Use integral power function rather than floating point version. The integral version...
authorA. Unique TensorFlower <gardener@tensorflow.org>
Thu, 17 May 2018 18:13:00 +0000 (11:13 -0700)
committerTensorFlower Gardener <gardener@tensorflow.org>
Thu, 17 May 2018 18:16:01 +0000 (11:16 -0700)
commit9e2ce8f4c483e68309a60dc89739bb1b79b4a12e
tree3115e1fdacc23de00326f97b77ea5fa1e7121461
parent0f8be44de22a344ce6aac1e2cee8595b7c89d9f8
Use integral power function rather than floating point version. The integral version is faster.

PiperOrigin-RevId: 197021020
tensorflow/core/grappler/optimizers/memory_optimizer.cc