llama.cpp
ggml : Implementations for Q4_0_8_8 quantization based functions - RISC-V vector version
#10029
Merged
Merged
ggml : Implementations for Q4_0_8_8 quantization based functions - RISC-V vector version #10029
ggerganov
merged
7 commits into
ggml-org:master
from
xctan:rvv_q4_0_8x8
ggml : RISC-V vector gemv for q4_0_8x8
9bfecf42
ggml : Added WIP rvv q4_0_8x8 gemm
3f7fdf24
ggml : Added initial implementation of rvv gemm
238cd667
ggml : optimize gemm to avoid register spillover
c039415e
ggml : Fix GCC rvv load alignment issue
78c78e2a
ggerganov
approved these changes
on 2024-10-25
ggml : Format gemm rvv code
37057a0f
ggml : Fix a typo in RVV q4_0_8_8 GEMM
274a7720
Login to write a write a comment.
Login via GitHub