UT //tensorflow/python:saver_test failed

[candyzone]

System information

• Have I written custom code (as opposed to using a stock example script provided in TensorFlow):N
• OS Platform and Distribution (e.g., Linux Ubuntu 16.04):Linux Ubuntu 16.04
• Mobile device (e.g. iPhone 8, Pixel 2, Samsung Galaxy) if the issue happens on mobile device:N
• TensorFlow installed from (source or binary):source
• TensorFlow version (use command below):1.15
• Python version:python3.6
• Bazel version (if compiling from source):0.26.1
• GCC/Compiler version (if compiling from source):gcc7.5.0
• CUDA/cuDNN version:-
• GPU model and memory:-

Describe the current behavior

bazel test --cxxopt=-D_GLIBCXX_USE_CXX11_ABI=0 --copt="-march=native" //tensorflow/python:saver_test
FAILED

Traceback (most recent call last):
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/training/saver_test.py", line 2419, in testClearDevicesOnExport
10, size=[1, 10])
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 956, in run
run_metadata_ptr)
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 1165, in _run
self._graph, fetches, feed_dict_tensor, feed_handles=feed_handles)
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 474, in init
self._fetch_mapper = _FetchMapper.for_fetch(fetches)
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 266, in for_fetch
return _ListFetchMapper(fetch)
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 375, in init
self._mappers = [_FetchMapper.for_fetch(fetch) for fetch in fetches]
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 375, in
self._mappers = [_FetchMapper.for_fetch(fetch) for fetch in fetches]
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 276, in for_fetch
return _ElementFetchMapper(fetches, contraction_fn)
File "/home/admin/chen.ding/gitlab/code/tensorflow_2021/bazel-bin/tensorflow/python/saver_test.runfiles/org_tensorflow/tensorflow/python/client/session.py", line 315, in init
'Tensor. (%s)' % (fetch, str(e)))
ValueError: Fetch argument 'new_model/optimize' cannot be interpreted as a Tensor. ("The name 'new_model/optimize' refers to an Operation not in the graph.")

Describe the expected behavior

UT pass

root casse
https://github.com/NVIDIA/tensorflow/blob/r1.15.5%2Bnv21.05/tensorflow/python/training/optimizer.py#L659
It changes NoOp name here(suffix with '-apply') , but without change in 'saver_test‘ test case

[candyzone]

Why NV change NoOp name with suffix “-apply“ in TensorFlow source code? Any other reason?