llama.cpp
vulkan: dynamic subgroup size for the remaining k quants
#10745

Merged

vulkan: dynamic subgroup size for the remaining k quants #10745

0cc4m merged 2 commits into ggml-org:master from vulkan2

q5_k

c2aa654a

revert as multi row isnt faster for k quants

8ee6beea

netrunnereve requested a review from

0cc4m 283 days ago

github-actions added Vulkan

github-actions added ggml

0cc4m approved these changes on 2024-12-10

jeffbolznv approved these changes on 2024-12-10

0cc4m merged dafae66c into master 283 days ago

netrunnereve deleted the vulkan2 branch 282 days ago

Reviewers

jeffbolznv

0cc4m

Assignees

No one assigned

Labels

Vulkan ggml

Milestone

No milestone