Skip to content
Discussion options

You must be logged in to vote

感谢反馈!缓存命中率主要取决于每轮请求的前缀是否稳定——版本升级如果调整了系统提示、上下文拼接或消息顺序,就可能打断 prefix 缓存,导致命中率下降。为了定位问题,麻烦点右上角的反馈按钮上报,并附上:升级前 / 后的具体版本号、是「同一历史会话续聊」还是「新建会话」、以及是否经过中转 / 代理。我们据此排查。感谢!

Thanks for the report! Cache hit rate mainly depends on whether each request's prefix stays stable — a version upgrade that changes the system prompt, context assembly, or message ordering can break the prefix cache and lower the hit rate. To help us investigate, please click the feedback button (top-right) and include: the exact versions before/after, whether it's a continued history session vs. a new one, and whether a proxy/relay is involved. Thank you!

Replies: 1 comment

Comment options

You must be logged in to vote
0 replies
Answer selected by tcp404
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants