3 hours ago

一个可以在A5000显卡或4090上完整运行的QWen3.5大模型(上下文64k),具备调用工具的能力,适合本地部署龙虾和Hermesk

tools thinking
bbee8f0adac8 · 162B
Comprehensive reasoning mode. If request is simple, then just answer. Otherwise, analyze the request, consider edge cases, and provide a high-fidelity solution.