Attend the Women in ML Symposium on December 7 Register now


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

Experiments which can be enabled/disabled on a per version basis.

Experiments are designed to gradually apply changes to datasets while maintaining backward compatibility with previous versions. All experiments should eventually be deleted, once used by all versions of all datasets.


class Experiment(enum.Enum): EXP_A = # Short description of experiment.

class MyBuilder(...): VERSION = tfds.core.Version('1.2.3', experiments={ tfds.core.Experiment.EXP_A: True, })

DUMMY <Experiment.DUMMY: 1>