llama.cpp
Slight quantization improvement for Q4_K and Q5_K
#5361

Merged

Slight quantization improvement for Q4_K and Q5_K #5361

ikawrakow merged 2 commits into master from ik/q4k_tuning

Q4_K: slightly better quantization

f58d49e5

Q5_K: slightly better quantization

d3cc1533

ggerganov approved these changes on 2024-02-06

ikawrakow merged f57fadc0 into master 1 year ago

ikawrakow deleted the ik/q4k_tuning branch 1 year ago

Reviewers

ggerganov

Assignees

No one assigned

Labels

None yet

Milestone

No milestone