multiml.task.pytorch.pytorch_util module

multiml.task.pytorch.pytorch_util.compile(obj, obj_args, modules)

multiml.task.pytorch.pytorch_util.inputs_size(inputs)

multiml.task.pytorch.pytorch_util.is_master_process(rank=0)

Check if master process or not..

class multiml.task.pytorch.pytorch_util.EarlyStopping(patience=None, save=False, path=None)

Bases: object

__init__(patience=None, save=False, path=None)

save_checkpoint(val_loss, model)

Saves model when validation loss decrease.

class multiml.task.pytorch.pytorch_util.ASNG_EarlyStopping(patience=None, save=False, path=None)

Bases: EarlyStopping

save_checkpoint(val_loss, model)

Saves model when validation loss decrease.

get_thetas()