Rand function for complex dtype #34924

kostekIV · 2020-03-17T22:28:06Z

kostekIV · 2020-03-18T06:42:12Z

Hi @anjali411 can you take look at this? I can't edit reviewers, also not sure why ci stopped at only 6 checks. Side note didn't run into issue with view_complex_as_float in header file so I didn't make it inline.

dr-ci · 2020-03-18T06:58:05Z

💊 CircleCI build failures summary and remediations

As of commit c8e8e74 (more details on the Dr. CI page):

2/4 failures introduced in this PR
2/4 broken upstream at merge base 95ad94c from Mar 21 until Mar 23 (51 commits; 62f11f0 - 0e0386b)
Please rebase on the viable/strict branch (expand for instructions)

If your commit is newer than viable/strict, you can try basing on an older, stable commit:
```
git fetch https://github.com/pytorch/pytorch viable/strict
git rebase --onto FETCH_HEAD $(git merge-base origin/master HEAD)
```
If your commit is older than viable/strict:
```
git fetch https://github.com/pytorch/pytorch viable/strict
git rebase FETCH_HEAD
```
Check out the recency history of this "viable master" tracking branch.

🕵️ 2 new failures recognized by patterns

The following build failures do not appear to be due to upstream breakages (reran 1 job to discount flakiness):

pytorch_xla_linux_xenial_py3_6_clang7_build (1/2)

Step: "Build" (full log | pattern match details) <confirmed not flaky by 2 failures>

Mar 23 17:50:18 torch_xla/csrc/aten_xla_type_default.cpp:11460:8: error: no matching member function for call to 'impl_unboxedOnlyKernel'

Mar 23 17:50:18       ~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 
Mar 23 17:50:18 /opt/conda/lib/python3.6/site-packages/torch/include/ATen/core/op_registration/op_registration.h:294:74: note: candidate template ignored: invalid explicitly-specified argument for template parameter 'kernel_func' 
Mar 23 17:50:18     std::enable_if_t<guts::is_function_type<FuncType>::value, Options&&> impl_unboxedOnlyKernel(DispatchKey dispatch_key) && { 
Mar 23 17:50:18                                                                          ^ 
Mar 23 17:50:18 torch_xla/csrc/aten_xla_type_default.cpp:11199:8: error: no matching member function for call to 'impl_unboxedOnlyKernel' 
Mar 23 17:50:18       .impl_unboxedOnlyKernel<at::Tensor &(at::Tensor &, double, double, at::Generator *), &AtenXlaType::normal_>(at::DispatchKey::XLATensorId) 
Mar 23 17:50:18       ~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 
Mar 23 17:50:18 /opt/conda/lib/python3.6/site-packages/torch/include/ATen/core/op_registration/op_registration.h:294:74: note: candidate template ignored: invalid explicitly-specified argument for template parameter 'kernel_func' 
Mar 23 17:50:18     std::enable_if_t<guts::is_function_type<FuncType>::value, Options&&> impl_unboxedOnlyKernel(DispatchKey dispatch_key) && { 
Mar 23 17:50:18                                                                          ^ 
Mar 23 17:50:18 torch_xla/csrc/aten_xla_type_default.cpp:11460:8: error: no matching member function for call to 'impl_unboxedOnlyKernel' 
Mar 23 17:50:18       .impl_unboxedOnlyKernel<at::Tensor(const at::Tensor &, const at::Tensor &, at::Scalar, at::Scalar, bool, at::Generator *), &AtenXlaType::rrelu_with_noise>(at::DispatchKey::XLATensorId) 
Mar 23 17:50:18       ~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 
Mar 23 17:50:18 /opt/conda/lib/python3.6/site-packages/torch/include/ATen/core/op_registration/op_registration.h:294:74: note: candidate template ignored: invalid explicitly-specified argument for template parameter 'kernel_func' 
Mar 23 17:50:18     std::enable_if_t<guts::is_function_type<FuncType>::value, Options&&> impl_unboxedOnlyKernel(DispatchKey dispatch_key) && { 
Mar 23 17:50:18                                                                          ^ 
ackages/torch/include/torch/csrc/api/include -I/opt/conda/lib/python3.6/site-packages/torch/include/TH -I/opt/conda/lib/python3.6/site-packages/torch/include/THC -I/opt/conda/include/python3.6m -c torch_xla/csrc/batch_norm.cpp -o build/temp.linux-x86_64-3.6/torch_xla/csrc/batch_norm.o -std=c++14 -Wno-sign-compare -Wno-deprecated-declarations -Wno-return-type -Wno-macro-redefined -Wno-return-std-move -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H -DTORCH_EXTENSION_NAME=_XLAC -D_GLIBCXX_USE_CXX11_ABI=1 
Mar 23 17:50:26 11 errors generated. 
torch/include/torch/csrc/api/include -I/opt/conda/lib/python3.6/site-packages/torch/include/TH -I/opt/conda/lib/python3.6/site-packages/torch/include/THC -I/opt/conda/include/python3.6m -c torch_xla/csrc/layout_manager.cpp -o build/temp.linux-x86_64-3.6/torch_xla/csrc/layout_manager.o -std=c++14 -Wno-sign-compare -Wno-deprecated-declarations -Wno-return-type -Wno-macro-redefined -Wno-return-std-move -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H -DTORCH_EXTENSION_NAME=_XLAC -D_GLIBCXX_USE_CXX11_ABI=1 
on3.6/site-packages/torch/include/torch/csrc/api/include -I/opt/conda/lib/python3.6/site-packages/torch/include/TH -I/opt/conda/lib/python3.6/site-packages/torch/include/THC -I/opt/conda/include/python3.6m -c torch_xla/csrc/view.cpp -o build/temp.linux-x86_64-3.6/torch_xla/csrc/view.o -std=c++14 -Wno-sign-compare -Wno-deprecated-declarations -Wno-return-type -Wno-macro-redefined -Wno-return-std-move -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H -DTORCH_EXTENSION_NAME=_XLAC -D_GLIBCXX_USE_CXX11_ABI=1 
te-packages/torch/include/torch/csrc/api/include -I/opt/conda/lib/python3.6/site-packages/torch/include/TH -I/opt/conda/lib/python3.6/site-packages/torch/include/THC -I/opt/conda/include/python3.6m -c torch_xla/csrc/nll_loss.cpp -o build/temp.linux-x86_64-3.6/torch_xla/csrc/nll_loss.o -std=c++14 -Wno-sign-compare -Wno-deprecated-declarations -Wno-return-type -Wno-macro-redefined -Wno-return-std-move -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H -DTORCH_EXTENSION_NAME=_XLAC -D_GLIBCXX_USE_CXX11_ABI=1

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test (2/2)

Step: "Test" (full log | pattern match details) <confirmed not flaky by 2 failures>

Mar 23 19:59:46 AssertionError: 11 not less than or equal to 1e-05 :

Mar 23 19:59:46 ---------------------------------------------------------------------- 
Mar 23 19:59:46 Traceback (most recent call last): 
Mar 23 19:59:46   File "/opt/conda/lib/python3.6/site-packages/torch/testing/_internal/common_distributed.py", line 175, in wrapper 
Mar 23 19:59:46     self._join_processes(fn) 
Mar 23 19:59:46   File "/opt/conda/lib/python3.6/site-packages/torch/testing/_internal/common_distributed.py", line 285, in _join_processes 
Mar 23 19:59:46     self._check_return_codes(elapsed_time) 
Mar 23 19:59:46   File "/opt/conda/lib/python3.6/site-packages/torch/testing/_internal/common_distributed.py", line 324, in _check_return_codes 
Mar 23 19:59:46     self.assertEqual(p.exitcode, first_process.exitcode) 
Mar 23 19:59:46   File "/opt/conda/lib/python3.6/site-packages/torch/testing/_internal/common_utils.py", line 915, in assertEqual 
Mar 23 19:59:46     super(TestCase, self).assertLessEqual(abs(x - y), prec, message) 
Mar 23 19:59:46 AssertionError: 11 not less than or equal to 1e-05 :  
Mar 23 19:59:46  
Mar 23 19:59:46 ---------------------------------------------------------------------- 
Mar 23 19:59:46 Ran 61 tests in 65.284s 
Mar 23 19:59:46  
Mar 23 19:59:46 FAILED (failures=1, skipped=1) 
Mar 23 19:59:46  
Mar 23 19:59:46 Generating XML reports... 
Mar 23 19:59:46 Traceback (most recent call last): 
Mar 23 19:59:46   File "test/run_test.py", line 682, in <module> 
Mar 23 19:59:46     main()

🚧 2 upstream failures:

These were probably caused by upstream breakages:

pytorch_windows_vs2019_py36_cuda10.1_test2 from Mar 21 until Mar 23 (51 commits; 62f11f0 - 0e0386b)
pytorch_linux_xenial_py3_clang5_asan_test from Mar 21 until Mar 23 (51 commits; 62f11f0 - 0e0386b)

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.

This comment has been revised 11 times.

anjali411 · 2020-03-20T01:28:27Z

Hi @kostekIV thanks for working on this! :D I wanted to have randn work for complex tensors on cuda so I added a header here #35056. Once this PR is landed, do you mind updating your PR and also adding rand support for complex on cuda?

kostekIV · 2020-03-20T12:57:04Z

Sure @anjali411 I will do that. Correct me if I am wrong but doesn't rand function in aten/src/ATen/native/TensorFactories.cpp already is for both cpu and cuda, so the change in this pr (after syncing it with #35056) is sufficient for it to work on both cpu and cuda?

anjali411 · 2020-03-20T14:51:42Z

Sure @anjali411 I will do that. Correct me if I am wrong but doesn't rand function in aten/src/ATen/native/TensorFactories.cpp already is for both cpu and cuda, so the change in this pr (after syncing it with #35056) is sufficient for it to work on both cpu and cuda?

oh yeah you are right! randn calls into uniform_ which has separate cpu and cuda calls so we don't need to worry about it. This PR looks good overall. ~~I'll let you know once I land~~ #35056 landed. once you update the PR, we can land it. thanks for working on this!! :D

kostekIV · 2020-03-23T08:55:12Z

@anjali411 updated

facebook-github-bot

@anjali411 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

anjali411

LGTM :) thanks for the awesome work!

anjali411 · 2020-03-23T14:49:54Z

aten/src/ATen/native/TensorFactories.cpp

+  if (result.is_complex()) {
+    auto float_tensor = at::native::view_complex_as_float(result);
+    float_tensor.uniform_(0, 1, generator);
+    // we want to return complex tensor not underlaying float tensor.


nit: typo- not underlying

oops, fixed

facebook-github-bot

@anjali411 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

anjali411 · 2020-03-23T22:21:01Z

@kostekIV two tests seem to be failing due to this PR. can you look into it?

kostekIV · 2020-03-25T14:28:28Z

@anjali411 I am honestly at lost here, I was looking for an error and couldn't find what is causing those failures here. I looked into other prs and few of the already merged ones had this 2 failures, for example this had torch_xla/csrc/aten_xla_type_default.cpp:11460:8: error: no matching member function for call to 'impl_unboxedOnlyKernel' and this had AssertionError: 11 not less than or equal to 1e-05. Not sure how to approach this. Thank you for your help :)

facebook-github-bot · 2020-03-26T16:16:19Z

@anjali411 merged this pull request in ada4077.

pytorchbot added the open source label Mar 17, 2020

anjali411 self-requested a review March 20, 2020 01:25

yf225 added module: complex Related to complex number support in PyTorch triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module labels Mar 20, 2020

kostekIV added 2 commits March 21, 2020 11:41

Extract view_complex_as_float to header and apply to rand function.

dcd5e17

Delete duplicated ComplexHelper files.

28e7fe0

facebook-github-bot reviewed Mar 23, 2020

View reviewed changes

anjali411 approved these changes Mar 23, 2020

View reviewed changes

anjali411 changed the title ~~Extract view_complex_as_float to header and apply to rand function.~~ Rand function for complex dtype Mar 23, 2020

anjali411 reviewed Mar 23, 2020

View reviewed changes

Fix typo

c8e8e74

facebook-github-bot reviewed Mar 23, 2020

View reviewed changes

facebook-github-bot closed this in ada4077 Mar 26, 2020

facebook-github-bot added the merged label Mar 26, 2020

kostekIV deleted the complex-rand branch March 26, 2020 17:06

mruberry added the Merged label Oct 28, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Rand function for complex dtype #34924

Rand function for complex dtype #34924

kostekIV commented Mar 17, 2020

kostekIV commented Mar 18, 2020

dr-ci bot commented Mar 18, 2020 •

edited

anjali411 commented Mar 20, 2020

kostekIV commented Mar 20, 2020

anjali411 commented Mar 20, 2020 •

edited

kostekIV commented Mar 23, 2020

facebook-github-bot left a comment

anjali411 left a comment

anjali411 Mar 23, 2020

kostekIV Mar 23, 2020

facebook-github-bot left a comment

anjali411 commented Mar 23, 2020

kostekIV commented Mar 25, 2020

facebook-github-bot commented Mar 26, 2020

Rand function for complex dtype #34924

Rand function for complex dtype #34924

Conversation

kostekIV commented Mar 17, 2020

kostekIV commented Mar 18, 2020

dr-ci bot commented Mar 18, 2020 • edited

💊 CircleCI build failures summary and remediations

🕵️ 2 new failures recognized by patterns

pytorch_xla_linux_xenial_py3_6_clang7_build (1/2)

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test (2/2)

🚧 2 upstream failures:

anjali411 commented Mar 20, 2020

kostekIV commented Mar 20, 2020

anjali411 commented Mar 20, 2020 • edited

kostekIV commented Mar 23, 2020

facebook-github-bot left a comment

Choose a reason for hiding this comment

anjali411 left a comment

Choose a reason for hiding this comment

anjali411 Mar 23, 2020

Choose a reason for hiding this comment

kostekIV Mar 23, 2020

Choose a reason for hiding this comment

facebook-github-bot left a comment

Choose a reason for hiding this comment

anjali411 commented Mar 23, 2020

kostekIV commented Mar 25, 2020

facebook-github-bot commented Mar 26, 2020

dr-ci bot commented Mar 18, 2020 •

edited

anjali411 commented Mar 20, 2020 •

edited