experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Remove the optimizer v2 instance check, and replace with v1 object check.

Browse Source 
Optimizer_v2 instance doesn't share same base class with optimizer v1.

PiperOrigin-RevId: 302719205
Change-Id: I225c891e59e93d3b5386fa3682fb1bdc97e770de
This commit is contained in:
Scott Zhu 2020-03-24 12:14:00 -07:00 committed by TensorFlower Gardener
parent 7ae1992664
commit 8ed969c39d
1 changed files with 4 additions and 4 deletions

8
tensorflow/python/tpu/tpu_optimizer.py
View File

@ -21,7 +21,6 @@ from __future__ import print_function
from tensorflow.python.framework import ops from tensorflow.python.framework import ops
from tensorflow.python.keras.optimizer_v2 import optimizer_v2
from tensorflow.python.ops.losses import losses from tensorflow.python.ops.losses import losses
from tensorflow.python.platform import tf_logging as logging from tensorflow.python.platform import tf_logging as logging
from tensorflow.python.tpu import tpu_function from tensorflow.python.tpu import tpu_function
@ -55,10 +54,11 @@ class CrossShardOptimizer(optimizer.Optimizer):
""" """
if reduction not in (losses.Reduction.SUM, losses.Reduction.MEAN): if reduction not in (losses.Reduction.SUM, losses.Reduction.MEAN):
raise ValueError("Unsupported reduction: %s." % reduction) raise ValueError("Unsupported reduction: %s." % reduction)
if isinstance(opt, optimizer_v2.OptimizerV2): if not isinstance(opt, optimizer.Optimizer):
raise TypeError( raise TypeError(
"CrossShardOptimizer does not work with OptimizerV2. If you are " "CrossShardOptimizer only works with tf.training.Optimizer and not "
"using TPUStrategy, OptimizerV2 will sum gradients across replicas." "Optimizer_v2. If you are using TPUStrategy, OptimizerV2 will sum "
"gradients across replicas."
"If you are using TPUEstimator, you may instead sum your gradients " "If you are using TPUEstimator, you may instead sum your gradients "
"with: grads = [tf.compat.v1.tpu.cross_replica_sum(g) for g in grads]" "with: grads = [tf.compat.v1.tpu.cross_replica_sum(g) for g in grads]"
". If you want to average your gradients, rescale your loss with: " ". If you want to average your gradients, rescale your loss with: "