def _get_dense_tensor(self, inputs, weight_collections=None, trainable=None):
"""Returns a `Tensor`."""
del weight_collections
input_tensor = inputs.get(self)
text_module = module.Module(
self.module_spec, trainable=self.trainable and trainable)
return self._get_dense_tensor_for_input_tensor(input_tensor, text_module)
def testGetExpectedImageSizeFromImageModuleInfo(self):
with tf.Graph().as_default():
spec = native_module.create_module_spec(image_module_fn_with_info)
self.assertAllEqual(image_util.get_expected_image_size(spec),
[2, 4])
m = module.Module(spec)
self.assertAllEqual(image_util.get_expected_image_size(m), [2, 4])
def testModuleInNestedScope(self):
with tf.Graph().as_default():
with tf.compat.v1.variable_scope("foo"):
m = module.Module(_ModuleSpec())
result = m([1, 2])
with tf.compat.v1.Session() as session:
self.assertAllEqual(session.run(result), [2, 4])
def testModuleInterfaceGettersExplicitSignatureAndTags(self):
"""Tests that tags from Module(...) apply to module.get_*()."""
m = module.Module(_ModuleSpec(), tags={"special"})
self.assertItemsEqual(m.get_signature_names(), ["default", "extra"])
self.assertItemsEqual(
m.get_input_info_dict(signature="extra").keys(), ["x", "y"])
self.assertItemsEqual(
m.get_output_info_dict(signature="extra").keys(), ["z", "default"])
def _get_dense_tensor(self,
inputs,
weight_collections=None,
trainable=None):
del weight_collections, trainable # Unused.
images = inputs.get(self)
image_module = module.Module(self.module_spec)
return self._get_dense_tensor_for_images(images, image_module)
def create_state(self, state_manager):
"""Imports the module along with all variables."""
# Note: state_manager._trainable is not public but is the pattern used
# to propagate the "trainable" state that used to be received via
# self._get_dense_tensor.
trainable = self.trainable and state_manager._trainable # pylint: disable=protected-access
m = module.Module(self.module_spec, trainable=trainable)
state_manager.add_resource(self, _MODULE_RESOURCE_STRING, m)
def _get_dense_tensor(self,
inputs,
weight_collections=None,
trainable=None):
"""Returns a `Tensor` to represent this feature in the input_layer()."""
del weight_collections, trainable # Unused.
m = module.Module(self.module_spec, trainable=False)
images = inputs.get(self)
return m({"images": images})
def _get_dense_tensor(self,
inputs,
weight_collections=None,
trainable=None):
"""Returns a `Tensor`."""
del weight_collections
text_batch = tf.reshape(inputs.get(self), shape=[-1])
m = module.Module(self.module_spec,
trainable=self.trainable and trainable)
return m(text_batch)
def _get_dense_tensor_for_inputs(self, text_batch, trainable):
m = module.Module(self.module_spec, trainable=self.trainable and trainable)
if self.default_value is not None:
text_batch = tf.sparse.fill_empty_rows(text_batch, self.default_value)[0]
embedded_tokens = m(text_batch.values)
embedding_ids = tf.SparseTensor(
indices=text_batch.indices,
values=tf.range(tf.shape(text_batch.indices)[0], dtype=tf.int32),
dense_shape=text_batch.dense_shape)
return tf.nn.embedding_lookup_sparse(
params=embedded_tokens,
sp_ids=embedding_ids,
sp_weights=None,
combiner=self.combiner)
def testModuleDictInput(self):
with tf.Graph().as_default():
m = module.Module(_ModuleSpec())
result = m({"x": [1, 2]})
with tf.compat.v1.Session() as session:
self.assertAllEqual(session.run(result), [2, 4])
def create_state(self, state_manager):
"""Imports the module along with all variables."""
# Module is not trainable by default.
m = module.Module(self.module_spec)
state_manager.add_resource(self, _MODULE_RESOURCE_STRING, m)
def testModuleInterfaceGettersDefaultSignatureAndTags(self):
with tf.Graph().as_default():
m = module.Module(_ModuleSpec())
self.assertItemsEqual(m.get_signature_names(), ["default"])
self.assertItemsEqual(m.get_input_info_dict().keys(), ["x"])
self.assertItemsEqual(m.get_output_info_dict().keys(), ["default"])
def testModuleSingleInput(self):
m = module.Module(_ModuleSpec())
result = m([1, 2])
with tf.Session() as session:
self.assertAllEqual(session.run(result), [2, 4])
def testModuleDictInput(self):
m = module.Module(_ModuleSpec())
result = m({"x": [1, 2]})
with tf.Session() as session:
self.assertAllEqual(session.run(result), [2, 4])
def _get_dense_tensor_for_input_tensor(self, input_tensor, trainable): text_batch = tf.reshape(input_tensor, shape=[-1]) m = module.Module(self.module_spec, trainable=self.trainable and trainable) return m(text_batch)
def testGetExpectedImageSizeFromShape(self):
spec = native_module.create_module_spec(image_module_fn)
self.assertAllEqual(image_util.get_expected_image_size(spec), [2, 4])
m = module.Module(spec)
self.assertAllEqual(image_util.get_expected_image_size(m), [2, 4])
def testModuleDictOutput(self):
m = module.Module(_ModuleSpec())
result = m([1, 2], as_dict=True)
self.assertTrue(isinstance(result, dict))
self.assertAllEqual(list(result.keys()), ["default"])
def _get_dense_tensor_for_images(self, images):
m = module.Module(self.module_spec, trainable=False)
return m({"images": images})
def testModuleDictOutput(self):
with tf.Graph().as_default():
m = module.Module(_ModuleSpec())
result = m([1, 2], as_dict=True)
self.assertIsInstance(result, dict)
self.assertAllEqual(list(result.keys()), ["default"])
def testGetNumImageChannels(self):
with tf.Graph().as_default():
spec = native_module.create_module_spec(image_module_fn)
self.assertEqual(image_util.get_num_image_channels(spec), 3)
m = module.Module(spec)
self.assertEqual(image_util.get_num_image_channels(m), 3)
def _hub_module(self):
if not hasattr(self, '_module'):
self._module = module.Module(self.module_spec,
trainable=self.trainable)
return self._module
