transformers

c87deb21 - feat(integration): Add KubeflowCallback to enable automatic progress … (#44487)

feat(integration): Add KubeflowCallback to enable automatic progress … (#44487) * feat(integration): Add KubeflowCallback to enable automatic progress and metrics reporting for training jobs running on Kubeflow Trainer. Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * fix: make KubeflowCallback self-contained and align with controller spec Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * docs: enhance KubeflowCallback docstring with env vars and features Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * fix: Auto-enable KubeflowCallback when running inside TrainJob Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * fix: add explicit is_world_process_zero checks and fix token cache test Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * fix: remove auto-enable for KubeflowCallback Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * fix: align env var names with merged kubeflow/trainer#3227 Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * refactor: cache SSL context, add logging, and cleanup KubeflowCallback Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> * fix: re-add auto-enable for KubeflowCallback and add unit test Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> --------- Signed-off-by: abhijeet-dhumal <abhijeetdhumal652@gmail.com> Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com>