What the fuck is this model about?

Fiendish_LLAMA_3B is a 3.2 billion parameter language model from SicariusSicariiStuff, built on the Llama-3-Instruct architecture with a 32768 token context length. It's described as the "naughty sister" of Impish_LLAMA_3B, offering a more edgy and less wholesome experience. The model is particularly noted for its superb roleplay capabilities for its size, generating short (1-2 paragraph) responses in a CAI-style format.

What makes THIS different from all the other models?

This model stands out due to its explicit focus on "fiendish" and "naughty" content, characterized by low refusal rates in both roleplay and assistant tasks. It's designed to provide total freedom in roleplay scenarios that other models might refuse. It also excels at following character cards very well, making it suitable for detailed character-driven narratives. The model uses a specific "Classic Internet RP" format (*action* speech *narration*) and a custom "SICAtxt" format for character and adventure descriptions, which are highly recommended for optimal results.

Should I use this for my use case?

Use this model if:

• You need a compact (3.2B) model for roleplay and creative writing.
• Your use case involves "naughty" or "evil" themes and requires low content refusal.
• You prioritize a model that adheres strictly to character cards.
• You prefer short, formatted responses (1-3 paragraphs).
• You are comfortable using specific generation settings and the recommended "Classic Internet RP" or "SICAtxt" formats for best performance.

Consider alternatives if:

• You require a model with high censorship or strict adherence to ethical guidelines.
• Your primary need is for general-purpose assistant tasks without a focus on creative or edgy content.