Re-enable contrib/quantize test when unfused mean and variance updates are used.

Fix missed change from "executing_eagerly" to "executing_eagerly_outside_functions" that was accidentally dropped in cl/300392015.

PiperOrigin-RevId: 300407526
Change-Id: Iaa88ac039a440b4e1081bc210a647de61cfad675
This commit is contained in:
A. Unique TensorFlower 2020-03-11 14:29:52 -07:00 committed by TensorFlower Gardener
parent a056967714
commit 37e7693c78

View File

@ -20,7 +20,6 @@ from __future__ import print_function
from tensorflow.python.compat import compat
from tensorflow.python.distribute import distribution_strategy_context
from tensorflow.python.eager import context
from tensorflow.python.framework import constant_op
from tensorflow.python.framework import dtypes
from tensorflow.python.framework import ops
@ -544,8 +543,9 @@ class BatchNormalizationBase(Layer):
# TODO(rmlarsen): Support using fused avg updates for non-eager execution
# after fixing graph pattern matching and enabling fused_batch_norm to
# take exponential_avg_factor as a tensor input.
use_fused_avg_updates = (compat.forward_compatible(2020, 3, 6) and
context.executing_eagerly())
use_fused_avg_updates = (
compat.forward_compatible(2020, 3, 6) and
ops.executing_eagerly_outside_functions())
if use_fused_avg_updates:
exponential_avg_factor = 1.0 - self.momentum
else: