mirror of
https://github.com/deepseek-ai/DeepSeek-R1.git
synced 2026-11-05 00:47:42 +00:00
[GH-ISSUE #454] 使用建议中，如何理解：we recommend enforcing the model to initiate its response with "<think>\n" at the beginning of every output #361
Labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/DeepSeek-R1#361
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @zouchengzhuo on GitHub (Feb 19, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-R1/issues/454
用户能控制的是 query，不能控制 response 啊，所以这个使用建议应该如何理解呢？
@hhnqqq commented on GitHub (Feb 19, 2025):
这个是说给部署模型的人听的
@zouchengzhuo commented on GitHub (Feb 19, 2025):
那么用户调用的时候应该如何让模型强制 think 呢，只能通过 prompt 实现吗 - -
@LZWuuu commented on GitHub (Feb 19, 2025):
Same question. When I test some Qwen distill models locally, I use the template in table 1, but most of the response only contain at the end of CoT without at the begining.
@dltc commented on GitHub (Feb 20, 2025):
老铁，看这个文件：https://huggingface.co/deepseek-ai/DeepSeek-R1/blob/main/tokenizer_config.json 的chat_template
就是以"
<think>\n"结尾。简单点说，现在的大模型对话都是文本完成 就是我输如个 ： 中国首都是 ，模型就去完成这个句子：中国首都是北京。对话都是模拟出来的。
现在 我这样写了 中国首都是"
<think>\n",模型就必须顺着前面这个结构去完成后续的内容。@github-actions[bot] commented on GitHub (Apr 23, 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 (May 28, 2025):
false