def build_transform(cfg):
tfs = []
for icfg in cfg:
tf = build_from_cfg(icfg, TRANSFORMS)
tfs.append(tf)
aug = Compose(tfs)
return aug
def build_backbone(cfg, default_args=None):
backbone = build_from_cfg(cfg, BACKBONES, default_args)
return backbone
def build_model(cfg, default_args=None):
model = build_from_cfg(cfg, MODELS, default_args)
return model
def build_loss(cfg, default_args=None):
loss = build_from_cfg(cfg, LOSSES, default_args)
return loss
def build_criterion(cfg, default_args=None):
criterion = build_from_cfg(cfg, CRITERIA, default_args)
return criterion
def build_grid(cfg, default_args=None):
grid = build_from_cfg(cfg, GRIDS, default_args)
return grid
def build_head(cfg, default_args=None):
head = build_from_cfg(cfg, HEADS, default_args)
return head
def build_neck(cfg, default_args=None):
neck = build_from_cfg(cfg, NECKS, default_args)
return neck
def build_optimizer(cfg, default_args=None):
optimizer = build_from_cfg(cfg, torch_optim, default_args, 'module')
return optimizer
def build_lr_scheduler(cfg, default_args=None):
lr_scheduler = build_from_cfg(cfg, LR_SCHEDULERS, default_args)
return lr_scheduler
def build_runner(cfg, default_args=None):
runner = build_from_cfg(cfg, RUNNERS, default_args)
return runner
def build_dataset(cfg, default_args=None):
dataset = build_from_cfg(cfg, DATASETS, default_args)
return dataset
def build_sampler(cfg, default_args):
sampler = build_from_cfg(cfg, SAMPLERS, default_args)
return sampler
def build_dataloader(cfg, default_args):
loader = build_from_cfg(cfg, DATALOADERS, default_args)
return loader