-
Notifications
You must be signed in to change notification settings - Fork 24.9k
[simplefsdp auto-bucketing] manual bucketing with plan #158321
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: gh/ruisizhang123/6/base
Are you sure you want to change the base?
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/158321
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ❌ 3 New Failures, 2 Unrelated FailuresAs of commit af8c4ce with merge base 33c9b41 ( NEW FAILURES - The following jobs have failed:
BROKEN TRUNK - The following job failed but were 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. |
This PR needs a
|
This manual plan takes a list of module definition as input, and bucket all-gather/reduce-scatter based on the definition.
An example would be:
torch._inductor.config.simplefsdp.bucketing_plan= ["tok_embeddings", "layers.[0-31]", "norm+output"]
.tok_embeddings
,layers.0
,layers.1
, etc.Stack from ghstack (oldest at bottom):