精选理由
科大讯飞开源了百万token上下文的边缘模型，比普通模型能处理更长文本，适合设备端部署。
科大讯飞全资子公司发布并开源了Spark X2.5-4B和X2.5-1.7B模型。这两款模型原生支持高达100万token的上下文长度。它们是首批支持如此长上下文的边缘模型。这些模型专为设备端AI设计。
原文 · pandaily
iFLYTEK Open-Sources Million-Token Context Models for On-Device AI
iFLYTEK's wholly-owned subsidiary launched and open-sourced Spark X2.5-4B and X2.5-1.7B, which it says are the first edge models to natively support up to one million tokens of context.