[nacs]

What model and hardware powers this?

Is this a Google T5 based model?

[pella]

3bit hard-wired Llama 3.1 8B ( https://taalas.com/the-path-to-ubiquitous-ai/ )

[cyansmoker]

3bit is a bit ridiculous. From that page I am unclear if the current model is 3 or 4bit. If it’s 4bit… well, NVIDIA showed that a well organized model can perform almost as well as 8bit.