CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-02 09:12:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

metal : improve MUL_MAT_ID (#15541)

Browse Source 
* metal : mul_mm_id remove hdst

* metal : remove mul_mm_id hsrc1

* metal : mul_mm_id simplify + add test

* metal : opt mul_mm_id map0

* metal : optimize mul_mm_id id gathering

* metal : mul/div opt

* metal : optimize mul_mm_id_map0

ggml-ci
This commit is contained in:
Georgi Gerganov
2025-08-26 12:46:15 +03:00
committed by GitHub
parent c4e9239064
commit 1d8d83deaa
4 changed files with 180 additions and 228 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
ggml/src/ggml-metal/ggml-metal-impl.h
View File

@@ -320,40 +320,31 @@ typedef struct {
} ggml_metal_kargs_mul_mv_ext;
typedef struct {
int32_t ne02;
int32_t ne10;
int32_t ne11; // n_expert_used (bcast)
uint64_t nb11;
uint64_t nb12;
int32_t neh11; // n_tokens
uint64_t nbh11;
int32_t ne21; // n_tokens
int32_t ne20; // n_expert_used
uint64_t nb21;
} ggml_metal_kargs_mul_mm_id_map0;
typedef struct {
int32_t ne20; // n_expert_used
int32_t neh0;
int32_t neh1;
uint64_t nbh1;
uint64_t nbh2;
int32_t ne0;
uint64_t nb1;
uint64_t nb2;
} ggml_metal_kargs_mul_mm_id_map1;
typedef struct {
int32_t ne00;
int32_t ne02;
uint64_t nb01;
uint64_t nb02;
uint64_t nb03;
int32_t neh12;
uint64_t nbh10;
uint64_t nbh11;
uint64_t nbh12;
uint64_t nbh13;
int32_t neh0;
int32_t neh1;
int32_t ne11;
uint64_t nb10;
uint64_t nb11;
uint64_t nb12;
uint64_t nb13;
int32_t ne20;
int32_t ne21;
int32_t ne0;
int32_t ne1;
int16_t r2;
int16_t r3;
} ggml_metal_kargs_mul_mm_id;