Model Overview

Statuo/Deepseeker-Kunou-Qwen2.5-14b is a 14.8 billion parameter language model developed by Statuo, leveraging the Qwen2.5 architecture. This model is the result of a multi-stage linear merge process, combining deepseek-ai/DeepSeek-R1-Distill-Qwen-14B and Sao10K/14B-Qwen2.5-Kunou-v1 to enhance its overall intelligence and creative writing proficiency.

Key Characteristics

• Architecture: Based on the Qwen2.5 family, known for its strong performance in various language tasks.
• Merge Strategy: Utilizes a linear merge method, carefully balancing the contributions of its constituent models.
• Context Length: Features a substantial context window of 131072 tokens, supporting extensive conversational and document processing tasks.
• Performance Focus: Aims to improve upon the base intelligence of the Kunou model, particularly in areas where DeepSeek-R1-Distill showed strengths beyond creative writing.

Considerations for Use

While designed for enhanced intelligence, the model may occasionally exhibit the Qwen tendency to prematurely output the End-of-Sequence (EoS) tag. It is released under the Qwen License.