-
Notifications
You must be signed in to change notification settings - Fork 25.3k
[ONNX] Add new ops and enable the MNIST test #99284
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.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. Weβll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
π Helpful Linksπ§ͺ See artifacts and rendered test results at hud.pytorch.org/pr/99284
Note: Links to docs will display an error until the docs builds have been completed. β 4 FailuresAs of commit d68419f: NEW FAILURES - The following jobs have failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This comment was marked as resolved.
This comment was marked as resolved.
da76929
to
d490881
Compare
Requires microsoft/onnxscript#645 |
Previously it was passed in as an empty list and graph_builder fails when the list is empty. Discovered in pytorch/pytorch#99284 - Added maxpool_2d/3d custom OpInfo tests. Skipping 3d for now due to shape mismatch. - Pin torch-nightly version to a requirements file
b2d2cc3
to
378b278
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, mergeable after consensus with @titaiwangms on test_mnist
I guess?
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
@pytorchbot merge |
Merge failedReason: Approval needed from one of the following: |
@pytorchbot -h |
PyTorchBot Help
Merge
Revert
Rebase
Label
Dr CI
|
@pytorchbot merge -ic |
Merge startedYour change will be merged while ignoring the following 4 checks: linux-binary-manywheel / manywheel-py3_8-cuda11_7-with-pypi-cudnn-build / build, linux-binary-manywheel / manywheel-py3_8-cuda12_1-build / build, linux-binary-manywheel / manywheel-py3_8-cuda11_7-with-pypi-cudnn-test, linux-binary-manywheel / manywheel-py3_8-cuda12_1-test Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
@pytorchbot merge -f "onnx checks succeeded" |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
π€ Generated by Copilot at d490881
Summary
π¦ππΊοΈ
This pull request adds support for converting more PyTorch FX operators to ONNX using the
onnxscript
package. It updates the package installation script and the operator mappings infunction_dispatcher.py
. It also enables a test case for themax_pool2d
operator.Walkthrough
onnxscript
package (link)_ATENLIB_FUNCTIONS
dictionary in thefunction_dispatcher.py
module (link, link, link, link, link)test_max_pool2d
method in theTestFxToOnnx
class in thetest_fx_to_onnx.py
file, since themax_pool2d
operator is now supported by theonnxscript
package (link)