def test_training_sampler_seed(self):
seed_all_rng(42)
sampler = TrainingSampler(30)
data = list(itertools.islice(sampler, 65))
seed_all_rng(42)
sampler = TrainingSampler(30)
seed_all_rng(999) # should be ineffective
data2 = list(itertools.islice(sampler, 65))
self.assertEqual(data, data2)
def build_inference_based_loader(
cfg: CfgNode, dataset_cfg: CfgNode,
model: torch.nn.Module) -> InferenceBasedLoader:
"""
Constructs data loader based on inference results of a model.
"""
dataset = build_bootstrap_dataset(dataset_cfg.DATASET,
dataset_cfg.IMAGE_LOADER)
training_sampler = TrainingSampler(len(dataset))
data_loader = torch.utils.data.DataLoader(
dataset,
batch_size=dataset_cfg.IMAGE_LOADER.BATCH_SIZE,
sampler=training_sampler,
num_workers=dataset_cfg.IMAGE_LOADER.NUM_WORKERS,
collate_fn=trivial_batch_collator,
)
return InferenceBasedLoader(
model,
data_loader=data_loader,
data_sampler=build_data_sampler(dataset_cfg.DATA_SAMPLER),
data_filter=build_data_filter(dataset_cfg.FILTER),
shuffle=True,
batch_size=dataset_cfg.INFERENCE.OUTPUT_BATCH_SIZE,
inference_batch_size=dataset_cfg.INFERENCE.INPUT_BATCH_SIZE,
)
def build_custom_train_loader(cfg, mapper=None):
"""
Modified from detectron2.data.build.build_custom_train_loader, but supports
different samplers
"""
source_aware = cfg.DATALOADER.SOURCE_AWARE
if source_aware:
dataset_dicts = get_detection_dataset_dicts_with_source(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
sizes = [0 for _ in range(len(cfg.DATASETS.TRAIN))]
for d in dataset_dicts:
sizes[d['dataset_source']] += 1
print('dataset sizes', sizes)
else:
dataset_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
dataset = DatasetFromList(dataset_dicts, copy=False)
if mapper is None:
assert 0
# mapper = DatasetMapper(cfg, True)
dataset = MapDataset(dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
# TODO avoid if-else?
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
elif sampler_name == "MultiDatasetSampler":
assert source_aware
sampler = MultiDatasetSampler(cfg, sizes, dataset_dicts)
elif sampler_name == "RepeatFactorTrainingSampler":
repeat_factors = RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
dataset_dicts, cfg.DATALOADER.REPEAT_THRESHOLD)
sampler = RepeatFactorTrainingSampler(repeat_factors)
elif sampler_name == "ClassAwareSampler":
sampler = ClassAwareSampler(dataset_dicts)
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def build_inference_based_loader(
cfg: CfgNode,
dataset_cfg: CfgNode,
model: torch.nn.Module,
embedder: Optional[torch.nn.Module] = None,
) -> InferenceBasedLoader:
"""
Constructs data loader based on inference results of a model.
"""
dataset = build_bootstrap_dataset(dataset_cfg.DATASET, dataset_cfg.IMAGE_LOADER)
meta = MetadataCatalog.get(dataset_cfg.DATASET)
training_sampler = TrainingSampler(len(dataset))
data_loader = torch.utils.data.DataLoader(
dataset, # pyre-ignore[6]
batch_size=dataset_cfg.IMAGE_LOADER.BATCH_SIZE,
sampler=training_sampler,
num_workers=dataset_cfg.IMAGE_LOADER.NUM_WORKERS,
collate_fn=trivial_batch_collator,
worker_init_fn=worker_init_reset_seed,
)
return InferenceBasedLoader(
model,
data_loader=data_loader,
data_sampler=build_data_sampler(cfg, dataset_cfg.DATA_SAMPLER, embedder),
data_filter=build_data_filter(dataset_cfg.FILTER),
shuffle=True,
batch_size=dataset_cfg.INFERENCE.OUTPUT_BATCH_SIZE,
inference_batch_size=dataset_cfg.INFERENCE.INPUT_BATCH_SIZE,
category_to_class_mapping=meta.category_to_class_mapping,
)
def build_detection_train_loader(cfg):
"""Builds a data loader for the baseline trainer with support of training on the subset of labeled data only.
Most of code comes from `d2.data.build.build_detection_train_loader()`, see it for more details.
"""
# CSD: check config is supported
assert cfg.DATALOADER.SAMPLER_TRAIN == "TrainingSampler", "Unsupported training sampler: {}".format(
cfg.DATALOADER.SAMPLER_TRAIN)
# Original code
dataset = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
# CSD: subsample the dataset if needed
dataset = check_subsample_dataset(dataset, cfg)
if comm.is_main_process(): # Log counts
logger = setup_logger(name=__name__)
logger.debug("Number of images in the dataset: {}".format(
len(dataset)))
_log_api_usage("dataset." + cfg.DATASETS.TRAIN[0])
# Original code
mapper = DatasetMapper(cfg, True)
sampler = TrainingSampler(len(dataset))
dataset = DatasetFromList(dataset, copy=False)
dataset = MapDataset(dataset, mapper)
sampler = TrainingSampler(len(dataset))
assert isinstance(sampler, torch.utils.data.sampler.Sampler)
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def build_detection_train_loader(cfg: CfgNode, mapper=None):
"""
A data loader is created in a way similar to that of Detectron2.
The main differences are:
- it allows to combine datasets with different but compatible object category sets
The data loader is created by the following steps:
1. Use the dataset names in config to query :class:`DatasetCatalog`, and obtain a list of dicts.
2. Start workers to work on the dicts. Each worker will:
* Map each metadata dict into another format to be consumed by the model.
* Batch them by simply putting dicts into a list.
The batched ``list[mapped_dict]`` is what this dataloader will return.
Args:
cfg (CfgNode): the config
mapper (callable): a callable which takes a sample (dict) from dataset and
returns the format to be consumed by the model.
By default it will be `DatasetMapper(cfg, True)`.
Returns:
an infinite iterator of training data
"""
_add_category_whitelists_to_metadata(cfg)
_add_category_maps_to_metadata(cfg)
dataset_dicts = combine_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
keep_instance_predicate=_get_train_keep_instance_predicate(cfg),
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN if cfg.MODEL.LOAD_PROPOSALS else None,
)
dataset = DatasetFromList(dataset_dicts, copy=False)
if mapper is None:
mapper = DatasetMapper(cfg, True)
dataset = MapDataset(dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
elif sampler_name == "RepeatFactorTrainingSampler":
repeat_factors = RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
dataset_dicts, cfg.DATALOADER.REPEAT_THRESHOLD
)
sampler = RepeatFactorTrainingSampler(repeat_factors)
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def build_detection_train_loader(cfg, mapper=None):
"""
A data loader is created by the following steps:
1. Use the dataset names in config to query :class:`DatasetCatalog`, and obtain a list of dicts.
2. Coordinate a random shuffle order shared among all processes (all GPUs)
3. Each process spawn another few workers to process the dicts. Each worker will:
* Map each metadata dict into another format to be consumed by the model.
* Batch them by simply putting dicts into a list.
The batched ``list[mapped_dict]`` is what this dataloader will yield.
Args:
cfg (CfgNode): the config
mapper (callable): a callable which takes a sample (dict) from dataset and
returns the format to be consumed by the model.
By default it will be ``DatasetMapper(cfg, True)``.
Returns:
an infinite iterator of training data
"""
dataset_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON
else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN if cfg.MODEL.LOAD_PROPOSALS else None,
)
dataset = DatasetFromList(dataset_dicts, copy=False)
if mapper is None:
mapper = DatasetMapper(cfg, True)
dataset = MapDataset(dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
# TODO avoid if-else?
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
elif sampler_name == "RepeatFactorTrainingSampler":
repeat_factors = RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
dataset_dicts, cfg.DATALOADER.REPEAT_THRESHOLD
)
sampler = RepeatFactorTrainingSampler(repeat_factors)
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def build_detection_semisup_train_loader(cfg, mapper=None):
dataset_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
# Divide into labeled and unlabeled sets according to supervision percentage
label_dicts, unlabel_dicts = divide_label_unlabel(
dataset_dicts,
cfg.DATALOADER.SUP_PERCENT,
cfg.DATALOADER.RANDOM_DATA_SEED,
cfg.DATALOADER.RANDOM_DATA_SEED_PATH,
)
dataset = DatasetFromList(label_dicts, copy=False)
if mapper is None:
mapper = DatasetMapper(cfg, True)
dataset = MapDataset(dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
elif sampler_name == "RepeatFactorTrainingSampler":
repeat_factors = (
RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
label_dicts, cfg.DATALOADER.REPEAT_THRESHOLD))
sampler = RepeatFactorTrainingSampler(repeat_factors)
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
# list num of labeled and unlabeled
logger.info("Number of training samples " + str(len(dataset)))
logger.info("Supervision percentage " + str(cfg.DATALOADER.SUP_PERCENT))
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def build_simple_dataloader(dataset_name: list, batch_size):
dataset_dicts = get_detection_dataset_dicts(dataset_name)
dataset = DatasetFromList(dataset_dicts, copy=False)
cfg = get_cfg()
cfg["aug_kwargs"] = {}
dataset = MapDataset(dataset, AlbumentationsMapper(cfg, False))
# set the shuffle to False in debugging mode
sampler = TrainingSampler(len(dataset), shuffle=False, seed=42)
dataloader = build_batch_data_loader(dataset=dataset, sampler=sampler,
total_batch_size=batch_size)
return dataloader
def build_hand_train_loader(cfg):
dataset_dicts, num_per_epoch = load_hand(cfg.HAND_PROJECT.DATA.MODE, cfg.HAND_PROJECT.DATA.ANNOT_SUBSET_TRAIN, cfg.HAND_PROJECT.DATA.BASE_PATH, selects=cfg.HAND_PROJECT.DATA.SELECTS)
# pdb.set_trace()
dataset = DatasetFromList(dataset_dicts, copy=False)
mapper = DatasetMapper(cfg, True)
dataset = MapDataset(dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
sampler = TrainingSampler(len(dataset))
return build_batch_data_loader(dataset, sampler, cfg.SOLVER.IMS_PER_BATCH, num_workers=cfg.DATALOADER.NUM_WORKERS), num_per_epoch
def build_detection_train_loader_with_train_sampler(cfg, mapper, seed=42, shuffle=True):
dataset_dicts = get_detection_dataset_dicts(cfg.DATASETS.TRAIN)
dataset = DatasetFromList(dataset_dicts, copy=False)
dataset = MapDataset(dataset, mapper)
logger = logging.getLogger(__name__)
logger.info("Using training sampler TrainingSampler with shuffle=False")
sampler = TrainingSampler(len(dataset), shuffle=shuffle, seed=seed)
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def build_x_train_loader(
dataset, *, mapper, sampler=None, total_batch_size, aspect_ratio_grouping=True, num_workers=0
):
if isinstance(dataset, list):
dataset = DatasetFromList(dataset, copy=False)
if mapper is not None:
dataset = MapDataset(dataset, mapper)
if sampler is None:
sampler = TrainingSampler(len(dataset))
assert isinstance(sampler, torch.utils.data.sampler.Sampler)
return build_batch_data_loader(
dataset,
sampler,
total_batch_size,
aspect_ratio_grouping=aspect_ratio_grouping,
num_workers=num_workers,
)
def test_to_iterable(self):
sampler = TrainingSampler(100, seed=10)
dataset = DatasetFromList(list(range(100)))
dataset = ToIterableDataset(dataset, sampler)
data_loader = data.DataLoader(dataset, num_workers=0, collate_fn=operator.itemgetter(0))
output = list(itertools.islice(data_loader, 100))
self.assertEqual(set(output), set(range(100)))
data_loader = data.DataLoader(
dataset,
num_workers=2,
collate_fn=operator.itemgetter(0),
worker_init_fn=worker_init_reset_seed,
# reset seed should not affect behavior of TrainingSampler
)
output = list(itertools.islice(data_loader, 100))
# multiple workers should not lead to duplicate or different data
self.assertEqual(set(output), set(range(100)))
def build_train_loader(cfg, mapper=None):
if mapper is None:
mapper = get_dataset_mapper(cfg.DATASETS.TRAIN[0])(cfg, True)
dataset_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
dataset = DatasetFromList(dataset_dicts, copy=False)
dataset = MapDataset(dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
# TODO avoid if-else?
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
elif sampler_name == "RepeatFactorTrainingSampler":
repeat_factors = RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
dataset_dicts, cfg.DATALOADER.REPEAT_THRESHOLD)
sampler = RepeatFactorTrainingSampler(repeat_factors)
elif sampler_name == "RatioFactorTrainingSampler":
repeat_factors = repeat_factors_from_ratios(dataset_dicts)
sampler = RepeatFactorTrainingSampler(repeat_factors)
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
return build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
def _train_loader_from_config(cfg, mapper=None, *, dataset=None, sampler=None):
if dataset is None:
dataset = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON
else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN if cfg.MODEL.LOAD_PROPOSALS else None,
)
if mapper is None:
mapper = DatasetMapper(cfg, True)
if sampler is None:
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger("detectron2.trainer")
logger.info("Using training sampler {}".format(sampler_name))
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
elif sampler_name == "RepeatFactorTrainingSampler":
repeat_factors = RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
dataset, cfg.DATALOADER.REPEAT_THRESHOLD
)
sampler = RepeatFactorTrainingSampler(repeat_factors)
else:
raise ValueError(
"Unknown training sampler: {}".format(sampler_name))
return {
"dataset": dataset,
"sampler": sampler,
"mapper": mapper,
"total_batch_size": cfg.SOLVER.IMS_PER_BATCH,
"aspect_ratio_grouping": cfg.DATALOADER.ASPECT_RATIO_GROUPING,
"num_workers": cfg.DATALOADER.NUM_WORKERS,
}
def build_train_dataloader(cfg): # like 'build_detection_train_loader'
if 'coco_2017_train' in cfg.DATASETS.TRAIN:
descs_train: List[Dict] = DatasetCatalog.get("coco_2017_train")
ds_train = DatasetFromList(descs_train, copy=False)
mapper = DatasetMapper(cfg, True)
else: # Open-Image-Dataset
if 'get_detection_dataset_dicts':
all_descs_train: List[Dict] = DatasetCatalog.get("oid_train")
if 'rebalancing':
image_id_vs_idx = {}
for idx, desc in enumerate(all_descs_train):
image_id_vs_idx[desc['image_id']] = idx
descs_train = list(map(lambda img_id: all_descs_train[image_id_vs_idx[img_id]], sample_image_ids()))
print('_' * 50 + f'train dataset len: {len(descs_train)}')
ds_train = DatasetFromList(descs_train, copy=False)
if 'DatasetMapper':
augs = [RandomContrast(0.8, 1.2),
RandomBrightness(0.8, 1.2),
RandomSaturation(0.8, 1.2)]
augs.extend(build_augmentation(cfg, is_train=True))
mapper = make_mapper('oid_train', is_train=True, augmentations=T.AugmentationList(augs))
ds_train = MapDataset(ds_train, mapper)
sampler = TrainingSampler(len(ds_train))
data_loader = build_batch_data_loader(
ds_train,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
global DATA_LOADER
DATA_LOADER = data_loader
return data_loader
def build_ss_train_loader(cfg, mapper):
"""Builds a semi-supervised data loader that yields both labeled and unlabeled images.
Data can be loaded in two modes (defined in `cfg.DATASETS.MODE`):
- "CROSS_DATASET": labeled and unlabeled images come from two disparate datasets, e.g.
VOCtrain and VOCtest
- "RANDOM_SPLIT": labeled and unlabeled images come from the same dataset by splitting it
into the labeled and unlabeled parts
For more details see `build_ss_datasets()`.
Each batch consists of `cfg.SOLVER.IMS_PER_BATCH_LABELED` labeled and
`cfg.SOLVER.IMS_PER_BATCH_UNLABELED` unlabeled images, which can be modified
in `csd/config/config.py` or in a custom `configs/*.yaml` config file
supplied to your training script.
The actual x-flips happen inside `AspectRatioGroupedSSDataset` that is instantiated by
`build_ss_batch_data_loader`
The returned tuple contains (1) a tuple of lists with dicts for labeled and unlabeled images
and (2) a DataLoader with infinite sampling yielding a pair of batches with labeled and unlabeled
images with the same aspect ratio within batch.
Specifically, the returned DataLoader yields a tuple of lists:
([labeled_img, labeled_img_xflip], [unlabeled_im, unlabeled_img_xflip]).
"""
# Load labeled and unlabeled dataset dicts (either use two separate ones or perform a random split)
labeled_dataset_dicts, unlabeled_dataset_dicts = build_ss_datasets(cfg)
# Log the datasets sizes
if comm.is_main_process():
logger = setup_logger(name=__name__)
logger.debug(
"Number of images in the labeled and unlabeled datasets: {}, {}".
format(len(labeled_dataset_dicts), len(unlabeled_dataset_dicts)))
# Print updated metadata counts
print_instances_class_histogram(
labeled_dataset_dicts,
MetadataCatalog.get(cfg.DATASETS.TRAIN[0]).thing_classes)
# Map metadata into actual objects (note: data augmentations also take place here)
labeled_dataset = MapDataset(labeled_dataset_dicts, mapper)
unlabeled_dataset = MapDataset(unlabeled_dataset_dicts, mapper)
# Define data samplers
assert cfg.DATALOADER.SAMPLER_TRAIN == "TrainingSampler", "Unsupported training sampler: {}".format(
cfg.DATALOADER.SAMPLER_TRAIN)
labeled_sampler = TrainingSampler(len(labeled_dataset))
unlabeled_sampler = TrainingSampler(len(unlabeled_dataset))
return (
labeled_dataset_dicts,
unlabeled_dataset_dicts,
), build_ss_batch_data_loader( # Initialize actual dataloaders
(labeled_dataset, unlabeled_dataset),
(labeled_sampler, unlabeled_sampler),
cfg.SOLVER.IMS_PER_BATCH_LABELED,
cfg.SOLVER.IMS_PER_BATCH_UNLABELED,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON
else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN if cfg.MODEL.LOAD_PROPOSALS else None,
)
dataset = DatasetFromList(dataset_dicts, copy=False)
dataset = MapDataset(dataset, DatasetMapper(cfg, True))
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
# TODO avoid if-else?
if sampler_name == "TrainingSampler":
sampler = TrainingSampler(len(dataset))
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
data_loader = build_batch_data_loader(
dataset,
sampler,
cfg.SOLVER.IMS_PER_BATCH,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
# logger.info("Starting training from iteration {}".format(start_iter))
# with EventStorage(start_iter) as storage:
# for data, iteration in zip(data_loader, range(start_iter, max_iter)):
# iteration = iteration + 1
# storage.step()
def test_build_iterable_dataloader_train(self):
kwargs = self._get_kwargs()
ds = DatasetFromList(kwargs.pop("dataset"))
ds = ToIterableDataset(ds, TrainingSampler(len(ds)))
dl = build_detection_train_loader(dataset=ds, **kwargs)
next(iter(dl))
def build_detection_semisup_train_loader_two_crops(cfg, mapper=None):
if cfg.DATASETS.CROSS_DATASET: # cross-dataset (e.g., coco-additional)
label_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN_LABEL,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
unlabel_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN_UNLABEL,
filter_empty=False,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
else: # different degree of supervision (e.g., COCO-supervision)
dataset_dicts = get_detection_dataset_dicts(
cfg.DATASETS.TRAIN,
filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
if cfg.MODEL.KEYPOINT_ON else 0,
proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN
if cfg.MODEL.LOAD_PROPOSALS else None,
)
# Divide into labeled and unlabeled sets according to supervision percentage
label_dicts, unlabel_dicts = divide_label_unlabel(
dataset_dicts,
cfg.DATALOADER.SUP_PERCENT,
cfg.DATALOADER.RANDOM_DATA_SEED,
cfg.DATALOADER.RANDOM_DATA_SEED_PATH,
)
label_dataset = DatasetFromList(label_dicts, copy=False)
# exclude the labeled set from unlabeled dataset
unlabel_dataset = DatasetFromList(unlabel_dicts, copy=False)
# include the labeled set in unlabel dataset
# unlabel_dataset = DatasetFromList(dataset_dicts, copy=False)
if mapper is None:
mapper = DatasetMapper(cfg, True)
label_dataset = MapDataset(label_dataset, mapper)
unlabel_dataset = MapDataset(unlabel_dataset, mapper)
sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
logger = logging.getLogger(__name__)
logger.info("Using training sampler {}".format(sampler_name))
if sampler_name == "TrainingSampler":
label_sampler = TrainingSampler(len(label_dataset))
unlabel_sampler = TrainingSampler(len(unlabel_dataset))
elif sampler_name == "RepeatFactorTrainingSampler":
raise NotImplementedError("{} not yet supported.".format(sampler_name))
else:
raise ValueError("Unknown training sampler: {}".format(sampler_name))
return build_semisup_batch_data_loader_two_crop(
(label_dataset, unlabel_dataset),
(label_sampler, unlabel_sampler),
cfg.SOLVER.IMG_PER_BATCH_LABEL,
cfg.SOLVER.IMG_PER_BATCH_UNLABEL,
aspect_ratio_grouping=cfg.DATALOADER.ASPECT_RATIO_GROUPING,
num_workers=cfg.DATALOADER.NUM_WORKERS,
)
