def load_training_data(params, config):
"""Load and return the training data."""
# Load data
if params['is_conditional']:
raise ValueError("Not supported yet.")
else:
labels = None
# Load data
LOGGER.info("Loading training data.")
data = load_data(config['data_source'], config['data_filename'])
LOGGER.info("Training data size: %d", len(data))
# Build dataset
LOGGER.info("Building dataset.")
dataset = get_dataset(data, labels, config['batch_size'],
params['data_shape'], config['use_random_transpose'],
config['n_jobs'])
# Create iterator
if params['is_conditional']:
train_x, train_y = dataset.make_one_shot_iterator().get_next()
else:
train_x, train_y = dataset.make_one_shot_iterator().get_next(), None
return train_x, train_y
def load_training_data(params, config):
"""Load and return the training data."""
# Load data
if params['is_conditional']: # greg: ¿Qué significa este error??
raise ValueError("Not supported yet.")
else:
labels = None
LOGGER.info("Loading training data.")
data = load_data(config['data_source'], config['data_filename'])
LOGGER.info("Training data size: %d", len(data))
# Build dataset
LOGGER.info("Building dataset.")
dataset = get_dataset(data, labels, config['batch_size'],
params['data_shape'], config['use_random_transpose'],
config['n_jobs'])
# Create iterator
if params[
'is_conditional']: # greg: ESTO NUNCA PASA, YA QUE is_conditional arroja Error Not supported yet en el método load_training_data
# greg: ¿Que hace make_one_shot_iterator.get_next()?
train_x, train_y = dataset.make_one_shot_iterator().get_next()
else:
train_x, train_y = dataset.make_one_shot_iterator().get_next(), None
return train_x, train_y
def load_training_data(params, config):
"""Load and return the training data."""
# Load data
if params['is_conditional']:
raise ValueError("Not supported yet.")
else:
labels = None
LOGGER.info("Loading training data.")
data = load_data(config['data_source'], config['data_filename'])
LOGGER.info("Training data size: %d", len(data))
# Build dataset
LOGGER.info("Building dataset.")
dataset = get_dataset(
data, labels, config['batch_size'], params['data_shape'],
config['use_random_transpose'], config['n_jobs'])
# Create iterator
if params['is_conditional']:
train_x, train_y = dataset.make_one_shot_iterator().get_next()
else:
train_x, train_y = dataset.make_one_shot_iterator().get_next(), None
return train_x, train_y