def build(self, hp, inputs=None):
inputs = nest.flatten(inputs)
utils.validate_num_inputs(inputs, 1)
input_node = inputs[0]
output_node = input_node
kernel_size = self.kernel_size or hp.Choice("kernel_size", [3, 5, 7],
default=3)
num_blocks = self.num_blocks or hp.Choice("num_blocks", [1, 2, 3],
default=2)
num_layers = self.num_layers or hp.Choice("num_layers", [1, 2],
default=2)
separable = self.separable
if separable is None:
separable = hp.Boolean("separable", default=False)
if separable:
conv = layer_utils.get_sep_conv(input_node.shape)
else:
conv = layer_utils.get_conv(input_node.shape)
max_pooling = self.max_pooling
if max_pooling is None:
max_pooling = hp.Boolean("max_pooling", default=True)
pool = layer_utils.get_max_pooling(input_node.shape)
if self.dropout is not None:
dropout = self.dropout
else:
dropout = hp.Choice("dropout", [0.0, 0.25, 0.5], default=0)
for i in range(num_blocks):
for j in range(num_layers):
output_node = conv(
hp.Choice(
"filters_{i}_{j}".format(i=i, j=j),
[16, 32, 64, 128, 256, 512],
default=32,
),
kernel_size,
padding=self._get_padding(kernel_size, output_node),
activation="relu",
)(output_node)
if max_pooling:
output_node = pool(
kernel_size - 1,
padding=self._get_padding(kernel_size - 1, output_node),
)(output_node)
if dropout > 0:
output_node = layers.Dropout(dropout)(output_node)
return output_node
def build(self, hp, inputs=None):
inputs = nest.flatten(inputs)
utils.validate_num_inputs(inputs, 1)
input_node = inputs[0]
output_node = input_node
kernel_size = utils.add_to_hp(self.kernel_size, hp)
separable = self.separable
if separable is None:
separable = hp.Boolean("separable", default=False)
if separable:
conv = layer_utils.get_sep_conv(input_node.shape)
else:
conv = layer_utils.get_conv(input_node.shape)
max_pooling = self.max_pooling
if max_pooling is None:
max_pooling = hp.Boolean("max_pooling", default=True)
pool = layer_utils.get_max_pooling(input_node.shape)
for i in range(utils.add_to_hp(self.num_blocks, hp)):
for j in range(utils.add_to_hp(self.num_layers, hp)):
output_node = conv(
utils.add_to_hp(
self.filters, hp, "filters_{i}_{j}".format(i=i, j=j)
),
kernel_size,
padding=self._get_padding(kernel_size, output_node),
activation="relu",
)(output_node)
if max_pooling:
output_node = pool(
kernel_size - 1,
padding=self._get_padding(kernel_size - 1, output_node),
)(output_node)
if utils.add_to_hp(self.dropout, hp) > 0:
output_node = layers.Dropout(utils.add_to_hp(self.dropout, hp))(
output_node
)
return output_node
