cycle_lens = list(map(int, cycle_lens))
if len(
cycle_lens
) == 2: # handles option of specifying cycles as pair (n_cycles, cycle_len)
cycle_lens = cycle_lens[0] * [cycle_lens[1]]
im_size = tuple([int(item) for item in args.im_size.split(',')])
if isinstance(im_size, tuple) and len(im_size) == 1:
tg_size = (im_size[0], im_size[0])
elif isinstance(im_size, tuple) and len(im_size) == 2:
tg_size = (im_size[0], im_size[1])
else:
sys.exit('im_size should be a number or a tuple of two numbers')
do_not_save = str2bool(args.do_not_save)
if do_not_save is False:
save_path = args.save_path
if save_path == 'date_time':
save_path = datetime.now().strftime("%Y-%m-%d-%H:%M:%S")
experiment_path = osp.join('experiments', save_path)
args.experiment_path = experiment_path
os.makedirs(experiment_path, exist_ok=True)
config_file_path = osp.join(experiment_path, 'config.cfg')
with open(config_file_path, 'w') as f:
json.dump(vars(args), f, indent=2)
else:
experiment_path = None
csv_train = args.csv_train
if args.device.startswith("cuda"):
# In case one has multiple devices, we must first set the one
# we would like to use so pytorch can find it.
os.environ['CUDA_VISIBLE_DEVICES'] = args.device.split(":", 1)[1]
if not torch.cuda.is_available():
raise RuntimeError("cuda is not currently available!")
print(f"* Running prediction on device '{args.device}'...")
device = torch.device("cuda")
else: #cpu
device = torch.device(args.device)
dataset = args.dataset
binarize = args.binarize
tta = args.tta
public = str2bool(args.public)
# parse config file if provided
config_file = args.config_file
if config_file is not None:
if not osp.isfile(config_file):
raise Exception('non-existent config file')
with open(args.config_file, 'r') as f:
args.__dict__.update(json.load(f))
experiment_path = args.experiment_path # these should exist in a config file
model_name = args.model_name
in_c = args.in_c
if experiment_path is None:
raise Exception('must specify path to experiment')