Qwen 2.5 72B Instruct

Qwen2.5 72B is the latest in the Qwen large language model series, offering significant improvements in knowledge, coding, and mathematics. It features specialized expert models, improved instruction following, long-text generation (over 8K tokens), structured data understanding, and robust multilingual support for over 29 languages. The model is optimized for resilience to diverse system prompts and enhanced role-play implementation.