ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-02-25 07:34:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

Minor

Browse Source
This commit is contained in:
Iwan Kawrakow
2025-06-27 19:00:58 +03:00
parent 1e6e4833bc
commit 036b5bb822
1 changed files with 0 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
ggml/src/ggml-cuda/mmq.cuh
View File

@@ -2789,7 +2789,6 @@ template <int mmq_y, int nwarps, bool need_check> static __device__ __forceinlin
for (int j = 0; j < 4; ++j) {
const int aux_q4 = get_int_b4(bxi->qs, 4*kqsx+j);
const int2 v = get_int_from_table_16(aux_q4, values);
const int k0 = 8 * (threadIdx.x / 4) + threadIdx.x % 4;
#ifdef INT8_MMA_AVAILABLE
x_qs[i*MMQ_MMA_TILE_X_K_Q8_0 + 8*kqsx + j + 0] = v.x;
x_qs[i*MMQ_MMA_TILE_X_K_Q8_0 + 8*kqsx + j + 4] = v.y;