if 'train' in cfg.data:
cfg.data.train.data_dir = args.data_dir
if 'val' in cfg.data:
cfg.data.val.data_dir = args.data_dir
if 'test' in cfg.data:
cfg.data.test.data_dir = args.data_dir
# init distributed env first, since logger depends on the dist info.
if args.launcher == 'none':
distributed = False
else:
distributed = True
init_dist(args.launcher, **cfg.dist_params)
# init logger before other steps
logger = get_root_logger(log_level=cfg.log_level)
# set random seeds
if args.seed is not None:
logger.info('Set random seed to {}'.format(args.seed))
set_random_seed(args.seed)
model = build_model(cfg.model)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
test_dataset = build_dataset(cfg.data.val)
test_network(model,
test_dataset,
cfg=cfg,
dict(type='GroupScale',
scales=[(171, 128)]),
dict(type='GroupCenterCrop',
out_size=112),
dict(type='GroupToTensor',
switch_rgb_channels=True,
div255=True,
mean=(0.485, 0.456, 0.406),
std=(0.229, 0.224, 0.225))
])),
)
return Dict(data)
if __name__ == '__main__':
logger = get_root_logger(log_level='INFO')
args = parse_args()
_cfg = Config.fromfile(args.cfg)
cfg = dict(model=prepare_model_config(_cfg['model']['backbone']),
data=prepare_data_config(args.dataset_name, args.data_dir))
cfg = Dict(cfg)
cfg.gpus = args.gpus
cfg.data.videos_per_gpu = args.batchsize
cfg.work_dir = args.work_dir
if args.checkpoint is None:
load_from = cfg.work_dir
else:
load_from = args.checkpoint