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

Fix q_xxs using mul_mat_q

Browse source
This commit is contained in:
Aidan 2024-05-22 11:46:22 +01:00
parent 62bfef5194
commit 4b1770109c
1 changed files with 1 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

1
ggml-sycl.cpp
View file

@ -15243,6 +15243,7 @@ static void ggml_sycl_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
}
} else {
bool use_mul_mat_q = min_compute_capability >= VER_4VEC && ggml_is_quantized(src0->type);
use_mul_mat_q = use_mul_mat_q && (src0->type != GGML_TYPE_IQ2_XXS);
if (use_xmx && min_compute_capability >= VER_GEN9 && src1->ne[1] > XMX_MAX_BATCH_SIZE) {
use_mul_mat_q = false;