Anyone got enough ram to benchmark deepseek r1 quants in llama cpp, with some layers offloaded to gpu? im debating buying more ram for this but dont wanty to buy it then the performance be trash lol unsloth/DeepSeek-R1-GGUF · Hugging Face you need like 200gb
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.