phlower.settings.PhlowerPredictorSetting

class phlower.settings.PhlowerPredictorSetting(selection_mode, device='cpu', log_file_name='log', saved_setting_filename='model', batch_size=1, num_workers=1, non_blocking=False, random_seed=0)

Bases: object

Methods

__init__(*args, **kwargs)
check_valid_selection_mode(name)

Attributes

batch_size	batch size.
device	device name.
log_file_name	name of log file.
non_blocking
num_workers	the number of cores.
random_seed	random seed.
saved_setting_filename	file name of pretrained model setting.
selection_mode	Define method to select checkpoint file.

Parameters:

• selection_mode (str)

• device (str)

• log_file_name (str)

• saved_setting_filename (str)

• batch_size (int)

• num_workers (int)

• non_blocking (bool)

• random_seed (int)

batch_size: int = 1

batch size. Defaults to 1

device: str = 'cpu'

device name. Defaults to cpu

log_file_name: str = 'log'

name of log file. Defaults to “log”

num_workers: int = 1

the number of cores. Defaults to 1.

random_seed: int = 0

random seed. Defaults to 0

saved_setting_filename: str = 'model'

file name of pretrained model setting. Defaults to “model”

selection_mode: str

Define method to select checkpoint file. Choose from “best”, “latest”, “train_best”, “specified”