Megatron-DeepSpeed
c5346794 - Merge branch 'bfloat_fused_softmax' into 'main'

Commit
4 years ago
Merge branch 'bfloat_fused_softmax' into 'main' Bfloat fused softmax + fused layer norm See merge request ADLR/megatron-lm!251
Author
Loading