Iwan Kawrakow a16f961033 Repack q4_0 and q8_0 to q8_0_R8 ...

q8_0 is fine, but I observe a very significant PPL increase
for q4_0. Best guess: precision loss with the 32 bit <-> 16 bit
scale conversions.

2025-06-18 08:46:47 +03:00

..

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

Repack q4_0 and q8_0 to q8_0_R8

2025-06-18 08:46:47 +03:00

.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