for e in range(1, config.num_epochs + 1):
print('-' * 15, 'Epoch %d' % e, '-' * 15)
for _ in range(config.steps_per_epoch):
image_batch_lr, image_batch_hr = next(
image_generator_g(config.batch_size, train_dir))
generated_images_sr = generator.predict(image_batch_lr)
real_data_Y = np.ones(config.batch_size) - np.random.random_sample(
config.batch_size) * 0.2
fake_data_Y = np.random.random_sample(config.batch_size) * 0.2
discriminator.trainable = True
d_loss_real = discriminator.train_on_batch(image_batch_hr, real_data_Y)
d_loss_fake = discriminator.train_on_batch(generated_images_sr,
fake_data_Y)
#d_loss = 0.5 * np.add(d_loss_fake, d_loss_real)
image_batch_lr, image_batch_hr = next(
image_generator_d(config.batch_size, train_dir))
gan_Y = np.ones(config.batch_size) - np.random.random_sample(
config.batch_size) * 0.2
discriminator.trainable = False
loss_gan = gan.train_on_batch(image_batch_lr, [image_batch_hr, gan_Y])
print("Loss HR , Loss LR, Loss GAN", "Perceptual Distance")
print(d_loss_real, d_loss_fake, loss_gan)
def train(epochs, batch_size, input_dir, output_dir, model_save_dir,
number_of_images, train_test_ratio):
# train_test_ratio 基本上是控制train data 的數量
x_train_lr, x_train_hr, x_test_lr, x_test_hr = Utils.load_training_data(
input_dir,
'.png',
number_of_images=number_of_images,
train_test_ratio=train_test_ratio)
loss = VGG_LOSS(image_shape)
batch_count = int(x_train_hr.shape[0] / batch_size)
shape = (image_shape[0] // downscale_factor,
image_shape[1] // downscale_factor, image_shape[2])
generator = Generator(shape).generator()
discriminator = Discriminator(image_shape).discriminator()
plot_model(discriminator, "d-model-2.png")
optimizer = Utils_model.get_optimizer()
generator.compile(loss=loss.vgg_loss, optimizer=optimizer)
discriminator.compile(loss="binary_crossentropy", optimizer=optimizer)
gan = get_gan_network(discriminator, shape, generator, optimizer,
loss.vgg_loss)
loss_file = open(model_save_dir + 'losses.txt', 'w+')
loss_file.close()
for e in range(1, epochs + 1):
print('-' * 15, 'Epoch %d' % e, '-' * 15)
for _ in tqdm(range(batch_count)):
rand_nums = np.random.randint(0,
x_train_hr.shape[0],
size=batch_size)
image_batch_hr = x_train_hr[rand_nums]
image_batch_lr = x_train_lr[rand_nums]
generated_images_sr = generator.predict(image_batch_lr)
real_data_Y = np.ones(
batch_size) - np.random.random_sample(batch_size) * 0.2
fake_data_Y = np.random.random_sample(batch_size) * 0.2
discriminator.trainable = True
d_loss_real = discriminator.train_on_batch(image_batch_hr,
real_data_Y)
d_loss_fake = discriminator.train_on_batch(generated_images_sr,
fake_data_Y)
discriminator_loss = 0.5 * np.add(d_loss_fake, d_loss_real)
rand_nums = np.random.randint(0,
x_train_hr.shape[0],
size=batch_size)
image_batch_hr = x_train_hr[rand_nums]
image_batch_lr = x_train_lr[rand_nums]
gan_Y = np.ones(
batch_size) - np.random.random_sample(batch_size) * 0.2
discriminator.trainable = False
gan_loss = gan.train_on_batch(image_batch_lr,
[image_batch_hr, gan_Y])
gan_names = gan.metrics_names
print(gan_names)
break
print("discriminator loss : %f" % discriminator_loss)
print("gan_loss :", gan_loss)
gan_loss = str(gan_loss)
loss_file = open(model_save_dir + 'losses.txt', 'a')
loss_file.write('epoch%d : gan_loss = %s ; discriminator loss = %f\n' %
(e, gan_loss, discriminator_loss))
loss_file.close()
break
if e == 1 or e % 20 == 0:
Utils.plot_generated_images(output_dir, e, generator, x_test_hr,
x_test_lr)
if e % 10 == 0:
generator.save(model_save_dir + 'gen_model%d.h5' % e)
discriminator.save(model_save_dir + 'dis_model%d.h5' % e)
