{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b3c8e93d",
   "metadata": {},
   "source": [
    "# Synthesizing Handwritten Digits from the MNIST Dataset using Unconditional Generative Adversarial Networks (GANs)\n",
    "\n",
    "This notebook is a _very simple demonstration_ of Generative Adversarial Networks (GANs).  We use the MNIST handwritten digit dataset and a dense MLP-style architecture for both the generator and descriminator.  The training approach that we follow is similar to the [early work on GANs by Goodfellow, et al](https://arxiv.org/abs/1406.2661).\n",
    "\n",
    "Note that this is intended to be a simple domonstration only and does not include a number of advanced features that can dramatically improve the performance of GANs.  See some of my other GAN notebooks for demonstrations of more advanced examples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "672cfc1f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib\n",
    "%matplotlib inline\n",
    "matplotlib.rcParams['figure.figsize'] = (11, 11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "7fee062d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import functools\n",
    "import gzip\n",
    "import operator\n",
    "import struct\n",
    "\n",
    "import tqdm.notebook as tqdm\n",
    "import munch\n",
    "\n",
    "import skimage as ski\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import torch as th"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85983730",
   "metadata": {},
   "source": [
    "## MNIST Dataset\n",
    "\n",
    "This is just a Torch dataset for the standard MNIST data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7061e24c",
   "metadata": {},
   "outputs": [],
   "source": [
    "class MNIST(th.utils.data.Dataset):\n",
    "    '''Simple MNIST dataset containing only the images.\n",
    "    '''\n",
    "    def __init__(self):\n",
    "        '''Initialize a new MNIST dataset.\n",
    "        '''\n",
    "        super().__init__()\n",
    "        self.imgs = self._load_imgs()\n",
    "        \n",
    "    @staticmethod\n",
    "    def _load_imgs():\n",
    "        '''Load all of the training images.  This dataset is relatively small,\n",
    "        so we just keep everything in memory on the CPU side.\n",
    "\n",
    "        Thanks:\n",
    "            https://stackoverflow.com/questions/39969045/parsing-yann-lecuns-mnist-idx-file-format\n",
    "        '''\n",
    "        filename = '../data/torchvision/MNIST/raw/train-images-idx3-ubyte.gz'\n",
    "        with gzip.open(filename, mode='rb') as fh:\n",
    "            _magic, size = struct.unpack('>II', fh.read(8))\n",
    "            nrows, ncols = struct.unpack('>II', fh.read(8))\n",
    "            data = np.frombuffer(fh.read(), dtype=np.dtype(np.uint8).newbyteorder('>'))\n",
    "            imgs = data.reshape((size, nrows, ncols)) / 255.\n",
    "\n",
    "        return th.as_tensor(imgs, dtype=th.float32)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        '''Return a single image specified by the given index.\n",
    "        '''\n",
    "        return self.imgs[idx]\n",
    "    \n",
    "    def __len__(self):\n",
    "        '''Return the number of images in the dataset.\n",
    "        '''\n",
    "        return self.imgs.shape[0]\n",
    "    \n",
    "    def plot(self, idx):\n",
    "        '''Plot the image specified by the given index.\n",
    "        '''\n",
    "        fig, ax = plt.subplots()\n",
    "        \n",
    "        ax.imshow(self.imgs[idx], cmap=plt.cm.gray)\n",
    "        ax.axis('off')\n",
    "        \n",
    "        return munch.Munch(fig=fig, ax=ax)\n",
    "        \n",
    "    def plot_montage(self, n=36):\n",
    "        '''Plot a montage of `n` randomly selected images.\n",
    "        '''\n",
    "        idxs = th.randperm(len(self))[:n]\n",
    "        imgs = self.imgs[idxs]\n",
    "        \n",
    "        montage = ski.util.montage(imgs)\n",
    "        \n",
    "        fig, ax = plt.subplots()\n",
    "        ax.imshow(montage, cmap=plt.cm.gray)\n",
    "        ax.axis('off')\n",
    "        \n",
    "        return munch.Munch(fig=fig, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6c88649a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60000"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = MNIST()\n",
    "len(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "85ec83ed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.plot(3);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c7f8e158",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.plot_montage();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d621e16",
   "metadata": {},
   "source": [
    "## Generator\n",
    "\n",
    "Next, we define a generator that maps a flat, random-normal latent vector to an image with the specified size.  The architecture we use here is just a dense, MLP-style network with hyperbolic tangent activations since we're just building a simple demonstration.  Note that we __do not use class labels for conditioning__ the generator in this example.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5493c162",
   "metadata": {},
   "outputs": [],
   "source": [
    "class _Tanh(th.nn.Module):\n",
    "    '''Hyperbolic tangent, scaled as described in Lecun's backprop tricks paper.\n",
    "    '''\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.register_buffer('_a', th.as_tensor(1.7159), persistent=False)\n",
    "        self.register_buffer('_b', th.as_tensor(2. / 3.), persistent=False)\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self._a * th.tanh(self._b * x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f95f009e",
   "metadata": {},
   "outputs": [],
   "source": [
    "class _Linear(th.nn.Linear):\n",
    "    '''Dense linear layer with linear Kaiming weight initialization.\n",
    "    '''\n",
    "    def reset_parameters(self):\n",
    "        th.nn.init.kaiming_normal_(self.weight, nonlinearity='linear')\n",
    "        if hasattr(self, 'bias'):\n",
    "            th.nn.init.zeros_(self.bias)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "008fc526",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Generator(th.nn.Module):\n",
    "    '''MLP-style generator.\n",
    "    '''\n",
    "    def __init__(self, latent_dim=32, img_size=(28, 28), layer_specs=(64, 128, 256)):\n",
    "        '''Initialize a new generator.\n",
    "        \n",
    "        latent_dim (int):\n",
    "            Expected number of dimensions for the input latent vector\n",
    "            used for generating images.\n",
    "        img_size (tuple(int)):\n",
    "            Two-tuple of integers specifying the height and width of\n",
    "            the output images in pixels.\n",
    "        layer_specs (tuple(int)):\n",
    "            A tuple where each value specifies the number of hidden\n",
    "            units in each layer.  The number of values in this tuple\n",
    "            determines the number of layers.\n",
    "        '''\n",
    "        super().__init__()\n",
    "        \n",
    "        self.latent_dim = latent_dim\n",
    "        self.img_size = img_size\n",
    "        \n",
    "        self.hidden = th.nn.Sequential()\n",
    "        layer_in, layer_out = None, latent_dim\n",
    "        for num_units in layer_specs:\n",
    "            layer_in, layer_out = layer_out, num_units\n",
    "            self.hidden.append(\n",
    "                th.nn.Sequential(\n",
    "                    _Linear(layer_in, layer_out),\n",
    "                    _Tanh(),\n",
    "                )\n",
    "            )\n",
    "\n",
    "        out_dim = functools.reduce(operator.mul, img_size)\n",
    "        layer_in, layer_out = layer_out, out_dim\n",
    "        self.visible = _Linear(layer_in, layer_out)\n",
    "        \n",
    "    def forward(self, z):\n",
    "        '''Forward pass generates synthetic images from\n",
    "        the latent vector `z`.\n",
    "        '''\n",
    "        batch_size, latent_dim = z.shape\n",
    "        assert latent_dim == self.latent_dim\n",
    "        \n",
    "        x = self.visible(self.hidden(z))\n",
    "        \n",
    "        return th.sigmoid(x + 0.5).view(batch_size, *self.img_size)\n",
    "    \n",
    "    @th.inference_mode()\n",
    "    def plot_montage(self, n=36, z=None):\n",
    "        '''Plot a montage of `n` synthetic images.  If `z` is ``None`` then\n",
    "        the latent vectors will be drawn from the random normal distribution.\n",
    "        '''\n",
    "        if z is None:\n",
    "            z = th.randn(n, self.latent_dim)\n",
    "        imgs = self(z).cpu()\n",
    "        \n",
    "        montage = ski.util.montage(imgs)\n",
    "        fig, ax = plt.subplots()\n",
    "        ax.imshow(montage, cmap=plt.cm.gray)\n",
    "        ax.axis('off')\n",
    "        \n",
    "        return munch.Munch(fig=fig, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d6c76aec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Generator(\n",
       "  (hidden): Sequential(\n",
       "    (0): Sequential(\n",
       "      (0): _Linear(in_features=32, out_features=64, bias=True)\n",
       "      (1): _Tanh()\n",
       "    )\n",
       "    (1): Sequential(\n",
       "      (0): _Linear(in_features=64, out_features=128, bias=True)\n",
       "      (1): _Tanh()\n",
       "    )\n",
       "    (2): Sequential(\n",
       "      (0): _Linear(in_features=128, out_features=256, bias=True)\n",
       "      (1): _Tanh()\n",
       "    )\n",
       "  )\n",
       "  (visible): _Linear(in_features=256, out_features=784, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "generator = Generator()\n",
    "generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "98941365",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "244944"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The number of parameters in our generator\n",
    "sum(p.numel() for p in generator.parameters())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "da42061c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([32, 28, 28]),\n",
       " tensor(0.0813),\n",
       " tensor(0.9768),\n",
       " tensor(0.6089),\n",
       " tensor(0.1613))"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Make sure that we get reasonable outputs for a batch of latent input vectors\n",
    "with th.inference_mode():\n",
    "    im = generator(th.randn(32, generator.latent_dim))\n",
    "\n",
    "plt.imshow(im[0], cmap=plt.cm.gray);\n",
    "plt.axis('off');\n",
    "plt.colorbar(shrink=0.8);\n",
    "plt.tight_layout();\n",
    "\n",
    "im.shape, im.min(), im.max(), im.mean(), im.std()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee9a2e91",
   "metadata": {},
   "source": [
    "## Discriminator\n",
    "\n",
    "Now we define our discriminator.  Again, we use an MLP-style network with a design that is very similar to our generator except that it takes images as it's inputs and outputs logits that indicate real or fake.  We also inject the mean image across each bach into the network as a simple form of minibatch discrimination.  This helps to prevent mode collapse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bdcf7585",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Discriminator(th.nn.Module):\n",
    "    '''MLP-style discriminator.  This is a binary classifier, i.e.,\n",
    "    there is no class conditioning.  This generator does, however,\n",
    "    use the mean images for each batch in order to perform a simple\n",
    "    version of \"minibatch discrimination\" that helps to prevent\n",
    "    mode collapse.\n",
    "    '''\n",
    "    def __init__(self, img_size=(28, 28), layer_specs=(256, 128, 64)):\n",
    "        '''Initialize a new discriminator.\n",
    "\n",
    "        img_size (tuple(int)):\n",
    "            Two-tuple of integers specifying the height and width of\n",
    "            the input images in pixels.\n",
    "        layer_specs (tuple(int)):\n",
    "            A tuple where each value specifies the number of hidden\n",
    "            units in each layer.  The number of values in this tuple\n",
    "            determines the number of layers.\n",
    "        '''\n",
    "        super().__init__()\n",
    "\n",
    "        self.img_size = img_size\n",
    "        self.in_dim = functools.reduce(operator.mul, img_size)\n",
    "\n",
    "        self.hidden = th.nn.Sequential()\n",
    "        layer_in, layer_out = None, self.in_dim * 2\n",
    "        for num_units in layer_specs:\n",
    "            layer_in, layer_out = layer_out, num_units\n",
    "            self.hidden.append(\n",
    "                th.nn.Sequential(\n",
    "                    _Linear(layer_in, layer_out),\n",
    "                    _Tanh(),\n",
    "                )\n",
    "            )\n",
    "\n",
    "        layer_in, layer_out = layer_out, 1\n",
    "        self.visible = _Linear(layer_in, layer_out)\n",
    "        \n",
    "    def forward(self, x):\n",
    "        '''Forward pass generates predicted class membership\n",
    "        probabilities for two classes: real and fake.\n",
    "        '''\n",
    "        # Means and variances are hard-coded for MNIST\n",
    "        x = (x - 0.1307) / 0.3081\n",
    "        \n",
    "        x = x.flatten(1)\n",
    "        batch_size, in_dim = x.shape\n",
    "        assert in_dim == self.in_dim\n",
    "\n",
    "        # We inject the mean images for each batch.  This is a\n",
    "        # simplistic form of minibatch discrimination that helps\n",
    "        # to prevent mode collapse.\n",
    "        x_mean = x.mean(dim=0, keepdim=True).expand(batch_size, -1)\n",
    "        x = th.cat((x, x_mean), dim=1)\n",
    "\n",
    "        return self.visible(self.hidden(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "112dd8c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Discriminator(\n",
       "  (hidden): Sequential(\n",
       "    (0): Sequential(\n",
       "      (0): _Linear(in_features=1568, out_features=256, bias=True)\n",
       "      (1): _Tanh()\n",
       "    )\n",
       "    (1): Sequential(\n",
       "      (0): _Linear(in_features=256, out_features=128, bias=True)\n",
       "      (1): _Tanh()\n",
       "    )\n",
       "    (2): Sequential(\n",
       "      (0): _Linear(in_features=128, out_features=64, bias=True)\n",
       "      (1): _Tanh()\n",
       "    )\n",
       "  )\n",
       "  (visible): _Linear(in_features=64, out_features=1, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discriminator = Discriminator()\n",
    "discriminator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "0f498184",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "442881"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The number of parameters in the discriminator\n",
    "sum(p.numel() for p in discriminator.parameters())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "af696c73",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(tensor(-0.7343),\n",
       " tensor(0.3366),\n",
       " tensor(-0.2641),\n",
       " tensor(0.2429),\n",
       " tensor([0.4252]))"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verify that we get a reasonable output\n",
    "with th.inference_mode():\n",
    "    logits = discriminator(im)\n",
    "\n",
    "logits.min(), logits.max(), logits.mean(), logits.std(), th.sigmoid(logits[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91531b9d",
   "metadata": {},
   "source": [
    "## Adversarial Training\n",
    "\n",
    "Now that the generator and discriminator are defined, we setup the training procedure.  This is a very simple adversarial training method where the discriminator minimizes cross-entropy for labeling real images as real and fake images as fake.  Simultaneously, the generator is trained produce fake images that the discriminator will label as real."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d2467427",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(generator, discriminator, data, epochs=350, batch_size=128, lr=0.0002,\n",
    "          lr_decay=0.99, label_smooth=0.9, device=0):\n",
    "    '''Simple, unconditioned adversarial training with cross-entropy loss.\n",
    "    \n",
    "    generator (torch.Module):\n",
    "        The generator module should take a batch of random latent vectors as\n",
    "        inputs and produce a batch images as outputs.\n",
    "    discriminator (torch.Module):\n",
    "        The discriminator should take a batch of images as inputs and assign\n",
    "        a binary class labels, fake or real, as outputs.\n",
    "    data (torch.utils.data.Dataset):\n",
    "        A Torch dataset that produces greyscale images in HW format.\n",
    "    epochs (int):\n",
    "        The number of training epochs.\n",
    "    batch_size (int):\n",
    "        The batch size.\n",
    "    lr (float):\n",
    "        The learning rate to use.\n",
    "    lr_decay (float):\n",
    "        Exponential decay rate for the learning rate, applied after\n",
    "        each training epoch.\n",
    "    label_smooth (float):\n",
    "        The amount of one-sided label smoothing to apply to the\n",
    "        discriminator.  This prevents very high confidence in\n",
    "        \"real\" labels, which helps to prevent very small gradients\n",
    "        that can slow training and result in mode collapse.\n",
    "    device (int):\n",
    "        The GPU device ID to use.\n",
    "    '''\n",
    "    generator.to(device)\n",
    "    discriminator.to(device)\n",
    "\n",
    "    generator.train()\n",
    "    discriminator.train()\n",
    "    \n",
    "    g_opt = th.optim.RMSprop(generator.parameters(), alpha=0.999, lr=lr)\n",
    "    d_opt = th.optim.RMSprop(discriminator.parameters(), alpha=0.999, lr=lr)\n",
    "    \n",
    "    g_sched = th.optim.lr_scheduler.ExponentialLR(g_opt, gamma=lr_decay)\n",
    "    d_sched = th.optim.lr_scheduler.ExponentialLR(d_opt, gamma=lr_decay)\n",
    "    \n",
    "    g_losses, d_losses = [], []\n",
    "    \n",
    "    dataloader = th.utils.data.DataLoader(\n",
    "        data,\n",
    "        batch_size=batch_size,\n",
    "        shuffle=True,\n",
    "        num_workers=4,\n",
    "        pin_memory=True,\n",
    "        drop_last=True,\n",
    "    )\n",
    "\n",
    "    for epoch in tqdm.trange(epochs):\n",
    "        for i, real_imgs in enumerate(dataloader):\n",
    "            real_imgs = real_imgs.to(device, non_blocking=True)\n",
    "            real_targs = th.ones(batch_size, 1, device=device)\n",
    "            fake_targs = th.zeros(batch_size, 1, device=device)\n",
    "\n",
    "            ## Generator\n",
    "\n",
    "            z = th.randn(batch_size, generator.latent_dim, device=device)\n",
    "            fake_imgs = generator(z)\n",
    "\n",
    "            g_loss = th.nn.functional.binary_cross_entropy_with_logits(\n",
    "                discriminator(fake_imgs), real_targs)\n",
    "\n",
    "            g_losses.append(g_loss.item())\n",
    "\n",
    "            g_opt.zero_grad()\n",
    "            g_loss.backward()\n",
    "            g_opt.step()\n",
    "\n",
    "            ## Discriminator\n",
    "\n",
    "            with th.no_grad():\n",
    "                z = th.randn(batch_size, generator.latent_dim, device=device)\n",
    "                fake_imgs = generator(z)\n",
    "\n",
    "            fake_loss = th.nn.functional.binary_cross_entropy_with_logits(\n",
    "                discriminator(fake_imgs), fake_targs)\n",
    "            real_loss = th.nn.functional.binary_cross_entropy_with_logits(\n",
    "                discriminator(real_imgs), real_targs * label_smooth)  \n",
    "            d_loss = (fake_loss + real_loss) / 2.\n",
    "\n",
    "            d_losses.append(d_loss.item())\n",
    "\n",
    "            d_opt.zero_grad()\n",
    "            d_loss.backward()            \n",
    "            d_opt.step()\n",
    "                \n",
    "        g_sched.step()\n",
    "        d_sched.step()\n",
    "            \n",
    "    generator.eval()\n",
    "    discriminator.eval()\n",
    "    \n",
    "    generator.cpu()\n",
    "    discriminator.cpu()\n",
    "    \n",
    "    return munch.Munch(\n",
    "        g_losses=g_losses,\n",
    "        d_losses=d_losses,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "8abfd99b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "adc44ff8da9b40d7858e43fe8f9db7be",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/350 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result = train(generator, discriminator, data, device=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8f90672e",
   "metadata": {},
   "source": [
    "## Results\n",
    "\n",
    "Here we see the loss for both the generator and discriminator.  Things look fairly stable, which is good.  The generator typically has higher loss, meaning that the generator is somewhat \"chasing\" the discriminator.  The variation in the loss settles down over time because of the learning rate scheduler.  These observations are generally good but it is often difficult to tell how well things are working by analyzing the losses alone.  Tracking a metric like FID would likely be helpful here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "a5195a98",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(result.g_losses);\n",
    "plt.title('Generator Loss During Training');\n",
    "plt.xlabel('Training Step');\n",
    "plt.ylabel('Loss');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "7610dcb6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(result.d_losses);\n",
    "plt.title('Discriminator Loss During Training');\n",
    "plt.xlabel('Training Step');\n",
    "plt.ylabel('Loss');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a3ec6da",
   "metadata": {},
   "source": [
    "## Examples of synthetic images\n",
    "\n",
    "Next, we see some images produced by our generator.  The images are a bit messy but we can clearly identify which digit is represented.  Many of these images appear to be believable as genuine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c86c0699",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generator.plot_montage();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "00b1bec9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generator.plot_montage();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90e50517",
   "metadata": {},
   "source": [
    "## Latent-space interpolation\n",
    "\n",
    "Next, we experiment with performing a linear interpolation between two latent space vectors and passing these interpolated latent vectors through our generator.  This shows how the images evolve from one digit to another as the latent-space input vectors vary."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "8a8f9d9d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "z0 = th.randn(1, generator.latent_dim)\n",
    "z1 = th.randn(1, generator.latent_dim)\n",
    "\n",
    "n = 36\n",
    "alpha = th.linspace(0., 1., n)[:, None]\n",
    "\n",
    "z = alpha * z0 + (1. - alpha) * z1\n",
    "\n",
    "generator.plot_montage(n=n, z=z);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "398d1186",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "z0 = th.randn(1, generator.latent_dim)\n",
    "z1 = th.randn(1, generator.latent_dim)\n",
    "\n",
    "n = 36\n",
    "alpha = th.linspace(0., 1., n)[:, None]\n",
    "\n",
    "z = alpha * z0 + (1. - alpha) * z1\n",
    "\n",
    "generator.plot_montage(n=n, z=z);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "d845079c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "z0 = th.randn(1, generator.latent_dim)\n",
    "z1 = th.randn(1, generator.latent_dim)\n",
    "\n",
    "n = 36\n",
    "alpha = th.linspace(0., 1., n)[:, None]\n",
    "\n",
    "z = alpha * z0 + (1. - alpha) * z1\n",
    "\n",
    "generator.plot_montage(n=n, z=z);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "47a9237e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x792 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "z0 = th.randn(1, generator.latent_dim)\n",
    "z1 = th.randn(1, generator.latent_dim)\n",
    "\n",
    "n = 36\n",
    "alpha = th.linspace(0., 1., n)[:, None]\n",
    "\n",
    "z = alpha * z0 + (1. - alpha) * z1\n",
    "\n",
    "generator.plot_montage(n=n, z=z);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7b6275a",
   "metadata": {},
   "source": [
    "And that concludes my simple demonstration of unconditioned GANs on the MNIST dataset!  Although this approach seems to work, we can definitely do better.  Check out my next notebook on conditioned GANs with MNIST in order to see how adding in class label information improves performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "995fc4bb",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}