[onnx] Fix output shape mismatch issue of max_pool #106270
Closed
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/106270
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New FailureAs of commit 4937175: NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
For onnx MaxPool with ceil_mode=1, the sliding windows that starts in the right padded region won't be ignored, which causes different output shape with torch. Therefore, need to add Pad op before and not to set ceil_mode for MaxPool op like what is done in symbolic_opset9.
thiagocrepaldi
approved these changes
Jul 31, 2023
|
@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 |
Merge failedReason: 1 jobs have failed, first few of them are: trunk / macos-12-py3-arm64 / test (default, 2, 3, macos-m1-12) Details for Dev Infra teamRaised by workflow job |
|
@pytorchbot merge -f "Unrelated inductor failure" |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
@pytorchbot drci |
This was referenced Nov 7, 2023
|
@thiagocrepaldi |
titaiwangms
added a commit
that referenced
this pull request
Nov 9, 2023
In #106270, the solution managed to solve the [`ceil_model` corner issue](onnx/onnx#5711) with the usage of `get_pool_ceil_padding`. However, padding the ceil in converter side only works when we already know the input shapes, therefore, a regression happens when users want to do dynamic inputs. This PR provides (1) refactor codes with torchlib implementation, (2) add dynamic shapes test, and (3) disable the corner tests with comments saying re-enable it when the [real fix from ONNX](onnx/onnx#5741) is merged. [ghstack-poisoned]
titaiwangms
added a commit
that referenced
this pull request
Nov 9, 2023
In #106270, the solution managed to solve the [`ceil_model` corner issue](onnx/onnx#5711) with the usage of `get_pool_ceil_padding`. However, padding the ceil in converter side only works when we already know the input shapes, therefore, a regression happens when users want to do dynamic inputs. This PR provides (1) refactor codes with torchlib implementation, (2) add dynamic shapes test, and (3) disable the corner tests with comments saying re-enable it when the [real fix from ONNX](onnx/onnx#5741) is merged. [ghstack-poisoned]
titaiwangms
added a commit
that referenced
this pull request
Nov 9, 2023
… inputs" In #106270, the solution managed to solve the [`ceil_model` corner issue](onnx/onnx#5711) with the usage of `get_pool_ceil_padding`. However, padding the ceil in converter side only works when we already know the input shapes, therefore, a regression happens when users want to do dynamic inputs. This PR provides (1) refactor codes with torchlib implementation, (2) add dynamic shapes test, and (3) disable the corner tests with comments saying re-enable it when the [real fix from ONNX](onnx/onnx#5741) is merged. [ghstack-poisoned]
titaiwangms
added a commit
that referenced
this pull request
Nov 9, 2023
In #106270, the solution managed to solve the [`ceil_model` corner issue](onnx/onnx#5711) with the usage of `get_pool_ceil_padding`. However, padding the ceil in converter side only works when we already know the input shapes, therefore, a regression happens when users want to do dynamic inputs. This PR provides (1) refactor codes with torchlib implementation, (2) add dynamic shapes test, and (3) disable the corner tests with comments saying re-enable it when the [real fix from ONNX](onnx/onnx#5741) is merged. [ghstack-poisoned]
pytorchmergebot
pushed a commit
that referenced
this pull request
Nov 10, 2023
In #106270, the solution managed to solve the [`ceil_model` corner issue](onnx/onnx#5711) with the usage of `get_pool_ceil_padding`. However, padding the ceil in converter side only works when we already know the input shapes, therefore, a regression happens when users want to do dynamic inputs. This PR provides (1) refactor codes with torchlib implementation, (2) add dynamic shapes test, and (3) disable the corner tests with comments saying re-enable it when the [real fix from ONNX](onnx/onnx#5741) is merged. Pull Request resolved: #113318 Approved by: https://github.com/thiagocrepaldi
Skylion007
pushed a commit
to Skylion007/pytorch
that referenced
this pull request
Nov 14, 2023
In pytorch#106270, the solution managed to solve the [`ceil_model` corner issue](onnx/onnx#5711) with the usage of `get_pool_ceil_padding`. However, padding the ceil in converter side only works when we already know the input shapes, therefore, a regression happens when users want to do dynamic inputs. This PR provides (1) refactor codes with torchlib implementation, (2) add dynamic shapes test, and (3) disable the corner tests with comments saying re-enable it when the [real fix from ONNX](onnx/onnx#5741) is merged. Pull Request resolved: pytorch#113318 Approved by: https://github.com/thiagocrepaldi
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
For onnx MaxPool with ceil_mode=1, the sliding windows that starts in the right padded region won't be ignored, which causes different output shape with torch.
Therefore, need to add Pad op before and not to set ceil_mode for MaxPool op like what is done in symbolic_opset9 when convertting torch max_pool to onnx MaxPool.