def load_model(self, file_path):
""" This function load already saved model and sets cuda parameters.
:param file_path: File path of a model to loaded.
:type file_path: string
"""
model_dict = torch.load(file_path)
model = RNNClassifier(
model_dict["input_size"],
model_dict["hidden_size"],
model_dict["output_size"],
model_dict["n_layers"],
)
model.load_state_dict(model_dict["state_dict"])
super()._load_model(model)
def load_checkpoint(self, file_path):
""" This function load already saved model checkpoint and sets cuda parameters.
:param file_path: File path of a model checkpoint to be loaded.
:type file_path: string
"""
checkpoint = torch.load(file_path)
model = RNNClassifier(
checkpoint["input_size"],
checkpoint["hidden_size"],
checkpoint["output_size"],
checkpoint["n_layers"],
)
model.load_state_dict(checkpoint["state_dict"])
super().leverage_model(model)
def init_model(self, char_vocab=128, hidden_size=100, n_domain_type=2, n_layers=3):
"""This function instantiates RNNClassifier model to train. And also optimizes to scale it and keep running on parallelism.
:param char_vocab: Vocabulary size is set to 128 ASCII characters.
:type char_vocab: int
:param hidden_size: Hidden size of the network.
:type hidden_size: int
:param n_domain_type: Number of domain types.
:type n_domain_type: int
:param n_layers: Number of network layers.
:type n_layers: int
"""
if self.model is None:
model = RNNClassifier(char_vocab, hidden_size, n_domain_type, n_layers)
self.leverage_model(model)
def init_model(self, char_vocab=128, hidden_size=100, n_domain_type=2, n_layers=3):
if self.model is None:
model = RNNClassifier(char_vocab, hidden_size, n_domain_type, n_layers)
self.leverage_model(model)