llama.cpp
ggml : switch to F32 src1 in the CPU backend
#31854
Merged

ggml : switch to F32 src1 in the CPU backend #31854

ggerganov merged 3 commits into master from gg/cpu-f32-src1-f16-matmul
ggerganov
github-actions github-actions added ggml
github-actions github-actions added testing
ggerganov ggml : switch to F32 src1 in the CPU backend
b1b26cce
ggerganov ggerganov force pushed from 657298be to b1b26cce 3 days ago
ServeurpersoCom
ggerganov
ServeurpersoCom
ServeurpersoCom
ServeurpersoCom cont : add some F16 x F32 vectorized paths
a16ab109
ggerganov ggerganov marked this pull request as ready for review 3 days ago
ServeurpersoCom
ServeurpersoCom approved these changes on 2026-11-07
ngxson
ngxson approved these changes on 2026-11-07
ServeurpersoCom cont : enable the fused F16 x F32 dot on AVX without FMA
ca46ea6c
ServeurpersoCom
ServeurpersoCom
ggerganov
ServeurpersoCom
ggerganov ggerganov merged 12ab5979 into master 3 days ago
ggerganov ggerganov deleted the gg/cpu-f32-src1-f16-matmul branch 3 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone