Together AI 发布 Aurora：基于强化学习的自适应投机解码框架

AI 摘要

Together AI 发布了开源框架 Aurora，基于强化学习，从实时推理轨迹中学习，持续更新投机解码中的草稿模型，无需中断服务。实验表明，Aurora 在 Qwen3 和 Llama3 等模型上比静态草稿模型额外获得 1.25 倍加速，并降低了基础设施成本。该框架支持异步训练和热替换，将投机解码转变为动态自改进的飞轮。

正文节选

Speculative decoding goes stale in production — draft models can drift and offline retraining can't always keep pace. Aurora fixes this. It's an open-source, RL-based framework that learns directly from live inference traces and continuously updates the speculator without interrupting serving. Key results: → Real-time adaptation across shifting traffic domains → 1.25x additional speedup over a well-trained static speculator The headline finding: online training from scratch can outperform a care

发布时间：—

抓取时间：2026-08-03 01:13

来源机构：Together AI