Skip to content

webgpu: Fuse FlashAttention decode kernels and extend to any sequence length#28389

Open
qjia7 wants to merge 5 commits into
mainfrom
webgpu-flash-attention-relax-subgroups
Open

webgpu: Fuse FlashAttention decode kernels and extend to any sequence length#28389
qjia7 wants to merge 5 commits into
mainfrom
webgpu-flash-attention-relax-subgroups

Commits

Commits on May 7, 2026

Commits on May 11, 2026

Commits on May 12, 2026