support AveragePool19 and Pad19 #15597
Merged
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
Signed-off-by: Liqun Fu <[email protected]>
xadupre
previously approved these changes
Apr 20, 2023
justinchuby
reviewed
Apr 20, 2023
Signed-off-by: Liqun Fu <[email protected]>
|
Nit: update the title and description to indicate that this PR also covers "Pad-19" ? |
gramalingam
reviewed
Apr 24, 2023
| if (mode == Mode::Reflect) { | ||
| PadInnermostAxis(axisStart - prePad, axisStart + prePad, -1 /* inputDelta */, onnxruntime::narrow<size_t>(prePad)); | ||
| PadInnermostAxis(output, output - 2, -1 /* inputDelta */, onnxruntime::narrow<size_t>(postPad)); | ||
| } else { |
There was a problem hiding this comment.
Nit: a comment like // mode == Mode::Wrap would be helpful in the else-line (here and similarly below).
gramalingam
previously approved these changes
Apr 24, 2023
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
Signed-off-by: Liqun Fu <[email protected]>
gramalingam
approved these changes
May 15, 2023
prathikr
pushed a commit
that referenced
this pull request
May 16, 2023
titaiwangms
added a commit
that referenced
this pull request
Jan 24, 2025
Fix #16203 Previous to this PR, if `ceil_mode` is on, the calculation of a value would divide the kernel size, even if remaining pixels is less than the kernel size, which causes the difference in this operator between ORT and torch. However, this fix only applies to the change in #15597, which only supports AvgPool since 19. The older opset version is remain the same, as it's using mlas files. Also, the PR fixes the shape mismatch caused by sliding window starting from padding. More detail: onnx/onnx#6650 (And this PR is also validated with the tests added in onnx/onnx#6650)
ashrit-ms
pushed a commit
that referenced
this pull request
Feb 11, 2025
Fix #16203 Previous to this PR, if `ceil_mode` is on, the calculation of a value would divide the kernel size, even if remaining pixels is less than the kernel size, which causes the difference in this operator between ORT and torch. However, this fix only applies to the change in #15597, which only supports AvgPool since 19. The older opset version is remain the same, as it's using mlas files. Also, the PR fixes the shape mismatch caused by sliding window starting from padding. More detail: onnx/onnx#6650 (And this PR is also validated with the tests added in onnx/onnx#6650)
guschmue
pushed a commit
that referenced
this pull request
Mar 6, 2025
Fix #16203 Previous to this PR, if `ceil_mode` is on, the calculation of a value would divide the kernel size, even if remaining pixels is less than the kernel size, which causes the difference in this operator between ORT and torch. However, this fix only applies to the change in #15597, which only supports AvgPool since 19. The older opset version is remain the same, as it's using mlas files. Also, the PR fixes the shape mismatch caused by sliding window starting from padding. More detail: onnx/onnx#6650 (And this PR is also validated with the tests added in onnx/onnx#6650)
ashrit-ms
pushed a commit
that referenced
this pull request
Mar 17, 2025
Fix #16203 Previous to this PR, if `ceil_mode` is on, the calculation of a value would divide the kernel size, even if remaining pixels is less than the kernel size, which causes the difference in this operator between ORT and torch. However, this fix only applies to the change in #15597, which only supports AvgPool since 19. The older opset version is remain the same, as it's using mlas files. Also, the PR fixes the shape mismatch caused by sliding window starting from padding. More detail: onnx/onnx#6650 (And this PR is also validated with the tests added in onnx/onnx#6650)
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.
Description
Motivation and Context
AveragePoolXDTask can be refactored with LpPoolXDTask. However, there has been an issue with LpPool in ONNX (#6302). with the issue fixed, LpPoolXDTask can be implemented using AveragePoolXDTask. This is not in the scope of this PR.