[DEV] torch_dist fixes, speed improvements and memory reduction for layerwise distopt - #1965
Merged
Merged
Conversation
FDecaYed
changed the title
Contributor
Author
|
/ok to test bb76f7f |
FDecaYed
force-pushed
the
deyuf/layerwise_improvements
branch
from
October 27, 2025 11:22
09971f8 to
f44b3fa
Compare
Contributor
|
Could we add more tests with current broadcast path as reference? They should exact match after all communication is done. |
Contributor
Author
|
/ok to test 572cf0a |
Contributor
Author
|
/ok to test c30e51d |
skyw
requested changes
Nov 4, 2025
skyw
left a comment
skyw
left a comment
Contributor
There was a problem hiding this comment.
Would be good to parameterize and test different combination of parallelisms.
Looks good overall.
skyw
reviewed
Nov 5, 2025
- switch to faster all gather - sort param size and assign to rank in a ping-pong loop for balanced memory usage Signed-off-by: Deyu Fu <deyuf@nvidia.com>
…nsorFactory replica_id change
Contributor
Author
|
/ok to test 325366c |
Contributor
Author
|
/ok to test 2c01a2d |
Contributor
Author
|
/ok to test ac64381 |
skyw
approved these changes
Nov 5, 2025
Contributor
Author
|
/ok to test 75a42dc |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do ?
PR to main
Adding some optimizations that misses initial PR adding layerwise dist opt:
Also added 3 minor fixes for layerwise:
Code review
The following process is enforced via the CODEOWNERS file for changes into
megatron/core. For changes outside ofmegatron/core, it is up to the PR author whether or not to tag the Final Reviewer team.For MRs into `main` branch
(Step 1): Add PR label
Expert Review(Step 2): Collect the expert reviewers reviews
Expert Reviewlabel when your PR is ready for review.Final Review might get declined if these requirements are not fulfilled.
(Step 3): Final Review
Final Reviewlabel(Optional Step 4): Cherry-pick into release branch
If this PR also needs to be merged into
core_r*release branches, after this PR has been merged, selectCherry-pickto open a new PR into the release branch.For MRs into `dev` branch
The proposed review process for `dev` branch is under active discussion.MRs are mergable after one approval by either
eharper@nvidia.comorzijiey@nvidia.com.Merging your PR
Any member of core-adlr and
core-nemowill be able to merge your PR.