New RunLLM response generated: >>>*# Question* I want to use Kueue as the batch scheduler for Flyte with PyTorch. The documentation mentions Yunikorn and K8s scheduler plugins with co-scheduling. I don't see any reason why Kueue won't work. https://kueue.sigs.k8s.io/docs/tasks/run/kubeflow/pytorchjobs/ There is a slight difference though. In order to send a job to Kueue, one needs a PyTorchJob metadata.label that specifies the queue. However, the PyTorch plugin does expose a way to customize the labels. This capability should be relatively simple to add. I am willing to contribute this to the PyTorch plugin backend and flytekit. Has anyone used Kueue before with Flyte? # Answer No direct evidence of Kueue being used with Flyte is available. However, Flyte supports various schedulers like Yunikorn and Volcano, and it is feasible to integrate Kueue by customizing pod templates and annotations. According to this discussion, Kueue can manage pods with specific annotations, and Flyte's Propeller should handle pods in a pending state until Kueue schedules them. Would you like more details on how to proceed with the integration or contribute to the PyTorch plugin backend and Flytekit? Sources: - Discussion on Kueue integration Slack Link https://flyte-org.slack.com/archives/CP2HDHKE1/p1724131715.536619 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.