metterian

/

llama-3-pro-ko-9b

like 0

Hi @metterian

I am trying to use this model for a multi-lingual korean capable application.
The outputs seem a bit weird though.

Is this a completions model rather than a chat model?
How big was the training corpus?

Is there an associated research paper with it?
I am guessing it is related in training technique and approach to your other model:
https://huggingface.co/metterian/llama-pro-ko-8b

Thanks heaps!