Skip to content

CUDA: gated_delta_net - share per-token k/q via shared memory for long prefills#54

Draft
roberteg16 wants to merge 1 commit into
gfx11from
rogarcia.gdn-shared-kq
Draft

CUDA: gated_delta_net - share per-token k/q via shared memory for long prefills#54
roberteg16 wants to merge 1 commit into
gfx11from
rogarcia.gdn-shared-kq