Fix for layer distillation with sequence parallel training#431

Merged

oleksost merged 2 commits intomainfrom

sp_layer_distillation

Dec 23, 2025

Commits on Dec 19, 2025

layer distillation loss with masking and sequence parallelism
oleksost
committed

Commits on Dec 23, 2025

Merge branch 'main' into sp_layer_distillation
oleksost
authored