def main(args):
del args
dataset_name = FLAGS.dataset or os.path.basename(
os.path.dirname(os.path.dirname(FLAGS.ckpt)))
try:
dataset = data.get_dataset(dataset_name)
except KeyError:
dataset = data.get_dataset('lsun_' + dataset_name)
ops = get_ops(dataset)
images = load_hires(dataset, get_samples_indexes(FLAGS.samples))
image_grid = get_candidates(ops, images)
img = utils.images_to_grid(image_grid)
output_file = os.path.abspath(FLAGS.save_to)
os.makedirs(os.path.dirname(output_file), exist_ok=True)
open(output_file, 'wb').write(utils.to_png(img))
print('Saved', output_file)
def main(argv):
del argv # Unused.
dataset = data.get_dataset(FLAGS.dataset)
schedule = TrainSchedule(2, FLAGS.scale, FLAGS.transition_kimg, FLAGS.training_kimg, FLAGS.total_kimg)
if FLAGS.memtest:
schedule.schedule = schedule.schedule[-2:]
model = LAG(
os.path.join(FLAGS.train_dir, dataset.name),
lr=FLAGS.lr,
batch=FLAGS.batch,
lod_min=1,
scale=FLAGS.scale,
downscaler=FLAGS.downscaler,
blocks=FLAGS.blocks,
filters=FLAGS.filters,
filters_min=FLAGS.filters_min,
mse_weight=FLAGS.mse_weight,
noise_dim=FLAGS.noise_dim,
transition_kimg=FLAGS.transition_kimg,
training_kimg=FLAGS.training_kimg,
ttur=FLAGS.ttur,
wass_target=FLAGS.wass_target,
weight_avg=FLAGS.weight_avg)
if FLAGS.reset:
model.reset_files()
model.train(dataset, schedule)
def main(argv):
del argv
nbatch = FLAGS.samples // FLAGS.batch
dataset = data.get_dataset(FLAGS.dataset)
train_data = dataset.train.batch(FLAGS.batch)
train_data = train_data.prefetch(32)
train_data = train_data.make_one_shot_iterator().get_next()
with tf.train.MonitoredSession() as sess:
for _ in trange(nbatch, leave=True):
sess.run(train_data)
def main(argv):
del argv # Unused.
dataset = data.get_dataset(FLAGS.dataset)
model = EDSR(os.path.join(FLAGS.train_dir, dataset.name),
lr=FLAGS.lr,
batch=FLAGS.batch,
scale=FLAGS.scale,
downscaler=FLAGS.downscaler,
filters=FLAGS.filters,
repeat=FLAGS.repeat)
model.train(dataset)
def main(argv):
del argv # Unused.
dataset = data.get_dataset(FLAGS.dataset)
decay_start = (FLAGS.total_kimg << 9) // FLAGS.batch
decay_stop = (FLAGS.total_kimg << 10) // FLAGS.batch
model = cGAN(os.path.join(FLAGS.train_dir, dataset.name),
scale=FLAGS.scale,
downscaler=FLAGS.downscaler,
blocks=FLAGS.blocks,
filters=FLAGS.filters,
noise=FLAGS.noise,
decay_start=decay_start,
decay_stop=decay_stop,
lr_decay=FLAGS.lr_decay)
if FLAGS.reset:
model.reset_files()
model.train(dataset)
def main(argv):
del argv # Unused.
dataset = data.get_dataset(FLAGS.dataset)
decay_start = (FLAGS.total_kimg << 9) // FLAGS.batch
decay_stop = (FLAGS.total_kimg << 10) // FLAGS.batch
model = SRGAN(os.path.join(FLAGS.train_dir, dataset.name),
scale=FLAGS.scale,
downscaler=FLAGS.downscaler,
filters=FLAGS.filters,
blocks=FLAGS.blocks,
decay_start=decay_start,
decay_stop=decay_stop,
lr_decay=FLAGS.lr_decay,
adv_weight=FLAGS.adv_weight,
pcp_weight=FLAGS.pcp_weight,
layer_name=FLAGS.layer_name)
if FLAGS.reset:
model.reset_files()
model.train(dataset)