[fsdp, model]{fix} Allow flash attention 2 to be used for NemotronH model on FSDP#5419
Open
thvasilo wants to merge 1 commit intoverl-project:mainfrom
Open
[fsdp, model]{fix} Allow flash attention 2 to be used for NemotronH model on FSDP#5419thvasilo wants to merge 1 commit intoverl-project:mainfrom
thvasilo wants to merge 1 commit intoverl-project:mainfrom