def testRoundTrip(self):
self.assertEqual(
tfs_flavor.parse_model_path(
tfs_flavor.make_model_path('/foo/bar', 'my-model', 123)),
('/foo/bar', 'my-model', 123))
self.assertEqual(
tfs_flavor.make_model_path(
*tfs_flavor.parse_model_path('/foo/bar/my-model/123')),
'/foo/bar/my-model/123')
def _PrepareModelPath(
self, model_uri: Text,
serving_spec: infra_validator_pb2.ServingSpec) -> Text:
model_path = path_utils.serving_model_path(model_uri)
serving_binary = serving_spec.WhichOneof('serving_binary')
if serving_binary == 'tensorflow_serving':
# TensorFlow Serving requires model to be stored in its own directory
# structure flavor. If current model_path does not conform to the flavor,
# we need to make a copy to the temporary path.
try:
# Check whether current model_path conforms to the tensorflow serving
# model path flavor. (Parsed without exception)
tf_serving_flavor.parse_model_path(
model_path, expected_model_name=serving_spec.model_name)
except ValueError:
# Copy the model to comply with the tensorflow serving model path
# flavor.
temp_model_path = tf_serving_flavor.make_model_path(
model_base_path=self._get_tmp_dir(),
model_name=serving_spec.model_name,
version=int(time.time()))
io_utils.copy_dir(src=model_path, dst=temp_model_path)
return temp_model_path
return model_path
def _PrepareModelPath(
self, model: types.Artifact,
serving_spec: infra_validator_pb2.ServingSpec) -> str:
model_path = path_utils.serving_model_path(
model.uri, path_utils.is_old_model_artifact(model))
serving_binary = serving_spec.WhichOneof('serving_binary')
if serving_binary == _TENSORFLOW_SERVING:
# TensorFlow Serving requires model to be stored in its own directory
# structure flavor. If current model_path does not conform to the flavor,
# we need to make a copy to the temporary path.
try:
# Check whether current model_path conforms to the tensorflow serving
# model path flavor. (Parsed without exception)
tf_serving_flavor.parse_model_path(
model_path, expected_model_name=serving_spec.model_name)
except ValueError:
# Copy the model to comply with the tensorflow serving model path
# flavor.
temp_model_path = tf_serving_flavor.make_model_path(
model_base_path=self._get_tmp_dir(),
model_name=serving_spec.model_name,
version=int(time.time()))
io_utils.copy_dir(src=model_path, dst=temp_model_path)
self._AddCleanup(io_utils.delete_dir,
self._context.get_tmp_path())
return temp_model_path
return model_path
def testMakeModelPath(self):
self.assertEqual(
tfs_flavor.make_model_path(model_base_path='/foo/bar',
model_name='my-model',
version=123), '/foo/bar/my-model/123')
self.assertEqual(
tfs_flavor.make_model_path(
model_base_path='s3://bucket-name/foo/bar',
model_name='my-model',
version=123), 's3://bucket-name/foo/bar/my-model/123')
self.assertEqual(
tfs_flavor.make_model_path(
model_base_path='gs://bucket-name/foo/bar',
model_name='my-model',
version=123), 'gs://bucket-name/foo/bar/my-model/123')
def MakeDockerRunParams(self, host_port: int, model_path: Text,
needs_mount: bool) -> Dict[Text, Any]:
"""Make parameters for docker `client.containers.run`.
Args:
host_port: Available port in the host to bind with container port.
model_path: A path to the model.
needs_mount: If True, model_path will be mounted to the container.
Returns:
A dictionary of docker run parameters.
"""
result = dict(self._BASE_DOCKER_RUN_PARAMS,
image=self._image,
ports={'{}/tcp'.format(self.container_port): host_port})
if needs_mount:
# model_path should be a local directory. In order to make TF Serving see
# the host model path, we need to mount model path volume to the
# container.
assert os.path.isdir(model_path), '{} does not exist'.format(
model_path)
container_model_path = tf_serving_flavor.make_model_path(
model_base_path=self._DEFAULT_MODEL_BASE_PATH,
model_name=self._model_name,
version=1)
result.update(environment=self.MakeEnvVars(),
mounts=[
docker_types.Mount(type='bind',
target=container_model_path,
source=model_path,
read_only=True)
])
else:
# model_path is presumably a remote URI. TF Serving is able to pickup
# model in remote directly using gfile, so all we need to do is setting
# environment variables correctly.
result.update(environment=self.MakeEnvVars(model_path=model_path))
return result