Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Bug]: deepseek-r1 模型回答一个较复杂的数学问题,会一直处于思考状态转圈,无法给到完整的答案。 #1156

Open
WildStrom opened this issue Feb 7, 2025 · 2 comments
Labels
bug Something isn't working

Comments

@WildStrom
Copy link

Platform

macOS

Version

0.9.19

Bug Description

deepseek-r1 模型回答一个较复杂的数学问题,会一直处于思考状态转圈,无法给到完整的答案。官方的app可以给到完整的回答,就是很长。

Steps To Reproduce

让接入硅基流动的deepseek-r1 模型回答一个数学问题如下:
任意 39个连续的正整数当中,肯定存在一个数,它的全部数字和能够被11整除。

然后它就一直处于思考状态中,可能是模型的输出太长了,但是结果还没有给出来,输出就结束了。

调整消息的长度到8100,依然没有用,再长请求接口就会报错。

Expected Behavior

能够输出完整的答案

Relevant Log Output

Additional Context

No response

@WildStrom WildStrom added the bug Something isn't working label Feb 7, 2025
@FuzGuo
Copy link

FuzGuo commented Feb 7, 2025

硅基流动中deepseek r1的max_tokens output最大值只能取到8192,再长硅基流动方会停止生成,这是api的问题。

@riverai
Copy link

riverai commented Feb 7, 2025

现在诡计流动在1400token的时候会主动截断推理过程,所以你这个问题是无解的。别问,问就是拉新已经完成了,可以过河拆桥弄死乞丐服了。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

3 participants