Skip to content

Arm backend: Add 16A8W support and test for sigmoid operation #13796

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.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 8 commits into from
Sep 11, 2025
Merged

Arm backend: Add 16A8W support and test for sigmoid operation #13796

merged 8 commits into from
Sep 11, 2025

Conversation

Ninja91
Copy link
Contributor

@Ninja91 Ninja91 commented Aug 29, 2025 •
edited
Loading

Stack from ghstack (oldest at bottom):

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:

  • Add INT16 dtype validation support in op_sigmoid.py
  • Add test_sigmoid_tensor_16a8w_tosa_INT test function
  • Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: D80510729

cc @digantdesai @freddan80 @per @zingo @oscarandersson8218

@Ninja91
Add 16A8W support and test for sigmoid operation
a5d989a 
Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

[ghstack-poisoned]
@Ninja91 Ninja91 requested a review from digantdesai as a code owner August 29, 2025 06:42
@pytorch-bot PyTorch Bot
Copy link

pytorch-bot bot commented Aug 29, 2025 •
edited
Loading

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/13796

Note: Links to docs will display an error until the docs builds have been completed.

❌ 3 New Failures, 6 Unrelated Failures

As of commit 2b2f943 with merge base 1d37845 (image):

NEW FAILURES - The following jobs have failed:

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Aug 29, 2025
This was referenced Aug 29, 2025
Merged
Merged
Merged
Merged
Merged
Merged
Merged
Merged
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@zingo zingo added ciflow/trunk partner: arm For backend delegation, kernels, demo, etc. from the 3rd-party partner, Arm module: arm Issues related to arm backend labels Aug 29, 2025
@zingo zingo changed the title Add 16A8W support and test for sigmoid operation Arm backend: Add 16A8W support and test for sigmoid operation Aug 29, 2025
digantdesai
digantdesai reviewed Aug 29, 2025
View reviewed changes
@digantdesai digantdesai requested a review from per August 29, 2025 20:03
@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
e3172ab 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@Ninja91 Ninja91 mentioned this pull request Sep 4, 2025
Open
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@Ninja91 Ninja91 mentioned this pull request Sep 4, 2025
Open
@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
2e14a35 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
b2f5604 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@Ninja91 Ninja91 added the release notes: arm Changes to the ARM backend delegate label Sep 8, 2025
@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
d0f8714 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
0000147 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
c342e21 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@Ninja91
Update on "Arm backend: Add 16A8W support and test for sigmoid operat…
2b2f943 
…ion"

Add 16A8W quantization support and test for the sigmoid operation in ExecutorTorch ARM backend.

This follows the pattern established for linear and mul operations, extending int16 support to sigmoid operations.

Changes:
- Add INT16 dtype validation support in op_sigmoid.py
- Add test_sigmoid_tensor_16a8w_tosa_INT test function
- Enable test_sigmoid.py in test targets configuration

The 16A8W configuration uses 16-bit activations with 8-bit weights, enabling higher precision for activations while maintaining weight efficiency.

Differential Revision: [D80510729](https://our.internmc.facebook.com/intern/diff/D80510729/)

cc digantdesai freddan80 per zingo oscarandersson8218

[ghstack-poisoned]
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D80510729

@facebook-github-bot facebook-github-bot merged commit bdfd22e into gh/Ninja91/9/base Sep 11, 2025
293 of 303 checks passed
@facebook-github-bot facebook-github-bot deleted the gh/Ninja91/9/head branch September 11, 2025 15:18
@pytorchbot pytorchbot mentioned this pull request Sep 11, 2025
Merged
Ninja91 added a commit that referenced this pull request Sep 11, 2025
@pytorchbot @Ninja91
Arm backend: Add 16A8W support and test for sigmoid operation (#14213)
6b9c0a6 
This PR was created by the merge bot to help merge the original PR into
the main branch.
ghstack PR number: #13796 by
@Ninja91
^ Please use this as the source of truth for the PR details, comments,
and reviews
ghstack PR base:
https://github.com/pytorch/executorch/tree/gh/Ninja91/9/base
ghstack PR head:
https://github.com/pytorch/executorch/tree/gh/Ninja91/9/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/main
Merge bot PR head:
https://github.com/pytorch/executorch/tree/gh/Ninja91/9/orig
@diff-train-skip-merge

Differential Revision:
[D80510729](https://www.internalfb.com/diff/D80510729)

Co-authored-by: Nitin Jain <jainnitin@meta.com>
StrycekSimon pushed a commit to nxp-upstream/executorch that referenced this pull request Sep 23, 2025
@pytorchbot @Ninja91 @StrycekSimon
Arm backend: Add 16A8W support and test for sigmoid operation (pytorc…
0247b61 
…h#14213)

This PR was created by the merge bot to help merge the original PR into
the main branch.
ghstack PR number: pytorch#13796 by
@Ninja91
^ Please use this as the source of truth for the PR details, comments,
and reviews
ghstack PR base:
https://github.com/pytorch/executorch/tree/gh/Ninja91/9/base
ghstack PR head:
https://github.com/pytorch/executorch/tree/gh/Ninja91/9/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/main
Merge bot PR head:
https://github.com/pytorch/executorch/tree/gh/Ninja91/9/orig
@diff-train-skip-merge

Differential Revision:
[D80510729](https://www.internalfb.com/diff/D80510729)

Co-authored-by: Nitin Jain <jainnitin@meta.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@digantdesai digantdesai digantdesai approved these changes

@per per Awaiting requested review from per

Assignees

No one assigned

Labels

ciflow/trunk CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported module: arm Issues related to arm backend partner: arm For backend delegation, kernels, demo, etc. from the 3rd-party partner, Arm release notes: arm Changes to the ARM backend delegate

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@Ninja91 @facebook-github-bot @digantdesai @zingo