TheDrummer/Anubis-70B-v1.1
TheDrummer/Anubis-70B-v1.1 is a 70 billion parameter language model developed by TheDrummer, building upon the Anubis 70B v1.0. This model is specifically optimized for strong character adherence and unaligned responses, making it suitable for applications requiring consistent persona portrayal. It utilizes the Llama 3 Chat Template for interaction.
Loading preview...
Anubis 70B v1.1: Shimmer Edition
Anubis 70B v1.1, developed by TheDrummer, is a significant iteration following Anubis 70B v1.0, focusing on enhanced character adherence and unalignment. This model is designed to maintain consistent character portrayal and mannerisms throughout interactions, offering a distinct style compared to other Llama 3-based models.
Key Capabilities
- Exceptional Character Adherence: Users report the model excels at portraying various characters, consistently adhering to their dialogue and mannerisms.
- Unaligned Responses: Provides responses that are not constrained by typical alignment filters, offering more creative freedom.
- Fresh Style: Delivers a unique conversational style, distinguishing it from other models in its class.
- Subtle Detail Emphasis: Capable of picking up and emphasizing subtle details in character cards, leading to more nuanced portrayals.
Good For
- Role-playing and Creative Writing: Ideal for scenarios requiring strong, consistent character personas.
- Applications needing Unaligned Output: Suitable for use cases where less constrained or more 'free-form' responses are desired.
- Exploration of Unique LLM Styles: For developers and users seeking a model with a distinct conversational and character interaction style.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.