fix(zenflow): restore assert and remove duplicated identical branch in gradient copy - #8445
Open
simpleqt wants to merge 1 commit into
Open
fix(zenflow): restore assert and remove duplicated identical branch in gradient copy#8445simpleqt wants to merge 1 commit into
simpleqt wants to merge 1 commit into
Conversation
…n gradient copy The ZenFlow fork of async_inplace_copy_grad_to_fp32_buffer_from_gpu mangled the original logic into two byte-identical if/else arms. If grad_accum were ever None, the None branch would itself call grad_accum.view(-1) and raise AttributeError, defeating the check. Restored the upstream assert + single-assignment shape from runtime/zero/stage_1_and_2.py. Signed-off-by: simpleqt <89645338+simpleqt@users.noreply.github.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
In
deepspeed/runtime/zenflow/zenflow_stage_1_and_2.py(ZenFlow overlap path), the fork ofasync_inplace_copy_grad_to_fp32_buffer_from_gpumangled the original logic into two byte-identical if/else arms:If the condition were ever true, the None branch itself calls
.view()onNone. Restored the upstream shape fromruntime/zero/stage_1_and_2.py(assert grad_accum is not None+ single assignment). No test covers this method (grep -rn async_inplace_copy tests/is empty), which is why the mangle survived.Signed-off-by: simpleqt 89645338+simpleqt@users.noreply.github.com