ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-04-29 02:41:47 +00:00
Code Issues Packages Projects Releases Wiki Activity

IQ4_NL_X4 (#118)

Browse Source 
* Adding iq4_nl_x4

Looks very promising - I get PP-512(LLaMA-3.1-8B) = 230 t/s
on the Ryzen-7950X! This is faster than any other quant and
~40% faster than iq4_nl.

* iq4_nl_x4: getting amazing

This Zen4 variant gets us to PP-512(LLaMA-3.1-8B) = 263 t/s!

* iq4_nl_x4: AVX2

Here we gain only 25% compared to iq4_nl

* iq4_nl_x4: NEON

On M2-Max we get PP-512(LLaMA-3.1-8B) = 109.7 t/s, up from
82.4 t/s for iq4_nl.

* iq4_nl_x4: minor NEON improvement and cleanup

This gets us to 110.3 t/s. In comparison,
IQ4_NL_4_4 in mainline llama.cpp achieves 92.3 t/s.

* iq4_nl_x4: NEON specialization for matrix x vector

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2024-12-02 07:25:39 +01:00
committed by GitHub
parent 6c73f704ca
commit 72d94fbf22
10 changed files with 398 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/include/ggml.h
View File

@@ -406,6 +406,8 @@ extern "C" {
GGML_TYPE_IQ4_KS = 144,
GGML_TYPE_IQ2_KS = 145,
GGML_TYPE_IQ4_KSS = 146,
GGML_TYPE_IQ4_NL_X4 = 220,
GGML_TYPE_COUNT,
};
@@ -464,6 +466,8 @@ extern "C" {
GGML_FTYPE_MOSTLY_IQ4_KS = 137, // except 1d tensors
GGML_FTYPE_MOSTLY_IQ2_KS = 138, // except 1d tensors
GGML_FTYPE_MOSTLY_IQ4_KSS = 139, // except 1d tensors
//
GGML_FTYPE_MOSTLY_IQ4_NL_X4 = 219, // except 1d tensors
};
// available tensor operations: