoch [   44/   50] | d_loss: 0.5143 | g_loss: 0.2451\n",
      "Epoch [   44/   50] | d_loss: 0.5086 | g_loss: 0.2614\n",
      "Epoch [   44/   50] | d_loss: 0.5210 | g_loss: 0.2486\n",
      "Epoch [   44/   50] | d_loss: 0.4951 | g_loss: 0.2507\n",
      "Epoch [   44/   50] | d_loss: 0.5205 | g_loss: 0.2602\n",
      "Epoch [   44/   50] | d_loss: 0.5270 | g_loss: 0.2157\n",
      "Epoch [   44/   50] | d_loss: 0.5074 | g_loss: 0.2476\n",
      "Epoch [   44/   50] | d_loss: 0.4907 | g_loss: 0.2618\n",
      "Epoch [   44/   50] | d_loss: 0.4572 | g_loss: 0.2410\n",
      "Epoch [   44/   50] | d_loss: 0.5493 | g_loss: 0.2508\n",
      "Epoch [   44/   50] | d_loss: 0.4983 | g_loss: 0.2466\n",
      "Epoch [   44/   50] | d_loss: 0.4935 | g_loss: 0.2807\n",
      "Epoch [   44/   50] | d_loss: 0.5481 | g_loss: 0.2508\n",
      "Epoch [   44/   50] | d_loss: 0.4821 | g_loss: 0.2420\n",
      "Epoch [   44/   50] | d_loss: 0.5007 | g_loss: 0.2574\n",
      "Epoch [   44/   50] | d_loss: 0.5082 | g_loss: 0.2412\n",
      "Epoch [   44/   50] | d_loss: 0.5012 | g_loss: 0.2614\n",
      "Epoch [   44/   50] | d_loss: 0.4847 | g_loss: 0.2618\n",
      "Epoch [   44/   50] | d_loss: 0.4781 | g_loss: 0.2523\n",
      "Epoch [   44/   50] | d_loss: 0.5512 | g_loss: 0.2576\n",
      "Epoch [   44/   50] | d_loss: 0.4762 | g_loss: 0.2554\n",
      "Epoch [   44/   50] | d_loss: 0.4697 | g_loss: 0.2604\n",
      "Epoch [   44/   50] | d_loss: 0.4962 | g_loss: 0.2551\n",
      "Epoch [   44/   50] | d_loss: 0.4729 | g_loss: 0.2351\n",
      "Epoch [   44/   50] | d_loss: 0.4717 | g_loss: 0.2614\n",
      "Epoch [   44/   50] | d_loss: 0.5168 | g_loss: 0.2931\n",
      "Epoch [   44/   50] | d_loss: 0.5054 | g_loss: 0.2445\n",
      "Epoch [   44/   50] | d_loss: 0.5041 | g_loss: 0.2486\n",
      "Epoch [   44/   50] | d_loss: 0.4598 | g_loss: 0.2631\n",
      "Epoch [   44/   50] | d_loss: 0.5243 | g_loss: 0.2652\n",
      "Epoch [   44/   50] | d_loss: 0.5420 | g_loss: 0.2621\n",
      "Epoch [   44/   50] | d_loss: 0.4893 | g_loss: 0.2472\n",
      "Epoch [   44/   50] | d_loss: 0.4762 | g_loss: 0.2486\n",
      "Epoch [   44/   50] | d_loss: 0.4859 | g_loss: 0.2231\n",
      "Epoch [   44/   50] | d_loss: 0.5395 | g_loss: 0.2472\n",
      "Epoch [   44/   50] | d_loss: 0.4936 | g_loss: 0.3188\n",
      "Epoch [   44/   50] | d_loss: 0.4644 | g_loss: 0.2684\n",
      "Epoch [   44/   50] | d_loss: 0.4708 | g_loss: 0.2731\n",
      "Epoch [   44/   50] | d_loss: 0.4802 | g_loss: 0.2751\n",
      "Epoch [   44/   50] | d_loss: 0.4894 | g_loss: 0.2584\n",
      "Epoch [   44/   50] | d_loss: 0.5245 | g_loss: 0.2523\n",
      "Epoch [   44/   50] | d_loss: 0.4828 | g_loss: 0.2517\n",
      "Epoch [   44/   50] | d_loss: 0.4862 | g_loss: 0.2662\n",
      "Epoch [   44/   50] | d_loss: 0.5027 | g_loss: 0.2343\n",
      "Epoch [   44/   50] | d_loss: 0.5109 | g_loss: 0.2487\n",
      "Epoch [   44/   50] | d_loss: 0.5036 | g_loss: 0.2551\n",
      "Epoch [   44/   50] | d_loss: 0.5186 | g_loss: 0.2369\n",
      "Epoch [   44/   50] | d_loss: 0.5126 | g_loss: 0.2820\n",
      "Epoch [   44/   50] | d_loss: 0.4966 | g_loss: 0.2785\n",
      "Epoch [   44/   50] | d_loss: 0.5237 | g_loss: 0.2610\n",
      "Epoch [   44/   50] | d_loss: 0.5374 | g_loss: 0.2448\n",
      "Epoch [   44/   50] | d_loss: 0.4999 | g_loss: 0.2785\n",
      "Epoch [   44/   50] | d_loss: 0.5117 | g_loss: 0.2744\n",
      "Epoch [   44/   50] | d_loss: 0.5028 | g_loss: 0.2561\n",
      "Epoch [   44/   50] | d_loss: 0.5154 | g_loss: 0.2346\n",
      "Epoch [   44/   50] | d_loss: 0.5210 | g_loss: 0.2835\n",
      "Epoch [   44/   50] | d_loss: 0.4792 | g_loss: 0.2305\n",
      "Epoch [   44/   50] | d_loss: 0.4770 | g_loss: 0.2657\n",
      "Epoch [   44/   50] | d_loss: 0.5286 | g_loss: 0.2444\n",
      "Epoch [   44/   50] | d_loss: 0.5084 | g_loss: 0.2413\n",
      "Epoch [   44/   50] | d_loss: 0.5140 | g_loss: 0.3056\n",
      "Epoch [   44/   50] | d_loss: 0.5017 | g_loss: 0.2551\n",
      "Epoch [   45/   50] | d_loss: 0.5167 | g_loss: 0.2699\n",
      "Epoch [   45/   50] | d_loss: 0.5064 | g_loss: 0.2494\n",
      "Epoch [   45/   50] | d_loss: 0.5096 | g_loss: 0.2767\n",
      "Epoch [   45/   50] | d_loss: 0.5321 | g_loss: 0.2541\n",
      "Epoch [   45/   50] | d_loss: 0.4982 | g_loss: 0.2563\n",
      "Epoch [   45/   50] | d_loss: 0.4843 | g_loss: 0.2372\n",
      "Epoch [   45/   50] | d_loss: 0.4834 | g_loss: 0.2554\n",
      "Epoch [   45/   50] | d_loss: 0.5199 | g_loss: 0.2470\n",
      "Epoch [   45/   50] | d_loss: 0.5360 | g_loss: 0.2519\n",
      "Epoch [   45/   50] | d_loss: 0.5216 | g_loss: 0.2729\n",
      "Epoch [   45/   50] | d_loss: 0.5542 | g_loss: 0.2609\n",
      "Epoch [   45/   50] | d_loss: 0.5377 | g_loss: 0.2479\n",
      "Epoch [   45/   50] | d_loss: 0.5261 | g_loss: 0.2246\n",
      "Epoch [   45/   50] | d_loss: 0.4983 | g_loss: 0.2609\n",
      "Epoch [   45/   50] | d_loss: 0.4664 | g_loss: 0.2315\n",
      "Epoch [   45/   50] | d_loss: 0.5331 | g_loss: 0.2581\n",
      "Epoch [   45/   50] | d_loss: 0.4985 | g_loss: 0.2697\n",
      "Epoch [   45/   50] | d_loss: 0.5081 | g_loss: 0.3109\n",
      "Epoch [   45/   50] | d_loss: 0.5319 | g_loss: 0.2498\n",
      "Epoch [   45/   50] | d_loss: 0.4727 | g_loss: 0.2397\n",
      "Epoch [   45/   50] | d_loss: 0.4808 | g_loss: 0.2523\n",
      "Epoch [   45/   50] | d_loss: 0.5020 | g_loss: 0.2407\n",
      "Epoch [   45/   50] | d_loss: 0.5379 | g_loss: 0.2401\n",
      "Epoch [   45/   50] | d_loss: 0.4899 | g_loss: 0.2447\n",
      "Epoch [   45/   50] | d_loss: 0.5163 | g_loss: 0.2503\n",
      "Epoch [   45/   50] | d_loss: 0.5156 | g_loss: 0.2877\n",
      "Epoch [   45/   50] | d_loss: 0.5424 | g_loss: 0.2353\n",
      "Epoch [   45/   50] | d_loss: 0.4672 | g_loss: 0.2553\n",
      "Epoch [   45/   50] | d_loss: 0.5471 | g_loss: 0.2674\n",
      "Epoch [   45/   50] | d_loss: 0.4855 | g_loss: 0.2474\n",
      "Epoch [   45/   50] | d_loss: 0.5246 | g_loss: 0.2759\n",
      "Epoch [   45/   50] | d_loss: 0.4925 | g_loss: 0.2813\n",
      "Epoch [   45/   50] | d_loss: 0.4928 | g_loss: 0.2698\n",
      "Epoch [   45/   50] | d_loss: 0.5131 | g_loss: 0.2390\n",
      "Epoch [   45/   50] | d_loss: 0.4914 | g_loss: 0.2797\n",
      "Epoch [   45/   50] | d_loss: 0.4560 | g_loss: 0.2634\n",
      "Epoch [   45/   50] | d_loss: 0.4767 | g_loss: 0.2494\n",
      "Epoch [   45/   50] | d_loss: 0.4878 | g_loss: 0.2596\n",
      "Epoch [   45/   50] | d_loss: 0.4992 | g_loss: 0.2401\n",
      "Epoch [   45/   50] | d_loss: 0.4953 | g_loss: 0.2417\n",
      "Epoch [   45/   50] | d_loss: 0.4841 | g_loss: 0.2569\n",
      "Epoch [   45/   50] | d_loss: 0.5173 | g_loss: 0.2410\n",
      "Epoch [   45/   50] | d_loss: 0.5000 | g_loss: 0.2621\n",
      "Epoch [   45/   50] | d_loss: 0.5196 | g_loss: 0.2667\n",
      "Epoch [   45/   50] | d_loss: 0.5149 | g_loss: 0.2601\n",
      "Epoch [   45/   50] | d_loss: 0.5155 | g_loss: 0.2710\n",
      "Epoch [   45/   50] | d_loss: 0.5016 | g_loss: 0.2339\n",
      "Epoch [   45/   50] | d_loss: 0.5149 | g_loss: 0.2390\n",
      "Epoch [   45/   50] | d_loss: 0.4967 | g_loss: 0.2419\n",
      "Epoch [   45/   50] | d_loss: 0.5009 | g_loss: 0.2262\n",
      "Epoch [   45/   50] | d_loss: 0.4944 | g_loss: 0.2496\n",
      "Epoch [   45/   50] | d_loss: 0.4518 | g_loss: 0.2688\n",
      "Epoch [   45/   50] | d_loss: 0.4362 | g_loss: 0.2947\n",
      "Epoch [   45/   50] | d_loss: 0.4928 | g_loss: 0.2541\n",
      "Epoch [   45/   50] | d_loss: 0.5190 | g_loss: 0.2547\n",
      "Epoch [   45/   50] | d_loss: 0.4721 | g_loss: 0.2524\n",
      "Epoch [   45/   50] | d_loss: 0.4975 | g_loss: 0.2895\n",
      "Epoch [   45/   50] | d_loss: 0.5024 | g_loss: 0.2308\n",
      "Epoch [   45/   50] | d_loss: 0.5105 | g_loss: 0.2511\n",
      "Epoch [   45/   50] | d_loss: 0.5165 | g_loss: 0.2713\n",
      "Epoch [   45/   50] | d_loss: 0.5215 | g_loss: 0.2541\n",
      "Epoch [   45/   50] | d_loss: 0.4964 | g_loss: 0.2612\n",
      "Epoch [   45/   50] | d_loss: 0.5179 | g_loss: 0.2344\n",
      "Epoch [   45/   50] | d_loss: 0.4931 | g_loss: 0.2596\n",
      "Epoch [   45/   50] | d_loss: 0.4861 | g_loss: 0.2616\n",
      "Epoch [   45/   50] | d_loss: 0.4631 | g_loss: 0.2350\n",
      "Epoch [   45/   50] | d_loss: 0.4894 | g_loss: 0.2684\n",
      "Epoch [   45/   50] | d_loss: 0.5261 | g_loss: 0.2402\n",
      "Epoch [   45/   50] | d_loss: 0.5132 | g_loss: 0.2733\n",
      "Epoch [   45/   50] | d_loss: 0.4849 | g_loss: 0.2268\n",
      "Epoch [   45/   50] | d_loss: 0.4783 | g_loss: 0.2468\n",
      "Epoch [   45/   50] | d_loss: 0.4787 | g_loss: 0.2598\n",
      "Epoch [   45/   50] | d_loss: 0.5044 | g_loss: 0.2634\n",
      "Epoch [   45/   50] | d_loss: 0.4953 | g_loss: 0.2762\n",
      "Epoch [   45/   50] | d_loss: 0.4828 | g_loss: 0.2402\n",
      "Epoch [   45/   50] | d_loss: 0.5161 | g_loss: 0.2395\n",
      "Epoch [   45/   50] | d_loss: 0.5156 | g_loss: 0.2722\n",
      "Epoch [   45/   50] | d_loss: 0.5104 | g_loss: 0.2513\n",
      "Epoch [   45/   50] | d_loss: 0.4801 | g_loss: 0.2688\n",
      "Epoch [   45/   50] | d_loss: 0.5048 | g_loss: 0.2422\n",
      "Epoch [   45/   50] | d_loss: 0.5062 | g_loss: 0.2928\n",
      "Epoch [   45/   50] | d_loss: 0.4709 | g_loss: 0.2350\n",
      "Epoch [   45/   50] | d_loss: 0.5125 | g_loss: 0.2464\n",
      "Epoch [   45/   50] | d_loss: 0.4926 | g_loss: 0.2491\n",
      "Epoch [   45/   50] | d_loss: 0.5272 | g_loss: 0.2500\n",
      "Epoch [   45/   50] | d_loss: 0.5102 | g_loss: 0.2858\n",
      "Epoch [   45/   50] | d_loss: 0.4712 | g_loss: 0.2675\n",
      "Epoch [   45/   50] | d_loss: 0.4712 | g_loss: 0.2651\n",
      "Epoch [   45/   50] | d_loss: 0.5343 | g_loss: 0.2721\n",
      "Epoch [   45/   50] | d_loss: 0.5107 | g_loss: 0.2781\n",
      "Epoch [   46/   50] | d_loss: 0.4724 | g_loss: 0.2411\n",
      "Epoch [   46/   50] | d_loss: 0.4675 | g_loss: 0.2290\n",
      "Epoch [   46/   50] | d_loss: 0.4986 | g_loss: 0.2846\n",
      "Epoch [   46/   50] | d_loss: 0.5214 | g_loss: 0.2542\n",
      "Epoch [   46/   50] | d_loss: 0.4942 | g_loss: 0.2351\n",
      "Epoch [   46/   50] | d_loss: 0.5417 | g_loss: 0.2739\n",
      "Epoch [   46/   50] | d_loss: 0.5035 | g_loss: 0.2763\n",
      "Epoch [   46/   50] | d_loss: 0.5332 | g_loss: 0.2502\n",
      "Epoch [   46/   50] | d_loss: 0.5016 | g_loss: 0.2570\n",
      "Epoch [   46/   50] | d_loss: 0.5443 | g_loss: 0.2522\n",
      "Epoch [   46/   50] | d_loss: 0.4952 | g_loss: 0.2634\n",
      "Epoch [   46/   50] | d_loss: 0.5041 | g_loss: 0.2506\n",
      "Epoch [   46/   50] | d_loss: 0.5187 | g_loss: 0.2536\n",
      "Epoch [   46/   50] | d_loss: 0.5370 | g_loss: 0.2764\n",
      "Epoch [   46/   50] | d_loss: 0.4875 | g_loss: 0.2660\n",
      "Epoch [   46/   50] | d_loss: 0.5088 | g_loss: 0.2405\n",
      "Epoch [   46/   50] | d_loss: 0.5433 | g_loss: 0.2235\n",
      "Epoch [   46/   50] | d_loss: 0.5299 | g_loss: 0.2565\n",
      "Epoch [   46/   50] | d_loss: 0.5129 | g_loss: 0.2574\n",
      "Epoch [   46/   50] | d_loss: 0.4965 | g_loss: 0.2701\n",
      "Epoch [   46/   50] | d_loss: 0.5092 | g_loss: 0.2788\n",
      "Epoch [   46/   50] | d_loss: 0.5032 | g_loss: 0.2769\n",
      "Epoch [   46/   50] | d_loss: 0.4756 | g_loss: 0.2621\n",
      "Epoch [   46/   50] | d_loss: 0.5281 | g_loss: 0.2827\n",
      "Epoch [   46/   50] | d_loss: 0.5484 | g_loss: 0.2444\n",
      "Epoch [   46/   50] | d_loss: 0.5120 | g_loss: 0.2323\n",
      "Epoch [   46/   50] | d_loss: 0.5021 | g_loss: 0.2652\n",
      "Epoch [   46/   50] | d_loss: 0.4877 | g_loss: 0.2518\n",
      "Epoch [   46/   50] | d_loss: 0.5099 | g_loss: 0.2585\n",
      "Epoch [   46/   50] | d_loss: 0.5122 | g_loss: 0.2706\n",
      "Epoch [   46/   50] | d_loss: 0.5249 | g_loss: 0.2519\n",
      "Epoch [   46/   50] | d_loss: 0.5073 | g_loss: 0.2620\n",
      "Epoch [   46/   50] | d_loss: 0.4855 | g_loss: 0.2474\n",
      "Epoch [   46/   50] | d_loss: 0.5340 | g_loss: 0.2640\n",
      "Epoch [   46/   50] | d_loss: 0.4851 | g_loss: 0.2545\n",
      "Epoch [   46/   50] | d_loss: 0.5244 | g_loss: 0.2589\n",
      "Epoch [   46/   50] | d_loss: 0.5122 | g_loss: 0.2695\n",
      "Epoch [   46/   50] | d_loss: 0.5438 | g_loss: 0.2710\n",
      "Epoch [   46/   50] | d_loss: 0.5370 | g_loss: 0.2395\n",
      "Epoch [   46/   50] | d_loss: 0.4910 | g_loss: 0.2442\n",
      "Epoch [   46/   50] | d_loss: 0.4986 | g_loss: 0.2652\n",
      "Epoch [   46/   50] | d_loss: 0.5070 | g_loss: 0.2440\n",
      "Epoch [   46/   50] | d_loss: 0.5133 | g_loss: 0.2479\n",
      "Epoch [   46/   50] | d_loss: 0.5544 | g_loss: 0.2804\n",
      "Epoch [   46/   50] | d_loss: 0.4841 | g_loss: 0.2703\n",
      "Epoch [   46/   50] | d_loss: 0.5016 | g_loss: 0.2632\n",
      "Epoch [   46/   50] | d_loss: 0.5033 | g_loss: 0.2545\n",
      "Epoch [   46/   50] | d_loss: 0.5196 | g_loss: 0.2738\n",
      "Epoch [   46/   50] | d_loss: 0.5048 | g_loss: 0.2657\n",
      "Epoch [   46/   50] | d_loss: 0.5030 | g_loss: 0.2797\n",
      "Epoch [   46/   50] | d_loss: 0.4960 | g_loss: 0.2593\n",
      "Epoch [   46/   50] | d_loss: 0.4688 | g_loss: 0.2415\n",
      "Epoch [   46/   50] | d_loss: 0.5269 | g_loss: 0.2456\n",
      "Epoch [   46/   50] | d_loss: 0.4980 | g_loss: 0.2673\n",
      "Epoch [   46/   50] | d_loss: 0.5280 | g_loss: 0.2279\n",
      "Epoch [   46/   50] | d_loss: 0.5281 | g_loss: 0.2744\n",
      "Epoch [   46/   50] | d_loss: 0.4966 | g_loss: 0.2366\n",
      "Epoch [   46/   50] | d_loss: 0.5123 | g_loss: 0.2250\n",
      "Epoch [   46/   50] | d_loss: 0.4956 | g_loss: 0.2600\n",
      "Epoch [   46/   50] | d_loss: 0.4922 | g_loss: 0.2687\n",
      "Epoch [   46/   50] | d_loss: 0.5107 | g_loss: 0.2906\n",
      "Epoch [   46/   50] | d_loss: 0.4749 | g_loss: 0.2448\n",
      "Epoch [   46/   50] | d_loss: 0.5068 | g_loss: 0.3052\n",
      "Epoch [   46/   50] | d_loss: 0.4868 | g_loss: 0.2343\n",
      "Epoch [   46/   50] | d_loss: 0.5135 | g_loss: 0.2388\n",
      "Epoch [   46/   50] | d_loss: 0.5179 | g_loss: 0.2747\n",
      "Epoch [   46/   50] | d_loss: 0.5071 | g_loss: 0.2431\n",
      "Epoch [   46/   50] | d_loss: 0.4898 | g_loss: 0.2567\n",
      "Epoch [   46/   50] | d_loss: 0.4824 | g_loss: 0.2744\n",
      "Epoch [   46/   50] | d_loss: 0.5306 | g_loss: 0.2594\n",
      "Epoch [   46/   50] | d_loss: 0.5057 | g_loss: 0.2534\n",
      "Epoch [   46/   50] | d_loss: 0.5148 | g_loss: 0.2458\n",
      "Epoch [   46/   50] | d_loss: 0.4868 | g_loss: 0.2562\n",
      "Epoch [   46/   50] | d_loss: 0.5186 | g_loss: 0.2829\n",
      "Epoch [   46/   50] | d_loss: 0.5120 | g_loss: 0.2384\n",
      "Epoch [   46/   50] | d_loss: 0.5038 | g_loss: 0.2884\n",
      "Epoch [   46/   50] | d_loss: 0.4876 | g_loss: 0.2512\n",
      "Epoch [   46/   50] | d_loss: 0.4926 | g_loss: 0.2494\n",
      "Epoch [   46/   50] | d_loss: 0.4858 | g_loss: 0.2666\n",
      "Epoch [   46/   50] | d_loss: 0.4918 | g_loss: 0.2503\n",
      "Epoch [   46/   50] | d_loss: 0.5023 | g_loss: 0.2946\n",
      "Epoch [   46/   50] | d_loss: 0.5278 | g_loss: 0.2354\n",
      "Epoch [   46/   50] | d_loss: 0.5068 | g_loss: 0.2724\n",
      "Epoch [   46/   50] | d_loss: 0.4978 | g_loss: 0.2610\n",
      "Epoch [   46/   50] | d_loss: 0.4895 | g_loss: 0.2418\n",
      "Epoch [   46/   50] | d_loss: 0.4929 | g_loss: 0.2545\n",
      "Epoch [   46/   50] | d_loss: 0.5540 | g_loss: 0.2586\n",
      "Epoch [   46/   50] | d_loss: 0.5185 | g_loss: 0.2466\n",
      "Epoch [   46/   50] | d_loss: 0.4920 | g_loss: 0.2562\n",
      "Epoch [   46/   50] | d_loss: 0.5527 | g_loss: 0.2335\n",
      "Epoch [   47/   50] | d_loss: 0.4799 | g_loss: 0.2315\n",
      "Epoch [   47/   50] | d_loss: 0.5089 | g_loss: 0.2573\n",
      "Epoch [   47/   50] | d_loss: 0.4892 | g_loss: 0.2446\n",
      "Epoch [   47/   50] | d_loss: 0.5017 | g_loss: 0.2458\n",
      "Epoch [   47/   50] | d_loss: 0.5294 | g_loss: 0.2787\n",
      "Epoch [   47/   50] | d_loss: 0.4643 | g_loss: 0.2491\n",
      "Epoch [   47/   50] | d_loss: 0.5630 | g_loss: 0.2512\n",
      "Epoch [   47/   50] | d_loss: 0.5054 | g_loss: 0.2399\n",
      "Epoch [   47/   50] | d_loss: 0.5015 | g_loss: 0.2613\n",
      "Epoch [   47/   50] | d_loss: 0.4931 | g_loss: 0.2551\n",
      "Epoch [   47/   50] | d_loss: 0.5065 | g_loss: 0.2483\n",
      "Epoch [   47/   50] | d_loss: 0.5148 | g_loss: 0.2455\n",
      "Epoch [   47/   50] | d_loss: 0.4869 | g_loss: 0.2673\n",
      "Epoch [   47/   50] | d_loss: 0.5270 | g_loss: 0.2634\n",
      "Epoch [   47/   50] | d_loss: 0.5494 | g_loss: 0.2604\n",
      "Epoch [   47/   50] | d_loss: 0.4858 | g_loss: 0.2696\n",
      "Epoch [   47/   50] | d_loss: 0.5039 | g_loss: 0.2636\n",
      "Epoch [   47/   50] | d_loss: 0.4894 | g_loss: 0.2334\n",
      "Epoch [   47/   50] | d_loss: 0.5101 | g_loss: 0.2561\n",
      "Epoch [   47/   50] | d_loss: 0.5015 | g_loss: 0.2409\n",
      "Epoch [   47/   50] | d_loss: 0.5085 | g_loss: 0.2816\n",
      "Epoch [   47/   50] | d_loss: 0.5020 | g_loss: 0.2653\n",
      "Epoch [   47/   50] | d_loss: 0.4559 | g_loss: 0.2707\n",
      "Epoch [   47/   50] | d_loss: 0.5132 | g_loss: 0.2520\n",
      "Epoch [   47/   50] | d_loss: 0.5265 | g_loss: 0.2397\n",
      "Epoch [   47/   50] | d_loss: 0.4929 | g_loss: 0.2475\n",
      "Epoch [   47/   50] | d_loss: 0.5268 | g_loss: 0.2691\n",
      "Epoch [   47/   50] | d_loss: 0.4905 | g_loss: 0.2672\n",
      "Epoch [   47/   50] | d_loss: 0.4724 | g_loss: 0.2655\n",
      "Epoch [   47/   50] | d_loss: 0.5319 | g_loss: 0.2491\n",
      "Epoch [   47/   50] | d_loss: 0.5279 | g_loss: 0.2569\n",
      "Epoch [   47/   50] | d_loss: 0.5126 | g_loss: 0.2535\n",
      "Epoch [   47/   50] | d_loss: 0.4779 | g_loss: 0.2325\n",
      "Epoch [   47/   50] | d_loss: 0.4893 | g_loss: 0.2713\n",
      "Epoch [   47/   50] | d_loss: 0.4957 | g_loss: 0.2326\n",
      "Epoch [   47/   50] | d_loss: 0.5109 | g_loss: 0.2691\n",
      "Epoch [   47/   50] | d_loss: 0.4424 | g_loss: 0.2495\n",
      "Epoch [   47/   50] | d_loss: 0.5318 | g_loss: 0.2578\n",
      "Epoch [   47/   50] | d_loss: 0.4854 | g_loss: 0.2617\n",
      "Epoch [   47/   50] | d_loss: 0.5162 | g_loss: 0.2531\n",
      "Epoch [   47/   50] | d_loss: 0.5245 | g_loss: 0.2622\n",
      "Epoch [   47/   50] | d_loss: 0.4965 | g_loss: 0.2586\n",
      "Epoch [   47/   50] | d_loss: 0.5341 | g_loss: 0.2095\n",
      "Epoch [   47/   50] | d_loss: 0.4689 | g_loss: 0.2554\n",
      "Epoch [   47/   50] | d_loss: 0.4746 | g_loss: 0.2626\n",
      "Epoch [   47/   50] | d_loss: 0.5050 | g_loss: 0.2594\n",
      "Epoch [   47/   50] | d_loss: 0.4717 | g_loss: 0.2614\n",
      "Epoch [   47/   50] | d_loss: 0.5113 | g_loss: 0.2545\n",
      "Epoch [   47/   50] | d_loss: 0.5086 | g_loss: 0.2705\n",
      "Epoch [   47/   50] | d_loss: 0.5090 | g_loss: 0.2818\n",
      "Epoch [   47/   50] | d_loss: 0.4682 | g_loss: 0.2935\n",
      "Epoch [   47/   50] | d_loss: 0.4990 | g_loss: 0.2464\n",
      "Epoch [   47/   50] | d_loss: 0.4827 | g_loss: 0.2665\n",
      "Epoch [   47/   50] | d_loss: 0.4963 | g_loss: 0.2592\n",
      "Epoch [   47/   50] | d_loss: 0.5283 | g_loss: 0.2529\n",
      "Epoch [   47/   50] | d_loss: 0.5420 | g_loss: 0.2432\n",
      "Epoch [   47/   50] | d_loss: 0.5041 | g_loss: 0.2470\n",
      "Epoch [   47/   50] | d_loss: 0.4994 | g_loss: 0.2397\n",
      "Epoch [   47/   50] | d_loss: 0.5099 | g_loss: 0.2530\n",
      "Epoch [   47/   50] | d_loss: 0.5029 | g_loss: 0.2779\n",
      "Epoch [   47/   50] | d_loss: 0.4648 | g_loss: 0.2371\n",
      "Epoch [   47/   50] | d_loss: 0.4993 | g_loss: 0.2586\n",
      "Epoch [   47/   50] | d_loss: 0.5250 | g_loss: 0.2752\n",
      "Epoch [   47/   50] | d_loss: 0.4607 | g_loss: 0.2502\n",
      "Epoch [   47/   50] | d_loss: 0.4993 | g_loss: 0.2406\n",
      "Epoch [   47/   50] | d_loss: 0.5395 | g_loss: 0.2514\n",
      "Epoch [   47/   50] | d_loss: 0.4867 | g_loss: 0.2736\n",
      "Epoch [   47/   50] | d_loss: 0.4861 | g_loss: 0.2929\n",
      "Epoch [   47/   50] | d_loss: 0.4904 | g_loss: 0.2427\n",
      "Epoch [   47/   50] | d_loss: 0.5102 | g_loss: 0.2504\n",
      "Epoch [   47/   50] | d_loss: 0.5050 | g_loss: 0.2394\n",
      "Epoch [   47/   50] | d_loss: 0.5306 | g_loss: 0.2432\n",
      "Epoch [   47/   50] | d_loss: 0.4892 | g_loss: 0.2596\n",
      "Epoch [   47/   50] | d_loss: 0.4748 | g_loss: 0.2669\n",
      "Epoch [   47/   50] | d_loss: 0.5088 | g_loss: 0.2449\n",
      "Epoch [   47/   50] | d_loss: 0.4746 | g_loss: 0.2391\n",
      "Epoch [   47/   50] | d_loss: 0.5070 | g_loss: 0.2611\n",
      "Epoch [   47/   50] | d_loss: 0.5057 | g_loss: 0.2741\n",
      "Epoch [   47/   50] | d_loss: 0.5243 | g_loss: 0.2475\n",
      "Epoch [   47/   50] | d_loss: 0.4830 | g_loss: 0.2604\n",
      "Epoch [   47/   50] | d_loss: 0.4693 | g_loss: 0.2732\n",
      "Epoch [   47/   50] | d_loss: 0.4995 | g_loss: 0.2899\n",
      "Epoch [   47/   50] | d_loss: 0.5376 | g_loss: 0.2593\n",
      "Epoch [   47/   50] | d_loss: 0.5011 | g_loss: 0.2508\n",
      "Epoch [   47/   50] | d_loss: 0.5028 | g_loss: 0.2763\n",
      "Epoch [   47/   50] | d_loss: 0.5272 | g_loss: 0.2752\n",
      "Epoch [   47/   50] | d_loss: 0.4765 | g_loss: 0.2460\n",
      "Epoch [   47/   50] | d_loss: 0.5042 | g_loss: 0.3148\n",
      "Epoch [   47/   50] | d_loss: 0.5009 | g_loss: 0.2648\n",
      "Epoch [   47/   50] | d_loss: 0.4860 | g_loss: 0.2554\n",
      "Epoch [   48/   50] | d_loss: 0.5310 | g_loss: 0.2565\n",
      "Epoch [   48/   50] | d_loss: 0.4676 | g_loss: 0.2458\n",
      "Epoch [   48/   50] | d_loss: 0.5093 | g_loss: 0.2731\n",
      "Epoch [   48/   50] | d_loss: 0.5093 | g_loss: 0.2558\n",
      "Epoch [   48/   50] | d_loss: 0.4778 | g_loss: 0.2434\n",
      "Epoch [   48/   50] | d_loss: 0.4854 | g_loss: 0.2714\n",
      "Epoch [   48/   50] | d_loss: 0.5045 | g_loss: 0.2434\n",
      "Epoch [   48/   50] | d_loss: 0.4895 | g_loss: 0.2502\n",
      "Epoch [   48/   50] | d_loss: 0.4858 | g_loss: 0.2617\n",
      "Epoch [   48/   50] | d_loss: 0.5020 | g_loss: 0.2783\n",
      "Epoch [   48/   50] | d_loss: 0.4975 | g_loss: 0.2619\n",
      "Epoch [   48/   50] | d_loss: 0.4969 | g_loss: 0.2532\n",
      "Epoch [   48/   50] | d_loss: 0.5254 | g_loss: 0.2329\n",
      "Epoch [   48/   50] | d_loss: 0.5055 | g_loss: 0.2583\n",
      "Epoch [   48/   50] | d_loss: 0.5375 | g_loss: 0.2448\n",
      "Epoch [   48/   50] | d_loss: 0.5253 | g_loss: 0.2399\n",
      "Epoch [   48/   50] | d_loss: 0.4595 | g_loss: 0.2612\n",
      "Epoch [   48/   50] | d_loss: 0.4990 | g_loss: 0.2524\n",
      "Epoch [   48/   50] | d_loss: 0.4647 | g_loss: 0.2286\n",
      "Epoch [   48/   50] | d_loss: 0.5083 | g_loss: 0.2463\n",
      "Epoch [   48/   50] | d_loss: 0.5128 | g_loss: 0.2471\n",
      "Epoch [   48/   50] | d_loss: 0.5317 | g_loss: 0.2691\n",
      "Epoch [   48/   50] | d_loss: 0.5038 | g_loss: 0.2613\n",
      "Epoch [   48/   50] | d_loss: 0.4837 | g_loss: 0.2477\n",
      "Epoch [   48/   50] | d_loss: 0.5214 | g_loss: 0.2519\n",
      "Epoch [   48/   50] | d_loss: 0.4873 | g_loss: 0.2361\n",
      "Epoch [   48/   50] | d_loss: 0.5215 | g_loss: 0.2547\n",
      "Epoch [   48/   50] | d_loss: 0.5123 | g_loss: 0.2373\n",
      "Epoch [   48/   50] | d_loss: 0.4901 | g_loss: 0.2655\n",
      "Epoch [   48/   50] | d_loss: 0.5190 | g_loss: 0.2921\n",
      "Epoch [   48/   50] | d_loss: 0.5375 | g_loss: 0.2534\n",
      "Epoch [   48/   50] | d_loss: 0.4962 | g_loss: 0.2458\n",
      "Epoch [   48/   50] | d_loss: 0.4905 | g_loss: 0.2559\n",
      "Epoch [   48/   50] | d_loss: 0.5007 | g_loss: 0.2446\n",
      "Epoch [   48/   50] | d_loss: 0.4904 | g_loss: 0.2418\n",
      "Epoch [   48/   50] | d_loss: 0.5239 | g_loss: 0.2668\n",
      "Epoch [   48/   50] | d_loss: 0.5016 | g_loss: 0.2352\n",
      "Epoch [   48/   50] | d_loss: 0.4738 | g_loss: 0.2344\n",
      "Epoch [   48/   50] | d_loss: 0.5181 | g_loss: 0.2682\n",
      "Epoch [   48/   50] | d_loss: 0.5174 | g_loss: 0.2522\n",
      "Epoch [   48/   50] | d_loss: 0.5387 | g_loss: 0.2561\n",
      "Epoch [   48/   50] | d_loss: 0.4868 | g_loss: 0.2690\n",
      "Epoch [   48/   50] | d_loss: 0.4939 | g_loss: 0.2494\n",
      "Epoch [   48/   50] | d_loss: 0.4961 | g_loss: 0.2906\n",
      "Epoch [   48/   50] | d_loss: 0.5018 | g_loss: 0.2842\n",
      "Epoch [   48/   50] | d_loss: 0.5039 | g_loss: 0.2670\n",
      "Epoch [   48/   50] | d_loss: 0.4706 | g_loss: 0.2513\n",
      "Epoch [   48/   50] | d_loss: 0.5065 | g_loss: 0.2685\n",
      "Epoch [   48/   50] | d_loss: 0.5066 | g_loss: 0.2526\n",
      "Epoch [   48/   50] | d_loss: 0.4781 | g_loss: 0.2651\n",
      "Epoch [   48/   50] | d_loss: 0.5137 | g_loss: 0.2384\n",
      "Epoch [   48/   50] | d_loss: 0.5349 | g_loss: 0.2629\n",
      "Epoch [   48/   50] | d_loss: 0.4974 | g_loss: 0.2843\n",
      "Epoch [   48/   50] | d_loss: 0.4923 | g_loss: 0.2466\n",
      "Epoch [   48/   50] | d_loss: 0.4832 | g_loss: 0.3011\n",
      "Epoch [   48/   50] | d_loss: 0.5192 | g_loss: 0.2384\n",
      "Epoch [   48/   50] | d_loss: 0.5043 | g_loss: 0.2301\n",
      "Epoch [   48/   50] | d_loss: 0.5373 | g_loss: 0.2291\n",
      "Epoch [   48/   50] | d_loss: 0.4824 | g_loss: 0.3002\n",
      "Epoch [   48/   50] | d_loss: 0.4848 | g_loss: 0.2628\n",
      "Epoch [   48/   50] | d_loss: 0.5041 | g_loss: 0.2440\n",
      "Epoch [   48/   50] | d_loss: 0.5079 | g_loss: 0.2450\n",
      "Epoch [   48/   50] | d_loss: 0.4637 | g_loss: 0.2519\n",
      "Epoch [   48/   50] | d_loss: 0.5117 | g_loss: 0.2715\n",
      "Epoch [   48/   50] | d_loss: 0.4938 | g_loss: 0.2353\n",
      "Epoch [   48/   50] | d_loss: 0.4761 | g_loss: 0.2389\n",
      "Epoch [   48/   50] | d_loss: 0.5095 | g_loss: 0.2606\n",
      "Epoch [   48/   50] | d_loss: 0.5158 | g_loss: 0.2334\n",
      "Epoch [   48/   50] | d_loss: 0.4990 | g_loss: 0.2376\n",
      "Epoch [   48/   50] | d_loss: 0.4936 | g_loss: 0.2612\n",
      "Epoch [   48/   50] | d_loss: 0.4621 | g_loss: 0.2745\n",
      "Epoch [   48/   50] | d_loss: 0.4950 | g_loss: 0.2521\n",
      "Epoch [   48/   50] | d_loss: 0.5368 | g_loss: 0.2913\n",
      "Epoch [   48/   50] | d_loss: 0.4909 | g_loss: 0.2569\n",
      "Epoch [   48/   50] | d_loss: 0.5213 | g_loss: 0.2670\n",
      "Epoch [   48/   50] | d_loss: 0.4876 | g_loss: 0.2427\n",
      "Epoch [   48/   50] | d_loss: 0.5097 | g_loss: 0.2696\n",
      "Epoch [   48/   50] | d_loss: 0.4909 | g_loss: 0.2595\n",
      "Epoch [   48/   50] | d_loss: 0.5100 | g_loss: 0.2605\n",
      "Epoch [   48/   50] | d_loss: 0.5530 | g_loss: 0.2579\n",
      "Epoch [   48/   50] | d_loss: 0.5107 | g_loss: 0.2522\n",
      "Epoch [   48/   50] | d_loss: 0.5125 | g_loss: 0.2331\n",
      "Epoch [   48/   50] | d_loss: 0.5206 | g_loss: 0.2581\n",
      "Epoch [   48/   50] | d_loss: 0.4790 | g_loss: 0.2496\n",
      "Epoch [   48/   50] | d_loss: 0.4843 | g_loss: 0.2359\n",
      "Epoch [   48/   50] | d_loss: 0.5229 | g_loss: 0.2606\n",
      "Epoch [   48/   50] | d_loss: 0.5138 | g_loss: 0.2611\n",
      "Epoch [   48/   50] | d_loss: 0.5225 | g_loss: 0.2469\n",
      "Epoch [   48/   50] | d_loss: 0.4909 | g_loss: 0.2578\n",
      "Epoch [   48/   50] | d_loss: 0.5189 | g_loss: 0.2490\n",
      "Epoch [   49/   50] | d_loss: 0.5104 | g_loss: 0.2361\n",
      "Epoch [   49/   50] | d_loss: 0.4738 | g_loss: 0.2670\n",
      "Epoch [   49/   50] | d_loss: 0.5212 | g_loss: 0.2492\n",
      "Epoch [   49/   50] | d_loss: 0.4984 | g_loss: 0.2394\n",
      "Epoch [   49/   50] | d_loss: 0.5265 | g_loss: 0.2236\n",
      "Epoch [   49/   50] | d_loss: 0.5062 | g_loss: 0.2924\n",
      "Epoch [   49/   50] | d_loss: 0.4794 | g_loss: 0.2143\n",
      "Epoch [   49/   50] | d_loss: 0.4959 | g_loss: 0.2563\n",
      "Epoch [   49/   50] | d_loss: 0.5062 | g_loss: 0.2493\n",
      "Epoch [   49/   50] | d_loss: 0.4776 | g_loss: 0.2550\n",
      "Epoch [   49/   50] | d_loss: 0.5258 | g_loss: 0.2798\n",
      "Epoch [   49/   50] | d_loss: 0.5124 | g_loss: 0.2591\n",
      "Epoch [   49/   50] | d_loss: 0.4841 | g_loss: 0.2736\n",
      "Epoch [   49/   50] | d_loss: 0.4857 | g_loss: 0.3020\n",
      "Epoch [   49/   50] | d_loss: 0.4819 | g_loss: 0.2654\n",
      "Epoch [   49/   50] | d_loss: 0.5088 | g_loss: 0.2500\n",
      "Epoch [   49/   50] | d_loss: 0.5116 | g_loss: 0.2609\n",
      "Epoch [   49/   50] | d_loss: 0.5114 | g_loss: 0.2537\n",
      "Epoch [   49/   50] | d_loss: 0.5293 | g_loss: 0.2474\n",
      "Epoch [   49/   50] | d_loss: 0.4989 | g_loss: 0.2605\n",
      "Epoch [   49/   50] | d_loss: 0.4993 | g_loss: 0.2497\n",
      "Epoch [   49/   50] | d_loss: 0.4983 | g_loss: 0.2777\n",
      "Epoch [   49/   50] | d_loss: 0.5001 | g_loss: 0.2592\n",
      "Epoch [   49/   50] | d_loss: 0.4971 | g_loss: 0.2760\n",
      "Epoch [   49/   50] | d_loss: 0.4775 | g_loss: 0.2403\n",
      "Epoch [   49/   50] | d_loss: 0.5624 | g_loss: 0.2650\n",
      "Epoch [   49/   50] | d_loss: 0.4964 | g_loss: 0.2523\n",
      "Epoch [   49/   50] | d_loss: 0.4880 | g_loss: 0.2665\n",
      "Epoch [   49/   50] | d_loss: 0.5140 | g_loss: 0.2336\n",
      "Epoch [   49/   50] | d_loss: 0.4832 | g_loss: 0.2766\n",
      "Epoch [   49/   50] | d_loss: 0.5266 | g_loss: 0.2690\n",
      "Epoch [   49/   50] | d_loss: 0.5448 | g_loss: 0.2455\n",
      "Epoch [   49/   50] | d_loss: 0.4750 | g_loss: 0.2597\n",
      "Epoch [   49/   50] | d_loss: 0.5309 | g_loss: 0.2583\n",
      "Epoch [   49/   50] | d_loss: 0.5035 | g_loss: 0.2804\n",
      "Epoch [   49/   50] | d_loss: 0.5032 | g_loss: 0.2782\n",
      "Epoch [   49/   50] | d_loss: 0.5094 | g_loss: 0.2849\n",
      "Epoch [   49/   50] | d_loss: 0.5149 | g_loss: 0.2643\n",
      "Epoch [   49/   50] | d_loss: 0.4943 | g_loss: 0.2583\n",
      "Epoch [   49/   50] | d_loss: 0.4786 | g_loss: 0.2248\n",
      "Epoch [   49/   50] | d_loss: 0.4965 | g_loss: 0.2434\n",
      "Epoch [   49/   50] | d_loss: 0.5144 | g_loss: 0.2596\n",
      "Epoch [   49/   50] | d_loss: 0.5384 | g_loss: 0.2566\n",
      "Epoch [   49/   50] | d_loss: 0.4833 | g_loss: 0.2762\n",
      "Epoch [   49/   50] | d_loss: 0.5145 | g_loss: 0.2450\n",
      "Epoch [   49/   50] | d_loss: 0.4976 | g_loss: 0.2755\n",
      "Epoch [   49/   50] | d_loss: 0.5220 | g_loss: 0.2572\n",
      "Epoch [   49/   50] | d_loss: 0.5058 | g_loss: 0.2543\n",
      "Epoch [   49/   50] | d_loss: 0.5152 | g_loss: 0.2798\n",
      "Epoch [   49/   50] | d_loss: 0.5077 | g_loss: 0.2483\n",
      "Epoch [   49/   50] | d_loss: 0.4906 | g_loss: 0.2809\n",
      "Epoch [   49/   50] | d_loss: 0.4663 | g_loss: 0.2359\n",
      "Epoch [   49/   50] | d_loss: 0.5215 | g_loss: 0.2684\n",
      "Epoch [   49/   50] | d_loss: 0.4904 | g_loss: 0.2566\n",
      "Epoch [   49/   50] | d_loss: 0.4885 | g_loss: 0.2191\n",
      "Epoch [   49/   50] | d_loss: 0.5312 | g_loss: 0.2604\n",
      "Epoch [   49/   50] | d_loss: 0.5218 | g_loss: 0.2378\n",
      "Epoch [   49/   50] | d_loss: 0.5118 | g_loss: 0.2667\n",
      "Epoch [   49/   50] | d_loss: 0.5067 | g_loss: 0.2928\n",
      "Epoch [   49/   50] | d_loss: 0.4807 | g_loss: 0.2861\n",
      "Epoch [   49/   50] | d_loss: 0.5314 | g_loss: 0.2616\n",
      "Epoch [   49/   50] | d_loss: 0.5135 | g_loss: 0.2642\n",
      "Epoch [   49/   50] | d_loss: 0.5235 | g_loss: 0.2644\n",
      "Epoch [   49/   50] | d_loss: 0.5096 | g_loss: 0.2575\n",
      "Epoch [   49/   50] | d_loss: 0.4982 | g_loss: 0.2256\n",
      "Epoch [   49/   50] | d_loss: 0.4853 | g_loss: 0.2898\n",
      "Epoch [   49/   50] | d_loss: 0.5172 | g_loss: 0.2976\n",
      "Epoch [   49/   50] | d_loss: 0.5136 | g_loss: 0.2242\n",
      "Epoch [   49/   50] | d_loss: 0.5391 | g_loss: 0.2270\n",
      "Epoch [   49/   50] | d_loss: 0.4806 | g_loss: 0.2888\n",
      "Epoch [   49/   50] | d_loss: 0.5063 | g_loss: 0.2548\n",
      "Epoch [   49/   50] | d_loss: 0.4907 | g_loss: 0.2468\n",
      "Epoch [   49/   50] | d_loss: 0.4948 | g_loss: 0.2523\n",
      "Epoch [   49/   50] | d_loss: 0.4873 | g_loss: 0.2715\n",
      "Epoch [   49/   50] | d_loss: 0.5237 | g_loss: 0.2232\n",
      "Epoch [   49/   50] | d_loss: 0.5239 | g_loss: 0.2357\n",
      "Epoch [   49/   50] | d_loss: 0.4621 | g_loss: 0.2492\n",
      "Epoch [   49/   50] | d_loss: 0.5240 | g_loss: 0.2725\n",
      "Epoch [   49/   50] | d_loss: 0.5094 | g_loss: 0.2234\n",
      "Epoch [   49/   50] | d_loss: 0.4791 | g_loss: 0.2500\n",
      "Epoch [   49/   50] | d_loss: 0.4953 | g_loss: 0.2443\n",
      "Epoch [   49/   50] | d_loss: 0.5084 | g_loss: 0.2539\n",
      "Epoch [   49/   50] | d_loss: 0.4901 | g_loss: 0.2445\n",
      "Epoch [   49/   50] | d_loss: 0.5103 | g_loss: 0.2698\n",
      "Epoch [   49/   50] | d_loss: 0.4595 | g_loss: 0.2676\n",
      "Epoch [   49/   50] | d_loss: 0.5157 | g_loss: 0.2305\n",
      "Epoch [   49/   50] | d_loss: 0.5002 | g_loss: 0.2440\n",
      "Epoch [   49/   50] | d_loss: 0.5206 | g_loss: 0.2733\n",
      "Epoch [   49/   50] | d_loss: 0.4749 | g_loss: 0.2624\n",
      "Epoch [   49/   50] | d_loss: 0.4865 | g_loss: 0.2589\n",
      "Epoch [   50/   50] | d_loss: 0.5005 | g_loss: 0.2717\n",
      "Epoch [   50/   50] | d_loss: 0.5049 | g_loss: 0.2526\n",
      "Epoch [   50/   50] | d_loss: 0.5137 | g_loss: 0.2549\n",
      "Epoch [   50/   50] | d_loss: 0.4982 | g_loss: 0.2363\n",
      "Epoch [   50/   50] | d_loss: 0.5404 | g_loss: 0.2487\n",
      "Epoch [   50/   50] | d_loss: 0.4773 | g_loss: 0.2499\n",
      "Epoch [   50/   50] | d_loss: 0.4895 | g_loss: 0.2588\n",
      "Epoch [   50/   50] | d_loss: 0.5224 | g_loss: 0.2605\n",
      "Epoch [   50/   50] | d_loss: 0.4968 | g_loss: 0.2465\n",
      "Epoch [   50/   50] | d_loss: 0.4627 | g_loss: 0.2682\n",
      "Epoch [   50/   50] | d_loss: 0.4973 | g_loss: 0.2594\n",
      "Epoch [   50/   50] | d_loss: 0.5148 | g_loss: 0.2632\n",
      "Epoch [   50/   50] | d_loss: 0.5166 | g_loss: 0.2697\n",
      "Epoch [   50/   50] | d_loss: 0.5209 | g_loss: 0.2533\n",
      "Epoch [   50/   50] | d_loss: 0.4980 | g_loss: 0.2588\n",
      "Epoch [   50/   50] | d_loss: 0.5331 | g_loss: 0.2546\n",
      "Epoch [   50/   50] | d_loss: 0.5348 | g_loss: 0.2536\n",
      "Epoch [   50/   50] | d_loss: 0.5033 | g_loss: 0.2589\n",
      "Epoch [   50/   50] | d_loss: 0.5110 | g_loss: 0.2507\n",
      "Epoch [   50/   50] | d_loss: 0.5031 | g_loss: 0.2598\n",
      "Epoch [   50/   50] | d_loss: 0.4995 | g_loss: 0.2485\n",
      "Epoch [   50/   50] | d_loss: 0.4952 | g_loss: 0.2227\n",
      "Epoch [   50/   50] | d_loss: 0.4847 | g_loss: 0.2505\n",
      "Epoch [   50/   50] | d_loss: 0.5195 | g_loss: 0.2866\n",
      "Epoch [   50/   50] | d_loss: 0.5012 | g_loss: 0.2840\n",
      "Epoch [   50/   50] | d_loss: 0.5011 | g_loss: 0.2432\n",
      "Epoch [   50/   50] | d_loss: 0.4968 | g_loss: 0.2513\n",
      "Epoch [   50/   50] | d_loss: 0.5037 | g_loss: 0.2411\n",
      "Epoch [   50/   50] | d_loss: 0.5414 | g_loss: 0.2611\n",
      "Epoch [   50/   50] | d_loss: 0.4661 | g_loss: 0.2674\n",
      "Epoch [   50/   50] | d_loss: 0.5337 | g_loss: 0.2252\n",
      "Epoch [   50/   50] | d_loss: 0.5218 | g_loss: 0.2586\n",
      "Epoch [   50/   50] | d_loss: 0.5179 | g_loss: 0.2584\n",
      "Epoch [   50/   50] | d_loss: 0.4822 | g_loss: 0.2793\n",
      "Epoch [   50/   50] | d_loss: 0.5090 | g_loss: 0.2702\n",
      "Epoch [   50/   50] | d_loss: 0.4977 | g_loss: 0.2797\n",
      "Epoch [   50/   50] | d_loss: 0.5191 | g_loss: 0.2520\n",
      "Epoch [   50/   50] | d_loss: 0.4886 | g_loss: 0.2584\n",
      "Epoch [   50/   50] | d_loss: 0.5111 | g_loss: 0.2876\n",
      "Epoch [   50/   50] | d_loss: 0.5009 | g_loss: 0.2567\n",
      "Epoch [   50/   50] | d_loss: 0.4874 | g_loss: 0.2742\n",
      "Epoch [   50/   50] | d_loss: 0.5412 | g_loss: 0.2349\n",
      "Epoch [   50/   50] | d_loss: 0.5066 | g_loss: 0.2325\n",
      "Epoch [   50/   50] | d_loss: 0.4842 | g_loss: 0.2629\n",
      "Epoch [   50/   50] | d_loss: 0.5149 | g_loss: 0.2895\n",
      "Epoch [   50/   50] | d_loss: 0.5039 | g_loss: 0.2368\n",
      "Epoch [   50/   50] | d_loss: 0.5144 | g_loss: 0.2498\n",
      "Epoch [   50/   50] | d_loss: 0.5385 | g_loss: 0.2934\n",
      "Epoch [   50/   50] | d_loss: 0.5119 | g_loss: 0.2829\n",
      "Epoch [   50/   50] | d_loss: 0.4794 | g_loss: 0.2740\n",
      "Epoch [   50/   50] | d_loss: 0.4543 | g_loss: 0.2768\n",
      "Epoch [   50/   50] | d_loss: 0.5072 | g_loss: 0.2451\n",
      "Epoch [   50/   50] | d_loss: 0.5087 | g_loss: 0.2691\n",
      "Epoch [   50/   50] | d_loss: 0.5101 | g_loss: 0.2673\n",
      "Epoch [   50/   50] | d_loss: 0.4935 | g_loss: 0.2731\n",
      "Epoch [   50/   50] | d_loss: 0.4900 | g_loss: 0.2456\n",
      "Epoch [   50/   50] | d_loss: 0.5067 | g_loss: 0.2555\n",
      "Epoch [   50/   50] | d_loss: 0.4907 | g_loss: 0.2725\n",
      "Epoch [   50/   50] | d_loss: 0.5006 | g_loss: 0.2451\n",
      "Epoch [   50/   50] | d_loss: 0.5124 | g_loss: 0.2745\n",
      "Epoch [   50/   50] | d_loss: 0.4895 | g_loss: 0.2683\n",
      "Epoch [   50/   50] | d_loss: 0.5189 | g_loss: 0.2688\n",
      "Epoch [   50/   50] | d_loss: 0.5175 | g_loss: 0.2411\n",
      "Epoch [   50/   50] | d_loss: 0.4840 | g_loss: 0.2444\n",
      "Epoch [   50/   50] | d_loss: 0.5068 | g_loss: 0.2371\n",
      "Epoch [   50/   50] | d_loss: 0.4783 | g_loss: 0.2641\n",
      "Epoch [   50/   50] | d_loss: 0.5375 | g_loss: 0.2795\n",
      "Epoch [   50/   50] | d_loss: 0.4723 | g_loss: 0.2573\n",
      "Epoch [   50/   50] | d_loss: 0.5192 | g_loss: 0.2483\n",
      "Epoch [   50/   50] | d_loss: 0.5081 | g_loss: 0.2765\n",
      "Epoch [   50/   50] | d_loss: 0.5253 | g_loss: 0.2529\n",
      "Epoch [   50/   50] | d_loss: 0.5305 | g_loss: 0.2147\n",
      "Epoch [   50/   50] | d_loss: 0.5085 | g_loss: 0.3054\n",
      "Epoch [   50/   50] | d_loss: 0.4855 | g_loss: 0.2541\n",
      "Epoch [   50/   50] | d_loss: 0.5267 | g_loss: 0.2819\n",
      "Epoch [   50/   50] | d_loss: 0.5311 | g_loss: 0.2256\n",
      "Epoch [   50/   50] | d_loss: 0.5217 | g_loss: 0.2475\n",
      "Epoch [   50/   50] | d_loss: 0.5134 | g_loss: 0.2321\n",
      "Epoch [   50/   50] | d_loss: 0.5205 | g_loss: 0.2671\n",
      "Epoch [   50/   50] | d_loss: 0.4870 | g_loss: 0.2455\n",
      "Epoch [   50/   50] | d_loss: 0.5070 | g_loss: 0.2442\n",
      "Epoch [   50/   50] | d_loss: 0.4864 | g_loss: 0.2408\n",
      "Epoch [   50/   50] | d_loss: 0.4799 | g_loss: 0.2719\n",
      "Epoch [   50/   50] | d_loss: 0.4920 | g_loss: 0.2558\n",
      "Epoch [   50/   50] | d_loss: 0.5368 | g_loss: 0.2302\n",
      "Epoch [   50/   50] | d_loss: 0.5021 | g_loss: 0.2614\n",
      "Epoch [   50/   50] | d_loss: 0.5143 | g_loss: 0.2157\n",
      "Epoch [   50/   50] | d_loss: 0.4825 | g_loss: 0.2596\n",
      "Epoch [   50/   50] | d_loss: 0.5349 | g_loss: 0.2746\n",
      "Epoch [   50/   50] | d_loss: 0.4815 | g_loss: 0.2675\n"
     ]
    }
   ],
   "source": [
    "from workspace_utils import active_session\n",
    "\n",
    "# set number of epochs \n",
    "n_epochs = 50\n",
    "\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL\n",
    "\"\"\"\n",
    "# call training function\n",
    "\n",
    "with active_session():\n",
    "    losses = train(D, G, n_epochs=n_epochs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Training loss\n",
    "\n",
    "Plot the training losses for the generator and discriminator, recorded after each epoch."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f69c4e48630>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f69ca3fe7b8>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "losses = np.array(losses)\n",
    "plt.plot(losses.T[0], label='Discriminator', alpha=0.5)\n",
    "plt.plot(losses.T[1], label='Generator', alpha=0.5)\n",
    "plt.title(\"Training Losses\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Generator samples from training\n",
    "\n",
    "View samples of images from the generator, and answer a question about the strengths and weaknesses of your trained models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# helper function for viewing a list of passed in sample images\n",
    "def view_samples(epoch, samples):\n",
    "    fig, axes = plt.subplots(figsize=(16,4), nrows=2, ncols=8, sharey=True, sharex=True)\n",
    "    for ax, img in zip(axes.flatten(), samples[epoch]):\n",
    "        img = img.detach().cpu().numpy()\n",
    "        img = np.transpose(img, (1, 2, 0))\n",
    "        img = ((img + 1)*255 / (2)).astype(np.uint8)\n",
    "        ax.xaxis.set_visible(False)\n",
    "        ax.yaxis.set_visible(False)\n",
    "        im = ax.imshow(img.reshape((32,32,3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load samples from generator, taken while training\n",
    "with open('train_samples.pkl', 'rb') as f:\n",
    "    samples = pkl.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f69ca3f9668>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "_ = view_samples(-1, samples)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Question: What do you notice about your generated samples and how might you improve this model?\n",
    "When you answer this question, consider the following factors:\n",
    "* The dataset is biased; it is made of \"celebrity\" faces that are mostly white\n",
    "* Model size; larger models have the opportunity to learn more features in a data feature space\n",
    "* Optimization strategy; optimizers and number of epochs affect your final result\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Answer:** (Write your answer in this cell)\n",
    "First of all, the resolution of the generated samples is very low. Ofcourse, we can't expect a high resolution since our input images where only of size 32x32. So, one option for getting better generated images would be to feed the discriminator high resolution images. We would then also have to make the generator and discriminator deeper (i.e. choose a larger number of convolutional layers for the two models) so that it is possible for those two networks to learn more detailed high-resolution features; and really benefit from the higher quality of the input in the end.\n",
    "\n",
    "Then ofcourse, since the discriminator is just fed images of mostly white people we can't expect the generator to generate average human faces. If we wanted the generator to do this we would have to find a dataset of images of human faces that show random people from the whole human population in the world.\n",
    "\n",
    "I trained the GAN for 10 and 50 epochs and the generated sample images got better with increasing number of training epochs. I would therefore suggest to train for an even longer time. There are still some images in the generated samples which are quite far away from realistic human faces. With longer training time these bad results would probably not show up any more.\n",
    "\n",
    "Also, the hyper-parameters for the Adam optimizers for the generator and discriminator are chosen analogously to the examples we saw in the lecture on DCGANs. But maybe there are better choices for our specific task of generating face images. This is something one would have to try out and investigate more carefully but I would guess that there is some room for improvement there."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Submitting This Project\n",
    "When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as \"dlnd_face_generation.ipynb\" and save it as a HTML file under \"File\" -> \"Download as\". Include the \"problem_unittests.py\" files in your submission."
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}