🐛 fix: fix `enable_search` parameter intro condition in Qwen #5297

hezhijie0327 · 2025-01-05T00:00:36Z

💻 变更类型 | Change Type

🔀 变更说明 | Description of Change

enable_search 仅对商业模型 qwen-max, qwen-plus, qwen-turbo 系列生效
引入两个环境变量

变量名	描述	默认值	可选值
QWEN_ENABLE_SEARCH	模型在生成文本时是否使用互联网搜索结果进行参考	0	1
QWEN_SEARCH_STRATEGY	搜索互联网信息的数量	standard	pro

📝 补充信息 | Additional Information

ref: https://help.aliyun.com/zh/model-studio/developer-reference/use-qwen-by-calling-api

vercel · 2025-01-05T00:00:39Z

@hezhijie0327 is attempting to deploy a commit to the LobeHub Team on Vercel.

A member of the Team first needs to authorize it.

gru-agent · 2025-01-05T00:00:47Z

TestGru Assignment

Summary

Link	CommitId	Status	Reason
Detail	`defa656`	🚫 Skipped	No files need to be tested {"src/libs/agent-runtime/qwen/index.ts":"no test value"}

Tip

You can @gru-agent and leave your feedback. TestGru will make adjustments based on your input

lobehubbot · 2025-01-05T00:00:47Z

👍 @hezhijie0327

Thank you for raising your pull request and contributing to our Community
Please make sure you have followed our contributing guidelines. We will review it as soon as possible.
If you encounter any problems, please feel free to connect with us.
非常感谢您提出拉取请求并为我们的社区做出贡献，请确保您已经遵循了我们的贡献指南，我们会尽快审查它。
如果您遇到任何问题，请随时与我们联系。

codecov · 2025-01-05T00:03:23Z

Codecov Report

Attention: Patch coverage is 68.75000% with 5 lines in your changes missing coverage. Please review.

Project coverage is 91.72%. Comparing base (bde6d64) to head (9f16d4b).
Report is 48 commits behind head on main.

Files with missing lines	Patch %	Lines
src/libs/agent-runtime/qwen/index.ts	68.75%	5 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5297      +/-   ##
==========================================
- Coverage   91.73%   91.72%   -0.01%     
==========================================
  Files         591      591              
  Lines       46027    46042      +15     
  Branches     2833     2713     -120     
==========================================
+ Hits        42221    42231      +10     
- Misses       3806     3811       +5

Flag	Coverage Δ
app	`91.72% <68.75%> (-0.01%)`	⬇️
server	`?`

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

…earch` parameter

PatriciaLee3 · 2025-01-05T08:43:41Z

Thank you very much for your PR.
Perhaps we should provide an option for people to choose whether to enable this feature, rather than having it on in all cases. IMO, it is unclear how enable_search affects the model's input, and it may lead to difficulties in calculating the number of tokens and unexpected usage costs.

hezhijie0327 · 2025-01-05T10:10:39Z

@PatriciaLee3 added QWEN_ENABLE_SEARCH & QWEN_SEARCH_STRATEGY, should suit your requirement

arvinxx

感觉不能直接在这里用 process.env.QWEN_ENABLE_SEARCH 吧？客户端请求模式下这个方案就失效了？

hezhijie0327 · 2025-01-13T05:13:06Z

感觉不能直接在这里用 process.env.QWEN_ENABLE_SEARCH 吧？客户端请求模式下这个方案就失效了？

确实，但是 Qwen 客户端请求是禁用的

lobe-chat/src/config/modelProviders/qwen.ts

Lines 258 to 262 in 9c5822a

    
           checkModel: 'qwen-turbo-latest', 
        
           description: 
        
             '通义千问是阿里云自主研发的超大规模语言模型，具有强大的自然语言理解和生成能力。它可以回答各种问题、创作文字内容、表达观点看法、撰写代码等，在多个领域发挥作用。', 
        
           disableBrowserRequest: true, 
        
           id: 'qwen',

后续有计划实现一个开关控制这类扩展功能嘛，Hunyuan 和 Minimax 其实也有类似功能，如 Hunyuan 默认是开着的

lobehubbot · 2025-01-13T05:13:19Z

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

I feel like I can’t use process.env.QWEN_ENABLE_SEARCH directly here, right? Does this solution fail in client request mode?

True, but Qwen client requests are disabled

lobe-chat/src/config/modelProviders/qwen.ts

Lines 258 to 262 in 9c5822a

    
           checkModel: 'qwen-turbo-latest', 
        
           description: 
        
             '通义千问是阿里云自主研发的超大规模语言模型，具有强大的自然语言理解和生成能力。它可以回答各种问题、创作文字内容、表达观点看法、撰写代码等，在多个领域发挥作用。', 
        
           disableBrowserRequest: true, 
        
           id: 'qwen',

There are plans to implement an extended function such as switch control in the future. Hunyuan and Minimax actually have similar functions. For example, Hunyuan is on by default.

Sun-drenched · 2025-01-13T07:01:00Z

感觉不能直接在这里用 process.env.QWEN_ENABLE_SEARCH 吧？客户端请求模式下这个方案就失效了？

确实，但是 Qwen 客户端请求是禁用的

lobe-chat/src/config/modelProviders/qwen.ts

Lines 258 to 262 in 9c5822a

checkModel: 'qwen-turbo-latest',

description:

'通义千问是阿里云自主研发的超大规模语言模型，具有强大的自然语言理解和生成能力。它可以回答各种问题、创作文字内容、表达观点看法、撰写代码等，在多个领域发挥作用。',

disableBrowserRequest: true,

id: 'qwen',

后续有计划实现一个开关控制这类扩展功能嘛，Hunyuan 和 Minimax 其实也有类似功能，如 Hunyuan 默认是开着的

感觉可以把扩展功能选项、随机性参数、历史消息数限制参数和模型绑定。有两个场景需求
1.用户在随便聊聊或其他助手处混用不同提供商的大模型，这些大模型对随机性参数的敏感性不同，比如openai的模型随机性超过1就容易出输出乱码、胡言乱语的问题，但是deepseek官方指引通用对话应当把temperature设置为1.3，翻译1.5、（当然这个问题也可以合并到#2726 中提出的预设输出风格配置交互解决）
2.如果一般模型无法解决问题需要调用某些高价模型时继续沿用一般模型的历史消息限制，容易产生过多消费；更改模型前后手动调整限制又略显繁琐。

lobehubbot · 2025-01-13T07:01:15Z

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

I feel like I can’t use process.env.QWEN_ENABLE_SEARCH directly here, right? Does this solution fail in client request mode?

True, but Qwen client requests are disabled

lobe-chat/src/config/modelProviders/qwen.ts

Lines 258 to 262 in 9c5822a

checkModel: 'qwen-turbo-latest',

description:

'通义千问是阿里云自主研发的超大规模语言模型，具有强大的自然语言理解和生成能力。它可以回答各种问题、创作文字内容、表达观点看法、撰写代码等，在多个领域发挥作用。',

disableBrowserRequest: true,

id: 'qwen',

There are plans to implement an extended function such as switch control in the future. Hunyuan and Minimax actually have similar functions. For example, Hunyuan is on by default. ![image](https://private-user-images.githubusercontent.com/58100052/402430282-bed5495e-0f92-49f0-8 ae3-58d7de1cf447.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoic mF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3MzY3NTE1MTUsIm5iZiI6MTczNjc1MTIxNSw icGF0aCI6Ii81ODEwMDA1Mi80MDI0MzAyODItYmVkNTQ5NWUtMGY5Mi00OWYwLThhZTMtNThkN2RlMWNmNDQ3LnBuZz9YLUFtei 1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNTAx MTMlMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjUwMTEzVDA2NTMzNVomWC1BbXotRXhwaXJ lcz0zMDAmWC1BbXotU2lnbmF0dXJlPTBhNTAyZjAxOWUxMGE4NGVmMzE4NmEzM2Y0ZWEyMjM4Y2E4MGY5ZDhjMWM0ODE4ZDkyZ mRlYWQzNTVjZjhlMTImWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0In0.HBYV9E0D3QqMJ50qTMwEsBUGo2dYrbFzgfH4E0w9yOA)

It seems that the extended function options, randomness parameters, and historical message number limit parameters can be bound to the model. There are two scenario requirements

Users use large models from different providers in Random Chat or other assistants. These large models have different sensitivities to randomness parameters. For example, if the randomness of OpenAI’s model exceeds 1, it is easy to output garbled codes and nonsense. , but deepseek official guideline general dialogue should set the temperature to 1.3 and the translation to 1.5. (Of course, this problem can also be merged into the default output style configuration proposed in [Request] 随便聊聊界面和模型随机性配置的改进建议 #2726 to solve the problem interactively)
If the general model cannot solve the problem and you need to call some high-priced models, the historical message limits of the general model will be used, which will easily lead to excessive consumption; it is also cumbersome to manually adjust the limits before and after changing the model.

arvinxx · 2025-01-14T16:11:43Z

Qwen 客户端请求是禁用的

如果是这样的话倒没问题

lobehubbot · 2025-01-14T16:11:57Z

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

Qwen client requests are disabled

If that's the case, that's fine.

lobehubbot · 2025-01-14T16:12:35Z

❤️ Great PR @hezhijie0327 ❤️

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.
项目的成长离不开用户反馈和贡献，感谢您的贡献! 如果您对 LobeHub 开发者社区感兴趣，请加入我们的 discord，然后私信 @arvinxx 或 @canisminor1990。他们会邀请您加入我们的私密开发者频道。我们将会讨论关于 Lobe Chat 的开发，分享和讨论全球范围内的 AI 消息。

### [Version 1.45.12](v1.45.11...v1.45.12) <sup>Released on **2025-01-14**</sup> #### 🐛 Bug Fixes - **misc**: Fix `enable_search` parameter intro condition in Qwen. <br/> <details> <summary><kbd>Improvements and Fixes</kbd></summary> #### What's fixed * **misc**: Fix `enable_search` parameter intro condition in Qwen, closes [#5297](#5297) ([0b19c20](0b19c20)) </details> <div align="right"> [![](https://img.shields.io/badge/-BACK_TO_TOP-151515?style=flat-square)](#readme-top) </div>

lobehubbot · 2025-01-14T16:21:17Z

🎉 This PR is included in version 1.45.12 🎉

The release is available on:

Your semantic-release bot 📦🚀

### [Version 1.87.7](v1.87.6...v1.87.7) <sup>Released on **2025-01-14**</sup> #### ♻ Code Refactoring - **plugin**: Update plugin prompt. - **misc**: Improve ai provider code. #### 🐛 Bug Fixes - **misc**: Fix `enable_search` parameter intro condition in Qwen, fix pull models error in new ai provider, fix some providers issues, support Gemini 2.0 HarmBlockThreshold. #### 💄 Styles - **misc**: Improve model config form modal. <br/> <details> <summary><kbd>Improvements and Fixes</kbd></summary> #### Code refactoring * **plugin**: Update plugin prompt, closes [lobehub#5083](https://github.com/bentwnghk/lobe-chat/issues/5083) ([ddd3b53](ddd3b53)) * **misc**: Improve ai provider code, closes [lobehub#5442](https://github.com/bentwnghk/lobe-chat/issues/5442) ([32013b4](32013b4)) #### What's fixed * **misc**: Fix `enable_search` parameter intro condition in Qwen, closes [lobehub#5297](https://github.com/bentwnghk/lobe-chat/issues/5297) ([0b19c20](0b19c20)) * **misc**: Fix pull models error in new ai provider, closes [lobehub#5441](https://github.com/bentwnghk/lobe-chat/issues/5441) ([1726d85](1726d85)) * **misc**: Fix some providers issues, closes [lobehub#5429](https://github.com/bentwnghk/lobe-chat/issues/5429) ([b7edbd3](b7edbd3)) * **misc**: Support Gemini 2.0 HarmBlockThreshold, closes [lobehub#5422](https://github.com/bentwnghk/lobe-chat/issues/5422) ([4d1a4cd](4d1a4cd)) #### Styles * **misc**: Improve model config form modal, closes [lobehub#5438](https://github.com/bentwnghk/lobe-chat/issues/5438) ([9b303e0](9b303e0)) </details> <div align="right"> [![](https://img.shields.io/badge/-BACK_TO_TOP-151515?style=flat-square)](#readme-top) </div>

…#5297) * 🐛 fix: fix `enable_search`parameter intro condition in Qwen * 🔨 chore: add some notes * ♻️ refactor: intro `QwenEnableSearchModelSeries` to control `enable_search` parameter * 🔨 chore: update comments * 🔨 chore: add env to control `enable_search` & `search_strategy`

### [Version 1.45.12](lobehub/lobe-chat@v1.45.11...v1.45.12) <sup>Released on **2025-01-14**</sup> #### 🐛 Bug Fixes - **misc**: Fix `enable_search` parameter intro condition in Qwen. <br/> <details> <summary><kbd>Improvements and Fixes</kbd></summary> #### What's fixed * **misc**: Fix `enable_search` parameter intro condition in Qwen, closes [lobehub#5297](lobehub#5297) ([0b19c20](lobehub@0b19c20)) </details> <div align="right"> [![](https://img.shields.io/badge/-BACK_TO_TOP-151515?style=flat-square)](#readme-top) </div>

hezhijie0327 added 2 commits January 5, 2025 07:54

🐛 fix: fix enable_searchparameter intro condition in Qwen

cc4d6a2

🔨 chore: add some notes

defa656

dosubot bot added the size:XS This PR changes 0-9 lines, ignoring generated files. label Jan 5, 2025

dosubot bot added the 🐛 Bug Something isn't working | 缺陷 label Jan 5, 2025

♻️ refactor: intro QwenEnableSearchModelSeries to control `enable_s…

78b3100

…earch` parameter

dosubot bot added size:S This PR changes 10-29 lines, ignoring generated files. and removed size:XS This PR changes 0-9 lines, ignoring generated files. labels Jan 5, 2025

🔨 chore: update comments

e13b28f

🔨 chore: add env to control enable_search & search_strategy

204eb12

Merge branch 'lobehub:main' into qwen

9f16d4b

SpeedupMaster mentioned this pull request Jan 11, 2025

[Bug] 阿里云千问/Qwen 2.5 模型请求错误 #5400

Closed

arvinxx reviewed Jan 13, 2025

View reviewed changes

arvinxx merged commit 0b19c20 into lobehub:main Jan 14, 2025
4 of 7 checks passed

lobehubbot added the released label Jan 14, 2025

hezhijie0327 deleted the qwen branch January 15, 2025 11:42

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

🐛 fix: fix `enable_search` parameter intro condition in Qwen #5297

🐛 fix: fix `enable_search` parameter intro condition in Qwen #5297

hezhijie0327 commented Jan 5, 2025 •

edited by arvinxx

Loading

vercel bot commented Jan 5, 2025

gru-agent bot commented Jan 5, 2025 •

edited

Loading

lobehubbot commented Jan 5, 2025

codecov bot commented Jan 5, 2025 •

edited

Loading

PatriciaLee3 commented Jan 5, 2025

hezhijie0327 commented Jan 5, 2025

arvinxx left a comment

hezhijie0327 commented Jan 13, 2025

lobehubbot commented Jan 13, 2025

Sun-drenched commented Jan 13, 2025 •

edited

Loading

lobehubbot commented Jan 13, 2025

arvinxx commented Jan 14, 2025

lobehubbot commented Jan 14, 2025

lobehubbot commented Jan 14, 2025

lobehubbot commented Jan 14, 2025

🐛 fix: fix enable_search parameter intro condition in Qwen #5297

🐛 fix: fix enable_search parameter intro condition in Qwen #5297

Conversation

hezhijie0327 commented Jan 5, 2025 • edited by arvinxx Loading

💻 变更类型 | Change Type

🔀 变更说明 | Description of Change

📝 补充信息 | Additional Information

vercel bot commented Jan 5, 2025

gru-agent bot commented Jan 5, 2025 • edited Loading

TestGru Assignment

Summary

lobehubbot commented Jan 5, 2025

codecov bot commented Jan 5, 2025 • edited Loading

Codecov Report

PatriciaLee3 commented Jan 5, 2025

hezhijie0327 commented Jan 5, 2025

arvinxx left a comment

Choose a reason for hiding this comment

hezhijie0327 commented Jan 13, 2025

lobehubbot commented Jan 13, 2025

Sun-drenched commented Jan 13, 2025 • edited Loading

lobehubbot commented Jan 13, 2025

arvinxx commented Jan 14, 2025

lobehubbot commented Jan 14, 2025

lobehubbot commented Jan 14, 2025

lobehubbot commented Jan 14, 2025

🐛 fix: fix `enable_search` parameter intro condition in Qwen #5297

🐛 fix: fix `enable_search` parameter intro condition in Qwen #5297

hezhijie0327 commented Jan 5, 2025 •

edited by arvinxx

Loading

gru-agent bot commented Jan 5, 2025 •

edited

Loading

codecov bot commented Jan 5, 2025 •

edited

Loading

Sun-drenched commented Jan 13, 2025 •

edited

Loading