vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

1.5 bit quantization (#5453)

Browse source 
* iq1_s: WIP basics

* iq1_s: CUDA is working

* iq1_s: scalar CPU dot product

* iq1_s: WIP AVX2 dot product - something is not right

* Fix tests

* Fix shadow warnings

* Fix after merge with latest master

* iq1_s: AVX2 finally works

* iq1_s: ARM_NEON dot product. Works, but not very fast

* iq1_s: better grid

* iq1_s: use IQ2_XXS for attn_output

At a cost of 0.04 extra bpw this gives a big improvement in PPL.

* iq1_s: Metal basics

Dequantize works, but not dot product

* iq1_s: Metal works, but quite slow

As usual, Apple Silicon does not like the code I write.

* iq1_s: Tests

* iq1_s: slightly faster dot product

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow 2024-02-18 18:16:55 +02:00 • committed by GitHub
parent c8e0d7efeb
commit bd2d4e393b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
12 changed files with 1286 additions and 48 deletions
Download patch file Download diff file Expand all files Collapse all files

6
examples/quantize/quantize.cpp
View file

@ -23,6 +23,7 @@ static const std::vector<struct quant_option> QUANT_OPTIONS = {
{ "Q5_1", LLAMA_FTYPE_MOSTLY_Q5_1, " 4.70G, +0.0349 ppl @ LLaMA-v1-7B", },
{ "IQ2_XXS",LLAMA_FTYPE_MOSTLY_IQ2_XXS," 2.06 bpw quantization", },
{ "IQ2_XS", LLAMA_FTYPE_MOSTLY_IQ2_XS, " 2.31 bpw quantization", },
{ "IQ1_S", LLAMA_FTYPE_MOSTLY_IQ1_S, " 1.56 bpw quantization", },
{ "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", },
{ "IQ3_XXS",LLAMA_FTYPE_MOSTLY_IQ3_XXS," 3.06 bpw quantization", },
@ -287,9 +288,10 @@ int main(int argc, char ** argv) {
}
}
if ((params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XS || params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XXS || params.ftype == LLAMA_FTYPE_MOSTLY_Q2_K_S) && imatrix_data.empty()) {
if ((params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XS || params.ftype == LLAMA_FTYPE_MOSTLY_IQ2_XXS ||
params.ftype == LLAMA_FTYPE_MOSTLY_Q2_K_S || params.ftype == LLAMA_FTYPE_MOSTLY_IQ1_S) && imatrix_data.empty()) {
fprintf(stderr, "\n===============================================================================================\n");
fprintf(stderr, "Please do not use IQ2_XXS, IQ2_XS or Q2_K_S quantization without an importance matrix\n");
fprintf(stderr, "Please do not use IQ1_S, IQ2_XXS, IQ2_XS or Q2_K_S quantization without an importance matrix\n");
fprintf(stderr, "===============================================================================================\n\n\n");
return 1;
}