nexuml.data.export.runner¶
Loader-driven dataset export utilities.
export_data_module(
data_module: NexuDataModule,
path: Path,
*,
backend: str | WriterSpec = "numpy",
splits: Sequence[str] | None = None,
transform: BatchTransform | None = None,
x_keys: Sequence[str] | None = None,
y_keys: Sequence[str] | None = None,
include_labels: bool = True,
label_prefix: str = "label__",
dtype: object | None = None,
device: device | str | None = None,
**backend_kwargs: object,
) -> Path
export_data_module(
data_module: NexuDataModule,
path: str,
*,
backend: str | WriterSpec = "numpy",
splits: Sequence[str] | None = None,
transform: BatchTransform | None = None,
x_keys: Sequence[str] | None = None,
y_keys: Sequence[str] | None = None,
include_labels: bool = True,
label_prefix: str = "label__",
dtype: object | None = None,
device: device | str | None = None,
**backend_kwargs: object,
) -> Path | str
export_data_module(
data_module: NexuDataModule,
path: str | Path,
*,
backend: str | WriterSpec = "numpy",
splits: Sequence[str] | None = None,
transform: BatchTransform | None = None,
x_keys: Sequence[str] | None = None,
y_keys: Sequence[str] | None = None,
include_labels: bool = True,
label_prefix: str = "label__",
dtype: object | None = None,
device: device | str | None = None,
**backend_kwargs: object,
) -> Path | str
Export the data exactly as seen through a configured data module.
Returns:
| Type | Description |
|---|---|
Path | str
|
Local export path, or the original S3 URI for a remote WebDataset export. |
Raises:
| Type | Description |
|---|---|
ValueError
|
If there are no samples or an unsupported remote backend is requested. |