tf.math.add_n | TensorFlow v2.16.1 (original) (raw)
tf.math.add_n
Stay organized with collections Save and categorize content based on your preferences.
Returns the element-wise sum of a list of tensors.
View aliases
Main aliases
Compat aliases for migration
SeeMigration guide for more details.
tf.compat.v1.add_n, tf.compat.v1.math.add_n
tf.math.add_n(
inputs, name=None
)
Used in the notebooks
All inputs in the list must have the same shape. This op does notbroadcastits inputs. If you need broadcasting, use tf.math.add (or the +
operator) instead.
For example:
a = tf.constant([[3, 5], [4, 8]])
b = tf.constant([[1, 6], [2, 9]])
tf.math.add_n([a, b, a]).numpy()
array([[ 7, 16],
[10, 25]], dtype=int32)
See Also:
- tf.reduce_sum(inputs, axis=0) - This performs the same mathematical operation, but tf.add_n may be more efficient because it sums the tensors directly.
reduce_sum
on the other hand callstf.convert_to_tensor on the list of tensors, unnecessarily stacking them into a single tensor before summing.
Args | |
---|---|
inputs | A list of tf.Tensor or tf.IndexedSlices objects, each with the same shape and type. tf.IndexedSlices objects will be converted into dense tensors prior to adding. |
name | A name for the operation (optional). |
Returns |
---|
A tf.Tensor of the same shape and type as the elements of inputs. |
Raises | |
---|---|
ValueError | If inputs don't all have same shape and dtype or the shape cannot be inferred. |
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.