def setup_runner(model, cfg, optimizer, logger, meta, timestamp):
if "runner" not in cfg:
cfg.runner = {
"type": "DSEpochBasedRunner",
"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
return runner
def test_build_runner():
temp_root = tempfile.gettempdir()
dir_name = ''.join(
[random.choice(string.ascii_letters) for _ in range(10)])
default_args = dict(model=Model(),
work_dir=osp.join(temp_root, dir_name),
logger=logging.getLogger())
cfg = dict(type='EpochBasedRunner', max_epochs=1)
runner = build_runner(cfg, default_args=default_args)
assert runner._max_epochs == 1
cfg = dict(type='IterBasedRunner', max_iters=1)
runner = build_runner(cfg, default_args=default_args)
assert runner._max_iters == 1
with pytest.raises(ValueError, match='Only one of'):
cfg = dict(type='IterBasedRunner', max_epochs=1, max_iters=1)
runner = build_runner(cfg, default_args=default_args)
def _build_demo_runner(runner_type='EpochBasedRunner',
max_epochs=1,
max_iters=None,
multi_optimziers=False):
class Model(nn.Module):
def __init__(self):
super().__init__()
self.linear = nn.Linear(2, 1)
self.conv = nn.Conv2d(3, 3, 3)
def forward(self, x):
return self.linear(x)
def train_step(self, x, optimizer, **kwargs):
return dict(loss=self(x))
def val_step(self, x, optimizer, **kwargs):
return dict(loss=self(x))
model = Model()
if multi_optimziers:
optimizer = {
'model1':
torch.optim.SGD(model.linear.parameters(), lr=0.02, momentum=0.95),
'model2':
torch.optim.SGD(model.conv.parameters(), lr=0.01, momentum=0.9),
}
else:
optimizer = torch.optim.SGD(model.parameters(), lr=0.02, momentum=0.95)
log_config = dict(
interval=1, hooks=[
dict(type='TextLoggerHook'),
])
tmp_dir = tempfile.mkdtemp()
runner = build_runner(
dict(type=runner_type),
default_args=dict(
model=model,
work_dir=tmp_dir,
optimizer=optimizer,
logger=logging.getLogger(),
max_epochs=max_epochs,
max_iters=max_iters))
runner.register_checkpoint_hook(dict(interval=1))
runner.register_logger_hooks(log_config)
return runner
def _build_demo_runner(workdir,
runner_type="EpochBasedRunner",
max_epochs=1,
max_iters=None):
class Model(torch.nn.Module):
def __init__(self):
super().__init__()
self.linear = torch.nn.Linear(2, 1)
self.conv = torch.nn.Conv2d(3, 3, 3)
def forward(self, x):
return self.linear(x)
def train_step(self, x, optimizer, **kwargs):
return dict(loss=self(x))
def val_step(self, x, optimizer, **kwargs):
return dict(loss=self(x))
model = Model()
optimizer = torch.optim.SGD(model.parameters(), lr=0.02, momentum=0.95)
runner = build_runner(
dict(type=runner_type),
default_args=dict(
model=model,
work_dir=workdir,
optimizer=optimizer,
logger=logging.getLogger(),
max_epochs=max_epochs,
max_iters=max_iters,
),
)
log_config = dict(interval=1, hooks=[dict(type="TextLoggerHook")])
runner.register_logger_hooks(log_config)
return runner
def train_model(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
device='cuda',
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.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
round_up=True,
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:
if device == 'cuda':
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
elif device == 'cpu':
model = model.cpu()
else:
raise ValueError(F'unsupported device name {device}.')
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
if cfg.get('runner') is None:
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)
runner = build_runner(cfg.runner,
default_args=dict(model=model,
batch_processor=None,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
# 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,
cfg.get('momentum_config', None),
custom_hooks_config=cfg.get(
'custom_hooks', 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=cfg.data.samples_per_gpu,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False,
round_up=True)
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,
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.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)
use_ddp_wrapper = cfg.get('use_ddp_wrapper', False)
# Sets the `find_unused_parameters` parameter in
# torch.nn.parallel.DistributedDataParallel
if use_ddp_wrapper:
mmcv.print_log('Use DDP Wrapper.', 'mmgen')
model = DistributedDataParallelWrapper(
model.cuda(),
device_ids=[torch.cuda.current_device()],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
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
if cfg.optimizer:
optimizer = build_optimizers(model, cfg.optimizer)
# In GANs, we allow building optimizer in GAN model.
else:
optimizer = None
# allow users to define the runner
if cfg.get('runner', None):
runner = build_runner(
cfg.runner,
dict(model=model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
else:
runner = IterBasedRunner(model,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta)
# set if use dynamic ddp in training
# is_dynamic_ddp=cfg.get('is_dynamic_ddp', False))
# an ugly walkaround to make the .log and .log.json filenames the same
runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
# In GANs, we can directly optimize parameter in `train_step` function.
if cfg.get('optimizer_cfg', None) is None:
optimizer_config = None
elif fp16_cfg is not None:
raise NotImplementedError('Fp16 has not been supported.')
# optimizer_config = Fp16OptimizerHook(
# **cfg.optimizer_config, **fp16_cfg, distributed=distributed)
# default to use OptimizerHook
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = OptimizerHook(**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
# update `out_dir` in ckpt hook
if cfg.checkpoint_config is not None:
cfg.checkpoint_config['out_dir'] = os.path.join(
cfg.work_dir, cfg.checkpoint_config.get('out_dir', 'ckpt'))
# register hooks
runner.register_training_hooks(cfg.lr_config, optimizer_config,
cfg.checkpoint_config, cfg.log_config,
cfg.get('momentum_config', None))
# # DistSamplerSeedHook should be used with EpochBasedRunner
# if distributed:
# runner.register_hook(DistSamplerSeedHook())
# In general, we do NOT adopt standard evaluation hook in GAN training.
# Thus, if you want a eval hook, you need further define the key of
# 'evaluation' in the config.
# register eval hooks
if validate and cfg.get('evaluation', None) is not None:
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
# Support batch_size > 1 in validation
val_loader_cfg = {
'samples_per_gpu': 1,
'shuffle': False,
'workers_per_gpu': cfg.data.workers_per_gpu,
**cfg.data.get('val_data_loader', {})
}
val_dataloader = build_dataloader(val_dataset,
dist=distributed,
**val_loader_cfg)
eval_cfg = deepcopy(cfg.get('evaluation'))
eval_cfg.update(dict(dist=distributed, dataloader=val_dataloader))
eval_hook = build_from_cfg(eval_cfg, HOOKS)
priority = eval_cfg.pop('priority', 'NORMAL')
runner.register_hook(eval_hook, priority=priority)
# 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_iters)
def train_detector(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
meta=None):
cfg = compat_cfg(cfg)
logger = get_root_logger(log_level=cfg.log_level)
use_apex = cfg.optimizer_config.get('type', None) == 'ApexOptimizerHook'
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
runner_type = 'EpochBasedRunner' if 'runner' not in cfg else cfg.runner[
'type']
train_dataloader_default_args = dict(
samples_per_gpu=2,
workers_per_gpu=2,
# `num_gpus` will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed,
runner_type=runner_type,
persistent_workers=False)
train_loader_cfg = {
**train_dataloader_default_args,
**cfg.data.get('train_dataloader', {})
}
data_loaders = [build_dataloader(ds, **train_loader_cfg) for ds in dataset]
auto_scale_lr(cfg, distributed, logger)
# use apex fp16 optimizer
if use_apex:
if apex is None:
raise RuntimeError('apex is not installed')
optimizer = build_optimizer(model, cfg.optimizer)
if cfg.optimizer_config.get('use_fp16', False):
model, optimizer = apex.amp.initialize(model.cuda(),
optimizer,
opt_level='O1')
for m in model.modules():
if hasattr(m, 'fp16_enabled'):
m.fp16_enabled = True
# 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 = build_ddp(model,
cfg.device,
device_ids=[int(os.environ['LOCAL_RANK'])],
broadcast_buffers=False,
find_unused_parameters=find_unused_parameters)
else:
model = build_dp(model, cfg.device, device_ids=cfg.gpu_ids)
# build optimizer
if not use_apex:
optimizer = build_optimizer(model, cfg.optimizer)
# build runner
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)
# gradient accumulation
if 'cumulative_iters' in cfg.optimizer_config:
if fp16_cfg is not None:
optimizer_config = GradientCumulativeFp16OptimizerHook(
**cfg.optimizer_config, **fp16_cfg, distributed=distributed)
elif distributed and 'type' not in cfg.optimizer_config:
optimizer_config = DebugGradientCumulativeOptimizerHook(
**cfg.optimizer_config)
else:
optimizer_config = cfg.optimizer_config
else:
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))
if distributed:
if isinstance(runner, EpochBasedRunner):
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_dataloader_default_args = dict(samples_per_gpu=1,
workers_per_gpu=2,
dist=distributed,
shuffle=False,
persistent_workers=False)
val_dataloader_args = {
**val_dataloader_default_args,
**cfg.data.get('val_dataloader', {})
}
# Support batch_size > 1 in validation
if val_dataloader_args['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, **val_dataloader_args)
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')
resume_from = None
if cfg.resume_from is None and cfg.get('auto_resume'):
resume_from = find_latest_checkpoint(cfg.work_dir)
if resume_from is not None:
cfg.resume_from = resume_from
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,
timestamp=None,
device=None,
meta=None):
logger = get_root_logger()
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
sampler_cfg = cfg.data.get('sampler', None)
data_loaders = [
build_dataloader(
ds,
cfg.data.samples_per_gpu,
cfg.data.workers_per_gpu,
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
round_up=True,
seed=cfg.seed,
sampler_cfg=sampler_cfg) 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:
if device == 'cpu':
warnings.warn(
'The argument `device` is deprecated. To use cpu to train, '
'please refers to https://mmclassification.readthedocs.io/en'
'/latest/getting_started.html#train-a-model')
model = model.cpu()
else:
model = MMDataParallel(model, device_ids=cfg.gpu_ids)
if not model.device_ids:
from mmcv import __version__, digit_version
assert digit_version(__version__) >= (1, 4, 4), \
'To train with CPU, please confirm your mmcv version ' \
'is not lower than v1.4.4'
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
if cfg.get('runner') is None:
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)
runner = build_runner(cfg.runner,
default_args=dict(model=model,
batch_processor=None,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
# 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,
cfg.get('momentum_config', None),
custom_hooks_config=cfg.get(
'custom_hooks', None))
if distributed and cfg.runner['type'] == 'EpochBasedRunner':
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=cfg.data.samples_per_gpu,
workers_per_gpu=cfg.data.workers_per_gpu,
dist=distributed,
shuffle=False,
round_up=True)
eval_cfg = cfg.get('evaluation', {})
eval_cfg['by_epoch'] = cfg.runner['type'] != 'IterBasedRunner'
eval_hook = DistEvalHook if distributed else EvalHook
# `EvalHook` needs to be executed after `IterTimerHook`.
# Otherwise, it will cause a bug if use `IterBasedRunner`.
# Refers to https://github.com/open-mmlab/mmcv/issues/1261
runner.register_hook(eval_hook(val_dataloader, **eval_cfg),
priority='LOW')
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_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', True)
# 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)
# model.ddp = model
else:
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
# print(model)
# 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))
# print(cfg.optimizer)
# print(cfg.optimizer_config)
optimizer_config = OptimizerHookLW(**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))
# 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_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)
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
# 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 train_model(model,
dataset,
cfg,
distributed=False,
validate=False,
timestamp=None,
device=None,
meta=None):
"""Train a model.
This method will build dataloaders, wrap the model and build a runner
according to the provided config.
Args:
model (:obj:`torch.nn.Module`): The model to be run.
dataset (:obj:`mmcls.datasets.BaseDataset` | List[BaseDataset]):
The dataset used to train the model. It can be a single dataset,
or a list of dataset with the same length as workflow.
cfg (:obj:`mmcv.utils.Config`): The configs of the experiment.
distributed (bool): Whether to train the model in a distributed
environment. Defaults to False.
validate (bool): Whether to do validation with
:obj:`mmcv.runner.EvalHook`. Defaults to False.
timestamp (str, optional): The timestamp string to auto generate the
name of log files. Defaults to None.
device (str, optional): TODO
meta (dict, optional): A dict records some import information such as
environment info and seed, which will be logged in logger hook.
Defaults to None.
"""
logger = get_root_logger()
# prepare data loaders
dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
# The default loader config
loader_cfg = dict(
# cfg.gpus will be ignored if distributed
num_gpus=cfg.ipu_replicas if device == 'ipu' else len(cfg.gpu_ids),
dist=distributed,
round_up=True,
seed=cfg.get('seed'),
sampler_cfg=cfg.get('sampler', None),
)
# The overall dataloader settings
loader_cfg.update({
k: v
for k, v in cfg.data.items() if k not in [
'train', 'val', 'test', 'train_dataloader', 'val_dataloader',
'test_dataloader'
]
})
# The specific dataloader settings
train_loader_cfg = {**loader_cfg, **cfg.data.get('train_dataloader', {})}
data_loaders = [build_dataloader(ds, **train_loader_cfg) 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:
if device == 'cpu':
warnings.warn(
'The argument `device` is deprecated. To use cpu to train, '
'please refers to https://mmclassification.readthedocs.io/en'
'/latest/getting_started.html#train-a-model')
model = model.cpu()
elif device == 'ipu':
model = model.cpu()
else:
model = MMDataParallel(model, device_ids=cfg.gpu_ids)
if not model.device_ids:
from mmcv import __version__, digit_version
assert digit_version(__version__) >= (1, 4, 4), \
'To train with CPU, please confirm your mmcv version ' \
'is not lower than v1.4.4'
# build runner
optimizer = build_optimizer(model, cfg.optimizer)
if cfg.get('runner') is None:
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)
if device == 'ipu':
if not cfg.runner['type'].startswith('IPU'):
cfg.runner['type'] = 'IPU' + cfg.runner['type']
if 'options_cfg' not in cfg.runner:
cfg.runner['options_cfg'] = {}
cfg.runner['options_cfg']['replicationFactor'] = cfg.ipu_replicas
cfg.runner['fp16_cfg'] = cfg.get('fp16', None)
runner = build_runner(cfg.runner,
default_args=dict(model=model,
batch_processor=None,
optimizer=optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
# 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:
if device == 'ipu':
from mmcv.device.ipu import IPUFp16OptimizerHook
optimizer_config = IPUFp16OptimizerHook(
**cfg.optimizer_config,
loss_scale=fp16_cfg['loss_scale'],
distributed=distributed)
else:
optimizer_config = Fp16OptimizerHook(
**cfg.optimizer_config,
loss_scale=fp16_cfg['loss_scale'],
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,
cfg.get('momentum_config', None),
custom_hooks_config=cfg.get(
'custom_hooks', None))
if distributed and cfg.runner['type'] == 'EpochBasedRunner':
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
# The specific dataloader settings
val_loader_cfg = {
**loader_cfg,
'shuffle': False, # Not shuffle by default
'sampler_cfg': None, # Not use sampler by default
**cfg.data.get('val_dataloader', {}),
}
val_dataloader = build_dataloader(val_dataset, **val_loader_cfg)
eval_cfg = cfg.get('evaluation', {})
eval_cfg['by_epoch'] = cfg.runner['type'] != 'IterBasedRunner'
eval_hook = DistEvalHook if distributed else EvalHook
# `EvalHook` needs to be executed after `IterTimerHook`.
# Otherwise, it will cause a bug if use `IterBasedRunner`.
# Refers to https://github.com/open-mmlab/mmcv/issues/1261
runner.register_hook(eval_hook(val_dataloader, **eval_cfg),
priority='LOW')
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_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]
if not torch.cuda.is_available():
len_gpu_ids = 2 # need to be changed
else:
len_gpu_ids = len(cfg.gpu_ids)
data_loaders = [
build_dataloader(
ds, # A PyTorch dataset.
cfg.data.
samples_per_gpu, # Number of training samples on each GPU, i.e., batch size of each GPU.
cfg.data.
workers_per_gpu, # How many subprocesses to use for data loading for each GPU.
# cfg.gpus will be ignored if distributed
len_gpu_ids,
# len(cfg.gpu_ids), # Number of GPUs. Only used in non-distributed training.
dist=distributed, # Distributed training/test or not. Default: True.
seed=cfg.seed,
drop_last=True) for ds in dataset
]
''' About build_dataloader
shuffle (bool): Whether to shuffle the data at every epoch.
Default: True.
seed (int | None): Seed to be used. Default: None.
drop_last (bool): Whether to drop the last incomplete batch in epoch.
Default: False
pin_memory (bool): Whether to use pin_memory in DataLoader.
Default: True
dataloader_type (str): Type of dataloader. Default: 'PoolDataLoader'
kwargs: any keyword argument to be used to initialize DataLoader'''
# 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:
if torch.cuda.is_available():
model = MMDataParallel(model.cuda(cfg.gpu_ids[0]),
device_ids=cfg.gpu_ids)
else:
model = MMDataParallel(model.to('cpu'))
# 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=len_gpu_ids,
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_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,
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)
# 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),
custom_hooks_config=cfg.get(
'custom_hooks', 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')
resume_from = None
if cfg.resume_from is None and cfg.get('auto_resume'):
resume_from = find_latest_checkpoint(cfg.work_dir)
if resume_from is not None:
cfg.resume_from = resume_from
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_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]
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)
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
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)
anchor_generator = build_anchor_generator(cfg.model.rpn_head.anchor_generator)
assigner = build_assigner(cfg.model.train_cfg.rpn.assigner)
total_num_targets = torch.tensor([0] * 5)
for iteration, data in enumerate(data_loaders):
for i in data:
# print(i.keys())
img_metas = i['img_metas']._data
# print(img_metas)
num_imgs = len(img_metas)
images = i['img']._data
gt_bboxes = i['gt_bboxes']._data
h, w = images[0].size()[-2:]
features_shape = []
for i in range(2, 7):
f_shape = [int(h/(2**i)), int(w/(2**i))]
features_shape.append(f_shape)
multi_level_anchors = anchor_generator.grid_anchors(
features_shape)
anchor_list = [multi_level_anchors for _ in range(num_imgs)]
# for each image, we compute valid flags of multi level anchors
valid_flag_list = []
for img_id, img_meta in enumerate(img_metas):
multi_level_flags = anchor_generator.valid_flags(
features_shape, img_meta[0]['pad_shape'])
valid_flag_list.append(multi_level_flags)
# print(anchor_list, valid_flag_list)
assert len(anchor_list) == len(valid_flag_list) == num_imgs
# anchor number of multi levels
num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
# concat all level anchors to a single tensor
concat_anchor_list = []
concat_valid_flag_list = []
for i in range(num_imgs):
assert len(anchor_list[i]) == len(valid_flag_list[i])
concat_anchor_list.append(torch.cat(anchor_list[i]))
concat_valid_flag_list.append(torch.cat(valid_flag_list[i]))
gt_bboxes_ignore_list= None
# compute targets for each image
if gt_bboxes_ignore_list is None:
gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
inside_flags = anchor_inside_flags(concat_anchor_list[0], concat_valid_flag_list[0],
img_metas[0][0]['img_shape'][:2],
0)
if not inside_flags.any():
return (None, ) * 7
# assign gt and sample anchors
anchors = concat_anchor_list[0][inside_flags, :]
assign_result = assigner.assign(
anchors.cpu(), gt_bboxes[0][0], gt_bboxes_ignore_list[0],
None)
print(assign_result.pos_gt_bboxes)
pos_inds = torch.nonzero(assign_result.gt_inds > 0, as_tuple=False)
labels = anchors.new_full((anchors.shape[0], ),
-1,
dtype=torch.long)
labels[pos_inds] = 1
num_total_anchors = concat_anchor_list[0].size(0)
labels = unmap(
labels, num_total_anchors, inside_flags,
fill=-1) # fill bg label
match_results = images_to_levels([labels], num_level_anchors)
# print(match_results)
for idx, match_result in enumerate(match_results):
num = torch.where(match_result==1)[0].numel()
total_num_targets[idx] += num
# print(total_num_targets)
print(total_num_targets)
print(total_num_targets)
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]
# step 1: give default values and override (if exist) from cfg.data
default_loader_cfg = {
**dict(num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.get('seed'),
drop_last=False,
persistent_workers=False),
**({} if torch.__version__ != 'parrots' else dict(
prefetch_num=2,
pin_memory=False,
)),
}
# update overall dataloader(for train, val and test) setting
default_loader_cfg.update({
k: v
for k, v in cfg.data.items() if k not in [
'train', 'val', 'test', 'train_dataloader', 'val_dataloader',
'test_dataloader'
]
})
# step 2: cfg.data.train_dataloader has highest priority
train_loader_cfg = dict(default_loader_cfg,
**cfg.data.get('train_dataloader', {}))
data_loaders = [build_dataloader(ds, **train_loader_cfg) 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:
if not torch.cuda.is_available():
assert digit_version(mmcv.__version__) >= digit_version('1.4.4'), \
'Please use MMCV >= 1.4.4 for CPU training!'
model = MMDataParallel(model, 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),
custom_hooks_config=cfg.get(
'custom_hooks', None))
if distributed:
if isinstance(runner, EpochBasedRunner):
runner.register_hook(DistSamplerSeedHook())
# register eval hooks
if validate:
val_samples_per_gpu = (cfg.data.get('val_dataloader', {})).get(
'samples_per_gpu', cfg.data.get('samples_per_gpu', 1))
if val_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)
cfg = replace_image_to_tensor(cfg)
val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
val_loader_cfg = {
**default_loader_cfg,
**dict(shuffle=False, drop_last=False),
**cfg.data.get('val_dataloader', {}),
**dict(samples_per_gpu=val_samples_per_gpu)
}
val_dataloader = build_dataloader(val_dataset, **val_loader_cfg)
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_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]
# step 1: give default values and override (if exist) from cfg.data
loader_cfg = {
**dict(
seed=cfg.get('seed'),
drop_last=False,
dist=distributed,
num_gpus=len(cfg.gpu_ids)),
**({} if torch.__version__ != 'parrots' else dict(
prefetch_num=2,
pin_memory=False,
)),
**dict((k, cfg.data[k]) for k in [
'samples_per_gpu',
'workers_per_gpu',
'shuffle',
'seed',
'drop_last',
'prefetch_num',
'pin_memory',
] if k in cfg.data)
}
# step 2: cfg.data.train_dataloader has highest priority
train_loader_cfg = dict(loader_cfg, **cfg.data.get('train_dataloader', {}))
data_loaders = [build_dataloader(ds, **train_loader_cfg) 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:
val_samples_per_gpu = (cfg.data.get('val_dataloader', {})).get(
'samples_per_gpu', cfg.data.get('samples_per_gpu', 1))
if val_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.val.get('pipeline', None) is not None:
# 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_loader_cfg = {
**loader_cfg,
**dict(shuffle=False, drop_last=False),
**cfg.data.get('val_dataloader', {}),
**dict(samples_per_gpu=val_samples_per_gpu)
}
val_dataloader = build_dataloader(val_dataset, **val_loader_cfg)
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))
# 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 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]
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)
# if just swa training is performed,
# skip building the runner for the traditional training
if not cfg.get('only_swa_training', False):
# 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
runner.register_hook(
eval_hook(val_dataloader, save_best='bbox_mAP', **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)
else:
# if just swa training is performed, there should be a starting model
assert cfg.swa_resume_from is not None or cfg.swa_load_from is not None
# perform swa training
# build swa training runner
if not cfg.get('swa_training', False):
return
from mmdet.core import SWAHook
logger.info('Start SWA training')
swa_optimizer = build_optimizer(model, cfg.swa_optimizer)
swa_runner = build_runner(cfg.swa_runner,
default_args=dict(model=model,
optimizer=swa_optimizer,
work_dir=cfg.work_dir,
logger=logger,
meta=meta))
# an ugly workaround to make .log and .log.json filenames the same
swa_runner.timestamp = timestamp
# fp16 setting
fp16_cfg = cfg.get('fp16', None)
if fp16_cfg is not None:
swa_optimizer_config = Fp16OptimizerHook(**cfg.swa_optimizer_config,
**fp16_cfg,
distributed=distributed)
elif distributed and 'type' not in cfg.swa_optimizer_config:
swa_optimizer_config = OptimizerHook(**cfg.swa_optimizer_config)
else:
swa_optimizer_config = cfg.swa_optimizer_config
# register hooks
swa_runner.register_training_hooks(cfg.swa_lr_config, swa_optimizer_config,
cfg.swa_checkpoint_config,
cfg.log_config,
cfg.get('momentum_config', None))
if distributed:
if isinstance(swa_runner, EpochBasedRunner):
swa_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
swa_runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
swa_eval = True
swa_eval_hook = eval_hook(val_dataloader,
save_best='bbox_mAP',
**eval_cfg)
else:
swa_eval = False
swa_eval_hook = None
# register swa hook
swa_hook = SWAHook(swa_eval=swa_eval,
eval_hook=swa_eval_hook,
swa_interval=cfg.swa_interval)
swa_runner.register_hook(swa_hook, priority='LOW')
# register 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)
swa_runner.register_hook(hook, priority=priority)
if cfg.swa_resume_from:
swa_runner.resume(cfg.swa_resume_from)
elif cfg.swa_load_from:
# use the best pretrained model as the starting model for swa training
if cfg.swa_load_from == 'best_bbox_mAP.pth':
best_model_path = os.path.join(cfg.work_dir, cfg.swa_load_from)
# avoid the best pretrained model being overwritten
new_best_model_path = os.path.join(cfg.work_dir,
'best_bbox_mAP_pretrained.pth')
if swa_runner.rank == 0:
import shutil
assert os.path.exists(best_model_path)
shutil.copy(best_model_path,
new_best_model_path,
follow_symlinks=False)
cfg.swa_load_from = best_model_path
swa_runner.load_checkpoint(cfg.swa_load_from)
swa_runner.run(data_loaders, cfg.workflow)
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]
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
]
# build optimizer
optimizer = build_optimizer(model, cfg.optimizer)
# use apex fp16 optimizer
if cfg.optimizer_config.get(
"type",
None) and cfg.optimizer_config["type"] == "DistOptimizerHook":
if cfg.optimizer_config.get("use_fp16", False):
model, optimizer = apex.amp.initialize(model.cuda(),
optimizer,
opt_level="O1")
for m in model.modules():
if hasattr(m, "fp16_enabled"):
m.fp16_enabled = True
# 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)
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
# build runner
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
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)
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]
# The default loader config
loader_cfg = dict(
# cfg.gpus will be ignored if distributed
num_gpus=len(cfg.gpu_ids),
dist=distributed,
seed=cfg.seed,
drop_last=True)
# The overall dataloader settings
loader_cfg.update({
k: v
for k, v in cfg.data.items() if k not in [
'train', 'val', 'test', 'train_dataloader', 'val_dataloader',
'test_dataloader'
]
})
# The specific dataloader settings
train_loader_cfg = {**loader_cfg, **cfg.data.get('train_dataloader', {})}
data_loaders = [build_dataloader(ds, **train_loader_cfg) 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:
if not torch.cuda.is_available():
assert digit_version(mmcv.__version__) >= digit_version('1.4.4'), \
'Please use MMCV >= 1.4.4 for CPU training!'
model = MMDataParallel(model, device_ids=cfg.gpu_ids)
# 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))
if distributed:
# when distributed training by epoch, using`DistSamplerSeedHook` to set
# the different seed to distributed sampler for each epoch, it will
# shuffle dataset at each epoch and avoid overfitting.
if isinstance(runner, EpochBasedRunner):
runner.register_hook(DistSamplerSeedHook())
# 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))
# The specific dataloader settings
val_loader_cfg = {
**loader_cfg,
'samples_per_gpu': 1,
'shuffle': False, # Not shuffle by default
**cfg.data.get('val_dataloader', {}),
}
val_dataloader = build_dataloader(val_dataset, **val_loader_cfg)
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 is None and cfg.get('auto_resume'):
resume_from = find_latest_checkpoint(cfg.work_dir)
if resume_from is not None:
cfg.resume_from = resume_from
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)