CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-06 09:46:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: Add s_off as a parameter in the args struct

Browse Source 
This may not be necessary, but it more closely mirrors the CUDA kernel

Branch: GraniteFourPerf

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
This commit is contained in:
Gabe Goodhart
2025-07-15 17:03:38 -06:00
parent 01612b7409
commit ba74a24730
1 changed files with 1 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
ggml/src/ggml-metal/ggml-metal-impl.h
View File

@@ -519,6 +519,7 @@ typedef struct {
int64_t n_group; int64_t n_group;
int64_t n_seq_tokens; int64_t n_seq_tokens;
int64_t n_seqs; int64_t n_seqs;
int64_t s_off;
uint64_t nb01; uint64_t nb01;
uint64_t nb02; uint64_t nb02;
uint64_t nb03; uint64_t nb03;