[ET-VK] benchmarking linear_qta8a_qga4w_qta8o #12007
Conversation
# Context In order to benchmark the linear_qta8a_qga4w_qta8o operator, this diff adds functionality for int32 in the operator since Android does not support int8 operations. Functionally, this change should not be used, and this is only being provided in order to add support for int32 in android cases. # Changes Created a new benchmarking file for the linear_qta8a_qga4w_qta8o operator, which tests against various test cases. Also made a few changes to the existing operator with support for int32 along with a new variant. Differential Revision: [D77384894](https://our.internmc.facebook.com/intern/diff/D77384894/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/12007
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ No FailuresAs of commit 9b79f85 with merge base 85cf6ce ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
# Context In order to benchmark the linear_qta8a_qga4w_qta8o operator, this diff adds functionality for int32 in the operator since Android does not support int8 operations. Functionally, this change should not be used, and this is only being provided in order to add support for int32 in android cases. # Changes Created a new benchmarking file for the linear_qta8a_qga4w_qta8o operator, which tests against various test cases. Also made a few changes to the existing operator with support for int32 along with a new variant. Differential Revision: [D77384894](https://our.internmc.facebook.com/intern/diff/D77384894/) ghstack-source-id: 292886977 Pull Request resolved: #12007
|
This pull request was exported from Phabricator. Differential Revision: D77384894 |
This PR needs a
|
Stack from ghstack (oldest at bottom):
Context
In order to benchmark the linear_qta8a_qga4w_qta8o operator, this diff adds functionality for int32 in the operator since Android does not support int8 operations. Functionally, this change should not be used, and this is only being provided in order to add support for int32 in android cases.
Changes
Created a new benchmarking file for the linear_qta8a_qga4w_qta8o operator, which tests against various test cases. Also made a few changes to the existing operator with support for int32 along with a new variant.
Differential Revision: D77384894