def persist(self, file_name: Text,
model_dir: Text) -> Optional[Dict[Text, Any]]:
"""Persist this model into the passed directory.
Returns the metadata necessary to load the model again.
"""
file_name = file_name + ".pkl"
if self.vectorizers:
# vectorizer instance was not None, some models could have been trained
attribute_vocabularies = self._collect_vectorizer_vocabularies()
if self._is_any_model_trained(attribute_vocabularies):
# Definitely need to persist some vocabularies
featurizer_file = os.path.join(model_dir, file_name)
if self.use_shared_vocab:
# Only persist vocabulary from one attribute. Can be loaded and
# distributed to all attributes.
vocab = attribute_vocabularies[TEXT]
else:
vocab = attribute_vocabularies
io_utils.json_pickle(featurizer_file, vocab)
return {"file": file_name}
def persist(self) -> None:
"""Persist this model into the passed directory.
Returns the metadata necessary to load the model again.
"""
if not self.vectorizers:
return
with self._model_storage.write_to(self._resource) as model_dir:
# vectorizer instance was not None, some models could have been trained
attribute_vocabularies = self._collect_vectorizer_vocabularies()
if self._is_any_model_trained(attribute_vocabularies):
# Definitely need to persist some vocabularies
featurizer_file = model_dir / "vocabularies.pkl"
if self.use_shared_vocab:
# Only persist vocabulary from one attribute. Can be loaded and
# distributed to all attributes.
vocab = attribute_vocabularies[TEXT]
else:
vocab = attribute_vocabularies
io_utils.json_pickle(featurizer_file, vocab)
# Dump OOV words separately as they might have been modified during
# training
rasa.shared.utils.io.dump_obj_as_json_to_file(
model_dir / "oov_words.json", self.OOV_words)
def persist(self, file_name: Text, model_dir: Text) -> Dict[Text, Any]:
"""Persist this model into the passed directory.
Return the metadata necessary to load the model again.
"""
if self.model is None:
return {"file": None}
model_dir = Path(model_dir)
tf_model_file = model_dir / f"{file_name}.tf_model"
io_utils.create_directory_for_file(tf_model_file)
self.model.save(str(tf_model_file))
io_utils.pickle_dump(model_dir / f"{file_name}.data_example.pkl",
self.data_example)
io_utils.pickle_dump(model_dir / f"{file_name}.label_data.pkl",
self._label_data)
io_utils.json_pickle(
model_dir / f"{file_name}.index_label_id_mapping.pkl",
self.index_label_id_mapping,
)
io_utils.json_pickle(
model_dir / f"{file_name}.index_tag_id_mapping.pkl",
self.index_tag_id_mapping,
)
return {"file": file_name}
def persist(self, path: Union[Text, Path]) -> None:
"""Persists the policy to a storage."""
if self.model is None:
logger.debug(
"Method `persist(...)` was called without a trained model present. "
"Nothing to persist then!")
return
model_path = Path(path)
tf_model_file = model_path / f"{SAVE_MODEL_FILE_NAME}.tf_model"
rasa.shared.utils.io.create_directory_for_file(tf_model_file)
self.featurizer.persist(path)
if self.model.checkpoint_model:
self.model.copy_best(str(tf_model_file))
else:
self.model.save(str(tf_model_file))
io_utils.json_pickle(
model_path / f"{SAVE_MODEL_FILE_NAME}.priority.pkl", self.priority)
io_utils.pickle_dump(model_path / f"{SAVE_MODEL_FILE_NAME}.meta.pkl",
self.config)
io_utils.pickle_dump(
model_path / f"{SAVE_MODEL_FILE_NAME}.data_example.pkl",
self.data_example)
io_utils.pickle_dump(
model_path / f"{SAVE_MODEL_FILE_NAME}.zero_state_features.pkl",
self.zero_state_features,
)
io_utils.pickle_dump(
model_path / f"{SAVE_MODEL_FILE_NAME}.label_data.pkl",
dict(self._label_data.data),
)
def persist_model_utilities(self, model_path: Path) -> None:
"""Persists model's utility attributes like model weights, etc.
Args:
model_path: Path where model is to be persisted
"""
model_filename = self._metadata_filename()
io_utils.json_pickle(model_path / f"{model_filename}.priority.pkl",
self.priority)
io_utils.pickle_dump(model_path / f"{model_filename}.meta.pkl",
self.config)
io_utils.pickle_dump(
model_path / f"{model_filename}.data_example.pkl",
self.data_example,
)
io_utils.pickle_dump(
model_path / f"{model_filename}.fake_features.pkl",
self.fake_features,
)
io_utils.pickle_dump(
model_path / f"{model_filename}.label_data.pkl",
dict(self._label_data.data),
)
entity_tag_specs = ([
tag_spec._asdict() for tag_spec in self._entity_tag_specs
] if self._entity_tag_specs else [])
rasa.shared.utils.io.dump_obj_as_json_to_file(
model_path / f"{model_filename}.entity_tag_specs.json",
entity_tag_specs,
)
def persist(self, file_name: Text, model_dir: Text) -> Optional[Dict[Text, Any]]:
"""Persist this model into the passed directory."""
class_encoder_file_name = file_name + "class_encoder.pkl"
if self.class_encoder:
io_utils.json_pickle(os.path.join(model_dir, class_encoder_file_name), self.class_encoder)
return {"class_encoder": class_encoder_file_name}
def persist(self, file_name: Text, model_dir: Text) -> Optional[Dict[Text, Any]]:
"""Persist this model into the passed directory.
Return the metadata necessary to load the model again."""
feature_to_idx_file = Path(model_dir) / f"{file_name}.feature_to_idx_dict.pkl"
io_utils.json_pickle(feature_to_idx_file, self.feature_to_idx_dict)
return {"file": file_name}
def persist(self) -> None:
"""Persist this model into the passed directory."""
with self._model_storage.write_to(self._resource) as model_dir:
file_name = self.__class__.__name__
classifier_file_name = model_dir / f"{file_name}_classifier.pkl"
encoder_file_name = model_dir / f"{file_name}_encoder.pkl"
if self.clf and self.le:
io_utils.json_pickle(encoder_file_name, self.le.classes_)
io_utils.json_pickle(classifier_file_name, self.clf.best_estimator_)
def persist(self, file_name: Text, model_dir: Text) -> Optional[Dict[Text, Any]]:
"""Persist this model into the passed directory."""
classifier_file_name = file_name + "_classifier.pkl"
encoder_file_name = file_name + "_encoder.pkl"
if self.clf and self.le:
io_utils.json_pickle(
os.path.join(model_dir, encoder_file_name), self.le.classes_
)
io_utils.json_pickle(
os.path.join(model_dir, classifier_file_name), self.clf.best_estimator_
)
return {"classifier": classifier_file_name, "encoder": encoder_file_name}
def persist(self, file_name: Text, model_dir: Text) -> Dict[Text, Any]:
"""Persist this model into the passed directory.
Return the metadata necessary to load the model again.
"""
if self.model is None:
return {"file": None}
super().persist(file_name, model_dir)
model_dir = Path(model_dir)
io_utils.json_pickle(
model_dir / f"{file_name}.retrieval_intent_mapping.pkl",
self.retrieval_intent_mapping,
)
return {"file": file_name}
def persist(self, path: Text) -> None:
"""Persists the policy to a storage."""
if self.model is None:
logger.debug("Method `persist(...)` was called "
"without a trained model present. "
"Nothing to persist then!")
return
model_path = Path(path)
tf_model_file = model_path / f"{SAVE_MODEL_FILE_NAME}.tf_model"
io_utils.create_directory_for_file(tf_model_file)
self.featurizer.persist(path)
self.model.save(str(tf_model_file))
io_utils.json_pickle(
model_path / f"{SAVE_MODEL_FILE_NAME}.priority.pkl", self.priority)
io_utils.pickle_dump(model_path / f"{SAVE_MODEL_FILE_NAME}.meta.pkl",
self.config)
io_utils.json_pickle(
model_path / f"{SAVE_MODEL_FILE_NAME}.data_example.pkl",
self.data_example)
io_utils.json_pickle(
model_path / f"{SAVE_MODEL_FILE_NAME}.label_data.pkl",
self._label_data)
def persist(self, file_name, model_dir):
"""Persist this model into the passed directory."""
classifier_file = os.path.join(model_dir, SENTIMENT_MODEL_FILE_NAME)
json_pickle(classifier_file, self)
return {"classifier_file": SENTIMENT_MODEL_FILE_NAME}