This product is a deeply optimized model based on the Deepseek-R1-14B architecture. By injecting a vast amount of novel and role-playing data, combined with chain-of-thought (CoT) training strategies, it significantly enhances the capabilities of role-playing, novel text generation, and long-term contextual coherence. Key advantages include strong contextual consistency, a rich vocabulary for role-playing, and support for complex logical reasoning. This model is tailored for scenarios that require deep creativity and dialogue, such as novel creation and scriptwriting. It is licensed under Apache 2.0 and is currently available for free use, pending adherence to copyright agreements.