Add on-demand full-state checkpointing for OpenShift AI / KubeFlow preemption#686

Open

RobotSail wants to merge 14 commits intomainfrom

claude/on-demand-checkpointing-dCsmo

Commits on Feb 26, 2026

Add on-demand full-state checkpointing for OpenShift AI / KubeFlow preemption
claude
committed

Commits on Mar 20, 2026

Add checkpoint checks at 5 synchronization points per training step

RobotSail
and
claude
committed

Commits on Mar 24, 2026