Kawrakow 5236c98b41 CUDA: MMQ for iqX_r4 quants (#557) ...

* cuda: MMQ for iq2_k_r4

* cuda: MMQ for iq3_k_r4

* cuda: MMQ for iq4_k_r4

* cuda: MMQ for iq5_k_r4

* iqk_r4 quants: use MMQ only for batches < 1024 tokens

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>

2025-06-26 08:50:49 +02:00

..

cmake

Merge mainline llama.cpp (#3)

2024-07-27 07:55:01 +02:00

include

Fix non rpc build error (#506)

2025-06-08 17:27:00 +03:00

src

CUDA: MMQ for iqX_r4 quants (#557)

2025-06-26 08:50:49 +02:00

.gitignore

Merge mainline llama.cpp (#3)

2024-07-27 07:55:01 +02:00

CMakeLists.txt

Better strategy for GPU offload (#520)

2025-06-12 19:25:11 +03:00