tf.raw_ops.XlaSparseCoreFtrl

View aliases

Compat aliases for migration

tf.compat.v1.raw_ops.XlaSparseCoreFtrl

tf.raw_ops.XlaSparseCoreFtrl(
    embedding_table,
    accumulator,
    linear,
    learning_rate,
    indices,
    gradient,
    beta,
    learning_rate_power,
    l2_regularization_strength,
    feature_width,
    multiply_linear_by_learning_rate,
    l1_regularization_strength,
    name=None
)

Args
`embedding_table`	A `Tensor` of type `float32`.
`accumulator`	A `Tensor` of type `float32`.
`linear`	A `Tensor` of type `float32`.
`learning_rate`	A `Tensor` of type `float32`.
`indices`	A `Tensor` of type `int32`.
`gradient`	A `Tensor` of type `float32`.
`beta`	A `Tensor` of type `float32`.
`learning_rate_power`	A `Tensor` of type `float32`.
`l2_regularization_strength`	A `Tensor` of type `float32`.
`feature_width`	An `int`.
`multiply_linear_by_learning_rate`	A `bool`.
`l1_regularization_strength`	A `float`.
`name`	A name for the operation (optional).

Returns
A tuple of `Tensor` objects (updated_embedding_table, updated_accumulator, updated_linear).
`updated_embedding_table`	A `Tensor` of type `float32`.
`updated_accumulator`	A `Tensor` of type `float32`.
`updated_linear`	A `Tensor` of type `float32`.

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. Some content is licensed under the numpy license.

Last updated 2024-04-26 UTC.

tf.raw_ops.XlaSparseCoreFtrl

View aliases

Args

Returns