millaguie/llama.cpp
1
0 Fork 0
Code Issues Pull Requests Actions 32 Packages Projects Releases Wiki Activity

CUDA: fix half2 -> half conversion for HIP (#15529)

Browse Source
This commit is contained in:
Johannes Gäßler authored and GitHub committed 2025-08-23 21:37:06 +02:00
1 parent 611f419cff
commit 710dfc465a
1 file changed
+1 -1
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-cuda/fattn-tile-f16.cu
+1 -1
View File
@@ -258,7 +258,7 @@ static __global__ void flash_attn_tile_ext_f16(
const half val = hexp(sink - kqmax[j0/nwarps]);
kqsum[j0/nwarps] = kqsum[j0/nwarps] * KQ_max_scale;
if (threadIdx.x == 0) {
kqsum[j0/nwarps].x = __hadd(kqsum[j0/nwarps].x, val);
kqsum[j0/nwarps].x = __hadd(__low2half(kqsum[j0/nwarps]), val);
}
#pragma unroll