[XLA] Better support for mul reductions in MakeFakeArguments()
authorMichael Kuperstein <mkuper@google.com>
Thu, 5 Apr 2018 21:54:36 +0000 (14:54 -0700)
committerTensorFlower Gardener <gardener@tensorflow.org>
Thu, 5 Apr 2018 21:56:57 +0000 (14:56 -0700)
commitb723cea002b3e612879291faaedb13eb702a5562
treec1ed2fd42837cc0507ecf70e7ea14d719faf7c88
parent051dd6cf2f805f12f6e4efb5ad91d57de88481fc
[XLA] Better support for mul reductions in MakeFakeArguments()

Mul reductions want a 1 as their init value, not a 0 or a random value.

PiperOrigin-RevId: 191802819
tensorflow/compiler/xla/tests/test_utils.cc