Skip to content

[ET-VK] Using TmpTensor for width packed versions of q_linear op shader to reduce memory usage. #8009

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.

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

Merged
merged 1 commit into from
Jan 28, 2025

Conversation

pytorchbot
Copy link
Collaborator

This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #7929
^ 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/trivedivivek/53/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/trivedivivek/53/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/trivedivivek/52/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/trivedivivek/53/orig
@diff-train-skip-merge

…er to reduce memory usage.

Pull Request resolved: #7929

This diff introduces the use of temporary tensors to reduce memory usage in the width packed versions of the q_linear op shader.
ghstack-source-id: 263456691
@exported-using-ghexport

Differential Revision: [D68561647](https://our.internmc.facebook.com/intern/diff/D68561647/)
Copy link

pytorch-bot bot commented Jan 28, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/8009

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

✅ No Failures

As of commit f28179e with merge base 7bc06d1 (image):
💚 Looks good so far! There are no failures yet. 💚

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

@facebook-github-bot facebook-github-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 Jan 28, 2025
@manuelcandales manuelcandales merged commit d1104a7 into gh/trivedivivek/52/orig Jan 28, 2025
44 checks passed
@manuelcandales manuelcandales deleted the gh/trivedivivek/53/orig branch January 28, 2025 21:27
trivedivivek added a commit that referenced this pull request Jan 29, 2025
* [ET-VK] Using shared memory to save position in conv2d dw output op.

Pull Request resolved: #7923

This diff introduces a change to conv2d dw op to save output positions in shared memory, which reduces register usage and improves performance.
ghstack-source-id: 263440666
@exported-using-ghexport

Differential Revision: [D68400890](https://our.internmc.facebook.com/intern/diff/D68400890/)

* [ET-VK] Using push constants for conv2d dw.

Pull Request resolved: #7928

This diff is related to the use of push constants for convolutional dw (depthwise) in Executorch's Vulkan backend. This optimization improves memory usage.
ghstack-source-id: 263440665
@exported-using-ghexport

Differential Revision: [D68493849](https://our.internmc.facebook.com/intern/diff/D68493849/)

* [ET-VK] Using TmpTensor for width packed versions of q_linear op shader to reduce memory usage. (#8009)

Pull Request resolved: #7929

This diff introduces the use of temporary tensors to reduce memory usage in the width packed versions of the q_linear op shader.
ghstack-source-id: 263456691
@exported-using-ghexport

Differential Revision: [D68561647](https://our.internmc.facebook.com/intern/diff/D68561647/)

Co-authored-by: Vivek Trivedi <[email protected]>

---------

Co-authored-by: Vivek Trivedi <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants