def main(configuration, init_distributed=False, predict=False):
# A reload might be needed for imports
setup_imports()
configuration.import_user_dir()
config = configuration.get_config()
if torch.cuda.is_available():
torch.cuda.set_device(config.device_id)
torch.cuda.init()
if init_distributed:
distributed_init(config)
seed = config.training.seed
config.training.seed = set_seed(seed if seed == -1 else seed + get_rank())
registry.register("seed", config.training.seed)
config = build_config(configuration)
setup_logger(color=config.training.colored_logs,
disable=config.training.should_not_log)
logger = logging.getLogger("mmf_cli.run")
# Log args for debugging purposes
logger.info(configuration.args)
logger.info(f"Torch version: {torch.__version__}")
log_device_names()
logger.info(f"Using seed {config.training.seed}")
trainer = build_trainer(config)
trainer.load()
if predict:
trainer.inference()
else:
trainer.train()
def setUp(self):
test_utils.setup_proxy()
setup_imports()
self.model_name = "mmf_transformer"
args = test_utils.dummy_args(model=self.model_name)
configuration = Configuration(args)
self.config = configuration.get_config()
self.config.model_config[self.model_name].model = self.model_name
self.finetune_model = build_model(self.config.model_config[self.model_name])
def run(opts: typing.Optional[typing.List[str]] = None, predict: bool = False):
"""Run starts a job based on the command passed from the command line.
You can optionally run the mmf job programmatically by passing an optlist as opts.
Args:
opts (typing.Optional[typing.List[str]], optional): Optlist which can be used.
to override opts programmatically. For e.g. if you pass
opts = ["training.batch_size=64", "checkpoint.resume=True"], this will
set the batch size to 64 and resume from the checkpoint if present.
Defaults to None.
predict (bool, optional): If predict is passed True, then the program runs in
prediction mode. Defaults to False.
"""
setup_imports()
if opts is None:
parser = flags.get_parser()
args = parser.parse_args()
else:
args = argparse.Namespace(config_override=None)
args.opts = opts
perturbation_arguments.args = args
configuration = Configuration(args)
# Do set runtime args which can be changed by MMF
configuration.args = args
config = configuration.get_config()
config.start_rank = 0
if config.distributed.init_method is None:
infer_init_method(config)
if config.distributed.init_method is not None:
if torch.cuda.device_count() > 1 and not config.distributed.no_spawn:
config.start_rank = config.distributed.rank
config.distributed.rank = None
torch.multiprocessing.spawn(
fn=distributed_main,
args=(configuration, predict),
nprocs=torch.cuda.device_count(),
)
else:
distributed_main(0, configuration, predict)
elif config.distributed.world_size > 1:
assert config.distributed.world_size <= torch.cuda.device_count()
port = random.randint(10000, 20000)
config.distributed.init_method = f"tcp://localhost:{port}"
config.distributed.rank = None
torch.multiprocessing.spawn(
fn=distributed_main,
args=(configuration, predict),
nprocs=config.distributed.world_size,
)
else:
config.device_id = 0
main(configuration, predict=predict)
def setUp(self):
test_utils.setup_proxy()
setup_imports()
replace_with_jit()
model_name = "visual_bert"
args = test_utils.dummy_args(model=model_name)
configuration = Configuration(args)
config = configuration.get_config()
model_config = config.model_config[model_name]
model_config.model = model_name
self.pretrain_model = build_model(model_config)
def setUp(self):
test_utils.setup_proxy()
setup_imports()
model_name = "mmbt"
args = test_utils.dummy_args(model=model_name)
configuration = Configuration(args)
config = configuration.get_config()
model_config = config.model_config[model_name]
model_config["training_head_type"] = "classification"
model_config["num_labels"] = 2
model_config.model = model_name
self.finetune_model = build_model(model_config)
def setUp(self):
setup_imports()
torch.manual_seed(1234)
config_path = os.path.join(
get_mmf_root(),
"..",
"projects",
"butd",
"configs",
"coco",
"beam_search.yaml",
)
config_path = os.path.abspath(config_path)
args = dummy_args(model="butd", dataset="coco")
args.opts.append(f"config={config_path}")
configuration = Configuration(args)
configuration.config.datasets = "coco"
configuration.freeze()
self.config = configuration.config
registry.register("config", self.config)
def setUp(self):
test_utils.setup_proxy()
setup_imports()
model_name = "vilbert"
args = test_utils.dummy_args(model=model_name)
configuration = Configuration(args)
config = configuration.get_config()
self.vision_feature_size = 1024
self.vision_target_size = 1279
model_config = config.model_config[model_name]
model_config["training_head_type"] = "pretraining"
model_config["visual_embedding_dim"] = self.vision_feature_size
model_config["v_feature_size"] = self.vision_feature_size
model_config["v_target_size"] = self.vision_target_size
model_config["dynamic_attention"] = False
model_config.model = model_name
self.pretrain_model = build_model(model_config)
model_config["training_head_type"] = "classification"
model_config["num_labels"] = 2
self.finetune_model = build_model(model_config)
def setUp(self):
setup_imports()
torch.manual_seed(1234)
config_path = os.path.join(
get_mmf_root(),
"..",
"projects",
"butd",
"configs",
"coco",
"nucleus_sampling.yaml",
)
config_path = os.path.abspath(config_path)
args = dummy_args(model="butd", dataset="coco")
args.opts.append(f"config={config_path}")
configuration = Configuration(args)
configuration.config.datasets = "coco"
configuration.config.model_config.butd.inference.params.sum_threshold = 0.5
configuration.freeze()
self.config = configuration.config
registry.register("config", self.config)
):
cls.mapping["state"]["writer"].warning(
"Key {} is not present in registry, returning default value "
"of {}".format(original_name, default)
)
return value
@classmethod
def unregister(cls, name):
r"""Remove an item from registry with key 'name'
Args:
name: Key which needs to be removed.
Usage::
from VisualBERT.mmf.common.registry import registry
config = registry.unregister("config")
"""
return cls.mapping["state"].pop(name, None)
registry = Registry()
# Only setup imports in main process, this means registry won't be
# fully available in spawned child processes (such as dataloader processes)
# but instantiated. This is to prevent issues such as
# https://github.com/facebookresearch/mmf/issues/355
if __name__ == "__main__":
setup_imports()
def setUp(self):
setup_imports()