def local_predict(input_data, model_dir):
"""Runs prediction locally.
Args:
input_data: list of input files to run prediction on.
model_dir: path to Tensorflow model folder.
"""
session, _ = session_bundle.load_session_bundle_from_path(model_dir)
# get the mappings between aliases and tensor names
# for both inputs and outputs
input_alias_map = json.loads(session.graph.get_collection('inputs')[0])
output_alias_map = json.loads(session.graph.get_collection('outputs')[0])
aliases, tensor_names = zip(*output_alias_map.items())
metadata_path = os.path.join(model_dir, 'metadata.yaml')
transformer = features.FeatureProducer(metadata_path)
for input_file in input_data:
with open(input_file) as f:
feed_dict = collections.defaultdict(list)
for line in f:
preprocessed = transformer.preprocess(line)
feed_dict[input_alias_map.values()[0]].append(
preprocessed.SerializeToString())
result = session.run(fetches=tensor_names, feed_dict=feed_dict)
for row in zip(*result):
print json.dumps({
name: (value.tolist()
if getattr(value, 'tolist', None) else value)
for name, value in zip(aliases, row)
})
def from_client(cls, client, model_path, skip_preprocessing=False):
preprocess_fn = None
if not skip_preprocessing:
metadata_path = _get_metadata_path(model_path)
if metadata_path:
feature_producer = features.FeatureProducer(metadata_path)
if feature_producer:
preprocess_fn = feature_producer.preprocess
return cls(client, preprocess_fn)