def main(params):
config = io_utils.load_yaml(params["config"])
# prepare dataset
D = cmf.get_dataset(params["dataset"])
dsets, L = cmf.get_loader(D,
split=["test"],
loader_configs=[config["test_loader"]],
num_workers=params["num_workers"])
# Build network
M = cmf.get_method(params["method"])
net = M(config, logger=None)
net.load_checkpoint(params["checkpoint"], True)
if config["model"]["use_gpu"]: net.gpu_mode()
# Evaluating networks
cmf.test(config, L["test"], net, -1, None, mode="Test")
def main():
# get parameters from cmd
params = _get_argument_params()
global M, dataset
M = cmf.get_model(params["model_type"])
dataset = cmf.get_dataset(params["dataset"])
# loading configuration and setting environment
config = io_utils.load_yaml(params["config_path"])
config = M.override_config_from_params(config, params)
cmf.create_save_dirs(config["misc"])
# create loggers
global logger
logger = cmf.create_logger(config)
# train network
train(config)
parser.add_argument("--debug_mode",
action="store_true",
default=False,
help="Train the model in debug mode.")
params = vars(parser.parse_args())
print(json.dumps(params, indent=4))
return params
if __name__ == "__main__":
# load parameters
params = _get_argument_params()
global M, dataset
M = cmf.get_model(params["model_type"])
dataset = cmf.get_dataset(params["dataset"])
# loading configuration and setting environment
config = io_utils.load_yaml(params["config_path"])
config = M.override_config_from_params(config, params)
cmf.create_save_dirs(config["misc"])
""" Build data loader """
if params["mode"] == "train":
dset = dataset.DataSet(config["train_loader"])
else:
dset = dataset.DataSet(config["test_loader"])
L = data.DataLoader(dset, batch_size=64, \
num_workers=config["misc"]["num_workers"], \
shuffle=False, collate_fn=dataset.collate_fn)
config = M.override_config_from_loader(config, dset)