Type DynamicLossScale
Namespace tensorflow.train.experimental
Parent LossScale
Interfaces IDynamicLossScale
Loss scale that dynamically adjusts itself. Dynamic loss scaling works by adjusting the loss scale as training progresses.
The goal is to keep the loss scale as high as possible without overflowing the
gradients. As long as the gradients do not overflow, raising the loss scale
never hurts. The algorithm starts by setting the loss scale to an initial value. Every N
steps that the gradients are finite, the loss scale is increased by some
factor. However, if a NaN or Inf gradient is found, the gradients for that
step are not applied, and the loss scale is decreased by the factor. This
process tends to keep the loss scale as high as possible without gradients
overflowing.