RandomSampler

class braintools.trainer.RandomSampler(data_source, replacement=False, num_samples=None, seed=None)

Samples elements randomly.

Parameters:

• data_source (Dataset) – Dataset to sample from.

• replacement (bool) – Whether to sample with replacement.

• num_samples (int | None) – Number of samples to draw. Default is len(data_source).

• seed (int | None) – Random seed for reproducibility.

Examples

>>> sampler = RandomSampler(dataset, seed=42)
>>> indices = list(sampler)  # Random permutation

reset(seed=None)

Reset the random state.