v5.8.0
Fix ZeRO-3 from_pretrained: load registered buffers in _load_state_dict_into_zero3_model (#45402)
Update distributed tests (#44338)
[Test refactor 1/5] Per-folder tests reorganization (#15725)
Refactor trainer data_collator and callbacks tests (#43776)
[Trainer] Add ddp_static_graph option (#45519)
Update common tests Trainer (#44260)
Fix: NotebookProgressCallback crash when evaluating with the Trainer (#44949)
Ensure final evaluation runs with step-based evaluation strategy (#44146)
Add GreedyLR adaptive learning rate scheduler (#44271)
Fix training ci and clean some tests (#44491)
feat(integration): Add KubeflowCallback to enable automatic progress … (#44487)