diff options
| author | Georgi Gerganov <ggerganov@gmail.com> | 2024-01-02 21:07:47 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-01-02 21:07:47 +0200 |
| commit | f3f62f0d835d559e80714bbeb05d03125574e3dd (patch) | |
| tree | f92c9a828a38a6f4c681635ba08a4e32926434f3 /tests/test-backend-ops.cpp | |
| parent | 0ef3ca2ac62016c0c545de1c89dc2e3e130f4a99 (diff) | |
metal : optimize ggml_mul_mat_id (faster Mixtral PP) (#4725)
* ggml : disable fast-math for Metal (cmake build only)
ggml-ci
* metal : fix Metal API debug warnings
* cmake : add -fno-inline for Metal build (#4545)
* metal : fix API debug warnings
* metal : fix compile warnings
* metal : use uint64_t for strides
* cmake : rename option to LLAMA_METAL_SHADER_DEBUG
* metal : fix mat-vec Q8_0 kernel for BS > 1
* metal : normalize mat-vec kernel signatures
* cmake : respect LLAMA_QKK_64 option
* metal : fix mat-vec Q4_K kernel for QK_K == 64
* metal : optimizing ggml_mul_mat_id (wip)
* metal : minor fix
* metal : opt mul_mm_id
Diffstat (limited to 'tests/test-backend-ops.cpp')
0 files changed, 0 insertions, 0 deletions