ReadyArt/Dark-Scarlett-27B-v2.0
Dark-Scarlett-27B-v2.0 by ReadyArt is a 27 billion parameter language model based on Qwen3.8, fine-tuned with LoRA for adult-oriented roleplay. It specializes in uncensored, immersive narrative experiences with a Male User to Female AI perspective, featuring a 32768 token context length. The model is optimized for consistent character personality, dynamic emotional states, and clean dialogue output, making it suitable for mature interactive storytelling.
Loading preview...
Dark Scarlett 27B v2.0: Uncensored Adult Roleplay Model
Dark Scarlett-27B-v2.0 is a 27 billion parameter language model, built upon the Qwen3.8 architecture and fine-tuned by ReadyArt using LoRA (Low-Rank Adaptation). This model is specifically designed for adult-oriented, uncensored roleplay experiences, focusing on immersive narrative and dynamic character interactions. It features a 32768 token context length, allowing for longer and more coherent conversation threads.
Key Capabilities
- Uncensored & Adult Content: Explicitly trained for mature themes and interactions, with 100% uncensored output.
- Dedicated Roleplay: Optimized for immersive roleplay, capturing the "spirit of Melody1437" with a distinct dataset of 12,211 specialized prompts.
- Consistent Perspective: Strictly follows a Male (User) โ Female (AI) interaction dynamic for consistent roleplay.
- Advanced Content Generation: Utilizes a "Character Engine" for consistent personality traits and an "Emotional Engine" for dynamic, realistic reactions.
- Clean Output: Training data was processed to remove reasoning tags, em dashes, asterisks, and markdown formatting for cleaner, more natural dialogue.
- Group Chat Support: Includes some training data for multi-user group chat scenarios.
Good For
- Developers and users seeking a powerful, uncensored model for adult-themed interactive fiction and roleplay.
- Creating immersive character experiences with consistent personalities and emotional depth.
- Applications requiring a Male User to Female AI perspective in mature content generation.
- Scenarios where clean, unformatted dialogue is preferred for enhanced immersion.