ONNX Update training ops and training amenable export API#32950
ONNX Update training ops and training amenable export API#32950lara-hdr wants to merge 32 commits intopytorch:masterfrom
Conversation
lara-hdr
commented
Feb 3, 2020
- Update Dropout and Batchnorm in opset 12 : Update Dropout and BatchNorm to be Training Friendly onnx/onnx#2568
- Update api logic for exporting to ONNX training amenable models
💊 CircleCI build failures summary and remediationsAs of commit bf610bf (more details on the Dr. CI page):
🕵️ 2 new failures recognized by patternsThe following build failures do not appear to be due to upstream breakages (reran 1 job to discount flakiness):
|
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
houseroad
left a comment
There was a problem hiding this comment.
Could you rebase to master and trigger the test again?
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
|
https://github.com/pytorch/pytorch/pull/32950/checks?check_run_id=512600263 clang-tidy is broken, could you fix it? |
|
@houseroad it seems like after adding the "Training" enum in torch/csrc/onnx/init.cpp, clang-tidy is trying to include onnx-ml.pb.h (instead of onnx.ph.h) which is the reason of the failure. I am not sure why this is happening since ONNX_ML is not defined.. |
This reverts commit 4f0b292.
…aidar/onnx_training
| -g"-torch/csrc/jit/export.cpp" \ | ||
| -g"-torch/csrc/jit/import.cpp" \ | ||
| -g"-torch/csrc/jit/netdef_converter.cpp" \ | ||
| -g"-torch/csrc/onnx/init.cpp" \ |
There was a problem hiding this comment.
@houseroad added this line to exclude init.cpp (which includes onnx_pb.h) to fix the linter error, as explained in the comment above (line 191-193)
|
cc @houseroad clang-tidy issue is fixed |
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
There was a problem hiding this comment.
@houseroad has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
|
@houseroad merged this pull request in 025a0ab. |
) Summary: - Update Dropout and Batchnorm in opset 12 : onnx/onnx#2568 - Update api logic for exporting to ONNX training amenable models Pull Request resolved: pytorch#32950 Reviewed By: hl475 Differential Revision: D19710370 Pulled By: houseroad fbshipit-source-id: e5e79d38552936966662c41d39ddf33be1ba3e35
) Summary: - Update Dropout and Batchnorm in opset 12 : onnx/onnx#2568 - Update api logic for exporting to ONNX training amenable models Pull Request resolved: pytorch#32950 Reviewed By: hl475 Differential Revision: D19710370 Pulled By: houseroad fbshipit-source-id: e5e79d38552936966662c41d39ddf33be1ba3e35