happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Update ggml/src/ggml-cuda/mmq.cuh 

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
iacopPBK 2026-04-01 13:24:19 +02:00 committed by GitHub
parent 777f5943a4
commit fbc4cfcdde
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ggml/src/ggml-cuda/mmq.cuh
View File

@ -496,6 +496,8 @@ static __device__ __forceinline__ void vec_dot_q4_1_q8_1_dp4a(
constexpr int max_cpy = ggml_cuda_get_max_cpy_bytes(); constexpr int max_cpy = ggml_cuda_get_max_cpy_bytes();
constexpr int mcpy_int = max_cpy / sizeof(int); constexpr int mcpy_int = max_cpy / sizeof(int);
static_assert(VDR_Q4_0_Q8_1_MMQ == 4, "bad VDR_Q4_0_Q8_1_MMQ");
int tmp0[4], tmp1[4]; int tmp0[4], tmp1[4];