Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Support symbolic for conv_tbc (#58359) #58577

Closed
wants to merge 1 commit into from
Closed

Support symbolic for conv_tbc (#58359) #58577

wants to merge 1 commit into from

Conversation

BowenBao
Copy link
Collaborator

@BowenBao BowenBao commented May 19, 2021 •
edited
Loading

Stack from ghstack:

This is a fix for exporting fairseq models, see:

model = torch.hub.load(github, 'conv.wmt14.en-fr', tokenizer='moses', bpe='subword_nmt')
model = torch.hub.load(github, 'conv.wmt17.en-de', tokenizer='moses', bpe='subword_nmt')

With this fix, and comment out model script one line GradMultiply, these two models can be exported successfully with perf met.

The original PR #57708 has merging issue, use this one instead.

Co-authored-by: David jiafa@microsoft.com

@jiafatom @BowenBao
Support symbolic for conv_tbc (#58359)
7040e55 
This is a fix for exporting fairseq models, see:
```python
model = torch.hub.load(github, 'conv.wmt14.en-fr', tokenizer='moses', bpe='subword_nmt')
model = torch.hub.load(github, 'conv.wmt17.en-de', tokenizer='moses', bpe='subword_nmt')
```
With this fix, and comment out model script one line `GradMultiply`, these two models can be exported successfully with perf met.

The original PR #57708 has merging issue, use this one instead.

Co-authored-by: David <jiafa@microsoft.com>

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

facebook-github-bot commented May 19, 2021 •
edited
Loading

💊 CI failures summary and remediations

As of commit 7040e55 (more details on the Dr. CI page):

  • 3/4 failures introduced in this PR
  • 1/4 broken upstream at merge base 352ab7f on May 19 from 8:24am to 2:23pm

🕵️ 1 new failure recognized by patterns

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

See CircleCI build pytorch_linux_xenial_py3_clang7_onnx_ort_test1 (1/1)

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

May 19 20:55:31 AssertionError: 
May 19 20:55:31   File "/var/lib/jenkins/workspace/test/onnx/test_pytorch_onnx_onnxruntime.py", line 92, in ort_compare_with_pytorch
May 19 20:55:31     [np.testing.assert_allclose(out, ort_out, rtol=rtol, atol=atol) for out, ort_out in zip(outputs, ort_outs)]
May 19 20:55:31   File "/var/lib/jenkins/workspace/test/onnx/test_pytorch_onnx_onnxruntime.py", line 92, in <listcomp>
May 19 20:55:31     [np.testing.assert_allclose(out, ort_out, rtol=rtol, atol=atol) for out, ort_out in zip(outputs, ort_outs)]
May 19 20:55:31   File "/opt/conda/lib/python3.6/site-packages/numpy/testing/_private/utils.py", line 1533, in assert_allclose
May 19 20:55:31     verbose=verbose, header=header, equal_nan=equal_nan)
May 19 20:55:31   File "/opt/conda/lib/python3.6/site-packages/numpy/testing/_private/utils.py", line 770, in assert_array_compare
May 19 20:55:31     flagged = func_assert_same_pos(x, y, func=isnan, hasval='nan')
May 19 20:55:31   File "/opt/conda/lib/python3.6/site-packages/numpy/testing/_private/utils.py", line 746, in func_assert_same_pos
May 19 20:55:31     raise AssertionError(msg)
May 19 20:55:31 AssertionError: 
May 19 20:55:31 Not equal to tolerance rtol=0.001, atol=1e-07
May 19 20:55:31 
May 19 20:55:31 x and y nan location mismatch:
May 19 20:55:31  x: array([[[          inf,           nan,           nan, -1.719250e+38,
May 19 20:55:31                    nan],
May 19 20:55:31         [          nan,           nan,           nan,           inf,...
May 19 20:55:31  y: array([[[          inf,           nan,           nan, -1.719250e+38,
May 19 20:55:31                    nan],
May 19 20:55:31         [          inf,           nan,           nan,           inf,...
May 19 20:55:31 =============================== warnings summary ===============================

2 failures not recognized by patterns:

Job Step Action
CircleCI pytorch_linux_bionic_cuda10_2_cudnn7_py3_9_gcc7_test1 Run tests 🔁 rerun
CircleCI pytorch_linux_bionic_py3_8_gcc9_coverage_test1 Report results 🔁 rerun

1 job timed out:

  • pytorch_linux_bionic_py3_8_gcc9_coverage_test1

🚧 1 fixed upstream failure:

These were probably caused by upstream breakages that were already fixed.

Please rebase on the viable/strict branch (expand for instructions)

If your commit is older than viable/strict, run these commands:

git fetch https://github.com/pytorch/pytorch viable/strict
git rebase FETCH_HEAD
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 to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

This was referenced May 19, 2021
Closed
Closed
Closed
Closed
Closed
Closed
Closed
@BowenBao BowenBao closed this May 20, 2021
@facebook-github-bot facebook-github-bot deleted the gh/BowenBao/66/head branch June 20, 2021 14:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@neginraoof neginraoof Awaiting requested review from neginraoof

@spandantiwari spandantiwari Awaiting requested review from spandantiwari

Assignees
No one assigned
Labels
cla signed open source
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@BowenBao @facebook-github-bot @pytorchbot @jiafatom