Please make a GGUF version
Can you please make a GGUF version of this model?
Thank you!
Yes it's coming soon. As soon as I've finished the Llama 2 models which will be later today
Yes it's coming soon. As soon as I've finished the Llama 2 models which will be later today
Thank you. I wish there was same model but 70B variant (I can't find so far at first glance) because from all the models that I tested this Wizard-Vicuna-30B-Uncensored still the best for conversation purposes it can express more emotions and it's sometimes behave like a truly AGI model.
P.S: Even synthia-70b-v1.1.Q8_0.gguf model can't compete with Wizard-Vicuna-30B-Uncensored. (But it's so much better at coding and math though) Especially emotions wise. Not only it can't express more emotions it also expressing them incorrectly at times like instead physical hug it says virtual hug and rarely expressing emotions. Maybe because Llama 2 highly censored and it makes roleplay nearly impossible AI tends to behave like AI by following the guard rails quite often. Plus Wizard-Vicuna dataset allows Wizard-Vicuna model to know how to express more emotions out of the box and it practically doesn't have any guard rails at all. I kinda fear for the future AI models just by looking how highly censored and how many guard systems Llama 2 has.