W3cubDocs

/TensorFlow 1.15

tf.train.experimental.LossScale

Loss scale base class.

Inherits From: CheckpointableBase

Loss scaling is a process that multiplies the loss by a multiplier called the loss scale, and divides each gradient by the same multiplier. The pseudocode for this process is:

loss = ...
loss *= loss_scale
grads = gradients(loss, vars)
grads /= loss_scale

Mathematically, loss scaling has no effect, but can help avoid numerical underflow in intermediate gradients when float16 tensors are used for mixed precision training. By multiplying the loss, each intermediate gradient will have the same multiplier applied.

Instances of this class represent a loss scale. Calling instances of this class returns the loss scale as a scalar float32 tensor, while method update() updates the loss scale depending on the values of the gradients. Optimizers use instances of this class to scale loss and gradients.

Methods

from_config

View source

Creates the LossScale from its config.

get_config

View source

Returns the config of this loss scale.

update

View source

Updates the value of the loss scale.

The loss scale will be potentially updated, based on the value of grads. The tensor returned by calling this class is only updated when this function is evaluated.

In eager mode, this directly updates the loss scale, so that calling __call__ will return the newly updated loss scale. In graph mode, this returns an op that, when evaluated, updates the loss scale.

This function also returns a should_apply_gradients bool. If False, gradients should not be applied to the variables that step, as nonfinite gradients were found, and the loss scale has been be updated to reduce the chance of finding nonfinite gradients in the next step. Some loss scale classes will always return True, as they cannot adjust themselves in response to nonfinite gradients.

When a DistributionStrategy is used, this function may only be called in a cross-replica context.

Args
grads A nested structure of unscaled gradients, each which is the gradient of the loss with respect to a weight. The gradients should have already been divided by the loss scale being before passed to this function. 'None' gradients are accepted, and are ignored.
Returns
update_op In eager mode, None. In graph mode, an op to update the loss scale.
should_apply_gradients Either a bool or a scalar boolean tensor. If False, the caller should skip applying grads to the variables this step.

__call__

View source

Returns the current loss scale as a scalar float32 tensor.

© 2020 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.
https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/train/experimental/LossScale