Kawrakow 1eac9e8487 NEON implementation for trellis quants (#471) ...

* iq2_kt: NEON implementation

* iq3_kt: NEON implementation

* iq4_kt: not working NEON implementation

* iq4_kt: NEON implementation

Have to use f32 arithmetic else I get gibberish?
Correspondigly ridiculously slow.

* Cleanup

* iq4_kt: slightly faster TG on NEON

---------

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

2025-05-29 18:57:41 +03:00

..

cmake

Merge mainline llama.cpp (#3)

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

include

Trellis quants with CPU inference (#441)

2025-05-23 09:17:52 +03:00

src

NEON implementation for trellis quants (#471)

2025-05-29 18:57:41 +03:00

.gitignore

Merge mainline llama.cpp (#3)

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

CMakeLists.txt

Option to enable disable the IQK CPU FA kernels (#429)

2025-05-17 11:21:58 +03:00