TorchActiveLearningPool¶

class probly.evaluation.active_learning.pool.torch.TorchActiveLearningPool(x_labeled: Tensor, y_labeled: Tensor, x_unlabeled: Tensor, y_unlabeled: Tensor, x_test: Tensor, y_test: Tensor)[source]¶

Bases: object

Active learning pool backed by PyTorch tensors.

Variables:

x_labeled (torch.Tensor) – Feature tensor for the currently labeled training samples.
y_labeled (torch.Tensor) – Label tensor for the currently labeled training samples.
x_unlabeled (torch.Tensor) – Feature tensor for the unlabeled pool.
y_unlabeled (torch.Tensor) – Ground-truth label tensor for the unlabeled pool (revealed only on query).
x_test (torch.Tensor) – Feature tensor for the held-out test set.
y_test (torch.Tensor) – Label tensor for the held-out test set.

property n_labeled: int¶: Number of currently labeled training samples.

property n_unlabeled: int¶: Number of samples remaining in the unlabeled pool.

x_labeled: Tensor¶

x_test: Tensor¶

x_unlabeled: Tensor¶

y_labeled: Tensor¶

y_test: Tensor¶

y_unlabeled: Tensor¶

Examples using `probly.evaluation.active_learning.pool.torch.TorchActiveLearningPool`¶

Active Learning with PyTorch - BADGE Selection

TorchActiveLearningPool¶

Examples using probly.evaluation.active_learning.pool.torch.TorchActiveLearningPool¶

Examples using `probly.evaluation.active_learning.pool.torch.TorchActiveLearningPool`¶