✨ feat(prompts): use XML format for topic title generation to improve DeepSeek compatibility#15413
Conversation
… DeepSeek compatibility
|
@hardy-one is attempting to deploy a commit to the LobeHub OSS Team on Vercel. A member of the Team first needs to authorize it. |
|
@tjx666 - This PR improves DeepSeek model compatibility for topic title generation by switching to XML-structured prompts. Please take a look. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 5710c97618
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## canary #15413 +/- ##
========================================
Coverage 71.02% 71.02%
========================================
Files 3212 3212
Lines 321285 321290 +5
Branches 29240 28292 -948
========================================
+ Hits 228179 228185 +6
+ Misses 92931 92929 -2
- Partials 175 176 +1
Flags with carried forward coverage won't be shown. Click here to find out more.
🚀 New features to boost your workflow:
|
|
❤️ Great PR @hardy-one ❤️ The growth of project is inseparable from user feedback and contribution, thanks for your contribution! If you are interesting with the lobehub developer community, please join our discord and then dm @arvinxx or @canisminor1990. They will invite you to our private developer channel. We are talking about the lobe-chat development or sharing ai newsletter around the world. |
💻 Change Type
🔗 Related Issue
🔀 Description of Change
DeepSeek V4 Flash/Pro 等模型在生成会话标题时,会忽略 system prompt 的指令,将标题生成任务误认为是对话任务。
使用 XML 结构化 prompt,在 user message 中用
<task>块明确指令,用<conversation>块包裹对话内容,提升模型对指令的遵循率。🧪 How to Test
📸 Screenshots / Videos
📝 Additional Information