def train_reid(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
logger = get_root_logger()
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ingored if distributed
len(cfg.gpu_ids),
dist=distributed,
sampler=cfg.data.get('sampler', None),
seed=cfg.seed) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Set the `find_unused_parameters` parameter in torch's DDP
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = EpochBasedRunner(model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
runner.timestamp = timestamp
# register hooks
runner.register_training_hooks(cfg.lr_config, cfg.optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
runner.register_hook(DistSamplerSeedHook())
# register eval hook
if validate:
test_dataset = build_dataset(cfg.data.test)
test_dataloader = build_dataloader(
test_dataset,
samples_per_gpu=cfg.data.samples_per_gpu,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
eval_cfg = cfg.get('evaluation', {})
runner.register_hook(EvalHook(test_dataloader, **eval_cfg))
# user-defined hooks
if cfg.get('custom_hooks', None):
custom_hooks = cfg.custom_hooks
assert isinstance(custom_hooks, list), \
f'custom_hooks expect list type, but got: {type(custom_hooks)}'
for hook_cfg in cfg.custom_hooks:
assert isinstance(hook_cfg, dict), \
'each item in custom hooks expects dict type, but got: ' \
f'{type(hook_cfg)}'
hook_cfg = hook_cfg.copy()
priority = hook_cfg.pop('priority', 'NORMAL')
hook = build_from_cfg(hook_cfg, HOOKS)
logger.info(f'Register custom hook: {hook_cfg.type}')
runner.register_hook(hook, priority=priority)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def main():
args = parse_args()
assert args.out or args.eval or args.format_only or args.show \
or args.show_dir, \
('Please specify at least one operation (save/eval/format/show the '
'results / save the results) with the argument "--out", "--eval"'
', "--format-only", "--show" or "--show-dir"')
if args.eval and args.format_only:
raise ValueError('--eval and --format_only cannot be both specified')
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = Config.fromfile(args.config)
if cfg.get('USE_MMDET', False):
from mmdet.apis import multi_gpu_test, single_gpu_test
from mmdet.datasets import build_dataloader
from mmdet.models import build_detector as build_model
if 'detector' in cfg.model:
cfg.model = cfg.model.detector
elif cfg.get('TRAIN_REID', False):
from mmdet.apis import multi_gpu_test, single_gpu_test
from mmdet.datasets import build_dataloader
from mmtrack.models import build_reid as build_model
if 'reid' in cfg.model:
cfg.model = cfg.model.reid
else:
from mmtrack.apis import multi_gpu_test, single_gpu_test
from mmtrack.datasets import build_dataloader
from mmtrack.models import build_model
if args.cfg_options is not None:
cfg.merge_from_dict(args.cfg_options)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
# cfg.model.pretrains = None
if hasattr(cfg.model, 'detector'):
cfg.model.detector.pretrained = None
cfg.data.test.test_mode = True
# 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)
# build the dataloader
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(dataset,
samples_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
if cfg.get('test_cfg', False):
model = build_model(cfg.model,
train_cfg=cfg.train_cfg,
test_cfg=cfg.test_cfg)
else:
model = build_model(cfg.model)
# We need call `init_weights()` to load pretained weights in MOT task.
model.init_weights()
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
if args.checkpoint is not None:
checkpoint = load_checkpoint(model,
args.checkpoint,
map_location='cpu')
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
if not hasattr(model, 'CLASSES'):
model.CLASSES = dataset.CLASSES
if args.fuse_conv_bn:
model = fuse_conv_bn(model)
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model,
data_loader,
args.show,
args.show_dir,
show_score_thr=args.show_score_thr)
else:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = multi_gpu_test(model, data_loader, args.tmpdir,
args.gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
if args.out:
print(f'\nwriting results to {args.out}')
mmcv.dump(outputs, args.out)
kwargs = {} if args.eval_options is None else args.eval_options
if args.format_only:
dataset.format_results(outputs, **kwargs)
if args.eval:
eval_kwargs = cfg.get('evaluation', {}).copy()
# hard-code way to remove EvalHook args
eval_hook_args = [
'interval', 'tmpdir', 'start', 'gpu_collect', 'save_best',
'rule', 'by_epoch'
]
for key in eval_hook_args:
eval_kwargs.pop(key, None)
eval_kwargs.update(dict(metric=args.eval, **kwargs))
print(dataset.evaluate(outputs, **eval_kwargs))
def train_model(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
"""Train model entry function.
Args:
model (nn.Module): The model to be trained.
dataset (:obj:`Dataset`): Train dataset.
cfg (dict): The config dict for training.
distributed (bool): Whether to use distributed training.
Default: False.
validate (bool): Whether to do evaluation. Default: False.
timestamp (str | None): Local time for runner. Default: None.
meta (dict | None): Meta dict to record some important information.
Default: None
"""
logger = get_root_logger(log_level=cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', {}),
workers_per_gpu=cfg.data.get('workers_per_gpu', {}),
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('train_dataloader', {}))
data_loaders = [
build_dataloader(ds, **dataloader_setting) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(
model.cuda(cfg.gpu_ids[0]), device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = EpochBasedRunner(
model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# an ugly workaround to make .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(
**cfg.optimizer_config, **fp16_cfg, distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = OptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
runner.register_hook(DistSamplerSeedHook())
if validate:
eval_cfg = cfg.get('evaluation', {})
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', {}),
workers_per_gpu=cfg.data.get('workers_per_gpu', {}),
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
shuffle=False)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('val_dataloader', {}))
val_dataloader = build_dataloader(val_dataset, **dataloader_setting)
eval_hook = DistEvalHook if distributed else EvalHook
runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def train_model(model,
dataset,
cfg,
distributed=False,
validate=False,
test=dict(test_best=False, test_last=False),
timestamp=None,
meta=None):
"""Train model entry function.
Args:
model (nn.Module): The model to be trained.
dataset (:obj:`Dataset`): Train dataset.
cfg (dict): The config dict for training.
distributed (bool): Whether to use distributed training.
Default: False.
validate (bool): Whether to do evaluation. Default: False.
test (dict): The testing option, with two keys: test_last & test_best.
The value is True or False, indicating whether to test the
corresponding checkpoint.
Default: dict(test_best=False, test_last=False).
timestamp (str | None): Local time for runner. Default: None.
meta (dict | None): Meta dict to record some important information.
Default: None
"""
logger = get_root_logger(log_level=cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=cfg.data.get('workers_per_gpu', 1),
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('train_dataloader', {}))
if cfg.omnisource:
# The option can override videos_per_gpu
train_ratio = cfg.data.get('train_ratio', [1] * len(dataset))
omni_videos_per_gpu = cfg.data.get('omni_videos_per_gpu', None)
if omni_videos_per_gpu is None:
dataloader_settings = [dataloader_setting] * len(dataset)
else:
dataloader_settings = []
for videos_per_gpu in omni_videos_per_gpu:
this_setting = cp.deepcopy(dataloader_setting)
this_setting['videos_per_gpu'] = videos_per_gpu
dataloader_settings.append(this_setting)
data_loaders = [
build_dataloader(ds, **setting)
for ds, setting in zip(dataset, dataloader_settings)
]
else:
data_loaders = [
build_dataloader(ds, **dataloader_setting) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(
model.cuda(cfg.gpu_ids[0]), device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
Runner = OmniSourceRunner if cfg.omnisource else EpochBasedRunner
runner = Runner(
model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# an ugly workaround to make .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(
**cfg.optimizer_config, **fp16_cfg, distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = OptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
if cfg.omnisource:
runner.register_hook(OmniSourceDistSamplerSeedHook())
else:
runner.register_hook(DistSamplerSeedHook())
# precise bn setting
if cfg.get('precise_bn', False):
precise_bn_dataset = build_dataset(cfg.data.train)
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=1, # save memory and time
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed)
data_loader_precise_bn = build_dataloader(precise_bn_dataset,
**dataloader_setting)
precise_bn_hook = PreciseBNHook(data_loader_precise_bn,
**cfg.get('precise_bn'))
runner.register_hook(precise_bn_hook)
if validate:
eval_cfg = cfg.get('evaluation', {})
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=cfg.data.get('workers_per_gpu', 1),
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
shuffle=False)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('val_dataloader', {}))
val_dataloader = build_dataloader(val_dataset, **dataloader_setting)
eval_hook = DistEvalHook(val_dataloader, **eval_cfg) if distributed \
else EvalHook(val_dataloader, **eval_cfg)
runner.register_hook(eval_hook)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner_kwargs = dict()
if cfg.omnisource:
runner_kwargs = dict(train_ratio=train_ratio)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs, **runner_kwargs)
if test['test_last'] or test['test_best']:
best_ckpt_path = None
if test['test_best']:
if hasattr(eval_hook, 'best_ckpt_path'):
best_ckpt_path = eval_hook.best_ckpt_path
if best_ckpt_path is None or not osp.exists(best_ckpt_path):
test['test_best'] = False
if best_ckpt_path is None:
runner.logger.info('Warning: test_best set as True, but '
'is not applicable '
'(eval_hook.best_ckpt_path is None)')
else:
runner.logger.info('Warning: test_best set as True, but '
'is not applicable (best_ckpt '
f'{best_ckpt_path} not found)')
if not test['test_last']:
return
test_dataset = build_dataset(cfg.data.test, dict(test_mode=True))
gpu_collect = cfg.get('evaluation', {}).get('gpu_collect', False)
tmpdir = cfg.get('evaluation', {}).get('tmpdir',
osp.join(cfg.work_dir, 'tmp'))
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=cfg.data.get('workers_per_gpu', 1),
num_gpus=len(cfg.gpu_ids),
dist=distributed,
shuffle=False)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('test_dataloader', {}))
test_dataloader = build_dataloader(test_dataset, **dataloader_setting)
names, ckpts = [], []
if test['test_last']:
names.append('last')
ckpts.append(None)
if test['test_best']:
names.append('best')
ckpts.append(best_ckpt_path)
for name, ckpt in zip(names, ckpts):
if ckpt is not None:
runner.load_checkpoint(ckpt)
outputs = multi_gpu_test(runner.model, test_dataloader, tmpdir,
gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
out = osp.join(cfg.work_dir, f'{name}_pred.pkl')
test_dataset.dump_results(outputs, out)
eval_cfg = cfg.get('evaluation', {})
for key in [
'interval', 'tmpdir', 'start', 'gpu_collect',
'save_best', 'rule', 'by_epoch', 'broadcast_bn_buffers'
]:
eval_cfg.pop(key, None)
eval_res = test_dataset.evaluate(outputs, **eval_cfg)
runner.logger.info(f'Testing results of the {name} checkpoint')
for metric_name, val in eval_res.items():
runner.logger.info(f'{metric_name}: {val:.04f}')
def main():
args = parse_args()
modele = osp.splitext(osp.basename(args.config))[0]
print(modele)
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# init distributed env first, since logger depends on the dist info.
if args.launcher == 'none':
distributed = False
print("not distributed")
else:
distributed = True
init_dist(args.launcher, **cfg.dist_params)
print("distributed")
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
nrows = 100
chunksize = int(4 * 1e3)
dataset = 'vit_files_img_xml_trunc'
log_dataset = 'log_%s_test' % modele
box_dataset = 'box_%s_test' % modele
img_df = load_data(dataset, nrows=nrows)
print(img_df.shape)
# Reprise sur incident
img_db = read_dataframe(API_KEY_VIT,
VERTICA_HOST,
PROJECT_KEY_VIT,
log_dataset,
columns=['path', 'img_name'])
img_db = img_db.eval('path + img_name').unique()
print('there is %s images seen' % len(img_db))
img_df = img_df.loc[~img_df.eval('path + img_name').isin(img_db), :]
print("Need to process : %s images" % img_df.shape[0])
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
load_checkpoint(model, args.checkpoint, map_location='cpu')
if not distributed:
model = MMDataParallel(model, device_ids=[0])
#outputs = single_gpu_test(model, data_loader, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
#outputs = multi_gpu_test(model, data_loader, args.tmpdir)
for chunk_df in chunker(img_df, chunksize):
img_seg = pd.DataFrame(columns=col_seg)
log_seg = pd.DataFrame(columns=['img_name', 'path', 'traceback'])
car_dataset = CustomDataset(chunk_df, ROOT_DIR, **cfg.data.val)
print('length of dataset :')
print(len(car_dataset))
data_loader = build_dataloader(
car_dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
if not distributed:
outputs, logs = single_gpu_test(model, data_loader, args.show)
else:
outputs, logs = multi_gpu_test(model, data_loader)
indices = list(data_loader.sampler)
# Convert and save
for (_, row), results, traceback in zip(
chunk_df.iloc[indices, :].iterrows(), outputs, logs):
to_save = save_result(results,
class_to_keep=class_to_keep,
dataset='coco',
score_thr=0.3)
to_save_df = pd.DataFrame(to_save)
to_save_df['img_name'] = row['img_name']
to_save_df['path'] = row['path']
img_seg = pd.concat([img_seg, to_save_df],
ignore_index=True,
sort=False)
log_seg = log_seg.append(dict(img_name=row['img_name'],
path=row['path'],
traceback=traceback),
ignore_index=True)
img_seg['score'] = img_seg['score'].round(decimals=2)
print('%s images processed' % log_seg.shape[0])
write_dataframe(VERTICA_HOST, PROJECT_KEY_VIT, log_dataset, log_seg)
write_dataframe(VERTICA_HOST, PROJECT_KEY_VIT, box_dataset,
img_seg[col_seg])
def main():
args = parse_args()
cfg = mmcv.Config.fromfile(args.config)
# Wudi change the args.out directly related to the model checkpoint file data
args.out = os.path.join(
cfg.work_dir,
cfg.data.test.img_prefix.split('/')[-3] +
args.checkpoint.split('/')[-2] + '_' +
args.checkpoint.split('/')[-1][:-4] + '.pkl')
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# 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)
# build the dataloader
dataset = build_dataset(cfg.data.test)
if not os.path.exists(args.out):
data_loader = build_dataloader(
dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model,
train_cfg=None,
test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model,
args.checkpoint,
map_location='cpu')
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
mmcv.dump(outputs, args.out)
else:
outputs = mmcv.load(args.out)
if distributed:
rank, _ = get_dist_info()
if rank != 0:
return
# Now we visualise the output here
dataset.visualise_pred(outputs, args)
def main():
args = parse_args()
assert args.out or args.show, \
('Please specify at least one operation (save or show the results) '
'with the argument "--out" or "--show"')
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
if args.workers == 0:
args.workers = cfg.data.workers_per_gpu
# 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)
# set random seeds
if args.seed is not None:
set_random_seed(args.seed)
if 'all' in args.corruptions:
corruptions = [
'gaussian_noise', 'shot_noise', 'impulse_noise', 'defocus_blur',
'glass_blur', 'motion_blur', 'zoom_blur', 'snow', 'frost', 'fog',
'brightness', 'contrast', 'elastic_transform', 'pixelate',
'jpeg_compression', 'speckle_noise', 'gaussian_blur', 'spatter',
'saturate'
]
elif 'benchmark' in args.corruptions:
corruptions = [
'gaussian_noise', 'shot_noise', 'impulse_noise', 'defocus_blur',
'glass_blur', 'motion_blur', 'zoom_blur', 'snow', 'frost', 'fog',
'brightness', 'contrast', 'elastic_transform', 'pixelate',
'jpeg_compression'
]
elif 'noise' in args.corruptions:
corruptions = ['gaussian_noise', 'shot_noise', 'impulse_noise']
elif 'blur' in args.corruptions:
corruptions = [
'defocus_blur', 'glass_blur', 'motion_blur', 'zoom_blur'
]
elif 'weather' in args.corruptions:
corruptions = ['snow', 'frost', 'fog', 'brightness']
elif 'digital' in args.corruptions:
corruptions = [
'contrast', 'elastic_transform', 'pixelate', 'jpeg_compression'
]
elif 'holdout' in args.corruptions:
corruptions = ['speckle_noise', 'gaussian_blur', 'spatter', 'saturate']
elif 'None' in args.corruptions:
corruptions = ['None']
args.severities = [0]
else:
corruptions = args.corruptions
aggregated_results = {}
for corr_i, corruption in enumerate(corruptions):
aggregated_results[corruption] = {}
for sev_i, corruption_severity in enumerate(args.severities):
# evaluate severity 0 (= no corruption) only once
if corr_i > 0 and corruption_severity == 0:
aggregated_results[corruption][0] = \
aggregated_results[corruptions[0]][0]
continue
test_data_cfg = copy.deepcopy(cfg.data.test)
# assign corruption and severity
if corruption_severity > 0:
corruption_trans = dict(type='Corrupt',
corruption=corruption,
severity=corruption_severity)
# TODO: hard coded "1", we assume that the first step is
# loading images, which needs to be fixed in the future
test_data_cfg['pipeline'].insert(1, corruption_trans)
# print info
print('\nTesting {} at severity {}'.format(corruption,
corruption_severity))
# build the dataloader
# TODO: support multiple images per gpu
# (only minor changes are needed)
dataset = build_dataset(test_data_cfg)
data_loader = build_dataloader(dataset,
imgs_per_gpu=1,
workers_per_gpu=args.workers,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model,
train_cfg=None,
test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model,
args.checkpoint,
map_location='cpu')
# old versions did not save class info in checkpoints,
# this walkaround is for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
rank, _ = get_dist_info()
if args.out and rank == 0:
eval_results_filename = (osp.splitext(args.out)[0] +
'_results' +
osp.splitext(args.out)[1])
mmcv.dump(outputs, args.out)
eval_types = args.eval
if cfg.dataset_type == 'VOCDataset':
if eval_types:
for eval_type in eval_types:
if eval_type == 'bbox':
test_dataset = mmcv.runner.obj_from_dict(
cfg.data.test, datasets)
logger = 'print' if args.summaries else None
mean_ap, eval_results = \
voc_eval_with_return(
args.out, test_dataset,
args.iou_thr, logger)
aggregated_results[corruption][
corruption_severity] = eval_results
else:
print('\nOnly "bbox" evaluation \
is supported for pascal voc')
else:
if eval_types:
print('Starting evaluate {}'.format(
' and '.join(eval_types)))
if eval_types == ['proposal_fast']:
result_file = args.out
else:
if not isinstance(outputs[0], dict):
result_files = results2json(
dataset, outputs, args.out)
else:
for name in outputs[0]:
print('\nEvaluating {}'.format(name))
outputs_ = [out[name] for out in outputs]
result_file = args.out
+ '.{}'.format(name)
result_files = results2json(
dataset, outputs_, result_file)
eval_results = coco_eval_with_return(
result_files, eval_types, dataset.coco)
aggregated_results[corruption][
corruption_severity] = eval_results
else:
print('\nNo task was selected for evaluation;'
'\nUse --eval to select a task')
# save results after each evaluation
mmcv.dump(aggregated_results, eval_results_filename)
# print filan results
print('\nAggregated results:')
prints = args.final_prints
aggregate = args.final_prints_aggregate
if cfg.dataset_type == 'VOCDataset':
get_results(eval_results_filename,
dataset='voc',
prints=prints,
aggregate=aggregate)
else:
get_results(eval_results_filename,
dataset='coco',
prints=prints,
aggregate=aggregate)
def _dist_train(model, dataset, cfg, logger=None, timestamp=None, meta=None):
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
data_loaders = [
build_dataloader(
ds,
cfg.data.imgs_per_gpu,
cfg.data.workers_per_gpu,
dist=True,
shuffle=True,
replace=getattr(cfg.data, 'sampling_replace', False),
seed=cfg.seed,
drop_last=getattr(cfg.data, 'drop_last', False),
# my
repeat=getattr(cfg.data, 'repeat', 1)) for ds in dataset
]
optimizer = build_optimizer(model, cfg.optimizer)
if 'use_fp16' in cfg and cfg.use_fp16 == True:
model, optimizer = apex.amp.initialize(model.cuda(),
optimizer,
opt_level="O1")
print_log('**** Initializing mixed precision done. ****')
# put model on gpus
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
# my
find_unused_parameters=True)
#
# build runner
# my
global Runner
if cfg.workflow[0][0] is 'search':
Runner = SearchRunner
#
runner = Runner(model,
batch_processor,
optimizer,
cfg.work_dir,
logger=logger,
meta=meta)
# an ugly walkaround to make the .log and .log.json filenames the same
runner.timestamp = timestamp
optimizer_config = DistOptimizerHook(**cfg.optimizer_config)
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config)
runner.register_hook(DistSamplerSeedHook())
# register custom hooks
for hook in cfg.get('custom_hooks', ()):
priority = hook.pop('priority', None)
if hook.type == 'DeepClusterHook':
common_params = dict(dist_mode=True, data_loaders=data_loaders)
else:
common_params = dict(dist_mode=True)
if priority is None:
runner.register_hook(build_hook(hook, common_params))
else:
runner.register_hook(build_hook(hook, common_params), priority)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def train_yolo(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
logger = get_root_logger(cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
if 'imgs_per_gpu' in cfg.data:
logger.warning('"imgs_per_gpu" is deprecated in MMDet V2.0. '
'Please use "samples_per_gpu" instead')
if 'samples_per_gpu' in cfg.data:
logger.warning(
f'Got "imgs_per_gpu"={cfg.data.imgs_per_gpu} and '
f'"samples_per_gpu"={cfg.data.samples_per_gpu}, "imgs_per_gpu"'
f'={cfg.data.imgs_per_gpu} is used in this experiments')
else:
logger.warning(
'Automatically set "samples_per_gpu"="imgs_per_gpu"='
f'{cfg.data.imgs_per_gpu} in this experiments')
cfg.data.samples_per_gpu = cfg.data.imgs_per_gpu
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = EpochBasedRunner(model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# an ugly workaround to make .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(**cfg.optimizer_config,
**fp16_cfg,
distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = OptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
# Support batch_size > 1 in validation
val_samples_per_gpu = cfg.data.val.pop('samples_per_gpu', 1)
if val_samples_per_gpu > 1:
# Replace 'ImageToTensor' to 'DefaultFormatBundle'
cfg.data.val.pipeline = replace_ImageToTensor(
cfg.data.val.pipeline)
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_dataloader = build_dataloader(
val_dataset,
samples_per_gpu=val_samples_per_gpu,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
eval_cfg = cfg.get('evaluation', {})
eval_hook = DistEvalHook if distributed else EvalHook
runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
# user-defined hooks
if cfg.get('custom_hooks', None):
custom_hooks = cfg.custom_hooks
assert isinstance(custom_hooks, list), \
f'custom_hooks expect list type, but got {type(custom_hooks)}'
for hook_cfg in cfg.custom_hooks:
assert isinstance(hook_cfg, dict), \
'Each item in custom_hooks expects dict type, but got ' \
f'{type(hook_cfg)}'
hook_cfg = hook_cfg.copy()
priority = hook_cfg.pop('priority', 'NORMAL')
hook = build_from_cfg(hook_cfg, HOOKS)
runner.register_hook(hook, priority=priority)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def train_segmentor(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
"""Launch segmentor training."""
logger = get_root_logger(cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed,
drop_last=True) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = IterBasedRunner(model=model,
batch_processor=None,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# register hooks
runner.register_training_hooks(cfg.lr_config, cfg.optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
# an ugly walkaround to make the .log and .log.json filenames the same
runner.timestamp = timestamp
# register eval hooks
if validate:
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_dataloader = build_dataloader(
val_dataset,
samples_per_gpu=4,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
eval_cfg = cfg.get('evaluation', {})
eval_hook = DistEvalHook if distributed else EvalHook
runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_iters)
def main(): # noqa: C901
"""Start test."""
args = parse_args()
if args.work_dir is not None:
mmcv.mkdir_or_exist(args.work_dir)
if args.tmpdir is None:
args.tmpdir = osp.join(args.work_dir, 'tmp_dir')
mmcv.mkdir_or_exist(args.tmpdir)
if args.out is None:
args.out = osp.join(args.work_dir, 'result.pkl')
if args.checkpoint is None:
args.checkpoint = osp.join(args.work_dir, 'latest.pth')
fps_file = osp.join(args.work_dir, 'fps.pkl')
mAP_file = osp.join(args.work_dir, 'mAP.pkl')
else:
mAP_file, fps_file = None, None
if args.checkpoint is None:
raise ValueError('Checkpoint file cannot be empty.')
if args.config.endswith(".json"):
load_method = mmcv.load
mmcv.load = json_to_dict
cfg = mmcv.Config.fromfile(args.config)
mmcv.load = load_method
else:
cfg = mmcv.Config.fromfile(args.config)
cfg.model.pretrained = None
cfg.data.test.test_mode = True
if args.dist:
init_dist('pytorch', **cfg.dist_params)
# build the dataloader
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=True,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
load_checkpoint(model, args.checkpoint, map_location='cpu')
model.CLASSES = dataset.CLASSES
if args.dist:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
else:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, fps_file)
rank, _ = get_dist_info()
if args.out and rank == 0:
print('\nwriting results to {}'.format(args.out))
mmcv.dump(outputs, args.out)
eval_types = args.eval
if eval_types:
print('Starting evaluate {}'.format(' and '.join(eval_types)))
assert not isinstance(outputs[0], dict)
result_files = results2json(dataset, outputs, args.out)
coco_eval(result_files,
eval_types,
dataset.coco,
dump_file=mAP_file)
def _dist_train(model, dataset, cfg, validate=False):
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
if 'use_img_sampling' not in cfg.data:
cfg.data.update({'use_img_sampling': False})
if 'use_sample_out' not in cfg.data:
cfg.data.update({'use_sample_out': False})
print('--Dist-train--IS:{}--ISout:{}'.format(cfg.data.use_img_sampling,
cfg.data.use_sample_out))
data_loaders = [
build_dataloader(ds,
cfg.data.imgs_per_gpu,
cfg.data.workers_per_gpu,
dist=True,
use_img_sampling=cfg.data.use_img_sampling,
use_sample_out=cfg.data.use_sample_out)
for ds in dataset
]
# put model on gpus
model = MMDistributedDataParallel(model.cuda())
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = Runner(model, batch_processor, optimizer, cfg.work_dir,
cfg.log_level)
# Add for LVIS by LiYu
import logging
runner.logger.setLevel(logging.INFO)
# ====================
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(**cfg.optimizer_config,
**fp16_cfg)
else:
optimizer_config = DistOptimizerHook(**cfg.optimizer_config)
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config)
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_dataset_cfg = cfg.data.val
eval_cfg = cfg.get('evaluation', {})
if isinstance(model.module, RPN):
# TODO: implement recall hooks for other datasets
runner.register_hook(
CocoDistEvalRecallHook(val_dataset_cfg, **eval_cfg))
else:
dataset_type = DATASETS.get(val_dataset_cfg.type)
if issubclass(dataset_type, datasets.CocoDataset):
runner.register_hook(
CocoDistEvalmAPHook(val_dataset_cfg, **eval_cfg))
else:
runner.register_hook(
DistEvalmAPHook(val_dataset_cfg, **eval_cfg))
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def main():
args = parse_args()
cfg = Config.fromfile(args.config)
if args.work_dir is not None:
cfg.work_dir = args.work_dir
pathlib.Path(cfg.work_dir).mkdir(parents=True, exist_ok=True)
cfg.gpus = args.gpus
# 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(cfg.work_dir)
logger.info('Distributed training: {}'.format(distributed))
# 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_detector(cfg.model,
train_cfg=cfg.train_cfg,
test_cfg=cfg.test_cfg)
if distributed:
model = MMDistributedDataParallel(model).cuda()
else:
model = MMDataParallel(model, device_ids=range(cfg.gpus)).cuda()
train_dataset = get_dataset(cfg.data.train)
optimizer = build_optimizer(model, cfg.optimizer)
train_loader = build_dataloader(train_dataset,
cfg.data.imgs_per_gpu,
cfg.data.workers_per_gpu,
dist=distributed)
start_epoch = it = 0
last_epoch = -1
lr_scheduler, lr_warmup_scheduler = build_scheduler(
optimizer,
total_iters_each_epoch=len(train_loader),
total_epochs=cfg.total_epochs,
last_epoch=last_epoch,
optim_cfg=cfg.optimizer,
lr_cfg=cfg.lr_config)
# -----------------------start training---------------------------
logger.info('**********************Start training**********************')
train_model(model,
optimizer,
train_loader,
lr_scheduler=lr_scheduler,
optim_cfg=cfg.optimizer,
start_epoch=start_epoch,
total_epochs=cfg.total_epochs,
start_iter=it,
rank=args.local_rank,
logger=logger,
ckpt_save_dir=cfg.work_dir,
lr_warmup_scheduler=lr_warmup_scheduler,
ckpt_save_interval=cfg.checkpoint_config.interval,
max_ckpt_save_num=args.max_ckpt_save_num,
log_interval=cfg.log_config.interval)
logger.info('**********************End training**********************')
def main(
config=None,
checkpoint=None,
img_prefix=None,
ann_file=None,
flip=False,
out=None,
format_only=False,
eval=None,
show=False,
show_dir=None,
log_file=None,
fold=None,
show_score_thr=0.3,
gpu_collect=False,
tmpdir=None,
options=None,
launcher="none",
local_rank=0,
score_thr=None,
iou_thr=None,
) -> Tuple[list, WheatDataset]:
logger = get_logger("inference", log_file=log_file, log_mode="a")
assert out or eval or format_only or show or show_dir, (
"Please specify at least one operation (save/eval/format/show the "
'results / save the results) with the argument "--out", "--eval"'
', "--format-only", "--show" or "--show-dir"')
if eval and format_only:
raise ValueError("--eval and --format_only cannot be both specified")
if out is not None and not out.endswith((".pkl", ".pickle")):
raise ValueError("The output file must be a pkl file.")
cfg = Config.fromfile(config)
if fold is not None:
cfg.data.test.ann_file = cfg.data.test.ann_file.format(fold=fold)
if img_prefix is not None:
cfg.data.test.img_prefix = img_prefix
if ann_file is not None:
cfg.data.test.ann_file = ann_file
if flip:
print(cfg.data.test.pipeline[-1])
print(f"flip: {cfg.data.test.pipeline[-1].flip}")
cfg.data.test.pipeline[-1].flip_direction = ["horizontal", "vertical"]
cfg.data.test.pipeline[-1].flip = flip
if score_thr is not None:
if "rcnn" in cfg.test_cfg:
cfg.test_cfg.rcnn.score_thr = score_thr
else:
cfg.test_cfg.score_thr = score_thr
if iou_thr is not None:
if "rcnn" in cfg.test_cfg:
cfg.test_cfg.rcnn.nms.iou_thr = iou_thr
else:
cfg.test_cfg.nms.iou_thr = iou_thr
# set cudnn_benchmark
if cfg.get("cudnn_benchmark", False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# init distributed env first, since logger depends on the dist info.
if launcher == "none":
distributed = False
else:
distributed = True
init_dist(launcher, **cfg.dist_params)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(dataset,
samples_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get("fp16", None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, checkpoint, map_location="cpu")
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if "CLASSES" in checkpoint["meta"]:
model.CLASSES = checkpoint["meta"]["CLASSES"]
else:
model.CLASSES = dataset.CLASSES
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, show, show_dir,
show_score_thr)
else:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = multi_gpu_test(model, data_loader, tmpdir, gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
if out:
print(f"\nwriting results to {out}")
mmcv.dump(outputs, out)
kwargs = {} if options is None else options
if format_only:
dataset.format_results(outputs, **kwargs)
if eval:
print_log(f"config: {config}", logger)
print_log(f"score_thr: {score_thr}", logger)
print_log(f"iou_thr: {iou_thr}", logger)
dataset.evaluate(outputs, logger=logger, **kwargs)
return outputs, dataset
def evaluate_model(model_name, paper_arxiv_id, weights_url, weights_name,
paper_results, config):
print('---')
print('Now Evaluating %s' % model_name)
evaluator = COCOEvaluator(root='./.data/vision/coco',
model_name=model_name,
paper_arxiv_id=paper_arxiv_id,
paper_results=paper_results)
out = 'results.pkl'
launcher = 'none'
if out is not None and not out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = mmcv.Config.fromfile(config)
cfg.data.test[
'ann_file'] = './.data/vision/coco/annotations/instances_val2017.json'
cfg.data.test['img_prefix'] = './.data/vision/coco/val2017/'
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# init distributed env first, since logger depends on the dist info.
if launcher == 'none':
distributed = False
else:
distributed = True
init_dist(launcher, **cfg.dist_params)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
local_checkpoint, _ = urllib.request.urlretrieve(
weights_url, '%s/.cache/torch/%s' % (str(Path.home()), weights_name))
print(local_checkpoint)
# '/home/ubuntu/GCNet/mask_rcnn_r50_fpn_1x_20181010-069fa190.pth'
checkpoint = load_checkpoint(model, local_checkpoint, map_location='cpu')
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
evaluator.reset_time()
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs, cache_exists = single_gpu_test(model, data_loader, False,
evaluator)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
if cache_exists:
print('Cache exists: %s' % (evaluator.batch_hash))
evaluator.save()
else:
from mmdetection.mmdet.core import results2json
rank, _ = get_dist_info()
if out and rank == 0:
print('\nwriting results to {}'.format(out))
mmcv.dump(outputs, out)
eval_types = ['bbox']
if eval_types:
print('Starting evaluate {}'.format(' and '.join(eval_types)))
if eval_types == ['proposal_fast']:
result_file = out
else:
if not isinstance(outputs[0], dict):
result_files = results2json(dataset, outputs, out)
else:
for name in outputs[0]:
print('\nEvaluating {}'.format(name))
outputs_ = [out[name] for out in outputs]
result_file = out + '.{}'.format(name)
result_files = results2json(
dataset, outputs_, result_file)
anns = json.load(open(result_files['bbox']))
evaluator.detections = []
evaluator.add(anns)
evaluator.save()
def main():
args = parse_args()
assert (
args.out or args.eval or args.format_only or args.show
or args.show_dir), (
'Please specify at least one operation (save/eval/format/show the '
'results / save the results) with the argument "--out", "--eval"'
', "--format-only", "--show" or "--show-dir".')
if args.eval and args.format_only:
raise ValueError('--eval and --format_only cannot be both specified.')
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = Config.fromfile(args.config)
if args.cfg_options is not None:
cfg.merge_from_dict(args.cfg_options)
# import modules from string list.
if cfg.get('custom_imports', None):
from mmcv.utils import import_modules_from_strings
import_modules_from_strings(**cfg['custom_imports'])
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
if cfg.model.get('neck'):
if isinstance(cfg.model.neck, list):
for neck_cfg in cfg.model.neck:
if neck_cfg.get('rfp_backbone'):
if neck_cfg.rfp_backbone.get('pretrained'):
neck_cfg.rfp_backbone.pretrained = None
elif cfg.model.neck.get('rfp_backbone'):
if cfg.model.neck.rfp_backbone.get('pretrained'):
cfg.model.neck.rfp_backbone.pretrained = None
# in case the test dataset is concatenated
samples_per_gpu = 1
if isinstance(cfg.data.test, dict):
samples_per_gpu = (cfg.data.get('test_dataloader', {})).get(
'samples_per_gpu', cfg.data.get('samples_per_gpu', 1))
if samples_per_gpu > 1:
# Support batch_size > 1 in test for text recognition
# by disable MultiRotateAugOCR since it is useless for most case
cfg = disable_text_recog_aug_test(cfg)
if cfg.data.test.get('pipeline', None) is not None:
# Replace 'ImageToTensor' to 'DefaultFormatBundle'
cfg.data.test.pipeline = replace_ImageToTensor(
cfg.data.test.pipeline)
elif isinstance(cfg.data.test, list):
for ds_cfg in cfg.data.test:
ds_cfg.test_mode = True
samples_per_gpu = max(
[ds_cfg.pop('samples_per_gpu', 1) for ds_cfg in cfg.data.test])
if samples_per_gpu > 1:
for ds_cfg in cfg.data.test:
ds_cfg.pipeline = replace_ImageToTensor(ds_cfg.pipeline)
# 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)
# build the dataloader
dataset = build_dataset(cfg.data.test, dict(test_mode=True))
# step 1: give default values and override (if exist) from cfg.data
loader_cfg = {
**dict(seed=cfg.get('seed'), drop_last=False, dist=distributed),
**({} if torch.__version__ != 'parrots' else dict(
prefetch_num=2,
pin_memory=False,
)),
**dict((k, cfg.data[k]) for k in [
'workers_per_gpu',
'seed',
'prefetch_num',
'pin_memory',
] if k in cfg.data)
}
test_loader_cfg = {
**loader_cfg,
**dict(shuffle=False, drop_last=False),
**cfg.data.get('test_dataloader', {}),
**dict(samples_per_gpu=samples_per_gpu)
}
data_loader = build_dataloader(dataset, **test_loader_cfg)
# build the model and load checkpoint
cfg.model.train_cfg = None
model = build_detector(cfg.model, test_cfg=cfg.get('test_cfg'))
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
load_checkpoint(model, args.checkpoint, map_location='cpu')
if args.fuse_conv_bn:
model = fuse_conv_bn(model)
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show, args.show_dir,
args.show_score_thr)
else:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = multi_gpu_test(model, data_loader, args.tmpdir,
args.gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
if args.out:
print(f'\nwriting results to {args.out}')
mmcv.dump(outputs, args.out)
kwargs = {} if args.eval_options is None else args.eval_options
if args.format_only:
dataset.format_results(outputs, **kwargs)
if args.eval:
eval_kwargs = cfg.get('evaluation', {}).copy()
# hard-code way to remove EvalHook args
for key in [
'interval', 'tmpdir', 'start', 'gpu_collect', 'save_best',
'rule'
]:
eval_kwargs.pop(key, None)
eval_kwargs.update(dict(metric=args.eval, **kwargs))
print(dataset.evaluate(outputs, **eval_kwargs))
def main():
args = parse_args()
# assert args.show or args.json_out, \
# ('Please specify at least one operation (save or show the results) '
# 'with the argument "--out" or "--show" or "--json_out"')
if args.json_out is not None and args.json_out.endswith('.json'):
args.json_out = args.json_out[:-5]
cfg = mmcv.Config.fromfile(args.config)
checkpoint_file = args.checkpoint
if not checkpoint_file:
def _epoch_num(name):
return int(
re.findall('epoch_[0-9]*.pth',
name)[0].replace('epoch_', '').replace('.pth', ''))
pths = sorted(glob.glob(os.path.join(cfg.work_dir, 'epoch_*.pth')),
key=_epoch_num)
if len(pths) > 0:
print("Found {}, use it as checkpoint by default.".format(
pths[-1]))
checkpoint_file = pths[-1]
if not checkpoint_file:
raise ValueError("Checkpoints not found, check work_dir non empty.")
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# 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)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(
dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=args.shuffle) # TODO: hack shuffle True
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, checkpoint_file, map_location='cpu')
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
num_evals = args.num_evals
if num_evals < 0:
num_evals = len(data_loader)
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, num_evals, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, num_evals, args.tmpdir)
rank, _ = get_dist_info()
if rank == 0:
gt_bboxes, gt_labels, gt_ignore, dataset_name = get_seq_gts(dataset)
print('\nStarting evaluate {}'.format(dataset_name))
eval_map(outputs,
gt_bboxes,
gt_labels,
gt_ignore,
scale_ranges=None,
iou_thr=0.5,
dataset=dataset_name,
print_summary=True)
# Save predictions in the COCO json format
if args.json_out and rank == 0:
if not isinstance(outputs[0], dict):
results2json(dataset, outputs, args.json_out)
else:
for name in outputs[0]:
outputs_ = [out[name] for out in outputs]
result_file = args.json_out + '.{}'.format(name)
results2json(dataset, outputs_, result_file)
def train_model(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
"""Train model entry function.
Args:
model (nn.Module): The model to be trained.
dataset (:obj:`Dataset`): Train dataset.
cfg (dict): The config dict for training.
distributed (bool): Whether to use distributed training.
Default: False.
validate (bool): Whether to do evaluation. Default: False.
timestamp (str | None): Local time for runner. Default: None.
meta (dict | None): Meta dict to record some important information.
Default: None
"""
logger = get_root_logger(cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
if 'imgs_per_gpu' in cfg.data:
logger.warning('"imgs_per_gpu" is deprecated in MMDet V2.0. '
'Please use "samples_per_gpu" instead')
if 'samples_per_gpu' in cfg.data:
logger.warning(
f'Got "imgs_per_gpu"={cfg.data.imgs_per_gpu} and '
f'"samples_per_gpu"={cfg.data.samples_per_gpu}, "imgs_per_gpu"'
f'={cfg.data.imgs_per_gpu} is used in this experiments')
else:
logger.warning(
'Automatically set "samples_per_gpu"="imgs_per_gpu"='
f'{cfg.data.imgs_per_gpu} in this experiments')
cfg.data.samples_per_gpu = cfg.data.imgs_per_gpu
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
len(cfg.gpu_ids),
samples_per_epoch=cfg.data.get('samples_per_epoch', None),
dist=distributed,
seed=cfg.seed,
persistent_workers=cfg.data.get('persistent_workers', False))
for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
if find_unused_parameters:
logger.info('set find_unused_parameters = True in DDP')
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model, device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = EpochBasedRunner(model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# an ugly workaround to make .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
optimizer_config = cfg.optimizer_config
if 'type' not in cfg.optimizer_config:
optimizer_config.type = 'Fp16OptimizerHook' \
if fp16_cfg else 'OptimizerHook'
if fp16_cfg:
optimizer_config.update(fp16_cfg)
if 'Fp16' in optimizer_config.type:
optimizer_config.update(distributed=distributed)
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_dataloader = build_dataloader(
val_dataset,
samples_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False,
persistent_workers=cfg.data.get('persistent_workers', False))
eval_cfg = cfg.get('evaluation', {})
eval_hook = DistEvalHook if distributed else EvalHook
runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
# user-defined hooks
if cfg.get('custom_hooks', None):
custom_hooks = cfg.custom_hooks
assert isinstance(custom_hooks, list), \
f'custom_hooks expect list type, but got {type(custom_hooks)}'
for hook_cfg in cfg.custom_hooks:
assert isinstance(hook_cfg, dict), \
'Each item in custom_hooks expects dict type, but got ' \
f'{type(hook_cfg)}'
hook_cfg = hook_cfg.copy()
priority = hook_cfg.pop('priority', 'NORMAL')
hook = build_from_cfg(hook_cfg, HOOKS)
runner.register_hook(hook, priority=priority)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def main():
args = parse_args()
cfg = mmcv.Config.fromfile(args.config)
# Load output_config from cfg
output_config = cfg.get('output_config', {})
# Overwrite output_config from args.out
output_config = merge_configs(output_config, dict(out=args.out))
# Load eval_config from cfg
eval_config = cfg.get('eval_config', {})
# Overwrite eval_config from args.eval
eval_config = merge_configs(eval_config, dict(metrics=args.eval))
# Add options from args.option
eval_config = merge_configs(eval_config, args.options)
assert output_config or eval_config, \
('Please specify at least one operation (save or eval the '
'results) with the argument "--out" or "--eval"')
# set cudnn benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.data.test.test_mode = True
if cfg.test_cfg is None:
cfg.test_cfg = dict(average_clips=args.average_clips)
else:
# You can set average_clips during testing, it will override the
# original settting
if args.average_clips is not None:
cfg.test_cfg.average_clips = args.average_clips
# 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)
# create work_dir
mmcv.mkdir_or_exist(osp.abspath(cfg.work_dir))
# build the dataloader
dataset = build_dataset(cfg.data.test, dict(test_mode=True))
dataloader_setting = dict(videos_per_gpu=cfg.data.get('videos_per_gpu', 2),
workers_per_gpu=cfg.data.get(
'workers_per_gpu', 0),
dist=distributed,
shuffle=False)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('test_dataloader', {}))
data_loader = build_dataloader(dataset, **dataloader_setting)
# build the model and load checkpoint
model = build_model(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
load_checkpoint(model, args.checkpoint, map_location='cpu')
if args.fuse_conv_bn:
model = fuse_conv_bn(model)
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader)
else:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = multi_gpu_test(model, data_loader, args.tmpdir,
args.gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
if output_config:
out = output_config['out']
print(f'\nwriting results to {out}')
dataset.dump_results(outputs, **output_config)
if eval_config:
eval_res = dataset.evaluate(outputs, **eval_config)
for name, val in eval_res.items():
print(f'{name}: {val:.04f}')
def train_detector(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
logger = get_root_logger(log_level=cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
if 'imgs_per_gpu' in cfg.data:
logger.warning('"imgs_per_gpu" is deprecated in MMDet V2.0. '
'Please use "samples_per_gpu" instead')
if 'samples_per_gpu' in cfg.data:
logger.warning(
f'Got "imgs_per_gpu"={cfg.data.imgs_per_gpu} and '
f'"samples_per_gpu"={cfg.data.samples_per_gpu}, "imgs_per_gpu"'
f'={cfg.data.imgs_per_gpu} is used in this experiments')
else:
logger.warning(
'Automatically set "samples_per_gpu"="imgs_per_gpu"='
f'{cfg.data.imgs_per_gpu} in this experiments')
cfg.data.samples_per_gpu = cfg.data.imgs_per_gpu
runner_type = 'EpochBasedRunner' if 'runner' not in cfg else cfg.runner[
'type']
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# `num_gpus` will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed,
runner_type=runner_type) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(
model.cuda(cfg.gpu_ids[0]), device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
if 'runner' not in cfg:
cfg.runner = {
'type': 'EpochBasedRunner',
'max_epochs': cfg.total_epochs
}
warnings.warn(
'config is now expected to have a `runner` section, '
'please set `runner` in your config.', UserWarning)
else:
if 'total_epochs' in cfg:
assert cfg.total_epochs == cfg.runner.max_epochs
runner = build_runner(
cfg.runner,
default_args=dict(
model=model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
# an ugly workaround to make .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(
**cfg.optimizer_config, **fp16_cfg, distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = OptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
if isinstance(runner, EpochBasedRunner):
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
# Support batch_size > 1 in validation
val_samples_per_gpu = cfg.data.val.pop('samples_per_gpu', 1)
if val_samples_per_gpu > 1:
# Replace 'ImageToTensor' to 'DefaultFormatBundle'
cfg.data.val.pipeline = replace_ImageToTensor(
cfg.data.val.pipeline)
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_dataloader = build_dataloader(
val_dataset,
samples_per_gpu=val_samples_per_gpu,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
eval_cfg = cfg.get('evaluation', {})
eval_cfg['by_epoch'] = cfg.runner['type'] != 'IterBasedRunner'
eval_hook = DistEvalHook if distributed else EvalHook
# In this PR (https://github.com/open-mmlab/mmcv/pull/1193), the
# priority of IterTimerHook has been modified from 'NORMAL' to 'LOW'.
runner.register_hook(
eval_hook(val_dataloader, **eval_cfg), priority='LOW')
# user-defined hooks
if cfg.get('custom_hooks', None):
custom_hooks = cfg.custom_hooks
assert isinstance(custom_hooks, list), \
f'custom_hooks expect list type, but got {type(custom_hooks)}'
for hook_cfg in cfg.custom_hooks:
assert isinstance(hook_cfg, dict), \
'Each item in custom_hooks expects dict type, but got ' \
f'{type(hook_cfg)}'
hook_cfg = hook_cfg.copy()
priority = hook_cfg.pop('priority', 'NORMAL')
hook = build_from_cfg(hook_cfg, HOOKS)
runner.register_hook(hook, priority=priority)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow)
def main():
args = parse_args()
assert args.out or args.show or args.json_out, \
('Please specify at least one operation (save or show the results) '
'with the argument "--out" or "--show" or "--json_out"')
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
if args.json_out is not None and args.json_out.endswith('.json'):
args.json_out = args.json_out[:-5]
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# 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)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(
dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, args.checkpoint, map_location='cpu')
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
rank, _ = get_dist_info()
if args.out and rank == 0:
print('\nwriting results to {}'.format(args.out))
mmcv.dump(outputs, args.out)
eval_types = args.eval
if eval_types:
print('Starting evaluate {}'.format(' and '.join(eval_types)))
if eval_types == ['proposal_fast']:
result_file = args.out
#coco_eval(result_file, eval_types, dataset.coco)
voc_eval(args.out, dataset)
else:
if not isinstance(outputs[0], dict):
result_files = results2json(dataset, outputs, args.out)
#coco_eval(result_files, eval_types, dataset.coco)
voc_eval(args.out, dataset)
else:
for name in outputs[0]:
print('\nEvaluating {}'.format(name))
outputs_ = [out[name] for out in outputs]
result_file = args.out + '.{}'.format(name)
result_files = results2json(dataset, outputs_,
result_file)
#coco_eval(result_files, eval_types, dataset.coco)
voc_eval(args.out, dataset)
# Save predictions in the COCO json format
if args.json_out and rank == 0:
if not isinstance(outputs[0], dict):
results2json(dataset, outputs, args.json_out)
else:
for name in outputs[0]:
outputs_ = [out[name] for out in outputs]
result_file = args.json_out + '.{}'.format(name)
results2json(dataset, outputs_, result_file)
def main():
args = parse_args()
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
for t in cfg.data.test:
t.test_mode = True
cfg.out_path = args.out.split('.pkl')[0] if args.out else None
# 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)
# build the model and load checkpoint
model = build_detector(
cfg.model, train_cfg=None, test_cfg=cfg.test_cfg, global_cfg=cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, args.checkpoint, map_location='cpu')
if args.det_ckpt is not None:
print('Loading detection models...')
det_ckpt = load_checkpoint(model, args.det_ckpt, map_location='cpu')
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
if not type(cfg.data.test) == list:
cfg.data.test = [cfg.data.test]
if not distributed:
model = MMDataParallel(model, device_ids=[0])
else:
model = MMDistributedDataParallel(model.cuda())
outputs = dict()
for c in cfg.data.test:
dataset = build_dataset(c)
data_loader = build_dataloader(
dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
if not distributed:
results = single_gpu_test(model, data_loader, args.out, args.show)
if results is not None:
outputs.update(results)
else:
outputs.update(multi_gpu_test(model, data_loader, args.tmpdir))
rank, _ = get_dist_info()
if len(outputs.keys()) > 0 and args.out and rank == 0:
print('\nwriting results to {}'.format(args.out))
if not (args.out.endswith('.pkl') or args.out.endswith('.json')):
args.out += '.pkl'
if 'track_results' in outputs.keys():
mmcv.dump(outputs['track_results'], args.out)
else:
mmcv.dump(outputs, args.out)
if 'bbox_results' in outputs.keys():
result_files = results2json(dataset, outputs['bbox_results'], args.out)
coco_eval(result_files, ['bbox', 'segm'], cfg.data.test[0].ann_file)
if 'segm_results' in outputs.keys():
result_files = results2json(dataset, [(b, s) for b, s in zip(outputs['bbox_results'], outputs['segm_results'])], args.out)
coco_eval(result_files, ['segm'], cfg.data.test[0].ann_file)
# if 'new_bbox_results' in outputs.keys():
# # For tracking
# result_files = results2json(dataset, outputs['new_bbox_results'],
# args.out)
# coco_eval(result_files, ['bbox'], cfg.data.test[0].ann_file)
if 'track_results' in outputs.keys():
print("Evaluating box tracking...")
mdat_eval(outputs['track_results'], dataset, args.out, cfg)
if 'segm_track_results' in outputs.keys():
print("Evaluating segmentation tracking...")
mdat_eval(outputs['segm_track_results'], dataset, args.out, cfg, with_mask=True)
def train_detector(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
logger = get_root_logger(cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
data_loaders = [
build_dataloader(
ds,
cfg.data.imgs_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = Runner(
model,
batch_processor,
optimizer,
cfg.work_dir,
logger=logger,
)
# an ugly walkaround to make the .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(**cfg.optimizer_config,
**fp16_cfg,
distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = DistOptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config)
if distributed:
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_dataloader = build_dataloader(
val_dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
eval_cfg = cfg.get('evaluation', {})
eval_hook = DistEvalHook if distributed else EvalHook
runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def main():
args = parse_args()
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
if args.ann_file is not None:
cfg.data.test.ann_file = args.ann_file
if args.img_prefix is not None:
cfg.data.test.img_prefix = args.img_prefix
if args.flip:
cfg.data.test.flip_ratio = 1
# 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)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
dataset = get_dataset(cfg.data.test)
data_loader = build_dataloader(dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
load_checkpoint(model, args.checkpoint, map_location='cpu')
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
rank, _ = get_dist_info()
if args.out and rank == 0:
print('\nwriting results to {}'.format(args.out))
mmcv.dump(outputs, args.out)
eval_types = args.eval
if eval_types:
print('Starting evaluate {}'.format(' and '.join(eval_types)))
if eval_types == ['proposal_fast']:
result_file = args.out
coco_eval(result_file, eval_types, dataset.coco)
else:
if not isinstance(outputs[0], dict):
result_file = args.out + '.json'
results2json(dataset, outputs, result_file)
coco_eval(result_file, eval_types, dataset.coco)
else:
for name in outputs[0]:
print('\nEvaluating {}'.format(name))
outputs_ = [out[name] for out in outputs]
result_file = args.out + '.{}.json'.format(name)
results2json(dataset, outputs_, result_file)
coco_eval(result_file, eval_types, dataset.coco)
def main():
args = parse_args()
assert args.out or args.eval or args.format_only or args.show, \
('Please specify at least one operation (save/eval/format/show the '
'results) with the argument "--out", "--eval", "--format_only" '
'or "--show"')
if args.eval and args.format_only:
raise ValueError('--eval and --format_only cannot be both specified')
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# 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)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(
dataset,
imgs_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, args.checkpoint, map_location='cpu')
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir,
args.gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
if args.out:
print('\nwriting results to {}'.format(args.out))
mmcv.dump(outputs, args.out)
kwargs = {} if args.options is None else args.options
if args.format_only:
dataset.format_results(outputs, **kwargs)
if args.eval:
dataset.evaluate(outputs, args.eval, **kwargs)
def _dist_train(model,
dataset,
cfg,
validate=False,
logger=None,
timestamp=None):
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
data_loaders = [
build_dataloader(ds,
cfg.data.imgs_per_gpu,
cfg.data.workers_per_gpu,
dist=True) for ds in dataset
]
# put model on gpus
model = MMDistributedDataParallel(model.cuda())
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
runner = Runner(model,
batch_processor,
optimizer,
cfg.work_dir,
logger=logger)
# an ugly walkaround to make the .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(**cfg.optimizer_config,
**fp16_cfg)
else:
optimizer_config = DistOptimizerHook(**cfg.optimizer_config)
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config)
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_dataset_cfg = cfg.data.val
eval_cfg = cfg.get('evaluation', {})
if isinstance(model.module, RPN):
# TODO: implement recall hooks for other datasets
runner.register_hook(
CocoDistEvalRecallHook(val_dataset_cfg, **eval_cfg))
else:
dataset_type = DATASETS.get(val_dataset_cfg.type)
if issubclass(dataset_type, datasets.CocoDataset):
runner.register_hook(
CocoDistEvalmAPHook(val_dataset_cfg, **eval_cfg))
else:
runner.register_hook(
DistEvalmAPHook(val_dataset_cfg, **eval_cfg))
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs)
def train_segmentor(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
"""Launch segmentor training."""
logger = get_root_logger(cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed,
drop_last=True) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
# NEW: support paramwise_cfg (untested)
try:
if cfg.paramwise_cfg is not None:
cfg.optimizer.paramwise_cfg = cfg.paramwise_cfg
except:
pass
# END NEW
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
if cfg.get('runner') is None:
cfg.runner = {'type': 'IterBasedRunner', 'max_iters': cfg.total_iters}
warnings.warn(
'config is now expected to have a `runner` section, '
'please set `runner` in your config.', UserWarning)
runner = build_runner(cfg.runner,
default_args=dict(model=model,
batch_processor=None,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
# register hooks
runner.register_training_hooks(cfg.lr_config, cfg.optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
# an ugly walkaround to make the .log and .log.json filenames the same
runner.timestamp = timestamp
# register eval hooks
if validate:
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_dataloader = build_dataloader(
val_dataset,
samples_per_gpu=1,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
eval_cfg = cfg.get('evaluation', {})
eval_cfg['by_epoch'] = cfg.runner['type'] != 'IterBasedRunner'
eval_hook = DistEvalHook if distributed else EvalHook
runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner.run(data_loaders, cfg.workflow)
def train_model(model,
dataset,
cfg,
distributed=False,
validate=False,
test=dict(test_best=False, test_last=False),
timestamp=None,
meta=None):
"""Train model entry function.
Args:
model (nn.Module): The model to be trained.
dataset (:obj:`Dataset`): Train dataset.
cfg (dict): The config dict for training.
distributed (bool): Whether to use distributed training.
Default: False.
validate (bool): Whether to do evaluation. Default: False.
test (dict): The testing option, with two keys: test_last & test_best.
The value is True or False, indicating whether to test the
corresponding checkpoint.
Default: dict(test_best=False, test_last=False).
timestamp (str | None): Local time for runner. Default: None.
meta (dict | None): Meta dict to record some important information.
Default: None
"""
logger = get_root_logger(log_level=cfg.log_level)
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=cfg.data.get('workers_per_gpu', 1),
persistent_workers=cfg.data.get('persistent_workers', False),
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('train_dataloader', {}))
if cfg.omnisource:
# The option can override videos_per_gpu
train_ratio = cfg.data.get('train_ratio', [1] * len(dataset))
omni_videos_per_gpu = cfg.data.get('omni_videos_per_gpu', None)
if omni_videos_per_gpu is None:
dataloader_settings = [dataloader_setting] * len(dataset)
else:
dataloader_settings = []
for videos_per_gpu in omni_videos_per_gpu:
this_setting = cp.deepcopy(dataloader_setting)
this_setting['videos_per_gpu'] = videos_per_gpu
dataloader_settings.append(this_setting)
data_loaders = [
build_dataloader(ds, **setting)
for ds, setting in zip(dataset, dataloader_settings)
]
else:
data_loaders = [
build_dataloader(ds, **dataloader_setting) for ds in dataset
]
# put model on gpus
if distributed:
find_unused_parameters = cfg.get('find_unused_parameters', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = MMDataParallel(model, device_ids=cfg.gpu_ids)
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
Runner = OmniSourceRunner if cfg.omnisource else EpochBasedRunner
runner = Runner(model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# an ugly workaround to make .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
optimizer_config = Fp16OptimizerHook(**cfg.optimizer_config,
**fp16_cfg,
distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = OptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# register hooks
runner.register_training_hooks(cfg.lr_config,
optimizer_config,
cfg.checkpoint_config,
cfg.log_config,
cfg.get('momentum_config', None),
custom_hooks_config=cfg.get(
'custom_hooks', None))
# multigrid setting
multigrid_cfg = cfg.get('multigrid', None)
if multigrid_cfg is not None:
from mmaction.utils.multigrid import LongShortCycleHook
multigrid_scheduler = LongShortCycleHook(cfg)
runner.register_hook(multigrid_scheduler)
logger.info('Finish register multigrid hook')
# subbn3d aggregation is HIGH, as it should be done before
# saving and evaluation
from mmaction.utils.multigrid import SubBatchNorm3dAggregationHook
subbn3d_aggre_hook = SubBatchNorm3dAggregationHook()
runner.register_hook(subbn3d_aggre_hook, priority='VERY_HIGH')
logger.info('Finish register subbn3daggre hook')
# precise bn setting
if cfg.get('precise_bn', False):
precise_bn_dataset = build_dataset(cfg.data.train)
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=1, # save memory and time
persistent_workers=cfg.data.get('persistent_workers', False),
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed)
data_loader_precise_bn = build_dataloader(precise_bn_dataset,
**dataloader_setting)
precise_bn_hook = PreciseBNHook(data_loader_precise_bn,
**cfg.get('precise_bn'))
runner.register_hook(precise_bn_hook, priority='HIGHEST')
logger.info('Finish register precisebn hook')
if distributed:
if cfg.omnisource:
runner.register_hook(OmniSourceDistSamplerSeedHook())
else:
runner.register_hook(DistSamplerSeedHook())
if validate:
eval_cfg = cfg.get('evaluation', {})
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=cfg.data.get('workers_per_gpu', 1),
persistent_workers=cfg.data.get('persistent_workers', False),
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
shuffle=False)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('val_dataloader', {}))
val_dataloader = build_dataloader(val_dataset, **dataloader_setting)
eval_hook = DistEvalHook(val_dataloader, **eval_cfg) if distributed \
else EvalHook(val_dataloader, **eval_cfg)
runner.register_hook(eval_hook)
if cfg.resume_from:
runner.resume(cfg.resume_from)
elif cfg.load_from:
runner.load_checkpoint(cfg.load_from)
runner_kwargs = dict()
if cfg.omnisource:
runner_kwargs = dict(train_ratio=train_ratio)
runner.run(data_loaders, cfg.workflow, cfg.total_epochs, **runner_kwargs)
if distributed:
dist.barrier()
time.sleep(5)
if test['test_last'] or test['test_best']:
best_ckpt_path = None
if test['test_best']:
ckpt_paths = [x for x in os.listdir(cfg.work_dir) if 'best' in x]
ckpt_paths = [x for x in ckpt_paths if x.endswith('.pth')]
if len(ckpt_paths) == 0:
runner.logger.info('Warning: test_best set, but no ckpt found')
test['test_best'] = False
if not test['test_last']:
return
elif len(ckpt_paths) > 1:
epoch_ids = [
int(x.split('epoch_')[-1][:-4]) for x in ckpt_paths
]
best_ckpt_path = ckpt_paths[np.argmax(epoch_ids)]
else:
best_ckpt_path = ckpt_paths[0]
if best_ckpt_path:
best_ckpt_path = osp.join(cfg.work_dir, best_ckpt_path)
test_dataset = build_dataset(cfg.data.test, dict(test_mode=True))
gpu_collect = cfg.get('evaluation', {}).get('gpu_collect', False)
tmpdir = cfg.get('evaluation', {}).get('tmpdir',
osp.join(cfg.work_dir, 'tmp'))
dataloader_setting = dict(
videos_per_gpu=cfg.data.get('videos_per_gpu', 1),
workers_per_gpu=cfg.data.get('workers_per_gpu', 1),
persistent_workers=cfg.data.get('persistent_workers', False),
num_gpus=len(cfg.gpu_ids),
dist=distributed,
shuffle=False)
dataloader_setting = dict(dataloader_setting,
**cfg.data.get('test_dataloader', {}))
test_dataloader = build_dataloader(test_dataset, **dataloader_setting)
names, ckpts = [], []
if test['test_last']:
names.append('last')
ckpts.append(None)
if test['test_best'] and best_ckpt_path is not None:
names.append('best')
ckpts.append(best_ckpt_path)
for name, ckpt in zip(names, ckpts):
if ckpt is not None:
runner.load_checkpoint(ckpt)
outputs = multi_gpu_test(runner.model, test_dataloader, tmpdir,
gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
out = osp.join(cfg.work_dir, f'{name}_pred.pkl')
test_dataset.dump_results(outputs, out)
eval_cfg = cfg.get('evaluation', {})
for key in [
'interval', 'tmpdir', 'start', 'gpu_collect',
'save_best', 'rule', 'by_epoch', 'broadcast_bn_buffers'
]:
eval_cfg.pop(key, None)
eval_res = test_dataset.evaluate(outputs, **eval_cfg)
runner.logger.info(f'Testing results of the {name} checkpoint')
for metric_name, val in eval_res.items():
runner.logger.info(f'{metric_name}: {val:.04f}')
def main():
args = parse_args()
assert args.out or args.eval or args.format_only or args.show \
or args.show_dir, \
('Please specify at least one operation (save/eval/format/show the '
'results / save the results) with the argument "--out", "--eval"'
', "--format-only", "--show" or "--show-dir"')
if args.eval and args.format_only:
raise ValueError('--eval and --format_only cannot be both specified')
if args.out is not None and not args.out.endswith(('.pkl', '.pickle')):
raise ValueError('The output file must be a pkl file.')
cfg = Config.fromfile(args.config)
if args.cfg_options is not None:
cfg.merge_from_dict(args.cfg_options)
# import modules from string list.
if cfg.get('custom_imports', None):
from mmcv.utils import import_modules_from_strings
import_modules_from_strings(**cfg['custom_imports'])
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
if cfg.model.get('neck'):
if isinstance(cfg.model.neck, list):
for neck_cfg in cfg.model.neck:
if neck_cfg.get('rfp_backbone'):
if neck_cfg.rfp_backbone.get('pretrained'):
neck_cfg.rfp_backbone.pretrained = None
elif cfg.model.neck.get('rfp_backbone'):
if cfg.model.neck.rfp_backbone.get('pretrained'):
cfg.model.neck.rfp_backbone.pretrained = None
# in case the test dataset is concatenated
samples_per_gpu = 1
if isinstance(cfg.data.test, dict):
cfg.data.test.test_mode = True
samples_per_gpu = cfg.data.test.pop('samples_per_gpu', 1)
if samples_per_gpu > 1:
# Replace 'ImageToTensor' to 'DefaultFormatBundle'
cfg.data.test.pipeline = replace_ImageToTensor(
cfg.data.test.pipeline)
elif isinstance(cfg.data.test, list):
for ds_cfg in cfg.data.test:
ds_cfg.test_mode = True
samples_per_gpu = max(
[ds_cfg.pop('samples_per_gpu', 1) for ds_cfg in cfg.data.test])
if samples_per_gpu > 1:
for ds_cfg in cfg.data.test:
ds_cfg.pipeline = replace_ImageToTensor(ds_cfg.pipeline)
# 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)
# build the dataloader
dataset = build_dataset(cfg.data.test)
data_loader = build_dataloader(
dataset,
samples_per_gpu=samples_per_gpu,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
cfg.model.train_cfg = None
model = build_detector(cfg.model, test_cfg=cfg.get('test_cfg'))
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, args.checkpoint, map_location='cpu')
if args.fuse_conv_bn:
model = fuse_conv_bn(model)
# old versions did not save class info in checkpoints, this walkaround is
# for backward compatibility
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = dataset.CLASSES
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, args.show, args.show_dir,
args.show_score_thr)
else:
model = MMDistributedDataParallel(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False)
outputs = multi_gpu_test(model, data_loader, args.tmpdir,
args.gpu_collect)
rank, _ = get_dist_info()
if rank == 0:
if args.out:
print(f'\nwriting results to {args.out}')
mmcv.dump(outputs, args.out)
kwargs = {} if args.eval_options is None else args.eval_options
if args.format_only:
dataset.format_results(outputs, **kwargs)
if args.eval:
eval_kwargs = cfg.get('evaluation', {}).copy()
# hard-code way to remove EvalHook args
for key in [
'interval', 'tmpdir', 'start', 'gpu_collect', 'save_best',
'rule'
]:
eval_kwargs.pop(key, None)
eval_kwargs.update(dict(metric=args.eval, **kwargs))
print(dataset.evaluate(outputs, **eval_kwargs))
def main():
args = parse_args()
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpuid
img_dir = args.img_dir
out_dir = args.out_dir
batch_size = args.batch_size
cfg = mmcv.Config.fromfile(args.config)
# set cudnn_benchmark
if cfg.get('cudnn_benchmark', False):
torch.backends.cudnn.benchmark = True
cfg.model.pretrained = None
cfg.data.test.test_mode = True
# 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)
# build the dataloader
# TODO: support multiple images per gpu (only minor changes are needed)
file_list = common.load_filepaths(img_dir,
suffix=('.jpg', '.png', '.jpeg'),
recursive=True)
print(file_list[:10])
print('imgs: ', len(file_list))
#
print(file_list[0].replace(img_dir, ''))
dataset = FilesDataset(file_list, cfg.test_pipeline)
data_loader = build_dataloader(dataset,
imgs_per_gpu=batch_size,
workers_per_gpu=batch_size,
dist=distributed,
shuffle=False)
# build the model and load checkpoint
model = build_detector(cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
wrap_fp16_model(model)
checkpoint = load_checkpoint(model, args.checkpoint, map_location='cpu')
model = reweight_cls(model, args.tau)
if not distributed:
model = MMDataParallel(model, device_ids=[0])
outputs = single_gpu_test(model, data_loader, False, cfg)
else:
model = MMDistributedDataParallel(model.cuda())
outputs = multi_gpu_test(model, data_loader, args.tmpdir)
print(type(outputs))
print(len(outputs))
print(len(outputs[0]))
exit()
# save outputs
for file_path, mmdet_ret in zip(file_list, outputs):
save_name = file_path.replace(img_dir, '')
save_path = os.path.join(out_dir, save_name)
common.makedirs(os.path.dirname(save_path))
save_in_tao_format(mmdet_ret, save_path)
