Skip to main content

Llama 3.3 70B too restrictive

i find that modèle way too much restrictive and go against all your philosophy at the studio why even allow that model to be in the character features mode if it got the most restrictive guidelines i ever seen,that model don’t even follow it’s prompts and it go everything your team believe in,the existence of that version of Llama is a disgrace on itself,do something about that model because you are an unrestricted and uncensored platform.

Status: Completed3 comments

Log in to comment and vote

Comments3

  • Red Tape

    •

    Jan 27, 2025

    So Jocelyn did it work? What jack said?

    • Azure Leaf

      •

      Jan 28, 2025

      well not really with the default llama modele it’s way too much rigid and not user friendly if you want to prompt that Model it will still follow his rigid and strict guideline the maker of that model,i use other version way more user friendly like llama 3.1.405B which clearly has been made to follw users prompts at all cost and is more smarter and maliable,i don’t know if someone of venice team can confirm what i’m saying but yea i avoid at all cost the default modele because it remind me of everything i despise about openai chatgpt strict guideline.

  • Jack

    Team•

    Jan 23, 2025

    Have you tried out any uncensored system prompts?

    https://venice.ai/blog/god-mode-the-power-of-ai-system-prompts

    Marking this problem as solved. Reach out via support channel if you need further help on prompting.