Skip to content

Fix gradient accumulation under FSDP2: defer cross-replica all-reduce - #162

Open
Sullivan07043 wants to merge 2 commits into
NVIDIA:mainfrom
Sullivan07043:fix-fsdp2-accum
Open

Fix gradient accumulation under FSDP2: defer cross-replica all-reduce#162
Sullivan07043 wants to merge 2 commits into
NVIDIA:mainfrom
Sullivan07043:fix-fsdp2-accum

Merge branch 'main' into fix-fsdp2-accum

fa98cc3
Select commit
Loading
Failed to load commit list.

Workflow runs completed with no jobs