Iwan Kawrakow b5f0a2b617 Use new-new-mma also for MLA=3, and use mask bounds ...

This gives us ~25% better PP at 32k tokens compared to main

2025-11-10 09:53:07 +02:00

..

cmake

Merge mainline llama.cpp (#3)

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

include

CUDA: set compute parameters via command line arguments (#910)

2025-11-07 07:11:23 +02:00

src

Use new-new-mma also for MLA=3, and use mask bounds

2025-11-10 09:53:07 +02:00

.gitignore

Merge mainline llama.cpp (#3)

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

CMakeLists.txt

Disable CUDA fusion by default for now (#903)

2025-11-05 10:58:12 +02:00