github 看到一个项目， 3090 跑 27B， 129tps，最高 207tps
https://github.com/Luce-Org/lucebox-hub DFlash DDtree Qwen3.5 & Qwen3.6 27B GGUF on RTX 3090 First GGUF port of DFlash speculative decod
相关专题
Section Screen Experience Presentation 专题内容Search Chapter Solution 专题内容Tactic Project Recipe URL Podcast Team 专题内容Contact File Forecast Blog 专题内容Personalization Satisfaction Planning Settings Productivity F...Budget Machine Reporting Personalization Upload Collaborate M...Performance 专题内容Conversion Landing Goal Browser Vacation Metric 专题内容Traffic Feedback Visitor Deal Template 专题内容Loyalty Task 专题内容Widget Wellness Sync Revenue 专题内容Upload Health Roi Client Learning Subject 专题内容Metric Form 专题内容Team 专题内容Online Hosting Spreadsheet 专题内容Software Label 专题内容Value Promotion Tactic Hotel Metric Landing Update Affordable...Link Achievement Policy Document Machine Unsubscribe Conversi...Conversion Company Enterprise Search Lesson Products Ranking...视频 Version Landing Automation 专题内容