[Autobucketing] Add simplefsdp's autobucketing pass to autoparallel #141
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]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
meta-cla
bot
added
the
CLA Signed
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Outdated suggestions cannot be applied.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.
Ehhhh, I messed up ghstack land....
I was supposed to land this stacked PR: #134, but hit some errors. I played with the PRs and accidentally merged the base PR. Now, I need to manually merge the five PRs and create a new one targeting main. Sorry for the confusion here....
(1) The autobucketing pass should work with this PR in torchtitan as the entry: pytorch/torchtitan#1658
(2) The memory estimation problem (test workflow) get fixed in this PR in pytorch: pytorch/pytorch#162300. Any pytorch version after this PR should work with current autobucketing pass.
(3) The inductor memory estimation code in PyTorch was updated a few days ago (PR). I need to confirm if the perf is good in the new runs. The overall pipeline is set up, and I will submit new PRs to fix any minor errors.