LLM-Inference/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-08-12 22:31:22 +04:00
Code Issues Packages Projects Releases Wiki Activity

ggml-cuda: Allow transpose-free gemmv computation (llama/26171)

Browse Source 
When matrix's weights are shaped 1xK is leverage a transpose-free
computation to use mat_mul_vec_f.
This commit is contained in:
Robert Esclapez
2026-08-04 13:37:47 +03:00
committed by Georgi Gerganov
parent 64d57d3df5
commit 1a8c5a62f5
1 changed files with 14 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-cuda/ggml-cuda.cu
+14
View File
@@ -1836,6 +1836,20 @@ static void ggml_cuda_mul_mat(ggml_backend_cuda_context & ctx, const ggml_tensor
ggml_cuda_mul_mat_vec_f(ctx, src0, src1, nullptr, dst);
return;
}
// A transposed vector can still use MMVQ (i.e. ne01 == 1)
if (ne01 == 1 && ne11 > MMVF_MAX_BATCH_SIZE && ne2 == 1 && ne3 == 1
&& src0->type == GGML_TYPE_F32
&& ggml_is_contiguous(src0) && ggml_is_contiguous(src1) && ggml_is_contiguous(dst)
&& ggml_cuda_should_use_mmvf(src1->type, cc, src1->ne, src1->nb, /*ne11 =*/ 1)) {
ggml_tensor dst_vec = *dst;
dst_vec.ne[0] = ne11;
dst_vec.ne[1] = 1;
dst_vec.nb[1] = dst_vec.nb[0]*ne11;
dst_vec.nb[2] = dst_vec.nb[1];
dst_vec.nb[3] = dst_vec.nb[1];
ggml_cuda_mul_mat_vec_f(ctx, src1, src0, nullptr, &dst_vec);
return;
}
if (ggml_cuda_should_use_mmf(src0->type, cc, warp_size, src0->ne, src0->nb, ne11, /*mul_mat_id =*/ false)) {
ggml_cuda_mul_mat_f(ctx, src0, src1, nullptr, dst);
return;