p_transform=p_transform,
data_prep_fun=data_prep_function_valid,
label_prep_fun=label_prep_function,
rng=rng,
full_batch=False,
random=False,
infinite=False)
import tta
tta = tta.LosslessTTA(p_augmentation)
tta_test_data_iterator = data_iterators.TTADataGenerator(
dataset='test-jpg',
tta=tta,
duplicate_label=False,
img_ids=test_ids,
p_transform=p_transform,
data_prep_fun=data_prep_function_valid,
label_prep_fun=label_prep_function,
rng=rng,
full_batch=False,
random=False,
infinite=False)
tta_test2_data_iterator = data_iterators.TTADataGenerator(
dataset='test2-jpg',
tta=tta,
duplicate_label=False,
img_ids=test2_ids,
p_transform=p_transform,
data_prep_fun=data_prep_function_valid,
label_prep_fun=label_prep_function,
rng=rng,
img_ids=valid_ids,
p_transform=p_transform,
data_prep_fun=data_prep_function_valid,
label_prep_fun=label_prep_function,
rng=rng,
full_batch=False,
random=False,
infinite=False)
tta = tta.LosslessTTA(p_augmentation)
tta_valid_data_iterator = data_iterators.TTADataGenerator(
dataset='train-jpg',
tta=tta,
img_ids=valid_ids,
p_transform=p_transform,
data_prep_fun=data_prep_function_valid,
label_prep_fun=label_prep_function,
rng=rng,
full_batch=False,
random=False,
infinite=False)
test_data_iterator = data_iterators.DataGenerator(
dataset='test-jpg',
batch_size=chunk_size,
img_ids=test_ids,
p_transform=p_transform,
data_prep_fun=data_prep_function_valid,
label_prep_fun=label_prep_function,
rng=rng,
full_batch=False,
