Are GPTQ's group-size, --true-sequential and act-order supported by llama.cpp?

[x02Sylvie]

As mentioned here https://github.com/qwopqwop200/GPTQ-for-LLaMa

These three can greatly improve results, so im wondering

[BadisG]

Groupsize yeah #545
For the two other we haven't tested it yet, someone has to convert the HF weights into .pt with those two parameters first, I can't I don't have enough vram :'(