llama.cpp
Q6_K AVX improvements
#10118
Merged

Q6_K AVX improvements #10118

slaren merged 10 commits into ggml-org:master from q6_k
netrunnereve
netrunnereve q6_k instruction reordering attempt
499e9f2f
netrunnereve better subtract method
e3a34321
netrunnereve should be theoretically faster
0b75215f
netrunnereve optimize bit fiddling
a420e4cd
netrunnereve handle -32 offset separately. bsums exists for a reason!
35255d64
netrunnereve use shift
5b367158
netrunnereve Merge branch 'ggerganov:master' into q6_k
ed6f845a
netrunnereve Update ggml-quants.c
d84c372b
github-actions github-actions added ggml
slaren
slaren
slaren approved these changes on 2024-11-01
Nexesenex
netrunnereve
netrunnereve Merge branch 'ggerganov:master' into q6_k
4ec3e4a5
netrunnereve have to update ci macos version to 13 as 12 doesnt work now. 13 is st…
f85336e2
github-actions github-actions added devops
netrunnereve
slaren slaren merged 34073647 into master 1 year ago
SmallAndSoft
netrunnereve netrunnereve deleted the q6_k branch 1 year ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone