[FSDP2] Added module, mp policy to fsdp_pre_all_gather
#136129
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
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/136129
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit b0ea438 with merge base c64ae60 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
oncall: distributed
awgu
added
release notes: distributed (fsdp2)
weifengpy
approved these changes
Sep 22, 2024
Collaborator
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 |
pytorchmergebot
pushed a commit
that referenced
this pull request
Sep 23, 2024
I messed up the test before. The extensions were not running :/ Pull Request resolved: #136130 Approved by: https://github.com/weifengpy ghstack dependencies: #136129
BoyuanFeng
pushed a commit
to BoyuanFeng/pytorch
that referenced
this pull request
Sep 25, 2024
) - Sometimes having access to the `MixedPrecisionPolicy` in the `fsdp_pre_all_gather` is useful. See [here](https://github.com/pytorch/ao/pull/748/files#r1760375325) in the torchao INT8 mixed precision training PR. - Sometimes having access to the owning `nn.Module` allows for using it for saving state. See [here](pytorch#114299 (comment)) for an example. The major paint point here is how to deal with backward compatibility. For now, we use `signature.inspect` to check if the user subclass follows the old vs. new signature. However, for the new signature, the `param_dtype` in the post-all-gather is redundant, as if the user needed it, the user could save it from the `mp_policy` passed in the pre-all-gather now. Pull Request resolved: pytorch#136129 Approved by: https://github.com/weifengpy
BoyuanFeng
pushed a commit
to BoyuanFeng/pytorch
that referenced
this pull request
Sep 25, 2024
I messed up the test before. The extensions were not running :/ Pull Request resolved: pytorch#136130 Approved by: https://github.com/weifengpy ghstack dependencies: pytorch#136129
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):
test_all_gather_extensions_monkey_patch#136130fsdp_pre_all_gather#136129MixedPrecisionPolicyin thefsdp_pre_all_gatheris useful. See here in the torchao INT8 mixed precision training PR.nn.Moduleallows for using it for saving state. See here for an example.The major paint point here is how to deal with backward compatibility. For now, we use
signature.inspectto check if the user subclass follows the old vs. new signature. However, for the new signature, theparam_dtypein the post-all-gather is redundant, as if the user needed it, the user could save it from themp_policypassed in the pre-all-gather now.cc @XilunWu @H-Huang @kwen2501 @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @c-p-i-o