Kawrakow 9119023a4b IQ4_XS_R4 (#123) ...

* Adding iq4_xs_r4

This is a 1st working version on Zen4.
We get PP-512(LLaMA-3.1-8B) = 226 t/s, so 16% slower
than iq4_nl_x4.

* iq4_xs_r4: WIP

* iq4_xs_r4: Use AVX2 version for matrix x vector on Zen4

* iq4_xs_r4: NEON

We get PP-512(LLaMA-3.1-8B) = 115.6 t/s on M2-Max,
up from 68.2 t/s for iq4_xs!

* DRY

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>

2024-12-04 15:20:07 +01:00

..

llama.h

IQ4_XS_R4 (#123)

2024-12-04 15:20:07 +01:00