Xin Wang comments

Results 15 comments of


                                            Xin Wang

[pytorch][sagemaker]Port https://github.com/aws/deep-learning-containers/pull/3293 to Cu121

> I guess we are not running the tests from #3293 on this PR. We would need to do that to ensure package has been installed properly. I have enabled...

[pytorch][sagemaker]Port https://github.com/aws/deep-learning-containers/pull/3293 to Cu121

Hi, what is latest for this PR?

feature: add python module entrypoint type, add python module support…

/Bot run all

Qn about GPU Hardware Failure Detection feature

Could anyone help comment on this? @Jeffwan Thanks

Qn about GPU Hardware Failure Detection feature

Thanks for the info. https://github.com/aibrix/ai-accelerator-tool seems very basic and we are not interested using that, it would be better when you plan for the diagnose integration interface, Aibrix could allow...

Qn about GPU Hardware Failure Detection feature

Hardware will still be Nvidia GPU, but we are looking for some alternatives for https://github.com/aibrix/ai-accelerator-tool since its logic is very basic. Yes, general interface will help, and we also try...

Training job restart enhancement

Thanks. Could you please provide some contexts about the difference between Kubeflow training operator v2 vs. JobSet? Is JobSet expected to eventually replace Kubeflow training operator in terms of training...

Training job restart enhancement

Thanks. @tenzen-y Could you please help comment for above questions?

Training job restart enhancement

Any update on this?

Training job restart enhancement

When GPU failure happens, training job might just stuck and pod does not exit with failure.