TFDS now supports the Croissant 🥐 format! Read the documentation to know more.

tfds.features.Encoding

For higher dimension tensors, it is recommended to define the encoding as zlib or bytes to save space on disk.

Attributes
`NONE`	No compression (default). bools/integers will be upcasted to int64 as this is the only integer format supported by the `tf.train.Example` protobufs in which examples are saved.
`BYTES`	Stored as raw bytes (avoid the upcasting from above).
`ZLIB`	The raw bytes are compressed using zlib.

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.

Attributes