Skip to content

Specify nodes for gpu metrics collection and split data to each rank #320

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.

Sign up for GitHub

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

Merged
hemildesai merged 2 commits into from
Aug 18, 2025
Merged

Specify nodes for gpu metrics collection and split data to each rank #320

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
0412bf7
Specify nodes for gpu metrics collection and split data to each rank
ashbhandare Aug 15, 2025
547b8c9
Fix unit test
ashbhandare Aug 15, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion nemo_run/core/execution/slurm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -560,7 +560,7 @@ def get_nsys_entrypoint(self) -> str:
launcher = self.get_launcher()
entrypoint, postfix = "nsys", ""
if launcher.nsys_gpu_metrics:
entrypoint = 'bash -c \'GPU_METRICS_FLAG=""; if [ "$SLURM_PROCID" -eq 0 ]; then GPU_METRICS_FLAG="--gpu-metrics-devices=all"; fi; nsys'
entrypoint = 'bash -c \'GPU_METRICS_FLAG=""; if echo "${GPU_METRICS_NODES}" | grep -q -w "${SLURM_NODEID}"; then GPU_METRICS_FLAG="--gpu-metrics-devices=${SLURM_LOCALID}"; fi; nsys'
postfix = "'"
return (entrypoint, postfix)

Expand Down
2 changes: 1 addition & 1 deletion test/core/execution/test_slurm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,7 +198,7 @@ def test_get_nsys_entrypoint(self):

with patch.object(executor, "get_launcher", return_value=launcher_mock):
assert executor.get_nsys_entrypoint() == (
'bash -c \'GPU_METRICS_FLAG=""; if [ "$SLURM_PROCID" -eq 0 ]; then GPU_METRICS_FLAG="--gpu-metrics-devices=all"; fi; nsys',
'bash -c \'GPU_METRICS_FLAG=""; if echo "${GPU_METRICS_NODES}" | grep -q -w "${SLURM_NODEID}"; then GPU_METRICS_FLAG="--gpu-metrics-devices=${SLURM_LOCALID}"; fi; nsys',
"'",
)

Expand Down
Loading