llama.cpp
CUDA : faster k-quant dot kernels
#1862
Merged

CUDA : faster k-quant dot kernels #1862

ikawrakow merged 5 commits into master from ik/cuda-faster-k-quants
ikawrakow
cuda : faster k-quant dot kernels
dc67f1a0
ikawrakow ikawrakow requested a review from slaren slaren 2 years ago
ikawrakow ikawrakow requested a review from JohannesGaessler JohannesGaessler 2 years ago
JohannesGaessler
JohannesGaessler commented on 2023-06-14
JohannesGaessler
slaren
slaren approved these changes on 2023-06-14
ikawrakow
JohannesGaessler
ggerganov
ikawrakow
johnson442
maddes8cht
ikawrakow
JohannesGaessler
Imrove Q2_K dot kernel on older GPUs
7ced1971
Imrove Q6_K dot kernel on older GPUs
3edee085
Add LLAMA_CUDA_KQUANTS_ITER to CMakeLists.txt and Makefile
31b20758
ikawrakow
ggerganov
ggerganov commented on 2023-06-16
ggerganov
ggerganov commented on 2023-06-16
johnson442
PR comments
0dc0b699
ikawrakow ikawrakow merged 3d011226 into master 2 years ago
ikawrakow ikawrakow deleted the ik/cuda-faster-k-quants branch 2 years ago
johnson442
mirek190
dragonfyre13
cmp-nct
mirek190

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone