Gino Lu fb75da2467 sparse_attn: wire -mask and -attention_sink (block-map prune + attn mask)

2026-05-19 23:22:00 -04:00

..

block_fmha_pipeline_qr_ks_vs_async_jenga.hpp

fix(sparse_attn): backport PR #4742 LDS s_barrier

2026-05-17 02:30:48 -04:00

block_fmha_pipeline_qr_ks_vs_async_sparge.hpp

sparse_attn: split-launch dispatch + 3-mode PV-skip

2026-05-19 21:45:23 -04:00

block_fmha_pipeline_qr_ks_vs_async_vsa.hpp

fix(sparse_attn): backport PR #4742 LDS s_barrier

2026-05-17 02:30:48 -04:00

sparge_blockmap_pipeline.hpp

sparse_attn: wire -mask and -attention_sink (block-map prune + attn mask)

2026-05-19 23:22:00 -04:00

sparge_kstats_pipeline.hpp

refactor(sparse_attn): caller-owned workspace + dtype-aware sizing

2026-05-17 02:34:23 -04:00