[fx graph cache] Refactor FxGraphCachePickler #138682
Closed
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
Summary: In an upcoming change, we need to modify how FxGraphCachePickler pickles tensor constants. Specifically, we'll pickle differently depending on whether parameters have been frozen. To do that, it will be convenient to change FxGraphCachePickler into a regular object instead of a collection of classmethods. Test Plan: unit tests [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/138682
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit dbb15b8 with merge base 8aedc64 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This was referenced Oct 23, 2024
masnesral
commented
Oct 23, 2024
| pickler.dumps(details3), | ||
| ) | ||
|
|
||
| def test_bypass_unsupported(self): |
There was a problem hiding this comment.
Noticed this code path wasn't being tested.
masnesral
commented
Oct 23, 2024
| # Sometimes inductor configs are unpickleable and can fail | ||
| raise BypassAOTAutogradCache from e | ||
|
|
||
| def debug_lines(self) -> List[str]: |
There was a problem hiding this comment.
Couldn't figure out why we need this since we can just do AOTAutogradCachePickler.debug_lines(details). Same in codecache.py
masnesral
commented
Oct 23, 2024
| dispatch_table[torch.Tensor] = _reduce_tensor | ||
| def __init__(self): | ||
| super().__init__() | ||
| self.dispatch_table: Dict |
There was a problem hiding this comment.
So mypy doesn't complain.
masnesral
commented
Oct 23, 2024
|
|
||
| @classmethod | ||
| def dumps(cls, obj: Any) -> bytes: | ||
| # Run with pickler.fast so it doesn't intern strings, making the hash result more predictable |
There was a problem hiding this comment.
This didn't have to move to __init__, but seemed like the more natural place for it
Summary: In an upcoming change, we need to modify FxGraphCachePickler to behave differently depending on whether the graph has frozen parameters (whether or not we have frozen parameters). To do that, it will be convenient to change FxGraphCachePickler into a regular object instead of a collection of classmethods. Test Plan: unit tests cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang aakhundov [ghstack-poisoned]
eellison
approved these changes
Oct 29, 2024
|
@pytorchbot merge |
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 |
pytorchmergebot
pushed a commit
that referenced
this pull request
Oct 31, 2024
Summary: Move all the custom `_reduce_*` functions inside the FxGraphCachePickler class. This is mostly a cosmetic change since they're conceptually members of FxGraphCachePickler. But also in an upcoming diff, I'll add a member variable to the class to control how we handle constant tensors, so it will be convenient to be able to query that setting via `self`. I made the analogous changes to AOTAutogradCachePickler for consistency. Test Plan: unit tests Pull Request resolved: #138683 Approved by: https://github.com/eellison ghstack dependencies: #138681, #138682
rahulsingh-intel
pushed a commit
to rahulsingh-intel/pytorch
that referenced
this pull request
Nov 5, 2024
Summary: In an upcoming change, we need to modify FxGraphCachePickler to behave differently depending on whether the graph has frozen parameters (whether or not we have frozen parameters). To do that, it will be convenient to change FxGraphCachePickler into a regular object instead of a collection of classmethods. Test Plan: unit tests Pull Request resolved: pytorch#138682 Approved by: https://github.com/eellison ghstack dependencies: pytorch#138681
rahulsingh-intel
pushed a commit
to rahulsingh-intel/pytorch
that referenced
this pull request
Nov 5, 2024
Summary: Move all the custom `_reduce_*` functions inside the FxGraphCachePickler class. This is mostly a cosmetic change since they're conceptually members of FxGraphCachePickler. But also in an upcoming diff, I'll add a member variable to the class to control how we handle constant tensors, so it will be convenient to be able to query that setting via `self`. I made the analogous changes to AOTAutogradCachePickler for consistency. Test Plan: unit tests Pull Request resolved: pytorch#138683 Approved by: https://github.com/eellison ghstack dependencies: pytorch#138681, pytorch#138682
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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):
Summary: In an upcoming change, we need to modify FxGraphCachePickler to behave differently depending on whether the graph has frozen parameters (whether or not we have frozen parameters). To do that, it will be convenient to change FxGraphCachePickler into a regular object instead of a collection of classmethods.
Test Plan: unit tests
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @chauhang @aakhundov