Prompt Format

#3
by Ransom - opened

Hi,
Thanks for releasing this model into the wild!
A couple requests:

  1. Could you please add a prompt format to your model card? Are we to assume that your prompt format followed standard Llama2, or is it a different one?
  2. Could you release 4 and 5 bit quantizations as per Llama.cpp? This will increase uptake of your model in the community.

Thanks Kindly

Also: Does this work with Llama.cpp out of the box at all? Or is your interpolation method completely custom?

Sign up or log in to comment