root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 10:24:35 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

remove row=1 cond (#6532)

Browse Source
This commit is contained in:
Abhilash Majumder 2024-04-08 13:56:01 +05:30 committed by GitHub
parent d752327c33
commit 87fb5b4234
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
ggml-sycl.cpp
View File

@ -15776,7 +15776,7 @@ static void ggml_sycl_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
#ifdef GGML_SYCL_FORCE_DMMV
const bool use_mul_mat_vec_q = false;
#else
bool use_mul_mat_vec_q = min_compute_capability >= VER_4VEC && ggml_is_quantized(src0->type) && ggml_nrows(src1) == 1;
bool use_mul_mat_vec_q = min_compute_capability >= VER_4VEC && ggml_is_quantized(src0->type);
use_mul_mat_vec_q = use_mul_mat_vec_q ||
(src0->type == GGML_TYPE_IQ2_XXS) || (src0->type == GGML_TYPE_IQ2_XS) || (src0->type == GGML_TYPE_IQ2_S) ||
(src0->type == GGML_TYPE_IQ3_XXS) || (src0->type == GGML_TYPE_IQ3_S) ||
@ -15787,7 +15787,6 @@ static void ggml_sycl_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
#endif // GGML_SYCL_FORCE_DMMV
if (use_mul_mat_vec_q) {
// NOTE: this kernel does not support ggml_nrows(src1) > 1
// GGML_SYCL_DEBUG("ggml_sycl_mul_mat ggml_sycl_op_mul_mat_vec_q path\n");
ggml_sycl_op_mul_mat(src0, src1, dst, ggml_sycl_op_mul_mat_vec_q, true);
} else {