跳到正文
RCreddit.com·
暂不在当前实时榜单

UkisAI Swift-Qwen3.8-27B / -58.3% thinking, x1.95 speed while keeping the accuracy of xhigh

AI 摘要

UkisAI has post-trained Qwen 3.8 27B, creating Swift-Qwen3.8-27B, which achieves a 1.95x speed-up and 58% fewer thinking tokens while maintaining xhigh accuracy. This was accomplished by penalizing tokens linked to overthinking and using On-Policy Distillation. Benchmarks show significant reductions in thinking tokens across various tasks, with minimal accuracy changes. UkisAI is also working on Swift 3.8 Flash Next, aiming for further reductions in thinking token usage.

为什么是这条

This report details a method to reduce thinking tokens by 58% and increase speed by 1.95x in Qwen 3.8 27B, unlike other optimizations that often sacrifice accuracy.

时间与来源

时间显示为 UTC

显示时区:UTC

本地时区尚不可用,暂时显示 UTC。

收录当时偏移:UTC+02026年9月14日 17:01 UTC

收录
2026年9月14日 17:01
来源类型
开发者社区

本站未收录正文。

前往源站阅读 →
来源·reddit.com