ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-02-23 14:44:09 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove forgotten printf

Browse Source
This commit is contained in:
Iwan Kawrakow
2025-11-09 18:27:59 +02:00
parent bcdc12d02b
commit 19ecaaad42
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.cu
View File

@@ -2145,7 +2145,6 @@ static int ggml_cuda_mul_mat_q(ggml_backend_cuda_context & ctx, const ggml_tenso
cgraph->nodes[node_n+2]->src[1]->type == GGML_TYPE_F32 &&
ggml_nrows(cgraph->nodes[node_n+2]->src[1]) == 1) {
// We have a bias applied after the matrix multiplication and we can fuse it
printf("Fusing %s(%s) and %s(%s)\n", ggml_op_name(dst->op), dst->name, ggml_op_name(cgraph->nodes[node_n+2]->op), cgraph->nodes[node_n+2]->name);
ggml_cuda_op_mul_mat_vec_q_biased(ctx, dst->src[0], src1, cgraph->nodes[node_n+2], cgraph->nodes[node_n+2]->src[1],
(const char *)dst->src[0]->data, nullptr, src1_quantized.get(), (float *)cgraph->nodes[node_n+2]->data,
0, dst->src[0]->ne[1], src1->ne[1], ne10_padded, stream);