DCGan.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import time\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras import layers\n",
    "import PIL\n",
    "from IPython import display\n",
    "from tensorflow.keras.datasets import mnist\n",
    "from tensorflow import GradientTape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "(X_train,y_train),(X_test,y_test)=mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28)\n",
      "(10000, 28, 28)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1)\n"
     ]
    }
   ],
   "source": [
    "train_images=X_train.reshape(X_train.shape[0],28,28,1).astype('float32')\n",
    "print(train_images.shape)\n",
    "train_images=(train_images-127.5)/127.5 #Normalise to -1 to +1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "BUFFER_SIZE=60000\n",
    "Batch_size=256\n",
    "\n",
    "#Convert train_images to a tf.data.Dataset\n",
    "train_dataset=tf.data.Dataset.from_tensor_slices(train_images).shuffle(BUFFER_SIZE).batch(Batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<BatchDataset shapes: (None, 28, 28, 1), types: tf.float32>\n"
     ]
    }
   ],
   "source": [
    "print(train_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_Genrator_Model():\n",
    "    model=tf.keras.Sequential()\n",
    "    model.add(layers.Dense(7*7*256,use_bias=False,input_shape=(100,)))\n",
    "    model.add(layers.BatchNormalization())\n",
    "    model.add(layers.LeakyReLU())\n",
    "    model.add(layers.Reshape((7,7,256)))\n",
    "    \n",
    "    assert model.output_shape==(None,7,7,256) #Debug\n",
    "    \n",
    "    model.add(layers.Conv2DTranspose(128,(5,5),strides=(1,1),padding='same',use_bias=False))\n",
    "    assert model.output_shape==(None,7,7,128)\n",
    "    model.add(layers.BatchNormalization())\n",
    "    model.add(layers.LeakyReLU())\n",
    "    \n",
    "    model.add(layers.Conv2DTranspose(64,(5,5),strides=(2,2),padding='same',use_bias=False))\n",
    "    assert model.output_shape==(None,14,14,64)\n",
    "    model.add(layers.BatchNormalization())\n",
    "    model.add(layers.LeakyReLU())\n",
    "    \n",
    "    model.add(layers.Conv2DTranspose(1,(5,5),strides=(2,2),padding='same',use_bias=False))\n",
    "    assert model.output_shape==(None,28,28,1)\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator=make_Genrator_Model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2a3d11a9940>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "noise=tf.random.normal([1,100])\n",
    "generated_image=generator(noise,training=False)\n",
    "plt.imshow(generated_image[0,:,:,0],cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_decriminator_model():\n",
    "    model=tf.keras.Sequential()\n",
    "    \n",
    "    model.add(layers.Conv2D(64,(5,5),strides=(2,2),padding='same',input_shape=[28,28,1]))\n",
    "    model.add(layers.LeakyReLU())\n",
    "    model.add(layers.Dropout(0.3))\n",
    "    \n",
    "    model.add(layers.Conv2D(128,(5,5),strides=(2,2),padding='same'))\n",
    "    model.add(layers.LeakyReLU())\n",
    "    model.add(layers.Dropout(0.3))\n",
    "          \n",
    "    model.add(layers.Flatten())\n",
    "    model.add(layers.Dense(1))\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_5\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_2 (Conv2D)            (None, 14, 14, 64)        1664      \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_10 (LeakyReLU)   (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 7, 7, 128)         204928    \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_11 (LeakyReLU)   (None, 7, 7, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 7, 7, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 6272)              0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 1)                 6273      \n",
      "=================================================================\n",
      "Total params: 212,865\n",
      "Trainable params: 212,865\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "discriminator=make_decriminator_model()\n",
    "discriminator.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "cross_entropy=tf.keras.losses.BinaryCrossentropy(from_logits=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "def discriminator_loss(real_output,fake_output):\n",
    "    real_loss=cross_entropy(tf.ones_like(real_output),real_output)\n",
    "    fake_loss=cross_entropy(tf.zeros_like(fake_output),fake_output)\n",
    "    total_loss=real_loss+fake_loss\n",
    "    return total_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generator_loss(fake_output):\n",
    "       return cross_entropy(tf.ones_like(fake_output),fake_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator_optimizer=tf.keras.optimizers.Adam(1e-4)\n",
    "discriminator_optimizer=tf.keras.optimizers.Adam(1e-4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "checkpoint_dir='./training_checkpoints_Gan'\n",
    "checkpoint_prefix=os.path.join(checkpoint_dir,'ckpt')\n",
    "checkpoint=tf.train.Checkpoint(generator_optimizer=generator_optimizer,discriminator_optimizer=discriminator_optimizer,\n",
    "                              generator=generator,discriminator=discriminator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16, 100)\n"
     ]
    }
   ],
   "source": [
    "EPOCHS=100\n",
    "noise_dim=100\n",
    "num_ex_to_generate=16\n",
    "test_random_vectors=tf.random.normal([num_ex_to_generate,noise_dim])\n",
    "print(test_random_vectors.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "@tf.function\n",
    "def train_step(images):\n",
    "    noise=tf.random.normal([Batch_size,noise_dim])\n",
    "    \n",
    "    with GradientTape(persistent=True) as gen_tape,GradientTape(persistent=True) as disc_tape:\n",
    "        genrated_images=generator(noise,training=True)\n",
    "        \n",
    "        real_output=discriminator(images,training=True)\n",
    "        fake_output=discriminator(genrated_images,training=True)\n",
    "        \n",
    "        gen_loss=generator_loss(fake_output)\n",
    "        disc_loss=discriminator_loss(real_output,fake_output)\n",
    "        \n",
    "    gradient_of_generator=gen_tape.gradient(gen_loss,generator.trainable_variables)\n",
    "    gradient_of_discriminator=gen_tape.gradient(disc_loss,discriminator.trainable_variables)\n",
    "    \n",
    "    generator_optimizer.apply_gradients(zip(gradient_of_generator,generator.trainable_variables))\n",
    "    generator_optimizer.apply_gradients(zip(gradient_of_discriminator,discriminator.trainable_variables))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(dataset,epochs):\n",
    "    for epoch in range(epochs):\n",
    "        start=time.time()\n",
    "\n",
    "        for image_batch in dataset:\n",
    "            train_step(image_batch)\n",
    "            \n",
    "        display.clear_output(wait=True)\n",
    "        generate_and_save_images(generator,epoch+1,test_random_vectors)\n",
    "        \n",
    "        if (epoch+1)%10==0:\n",
    "            checkpoint.save(file_prefix=checkpoint_prefix)\n",
    "            \n",
    "        print('Time for epoch {} is {} sec'.format(epoch+1,time.time()-start))\n",
    "    display.clear_output(wait=True)\n",
    "    generate_and_save_images(generator,epochs,test_random_vectors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_and_save_images(model,epoch,test_input):\n",
    "    predictions=model(test_input,training=False)\n",
    "    fig=plt.figure(figsize=(4,4))\n",
    "    for i in range(predictions.shape[0]):\n",
    "        plt.subplot(4,4,i+1)\n",
    "        plt.imshow(predictions[i,:,:,0]*127.5+127.5,cmap='gray')\n",
    "        plt.axis('off')\n",
    "    plt.savefig('image_at_epoch{:04d}.png'.format(epoch))\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train(train_dataset,EPOCHS)"
   ]
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}