Skip to content

Add a16w8 per-op test for gelu (#19598)#19598

Open
christine-long-meta wants to merge 3 commits into
pytorch:mainfrom
christine-long-meta:export-D104532359
Open

Add a16w8 per-op test for gelu (#19598)#19598
christine-long-meta wants to merge 3 commits into
pytorch:mainfrom
christine-long-meta:export-D104532359

Conversation

@christine-long-meta
Copy link
Copy Markdown
Contributor

@christine-long-meta christine-long-meta commented May 14, 2026 •
edited by meta-codesync Bot
Loading

Summary:

Add int16 activation / int8 weight (a16w8) quantization tests for aten.gelu on Ethos-U55 and Ethos-U85.

Changes

  • Add test_gelu_a16w8_u55_INT using EthosU55PipelineINT with a16w8_quantization=True, symmetric_io_quantization=True, qtol=128, epsilon=2**-16, reusing existing Gelu.test_data parameters (12 test configurations covering both `none` and `tanh` approximation modes)
  • Add test_gelu_a16w8_u85_INT using EthosU85PipelineINT with same kwargs
  • Register ops/test_gelu.py in fbcode/ and xplat/ targets.bzl

bypass-pytorch-oss-checks

Differential Revision: D104532359

@pytorch-bot
Copy link
Copy Markdown

pytorch-bot Bot commented May 14, 2026 •
edited
Loading

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19598

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

❌ 1 New Failure, 1 Unrelated Failure

As of commit f281d17 with merge base 42d87c4 (image):

NEW FAILURE - The following job has failed:

FLAKY - The following job failed but was likely due to flakiness present on trunk:

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

@github-actions github-actions Bot added ciflow/trunk module: arm Issues related to arm backend labels May 14, 2026
@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 May 14, 2026
@pytorch-bot
Copy link
Copy Markdown

pytorch-bot Bot commented May 14, 2026 •
edited
Loading

Workflows were awaiting approval. CI has now been triggered for the ciflow labels on this PR.

@github-actions
Copy link
Copy Markdown

github-actions Bot commented May 14, 2026

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

christine-long-meta added a commit to christine-long-meta/executorch that referenced this pull request May 14, 2026
@christine-long-meta @facebook-github-bot
Add a16w8 per-op test for gelu (pytorch#19598)
42c2abe 
Summary:

Add int16 activation / int8 weight (a16w8) quantization tests for `aten.gelu` on Ethos-U55 and Ethos-U85.


## Changes
- Add `a16w8_gelu_test_parameters` dict with 3 test configurations covering rank-1, rank-2, and rank-3 tensors
- Add `test_gelu_a16w8_u55_INT` using `EthosU55PipelineINT` with `a16w8_quantization=True, symmetric_io_quantization=True, qtol=128, epsilon=2**-16`
- Add `test_gelu_a16w8_u85_INT` using `EthosU85PipelineINT` with same kwargs
- Register `ops/test_gelu.py` in `fbcode/` and `xplat/` `targets.bzl`

bypass-pytorch-oss-checks

Differential Revision: D104532359
@meta-codesync meta-codesync Bot changed the title Add a16w8 per-op test for gelu Add a16w8 per-op test for gelu (#19598) May 14, 2026
@christine-long-meta christine-long-meta force-pushed the export-D104532359 branch 2 times, most recently from 15ffc91 to 1371cae Compare May 14, 2026 16:53
@meta-codesync
Copy link
Copy Markdown
Contributor

meta-codesync Bot commented May 14, 2026

@christine-long-meta has exported this pull request. If you are a Meta employee, you can view the originating Diff in D104532359.

2 similar comments
@meta-codesync
Copy link
Copy Markdown
Contributor

meta-codesync Bot commented May 14, 2026

@christine-long-meta has exported this pull request. If you are a Meta employee, you can view the originating Diff in D104532359.

@meta-codesync
Copy link
Copy Markdown
Contributor

meta-codesync Bot commented May 14, 2026

@christine-long-meta has exported this pull request. If you are a Meta employee, you can view the originating Diff in D104532359.

@meta-codesync meta-codesync Bot changed the title Add a16w8 per-op test for gelu (#19598) Add a16w8 per-op test for gelu May 14, 2026
@christine-long-meta christine-long-meta force-pushed the export-D104532359 branch 2 times, most recently from e344675 to 1fbd10a Compare May 16, 2026 02:07
@christine-long-meta @facebook-github-bot
Add a16w8 per-op test for var (pytorch#19596)
3879789 
Summary:

Add int16 activation / int8 weight (a16w8) quantization tests for `aten.var` on Ethos-U55 and Ethos-U85.

## Changes
- Add `test_parameters_ethosu` class attribute to `Var` with 2 test configurations (4D tensors with correction=0 and correction=1)
- Switch existing `test_var_dim_u55_INT_no_dim` and `test_var_dim_u85_INT_no_dim` from `Var.test_parameters` to `Var.test_parameters_ethosu` for Ethos-U compatible tensor shapes
- Add `test_var_a16w8_u55_INT` using `EthosU55PipelineINT` with `a16w8_quantization=True, symmetric_io_quantization=True`
- Add `test_var_a16w8_u85_INT` using `EthosU85PipelineINT` with same kwargs
- Register `ops/test_var.py` in `fbcode/` and `xplat/` `targets.bzl`

Differential Revision: D104532362
@christine-long-meta @facebook-github-bot
Add a16w8 per-op test for conv1d (pytorch#19597)
09b0ebf 
Summary:

Add int16 activation / int8 weight (a16w8) quantization tests for `aten.conv1d` on Ethos-U55 and Ethos-U85.

## Changes
- Add `test_conv1d_a16w8_u55_INT` using `EthosU55PipelineINT` with `a16w8_quantization=True, symmetric_io_quantization=True`, reusing existing `test_data_INT` parameters
- Add `test_conv1d_a16w8_u85_INT` using `EthosU85PipelineINT` with same kwargs
- Register `ops/test_conv1d.py` in `fbcode/` and `xplat/` `targets.bzl`

Reviewed By: Ninja91

Differential Revision: D104532360
@christine-long-meta @facebook-github-bot
Add a16w8 per-op test for gelu (pytorch#19598)
f281d17 
Summary:

Add int16 activation / int8 weight (a16w8) quantization tests for `aten.gelu` on Ethos-U55 and Ethos-U85.

## Changes
- Add `test_gelu_a16w8_u55_INT` using `EthosU55PipelineINT` with `a16w8_quantization=True, symmetric_io_quantization=True, qtol=128, epsilon=2**-16`, reusing existing `Gelu.test_data` parameters (12 test configurations covering both \`none\` and \`tanh\` approximation modes)
- Add `test_gelu_a16w8_u85_INT` using `EthosU85PipelineINT` with same kwargs
- Register `ops/test_gelu.py` in `fbcode/` and `xplat/` `targets.bzl`

bypass-pytorch-oss-checks

Differential Revision: D104532359
@meta-codesync meta-codesync Bot changed the title Add a16w8 per-op test for gelu Add a16w8 per-op test for gelu (#19598) May 16, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@digantdesai digantdesai Awaiting requested review from digantdesai digantdesai is a code owner

At least 1 approving review is required to merge this pull request.

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 meta-exported module: arm Issues related to arm backend

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@christine-long-meta