def load(path, fields_tuples, current_gpu_id):
options = opts.load(path)
# set gpu device to the current device
options.gpu_id = current_gpu_id
# hack: set dummy loss_weights (the correct values are going to be loaded)
tags_field = dict(fields_tuples)['tags']
loss_weights = None
if options.loss_weights == 'balanced':
loss_weights = [0] * (len(tags_field.vocab) - 1)
model = build(options, fields_tuples, loss_weights)
load_state(path, model)
return model
def load(self, dir_path):
# load options from the json file
self.options = opts.load(dir_path)
# load vocabularies for each field
fields.load_vocabs(dir_path, self.fields_tuples)
# set the current gpu
self.options.gpu_id = self.gpu_id
# load model, optimizer and scheduler
self.model = models.load(dir_path, self.fields_tuples, self.gpu_id)
self.optimizer = optimizer.load(dir_path, self.model.parameters())
self.scheduler = scheduler.load(dir_path, self.optimizer)
# now we have a loaded tagger
self._loaded = True
def load(path, model_parameters):
options = opts.load(path)
optim = build(options, model_parameters)
load_state(path, optim)
return optim
def load(path):
options = opts.load(path)
return build(options)
def load(path, optim):
options = opts.load(path)
scheduler = build(options, optim)
load_state(path, scheduler)
return scheduler