def build(self, hp, inputs=None):
input_node = nest.flatten(inputs)[0]
output_node = input_node
block_type = self.block_type or hp.Choice(
'block_type', ['resnet', 'xception', 'vanilla'], default='resnet')
normalize = self.normalize
if normalize is None:
normalize = hp.Choice('normalize', [True, False], default=True)
augment = self.augment
if augment is None:
augment = hp.Choice('augment', [True, False], default=True)
if normalize:
output_node = preprocessor.Normalization()(output_node)
if augment:
output_node = preprocessor.ImageAugmentation(
seed=self.seed)(output_node)
sub_block_name = self.name + '_' + block_type
if block_type == 'resnet':
output_node = block.ResNetBlock(name=sub_block_name)(output_node)
elif block_type == 'xception':
output_node = block.XceptionBlock(name=sub_block_name)(output_node)
elif block_type == 'vanilla':
output_node = block.ConvBlock(name=sub_block_name)(output_node)
return output_node
def test_augment():
dataset = common.generate_data(dtype='dataset')
new_dataset = run_preprocessor(preprocessor.ImageAugmentation(seed=common.SEED),
dataset,
common.generate_data(dtype='dataset'),
tf.float32)
assert isinstance(new_dataset, tf.data.Dataset)
def test_augment():
raw_images = tf.random.normal([1000, 32, 32, 3], mean=-1, stddev=4)
augment = preprocessor.ImageAugmentation(seed=5)
dataset = tf.data.Dataset.from_tensor_slices(raw_images)
hp = kerastuner.HyperParameters()
augment.set_hp(hp)
augment.set_config(augment.get_config())
for a in dataset:
augment.transform(a, True)
def map_func(x):
return tf.py_function(functools.partial(augment.transform, fit=True),
inp=[x],
Tout=(tf.float32, ))
new_dataset = dataset.map(map_func)
for _ in new_dataset:
pass
assert isinstance(new_dataset, tf.data.Dataset)