NostraEmpire/mirror-qwen2.5-coder-7b-instruct
NostraEmpire/mirror-qwen2.5-coder-7b-instruct is a 7.61 billion parameter instruction-tuned causal language model from the Qwen2.5-Coder series, developed by Qwen. This model is specifically optimized for code generation, code reasoning, and code fixing, building upon the strong Qwen2.5 foundation with 5.5 trillion training tokens including source code. It features a full 131,072 token context length with YaRN support for long texts, making it highly suitable for complex coding tasks and code agent applications.
Loading preview...
Overview
NostraEmpire/mirror-qwen2.5-coder-7b-instruct is a 7.61 billion parameter instruction-tuned model from the Qwen2.5-Coder family, developed by Qwen. This series, formerly known as CodeQwen, represents the latest advancements in code-specific large language models. It is built on the robust Qwen2.5 architecture and has been extensively trained on 5.5 trillion tokens, incorporating source code, text-code grounding, and synthetic data.
Key Capabilities
- Enhanced Code Performance: Significantly improves upon previous versions in code generation, reasoning, and fixing.
- Long-Context Support: Features a full 131,072 token context length, utilizing YaRN for optimal performance with extensive inputs.
- General Competency: While specialized for code, it maintains strong capabilities in mathematics and general language understanding.
- Architecture: Employs transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias.
Good For
- Code Agents: Provides a comprehensive foundation for real-world applications requiring advanced coding abilities.
- Complex Coding Tasks: Excels in scenarios demanding high-quality code generation and debugging.
- Long Codebase Analysis: Its extended context window is ideal for processing and understanding large code files or projects.