CUDA: faster tile FA, add oob checks, more HSs (#16492)

2025-10-11 20:54:32 +02:00
parent a3cb04744f
commit 11f0af5504
18 changed files with 1358 additions and 784 deletions
@@ -0,0 +1,5 @@
+// This file has been autogenerated by generate_cu_files.py, do not edit manually.
+
+#include "../fattn-tile.cuh"
+
+DECL_FATTN_TILE_CASE(40, 40);