Please gemma 4 queen 31b qat imatrix uncensored gguf
ufff, sorry, i just figured out mrademacher only does quants and imatrix... and you have to link the already made model to him. I´m new, will not happen again.
Yes of course, how else would I quantize hehe? I cant get a model from a thin air, you need to give me an existing model so I can convert it to gguf
It would be possible for you to convert the MTP header of a model you already converted?
https://huggingface.co/Avesed/Qwopus3.6-27B-v2-abliterated/tree/main
the model-mtp.safetensors to model-mtp.gguf to use it on llama. Allegedly it seems easy to do with the tools the folder already has, but I can´t figure how.
I dont know if we can do it without redoing the whole model, and a lot of mtp fails as well, and we quanted this model a day ago
I understand, thanks anyways.