vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Reduced verbosity of comment

Browse source
This commit is contained in:
Alberto Cabrera 2024-07-09 10:47:43 +01:00
parent f288ae1a14
commit 0a00d6e4b8
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/ggml-sycl.cpp
View file

@ -3658,7 +3658,7 @@ static void ggml_sycl_mul_mat(ggml_backend_sycl_context & ctx, const ggml_tensor
use_mul_mat_q = use_mul_mat_q && (src1->ne[1] <= MMQ_MAX_BATCH_SIZE);
#endif // SYCL_USE_XMX
// mmvq path is faster in the Nvidia backend but slower on the Intel backend
// mmvq path is faster in the CUDA backend.
if (ctx.stream()->get_backend() == sycl::backend::ext_oneapi_cuda)
use_dequantize_mul_mat_vec = use_dequantize_mul_mat_vec && !use_mul_mat_vec_q;