llama.cpp
vulkan: Slang flash attention shader
#20451
Open

Commits
  • vulkan: port Flash Attention shader to Slang
    0cc4m committed 100 days ago
  • fix slang issues
    0cc4m committed 100 days ago
  • generic reductions
    0cc4m committed 100 days ago
  • move kv shmem staging to function
    0cc4m committed 100 days ago
  • Revert "move kv shmem staging to function"
    0cc4m committed 100 days ago
  • unify scalar+vector and fix reduce function
    0cc4m committed 100 days ago
Loading