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": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Archive:  processed_celeba_small.zip\n",
      "replace processed_celeba_small/.DS_Store? [y]es, [n]o, [A]ll, [N]one, [r]ename: ^C\n"
     ]
    }
   ],
   "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 torch.utils.data import DataLoader\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",
    "    # resize and normalize the images\n",
    "    transform = transforms.Compose([transforms.Resize(image_size), # resize to 128x128\n",
    "                                    transforms.ToTensor()])\n",
    "    \n",
    "    # define datasets using ImageFolder\n",
    "    img_dataset = datasets.ImageFolder(data_dir, transform)\n",
    "    \n",
    "    # create and return DataLoaders\n",
    "    data_loader = DataLoader(dataset=img_dataset, batch_size=batch_size, shuffle=True)\n",
    "    \n",
    "    return data_loader\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 = 128\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 0x7fe06f7b4470>"
      ]
     },
     "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",
    "    # scale from 0-1 to feature_range\n",
    "    min, max = feature_range\n",
    "    x = x * (max - min) + min\n",
    "    return x\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min:  tensor(-0.9451)\n",
      "Max:  tensor(0.5294)\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",
    "# helper conv function\n",
    "def conv(in_channels, out_channels, kernel_size, stride=2, padding=1, batch_norm=True):\n",
    "    \"\"\"Creates a convolutional layer, with optional batch normalization.\n",
    "    \"\"\"\n",
    "    layers = []\n",
    "    conv_layer = nn.Conv2d(in_channels=in_channels, out_channels=out_channels, \n",
    "                           kernel_size=kernel_size, stride=stride, padding=padding, bias=False)\n",
    "    \n",
    "    layers.append(conv_layer)\n",
    "\n",
    "    if batch_norm:\n",
    "        layers.append(nn.BatchNorm2d(out_channels))\n",
    "    return nn.Sequential(*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",
    "        self.conv_dim = conv_dim\n",
    "\n",
    "        # 32x32 input\n",
    "        self.conv1 = conv(3, conv_dim, 4, batch_norm=False) # first layer, no batch_norm\n",
    "        # 16x16 out\n",
    "        self.conv2 = conv(conv_dim, conv_dim*2, 4)\n",
    "        # 8x8 out\n",
    "        self.conv3 = conv(conv_dim*2, conv_dim*4, 4)\n",
    "        # 4x4 out\n",
    "        \n",
    "        # final, fully-connected layer\n",
    "        self.fc = nn.Linear(conv_dim*4*4*4, 1)\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",
    "        # all hidden layers + leaky relu activation\n",
    "        out = F.leaky_relu(self.conv1(x), 0.2)\n",
    "        out = F.leaky_relu(self.conv2(out), 0.2)\n",
    "        out = F.leaky_relu(self.conv3(out), 0.2)\n",
    "        \n",
    "        # flatten\n",
    "        out = out.view(-1, self.conv_dim*4*4*4)\n",
    "        \n",
    "        # final output layer\n",
    "        out = self.fc(out)        \n",
    "        return out\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": [
    "# helper deconv function\n",
    "def deconv(in_channels, out_channels, kernel_size, stride=2, padding=1, batch_norm=True):\n",
    "    \"\"\"Creates a transposed-convolutional layer, with optional batch normalization.\n",
    "    \"\"\"\n",
    "    # create a sequence of transpose + optional batch norm layers\n",
    "    layers = []\n",
    "    transpose_conv_layer = nn.ConvTranspose2d(in_channels, out_channels, \n",
    "                                              kernel_size, stride, padding, bias=False)\n",
    "    # append transpose convolutional layer\n",
    "    layers.append(transpose_conv_layer)\n",
    "    \n",
    "    if batch_norm:\n",
    "        # append batchnorm layer\n",
    "        layers.append(nn.BatchNorm2d(out_channels))\n",
    "        \n",
    "    return nn.Sequential(*layers)\n"
   ]
  },
  {
   "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",
    "        # first, fully-connected layer\n",
    "        self.fc = nn.Linear(z_size, conv_dim*4*4*4)\n",
    "\n",
    "        # transpose conv layers\n",
    "        self.t_conv1 = deconv(conv_dim*4, conv_dim*2, 4)\n",
    "        self.t_conv2 = deconv(conv_dim*2, conv_dim, 4)\n",
    "        self.t_conv3 = deconv(conv_dim, 3, 4, batch_norm=False)\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",
    "        # fully-connected + reshape \n",
    "        out = self.fc(x)\n",
    "        out = out.view(-1, self.conv_dim*4, 4, 4) # (batch_size, depth, 4, 4)\n",
    "        \n",
    "        # hidden transpose conv layers + relu\n",
    "        out = F.relu(self.t_conv1(out))\n",
    "        out = F.relu(self.t_conv2(out))\n",
    "        \n",
    "        # last layer + tanh activation\n",
    "        out = self.t_conv3(out)\n",
    "        out = F.tanh(out)\n",
    "        \n",
    "        return out\n",
    "\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": 19,
   "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",
    "    if classname.find('Linear') != -1 or classname.find('Conv') != -1:\n",
    "        # get the number of the inputs\n",
    "        m.weight.data.normal_(0, 0.02)"
   ]
  },
  {
   "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": 20,
   "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": 21,
   "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",
      ")\n",
      "\n",
      "Generator(\n",
      "  (fc): Linear(in_features=100, out_features=2048, bias=True)\n",
      "  (t_conv1): 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",
      "  (t_conv2): Sequential(\n",
      "    (0): ConvTranspose2d(64, 32, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (t_conv3): Sequential(\n",
      "    (0): ConvTranspose2d(32, 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 = 32\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": 22,
   "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": 29,
   "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",
    "    batch_size = D_out.size(0)\n",
    "    # label smoothing\n",
    "    #if smooth:\n",
    "        # smooth, real labels = 0.9\n",
    "        #labels = torch.ones(batch_size)*0.9\n",
    "    #else:\n",
    "    labels = torch.ones(batch_size) # real labels = 1\n",
    "    # move labels to GPU if available     \n",
    "    if train_on_gpu:\n",
    "        labels = labels.cuda()\n",
    "    # binary cross entropy with logits loss\n",
    "    criterion = nn.BCEWithLogitsLoss()\n",
    "    # calculate loss\n",
    "    loss = criterion(D_out.squeeze(), labels)\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",
    "    batch_size = D_out.size(0)\n",
    "    labels = torch.zeros(batch_size) # fake labels = 0\n",
    "    if train_on_gpu:\n",
    "        labels = labels.cuda()\n",
    "    criterion = nn.BCEWithLogitsLoss()\n",
    "    # calculate loss\n",
    "    loss = criterion(D_out.squeeze(), labels)\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": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.optim as optim\n",
    "\n",
    "# params\n",
    "lr = 0.0002\n",
    "beta1=0.5\n",
    "beta2=0.999 # default value\n",
    "\n",
    "# Create optimizers for the discriminator D and generator G\n",
    "d_optimizer = optim.Adam(D.parameters(), lr, [beta1, beta2])\n",
    "g_optimizer = optim.Adam(G.parameters(), lr, [beta1, beta2])"
   ]
  },
  {
   "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": 31,
   "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",
    "            #         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",
    "            # Compute the discriminator losses on real images \n",
    "            if train_on_gpu:\n",
    "                real_images = real_images.cuda()\n",
    "            D_real = D(real_images)\n",
    "            d_real_loss = real_loss(D_real)\n",
    "\n",
    "            # 2. Train with fake images\n",
    "        \n",
    "            # Generate fake images\n",
    "            z = np.random.uniform(-1, 1, size=(batch_size, z_size))\n",
    "            z = torch.from_numpy(z).float()\n",
    "            # move x to GPU, if available\n",
    "            if train_on_gpu:\n",
    "                z = z.cuda()\n",
    "            fake_images = G(z)\n",
    "            \n",
    "            # Compute the discriminator losses on fake images            \n",
    "            D_fake = D(fake_images)\n",
    "            d_fake_loss = fake_loss(D_fake)\n",
    "        \n",
    "            # add up loss and perform backprop\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",
    "            g_optimizer.zero_grad()\n",
    "            \n",
    "            # 1. Train with fake images and flipped labels\n",
    "        \n",
    "            # Generate fake images\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",
    "            # Compute the discriminator losses on fake images \n",
    "            # using flipped labels!\n",
    "            D_fake = D(fake_images)\n",
    "            g_loss = real_loss(D_fake) # use real loss to flip labels\n",
    "        \n",
    "            # perform backprop\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": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [    1/   30] | d_loss: 0.7290 | g_loss: 1.7826\n",
      "Epoch [    1/   30] | d_loss: 0.3261 | g_loss: 2.1924\n",
      "Epoch [    1/   30] | d_loss: 0.4505 | g_loss: 2.3533\n",
      "Epoch [    1/   30] | d_loss: 0.2516 | g_loss: 2.6897\n",
      "Epoch [    1/   30] | d_loss: 0.1796 | g_loss: 3.4187\n",
      "Epoch [    1/   30] | d_loss: 0.7460 | g_loss: 3.8070\n",
      "Epoch [    1/   30] | d_loss: 0.2681 | g_loss: 2.6202\n",
      "Epoch [    1/   30] | d_loss: 0.7060 | g_loss: 1.6240\n",
      "Epoch [    1/   30] | d_loss: 0.4663 | g_loss: 3.6797\n",
      "Epoch [    1/   30] | d_loss: 0.3438 | g_loss: 2.6912\n",
      "Epoch [    1/   30] | d_loss: 0.3724 | g_loss: 2.6982\n",
      "Epoch [    1/   30] | d_loss: 0.3478 | g_loss: 2.5589\n",
      "Epoch [    1/   30] | d_loss: 0.2575 | g_loss: 2.5226\n",
      "Epoch [    1/   30] | d_loss: 0.3071 | g_loss: 2.6277\n",
      "Epoch [    1/   30] | d_loss: 0.3507 | g_loss: 2.5574\n",
      "Epoch [    2/   30] | d_loss: 0.2460 | g_loss: 3.1215\n",
      "Epoch [    2/   30] | d_loss: 0.3376 | g_loss: 2.0572\n",
      "Epoch [    2/   30] | d_loss: 0.2711 | g_loss: 2.3013\n",
      "Epoch [    2/   30] | d_loss: 0.3047 | g_loss: 2.2781\n",
      "Epoch [    2/   30] | d_loss: 0.2200 | g_loss: 2.8619\n",
      "Epoch [    2/   30] | d_loss: 1.2846 | g_loss: 1.6172\n",
      "Epoch [    2/   30] | d_loss: 0.2968 | g_loss: 2.7878\n",
      "Epoch [    2/   30] | d_loss: 0.2994 | g_loss: 2.4334\n",
      "Epoch [    2/   30] | d_loss: 0.2406 | g_loss: 3.0048\n",
      "Epoch [    2/   30] | d_loss: 0.8923 | g_loss: 3.8936\n",
      "Epoch [    2/   30] | d_loss: 0.2887 | g_loss: 3.1527\n",
      "Epoch [    2/   30] | d_loss: 0.2451 | g_loss: 2.2879\n",
      "Epoch [    2/   30] | d_loss: 0.2407 | g_loss: 2.6732\n",
      "Epoch [    2/   30] | d_loss: 0.3887 | g_loss: 2.1842\n",
      "Epoch [    2/   30] | d_loss: 0.3109 | g_loss: 2.2864\n",
      "Epoch [    3/   30] | d_loss: 0.2728 | g_loss: 2.7189\n",
      "Epoch [    3/   30] | d_loss: 0.3431 | g_loss: 3.3116\n",
      "Epoch [    3/   30] | d_loss: 0.2561 | g_loss: 2.6526\n",
      "Epoch [    3/   30] | d_loss: 0.3488 | g_loss: 3.7393\n",
      "Epoch [    3/   30] | d_loss: 0.2712 | g_loss: 2.8543\n",
      "Epoch [    3/   30] | d_loss: 0.4443 | g_loss: 4.8208\n",
      "Epoch [    3/   30] | d_loss: 0.6276 | g_loss: 2.4374\n",
      "Epoch [    3/   30] | d_loss: 0.4217 | g_loss: 3.1432\n",
      "Epoch [    3/   30] | d_loss: 0.4045 | g_loss: 2.9584\n",
      "Epoch [    3/   30] | d_loss: 0.3157 | g_loss: 3.8814\n",
      "Epoch [    3/   30] | d_loss: 0.4271 | g_loss: 1.9916\n",
      "Epoch [    3/   30] | d_loss: 0.2743 | g_loss: 2.5599\n",
      "Epoch [    3/   30] | d_loss: 0.2124 | g_loss: 2.6361\n",
      "Epoch [    3/   30] | d_loss: 0.2760 | g_loss: 2.4864\n",
      "Epoch [    3/   30] | d_loss: 0.3172 | g_loss: 2.8209\n",
      "Epoch [    4/   30] | d_loss: 0.2444 | g_loss: 2.4681\n",
      "Epoch [    4/   30] | d_loss: 3.4188 | g_loss: 3.7397\n",
      "Epoch [    4/   30] | d_loss: 0.7594 | g_loss: 0.9980\n",
      "Epoch [    4/   30] | d_loss: 0.3248 | g_loss: 3.4854\n",
      "Epoch [    4/   30] | d_loss: 0.3566 | g_loss: 1.9719\n",
      "Epoch [    4/   30] | d_loss: 0.2475 | g_loss: 2.7677\n",
      "Epoch [    4/   30] | d_loss: 0.2363 | g_loss: 2.7853\n",
      "Epoch [    4/   30] | d_loss: 0.3957 | g_loss: 2.1729\n",
      "Epoch [    4/   30] | d_loss: 0.6338 | g_loss: 1.6278\n",
      "Epoch [    4/   30] | d_loss: 0.3590 | g_loss: 2.2367\n",
      "Epoch [    4/   30] | d_loss: 0.3093 | g_loss: 3.0272\n",
      "Epoch [    4/   30] | d_loss: 0.2650 | g_loss: 2.7781\n",
      "Epoch [    4/   30] | d_loss: 0.1641 | g_loss: 2.5929\n",
      "Epoch [    4/   30] | d_loss: 0.2598 | g_loss: 3.7315\n",
      "Epoch [    4/   30] | d_loss: 0.3274 | g_loss: 2.0579\n",
      "Epoch [    5/   30] | d_loss: 0.4083 | g_loss: 1.9168\n",
      "Epoch [    5/   30] | d_loss: 0.2382 | g_loss: 3.2634\n",
      "Epoch [    5/   30] | d_loss: 0.2854 | g_loss: 2.1707\n",
      "Epoch [    5/   30] | d_loss: 0.2923 | g_loss: 2.7066\n",
      "Epoch [    5/   30] | d_loss: 0.2424 | g_loss: 3.1382\n",
      "Epoch [    5/   30] | d_loss: 0.7944 | g_loss: 4.5619\n",
      "Epoch [    5/   30] | d_loss: 0.4576 | g_loss: 1.8881\n",
      "Epoch [    5/   30] | d_loss: 0.9619 | g_loss: 3.2556\n",
      "Epoch [    5/   30] | d_loss: 0.2932 | g_loss: 2.9921\n",
      "Epoch [    5/   30] | d_loss: 0.2799 | g_loss: 2.7907\n",
      "Epoch [    5/   30] | d_loss: 0.2865 | g_loss: 2.4909\n",
      "Epoch [    5/   30] | d_loss: 0.2722 | g_loss: 2.5653\n",
      "Epoch [    5/   30] | d_loss: 0.1728 | g_loss: 3.5271\n",
      "Epoch [    5/   30] | d_loss: 0.2334 | g_loss: 2.9818\n",
      "Epoch [    5/   30] | d_loss: 0.2069 | g_loss: 2.4808\n",
      "Epoch [    6/   30] | d_loss: 0.3631 | g_loss: 4.3535\n",
      "Epoch [    6/   30] | d_loss: 0.9522 | g_loss: 1.4211\n",
      "Epoch [    6/   30] | d_loss: 0.2342 | g_loss: 2.7081\n",
      "Epoch [    6/   30] | d_loss: 1.1785 | g_loss: 2.4041\n",
      "Epoch [    6/   30] | d_loss: 0.4590 | g_loss: 3.3454\n",
      "Epoch [    6/   30] | d_loss: 0.2586 | g_loss: 2.8696\n",
      "Epoch [    6/   30] | d_loss: 0.2410 | g_loss: 2.9200\n",
      "Epoch [    6/   30] | d_loss: 0.3046 | g_loss: 2.5969\n",
      "Epoch [    6/   30] | d_loss: 0.2189 | g_loss: 2.7273\n",
      "Epoch [    6/   30] | d_loss: 0.2095 | g_loss: 2.3615\n",
      "Epoch [    6/   30] | d_loss: 0.2242 | g_loss: 3.2407\n",
      "Epoch [    6/   30] | d_loss: 0.5488 | g_loss: 4.3869\n",
      "Epoch [    6/   30] | d_loss: 0.3750 | g_loss: 1.9196\n",
      "Epoch [    6/   30] | d_loss: 0.2059 | g_loss: 3.6216\n",
      "Epoch [    6/   30] | d_loss: 0.1692 | g_loss: 3.1028\n",
      "Epoch [    7/   30] | d_loss: 0.2568 | g_loss: 3.1606\n",
      "Epoch [    7/   30] | d_loss: 0.2591 | g_loss: 3.5917\n",
      "Epoch [    7/   30] | d_loss: 2.4375 | g_loss: 3.1540\n",
      "Epoch [    7/   30] | d_loss: 0.3731 | g_loss: 1.7115\n",
      "Epoch [    7/   30] | d_loss: 0.2381 | g_loss: 2.9700\n",
      "Epoch [    7/   30] | d_loss: 0.1703 | g_loss: 3.5770\n",
      "Epoch [    7/   30] | d_loss: 0.2497 | g_loss: 2.8139\n",
      "Epoch [    7/   30] | d_loss: 0.2471 | g_loss: 2.9624\n",
      "Epoch [    7/   30] | d_loss: 0.2240 | g_loss: 3.0318\n",
      "Epoch [    7/   30] | d_loss: 0.3200 | g_loss: 3.1266\n",
      "Epoch [    7/   30] | d_loss: 0.2828 | g_loss: 3.0792\n",
      "Epoch [    7/   30] | d_loss: 0.2734 | g_loss: 3.3034\n",
      "Epoch [    7/   30] | d_loss: 0.8084 | g_loss: 2.7376\n",
      "Epoch [    7/   30] | d_loss: 0.2954 | g_loss: 2.5609\n",
      "Epoch [    7/   30] | d_loss: 0.2187 | g_loss: 3.6280\n",
      "Epoch [    8/   30] | d_loss: 0.3868 | g_loss: 3.2732\n",
      "Epoch [    8/   30] | d_loss: 0.2885 | g_loss: 3.1606\n",
      "Epoch [    8/   30] | d_loss: 0.1543 | g_loss: 3.1041\n",
      "Epoch [    8/   30] | d_loss: 0.1727 | g_loss: 2.7337\n",
      "Epoch [    8/   30] | d_loss: 0.1725 | g_loss: 3.4118\n",
      "Epoch [    8/   30] | d_loss: 0.1710 | g_loss: 2.9059\n",
      "Epoch [    8/   30] | d_loss: 7.0205 | g_loss: 6.7746\n",
      "Epoch [    8/   30] | d_loss: 0.7710 | g_loss: 4.2787\n",
      "Epoch [    8/   30] | d_loss: 0.2563 | g_loss: 2.9897\n",
      "Epoch [    8/   30] | d_loss: 0.3199 | g_loss: 3.2894\n",
      "Epoch [    8/   30] | d_loss: 0.2579 | g_loss: 2.6081\n",
      "Epoch [    8/   30] | d_loss: 0.1420 | g_loss: 3.5035\n",
      "Epoch [    8/   30] | d_loss: 0.2745 | g_loss: 3.0487\n",
      "Epoch [    8/   30] | d_loss: 0.8058 | g_loss: 1.4188\n",
      "Epoch [    8/   30] | d_loss: 0.2747 | g_loss: 2.6766\n",
      "Epoch [    9/   30] | d_loss: 0.1741 | g_loss: 3.3952\n",
      "Epoch [    9/   30] | d_loss: 0.2495 | g_loss: 3.0065\n",
      "Epoch [    9/   30] | d_loss: 0.3860 | g_loss: 2.1780\n",
      "Epoch [    9/   30] | d_loss: 0.1635 | g_loss: 3.0496\n",
      "Epoch [    9/   30] | d_loss: 0.1584 | g_loss: 2.8408\n",
      "Epoch [    9/   30] | d_loss: 0.1331 | g_loss: 3.4481\n",
      "Epoch [    9/   30] | d_loss: 0.2020 | g_loss: 3.4276\n",
      "Epoch [    9/   30] | d_loss: 0.4212 | g_loss: 1.8130\n",
      "Epoch [    9/   30] | d_loss: 0.2543 | g_loss: 2.4408\n",
      "Epoch [    9/   30] | d_loss: 0.1564 | g_loss: 3.8239\n",
      "Epoch [    9/   30] | d_loss: 0.3283 | g_loss: 2.4479\n",
      "Epoch [    9/   30] | d_loss: 0.5269 | g_loss: 2.2724\n",
      "Epoch [    9/   30] | d_loss: 0.3588 | g_loss: 2.7273\n",
      "Epoch [    9/   30] | d_loss: 0.5652 | g_loss: 1.6345\n",
      "Epoch [    9/   30] | d_loss: 0.1968 | g_loss: 3.6721\n",
      "Epoch [   10/   30] | d_loss: 0.6651 | g_loss: 1.3876\n",
      "Epoch [   10/   30] | d_loss: 0.8082 | g_loss: 2.0962\n",
      "Epoch [   10/   30] | d_loss: 0.3957 | g_loss: 2.7728\n",
      "Epoch [   10/   30] | d_loss: 0.2240 | g_loss: 2.7393\n",
      "Epoch [   10/   30] | d_loss: 0.1820 | g_loss: 2.5547\n",
      "Epoch [   10/   30] | d_loss: 0.1132 | g_loss: 2.6301\n",
      "Epoch [   10/   30] | d_loss: 0.1579 | g_loss: 2.9077\n",
      "Epoch [   10/   30] | d_loss: 0.8571 | g_loss: 1.4262\n",
      "Epoch [   10/   30] | d_loss: 0.1209 | g_loss: 3.6752\n",
      "Epoch [   10/   30] | d_loss: 0.2777 | g_loss: 3.4196\n",
      "Epoch [   10/   30] | d_loss: 0.2347 | g_loss: 2.9763\n",
      "Epoch [   10/   30] | d_loss: 0.3705 | g_loss: 4.3815\n",
      "Epoch [   10/   30] | d_loss: 0.2636 | g_loss: 2.6221\n",
      "Epoch [   10/   30] | d_loss: 0.1908 | g_loss: 3.7146\n",
      "Epoch [   10/   30] | d_loss: 0.1878 | g_loss: 2.9388\n",
      "Epoch [   11/   30] | d_loss: 0.1233 | g_loss: 3.6688\n",
      "Epoch [   11/   30] | d_loss: 0.1983 | g_loss: 2.5277\n",
      "Epoch [   11/   30] | d_loss: 0.1763 | g_loss: 3.3281\n",
      "Epoch [   11/   30] | d_loss: 0.7991 | g_loss: 1.4691\n",
      "Epoch [   11/   30] | d_loss: 0.2214 | g_loss: 3.6789\n",
      "Epoch [   11/   30] | d_loss: 5.7446 | g_loss: 0.3312\n",
      "Epoch [   11/   30] | d_loss: 0.8791 | g_loss: 1.9614\n",
      "Epoch [   11/   30] | d_loss: 0.3035 | g_loss: 3.0388\n",
      "Epoch [   11/   30] | d_loss: 0.2117 | g_loss: 2.8141\n",
      "Epoch [   11/   30] | d_loss: 0.2284 | g_loss: 3.4510\n",
      "Epoch [   11/   30] | d_loss: 0.1466 | g_loss: 2.5511\n",
      "Epoch [   11/   30] | d_loss: 0.1966 | g_loss: 3.1599\n",
      "Epoch [   11/   30] | d_loss: 0.8744 | g_loss: 2.5540\n",
      "Epoch [   11/   30] | d_loss: 0.2947 | g_loss: 2.9614\n",
      "Epoch [   11/   30] | d_loss: 0.1055 | g_loss: 3.0904\n",
      "Epoch [   12/   30] | d_loss: 0.1391 | g_loss: 3.5646\n",
      "Epoch [   12/   30] | d_loss: 0.1603 | g_loss: 3.0300\n",
      "Epoch [   12/   30] | d_loss: 0.1592 | g_loss: 2.7402\n",
      "Epoch [   12/   30] | d_loss: 0.1691 | g_loss: 2.3469\n",
      "Epoch [   12/   30] | d_loss: 1.2466 | g_loss: 0.9854\n",
      "Epoch [   12/   30] | d_loss: 0.2592 | g_loss: 2.8770\n",
      "Epoch [   12/   30] | d_loss: 0.2328 | g_loss: 3.7164\n",
      "Epoch [   12/   30] | d_loss: 0.3525 | g_loss: 4.0776\n",
      "Epoch [   12/   30] | d_loss: 0.1600 | g_loss: 3.3871\n",
      "Epoch [   12/   30] | d_loss: 0.1901 | g_loss: 3.6896\n",
      "Epoch [   12/   30] | d_loss: 0.1695 | g_loss: 3.0002\n",
      "Epoch [   12/   30] | d_loss: 0.1623 | g_loss: 3.1699\n",
      "Epoch [   12/   30] | d_loss: 0.1608 | g_loss: 2.9888\n",
      "Epoch [   12/   30] | d_loss: 0.7541 | g_loss: 2.6076\n",
      "Epoch [   12/   30] | d_loss: 0.0978 | g_loss: 4.2800\n",
      "Epoch [   13/   30] | d_loss: 0.1816 | g_loss: 3.4704\n",
      "Epoch [   13/   30] | d_loss: 0.1156 | g_loss: 3.8711\n",
      "Epoch [   13/   30] | d_loss: 0.1756 | g_loss: 3.1564\n",
      "Epoch [   13/   30] | d_loss: 0.1647 | g_loss: 3.3169\n",
      "Epoch [   13/   30] | d_loss: 1.1537 | g_loss: 0.9847\n",
      "Epoch [   13/   30] | d_loss: 0.1295 | g_loss: 3.5252\n",
      "Epoch [   13/   30] | d_loss: 0.5226 | g_loss: 2.2871\n",
      "Epoch [   13/   30] | d_loss: 0.2808 | g_loss: 3.8364\n",
      "Epoch [   13/   30] | d_loss: 0.2297 | g_loss: 2.6796\n",
      "Epoch [   13/   30] | d_loss: 0.1473 | g_loss: 3.3680\n",
      "Epoch [   13/   30] | d_loss: 0.1804 | g_loss: 3.3261\n",
      "Epoch [   13/   30] | d_loss: 1.4931 | g_loss: 0.8673\n",
      "Epoch [   13/   30] | d_loss: 0.2538 | g_loss: 2.7716\n",
      "Epoch [   13/   30] | d_loss: 0.3037 | g_loss: 3.6563\n",
      "Epoch [   13/   30] | d_loss: 0.1745 | g_loss: 3.1626\n",
      "Epoch [   14/   30] | d_loss: 0.2268 | g_loss: 2.7801\n",
      "Epoch [   14/   30] | d_loss: 0.1713 | g_loss: 3.2365\n",
      "Epoch [   14/   30] | d_loss: 0.1969 | g_loss: 3.5426\n",
      "Epoch [   14/   30] | d_loss: 0.1150 | g_loss: 3.3277\n",
      "Epoch [   14/   30] | d_loss: 0.6010 | g_loss: 1.5924\n",
      "Epoch [   14/   30] | d_loss: 0.5024 | g_loss: 2.9438\n",
      "Epoch [   14/   30] | d_loss: 0.2322 | g_loss: 3.3428\n",
      "Epoch [   14/   30] | d_loss: 0.1423 | g_loss: 2.4906\n",
      "Epoch [   14/   30] | d_loss: 0.1704 | g_loss: 3.5676\n",
      "Epoch [   14/   30] | d_loss: 0.1383 | g_loss: 3.4518\n",
      "Epoch [   14/   30] | d_loss: 0.2031 | g_loss: 2.4259\n",
      "Epoch [   14/   30] | d_loss: 0.3934 | g_loss: 4.9195\n",
      "Epoch [   14/   30] | d_loss: 0.2013 | g_loss: 3.7812\n",
      "Epoch [   14/   30] | d_loss: 0.1613 | g_loss: 2.8166\n",
      "Epoch [   14/   30] | d_loss: 0.2188 | g_loss: 2.9043\n",
      "Epoch [   15/   30] | d_loss: 0.0979 | g_loss: 3.2491\n",
      "Epoch [   15/   30] | d_loss: 0.1494 | g_loss: 3.5833\n",
      "Epoch [   15/   30] | d_loss: 0.1518 | g_loss: 2.9019\n",
      "Epoch [   15/   30] | d_loss: 0.7351 | g_loss: 1.2729\n",
      "Epoch [   15/   30] | d_loss: 0.2975 | g_loss: 3.0944\n",
      "Epoch [   15/   30] | d_loss: 0.2916 | g_loss: 3.7312\n",
      "Epoch [   15/   30] | d_loss: 0.4443 | g_loss: 1.8920\n",
      "Epoch [   15/   30] | d_loss: 0.2334 | g_loss: 2.8465\n",
      "Epoch [   15/   30] | d_loss: 0.1596 | g_loss: 3.7483\n",
      "Epoch [   15/   30] | d_loss: 0.1963 | g_loss: 3.2950\n",
      "Epoch [   15/   30] | d_loss: 0.2259 | g_loss: 2.4213\n",
      "Epoch [   15/   30] | d_loss: 0.1387 | g_loss: 3.4703\n",
      "Epoch [   15/   30] | d_loss: 0.1653 | g_loss: 3.5590\n",
      "Epoch [   15/   30] | d_loss: 0.1990 | g_loss: 2.1023\n",
      "Epoch [   15/   30] | d_loss: 0.1890 | g_loss: 4.7166\n",
      "Epoch [   16/   30] | d_loss: 0.1982 | g_loss: 3.2133\n",
      "Epoch [   16/   30] | d_loss: 0.3874 | g_loss: 3.8540\n",
      "Epoch [   16/   30] | d_loss: 0.1358 | g_loss: 4.4402\n",
      "Epoch [   16/   30] | d_loss: 1.3114 | g_loss: 1.0862\n",
      "Epoch [   16/   30] | d_loss: 0.2198 | g_loss: 3.3972\n",
      "Epoch [   16/   30] | d_loss: 0.1436 | g_loss: 3.7951\n",
      "Epoch [   16/   30] | d_loss: 0.2083 | g_loss: 2.3285\n",
      "Epoch [   16/   30] | d_loss: 0.7412 | g_loss: 1.7660\n",
      "Epoch [   16/   30] | d_loss: 0.1892 | g_loss: 3.9705\n",
      "Epoch [   16/   30] | d_loss: 0.1389 | g_loss: 2.6107\n",
      "Epoch [   16/   30] | d_loss: 0.1814 | g_loss: 2.8494\n",
      "Epoch [   16/   30] | d_loss: 1.1376 | g_loss: 1.6617\n",
      "Epoch [   16/   30] | d_loss: 1.6824 | g_loss: 5.1579\n",
      "Epoch [   16/   30] | d_loss: 0.1871 | g_loss: 2.9017\n",
      "Epoch [   16/   30] | d_loss: 0.2469 | g_loss: 3.1842\n",
      "Epoch [   17/   30] | d_loss: 0.2180 | g_loss: 2.8790\n",
      "Epoch [   17/   30] | d_loss: 0.1360 | g_loss: 3.5892\n",
      "Epoch [   17/   30] | d_loss: 0.1504 | g_loss: 3.7533\n",
      "Epoch [   17/   30] | d_loss: 0.1276 | g_loss: 3.5961\n",
      "Epoch [   17/   30] | d_loss: 0.1565 | g_loss: 4.1215\n",
      "Epoch [   17/   30] | d_loss: 0.4976 | g_loss: 1.2910\n",
      "Epoch [   17/   30] | d_loss: 0.2603 | g_loss: 2.8671\n",
      "Epoch [   17/   30] | d_loss: 0.7441 | g_loss: 1.6406\n",
      "Epoch [   17/   30] | d_loss: 0.1881 | g_loss: 3.7249\n",
      "Epoch [   17/   30] | d_loss: 0.1418 | g_loss: 3.4993\n",
      "Epoch [   17/   30] | d_loss: 0.1271 | g_loss: 3.4580\n",
      "Epoch [   17/   30] | d_loss: 0.3554 | g_loss: 3.7401\n",
      "Epoch [   17/   30] | d_loss: 0.2423 | g_loss: 2.6306\n",
      "Epoch [   17/   30] | d_loss: 0.1836 | g_loss: 3.2250\n",
      "Epoch [   17/   30] | d_loss: 0.2276 | g_loss: 5.1521\n",
      "Epoch [   18/   30] | d_loss: 0.1321 | g_loss: 4.1996\n",
      "Epoch [   18/   30] | d_loss: 0.2166 | g_loss: 3.3663\n",
      "Epoch [   18/   30] | d_loss: 0.1381 | g_loss: 3.4589\n",
      "Epoch [   18/   30] | d_loss: 0.2366 | g_loss: 2.6822\n",
      "Epoch [   18/   30] | d_loss: 0.7279 | g_loss: 0.7507\n",
      "Epoch [   18/   30] | d_loss: 0.5169 | g_loss: 2.2190\n",
      "Epoch [   18/   30] | d_loss: 0.1826 | g_loss: 2.8608\n",
      "Epoch [   18/   30] | d_loss: 0.1180 | g_loss: 2.7186\n",
      "Epoch [   18/   30] | d_loss: 0.2203 | g_loss: 2.5045\n",
      "Epoch [   18/   30] | d_loss: 0.1435 | g_loss: 3.2361\n",
      "Epoch [   18/   30] | d_loss: 0.2321 | g_loss: 2.5059\n",
      "Epoch [   18/   30] | d_loss: 0.1796 | g_loss: 3.9955\n",
      "Epoch [   18/   30] | d_loss: 0.2956 | g_loss: 4.4201\n",
      "Epoch [   18/   30] | d_loss: 1.7701 | g_loss: 0.6941\n",
      "Epoch [   18/   30] | d_loss: 0.8767 | g_loss: 2.1437\n",
      "Epoch [   19/   30] | d_loss: 0.7314 | g_loss: 2.3106\n",
      "Epoch [   19/   30] | d_loss: 0.3780 | g_loss: 3.2477\n",
      "Epoch [   19/   30] | d_loss: 3.0759 | g_loss: 0.3455\n",
      "Epoch [   19/   30] | d_loss: 0.3084 | g_loss: 2.6326\n",
      "Epoch [   19/   30] | d_loss: 0.2368 | g_loss: 3.6684\n",
      "Epoch [   19/   30] | d_loss: 0.1967 | g_loss: 3.5182\n",
      "Epoch [   19/   30] | d_loss: 0.1826 | g_loss: 2.8467\n",
      "Epoch [   19/   30] | d_loss: 0.1417 | g_loss: 3.6918\n",
      "Epoch [   19/   30] | d_loss: 0.3379 | g_loss: 2.4612\n",
      "Epoch [   19/   30] | d_loss: 0.3162 | g_loss: 2.8847\n",
      "Epoch [   19/   30] | d_loss: 0.2013 | g_loss: 2.7380\n",
      "Epoch [   19/   30] | d_loss: 0.1792 | g_loss: 2.9908\n",
      "Epoch [   19/   30] | d_loss: 0.1353 | g_loss: 3.1807\n",
      "Epoch [   19/   30] | d_loss: 0.2616 | g_loss: 3.4183\n",
      "Epoch [   19/   30] | d_loss: 0.1343 | g_loss: 3.8637\n",
      "Epoch [   20/   30] | d_loss: 0.1303 | g_loss: 4.0135\n",
      "Epoch [   20/   30] | d_loss: 0.2333 | g_loss: 3.8806\n",
      "Epoch [   20/   30] | d_loss: 0.1733 | g_loss: 3.1555\n",
      "Epoch [   20/   30] | d_loss: 1.1729 | g_loss: 1.1328\n",
      "Epoch [   20/   30] | d_loss: 0.3663 | g_loss: 1.5363\n",
      "Epoch [   20/   30] | d_loss: 0.1887 | g_loss: 2.8453\n",
      "Epoch [   20/   30] | d_loss: 0.2529 | g_loss: 2.2912\n",
      "Epoch [   20/   30] | d_loss: 0.2237 | g_loss: 3.6784\n",
      "Epoch [   20/   30] | d_loss: 0.4736 | g_loss: 1.3346\n",
      "Epoch [   20/   30] | d_loss: 0.2447 | g_loss: 2.7077\n",
      "Epoch [   20/   30] | d_loss: 0.9609 | g_loss: 4.1323\n",
      "Epoch [   20/   30] | d_loss: 0.1366 | g_loss: 4.0901\n",
      "Epoch [   20/   30] | d_loss: 0.2880 | g_loss: 4.4135\n",
      "Epoch [   20/   30] | d_loss: 0.2897 | g_loss: 4.0469\n",
      "Epoch [   20/   30] | d_loss: 0.1580 | g_loss: 4.3309\n",
      "Epoch [   21/   30] | d_loss: 0.1676 | g_loss: 3.1752\n",
      "Epoch [   21/   30] | d_loss: 0.2045 | g_loss: 3.1062\n",
      "Epoch [   21/   30] | d_loss: 0.2465 | g_loss: 4.6013\n",
      "Epoch [   21/   30] | d_loss: 0.2308 | g_loss: 2.3959\n",
      "Epoch [   21/   30] | d_loss: 0.2599 | g_loss: 5.2230\n",
      "Epoch [   21/   30] | d_loss: 2.1694 | g_loss: 2.6648\n",
      "Epoch [   21/   30] | d_loss: 0.1751 | g_loss: 4.2369\n",
      "Epoch [   21/   30] | d_loss: 0.2016 | g_loss: 3.3950\n",
      "Epoch [   21/   30] | d_loss: 0.7030 | g_loss: 1.3690\n",
      "Epoch [   21/   30] | d_loss: 0.3239 | g_loss: 2.2442\n",
      "Epoch [   21/   30] | d_loss: 0.1159 | g_loss: 3.2581\n",
      "Epoch [   21/   30] | d_loss: 0.1638 | g_loss: 2.9453\n",
      "Epoch [   21/   30] | d_loss: 0.1558 | g_loss: 2.8158\n",
      "Epoch [   21/   30] | d_loss: 0.0927 | g_loss: 3.5629\n",
      "Epoch [   21/   30] | d_loss: 0.1562 | g_loss: 3.3138\n",
      "Epoch [   22/   30] | d_loss: 0.1379 | g_loss: 4.2298\n",
      "Epoch [   22/   30] | d_loss: 0.0922 | g_loss: 4.0105\n",
      "Epoch [   22/   30] | d_loss: 0.1402 | g_loss: 3.4768\n",
      "Epoch [   22/   30] | d_loss: 0.5672 | g_loss: 2.7118\n",
      "Epoch [   22/   30] | d_loss: 0.2081 | g_loss: 3.1936\n",
      "Epoch [   22/   30] | d_loss: 0.1706 | g_loss: 3.4707\n",
      "Epoch [   22/   30] | d_loss: 0.1877 | g_loss: 1.9814\n",
      "Epoch [   22/   30] | d_loss: 0.1314 | g_loss: 3.5394\n",
      "Epoch [   22/   30] | d_loss: 0.2181 | g_loss: 2.9443\n",
      "Epoch [   22/   30] | d_loss: 0.1308 | g_loss: 4.0671\n",
      "Epoch [   22/   30] | d_loss: 0.1525 | g_loss: 2.9395\n",
      "Epoch [   22/   30] | d_loss: 0.2687 | g_loss: 2.4907\n",
      "Epoch [   22/   30] | d_loss: 0.4062 | g_loss: 3.1362\n",
      "Epoch [   22/   30] | d_loss: 0.1309 | g_loss: 3.9283\n",
      "Epoch [   22/   30] | d_loss: 0.2492 | g_loss: 4.0989\n",
      "Epoch [   23/   30] | d_loss: 0.2178 | g_loss: 3.7527\n",
      "Epoch [   23/   30] | d_loss: 0.2011 | g_loss: 2.7677\n",
      "Epoch [   23/   30] | d_loss: 0.3121 | g_loss: 2.7061\n",
      "Epoch [   23/   30] | d_loss: 0.2203 | g_loss: 3.3783\n",
      "Epoch [   23/   30] | d_loss: 0.2295 | g_loss: 3.2184\n",
      "Epoch [   23/   30] | d_loss: 0.4296 | g_loss: 2.9264\n",
      "Epoch [   23/   30] | d_loss: 0.1460 | g_loss: 3.2053\n",
      "Epoch [   23/   30] | d_loss: 0.2244 | g_loss: 3.3783\n",
      "Epoch [   23/   30] | d_loss: 0.1598 | g_loss: 3.4205\n",
      "Epoch [   23/   30] | d_loss: 0.1098 | g_loss: 3.3806\n",
      "Epoch [   23/   30] | d_loss: 0.1532 | g_loss: 2.9260\n",
      "Epoch [   23/   30] | d_loss: 0.9103 | g_loss: 3.4257\n",
      "Epoch [   23/   30] | d_loss: 0.1738 | g_loss: 3.0835\n",
      "Epoch [   23/   30] | d_loss: 0.1403 | g_loss: 3.5967\n",
      "Epoch [   23/   30] | d_loss: 1.1494 | g_loss: 1.4626\n",
      "Epoch [   24/   30] | d_loss: 0.8828 | g_loss: 1.8782\n",
      "Epoch [   24/   30] | d_loss: 0.1980 | g_loss: 3.4863\n",
      "Epoch [   24/   30] | d_loss: 0.1517 | g_loss: 2.5725\n",
      "Epoch [   24/   30] | d_loss: 0.3637 | g_loss: 2.9132\n",
      "Epoch [   24/   30] | d_loss: 0.1279 | g_loss: 3.3873\n",
      "Epoch [   24/   30] | d_loss: 0.2026 | g_loss: 2.4419\n",
      "Epoch [   24/   30] | d_loss: 0.1337 | g_loss: 4.1268\n",
      "Epoch [   24/   30] | d_loss: 0.1654 | g_loss: 3.3387\n",
      "Epoch [   24/   30] | d_loss: 0.4493 | g_loss: 2.9678\n",
      "Epoch [   24/   30] | d_loss: 0.1278 | g_loss: 3.7281\n",
      "Epoch [   24/   30] | d_loss: 0.1414 | g_loss: 3.8877\n",
      "Epoch [   24/   30] | d_loss: 0.1204 | g_loss: 4.3009\n",
      "Epoch [   24/   30] | d_loss: 0.1567 | g_loss: 3.1514\n",
      "Epoch [   24/   30] | d_loss: 0.6430 | g_loss: 1.3034\n",
      "Epoch [   24/   30] | d_loss: 0.2588 | g_loss: 2.7439\n",
      "Epoch [   25/   30] | d_loss: 0.2102 | g_loss: 3.5018\n",
      "Epoch [   25/   30] | d_loss: 0.1957 | g_loss: 3.3475\n",
      "Epoch [   25/   30] | d_loss: 0.1575 | g_loss: 3.3975\n",
      "Epoch [   25/   30] | d_loss: 0.1506 | g_loss: 2.3757\n",
      "Epoch [   25/   30] | d_loss: 0.1520 | g_loss: 3.4028\n",
      "Epoch [   25/   30] | d_loss: 0.1845 | g_loss: 4.1216\n",
      "Epoch [   25/   30] | d_loss: 0.1855 | g_loss: 3.9626\n",
      "Epoch [   25/   30] | d_loss: 0.1549 | g_loss: 2.6266\n",
      "Epoch [   25/   30] | d_loss: 0.7928 | g_loss: 1.1433\n",
      "Epoch [   25/   30] | d_loss: 0.2862 | g_loss: 3.8807\n",
      "Epoch [   25/   30] | d_loss: 0.2179 | g_loss: 4.2092\n",
      "Epoch [   25/   30] | d_loss: 0.1206 | g_loss: 3.2727\n",
      "Epoch [   25/   30] | d_loss: 0.0816 | g_loss: 4.2075\n",
      "Epoch [   25/   30] | d_loss: 0.1529 | g_loss: 3.6796\n",
      "Epoch [   25/   30] | d_loss: 0.1570 | g_loss: 4.0294\n",
      "Epoch [   26/   30] | d_loss: 0.1499 | g_loss: 4.0514\n",
      "Epoch [   26/   30] | d_loss: 0.1726 | g_loss: 2.2888\n",
      "Epoch [   26/   30] | d_loss: 0.1407 | g_loss: 3.3059\n",
      "Epoch [   26/   30] | d_loss: 0.1262 | g_loss: 2.9996\n",
      "Epoch [   26/   30] | d_loss: 0.8832 | g_loss: 0.9532\n",
      "Epoch [   26/   30] | d_loss: 0.1443 | g_loss: 2.8453\n",
      "Epoch [   26/   30] | d_loss: 0.1933 | g_loss: 2.7911\n",
      "Epoch [   26/   30] | d_loss: 0.1305 | g_loss: 4.0298\n",
      "Epoch [   26/   30] | d_loss: 0.1723 | g_loss: 4.2380\n",
      "Epoch [   26/   30] | d_loss: 0.0875 | g_loss: 3.3749\n",
      "Epoch [   26/   30] | d_loss: 0.1363 | g_loss: 3.5386\n",
      "Epoch [   26/   30] | d_loss: 0.8083 | g_loss: 5.7261\n",
      "Epoch [   26/   30] | d_loss: 0.1340 | g_loss: 3.7081\n",
      "Epoch [   26/   30] | d_loss: 0.3815 | g_loss: 2.8648\n",
      "Epoch [   26/   30] | d_loss: 0.1506 | g_loss: 3.8215\n",
      "Epoch [   27/   30] | d_loss: 0.2193 | g_loss: 3.9564\n",
      "Epoch [   27/   30] | d_loss: 0.1371 | g_loss: 3.2263\n",
      "Epoch [   27/   30] | d_loss: 0.1878 | g_loss: 4.2160\n",
      "Epoch [   27/   30] | d_loss: 1.7126 | g_loss: 0.7238\n",
      "Epoch [   27/   30] | d_loss: 0.2170 | g_loss: 3.4780\n",
      "Epoch [   27/   30] | d_loss: 1.0602 | g_loss: 1.1390\n",
      "Epoch [   27/   30] | d_loss: 1.0974 | g_loss: 1.0578\n",
      "Epoch [   27/   30] | d_loss: 0.1041 | g_loss: 4.1185\n",
      "Epoch [   27/   30] | d_loss: 0.1983 | g_loss: 4.0032\n",
      "Epoch [   27/   30] | d_loss: 0.3187 | g_loss: 2.8582\n",
      "Epoch [   27/   30] | d_loss: 0.1935 | g_loss: 3.8032\n",
      "Epoch [   27/   30] | d_loss: 0.1670 | g_loss: 3.3256\n",
      "Epoch [   27/   30] | d_loss: 0.1740 | g_loss: 3.6073\n",
      "Epoch [   27/   30] | d_loss: 0.1319 | g_loss: 3.4902\n",
      "Epoch [   27/   30] | d_loss: 0.1508 | g_loss: 2.6407\n",
      "Epoch [   28/   30] | d_loss: 0.1224 | g_loss: 3.7781\n",
      "Epoch [   28/   30] | d_loss: 0.1994 | g_loss: 3.4659\n",
      "Epoch [   28/   30] | d_loss: 0.1782 | g_loss: 4.3758\n",
      "Epoch [   28/   30] | d_loss: 0.1537 | g_loss: 2.9964\n",
      "Epoch [   28/   30] | d_loss: 0.1318 | g_loss: 4.1370\n",
      "Epoch [   28/   30] | d_loss: 0.1464 | g_loss: 4.4447\n",
      "Epoch [   28/   30] | d_loss: 0.1550 | g_loss: 3.9640\n",
      "Epoch [   28/   30] | d_loss: 0.3456 | g_loss: 2.6444\n",
      "Epoch [   28/   30] | d_loss: 0.1008 | g_loss: 3.9858\n",
      "Epoch [   28/   30] | d_loss: 0.0789 | g_loss: 4.4411\n",
      "Epoch [   28/   30] | d_loss: 0.1386 | g_loss: 3.6904\n",
      "Epoch [   28/   30] | d_loss: 0.1272 | g_loss: 3.8551\n",
      "Epoch [   28/   30] | d_loss: 4.5994 | g_loss: 0.4465\n",
      "Epoch [   28/   30] | d_loss: 0.1451 | g_loss: 3.5280\n",
      "Epoch [   28/   30] | d_loss: 0.1665 | g_loss: 3.6558\n",
      "Epoch [   29/   30] | d_loss: 0.1690 | g_loss: 4.1791\n",
      "Epoch [   29/   30] | d_loss: 0.0679 | g_loss: 3.3379\n",
      "Epoch [   29/   30] | d_loss: 0.1806 | g_loss: 2.9764\n",
      "Epoch [   29/   30] | d_loss: 0.2133 | g_loss: 2.7061\n",
      "Epoch [   29/   30] | d_loss: 0.2077 | g_loss: 3.1171\n",
      "Epoch [   29/   30] | d_loss: 2.2565 | g_loss: 1.5785\n",
      "Epoch [   29/   30] | d_loss: 0.1305 | g_loss: 4.7251\n",
      "Epoch [   29/   30] | d_loss: 0.2328 | g_loss: 4.3972\n",
      "Epoch [   29/   30] | d_loss: 0.8210 | g_loss: 1.8320\n",
      "Epoch [   29/   30] | d_loss: 0.1534 | g_loss: 3.8942\n",
      "Epoch [   29/   30] | d_loss: 0.1433 | g_loss: 3.6508\n",
      "Epoch [   29/   30] | d_loss: 0.7525 | g_loss: 5.5288\n",
      "Epoch [   29/   30] | d_loss: 0.9313 | g_loss: 3.4293\n",
      "Epoch [   29/   30] | d_loss: 0.2765 | g_loss: 3.6661\n",
      "Epoch [   29/   30] | d_loss: 0.1793 | g_loss: 2.7599\n",
      "Epoch [   30/   30] | d_loss: 0.1608 | g_loss: 3.5224\n",
      "Epoch [   30/   30] | d_loss: 0.1206 | g_loss: 4.0854\n",
      "Epoch [   30/   30] | d_loss: 0.1086 | g_loss: 3.5448\n",
      "Epoch [   30/   30] | d_loss: 0.0896 | g_loss: 5.0674\n",
      "Epoch [   30/   30] | d_loss: 0.1284 | g_loss: 2.9921\n",
      "Epoch [   30/   30] | d_loss: 0.1320 | g_loss: 3.4551\n",
      "Epoch [   30/   30] | d_loss: 0.2855 | g_loss: 3.1429\n",
      "Epoch [   30/   30] | d_loss: 2.0520 | g_loss: 0.5592\n",
      "Epoch [   30/   30] | d_loss: 0.1644 | g_loss: 4.0527\n",
      "Epoch [   30/   30] | d_loss: 0.1196 | g_loss: 3.7227\n",
      "Epoch [   30/   30] | d_loss: 0.1088 | g_loss: 4.1531\n",
      "Epoch [   30/   30] | d_loss: 0.2013 | g_loss: 2.8714\n",
      "Epoch [   30/   30] | d_loss: 0.2321 | g_loss: 3.6401\n",
      "Epoch [   30/   30] | d_loss: 0.2183 | g_loss: 3.2349\n",
      "Epoch [   30/   30] | d_loss: 0.0693 | g_loss: 4.0281\n"
     ]
    }
   ],
   "source": [
    "# set number of epochs \n",
    "n_epochs = 30\n",
    "\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL\n",
    "\"\"\"\n",
    "# call training function\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": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fe06837db70>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe06dbf4438>"
      ]
     },
     "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": 45,
   "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": 46,
   "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": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe06dd8b2b0>"
      ]
     },
     "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",
    "\n",
    "Optimization strategy -- In the paper(UNSUPERVISED REPRESENTATION LEARNING WITH DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORKS), the writers give a lot advice about the hyperparameters, so I tried to change Epoch between 10, 20, and 30. I found that the Epoch will affect the loss graph, and generated faces. The result of 30 is better than 20 and 10. I think that train the model with more Epochs maybe a optimization strategy.\n",
    "\n",
    "In some generated faces, the cheek and eyes are weird, and not very clear. Deeper model size can extract more features, maybe can make the model better.\n",
    "\n",
    "I trained the model 3 times with Epoch 10, 20, and 30, and there are 48 faces images. In these 48 face images, I found that there are only 4 or 5 images are not white. I guess the \"celebrity\" faces are most white, so make Discriminator think white is more real. If there are more different kind of people faces in database, I think the model will be better.\n",
    "\n"
   ]
  },
  {
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}