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-blas.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-blas.h')
| -rw-r--r-- | ggml-blas.h | 23 |
1 files changed, 0 insertions, 23 deletions
diff --git a/ggml-blas.h b/ggml-blas.h deleted file mode 100644 index f2e37de0..00000000 --- a/ggml-blas.h +++ /dev/null @@ -1,23 +0,0 @@ -#pragma once - -#include "ggml.h" -#include "ggml-backend.h" - - -#ifdef __cplusplus -extern "C" { -#endif - -// backend API -GGML_API GGML_CALL ggml_backend_t ggml_backend_blas_init(void); - -GGML_API GGML_CALL bool ggml_backend_is_blas(ggml_backend_t backend); - -// number of threads used for conversion to float -// for openblas and blis, this will also set the number of threads used for blas operations -GGML_API GGML_CALL void ggml_backend_blas_set_n_threads(ggml_backend_t backend_blas, int n_threads); - - -#ifdef __cplusplus -} -#endif |