llama.cpp
CUDA: Add launch bounds for Pascal, small q4_K, q5_K refactor
#2596

Merged

CUDA: Add launch bounds for Pascal, small q4_K, q5_K refactor #2596

JohannesGaessler merged 1 commit into ggml-org:master from JohannesGaessler:cuda-faster-mmq-10

JohannesGaessler force pushed 3 years ago

slaren commented on 2023-08-14

slaren approved these changes on 2023-08-14

CUDA: launch_bounds, small q4_K, q5_K mmq refactor

e9505187

JohannesGaessler force pushed to e9505187 3 years ago

JohannesGaessler merged 1cd06fa2 into master 3 years ago

Reviewers

slaren

Assignees

No one assigned

Labels

None yet

Milestone

No milestone