Turn on AOTAutogradCache by default on open source #141981
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/141981
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit d6b6f62 with merge base 1d3b010 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
jamesjwu
added
topic: not user facing
[ghstack-poisoned]
[ghstack-poisoned]
| @@ -95,10 +95,6 @@ def should_use_local_autograd_cache(): | |||
| return config.enable_autograd_cache | |||
|
|
|||
|
|
|||
| def autograd_cache_enabled(): | |||
There was a problem hiding this comment.
Realized we don't need this
| remote, | ||
| ) | ||
| if isinstance(fw_compiler, SerializableAOTDispatchCompiler): | ||
| local = should_use_local_autograd_cache() |
There was a problem hiding this comment.
This makes it so that we don't spend time calling these should_use_local_autograd_cache functions if there's no way we're caching anyway
[ghstack-poisoned]
|
@laithsakka I updated 2 benchmarks which regressed slightly more than the requirement (~1.5%). This is expected since AOTAutogradCache will take a little bit of extra time with inductor enabled just due to cache key calculations and logging. Actual walltime on non trivial model benchmarks will not be affected. |
|
@pytorchbot rebase |
|
@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here |
|
Successfully rebased |
|
@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 |
|
The merge job was canceled or timed out. This most often happen if two merge requests were issued for the same PR, or if merge job was waiting for more than 6 hours for tests to finish. In later case, please do not hesitate to reissue the merge command |
|
@jamesjwu Sorry I've cancelled the merge. We're doing release branch cut today, please wait before landing. |
|
@pytorchbot rebase |
|
@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here |
|
Successfully rebased |
[ghstack-poisoned]
|
@pytorchbot merge -i |
Merge startedYour change will be merged while ignoring the following 0 checks: Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
i guess there was no runaway from the regressiob |
Pull Request resolved: pytorch#141981 Approved by: https://github.com/bdhirsh, https://github.com/oulgen
Stack from ghstack (oldest at bottom):
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @chenyang78 @kadeng @chauhang @amjames