NXP backend: Replace move relu before concat optimization#15394
Merged
roman-janik-nxp merged 2 commits intopytorch:mainfrom Nov 6, 2025
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/15394
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 9 PendingAs of commit 73fb5e2 with merge base acf5b4b ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
Collaborator
Author
|
@pytorchbot label "module: nxp" "release notes: nxp" |
pytorch-bot
bot
added
module: nxp
roman-janik-nxp
requested review from
MartinPavella,
StrycekSimon,
jirioc and
robert-kalmar
and removed request for
robert-kalmar
MartinPavella
approved these changes
Nov 5, 2025
roman-janik-nxp
force-pushed
the
feature/nxg11066/EIEX-465-Replace-move_relu_before_concat-optimization_github
branch
from
cf63bbc to
73fb5e2
Compare
Collaborator
Author
|
The failing tests are not caused by this PR. |
roman-janik-nxp
deleted the
feature/nxg11066/EIEX-465-Replace-move_relu_before_concat-optimization_github
branch
abhinaykukkadapu
pushed a commit
to abhinaykukkadapu/executorch
that referenced
this pull request
Nov 6, 2025
) ### Summary This PR replaces optimization in `move_relu_before_concat.py` by `MoveActivationBeforeConcat` aten pass. The pass moves selected activations that are supported for fusion on Neutron (Relu, Relu6, Sigmoid, Tanh) before the `concat` node if concat input nodes are either Conv 2D or Linear 2D. The whole node Logic is determined by target specs, now supporting Neutron-C. Tests updated. ### Test plan Unit tests provided (test_move_activation_before_concatenation.py). cc @robert-kalmar
JacobSzwejbka
pushed a commit
that referenced
this pull request
Nov 6, 2025
### Summary This PR replaces optimization in `move_relu_before_concat.py` by `MoveActivationBeforeConcat` aten pass. The pass moves selected activations that are supported for fusion on Neutron (Relu, Relu6, Sigmoid, Tanh) before the `concat` node if concat input nodes are either Conv 2D or Linear 2D. The whole node Logic is determined by target specs, now supporting Neutron-C. Tests updated. ### Test plan Unit tests provided (test_move_activation_before_concatenation.py). cc @robert-kalmar
jirioc
pushed a commit
to nxp-upstream/executorch
that referenced
this pull request
Dec 19, 2025
) ### Summary This PR replaces optimization in `move_relu_before_concat.py` by `MoveActivationBeforeConcat` aten pass. The pass moves selected activations that are supported for fusion on Neutron (Relu, Relu6, Sigmoid, Tanh) before the `concat` node if concat input nodes are either Conv 2D or Linear 2D. The whole node Logic is determined by target specs, now supporting Neutron-C. Tests updated. ### Test plan Unit tests provided (test_move_activation_before_concatenation.py). cc @robert-kalmar
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
Summary
This PR replaces optimization in
move_relu_before_concat.pybyMoveActivationBeforeConcataten pass. The pass moves selected activations that are supported for fusion on Neutron (Relu, Relu6, Sigmoid, Tanh) before theconcatnode if concat input nodes are either Conv 2D or Linear 2D. The whole node Logic is determined by target specs, now supporting Neutron-C. Tests updated.Test plan
Unit tests provided (test_move_activation_before_concatenation.py).
cc @robert-kalmar