Skip to content
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.

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

Add DGXCloud executor based on Run.ai REST API #141

Merged
merged 6 commits into from
Feb 5, 2025

Conversation

hemildesai
Copy link
Collaborator

No description provided.

Signed-off-by: Hemil Desai <[email protected]>
Signed-off-by: Hemil Desai <[email protected]>
src/nemo_run/core/execution/dgxcloud.py Fixed Show fixed Hide fixed
return None
executor.delete(job_id)

def list(self) -> list[ListAppResponse]: ...

Check notice

Code scanning / CodeQL

Statement has no effect Note

This statement has no effect.
Signed-off-by: Hemil Desai <[email protected]>
"Logs not available for DGXCloudExecutor based jobs. Please visit the cluster UI to view the logs."
)

def cleanup(self, handle: str): ...

Check notice

Code scanning / CodeQL

Statement has no effect Note

This statement has no effect.
@pablo-garay pablo-garay marked this pull request as ready for review February 4, 2025 22:26
pablo-garay
pablo-garay previously approved these changes Feb 4, 2025
roclark and others added 2 commits February 5, 2025 16:29
DGX Cloud uses the PyTorch Training Operator from KubeFlow under the
hood to launch jobs. This handles many of the variables necessary for
running distributed PyTorch jobs with torchrun and only a subset of
settings are required to launch the job as the original default
settings will conflict with the auto-configured setup from DGX Cloud.

Signed-Off-By: Robert Clark <[email protected]>
Add support for torchrun to DGXC Executor
@pablo-garay pablo-garay merged commit 463d6e1 into main Feb 5, 2025
7 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants