GAN in PyTorch.ipynb

{
 "cells": [
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Implementation of a GAN to generate hand written digits from MNIST dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn.functional as nn\n",
    "import torch.autograd as autograd\n",
    "import torch.optim as optim\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.gridspec as gridspec\n",
    "import os\n",
    "from torch.autograd import Variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# So we need MNIST data , tensorflow provides a MNIST data\n",
    "from tensorflow.examples.tutorials.mnist import input_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "mnist = input_data.read_data_sets('../../MNIST_data', one_hot=True)\n",
    "# set the min batch size\n",
    "mb_size = 64\n",
    "# the number of samples to take from a random distribution\n",
    "Z_dim = 100\n",
    "\n",
    "# The number of training examples\n",
    "X_dim = mnist.train.images.shape[1]\n",
    "y_dim = mnist.train.labels.shape[1]\n",
    "\n",
    "# The size of hidden layer\n",
    "h_dim = 128\n",
    "\n",
    "# Set the initial cost to zero\n",
    "c = 0\n",
    "\n",
    "# Set the learning rate\n",
    "lr = 1e-3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lets define the 'Xavier Initialization function'\n",
    "def xavier_init(size):\n",
    "    in_dim = size[0]\n",
    "    xavier_stddev = 1. / np.sqrt(in_dim / 2.)\n",
    "    return Variable(torch.randn(*size) * xavier_stddev, requires_grad=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let's start with the generator network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### First the variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Wzh = xavier_init(size=[Z_dim, h_dim])\n",
    "bzh = Variable(torch.zeros(h_dim), requires_grad=True)\n",
    "\n",
    "Whx = xavier_init(size=[h_dim, X_dim])\n",
    "bhx = Variable(torch.zeros(X_dim), requires_grad=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Then the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def G(z):\n",
    "    h = nn.relu(z @ Wzh + bzh.repeat(z.size(0), 1))\n",
    "    X = nn.sigmoid(h @ Whx + bhx.repeat(h.size(0), 1))\n",
    "    return X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The Discriminator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### The variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Wxh = xavier_init(size=[X_dim, h_dim])\n",
    "bxh = Variable(torch.zeros(h_dim), requires_grad=True)\n",
    "\n",
    "Why = xavier_init(size=[h_dim, 1])\n",
    "bhy = Variable(torch.zeros(1), requires_grad=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### The discriminator network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def D(X):\n",
    "    h = nn.relu(X @ Wxh + bxh.repeat(X.size(0), 1))\n",
    "    y = nn.sigmoid(h @ Why + bhy.repeat(h.size(0), 1))\n",
    "    return y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Store all the parameters under a single roof\n",
    "G_params = [Wzh, bzh, Whx, bhx]\n",
    "D_params = [Wxh, bxh, Why, bhy]\n",
    "params = G_params + D_params"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets begin to code the training process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def reset_grad():\n",
    "    for p in params:\n",
    "        if p.grad is not None:\n",
    "            data = p.grad.data\n",
    "            p.grad = Variable(data.new().resize_as_(data).zero_())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Declare the optimizers for generator\n",
    "\n",
    "G_solver = optim.Adam(G_params, lr=1e-3)\n",
    "\n",
    "# Declare the optimizers for discriminator\n",
    "D_solver = optim.Adam(D_params, lr=1e-3)\n",
    "\n",
    "\n",
    "# The targets for the discriminator.\n",
    "# Zeros for the samples coming from generator and one for the data coming from sample space\n",
    "ones_label = Variable(torch.ones(mb_size))\n",
    "zeros_label = Variable(torch.zeros(mb_size))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The actual training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "gen_loss=[]\n",
    "dis_loss=[]\n",
    "\n",
    "for it in range(100000):\n",
    "    # Sample data\n",
    "    z = Variable(torch.randn(mb_size, Z_dim))\n",
    "    X, _ = mnist.train.next_batch(mb_size)\n",
    "    \n",
    "    #The iputs should be a troch tensor\n",
    "    X = Variable(torch.from_numpy(X))\n",
    "\n",
    "    # Dicriminator forward-loss-backward-update\n",
    "    G_sample = G(z)\n",
    "    D_real = D(X)\n",
    "    D_fake = D(G_sample)\n",
    "\n",
    "    D_loss_real = nn.binary_cross_entropy(D_real, ones_label)\n",
    "    D_loss_fake = nn.binary_cross_entropy(D_fake, zeros_label)\n",
    "    D_loss = D_loss_real + D_loss_fake\n",
    "\n",
    "    D_loss.backward()\n",
    "    D_solver.step()\n",
    "\n",
    "    # Housekeeping - reset gradient\n",
    "    reset_grad()\n",
    "\n",
    "    # Generator forward-loss-backward-update\n",
    "    z = Variable(torch.randn(mb_size, Z_dim))\n",
    "    G_sample = G(z)\n",
    "    D_fake = D(G_sample)\n",
    "\n",
    "    G_loss = nn.binary_cross_entropy(D_fake, ones_label)\n",
    "\n",
    "    G_loss.backward()\n",
    "    G_solver.step()\n",
    "\n",
    "    # Housekeeping - reset gradient\n",
    "    reset_grad()\n",
    "\n",
    "    # Print and plot every now and then\n",
    "    if it % 1000 == 0:\n",
    "        print('Iter-{}; D_loss: {}; G_loss: {}'.format(it, D_loss.data.numpy(), G_loss.data.numpy()))\n",
    "        dis_loss.append(D_loss.data.numpy())\n",
    "        gen_loss.append(G_loss.data.numpy())\n",
    "\n",
    "        samples = G(z).data.numpy()[:16]\n",
    "\n",
    "        fig = plt.figure(figsize=(4, 4))\n",
    "        gs = gridspec.GridSpec(4, 4)\n",
    "        gs.update(wspace=0.05, hspace=0.05)\n",
    "\n",
    "        for i, sample in enumerate(samples):\n",
    "            ax = plt.subplot(gs[i])\n",
    "            plt.axis('off')\n",
    "            ax.set_xticklabels([])\n",
    "            ax.set_yticklabels([])\n",
    "            ax.set_aspect('equal')\n",
    "            plt.imshow(sample.reshape(28, 28), cmap='Greys_r')\n",
    "\n",
    "        plt.savefig('out-pytorch-gan/{}.png'.format(str(c).zfill(3)), bbox_inches='tight')\n",
    "        c += 1\n",
    "        plt.close(fig)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "with open(\"./out-pytorch-gan/losses_pickles/genloss.pkl\",\"wb\") as f:\n",
    "    pickle.dump(gen_loss,f)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"./out-pytorch-gan/losses_pickles/disloss.pkl\",\"wb\") as f:\n",
    "    pickle.dump(dis_loss,f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW4AAAD8CAYAAABXe05zAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f611b1d0a20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with open(\"./out-pytorch-gan/losses_pickles/genloss.pkl\",\"rb\") as f:\n",
    "    gen_loss=pickle.load(f)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(gen_loss)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f60edbcceb8>]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(\"./out-pytorch-gan/losses_pickles/disloss.pkl\",\"rb\") as f:\n",
    "    dis_loss=pickle.load(f)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(dis_loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6118254780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}