def __init__(self, preprocessors=None, **kwargs):
super().__init__(**kwargs)
self.preprocessors = nest.flatten(preprocessors)
self._finished = False
# Save or load the HyperModel.
utils.save_json(os.path.join(self.project_dir, 'graph'),
graph_module.serialize(self.hypermodel.hypermodel))
def test_text_block():
block = wrapper.TextBlock()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.TextInput(shape=(1, )).build())
assert utils.name_in_hps('vectorizer', hp)
def test_spatial_reduction():
input_shape = (32, 32, 3)
block = reduction.SpatialReduction()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('reduction_type', hp)
def test_time_series_input_node():
# TODO. Change test once TimeSeriesBlock is added.
node = ak.TimeseriesInput(shape=(32, ), lookback=2)
output = node.build()
assert isinstance(output, tf.Tensor)
node = graph_module.deserialize(graph_module.serialize(node))
output = node.build()
assert isinstance(output, tf.Tensor)
def block_basic_exam(block, inputs, hp_names):
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
outputs = block.build(hp, inputs)
for hp_name in hp_names:
assert name_in_hps(hp_name, hp)
return outputs
def test_image_block():
block = wrapper.ImageBlock(normalize=None, augment=None)
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.ImageInput(shape=(32, 32, 3)).build())
assert utils.name_in_hps('block_type', hp)
assert utils.name_in_hps('normalize', hp)
assert utils.name_in_hps('augment', hp)
def test_imag_augmentation():
input_shape = (32, 32, 3)
block = preprocessing.ImageAugmentation()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('vertical_flip', hp)
assert utils.name_in_hps('horizontal_flip', hp)
def test_segmentation():
y = np.array(['a', 'a', 'c', 'b'])
head = head_module.SegmentationHead(name='a')
adapter = head.get_adapter()
adapter.fit_transform(y)
head.config_from_adapter(adapter)
input_shape = (64, 64, 21)
hp = kerastuner.HyperParameters()
head = graph_module.deserialize(graph_module.serialize(head))
head.build(hp, ak.Input(shape=input_shape).build())
def test_dense_block():
input_shape = (32, )
block = basic.DenseBlock()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('num_layers', hp)
assert utils.name_in_hps('use_batchnorm', hp)
def test_embedding_block():
input_shape = (32, )
block = basic.Embedding()
block.max_features = 100
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('pretraining', hp)
assert utils.name_in_hps('embedding_dim', hp)
def test_rnn_block():
input_shape = (32, 10)
block = basic.RNNBlock()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('bidirectional', hp)
assert utils.name_in_hps('layer_type', hp)
assert utils.name_in_hps('num_layers', hp)
def test_conv_block():
input_shape = (32, 32, 3)
block = basic.ConvBlock()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('kernel_size', hp)
assert utils.name_in_hps('num_blocks', hp)
assert utils.name_in_hps('separable', hp)
def test_resnet_block(init, build):
input_shape = (32, 32, 3)
block = basic.ResNetBlock()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('version', hp)
assert utils.name_in_hps('pooling', hp)
assert init.called
assert build.called
def test_merge():
input_shape_1 = (32, )
input_shape_2 = (4, 8)
block = reduction.Merge()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, [
ak.Input(shape=input_shape_1).build(),
ak.Input(shape=input_shape_2).build()
])
assert utils.name_in_hps('merge_type', hp)
def test_xception_block(init, build):
input_shape = (32, 32, 3)
block = basic.XceptionBlock()
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.build(hp, ak.Input(shape=input_shape).build())
assert utils.name_in_hps('activation', hp)
assert utils.name_in_hps('initial_strides', hp)
assert utils.name_in_hps('num_residual_blocks', hp)
assert utils.name_in_hps('pooling', hp)
assert init.called
assert build.called
def test_structured_data_block():
block = wrapper.StructuredDataBlock()
block.column_names = ['0', '1']
block.column_types = {
'0': adapters.CATEGORICAL,
'1': adapters.CATEGORICAL,
}
hp = kerastuner.HyperParameters()
block = graph_module.deserialize(graph_module.serialize(block))
block.column_names = ['0', '1']
block.column_types = {
'0': adapters.CATEGORICAL,
'1': adapters.CATEGORICAL,
}
output = block.build(hp, ak.StructuredDataInput(shape=(2, )).build())
assert isinstance(output, tf.Tensor)
def test_timeseries_block():
block = wrapper.TimeseriesBlock()
hp = kerastuner.HyperParameters()
block.column_names = ['0', '1']
block.column_types = {
'0': adapters.NUMERICAL,
'1': adapters.NUMERICAL,
}
block = graph_module.deserialize(graph_module.serialize(block))
block.column_names = ['0', '1']
block.column_types = {
'0': adapters.NUMERICAL,
'1': adapters.NUMERICAL,
}
output = block.build(hp,
ak.TimeseriesInput(shape=(32, ), lookback=2).build())
assert isinstance(output, tf.Tensor)
def __init__(self, **kwargs):
super().__init__(**kwargs)
self._finished = False
# Save or load the HyperModel.
utils.save_json(os.path.join(self.directory, 'graph'),
graph_module.serialize(self.hypermodel.hypermodel))
