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",
    "    transform=transforms.Compose([  transforms.Resize(image_size),\n",
    "                                  transforms.CenterCrop(image_size),\n",
    "                                transforms.ToTensor()\n",
    "                      \n",
    "                             ])\n",
    "    data= datasets.ImageFolder(data_dir,transform=transform)\n",
    "    data_loader=torch.utils.data.DataLoader(data,batch_size=batch_size,shuffle=True)\n",
    "    \n",
    "    return data_loader\n",
    "\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 = 256\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 0x7f9bc3702d30>"
      ]
     },
     "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",
    "    min,max=feature_range\n",
    "    x=x*(max-min)+min\n",
    "    \n",
    "    return x\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min:  tensor(-0.8353)\n",
      "Max:  tensor(0.9686)\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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 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, out_channels, \n",
    "                           kernel_size, stride, padding, bias=False)\n",
    "    \n",
    "    # append conv layer\n",
    "    layers.append(conv_layer)\n",
    "\n",
    "    if batch_norm:\n",
    "        # append batchnorm layer\n",
    "        layers.append(nn.BatchNorm2d(out_channels))\n",
    "     \n",
    "    # using Sequential container\n",
    "    return nn.Sequential(*layers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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",
    "        self.conv_dim = conv_dim\n",
    "        self.layer_1 = conv(3,conv_dim,4,batch_norm = False) #16\n",
    "        self.layer_2 = conv(conv_dim,conv_dim*2,4) #8\n",
    "        self.layer_3 = conv(conv_dim*2,conv_dim*4,4) #4\n",
    "        self.fc = nn.Linear(conv_dim*4*4*4,1)\n",
    "        # complete init function\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",
    "        x = F.leaky_relu(self.layer_1(x))\n",
    "        x = F.leaky_relu(self.layer_2(x))\n",
    "        x = F.leaky_relu(self.layer_3(x))\n",
    "        x = x.view(-1,self.conv_dim*4*4*4)\n",
    "        x = self.fc(x)\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)\n"
   ]
  },
  {
   "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": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def deconv(input_dim,output_dim,kernel_size,stride=2,batch_norm=True):\n",
    "    layers=[]\n",
    "    conv_t=nn.ConvTranspose2d(input_dim,output_dim,kernel_size,stride,padding=1,bias=False)\n",
    "    layers.append(conv_t)\n",
    "    if batch_norm:\n",
    "        layers.append(nn.BatchNorm2d(output_dim))\n",
    "    return nn.Sequential(*layers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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",
    "        self.conv_dim = conv_dim\n",
    "        self.fc = nn.Linear(z_size,conv_dim*8*2*2)\n",
    "        self.layer_1 = deconv(conv_dim*8,conv_dim*4,4)\n",
    "        self.layer_2 = deconv(conv_dim*4,conv_dim*2,4)\n",
    "        self.layer_3 = deconv(conv_dim*2,conv_dim,4)\n",
    "        self.layer_4 = deconv(conv_dim,3,4,batch_norm = False)\n",
    "        # complete init function\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",
    "        x = self.fc(x)\n",
    "        x = x.view(-1,self.conv_dim*8,2,2) #(batch_size,depth,width,height)\n",
    "        x = F.relu(self.layer_1(x))\n",
    "        x = F.relu(self.layer_2(x))\n",
    "        x = F.relu(self.layer_3(x))\n",
    "        x = torch.tanh(self.layer_4(x))\n",
    "        return x\n",
    "\n",
    "\"\"\"\n",
    "DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n",
    "\"\"\"\n",
    "tests.test_generator(Generator)\n"
   ]
  },
  {
   "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": 14,
   "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",
    "    # TODO: Apply initial weights to convolutional and linear layers\n",
    "    if hasattr(m, 'weight') and (classname.find('Conv') != -1 or classname.find('Linear') != -1):\n",
    "        m.weight.data.normal_(0.0, 0.02)\n",
    "        \n",
    "        # The bias terms, if they exist, set to 0\n",
    "        if hasattr(m, 'bias') and m.bias is not None:\n",
    "            m.bias.data.zero_()"
   ]
  },
  {
   "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": 15,
   "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": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Discriminator(\n",
      "  (layer_1): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "  )\n",
      "  (layer_2): 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",
      "  (layer_3): Sequential(\n",
      "    (0): Conv2d(128, 256, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (fc): Linear(in_features=4096, out_features=1, bias=True)\n",
      ")\n",
      "\n",
      "Generator(\n",
      "  (fc): Linear(in_features=100, out_features=2048, bias=True)\n",
      "  (layer_1): Sequential(\n",
      "    (0): ConvTranspose2d(512, 256, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  )\n",
      "  (layer_2): 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",
      "  (layer_3): 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",
      "  (layer_4): 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 = 64\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": 17,
   "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": 18,
   "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",
    "    \n",
    "    labels = torch.ones(batch_size) * 0.9\n",
    "    \n",
    "    if train_on_gpu:\n",
    "        labels = labels.cuda()\n",
    "    \n",
    "    criterion = nn.BCEWithLogitsLoss()\n",
    "    loss = criterion(D_out.squeeze(), labels)\n",
    "    return loss\n",
    "\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": 19,
   "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(),lr=0.0002,betas=(0.5,0.999))\n",
    "g_optimizer = optim.Adam(G.parameters(),lr=0.0002,betas=(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": 20,
   "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",
    "            if train_on_gpu:\n",
    "                real_images = real_images.cuda()\n",
    "            # ===============================================\n",
    "            #         YOUR CODE HERE: TRAIN THE NETWORKS\n",
    "            # ===============================================\n",
    "            \n",
    "            # 1. Train the discriminator on real and fake ima.ges\n",
    "            d_optimizer.zero_grad()\n",
    "            d_out_real = D(real_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",
    "            d_loss = real_loss(d_out_real) + fake_loss(D(G(z)))\n",
    "            d_loss.backward()\n",
    "            d_optimizer.step()\n",
    "            # 2. Train the generator with an adversarial loss\n",
    "            G.train()\n",
    "            g_optimizer.zero_grad()\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",
    "            g_loss = real_loss(D(G(z)))\n",
    "            g_loss.backward()\n",
    "            g_optimizer.step()\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/   25] | d_loss: 1.4579 | g_loss: 1.1128\n",
      "Epoch [    1/   25] | d_loss: 0.5452 | g_loss: 4.2095\n",
      "Epoch [    1/   25] | d_loss: 0.5674 | g_loss: 2.4376\n",
      "Epoch [    1/   25] | d_loss: 0.7899 | g_loss: 1.5971\n",
      "Epoch [    1/   25] | d_loss: 0.7874 | g_loss: 2.6852\n",
      "Epoch [    1/   25] | d_loss: 0.7902 | g_loss: 2.0172\n",
      "Epoch [    1/   25] | d_loss: 0.9270 | g_loss: 3.0911\n",
      "Epoch [    1/   25] | d_loss: 0.9003 | g_loss: 1.9582\n",
      "Epoch [    2/   25] | d_loss: 0.8302 | g_loss: 2.2965\n",
      "Epoch [    2/   25] | d_loss: 1.0267 | g_loss: 2.4829\n",
      "Epoch [    2/   25] | d_loss: 1.0317 | g_loss: 1.2038\n",
      "Epoch [    2/   25] | d_loss: 0.9868 | g_loss: 1.4817\n",
      "Epoch [    2/   25] | d_loss: 1.1043 | g_loss: 2.0252\n",
      "Epoch [    2/   25] | d_loss: 0.9735 | g_loss: 1.2955\n",
      "Epoch [    2/   25] | d_loss: 1.5399 | g_loss: 1.0016\n",
      "Epoch [    2/   25] | d_loss: 1.3520 | g_loss: 2.8295\n",
      "Epoch [    3/   25] | d_loss: 1.1370 | g_loss: 2.0188\n",
      "Epoch [    3/   25] | d_loss: 1.0015 | g_loss: 1.8332\n",
      "Epoch [    3/   25] | d_loss: 1.1207 | g_loss: 1.4502\n",
      "Epoch [    3/   25] | d_loss: 1.0967 | g_loss: 1.3062\n",
      "Epoch [    3/   25] | d_loss: 1.0249 | g_loss: 1.6712\n",
      "Epoch [    3/   25] | d_loss: 1.0748 | g_loss: 1.5870\n",
      "Epoch [    3/   25] | d_loss: 1.1265 | g_loss: 1.9862\n",
      "Epoch [    3/   25] | d_loss: 1.1000 | g_loss: 1.1371\n",
      "Epoch [    4/   25] | d_loss: 1.1358 | g_loss: 1.0845\n",
      "Epoch [    4/   25] | d_loss: 1.0709 | g_loss: 1.1239\n",
      "Epoch [    4/   25] | d_loss: 1.0313 | g_loss: 1.3408\n",
      "Epoch [    4/   25] | d_loss: 0.9556 | g_loss: 1.7610\n",
      "Epoch [    4/   25] | d_loss: 1.0778 | g_loss: 1.0321\n",
      "Epoch [    4/   25] | d_loss: 1.1424 | g_loss: 1.3123\n",
      "Epoch [    4/   25] | d_loss: 1.0417 | g_loss: 1.2256\n",
      "Epoch [    4/   25] | d_loss: 1.0617 | g_loss: 0.9800\n",
      "Epoch [    5/   25] | d_loss: 1.1406 | g_loss: 0.8960\n",
      "Epoch [    5/   25] | d_loss: 1.0578 | g_loss: 1.1051\n",
      "Epoch [    5/   25] | d_loss: 1.0247 | g_loss: 1.3127\n",
      "Epoch [    5/   25] | d_loss: 1.0622 | g_loss: 1.2949\n",
      "Epoch [    5/   25] | d_loss: 1.0252 | g_loss: 1.4100\n",
      "Epoch [    5/   25] | d_loss: 1.1831 | g_loss: 1.9247\n",
      "Epoch [    5/   25] | d_loss: 1.0154 | g_loss: 1.4340\n",
      "Epoch [    5/   25] | d_loss: 1.0544 | g_loss: 1.9588\n",
      "Epoch [    6/   25] | d_loss: 1.2598 | g_loss: 2.1201\n",
      "Epoch [    6/   25] | d_loss: 1.1289 | g_loss: 2.0916\n",
      "Epoch [    6/   25] | d_loss: 1.0049 | g_loss: 1.0800\n",
      "Epoch [    6/   25] | d_loss: 1.0845 | g_loss: 1.5002\n",
      "Epoch [    6/   25] | d_loss: 1.0676 | g_loss: 1.6526\n",
      "Epoch [    6/   25] | d_loss: 1.0987 | g_loss: 1.1476\n",
      "Epoch [    6/   25] | d_loss: 1.0511 | g_loss: 1.0694\n",
      "Epoch [    6/   25] | d_loss: 1.0282 | g_loss: 1.2154\n",
      "Epoch [    7/   25] | d_loss: 1.1444 | g_loss: 1.0752\n",
      "Epoch [    7/   25] | d_loss: 1.0753 | g_loss: 1.2638\n",
      "Epoch [    7/   25] | d_loss: 1.0719 | g_loss: 1.0463\n",
      "Epoch [    7/   25] | d_loss: 1.0733 | g_loss: 1.1955\n",
      "Epoch [    7/   25] | d_loss: 1.3881 | g_loss: 0.9780\n",
      "Epoch [    7/   25] | d_loss: 0.9745 | g_loss: 1.4470\n",
      "Epoch [    7/   25] | d_loss: 1.1750 | g_loss: 0.7933\n",
      "Epoch [    7/   25] | d_loss: 1.0512 | g_loss: 1.7027\n",
      "Epoch [    8/   25] | d_loss: 1.1496 | g_loss: 2.1486\n",
      "Epoch [    8/   25] | d_loss: 1.0494 | g_loss: 0.9990\n",
      "Epoch [    8/   25] | d_loss: 1.1911 | g_loss: 0.9998\n",
      "Epoch [    8/   25] | d_loss: 1.0971 | g_loss: 1.2704\n",
      "Epoch [    8/   25] | d_loss: 1.0090 | g_loss: 1.3896\n",
      "Epoch [    8/   25] | d_loss: 1.0320 | g_loss: 1.3367\n",
      "Epoch [    8/   25] | d_loss: 1.0203 | g_loss: 1.3294\n",
      "Epoch [    8/   25] | d_loss: 0.9840 | g_loss: 1.2241\n",
      "Epoch [    9/   25] | d_loss: 0.9481 | g_loss: 1.5084\n",
      "Epoch [    9/   25] | d_loss: 0.8889 | g_loss: 1.2174\n",
      "Epoch [    9/   25] | d_loss: 1.1182 | g_loss: 1.4233\n",
      "Epoch [    9/   25] | d_loss: 0.9739 | g_loss: 1.1509\n",
      "Epoch [    9/   25] | d_loss: 1.1346 | g_loss: 1.8107\n",
      "Epoch [    9/   25] | d_loss: 1.0120 | g_loss: 1.5772\n",
      "Epoch [    9/   25] | d_loss: 1.0002 | g_loss: 1.6812\n",
      "Epoch [    9/   25] | d_loss: 1.0635 | g_loss: 1.5476\n",
      "Epoch [   10/   25] | d_loss: 1.0300 | g_loss: 1.5032\n",
      "Epoch [   10/   25] | d_loss: 1.0422 | g_loss: 1.3814\n",
      "Epoch [   10/   25] | d_loss: 1.1421 | g_loss: 0.8624\n",
      "Epoch [   10/   25] | d_loss: 1.1266 | g_loss: 1.9774\n",
      "Epoch [   10/   25] | d_loss: 1.0056 | g_loss: 1.3740\n",
      "Epoch [   10/   25] | d_loss: 0.9726 | g_loss: 1.2757\n",
      "Epoch [   10/   25] | d_loss: 1.0160 | g_loss: 1.3375\n",
      "Epoch [   10/   25] | d_loss: 0.9894 | g_loss: 1.1755\n",
      "Epoch [   11/   25] | d_loss: 0.9773 | g_loss: 1.4342\n",
      "Epoch [   11/   25] | d_loss: 1.0017 | g_loss: 1.4789\n",
      "Epoch [   11/   25] | d_loss: 1.0380 | g_loss: 1.0156\n",
      "Epoch [   11/   25] | d_loss: 1.0900 | g_loss: 1.1872\n",
      "Epoch [   11/   25] | d_loss: 1.0055 | g_loss: 1.2248\n",
      "Epoch [   11/   25] | d_loss: 1.0432 | g_loss: 1.1342\n",
      "Epoch [   11/   25] | d_loss: 1.0468 | g_loss: 1.6432\n",
      "Epoch [   11/   25] | d_loss: 1.0058 | g_loss: 1.2480\n",
      "Epoch [   12/   25] | d_loss: 1.0383 | g_loss: 1.1389\n",
      "Epoch [   12/   25] | d_loss: 1.0875 | g_loss: 1.8719\n",
      "Epoch [   12/   25] | d_loss: 1.1790 | g_loss: 0.9340\n",
      "Epoch [   12/   25] | d_loss: 0.8521 | g_loss: 2.3588\n",
      "Epoch [   12/   25] | d_loss: 1.0508 | g_loss: 1.2711\n",
      "Epoch [   12/   25] | d_loss: 1.0542 | g_loss: 1.5053\n",
      "Epoch [   12/   25] | d_loss: 1.0671 | g_loss: 1.0347\n",
      "Epoch [   12/   25] | d_loss: 1.0819 | g_loss: 1.2309\n",
      "Epoch [   13/   25] | d_loss: 0.9456 | g_loss: 1.5369\n",
      "Epoch [   13/   25] | d_loss: 1.1099 | g_loss: 0.9057\n",
      "Epoch [   13/   25] | d_loss: 1.0374 | g_loss: 1.0446\n",
      "Epoch [   13/   25] | d_loss: 1.0829 | g_loss: 1.1577\n",
      "Epoch [   13/   25] | d_loss: 0.9917 | g_loss: 1.2094\n",
      "Epoch [   13/   25] | d_loss: 0.9320 | g_loss: 1.5573\n",
      "Epoch [   13/   25] | d_loss: 0.9934 | g_loss: 1.2711\n",
      "Epoch [   13/   25] | d_loss: 1.0111 | g_loss: 1.7381\n",
      "Epoch [   14/   25] | d_loss: 1.1568 | g_loss: 2.0339\n",
      "Epoch [   14/   25] | d_loss: 1.0560 | g_loss: 1.1236\n",
      "Epoch [   14/   25] | d_loss: 0.9802 | g_loss: 1.4759\n",
      "Epoch [   14/   25] | d_loss: 1.0605 | g_loss: 1.1891\n",
      "Epoch [   14/   25] | d_loss: 1.0255 | g_loss: 1.1865\n",
      "Epoch [   14/   25] | d_loss: 1.0618 | g_loss: 1.7552\n",
      "Epoch [   14/   25] | d_loss: 0.9262 | g_loss: 1.0867\n",
      "Epoch [   14/   25] | d_loss: 1.1536 | g_loss: 0.9286\n",
      "Epoch [   15/   25] | d_loss: 1.2354 | g_loss: 0.8037\n",
      "Epoch [   15/   25] | d_loss: 0.9457 | g_loss: 1.4282\n",
      "Epoch [   15/   25] | d_loss: 0.9795 | g_loss: 1.3640\n",
      "Epoch [   15/   25] | d_loss: 1.0304 | g_loss: 1.3620\n",
      "Epoch [   15/   25] | d_loss: 1.0658 | g_loss: 0.9704\n",
      "Epoch [   15/   25] | d_loss: 0.9100 | g_loss: 1.7108\n",
      "Epoch [   15/   25] | d_loss: 1.0390 | g_loss: 1.2356\n",
      "Epoch [   15/   25] | d_loss: 0.9646 | g_loss: 1.6543\n",
      "Epoch [   16/   25] | d_loss: 1.0421 | g_loss: 1.3634\n",
      "Epoch [   16/   25] | d_loss: 1.1691 | g_loss: 1.2419\n",
      "Epoch [   16/   25] | d_loss: 1.0901 | g_loss: 1.3929\n",
      "Epoch [   16/   25] | d_loss: 1.0327 | g_loss: 1.1793\n",
      "Epoch [   16/   25] | d_loss: 0.9820 | g_loss: 1.6226\n",
      "Epoch [   16/   25] | d_loss: 0.9678 | g_loss: 1.4290\n",
      "Epoch [   16/   25] | d_loss: 1.4768 | g_loss: 0.8431\n",
      "Epoch [   16/   25] | d_loss: 0.9982 | g_loss: 1.2954\n",
      "Epoch [   17/   25] | d_loss: 0.9211 | g_loss: 1.4981\n",
      "Epoch [   17/   25] | d_loss: 1.0190 | g_loss: 1.2831\n",
      "Epoch [   17/   25] | d_loss: 1.1003 | g_loss: 1.9567\n",
      "Epoch [   17/   25] | d_loss: 1.0917 | g_loss: 1.5898\n",
      "Epoch [   17/   25] | d_loss: 0.9532 | g_loss: 1.2482\n",
      "Epoch [   17/   25] | d_loss: 1.0008 | g_loss: 1.3640\n",
      "Epoch [   17/   25] | d_loss: 1.5909 | g_loss: 1.2033\n",
      "Epoch [   17/   25] | d_loss: 0.9606 | g_loss: 1.5033\n",
      "Epoch [   18/   25] | d_loss: 1.1560 | g_loss: 1.7361\n",
      "Epoch [   18/   25] | d_loss: 1.0224 | g_loss: 1.2003\n",
      "Epoch [   18/   25] | d_loss: 0.9278 | g_loss: 1.3686\n",
      "Epoch [   18/   25] | d_loss: 0.9896 | g_loss: 1.3906\n",
      "Epoch [   18/   25] | d_loss: 0.9632 | g_loss: 1.5120\n",
      "Epoch [   18/   25] | d_loss: 0.9455 | g_loss: 1.2204\n",
      "Epoch [   18/   25] | d_loss: 0.8904 | g_loss: 1.5756\n",
      "Epoch [   18/   25] | d_loss: 1.0509 | g_loss: 0.9567\n",
      "Epoch [   19/   25] | d_loss: 0.9556 | g_loss: 1.3028\n",
      "Epoch [   19/   25] | d_loss: 1.1649 | g_loss: 1.1343\n",
      "Epoch [   19/   25] | d_loss: 0.9777 | g_loss: 1.5674\n",
      "Epoch [   19/   25] | d_loss: 1.1069 | g_loss: 1.4675\n",
      "Epoch [   19/   25] | d_loss: 1.0429 | g_loss: 1.5390\n",
      "Epoch [   19/   25] | d_loss: 1.0233 | g_loss: 1.1690\n",
      "Epoch [   19/   25] | d_loss: 1.0303 | g_loss: 1.6366\n",
      "Epoch [   19/   25] | d_loss: 1.0871 | g_loss: 1.5678\n",
      "Epoch [   20/   25] | d_loss: 0.9926 | g_loss: 1.0698\n",
      "Epoch [   20/   25] | d_loss: 0.9423 | g_loss: 1.4370\n",
      "Epoch [   20/   25] | d_loss: 0.8980 | g_loss: 1.6007\n",
      "Epoch [   20/   25] | d_loss: 1.0008 | g_loss: 1.0700\n",
      "Epoch [   20/   25] | d_loss: 1.1175 | g_loss: 1.8930\n",
      "Epoch [   20/   25] | d_loss: 1.0233 | g_loss: 1.5485\n",
      "Epoch [   20/   25] | d_loss: 0.9690 | g_loss: 1.2459\n",
      "Epoch [   20/   25] | d_loss: 1.0172 | g_loss: 1.6190\n",
      "Epoch [   21/   25] | d_loss: 0.9230 | g_loss: 1.8275\n",
      "Epoch [   21/   25] | d_loss: 0.9934 | g_loss: 1.4873\n",
      "Epoch [   21/   25] | d_loss: 0.9591 | g_loss: 1.3180\n",
      "Epoch [   21/   25] | d_loss: 0.9710 | g_loss: 1.1797\n",
      "Epoch [   21/   25] | d_loss: 1.0470 | g_loss: 1.1834\n",
      "Epoch [   21/   25] | d_loss: 1.0307 | g_loss: 1.3608\n",
      "Epoch [   21/   25] | d_loss: 1.1134 | g_loss: 1.2313\n",
      "Epoch [   21/   25] | d_loss: 0.8243 | g_loss: 1.6580\n",
      "Epoch [   22/   25] | d_loss: 0.9671 | g_loss: 0.7970\n",
      "Epoch [   22/   25] | d_loss: 1.1990 | g_loss: 1.6832\n",
      "Epoch [   22/   25] | d_loss: 1.1031 | g_loss: 1.8293\n",
      "Epoch [   22/   25] | d_loss: 0.9255 | g_loss: 1.5142\n",
      "Epoch [   22/   25] | d_loss: 0.9672 | g_loss: 1.0766\n",
      "Epoch [   22/   25] | d_loss: 1.0521 | g_loss: 1.6995\n",
      "Epoch [   22/   25] | d_loss: 1.3370 | g_loss: 0.7649\n",
      "Epoch [   22/   25] | d_loss: 0.9321 | g_loss: 1.2227\n",
      "Epoch [   23/   25] | d_loss: 0.9802 | g_loss: 1.0647\n",
      "Epoch [   23/   25] | d_loss: 1.6820 | g_loss: 0.5917\n",
      "Epoch [   23/   25] | d_loss: 0.8749 | g_loss: 1.6819\n",
      "Epoch [   23/   25] | d_loss: 0.9148 | g_loss: 1.6736\n",
      "Epoch [   23/   25] | d_loss: 0.9992 | g_loss: 1.4947\n",
      "Epoch [   23/   25] | d_loss: 0.9213 | g_loss: 1.7286\n",
      "Epoch [   23/   25] | d_loss: 0.9878 | g_loss: 1.6397\n",
      "Epoch [   23/   25] | d_loss: 1.0047 | g_loss: 1.1360\n",
      "Epoch [   24/   25] | d_loss: 0.9860 | g_loss: 2.4115\n",
      "Epoch [   24/   25] | d_loss: 0.9545 | g_loss: 1.1552\n",
      "Epoch [   24/   25] | d_loss: 1.0089 | g_loss: 1.3185\n",
      "Epoch [   24/   25] | d_loss: 1.1005 | g_loss: 0.9129\n",
      "Epoch [   24/   25] | d_loss: 0.9066 | g_loss: 1.4636\n",
      "Epoch [   24/   25] | d_loss: 1.0094 | g_loss: 1.6213\n",
      "Epoch [   24/   25] | d_loss: 1.0889 | g_loss: 1.2226\n",
      "Epoch [   24/   25] | d_loss: 0.8893 | g_loss: 1.7690\n",
      "Epoch [   25/   25] | d_loss: 0.9054 | g_loss: 1.3099\n",
      "Epoch [   25/   25] | d_loss: 0.8879 | g_loss: 1.6193\n",
      "Epoch [   25/   25] | d_loss: 0.8905 | g_loss: 1.8119\n",
      "Epoch [   25/   25] | d_loss: 1.0433 | g_loss: 1.0894\n",
      "Epoch [   25/   25] | d_loss: 1.0102 | g_loss: 1.2913\n",
      "Epoch [   25/   25] | d_loss: 0.9769 | g_loss: 1.2147\n",
      "Epoch [   25/   25] | d_loss: 0.9372 | g_loss: 2.0663\n",
      "Epoch [   25/   25] | d_loss: 0.9771 | g_loss: 1.9819\n"
     ]
    }
   ],
   "source": [
    "# set number of epochs \n",
    "n_epochs = 25\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": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f9bbcf54a90>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f9bbf8f5828>"
      ]
     },
     "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": null,
   "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": null,
   "metadata": {},
   "outputs": [],
   "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",
    "The quality of pictures generated is not too great, in my opinion we could have included data that was clearer.for example we used pictures in which portion of face was covered with glasses and hats , we could use data containing full face.\n",
    "Also since facial features are complex , using more than 4 layers can help.\n",
    "\n",
    "Along with DCGAN we can use CycleGAN which will allow us to generate pictures by learning swapping between faces.\n",
    "\n",
    "I sticked with using Adam as an optimizer beacause that's what mentioned to be efficient in the DCGAN paper."
   ]
  },
  {
   "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
}