deepseek-ai/DeepSeek-R1

Fork 0

mirror of https://github.com/deepseek-ai/DeepSeek-R1.git synced 2026-11-05 00:47:42 +00:00

[GH-ISSUE #565] [Question] 有办法控制DS-R1 的思考时间么？ #451

New issue

Closed

opened 2026-10-26 21:21:32 +00:00 by bsakaguchi · 7 comments

bsakaguchi commented

2026-10-26 21:21:32 +00:00

Owner

Originally created by @Kingtous on GitHub (Mar 12, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-R1/issues/565

有些任务思考时长过长，长达 4 分钟，一直在来回 thinking。想问一下有没有好的实践，可以减轻Ds-R1 的思考时间，达到一个可控范围？

Originally created by @Kingtous on GitHub (Mar 12, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-R1/issues/565 有些任务思考时长过长，长达 4 分钟，一直在来回 thinking。想问一下有没有好的实践，可以减轻Ds-R1 的思考时间，达到一个可控范围？

bsakaguchi

2026-10-26 21:21:32 +00:00

closed this issue
added the
stale

closed-as-stale
labels

bsakaguchi commented

2026-10-26 21:21:33 +00:00

Author

Owner

@fearry commented on GitHub (Mar 12, 2025):

api有控思维链的长度的参数（reasoning_effort）~~有计划会开放使用~~
@https://api-docs.deepseek.com/zh-cn/guides/reasoning_model
在此之前其他手段只能辅助，比如拆分问题，分好几步问之类的

@fearry commented on GitHub (Mar 12, 2025): api有控思维链的长度的参数（reasoning_effort）~~有计划会开放使用~~ @https://api-docs.deepseek.com/zh-cn/guides/reasoning_model 在此之前其他手段只能辅助，比如拆分问题，分好几步问之类的

bsakaguchi commented

2026-10-26 21:21:33 +00:00

Author

Owner

@Kingtous commented on GitHub (Mar 12, 2025):

api有控思维链的长度的参数（reasoning_effort）有计划会开放使用……
@https://api-docs.deepseek.com/zh-cn/guides/reasoning_model
在此之前其他手段只能辅助，比如拆分问题，分好几步问之类的

了解，有这个参数，体验会变得更好。自部署的 ds，有办法怎么控制思维链长度么？

@Kingtous commented on GitHub (Mar 12, 2025): > api有控思维链的长度的参数（reasoning_effort）有计划会开放使用…… > @https://api-docs.deepseek.com/zh-cn/guides/reasoning_model > 在此之前其他手段只能辅助，比如拆分问题，分好几步问之类的了解，有这个参数，体验会变得更好。自部署的 ds，有办法怎么控制思维链长度么？

bsakaguchi commented

2026-10-26 21:21:33 +00:00

Author

Owner

@github-actions[bot] commented on GitHub (Apr 14, 2025):

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!

@github-actions[bot] commented on GitHub (Apr 14, 2025): This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!

bsakaguchi commented

2026-10-26 21:21:34 +00:00

Author

Owner

@lastlastsummer commented on GitHub (Apr 24, 2025):

现在有办法控制思维链的长度吗，我调用vllm部署的deepseek-r1，经常会因为思考太久导致一些问题。

@lastlastsummer commented on GitHub (Apr 24, 2025): 现在有办法控制思维链的长度吗，我调用vllm部署的deepseek-r1，经常会因为思考太久导致一些问题。

bsakaguchi commented

2026-10-26 21:21:34 +00:00

Author

Owner

@Evi233 commented on GitHub (Jul 1, 2025):

Directly truncate the thought chain after reaching a certain number of tokens/time, add </think> and then continue generating. Incomplete thought chains will not cause issues, at most they may result in less precise answers.

@Evi233 commented on GitHub (Jul 1, 2025): Directly truncate the thought chain after reaching a certain number of tokens/time, add `</think>` and then continue generating. Incomplete thought chains will not cause issues, at most they may result in less precise answers.

bsakaguchi commented

2026-10-26 21:21:34 +00:00

Author

Owner

@github-actions[bot] commented on GitHub (Aug 2, 2025):

@github-actions[bot] commented on GitHub (Aug 2, 2025): This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!

bsakaguchi commented

2026-10-26 21:21:34 +00:00

Author

Owner

@github-actions[bot] commented on GitHub (Aug 18, 2025):

false

@github-actions[bot] commented on GitHub (Aug 18, 2025): false