MINST GAN Example

minst_gan_batch_train.py

# MINST GAN Example
# based on https://www.tensorflow.org/tutorials/generative/dcgan

import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
from tensorflow.keras import layers, models
import time

BUFFER_SIZE = 60000
BATCH_SIZE = 300

(train_images, train_labels), (_, _) = tf.keras.datasets.mnist.load_data()
train_images = train_images.reshape(train_images.shape[0], 28, 28, 1).astype('float32')
train_images = (train_images - 127.5) / 127.5  # Normalize the images to [-1, 1]

train_dataset = tf.data.Dataset.from_tensor_slices(train_images).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)

discriminator = models.Sequential([
    layers.Conv2D(64, (5, 5), strides=(2, 2), padding='same', input_shape=[28, 28, 1]),
    layers.LeakyReLU(),
    layers.Dropout(0.3),
    layers.Conv2D(128, (5, 5), strides=(2, 2), padding='same'),
    layers.LeakyReLU(),
    layers.Dropout(0.3),
    layers.Flatten(),
    layers.Dense(1)
])
discriminator.compile(loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
                      optimizer=tf.keras.optimizers.Adam(1e-4))

generator = models.Sequential([
    layers.Dense(7 * 7 * 256, use_bias=False, input_shape=(100,)),
    layers.BatchNormalization(),
    layers.LeakyReLU(),
    layers.Reshape((7, 7, 256)),
    layers.Conv2DTranspose(128, (5, 5), strides=(1, 1), padding='same', use_bias=False),
    layers.BatchNormalization(),
    layers.LeakyReLU(),
    layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False),
    layers.BatchNormalization(),
    layers.LeakyReLU(),
    layers.Conv2DTranspose(1, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh')
])

discriminator.trainable = False
gan = models.Sequential([
    generator,
    discriminator
])
gan.compile(loss=tf.keras.losses.BinaryCrossentropy(from_logits=True), optimizer=tf.keras.optimizers.Adam(1e-4))

EPOCHS = 50
noise_dim = 100
num_examples_to_generate = 16

seed = tf.random.normal([num_examples_to_generate, noise_dim])


def train_step(images):
    noise = tf.random.normal([BATCH_SIZE, noise_dim])
    fake_images = generator.predict(noise)

    real_images_y = np.ones((BATCH_SIZE, 1))
    discriminator.train_on_batch(images, real_images_y)

    fake_images_y = np.zeros((BATCH_SIZE, 1))
    discriminator.train_on_batch(fake_images, fake_images_y)

    gan_y = np.ones((BATCH_SIZE, 1))
    gan.train_on_batch(noise, gan_y)


for epoch in range(EPOCHS):
    start = time.time()

    for image_batch in train_dataset:
        train_step(image_batch)

    predictions = generator(seed, training=False)
    plt.figure(figsize=(4, 4))
    for i in range(predictions.shape[0]):
        plt.subplot(4, 4, i + 1)
        plt.imshow(predictions[i, :, :, 0] * 127.5 + 127.5, cmap='gray')
        plt.axis('off')
    plt.show()

    print('Time for epoch {} is {} sec'.format(epoch + 1, time.time() - start))