[ET-VK][int4] Wrap int4 linear calls with view_copy nodes to squeeze/unsqueeze inputs #8226
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
…unsqueeze inputs This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/8226
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 30e966c with merge base 7805229 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
This was referenced Feb 5, 2025
|
This pull request was exported from Phabricator. Differential Revision: D69065866 |
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 5, 2025
…unsqueeze inputs This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/) ghstack-source-id: 264874724 Pull Request resolved: #8226
nathanaelsee
added
the
release notes: vulkan
…to squeeze/unsqueeze inputs" This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D69065866 |
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 5, 2025
…unsqueeze inputs Pull Request resolved: #8226 This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. ghstack-source-id: 264908152 @exported-using-ghexport Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/)
…to squeeze/unsqueeze inputs" This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D69065866 |
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 5, 2025
…unsqueeze inputs Pull Request resolved: #8226 This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. ghstack-source-id: 264915710 @exported-using-ghexport Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/)
jorgep31415
approved these changes
Feb 5, 2025
…to squeeze/unsqueeze inputs" This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D69065866 |
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 6, 2025
…unsqueeze inputs Pull Request resolved: #8226 This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. ghstack-source-id: 264952606 @exported-using-ghexport Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/)
facebook-github-bot
merged commit 2ba4ab2
into
gh/nathanaelsee/3/base
45 of 47 checks passed
facebook-github-bot
temporarily deployed
to
cherry-pick-bot
GitHub Actions
Inactive
kirklandsign
pushed a commit
that referenced
this pull request
Feb 6, 2025
…unsqueeze inputs Pull Request resolved: #8226 This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op. The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes. ghstack-source-id: 264952606 @exported-using-ghexport Differential Revision: [D69065866](https://our.internmc.facebook.com/intern/diff/D69065866/) --------- Co-authored-by: Nathanael See <[email protected]>
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 6, 2025
See T214560872 #8226 added the pass to the partition preprocess pass list, so now it runs on all exports. This uncovered a bug in the squeeze dims finding function in the mobilenet test case. Differential Revision: [D69254910](https://our.internmc.facebook.com/intern/diff/D69254910/) [ghstack-poisoned]
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 6, 2025
See T214560872 #8226 added the pass to the partition preprocess pass list, so now it runs on all exports. This uncovered a bug in the squeeze dims finding function in the mobilenet test case. Differential Revision: [D69254910](https://our.internmc.facebook.com/intern/diff/D69254910/) ghstack-source-id: 265078517 Pull Request resolved: #8281
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 7, 2025
…queezeUnsqueezePass" See T214560872 #8226 added the pass to the partition preprocess pass list, so now it runs on all exports. This uncovered a bug in the squeeze dims finding function in the mobilenet test case. Differential Revision: [D69254910](https://our.internmc.facebook.com/intern/diff/D69254910/) [ghstack-poisoned]
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 7, 2025
…ass" See T214560872 #8226 added the pass to the partition preprocess pass list, so now it runs on all exports. This uncovered a bug in the squeeze dims finding function in the mobilenet test case. Differential Revision: [D69254910](https://our.internmc.facebook.com/intern/diff/D69254910/) [ghstack-poisoned]
nathanaelsee
pushed a commit
that referenced
this pull request
Feb 7, 2025
Pull Request resolved: #8281 See T214560872 #8226 added the pass to the partition preprocess pass list, so now it runs on all exports. This uncovered a bug in the squeeze dims finding function in the mobilenet test case. ghstack-source-id: 265183421 @exported-using-ghexport Differential Revision: [D69254910](https://our.internmc.facebook.com/intern/diff/D69254910/)
This was referenced Feb 11, 2025
This was referenced Feb 24, 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.
Stack from ghstack (oldest at bottom):
This is done automatically for full-precision linear/mm nodes in the graph at torch.export graph tracing time, but is not done for the int4 op.
The new pass adds view_copy nodes, as there are subsequent passes which can fuse view_copy nodes if redundant, and convert view_copy nodes to squeeze/unsqueeze nodes.
Differential Revision: D69065866