Skip to content
Discussion options

You must be logged in to vote

"思考强度"(reasoning effort)在 DSH 里的控制位置:

官方模型(deepseek-official)

  • 全局默认:~/.dsh/settings.yamlagent-default-model.reasoningEffort,取值 min | low | medium | high | max
  • 也可以直接在 Web GUI 的模型设置页选
  • 这是会话级偏好,修改后新会话生效

llama-server / OpenAI 兼容端点
DSH 走"自定义 provider"表单(Providers 页的 Custom 表单,支持自定义 base URL + 模型名)。这里有个现实边界需要说明:
reasoningEffort 是 DSH 的会话偏好参数,最终会不会变成请求里的 reasoning_effort/相应字段透传给 llama-server,取决于对应 LLM adapter 的实现——DSH 官方 adapter 是为 DeepSeek 官方 API 写的,第三方端点不一定映射这个参数。llama-server 侧的"思考强度"目前更直接的控制是它自己的参数(如 --reasoning--jinja、chat template 或温度/采样参数)。

可行路径

  1. 在自定义 provider 表单把 base URL 指向 llama-server,先用默认 reasoning 跑通;
  2. 若需要压思考:优先调 llama-server 启动参数(采样/模板层);reasoningEffort 只作为 DSH 侧偏好设置,能否生效以请求抓包为准。

Replies: 6 comments 2 replies

Comment options

You must be logged in to vote
0 replies
Answer selected by livelylife
Comment options

You must be logged in to vote
0 replies
Comment options

You must be logged in to vote
0 replies
Comment options

You must be logged in to vote
1 reply
@drscrewdriver
Comment options

Comment options

You must be logged in to vote
1 reply
@drscrewdriver
Comment options

Comment options

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