Files
ik_llama.cpp/iqk_mul_mat.cpp
Iwan Kawrakow 29d9bf65f3 iqk_mul_mat: improve iq1_bn (bitnet) on vanilla AVX2
I now get PP-512 = 270 t/s on the Ryzen-5975WX
2024-06-22 12:02:51 +03:00

173 KiB