ArnavM3434/swe-3b-backdoor-base
Qwen2.5-Coder-3B-Instruct is a 3.09 billion parameter instruction-tuned causal language model developed by Qwen, part of the Qwen2.5-Coder series. This model features a 32,768 token context length and is specifically optimized for code generation, code reasoning, and code fixing. It builds upon the Qwen2.5 architecture, with training scaled up to 5.5 trillion tokens including source code and synthetic data, making it suitable for real-world coding applications and Code Agents.
Loading preview...
Qwen2.5-Coder-3B-Instruct Overview
Qwen2.5-Coder-3B-Instruct is an instruction-tuned causal language model from the Qwen2.5-Coder series, developed by Qwen. This model, with 3.09 billion parameters and a 32,768 token context length, represents an advancement over CodeQwen1.5, focusing on enhanced coding capabilities.
Key Capabilities
- Significant improvements in code generation, code reasoning, and code fixing.
- Trained on an extensive dataset of 5.5 trillion tokens, including source code, text-code grounding, and synthetic data.
- Maintains strong performance in mathematics and general competencies alongside its coding enhancements.
- Designed as a comprehensive foundation for real-world applications such as Code Agents.
- Utilizes a transformer architecture with RoPE, SwiGLU, RMSNorm, Attention QKV bias, and tied word embeddings.
Good For
- Developers requiring a robust model for code-centric tasks.
- Applications involving automated code generation, debugging, or refactoring.
- Building Code Agents that require strong coding and general reasoning abilities.
- Use cases benefiting from a large context window for complex codebases.