Updates tests for integer division deprecation by mruberry · Pull Request #38621 · pytorch/pytorch

mruberry · 2020-05-17T09:48:37Z

Updates our tests in preparation of integer division using torch.div and torch.addcdiv throwing a runtime error by avoiding integer division using torch.div. This creates a brief period where integer division using torch.div is untested, but that should be OK (since it will soon throw a runtime error).

These callsites were identified using #36897.

dr-ci · 2020-05-17T10:12:15Z

💊 CI failures summary and remediations

As of commit b7d065c (more details on the Dr. CI page):

4/4 failures possibly* introduced in this PR
- 1/4 non-CircleCI failure(s)

🕵️ 3 new failures recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

pytorch_macos_10_13_py3_test (1/3)

Step: "Test" (full log | diagnosis details | 🔁 rerun)

May 19 12:12:05 AssertionError: False is not true : Tensors failed to compare as equal! Found 1 different element(s) (out of 9), with the greatest difference of 32768 (-32768 vs. 0) occuring at index 0.

May 19 12:12:05   File "/Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/testing/_internal/common_device_type.py", line 234, in instantiated_test 
May 19 12:12:05     result = test(self, device_arg, dtype) 
May 19 12:12:05   File "/Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/testing/_internal/common_device_type.py", line 501, in only_fn 
May 19 12:12:05     return fn(self, device, *args, **kwargs) 
May 19 12:12:05   File "test_torch.py", line 16632, in test_float_to_int_conversion_finite 
May 19 12:12:05     self._float_to_int_conversion_helper(vals, device, dtype) 
May 19 12:12:05   File "test_torch.py", line 16608, in _float_to_int_conversion_helper 
May 19 12:12:05     self.assertEqual(torch.from_numpy(a), t.cpu()) 
May 19 12:12:05   File "/Users/distiller/workspace/miniconda3/lib/python3.7/site-packages/torch/testing/_internal/common_utils.py", line 1083, in assertEqual 
May 19 12:12:05     self.assertTrue(result, msg=message) 
May 19 12:12:05 AssertionError: False is not true : Tensors failed to compare as equal! Found 1 different element(s) (out of 9), with the greatest difference of 32768 (-32768 vs. 0) occuring at index 0. 
May 19 12:12:05  
May 19 12:12:05 ---------------------------------------------------------------------- 
May 19 12:12:05 Ran 2305 tests in 118.089s 
May 19 12:12:05  
May 19 12:12:05 FAILED (failures=1, skipped=135) 
May 19 12:12:05  
May 19 12:12:05 Generating XML reports... 
May 19 12:12:05 Generated XML report: test-reports/python-unittest/TEST-TestTensorDeviceOpsCPU-20200519121007.xml 
May 19 12:12:05 Generated XML report: test-reports/python-unittest/TEST-TestTorch-20200519121007.xml 
May 19 12:12:05 Generated XML report: test-reports/python-unittest/TEST-TestTorchDeviceTypeCPU-20200519121007.xml

pytorch_linux_xenial_py3_6_gcc5_4_ge_config_simple_test (2/3)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

May 19 19:37:34 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" }

May 19 19:37:34 Traceback (most recent call last): 
May 19 19:37:34   File "test/run_test.py", line 690, in <module> 
May 19 19:37:34     main() 
May 19 19:37:34   File "test/run_test.py", line 683, in main 
May 19 19:37:34     raise RuntimeError(message) 
May 19 19:37:34 RuntimeError: test_type_hints failed! 
May 19 19:37:34 + cleanup 
May 19 19:37:34 + retcode=1 
May 19 19:37:34 + set +x 
May 19 19:37:34 =================== sccache compilation log =================== 
May 19 19:37:34 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" } 
May 19 19:37:34  
May 19 19:37:34 =========== If your build fails, please take a look at the log above for possible reasons =========== 
May 19 19:37:34 Compile requests                 64 
May 19 19:37:34 Compile requests executed        35 
May 19 19:37:34 Cache hits                       27 
May 19 19:37:34 Cache misses                      7 
May 19 19:37:34 Cache timeouts                    0 
May 19 19:37:34 Cache read errors                 0 
May 19 19:37:34 Forced recaches                   0 
May 19 19:37:34 Cache write errors                0

pytorch_linux_xenial_py3_6_gcc5_4_test (3/3)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

May 19 19:41:13 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" }

May 19 19:41:13 Traceback (most recent call last): 
May 19 19:41:13   File "test/run_test.py", line 690, in <module> 
May 19 19:41:13     main() 
May 19 19:41:13   File "test/run_test.py", line 683, in main 
May 19 19:41:13     raise RuntimeError(message) 
May 19 19:41:13 RuntimeError: test_type_hints failed! 
May 19 19:41:13 + cleanup 
May 19 19:41:13 + retcode=1 
May 19 19:41:13 + set +x 
May 19 19:41:13 =================== sccache compilation log =================== 
May 19 19:41:13 ERROR:sccache::server: Compilation failed: Output { status: ExitStatus(ExitStatus(256)), stdout: "", stderr: "/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp: In function \'int main()\':\n/var/lib/jenkins/.cache/torch_extensions/test_compilation_error_formatting/main.cpp:2:23: error: expected \';\' before \'}\' token\n int main() { return 0 }\n                       ^\n" } 
May 19 19:41:13  
May 19 19:41:13 =========== If your build fails, please take a look at the log above for possible reasons =========== 
May 19 19:41:13 Compile requests                 64 
May 19 19:41:13 Compile requests executed        35 
May 19 19:41:13 Cache hits                       27 
May 19 19:41:13 Cache misses                      7 
May 19 19:41:13 Cache timeouts                    0 
May 19 19:41:13 Cache read errors                 0 
May 19 19:41:13 Forced recaches                   0 
May 19 19:41:13 Cache write errors                0

1 failure confirmed as flaky and can be ignored:

pytorch_xla_linux_bionic_py3_6_clang9_build

ci.pytorch.org: 1 failed

Failed: pr/py3.6-clang7-rocmdeb-ubuntu16.04

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker.

See how this bot performed.

This comment has been revised 9 times.

facebook-github-bot

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

gchanan · 2020-05-19T17:31:24Z

            self.assertRaises(RuntimeError, lambda: op(s1, d2.view(d2.numel())[0].item()))

-    def _run_all_tests_for_sparse_op(self, op_name, device):
+    def _run_all_tests_for_sparse_op(self, op_name, device, dtypes=None):


nit: no default arguments in tests.

gchanan · 2020-05-19T17:33:11Z

        self.assertEqual(2 - torch.tensor(3), 2 - torch.tensor(3).to(devices[1]))    # __rsub__
        self.assertEqual(2 * torch.tensor(3), 2 * torch.tensor(3).to(devices[1]))    # __rmul__
-        self.assertEqual(2 / torch.tensor(3), 2 / torch.tensor(3).to(devices[1]))    # __rtruediv__
+        self.assertEqual(2 / torch.tensor(3.), 2 / torch.tensor(3.).to(devices[1]))  # __rtruediv__


for these that you changed from ints to floats, can you just check that there are places here that test floor_div on ints.

Yep (see the line just below this one, for example).

facebook-github-bot

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

facebook-github-bot · 2020-05-20T04:13:30Z

@mruberry merged this pull request in 6458457.

jeffdaily · 2020-05-20T20:08:31Z

@ezyang @mruberry This PR is breaking some ROCm CI tests. Looks like this PR changed a few tests in test_torch.py to use _float_types2 which added the bfloat16 when testing on ROCm.

mruberry · 2020-05-20T20:12:49Z

I see. That issue was probably concealed by other ROCm build failures. Would you submit a PR updating the types to not include bfloat16 on ROCm?

jeffdaily · 2020-05-20T20:20:56Z

@rohithkrn is working on a PR.

mruberry · 2020-05-20T20:22:42Z

Thank you, @jeffdaily! Sorry for the inconvenience.

Summary: This PR fixes the tolerance values for some of the bfloat16 div tests that were enabled on ROCm with incorrect tolerance values in the PR #38621 Also disabled(to unblock CI) `test_addcdiv*` for which the error is large when absolute values in the tensor are higher. This will have to be investigated further. ezyang jeffdaily sunway513 Pull Request resolved: #38823 Differential Revision: D21686290 Pulled By: ezyang fbshipit-source-id: 85472680e1886bdc7c227ed2656e0b4fd5328e46

Summary: Updates our tests in preparation of integer division using torch.div and torch.addcdiv throwing a runtime error by avoiding integer division using torch.div. This creates a brief period where integer division using torch.div is untested, but that should be OK (since it will soon throw a runtime error). These callsites were identified using pytorch#36897. Pull Request resolved: pytorch#38621 Differential Revision: D21612823 Pulled By: mruberry fbshipit-source-id: 749c03a69feae02590b4395335163d9bf047e162

…8823) Summary: This PR fixes the tolerance values for some of the bfloat16 div tests that were enabled on ROCm with incorrect tolerance values in the PR pytorch#38621 Also disabled(to unblock CI) `test_addcdiv*` for which the error is large when absolute values in the tensor are higher. This will have to be investigated further. ezyang jeffdaily sunway513 Pull Request resolved: pytorch#38823 Differential Revision: D21686290 Pulled By: ezyang fbshipit-source-id: 85472680e1886bdc7c227ed2656e0b4fd5328e46

Updates tests for integer division deprecation

c0dc29e

mruberry added module: numpy Related to numpy support, and also numpy compatibility of our operators and removed module: numpy Related to numpy support, and also numpy compatibility of our operators labels May 17, 2020

facebook-github-bot reviewed May 17, 2020

View reviewed changes

mruberry requested a review from gchanan May 18, 2020 18:05

gchanan reviewed May 19, 2020

View reviewed changes

gchanan approved these changes May 19, 2020

View reviewed changes

Removes default test argument

b7d065c

facebook-github-bot reviewed May 19, 2020

View reviewed changes

facebook-github-bot closed this in 6458457 May 20, 2020

facebook-github-bot added the merged label May 20, 2020

mruberry deleted the integer_div_test_callsites branch May 20, 2020 20:22

rohithkrn mentioned this pull request May 20, 2020

[ROCm] Set correct tolerance values for bfloat16 div tests #38823

Closed

mruberry added the Merged label Oct 28, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Updates tests for integer division deprecation#38621

Updates tests for integer division deprecation#38621
mruberry wants to merge 2 commits intomasterfrom
integer_div_test_callsites

mruberry commented May 17, 2020

Uh oh!

dr-ci Bot commented May 17, 2020 •

edited

Loading

Uh oh!

facebook-github-bot left a comment

Uh oh!

gchanan May 19, 2020 •

edited

Loading

Uh oh!

gchanan May 19, 2020

Uh oh!

mruberry May 19, 2020

Uh oh!

facebook-github-bot left a comment

Uh oh!

facebook-github-bot commented May 20, 2020

Uh oh!

jeffdaily commented May 20, 2020

Uh oh!

mruberry commented May 20, 2020

Uh oh!

jeffdaily commented May 20, 2020

Uh oh!

mruberry commented May 20, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

mruberry commented May 17, 2020

Uh oh!

dr-ci Bot commented May 17, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

💊 CI failures summary and remediations

🕵️ 3 new failures recognized by patterns

pytorch_macos_10_13_py3_test (1/3)

pytorch_linux_xenial_py3_6_gcc5_4_ge_config_simple_test (2/3)

pytorch_linux_xenial_py3_6_gcc5_4_test (3/3)

ci.pytorch.org: 1 failed

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

gchanan May 19, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

gchanan May 19, 2020

Choose a reason for hiding this comment

Uh oh!

mruberry May 19, 2020

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented May 20, 2020

Uh oh!

jeffdaily commented May 20, 2020

Uh oh!

mruberry commented May 20, 2020

Uh oh!

jeffdaily commented May 20, 2020

Uh oh!

mruberry commented May 20, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

dr-ci Bot commented May 17, 2020 •

edited

Loading

gchanan May 19, 2020 •

edited

Loading