nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 19 Packages Projects Releases Wiki Activity

CUDA: fix FA kernel selection logic (#21271)

Browse Source
This commit is contained in:
Johannes Gäßler
2026-04-01 22:28:19 +03:00
committed by GitHub
parent 6de97b9d3e
commit 86221cf6da
1 changed files with 7 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-cuda/fattn.cu
+7
View File
@@ -340,7 +340,14 @@ static best_fattn_kernel ggml_cuda_get_best_fattn_kernel(const int device, const
case 128: case 128:
case 112: case 112:
case 256: case 256:
if (V->ne[0] != K->ne[0]) {
return BEST_FATTN_KERNEL_NONE;
}
break;
case 512: case 512:
if (V->ne[0] != K->ne[0]) {
return BEST_FATTN_KERNEL_NONE;
}
if (!gqa_opt_applies) { if (!gqa_opt_applies) {
return BEST_FATTN_KERNEL_NONE; return BEST_FATTN_KERNEL_NONE;
} }