def testLayerProperties(self):
layer = base_layers.Layer(name='my_layer')
self.assertListEqual(layer.variables, [])
self.assertListEqual(layer.trainable_variables, [])
self.assertListEqual(layer.non_trainable_variables, [])
self.assertListEqual(layer.updates, [])
self.assertListEqual(layer.losses, [])
self.assertEqual(layer.built, False)
layer = base_layers.Layer(name='my_layer', trainable=False)
self.assertEqual(layer.trainable, False)
def testLayerProperties(self):
layer = base_layers.Layer(name='my_layer')
self.assertEqual(layer.variables, [])
self.assertEqual(layer.trainable_variables, [])
self.assertEqual(layer.non_trainable_variables, [])
if not context.executing_eagerly():
# updates, losses only supported in GRAPH mode
self.assertEqual(layer.updates, [])
self.assertEqual(layer.losses, [])
self.assertEqual(layer.built, False)
layer = base_layers.Layer(name='my_layer', trainable=False)
self.assertEqual(layer.trainable, False)
def testLayerProperties(self):
layer = base_layers.Layer(name='my_layer')
self.assertListEqual(layer.variables, [])
self.assertListEqual(layer.trainable_variables, [])
self.assertListEqual(layer.non_trainable_variables, [])
if context.in_graph_mode():
# updates, losses only suppported in GRAPH mode
self.assertListEqual(layer.updates, [])
self.assertListEqual(layer.losses, [])
self.assertEqual(layer.built, False)
layer = base_layers.Layer(name='my_layer', trainable=False)
self.assertEqual(layer.trainable, False)
def testAddWeight(self):
with self.test_session():
layer = base_layers.Layer(name='my_layer')
# Test basic variable creation.
variable = layer.add_variable(
'my_var', [2, 2], initializer=init_ops.zeros_initializer())
self.assertEqual(variable.name, 'my_layer/my_var:0')
self.assertListEqual(layer.variables, [variable])
self.assertListEqual(layer.trainable_variables, [variable])
self.assertListEqual(layer.non_trainable_variables, [])
self.assertListEqual(
layer.variables,
ops.get_collection(ops.GraphKeys.TRAINABLE_VARIABLES))
# Test non-trainable variable creation.
# layer.add_variable should work even outside `build` and `call`.
variable_2 = layer.add_variable(
'non_trainable_var', [2, 2],
initializer=init_ops.zeros_initializer(),
trainable=False)
self.assertListEqual(layer.variables, [variable, variable_2])
self.assertListEqual(layer.trainable_variables, [variable])
self.assertListEqual(layer.non_trainable_variables, [variable_2])
self.assertEqual(
len(ops.get_collection(ops.GraphKeys.TRAINABLE_VARIABLES)), 1)
# Test with regularizer.
regularizer = lambda x: math_ops.reduce_sum(x) * 1e-3
variable = layer.add_variable(
'reg_var', [2, 2],
initializer=init_ops.zeros_initializer(),
regularizer=regularizer)
self.assertEqual(len(layer.losses), 1)
def testAddWeight(self):
layer = base_layers.Layer(name='my_layer')
# Test basic variable creation.
variable = layer.add_variable(
'my_var', [2, 2], initializer=init_ops.zeros_initializer())
self.assertEqual(variable.name, 'my_layer/my_var:0')
self.assertEqual(layer.variables, [variable])
self.assertEqual(layer.trainable_variables, [variable])
self.assertEqual(layer.non_trainable_variables, [])
if not context.executing_eagerly():
self.assertEqual(
layer.variables,
ops.get_collection(ops.GraphKeys.TRAINABLE_VARIABLES))
# Test non-trainable variable creation.
# layer.add_variable should work even outside `build` and `call`.
variable_2 = layer.add_variable(
'non_trainable_var', [2, 2],
initializer=init_ops.zeros_initializer(),
trainable=False)
self.assertEqual(layer.variables, [variable, variable_2])
self.assertEqual(layer.trainable_variables, [variable])
self.assertEqual(layer.non_trainable_variables, [variable_2])
if not context.executing_eagerly():
self.assertEqual(
len(ops.get_collection(ops.GraphKeys.TRAINABLE_VARIABLES)), 1)
regularizer = lambda x: math_ops.reduce_sum(x) * 1e-3
_ = layer.add_variable(
'reg_var', [2, 2],
initializer=init_ops.zeros_initializer(),
regularizer=regularizer)
self.assertEqual(len(layer.losses), 1)
added_variable = [False]
# Test that sync `ON_READ` variables are defaulted to be non-trainable.
variable_3 = layer.add_variable(
'sync_on_read_var', [2, 2],
initializer=init_ops.zeros_initializer(),
synchronization=variable_scope.VariableSynchronization.ON_READ,
aggregation=variable_scope.VariableAggregation.SUM)
self.assertEqual(layer.non_trainable_variables, [variable_2, variable_3])
@def_function.function
def function_adds_weight():
if not added_variable[0]:
layer.add_variable(
'reg_var_from_function', [2, 2],
initializer=init_ops.zeros_initializer(),
regularizer=regularizer)
added_variable[0] = True
function_adds_weight()
self.assertEqual(len(layer.losses), 2)
def testKerasStyleAddWeight(self):
keras_layer = keras_base_layer.Layer(name='keras_layer')
with ops.name_scope('foo', skip_on_eager=False):
keras_variable = keras_layer.add_variable(
'my_var', [2, 2], initializer=init_ops.zeros_initializer())
self.assertEqual(keras_variable.name, 'foo/my_var:0')
with ops.name_scope('baz', skip_on_eager=False):
old_style_layer = base_layers.Layer(name='my_layer')
# Test basic variable creation.
variable = old_style_layer.add_variable(
'my_var', [2, 2], initializer=init_ops.zeros_initializer())
self.assertEqual(variable.name, 'my_layer/my_var:0')
with base_layers.keras_style_scope():
layer = base_layers.Layer(name='my_layer')
# Test basic variable creation.
with ops.name_scope('bar', skip_on_eager=False):
variable = layer.add_variable(
'my_var', [2, 2], initializer=init_ops.zeros_initializer())
self.assertEqual(variable.name, 'bar/my_var:0')
def testInvalidTrainableSynchronizationCombination(self):
layer = base_layers.Layer(name='my_layer')
with self.assertRaisesRegexp(
ValueError, 'Synchronization value can be set to '
'VariableSynchronization.ON_READ only for non-trainable variables. '
'You have specified trainable=True and '
'synchronization=VariableSynchronization.ON_READ.'):
_ = layer.add_variable(
'v', [2, 2],
initializer=init_ops.zeros_initializer(),
synchronization=variable_scope.VariableSynchronization.ON_READ,
trainable=True)
def testReusePartitionedVaraiblesAndRegularizers(self):
regularizer = lambda x: math_ops.reduce_sum(x) * 1e-3
partitioner = partitioned_variables.fixed_size_partitioner(3)
for reuse in [False, True]:
with variable_scope.variable_scope(variable_scope.get_variable_scope(),
partitioner=partitioner,
reuse=reuse):
layer = base_layers.Layer(name='my_layer')
_ = layer.add_variable(
'reg_part_var', [4, 4],
initializer=init_ops.zeros_initializer(),
regularizer=regularizer)
self.assertEqual(
len(ops.get_collection(ops.GraphKeys.REGULARIZATION_LOSSES)), 3)
def testInt64Layer(self):
layer = base_layers.Layer(name='my_layer', dtype='int64')
layer.add_variable('my_var', [2, 2])
self.assertEqual(layer.name, 'my_layer')
def testActivityRegularizer(self):
regularizer = math_ops.reduce_sum
layer = base_layers.Layer(activity_regularizer=regularizer)
x = array_ops.placeholder('int32')
layer.apply(x)
self.assertEqual(len(layer.get_losses_for(x)), 1)
from tensorflow.python.layers import base as base_layer print(dir(base_layer.Layer)) print(dir(base_layer.Layer()))
