qxmt.datasets.tfds package

qxmt.datasets.tfds package#

Submodules#

Module contents#

class qxmt.datasets.tfds.TFDSDataLoader(name, split='train', save_path=None, return_format='numpy', data_dir=None, download=True, shuffle_files=False, flatten=True)#

Bases: object

This class loads a supervised dataset from TensorFlow Datasets (tensorflow/datasets) and converts it to numpy arrays.

Supported return formats: - numpy: return as a tuple of numpy arrays.

Supported save formats: - numpy: .npz, .npy

Parameters:

  • name (str)

  • split (str | Sequence[str])

  • save_path (str | Path | None)

  • return_format (str)

  • data_dir (str | Path | None)

  • download (bool)

  • shuffle_files (bool)

  • flatten (bool)

load()#

Load a supervised TensorFlow Dataset and return features and labels as numpy arrays.

Return type:

tuple[ndarray, ndarray]