ik_llama_opt/ggml
Kawrakow 6b809ca0e1 Bug fix in activation quantization
I added a change in the last PR how activations are quantized.
It looked like it is working and slightly improving performance.
But I now hit an edge case where I get gibberish that goes away if
I remove the change. I absolutely don't see what goes wrong, so
leaving the change in commented out for now.
2025-02-15 19:50:53 +02:00
..
cmake Merge mainline llama.cpp (#3) 2024-07-27 07:55:01 +02:00
include Use Q8_K_128 for IQ1_S_R4 and IQ1_M_R4 matrix multiplications (#194) 2025-02-09 09:14:52 +02:00
src Bug fix in activation quantization 2025-02-15 19:50:53 +02:00
.gitignore Merge mainline llama.cpp (#3) 2024-07-27 07:55:01 +02:00
CMakeLists.txt FA: Add option to build all FA kernels (#197) 2025-02-09 18:59:33 +02:00