Reuse profiler helpers in mFSDP v2 symmetric memory tests#5873
Merged
Conversation
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
2 times, most recently
from
July 18, 2026 23:11
cbf1c5a to
1383435
Compare
copy-pr-bot
Bot
force-pushed
the
pull-request/5859
branch
from
July 18, 2026 23:12
9153657 to
fb5bf82
Compare
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
from
July 18, 2026 23:14
1383435 to
a16d05a
Compare
copy-pr-bot
Bot
force-pushed
the
pull-request/5859
branch
from
July 18, 2026 23:16
fb5bf82 to
be3cbc6
Compare
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
from
July 18, 2026 23:17
a16d05a to
67a5adf
Compare
copy-pr-bot
Bot
force-pushed
the
pull-request/5859
branch
from
July 18, 2026 23:27
be3cbc6 to
949a9df
Compare
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
from
July 18, 2026 23:28
67a5adf to
af677a8
Compare
wujingyue
marked this pull request as ready for review
July 18, 2026 23:28
Autumn1998
approved these changes
Jul 21, 2026
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
2 times, most recently
from
July 21, 2026 15:39
adab7f6 to
46b3e02
Compare
Signed-off-by: Jingyue Wu <jingyuew@nvidia.com>
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
from
July 21, 2026 16:08
8a049c8 to
85bc135
Compare
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
from
July 21, 2026 16:09
85bc135 to
4a06492
Compare
wujingyue
force-pushed
the
agent/reuse-profiler-utils-symm
branch
from
July 21, 2026 16:10
4a06492 to
1b1a649
Compare
wujingyue
enabled auto-merge
July 21, 2026 17:00
Contributor
Author
|
/ok to test |
Contributor
|
🔄 Merge queue validation started! You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/29920736154 |
Contributor
|
🔄 Merge queue validation started! You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/29935237896 |
Contributor
|
🔄 Merge queue validation started! You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/29960716719 |
Contributor
|
🔄 Merge queue validation started! You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/29968252116 |
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
3 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 changed
test_symmetric_memory.pyto profile CPU+CUDA events and reusecollect_linked_kernelsfor all-gather and reduce-scatter kernel attribution.ncclSymkdirectly on the linked collective kernel names instead of keeping local profiler-name/count helpers.Why
This keeps the mFSDP v2 profiler parsing logic in one helper path after #5859 added linked CPU-op attribution for profiler kernels.
Impact
Test-only refactor. There is no production-code behavior change.
Stacking
Stacked on #5859 with base branch
pull-request/5859.Validation
uv run --no-sync isort tests/unit_tests/distributed/mfsdp_v2/profiler_utils.py tests/unit_tests/distributed/mfsdp_v2/test_symmetric_memory.pyuv run --no-sync python -m compileall tests/unit_tests/distributed/mfsdp_v2/profiler_utils.py tests/unit_tests/distributed/mfsdp_v2/test_symmetric_memory.pygit diff --checkuv run --no-sync python -m pytest --collect-only -q tests/unit_tests/distributed/mfsdp_v2/test_symmetric_memory.pyFull distributed CUDA execution is left to CI because this refactor depends on linked profiler correlations from the CI PyTorch profiler build.