nicolas
/
ik_llama_opt
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix q4_1 and q5_1 on Arm (#348) 

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow 2025-04-25 19:48:08 +02:00 committed by GitHub
parent fb98619852
commit 4413f17b58
1 changed files with 8 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
ggml/src/ggml.c
View File

@ -741,7 +741,11 @@ static const ggml_type_traits_t type_traits[GGML_TYPE_COUNT] = {
.from_float_ref = (ggml_from_float_t) quantize_row_q4_1_ref,
.vec_dot = ggml_vec_dot_q4_1_q8_1,
#if GGML_USE_IQK_MULMAT
#if defined __AVX2__
.vec_dot_type = GGML_TYPE_Q8_2_X4,
#else
.vec_dot_type = GGML_TYPE_Q8_1_X4,
#endif
#else
.vec_dot_type = GGML_TYPE_Q8_1,
#endif
@ -809,7 +813,11 @@ static const ggml_type_traits_t type_traits[GGML_TYPE_COUNT] = {
.from_float_ref = (ggml_from_float_t) quantize_row_q5_1_ref,
.vec_dot = ggml_vec_dot_q5_1_q8_1,
#if GGML_USE_IQK_MULMAT
#ifdef __AVX2__
.vec_dot_type = GGML_TYPE_Q8_2_X4,
#else
.vec_dot_type = GGML_TYPE_Q8_1_X4,
#endif
#else
.vec_dot_type = GGML_TYPE_Q8_1,
#endif