GPTQ 4bit 128g

#7
by pszemraj - opened

Hi! In case anyone finds useful, I made a 4bit quantized version of this using GPTQ and 7,500 examples from the open assistant dataset to guide the quantization. Check it out below & there’s a demo/usage guide on the model card:

https://huggingface.co/pszemraj/stablelm-7b-sft-v7e3-autogptq-4bit-128g

Sign up or log in to comment