#
# Copyright 2018 Analytics Zoo Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
from zoo.tfpark import KerasModel
# TODO: add word embedding file support
[docs]class TextKerasModel(KerasModel):
"""
The base class for text models in tfpark.
"""
def __init__(self, labor, optimizer=None, **kwargs):
self.labor = labor
self.labor.build(**kwargs)
model = self.labor.model
# Recompile the model if user uses a different optimizer other than the default one.
if optimizer:
model.compile(loss=model.loss, optimizer=optimizer, metrics=model.metrics)
super(TextKerasModel, self).__init__(model)
# Remark: nlp-architect CRF layer has error when directly loaded by tf.keras.models.load_model.
# Thus we keep the nlp-architect class as labor and uses its save/load,
# which only saves the weights with model parameters
# and reconstruct the model using the exact parameters and setting weights when loading.
[docs] def save_model(self, path):
"""
Save the model to a single HDF5 file.
:param path: String. The path to save the model.
"""
self.labor.save(path)
@staticmethod
def _load_model(labor, path):
labor.load(path)
model = KerasModel(labor.model)
model.labor = labor
return model