Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-14 14:28:58 +01:00
Code Issues Packages Projects Releases Wiki Activity

iq3_s_mult: back to blocks of 32

Browse Source
This commit is contained in:
Iwan Kawrakow 2024-03-03 18:50:26 +02:00
parent f4cb4eac45
commit e5e72562c5
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ggml-quants.h
View File

@ -201,7 +201,7 @@ typedef struct {
static_assert(sizeof(block_iq3_xxs) == sizeof(ggml_fp16_t) + 3*(QK_K/8), "wrong iq3_xxs block size/padding"); static_assert(sizeof(block_iq3_xxs) == sizeof(ggml_fp16_t) + 3*(QK_K/8), "wrong iq3_xxs block size/padding");
// 3.4375 bpw // 3.4375 bpw
#define IQ3S_BLOCK_SIZE 16 #define IQ3S_BLOCK_SIZE 32
#if QK_K == 64 #if QK_K == 64
#define IQ3S_N_SCALE 2 #define IQ3S_N_SCALE 2
#else #else