CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-03 09:22:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

metal : reduce command encoding overhead (#9698)

Browse Source 
* metal : reduce command encoding overhead

ggml-ci

* metal : add comments
This commit is contained in:
Georgi Gerganov
2024-10-01 16:00:25 +03:00
committed by GitHub
parent a90484c6d9
commit cad341d889
5 changed files with 2000 additions and 1912 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/llama.cpp
View File

@@ -17025,12 +17025,6 @@ static void llama_graph_compute(
ggml_cgraph * gf,
int n_threads,
ggml_threadpool * threadpool) {
#ifdef GGML_USE_METAL
if (ggml_backend_is_metal(lctx.backend_metal)) {
ggml_backend_metal_set_n_cb(lctx.backend_metal, n_threads);
}
#endif
if (lctx.backend_cpu != nullptr) {
ggml_backend_cpu_set_n_threads(lctx.backend_cpu, n_threads);
ggml_backend_cpu_set_threadpool(lctx.backend_cpu, threadpool);