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.

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

Jump to bottom

[FEATURE] Option for easily assigning a GPU to an execution flow when multiple GPUs are present #238

Closed
aristizabal95 opened this issue Aug 27, 2022 · 3 comments · Fixed by #420
Closed

[FEATURE] Option for easily assigning a GPU to an execution flow when multiple GPUs are present #238

aristizabal95 opened this issue Aug 27, 2022 · 3 comments · Fixed by #420
Assignees
@aristizabal95
Labels
component: client issues regarding the CLI component: dependencies issues regarding dependencies type: enhancement New feature or request
Milestone
2023 Q1 (v1.0.0)

Comments

@aristizabal95
Copy link
Contributor

Is your feature request related to a problem? Please describe.
Currently, medperf doesn't handle anything GPU related, leveraging this to MLCube. This has caused some issues during the FeTS Challenge, where some contributors want to be specific about which GPU to use.

Describe the solution you'd like
There should be an option that can be passed to MLCube to specify which GPU to use. Something like --gpu=0 which would be piped to Cube.run and handled by MLCube. We must first determine what MLCube provides for this kind of scenarios.
@aristizabal95 aristizabal95 added type: enhancement New feature or request component: client issues regarding the CLI labels Aug 27, 2022
@hasan7n hasan7n added the component: dependencies issues regarding dependencies label Sep 3, 2022
@hasan7n hasan7n added this to the 2023 Q1 milestone Dec 20, 2022
@hasan7n
Copy link
Contributor

hasan7n commented Mar 20, 2023

depends on mlcommons/mlcube#277

@aristizabal95 aristizabal95 mentioned this issue Apr 21, 2023
Merged
@aristizabal95
Copy link
Contributor Author

There seems to be a problem with passing this parameter to MLCube: mlcommons/mlcube#307

@aristizabal95
Copy link
Contributor Author

We must test this at least manually, hopefully with CI/CD

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@aristizabal95 aristizabal95

Labels
component: client issues regarding the CLI component: dependencies issues regarding dependencies type: enhancement New feature or request
Projects
None yet
Milestone
2023 Q1 (v1.0.0)
Development

Successfully merging a pull request may close this issue.

Provide a parameter for specifying GPUs to use aristizabal95/medperf-2
2 participants
@aristizabal95 @hasan7n