tfds.core.BuilderConfig  |  TensorFlow Datasets (original) (raw)

tfds.core.BuilderConfig

Stay organized with collections Save and categorize content based on your preferences.

Base class for DatasetBuilder data configuration.

tfds.core.BuilderConfig(
    name: str,
    version: Optional[VersionOrStr] = None,
    release_notes: Optional[Dict[str, str]] = None,
    supported_versions: List[VersionOrStr] = dataclasses.field(default_factory=list),
    description: Optional[str] = None,
    tags: List[str] = dataclasses.field(default_factory=list)
)

DatasetBuilder subclasses with data configuration options should subclassBuilderConfig and add their own properties.

Attributes
name The name of the config.
version The version of the config.
release_notes A dictionary associating versions to changes.
supported_versions A list of versions which this Builder Config supports.
description a human description of the config.
tags [Experimental] a list of freeform tags applying to the config. This is not used by TFDS, but can be retrieved later from a ConfigBuilder instance.

Methods

from_dataset_info

View source

@classmethod from_dataset_info( info_proto: dataset_info_pb2.DatasetInfo ) -> Optional['BuilderConfig']

Instantiates a BuilderConfig from the given proto.

Args
info_proto DatasetInfo proto which documents the requested dataset config, including its name, version, and features.
Returns
A BuilderConfig for the requested config.
Class Variables
description None
release_notes None
version None

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2024-04-26 UTC.