ik_llama_opt/ggml
Kawrakow 250c325e7e iq3_k: fix and optimize Metal dot product (#87)
* iq3_k: fix Metal dot product

I was accessing the scales as 4-byte aligned, but iq3_k is
not 4-byte aligned. Instead of throwing an error (as it happens
on CUDA when one makes this mistake), Metal silently accepts
and we get garbage.

* iq3_k: slightly faster Metal dot product

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
2024-10-14 10:46:41 +03:00
..
cmake Merge mainline llama.cpp (#3) 2024-07-27 07:55:01 +02:00
include IQ2_KS: 2.1875 bpw non-linear quantization (#85) 2024-10-13 13:34:30 +03:00
src iq3_k: fix and optimize Metal dot product (#87) 2024-10-14 10:46:41 +03:00
.gitignore Merge mainline llama.cpp (#3) 2024-07-27 07:55:01 +02:00
CMakeLists.txt Move to c++17 projectwide (#80) 2024-10-04 14:43:26 +03:00