Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ggml-metal: FWHT kernel for metal backend (#25924)

Browse Source 
* metal fwht wip

* shape guard and formatting

* formatting

* Formatting and typos

Co-authored-by: YiChen Lv <63285796+forforever73@users.noreply.github.com>

* fix narrowing issue

Co-authored-by: YiChen Lv <63285796+forforever73@users.noreply.github.com>

* cont : minor style

---------

Co-authored-by: YiChen Lv <63285796+forforever73@users.noreply.github.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Nick Lafleur
2026-07-28 10:44:06 +03:00
committed by GitHub
co-authored by YiChen Lv Georgi Gerganov
parent f87067841b
commit f95de9776b
7 changed files with 143 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-metal/ggml-metal-device.cpp
+15
View File
@@ -1252,6 +1252,21 @@ ggml_metal_pipeline_with_params ggml_metal_library_get_pipeline_argsort_merge(gg
return res;
}
ggml_metal_pipeline_with_params ggml_metal_library_get_pipeline_fwht(ggml_metal_library_t lib, int n) {
char base[256];
char name[256];
snprintf(base, 256, "kernel_fwht_f32_%d", n);
snprintf(name, 256, "%s", base);
ggml_metal_pipeline_with_params res = ggml_metal_library_get_pipeline(lib, name);
if (!res.pipeline) {
res = ggml_metal_library_compile_pipeline(lib, base, name, nullptr);
}
return res;
}
// note: reuse the argsort kernel for top_k
ggml_metal_pipeline_with_params ggml_metal_library_get_pipeline_top_k(ggml_metal_library_t lib, const ggml_tensor * op) {
assert(op->op == GGML_OP_TOP_K);