def __init__(self,
dataset: ScanObjectNNDataset,
use_potential=True,
balance_labels=False):
Sampler.__init__(self, dataset)
# Does the sampler use potential for regular sampling
self.use_potential = use_potential
# Should be balance the classes when sampling
self.balance_labels = balance_labels
# Dataset used by the sampler (no copy is made in memory)
self.dataset = dataset
# Create potentials
if self.use_potential:
self.potentials = np.random.rand(len(
dataset.input_labels)) * 0.1 + 0.1
else:
self.potentials = None
# Initialize value for batch limit (max number of points per batch).
self.batch_limit = 100000
return
def __init__(self, num_samples):
"""
初始化权重均相同，跟据该权重初次抽样进入网络
:param num_samples:
"""
Sampler.__init__(self, data_source=None)
self.weights = torch.zeros(num_samples).fill_(1 / num_samples)
self.num_samples = num_samples
self.sampled_index = None
def __init__(
self,
sequence_strs: List[str],
toks_per_batch: int,
crop_sizes: Tuple[int, int] = (512, 1024),
):
Sampler.__init__(self, data_source=None)
self._sequence_strs = sequence_strs
self._toks_per_batch = toks_per_batch
self._crop_sizes = crop_sizes
self._init_batches = get_batch_indices(
sequence_strs=sequence_strs,
toks_per_batch=toks_per_batch,
crop_sizes=crop_sizes,
)
def __init__(self, dataset: ModelNet40Dataset, use_potential=True, balance_labels=False):
Sampler.__init__(self, dataset)
self.use_potential = use_potential
self.balance_labels = balance_labels # 采样时是否平衡各类别
self.dataset = dataset # 需要采样处理的dataset,内存中没有复制
if self.use_potential:
# 大小是数据集点云的数目
self.potentials = np.random.rand(len(dataset.labels)) * 0.1 + 0.1
else:
self.potentials = None
# 每个batch所能包含的最多的点数
self.batch_limit = 10000
return
def __init__(
self,
sequence_strs: List[str],
toks_per_batch: int,
crop_sizes: Tuple[int, int] = (512, 1024),
num_replicas: Optional[int] = None,
rank: Optional[int] = None,
seed: int = 0,
):
Sampler.__init__(self, data_source=None)
# Replicate Torch Distributed Sampler logic
if num_replicas is None:
if not dist.is_available():
raise RuntimeError(
"Requires distributed package to be available")
num_replicas = dist.get_world_size()
if rank is None:
if not dist.is_available():
raise RuntimeError(
"Requires distributed package to be available")
rank = dist.get_rank()
if rank >= num_replicas or rank < 0:
raise ValueError("Invalid rank {}, rank should be in the interval"
" [0, {}]".format(rank, num_replicas - 1))
self._num_replicas = num_replicas
self._rank = rank
self._epoch = 0
self._seed = seed
self._sequence_strs = sequence_strs
self._toks_per_batch = toks_per_batch
self._crop_sizes = crop_sizes
self._init_batches = get_batch_indices(
sequence_strs=sequence_strs,
toks_per_batch=toks_per_batch,
crop_sizes=crop_sizes,
seed=self._seed + self._epoch,
)
self._num_samples = math.ceil(
len(self._init_batches) / self._num_replicas)
self._total_size = self._num_samples * self._num_replicas
def __init__(self, datasets: MultiTaskDataset, batch_size: int, shuffle=True):
super(MultiTaskBatchSampler, self).__init__(sampler=Sampler(datasets), batch_size=batch_size,
drop_last=False)
self.datasets_length = {task_id: len(dataset) for
task_id, dataset in datasets.datasets.items()}
self.shuffle = shuffle
self.batch_size = batch_size
self.task_indexes = []
self.batch_indexes = {}
self.task_num_batches = {}
self.total_batches = 0
self.init()
def get_pattern_distribution(patterns: Tuple[str, ...] = (
"checkerboard", )) -> "Sampler":
"""
Returns a pattern sampler.
Args:
patterns: A tuple of patterns to use.
Returns:
Sampler: Sampling from the patterns
"""
pattern_arrays = [PATTERNS[pattern].flatten() for pattern in patterns]
X = np.stack(pattern_arrays, axis=0)
pattern_dist = Sampler.from_samples(X)
pattern_dist.visualize = plot_image_samples([2, 2], False)
return pattern_dist
def __init__(self, dataset, *args, **kwargs):
Sampler.__init__(self, dataset)
self.init(*args, **kwargs)