DCGAN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "DCGAN.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "BZm5pN8Q-7Es",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KcvNSARf_C_q",
        "colab_type": "code",
        "outputId": "4408d530-4a15-4649-8001-61cdabbb781d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tf.__version__"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'2.2.0-rc4'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fo-ChHDE_QFC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip install -q imageio"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "McpVGfuf_QpK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import glob\n",
        "import imageio\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import os\n",
        "import PIL\n",
        "from tensorflow.keras import layers\n",
        "import time\n",
        "\n",
        "from IPython import display"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DtTeOWM_BRlu",
        "colab_type": "text"
      },
      "source": [
        "***Loading data***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jrsug4WX_QmR",
        "colab_type": "code",
        "outputId": "000171ca-1dbf-4005-aed2-bafc0187adfe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "(train_img, train_labels), (_,_) = tf.keras.datasets.mnist.load_data()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 0s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PiU8kl9qBWjn",
        "colab_type": "text"
      },
      "source": [
        "***Reshaping the data and converting its type and normalizing the data***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mP0addTQ_QjX",
        "colab_type": "code",
        "outputId": "7acbee98-e1a6-4fce-be77-938fd95e385a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train_img.shape[0]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "60000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pNnh4fNH_Qew",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Reshaping the data and converting its type\n",
        "train_img = train_img.reshape(train_img.shape[0],28,28,1).astype('float32')\n",
        "#Normalization of data in the range [-1,1]\n",
        "train_img = (train_img - 127.5) / 127.5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C88ZxqQiCF7e",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "BUFFER_SIZE = 60000\n",
        "BATCH_SIZE = 256"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fDIFOCMgCaY9",
        "colab_type": "text"
      },
      "source": [
        "***Batch and shuffle the data***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VQpgFERbCi9v",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_dataset = tf.data.Dataset.from_tensor_slices(train_img).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v_AoXE3bFMih",
        "colab_type": "text"
      },
      "source": [
        "***Generator Model***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UmKDLAf7C_xL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def make_generator_model():\n",
        "  model = tf.keras.Sequential([\n",
        "              # Dense Layer\n",
        "              layers.Dense(7*7*256, use_bias=False, input_shape=(100,)),\n",
        "              layers.BatchNormalization(),\n",
        "              layers.LeakyReLU(),\n",
        "              layers.Reshape((7,7,256)),\n",
        "              # Conv2DTranspose Layers\n",
        "              layers.Conv2DTranspose(128,(5,5), strides=(1,1), padding='same', use_bias=False),\n",
        "              layers.BatchNormalization(),\n",
        "              layers.LeakyReLU(),\n",
        "\n",
        "              layers.Conv2DTranspose(64, (5,5), strides=(2,2), padding='same', use_bias=False),\n",
        "              layers.BatchNormalization(),\n",
        "              layers.LeakyReLU(),\n",
        "\n",
        "              layers.Conv2DTranspose(1, (5,5), strides=(2,2), padding='same', use_bias=False),\n",
        "              layers.BatchNormalization(),\n",
        "              layers.LeakyReLU()\n",
        "  ])\n",
        "  return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y22DxenzSCFr",
        "colab_type": "text"
      },
      "source": [
        "***Creating an image using Generator without training***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zbADo95TR2pp",
        "colab_type": "code",
        "outputId": "abe3ee36-5a8e-4ab4-a257-f095e6e39e94",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        }
      },
      "source": [
        "generator = make_generator_model()\n",
        "\n",
        "noise = tf.random.normal([1,100])\n",
        "generated_image = generator(noise, training=False)\n",
        "plt.imshow(generated_image[0,:,:,0], cmap='gray')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f6f7039c5f8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cahRRCV3Tiga",
        "colab_type": "text"
      },
      "source": [
        "***Discriminator Model***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nQKJkd8zS1EY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def make_discriminator_model():\n",
        "  model = tf.keras.Sequential([\n",
        "                # Conv2D layers aka 2D convolutions\n",
        "                layers.Conv2D(64, (5,5), strides=(2,2), padding='same', input_shape=[28,28,1]),\n",
        "                layers.LeakyReLU(),\n",
        "                layers.Dropout(0.3),\n",
        "                \n",
        "                layers.Conv2D(128, (5,5), strides=(2,2), padding='same'),\n",
        "                layers.LeakyReLU(),\n",
        "                layers.Dropout(0.3),\n",
        "                # Flattening the image\n",
        "                layers.Flatten(),\n",
        "                # Dense Layer\n",
        "                layers.Dense(1)\n",
        "  ])\n",
        "  return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jwt-9ARfWWFB",
        "colab_type": "text"
      },
      "source": [
        "***Classifying generated image as real or fake using Discriminator***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_Z8zigBwV7RZ",
        "colab_type": "code",
        "outputId": "7533bfa7-b048-4868-bc0e-6d3a63bfbe16",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "discriminator = make_discriminator_model()\n",
        "decision = discriminator(generated_image)\n",
        "print(decision)\n",
        "if decision < 0:\n",
        "  print('Fake Image')\n",
        "else:\n",
        "  print('Real Image')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tf.Tensor([[8.353386e-05]], shape=(1, 1), dtype=float32)\n",
            "Real Image\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mj8p88H-YFWe",
        "colab_type": "text"
      },
      "source": [
        "***Define Loss Function***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-dmbWEKXWRNV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "loss = tf.keras.losses.BinaryCrossentropy(from_logits=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3ax3P_scYmqP",
        "colab_type": "text"
      },
      "source": [
        "***Discriminator Loss***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "heehBBKdYu4m",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_discriminator_loss(real_op, fake_op):\n",
        "  real_loss = loss(tf.ones_like(real_op), real_op)\n",
        "  fake_loss = loss(tf.zeros_like(fake_op), fake_op)\n",
        "  total_loss = real_loss + fake_loss\n",
        "  return total_loss"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WpPncYXnZhvC",
        "colab_type": "text"
      },
      "source": [
        "***Generator Loss***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dviGpQAzZpsV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_generator_loss(fake_op):\n",
        "  return loss(tf.ones_like(fake_op), fake_op)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QHzIfwUxZ63m",
        "colab_type": "text"
      },
      "source": [
        "***Generator and Discriminator Optimizers***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "txB9xmmqaBZr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "generator_optimizer = tf.keras.optimizers.Adam(1e-4)\n",
        "discriminator_optimizer = tf.keras.optimizers.Adam(1e-4)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gvTt3Bgfl0YT",
        "colab_type": "text"
      },
      "source": [
        "***Checkpoints***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u097B45rl4zt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "checkpoint_dir = './training_checkpoints'\n",
        "checkpoint_prefix = os.path.join(checkpoint_dir,\"ckpt\")\n",
        "checkpoint = tf.train.Checkpoint(generator_optimizer = generator_optimizer,\n",
        "                                 discriminator_optimizer = discriminator_optimizer,\n",
        "                                 generator = generator,\n",
        "                                 discriminator = discriminator)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rcBIMtbVljpy",
        "colab_type": "text"
      },
      "source": [
        "***Training Loop***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JtbvD2GSlh19",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "EPOCHS = 500\n",
        "noise_dim = 100\n",
        "num_examples_to_generate = 16\n",
        "\n",
        "seed = tf.random.normal([num_examples_to_generate, noise_dim])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r39jy-VNo2C2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "@tf.function\n",
        "def train_step(images):\n",
        "  noise = tf.random.normal([BATCH_SIZE, noise_dim])\n",
        "\n",
        "  with tf.GradientTape() as gen_tape, tf.GradientTape() as dis_tape:\n",
        "    generated_image = generator(noise, training = True)\n",
        "\n",
        "    real_op = discriminator(images, training = True)\n",
        "    fake_op = discriminator(generated_image, training = True)\n",
        "    generator_loss = get_generator_loss(fake_op)\n",
        "    discriminator_loss = get_discriminator_loss(real_op, fake_op)\n",
        "\n",
        "    gradients_of_generator = gen_tape.gradient(generator_loss, generator.trainable_variables)\n",
        "    gradients_of_discriminator = dis_tape.gradient(discriminator_loss, discriminator.trainable_variables)\n",
        "\n",
        "    generator_optimizer.apply_gradients(zip(gradients_of_generator, generator.trainable_variables))\n",
        "    discriminator_optimizer.apply_gradients(zip(gradients_of_discriminator, discriminator.trainable_variables))\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-wS1x9Jntw7Y",
        "colab_type": "text"
      },
      "source": [
        "***Training***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hdftiKrHsXj7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(dataset, epochs):\n",
        "  for epoch in range(epochs):\n",
        "    start = time.time()\n",
        "    # for each batch of data, get gradients and generated image and its verification by discriminator\n",
        "    for img_batch in dataset:\n",
        "      train_step(img_batch)\n",
        "    # Produce gif on go\n",
        "    display.clear_output(wait=True)\n",
        "    # Generate image and save them\n",
        "    generate_and_save_images(generator, epoch+1, seed)\n",
        "    # Save model in 15 seconds\n",
        "    if (epoch + 1) % 15 == 0:\n",
        "      checkpoint.save(file_prefix=checkpoint_prefix)\n",
        "    print('Time for epoch {} is {} sec'.format(epoch+1, time.time()-start))\n",
        "  \n",
        "  # Genearate after final epoch\n",
        "  display.clear_output(wait=True)\n",
        "  generate_and_save_images(generator, epochs, seed)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PZRxbQhMwzWh",
        "colab_type": "text"
      },
      "source": [
        "***Generate and Save Images***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-rwsWcpnw-Wt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def generate_and_save_images(model, epoch, test_input):\n",
        "  predictions = model(test_input, training=False)\n",
        "\n",
        "  fig = plt.figure(figsize=(4,4))\n",
        "  # Plotting images \n",
        "  for i in range(predictions.shape[0]):\n",
        "    plt.plot(4,4,i+1)\n",
        "    plt.imshow(predictions[i,:,:,0] * 127.5 + 127.5, cmap='gray')\n",
        "    plt.axis('off')\n",
        "  # Saving figure\n",
        "  plt.savefig('image_at_epoch_{:04d}.png'.format(epoch))\n",
        "  plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NUClAJTbzCrW",
        "colab_type": "text"
      },
      "source": [
        "***Start Traing the model***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lq0Tod6KzIms",
        "colab_type": "code",
        "outputId": "e0ba0c24-ef8f-41ad-d918-88b031d7d7f6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "train(train_dataset, EPOCHS)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Time for epoch 239 is 13.093102931976318 sec\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Nr2J2thLztGx",
        "colab_type": "text"
      },
      "source": [
        "***Restore latest checkpoint***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ksLdw9tzzxzp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "checkpoint.restore(tf.train.latest_checkpoint(checkpoint_dir))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fj_5-kEg0JDE",
        "colab_type": "text"
      },
      "source": [
        "***Creating GIF***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cDLSsOJL0Pvz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def display_image(epoch_no):\n",
        "  return PIL.Image.open('image_at_epoch_{:04d}.png'.format(epoch_no))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TxTd2DZs0jx5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "display_image(EPOCHS)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GMt78Is70n1X",
        "colab_type": "text"
      },
      "source": [
        "***Creating animated GIF***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2wvLbjS00ri9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "anim_file = 'dcgan.gif'\n",
        "\n",
        "with imageio.get_writer(anim_file, mode='I') as writer:\n",
        "  filenames = glob.glob('image*.png')\n",
        "  filenames = sorted(filenames)\n",
        "  last = -1\n",
        "  for i,filename in enumerate(filenames):\n",
        "    frame = 2*(i**0.5)\n",
        "    if round(frame) > round(last):\n",
        "      last = frame\n",
        "    else:\n",
        "      continue\n",
        "    image = imageio.imread(filename)\n",
        "    writer.append_data(image)\n",
        "  image = imageio.imread(filename)\n",
        "  writer.append_data(image)\n",
        "\n",
        "import IPython\n",
        "if IPython.version_info > (6,2,0,''):\n",
        "  display.Image(filename=anim_file)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Se0I1gNd1F5h",
        "colab_type": "text"
      },
      "source": [
        "***Downloading GIF***"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HDjLIM8U1JUA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "try:\n",
        "  from google.colab import files\n",
        "except ImportError:\n",
        "   pass\n",
        "else:\n",
        "  files.download(anim_file)"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}