diff options
| author | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-07-31 08:49:44 +0200 |
|---|---|---|
| committer | Kawrakow <48489457+ikawrakow@users.noreply.github.com> | 2024-08-01 09:38:06 +0200 |
| commit | 394ed3913c16ad9baae24e93e126847030063fad (patch) | |
| tree | 3c1ee079d2edf1ba1d4ef297ea8d53aee4da6971 /ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu | |
| parent | 062313dab41381c6170175ea0c2075b2328b6f33 (diff) | |
iq3_k: slightly faster Metal dequantize kernel
PP-512 goes to 473 t/s up from 452 t/s.
Diffstat (limited to 'ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu')
0 files changed, 0 insertions, 0 deletions