ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-03-03 02:20:01 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
3b6597c7a180bcbd780ec3b6dcfa1f0e35321231
ik_llama.cpp/include
History
Iwan Kawrakow 3da565c9c9 iq1_kt: CUDA dequantize 
Testing with LlaMA-3.1-8B-Instruct, we get almost the same PPL
as iq2_xxs, so about 0.2 bpw fewer bits for the same quality.
2025-07-16 16:55:20 +03:00
..
llama.h
iq1_kt: CUDA dequantize
2025-07-16 16:55:20 +03:00