diff options
| author | Kawrakow <48489457+ikawrakow@users.noreply.github.com> | 2024-07-27 07:55:01 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-07-27 07:55:01 +0200 |
| commit | 154e0d75fccf1784fe9ff6fd76a630b66563da3d (patch) | |
| tree | 81ce6dbb5b1900c1aa78a879f0593c694cab9d27 /ggml/src/llamafile/sgemm.h | |
| parent | 0684c3e9c70d49323b4fc517128cbe222cab7f96 (diff) | |
Merge mainline llama.cpp (#3)
* Merging mainline - WIP
* Merging mainline - WIP
AVX2 and CUDA appear to work.
CUDA performance seems slightly (~1-2%) lower as it is so often
the case with llama.cpp/ggml after some "improvements" have been made.
* Merging mainline - fix Metal
* Remove check
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Diffstat (limited to 'ggml/src/llamafile/sgemm.h')
| -rw-r--r-- | ggml/src/llamafile/sgemm.h | 14 |
1 files changed, 14 insertions, 0 deletions
diff --git a/ggml/src/llamafile/sgemm.h b/ggml/src/llamafile/sgemm.h new file mode 100644 index 00000000..caf6dd55 --- /dev/null +++ b/ggml/src/llamafile/sgemm.h @@ -0,0 +1,14 @@ +#pragma once +#include <stdint.h> +#include <stdbool.h> +#ifdef __cplusplus +extern "C" { +#endif + +bool llamafile_sgemm(int64_t, int64_t, int64_t, const void *, int64_t, + const void *, int64_t, void *, int64_t, int, int, + int, int, int); + +#ifdef __cplusplus +} +#endif |