def test_serving_input_receiver_receiver_tensors_invalid(self):
features = {
"feature0": constant_op.constant([0]),
u"feature1": constant_op.constant([1]),
"feature2": sparse_tensor.SparseTensor(
indices=[[0, 0]], values=[1], dense_shape=[1, 1]),
}
with self.assertRaisesRegexp(
ValueError, "receiver_tensors must be defined"):
export.ServingInputReceiver(
features=features,
receiver_tensors=None)
with self.assertRaisesRegexp(
ValueError, "receiver_tensors keys must be strings"):
export.ServingInputReceiver(
features=features,
receiver_tensors={
1: array_ops.placeholder(dtypes.string, name="example0")})
with self.assertRaisesRegexp(
ValueError, "receiver_tensor example1 must be a Tensor"):
export.ServingInputReceiver(
features=features,
receiver_tensors={"example1": [1]})
def test_multi_feature_single_receiver(self):
features = {
"foo": constant_op.constant(5),
"bar": constant_op.constant(6)
}
receiver_tensor = array_ops.placeholder(dtypes.string)
_ = export.ServingInputReceiver(features, receiver_tensor)
def serving_input_receiver_with_asset_fn():
features, receiver_tensor = serving_input_receiver_fn()
filename = ops.convert_to_tensor(vocab_file_name,
dtypes.string,
name='asset_filepath')
ops.add_to_collection(ops.GraphKeys.ASSET_FILEPATHS, filename)
features['bogus_filename'] = filename
return export.ServingInputReceiver(features, receiver_tensor)
def test_multi_feature_multi_receiver(self):
features = {
"foo": constant_op.constant(5),
"bar": constant_op.constant(6)
}
receiver_tensors = {
"baz": array_ops.placeholder(dtypes.int64),
"qux": array_ops.placeholder(dtypes.float32)
}
_ = export.ServingInputReceiver(features, receiver_tensors)
def test_single_feature_single_receiver(self):
feature = constant_op.constant(5)
receiver_tensor = array_ops.placeholder(dtypes.string)
input_receiver = export.ServingInputReceiver(feature, receiver_tensor)
# single feature is automatically named
feature_key, = input_receiver.features.keys()
self.assertEqual("feature", feature_key)
# single receiver is automatically named
receiver_key, = input_receiver.receiver_tensors.keys()
self.assertEqual("input", receiver_key)
def test_serving_input_receiver_features_invalid(self):
receiver_tensors = {
"example0": array_ops.placeholder(dtypes.string, name="example0"),
u"example1": array_ops.placeholder(dtypes.string, name="example1"),
}
with self.assertRaisesRegexp(ValueError, "features must be defined"):
export.ServingInputReceiver(
features=None,
receiver_tensors=receiver_tensors)
with self.assertRaisesRegexp(ValueError, "feature keys must be strings"):
export.ServingInputReceiver(
features={1: constant_op.constant([1])},
receiver_tensors=receiver_tensors)
with self.assertRaisesRegexp(
ValueError, "feature feature1 must be a Tensor or SparseTensor"):
export.ServingInputReceiver(
features={"feature1": [1]},
receiver_tensors=receiver_tensors)
def test_serving_input_receiver_constructor(self):
"""Tests that no errors are raised when input is expected."""
features = {
"feature0": constant_op.constant([0]),
u"feature1": constant_op.constant([1]),
"feature2": sparse_tensor.SparseTensor(
indices=[[0, 0]], values=[1], dense_shape=[1, 1]),
}
receiver_tensors = {
"example0": array_ops.placeholder(dtypes.string, name="example0"),
u"example1": array_ops.placeholder(dtypes.string, name="example1"),
}
export.ServingInputReceiver(features, receiver_tensors)
def serving_input_receiver_fn():
"""An input function on serialized SequenceExample protos."""
serialized_sequence_example = array_ops.placeholder(
dtype=dtypes.string,
shape=[default_batch_size],
name="input_sequence_example_tensor")
receiver_tensors = {"sequence_example": serialized_sequence_example}
features = parse_from_sequence_example(
serialized_sequence_example,
input_size,
context_feature_spec=context_feature_spec,
example_feature_spec=example_feature_spec)
return export_lib.ServingInputReceiver(features, receiver_tensors)
def test_receiver_wrong_type(self):
feature = constant_op.constant(5)
receiver_tensor = "not a tensor"
with self.assertRaises(ValueError):
_ = export.ServingInputReceiver(feature, receiver_tensor)
def test_feature_wrong_type(self):
feature = "not a tensor"
receiver_tensor = array_ops.placeholder(dtypes.string)
with self.assertRaises(ValueError):
_ = export.ServingInputReceiver(feature, receiver_tensor)
def test_feature_labeled_tensor(self):
feature = LabeledTensorMock()
receiver_tensor = array_ops.placeholder(dtypes.string)
_ = export.ServingInputReceiver(feature, receiver_tensor)
def serving_input_receiver_fn():
return export.ServingInputReceiver(
{'test-features': constant_op.constant([[1], [1]])},
array_ops.placeholder(dtype=dtypes.string))
def serving_input_receiver_fn():
return export.ServingInputReceiver(
given_features, array_ops.placeholder(dtype=dtypes.string))
