[ET-VK] Fixing out_limits_scaled calculation for conv2d pw ops. #11271
Conversation
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
…o not squeeze if significant axis are all 1 and trailing axis are all > 1. Pull Request resolved: #11177 This diff modifies the `should_squeeze` function in `SqueezeUnsqueezeInputs` to not squeeze (return False) if significant axes are all 1 and trailing axes are all > 1. ghstack-source-id: 287222796 @exported-using-ghexport Differential Revision: [D75483587](https://our.internmc.facebook.com/intern/diff/D75483587/)
…to improve performance. Pull Request resolved: #11178 This diff removes shared memory usage in `conv2d_dw_output_tile.glsl` shader to improve performance. Makes sum a one dimensional array, and moves bias application before storing texel. ghstack-source-id: 287222799 @exported-using-ghexport Differential Revision: [D75499165](https://our.internmc.facebook.com/intern/diff/D75499165/)
Pull Request resolved: #11251 The fix changing the calculation of `out_limits_scaled` from: ```glsl const int out_limits_scaled[2] = {out_limits.x + (TILE_SIZE_X - 1) * TILE_SIZE_X, out_limits.y + (TILE_SIZE_Y - 1) * TILE_SIZE_Y}; ``` to: ```glsl const int out_limits_scaled[2] = {(out_limits.x + (TILE_SIZE_X - 1)) / TILE_SIZE_X, (out_limits.y + (TILE_SIZE_Y - 1)) / TILE_SIZE_Y}; ``` This change ensures that `out_limits_scaled` is calculated correctly, taking into account the tile size and the output limits of the convolution operation. ghstack-source-id: 287225747 @exported-using-ghexport Differential Revision: [D75575662](https://our.internmc.facebook.com/intern/diff/D75575662/)
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/11271
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (3 Unrelated Failures)As of commit ab7d6c0 with merge base f8a3fd8 ( FLAKY - The following jobs failed but were likely due to flakiness present on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
trivedivivek
added
the
release notes: vulkan
trivedivivek
approved these changes
May 30, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #11251 by @trivedivivek
^ 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/107/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/trivedivivek/107/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/trivedivivek/102/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/trivedivivek/107/orig
@diff-train-skip-merge