def parse_args():
parser = argparse.ArgumentParser(
description=
"Evaluate a model for image classification/segmentation (Chainer)",
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument(
"--dataset",
type=str,
default="ImageNet1K",
help=
"dataset name. options are ImageNet1K, CUB200_2011, CIFAR10, CIFAR100, SVHN, VOC2012, ADE20K, Cityscapes, "
"COCO")
parser.add_argument(
"--work-dir",
type=str,
default=os.path.join("..", "imgclsmob_data"),
help="path to working directory only for dataset root path preset")
args, _ = parser.parse_known_args()
dataset_metainfo = get_dataset_metainfo(dataset_name=args.dataset)
dataset_metainfo.add_dataset_parser_arguments(parser=parser,
work_dir_path=args.work_dir)
add_eval_parser_arguments(parser)
args = parser.parse_args()
return args
def test_model(args):
"""
Main test routine.
Parameters:
----------
args : ArgumentParser
Main script arguments.
Returns
-------
float
Main accuracy value.
"""
ds_metainfo = get_dataset_metainfo(dataset_name=args.dataset)
ds_metainfo.update(args=args)
assert (ds_metainfo.ml_type != "imgseg") or (args.batch_size == 1)
assert (ds_metainfo.ml_type != "imgseg") or args.disable_cudnn_autotune
global_config.train = False
use_gpus = prepare_ch_context(args.num_gpus)
net = prepare_model(
model_name=args.model,
use_pretrained=args.use_pretrained,
pretrained_model_file_path=args.resume.strip(),
use_gpus=use_gpus,
net_extra_kwargs=ds_metainfo.test_net_extra_kwargs,
num_classes=(args.num_classes if ds_metainfo.ml_type != "hpe" else None),
in_channels=args.in_channels)
assert (hasattr(net, "classes") or (ds_metainfo.ml_type == "hpe"))
assert (hasattr(net, "in_size"))
get_test_data_source_class = get_val_data_source if args.data_subset == "val" else get_test_data_source
test_data = get_test_data_source_class(
ds_metainfo=ds_metainfo,
batch_size=args.batch_size,
num_workers=args.num_workers)
if args.data_subset == "val":
test_metric = get_composite_metric(
metric_names=ds_metainfo.val_metric_names,
metric_extra_kwargs=ds_metainfo.val_metric_extra_kwargs)
else:
test_metric = get_composite_metric(
metric_names=ds_metainfo.test_metric_names,
metric_extra_kwargs=ds_metainfo.test_metric_extra_kwargs)
assert (args.use_pretrained or args.resume.strip())
acc_values = calc_model_accuracy(
net=net,
test_data=test_data,
metric=test_metric,
calc_weight_count=True,
calc_flops_only=args.calc_flops_only,
extended_log=True)
return acc_values[ds_metainfo.saver_acc_ind] if len(acc_values) > 0 else None
def main():
"""
Main body of script.
"""
args = parse_args()
if args.disable_cudnn_autotune:
os.environ["MXNET_CUDNN_AUTOTUNE_DEFAULT"] = "0"
_, log_file_exist = initialize_logging(
logging_dir_path=args.save_dir,
logging_file_name=args.logging_file_name,
script_args=args,
log_packages=args.log_packages,
log_pip_packages=args.log_pip_packages)
ds_metainfo = get_dataset_metainfo(dataset_name=args.dataset)
ds_metainfo.update(args=args)
assert (ds_metainfo.ml_type != "imgseg") or (args.batch_size == 1)
assert (ds_metainfo.ml_type != "imgseg") or args.disable_cudnn_autotune
global_config.train = False
use_gpus = prepare_ch_context(args.num_gpus)
net = prepare_model(model_name=args.model,
use_pretrained=args.use_pretrained,
pretrained_model_file_path=args.resume.strip(),
use_gpus=use_gpus,
net_extra_kwargs=ds_metainfo.net_extra_kwargs,
num_classes=args.num_classes,
in_channels=args.in_channels)
assert (hasattr(net, "classes"))
assert (hasattr(net, "in_size"))
if args.data_subset == "val":
get_test_data_source_class = get_val_data_source
test_metric = get_composite_metric(
metric_names=ds_metainfo.val_metric_names,
metric_extra_kwargs=ds_metainfo.val_metric_extra_kwargs)
else:
get_test_data_source_class = get_test_data_source
test_metric = get_composite_metric(
metric_names=ds_metainfo.test_metric_names,
metric_extra_kwargs=ds_metainfo.test_metric_extra_kwargs)
test_data = get_test_data_source_class(ds_metainfo=ds_metainfo,
batch_size=args.batch_size,
num_workers=args.num_workers)
assert (args.use_pretrained or args.resume.strip())
test(net=net,
test_data=test_data,
metric=test_metric,
calc_weight_count=True,
extended_log=True)
def main():
args = parse_args()
args.seed = init_rand(seed=args.seed)
_, log_file_exist = initialize_logging(
logging_dir_path=args.save_dir,
logging_file_name=args.logging_file_name,
script_args=args,
log_packages=args.log_packages,
log_pip_packages=args.log_pip_packages)
ds_metainfo = get_dataset_metainfo(dataset_name=args.dataset)
ds_metainfo.update(args=args)
use_gpus = prepare_ch_context(args.num_gpus)
# batch_size = args.batch_size
net = prepare_model(
model_name=args.model,
use_pretrained=args.use_pretrained,
pretrained_model_file_path=args.resume.strip(),
use_gpus=use_gpus,
num_classes=args.num_classes,
in_channels=args.in_channels)
assert (hasattr(net, "classes"))
assert (hasattr(net, "in_size"))
train_data = get_train_data_source(
ds_metainfo=ds_metainfo,
batch_size=args.batch_size,
num_workers=args.num_workers)
val_data = get_val_data_source(
ds_metainfo=ds_metainfo,
batch_size=args.batch_size,
num_workers=args.num_workers)
trainer = prepare_trainer(
net=net,
optimizer_name=args.optimizer_name,
lr=args.lr,
momentum=args.momentum,
num_epochs=args.num_epochs,
train_data=train_data,
val_data=val_data,
logging_dir_path=args.save_dir,
use_gpus=use_gpus)
trainer.run()
