root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-26 11:24:35 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Implemented vector length agnostic SVE using switch case for 512-bit, 256-bit, 128-bit vector lengths

Browse Source
This commit is contained in:
vithulep 2024-09-03 12:17:44 +05:30
parent 4dbdb6c82f
commit 6a6cfd6c6f
1 changed files with 0 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ggml/src/ggml-quants.c
View File

@ -3821,7 +3821,6 @@ void ggml_vec_dot_q4_0_q8_0(int n, float * restrict s, size_t bs, const void * r
svfloat32_t sumv0 = svdup_n_f32(0.0f); svfloat32_t sumv0 = svdup_n_f32(0.0f);
svfloat32_t sumv1 = svdup_n_f32(0.0f); svfloat32_t sumv1 = svdup_n_f32(0.0f);
assert(nb % 2 == 0); // TODO: handle odd nb
const int vector_length = ggml_sve_cnt_b*8; const int vector_length = ggml_sve_cnt_b*8;
// VLA Implementation using switch case // VLA Implementation using switch case
@ -5402,7 +5401,6 @@ void ggml_vec_dot_q8_0_q8_0(int n, float * restrict s, size_t bs, const void * r
svfloat32_t sumv0 = svdup_n_f32(0.0f); svfloat32_t sumv0 = svdup_n_f32(0.0f);
svfloat32_t sumv1 = svdup_n_f32(0.0f); svfloat32_t sumv1 = svdup_n_f32(0.0f);
assert(nb % 2 == 0); // TODO: handle odd nb
const int vector_length = ggml_sve_cnt_b*8; const int vector_length = ggml_sve_cnt_b*8;
//VLA Implemenation for SVE //VLA Implemenation for SVE