Parse and evaluate a standard result file. A standard result file is named as: {flag}{size_flag}_{split}|*|.csv (|*| can be anyth

Return a dummy prediction of correct shape.

Save the dataset as standard figures, which could be used for AutoML tools, e.g., Google AutoML Vision.

List all available datasets.

Download all available datasets.

Accuracy metric. :param y_true: the ground truth labels, shape: (n_samples, n_labels) or (n_samples,) if n_labels==1 :param y_score: the predi

AUC metric. :param y_true: the ground truth labels, shape: (n_samples, n_labels) or (n_samples,) if n_labels==1 :param y_score: the predicted

Print the dataset details given a subset flag.

return: (without transform/target_transofrm) img: PIL.Image target: np.array of `L` (L=1 for single-label)

return: (without transform/target_transofrm) img: an array of 1x28x28x28 or 3x28x28x28 (if `as_RGB=True`), in [0,1] t

return: (without transform/target_transofrm) img: PIL.Image target: np.array of `L` (L=1 for single-label)

return: (without transform/target_transofrm) img: an array of 1x28x28x28 or 3x28x28x28 (if `as_RGB=True`), in [0,1] t

Args: split (str, required): 'train', 'val' or 'test' transform (callable, optional): data transformation

Args: split (string): 'train', 'val' or 'test', required transform (callable, optional): A function/transform that t

Adapted from torchvision.ss

Adapted from torchvision.

Delete all downloaded npz from root.

Parse and evaluate a standard result file. A standard result file is named as: {flag}{size_flag}_{split}|*|.csv (|*| can be anything)

Save ground truth and scores :param y_true: the ground truth labels, shape: (n_samples, n_classes) for multi-label, and (n_samples,) for other tas

For developmemnt only.