ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-01-31 03:29:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

Adding IQ1_TN - 1.6875 bpw for TriLM ternary models (#44)

Browse Source 
* Adding iq1_tn - 1.6875 bpw for TriLM ternary models

* iq1_tn: NEON

* iq1_tn: faster NEON

* iq2_bn: improve performance on NEON

We now get TG-128 = 100 t/s for Bitnet-3B-1.58b!

* iq1_tn: improve AVX2

PP-512 goes to 533 t/s up from 455.
TG-128 @ 2 threads goes to 16.6 t/s up from 14.2.
However, we seem to have a bottleneck somewhere as
TG saturates at 8 threads.

* iq1_tn: improve Zen4

PP-512 goes to 485 t/s up from 352. With FA we get 545 t/s up from 380.
TG-128 @ 1 thread goes to 12.4 t/s up from 10.4.
However, we seem to have a bottleneck somewhere as
TG saturates at 8 threads.

* iq2_bn: improve on Zen4

We now get PP-512 = 614 t/s up from 542 t/s

* iq2_bn: improve AVX2 implementation

We now get PP-512 = 753 t/s up from 680 t/s.

* Remove unnecessary barrier in ggml_compute_forward_mul_mat

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2024-09-09 14:56:34 +03:00
committed by GitHub
parent f2ef628e4e
commit 4a5d5e207d
10 changed files with 304 additions and 148 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
examples/quantize/quantize.cpp
View File

@@ -28,6 +28,7 @@ static const std::vector<struct quant_option> QUANT_OPTIONS = {
{ "IQ1_M", LLAMA_FTYPE_MOSTLY_IQ1_M, " 1.75 bpw quantization", },
{ "IQ1_BN", LLAMA_FTYPE_MOSTLY_IQ1_BN, " 1.62 bpw quantization (Bitnet)", },
{ "IQ2_BN", LLAMA_FTYPE_MOSTLY_IQ2_BN, " 2.00 bpw quantization (Bitnet)", },
{ "IQ1_TN", LLAMA_FTYPE_MOSTLY_IQ1_TN, " 1.69 bpw quantization (TriLM)", },
{ "IQ2_TN", LLAMA_FTYPE_MOSTLY_IQ2_TN, " 2.06 bpw quantization (TriLM)", },
{ "Q2_K", LLAMA_FTYPE_MOSTLY_Q2_K, " 2.63G, +0.6717 ppl @ LLaMA-v1-7B", },
{ "Q2_K_S", LLAMA_FTYPE_MOSTLY_Q2_K_S, " 2.16G, +9.0634 ppl @ LLaMA-v1-7B", },