Quantization

#1
by shadow123 - opened

Can you make a GPTQ Int8 for it? Thx

Can you make a GPTQ Int8 for it? Thx

Maybe you are looking for this: https://huggingface.co/bartowski/huihui-ai_DeepSeek-R1-Distill-Llama-70B-abliterated-GGUF

The Q8_0 version takes 74.98GB, and the page also provide a model size range down to 16.75GB.

Sign up or log in to comment