Convert fake tensors to detected fake mode in pattern matcher replace_by_example#176938
Closed
frgossen wants to merge 4 commits intogh/frgossen/4/basefrom
Closed
Convert fake tensors to detected fake mode in pattern matcher replace_by_example#176938frgossen wants to merge 4 commits intogh/frgossen/4/basefrom
frgossen wants to merge 4 commits intogh/frgossen/4/basefrom
Conversation
…_by_example When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/176938
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 e616cfd with merge base e45dfba ( FLAKY - The following job failed but was likely due to flakiness present on trunk:
BROKEN TRUNK - The following job failed but was present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
UNSTABLE - The following job is marked as unstable, possibly due to flakiness on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
frgossen
added a commit
that referenced
this pull request
Mar 9, 2026
…_by_example When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. ghstack-source-id: c5c06e4 Pull Request resolved: #176938
This PR needs a
|
zou3519
reviewed
Mar 10, 2026
zou3519
reviewed
Mar 10, 2026
zou3519
reviewed
Mar 10, 2026
Contributor
zou3519
left a comment
There was a problem hiding this comment.
seems reasonable modulo the fx.map_arg comment
…her replace_by_example" When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy kadeng muchulee8 amjames chauhang aakhundov coconutruben jataylo [ghstack-poisoned]
frgossen
added a commit
that referenced
this pull request
Mar 10, 2026
…_by_example When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. ghstack-source-id: 33be38c Pull Request resolved: #176938
…her replace_by_example" When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy kadeng muchulee8 amjames chauhang aakhundov coconutruben jataylo [ghstack-poisoned]
frgossen
added a commit
that referenced
this pull request
Mar 10, 2026
…_by_example When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. ghstack-source-id: 6654245 Pull Request resolved: #176938
aorenste
reviewed
Mar 10, 2026
aorenste
approved these changes
Mar 10, 2026
zou3519
approved these changes
Mar 10, 2026
…her replace_by_example" When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy kadeng muchulee8 amjames chauhang aakhundov coconutruben jataylo [ghstack-poisoned]
frgossen
added a commit
that referenced
this pull request
Mar 10, 2026
…_by_example When running pre-grad passes outside the `fake_mode` wrapper in `compile_fx`, the fake tensors attached to graph nodes can belong to a different fake mode than the one active during pattern replacement tracing. This mismatch triggers the "Mixing fake modes NYI" assertion. Fix this by detecting the fake mode from the example values in `Match.replace_by_example` and converting any fake tensors into that mode before calling `trace_fn`. This removes the need for the `wrapped_run_pre_grad_passes` shim in `compile_fx` that previously papered over the problem by restoring the original fake mode. ghstack-source-id: e36bc0a Pull Request resolved: #176938
Contributor
Author
|
@pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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):
When running pre-grad passes outside the
fake_modewrapper incompile_fx, the fake tensors attached to graph nodes can belong to adifferent fake mode than the one active during pattern replacement
tracing. This mismatch triggers the "Mixing fake modes NYI" assertion.
Fix this by detecting the fake mode from the example values in
Match.replace_by_exampleand converting any fake tensors into thatmode before calling
trace_fn. This removes the need for thewrapped_run_pre_grad_passesshim incompile_fxthat previouslypapered over the problem by restoring the original fake mode.
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @kadeng @muchulee8 @amjames @chauhang @aakhundov @coconutruben @jataylo