Skip to content

feat: TBQ4_0 + TBQ3_0 CUDA flash attention for SM121 (DGX Spark)#1

Open
mihai-chiorean wants to merge 112 commits into
release/turbo3-cudafrom
feat/tbq4-cuda-fa-sm121
Open

feat: TBQ4_0 + TBQ3_0 CUDA flash attention for SM121 (DGX Spark)#1
mihai-chiorean wants to merge 112 commits into
release/turbo3-cudafrom
feat/tbq4-cuda-fa-sm121

refactor: remove TBQ rotation, use upstream Walsh-Hadamard (PR #21038)

9d4d0a0
Select commit
Loading
Failed to load commit list.

Select a check to view from the sidebar