[ONNX] Use the torchlib opset number and fix opset import logic #141413
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
justinchuby
requested review from
shubhambhokare1,
titaiwangms and
wschin
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/141413
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 9e84596 with merge base 259a00b ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
release notes: onnx
justinchuby
added
module: onnx
…logic" - Update the ONNX IR `add_opset_imports` pass to remove the heuristics of taking the `max` of the seen opsets. Instead, it uses the torchlib default opset version for the model's opset_import. The version converter is able to take the true opset versions in the nodes and convert the model to the correct version. - Update all hard coding of opset 18 to instead query the default torchlib opset from onnxscript, introduced in microsoft/onnxscript#1963 Fixes #141260 [ghstack-poisoned]
titaiwangms
reviewed
Nov 24, 2024
| # so we don't need to specifically process them. | ||
| with onnxscript.evaluator.default_as(tracer): | ||
| output = onnxscript.opset18.Concat(*sequence_mixed_elements, axis=0) # type: ignore[type-var] | ||
| output = onnxscript_apis.torchlib_opset().Concat( |
There was a problem hiding this comment.
Ops in _building.py also needs the changes?
There was a problem hiding this comment.
The opset version is used in https://github.com/pytorch/pytorch/pull/141413/files#diff-c2f19ccd7bc57c1edd3281f4d758f616c933599307fded5d3ad3d063a5313222R115 which is then used in
and . So we should be good
titaiwangms
approved these changes
Nov 25, 2024
|
@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 |
Ryo-not-rio
pushed a commit
to Ryo-not-rio/pytorch
that referenced
this pull request
Dec 2, 2024
…rch#141413) - Update the ONNX IR `add_opset_imports` pass to remove the heuristics of taking the `max` of the seen opsets. Instead, it uses the torchlib default opset version for the model's opset_import. The version converter is able to take the true opset versions in the nodes and convert the model to the correct version. - Update all hard coding of opset 18 to instead query the default torchlib opset from onnxscript, introduced in microsoft/onnxscript#1963 Fixes pytorch#141260 Pull Request resolved: pytorch#141413 Approved by: https://github.com/titaiwangms
pobin6
pushed a commit
to pobin6/pytorch
that referenced
this pull request
Dec 5, 2024
…rch#141413) - Update the ONNX IR `add_opset_imports` pass to remove the heuristics of taking the `max` of the seen opsets. Instead, it uses the torchlib default opset version for the model's opset_import. The version converter is able to take the true opset versions in the nodes and convert the model to the correct version. - Update all hard coding of opset 18 to instead query the default torchlib opset from onnxscript, introduced in microsoft/onnxscript#1963 Fixes pytorch#141260 Pull Request resolved: pytorch#141413 Approved by: https://github.com/titaiwangms
73 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
add_opset_importspass to remove the heuristics of taking themaxof the seen opsets. Instead, it uses the torchlib default opset version for the model's opset_import. The version converter is able to take the true opset versions in the nodes and convert the model to the correct version.Fixes #141260