Reject invalid ZenFlow top-k ratio endpoints - #8274
Open
tandede wants to merge 1 commit into
Open
Conversation
Signed-off-by: tandede <1090179959@qq.com>
tandede
force-pushed
the
agent/reject-zenflow-topk-boundaries
branch
from
August 19, 2026 01:17
9d6ed9a to
0bbc0ef
Compare
tandede
marked this pull request as ready for review
August 19, 2026 11:28
|
Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits. |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
zenflow.topk_ratioto be strictly between 0 and 1Why
ZenFlow's automatic update path normalizes selected gradients by
topk_ratioand unselected gradients by1 - topk_ratioin both the ZeRO-1/2 and ZeRO-3 implementations. The previous inclusive constraints accepted 0 and 1, so either endpoint could trigger division by zero after training had already initialized.Rejecting the endpoints during configuration parsing provides an immediate, actionable validation error while preserving all usable selective-update ratios.
Testing
python -m pytest -q tests/unit/runtime/zero/test_zero_config.py tests/unit/runtime/zenflow/test_zf_config.py tests/unit/runtime/zenflow/test_zf.py::test_split_affinity(19 passed)pre-commit run --files deepspeed/runtime/zenflow/zenflow_config.py tests/unit/runtime/zenflow/test_zf_config.py