NousResearch

/

Yarn-Mistral-7b-128k

like 568

Are there any plans to do this for mixtral? also with llama-cpp would 512GB of ram work for cpu inference or would I have to spend another $800 to move up to 1TB?

Thanks